UNDERSTANDING BREASTFEEDING SELF - EFFICACY & SOCIAL SUPPORT IN A BREASTFEEDING DURATION INTERVENTION FOCUSED ON AFRICAN AMERICAN WOMEN By Gayle Shipp A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Human Nutrition - Doctor of Philosophy 2020 ABSTRACT UNDERSTANDING B REASTFEEDING SELF - EFFICACY & SOCIAL SUPPORT IN A BREASTFEEDING INTERVENTION FOCUSED ON AFRICAN AMERICAN WOMEN By Gayle Shipp C ompelling evidence of breastfeeding (BF) mother and child benefits and ongoing efforts to promote BF , have led to s low inclines in BF initiation , but unfortunately not sustained or exclusive BF u p to 6 months , especially among African American women (AA) . Because of concern about associated mother and child health risks, the extent to which BF self - efficacy (BFSE) and perceived social support (PSS), tw o important potential influencers of BF decisions (initiation, duration and exclusivity ), influence BF behavior among AA was examined . We also qualitatively determined how these variables and other factors can impact success of AA in achieving BF goals . Fi fty - three AA enrolled in a 2 - arm randomized, controlled, feasibility trial of BF support and p ost - p artum (PP) weight management (MAMA Bear). Perceived social support, BFSE , and other key variables were collected longitudinally (prenatally, 2 - 8 weeks PP, an d 20 weeks PP) . Controlling for potential confounders (stepwise selection), impact on BF duration and intensity but not initiation (high rate of 75%), were assessed. BF self - efficacy and PSS, moderately correlated, were assessed independently. After completing the first part of the MAMA Bear study , women (n=14) were recruited around 20 weeks PP until data saturation, to participate in semi - structured qualitative interviews that were digitally r ecorded and transcribed. Inductive thematic content analysis was conducted in co njunction with a review of field notes . B reastfeeding self - .05 ; = .114, p .017) , and between 2 - 8 weeks PP p .01 ; =.188, p .001) for all participants , was positively associated with BF duration and BF intensity at 2 - 8 weeks respectively in fully adjusted models. At 20 weeks PP , BF self - efficacy was associated with BF duration (simple linear regression) , but no longer signific ant after adjusting for relevant covariates. Breastfeeding self - efficacy 2 - 8 weeks PP positively influenced BF intensity at 20 weeks PP ( B =.165, p .05). P erceived social support between 2 - 8 weeks PP was positively associated with BF duration in the usual care group .05), but there was no association for the combined group . Perceived social support prenatally in the combined group , was associated with BF intensity between 2 - 0.05) , but not at 20 weeks PP. T wo primary qualitative themes emerged for study participants regarding challenges and facilitators in meeting BF goals : BF c onfidence and discouragement related to BF - related issues . Key BF facilitators were : 1) d ual benef it for infant and mother, 2) p erseverance 3) c ommitment and s elf - d etermination, 4) f luctuating c onfidence , 5) a bility to p ump and, 6) s upport from identified s ocial s upport n etworks . S ome suggestions for BF promotion and support included: 1) t angible and i mmediate help, 2) p ositive n on - j udgmental s education , and 4) s elf - motivation/ w ill p ower/ p erseverance . In conclusion, our findings suggest that continual assessment of BF SE and PSS, particularly prenatally and 6 weeks PP , is beneficial in assisting mother s to achieve BF goals and outcomes, in our AA target group. Future intervention s can use th is strategy and the suggestions from the mothers themselves, for individua lized confidence - building to gain a better understanding of time - sensitive needs, and guide healthcare professionals and paraprofessionals on how to better help AA to achiev e BF goals. Therefore, c ollaborati on and active engag ement with AA and other mother s are en couraged to improve BF promotion, support, and outcomes . Copyright by GAYLE SHIPP 2020 v This dissertation is dedicated to my parents and my daughter: To my parents: You have always been a shining light in my life, encouraging and supporting me on my journey forward. I am appreciative for all that you have done and continue to do for me. To my daughter: My fearless inquisitor. You exude so many amazing qualities at such a young age. Know that you are encouraging me to utilize these qualities in various facets of my life. While I am teaching you, you are also teaching me. vi ACKNOWLEDGMENTS I would like to first thank God for giving me the strength, dedication, and ability to advance in the completion of my doctoral program and dissertation. Next, I would like to acknowledge with a deep sense of reverence, my gratitude towards my parents , Gwendolyn and George Crippen. I would not be where I am today without the ir constant sacrifice, unwavering support , and unconditional love they have and still continue to provide. I am also very thankful for my daughter, Gabrielle. Being a mother to her has given me the opportunity to learn many new things about myself , while shifting my perspective and realizing the significance of my life. I would also like to thank my siblings ( brothers and sister s ) and all of my close friends. Your support and encouragement has always been truly appreciated. I am forever grateful to my advisor Dr. Lorraine Weatherspoon for her dedicated support, guidance , and wisdom . I am fortunate to have been your student and thankful for your mentorship, effort , and time you have invested in me. I hope I will one day become an advisor to others as you have been to me. I would also like to thank Dr. Jean Kerver for believing in me and providing me the opportuni ty to work with her while receiving her invaluable guidance and support . To my remaining committee members , Dr. Sarah Comstock and Dr. Joseph Gardiner, I am so appreciative for your support , expertise and feedback through the completion of my dissertation. I am also thankful for the doctoral students that came before me and those that came into the program with me. Sharing our stories and experience s together encouraged me to continue , and made my day to day experie nce better , knowing that I had each and every one of you. I would also like to provide many thanks to the undergraduate students for their collaborative effort during data collection. I am also appreciative of all of the faculty and staff in vii t he Department of Food Science and Human Nutrition who I have encountered throughout my education al journey at MSU. As a recipient of Robert Wood Johnson Foundation - Health Policy Research Scholars (HPRS) , I first would like to thank the organization for having the v ision, focus and belief to crea te a program of this magnitude where the objective and goal is to make changes for the future by investing in the potential talent of future leaders. I am deeply grateful to have been selected as a scholar in the first cohort. The knowledge and skills accrued through the many networking conferences, workshops , and classes have exposed me to the many opportunities available. This has been of inestimable value towards my gro wth as a scholar ly educator . A ll of the staff, scholars , and mentors deserve the highest kudos. We have truly become a family! I thank all of you for your continuing support throughout this journey . I have so much gratitude for the Mama Bear research sta ff , Henry Ford Health Care System , Wayne State University , and most importantly all the participants (women) that took part in this study that enabled this research to be possible. Every person that I have acknowledged wh o has touched me , has also been touched by God ! He has been the guiding hand in my efforts. The journey of completing my dissertation has been educational , meaningful, and rewarding. I am thankful to the many individuals that ha ve provided guidance and assistance to me whi le navigating this path over the last 5 years. viii TABLE OF CONTENT S LIST OF TABLES ................................ ................................ ................................ ........................ xi LIST OF FIGURES ................................ ................................ ................................ .................... xiv 1.0 Chapter 1 - I ntroduction ................................ ................................ ................................ ............ 1 1.1 Background ................................ ................................ ................................ ......................... 1 1.2 Specific Aims ................................ ................................ ................................ ....................... 6 1.3 Significance of Research ................................ ................................ ................................ ..... 7 1.4 Organization of the Dissertation ................................ ................................ ........................ 8 1.5 Working Definition of Terms ................................ ................................ ............................. 8 2.0 Chapter 2 - Review of the Literature ................................ ................................ ..................... 10 2.1 Breastfeeding Significance and Recommendations ................................ ....................... 10 2.2 Racial and Ethnic differences in Breastfeedi ng Rates ................................ ................... 11 2.3 Determinants of Breastfeeding among African Americans ................................ .......... 13 2.3.1 Historical Factors ................................ ................................ ................................ ....... 15 2.3.2 Cultural and Ethnic Influences ................................ ................................ ................. 16 2.3.3 Intention to Breastfeed ................................ ................................ .............................. 17 2.3.4 Breastfeeding attitudes, beliefs and knowledge ................................ ....................... 18 2.3.5 Misconception about Formula ................................ ................................ .................. 19 2.3.6 Maternity Care Practices ................................ ................................ .......................... 20 2.3.7 Preference to Bottle Feed ................................ ................................ .......................... 21 2.3.8 Perceived Inconvenience ................................ ................................ ........................... 21 2.3.9 Family and Social Support ................................ ................................ ........................ 22 2.3.10 Lactation Experience ................................ ................................ ............................... 23 2.3.11 Employment and Child Care ................................ ................................ .................. 24 2.4 Social Support and Breastfeeding ................................ ................................ ................... 25 2.5 Breastfeeding Self Efficacy ................................ ................................ ............................... 29 2.6 Conceptual Framework ................................ ................................ ................................ .... 33 2.7 Gaps in the Literature ................................ ................................ ................................ ...... 34 3.0 Chapter 3 - Methods ................................ ................................ ................................ ................ 38 Overview ................................ ................................ ................................ ................................ .. 38 3.1 Aim 1 Methods ................................ ................................ ................................ .................. 42 3.1.1 Approach for Aim 1 ................................ ................................ ................................ ... 42 3.1.2 Research Design ................................ ................................ ................................ ......... 42 3.1.3 Sample and Data Collection Procedures ................................ ................................ .. 43 3.1.4 Survey Instruments and variables ................................ ................................ ............ 49 3.1.5 Analyses ................................ ................................ ................................ ...................... 50 3.2 Aim 2 Methods ................................ ................................ ................................ .................. 53 3.2.1 Approach for Aim 2 ................................ ................................ ................................ ... 53 3.2.2 Study Design ................................ ................................ ................................ ............... 54 3.2.3 Sample Procedures ................................ ................................ ................................ ..... 54 ix 3.2.4 Instruments and variables ................................ ................................ ......................... 55 3.2.5 Analyses ................................ ................................ ................................ ...................... 56 3.3 Aim 3 Methods ................................ ................................ ................................ .................. 58 3.3.1 Approach for Aim 3 ................................ ................................ ................................ ... 58 3.3.2 Study Design ................................ ................................ ................................ ............... 59 3.3.3 Sample Procedures ................................ ................................ ................................ ..... 59 3.3.4 Instruments and variables ................................ ................................ ......................... 61 3.3.5 Analyses ................................ ................................ ................................ ...................... 62 Chapter 4 - Breastfeeding Self - Efficacy Predicts Breastfeeding Outco mes in a Sample of African American Women ................................ ................................ ................................ ........... 65 4.1 Abstract: ................................ ................................ ................................ ............................ 65 4.2 Introduction ................................ ................................ ................................ ....................... 67 4.3 Methods ................................ ................................ ................................ .............................. 70 1. Measures: ................................ ................................ ................................ ......................... 72 2. Statistical Analyses: ................................ ................................ ................................ ........ 74 4.4 Results: ................................ ................................ ................................ ............................... 75 4.5 Discussion ................................ ................................ ................................ ......................... 104 4.6 Future Implications ................................ ................................ ................................ ........ 107 Chapter 5 - Perceived Social Support measured during late pregnancy, 6 and 20 weeks post - partum Predicts Breastfeeding Behavior in African American Women ................................ . 109 5.1 Abstract: ................................ ................................ ................................ .......................... 109 5.2 Introduction ................................ ................................ ................................ ..................... 111 5.3 Methods ................................ ................................ ................................ ............................ 116 1. Measures: ................................ ................................ ................................ ....................... 117 2. Statistical Analyses: ................................ ................................ ................................ ...... 119 5.4 Resu lts ................................ ................................ ................................ .............................. 120 5.5 Discussion ................................ ................................ ................................ ......................... 132 5.6 Future Implications ................................ ................................ ................................ ........ 135 Chapter 6 - Understanding Factors Influencing Breastfeeding Outcomes in a Sample of African American Women ................................ ................................ ................................ ......... 137 6.1 Abstract: ................................ ................................ ................................ .......................... 137 6.2 Introduction ................................ ................................ ................................ ..................... 139 6.3 Methods ................................ ................................ ................................ ............................ 141 1. Sample and Recruitment ................................ ................................ .............................. 141 2. Instruments and Data Collection Procedures ................................ ............................ 143 3. Data Analysis ................................ ................................ ................................ ................. 144 6.4 Results ................................ ................................ ................................ .............................. 145 6.5 Discussion ................................ ................................ ................................ ......................... 156 6.6 Conclusion ................................ ................................ ................................ ....................... 162 Chapter 7 - Summary and Conclusions ................................ ................................ ..................... 164 APPENDICES ................................ ................................ ................................ ........................... 170 Appendix A: Clinic Approval ................................ ................................ .............................. 171 x Appendix B: Participant Consent ................................ ................................ ........................ 172 Appendix C: Data Collector Telephone Script ................................ ................................ .. 181 Appendix D: Randomization Scheme ................................ ................................ ................. 185 Appendix E: Demographical Questions ................................ ................................ .............. 186 Appendix F: Breastfeeding Characteristics Questionnaire ................................ .............. 193 Appendix G: Breastfeeding Self Efficacy Scale - Short Form Instrument by Cindy Lee Dennis ................................ ................................ ................................ ................................ ..... 197 Appendix H: Breastfeeding Self Efficacy Scale - Short Form Instrument (present tense - during pregnancy) ................................ ................................ ................................ ................. 199 Appendix I: Breastfeeding Self Efficacy Scale - Short Form Instrument (past tense - post - partum after breastfeeding has stopped) ................................ ................................ ............ 201 Appendix J: Breastfeeding Intensity ................................ ................................ ................... 203 Appendix K: Breastfeeding Social Support Questionnaire by Robbie Hughes .............. 204 Appendix L: Modified Breastfeeding Social Support Questionnaire .............................. 205 REFERENCES ................................ ................................ ................................ .......................... 221 xi LIST OF TABLES Table 1.1 ..5 Table 3.1 Project Incentives .48 .48 52 . 5 3 .58 Table 3.6 Aim 3 Research Topic Areas/ Questions and Interview Questions 6 4 Table 4.1 Characteristics of A ll P articipants at E ..7 6 Table 4.2 Mean Pregnancy and Post - Partum Breastfeeding Self - Efficacy Scores of Participants ... Table 4.3 Mean Table 4.4 Characteristics of P articipants at E nrollment by B reastfeeding S tatus at 6 and 20 W eeks P ost - Table 4.5 Breastfeeding Intensity of participants at 6 and 20 W eeks P ost - P 79 Table 4.6: Breast feeding Self - Efficacy measured at L ate - P regnancy I nfluences B reastfeeding D uration of all P articipant s ( N=40 ) 1 Table 4.7 Predictors Measured at Late Pregnancy Influencing Breastfeeding Duration Separated by Group ( N=40 ) .. 8 3 Table 4.8 Breastfeeding Self - Efficacy Measured at 6 weeks Post - Partum Influences Breastfeeding Duration ( N=28 ) 5 Table 4.9 Influence of Breastfeeding Self - Efficacy M easured at 6 W eeks P ost - P artum on Breastfeeding Duration S eparated by G roup ( N=28 86 Table 4.10 Breastfeeding Self - Efficacy Measured at 20 W eeks Post - Partum Influences Breastfeeding Duration (N=21) .. 88 Table 4.11 Influence of Breastfeeding Self - Efficacy Measured at 20 weeks Postpartum on .. 89 xii Table 4.12 Breastfeeding Self Efficacy Measured Prenatally Influences Breastfeeding Intensity at 6 Weeks Pos t - 9 0 Table 4.13 Prenatal Breastfeeding Self - Efficacy Influences Breastfeeding Intensity at 6 weeks Post - .. 9 2 Table 4.14 Association Between Predictors Measured a t Late Pregnancy and Breastfeeding Intensity Measured at 20 Weeks Post - 9 3 Table 4.15 Predictors Influencing Breastfeeding Intensity at 20 W eeks Post - Partum S eparated by G roup ( N=32 ) 4 Table 4.16 Breastfeeding Self - Efficacy Influences Breastfeeding Intensity When Both Measured at 6 Weeks Post - Partum (N=28) Table 4.17 Breastfeeding Self - Efficacy Measured at 6 Weeks Post - Partum Influences Breastfeeding Intensity a t 6 Weeks Post - Table 4.18 Breastfeeding Self - Efficacy Measured at 6 Weeks Post - Partum Influences Breastfeeding Intensity at 20 Weeks Post - Table 4.1 9 Association Between Predictors Measured at 6 Weeks Post - Partum and Breastfeeding Intensity at 20 Wee ks Post - 10 0 Table 4.20 Breastfeeding Self - Efficacy Influences Breastfeeding Intensity M easured at 20 W eeks Post - Partum ( N=21 ) . 10 2 Table 4.21 Predictors Influencing Breastfeeding Intensity at 20 W eeks Post - Partum ( N=21 . 10 3 Table 5.1 Demographic Characteristics of A ll P articipants D uring L ate P 1 Table 5.2 Summary of Social Su 2 Table 5.3 Association of Predictors M easured at L ate P regnancy I nfluencing Breastfeeding 4 Table 5.4 Perceived Social Support M easured at 6 W eeks Post - Partum I nfluence s Breastfeeding 25 Table 5.5 Predictors M easured at 20 W eeks Post - Partum I nfluencing Breastfeeding Duration.1 26 Table 5.6 Perceived Social Support M easured at L ate P regnancy I nfluences Breastfeed ing Intensity at 6 W eeks Post - P 27 Table 5.7 Predictors M easured at L ate P regnancy Influences Breastfeeding Intensity at 20 W eeks Post - 28 xiii Table 5.8 Perceived Social Support I nfluences Breastfeeding Intensity at 6 W eeks Post - 29 Table 5.9 Association of Predictors Measured at 6 Weeks Post - Partum Influencing Breastfeeding Intensity at 20 Weeks Post - Par tum .. ...13 0 Table 5.10 Association of Predictors M easured at 20 W eeks Post - Partum I nfluencing Breastfeeding Intensity at 20 W eeks Post - ... 1 Table 6.1 Demographic Characteristics of African American Qualitative Study Participants ( N ... ...1 45 6 Table 6.3: Successes with Meeting Breastfeeding .. ... . ..1 47 Table 6.4: Suggestions for E xpanding B reastfeeding S upport and P romotion A mong African American W ... 48 xiv LIST OF FIGURES Figure 1.1 Social Ecological 37 1 Figure 5.1 Relationship B etween Breastfeeding Self - Efficacy and Perceived Social Support M easured D uring L ate P . .12 2 Figure 5.2 Relationship Between Breastfeeding Self - Efficacy and Perceived Social Support Measured Between 2 - 8 Weeks Post - .. 12 3 Figure 5.3 Relationship B etween Breastfeeding Self - Efficacy and Perceived Social Support M easured at 20 W eeks Post - 3 1 1.0 Chapter 1 - Introduction 1.1 Background Ev idence is compellingly in support of breastfeeding (BF) providing numerous benefits to both infant s and mothers. The positive effects range from nutritional, physiological, psychological, and even dev elopmental for mother and child . 1 2 Community benefits such as, decreased annual healt h care costs, decreased cost for public health programs, decreased environmental burden for disposal of formula bottles and cans associated with breastfeeding have also been documented . 3 4 In spite of all of the evidence in support of breastfeeding and the clear recommendations provided by the American Academy of Pediatrics, Institute of Medicine, and World Health Organization (WHO) , most women globally do not continue breastfeeding or exclusively breastfeed up to 6 months postpartum . 5 The United Nations Children Fund ( UNICEF ) and World Health Organization ( WHO ) note breastfeed ing as a, and reference the substantial impacts bre astfeeding makes worldwide . 6 Increasing the rate of exclusive b reastfeeding in the first six months for up to at least 50% of all women , has recently been specified by WHO as one of the six global nutrition targets for 2025 . While the WHO recommends exclusive breastfeeding for the first 6 months, th eir data between 2 0 13 - 20 18 suggest s that globally only 43% initiate breastfeeding and even less exclusively breastfeed for up to 6 months ( 37% ) . 7 Therefore , i exclusive breastfeeding rates to at least 6 months postpartum continue s to remain a primary aim of nutrition and public h ealth programs across the world . Additionally, in the US, the Healthy People 2020 objectives for breastfeeding are to increase the proportion of infants who are ever breastfed, increase the duration of breastfeeding to at least 6 months , and increase the proportion of infants who are breastfed ex clusively up to 3 and 6 months . The Healthy People 2020 objective target s are set at 81.9% for infants ever 2 breastfed, 60.6% for those who continue to breastfeed up to 6 months and 25.5% of mothers who exclusively breastfeed their infants through 6 months . According to the 2014 National Immunization Survey (Table 1 .1 ) , 82.5 % of infants were ever breastfed, 55.3 % were breastfed up to 6 months, and 24.9 % were exclusively breast fed for 6 months in the US . Due to the success of various breastf eeding initiatives , breastfeeding rates across the U.S. have modestly been rising. Michigan is not far behind compared to the U.S. national rates with 75.3% of infants ever breastfed, 46.6% breastf ed at 6 months and 16.2 % breast feeding exclusively at 6 months . A lthough, over the past decade there have been modest national increase s in breastfeeding initiation and among African Americans, Hispanics, and Non - Hispanic W hites ; none of the se subgroups have reached the recommendations set by Healthy People 2010. B reastfeeding practices are disproportionately lower for certain races 1 , which has been shown to be related to sociodemographic characteristics, cultural differences, and geographical location . Breastfeed ing rates for African American women have increased as evidence d by the National Immunization Survey of children born in 2014 3 ( Table 1 .1 ), but African American mothers are still 2. 5 times less likely to ever breastfeed or continue breastfeeding compared to W hite women . The rates of breastfeeding vary from state to state, but the disparity in rates is evident among African American s in almost all states . This disparity is disturbing when one considers the fact t hat nationally , African American infant mortality rates are double that compared to Non - Hispanic Whites , and breastfeeding has been shown to reduce infant mortality rate s. 8 9 T here has been substantial progress in breastfeeding rates from the implemen tation of national initiatives . However, the racial/ethnic gap in breastfeeding rates has not narrowed , despite all of the continuous efforts which continue to make this a public health issue . 1 10 3 Michigan focus ed on breastfeeding through the implemen tation of the Infant Mortality Reduction Plan 2016 - 2019, which not only intends to promote breastfeeding initiation and duration but also reduc e breastfeeding disparities among African Americans . 11 With Detroit, Michigan being pr edominantly African American demographically, the city has taken ste ps to improve BF rates among AA residents by hiring African American women to become breastfeeding peer counselors and training them to become skilled lactation consultants. However, it is critical to recognize that there are several potential race specif ic challenges that could influence the success of such programs. It has been shown in the literature that there are several factors that contribute to the choice of infant feeding, which account for the observed racial/ethnic differences in breastfeeding behavior. For example, a critical barrier of breastfeeding initiation, du ration and exclusivity , particularly for African American women is r eturning to work , e specially for those employed in low - income jobs where maternity leave is usually shorter , and the work environment is less supportive . 1 In addition, studies ha ve identified poor family and social support in general as additional barriers for continuing breastfeeding 8 and demonstrated that mothers and other caregivers require active support for establishing and sustaining breastfeeding practices . 12 However, i t is not well understood how African American interpersonal support systems can increase or maintain their breastfeed ing practices . Therefore, it is important to understand how breastfeeding social support is perceived, the elements th at women perceive as supportive, how or if they change over time and to gain a better understanding of the representations of support with in different spheres (hospital, family, work) in order to potentially maximize breastfeeding success. 4 Multiple studies have also shown that maternal confidence is a nother significant modifiable factor that influence s breastfeeding outcomes . 13 - 15 But, there is a paucity of studies that have analyzed breastfeeding self - efficacy specifically within the African American population. In addition, most of the studies in the literature have focused on breastfeeding sel f - efficacy primarily as it relates to breastfeeding duration and not other breastfeeding outcomes such as exclusivity or intensity . African American women and their infants are at a particularly high risk for health complications given the clear lag in achieving the Healthy People 2020 breastfeeding initiation, duration and exclusivity objectives compared to those in other race categorie s. 16 17 B reastfeeding self - efficacy and social support can be key influences for infant feeding decisions, s pecifically breastfeeding initiation and more importantly , breastfeeding duration . 15 18 19 It is therefore, imperative to understand how these modifiable factors, social support a nd breastfeeding self - efficacy , impact breastfeeding in African American women wh o have the lowest rates of BF initiation, duration and exclusivity . The overall goal of this proposed research is to understand if and how breastfeeding self - efficacy influences breastfeeding practices among African American women , as well as to gain a better understanding of the role of social support in BF outcomes . Specific objectives are : 1) to determine how breastfeeding self - efficacy impact s breastfeeding duration and intensity ; 2) asses s how social support impact s breastfeeding duration and int ensity and ; 3) understand how these and other modifiable factors in achieving her breastfeeding goals. This study was conducted in conjunction with Mothers Allied with Mothers Around Breastfeeding Encouragement and Return to Health After Baby ( Mama Bear ) pilot dual 5 intervention (R21HD085138 - 02). Mama Bear is a randomized control led trial that tested an intervention , which incorporated a postpartum weight management component into an effective breastfeeding peer - suppo rt program with up to 6 months postpartum follow - up . The overarching goal of this study wa s to identify and understand how the modifiable variables of interest (breastfeeding self - efficacy and social support) influence breastfeeding outcomes by also measur ing breastfeeding intensity (% of feeds from breastfeeding versus f ormula) in an effort to gain an understanding of how th ese predictors may need to be enhanced through Mama Bear and other intervention s . Gaining a better understanding of BF self - efficacy and perceived social support can provide an important theoretical basis for future planning and enhance our understand ing of the role they play in increasing or promoting initiation , as well as longer dura tion and exclusive patterns of breastfeeding among African Americans . The ultimate goal is to enhance the comprehensiveness of breastfeeding initiatives for this vulnerable target group t hat could lead to increased duration and intensity of breastfeeding a s well as exclusivity , and reduce the health disparity gap. Table 1.1 National Breastfeeding Initiation, Continuation, and Exclusivity Rates Healthy People 2020 Goals National Average Rates Hispanic Women Caucasian Women African American Women Initiation 81.9 82.5 84.8 85.7 68.0 Exclusive thru 3 months 46.2 46.6 45.5 51.5 32.7 Exclusive thru 6 months 25.5 24.9 24.5 27.9 15.0 Breastfed at 6 months 60.6 55.3 52.5 60.0 41.5 Breastfed at 12 months 34.1 33.7 31.7 37.8 21.5 Note. Adapted from - demographics among Prevention, 2014, & U.S. Department of Health and Human Services (2000). Healthy People 2020. Washington, DC. 6 1.2 Specific Aims Aim 1 Specific Aim 1 was to d etermine how breastfeeding self - efficacy (BSE) impacts breastfeeding initiation, duration, and intensity in African American women . This aim was guided by the following research questions: 1. Does breastfeeding self - efficacy (BSE) assessed in late pregnancy, prior to the breastfeeding support intervention trial , predict breastfeeding initiation regardles s of group assignment in the randomized control trial ? 2 . Does breastfeeding self - efficacy (BSE) predict breastfeeding duration for the participants in either intervention or control group ? 3. Does breastfeeding self - efficacy (BSE) predict breastfeeding intensity for the participants in either intervention or control group ? Aim 2 Identify perceived social support for breastfeeding and assess how it impacts breastfeeding self - efficacy, initiation, duration, and intensity . This aim was guided by the following research questions: 1. Does Breastfeeding Social Support (BSS) or Breastfeeding Self - Efficacy (BSE) , or the interaction between BSS and BSE measured in late pregnancy , prior to a breastfeeding intervention , predict breastfeeding init iation? 2. Do Social Support and Breastfeeding Self - Efficacy combined predict breastfeeding duration for the participants? 3. Do Social Support and Breastfeeding Self - Efficacy combined predict breastfeeding intensity for the participants? 7 Aim 3 Use a two - stage interviewing process to g ain an in depth , qualitative unde rstanding of the factors that influence African American women of their b reastfeeding goals. The s pecific in terest for this aim was to: 1. Utilize information provided by key informants (those who work within the community and with Mama Bear participants) to inform and guide questions used with Mama Bear participants (Stage 2) 2. Gain a better understanding of personal experiences of African American women and how breastfeeding social support and self - efficacy may influence the success of breastfeeding goals and overcoming barriers. 3. Identify new ideas or knowledge that can assist in the support and success of breastfeeding for the benefit of African American women who want to breastfeed. 1.3 Significance of Research While breastfeeding initiation rates among all groups of women within the US have increase d , the rises in breastfeeding duration and exclusivity remain disproportionately lower than national goals , especially for low - income African American women , making this a n issue of concern that warrants attention. T here has been minimal research with interventions aiming to enhance breastfeeding duration , especially among African American women . 1 Johnson et al 1 , conducted a systematic revi ew of interventions aimed at enhanc ing breastfeeding rates among African American women and suggested that b reastfeeding duration in the past may have been less of a focus, primarily because the re were low levels of breastfeeding initiation among all populations . Hence, this made breastfeeding initiation a primary focus of studies. A unique aspect of th e current study is measuring breastfeeding intensity in an effort to gain an in - depth 8 understanding o f how modifiable factors such as, s elf - efficacy and social support may impact breastfeeding behaviors. The information gained from this study provide s a better understanding of strategies and potential solutions needed to enhance breastf eeding beh aviors , specifically with the goal of increasing the duration and intensity of breastfeeding and subsequently breastfeeding exclusivity among African American mothers in order to narrow the gap in breastfeeding outcomes between AA and N on - Hispanic W hite women in the US . 1.4 Organization of the Dissertation This dissertation is organized into seven chapters addressing the specific aims . Chapter one present s a general introduction of the problem and rationale for completing the study. Chapter two provide s a review of the literature on breastfeeding significance and recommendations, racial and ethnic differences in breastfeeding rates, determinants of breastfe eding, factors influencing breastfeeding among African Americans, social support and breastfeeding, breastfeeding self - efficacy, breastfeeding intensity, and challenges of breastfeeding interventions . Chapter three present s the overall methods employed to achieve the study goals and objectives . Chapter four, manuscript one , addresses Aim 1 relative to breastfeeding self - efficacy. Chapter five, manuscript two, addresses the second aim of the study focused on social support. Chapter six, manuscript three, enc ompasses qualitative findings from African American women describing unique factors impact ing breastfeeding goals. Chapter seven provides an overall summary of the three studies including strengths and weaknesses, conclusions , and recommendations for futur e research studies. 1.5 Working Definition of Terms The following terms will be reference d throughout the dissertation : 1) Breastfeeding Initiation : i nitiating breastfeeding within the first hour of life . 20 9 2) Exclusive Breastfeeding : t he practice of feeding an infant breast milk only without any additional food or water. The American Academy of Pediatrics (AAP) and WHO recommend that infants be exclusively breastfed for the first six months of life. 3) Breastfeeding Duration : t he length of time breastfeeding occurs. The AAP recommends that babies should continue to breastfeed for a year and for as long as desired by the mother and baby. 4) Supplementation : feeding the infant as much breastmilk as the mother is able to provide as the primary means of nutrition and feeding the infant with formula occasionally to make up for the lac k of breastmilk provided. Mothers that have a lower production of milk that does not meet their infants needs often supplement feeding with infant formula to ensure their ba bies receive adequate nutrition. 5) Breastfeeding Intention: Having the goal or plan of following through with the behavior of breastfeeding. 6) Breastfeeding Intensity: The percentage of milk feeds from breastfeeding versus Formula . 21 - 23 7) Social Sup port: inter - personal interactions . 24 8) Breastfeeding Self - Efficacy: Self - perceived confidence in their ability to perform a specific behavior while regulating motivation, thoughts, emotions, and social environment. Breastfeeding self - efficacy in this dissertation is a her ability to breastfeed as measured by the Breastfeeding Self Efficacy Scale (BSES) . 13 10 2.0 Chapter 2 - Review of the Literature 2.1 Breastfeeding Significance and Recommendations Breast milk is as an optimal source of nutrition for infants because of its unique qualities and is known as an effective preventive health measure for infants. Infants who are breastfed have a lower risk of developing upper respiratory infections, diabetes mellitus, allergies, necrotizing enterocolitis, gastro - intestinal infections, and SIDS compared to infants who are formula fed . 2 25 - 27 Breastfeeding has also been associated with reduced mortality and morbidity in infants and children well into the second year of life . 28 - 30 Studies have concluded that breastfeeding provides a protective factor providing resistance to many illnesses . 31 In addition , there are many documented benefits for breastf eeding mothers such as, reduced rates of ovarian cancer, reduced premenopausal breast cancer, reduced obesity, typ e 2 diabetes, and heart disease . 2 32 T here is an extensive body of literature that expounds many long - term public health benefits that extend beyond the mother and child, specifically providing positive changes to communities globally and locally. 20 Multiple health organizations , including th e World Health Organization (WHO) , United UNICEF ) , Centers for Disease and Control ( CDC ) , and United States Breastfeeding Committee ( USBC ) all have recommendations for breastfeeding that are as follows; initiation of breastfeeding within the first hour after birth; exclusive breastfeeding for the first six months and complementary feeding starting thereafter ; and continued breastfeeding for at least one year or more. 18 20 25 33 All of these organizations recognize the importance of breastfeeding , stand firm in the recommend atio n s for breastfeeding and value the prominence of exclusive breastfeeding for the minimum specified time . While most mothers are initiating breastfeeding with U.S. National Breastfeeding rates at 81.1% , f ewer reach the duration targets 11 recommended by the World Health Organization (WHO) and the American Academy of Pediatrics (AAP) with 51.8% breastfeeding at 6 months, 30.7% breastfeeding at 12 months, and 22% exclusively breastfeeding at 6 months. 25 34 35 While breastfeeding promotion efforts have resulted in overall increases in women successfully initiating breastfeeding, there are still many women and certain ethnic groups who experience different barriers/challenges that may keep them from not only initiating , but also from continuing to breastfeed. 2.2 Racial and Ethnic differences in Breastfeeding Rat es While the benefits of breastfeeding are well known and supported , there are still fluctuations in rates throughout the U.S . . Rates are considerably lower among certain subgroups of women. Table 1 .1 shows from the 2009 National Immunization Survey , that Non - Hispanic Whites have the highest breastfeeding initiation rates at 84 % and 58 % breastfeeding until 6 months ; Hispanics have breastfeeding initiation rates at 83% but decrease to 45.6 % at 6 months, and blacks have a 66.3 % initiation that decreases to 39.1 % at 6 - month s . 3 While these numbers are encouraging and there have been noticeable improvements over the last 20 years , these rates still mask huge racial disparities that exist . S pecifically , among African American women , their rates are s ignificantly lower compared to their counterparts. African American women have consistently maintained the lowest breastfeeding rates of any race being 2.5 times less likely to bre astfeed compared to white women . While breastfeeding is beneficial to all mo thers and infants, these disparities are particularly troubling because African American women and their infants are at the highest risk of developing health complications and experiencing adverse health effects during the perinatal and post - partum periods . I nfant mortality rates are 2.2 times higher among African American infants compared to Non - Hispanic Whites. 36 African American women 12 have higher rates of diabetes, cardiovascular disease, and obesity compared to Non - Hispanic White women. 9 Although breastfeeding has been suggested as one of the means for addressing excess morbidity and mortality in underserved populations, t here are several factors that contribute to the choice of infant feeding , which may account for the observed raci al/ethnic differences in breastfeeding behavior. Factors include socio - demographic characteristics such as; age, marital status, family income, educational level, and employment. It has been suggested that m others with lower rates of breastfeeding tend to be young, low - income, African American, unmarried, less educated, participants in the Supplemental Nutrition Program for Women, Infants, and Children (WIC), overweight or obese before pregnancy, and more likely to report their pregnan cy was unintended . 37 Mitra and colleagues 38 concluded that women who are white, educated, from the middle class and hav e small families are more likely to breastfeed than their low - income, less educated counterparts and more likely to breastf eed than many African American women. Dubois and Girard 39 socioeconomic status increases , her likelihood of breastfeeding and exclusively breastfeeding also increases. In addition , to the factors that are commonly attributed to low breastfeeding rates, there are additional factors that may be specific to African American women. Along with ec onomic and educational factors, African American mothers face other breastfeeding barriers that include cultural, social an d psychological beliefs and factors, ideas of motherhood, ideas about and attitudes towards breastfeeding from social and personal su pport systems, life experiences, and beliefs about their bodies. 9 40 41 While there is debate in the literature about whi ch are the strongest predictors, none have been determined as having a causal relationship . 13 While breast milk is highly recommended, the causality of the racial disparity in breastfeeding initiation, duration, and exclusivity is not well known. H aving racial / ethnic disparities in breastfeeding initiation and duration may contribute to further disparities in maternal and infant health outcomes. African American children are more likely than average to experience obesity, SIDS, and other conditions associated with a lack of breastfeeding. 42 In addition , African American infants have the highest rates of infant mortality, premature birth, low birth weight, and very low birth weights in the United States . 43 African American infants can benefit greatly from increased breastfeeding initiation and duration by the mother that can help in decreasing poor health outcomes. There remains a strong need for determining why there is a disparity in breastfeeding rates between African American women and what cultural, social, and psychological factors are shaping breastfeeding behaviors. In addition, it is imperative to have their voice describe these factors to help inform how to promote and protect breastfeeding . It remains a priority because breastfeeding disparities still exist which having strategies and perspectives can help tailor breastfeeding interventions and could help i ncrease breastfeeding rates within the African American population. 10 2.3 Determinants of Breastfeeding among African Americans Although the benefits of breastfeeding are significant and well known, women's choices surrounding breastfeeding are influenced by the complexities of their social and cultural realities . The method of infant feeding is a personal decision made by the mother, but often s ocio - contextual factors across multiple levels influence the mother s decision . 44 In 2011, The Surgeon specific approaches that have been shown to improve breastfeeding rates for all women. 18 While there 14 have been modest strides towards the improvement of those rates, there are still barriers to breastfeeding experienced among all women. The literature indicates that breastfeeding is associated with sociodemographic characteristics, such as race, socio - eco nomic level, working status, immigrant status, educational level. Women who breastfeed their infants tend to be older, married, have more years of formal education, and have higher household incomes. Several studies noted that prenatal intentions are among the strongest predictors for initiating breastfeeding. 5 40 45 Even if a mother may intend to breastfeed, she still may face obstacles that render it difficult to convert the intention into a sustained practice. Breastfeeding rates show an increase in breastfeeding initiation among African Americans, but reasons women mention for limiting or terminating breastfeeding a lot sooner than their counterparts illuminate some of the challenges they face. Intention to breastfeed, 45 46 attitudes toward infant feeding methods, 46 47 lactation experiences, maternity care practices, 48 maternal obesity, breastfeeding self - efficacy, 49 social support, 49 50 health care, and employment 51 all influence breastfeeding initiation and duration. African American women tend to experience a culmination of these various factors and more affecting their choices and breastfeeding experiences , resulting in a persistent health disparity. Additional factors specific to African Americans that are prevalent in the community contributing to breastfeeding disparities and influencing their decision to and continue breastfeed ing include cultural knowledge, racism, 52 healthcare provide bias, 52 attitudes towards and about breastfee ding, stigma, lack of social support, beliefs of inconvenience, pain, and sexuality. 53 In the upcoming subsections, factors influencing breastfeeding among all women are addressed, but it is important to note that some of these fact ors may be more prevalent among African American women compared to others. The overall result is that while far fewer women 15 are breastfeeding exclusively at six months , African American mothers specifically are less likely than any other racial group to in itiate breastfeeding, continue breastfeeding to 6 months, and breastfeed exclusively. African American women have historically breastfed at lower rates than recommended and below national goals. The literature has established that African American women ar e significantly less likely to breastfeed and breastfeed for a shorter length of time compared to women of other groups. 12 40 41 43 54 Breastfeeding disparities are hence a public health problem for all, but especially for African Americans. 2.3.1 Historical Factors The history of breastfeeding in the African American community and factors related to why these disparities exist are ve ry complex and inherent. Many African American women slavery. 55 As wet nurses, African American women were forced to nurse and care for the children of thei r slave masters, while oftentimes weaning and neglecting their own children . If a white w oman had problems with breastfeeding or did not want to breastfeed, a wet nurse was enlisted . This resulted in insufficient car e , food , and higher infant mortality rates for African American children . In addition to the hardship of slavery faced by African American women, they also endured exploitation and demonization of their character, sexuality and bodi es, which le d them . 55 Reflecting through the distinct historical lens, of slavery and racial hatred had an emotional resonance that cont inue s to impact African American women, infants and greatly affect breastfeeding rates. 55 16 2.3.2 Cultural and Ethnic Influence s While many factors influence breastfeedi ng behaviors such as initiation and duration , s tudies alluded to the influence of culture . L eading researcher s culture, beyond her ethnicity, may influence breastfeeding behaviors . a cultural and structural dimension that reflect s various influences that directly , or indirectly affect health and well - bein g . 56 T hese influences include kinship and socia l factors. McKee and co llea gues 57 found from examining predictors of breastfeeding decisions and initiation that education and ethnicity were the most important predictors of planning to breastfeed. In addition, among Hispanic participants, greater identification with Hispanic culture was associated with increased likelihood of planning to breastfeed. M ore rese arch is needed however, to gain a better understanding of the influence of culture, values, and beliefs on breastfeeding. Underwood et al , 58 found from a small purposive sample of low income African American women that their values and beliefs were developed and learned primarily within the context of their cultural group , which influenced not only the type of feeding, but feeding schedule, amount of feeding, and the introduction of solid foods. One qualitative study con ducted by Bentley and colleagues mentioned how African American mothers felt that feeding decisions were intergenerational . 59 Their participants recalled bottle - type of the bottle, as well as the brands of formula as they grew up . These early life experiences that are recalled can leave impressions and impact infant feeding decisions. Lewallen et al, 12 found from a qualitative study with African American women that specific perceived cultural issues that arose were often related to comments received from famil y members and friends . These individuals are the support networks for these women and sharing opinions whether 17 positive or negative often influences breastfeeding behaviors. However , Street and Lewallen found within their qualitative study that 73 out of 245 responses mentioned the word culture . Of those responses mentioning culture by name, more white women indicated culture having an influence than black women. There could be several reasons and limitations when addressing cultural influence in stud ies ; i t may be that several other factors had more of an influence than culture, or the definition of culture used may be too narrow . Another option to take under consideration is that studies may not include or address how the participant may define culture and what specific elements define their culture. 2.3.3 Intention to B reastfeed Having the intention to breastfeed has been shown to be a predictor of initiating the breastfeeding behavior. Persad suggested that breastfeeding intention was associated with positive breastfeeding attitudes, higher household incomes, higher education, being born outside of the US, being A fro - Caribbean as opposed to African American, having social support, and attending breastfeeding classes . 45 Many studies have explor ed breastfeeding intention and have found that it is one of the strongest predictors of breastfeeding initiation, duration , intensity, and exclusivity , 38 46 60 but more work is needed with understanding intention to breastfeed can translate into sustained breastfeeding behavior . Stuebe and colleagues 46 found in thei r study that maternal knowledge about infant health benefits, as well as comfort with breastfeeding in social settings, was directly related to intention to exclusively breastfeed, particularly among minority women. Studies have shown that African American women however have intentions to breastfeed, but for some reason it is not being translated into a successful practice . 43 45 61 18 2.3.4 Breastfeeding a tti tudes, beliefs and knowledge While most women are aware of breastfeeding benefits, some women may have certain feelings, beliefs , and even a wide - range in knowledge about breastfeeding prior to becoming pregnant. Prior to pregnancy, women may receive little or no information about bre astfeeding , specifically addressing the importance of breast milk for normal infant growth, development , and he alth . In a 2013 study, Kornides reported that mothers with greater knowledge about breastfeeding benefits were 11.2 times more likely to initiate breastfeeding , and 5.6 times more likely to breastfeed at two months than those with lower levels of knowledge about breastfeeding . 47 While health organizations have focused efforts to educate women about the benefits of breastfeeding, knowledge of the health benefits of breastfeeding alone is clearly not enough to facilitate the process . One key challenge are the inconsistent messages that women get from health care professionals and their social network of family and friends . 62 Women also describe d receiving conflicting professional advice early in the postnatal period about mastering the technical nuances of breastfeeding . 63 Additionally, women may also recei ve different sources of education about breastfeeding. The level of influence exhibited by differing sources of education about breastfeeding needs further exploration . However, Chen and colleagues examined the educational factors that promote breastfeedin g initiation and duration , with a focus on the sources of education. 64 According to Chen et al , 64 shorter breastfeeding duration was associated with receiving breast pump educ ation from health professionals and longer duration was associated with receiving breastfeeding education from breastfeeding classes, support groups, friends and relatives. One study found that is a strong predictor of feeding choice more so than sociodemographic variables and having breastfeeding support . 65 Studies have also found that parents who choose to breastfeed their 19 infants had more positive breastfeeding attitudes and were more knowledgeable about the health b enefits . 12 45 66 It is hence evident that many factors beyond education and promotion of breastfeeding aff ect the infant feeding decisions and breastfeeding behaviors among mothers. 2 .3. 5 Misconception about F ormula Dating back to the 19 th century, formula manufacturers have been advertising and encouraging women to substitu te formula for breastmilk , implying that breastmilk alone is not enough for raising a healthy infant. Traditionally , they have equipped hospitals with free formula and n ewborn formula starter pack gifts for new mothers . From these marketing and advertisement schematics, some women have developed a misconception that formula is equivalent to breastmilk. Hannon et al , 67 found from their qualitative study that several Latina nutrition for their infants. Kaufman et al 54 found that a wom a was embedded from the generational use of formula , along with advertis ements and offerings encouraged by WIC and hospitals. Rosenberg and colleagues 68 found that women who received infant formula at discharge , exclusively breastfed for less time compared to women who had not received the formula. Additionally, s tudies have shown that African American women are more likely to receive formula in the hospital compared to white women . Many studies have concluded that receiving formula from the hospital is a modifiable risk factor for breastfeeding cessation. 60 69 A strategy that has been impl emented in an effort to reduce the unnecessary use of formula supplementation and increase the support of breastfeeding while in the hospital has been the creation of Baby Friendly Hospitals . 70 One study found that African American women had a higher comfort with formula feeding, suggesting that attitudes towards formula and breastfeeding should be considered when trying to understand breastfeeding intentions . 60 In addition, African 20 American communities have been targets of aggressive marketing by formula companies . While it is not thoroughly addressed within the literature , aggressive marketing an influence their decisions to breastfeed. 2 .3. 6 Maternity Care Practices Another determinant of breastfeeding behavior is the mother s experience during delivery , a major global movement launched the Baby Friendly Hospital Initiative that was established by WHO and UNICEF resulting in better breastfeeding outcomes. They implemented a set of maternity care practices known as the Ten Steps to Successful Breastfeeding . 71 The Joint Commission identifies these practices as a promising strategy to improve maternity care. Currently, 420 hospitals and birth ing centers in the United States have implemented the Baby Friendly Hospital Initiative practices to promote breastfeeding within the hospital. While maternity care practices are common and widespread, a mplementing the Baby Friendly Hospital with maternity care practices vary across age, income and race . Recent studies have found that healthcare providers with in the hospital are less likely to discuss breastfeeding with African American women and supportive maternity care practices are limited in communities with more African Americans. 8 37 41 In zip code s where more than 12.2% of residents were black, women were significantly less likely to have early initiation of breastfeeding and participate in r ooming in. They were more likely to have use of breastfeeding supplements, use pacifiers, and have limited post - discharge support compared to facilities in zip codes where less than 12% of residents were black . 72 These experiences influence breastfeeding behavior and may play a major role in the existence of the disparities in breastfeeding rates among racial groups. Although, there 21 and sustain breastfeeding, experiences during childbirth are one of the earliest influences and are critical for establishing breastfeeding. 2 .3. 7 Preference to Bottle Feed Mitra et al, 38 found that l ow - income, less educated women with larger families are more likely to bottle feed and that these women tend to be African American. Si milarly, Forste and colleagues 73 found that African American women were less likely to breastfeed and had a definite preference for bot tle - feeding . The authors understanding of this preference was limited given their survey structure. Besides preference for bottle - feeding , there are many other reasons studies have found that influence decisions to breastfeed. Wambach and Koehn found that a group of pregnant adolescents claimed that their decision to combine breast - and bottle - feeding was solely theirs, yet there was evidence that people from their social network played an influential r ole by either encouraging or discouraging breastfeeding . 74 circumstances, education and previous exper iences affe ct her choice of feeding method. 2.3. 8 Perceived Inconvenience ction to Support Breastfeeding stated that there is a large percentage of mothers perceiving br eastfeeding as an inconvenience . 18 Mothe rs believe that they may have to give up too many competing lifestyle habits with the commitment of breastfeeding. Brown and colleagues conducted a longitudinal study and found the most common reasons for not breastfeeding cited were inconvenience or fatig ue associated with breastfeeding . 75 In another study breastfeeding was perceived as time consuming and that bottle - feeding allows the mother an opportunity to regain a schedule compared to breastfeeding, making bottle - feeding more attractive . 76 It may not necessarily be that there is a preference to bottle feed, but a 22 bottle - feeding and a lack of practicality associated with breastfeed ing. 2.3. 9 Family and Social Support Successful breastfeeding depends on multiple factors related to the mother, infant, the community she resides and the level of support in her environment . Macro level factors often identified in the Social Ecological Model , influence breastfeeding and thos e include, but are not limited to community based programs, coalitions, organizations, schools, childcare centers, businesses, and the media . 40 T he Surgeon General s Call to Action to Support Breastfeeding confirms that these entities are important factors t hat either support or discourag e breastfeeding , and are crucial to /or sustain breastfeeding . 18 Fabiyi and colleagues conducted a qualitative study where 20 middle class African American wome n reported inconsistent support from health care providers and discouraging remarks from some family members and employers. 77 Authors reported that those discouraging remarks and diminished family support led some of the mothers to begin formula supplementation. S tudies have shown are important influence s on her decision to breastfeed and the continuance of breastfeeding . 12 78 - 81 Studies also continue to confirm that while the choice to breastfeed is ultimately the mothers , fath ers have an influential role in the choice to initiate or continue breastfeeding by acting as either a supporter , or a deterrent to breastfeeding by the mother . 40 81 82 For many women, social influences extend beyond the father , to the maternal grandmother and close fri ends . Peer support is an additional method for improving breastfeeding rates. Peer support can be provided in many different pract ices where trained mothers who have personal experience with breastfeeding , provide counseling, assistance, and support with breastfeeding . 50 A randomized control led trial using peer - based 23 (mother to mother) telephone support showed that peer volunteers provided assistance to mothers when difficulties were experienced, increased their confidence, decreased concerns, and assisted them in reaching their breastfeeding goals . 50 All of these individuals contribute to the breastfeeding success of that mother. Studies among socio - economically disadvantaged e of the most influential factors in the infant feeding decision . 40 83 2.3.10 Lactation Experience Evidence suggests that breastfeeding is a learned skill . W omen may struggle early on with the physical challenges associated with breastfeeding , such as latching issues that can result in painful nipples and uncomfortable feeding experiences. 84 Other women may overcome these physical issues but continue to struggle trying to integrate breastfeeding into modern life. but certain skills may need to be taught to assist in breastfeeding successfully. 85 86 W omen may initially expect breastfeeding to be relatively easy and then are faced with challenges. Scott et al ., 66 conducted a cohort study and found that women experiencing breastfeeding difficulties in the first 4 weeks was negatively associated with b reastfeeding duration . Another study found that women were more likely to cite within the first week postpartum (7.9%) or within one to six weeks (12.9%) than if they breastfed their infant for six weeks or more before stopping . 75 having with breastfeeding initiation, duration and formula supplementation . This misconception of an inadequate supply of breast milk may be due to a lack of confidenc e in breastfeeding. 18 24 In addition to breastfeeding difficulties, studies have cited a major chall enge with the discomfort many women experience while breastfeeding in public places . 67 87 88 Breastfeeding an infant in public has often been regarded as indecent and has a sexual connotation associated with it . Hannon et al . 67 found that there is not only a fear of pain, but also embarrassment with public exposure and uneas iness with the act of breastfeeding , all of which acted as barriers for minority teenagers who were considering breastfeeding . Other studies have stated women feeling uncomfortable and even embarrassed to feed in public . 67 87 88 The ability to breastfeed on demand is crucial for continued milk production , making this shaming an unfortunate experience that many women have to endure. This discomfort/ embarrassment to breastfeed in public may discourage women from continued breastfeeding even if they initially chose to do so , ultimately impeding their lactation experience. While educational materials focused on the benefits or health advantages of breastfeeding has great importance, knowing how to breastfeed , being comfortable with the act , and having education addressing the practical details of breastfeeding are also crucial to sustaining breastfeeding. 2.3. 11 Emp loyment and Child Care Many mothers return to the workplace during their infant s first year of life and face multiple challenges when they combine lactation with employment. The re is a considerable amount of literature indicating that m aternal employment is a critical barrier with a strong influence on breastfeeding duration and ofte n even breastfeeding initiation in all ethnic groups . 51 75 78 89 90 Gielen and colle a g u es 91 found that bre astfeeding cessation was significantly associated with women who worked more than 20 hours . Since many mothers are not able to stay at home for an extended period of time following birth , this further encourages them to stop exclusive breastfeeding in exch ange for formula. Women who are working full - time during pregnancy 25 often quit breastfeeding prior to returning to work, this is often around 6 weeks or 3 months, which corresponds to the U.S. maternity leave policies . 90 A study including in - depth interviews from 20 mothers , conducted by Gatrell , indicated that mothers who attempt to combine breastfeeding with paid work , con tinue to do so with difficulty. 92 Many African American s, in particular , experience difficulty or find it difficult to c ontinu e breastfeeding while returning to work , specifically because African American women tend to be in lower - income jobs, have shorter maternity leav e , and often return to less supportive work environments. 1 The National Conference of State Leg islatures reported that a s of 2017 , 28 states have laws related to breastfeeding in the workplace . P rograms are gradually being implemented across the nation to address this barrier. 93 While strides are being made to encourage the combin ation of work and breastfeeding, the timing of the return to work is still a concern where breastfeeding cessation is closely linked . This is true especially for low - incom e women working in occupations wit h short maternity leave options that make it difficult for them continue to breastfeed. 2.4 Social Support and Breastfeeding Social support is complex and multidimensional in nature. Early research has associated social support as a critical factor for individuals facing any challenge. Social support has been defined and measured in numerous ways. Kaplan defines social support as the degree to which a - personal interactions. 24 According to House (1981), social support is the functional content of relationships that can be categorized into four broad types of behavior: emotional, instrumental , informational, and appraisal support based on theoretical definitions. 94 The first type of support is e motional support that involves receiving empathy, love, trust, and care. Family members or friends may provide this experience to the individual. The second type of social support is instrumental support and it involves 26 receiving tangible aid or services that can directly support the person in need. The next ty pe of support is i nformational support that involves the receipt of advice, suggestions, and or information. This usually comes from a health care provider or professional. Appraisal support is the last , and it involves a person receiving affirmative words , or information that is considered as constructive feedback. While these broad areas of social support may be separated, relationships providing one type of support will also provide another type making it difficult to study them solely as separate constr ucts. F rom these original four broad types , the literature has adapted the use of three constructs of social support including tangible, emotional, and informational . Social support has been applied to several fields and topics including breastfeeding. Matich and Sims were one of the first to explore specific aspects of social support considered instrumental in the promotion and continuance of breastfeeding. Matich an d Sims 95 define tangible social support as the loaning of money, time, skills, and/or service to an individual, e motional support as providing affection , love, empathy, or acceptance, and informational support as providing guidance, advice, facts, and/ or knowledge to the individual. These three constructs were identified and are consistently mentioned within the literature as influential factor s and variable s for encouragement, success, and confidence in infant feeding outcomes . Early on , there had not been an exact consensus on the best approach for measuring breastfeeding social support, which led to the development of an instrument for measur ement in 1984 by Robbie Hughes . 96 Hughes developed an instrument to measure perceived emotiona l, instrumental, and informational support among breastfeeding mothers based on the theoretical definitions described by House and Cobb. 94 Following its developm ent , the content and face validity of this tool was evaluated in a pilot study of ten breastfeeding primiparous women . A panel of health care professionals established expert validity , and reliability scores were obtained 27 from a sample of thirty primiparous women. participants in this study. The questionnaire is comprised of 30 items. Of those questions, 10 relate to emotional support, 10 to instrumental support, and 10 to inf ormational support. McNatt et al, 97 utilized the Hughes Breastfeeding Social Support Instrument and found from examining a convenience sample of 45 primiparous women in Connecticut , that significant correlations existe d between the amount of health care provider support, informational support and perception of successful lactation. However, breastfeeding outcomes (duration, continuation, exclusivity) were not measured. This study assessed mothers perceived social suppor t in comparison to perceived feelings of breastfeeding at 4 - 6 weeks postpartum, which they found no statistical differences. This could be due to the lack of multiple evaluation periods, which may have been helpful in identifying differences between perce ived social support from networks in the early days of breastfeeding. In addition, this study could not be generalized because of the non - random convenience sample. However, this study does illuminate th e importance of pport network to include more qualified health care professionals which may increase women being satisfied with their breastfeeding experience. A number of studies conclude that social support play s a major role in the continuation or early cessation of breastfeeding and may vary depending on the wom a or culture . 8 95 98 social network such as the infant s father, family members, or friends to her formal social network , which includes support groups, lactation consultants, or physicians . Glanz and colleagues 99 noted that d iffering amounts and types of support may be experienced at different times and from different social support network members. In addition, the effectiveness of this support may depend on the actual source. S ocial support is likely t o vary depending on the 28 cultural context in which the woman participates , specifically her marital status, ethnic group, socioeconomic status, and employment . Bar ona - Vilar and coll eagues 100 found from their qualitative study that w omen from higher socioeconomic backgrounds highly considered their with decisions on breast - feeding . They also acknowledged great importance for formal health support, and employed mothers desired more institutional support. Among women from lower socioeconomic backgrounds, friends were the closest social network and had the greatest influence on feeding decisions. The study also found that y ounge r women without breast - feeding experience or the possibility of receiving tangible support from their mothers, wanted more practical health - care support such as providing skills in breast - feeding technique . 100 Another study determined that most women strongly agreed that while the support system was important, the decision to breastfe ed was still ultimately theirs and the decision to discontinue breastfeedin g was not due to lack of support, but because of the need to return to work or school. 82 Several studies have concluded that the plays an important role in influencing breastfeeding behavior. 40 82 95 Raj concluded from a r eview of t he literature , which assessed the role of social support in breastfeeding promotion, that social support that increases breastfeeding includes emotional, tangible, and educational components from informal and formal networks. 98 Raj also concluded that policies or programs pr oviding social support for breastfeeding should evaluate exi sting sources of support , as well as any barriers that the mother may face. This is also supported where the literature suggests that more p rograms , such as WIC , should focus their efforts on targeting the grandmother and male partners of African American women, since they influence breastfeeding. 83 More research and 29 programs will need to focus on the breastfeeding perceptions and experiences of significant others and family members . Much of the accumulated literature has utilized methodology where social support was primarily assessed qualitatively. While it is very important and may be most sufficient for identifying underlying relations , such as the quality of social support , i t is also very important to be able to quantify and illuminate which areas/ constructs of social support are p erceived as support ive . Perceptions of social support has been shown to be strongly linked to not only the he continuance of breastfeeding. It is necessary to continue to identify factors that may influence support perception as well as the timing/duration of such support . Work is still needed on identifying how this support is manifested. Many studies have l ooked at how perceived social support impacts breastfeeding intention, infant feeding decision and breastfeeding initiation. However, breastfeeding duration is a less commonly measured breastfeeding outcome . 101 With breastfeeding initiation rates beginning to meet the Healthy People 2020 goals , 17 more work is needed to increase breastfeeding duration and exclusivity rates to meet the Healthy People goals. With breastfeeding disparities still prevalent , this topic area is important specifically in understand ing how breastfeeding social support may be perceived, the elements that women perceive as supportiv e and how , or if they change over the course of pregnancy and the post - partum period , especially since it is a modifiable variable . 2.5 Breastfeeding Self Efficacy According to Bandura , 102 103 self - efficacy is defined as a motivat ion, thought processes, emotional states, and social environment in performing a specific 30 behavior. Self - efficacy represents the personal perception from outside experiences that is prominent in influencing the outcome of many events , making it an important aspect of . 103 Many studies within a variety of fields have used se lf - efficacy as a theoretical framework to explore, explain, and predict health behavior. Self - efficacy has been found to be a consistent predic tor of short term and /or long - term behavior change and maintenance, thus providing the brea stfeeding self - efficacy theory as a useful framework for studying and measuring the . 104 105 Bandura advocated that a measure of general self - efficacy was inadequate i areas associated with a specific behavior which resulted in Cindy Dennis creating the Breastfeeding Self - e fficacy Scale. Dennis defines breastfeeding self - efficacy, as a mother s perceived confidence in her ability to breastfeed , as measured by the Breastfeedi ng Self - Efficacy Scale (BSES). 13 106 The Breastfeeding Self - Efficacy Scale (BSES) is a 33 - item, 5 - point Likert - type scale instrument used to measure breastfeeding confidence . Coinciding with Bandura , 102 all of the items are aggregated creating a sum ranging from 33 to 165. Low scores indicate low levels of confidence in breastfeeding, while higher sc ores indicate a higher confidence in breastfeeding behavior. This instrument was tested and had a C .96. Items within the original BSES tool were found to be redundant and revised resulting in the creation of the Breastfeedin g Self - Efficacy Scale Short Form ( BSES - SF ) . It is a 14 - alpha coefficient of 0.94. Since the creation, multiple studies analyzing breastfeeding self - efficacy have utilized this tool created by De nnis. 15 19 104 107 - 109 31 In a recent quantitative study amo ng primiparous women examining breastfeeding self - e fficacy and breastfeeding outcomes, breastfeeding self - efficacy along with other factors was associated with higher levels of breastfeed ing exclusivity. 110 H igh breastfeeding self - efficacy was also associate d with having fewer depressive symptoms. Another study assessed 199 Chinese women at 6 weeks post - partum and found that women who were successful at breastfeeding exclusively till 6 months post - partum were (1) women who planned to breastfeed that long and (2) women with a high level of breastfeeding self - efficacy. 111 These studies illuminate that the modifiable factor, self - efficacy , is a strong predictor of breastfeeding duration as well as exclusivity. b reastfeeding s elf - efficacy and durati on; these studies are limited in the fact that the majority of their sample has been primarily Caucasian , and that breastfeeding duration has been the only outcome studied. Breastfeeding self - efficacy has been shown to predict breastfeeding outcomes (durat ion and exclusivity) in diverse po pulations of women including Canadian, 13 Australian , 15 Puerto Rican , 112 Asian, 113 and Pol ish 114 women. While these studi es have shown breastfeeding self - efficacy to be indicative of breastfeeding outcomes, the re is a paucity of data among African American women. McCarter - Spaulding 115 and colleagues examined breastfeeding self - efficacy among women of African descent and found that breastfeeding self - efficacy was a significant predictor of breastfeeding duration, which is consistent within the li terature. This was one of the first studies , and the only study to date , to examine predictability of breastfeeding self - efficacy on duration and breastfeeding pattern exclusively with Black women from multiple ethnic backgrounds. While 50% of their sample included African American women, their results were still specific to their entire sample and not generalizable to all African American women. 32 Interestingly, they did find that women who were born within the United States and identified themselves as Afri can American , had a lower mean of breastfeeding self - efficacy scores compared to the other women who identified as black. Overall, the majority of their sample had at least a college education, with only 21% having reported an education of high school or l ess. In addition, from their limited sample size they did not analyze differences from the ethnic groups. Lastly, McCarter - Spaulding assessed the relationship between breastfeeding self - efficacy a nd breastfeeding outcomes where breastfeeding outcomes was c lassified into six categories. The authors collapsed them down to three categories of exclusive breastfeeding , partial breastfeeding and bottle feeding/weaned. Almost exclusive and high breastfeeding were included in the category of exclusive breastfeeding causing ambiguity. Breastfeeding initiation or intensity was not addressed. Another study focusing on the African American population used a mixed - methods approach, which examined infant feeding choices and prenatal bre astfeeding self - efficacy from a black feminist perspective. Robinson et al. 116 found that breastfeeding self - efficacy has an impact on infant feeding intentions, which is similar to the results fo und within the literature. From their sample narratives of the black women , the results highlighted some of the factors that went into their decisions regarding infant feeding, which agreed with what has commonly been cited within the literature. The novelty of this study was the illumination of the perspectives from African American women. While, the narratives were novel , there was a small sample size of participants within the study and most of the wome n were multiparous with previous infant feeding experiences. In addition to the small sample size , it was not representative of all African American women with the majority of the participants being married, middle income ( $40,000 - $60,000 ), and having some college education. 33 Tools to measure breastfeeding self - efficacy have been developed, tested for validity and reliability, but they are limited to predominantly Caucasian women. Breastfeeding s elf - efficacy has been determined to be an important factor in breastfeeding behaviors, but it needs to be studied more among African American women. 2.6 Conceptual Framework Previous literature has proposed the use of the Socio - ecological model (SEM) in examining barriers related specifically to breastfeeding. 44 Bronfenbrenner (1986) originally proposed the human eco logical model as a theoretical perspective for understanding the relationship and underpinnings involved with breastfeeding. 117 The SEM model taken from Bentley and colleagues 40 , shown in Fig ure 1, provides a useful theoretical framework for addressing numerous challenges and obstacles that African American women face when integrated through a network of intr apersonal characteristics, interpersonal processes, institutional factors, community factors and public policy. 118 The model assumes that interactions between individuals and their environm ent are reciprocal and may overlap at levels. The intrapersonal level of the SEM is made up of individual knowledge, attitudes, beliefs and perceptions that influence behavior. The interpersonal level of the SEM involves cluding family, friends, peers and health care providers. The institutional level of the SEM comprises the health care system, policies, and structures working together to either assist or hinder breastfeeding. Multiple studies have utilized this framewor k to investigate and conceptualize breastfeeding linkages across the various levels within the SEM. 119 Using the SEM lens for this dissertation may lead to a better 34 understanding of how modifiable factors (perceived social support and breastfeeding self - efficacy) are involved and influenced from the micro to macro levels in the SEM. 2. 7 Gaps in the Literature The importance of breastfeeding, determinants of breastfeeding, racial and ethnic disparities in breastfeeding, breastfeeding social support and self - efficacy were discussed in the preceding in - depth literature review. The fact remains that African American women are least likely to initiate, continue, and exclusively breastfeed compared to all other ethnic groups within the US. On the surface, the act of breastfeeding seems straightforward and effortless. However, many underlying complex factors influence breastfeeding. Using the social ecological framework, factors within each level ranging from the micro level to the macro level can either support or discourage breastfeeding. These levels can interact with each other and influence the breastfeeding. The macro level factors include factors such as media, aggressive marketing of formula, and structural policies. Micro level factors addressed within the review of the literature include social support networks, workplace support , knowledge, and cultural beliefs. W hile we have gained a wealth of knowledge from the literature, more attention should be given to less studied, modifiable factors such as social support and self - efficacy. The literature has identified these modifiable factors and the impact on breastfeedi ng outcomes specific to breastfeeding intent and initiation, but more work is needed on understanding how they influence breastfeeding duration , exclusivity , and intensity . In addition, authors have suggested that studies of modifiable risk factors focusin g on African American women are particularly sparse . Thus, highlighting the need for more studies focused with in the context of this population and better understanding 35 breastfeeding from their perspective , thus a mixed - methods study can add significant value to the literature. While there is a growing amount of literature regarding the reasons why African American mothers cease breastfeeding earlier than recommended, there is a paucity of information about how those reasons are influenced by modifiable factors of social support and breastfeeding self - efficacy. With breastfeeding disparities still prevalent this topic area is especially important to improve our understanding of breastfeeding self - efficacy and soc ial support and how these variables could be utilized to improve breastfeeding outcomes. With regard to breastfeeding self - efficacy, many studies have shown that maternal confidence is a significant modifiable factor that influence s breastfeeding outcomes , but it has not been extensively studied among African American women. 13 Social support is also another variable reported as being modifiable . I t is important to gain a better understanding of the representations of support within the different socio - ecological spheres (hospital, family, work) , It is important to gain a better understanding of who African American women perceive as su pportive and how or if they change over the course of pregnancy and the post - partum period . Other studies have also concluded and suggested the importance of having prenatal and postpartum intervention studies designed to prolong breastfeeding duration, w hich may be particularly beneficial during the critical periods (early postpartum) of breastfeeding. Understanding how modifiable factors specifically breastfeeding self - efficacy and perceived social support influence breastfeeding patterns , including dur ation and impact on exclusivity by measuring breastfeeding intensity in an intervention study , is an important step in the development of effective breastfeeding promotion efforts, whether community wide or targeted 36 to focus on the individual woman before, during, and after pregnancy. By conducting a thorough review of the literature, gaps that jus tify the rationale for this study including the use of both quantitative and qualitative approaches have been identified. 37 Figure 1 . 1 Social Ecological Framework From Bentley and colleagues 40 38 3.0 Chapter 3 - Methods Overview T his study was done in conjunction with the study titled: Mothers Allied with Mothers Arou nd Breastfeeding Encouragement a nd Return to Health After Baby ( Mama Bear ), a pilot dual intervention (R21HD085138 - 02) . The Mama Bear study was funded by the National Institutes of Health (NIH R21) and is a gr oup project collaboration between Henry Ford Health System (HFHS) and Michigan State University (MSU). Mama Bear is a randomized control led trial that tested an intervention which incorporate d a postpartum weight management component and an established bre astfeeding peer - support program. The overall goal of this developmental/exploratory study was to gather pilot data to refine the intervention so it could be tested in a larger, longer study using a factorial design in a future R01 phase. This study was aff iliated with Henry Ford Health System (HFHS). Recruitment occurred in a large inner - city prenatal care clinic affiliated with HFHS in Detroit, MI . African American pregnant women (~28 gestation) who were considering breastfeeding (n= 53 ) were enrolled and randomize d into either the intervention or usual care group and were follow ed up to 20 weeks postpartum. Mama Bear has specific objectives and aims: 1) assess feasibility, 2) evaluate acceptability and, 3) estimate the preliminary effect size of the intervention at 20 weeks postpartum relative to the usual care group on breastfeeding duration and postpartum weight retention. This project is significant because the combined intervention , of postpartum weig ht management and breastfeeding, is designed to address two interrelated, highly prevalent problems that disproportionately affect African American families . This dissertation was conducted using a mixed - methods design that followed the Mama Bear study des ign including recruitment, data collection, intervention, project flow, and 39 incentives. Figure 3. 1 shows the Mama Bear study design flow chart displaying a timeline of consent, recruitment, data collection, intervention and incentives . The data obtained address es Aims 1 and 2 (quantitative) of the dissertation and was collected in conjunction with the Mama Bear Study . T he questionnaires for the se aims were submitted along with other Mama Bear study materials and approved I nstitutional Review Boards (IRB) and can be found in Appendix A . Aim 3 (qualitative) require d a revision to the current IRB where there w as a two - stage recruitment process. A im (3) includes a 6 - month follow up of the Mama Bear project with the aim to gain a better understanding o f personal experiences of African American women and how breastfeeding social support and self - efficacy may influence the success of breastfeeding goals and overcoming barriers . Among African American women, we were especially interested in who or what hel ps the process of breastfeeding. Community health workers (key informants) who have worked with Mama Bear participants and other African American women in the community were invited to participate in a semi - structured interview (Stage 1). Participants from the Mama Bear pilot intervention study, IRB#16 - 297, were invited to participate in individual semi - structured qualitative interviews after 20 weeks post - partum (Stage 2) . Originally it was proposed that Mama Bear participants would be invited back to part icipate in a focus group session via a letter of invitation outlining the study for the focus group session (Stage 3) . However, because of the loss due to follow up from the Mama Bear study , it was difficult to obtain enough participants to reach data satu ration for the semi - structured interviews . I t was hence decided to include the additional questions from stage 3 into the stage 2 semi - structured interview questionnaire. T his complementary mix of data collection provide d rich information from objective to broader level deep discussions that can be applied to efforts to advanc e bre astfeeding among African American women. 40 This chapter describes the sample and data collection procedures; the sur vey instruments, define s the outcome measures , discuss es analyses and address es any changes to the originally proposed analyses. This study was approved as an addendum to the originally approved parent study by the IRB at MSU and Henry Ford Health System . The primary researcher f rom the Department of Food Science and Human Nutrition at Michigan State University worked closely with the Primary Investigators from Mama Bear and Henry Ford Health System (HFHS) , in Detroit, MI. 41 Figure 3.1 Mama Bear Study Flowchart CHW= Community Health Workers NCO= New Center One Clinic 20 weeks postpartum visit (n=17) Location: Home Data collection completed Incentive: $50.00 Completed Step 1 of Enrollment (n=65) ~28 weeks gestation NCO Clinic: Assessed for Eligibility; Consent; Weight; Blood; Urine Incentive: $30.00 Usual Care Group (n=25) ~32 - 36 weeks gestation Incentive: $20.00 mailed Intervention Group (n=28) ~32 - 36 weeks gestation Incentive: $20.00 in person 2 - 6 weeks postpartum visit (n=16) Location: NCO Clinic Data collection completed Intervention with CHW/Peer Counselors Incentive: $30.00 2 - 6 weeks postpartum visit (n=18) Location: Home/NCO Clinic Data collection completed Incentive: $30.00 Appr oached for Screening (n=86) Completed Step 2 of Enrollment (n=53) ~29 - 32 weeks gestation Baseline Survey Data Collection: Phone Not interested/ eligible (n=21) Did Not Complete Baseline Survey Required for Randomization (N=12) 20 weeks postpartum visit (n=17) Location: Home Data collection completed Incentive: $50.00 42 3. 1 Aim 1 Methods 3.1 .1 Approach for Aim 1 The first a im of this study was to determine how breastfeeding self - efficacy impacts breastfeeding initiation, duration, and intensity. T he validated Breastfeeding Self - Efficacy Scale Short - Form ( BSE S - SF) questionnaire created by Cindy Lee Dennis 13 was used to measure breastfeeding self - efficacy of women enrolled in the Mama Bear program at three time points ( late - pregnancy , two - six weeks post - partum , and twenty weeks post - partum ) . The association s between self - efficacy and the outcome variables : breastfeeding initiation, duration, and intensity were of specific interest to observe the differences with in or between the control and intervention group. The rationale underlying this ai m is to gain a better understanding of breastfeeding self - efficacy and determine if there is a critical value of breastfeeding self - efficacy where breastfeeding exclusivity and/or weaning may occur . The goal was to gain a better understanding of breastfeeding behaviors that lead to exclusivity. The findings were also intended to inform interventions and clinical settings to identify and tailor messages for women at early risk of breastfeeding cessation by utilizing the BSE S - SF tool . 3. 1 .2 Research Design Aim 1 was done in conjunction with the Mama Bear dual intervention (peer support and weight retention) pilot study . Parallel to this longitudinal randomized control study; this aim specifically examine d breastfeeding self - efficacy over the Mama Bear study period. T he BSE S - SF was collected at three different time - points to examine the effects of the outcome variables breastfeeding initiation, duration, and intensity. 43 3. 1 .3 Sample and Data Collection Procedures Henry Ford Health S ystem (HFHS) and MSU staff members we re integral in the facilitation of the Mama Bear Study. They include d : 1)Two HFHS clinic and hospital registered nurses who were trained on the Mama Bear study and recruitment protocol to screen and enroll participants in the study; 2) Two experienced HFHS r esearch a ssistants and one MSU research assistant , with college level education , were trained on the Mama Bear study and data collection protocol for which they collect ed information over the phone and in home; 3) Two experienced HFHS community health workers , with a high school education or equivalent , were trained on the Mama Bear i ntervention protocol and were provided with ed ucation materials to implement the intervention arm by acting as peer counselors and health coaches to provide support for breastfeeding and weight management ; 4) One International Board of Lactation Consultant (IBCLC) was hired through MSU to provide brea stfeeding support and assist/answer any technical breastfeeding questions that peer counselors may have . The community health workers are trusted members of the community and were trained on home visiting methods, as well as additional ways to provide advi ce, resources and community referrals. Eligibility criteria include d pregnant African American women who : 1) were 24 - 32 gestation who are scheduled for the one hour glucose challenge test that visit, 2) had no known fetal or chromosomal anomalies, 3 ) were fluent in English, 4) were interested in breastfeeding, 5) and had reliable internet access ( via smartphone, tablet, or computer). If eligible , the woman was approached by a nurse during their prenatal visit at the Henry Ford New Center One Clinic in Detroit, MI. Exclusion c riteria include d breastfeeding contraindications i.e. (HIV positive status, chronic therapy with medications incompatible with breastfeedi ng such as ( cyclophosphamide, chloramphenicol, metronidazole, nitrofurantoin, antithyroid drugs, 44 psychotropic drugs, and radiopharmaceutical s ) ; known drug or alcohol abuse ; active untreated tuberculosis); underweight women with a pre - pregnancy BMI < 18.5 k g/m 2 ; and any acute or chronic illness for which general healthy eating and physical activity would be contraindicated e.g. as determined by the attending obstetrician. Aim 1 used the Mama Bear study design as approved by the IRB at M ichigan State University and Henry Ford Health System that is attached in Appendi x A . Mama Bear staff, including nurses and data collectors, screen ed for eligible women, recruit ed and obtained their co nsent to participate in the Mama Bear Study. The consent form can be found in Appendix B . Following recruitment and consent, baseline d ata collection occurred with two data collectors via phone using a telephone script. The telephone script drafted for data collectors is provided in Appendix C . D ata collectors ask ed participants a number of health questions, many related to pregnancy, diet, and breastfeeding that were recorded into Michigan State University Biomedical Res earch Informatics Core (BRIC) Research Electronic Data Capture ( REDCap ) system . REDCap 120 is a secure web application for building and managing online survey s and database , which was used to collect and store the Mama Bear data . The primary researcher of this study , from MSU, managed and maintained the data from in the REDCap system. The primary researcher provided trainings and was available to provide immediate assistance to others utilizing the REDCap system. A variety of data was collected for Mama Bear . In addition to the other data specific to the Mama Bear study , demogr aphics and BSE S - SF were collected which are specifically related to aim 1. Following the completion of the baseline survey, participants were randomized into either treatment or control group using a block randomization scheme, created by MSU BRIC data man agers . The r andomization scheme is attached in Appendix D . After randomization, a Community Health Worker (CHW) / Breastfeeding Peer Counselor 45 contact ed participants in the intervention group to begin the intervention . P articipants , randomized into the control group , receive d a letter including an incentive of $50 and continue d with usual care as displayed in Figure 3.1 . S eparately from the intervention , between two to six weeks and twenty weeks postpartum , all participants had their data collected in person by data collectors and receive d an incentive at each of those data collection time points . View Table 3.1 for the list of project incentives at various data collection points. The primary researcher of th is study was involved and attended all core meetings, managed the data, assisted with quality assurance of the data, and attended/ assisted in trainings for data collectors and intervention staff. Figure 3. 1 shows the Mama Bear study design that aim 1 foll ow ed . The data collected specific to this study include d : Demographic c haracteristics (age, parity, marital status, socioeconomic status, source of health care, education, employment status, mode of delivery ) , participation in the Supplemental Nutrition Program for Women, Infants, and Children (WIC), b reastfeeding c haracteristics (previous breastfeeding experience, intended breastfeeding duration, breastfeeding knowledge and attitudes) b reastfeeding self - effic acy , b reastfeeding s ocial s upport, and b reastfeeding i ntensity. All ques tionnaires can be found in Appendices E - L . Data was collected in - person, by telephone, text, and via medical records in the same way across the two study arms at multiple time points. At all the time points, the Breastfeeding Self - Efficacy Scale Short Form (BSE S - SF) and Breastfeeding i ntensity (once baby was born) was collected. Table 3.2 shows the Variable Measurement and Data Collection Schedule . The literature was limited with regard to power analyses for assessing breastfeeding self - efficacy on the main breastfeeding outcomes of interest (breastfeeding intensity and duration ) . Studies that have solely utilized the BSE S - SF have looked at reliability and predictive validity of 46 the instrument administered, but have not attempted to detect power or effect using priori or post hoc hypothese s . 13 106 - 108 Other studies assessing the impact of breastfeeding self - efficacy in interventions on breastfeeding outcomes did not conduct a power analysis based on the r easoning that they were pilot studies , with no previous research . Becaus e the re are minimal studies published within the li terature where power analyses were conducted using breastfeeding self - efficacy scale (BSE S - SF) to measure breastfeeding duration and intensity , the effect and power size are based on what is typically observed in the field. Originally proposed u sing G - power 3.1 software , the sample size of 85 was calculated for a medium effect size of 0.15 at 80% power and type I error of 0.05 for the linear regression model . A 10% attrition rate was estimated if a medium effect size was used resulting in 94 participa nts. The linear regression model is wher e Y= BF outcome, =exposure (BSE) , z= control variables. The hypothesis t o be tested is vs The parameter is related to the partial correlation between ( Y , ) controlling for z , . Therefore, when standardized, , the regression coefficient is precisely the partial correlation. Another interpretation of is comparing expected outcome for one unit increase in exposure, all else fixed: . It is customary t o , for example, / SD ( Y | z ) . 47 For the Mama Bear pilot study, a total sample size of 80 women was pre - determined to detect at least a moderate effect size of 0.65 with a power of 80% based on research objectives approved in the NIH funded grant. The context was a 2 - sample t - test, the effect size was the scaled mean difference between two intervention groups. Using labels =1 for one group, and =0, for the other group, the effect size is / SD ( Y ) where . Sample size ca lculation A two - sided test of vs at significance level (type I error) 0.05, would require a sample size of 82 subjects to detect a partial correlation of at least 0.30. To detect a partial correlation of at least 0.40, keeping all other design features the same, requires a sample of 44 subjects. However, during the pilot study there was a lull in recruitment, progress of the study and data collection determining the need to change overall study efforts. By the end of the Mama Bear study, we recruited less than originally planned (n= 65 ) and lost ad ditional participants due to follow up , as shown in Figure 3.1 . Although, priori hypothes e s was formulated as a proposed plan, this study was done in conjunction with the Mama Bear study and followed the same protocols resulting in a sample size of 5 3 completing baseline data collection . 48 Table 3.1: Project Incentives Incentive Description Value/ Type Delivery of Incentive Recruitment and Consent $30 gift card New Center One Clinic at Henry Ford Health System (HFHS) Baseline/ Prenatal Breastfeeding Support Meeting * $20 gift card Mailed/ Provided in - person 2 - 6 weeks post - partum $20 gift card In home/ New Center One Clinic (HFHS) 20 weeks post - partum $50 gift card In home *Incentive s were mailed to those randomized to the usual care group and provided in - person by the Community Health Worker/ Peer Counselors to participants assigned to the intervention group. Table 3.2 Aim s 1 and 2 Timeline s and Variable Measurement Schedule Variable Measurement ~28 weeks N=86 Baseline N=65 2 weeks post - partum N=53 6 weeks post - partum N=34 20 weeks post - partum N=34 Eligibility and Consent Breastfeeding Intention, Internet/ Mobile Device Access Demographics, Family History, Breastfeeding Knowledge and Experience Breastfeeding Self - Efficacy Scale Short Form (BSES - SF) Breastfeeding Social Support (HBSS) Breastfeeding Intensity 49 3. 1 .4 Survey Instruments and variables Demographic variables (Appendix E) collected were based on the review of literature showing previous linkages to breastfeeding outcomes for Aim 1 , which include: age, marital status, education, employment status, mode of delivery, and participation in the Supplemental Nutrition Program for Women, Infants, and Children (WIC) . B reastfeeding characteristics (Appendix F) such as previous breastfeeding experience, inte nded breastfeeding duration , breastfeeding knowledge and attitudes were also collected at baseline. Breastfeeding Self - Efficacy was measured at late pregnancy, two - six weeks and twenty weeks post - partum using the validated Breastfeeding Self - Efficacy Scale Short Form (BSES - SF) copyrighted by Cindy Lee Dennis . The original questionnaire can be found in Appendix G. P ermission was electronically obtained for use of this form by the creator/author Cindy Lee Dennis . The verb tenses on the form were modified only when assessed at late - pregnancy , prior to breastfeeding (future tense) (Appendix H) , and after breastfeeding ha d stopped (past tense) (Appendix I) only for the enhancement of participants comprehension and for the data collectors understanding during interviews . The BSES - SF includes 14 questions using a 5 - point Likert scale to measure breastfeeding self - efficacy as a single construct . Responses range from 1 (not at all confident) to 5 (very confident). Scoring was done by adding the rating for the items to achieve a total score. The total scor e produce d a range from 14 to 70 , with low scores indicating that parti ci pants have a lower confidence in the ability to breastfeed and high scores indicating that there is a higher confidence in the ability to breastfeed . Factor Analysis of the instrument was conducted by Dennis and it was determined that it was unifactor i al . 13 Predictive validity was determined and found to be significant (p=.04) . The authors reported a coefficient for the instrument of 0 .94 . Breastfeeding Self - Efficacy for aim 1 analys i s was left 50 continuous, as it was originally coded . Review Appendices G - I for original and modified versions of the instrument . The definition of breastfeeding duration within the literature has been suggested t o lack consistency and precision resulting in misinterpretation and problems with comparability between studies . 21 121 122 123 With the lack of consistency , s tudies have begun using breastfeeding intensity to provide a more accurate comparison for defining breastfeeding duration. 21 F or clarity and to gain a better understanding of breastfeeding behavior for these aims , breastfeeding intensity was used to measure the main outcome variables. Breastfeeding Intensity was determined at two data collection time point s ( between two - six, and twenty weeks post - partum) . ce the last interview), how much For this question, the outcome variabl e breastfeeding initiation was coded into a nominal variable . Breastfeeding duration was calculated from the your infant was last To assess breastfeeding i ntensity , it remained coded as an ordinal variable . While, t he items in this breastfeeding intensity scale are ordered and hav e seven levels theoretically, it can be analyzed as a continuous variable . 124 125 - 127 Table 3 .3 outlines the coding of the outcome variables from the breastfeeding intensity question . The b reastfeeding i ntensity question is provided in Appendix J. 3. 1 .5 Analyses Using Stata version 14.0, (College Station, TX) data normality (histograms, P - P plots, Shapiro Wilk test) and assumptions were checked, and results were reported with descriptive statistics. Descriptive statistics were utilized for demographic variables and breastfeeding 51 characteristics based on previous linkages established in the literature. Frequencies and the sample mean s, as approp ria te, of each breastfeeding outcome of interest was correlation s and scatter plots were used to examine the association and check for linearity. Binary Logistic Regression was an optional procedure t hat was utilized for research question 1 and was contingent upon acquiring the actual data. With 75 % of the participants initiating breastfeeding (n=40) ther e were not enough in each category to conduct the analyses . For research questions 2 and 3, a full factorial model of linear regression was used to determine if there was significance of the interaction effects and main outcomes of interest. S tepwise selection was used to decide which covariates would be the best explanatory variables for all the models. Following each regression model , assumptions of the linear regression were assessed . Th ey i ncluded checking the linear relationship, normality, checking for multicollinearity, and homoscedasticity. Research question 4 was originally proposed utilizing a mixed two - way ANOVA for analyses. Following proposal committee suggestions, it was decided to remove this question. See Table 3.4, which outlines Aim 1 research questions, statistical analyses, outcome variables, and the equation model. 52 Table 3.3 Breastfeeding Intensity Question (Variables and Coding) Outcome Variables Coding of Question Variable Type Breastfeeding Initiation No = 100% artificial milk or solids Yes = 100% breast milk, 80% breast milk, 50 - 80% breast milk, 50% breast milk, 20 - 25% breast milk, or 20% breast milk Yes or No (nominal) Breastfeeding Duration date when infant was last breastfed date when infant was born= Number of days infant was b reastfed Continuous - Number (Interval) Breastfeeding Intensity 0=100% artificial milk or solids, 1= 20% breast milk, 2= 20 - 25% breast milk, 3= 50% breast milk, 4= 50 - 80% breast milk, 5=80% breast milk, 6= 100% breast milk Ordinal or Continuous 53 3. 2 Aim 2 Methods 3 .2 .1 Approach for Aim 2 Specific aim 2 assess ed how social support impacts breastfeeding self - efficacy , initiation, duration, and intensity. T his aim also examine d whether social support and breastfeeding self - efficacy have a relationship, and how social support and breastfeeding self - efficacy together impact breastfeeding (initiation, duration , and intensity). Aim 2 use d the validated Hughes Breastfeeding Support S cale (HBSS) questionnaire created by Robbie Hughes 96 to measure breastfeeding social support f rom women enrolled in the Mama Bear pilot study at three time points ( late - pregnancy , two - s ix, and twenty weeks post - partum). The HBSS survey was used to examine the association s among the outcome variables breastfeeding initiation, duration, and Table 3.4: Aim 1 Research Questions and Statistical Analysis Research Question Statistical Analysis Outcome Variable Equation Model 1) Does breastfeeding self - efficacy assessed in late pregnancy, prior to the breastfeeding support intervention predict breastfeeding initiation, regardless of group assignment in the RCT? Binary Logistic Regression Y=Breastfeeding Initiation, p = probability of BF initiation + * (BSE) 2) Does breastfeeding self - efficacy predict breastf eeding duration for the participants in either group? Linear Regression Y=Breastfeeding Duration + * (BSE) + * (Group) + * (BSE * Group) + BSE + GROUP + BSE * GROUP 3) Does breastfeeding self - efficacy predict breastfeeding intensity for the participants in either group? Linear Regression Y=Breastfeeding Intensity + * (BSE) + * (Group) + * (BSE * Group) + BSE + GROUP + BSE * GROUP 54 intensity . In addition, the constructs from the b reastfee ding social s upport survey were analyzed separately to determine if any of the constructs influence d breastfeeding behavior. The rationale underlying this aim was to gain a better understanding of breastfeeding social support that is perceived to be helpful to the mother during pre gnancy and in the post - partum period . The completion of the research in this aim was intended to enhance our understanding of how breast feeding social support may be perceived, the elements that women perceive as supportive , a nd gain a better understanding of the representations of support within those different spheres (hospital, family, work) that impact a wom a 3. 2 .2 Study Design Aim 2 , completed in conjunction with the Mama Bear Project , follow ed the same recruitment and data collection procedures as described previously in Aim 1 . This aim specifically focus ed on Breastfeeding Social Support ( H BSS) , as well as H BSS in conjunction with Breastfeeding Se lf Efficacy . 3. 2 .3 Sample Procedures Aim 2 follow ed the same sample procedure s as described in Aim 1. At thr e e - different time points , the Breastfeeding Social Support Scale (HBSS) and Breastfeeding Intensity data (as described in aim 1) , were collected. Refer to Table 3. 2 where the timeline and the variable measurement schedule is displayed. 55 Sample size calculation T he sample size of 80 women was predetermined for the Mama Bear study , aim 2 originally proposed and use d an priori hypothesis to determine the sample size required to have a medium to large effect with a power of .80 based on the previous literature assessing social support and the main breastfeeding outcomes of duration a nd intensity . Using G - power 3.1 software , A two - sided test of vs at significance level (type I error) 0.05 , would require a sample size of 82 subjects to detect a partial correlation of at least 0.30. To detect a partial correlation of at least 0.40, keeping all other design features the same, require d a sample of 44 subjects. 3. 2 .4 Instruments and variables The Hughes Breastfeeding Support Scale (HBSS) was used to measure breastfeeding social support at three different time points ( late - pregnancy , two - six, and twenty weeks post - partum ) . This instrument is provided in Appendix K. The instrument cons ists of a 30 - item questionnaire with a Likert - type scale that focus on different types of breastfeeding support a mother is receiving . Of the 30 - item questionnaire, there are three different constructs each consisting of 10 questions (emotional, informati onal, and instrumental support) , which were analyzed together and separately . Responses range from 1 (none at all) to 4 (as much as I wanted). Scoring was done by adding the item ratings from each of the three categories achieving three sub - scores. In order to obtain a total support score of 30 for each participant, the three sub - scores were added together. The total score was used for data analysis in this study . Low scores indicate inad equate breastfeeding support and high scores indicate adequate breastfeeding support. 56 T o capture additional details for each question , the primary researcher adjusted the HBSS. For each social support question, participants were asked to choose a maximum o f three individuals who they consider ed most helpful / supportive. Participants had to choose them in order , supportive of the three. If the participant answer ed , then this additional question was not asked. The options available for participants to choose from are s ignificant other, mother, sister, close friend, nurse, other relative, lactation consultant, pediatrician, obstetrician , peer counselor, and other person (open field) . Open fields were also available on REDCap to obtain additional indiv iduals not provided in the list. Review Appendix L for modified HBSS questionnaire. Although , not included in Aim 2 as one of the main outcomes, data on the top three individuals who provide d breastfeeding support w ere assessed and descriptively reported. For this aim, breastfeeding social support and b reastfeeding self - efficacy were analyzed as predictor variables. Breastfeeding social support was aggregated to provide the total score and remain ed coded as a continuous variable. Breastfeeding self - efficacy was aggregated and remain ed coded as a continuous variable. The Breastfeeding Intensity Question was collected at time points (two - six weeks post - partum , and twenty weeks post - partum). This quest ion remained as coded and previously described in Aim 1 instruments and variables procedure , to provide the outcome variables (breastfeeding initiation, duration, and intensity). 3. 2 .5 Analyse s Using Stata version 14.0, data and assumptions were first checked . Frequencies and assumptions were tested for each breastfeeding outcome of interest and correlations and scatter plots were used to examine the association and check for linearity. T he 57 data was determined linear. Binary Logistic Regression was an optional procedure that w as utilized for research question 1 and was contingent upon acquiring the actual data. With 75 % of the initiating breastfeeding , ther e were not enough in each category to conduct the analyses ( n=40 ) . For research questions 2 and 3 a full factorial model of linear regression was utilized to determine if there is significance of the interaction effects and m ain outcomes of interest. If it was determined that the interaction effects were not significant, then the main outcomes were reported. S tepwise regressi on was used to decide which covariates would be the best explanatory variables for all the models. Foll owing each regression model , assumptions of the linear regression were assessed. T ests included checking the linear relationship, normality, checking for multicollinearity, and homoscedasticity. Research question 4 was originally proposed utilizing a mixed two - way ANOVA for analyses. Following proposal committee suggestions, it was decided to remove this question. See Table 3.5, which outlines Aim 2 research questions, statistical analyses, outcome variables, and the equation model s . 58 3. 3 Aim 3 Methods 3. 3 .1 Approach for Aim 3 For Aim 3 , a 6 - month follow up was conducted where the knowledge from Aim s #1 and #2 was utilized to gain a better understanding qualitatively of experiences , and specific factors influencing African American women to meet breastfeeding goals . Aim 3 determine d if and how breastfeeding social support and self - efficacy from the perspective of African American women is likely to influence the success of breastfeeding goals and overcoming barriers . There Table 3.5: Aim 2 Research Questions and Statistical Analysis Research Question Statistical Analysis Outcome Variable Equation Model 1) Does Social Support or Breastfeeding Self - Efficacy, or the interaction between BSS and BSE, measured in late pregnancy; prior to the breastfeeding support intervention predict breastfeeding initiation? Logistic Regression Y= Breastfeeding Initiation , p= probability of BF initiation + * (BSS) + * (BSE) + * (BSS * BSE) 2) Does Social Support and Breastfeeding Self - Efficacy predict breastfeeding duration for the participants? Linear Regression Y= Breastfeeding Duration + * (BSS) + * (BSE) + * (Group) + * (BSS*BSE*Group) + BSS + BSE + GROUP+ BSS*BSE*GROUP 3) Does Social Support and Breastfeeding Self - Efficacy predict breastfeeding intensity for the participants? Linear Regression Y= Breastfeeding Intensity + * (BSS) + * (BSE) + * (Group) + * (BSS*BSE*Group) + BSS + BSE + GROUP+ BSS*BSE*GROUP 59 w as a two - stage process of interviewing, starting with community health workers which based on responses , guide d additional questions for the individual interviews with Mama Bear partic ipants who completed the study. Hypothesis: It was hypothesized that African American women w ould have unique experienc es and factors relative to social support and self - efficacy influencing breastfeeding behavior (duratio n and exclusiv ity). Objectives: 1. Gain a better understanding of personal experiences of African American women and how breastfeeding social support and self - efficacy may influence the success of breastfeeding goals and overcoming barriers. 2. Identify new ideas or knowledge that can assist in the support and success of breastfeeding for the benefit of African American women who want to breastfeed. 3. 3 .2 Study Design Aim 3 us ed a qualitative approach with semi - structured interviews following the participation of the Mama Bear pilot in tervention. I nstitutional R eview B oard (IRB) approval from Michigan State University was obtained. 3. 3 .3 Sample Procedures Recruitment: There was a two - stage recruitment process for participating in this study. Stage 1 Community health workers (key informants) , who ha d worked with Mama Bear participants and other African American women in the community , were invited to participate in a semi - 60 structured interview via a letter of invitation ( Appendix M ) outlining the study. After the letters were sent, the key informants were contacted by the primary researcher using a script ( Appendix N ) invit ing them to participate in this study. If the key informant agree d to participate over the phone, an in - person meeting was scheduled for the primary researcher to obtain formal consent ( Appendix O ) and conduct interviews. The option was given to the key informant to conduct the interview over video conference, if preferred. Stage 2 Participants from the Mama Bear pilot intervention study, IRB#16 - 297, were invited to participate in individual semi - structured qualitative interviews after 20 weeks post - partum. The data collector at the 20 - week post - partum visit ask ed the participant for permission to be contacted for future studies ( Appendix P ) and the participants we re given an invitation outlining this study (Appendix Q ). Following the visit and approval of permission to make contact, the participant from the Mama Bear study was called by the primary researcher inviting them to participate in the study using a script ( Appendix R ). If the participant agree d to participate in the study , they were provided the option when scheduling for the interview to either have it conducted over the phone, video conference, or in person to allow the participant flexibility and comfor t. If the participant agreed to conduct the interview over video conference or phone, the formal consent was read verbatim during the time of meeting and the participant was allowed to agree verbally. If the participant scheduled an in - person meeting , the primary researcher obtain ed formal written consent ( Appendix S ) and conduct ed interviews around 6 months post - partum. A reminder call was made to the participant a day before the scheduled interview. Consent Procedures: 61 Verbal consent was audio reco rded and the primary researcher email ed and mailed a copy of the consent form to participants to keep for their records. Prior to beginning the interview, the primary researcher read the consent form to participants via video conference or phone and ask ed for verbal consent using a specified script ( Appendix S ) to participate in the study. Incentives: Incentives in the form of $30 gift cards (Walmart or Meijer) were provided following the completion of the semi - structured individual interviews (key infor mants and participants) . For the community health worker interviews (Stage 1), two participants were interviewed for the study. For the individual interviews (Stage 2), fourteen participants were interviewed for the study , which is also where data saturat ion was reached. F ocus group session s (Stage 3), were not conducted since we were concerned about obtaining enough participants for this component of the study. 3. 3 .4 Instruments and variables Utilizing grounded theory 128 s emi - structured questions were developed based on aspects of the breastfeeding social support and self - efficacy questionnaire s to gain a better understanding of Mama Bear participant experiences and s pecific factors affecting African American breastfeeding duration and exclusivity. The primary researcher conduct ed the interviews with a paper - based interview guide that was prepared with a list of questions and topics to be discussed (Appendi ces T and U) . Data Collection: At each visit, we beg an with obtaining informed consent and the primary researcher conduct ed the semi - structured interviews either in - person, over the phone, or via video conference with participants using a paper - based interview guide. All interview guides are 62 attached for review ( Appendices T and U) . A demographic questionnaire was completed with participants following the interview ( Appendix V ). E ach interview was recorded with a digital tape recorder and the primary researcher took additional field notes. Towards the end of the interview, the primary researcher debrief ed with the participant s allowing them the opportunity to add or clarify information and/or correct misunderstandings. Following the interview, the recording was transcribed verbatim by the primary researcher after data was stripped of any identifiers to maximiz e confidentiality and minimize risk to individual participants. All digital and document files were stored on enc rypted and password protected university sites. All data will be retained per university data retention policies. The objective behind this design for the qualitative portion of th e study was to gain more in - depth perceptions and enhance the interpretati on of the results from the quantitative analyses . The individual participant semi - structured interviews were appropriate since the timing of the interviews need ed to be staggered because not all women reach ed 6 months post - partum concurrently. It also all ow ed the participants the freedom and privacy to express their views in their own terms providing reliable, comparable qualitative data. It was difficult to find a time to convene a group of 6 - 10 post - partum women with an infant together within on e room to conduct a focus group ; hence S tage 3 was omitted . We were still able to have the participating women share knowledge on what is needed within the community to assist women to meet their breastfeeding goals during Stage 2 . 3. 3 .5 Analyse s All individual semi - structured interviews w ere digitally recorded along with hand notes. The recordings were independently reviewed and transcribed verbatim by the primary researc h er 63 and one IRB trained undergraduate research assistant. To ensure accuracy, the transcriptions were compared against the audiotaped interviews and the field notes taken. The primary researcher and another doctoral student (qualitatively trained) independently review ed all data line - by - line looking for patterns, recurrent words and th emes. Then, each reviewer generate d coding labels as potential categories based on an initial coding schema, along with highlighting exemplar quotes. The two reviewers discuss ed and compare d results, resolve d any ambiguities and coding discrepancies until consensus wa s reached. Reviewers independently review ed the transcripts/passages and add ed new categories where needed. NVivo 1 2 , a qualitative analysis software , was used to assist in content analysis of the transcripts. Data from each code was summarized and organized to be displayed according to each code. 64 Table 3.6 Aim 3 Research Topic Areas/ Questions and Interview Questions Aim 3: Research Topic Areas Research Questions: Interview Questions Opinions and Importance of meeting Breastfeeding Goals (Challenges/ Success) To what extent are challenges and facilitators involved in meeting breastfeeding goals ? 1) What made you successful in completing your breastfeeding goals? 2) What inhibited you from meeting your breastfeeding goals? 3) What were some challenges that you may have experienced during this time? 4) What strategies were employed that assisted in your success for breastfeeding goals that were established? 5) How did the feeding method chosen affect your success? Social Support to meet Breastfeeding Goals How does social support networks influence breastfeeding goals ? 1) What kind of support do you think would have been helpful for you? 2) What kind of support do you think is still needed for all women to meet their goals? 3) What do you think is needed to help more women as yourself to complete Breastfeeding goals ? 4) How did your social support networks impact your breastfeeding goals? 5) What do you think society or the community could do to support exclusive/ continued breastfe eding? 65 Chapter 4 - Breastfeeding Self - Efficacy Predicts Breastfeeding Outcomes in a Sample of African American Women Target Journal: Maternal and Child Health Journal 4 . 1 Abstract: Objectives: Breastfeeding (BF) is known to benefit both mother and infant. While breastfeeding has increased overall in the US, rates remain lower among African Americans (AA). BF self - efficacy, a predictor of breastfeeding initiation and duration, has seldom been assessed among AA. This study sought to measure BF self - ef ficacy during mid - late pregnancy, 6 and 20 w ee ks postpartum (PP) to determine its association with breastfeeding outcomes (duration and intensity (% of feeds from BF versus Formula)) in a sample of AA women . Methods: AA women were recruited in mid - late p regnancy from a prenatal clinic in Detroit, MI; then enrolled in a 2 - arm randomized, controlled, feasibility trial of BF support and PP weight management. Data was collected either in person or by phone at enrollment, 6 and 20 w ee ks PP. Surveys included qu estions about sociodemographic characteristics, breastfeeding self - efficacy, social support, and infant feeding practices. Stepwise selection determined predictors for inclusion in all multiple linear regression models. The results reflect data from enroll ed women (n=53) specifically with complete data on targeted variables. Results: Participant ages ranged from 18 - 43 yrs. Of the participants, 53% were WIC enrollees, ee ks PP were 67% and 4 7% respectively. The mean breastfeeding self - efficacy score among participants decreased over the study period when measured at late pregnancy (59.8) and 20 w ee ks PP (56.2). Breastfeeding self - ee ks P P p=.01) positively influenced breastfeeding duration (measured in days) in fully adjusted models for all participants. Breastfeeding self - = .114, p=.017) and at 6 66 w ee ks PP ( =.188, p =.001), showed an increase in breastfeeding intensity at 6 w ee ks postpartum in fully adjusted models for all participants. For every one unit increase in breastfeeding self - efficacy measured at 6 wks PP, breastfeeding intensity increased by .162 units when measured at 20 w ee ks PP ( =.162, p =.05). Conclusions for Practice: Assessment of BF Self - Efficacy provides direction for individualized confidence - building interventions that c ould help improve the disproportionately low rates of BF among AAs. Keywords: African American breastfeeding, breastfeeding self - efficacy, breastfeeding intensity 67 4.2 Introduction Breastfeeding (BF) is an excellent source of nutrition, and also offers protective factors for infants and mothers by providing resistance to many illnesses. 31 Despite the various recognized benefits, BF rates are suboptimal, and BF promotion continues to remain a worldwide public health priority. 18 Although BF rates in the United States (US) are slowly inclining, many infants are not being breastfed exclusively or for the recommended duration. 33 The American Academy of Pediatrics recommends exclusive BF for the first 6 months, and continued BF with the introduction of complementary foods through at least the first year of life. 20 While there has been a steady incline in BF rates throughout the US from the implementation of BF initiatives, AA women still have disproportionately lower rates of BF initiation and duration compared to other ethnic groups. 129 130 According to The Centers for Disease Control and Prevention (CDC), from 2011 to 2015 the percentage of US women who initiated BF was 64.3% for African Americ ans, 81.5% for Whites, and 81.9% for Hispanics. Exclusive BF through 6 months was reportedly at 14% for African Americans, 22.5 % for Whites, and 18.2% for Hispanics. 130 While these rates have improved among all women in the last decade, there remains a considerable concern about and disparity in BF rates among ethnic groups. Breastfe eding noted that even after adjusting for family income or education level, BF rates for African American infants were still lower than for White infants at birth, 6 months, and 12 months of age. 18 BF initiatives have traditionally focused on BF initiation, and have been rather successful. 131 However, low BF rates for all mother - infant dyads at 6 months postpartum persist, and the disparities between different ethnicities indicate that there is a clear need to shift the focus from BF initiation to other BF patterns/behaviors. It has been noted in the literature that it 68 may be better to assess the various types of BF behavior beyond the labeling of full exclusive BF, and utilize BF i ntensity (% of feeds from BF versus Formula) as a measure. 21 22 One reason which makes it challenging to determine a dose related relationship for comp arison across studies. 21 121 132 Measuring BF intensity provides a more useful outcome measure, which helps in understanding BF behavior and can be applied at the individual and/or program leve l. 22 Also, noticing the difficulty to reach breastfeeding goals among this group and in an effort to increase BF duration and reach exclusivity, it has been suggested tha t measuring BF intensity may be a better method of measurement to support an encouraging environment, along with reinforcing manageable accomplishments, and it also has been shown to reflect breastfeeding duration 133 making it an useful outcome measure. 22 Several studies have documented that multiple factors, including psychosocial factors, 134 - 136 African American women often have additional factors that affect not only their choices, but also their continued BF experiences, which may contribu te to the persistent health disparity. 129 BF rates are a public health concern particularly for African American families, and BF specifically has been suggested as one of the means for addressing excess morbidity and mortality in underserved populations. 18 43 It is hence important to assess and begin to address modifiable factors that are known contributors to the range of BF challenges experienced by Afric an American mothers. 129 Self - effic acy has been documented in several studies as being a psychosocial factor that can be used as a valuable tool for predicting BF initiation and duration. 13 15 106 129 136 Self - efficacy is defined a 102 Dennis - 69 efficacy theory and scale assessment. 13 Since Dennis created the Breastfeeding Self - Efficacy Short Form (BSES - SF) 106 , BF self - efficacy has been studied in vario us populations, specifically among Canadian 13 , Australian 15 , Puerto Rican 112 , Chinese 137 , African 115 and Polish 114 women. All of these studies reported similar findings of self - efficacy significantly influ encing BF initiation and duration, however it has been minimally studied among African American women. Breastfeeding self - efficacy among this population has often been studied within the literature and identified as a barrier or facilitator to breastfeedi ng. While low BF self - efficacy is not unique to African American mothers, it is disproportionally prevalent among this population, potentially contributing to low BF duration. 129 McCarter - Spaulding and Dennis 108 assessed the psychometric properties for use of the BSES - SF with black women from various ethnic backgrounds within the US. The results wer e consistent with the literature, where postpartum BF self - efficacy (around 3 weeks postpartum) predicted BF behavior at 4 and 24 weeks postpartum. Their study specifically tested the validity and reliability of the BSES - SF tool among Black women from seve ral ethnic backgrounds ; specifically 32% of their sample identified as African American. However, additional testing of this tool is needed within this group to ensure effective adaptation, and continue advancing our collective knowledge of if, and how it can increase multiple BF rates (duration and intensity) in this population. 138 In order to reduce existing BF disparities, it may be important to assess African Ameri - delivery to improve BF outcomes. Therefore, the primary purpose of this study was to determine if or how breastfeeding self - efficacy, from late pregnancy through a defined post - partum period, impacts breast feeding outcomes (duration and intensity) for African American women. This study/aim addresses the following research questions and hypotheses. 70 1. Does breastfeeding self - efficacy (BSE) assessed in late pregnancy, prior to a breastfeeding support interventio n, and predict breastfeeding initiation regardless of group assignment in a randomized control trial? Hypothesis: It was hypothesized that having a high breastfeeding self - efficacy in late pregnancy will be positively associated with breastfeeding initiati on for the participants. 2. Does breastfeeding self - efficacy (BSE) predict breastfeeding duration for the participants in either the intervention or control group? Hypothesis: It was hypothesized that higher breastfeeding self - efficacy would truly increase b reastfeeding duration for the participants in both the intervention and control groups. 3. Does breastfeeding self - efficacy (BSE) predict breastfeeding intensity for the participants in either the intervention or control group? Hypothesis: It was hypothesize d that higher breastfeeding self - efficacy w ould truly increase breastfeeding intensity for the participants in both the intervention and control groups. 4.3 Methods This study was completed in conjunction with a randomized control trial, dual intervention pilot study called: Mothers Allied with Mothers Around Breastfeeding Encouragement and Return to Health After Baby (Mama Bear - R21HD085138 - 02). Mama Bear was funded by the National Institutes of Health (NIH R21) and is a group project collaboration betwee n Henry Ford Health System (HFHS) and Michigan State University (MSU). The dual intervention focused on a postpartum weight management component and breastfeeding peer - support. Recruitment occurred in a prenatal care clinic in Detroit, MI, which is affilia ted with Henry Ford Health System (HFHS). Research staff screened pregnant women who were scheduled for glucose tolerance tests. The timing of study enrollment was chosen so that blood 71 could also be collected for other research objectives. Eligibility crit eria included pregnant African American women 18 years, who expressed the intent to breastfeed. Research staff from HFHS approached and consented eligible women for participation in the study. Following enrollment, study staff contacted participants (within one week) to complete interviewer administer ed surveys by phone before randomizing into either an intervention arm or usual care group. During the phone interview, questions concerning the following topics were asked: sociodemographic, health/pregnancy, maternal diet, and breastfeeding self - efficacy . Following randomization, a trained community health worker/ BF peer counselor contacted participants within the intervention group to meet in person where they provided one - time prenatal education focused on BF and weight management and a monetary incent ive (gift card) for partaking in the study. The interventionists continued to meet and provide support, technical help, and resources to participants through a combination of in person and mobile technology. Participants randomly assigned to the control gr oup received a letter stating that their usual care should be continued and included an incentive (gift card). Between two - six and twenty weeks post - partum, all participants were asked a series of questions and had their weight measured in person by resear ch staff. After each data collection visit, participants received a gift card (Meijer or Walmart) to thank them for their time. The institutional review boards at Michigan State University (16 - 297) and Henry Ford Health System (HFHS) approved all study pro tocols . For the Mama Bear pilot study, a total sample size of 80 women was pre - determined to detect at least a moderate effect size of 0.65 with a power of 80% based on the research objective to estimate the effect size rather than test for significance wh ich was approved in the NIH funded grant. However, due to recruitment and data collection challenges in the Mama Bear study, the final sample size at enrollment was n= 53. and n=34 at the conclusion of the study 72 1. Measures: Demographic variables collected for Aim 1 include: age, marital status, education, employment status, and participation in the Supplemental Nutrition Program for Women, Infants, and Children (WIC). Breastfeeding characteristics such as previous breastfeedi ng experience, intended breastfeeding duration, and breastfeeding knowledge and attitudes were also collected at enrollment. T he following demographic variables were coded as categorical: age, employment status, WIC enrollment, and number of pregnancies va riables. Marital status and education were dummy coded with three levels. The Breastfeeding Self - Efficacy Short Form (BSES - SF) 106 questionnaire, copyrighted by Cindy Lee Dennis, was adapted to be measured at three different time points (during late - pregnancy, two - six weeks postpartum, and twenty weeks post - partum). Permission was electronically obtained for use of this questionnaire by the creator/author Cindy Lee Dennis. The verb tenses on the form were modified to assess late pregnancy , which was prior to BF (future tense), and after BF had stopped (past tense). Modifications were made to help in assessing BF self - efficacy longitud collector understanding of participant responses. The BSES - SF included 14 questions, and used a 5 - point Likert scale to measure this single construct. Responses ranged from 1 (not at all confident) to 5 (very confident). Scoring was done by adding the rating for the items to achieve a total score ranging from 14 to 70. Low scores indicated participants had lower confidence in the ability to BF, and high scores indicated a higher con fidence in the ability to breastfeed. Authors reported a instrument of 0.94. Breastfeeding self - efficacy was analyzed as a continuous variable, and included as one of the cov ariates in the analyses for this paper (Aim 1) . 73 The BF intensity (% of feeds from BF versus f ormula) questionnaire was used as a measure for the main outcome variables. 21 22 This question was asked at two time points (2 - 8 weeks and 20 weeks post - par Breastfeeding Intensity Scale was coded as: 0= 100% artificial milk or solids (includes weaned), 1 = 20% breast milk combined with 80% artificial milk or solids (8 of every 10 feedings are NOT breast milk), 2= 20 - 25% breast milk and the rest artificial milk or solids (i.e. 5 - 8 of every 10 feedings are not breast milk), 3=50% breast milk and 50% artifici al milk or solids, 4= 50 - 80% breast milk and the rest artificial milk or solids (i.e. 2 - 5 of every 10 feedings are not breast milk), 5= 80% breast milk combined with 20% artificial milk or solids (2 of every 10 feedings are NOT breast milk), and 6=100% Bre astmilk. The items in this scale were ordered. With BF intensity having seven discrete levels, it remained coded and analyzed as a continuous variable. 124 Breastfeeding intensity was measured again between 2 - 6 weeks postpartum where 100% artificial milk or solids remained coded as 0 and any level of breastfeeding was co ded as 1. at 20 weeks postpartum, but due to difficulty conta cting participants, and/or the variation of time between when data collection was scheduled versus when it was actually collected, some data was collected later than 20 weeks postpartum. This flexibility of collecting data at later dates assisted in reduci ng attrition rates. However, we did have participants who were still breastfeeding at the final time of data collection, for whom we truncated the end date right at 5 months (140 days). 74 2. Statistical Analyses: Individual level descriptive statistics were completed on the socio - demographic variables of the study participants, including the following: age, marital status, employment, education, WIC enrollment, and number of pregnancies. T - tests were done with the predictor variable (breastfeeding self - effic acy) to show the mean and significant differences between usual care and intervention groups. Breastfeeding initiation was not analyzed since the majority ( 75% ) of the participants had initiated breastfeeding. It was proposed that this outcome variable would only be analyzed as a logistic regression if there were enough participants in both categories to conduct analyses. Since there were only four participants who did not breastfeed, l ogistic regression analyses was not feasible. Multiple linear regression models were used to determine if b reastfeeding self - efficacy measured at: 1) late pregnancy, 2) 6 weeks postpartum, 3) and 20 weeks postpartum predicted BF duration. Breastfeeding se lf - efficacy measured at: 1) late pregnancy, 2) 6 weeks postpartum, 3) and 20 weeks postpartum was also assessed in models to determine if it predicted breastfeeding intensity at 6 and 20 weeks postpartum. Linear regression models were determined by using s tepwise selection, specifically to reduce the large group of predictor variables and extract the best subset of variables for use in the model. All models conformed to linear regression assumptions. Unstandardized beta coefficients were reported in the tab les. However, standardized beta coefficients are reported and interpreted in the text. Stata version 15.1, 2017 was used for analyses. P values 0.05 were considered statistically significant for bivariate and multivariate analyses. 75 4 . 4 Results: A total of 53 participants enrolled in the randomized control trial dual intervention pilot study. Participants were predominantly between the ages of 18 - 30 years (67.92%), were unemployed/ looking for work (52.8%), received some college education or more (62.3%), and had never been married (61.76%). The majority were enrolled in the Women, Infants, and Children program (WIC) (52.83%), and were multiparous (52.83%) as depicted in Table 4.1. Parity was the only variable that was significantly different between the u sual care and intervention groups. All the participants had a moderately high breastfeeding self - efficacy mean that decreased slightly over the study period from enrollment with a mean of 59.8 and ending at 20 weeks post - partum with a mean of 56.2, which i s shown in Table 4. 2. There were no significant differences between the usual care and intervention groups regarding breastfeeding self - efficacy at each of the three time points. 76 Table 4.2 Mean Pregnancy and Post - Partum Breastfeeding Self - E fficacy Scores of Participants Breastfeeding Self - Efficacy Combined Mean (SE) Usual Care N Mean (SE) Intervention N Mean (SE) P - value Late Pregnancy n=50 59.8 (1.2) 24 59.3(1.9) 26 60.23(1.6) 0.71 6 weeks Post - Partum n=31 57.5 (2.1) 17 60.8(2.2) 14 53.5(3.6) 0.08 20 weeks Post - Partum n=31 56.2 (1.9) 16 56.2(2.6) 15 56.3(3.1) 0.97 Table 4.1 Characteristics of All Participants at Enrollment Baseline Characteristics Participants (n=53) n (%) Usual Care n=25 n (%) Intervention n=28 n (%) P - value Age (years) 0.07 18 - 30 36 (67.9) 20(80.0) 16(57.1) 31 - 45 17(32.1) 5(20.0) 12(42.9) Marital Status 0.74 Married 9 (16.9) 4(16.0) 5(17.9) Living with a Partner 14(26.4) 8(32.0) 6(21.4) Never been married 30(56.7) 13(52.0) 17(60.7) Employment Status 0.78 Employed 25(47.2) 11(44.0) 14(50.0) Unemployed, Looking for Work 28(52.8) 14(56.0) 14(50.0) Education 0.82 Equivalency 20(37.7) 9(36.0) 11(39.3) Some College 23(43.4) 12(48.0) 11(39.3) 4 Year College Degree, or More 10(18.9) 4(16.0) 6(21.4) WIC enrollment 0.58 Yes 28(52.8) 12(48.0) 16(57.1) No 25(47.2) 13(52.0) 12(42.9) Parity 0.02 Nulliparous 13(24.5) 10(40.0) 3(10.7) Multiparous 40(75.5) 15(60.0) 25(89.3) 77 Table 4. 3 depicts enrollment characteristics of the participants by breastfeeding duration in days. The majority of participants (n=30) between the ages of 18 - 30 yrs had a mean breastfeeding duration of 96 days. Participants who were married, unemployed , had a 4 - year coll ege degree or more, and/or not enrolled in WIC were more likely to breastfeed for a longer period of time . The mean duration of continued breastfeeding for the participants was 91.8 days . Table 4.3 Mean Breastfeeding Duration (days) of Participants Baseline Characteristics Breastfeeding Duration (days) N=43 Mean (SD) Age (years) 18 - 30 n=30 96.0(50.7) 31 - 45 n=13 82.2 (59.1) Marital Status Married n=7 96.7 (56.5) Living with a Partner n=12 87.1 (51.4) Never been married n=24 92.7(55.0) Employment Status Employed n=22 74.6 (55.6) Unemployed, Looking for Work n=21 109.8 (44.7) Education n=14 74.8 (57.2) Some College n=21 99.2 (52.1) 4 Year College Degree, or More n=8 102.1 (47.0) WIC enrollment Yes n=22 79.7 (54.9) No n=21 104.5(49.2) Parity One or two n=19 91.7 (56.2) Three or more n=24 91.9 (51.6) Breastfeeding Duration Mean SD 91.81 8.09 78 When comparing breastfeeding intensity o f participants at 6 and 20 weeks, participants in all categories were more likely to be breastfeeding at 6 weeks than at 20 weeks postpartum (Table 4 .4 ). Proportion breastfeeding intensity was slightly higher at 6 and 20 weeks PP (3.1 vs 2.7 respectively ) of those who were 31 - 45 years compared to participants between the ages of 18 - 30 yrs. Those who were married were more likely to be breastfeeding at 6 (4.2) and 20 (3.3) weeks compared to those who were either living with a partner or had never been married. Participants who were unemployed had a higher breastfeeding mean at 6 and 20 weeks PP compared to employed participants. Participants who were not enrolled in WIC had a higher proportion of breastfeeding at 6 (3.3) and 20 (3.3) weeks compared to their counterparts enrolled in WIC. Table 4.4 Characteristics of Participants at Enrollment by Breastfeeding Status at 6 and 20 Weeks Post - Partum Baseline Characteristics Proportion BF Intensity at 6 weeks PP (n=34) Mean (SD) Proportion BF Intensity at 20 weeks PP (n=34) Mean (SD) Age (years) 18 - 30 3.0(2.5) n=25 1.9 (2.43) n=25 31 - 45 3.1(2.7) n=9 2.7 (2.8) n=9 Marital Status Married 4.2(2.5) n=5 3.3 (3.0) n=6 Living with a Partner 1.8(2.3)n=8 1.0 (1.7) n=6 Never been married 3.2(2.5)n=21 2.0 (2.5) n=22 Employment Status Employed 2.9(2.5) n=17 1.38(2.4) n=13 Unemployed, Looking for Work 3.2(2.5) n=17 2.5(2.6) n=21 Education 3.4(2.4) n=10 1.5 (2.1) n=11 Some College 3.3(2.6) n=17 2.7 (2.7) n=16 4 Year College Degree, or More 1.7 (2.3) n=7 1.6 (2.7) n=7 WIC enrollment Yes 2.8(2.6) n=17 1.0(1.9) n=18 No 3.3 (2.5) n=17 3.3(2.6) n=16 Parity 79 Table 4. 5 shows the breastfeeding intensity percentage of all participants at 6 and 20 weeks postpartum. At six weeks post - partum, 32% of participants were exclusively breastfeeding, but this rate declined to 20% by 20 weeks postpartum. The majority of par ticipants at 6 weeks post - partum (67.6%) were providing their infants with some breastmilk along with other milk and/or solids. A t 20 weeks post - partum , the majority of participants (52.9%) were providing 100% artificial milk and/or solids. Table 4.4 One or two 3.3 (2.6) n=16 2.5 (2.8) n=15 Three or more 2.8(2.5) n=18 1.8(2.3) n=19 Table 4.5 Breastfeeding Intensity of P articipants at 6 and 20 W eeks P ost - P artum All Participants (%) Breastfeeding Intensity 6 weeks post - partum n=34 20 w eeks post - partum n=34 100% artificial milk or solids 11 (32.35) 18 (52.94) 20% breast milk combined with 80% artificial milk or solids --------- ---------- 20 - 25% breast milk and the rest artificial milk or solids 4 (11.76) 4 (11.76) 50% breast milk and 50% artificial milk or solids 4 (11.76) 1(2.94) 50 - 80% breast milk and the rest artificial milk or solids 3(8.82) 2(5.88) 80% breast milk combined with 20% artificial milk or solids 1(2.94) 2(5.88) 100% breast milk 11(32.35) 7(20.59) 80 Table 4.6 displays the results of breastfeeding self - efficacy (ranges from 14 - 70) measured at late pregnancy, and the additional predictors identified from stepwise selection , influencing breastfeeding duration (number of days) for all participants. Results show from the simple linear regression (Model 1) that as breastfeeding self - eff icacy increase d prenatally, breastfeeding duration also increased ( B .05). F rom standardizing the beta coefficient , results show that with every one unit i ncrease in prenatal breastfeeding self - efficacy, the n increased by .395 units. After controlling for additional covariates (education, WIC enrollment, and marital status), prenatal breastfeeding self - efficacy still remained a significant positive influence on breastfeeding duration ( B p .05) . With a one unit increase in prenatal breastfeeding self - efficacy, the mothers breastfeeding duration increased by .380 units, assuming all additional adjusted variables are held constant. As breastfeeding self - efficacy increases, days of breastfeeding sl ightly increase. As shown within the table women who had some college education breastfed on average 18.99 days more than women who completed high school or the equivalent. In addition, women who had a college degree breastfed on average 25.25 days more th an women who completed high school or the equivalent. However, these variables were not significant within the final adjusted model. After adjusting for all the variables (education, WIC, and marital status) the overall model was no longer significant i nd icating that these may not be the best explanatory variables for the regression model. However, breastfeeding self - efficacy remained significant having the same magnitude of effect across the unadjusted and adjusted models. 81 a b Married=referent Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4. 6: Breastfeeding Self - Efficacy M easured at L ate - P regnancy I nfluenc es B reastfeeding D uration of a ll P articipants ( N=40 ) Model 1 B (SE) Model 2 B (SE) Model 3 B (SE) Model 4 B (SE) Intercept - 48.64(53.6) - 64.40(54.90) - 52.22(59.03) - 52.64(64.13) Breastfeeding Self - Efficacy prenatally 2.40(.905)* 2.40(.910)* 2.32(.930)* 2.31(.957)* Education Some College a -------- 20.23(18.2) 18.07(18.71) 18.99(19.93) College Degree a -------- 28.75(22.71) 24.12(21.19) 25.25(25.25) WIC -------- -------- - 10.24(17.19) - 10.64(17.76) Marital Status Never been married b -------- -------- -------- - 3.34 (26.15 ) Living with a partner b -------- -------- -------- 2.12(23.46) R 2 0.16 0.20 0.21 0.21 Adjusted R 2 0.13 0.13 0.11 0.06 Omnibus 0.01 0.04 0.08 0.22 82 Table 4. 7 shows the results of breastfeeding self - efficacy (ranges from 14 - 70) measured at late pregnancy (adjusted for additional covariates) influencing breastfeeding duration (n=40). This table provides two distinct fully adjusted linear regression models using the same covariates identified from the stepwise selection for the usual care (n=19) and intervention group (n=21). Findings show that in the fully adjusted model breastfeeding self - efficacy measured during late pregnancy was positively associated with br eastfeeding duration in the usual care group ( B =3.60, .607, p 0.05). For every one unit increase in prenatal breastfeeding self - efficacy (usual care group) , the mothers breastfeeding duration (number of days) increase d by .607 units , assuming all variables are held constant. The overall regression model in the control group was approaching significance (p .05), which suggested that the variables within the model were the best explanatory variables for the regression model. No variabl es were associated with breastfeeding duration in the intervention group. Although not significant , women who were enrolled in WIC , breastfed 27.44 days less than women who were not enrolled in WIC. 83 Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4. 8 provides the results of breastfeeding self - efficacy (ranging from 14 - 70) measured at 6 weeks postpartum, and the additional predictors influencing breastfeeding duration (number of days) for all participants (n=28). Results depicted by the bivariate (sim ple) linear regression (Model 1) show that as breastfeeding self - efficacy at 6 weeks postpartum increase d breastfeeding duration (in days) also increase d ( B= 1.96, .481 , p .01). After adjusting for additional covariates (prenatal breastfeeding self - efficacy, group assignment, WIC enrollment, and number of pregnancies including current), breastfeeding self - efficacy at 6 weeks postpartum still remained significant with a positive influence on breastfeeding duration ( B = 3.10, =.752 , p .0 0 1). Results from standardizing the beta coefficient show that with every one unit increase in breastfeeding self - efficacy measured at 6 weeks postpartum, the mothers breastfeeding duration Table 4.7 Predictors M easured at L ate P regnancy I nfl uencing Breastfeeding Duration S eparated by G roup ( N=40 ) Usual Care N=19 B (SE) Intervention N=21 B (SE) Intercept - 155.87(89.57) - 1.65(108.1) Breastfeeding Self - Efficacy prenatally 3.60(1.17)* 1.69(1.73) Some College 42.09(26.38) - 6.57(34.27) College Degree 4.96(37.29) 20.64(38.45) WIC (enrolled) - 27.44(24.95) 5.11(28.51) Marital Status (referent= married) Never been married 34.8(37.14) - 9.96(46.26) Living with a partner 41.73(32.67) - 18.03(37.92) R 2 0.59 0.08 Adjusted R 2 0.39 - 0.30 Omnibus 0.05 0.96 84 increase d by .752 units , assuming all variables are held const ant. Prenatal breastfeeding self - efficacy, although not significant, was adjusted for in all models because participants were asked the same questions while pregnant. With the same questions being asked again at 6 weeks postpartum, this could have potentially influenced their responses . Group assignment was not significant, but was an important predictor to con trol for , since the intervention could have potentially influenced breastfeeding duration. WIC interesting ly was negatively associated with breastfeeding duration as shown in Model 5 ( B = - 45.27, - .465, p .05). From standardizing the beta coefficient, Mo del 5 displayed that for mothers enrolled in WIC, the mothers breastfeeding duration decrease d by - .465 units , assuming all additional adjusted variables are held constant. Breastfeeding self - efficacy and WIC participation were significant in Model 5. When , interaction effects between the two variables were tested, the interaction between the two variables was not significant. After adjusting for all the variables, the overall model (Model 5) was significant (p .01), which indicated that the explanatory variables provided the best fit in the regression model with an adjusted R 2 of 0.33 . 85 . a n =31 Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4.8 Breastfeeding S elf - E fficacy M easured at 6 weeks P ost - Partum I nfluences Breastfeeding Duration N=28 Model 1 a B (S E) Model 2 B (SE) Model 3 B (SE) Model 4 B (SE) Model 5 B (SE) Intercept - 12.41(38.95) - 17.78(69.26) - 48.76(74.89) - 50.29(71.33) - 85.55(71.69) Breastfeeding self - efficacy at 6 weeks post - partum 1.96(.664)** 2.10(.761)* 2.44(.822)** 2.87(.817)** 3.10(.797)** Breastfeeding Self - Efficacy prenatally -------- - .004(1.21) - .275(1.23) - .398(1.17) - .511(1.13) Group -------- -------- 19.39(18.13) 17.72(17.28) 10.51(17.16) WIC -------- -------- -------- - 31.36(16.86) - 45.27(18.18)* Pregnancies including current -------- -------- -------- -------- 29.59(17.48) R 2 0.23 0.26 0.29 0.39 0.46 Adjusted R 2 0.20 0.20 0.21 0.28 0.33 Omnibus 0.006 0.02 0.03 0.01 0.01 86 Table 4. 9 shows the impact of breastfeeding self - efficacy (ranging from 14 - 70) measured at six weeks postpartum (adjusted for additional covariates) on breastfeeding duration. This table provides two distinct fully adjusted linear regressions (usual care and intervention group). Findings from the fully adjusted model show that breast feeding self - efficacy measured at six weeks postpartum positively influenced breastfeeding duration in the usual care group ( B = 3.48, = .666, p 0.05). From standardizing the beta coefficients, the final model shows that with every one unit increase in brea stfeeding self - efficacy measured at 6 weeks postpartum, the mothers breastfeeding duration (number of days) increased by .666 units , assuming all additional adjusted variables are held constant. The overall regression model in the control group was signifi cant (p .03), which suggests that the explanatory variables within the model were the best variables to use for the regression model. No variables were associated with breastfeeding duration in the intervention group. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4.9 Influence of Breastfeeding Self - Efficacy M easured at 6 W eeks P ost - P artum on Breastfeeding Duration S eparated by G roup ( N=28 ) Usual Care N=16 B (SE) Intervention N=12 B (SE) Intercept - 150.68(88.22) 6.59(127.13) Breastfeeding Self - Efficacy at 6 weeks postpartum 3.48(1.51)* 2.87(1.33) Breastfeeding Self - Efficacy at late pregnancy .326(2.04) - 1.60(1.82) WIC - 39.40(20.06) - 56.91(42.65) Parity 29.91(20.12) 35.42(37.24) R 2 0.58 0.41 Adjusted R 2 0.43 0.07 Omnibus 0.03 0.38 87 Table 4. 10 provides the results of breastfeeding self - efficacy measured at 20 weeks postpartum, and the additional predictors influencing breastfeeding duration for all participants (n=21). Results from the bivariate linear regression (Model 1) show that as breastfeeding self - efficacy at 20 weeks postpartum increased, breastfeeding duration also increased ( B =2.29, .553, p .01). From stan dardizing the beta coefficients, model 1 shows that with every one unit increase in breastfeeding self - efficacy measured at 20 weeks postpartum, the mothers breastfeeding days increased by .553. After adjusting for additional covariates (breastfeeding self - efficacy prenatally and 6 weeks postpartum, group, WIC enrollment, and pregnancies including current), breastfeeding self - efficacy at 20 weeks postpartum was no longer significantly associated with breastfeeding, or related to BF duration. However, WIC ( - .581, p .05) and parity status ( p .05) was negatively associated with breastfeeding duration significant in Model 6. Interestingly, for the m others enrolled in WIC , breastfeeding duration decreased by - .581 units , assuming the other adjusted v ariables are held constant . As parity increase d , the number of days breastfeeding also increased by .439 units . After adjusting for all the variables, the overall model (Model 6) was significant (p .01), indicating that the included variables explain ed 5 9% of the variance in breastfeeding duration ( adjusted R 2 = 0 .59) 88 a n=31, b n=23 Standard errors in parentheses *** p<0.001, ** p<0.01, *p<0.05 Table 4.10 Breastfeeding S elf - E fficacy M easured at 20 W eeks Post - Partum I nfluences B reastfeeding D uration ( N=21 ) Model 1 a B (SE) Model 2 b B (SE) Model 3 B (SE) Model 4 B (SE) Model 5 B (SE) Model 6 B (SE) Intercept - 20.95(36.84) 19.36(42.22) - 29.46(57.91) - 80.76(62.14) - 50.75 (61.25) - 67.22(52.77) Breastfeeding Self - Efficacy at 20wks post - partum 2.29(.643)** 1.68(1.19) 2.11(1.09) 1.76(1.06) .733(1.16) .486(1.00) Breastfeeding Self - Efficacy at 6wks post - partum -------- .028(1.24) - .197(1.23) .636(1.26) 1.65(1.33) 1.93(1.14) Breastfeeding Self - Efficacy prenatally -------- -------- .686(1.00) .459(.957) .155(.921) - .055(.792) Group -------- -------- -------- 25.35(14.55) 23.93(13.76) 13.63(12.44) WIC -------- -------- -------- -------- - 25.20(14.72) - 43.13(14.41)* Pregnancies including current -------- -------- -------- -------- -------- 32.29(12.65)* R 2 0.30 0.25 0.40 0.50 0.58 .71 Adjusted R 2 0.28 0.17 0.30 0.37 0.44 .59 Omnibus .001 .059 .028 .02 .01 .003 89 Table 4. 11 shows findings for breastfeeding self - efficacy measured at 20 weeks postpartum (adjusted for additional covariates), and its influence on breastfeeding duration in the usual care and intervention groups separately. This table provides results from two distinct, fully adjust ed linear regressions (usual care and intervention group) using the covariates identified from the stepwise linear regression that were used in Table 10. In the fully adjusted model, breastfeeding self - efficacy measured at twenty weeks postpartum did not p redict breastfeeding duration in either the usual care or intervention group. In addition, no other covariates were associated with breastfeeding duration in either regression. Standard errors in parentheses *** p<0.001, ** p<0.01, *p<0.05 Table 4.11 Influence of Breastfeeding Self - E fficacy M easured at 20 weeks P ostp artum on Breastfeeding Duration S eparated by G roup ( N=2 1) Usual Care Group N=12 B (SE) Intervention N=9 B (SE) Intercept - 42.35(133.77) - 64.08(47.0) Breastfeeding Self - Efficacy at 20 weeks postpartum .650(2.34) 1.42(2.28) Breastfeeding Self - Efficacy at 6 weeks postpartum 1.72(2.06) .725(2.62) Breastfeeding Self - Efficacy prenatally - .012(2.15) .202(.706) WIC - 45.59(43.86) - 31.29(15.99) Parity 25.73(28.44) 43.90(14.08) R 2 0.64 0.94 Adjusted R 2 0.34 0.83 Omnibus 0.18 0.05 90 Table 4. 12 shows predictors of breastfeeding intensity at 6 weeks PP including BFSE and additional variables identified (education, WIC enrollment, and marital status) from the stepwise linear regression . Results from the bivariate linear regression (Model 1) show that as prenatal breastfeeding self - efficacy increased, participant breastfeeding intensity (ratio of breastfeeding vs formula) also increased at 6 weeks postpartum ( B = .124, 50, p .0 5 ). After controlling for addit ional covariates prenatal breastfeeding self - efficacy remained a significant positive influence on breastfeeding intensity at 6 weeks postpartum ( B = .114, =.412, p .0 5 ). With every one unit increase in prenatal self - efficacy , breastfeeding intensity increased by .412 units, assuming the other adjusted variables are held constant. After adjusting for all the variables, the overall model remained significant , indicating that these explanatory variables provided the best fit in the regression model. a b Married=referent Standard errors in parentheses *** p 0.001, ** p 0.01, * p 0.0 5 Table 4.12 Breastfeeding Self Efficacy M easured P renatally I nfluences Breastfeeding Intensity at 6 W eeks Post - Partum ( N=3 1) Model 1 B (SE) Model 2 B (SE) Model 3 B (SE) Model 4 B (SE) Intercept - 4.26(2.67) - 3.54(2.72) - 2.81(2.98) - 1.88(2.91) Breastfeeding Self - Efficacy prenatally .124(.045)* .123(.046)* .118(.047)* .114(.045)* Education Some College a -------- - .494(.984) - .714(1.05) - 1.02(1.07) College Degree a -------- - 1.70(1.17) - 2.00(1.27) - 2.04(1.25) WIC -------- -------- - .600(.929) - .751(.900) Marital Status Never been married b -------- -------- -------- - 2.80(1.40) Living with a partner b -------- -------- -------- - 1.42(1.20) R 2 0.20 0.26 0.27 0.35 Adjusted R 2 0.17 0.18 0.16 0.22 Omnibus 0.01 0.03 0.03 0.04 91 Table 4. 13 presents the results of the predictor , breastfeeding self - efficacy (ranging 14 - 70) , measured prenatally, and additional predictors identified from the stepwise linear regression influencing breastfeeding intensity at 6 weeks postpartum. This table provides results from two distinct, fully adjusted linear regressions (usual care and intervention group). In the usual care group, as prenatal breastfeeding self - efficacy increased, participant breastfeeding in tensity increased at 6 weeks postpartum ( B = .178, p .01). In the usual care group, among women who had a college degree, compared to those with a high school education or less, breastfeeding intensity decreased by - .946 units ( B = - 6.34, - .946 , p .05). In the fully adjusted linear regression , in the intervention group, as prenatal breastfeeding self - efficacy increase d, breastfeeding intensity increase d at 6 weeks postpartum by .901 units , assuming all other variables were held constant ( B = .256, =.901, p .01). In the intervention group, the for women who had never been married , breastfeeding intensity decreased by - .966 units , compared to those who are married ( B = - 10.60, - .966 , p .001). In addition, for women living with a partner, breastfe e d ing intens ity decreased by - .971 units , compared to those who were married , assuming all variables were held constant ( B = - 6.30, - .971 , p .01). The overall models in both regressions were significant, indicating that these explanatory variables provided the best fit in the regression model. 92 Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4. 13 Prenatal Breastfeeding Self - Efficacy Influences Breastfeeding Intensity at 6 weeks Post - Partum S eparated by G roup ( N=31 ) Usual Care N=17 B (SE) Intervention N=14 B (SE) Intercept - 3.52(3.97) - 6.89(2.97) Breastfeeding Self - Efficacy prenatally .178(.055)** .256(.051)** School) Some College - 2.72(1.43) .584(.977) College Degree - 6.34(2.07)* - .912(.958) WIC (enrolled) - 2.64(1.25) 1.78(.921) Marital Status (referent= married) Never been married .080(1.68) - 10.60(1.93)*** Living with a partner .729(1.45) - 6.30(1.33)** R 2 0.66 0.86 Adjusted R 2 0.46 0.73 Omnibus 0.04 0.01 93 Table 4. 14 displays the association between prenatal breastfeeding self - efficacy, and additional predictors that influenced breastfeeding intensity at 20 weeks post - partum . B reastfeeding self - efficacy measured prenatally was not associated with breastfeeding intensity at 20 weeks postpartum, and remained non - significant after adjusting for additional covariates. However, as WIC was added into Model 3 and 4, it was negatively associated with breastfeeding intensity at 20 weeks postpartum. None of the overall models were significant as shown by the omnibus. Although, not significant Models 1,3, and 4 were trending toward significa nce suggesting that having a larger sample size may increase the significance level. a b Married=referent Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4.14 Association B etween P redictors M easured at L ate P regnancy and Breastfeeding Intensity Me asured at 20 W eeks Post - Partum ( N=32 ) Model 1 B (SE) Model 2 B (SE) Model 3 B (SE) Model 4 B (SE) Intercept - 2.63(2.71) - 2.61(2.83) .205(2.86) 1.44(3.05) Breastfeeding Self - Efficacy prenatally .079(.045) .074(.047) .051(.044) .054(.044) Education Some Higher Education a -------- .579(1.01) .289(.943) - .118(1.01) College Degree a -------- - .118(1.21) - .989(1.17) - 1.01(1.20) WIC -------- -------- - 2.09(.865)* - 2.13(.859)* Marital Status Never been married b -------- -------- -------- - 2.20(1.42) Living with a partner b -------- -------- -------- - 1.28(1.13) R 2 0.09 0.11 0.27 0.33 Adjusted R 2 0.06 0.01 0.16 0.17 Omnibus 0.09 0.35 0.07 0.09 94 Table 4. 15 presents results from the fully adjusted model of breastfeeding self - efficacy measured prenatally, and its influence on breast feeding intensity at 20 weeks postpartum. This table provides results from the linear regressions (usual care and intervention group) where participants were separated by group randomization. None of the predictor variables were associated with breastfeed ing intensity at 20 weeks postpartum in either the usual care or intervention group. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4. 15 Predictors Influencing Breastfeeding Intensity at 20 weeks Post - Partum S eparated by G roup ( N=32 ) Usual Care N=16 B (SE) Intervention N=16 B (SE) Intercept - 2.36(4.66) 4.98(5.20) Breastfeeding Self - Efficacy prenatally .098(.063) .018(.080) Education Some College .533(1.36) - 1.29(1.85) College Degree - .700(2.08) - 2.09(1.91) WIC (enrolled) - 2.66(1.29) - 1.26(1.41) Marital Status (referent= married) Never been married - 1.20(2.01) - 2.77(2.59) Living with a partner .218(1.61) - 2.72(2.00) R 2 0.56 0.31 Adjusted R 2 0.27 0.14 Omnibus 0.18 0.66 95 Table 4. 16 displays results of breastfeeding self - efficacy (ranging from 14 - 70) measured at 6 weeks post - partum, along with additional covariates influencing breastfeeding intensity at 6 weeks post - partum (n=28). From the bivariate linear regression, breastfeeding self - efficacy measured at 6 weeks post - partum positively influenced breastfeedin g intensity (increase in breastmilk) at 6 weeks ( B = .146, =. 706 , p .001). After adjusting for additional variables, breastfeeding self - efficacy at 6 weeks postpartum ( B = .188, = .894 , p .001) remained significant, being positively associated with breastfeeding intensity in the final model. The standardized coefficient results show that with every one unit increase in breastfeeding self - efficacy measured at 6 weeks post - partum, breastfeedi ng intensity (use of breastmilk) at 6 weeks post - partum increased by .894 units , assuming the other adjusted variables are held constant. When group was controlled for in Model s 3 - 5, participants in the intervention were more likely to increase breastfeed ing intensity ( B = 1.6 = .324 , p .001). The standardized coefficient in model 5, showed that within the intervention group, mothers breastfeeding intensity increased by .324 units , assuming all other variables were held constant. Prenatal breastfeeding self - efficacy was not significant, but was important to adjust for in all models since participants were asked the same questions while pregnant, which could potentially influence how they responded when assessed at 6 weeks postpartum. The overall model was significant (p .001), indicating that the explanatory variables provide a good fit in th is regression mod el. 96 a n =31 Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4. 16 Breastfeeding Self - Efficacy Influences Breastfeeding Intensity When Both Measured at 6 W eeks Post - Partum ( N=28 ) Model 1 a B (SE) Model 2 B (SE) Model 3 B (SE) Model 4 B (SE) Model 5 B (SE) Intercept - 5.11(1.60)* - 6.63(2.82)* - 9.21(2.82)** - 9.25(2.74)** - 8.95(3.31)** Breastfeeding Self - Efficacy at 6wks post - partum .146(.027)*** .148(.031)*** .177(.031)*** .189(.031)*** .188(.032)*** Breastfeeding Self - Efficacy prenatally -------- .024(.049) .002(.047) - .002(.045) - .003(.046) Group -------- -------- 1.61(.682)* 1.56(.666)* 1.62(.763)* WIC -------- -------- -------- - .974(.649) - .957(.670) Age -------- -------- -------- -------- - .011(.062) R 2 0.50 0.53 0.62 0.65 0.65 Adjusted R 2 0.48 0.49 0.57 0.59 0.58 Omnibus .001 .001 .001 .001 .001 97 Table 4. 17 shows the association between breastfeeding self - efficacy measured at six weeks postpartum (adjusted for additional covariates), and breastfeed ing intensity at 6 weeks postpartum. This table shows results from two distinct fully adjusted linear regressions (usual care and intervention groups). In the fully adjusted model, breastfeeding self - efficacy measured at six weeks postpartum was positively associated with increased breastfeeding intensity (use in breastmilk) in both the usual care ( B= .188, = .669, p .05) and intervention groups ( B = .206, = .678 = p .01). Within the both groups, mothers on average were increasing the use of breast milk when measured at six weeks postpartum, assuming all other variables were held constant. The overall model for both usual care and intervention groups were significant (p .01) , suggesting that the variables within each model were the best explanatory variables for the regression. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4. 17 Breastfeeding Self - Efficacy M easured at 6 W eeks Post - Partum Influenc es Breastfeeding Intensity at 6 W eeks Post - Partum S eparated by G roup ( N =28 ) Usual Care N=16 B (SE) Intervention N=12 B (SE) Intercept - 6.66(3.89) - 9.42(5.73) Breastfeeding Self - Efficacy at 6 weeks post - partum .188(.081)* .206(.047)** Breastfeeding Self - Efficacy prenatally .004(.102) .030(.062) Education Some College - 1.49(1.48) .209(1.48) College Degree - 4.45(2.14) 1.19(1.46) WIC - 2.27(1.27) - 2.57(1.49) Pregnancies including current .778(.935) .329(1.38) R 2 0.78 0.83 Adjusted R 2 0.64 0.62 Omnibus 0.01 0.01 98 Table 4. 18 displays results of breastfeeding self - efficacy measured at 6 weeks postpartum, and its influence on breastfeeding inte nsity at 20 weeks post - partum (n=22). With every one unit increase in breastfeeding self - efficacy measured at 6 weeks post - partum, breastfeeding intensity at 20 weeks post - partum increased by .434 units ( B =. 111, =.434, p .05). In Model 3 within the intervention group, results showed for every one unit increase in breastfeeding self - efficacy, breastfeeding intensity increased by .443 units at 20 weeks postpartum , holding all other variables constant ( B =2.36, = .443 , p .05 ). In Models 4 and 5, participants enrolled in WIC services showed a decrease d use of breast milk when measured at 20 weeks post - partum ( B = - 2.65, = - .430 , p .01 ). In the final adjusted model, as breastfeeding self - efficacy measured at 6 weeks post - partum increased, breastfeeding intensity at 20 weeks post - partum increased, indicating an increase use of breastmilk ( B =.152, p .05). The overall model adjusted for all confounders was significant (p .001), indicating variables provided were the best fit for the linear regression model 99 a n=24 Standard errors in parentheses *** p<0.001, ** p<0.01, *p<0.0 5 Table 4.18 Breastfeeding Self - Efficacy M easured at 6 W eeks Post - Partum I nfluences Breastfeeding Intensity at 20 W eeks Post - Partum ( N=22 ) Model 1 a B (SE) Model 2 B (SE) Model 3 B (SE) Model 4 B (SE) Model 5 B (SE) Intercept - 3.97(2.94) - 5.89(4.35) - 8.87(4.19)* - 7.43(3.67) - 11.83(4.68)** Breastfeeding Self - Efficacy at 6wks post - partum .111(.049)* .107(.056) .158(.056)* .165(.048)** .152(.049)** Breastfeeding Self - Efficacy prenatally -------- .036(.077) .020(.071) .009(.062) .032(.0685) Group -------- -------- 2.36(1.08)* 1.94(.945) 1.20(1.17) WIC -------- -------- -------- - 2.27(.855)* - 2.50 (.881)* Age -------- -------- -------- -------- .800(.669) R 2 0.19 0.23 0.39 0.57 0.61 Adjusted R 2 0.15 0.15 0.29 0.47 0.48 Omnibus .033 .079 .025 .004 .007 100 Table 4. 19 shows the analysis of breastfeeding self - efficacy measured at six weeks postpartum (adjusted for additional covariates), and breastfeeding intensity at 20 weeks postpartum. This table provides res ults from two distinct fully adjusted linear regressions (usual care and intervention groups). In the fully adjusted model, breastfeeding self - efficacy measured at six weeks postpartum was not associated with breastfeeding intensity in either the usual car e or intervention groups. In addition, no other covariates were associated with breastfeeding intensity in either regression. Standard errors in parentheses *** p<0.001, ** p<0.01, *p<0.05 Table 4.19 Association B etween P redictors M easured at 6 W eeks Post - Partum and Breastfeeding Intensity at 20 W eeks Post - Partum ( N= 22 ) Usual Care N=13 B (SE) Intervention N=9 B (SE) Intercept - 13.78(7.27) - 382(8.86) Breastfeeding Self - Efficacy at 6 weeks post - partum .128(.161) .413(.177) Breastfeeding Self - Efficacy prenatally .106(.146) - .131(.110) High School) Some College - 2.23(2.44) - 8.44(6.72) College Degree - .247(3.74) .453(2.18) WIC (enrolled) - 2.33(2.65) - 5.84(2.76) Pregnancies including current .921(1.82) - 4.78(4.27) R 2 0.69 0.86 Adjusted R 2 0.37 0.45 Omnibus 0.18 0.36 101 Table 4. 20 provides the results of breastfeeding self - efficacy (ranging from 14 - 70) measured at 20 weeks postpartum of all participants, and its influence on breastfeeding intensity at 20 weeks postpartum (n=21). Results from the bivariate (simple) linear regressi on model showed that for every one unit increase in breastfeeding self - efficacy, breastfeeding intensity increased (increase in breast milk use) when measured at 20 weeks ( B = .122, = .530 , p .01). Breastfeeding self - efficacy remained significant (Models 2 - 4) after sequentially adjusting for additional predictors; with a positive influence on breastfeeding intensity. In model 6, after adjusting for all variables, breastfeeding self - efficacy measured at 20 weeks was no longer significant, but still showed a n increase in breast milk use . Breastfeeding self - efficacy that was measured prenatally and at 6 weeks post - partum, was adjusted for in Models 2 - 6, with the expectation that the questions asked at th ose specific time points may potentially influence how part icipants respond ed when assessed at 20 weeks postpartum. In Model 6, after adjusting for all covariates, the omnibus was significant (p .01), i ndicating that the explanatory variables provided the best fit . 102 a n=31 Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4.20 Breastfeeding Self - Efficacy Influences Breastfeeding Intensity M easured at 20 W eeks Post - Partum ( N=21 ) Model 1 a B ( SE) Model 2 B (SE) Model 3 B (SE) Model 4 B (SE) Model 5 B (SE) Model 6 B (SE) Intercept - 4.73(2.07)* - 4.80(2.86) - 8.15(3.99) - 10.17(3.81)* - 8.06(3.69)* - 11.68(4.52)* Breastfeeding Self - Efficacy at 20wks post - partum .122(.036)** .170(.081)* .204(.076)* .177(.071)* .101(.076) .081(.081) Breastfeeding Self - Efficacy at 6wks post - partum -------- - .042(.084) - .083(.085) - .018(.085) .057(.087) .066(.088) Breastfeeding Self - Efficacy prenatally -------- -------- .064 (.069) .047(.064) .024(.061) .039(.063) Group -------- -------- -------- 1.97(.978) 1.87(.906) 1.29(1.12) WIC -------- -------- -------- --------- - 1.87 (.969) - 2.21(1.05) Age -------- -------- -------- -------- -------- .087(.098) R 2 0.28 0.32 0.44 0.56 0.65 0.66 Adjusted R 2 0.26 0.25 0.35 0.45 0.53 0.52 Omnibus .002 .021 .016 .008 .004 .009 103 Table 4. 21 shows the results of breastfeeding self - efficacy measured at twenty weeks postpartum (adjusted for additional cov ariates), and its influence on breastfeeding intensity at 20 weeks postpartum. This table provides results from two distinct fully adjusted linear regression models (usual care and intervention groups). In the fully adjusted model, breastfeeding self - effic acy at twenty weeks postpartum was not associated with increased breastfeeding intensity in either the usual care or intervention group. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 4. 2 1 Predictors Influencing Breastfeeding Intensity at 20 W eeks Post - Partum ( N=21 ) Usual Care N=12 B (SE) Intervention N=9 B (SE) Intercept - 13.58(9.74) - 4.16(7.71) Breastfeeding Self - Efficacy at 20 weeks postpartum .029(.118) .135(.314) Breastfeeding Self - Efficacy at 6 weeks postpartum .028(.129) .047(.362) Breastfeeding Self - Efficacy prenatally .147(.128) - .024(.117) WIC - 3.00(2.06) - 2.47(2.77) Age .181(.220) - .009(.154) R 2 0.65 0.76 Adjusted R 2 0.36 0.37 Omnibus 0.17 0.31 104 4.5 Discussion The primary purpose of this study was to gain a better understanding of how breastfeeding self - efficacy influences breastfeeding behavior within a small sample of African American women. Results within our study showed that BF self - efficacy measured at var ious timepoints (late pregnancy, 6 and 20 weeks postpartum) was positively associated with increased BF behavior (breastfeeding duration and intensity measured at 6 and 20 weeks postpartum) in this sample of African American women. These results confirm pr ior published research. Several studies have reported the importance of measuring maternal confidence, which has been shown to be a key predictor of BF duration. 15 104 106 131 Continual assessment of this modifiable factor allows time to compare BF self - efficacy to the last assessment, which could then allow time for tailored individualized interventions. Our findings s howed that BF self - efficacy measured during late pregnancy and six weeks post - partum truly influenced BF behavior in all participants, specifically duration which was measured in days (Tables 6 and 8) and intensity at 6 weeks (Tables 12 and 16) and 20 weeks (Tables 18 and 20) postpartum. However, this did not persist when BF self - efficacy measured during late pregnancy was assessed a t 20 weeks postpartum (Table 14) . One explanation for not detecting an association of breastf eeding behavior at this time point could be that the tool used was not suitable for assessing BF self - efficacy during pregnancy; the questions asked may not have been appropriate for that particular timepoint . The breastfeeding self - efficacy tool was origi nally created to be assessed in postpartum women. Another reason may have been a - efficacy may change over time or be impacted within a certain timeline (pregnancy to postpartum) , hence the importance of continual testing and creating an asses sment that could be used repeatedly in order to detect 105 confidence while BF. Tuthill and colleagues discussed that there are no current tools to date that can be administered over time to assess BF self - efficacy, especially between anticipated versus actual experienced self - efficacy. 138 The BFSE - SF was originally developed to measure BF self - efficacy among postpartum populations. Although, we adapted the BSES - SF tool for use in our study, our results (from BF self - efficacy being measured during late pregnancy, six weeks and 20 weeks po stpartum) further reiterate the strong need for a new tool to adequately assess BF self - efficacy longitudinally. 138 139 138 Mckinley et al. 139 recently developed a scale to measure prenatal BF self - efficacy; their tool showed that prenatal BF self - efficacy among participants was highly correlated with BF intent ion. However, it could not confirm postpartum BF behavior since it was not measured within the study. Therefore, a strong need remains for the development of a tool that can assess BF self - efficacy prenatally as well as postpartum. Assessing BF self - efficacy during late pregnancy might be beneficial, by potentially allowing sufficient time for either an intervention or tailored individualized education to occur, in order to either increase or continue maintenance of self - efficacy in preparation f or postpartum BF. 139 Jager and colleagues identified from their systematic review that BF self - efficacy is not only an important psychosocial pre dictor in exclusive BF, but a modifiable factor that can be changed through intervention and experiences. 135 131 similarly suggested incorporating confidence building s trategies prenatally may also help women come up with solutions to common BF problems. By assessing the usual care and intervention groups, we did find that breastfeeding self - efficacy at late pregnancy and six weeks postpartum was positively associated with breastfeeding duration in the usual care group only. This group could have potentially driven the findings/analyses when assessed together. When breastfeeding self - efficacy was assessed by 106 group, it was found that this predictor was positively associ ated with breastfeeding intensity at six weeks in both groups even when controlling for other factors. Interestingly, WIC enrollment was negatively associated with increased BF duration and intensity measured at 20 weeks postpartum. This finding is consi stent with findings in the literature where BF rates among WIC participants remain below those of other low - income infants who may be eligible but not participating, and even those who are ineligible. 140 While Although WIC is a program to encourage breastfee ding, it is still well known for providing free formula to low income mothers. A mother may be experiencing breastfeeding issues, and when finally connected with WIC , could potentially be offered and accept the free formula since it is known that it is ava ilable if needed. Therefore , this pre - existing knowledge and expectations c ould be a contributor to the negative association observed in breastfeeding duration and intensity. Studies measuring BF self - efficacy have reported that measuring maternal BF confidence increases over time. 115 141 15 Specifically, one study found that BF self - efficacy continued to increase over the first 4 months among mothers who initiated exclusive BF, but BF self - efficacy declined among those who experienced difficulties. 142 We know based on the literature that the 13 15 19 104 106 - 108 143 , but this would also give health care providers, and/or other tra ined para - professionals a chance to intervene if BF is not - efficacy assessment. Tuthill and colleagues mentioned that BF self - efficacy allows for precise measurement of task and action - specific behaviors, and has the adva ntage of being a manipulated factor. 138 This would provide an opportunity based on the need , of going beyond providing general BF recommendations, and actually tailoring education to specified needs. 144 107 There is a multitude of risk factors documented within the literature that appear to be associated with the prevalence of low BF r ates among African American women. It is important to understand modifiable factors, such as BF self - efficacy, since it has not been studied extensively among this population. We need to gain a better understanding of how this tool can be utilized in vario us settings to assist African American women who have disproportionally low BF rates. Studying this predictor was a strength of this study, since it has not been studied extensively among this population. Another strength was the design of a randomized con trolled longitudinal pilot study. Randomization is the most rigorous method in determining whether there is a relationship between an intervention and outcome , which helps to reduce any bias. This design allowed us to follow the change over time with parti cipants. The main limitation of this study was not having a tool available to assess BF self - efficacy longitudinally (during pregnancy through post - partum). Although some associations were evident between late pregnancy and post - partum, small modification s were necessary for use between pre and post - partum. In addition , the exploratory nature of the study resulted in a small sample size, and the criteria for eligibility of participants who expressed their intention to breastfeed may have added to the limit ations of this study. 4 . 6 Future Implications The findings of this small study indicate the importance of assessing BF self - efficacy among African American mothers specifically as an agent to increase BF duration and BF intensity leading to exclusivity. W e observed that participants within the intervention group had increased breastfeeding intensity (meaning on an average were using breast milk more often) compared to participants in the usual care group. This increase could be due to the intervention wher 108 use the BF self - efficacy survey as a baseline of gauging maternal confidence and a guide for tailoring the intervention based on the score of the participants. Unfortun ately, that was not done within this study. However, by knowing that BF self - efficacy is a modifiable factor, and a significant predictor in successful BF, future studies utilizing interventions to increase breastfeeding duration or exclusivity targeting A frican American mothers should assess and tailor breastfeeding education continually based on the use of the breastfeeding self - efficacy survey, as well as continue monitoring maternal confidence in practical settings. Continual assessment gauging BF confi dence during pregnancy and postpartum is essential with this group because of their low BF rates and high risk for BF cessation. 109 Chapter 5 - Perceived Social Support measured during late pregnancy, 6 and 20 weeks post - partum Predicts Breastfeeding Behavior in African American Women Target Journal: Journal of the Academy of Breastfeeding Medicine 5 . 1 Abstract: Introduction: There are many barriers documented within the literature that can influence nd to have additional factors affecting not only their choices with infant feeding , but also breastfeeding experiences resulting in a persistent health disparity. The lack of social support has been identified as a key barrier in numerous studies. The purp ose of this study was to determine if there was an association between perceived social support by AA women during mid - late pregnancy and breastfeeding outcomes measured at 6 and 20 weeks postpartum (PP). Methods: AA women were recruited in mid - late pregn ancy from a prenatal clinic in Detroit, MI; then enrolled in a 2 - arm randomized, controlled, feasibility trial of BF support and PP weight retention. Data were collected at enrollment (n=53), 6 (n=30) and 20 (n=31) w ee ks PP. Surveys included socio - demograp hics, social support, breastfeeding self - efficacy, maternal diet, and infant feeding practices. Stepwise linear regression determined predictors for inclusion in a multiple linear regression model . Results: Data presented are from women with complete data on targeted variables. Age ranged from 18 - higher mean of perceived social support when measured at late pregnancy compar ed to 20 w ee ks PP. Perceived social support measured at 6 weeks PP was positively associated with BF duration (measured in days) both groups were combined. Perceived Social Suppor t measured prenatally was associated with BF Intensity at 6 w ee ks PP in the full adjusted model ( =.035, p<0.05), but not at 20 w ee ks PP. 110 Perceived Social Support when measured at 6 weeks PP was also associated with BF intensity in the usual care group ( =.070, p<0.05), but was no longer significant in the combined group. Conclusion: Assessment of existing sources of perceived social support during late pregnancy and 6 w ee ks PP could provide a better understanding of a mother s needs for achieving breastf eeding goals , particularly among African American women. 111 5 . 2 Introduction Breastfeeding is an optimal source of nutrition and an effective preventive health measure for infants because of its unique qualities. 2 25 - 27 Breastfeeding is associated with reduced mortality and morbidity in infants and children well into the second year of life. 28 - 30 Studies have concluded that breastfeeding provides many protective factors that increase resistance to many illnesses . 31 In addition, breastfeeding provides many physical, psychological, and social advantages for infants and moth ers. 145 With there being numerous short and long term benefits of breastfeeding, it remains a top public health priority world - wide. The World Health Organization, UNICEF, and American Academy of Pediatrics all recommend exclusive breastfeeding during the first 6 months of life, with continue d breastfeeding up to 2 years of age. 6 25 35 Despite these recommendations, rates of breastfeeding have remained suboptimal overall but e specially among African American (AA) women . While breastfeeding rates have increased substantially in last decade among all mothers, African American mothers still have the lowest rate of breastfeeding initiation and duration. The Centers for Disease Con trol and Prevention (CDC) noted in the National Immunization Survey from 2011 - 2015 130 that the percentage of women who initiated breastfeeding was 64.3% for African Americans, 81.5% for Whites, and 81.9% for Hispanics. Rates for exclusive breastfeeding through 6 months was 14.0% for African Americans, 22.5% for Whites, and 18.2% for Hispanics. Michigan WIC data showed that African American women had the lowest rates for breastfeeding initiation and duration measured at 3 months, 77.3% and 35.2% compared to Non - Hispanic whites 86.3% and 55.7% respectively. 146 These rates show that there is a persistent gap between African American and white women in all aspects of breastfeeding, especially continuation, which has remained consistent for over a decade. 112 Although the benefits of breastfeeding for the mother and infant are significant and well known, studies have shown that women face complex barriers, often influenced by many aspects from meeting the breastfeeding goals they set. 1 While all women experience var ious factors affecting breastfeeding and its sustainability, African American women tend to face complex dynamics and factors shaping not only their decisions, but also views and attitudes towards breastfeeding, which account for the observed and persisten t disparities in breastfeeding rates and associated health outcomes. 18 Along with economic and edu cational factors, African American mothers face additional barriers which include cultural, social and psychological beliefs and factors, perceptions of motherhood, ideas about and attitudes towards breastfeeding from social and personal support systems, l ife experiences, and beliefs about their bodies. All of these areas have been shown to impede breastfeeding. 1 51 93 147 Mitra and colleagues concluded that women who are white, college educated, and middle class with small families, are more likely to breastfeed than their low - income, less educated counterparts, and more likely to breastfeed than African American women. 38 Dubois and Girard 39 tional level and socioeconomic status with an increased likelihood of not only breastfeeding, but also doing so exclusively. These factors and specific characteristics of African American mothers should be considered when assessing breastfeeding outcomes in order to meet breastfeeding initiatives that have been set. In this regard, another factor that is essential and often plays an important role in influencing breastfeeding behavior , 40 82 95 A number of studie s concluded that social support plays a major role in the continuation or early cessation of breastfeeding, and may vary depending on age, income, ethnicity, or culture. 8 95 98 Social support 113 varies within the literature in terms of how it may be defined and measured. As defined by Robbie Hughes, perceived social support can be classified into thr ee groups: emotional (involves receiving empathy, love, trust, and care), tangible (aid or services that can directly support the person in need), and informational (receiving advice, directives, suggestions, and or information) support, which are based on theoretical definitions described by House and Cobb. 96 The source includes support groups, lactation consultants, or physicia ns. 98 99 Differing amounts, and the type of support, can be experienced at different times and from different members social network. 99 African American women may make infant feeding decisions based not only on their current lifestyles or comfort level, but also due to the influence of their social support networks. 8 41 8 41 Several studies have concluded that social support networks positively influences breastfeeding behavior; much of the accumulated literature utilized methodology where social support was primarily assessed qualitatively. Findings from the qualitative study by Barona - Vilar and colleagues 100 showed that women with higher socioeconomic status highly - feeding decisions. They also acknowledge d the significance of formal health support, and employed mothers expressed a need for more institutional support. In contrast, women from lower socioeconomic backgrounds stated that friends were their closest social network and had the greatest influence on their feeding decisions. 100 Asidou and colleagues 8 reported from their ethnography study concerning first time African American mothers, that while mothers had the intention to exclusively breastfeed, few mothers foll owed through with translating their intentions into behaviors. In reference to their qualitative interviews, participants reported having both positive and negative conversations 114 around breastfeeding with family members and friends. Such conversations lead to questions, doubts, and concerns over not only their desire, but also their ability to breastfeed. Assessing social support qualitatively is essential and helpful for identifying underlying associations, such as the quality of social support. However, it is also very important to be able to quantify and illuminate which areas/ constructs of social support received by African American women were perceived as supportive. In addition to illuminating these constructs of social support, it is also imperative to understand how they influence breastfeeding outcomes. Highlighting or being able to quantify these areas of support, will assist health care providers in for achieving their breastfeeding goals. also to the continuance of breastfeeding, 98 98 sup port system may be perceived as supportive, based on how often support is received, whether it is continued, and what person(s) are identified within her support network. Further research is needed to determine how breastfeeding social support is manifeste d specifically for African American women. Despite the growing research focused on reasons for low breastfeeding rates among African American mothers and methods to increase them, breastfeeding disparities still exist. As breastfeeding initiation rates b egin to meet the Healthy People 2020 goals, more effort is still needed to increase breastfeeding duration and exclusivity rates among African American women. The purpose of this study is to understand how perceived breastfeeding social support measured at three time points (late pregnancy, 6 and 20 weeks postpartum,) influences breastfeeding behavior. The study includes breastfeeding (duration and intensity) at 6 and 20 weeks 115 postpartum. Another purpose of this study is to gain a better understanding of th e individuals that African American women report and perceive as supportive within social support constructs. This will provide insight into the representations of support within the different spheres (from hospital, family, and work context) that impact A success. More specifically, the primary aim of this study was to determine if and how social support impacts breastfeeding self - efficacy, initiation, duration, and intensity. This study addresses the following research questions and hypotheses: 1. Does Breastfeeding Social Support (BSS) or Breastfeeding Self - Efficacy (BSE), or the interaction between BSS and BSE measured in late pregnancy, prior to a breastfeeding intervention, predict breastfeeding initiation? Hypothesis : It is hypothesized that Breastfeeding Social Support (BSS) or Breastfeeding Self - Efficacy (BSE), or the interaction between BSS and BSE measured in late pregnancy will positively influence breastfeeding initiation among the participants. 2. Does Social Sup port and Breastfeeding Self - Efficacy predict breastfeeding duration? Hypothesis: It is hypothesized that Breastfeeding Social Support (BSS) and Breastfeeding Self - Efficacy (BSE), or the interaction between BSS and BSE will positively truly influence breas tfeeding duration among the participants. 3. Does Social Support and Breastfeeding Self - Efficacy predict breastfeeding intensity for the participants? Hypothesis: It is hypothesized that Breastfeeding Social Support (BSS) and Breastfeeding Self - Efficacy (BSE), or the interaction between BSS and BSE will positively influence breastfeeding intensity among the participants. 116 5.3 Methods Participants were en rolled in conjunction with a small, randomized controlled trial dual intervention, with a focus on breastfeeding duration and decreasing postpartum weight retention labeled MAMA BEAR (R21HD085138 - 02). Eligibility criteria included pregnant African American women who 1) were 24 - challenge test that visit, 2) had no known history of fetal or chromosomal anomalies, 3) were fluent in English, 4) were interested in breastfeeding, 5) and had reliable internet access (via smartphone, tablet, or computer). Exclusion criteria included breastfeeding contraindications (HIV positive status, chronic therapy with medications incompatible with breastfeeding, known drug or alcohol abuse, and active untreated tuberculosi s); underweight with a pre - pregnancy BMI < 18.5; any acute or chronic illness for which general healthy eating and physical activity would be contraindicated e.g. as determined by the attending obstetrician. Research staff screened, recruited, and consente d eligible women to participate in the study during their prenatal visit at the Henry Ford New Center One Clinic in Detroit, MI. After enrollment, study staff contacted participants (within one week) to complete interviewer administered surveys by phone be fore randomizing active members into the intervention arm or usual care group. During the phone interview, sociodemographic, health/pregnancy, maternal diet, breastfeeding self - efficacy, and breastfeeding social support questions were asked. Following rand omization, trained research staff contacted participants within the intervention group to meet in person and provide an incentive in addition to prenatal education focused on breastfeeding and weight management. Participants in the control group received a n incentive and continued with their usual care. Apart from the intervention, (between two - six and twenty weeks post - partum) all participants received a gift card. The institutional review boards at Michigan State University 117 (R21HD085138 - 02) and Henry Ford Health System (HFHS) ( 10403) approved all study protocols. For the Mama Bear pilot study, a total sample size of 80 women was pre - determined to detect at least a moderate effect size of 0.65, with a power of 80% based on research objectives approved in the NIH funded grant. However, because of extenuating circumst ances, as described in Chapter 4, we recruited fewer participants than originally planned, and lost additional participants during follow up. Since this study was done in conjunction with the Mama Bear study, the same protocols were followed ending with a sample size of n=53 at recruitment. 1. Measures: The Hughes Breastfeeding Support Scale (HBSS) 9 6 was used to measure perceived breastfeeding social support at three different time points (during late - pregnancy, two - six weeks postpartum, and twenty weeks post - partum). The HBSS instrument includes a 30 - item questionnaire, using a 4 - point Likert - typ e scale that focuses on different types of breastfeeding support a mother is receiving. Within the 30 - item questionnaire, there are three different constructs (emotional, informational, and instrumental support), each consisting of 10 questions that can e ither be analyzed together or separate. Responses range from 1 (none at all) to 4 (as much as I wanted). Scoring was completed by adding ratings for the items from each of the three constructs, achieving three sub - scores. We then obtained a total support s core, ranging from 30 to 120 for each participant, by aggregating the three sub - scores together. The three sub - scores and total score were used for data analysis in this study. The total score aggregated remained as coded, providing a continuous variable f or analyses. Inadequate breastfeeding support was indicated with low scores, while high scores indicated adequate breastfeeding support, however there is no documented cut point. 118 To capture additional details from each participant regarding who they consid ered as supportive, the primary researcher adjusted the HBSS survey. For each social support question, participants were asked to choose a maximum of three individuals who they considered most helpful / supportive. Participants ranked them in order, with participants to choose from included: significant other, mother, sister, close friend, nurse, other relative, lactation consultant , pediatrician, Obstetrician, peer counselor, or other person (open field). The BSES - SF 106 is a single construct questionnaire including 14 questions using a 5 - point Liker t scale for measurement. Aggregating the scores for the items result in a total score with a possible range from 14 to 70. BF self - efficacy was analyzed as a continuous variable included as one of the covariates for aim 2 analyses. Breastfeeding Self - Effic acy is described in full detail in Chapter 4 when addressing Aim 1. The BF intensity (% of feeds from BF versus Formula) questionnaire was used as a measure for the main outcome variable. The following question was asked at 2 - 8 weeks and 20 weeks post - par variable as described in Chapter 4 addressing Aim 1. 124 Breastfeeding duration was another outcome variable, which was calculated from the lected later than 20 weeks postpartum, the end date was truncated to 5 months (140 days) for those participants who were still breastfeeding. 119 In reference to the breastfeeding intensity question, the outcome variable, breastfeeding initiation was coded in to a nominal variable. However, it was not a part of the analyses since the majority ( 75% ) of the participants had initiated breastfeeding. 2. Statistical Analyses: between social support and breastfeeding self - efficacy to address the component of aim 2 of understanding how perceived social support influences breastfeeding self - efficacy. Following the estimated regression, multi - collinearity was measured using varian ce inflation factors (VIF) for each of the models. Results showed both predictor variables, perceived social support and breastfeeding self - efficacy, as being moderately correlated with a VIF of 1.67. Since perceived social support was the primary predicto r of interest for aim 2 , self - efficacy was eliminated from the model(s). Therefore, the analysis in aim 2 only included perceived total social support and adjusted for additional variables to address the research questions. Individual level descriptive st atistics were completed based on the socio - demographic variables of the study participants, including the following: age, marital status, employment, education, WIC enrollment, and number of pregnancies. The demographic variables age, employment status, ma rital status, education, WIC enrollment, and number of pregnancies were all coded as categorical variables . The mean of social support was provided for participants at each time point (late pregnancy, 6 and 20 weeks post - partum). Stepwise linear regressi on was used to determine the best predictors to include in the model. Multiple linear regression was used to determine if total social support predicted breastfeeding duration and intensity measured at 6 and 20 weeks post - partum. All models conformed to the assumptions of linear regression. Unstandardized beta coefficients were 120 reported in the tables. However, significant standardized beta coefficients are also reported and interpreted in the text. Stata version 15.0 was used for analyses. P values 0.05 were considered statistically significant. 5 . 4 Results Table 5. 1 displays the characteristics of all the participants (n=53) enrolled in the study, and the distribution of variables of interest. Participants were predominantly between the ages of 18 - 30 (67.92%), and had received some college education or more (62.3%). Most of the participants had never been married (61.76%). The majority were currently enrolled in the Women, Infants, and Children program, WIC (52.83%) and had been pregnant three or more times (52.83%). Results from Table 2 show that all women responded with having reasonably high levels of perceived social support during late pregnancy (93.69), and at 6 weeks post - partum (93.77). T he mean perceived social support was a little lower when measured at 20 weeks (85.47) compared to the other time points. When the individual constructs were assessed, results showed emotional support with the highest mean points, and informational support with the lowest mean at all three time points. Informational support had a mean of 27.66 when measured at late pregnancy, and showed a decrease when measured at 6 and 20 weeks post - partum to 25.31 and 24.34, respectively. Scatter plots from figures 1 - 3 sho w a positive correlation of perceived social support and breastfeeding self - efficacy when measured at three time points. There was an increase in perceived social support when measured at late pregnancy as shown in Figure 5.1 Results from Figure 5.2 show t hat perceived social support was relatively stagnant when measured between 2 - 8 weeks postpartum. In Figure 5.3 , perceived social support - mom e nt correlation was run to assess the relationsh ip between perceived social support and breastfeeding self - efficacy in 121 African American women who participated in the study. Results showed that there is a moderately positive relationship (.442, p<.001) between social support and breastfeeding self - effica cy based on 48 observations. Table 5.1 Demographic Characteristics of All Participants During Late Pregnancy (N=53) Baseline Characteristics All Participants (%) Age in years 18 - 30 36 (67.9) 31 - 45 17(32.1) Marital Status Married 9 (17) Living with a Partner 14(26.4) Never been married 30(56.6) Employment Status Employed 25(47.2) Unemployed, Looking for Work 28(52.8) Education High School Diploma, GED, or Equivalency or less 20(37.7) Some College 23(43.4) 4 Year College Degree, or More 10(18.9) WIC enrollment Yes 28(52.8) No 25(47.2) Pregnancies including current One or two 25(47.2) Three or more 28(52.8) 122 Social Support Scale aggregation of emotional, tangible and informational constructs (scores range from 30 - 120) *Construct Scores range from 10 to 40 Figure 5. 1 Relationship B etween Breastfeeding Self - Efficacy and Perceived Social Support M easured D uring L ate P regnancy Table 5.2 Summary of Social Support Constructs Predictors Measured Mean (Standard Error) Late Pregnancy N=51 6 weeks N=33 20 weeks N=34 Perceived Social Support Total 93.69(3.42) 93.77(3.12) 85.47(3.74) Individual Constructs Emotional Support 33.86(1.31) 36.26(1.16) 34.03(1.55) Tangible Support 31.88(1.25) 30.60(1.32) 27.32(1.44) Informational Support 27.66(1.37) 25.31(1.31) 24.34(1.45) 123 Figure 5.2 Relationship B etween Breastfeeding Self - Efficacy and Perceived Social Support M easured Between 2 - 8 W eeks Post - Partum Figure 5. 3 Relationship B etween Breastfeeding Self - Efficacy and Perceived Social Support M easured at 20 W eeks Post - Partum 124 Table 5. 3 displays predictors measured at late pregnancy influencing breastfeeding duration for the participants, as specified by group arrangement and combined. Results from the fully adjusted model showed that none of the predictors (perceived social support or additional covariates ) were associated with breastfeeding duration in either the usual care or intervention group. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 5.3 Association of Predictors M easured at L ate P regnancy I nfluencing Breastfeeding Duration Usual Care N=20 B (SE) Intervention N=21 B (SE) Combined N=41 B (SE) Intercept 8.13(102.97) 82.36(64.37) 63.31(48.19) Perceived Social Support (at late pregnancy) .557(.636) .051(.585) .194(.372) Some College 47.45(34.01) - 10.97(30.72) 17.15(20.04) College Degree - .575(47.87) - 3.72(37.20) 8.33(27.08) WIC (enrolled) - 35.15(30.33) 5.53(26.52) - 17.07(17.66) Marital Status (ref category married) Never been married 27.95(53.37) 1.19(42.91) 3.13(28.29) Living with a partner 34.25(45.57) 10.68(39.41) 16.50(26.75) R 2 0.33 0.02 0.07 Adjusted R 2 0.03 - 0.39 - 0.09 Omnibus 0.42 0.99 0.84 125 Table 5. 4 displays results from the fully adjusted model where total social support was measured at 6 weeks postpartum, influencing breastfeeding duration for the participants, specified by group arrangement, a s well as a combination of both groups. Results from the fully adjusted model showed that none of the predictors (perceive d social support or additional covariates) were associated with breastfeeding duration in the intervention group or the combined group. Interestingly, perceived social support was positively associated with breastfeeding duration (increased days of being b reastfed) in the usual care group, but was no longer significant when both groups were combined ( B = 1.33, p<.05). The standardized coefficient results show that with every one increase in perceived social support measured at 6 weeks post - partum, breastfeeding duration increased by .584 units , assuming the other adjusted variables are held constant. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 5. 4 Perceived Social Support M easured at 6 W eeks Post - Partum I nfluenc es Breastfeeding Duration in the Usual Care Group Usual Care N=17 B (SE) Intervention N=16 B (SE) Combined N=33 B (SE) Intercept - 41.82(56.35) 131.54(239.40) 16.67(62.50) Perceived Social Support (at 6 weeks post - partum) 1.33(.478)* .210(2.22) .991(.535) Some College 25.91(28.74) - 27.95(39.96) 4.41(22.16) College Degree 38.85(42.30) - .947(50.37) 15.37(27.99) WIC (enrolled) - 40.42(24.89) - 25.12(38.77) - 36.91(18.79) Marital Status (ref category married) Never been married 17.82(31.33) - 72.14(65.45) - 19.89(30.17) Living with a partner 32.88(27.25) 44.17(59.78) 5.24(27.26) R 2 0.65 0.30 0.30 Adjusted R 2 0.45 - 0.16 0.14 Omnibus 0.05 0.69 0.12 126 Table 5. 5 depicts the models of the predictors: perceived total social support measured at 20 weeks postpartum, education, WIC enrollment, and marital status influencing breastfeeding duration for th e participants, specified by group arrangement and combined. None of the predictors (perceived social support or additional covariates) were associated with breastfeeding duration in either the intervention group or the combined group. Interestingly, WIC e nrollment was negatively associated with breastfeeding duration in the usual care group ( B = - 63.39, - .606, p<.05). P articipants enrolled in WIC showed a decrease in breastfeeding duration by .606 units , assuming the other adjusted variables are held co nstant. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 5.5 Predictors M easured at 20 W eeks Post - Partum I nfluencing Breastfeeding Duration Usual Care N=17 B (SE) Intervention N=17 B (SE) Combined N=34 B (SE) Intercept 33.46(48.34) 164.18(99.64) 71.13(46.37) Perceived Social Support (at 20 weeks post - partum) .783(.503) - .214(.910) .522(.438) Some College 26.74(29.51) - 21.14(37.12) 9.08(21.64) College Degree - 3.25(43.05) - 9.59(39.16) - 3.29(26.65) WIC (enrolled) - 63.39(24.10)* - 26.57(30.01) - 47.23(17.69) Marital Status (ref category married) Never been married 55.59(36.13) - 21.50(53.11) 26.07(29.65) Living with a partner 20.62(27.83) - 26.89(44.60) 5.95(24.22) R 2 0.67 0.12 0.28 Adjusted R 2 0.47 - 0.40 0.13 Omnibus 0.04 0.96 0.13 127 Table 5. 6 shows that with every one increase in perceived social support measured at late pregnancy, breastfeeding intensity in the combined group increased by .370 units , holding all other variables constant ( B = .034, =.370 , p<.05). With every one increase in p erceived social support measured at late pregnancy in the usual care group, breastfeeding intensity at 6 weeks post - partum increased by .595 units , assuming other variables were held constant ( B = .059, = .595 , p<.05). Interestingly, participants who had a college degree on average decreased the use of breast milk compared to those who completed high school in the usual care group ( B = - 6.32, = - .913 , p<.05), and remained significant when both groups were combined ( B = - 3.49 - .559 , p<.05). Standard errors in parenthes es *** p<0.001, ** p<0.01, * p<0.05 Table 5. 6 Perceived Social Support M easured at L ate P regnancy I nfluences Breastfeeding Intensity at 6 W eeks Post - partum Usual Care N=18 B (SE) Intervention N=14 B (SE) Combined N=32 B (SE) Intercept - .692(4.26) 3.65(3.62) 1.04(2.44) Perceived Social Support (at late pregnancy) .059(.026)* .021(.026) .034(.016)* Some College - 1.38(1.61) - .758(1.65) - .827(1.03) College Degree - 6.32(2.35)* - 3.44(1.82) - 3.49(1.26)* WIC (enrolled) - 2.29(1.39) 1.23(1.41) - .636(.844) Marital Status (ref category married) Never been married 1.06(2.16) - 3.90(2.94) - .817(1.46) Living with a partner 1.99(1.86) - 1.88(2.75) .791(1.39) R 2 0.56 0.43 0.35 Adjusted R 2 0.32 - .0.05 0.19 Omnibus 0.10 0.55 0.07 128 Table 5. 7 shows that perceived social support measured at late pregnancy was not associated with breastfeeding intensity at 20 weeks post - partum in either group, or when both groups were combined. Interestingly, participants enrolled in WIC had a decrease use of breast milk ( b reastfeeding intensity ) when measured at 20 weeks post - partum ( B = - 3.31, = - .651 , p<.05) in the usual care group, and this remained significant when both groups were combined ( B = - 2.32, = - .474 , p<.01) . Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 5.7 Predictors M easured at L ate P regnancy Influences Breastfeeding Intensity at 20 W eeks Post - Partum Usual Care N=17 B (SE) Intervention N=15 B (SE) Combined N=32 B (SE) Intercept 2.89(4.71) 2.75(3.52) 1.71(2.37) Perceived Social Support (at late pregnancy) .008(.031) .015(.027) .019(.016) Some College .925(1.47) - .766(1.74) .352(.99) College Degree - .847(2.39) - 2.62(1.730 - 1.78(1.20) WIC (enrolled) - 3.31(1.36)* - 1.11(1.38) - 2.32(.818)** Marital Status (ref category married) Never been married - 1.25(2.64) - 1.86(2.58) - .907(1.50) Living with a partner .054(1.99) - .690(2.27) .124(1.28) R 2 0.50 0.34 0.35 Adjusted R 2 0.20 - 0.15 0.19 Omnibus 0.22 0.66 0.06 129 Table 5. 8 provides results from the fully adjusted model which include s predictors: perceived total social support measured at 6 weeks post - partum, education, WIC enrollment, and marital status influencing breastfeeding intensity at 6 weeks post - partum for the participants, specified by group and combined. Results showed that perceived social support was positively associated with increased breastfeeding intensity at 6 weeks post - partum in the usual care group ( B = .070, =.591 , p<.05), but not in the intervention group. In addition, results showed that having a college degree versus having completed high school ( B = - 6.31, = - .801 , p<.05) and being enrolled in WIC ( B = - 3.88, = - .765 , p<.05) were negatively associated with breastfeeding intensity at 6 weeks postpartum in the usual care group. None of the predictors were fo und to be significant in either the intervention group, or combined groups. Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 5. 8 Perceived Social Support I nfluences Breastfeeding Intensity at 6 W eeks Post - Partum Usual Care N=17 B (SE) Intervention N=16 B (SE) Combined N=33 B (SE) Intercept 1.84(3.28) 5.38(10.50) 1.70(3.09) Perceived Social Support (at 6 weeks postpartum) .070(.028)* .011(.097) .039(.026) Some College - 3.30(1.67) - .242(1.75) - .885(1.09) College Degree - 6.31(2.46)* - 2.01(2.21) - 2.55(1.38) WIC (enrolled) - 3.88(1.45)* .651(1.70) - 1.19(.929) Marital Status (ref category married) Never been married - .783(1.82) - 4.50(2.87) - 1.69(1.49) Living with a partner .322(1.58) - 3.51(2.62) - .518(1.35) R 2 0.56 0.27 0.22 Adjusted R 2 0.31 - 0.21 0.04 Omnibus 0.12 0.75 0.32 130 Table 5. 9 displays results from the fully adjust ed multiple linear regression of the predictors: perceived total social support measured at 6 weeks post - partum, education, WIC enrollment, and marital status influencing breastfeeding intensity at 20 weeks post - partum for the participants, specified by gr oup and combined. Results from the fully adjusted model showed that perceived social support was not associated with breastfeeding intensity at 20 weeks post - partum in either group, or when both groups were combined. Interestingly, participants enrolled in WIC had a decrease d use of breast milk (breastfeeding intensity) when measured at 20 weeks post - partum ( B = - - .503, p<.05). Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 5.9 Association of P redictors M easured at 6 W eeks P ost - P artum I nfluencing Breastfeeding Intensity at 20 W eeks Post - Partum Usual Care N=15 B (SE) Intervention N=12 B (SE) Combined N=27 B (SE) Intercept 1.91(4.10) 8.78(13.46) 4.78(3.29) Perceived Social Support (at 6 weeks postpartum) .003(.037) - .017(.131) - .001(.029) Some College 2.02(2.24) .023(2.24) .706(1.16) College Degree 1.62(3.27) - 1.71(2.97) - .631(1.37) WIC (enrolled) - 2.37(1.70) - 1.22(2.46) - 2.66(.938)* Marital Status (ref category married) Never been married - 2.84(2.45) - 5.75(4.75) - 2.89(1.75) Living with a partner .233(1.78) - 4.15(2.80) - 1.20(1.29) R 2 0.53 0.56 0.44 Adjusted R 2 0.18 0.04 0.27 Omnibus 0.28 0.47 0.05 131 Table 5. 10 shows the results from the fully adjusted multiple linear regression of the predictors: perceived total social support measured at 20 weeks post - partum, education, WIC enrollment, and marital status influencing breastfeeding intensity at 20 weeks post - partum fo r the participants, specified by group arrangement and combined. P erceived social support was not associated with breastfeeding intensity at 20 weeks post - partum in either group or when both groups were combined. Participants enrolled in WIC had a decrease d use of breast milk (breastfeeding intensity) at 20 weeks post - partum in the usual care ( B= - 3.31 - .654 , p<.05) and this finding remained significant when both groups were combined . Standard errors in parentheses *** p<0.001, ** p<0.01, * p<0.05 Table 5.10 Association of Predictors M easured at 20 W eeks Post - Partum I nfluencing Breastfeeding Intensity at 20 W eeks Post - Partum Usual Care N=17 B (SE) Intervention N=17 B (SE) Combined N=34 B (SE) Intercept 4.36(2.89) 5.26(4.42) 4.57(2.21)* Perceived Social Support (at 20 weeks postpartum) - .005(.030) .004(.040) .001(.021) Some College 1.01(1.76) - .883(1.65) .159(1.03) College Degree - .401(2.57) - 1.86(1.74) - 1.16(1.27) WIC (enrolled) - 3.31(1.44)* - 1.44(1.33) - 2.54(.844)** Marital Status (ref category married) Never been married - 1.72(2.16) - 3.07(2.35) - 2.04(1.41) Living with a partner - .303(1.66) - 2.35(1.97) - 1.07(1.15) R 2 0.50 0.31 0.33 Adjusted R 2 0.20 - 0.09 0.18 Omnibus 0.22 0.61 0.07 132 5 . 5 Discussion Findings from this study showed that total social support during late pregnancy, and at six weeks post - partum, influence d breastfeeding behavior in African American mothers. These findings are corroborated by other studies where social support is a significant predictor of breastfeeding. 8 148 - 150 Specifically, our findings showed that total perceived social support measured at late pregnancy influenced breastfeeding intensity at six weeks post - partum in all participants. When usual care and intervention g roups were assessed separately , we found that total perceived social support measured at six weeks post - partum influenced breastfeeding behavior for both BF duration and breastfeeding intensity at 6 weeks post - partum in the usual care group , but not the co mbined group. There were no associations with total perceived social support measured at late pregnancy and breastfeeding intensity at 20 weeks postpartum. One possible explanation for not detecting results at that time point is that there was too much tim e from when the assessment occurred, during late pregnancy. The literature supports that a antepartum and postpartum periods. Asidou and colleagues , 8 conducted a qualitative study that focused on the influences of social support among first time African American mothers. They study; where some participants thought they would have support, they ended up completing the study without it, and others noted that their family members were unsupportive. These findings highlight the importance of continual assessment of social support is warranted. Interestingly, we did find that WIC enrollment was negatively associated with increased BF durati on and breastfeeding intensity at 6 - and 20 - weeks post - partum. This finding appeared significant several times throughout the analyses. This finding is consistent with findings in the 133 literature where BF rates among WIC participants remain below those of o ther low - income infants who may be eligible but not participating, and even among those who are ineligible. 140 Although this was not a predictor of interest, it is interesting that WIC could serve as, and is specifically, it was negatively associated with continued breastfeeding among this population. These findings could suggest that WIC may be viewed or utilized in a contradicting manner. While it may be viewed as a place for providing breastfeeding support, it is also commonly known as a place to receive infant formula. Therefore, our mothers may have relied more on the accessibility of free formula especially when there were other competing breastfeeding challenges. Several studies have recognized the importance of tailoring messages, having culturally appropriate education, and individualized counseling 1 80 151 but what resources to use, when to use them, and how they can be utilized when meeting with African American mothers in regards to breastfeeding support remains unclear . Using a social support tool , could help generate conversation between the health care provider and the mother about breastfeeding, and help her in establishing breastfeeding goals, and identify those areas where additional or any support is needed. Our findings demonstrate that during two critical periods (late pregnancy and six weeks post - partum) for not only establishing, but continuing breastfeeding is important and useful for identifying challenges a mother may face . Using a social support survey as an assessment tool during these time periods could help mothers begin to think about their social support networks, and who they identify within those to help assist and meet their breastfeeding g oals. The conversation around this topic of social support, and what it may look like for the 134 mother, could help ease the mother into asking additional questions, which could lead to her being provided with the specific resources and/or support she needs t o be successful. Health professionals providing breastfeeding counseling who include this important assessment, would have sufficient time for an intervention, as well as tailored individualized counseling to occur if they identified support needs by late pregnancy through the PP period. Strengths of this study include, the assessment of perceived social support of African American mothers in a longitudinal randomized control trial dual intervention, in order to gain a insight into how this modifiable vari able could potentially influence breastfeeding behavior. While social support has been identified as a significant predictor in breastfeeding behavior, it is important to gain a better understanding on how this predictor can be used to help increase breast feeding among African American mothers who continue to have low breastfeeding rates. One limitation to our study, and a possible explanation for the limited associations with breastfeeding behavior at the various time points, is that social support continu ally decreased from pregnancy to 20 - w ee k PP , which could explain the l imited effect at 20wk PP. Another factor to consider is that some of the questions asked within the survey used were potentially outdated (survey was created in 1984) and may not have been relevant to our mothers . For example, one assessing social support quantitatively exist, which prompt ed us to utilize this tool. 150 While we were able to detect some associations, it highlights the need for a current and culturally relevant breastfeeding social support tool where responses from the assessment can be used in intervention studies or practical health care settings. O ur small sample size and criteria for eligibility of participants with the intention to breastfeed may have added to the limitations of this study . 135 Boateng and colleagues 149 conducted a study in Uganda, assessing exclusive social support for which no scale existed. Their study utilized HBSS, but made modifications to the survey they created, and validated a scale to assess exclusive breastfeeding. Their r esults showed that mothers did not receive the desired emotional or instrumental support indicating the need for interventions to increase support for recommended infant feeding practices. Therefore, the need to create and validate social support tools tha t can assess any breastfeeding and exclusive breastfeeding among various populations is another important aspect for future studies . 5.6 Future Implications The primary purpose of the overarching study was to test the feasibility of implementing a dual - in tervention, which combines home visits from experienced BF peer counselors with text - based postpartum weight management encouragement. Interestingly, we observed perceived social support influence breastfeeding behavior in both the usual care group and combined groups. Although, we assessed perceived social support at multiple time points in the study, it was not the intent of the overarching study to use the survey as a guide for tailoring the intervention based on the score of the participants. However, knowing that social support is a modifiable factor and a significant predictor in successful BF, future studies focused on BF interventions targeting African American mothers should assess perceived social support during late pre gnancy and post - partum. Results from the survey could be used to tailor interventions and educational sessions based on responses to assist in increasing breastfeeding. It is also suggested that health care providers who provide breastfeeding counseling an d education , utilize perceived social support surveys as a critical part of their assessment. Utilizing a breastfeeding social support tool would be useful in determining effectiveness of interventions and programs longitudinally. 138 136 It is well known that positive social support is capable of foster ing confidence in a tfeed. 5 In addition, it has been noted as a public health objective, that health care providers and advocates need to offer more breastfeeding support to African American mothers; but to date, there has been no guidance or steps regarding how that support could be pr ovided using resources that are readily available. The current study findings illuminate s some critical time points (late pregnancy and 6 weeks PP) for assessing an African persons /organizations relative to the social support constructs, and offer insight on how to utilize existing evidence based tools in research and practical settings to help improve breastfeeding rates among African American women. 137 Chapter 6 - Understanding Factors Influencing Breastfeeding Outcomes in a Sample of African American Women Target Journal: Breastfeeding Medicine 6 . 1 Abstract: Background: Breastfeeding (BF) is recognized as the optimal feeding choice for infants because of the numerous health benefits. However, there are several complex factors involved in either encouraging or discouraging BF especially among African Americans (AA) resulting in a persistent disparity. We recognize that many environmental and other factors across the Socio - ecological model (SEM) can impact breastfeeding. Our research expands upon the current literature on barriers and facilitators to BF among AA women, by providing an understanding of how BF social support and self - efficacy influences BF goals by i mproving behavior (duration and exclusivity). In addition, the sharing of unique personal experiences provide s ideas of how AA women who want to breastf eed can be supported. Methods: Fourteen pregnant AA women who breastfed were recruited (from a larger study) around 20 weeks (w ee ks) postpartum (PP) to participate in semi - structured qualitative interviews. Interviews were digitally recorded and audiotape s were transcribed. Inductive thematic content analysis was conducted in combination with a review of field notes. Results: Factors within both micro to macro levels within the SEM either reinforced or discouraged BF as shared by participants. Specific cha llenges identified included : 1) Confidence with breastfeeding and 2) Discouragement from issues associated with breastfeeding . Several facilitators ( Dual benefit for infant and mother, Perseverance, Commitment and Self Determination Fluctuating Confidence , Ability to Pump , and Support from identified Social Support Networks) were identified that assisted women in meeting their breastfeeding goals . Some suggestions for expanding breastfeeding promotion and support included: 1) Tangible and 138 Immediate help, 2) Positive Non - Self - motivation/Will Power/Perseverance . In regard to providing help and support, it was apparent that the mode of education, support and help provided from those identified within their social support networks should be revisited and discussed as these may vary. Conclusion: Women who participated in this study expressed several suggestions for expanding BF support and promotion efforts with AA women. Findings showed that AA women face multiple obstacles when breastfeeding. Collaborating and engaging in discussions with specifically with AA women should be encouraged in order to improve BF support, outcomes and promotion . 139 6 . 2 Introduction Breastfeeding is extensively recognized as the optimal feeding choice for infants because of the numerous associated health benefits. 152 - 154 Breastfeeding rates show that the majority of women initiate breastfeeding, but significantly fewer women continue to breastfeed exclusively until six mo nths. 18 43 Reasons mothers mention for limiting or terminating breastfeeding illuminate some of the challenges they face and those tend to vary. Although, breastfeeding rates are improving through the launch of national breastfeeding initiatives and targeted efforts, they are still disproportionately lower than recommended, especially among certain racial groups. 155 African American women are the le ast likely to initiate breastfeeding, continue breastfeeding to 6 months, or breastfeed exclusively. They have historically breastfed at lower rates than recommended and continue to fall below national goals, but have the highest rates of maternal and infant mortality and morbid ity. 34 129 Improving the health of African American infants and mothers continues to remain an important public health goal for the United States. There are sever al factors known to influence breastfeeding behavior among African Americans. Some factors other than age, education, housing, marital status, and health care that tend to influence breastfeeding include but are not limited to: i ntention to breastfeed, 45 46 attitudes toward infant fee ding methods, 46 47 lactation experiences, maternity care practices, 48 limited lactation support, 155 breastfeeding self - efficacy, 49 social support, 49 50 health care, and non - supportive workplace and childcare environments . 51 It is important to note that these factors influence breastfeeding initiation and duration among all women, but have been shown to be more prevalent among African American women compared to others. Additional factors specific to African Americans that con tribute to breastfeeding disparities and decision 140 to breastfeed include cultural influences, attitudes, and behaviors, stigma, lack of social support, beliefs of inconvenience, pain, and sexuality. 53 155 Spencer and Grassley 10 specifically identified several factors that appear to have a negative impact on African American women and breastfeeding from an integrative review of the literature regarding Fac tors they mentioned include d : early return to work, short maternity leave, less information received about breastfeeding from health - care providers, less access to professional support, embarrassment with breastfeeding, and perceptions around poor milk sup ply. There are many quantitative studies in the literature recognizing factors known to influence breastfeeding in the African American population, but there are more limited qualitative studies that expand these findings to provide strategies/solutions t hat are modifiable such as (breastfeeding social support and self - efficacy) through the voices of African American themselves. Their experiential knowledge and lived experiences are important considerations for professionals who want to enhance breastfeedi ng outcomes. The theoretical foundation used for this qualitative study was the Socio - ecological model (SEM). The SEM dates back to the 1980s where it was proposed as a theoretical perspective for understanding the relationship and underpinnings involved with breastfeeding behavior. 44 117 156 The SEM provides a useful theoretical framework for understanding group level factors that either hinder or support mothers in their ability to breastfeed. Within the integrated con text of intrapersonal characteristics, interpersonal processes, institutional factors, community factors and public policy. It has been suggested that within the SEM there may be overlap at various levels between individuals and their environment. 40 Traditionally, studies often focus on the mother as the sole person responsible for any infant feeding 141 decision, but we recog nize that many environmental factors and different levels across the SEM can impact breastfeeding. 157 The numerous challenges and obstacles that African American women face when breastfeeding; using the SEM framework provides a basis for gaining insight into what group - level factors in the envir onment facilitate or discourage breastfeeding initiation and/or maintenance. Our research expands upon the current literature that has identified barriers and facilitators to breastfeeding among African American women and enhances the interpretation of th e quantitative findings from Aims 1 and 2. The objectives of this qualitative study were to 1) Gain a better understanding of personal experiences of African American women and how breastfeeding social support and self - efficacy may influence the success of breastfeeding goals and overcoming barriers; and 2) Identify new ideas or knowledge that can assist in the support and success of breastfeeding among African American women who want to breastfeed. It was hypothesized that African American women w ould provide insight into unique experiences and factors relative to social support and self - efficacy, which influence their breastfeeding goals and behavior (duration and exclusivity). 6.3 Methods 1. Sample and Recruitment Stage 1 in this study comprised of interviews with 2 key informants who worked with Mama Bear participants and worked with African American women in the community as a part of their jobs. Participants included a Community Health Worker and a Breastfeeding peer counselor. The trained Community Health Worker works within a hospital system where she meets with women to provide them with resources and support for all their needs, including breastfeeding. The Breastfeeding peer counselor, a Certified Lactation C ounselor (CLC) works 142 within the community providing foundational and skilled breastfeeding support. Both of these key informants also worked with Mama Bear participants. The participants were purposively chosen because they self - identified as being Africa n American and they assisted African American women in some capacity with breastfeeding. One interview occurred face to face and the other interview, because of traveling difficulty, was conducted over video conference. The second stage of this study used a convenience sampling approach to recruit 14 African American mothers who breastfed. Participants from the larger Mama Bear study, IRB#16 - 297 (n= 54) were called and invited to participate in individual semi - structured qualitative interviews after 20 wee ks post - partum. At the 20 - week post - partum visit, permission was sought from all Mama Bear participants to be contacted for future studies. Upon receipt of approval for permission to make future contact, the participants from the Mama Bear study were conta cted by the primary researcher using a script for participation in the study. All participants were offered the option to conduct interview over the phone, video conference, or in person to allow flexibility and comfortability. A sample size of 15 - 20 part icipants was sought, based on a previous qualitative study that used semi - structured interviews to determine (n=17) how African American women experience breastfeeding in the context of their day - to - day lives. 81 Recruitment continued until data saturation was reached, i.e. where sufficient data was gathered to fully develop and understand the depth and range of con cepts and relationships. 158 Verbal consent was audio recorded and the primary researcher emailed and mailed a copy of the consent form to participants to keep for their records. Gift cards of $30 were either mailed or provided as e - gift cards after completion of each interview which ranged from 30 to 60 minutes. 143 2. Instruments and Data Collection Procedures The study and procedures were approved by Michigan State University human research protection program. The semi - structured questions were developed based on the SEM as a concept ual framework. Questions were based on aspects of the breastfeeding social support and self - experiences through their voices, specific factors that influence breastfeeding, strategies employed to assist in breastfeeding, and steps needed to assist in increasing breastfeeding outcomes. Results from the interviews with the key informants informed the relevance of questions included in the questionnaire and provided additional questions to ask participants. To ensure content validity, an audit trail tracking each step of the qualitative process was conducted, after which questions were reviewed by four researchers with experience in qualitative research design, who also confirme d face validity of the questions. An individual semi - structured format was chosen and seemed most appropriate since the timing of the interviews need to be staggered because not all women reached 6 months post - partum simultaneously. In addition, it provid ed a more conversational approach, allowing the primary researcher an opportunity to establish trust with participants and elicit richer input. The primary researcher obtained verbal audio recorded informed consent and conducted the semi - structured intervi ews either in - person, over the phone, or via video conference with participants using a paper interview guide. All interviews were audio recorded and additional field notes recorded. At the end of each interview participants were allowed to debrief, where they could add or clarify information and/or correct misunderstandings. A demographic questionnaire was also completed with participants at the end of the interview. Following each interview, the primary researcher stripped the recording of any identifier s to maximize confidentiality and minimize risk for participants. Stripped recordings 144 were transcribed verbatim by an undergraduate research assistant. All digital and document files were stored on encrypted and password protected university sites. 3. Da ta Analysis Digitally recorded interviews were transcribed verbatim by an undergraduate research assistant. To ensure accuracy, transcriptions were reviewed by the primary researcher in conjunction with field notes . Inductive Thematic Analysis occurred ut ilizing six steps as described by Braun and Clark. 159 The primary researcher and another doctoral student (qualitatively trained) became familiar with the data by independently reviewing transcriptions. Each reviewer read transcripts independently, line by line, and generated initial axial codes for each research question. After reviewing each interview, the reviewers came together to discuss any ambiguities that may have arose while reviewing the data. Then, each reviewer generated potential themes from examining all quotes based on an initial coding schema along with highlighting exemplar quotes. The two reviewers discussed and compared results, then refined themes by collapsing or eliminating where necessary. The primary researcher with consensus f rom the other reviewer defined and named the themes by describing the essence of each theme and giving it a compelling name. The primary researcher provided the final report. NVivo 12, a qualitative analysis software was used to assist in content analysis of the transcripts. Descriptive statistics were generated using Microsoft Excel 2016. 145 6 . 4 Results Table 6.1: Demographic Characteristics of African American Qualitative Study Participants ( N =13) Variables N (%) Age 18 - 24 years 3 (23.08) 25 - 34 years 9(69.23) 35 - 45 years 1(7.69) Education Some high school, no diploma 2(15.38) High school graduate, diploma or the equivalent (GED) 5(38.46) Some college credit, no degree 4(30.77) 2(15.38) Number of Children Primiparous 4 (30.77) Multiparous 9(69.23) Marital Status Single, never married 9(69.23) Married or domestic partnership 4(30.77) Occupation Employed for wages 9(69.23) Out of work and looking for work 1(7.69) Out of work but not looking for work 1(7.69) Student 2(15.38) Annual household income Chose not to provide 1(7.69) Less than $10,000 5(38.46) $10,000 to less than $20,000 3(23.08) $20,000 to less than $35,000 2(15.38) $35,000 to less than $50,000 1(7.69) $50,000 to less than $75,000 1(7.69) Completely stopped breastfeeding and pumping milk for your baby Yes 7(53.85) No 6(46.15) Age of the baby when stopped breastfeeding Three months 1(14.29) Twelve months 3(42.86) Fourteen months 1(14.29) Sixteen months 2(28.57) 146 Table 6.2 Challenges with not M eeting B reastfeeding Goals Sub - theme Representative Quotes Confidence with breastfeeding when my breast were not filling with milk, getting heavy or leaking as much as it was at first. That discouraged me, like maybe it i ID 007 told me she was still gaining weight and was in the curve and all that, but she should be eating more ID 002 Discouragement from issues associated with breastfeeding been able - ID 010 you know make you thinking that your baby is not eating enough or ID 002 147 Table 6.3: Successes with M eeting B reastfeeding Goals Sub - theme Representative Quotes Dual benefit for infant and mother ID 014 Baby being healthy, gaining weight and getting nutrients ID 008, - ID 2042 Mother Weight loss ID 016 Infant weight gain and growing ID 003, - ID 2042 Less sickness ID 006 Fluctuating Confidence wellness checks and she hitting all her milestones; could see she ID 2042 ut I am very confident now. ID 008 Perseverance, Commitment and Self Determination I said I was overthinking with it, but at the e nd I said I am going to ID 014 ID 004 Ability to Pump ID 008 what I want to do. Not being a big issue for me to walk away and ID 002 ID 013 ime he emptied a bottle, I re - ID 010 - ID 006 Encouragement, help, and support from identified Social Support Networks ID 007 - ID 013 ID 005, ID 011, ID 003 148 Table 6.1 displays the characteristics of the Mama Bear participants (n=13) who enrolled in the qualitative component of the study, and the distribution of variables of interest. Participants Table 6.4: Suggestions for E xpanding B reastfeeding S upport and P romotion A mong African American W omen Sub - theme Representative Quotes Tangible, Immediate, Proactive help hospital. When you go home, the people around you are not doing - ID 011 , maybe - ID 010 - ID 016 Positive Non - Judgmental Support (Professional and Persona l) ID 009 - ID 006 Education about someone to show me how to use it instead of trying to mimic a ID 2042 Q2: It was the fact that the pumping was taking so long. If I would ID 014 Self - motivation/Will Power/Perseveranc e someo ne else persuade without thinking about it I feel like they will quit. - ID 002 - ID 014 do it. Try to ID 008 ID 004 149 were predominantly between the ages of 25 - 34 (69.23%), multiparo us (69.23%), and s ingle, never married (69.23%). The majority of the participants (53.85%), when interviewed, had completely stopped breastfeeding and pumping milk for their infant. Of those participants who stopped breastfeeding and pumping milk, their in fants where between 12 - 16 months. Two primary themes identified from participant interviews included challenges to and successes for not meeting or meeting various BF goals . Regardless of the challenges or successes described, all participants described ha multiple phrases to describe their most memorable moment while breastfeeding which included: mother and infant bonding, feelings of closeness, and eye contact. Many mothers also described remembering the initial latch of the baby . These moments shared seemed to enhance the breastfeeding experience. 1. Challe n ges with not meeting breastfeeding goals 1. Confidence with Breastfeeding: (about being able to breastfeed) at the beginning loss confidence as they experienced breastfeeding issues which remained unresolved despite attempts for help which lead to feelings of discouragement and ultimately decreased confidence. Participants described having feelings of discouragement and a loss of confidence which impacted their level of breastfeeding. This occurred around issues that were associated with br these kind of stuff makes me discouraged, you know make you thinking that your baby is not eating enough or something eve ID 002 Another participant 150 I think what discouraged me is when my breast were not filling with milk, getting heavy or lea ID 007 . Discouragement from issues associated with breastfeeding: Breastfeeding was described as being something new for several of the participants and something tha t they wanted to do particularly for the health benefits that both (mother and infant) would receive. However, during the process of breastfeeding, many of the participants described experiencing breastfeeding issues and difficulties (i.e. life issues (dea th), issues with pumping, milk supply, latching, and medical issues (preterm birth, thrush)) where they mentioned feeling discouraged and there was a trend where many of them then gave up. The most common challenges described by the participants were: maki ng sure baby was getting enough, latching, pumping, milk supply, ID 010 . It was noted that issues wit h the infant not latching on, mothers feeling like they were not pumping enough, and reduction in milk supply was a reoccurring theme. 1. Success with meeting breastfeeding goals 1. Dual benefit for infant and mother : Participants described the decision to exclusively breastfeed their infant was primarily from receiving education and being knowledgeable about the general benefits for both, mother and infant from breastfeeding. ID 014 ID 004 ID 008 151 ID 011 Most mothers mentioned receiving this information from t heir health care providers , which included but was not limited to: Doctor, Midwife, and WIC. 2. Fluctuating Confidence: Mothers who lacked confidence at the beginning , showed an increase/boost in their the infant was wellness checks and she hitting a ll her milestones; could see she was thriving, making the right the beginning, but I am very confident now. Once became familiar with it, it just became regul ID 008 3. Perseverance, Commitment, and Self - determination Participants also showed perseverance with breastfeeding despite having issues with could only fill the ID 003 It was noted how one participant experienced difficulty while breastfeeding but did not want to give her child formula. - ID 016 Self - determination was noted as a re - occurring theme which helped many mothers to with their own positive affirmations. One participant ID 2042 . 152 Com mitment to try and breastfeed. Participants mentioned being able to breastfeed and continue breastfeeding despite any issues/ concerns. Many mothers described making the commitment to breastfeed and at least trying to successfully. One participant shared, that I went through with it. That was a success. Like I said I was overthinking with it, but at the ID 014 . Another participant shared, I want to keep it at ID 004 . 4. Ability to Pump Having the ability to pump was mentioned several times as a n important component of being able to meet breastfeeding goals . Quotes noted below include: I want to do. Not ID 002 ID 013 emptied a bottle, I re - ID 010 5. Encouragement, help and support from identified social support networks Outside encouragement/support from support networks was another sub - theme identified for meeting breastfeeding goals. One participant shared, ID 007 . Mothers felt they received enough support and utilized their support system s within the community. People/ organizations most of ten mentioned were WIC, Mama Bear, social worker, and online resources (mother blogs). Mothers sought out help and asked questions from people and even to other mothers who also breastfed. Most of the help described that women received 153 was emotional (encou ragement and positive affirmations/reinforcement) and informational support (tips, advice). Mothers mentioned how they utilized and appreciated community programs that provided items for their infant. However, very few participants mentioned receiving hand s - breastfeeding counselor, f rom the WIC office, she invited me and gave me a breast pump. She invited me to her breastfeeding classes. She said she will come and show me how to help me wi - ID 016. Many of the participants were very positive and happy to receiv e some level of support, which enhanced their breastfeeding experience. 2. Suggestions for expanding breastfeeding support and promotion among African American women 1. Tangible , Immediate , and Proactive help: Study Participants mentioned knowing the importance of breastfeeding but wanted more , but like it needs to - ID 016 Study participants shared the need for African American women to receive help from d to be able to readily communicate to someone who c ould assist with breastfeeding. Quotes identified include d : home, the people around you are not doing that, then you are not gonna want to do it - 011 154 - 010 not every - 016 2. Positive Non - Judgmental Support (Professional and Personal) : In regard to breastfeeding promotion, mothers described the need for other African American mothers to receive positive, non - judgmental support from al l people within their social support ID 009 . The majority of participants felt that support should come from all support networks including family memb ers and extend to health care professionals. One mother ID 005. friends, local mother support groups, churches, doctors, midwife, lactation consultant, and WIC. ID 009 . One mothe r shared her thoughts ID 006 . Many participants mentioned that they were willing to accept encouragement from anyone. It was also noted th at need ed to feel comfortable with expressing challenges authentically. Another mother shared, ID 006 . M others specifically described the need for breastfeeding to be normalized for women in order for them to feel comfortable not only asking questions , but also feeling comfortable with breastfeed ing in public or at work and to 155 receive more encouragement around the practice. Another participant shar - ID 006 . 3. : Regarding education, participants felt it was important to have readily available and accessible assistance f rom health care professionals who focus on breastfeeding , especially when a mother needs breastfeeding help or has questions. It was mentioned that education should be available preconception, prenatally, and post - partum. Mothers described the need to have education centered on what to do regarding breastfeeding and healthy eating, Many mothers expressed being unsure of why they had a fluctuation or decrease in their milk supply. Along those same lines . which could play a role in the issue with milk supply , was the use of the pump. Mothers often described, in the same thought, having issues with their milk supply and using a pump. T hey did not connect the two t ogether as one possibly influencing the other. There was a and pumping more often. There was a lack of education and knowledge about how to use the pump pump and pump so that ID 2042. 4. Self - motivation, Will Power, and Perseverance : Study Participants suggested maintaining their own self - c onfidence and not being easily discouraged or persuaded. Many mothers described the need for African American women to without thinking about it I feel like they will quit - ID 002 . Another participant shared similar 156 - ID 014 . 6 . 5 Discussion There are complex factors either reinforcing or discouraging breastfeeding initiation and maintenance among African American women that contribute to the disparities as compared to other racial groups. Revisiting the Social Ecological framework, there are factors within each level contributing to these apparent disparities. At the individual level, study participants continuously shared awareness of ho w important it was for them to breastfeed. Many of the mothers were knowledgeable about the general benefits of breastfeeding and they wanted both mother and infant to receive those benefits. Many participants shared different resources that informed them of the benefits of breastfeeding. The resources and community organizations such as WIC, lactation consultants, physicians, and breastfeeding groups were consistent as documented in the literature. 18 Many consistently shared how self - motivation, self - determination and commitment influenced as well as pushed them to continue breastfeeding. This has ofte n been noted and described in other studies as being a common denominator for successful breastfeeding experiences. 81 However, in our study many participants consistently also shared multiple challenges that frequently compromised breastfeeding. These challenges are consistent with the barriers that have been reported as impeding African American women from breastfeeding. 18 With many of the study participants experiencing these challenges i.e. (making su re baby was getting enough, latching, pumping, adequacy of milk supply, breastfeeding around people (family and in public), and working) although wanting to continue breastfeeding, many shared feelings of discouragement and a fluctuation of their confiden ce and whe n these challenges were not resolved , they gave up. 157 In the midst of those challenges, many mothers reported receiving words of encouragement and support from their interpersonal networks and described how important it was for their breastfeeding su ccess journey. While mothers felt that it was important to have social support from all people identified within their social network, they emphasized the importance and the strong role fathers play ed in facilitating breastfeeding , consistently document ed within the literature. 18 98 160 Th is form of emotional support from social support networks has often been described as essential to breastfeeding practice and success. 98 c are providers, more accessible help from health c are providers who focus on breastfeeding to help them through challenges experienced. This finding is in line with previous research suggesting that offering complementary support in addit ion to standard care can aid in continued breastfeeding. 161 Many mothers mentioned receiving pamphlets/ education or calling and receiving assistance over the phone. While this mode of support and help may aid with breastfeeding, it may not be ideal for all, especially those who are most susceptible to breastfeeding cessation. It is important to note that breastfeeding is being discussed at some level; how ever . it is important to also recognize that breastfeeding education should be tailored to the individual and breastfeeding . but there should be more education regarding breastfeeding in addition to just the benefits. Overall, study participants were very knowledgeable and had been provided with some form of education around the benefits of breastfeeding. However, it was very apparent that a ssociated with breastfeeding. While women do not know how to handle all issues when it comes to breastfeeding , they should 158 : 1) know who to contact, 2) feel comfortable with contacting, and 3) have confidence in knowing that someone is readily available. T he American College of Obstetricians and Gynecologist (ACOG), recommended that obstetric care providers should develop and maintain skills for supporting normal breastfeeding physiology and management of common complications of lactation. 162 In addition, breastfeeding is a conversation that should b e continually discussed throughout the ante natal and post - partum period. This is also reflected in the recommendations from ACOG optimizing support for breastfeeding. 162 It is important to continue the discussion and education that is needed around breastfeeding so African American women can begin t o conceptualize how it can fit into their daily lives. It was apparent from some participants, specifically those who did not have any breastfeeding goals , that there was no discussion with them or in - depth thoughts or plans on how to incorporate breastfee ding into their daily lives. Traditionally, exclusive pumping was perceived as reserved for babies in the neonatal intensive care unit or for those medically unable to suckle directly at the breast. 163 With the widespread availabi lity of portable, personal , electric breast pumps because of the mandate in the Affordable Care Act, the practice of pumping frequently or exclusively has grown, increasing unseen and potential consequences associated with pumping. An unexpected theme that arose around making the breastfeeding experience more successful , was the need for education specifically regarding the use of the pump and milk supply. Many study participants mentioned either fluctuation s or decrease s in their milk supply, but they were also unsure why this occurred. It is well known that the volume of milk ( ) can vary between feedings and between breast. 164 However, it was noted that participants often li 159 More specifically, while mother s constantly mentioned pumping and pumping more often, there was a lack of knowledge about how to correctly use the pump and when/how often the pump should be used. The study participants were aware of various places to contact for breastfeeding support a t the community level. Participants mentioned contacting and utilizing these support systems when they needed help. The majority of the study participants mentioned utilizing WIC as a place of support and breastfeeding help. However, it was noted that the help and assistance provided from WIC varied based on location. None of the help, described by participants, was deemed consistent. In addition, to participants mentioning WIC as a place for receiving breastfeeding support and receiving help regarding technical breastfeeding questions; from further probing, it was noted that suggestions were provided over the phone, but there was no . Breastfeeding promotion was federally mandated for all WIC programs in 1 991. 165 Since implementation, WIC has been known for providing breastfeeding support to mothers. Our finding s show that through their efforts , they are motivating and supporting women who want to breastfeed, but it becomes important to make sure that the form of support provided (tangible, emotional, or informational) is consistent across all locations. Based on these findings, it may be important to consider assessing breastfeeding support provided across WIC locations within the state or region and determine/set clear standard guidelines for in conjunction with measurable meaningful outcomes, especially for groups such as AA women who are not meeting guidelines compared to other groups . 160 In order to assist with breastfeeding challenges, it becomes important to consider the being offered and determine what is really helpful in increasing breastfeeding outcomes. From the systematic review of support for breastfeeding mothers by Sikorski and colleagues, it was determined that different modes and timing of the support makes a s ignificant difference in breastfeeding outcomes. 161 Specifically, meeting face to - face showed a significant benefit in continued breastfeeding compared to advice received over the phone. Our study participants reinforced that this mode of support/ assistance was preferred, and might hence be more effective among AA. It was recognized from participant interviews that support should be multi - faceted. When pa rticipants were asked about breastfeeding support, they mentioned/ discussed the importance of having support in all aspects of their lives ( e.g. caring for children, transportation, money for groceries, time to rest). This was identified as being most i mportant to the study participants in order to focus on breastfeeding. Many of the participants expressed the willingness to receive support and/or encouragement from anyone, but it was especially important for them to receive this support from those ident ified within their social support network s . Receiving support from those identified within their social support networks (significant other, parents, sibling(s), friends, work) is consistently identified within the literature as being a factor to assist in enhancing breastfeeding, which seemed especially relevant for these AA women. 98 161 Study participants were also asked to share their thoughts, opinions and suggestions on what they felt African American women would need in order to facilitate breastfeeding and increase breastfee ding outcomes. Utilizing the social ecological model in this context, encouraging women to recognize and utilize their own will power and having self - determination w ere identified (inner sphere) as factors that enhanced the likelihood of breastfeed ing . Fac tors 161 identified within the middle and outer spheres included Tangible, Immediate, and Proactive help, Positive Non - All participants discussed the need to h ave genuine support with those identified in their social support networks at the interpersonal level extending to the macro levels which included the community, environment and multiple organizations. Noticing how African American mothers responses were centered on encompassing multiple levels of the social ecological model, was consistent with the study by Tiedje et al, 2002 , where authors suggested that breastfeeding promotion should occur at several levels reflective of the socio - ecological perspective to help ensure breastfeeding can continue beyond initiation. 44 At the intrapersonal level, the theme focused on willpower and self - determination as two important factors needed for reinforcing breastfeeding . Participants described it as factors at each level may play a role in influencing the intrapersonal level. While an African American woman may have the will power and determination to breastfeed, her external community and environment may impede breastfeeding, which is noted by many s tudies within the literature. 40 129 Study participants consistently mentioned that receiving encouragement to breastfeed was important and it should be from everyone in their environment. It was also recognized that the support provided needed to be c entered on normalizing breastfeeding. Participants mentioned the need for support, from these facets, to be genuine and positive to help make them feel comfortable not only with the act of breastfeeding, but also with asking questions and participating in open discussions centered around breastfeeding. At the community, environmental and organizational levels participants consistently mentioned that assistance in the form of education, resources, or actual technical help should be readily accessible and ava ilable. Research confirms these 162 suggestions, that mothers need continuity in care using a multi systems approach to address social and personal barriers to breastfeeding. 129 80 6 . 6 Conclusion There are many studies that expound perceived barriers and facilitators to breastfeeding with African American women. The perceived barriers and facilitators identified wi thin our study support the literature. To our knowledge, no other studies have included African American women shar ing their ideas by providing suggestions for breastfeeding promotion in an effort to overcom e barriers to successfully reach breastfeeding go als. Strengths of our study include using a sample with a diverse socioeconomic background (i.e. income and occupation) to provide a range of perspectives. Along with the socioeconomic diversity, there was additional diversity in the parity status of the p articipants interviewed. Only four of the women interviewed were first time mothers, meaning that these experiences are not exclusive to one particular group over another. Another strength of this study included interviewing key informants as they were abl e to provide us with suggestions and guidance for the questionnaire given their familiarity with and experiencing working in the community. It is however important to note there are several study limitations. These results are not generalizable to all African American women as the sample of participants had some interest in and initially planned on breastfeeding. Furthermore, some of the parti cipant interviews occurred over the phone. While this provided flexibility for the participants, there may have been other distractions inhibiting them from fully participating and/or being explicit in their answers if there were others around. Another lim itation was the limited target group since we only interviewed participants who completed the Mama Bear study , which might have also increased the likelihood of biased responses. 163 We believe the social ecological framework provides a useful context for dis cussions around factors enabling or inhibiting breastfeeding within the African American community at multiple levels. African American women from all socioeconomic backgrounds can successfully breastfeed if provided an equitable amount of breastfeeding su pport from all facets to help increase and maintain their breastfeeding self - efficacy. However, participants within this study , aligned with the literature, experienced gaps in overall breastfeeding support that inhibited them from continuing to breastfeed . Our participants offered suggestions on how to improve breastfeeding support and promotion for African American women. Breastfeeding rates within the African American community will unquestionably not increase based on one simple solution. Multiple conve rsations are needed where African American women are in the forefront and involved in those conversations, as they provide knowledge and personal experiences for the needed solutions, in order to shift cultural norms and structures at all levels. Findings from this qualitative research may suggest the opportunity for those who work at different levels to identify opportunities for further discussion, intervention, or advocacy around increasing equity for breastfeeding among African American women. 164 Chapter 7 - Summary and Conclusions B reastfeeding has been established as an effective way to provide infant nutrition, with numerous health benefits for both mother and baby. While there has been an upward trend in BF rates, AA women are still initiating and continuing BF at much lower rates compared to their counterparts (Caucasian, Asian and Hispanic). There is extensive evidence sug gesting that there are several factors which contribute to the choice of infant feeding, and that also account for the observed racial/ethnic differences in BF behavior. With the apparent BF disparity and incomplete understanding of how to close the BF gap , there remains a strong need for determining how certain modifiable factors such as, self - efficacy and social support may help shape BF behaviors. It is also important to have AA women provide insight on their experience s and suggestions on how these factors influence BF goals , especially since there is a lack of studies that investigate the perspectives of AA women and BF. The overall study, Mama Bear, was a randomized control led longitudinal pilot study that tested a du al intervention which incorporated a postpartum weight management component into an effective BF peer - support program. The first study, Chapter 4, utilized the Mama Bear population to determine how BF self - efficacy impacts BF outcomes (initiation, duration , and intensity). BF self - efficacy measured at late pregnancy was positively associated with BF duration and intensity at 6 weeks postpartum. However, these results were no longer present when measured at 20 weeks postpartum. BF self - efficacy measured at 6 weeks postpartum positively influenced BF duration and BF intensity measured at 20 weeks postpartum in this sample of AA women. The results, from Chapter 4, support the existing literature where maternal confidence is a key predictor of BF duration. BF self - efficacy measured during the antenatal and postnatal 165 period were positively associated with breastfeeding duration and intensity . T his suggests that continual assessment of this modifiable variable could increase BF duration and exclusivity . A ssessmen t of BF self - efficacy at multiple timepoints, specifically late pregnancy and re - assessment again at 6 weeks postpartum, allows time to compare current responses to the last assessment. This could then allow time to mov e beyond providing general BF recomme ndations and offering more tailored individualized education, counseling, and/ or interventions that are The second study, Chapter 5, utilized the Mama Bear study population to determine how social support impacts BF outcom es (duration and intensity). BF initiation was not assessed since 75% of the participants began BF. There were no associations with perceived social support when measured in late pregnancy and BF duration. In addition, we found that there were no associat ions with perceived social support when measured at 6 or 20 weeks postpartum and BF duration. However, when assessing the usual care and intervention groups separately , we found that perceived social support measured at six weeks post - partum positively influenced BF duration in the usual care group. P erceived social support measured at late pregnancy was positively associated with BF intensity at six weeks post - partum i n all participants. There were no associations with perceived social support measured at late pregnancy and BF intensity at 20 weeks postpartum. When perceived social support was measured at 6 weeks postpartum, we noticed that it was positively associated with BF intensity at 6 weeks postpartum in the usual care group, but los t significance when the usual care and intervention group s were combined in the model. There were no associations observed with perceived social support measured at 6 weeks postpartum and BF intensity at 20 weeks postpartum. The small associations observed within this study support the evidence of social support 166 perceived social support system during two critical periods (late pregnancy and six weeks post - partum) for continuing BF. Using a social support survey as an assessment tool during these two critical periods could help healthcare providers and other health professionals initiate the conv ersation for mothers to begin thinking about their social support networks, and who they goals . The results of these two pilot studies indicate the importance of assessing both B F self - efficacy and social support among AA mothers specifically as instruments to increase BF duration and BF intensity as well as exclusivity. Studies have often listed in their future implications the need to have more individualized education and couns eling, but there has been a lack of guidance on what factors would be most appropriate to assess in order to implement these suggestions. The novelty of these two quantitative studies is suggestions on how these two variables could be used at very critical timepoints (late pregnancy and 6 weeks postpartum) within interventions or clinical settings to help influence BF outcomes. The third study builds upon the first two studies, with in - person and virtual semi - structured individual interviews with Mama Bear participants . F actors identified involved either encouraging or discouraging the attainment BF goals. In addition, it was important to have the voices and personal experiences of AA women provide suggestions of how AA women who want to breastfeed can be s upported. The primary challenges to meeting BF goals included: participants expressing a fluctuation in confidence and discouragement from issues associated with BF. The most common issues with BF described by participants included: making sure baby was getting enough milk , latching, pumping, milk supply, BF around people (family and in public), and working. Perceived facilitators for meeting BF goals included: 1) d ual benefit for 167 mother and baby, 2) f luctuating confidence, 3) p erseverance, commitment, an d self - determination, 4) a bility to pump, and 5) e ncouragement, help and support from identified social support networks. Participants provided suggestions for expanding BF support and promotion among AA women which included: 1) t angible, i mmediate, and p roactive help; 2) p ositive n on - - motivation, will power, and perseverance . The perceived challenges and successes identified in the third study supported evidence currently in the literature . The factors reinforcing or discouraging BF initiation and continuation among AA women in this study were similar to those reported previously by others . The novelty of this study included recommendations from AA women on how they could be supported in meeting BF goals. Findings from this qualitative research also supported the current evidence that all levels of the SEM are involved and interconnected in BF and in order to address BF promotion among AA women, all of the factors identified can be a useful framework for stakeholders to continue these conversations. Strengths of the quantitative components of the studies include the assessment of BF self - efficacy and perceived social support of AA mothers in a lo ngitudinal randomized control led dual intervention trial . These two predictors have frequently been identified in the literature as predictors associated with BF behavior. The quantitative studies provided a better understanding of how breastfeeding self - e fficacy and perceived social support could be used to potentially influence BF behavior. Another strength was the exploratory nature of the studies and creative utiliz ation of tools already available. Even though small modifications were made for use betwe en pre - and post - partum, we were able to still identify associations with BF self - efficacy and perceived social support between late pregnancy and post - partum. While much of the 168 literature regarding AA women and BF focuses on mothers identified as being lo w income, our three studies used a sample with various socioeconomic backgrounds (i.e. income and occupation). In addition, the parity status of the participants interviewed varied. This provided a diverse range of BF perspectives and experiences. It is i mportant to note that there are several limitations to our studies. These results are not generalizable to all AA women since we had a small sample of participants , and they had some interest and initially planned on BF. While breastfeeding exclusivity was folded into breastfeeding intensity, our study did not specifically assess this outcome (breastfeeding exclusivity), because of our small sample size. While we cr eatively used the survey tools available, the BF self - efficacy tool was originally created to be assessed in post - partum populations. In addition, the perceived social support survey was created in 1984, and some of these questions may have been outdated. Although, we were still able to detect some associations because of the exploratory nature of the study, it may be valuable to assess whether these tools should be updated and perhaps culturally adapted for future use. In addition, it may be worthwhile to assess whether our findings are specific to only AA women. The challenges and successes AA women reported experiencing while trying to meet BF goals , which we identified within the qualitative component , continues to support the evidence in the literature . I t would be useful to compare these findings to women from other ethnicities, however the overall reason and key purpose of focusing on AA women is because they continue to have poor health outcomes post - partum and have disproportionally low BF rates. Some recommendations for future research studies include: 1) Determining whether the perceived social support tool is still appropriate in this era or if it needs to be updated and culturally adapted f or future use, 2) conducting an intervention study wher e the survey tools are 169 administered and counseling/education is tailored based on mothers assessment, and 3) conduct a similar study /studies to determine if these findings are specific to only this sample of women or a broad range of women. Our findings highlighted that AA women from all socioeconomic backgrounds in the Detroit, Michigan area who reported higher levels of social support and self - efficacy are able to breastfeed longer, more intensely and exclusively. However, participants within this study , as similarly found within the literature, experienced gaps in BF support and a fluctuation of maternal confidence at various times (prenatally and postpartum), which inhibited many of them from continuing to breastfeed. The knowledge gained from these th ree studies provide s a better understanding of important considerations for professionals (working at various levels) where they can create space for further discussion, intervention, or advocacy to not only enhance the BF experience , but also the goal of improving BF outcomes among AA mothers. 170 APPENDICE S 171 Appendix A: Clinic Approval 172 Appendix B: Participant Consent Gwen Alexander, Ph.D. Assistant Research Scientist Henry Ford Health System Public Health Sciences 1 Ford Place, 5C Detroit MI 48202 1. WHY IS TH I S RES E A R CH BE I NG DO N E? The purposes of this s t udy are to 1) learn what kinds of su p port will help women m e et their goa l s for breastfeedi n g and weight managem e nt after their baby is born and 2) to t est urine and bl o od for nutrition fac t ors that may help us un d erstand how breastfeeding and weight loss after a baby is born affects the health of mothers and their baby. You are being asked to take part in this research study because you are an African American woman in the later months of your pregnancy and you said you were interested in breastfeedi n g your baby. We will en r oll about 80 women in this research study at Henry Ford H e alth System (H F HS). This s t udy is funded by the Na t ional Ins t itu t es of Health (NIH R21) and is a gr o up project between HFHS and Mi c higan State University (MSU). 2. WHAT WI L L HAP P EN IF I TAKE P ART IN TH I S RES E A R CH ST U D Y ? There will b e two groups in the stu d y. The group you are assigned to wi l l be chosen by chance ( l ike flipping a co i n). Both groups will get educational information about how to get a heal t hy start for both mom and baby. One group will get writ t en information, and the other group will get info rm ation from an online program and will have in - person visits from a c o mmunity h e alth worker. Women in both groups will be ask e d to: 1) Answer survey questions four (4) times during the 6 - month study. CONSENT TO PARTIC I PATE IN A RES E AR C H STU D Y (HFH I RB form r ev: 02/2012) D A T E : M R N : NAME: APPROVAL PER I OD Feb 03, 20 1 7 Mar 12, 2017 INSTI T U T IONAL REVIEW BOARD PRO J ECT TITLE: Mothers Allied w i th Mothers Around Breastfeeding Encouragement And Ren e w e d H e alth (Ma m a Bea r ) 173 2) Have your height, weig h t and body composition measured four (4) times during the 6 - month study. Body comp os ition will be measured with a bioel e c trical impedance analysis (BIA) scale that se n ds a small curre n t through your body that you feel. This i n fo rm ation will be collec t ed f rom you at your home. 3) Have some urine (from the sample you give your doctor duri n g your visit) collected two (2) times during the 6 - month study (today and at your routine doctor visit that will b e about 6 weeks after your baby is born). Your urine will be test e d for iod i ne and some environment a l chemicals t hat interfere with iodine u ptake into t h e thyroid gland (includi n g perchlora t e which is f o und in many products a n d thioyanate which is fo u nd in cigare t te smoke and some foods). 4) Have some extra blood (2 - 4 teaspo o ns or 10 - 20 ml) collect e d and stored two (2) times during the 6 - month study (today and at your routine doctor visit that will b e about 6 weeks after your baby is born). Your blood will be tested for n utrition fact o r s (inclu d ing ferritin which measures your iron status and carotenoid c o ncentrations which estimates your fruit and vegetable intake), and other factors (i n c luding thyroid hormone). 5) Allow your medical records from pregnancy, l abor and delivery and y o ur medical reco r ds to be reviewed by the research team. 6) Researchers will c o llect health and o ther information concer n ing the birth of your child as reported to the Michigan Department of He a lth and Human S e rvices (MDHHS) at the time of your In or d er to locate your birth certif ic ate, research staff will r e quest your social s e curi t y number. All of the p r ocedures at these data collection vis i ts are extra these are t hings that you will not d o if you choose not t o participate in the study. None of the s e data collection pr oc edures are experiment a l. If your identifiable specimens or health info rm ation are selected for u s e in addi t i o nal researc h , you will be contacted f o r permissio n . All women: At visit 1, which will be t oday, if you choose to p articipate, y ou will have the following CONSENT TO PARTIC I PATE IN A RES E AR C H STU D Y (HFH I RB form r ev: 02/2012) D A T E : M R N : NAME: APPROVAL PER I OD Feb 03, 20 1 7 Mar 12, 2017 INSTI T U T IONAL REVIEW BOARD PRO J ECT TITLE: Mothers Allied w i th Mothers Around Breastfeeding Encouragement And Ren e w e d H e alth (Ma m a Bea r ) 174 procedures: we will measure your height, weight and body comp os ition, and collect some urine and b l ood. At visit 2, which will be by telephone, a researcher will ca l l you within the next w e ek or two and ask you some survey questions about your social si t uation, your diet and exercise habi t s (about 25 minutes to complete). At visits 3, 4 and 5, we will meet y o u at your office visit or at your home to measure your weight and ask you the same survey q u estions and a few more about what you are feeding your baby (about 25 minutes to complete). Visit 3 will b e when your baby is about 2 weeks old. Visit 4 will b e when your baby is about 6 weeks old. Visit 5 will b e when your baby is about 20 weeks (5 months old). We will try to schedule t hese study visits at the same time you come to see your doctor or your but if that work, we will ask your p ermission to schedule a visit at your home. For women i n the group that gets in f ormation from t he online program: You will rece iv e informat i on and be encouraged to use an o n line program to record activities relat e d to returni n g to your pre - baby wei g ht. You will also ha v e one - on - one visits, at your home and by t e lephone, with a community health worker to talk about infant feeding and your weight before and after the b a by is born. These visits are extra beyond your regular prenatal care at HFHS and a fter your baby is born. The visits b y the community health worker and the use of the online program are new ways to offer sup p ort to pregnant women and new m o ms. We are testing t hese ways of communicating with women to understand how best to su p port women like you after a new baby is born. 3. WHAT ARE THE R I SKS OF TH E STUDY? CONSENT TO PARTIC I PATE IN A RES E AR C H STU D Y (HFH I RB form r ev: 02/2012) D A T E : M R N : NAME: APPROVAL PER I OD Feb 03, 20 1 7 Mar 12, 2017 INSTI T U T IONAL REVIEW BOARD PRO J ECT TITLE: Mothers Allied w i th Mothers Around Breastfeeding Encouragement And Ren e w e d H e alth (Ma m a Bea r ) 175 We do not expect any increase in p h ysical, psychological, s o cial, leg a l n o r economic risks to you or your baby from b eing in this study. You should tell t h e person ob t aining your consent ab o ut any other medical research st u dies you are involved in right now. It is not expected that you will have any complications or discomforts from being in this study. There may be risks or discomforts that are not known at this time. 4. WHAT A RE THE B E N E FITS TO T A KING PART IN THE S T UDY? The potenti a l benefits to you for taking part in this s t udy inclu d e all of the many health and emotional benefits associated with either breas t feeding or weight management or both. You may also benefit from the feeling o f being involved in an important stu d y that m a y h elp improve two import a nt health ac t ivities among African American women breastfeedi n g and post p artum weig h t managem e nt. 5. WHAT O TH ER OPT I O N S ARE T H E RE? Your participation in this study is c o mpletely vo l untary. You have the right to refuse to be in the study or to stop at a n y time with o ut affecting your present or future medical car e . At this time, there is no known other study like this one. If you wish, you m a y talk to your doctor about your other c h oices before you decide if y o u will take p art in this s t udy. 6. WHAT A BOUT CONFI D ENTIALIT Y ? By signing this consent form, you a g ree that we may collect, use and rel e ase your personal and h ealth information for the purpose of this r e search stu d y. We on our study team may collect a n d use: Your existing medical records. New health information about you and your baby created during this stu d y. Health insu r ance and o t her billing i n formation. CONSENT TO PARTIC I PAT E IN A RES E AR C H STU D Y (HFH I RB form r ev: 02/2012) D A TE : M R N: 176 We (HFHS) may release this inform a tion, using y our study identification without names attached, to the following p e ople: The Princi p al Investiga t or and his/ h er associ a t e s who work on, or oversee the rese a rch activities. Gover n ment officials who oversee research. The research sponsor N ational Ins t i t utes of Health. Your insur a nce company or others responsible f or paying y o ur medical bills. Other researchers at ot h er institutio n s particip a ting in the research. The Michigan State University Hum a n Research Protection Program. Once your d e - identified information (no name attached) has been released accordi n g to this co n s ent form, it could be r eleased ag a in and may no longer be protected b y federal privacy regulations. This conse n t form, test results, medical reports a nd other inf o rmation about you from t his study may be placed into your medic a l record. Generally, you are allowed to look at your medical record. During the research st u dy, you will not be allo w ed to look at your research study in f ormation that is not in y o ur medical record. HFHS or ot h ers in this r e search gro u p may publish the resul t s of this st u dy. No names, identifyi n g pictures or other dir e ct identi f iers will be used in any publ i c presentat i on or publication about t his study un l ess you sign a sepa r ate consent allowing that use. This conse n t to use and release your personal a n d health inf o rmation will not expire at the end of t his research st u dy. The research records will be maintained for a minim u m of 3 years after the end of t he study. You do not have to sign this co n s e n t to release your medic a l information and may c a ncel it at a n y time. If you decide not to sign t his consent or cancel y o ur c onsent, you cannot participate in this study. If you notify us that you wish to stop participating in this study, we may c o ntinue to use and release the information that has already been coll e c t ed. If you would like to t a lk a bout this st u dy or wi t hdraw for any reason, you may contact the princip a l investigator, Gwen Alexander, Ph.D., at (313) 8 74 - 6737 or the HFHS IRB Coordinator (313) 916 - 2024 to cancel your consent and we will se n d you a form to sign or you can send a written and dated not i c e to the principal APPROVAL PER I OD Feb 03, 20 1 7 Mar 12, 2017 INSTI T U T IONAL REVIEW BOARD PRO J ECT TITLE: Mothers Allied w i th Mothers Around Breastfeeding Encouragement And Ren e w e d H e alth (Ma m a Bea r ) 177 investigator at t he address listed on the first page of this form. CONSENT TO PARTIC I PATE IN A RES E AR C H STU D Y (HFH I RB form r ev: 02/2012) D A T E : M R N : NAME: APPROVAL PER I OD Feb 03, 20 1 7 Mar 12, 2017 INSTI T U T IONAL REVIEW BOARD PRO J ECT TITLE: Mothers Allied w i th Mothers Around Breastfeeding Encouragement And Ren e w e d H e alth (Ma m a Bea r ) 7. WHAT IF I AM IN J UR E D? There is no f ederal, stat e , or other program that will compensate you or pay for your medical ca r e if you are injured as a result of par t icipating in t his study. You and/or your medical insurance may have to p ay for your medic a l care if you are injured a s a result of participating in this study. You are not giving up any of your legal rights by signing this co ns ent form. 8. WHO DO I CALL WITH QUEST I ONS ABOUT T HE STUDY OR T O REPORT AN INJURY? Gwen Alex a nder, Ph.D., or her staff member has explained this research study and has offered to answer any questions. If you have questions about the study procedures, or to report an injury you may contact Gwen Alex a nder at (313) 874 - 6737. If you have questions a bout your rights as a research subj e c t, you may contact the Henry Ford Health System IRB Coordinator at (313) 916 - 2024. The IRB (In s titutional Review Board) is a group of people at HFHS who review all of the research to protect your rights. 9. DO I HAVE TO PARTICIPATE IN T HIS STUDY? No, your participation in t his research study is voluntary. If y o u decide to participate, you can stop at any time. If this happens, you may be a s ked to retu r n for 178 a visit f or safety reasons. You will get the same medical ca r e from HFHS whether or not you participate in t his study. T here will be no penalties or loss of benefits to which you would otherwise be entitl e d if you choose not to p articipate or if you choose to stop your participation once you have sta r ted. You will be told ab o ut any significant information that is discovered t hat could r e asonably affect your will i ngness to continue bei n g in the stu d y. 10. WHO EL S E CAN S T OP MY P ART I CIPAT I ON? While this is not anticip a ted due to t h e nature of t his study, t h e Principal I nvestigator can end your participation in the rese a rch study at any time. T he Principal Investigator, sponsor or your doctor can end your participation in the r esearch st u dy at any t i me. If this h appens, you may be asked to return for a visit for safety reasons. 11. WILL IT COST ANY T H I NG T O PARTICIPATE? CONSENT TO PARTIC I PATE IN A RES E AR C H STU D Y (HFH I RB form r ev: 02/2012) D A T E : M R N : NAME: APPROVAL PER I OD Feb 03, 20 1 7 Mar 12, 2017 INSTI T U T IONAL REVIEW BOARD PRO J ECT TITLE: Mothers Allied w i th Mothers Around Breastfeeding Encouragement And Ren e w e d H e alth (Ma m a Bea r ) We do not expect there to be any additional c o s t s to you if you participa t e in this st u dy. Items related to the routine medical care that you would receive even if you did not participa t e in this st u dy will be bi l led to you or your insurance company. You have the right to ask w hat it will cost you to take part in this study. 12. WILL I BE PAID T O PARTICIPA T E? If you consent to partici p ate and data collec t ion b egins, you will be given g ift cards at fi ve different times during the s t udy to thank you for participating. Y o u will recei v e: 179 $30 gift card today $20 gift card after your phone interview that will be within a f ew weeks after visit $20 gift card after the st u dy visit wh e n your baby is about two weeks old $30 gift card after the st u dy visit wh e n your baby is about six weeks old a nd $50 gift card after completion of the f inal survey when your b aby is about five months old. We will also provide you with some non - monetary gifts (like a teddy bear). If you complete the st u dy, you will be rece iv e a total of $150 to tha n k you. If you do not finish the study, you will be given the gift cards for the parts of t he study that you did complete. 13. CONSENT You have r e ad this co ns ent form or it has been r ead and explained to you. You understand what you are being asked to do. Your questions h ave been answered. Any technical t erms you d i d not understand have been explained to you. You agree to be in this study. You will be given a copy of this cons e nt f orm. Signature of Subject D ate Time Print Name of Subject Witness to Signature Date Time Print Name of Person Obtaining C o nsent CONSENT TO PARTIC I PATE IN A RES E AR C H STU D Y (HFH I RB form r ev: 02/2012) D A T E : M R N 180 APPROVAL PER I OD Feb 03, 20 1 7 Mar 12, 2017 INSTI T U T IONAL REVIEW BOARD PRO J ECT TITLE: Mothers Allied w i th Mothers Around Breastfeeding Encouragement And Ren e w e d H e alth (Ma m a Bea r ) Signature of Person Obtaining Consent 181 Appendix C: Data Collector Telephone Script ) A. Procedures: Purpose of this document This document should be used when contacting participants for the study: Mama Bear . Direct phone contact with participants should be initiated as soon as possible after consent is obtained in the clinic during pregnancy around 28 weeks gestation. Steps in phone call contact Initiate call. Let the phone ring 7 times If answering machine appears, leave message explaining who you are and when you will con tact them again. When call is complete complete REDCap Checklist If the participant calls and reaches you, start from step 3 of the script B. Telephone script STEP 1 | INITIAL CONTACT Hello. I am first name from HFHS he Mama Bear Study Team. May I please speak to . When desired person is on the phone Hello [Ms/Mrs]_________ My name is First Name and I am a data collector calling on behalf of Henry Ford Health System with the Mama Bear Study. On "Time when consent was obtained" you agreed to participate in the Mama Bear study and the nurse told you that you would receive a call to collect some information from you. Is now a good time to talk more about the Mama Bear study and collect some information from you? IF NO, GO TO STE P 2 IF YES, GO TO STEP 3 If desired person is not available Is there a better day and time to reach [Ms]_________ ? Note days and times and enter into REDCap Thank you for your assistance. I will try to call back then. 182 If desired person is in distress or does not have capacity I am sorry to hear cause of distress and I hope this call has not caused you further distress. If the person becomes distressed, do not continue with question below. Do you feel comfortable continuing this call? IF YES, GO TO STEP 3 IF NO. Okay. Is there a better day and time I can call you back? I need to collect some information what is the best time and day to call you back in the next few days? Thank you. Make note in REDCap STEP 2 | INITAIL CONTACT CONTINUED If interrupted or strong immediate refusal Is there a better day and time I can call you back? IF YES, Thank you. Myself or include names of other data collectors of the team will try to call back then. Note days and times and enter into REDCap IF NO/NOT INTERESTED. Okay. Thank you for your time. End Call. Note and enter into REDCap Version Number 1 Date MAMA Bear Research Program Page 182 of 248 Telephone Script Draft 8 March 2017 ) STEP 3 | PARTICIPATION DETAILS STEP 3 The Mama Bear study would like your help to find the best ways to help moms and babies stay healthy. 183 Participation in this study is completely voluntary and will involve contact at multiple points over the course of 6 months. Initially, in the clinic, yo u signed a consent form and received a copy of it to take home. Now, we want to ask you some qustions during this phone call. Do you mind if I proceed? IF NO/NOT INTERESTED. Okay. May I ask why you are not interested anymore? Thank you for your t ime. End Call. Note and enter into REDCap IF YES, GO TO STEP 4 STEP 4 | USE OF INFORMATION STEP 4 Your answers will be kept completely confidential. We will make sure that the information we collect from you is kept private and used only for the research study we are discussing. If you do not agree to continue the phone call, or if you do not want to p articipate in the research project, it will not affect your care or any future care you may have at Henry Ford Health System. STEP 5 | QUESTIONNAIRE STEP 5 As a participant in the Mama Bear Study , you will be asked to complete questionnaires about feeding your baby and also about other topics related to your health and pregnancy. There are no right or wrong answers. We wan t to hear about your experiences. The questionnaire will take about 25 minutes of your time, depending on how many of the questions apply to your situation. IF YES Continue questionnaire and go to STEP 6 Ensure that you indicate how participants are to answer each question at the beginning of the questions (ie strongly disagree to strongly agree) Ensure you check the participant is happy to continue with the questionnaire at approximately 15 minute intervals. 184 BE PREPARED TO PROBE IF THE RESPONDENT ANSWERS OUTSIDE OF THE CATEGORIES PROVIDED. PROBE USING THE ANSWER CATEGORIES ONLY; DO NOT INTERPR ET FOR THE PARTICIPANT. IF NO Thank them for their time and indicate you plan to keep the information you have collected so far and request them to provide you another time to finish the survey questions. Also, encourage them that upon their completion t hey will receive a mailed incentive ($20 Walmart gift card). Version Number 1 Date MAMA Bear Research Program Page 184 of 248 Telephone Script Draft 8 March 2017 ) STEP 6 | ENDING THE CALL STEP 6 Thank you again for your time and interest in the Mama Bear study. Do you have any questions? As we mentioned when you enrolled, we will be calling you from time to time to collect more information from you and see how you are doing. Either I or my colleague will be calling you again in about [# of weeks] . Let me check again. Is [day/time] still the best time to reach you? We will keep trying, so thanks for connecting with us when we try to reach you. I would like to confirm I have your mailing address details correct. Depending on which group you are in, we will either mail your gift c ard or someone will call you to make arrangements to deliver it to you in - person. We are only able to mail this one time, so please make sure this is an address where you can pick up your mail for at least the next week or so. Again, my name is first name . If you have any questions after this phone call you can contact me on insert contact number based at HFHS . Thank you for your time today. Goodbye. End Call. 185 Appendix D: Randomization Scheme RANDOMIZATION PROCESS/LOGISTICS 1. BRIC Prepare randomization table Packet ID# (sequential)/ group assignment (T or C) Randomization scheme: block randomization with random block size 2. BRIC Email randomization table to Claire. [Claire is unblinded.] 3. Claire Create packets Insert appropriate material (T or C) based on the randomizatio n table. Label with Packet ID# on outside of packet. Deliver packets (in numerical order) to Data Collectors (DCs) 4. DC Complete baseline data collection. Retrieve next numbered packet. Use packets in sequential order. On Checklist Baseline Data Collection Phone Call FORM enter the Packet ID# for each Participant. Record on the paper log sheet for each Packet ready for mailing: PacketID#/Participant Initials/Date packet mailed 5. Claire View the Packet GroupAssnmnt REPORT on a daily basis. Open the Intervention FORM for each Participant listed in the report and enter the Group Assignment (T or C) for the given Packet ID#, by using the randomization table. Note: The report and form are bli nded to all but Claire and Interventionists 6. Interventionist View the Ready for Intervention Scheduling REPORT on a daily basis. Participants listed in the report are ready to be schedule for intervention visits. Note: This report is blinded to all but Cla ire and Interventioni sts. 186 Appendix E: Demographical Questions 187 188 189 190 191 192 193 Appendix F: Breastfeeding Characteristics Questionnaire 194 195 196 197 Appendix G: Breastfeeding Self Efficacy Scale - Short Form Instrument by Cindy Lee Dennis 198 199 Appendix H: Breastfeeding Self Efficacy Scale - Short Form Instrument (present tense - during pregnancy) 200 201 Appendix I: Breastfeeding Self Efficacy Scale - Short Form Instrument (past tense - post - partum after breastfeeding has stopped) 202 203 Appendix J: Breastfeeding Inte nsity nsity 204 Appendix K: Breastfeeding Social Support Questionnaire by Robbie Hughes 205 Appendix L: Modified Breastfeeding Social Support Questionnaire 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 REFERENCES 222 REFERENCES 1. Johnson A, Kirk R, Rosenblum KL, et al. Enhancing breastfeeding rates among African American women: a systematic review of current psychosocial interventions. Breastfeed Med 2015;10(1):45 - 62. doi: 10.1089/bfm.2014.0023 [published Online First: 2014/11/2 6] review. Early Child Development and Care 2010;180(6):703 - 18. 3. Prevention CfDCa. Racial and ethnic differences in breastfeeding initiation and duration, by state - National Immunization Survey, United States, 2004 - 2008. MMWR Morbidity and mortality weekly report 2010;59(11):327. 4. Anatolitou F. Human milk benefits and breastfeeding. Journal of Pediatric and Neonatal Individualized Medicine (JPNIM) 2012;1(1): 11 - 18. 5. Meedya S, Fahy K, Kable A. Factors that positively influence breastfeeding duration to 6 months: a literature review. Women Birth 2010;23(4):135 - 45. doi: 10.1016/j.wombi.2010.02.002 [published Online First: 2010/03/20] 6. (UNICEF) UNCsF. The state of 7. World Health Organization U. Increasing commitment to breastfeeding through funding and improved policies and programmes: Global breastfeeding scorecard 2019, 2019:4. 8. Asiodu IV, Waters CM, Dailey DE, et al. Infant feeding decision - making and the influences of social support persons among first - time African American mot hers. Maternal and Child Health Journal 2016:1 - 10. 9. Jones KM, Power ML, Queenan JT, et al. Racial and ethnic disparities in breastfeeding. Breastfeeding Medicine 2015;10(4):186 - 96. 10. Spencer BS, Grassley JS. African American women and breastfeeding: an integrative literature review. Health care for women international 2013;34(7):607 - 25. doi: 10.1080/07399332.2012.684813 [published Online First: 2013/03/01] 11. State of Michigan Infant Mortality Reduction Plan 2016 - 2019. In: Services MDoHaH, ed. Lansing, Michigan, 2016. 12. Lewallen LP, Street DJ. Initiating and sustaining breastfeeding in African American women. Journal of Obstetric, Gynecologic, & Neonatal Nursing 2010;39(6):667 - 74. 13. Dennis C - L, Faux S. Development and psychometric testing o f the Breastfeeding Self - Efficacy Scale. Research in Nursing and Health 1999;22(5):399 - 409. 223 14. Avery A, Zimmermann K, Underwood PW, et al. Confident commitment is a key factor for sustained breastfeeding. Birth 2009;36(2):141 - 48. 15. Blyth R, Creedy DK, Dennis CL, et al. Effect of maternal confidence on breastfeeding duration: An application of breastfeeding self efficacy theory. Birth 2002;29(4):278 - 84. 16. Committee USB. Healthy people 2020: Breastfeeding objectives. Healthy People 2020 2013 17. OoDPaHP. Maternal, Infant, and Child Health. In Healthy People 2020 2014 [Available from: https://www.healthypeople.gov/2020/topics - o bjectives/topic/maternal - infant - and - child - health/objectives accessed May 20 2016. 18. Services UDoHaH. The Surgeon General's call to action to support breastfeeding. 2011 19. Nichols J, Schutte NS, Brown RF, et al. The impact of a self - efficacy interventi on on short - term breast - feeding outcomes. Health Education & Behavior 2009;36(2):250 - 58. 20. Eidelman AI, Schanler RJ, Johnston M, et al. Breastfeeding and the Use of Human Milk. Pediatrics 2012;129(3):e827 - e41. doi: 10.1542/peds.2011 - 3552 21. Labbok M, Krasovec K. Toward consistency in breastfeeding definitions. Studies in family planning 1990;21(4):226 - 30. 22. Piper S, Parks PL. Use of an intensity ratio to describe breastfeeding exclusivity in a national sample. Journal of Human Lactation 2001;17(3):2 27 - 32. 23. Bonuck KA, Trombley M, Freeman K, et al. Randomized, controlled trial of a prenatal and postnatal lactation consultant intervention on duration and intensity of breastfeeding up to 12 months. Pediatrics 2005;116(6):1413 - 26. 24. Kaplan BH, Cass el JC, Gore S. Social support and health. Medical care 1977;15(5):47 - 58. 25. Onis M, Dewey KG, Borghi E, et al. The World Health Organization's global target for reducing childhood stunting by 2025: rationale and proposed actions. Maternal & child nutritio n 2013;9(S2):6 - 26. 26. Organization WH. Exclusive breastfeeding for six months best for babies everywhere. World Health Organization 2011 27. Howie PW, Forsyth JS, Ogston SA, et al. Protective effect of breast feeding against infection. Bmj 1990;300(6716):11 - 16. 28. Debes AK, Kohli A, Walker N, et al. Time to initiation of breastfeeding and neonatal mortality and morbidity: a systematic rev iew. BMC public health 2013;13(3):1. 224 29. Group NS. Timing of initiation, patterns of breastfeeding, and infant survival: prospective analysis of pooled data from three randomised trials. The Lancet Global Health 2016;4(4):e266 - e75. 30. Victora C, Barros A. Effect of breastfeeding on infant and child mortality due to infectious diseases in less developed countries. The Lancet 2000;355(9202):451 - 55. 31. Hirschman C, Butler M. Trends and differentials in breast feeding: an update. Demo graphy 1981;18(1):39 - 54. 32. Binns C, Lee M, Low WY. The long - term public health benefits of breastfeeding. Asia - Pacific Journal of Public Health 2016;28(1):7 - 14. 33. CDC breastfeeding report card; 2014. Web. 34. Prevention CfDCa. Breastfeeding among US children born 2002 2012, national immunization surveys. August 1, 2016 ed, 2012. 35. Pediatrics AAo. American Academy of Pediatrics Breastfeeding Initiatives. 2010, 2009. 36. Mathews T, MacDorman MF. Infant mortality statistics from the 2010 period linked birth/infant death data set. Natl Vital Stat Rep 2013;62(8):1 - 26. 37. McKinney CO, Hahn - Holbrook J, Chase - Lansdale PL, et al. Racial and Ethnic Differences in Breastfeeding. Pediatrics 2016 doi: 10.1542/peds.2015 - 2388 . 38. Mitra AK, Khoury AJ, Hinton AW, et al. Predictors of breastfeeding intention among low - income women. Maternal and child health journal 2004;8(2):65 - 70. 39. Dubois L, Girard M. Social determinants of initiation, duration and exclusivity of breastfeeding at the population level: the resu lts of the Longitudinal Study of Child Development in Quebec (ELDEQ 1998 - 2002). Canadian Journal of Public Health/Revue Canadienne de Sante'e Publique 2003:300 - 05. 40. Bentley ME, Dee DL, Jensen JL. Breastfeeding among low income, African - American women: power, beliefs and decision making. The Journal of nutrition 2003;133(1):305S - 09S. 41. Cricco Lizza R. Black Non Feeding Methods by Nurses and Physicians. Journal of Obstetric, Gynecologic, & Neo natal Nursing 2006;35(2):173 - 80. 42. Kramer MS, Kakuma R. The optimal duration of exclusive breastfeeding. Protecting Infants through Human Milk: Springer 2004:63 - 77. 43. Asiodu I, Flaskerud JH. Got milk? A look at breastfeeding from an African American perspective. Issues in mental health nursing 2011;32(8):544 - 46. 225 44. Tiedje LB, Schiffman R, Omar M, et al. An ecological approach to breastfeeding. MCN: The American Jo urnal of Maternal/Child Nursing 2002;27(3):154 - 61. 45. Persad MD, Mensinger JL. Maternal breastfeeding attitudes: association with breastfeeding intent and socio - demographics among urban primiparas. Journal of community health 2008;33(2):53 - 60. 46. Stueb e AM, Bonuck K. What predicts intent to breastfeed exclusively? Breastfeeding knowledge, attitudes, and beliefs in a diverse urban population. Breastfeeding Medicine 2011;6(6):413 - 20. 47. Kornides M, Kitsantas P. Evaluation of breastfeeding promotion, sup port, and knowledge of benefits on breastfeeding outcomes. Journal of child health care 2013:1367493512461460. experiences of women in different racial and ethnic gr oups: pregnancy risk assessment and monitoring system (PRAMS). Maternal and child health journal 2012;16(8):1672 - 78. 49. Thulier D, Mercer J. Variables associated with breastfeeding duration. Journal of Obstetric, Gynecologic, & Neonatal Nursing 2009;38(3 ):259 - 68. 50. Dennis CL. Breastfeeding peer support: maternal and volunteer perceptions from a randomized controlled trial. Birth 2002;29(3):169 - 76. 51. Grummer - Strawn LM, Shealy KR. Progress in protecting, promoting, and supporting breastfeeding: 1984 2 009. Breastfeeding Medicine 2009;4(S1):S - 31 - S - 39. 52. Robinson K, Fial A, Hanson L. Racism, Bias, and Discrimination as Modifiable Barriers to Breastfeeding for African American Women: A Scoping Review of the Literature. Journal of midwifery & women's hea lth 2019;64(6):734 - 42. 53. Ramos DE. Breastfeeding: A bridge to addressing disparities in obesity and health. Breastfeeding Medicine 2012;7(5):354 - 57. 54. Kaufman L, Deenadayalan S, Karpati A. Breastfeeding ambivalence among low - income African American a nd Puerto Rican women in north and central Brooklyn. Maternal and child health journal 2010;14(5):696 - 704. 55. Dunaway WA. The African - American family in slavery and emancipation: Cambridge University Press 2003. 56. Leininger MM, McFarland MR. Culture care diversity & universality: A worldwide nursing theory: Jones & Bartlett Learning 2006. 226 57. McKee MD, Zayas LH, Jankowski K. Breastfeeding intention and practice in an urban minority population: relationship to maternal depressive symptoms and mother infant closeness. Journal of reproductive and infant psychology 2004;22(3):167 - 81. 58. Underwood S, Pridham K, Brown L, et al. Infant feeding practices of low - income African American women in a central city community. Journal of Community Health Nu rsing 1997;14(3):189 - 205. 59. Bentley M, Gavin L, Black MM, et al. Infant feeding practices of low - income, African - American, adolescent mothers: an ecological, multigenerational perspective. Social science & medicine 1999;49(8):1085 - 100. 60. Nommsen - Rivers LA, Chantry CJ, Cohen RJ, et al. Comfort with the idea of formula feeding helps explain ethnic disparity in breastfeeding intentions among expectant first - time mothers. Breastfeeding Medicine 2010;5(1):25 - 33. 61. Mickens AD, Modeste N, Montgomery S, et al. Peer support and breastfeeding intentions among black WIC participants. Journal of Human Lactation 2009;25(2):157 - 62. - an interview study. International breastfeeding journal 2019;14(1):51. The Journal of perinatal educati on 2015;24(2):102 - 09. 64. Chen PG, Johnson LW, Rosenthal MS. Sources of education about breastfeeding and breast pump use: what effect do they have on breastfeeding duration? An analysis of the Infant Feeding Practices Survey II. Maternal and child health journal 2012;16(7):1421 - 30. 65. Scott JA, Shaker I, Reid M. Parental attitudes toward breastfeeding: their association with feeding outcome at hospital discharge. Birth 2004;31(2):125 - 31. 66. Scott JA, Binns CW, Oddy WH, et al. Predictors of breastfeedi ng duration: evidence from a cohort study. Pediatrics 2006;117(4):e646 - e55. 67. Hannon PR, Willis SK, Bishop - Townsend V, et al. African - American and Latina adolescent Jour nal of Adolescent Health 2000;26(6):399 - 407. 68. Rosenberg KD, Eastham CA, Kasehagen LJ, et al. Marketing infant formula through hospitals: the impact of commercial hospital discharge packs on breastfeeding. American Journal of Public Health 2008;98(2):290 - 95. 227 69. Chantry CJ, Dewey KG, Peerson JM, et al. In - Hospital Formula Use Increases Early Breastfeeding Cessation Among First - Time Mothers Intending to Exclusively Breastfeed. The Journal of pediatrics 2014;164(6):1339 - 45.e5. 70. Gomez - Pomar E, Blubaugh R. The Baby Friendly Hospital Initiative and t he ten steps for successful breastfeeding. a critical review of the literature. Journal of Perinatology 2018;38(6):623 - 32. 71. Organization WH. Evidence for the ten steps to successful breastfeeding. 1998 . 72. Lind JN, Perrine CG, Li R, et al. Racial dis parities in access to maternity care practices that support breastfeeding United States, 2011. MMWR Morb Mortal Wkly Rep 2014;63(33):725 - 28. 73. Forste R, Weiss J, Lippincott E. The Decision to Breastfeed in the United States: Does Race Matter? Pediatrics 2001;108(2):291 - 96. doi: 10.1542/peds.108.2.291 . 74. Wambach KA, Koehn M. Experiences of infant feeding decision making among urban economically disadvantaged pregnant adolescents. Journal of advanced nursing 2004;48(4):361 - 70. 75. Brown C, Dodds L, Leg ge A, et al. Factors influencing the reasons why mothers stop breastfeeding. Can J Public Health 2014;105(3):e179 - e85. 76. Brodribb W, Fallon AB, Hegney D, et al. Identifying predictors of the reasons women give for choosing to breastfeed. Journal of Human lactation 2007;23(4):338 - 44. 77. Fabiyi C, Peacock N, Hebert - Beirne J, et al. A qualitative study to understand nativ ity differences in breastfeeding behaviors among middle - class African American and African - born women. Maternal and child health journal 2016;20(10):2100 - 11. 78. Scott J, Landers M, Hughes RM, et al. Factors associated with breastfeeding at discharge and duration of breastfeeding. Journal of paediatrics and child health 2001;37(3):254 - 61. 79. Street DJ, Lewallen LP. The influence of culture on breast - feeding decisions by African American and white women. The Journal of perinatal & neonatal nursing 2013;27 (1):43 - 51. 80. Gross TT, Davis M, Anderson AK, et al. Long - Term Breastfeeding in African American Mothers: A Positive Deviance Inquiry of WIC Participants. Journal of Human Lactation 2017:0890334416680180. 81. Spencer B, Wambach K, Domain EW. African Ame experiences: Cultural, personal, and political voices. Qualitative health research 2015;25(7):974 - 87. 228 82. Reeves CC, Close F, Simmons MC, et al. Social support indicators that influence breastfeeding decisions in mothers of North Florida. Florida Public Health Review 2006;3:1 - 7. 83. Bentley ME, Caulfield LE, Gross SM, et al. Sources of influence on intent ion to breastfeed among African - American women at entry to WIC. Journal of Human Lactation 1999;15(1):27 - 34. 84. Cottrell BH, Detman LA. Breastfeeding concerns and experiences of African American mothers. MCN: The American Journal of Maternal/Child Nursin g 2013;38(5):297 - 304. 85. Dietrich Leurer M, Misskey E. The psychosocial and emotional experience of breastfeeding: reflections of mothers. Global qualitative nursing research 2015;2:2333393615611654. 86. Locke A. Natural versus Taught' Competing Discour ses in Antenatal Breastfeeding Workshops. Journal of Health Psychology 2009;14(3):435 - 46. 87. Robinson KM, VandeVusse L. African American women's infant feeding choices: prenatal breast - feeding self - efficacy and narratives from a black feminist perspectiv e. The Journal of perinatal & neonatal nursing 2011;25(4):320 - 28. 88. McCann MF, Baydar N, Williams RL. Breastfeeding attitudes and reported problems in a national sample of WIC participants. Journal of Human Lactation 2007;23(4):314 - 24. 89. Johnston ML, Esposito N. Barriers and facilitators for breastfeeding among working women in the United States. Journal of Obstetric, Gynecologic, & Neonatal Nursing 2007;36(1):9 - 20. 90. Kimbro RT. On - the - job moms: work and breastfeeding initiation and duration for a sample of low - income women. Maternal and child health journal 2006;10(1):19 - 26. 91. Gielen AC, Faden RR, O'Campo P, et al. Maternal employment during the early postpartum period: effects on initiation and continuation of breast - feeding. Pediatrics 1991;87 (3):298 - 305. 92. Gatrell CJ. Secrets and lies: breastfeeding and professional paid work. Social Science & Medicine 2007;65(2):393 - 404. 93. Hawkins SS, Dow - Fleisner S, Noble A. Breastfeeding and the affordable care act. Pediatric clinics of North America 2015;62(5):1071. 94. House J. Work Stress and Social Support. Phillippines: Addison - Wesley 1981:24 - 25. 95. Matich JR, Sims LS. A comparison of social support variables between women who intend to breast or bottle feed. Social science & medicine 1992;34(8):919 - 27. 229 96. Hughes RB. The Development of an Instrument to Measure Perceived Emotional, Instrumental, and Informational Support in Breastfeeding Mothers. Issues in comprehensive pediatric nursing 1984;7(6):357 - 62. 97. McNatt MH, Freston MS. Social support and lactation outcomes in postpartum women. Journal of Human Lactation 1992;8(2):73 - 77. 98. Raj VK, Plichta SB. The Role of Social Support in Breastfeeding Promotion: A Literature Review. Journal of Human Lactation 1998;14(1):41 - 45. doi: 10.1177/089033449801400114 . 99. Glanz K, Rimer BK, Viswanath K. Health behavior and health education: theory, researc h, and practice: John Wiley & Sons 2008. 100. Barona - Vilar C, Escribá - Agüir V, Ferrero - Gandía R. A qualitative approach to social support and breast - feeding decisions. Midwifery 2009;25(2):187 - 94. 101. Brand E, Kothari C, Stark MA. Factors related to bre astfeeding discontinuation between hospital discharge and 2 weeks postpartum. The Journal of perinatal education 2011;20(1):36. 102. Bandura A. Self - efficacy: toward a unifying theory of behavioral change. Psychological review 1977;84(2):191. 103. Bandur a A. Social foundations of thought and action: A social cognitive theory: Englewood Cliffs, NJ, US: Prentice - Hall, Inc 1986. 104. Dennis CLE. Identifying predictors of breastfeeding self efficacy in the immediate postpartum period. Research in nursing & h ealth 2006;29(4):256 - 68. 105. Bartle NC, Harvey K. Explaining infant feeding: The role of previous personal and vicarious experience on attitudes, subjective norms, self efficacy, and breastfeeding outcomes. British journal of health psychology 2017;22(4) :763 - 85. 106. Dennis CL. The breastfeeding self efficacy scale: Psychometric assessment of the short form. Journal of Obstetric, Gynecologic, & Neonatal Nursing 2003;32(6):734 - 44. 107. Kingston D, Dennis C - L, Sword W. Exploring breast feeding self effica cy. The Journal of perinatal & neonatal nursing 2007;21(3):207 - 15. 108. McCarter Spaulding DE, Dennis CL. Psychometric testing of the breastfeeding self efficacy scale short form in a sample of black women in the United States. Research in nursing & health 2010;33(2):111 - 19. 230 109. Otsuka K, Dennis CL, Tatsuoka H, et al. The relationship between breastfeeding self efficacy and perceived insufficient milk among Japanese mothers. Journal of Obstetric, Gynecologic, & Neonatal Nursing 2008;37(5):546 - 55. 110. Henshaw EJ, Fried R, Siskind E, et al. Breastfeeding self - efficacy, mood, and breastfeeding outcomes among primi parous women. Journal of Human Lactation 2015;31(3):511 - 18. 111. Loke AY, Chan LKS. Maternal Breastfeeding Self Efficacy and the Breastfeeding Behaviors of Newborns in the Practice of Exclusive Breastfeeding. Journal of Obstetric, Gynecologic, & Neonatal Nursing 2013;42(6):672 - 84. 112. Molina Torres M, Dávila Torres RR, Parrilla Rodríguez AM, et al. Translation and validation of the breastfeeding self - efficacy scale into Spanish: data from a Puerto Rican population. Journal of Human Lactation 2003;19(1):3 5 - 42. 113. Gregory A, Penrose K, Morrison C, et al. Psychometric properties of the Breastfeeding Self Efficacy Scale Short Form in an ethnically diverse UK sample. Public Health Nursing 2008;25(3):278 - 84. 114. Wutke K, Dennis C - L. The reliability and val idity of the Polish version of the Breastfeeding Self - Efficacy Scale - Short Form: Translation and psychometric assessment. International Journal of Nursing Studies 2007;44(8):1439 - 46. 115. McCarter Spaulding D, Gore R. Breastfeeding self efficacy in women of African descent. Journal of Obstetric, Gynecologic, & Neonatal Nursing 2009;38(2):230 - 43. 116. Robinson K. African American women's infant feeding choices: Analyzing self - efficacy & narratives from a black feminist perspective: Marquette University 201 0. 117. Bronfenbrenner U. Ecology of the family as a context for human development: Rearch perspectives. Dev Psychol 1986;22 doi: 10.1037/0012 - 1649.22.6.723 . 118. Stokols D. Translating social ecological theory into guidelines for community health promot ion. American journal of health promotion 1996;10(4):282 - 98. 119. Dunn RL, Kalich KA, Henning MJ, et al. Engaging field - based professionals in a qualitative assessment of barriers and positive contributors to breastfeeding using the social ecological mode l. Maternal and child health journal 2015;19(1):6 - 16. 120. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap) a metadata - driven methodology and workflow process for providing translational research informatics support. Journa l of biomedical informatics 2009;42(2):377 - 81. 121. Noel - Weiss J, Boersma S, Kujawa - Myles S. Questioning current definitions for breastfeeding research. International breastfeeding journal 2012;7(1):9. 231 122. Yourkavitch J, Chetwynd EM. Toward consistency: Updating lactation and breastfeeding terminology for population health research. Journal of Human Lactation 2019;35(3):418 - 23. 123. Labbok MH, Starling A. Definitions of breastfeeding: call for the deve lopment and use of consistent definitions in research and peer - reviewed literature. Breastfeeding Medicine 2012;7(6):397 - 402. 124. Crane PK, Gibbons LE, Jolley L, et al. Differential item functioning analysis with ordinal logistic regression techniques: D IFdetect and difwithpar. Medical care 2006;44(11):S115 - S23. 125. Johnson DR, Creech JC. Ordinal measures in multiple indicator models: A simulation study of categorization error. American Sociological Review 1983:398 - 407. 126. Norman G. Likert scales, le Advances in health sciences education 2010;15(5):625 - 32. 127. Sullivan GM, Artino Jr AR. Analyzing and interpreting data from Likert - type scales. Journal of graduate medical education 2013;5(4):541 - 42. 12 8. Glaser BG, Strauss AL. Discovery of grounded theory: Strategies for qualitative research: Routledge 2017. 129. Johnson A, Kirk R, Rosenblum KL, et al. Enhancing breastfeeding rates among African American women: A systematic review of current psychosoci al interventions. Breastfeeding Medicine 2015;10(1):45 - 62. 130. Anstey EH, Chen J, Elam - Evans LD, et al. Racial and geographic differences in breastfeeding United States, 2011 2015. MMWR Morbidity and mortality weekly report 2017;66(27):723. 131. O'Campo P, Faden RR, Gielen AC, et al. Prenatal factors associated with breastfeeding duration: recommendations for prenatal interventions. Birth 1992;19(4):195 - 201. 132. Hector DJ. Complexities and subtleties in the measurement and reporting of breastfeeding pr actices. International breastfeeding journal 2011;6(1):5. 133. Li R, Fein SB, Grummer - Strawn LM. Association of breastfeeding intensity and bottle - emptying behaviors at early infancy with infants' risk for excess weight at late infancy. Pediatrics 2008;12 2(Supplement 2):S77 - S84. duration. Journal of advanced nursing 2008;63(4):397 - 408. 232 135. de Jager E, Broadbent J, Fuller - Tyszkiewicz M, et al. The role of psychosocial factors in exclusive breastfeeding to six months postpartum. Midwifery 2014;30(6):657 - 66. 136. de Jager E, Broadbent J, Fuller - Tyszkiewicz M, et al. A longitudinal study o f the effect of psychosocial factors on exclusive breastfeeding duration. Midwifery 2015;31(1):103 - 11. 137. Zhu J, Chan WCS, Zhou X, et al. Predictors of breast feeding self - efficacy among Chinese mothers: a cross - sectional questionnaire survey. Midwifery 2014;30(6):705 - 11. 138. Tuthill EL, McGrath JM, Graber M, et al. Breastfeeding self - efficacy: a critical review of available instruments. Journal of Human Lactation 2016;32(1):35 - 45. 139. McKinley EM, Knol LL, Turner LW, et al. The Prenatal Rating of Efficacy in Preparation to Breastfeed Scale: A New Measurement Instrument for Prenatal Breastfeeding Self - efficacy. Journal of Human Lactation 2019;35(1):21 - 31. 140. Oliveira V, Prell M, Chneg X. Economic Implications of Increased Breastfeeding Rates in WIC. In: Assistance FaN, ed. Amber Waves Magazine, 2019. 141. McCarter - Spaulding D, Gore R. Social support improves breastfeeding self - efficacy in a sample of black women. Clinical lactati on 2012;3(3):112 - 15. 142. Semenic S, Loiselle C, Gottlieb L. Predictors of the duration of exclusive breastfeeding among first time mothers. Research in nursing & health 2008;31(5):428 - 4 1. 143. McQueen KA, Dennis CL, Stremler R, et al. A pilot randomized controlled trial of a breastfeeding self efficacy intervention with primiparous mothers. Journal of Obstetric, Gynecologic, & Neonatal Nursing 2011;40(1):35 - 46. 144. Alghamdi S, Horodynski M, Stommel M. Racial and ethnic differences in breastfeeding, mat ernal knowledge, and self - efficacy among low - income mothers. Applied Nursing Research 2017 . 145. Maleki - Saghooni N, Amel Barez M, Karimi FZ. Investigation of the relationship between social support and breastfeeding self - efficacy in primiparous breastfeed ing mothers. The Journal of Maternal - Fetal & Neonatal Medicine 2019:1 - 6. 146. WIC Breastfeeding Rate Report, Spring 2016. 147. DeVane - Johnson S, Woods - Giscombé C, Thoyre S, et al. Integrative literature review of factors related to breastfeeding in African American women: Evidence for a potential paradigm shift. Journal of Human Lactation 2017;33(2):435 - 47. 148. Brown A. Mat ernal trait personality and breastfeeding duration: the importance of confidence and social support. Journal of advanced nursing 2014;70(3):587 - 98. 233 149. Boateng GO, Martin SL, Collins SM, et al. Measuring exclusive breastfeeding social support: Scale deve lopment and validation in Uganda. Maternal & child nutrition 2018;14(3):e12579. 150. Casal CS, Lei A, Young SL, et al. A critical review of instruments measuring breastfeeding attitudes, knowledge, and social support. Journal of Human Lactation 2017;33(1) :21 - 47. 151. Reeves EA, Woods - Giscombé CL. Infant - feeding practices among African American women: Social - ecological analysis and implications for practice. Journal of Transcultural Nursing 2015;26(3):219 - 26. 152. McDowell MM, Wang C - Y, Kennedy - Stephenson J. Breastfeeding in the United States; findings from the National Health and Nutrition Examination Surveys, 1999 - 2006. 2008 . 153. Lederman SA, Akabas SR, Moore BJ, et al. Summary of the presentations at the conference on preventing childhood obesity, Dec ember 8, 2003. Pediatrics 2004;114(Supplement 3):1146 - 73. 154. Hauck FR, Thompson JM, Tanabe KO, et al. Breastfeeding and reduced risk of sudden infant death syndrome: a meta - analysis. Pediatrics 2011;128(1):103 - 10. 155. Reis - Reilly H, Fuller - Sankofa N, Tibbs C. Breastfeeding in the community: Addressing disparities through policy, systems, and environmental changes interventions. Journal of Human Lactation 2018;34(2):262 - 71. 156. Sallis JF, Owen N, Fisher EB. Ecolog ical models of health behavior. Health behavior and health education: Theory, research, and practice 2008;4:465 - 86. 157. Bueno - Gutierrez D, Chantry C. Using the socio - ecological framework to determine breastfeeding obstacles in a low - income population in Tijuana, Mexico: healthcare services. Breastfeeding Medicine 2015;10(2):124 - 31. 158. Corbin J, Strauss A. Techniques and procedures for developing grounded theory. Basics of Qualitative Research, 3rd ed; Sage: Thousand Oaks, CA, USA 2008 . 159. Braun V, C larke V, Hayfield N, et al. Thematic analysis. Handbook of research methods in health social sciences 2018:1 - 18. 160. Arora S, McJunkin C, Wehrer J, et al. Major factors influencing breastfeeding rates: Mother's perception of father's attitude and milk supply. Pediatrics 2000;106(5):e67 - e67. 161. Sikorski J, Renfrew MJ, Pindoria S, et al. Support for breastfeeding m others: a systematic review. Paediatric and perinatal epidemiology 2003;17(4):407 - 17. 234 162. Breastfeeding EWG, Gynecologists ACoOa, Practice CoO. Committee Opinion No. 658: optimizing support for breastfeeding as part of obstetric practice. Obstetrics and gynecology 2016;127(2):e86. 163. EGLASH A, MALLOY ML. Breastmilk expression and breast pump technology. Clinical obstetrics and gynecology 2015;58(4):855 - 67. 164. Daly S, Kent J, Huynh D, et al. The determination of short term breast volume changes and t he rate of synthesis of human milk using computerized breast measurement. Experimental Physiology: Translation and Integration 1992;77(1):79 - 87. 165. Caulfield LE, Gross SM, Bentley ME, et al. WIC - based interventions to promote breastfeeding among African - American Women in Baltimore: effects on breastfeeding initiation and continuation. Journal of Human Lactation 1998;14(1):15 - 22.