am. 1.. niqwmm «9 1m 1. in»: _..H . . z "2.1. A4 L; . .. J .3 2U? ‘ LIBRARY Michigan State University This is to certify that the dissertation entitled LEISURE-TIME PHYSICAL ACTIVITY DURING PREGNANCY AND OFFSPRING SIZE presented by LANAY M. MUDD has been accepted towards fulfillment of the requirements for the Doctoral degree in Kinesiolgqy and Epidemiology OW” 72 6M Major Professor’ 5 Signature g/s 0? Date MSU is an Affirmative Action/Equal Opportunity Employer PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE SEP 2 o 2013 10/4/13 @515; ll. IMP: 5- 1344 9913 14 5/08 K:IProg/Acc&Pres/ClRC/DaleDue indd LEISURE-TIME PHYSICAL ACTIVITY DURING PREGNANCY AND OFFSPRING SIZE By Lanay M. Mudd A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Kinesiology and Epidemiology 2009 Cor flit" tom preg lBll' for-g pregn IECOII‘ Some : “are c 0D 36); ABSTRACT LEISURE-TIME PHYSICAL ACTIVITY DURING PREGNANCY AND OFFSPRING SIZE By Lanay M. Mudd The aims of this dissertation were to determine the effect of maternal leisure-time physical activity (LTPA) on offspring size at birth, to determine the separate and combined effect of LTPA during pregnancy and child leisure-time behavior on odds of child obesity, and to determine the reliability of postpartum recall of LTPA during pregnancy. For Aims One and Two, women enrolled in the Pregnancy Outcomes and Community Health study (1998-2004) were followed up in 2007. Follow-up efforts were extensive for a subcohort and minimal for the remainder, or non—subcohort. This resulted in 596 subcohort and 418 non-subcohort women who delivered at term and provided complete information relative to our aims. Original data collection supplied demographic, pregnancy, and birth information. Sex and gestational age specific birth weight z-scores (BWz) were calculated and offspring were categorized as small-, appropriate—, or large- for—gestational-age (SGA, AGA, and LGA, respectively). At follow-up, women recalled pregnancy LTPA and were classified as inactive, insufficiently active, or meeting LTPA recommendations. Women also reported child leisure-time behavior (mostly sedentary, some sedentary/some active, or mostly active) and child height and weight. Children were classified as healthy weight (5-<85"‘), overweight (85-<95‘h), or obese (295‘1‘) based on sex and age specific BMI percentiles. III SD m; tilt 335i laO lent Chil their POSII Reca agnn list). all m Postpg Among the non-subcohort, meeting LTPA recommendations significantly decreased odds of LGA (aOR=0.30, 95%CI: 0.14-0.64) without affecting odds of SGA. In quantile regression analyses, meeting LTPA recommendations was unrelated to BWz until the 0.65 quantile, afier which it significantly reduced BWz. Results for the subcohort were similar but non-significant. Thus, LTPA during pregnancy may benefit maternal/child health by reducing BWz only among the upper parts of the distribution. Insufficient activity and meeting LTPA recommendations during pregnancy were each associated with borderline significant reduced odds of child obesity by ~50% within the non-subcohort and by~30% within the subcohort. The highest odds of obesity was associated with maternal inactivity during pregnancy combined with sedentary child activity within both the non-subcohort (aOR=2.47, 95%CI: 0.83-7.39) and subcohort (aOR=1.76, 95%CI: 0.84-3.71). Both LTPA during pregnancy and children’s own leisure-time behavior appear to contribute towards risk of child obesity. For Aim Three, 298 women enrolled in the Michigan Alliance for the National Children’s Study Pilot Study reported min/wk spent in moderate and vigorous LTPA at their first prenatal care visit (total min/wk of LTPA calculated). At 15 -30 months postpartum, 82 women recalled LTPA during the trimester of their original interview. Recalled vigorous (rs=0.34, p=0.002) and total (rs=0.28, p=0.016) LTPA values were significantly correlated with original reports, but this was not true for moderate LTPA (rs=0. 10, p=0.350). There was high percent agreement for original vs. recalled report of any moderate, vigorous, or total LTPA (70-79%), but kappa values were low (0.02-0.18). Postpartum recall of LTPA during pregnancy appears to be moderately reliable. Copyright by LANAY M. MUDD 2009 ACKNOWLEDGMENTS I’d like to acknowledge the many people who helped me complete this work. To all my friends in HERL, thank you for your laughter, stories and support. To the ladies of POUCH, thanks for jumping in at the end and helping me to finish it off by freely sharing all your expertise. To all my helpers with PAPOS, thanks for your hard work and energy as you collected data and helped me explore new research questions with it. I’d like to specifically thank Josh Ode, Jeremy Knous, Patti Bauer, and Sarah Nechuta for all their friendship and advice that helped me get through this entire process. Also thanks to my family who always encouraged me to pursue my dreams. And finally last, but not least, to my advisor and committee members, thanks for inspiring me to formulate and test new questions, to carefully consider the results, and to find the public significance in the work that I do. I have enjoyed the journey to the filIICSt and look forward to life—long collaborations. USTOI USICH (HAPT RESE 0R0. REFE CHAPI IXTR IER‘ Ema MAT ERI FAC' liil BUT SEN REF CHAP' THEB ABS INTI MEI RES DB( ACI APP REF TABLE OF CONTENTS LIST OF TABLES ............................................................................................................ IX LIST OF FIGURES .......................................................................................................... XI CHAPTER ONE: INTRODUCTION ................................................................................. 1 RESEARCH AIMS: ........................................................................................................ 3 ORGANIZATION OF THE DISSERTATION .............................................................. 5 REFERENCES ................................................................................................................ 7 CHAPTER TWO: REVIEW OF LITERATURE PART ONE: BIRTH SIZE ................... 8 INTRODUCTION ........................................................................................................... 8 TERMINOLOGY ............................................................................................................ 8 EPIDEMIOLOGY OF LARGE INFANTS ................................................................... 10 MATERNAL AND OFFSPRING HEALTH RISKS ASSOCIATED WITH HIGH BIRTH WEIGHT .......................................................................................................... 10 FACTORS RELATED TO INFANT SIZE AT BIRTH ................................................ 12 MATERNAL SIZE AND INFANT SIZE AT BIRTH .................................................. l4 Epidemiology of Maternal Overweight Status and Gestational Weight Gain ........... 14 Relationships Among Maternal Anthropometric Variables and Infant Birth Size 1 6 MATERN AL PHYSICAL ACTIVITY DURING PREGNANCY ............................... I7 Epidemiology of Physical Activity during Pregnancy ............................................... 18 Health Benefits of Physical Activity during Pregnancy ............................................ 20 MATERNAL PHYSICAL ACTIVITY AND INFANT SIZE AT BIRTH .................... 21 Mechanism of Aflect of Physical Activity on Birth Size ............................................. 24 MATERN AL SIZE, MATERNAL PHYSICAL ACTIVITY AND INFANT SIZE AT BIRTH ........................................................................................................................... 25 SUMMARY .................................................................................................................. 26 REFERENCES .............................................................................................................. 28 CHAPTER 3: MATERNAL PHYSICAL ACTIVITY DURING PREGNANCY AND THE BIRTH WEIGHT DISTRIBUTION: WHERE IS THE EFFECT? ......................... 40 ABSTRACT .................................................................................................................. 40 INTRODUCTION ......................................................................................................... 42 METHODS .................................................................................................................... 43 RESULTS ...................................................................................................................... 48 DISCUSSION ............................................................................................................... 5 1 ACKNOWLEDGEMENTS .......................................................................................... 65 APPENDIX: QUESTIONS FROM THE POUCH FOLLOW-UP SURVEY ............... 66 REFERENCES .............................................................................................................. 69 vi (HAPIEI NRO IERXII lain EPEDI‘.‘ HEAL" 5}]. Ir I102; CHILI (in; CM Sim: FARE Sun PERI? 8;"! Ila. Ila Us I J u' .lla Br. SUI RE Fl CHAP' SEDE} ABS [\"II MEI DlSl REF CHAP .‘lCTIi INI ME RES CHAPTER 4: REVIEW OF LITERATURE PART TWO: CHILD SIZE ....................... 72 INTRODUCTION ......................................................................................................... 72 TERMINOLOGY OF CHILD WEIGHT STATUS ...................................................... 72 Validity of Parentally Reported Child Body Size Values .......................................... 73 EPIDEMIOLOGY OF CHILDHOOD OVERWEIGHT STATUS ............................... 75 HEALTH RISKS OF OVERWEIGHT IN EARLY CHILDHOOD .............................. 77 Short- Term Health Risks ........................................................................................... 77 Long-Term Health Risks ............................................................................................ 78 CHILD BEHAVIORS AND CHILD BODY SIZE ....................................................... 79 Child Physical Activity .............................................................................................. 80 Child Dietary Patterns ............................................................................................... 83 Summary of Child Behaviors and Weight Status ....................................................... 85 PARENTAL CHARACTERISTICS AND CHILD BODY SIZE ................................. 85 Summary of Parental Influences and Child Weight .................................................. 87 PERWATAL EXPOSURES AND CHILD BODY SIZE ...................... ' ....................... 88 Birth Weight and Child Size ...................................................................................... 88 Maternal Size / Weight Gain during Pregnancy and Child Size ............................... 89 Maternal Diabetes in Pregnancy and Child Size ...................................................... 91 Maternal Smoking during Pregnancy and Child Size ............................................... 92 Maternal Nutrition during Pregnancy and Child Size .............................................. 94 Maternal Physical Activity during Pregnancy and Child Size .................................. 95 Breastfeeding and Child Body Size ........................................................................... 97 SUMMARY .................................................................................................................. 98 REFERENCES ............................................................................................................ 10 1 CHAPTER 5: MATERNAL PHYSICAL ACTIVITY DURING PREGNANCY, CHILD SEDENTARY BEHAVIOR, AND CHILD WEIGHT STATUS .................................. 111 ABSTRACT ................................................................................................................ 111 INTRODUCTION ....................................................................................................... l 13 METHODS .................................................................................................................. 1 14 RESULTS .................................................................................................................... 119 DISCUSSION ............................................................................................................. 124 REFERENCES ............................................................................................................ 136 CHAPTER SIX: RELIABILITY OF RECALLING LEISURE-TIME PHYSICAL ACTIVITY PERFORMED DURING PREGNANCY ................................................... 140 ABSTRACT ................................................................................................................ 140 INTRODUCTION ....................................................................................................... 141 METHODS .................................................................................................................. 142 RESULTS .................................................................................................................... 145 DISCUSSION ............................................................................................................. 147 vii ACKNOWLEDGEMENTS ........................................................................................ 156 REFERENCES ............................................................................................................ 157 CHAPTER SEVEN: SUMMARY AND RECOMMENDATIONS .............................. 160 SUMMARY OF RESULTS ........................................................................................ 160 RECOMMENDATIONS ............................................................................................ 162 viii Ia'c lab lab] Iabl LIST OF TABLES Table 2.1: 1990 IOM Recommended Gestational Weight Gain Ranges According to Pre— Pregnancy Weight Status .............................................................................................. 15 Table 3.1: Participant characteristics of the POUCH study non-subcohort and subcohort ....................................................................................................................................... 57 Table 3.2: Linear regression analyses for the unadjusted and adjusted association between maternal physical activity during pregnancy and birth weight z-score by subcohort status .............................................................................................................................. 59 Table 3.3: Associations among maternal physical activity during pregnancy and appropriateness of size-for-gestational-age at birth (reference: AGA) by subcohort status .............................................................................................................................. 60 Table 4.1: Prevalence of high BMI for age among US Children, 2003-2006 .................. 76 Table 5.1: Maternal characteristics of the POUCH study non-subcohort and subcohort ..................................................................................................................................... 130 Table 5.2: Child characteristics of the POUCH study non-subcohort and subcohort 131 Table 5.3: Associations among maternal LTPA during pregnancy, child leisure-time behavior, and child weight status at 3—9 years (Reference: Healthy Weight) within the POUCH study non-subcohort and subcohort .............................................................. 132 Table 5.4: Associations among maternal/child LTPA and child weight status at 3-9 years (Reference: Healthy Weight) within the POUCH study non-subcohort and subcohort ..................................................................................................................................... 133 Table 5.5: The effect of maternal LTPA during pregnancy and child leisure-time behavior on obesity status among children 3-9 years in the POUCH cohort (n=703§). ............ 134 Table 6.1: Participant characteristics during pregnancy according to information reported during the original interview. ...................................................................................... 153 Table 6.2: Average physical activity values and mean differences in originally reported and recalled values for moderate, vigorous, and total leisure-time physical activity among women who completed follow-up (n=82) ....................................................... 154 Table 6.3: Spearrnan correlations (rs, p-value) between originally reported and recalled min/wk of moderate, vigorous, and total leisure-time physical activity for the group as a whole and by participant subgroups ......................................................................... 154 ix It: rs: AC1 Table 6.4: Categorical agreement between originally reported and recalled participation in moderate, vigorous, and total leisure-time physical activity, and participation in ACOG recommended amounts of leisure-time physical activity ................................ 155 LIST OF FIGURES Figure 3.1: Unadjusted sasociations of LTPA during pregnancy and birth weight z-score among the POUCH study non-subcohort participants across quantiles ........................ 61 Figure 3.2: Adjusted associations of LTPA during pregnancy and birth weight z-score among the POUCH study non—subcohort participants across quantiles ........................ 62 Figure 3.3: Unadjusted associations of LTPA during pregnancy and birth weight z-score among the POUCH study subcohort participants across quantiles ............................... 63 Figure 3.4: Adjusted associations of LTPA during pregnancy and birth weight z-score among the POUCH study subcohort participants across quantiles ............................... 64 xi ‘1- ,3") L‘s.>\ ’" you w a‘ _ Mu ..'... w a.“ — t, .. “IC‘Q 4 ‘I 1.251, r CHAPTER ONE: INTRODUCTION Giving birth to an infant that is either too small or too large is associated with adverse health outcomes for both the mother and her child. While the causes and health risks of small infants have received considerable attention, few researchers have examined the etiology of excess fetal growth leading to large birth size. In addition to having an increased risk of birth trauma, it appears that large infants are predisposed towards becoming overweight children." 2 Thus, examining modifiable factors that influence birth size may uncover new pathways for preventing childhood obesity. Past research among select samples of women has shown that maternal leisure- time physical activity (LTPA) during pregnancy decreases birth size modestly,3 and reduces the risk of being born large for gestational age.4 However, the independent effects of pro-pregnancy LTPA, trimester-specific LTPA, and maternal body size on the distribution of offspring birth size are not easy to delineate. In addition, only a couple of small studies have evaluated the relationship between maternal LTPA and offspring Size during toddlerhood/childhood.5 Gaining more insight into the effect of maternal LTPA on birth size could inform future research concerning LTPA recommendations for women of childbearing age. Furthermore, understanding the role of maternal LTPA on offspring growth may also help to formulate new approaches for preventing childhood overweight and obesity. Finally, the reliability of recalling pregnancy-related LTPA needs to be established in order to facilitate future epidemiological research on the long-term effects of LTPA during pregnancy on maternal and child health outcomes. To date, only one small study has examined the validity of recalling LTPA performed during pregnancy at Ch; It‘ll oils Ill and LIP 6 years postpartum.6 Using a sample of white, upper—middle class women with mainly high levels of LTPA measured during pregnancy, Bauer et al. found moderate to high agreement between originally measured and recalled amounts of LTPA. However, LTPA had been measured precisely via accelerometry, heart rate telemetry, and LTPA surveys at two time-points during pregnancy among these women.6 Thus, their ability to recall LTPA may have been influenced by greater attention to their activity during pregnancy. More research is needed to determine whether women with more diverse personal characteristics and/or LTPA participation may have similarly high recall abilities and whether participant characteristics may influence recall ability. This dissertation evaluates the effects of maternal LTPA during pregnancy on offspring size at birth and in childhood. Additionally, the reliability of pregnancy-related LTPA recalled at two years postpartum is evaluated. Data from the Pregnancy Outcomes and Community Health (POUCH) Study were used to determine relationships among LTPA during pregnancy and offspring size. Women enrolled in the POUCH study were contacted at 3-9 yrs postpartum and asked to recall type, frequency, and duration of up to two leisure-time physical activities they performed most often during atypical week in their POUCH pregnancy. They also reported their children’s current height and weight at that time. Previously collected data provided birth weight, gestational age at delivery, and descriptive characteristics. Women enrolled in the Michigan Alliance for the National Children’s Study (MANCS) pilot study were followed-up to assess reliability of recalled LTPA. Women originally reported participation in moderate and vigorous LTPA while pregnant in the summer of 2006. These women were contacted in fall of 2008 to recall their pre- preg! mail: addr: RJZS Sheri birth. gesta (AGE 5pm] in 111 Ol'cnr. ObESQ‘ a firm- Sedi‘nlz pregnancy and trimester specific LTPA using the Modified Activity Questionnaire via a mailed survey or internet survey.7 Using data from these two cohorts, this dissertation addresses the following three Specific Aims. RESEARCH AIMS: Specific Aim 1: To evaluate the effect of LTPA during pregnancy on offspring size at birth, measured continuously as birth weight z-score and categorically as size for gestational age (i.e., small for gestational age (SGA), appropriate for gestational age (AGA), and large for gestational age (LGA)). H 1.1. We hypothesize that meeting LTPA recommendations during pregnancy will not be associated with a mean change in birth weight z-score in linear regression analyses. H 1.2. We hypothesize that meeting LTPA recommendations during pregnancy will reduce odds of delivering an LGA infant, without affecting the odds of delivering SGA in logistic regression analyses. H 1.3. We hypothesize that meeting LTPA recommendations during pregnancy will be associated with lower birth weight z-score only among the upper quantiles of the birth weight z-score distribution in quantile regression analyses. Specific Aim 2: To determine the separate and combined effect of maternal participation in LTPA during pregnancy and child leisure-time behavior on odds of childhood overweight (body mass index (BMI) 2 85th to <95th age and sex-specific percentile) and obese (BMI 2 95th age and sex-specific percentile) status. To assess the interactive effect, a four-category combined variable will be created: inactivity during pregnancy and sedentary child activity (-/—), inactivity during pregnancy and non-sedentary child activity dunr Shem apprt Whetl (-/+), any LTPA during pregnancy and sedentary child activity (+/-), and any LTPA during pregnancy and non-sedentary child activity (+/+). H 2.1. We hypothesize that meeting LTPA recommendations during pregnancy will reduce the odds of child overweight and obese status when compared to inactivity during pregnancy. H 2.2. We hypothesize that more active child leisure-time behavior will be have reduce the odds of child overweight and obese status when compared to sedentary leisure-time behavior. H 2.3. We hypothesize that inactivity during pregnancy combined with sedentary child activity (-/-) will be associated with the highest odds of child overweight and obese status when compared to the +/+ condition. Specific Aim 3: To determine the reliability of recalling LTPA during pregnancy at approximately two years postpartum among a diverse sample of women and examine whether participant characteristics are associated with recall ability. H 3.1. We hypothesize that moderate to strong correlations (i.e., r-values 20.4) will exist between originally reported and recalled moderate, vigorous and total LTPA. H 3.2. We hypothesize that strong kappa values (i.e., 20.8) will indicate good categorical agreement such that women will fall into similar categories of “none vs. any” LTPA and “meeting vs. not meeting” LTPA recommendations when using original and recalled reports of LTPA. H 3.3. Examining the influences of participant characteristics on recall ability is a descriptive aim, thus we have no hypothesis for the direction of effect. We will OR bin] of 3 Eat] refs: diret size. gestz app]: lx'nm1 Plfgn mimic the re create strata of women based on descriptive characteristics and compare correlation coefficients for original and recalled moderate, vigorous and total LTPA across strata to assess this portion of Aim Three. ORGANIZATION OF THE DISSERTATION This dissertation is divided into two parts, the first focusing on offspring size at birth and the second on child body size and reliability of LTPA recall. Both parts consist of a review of relevant literature followed by a paper(s) that evaluate the Specific Aims. Each paper is in manuscript form (abstract, introduction, methods, results, discussion, and references). The final chapter provides an overall summary of all results and discusses the direction of future research. Chapters Two and Three evaluate the effects of maternal LTPA on offspring birth size, measured as birth weight z-scores and categorical appropriateness of size for gestational age. Chapter Two is a literature review that introduces common terminology applied to birth size, reviews the epidemiology of large infants, and explores factors known to be related to birth size. It focuses primarily on the influences of maternal size and maternal LTPA on offspring birth size, while highlighting gaps in our current knowledge related to these topics. Chapter Three evaluates the effects of LTPA during pregnancy on the distribution of size at birth using linear, polytomous logistic, and quantile regression analyses (Specific Aim One). Chapters Four and Five consider the relationships among maternal LTPA during pregnancy, child LTPA, and child body size. Chapter Four is a literature review that introduces the epidemiology of overweight status during toddlerhood and the stability of the relationship between size at birth and toddlerhood. It also discusses the available htere Chat IIII iionr fiacto years sugge repres pregn. resear prerer literature linking maternal LTPA during pregnancy to offspring size in childhood. Chapter Five evaluates whether maternal LTPA during pregnancy combined with child LTPA level (measured as sedentary vs. non-sedentary activity choice) protects children from being classified as overweight or obese while controlling for possible confounding factors (Specific Aim Two). The reliability of recalled LTPA during pregnancy at two years postpartum is addressed in Chapter Six (Specific Aim Three). Finally, Chapter Seven provides an overall summary of results and offers suggestions for future research. The research questions addressed by this dissertation represent important steps in evaluating the role of maternal LTPA participation during pregnancy on offspring growth and development. Results of this study will inform future research on LTPA interventions among pregnant women and will aid childhood obesity prevention efforts. RIFEI .4 I -4 C REFERENCES l. Hediger ML, Overpeck MD, McGlynn A, Kuczmarski RJ, Maurer KR, Davis WW. Growth and fatness at three to six years of age of children born small- or large—for-gestational age. Pediatrics. Sep 1999;104(3):e33. Haram K, Pirhonen J, Bergsjo P. Suspected big baby: a difficult clinical problem in obstetrics. Acta Obstet Gynecol Scand. Mar 2002;81(3): 185-194. Perkins CC, Pivamik JM, Paneth N, Stein AD. Physical activity and fetal growth during pregnancy. Obstet Gynecol. Jan 2007;109(1):81-87. Alderman BW, Zhao H, Holt VL, Watts DH, Beresford SA. Maternal physical activity in pregnancy and infant size for gestational age. Ann Epidemiol. Nov 1998;8(8):513-519. Clapp J F . The morphometric and neurodevelopmental outcome at five years of the offspring of women who continued exercise throughout pregnancy. J Pediatr. Dec 1996;129(6):856-863. Bauer PW, Pivamik J M, F eltz DL, Paneth N, Womack CJ. Validation of a historical physical activity recall tool on postpartum women. Med Sci Sports Exerc. 2006; 38(Suppl): S561. Kriska AM. Modifiable Activity Questionnaire. Med Sci Sports Exerc. 1997;29(Suppl):S73-S78. INTI mesh estab OVCI: OUICI p016: heal pre‘. (int 531". We sit Ci CHAPTER TWO REVIEW OF LITERATURE PART ONE: BIRTH SIZE INTRODUCTION While several factors are known to be associated with birth size, causal mechanisms for giving birth to an infant either too small or too large have yet to be fully established. The etiology and health risks of being born small have received considerable attention; however, few researchers have focused on excess fetal growth. Fetal overgrth leading to large birth size is associated with increased risk of adverse birth outcomes and future childhood overweight status. M Thus, gaining more insight on potentially modifiable determinants of birth size could improve both short- and long-term health outcomes. Both pro-pregnancy body size and gestational weight gain are positively related to infant birth size, with the largest infants born to overweight/obese women who gain excess weight during pregnancy.5’ 6 Recent evidence suggests that LTPA during pregnancy may help control excess Size at birth7’ 8 and decrease the risk of overfatness during childhood.9 However, these results were based on small, non-representative samples of women. . Further investigation of the relationships among maternal size and weight gain, LTPA during pregnancy, birth size, and child body size has the potential to elucidate a practical intervention to promote healthy pregnancies, birth outcomes, and child weight status. TERMINOLOGY Before examining factors associated with birth size, some terminology must be established. Unadjusted birth weight is used to describe birth size in the crudest form. U5 US. he: 3P1 bit. this While a standard definition exists for low birth weight (birth weight < 2.5 kg), multiple definitions exist for high birth weight or macrosomia (birth weight > 4.0, 4.2, or 4.5 kg).”"'4 Since birth weight is irrevocably tied to the length of gestation, adjusting birth weight for gestational age provides a more precise description of the appropriateness of birth size. Birth weight is often compared to a population standard for the gestational age of the infant and categories are used to define infants born SGA (birth weight < the 10th percentile), AGA (birth weight between the 10th -< 90th percentile), and LGA (birth weight > the 90th or 95th percentile).15 ’ '6 Once again, differing cut-points have been used to designate LGA infants. Within US. populations, the most commonly accepted cut- point for LGA is the 90th percentile; however, some authors have argued for more stringent definitions (>2 standard deviations above mean birth weight), especially among diabetic pregnancies.” '6 Previous literature mainly used absolute terms to express birth size (e. g. low birth weight or macrosomia), while more recent literature has preferred using relative measures (e.g. SGA or LGA) to describe size at birth. Using either absolute or relative cut-points to designate in/appropriate birth size is useful for calculating and interpreting health risks; however these methods assume that health risks change dramatically at a given threshold value of birth weight. In reality, there is little evidence that such a threshold exists for excess fetal growth. Alternatively, approaching birth size as a continuum increases power to detect factors that influence birth size at the extremes, as well as within the normal range. For continuous analyses this dissertation uses birth weight z-scores (BWz), which are calculated as the observed minus the expected birth weight, divided by the standard deviation of birth weight from a population standard. The expected birth weight is derived from gestational age and 26111161 51"; A: 3 .. - .dtit‘l}. EPIDENI .I - l0 3 1 13,. Ltl .— Iiomer. gi percentile Whites. .\' \\ birth to n' I‘Th \t‘ei 1098?“ I from 6.5 birth Wei [firm bill: [lllS lime HATER BIRTH I bfllh [he RTOIOnge ceSflTfi‘an risks for gender specific growth curves, thereby controlling for these potentially confounding factors.'7 EPIDEMIOLOGY OF LARGE INFANTS The proportion of infants born macrosomic or LGA varies among populations from 5 to 20% and is highest in Nordic countries.“ '8’ '9 A sizable proportion of US. women give birth to large infants. Specifically, the incidence of LGA births (>90th percentile) ranges from ~6% among non-Hispanic Blacks, to 11% in non-Hispanic Whites, 8% in Hispanics, and 12% among Native Americans.20 Worldwide, there has been a 15-25% increase in the proportion of women giving birth to macrosomic/LGA infants over the past two decades. '8’ 2'45 In the US, both mean birth weight and BWz increased in term-born white and black populations from 1985 to 1998.26 The proportion of LGA births also increased from 11.5 to 12% for whites and from 6.5 to 7% among blacks (Hispanic ethnicity not considered).26 These increases in birth weight and rates of LGA were only observed among term births. Pre-term and post- term births demonstrated small reductions in birth weight, BWz, and rates of LGA during this time-frame in the US.26 MATERNAL AND OFFSPRING HEALTH RISKS ASSOCIATED WITH HIGH BIRTH WEIGHT Delivering a large infant is associated with several adverse health outcomes for both the mother and her offspring. Maternal risks associated with delivery include prolonged/assisted delivery, vaginal, cervical, and/or perineal lacerations, emergency cesarean delivery, and abnormal hemorrhage.” '6’ ”'29 Vital records data indicate that risks for obstetric complications increase with progressively large infants.30 For example, 10 their “hen kg.u‘ cesan fimes gener chfldl 13.31 ‘ theris those 132); l95°o( resuln Vagnia 2496a bIaChj; COmpa llSk 0f asphati beRVee the odds of cesarean delivery increases about 1.5 times (OR=1.6, 95%CI= 1.61-1.63) when giving birth to an infant of 4.00-4.49 kg compared to a birth weight of 300-399 kg, while a birth weight of 4.50—4.99 kg is associated with more than twice the risk of cesarean (OR=2.6, 95%CI: 2.58-2.64) and a birth weight 2 5.00 kg has more than four times the risk (0R4.7, 95%CI=4.54-4.83).3O Long—term maternal health outcomes associated with giving birth to a macrosomic/LGA infant include anal dysfunction and 13.31 general perineal defects. However, the permanence and/or clinical significance of childbirth-related pelvic floor trauma among the US population remain unclear.32 Macrosomic/LGA infants face an increased risk of birth trauma during delivery.4’ '3’ 2" 30’ 33 Vital data from 1995-1997 indicate that infants born at 4.00-4.50 kg have twice the risk of birth injuries as infants born 3.00-3.99 kg (OR=2.0, 95%CI=1.92-2.05), while those weighing 4.50-4.99 kg have more than three-times the risk (OR=3.1, 95%CI=2.96- 3.32) and birth weight 2 5.00 kg is associated with 4.5-fold increased risk of trauma (95%CI=3.95-5.19).30 The most common birth injuries include shoulder dystocia and resulting plexus injuries.4’ 34 Shoulder dystocia is estimated to complicate ~2% of all vaginal deliveries; however, when birth weight exceeds 4.5 kg this risk increases to 9- 24% among non-diabetic women and 20-50% among diabetic pregnancies.35 The risk of brachial plexus injury is approximately 18- to 21-fold higher among macrosomic infants compared to normal birth weight infants as well.35 Aside from injuries sustained at birth, macrosomic/LGA infants have an increased risk of congential anomalies, hypoglycemia, hyaline membrane disease, and meconium aspiration 13’ 30‘ 36 Additionally, national data indicate that optimum birth weight occurs between 3.0-4.0 kg, above which mortality rates begin to increase.37 Analyses of US. 11 dc may l‘atlo ll, an infant death files support that infants with a birth weight 2 5.00 kg have a 2.7 increased odds of neonatal death (95%CI=1.91-3.80) compared to normal weight infants.30 Most injuries sustained during birth resolve as the infant matures; however, there is evidence that macrosomic/LGA infants have increased body fatness at birth when compared to normal weight infants.5’ 38‘ 39 This increased fatness appears to persist into childhood, placing macrosomic/LGA infants at greater risk for obesity and metabolic disease later in life."3’ 4042 Other long-term outcomes associated with macrosomia/LGA have included high blood pressure/hypertension in adolescence and adulthood, as well as childhood cancer.4345 Thus, accumulating evidence suggests that size at birth, and in . particular, body composition at birth, represent important determinants of later health outcomes.“ 47 FACTORS RELATED TO INFANT SIZE AT BIRTH While several factors contribute towards explaining birth size, the true determinants are poorly understood. Non-modifiable factors related to birth size include parity, fetal sex, maternal age, maternal height, maternal race, and genetics.13 Mean birth weight increases with successive births and it appears that parity is also associated with increased neonatal fat mass.5’ 48 In addition, male sex, maternal age, maternal height, and Caucasian race or Hispanic ethnicity are all associated with increased mean birth weight and/or LGA. '6’ ”'53 Aside fiom the factors discussed in the previous paragraph, genetic influences may account for 30-80% of the variation in birth size.50’ 54 In attempts to isolate genetic factors, researchers have shown that fetal insulin, insulin-like growth factor (IGF)-I, IGF- II, and associated receptors are critical for normal fetal growth.” 54’ 55 Elevated levels of 12 ' I 9 ' 1 ~ 16‘3“ II . I u “ II.“ ;.\?‘n LAM-1&1. p725: D1536: lilCl-C.‘ Pr :m V \urI.I W111 i OVQI‘, tent Impk‘ URI: rec: fetal insulin and IGF—1 have been associated with increased risk for macrosomia.50 It is likely that complex interactions between parental, placental, and fetal genes exert influences on fetal growth. I 3‘ 54’ 55 Several modifiable factors have also been associated with size at birth, including pregnancy complications, maternal anthropometrics, and lifestyle-related behaviors. Diabetes during pregnancy (pre-gestational or gestational) increases risk of macrosomia/LGA and is also associated with increased neonatal fat mass.” 56 Even when examined on a continuous scale, increased maternal serum glucose levels are associated with increased birth weight.57 Overweight pre-pregnancy body mass index (BMI) and excessive weight gain are both related to increased birth weight and risk of LGA.5’ 6’ H’ 5 I In fact, population-based data indicate that once the increasing prevalence of maternal overweight and decreasing prevalence of maternal smoking are accounted for, temporal trends for increased incidence of LGA are no longer apparent, underscoring the importance of these two factors in the development of birth size.” 24 Aside from smoking during pregnancy, other lifestyle factors such as maternal nutrition and LTPA appear to influence birth size. Investigations on the Dutch famine of 1944-1945 have shown that undemutrition during pregnancy does not affect birth weight unless total caloric intake drops below 1500 kilocalories in the third trimester.” 59 A recent study of 553 women in a developed country showed that total energy intake and/or percent macronutrient intake were not significantly related to risk of extreme birth weight. ” However, the influence of maternal nutrition on birth weight variation within the normal range and subsequent adult health remains a subject of intense scientific debate.60 Relationships between LTPA during pregnancy and birth size are complex and 13 r11h3&\ nulgnuk l _ . Shit”. b 50 l ' .. 311*. >116. a. \ h (a. II endure. 31311210501 . a. 'WOIIIEII ‘ Ol em e Influen Inn filial», Ultra will be discussed in detail below. Overall, it appears that LTPA during pregnancy reduces fetal growth modestly, without increasing the risk of low birth weight.7’ 8 Other factors such as socioeconomic status/income and marital status have been hypothesized to affect birth size, but the evidence thus far is equivocals‘ 49‘ 6" 62 Despite a long list of variables associated with birth size, no single factor can accurately predict birth weight. The strongest modifiable risk factors for macrosomia/LGA across populations include pre-pregnant maternal overweight status, high pregnancy weight gain, and the presence of diabetes during pregnancy.”' 63’ 64 While women with decreased glucose tolerance deliver larger infants,“ 6568 diabetic pregnancies account for only a small percentage of LGA births due to the comparatively low incidence of gestational diabetes (~6%) and/or the prevalence of pre-pregnancy diabetes (~0.5%).63' 69 In contrast, Since ~52-60% of females of childbearing age are overweight,70 maternal size accounts for substantially more LGA births.63 Thus the influences of maternal size on birth size are presented in more detail below. MATERNAL SIZE AND INFANT SIZE AT BIRTH Epidemiology of Maternal Overweight Status and Gestational Weight Gain Despite known health risks associated with overweight, the prevalence of overweight/obesity among US. adults has continued to increase in recent years.“ Data from the Pregnancy Risk Assessment Monitoring System (PRAMS) from 26 states and New York City indicate that 23% of women are classified as overweight pre-pregnancy and 19% are obese according to BMI values calculated from self-reported data.72 State- specific prevalences range from 20-25% overweight and 14-24% obese.72 14 accor ilabl usoc indie. with and a rtsea than In 1990, the Institute of Medicine published guidelines for gestational weight gain according to pre-pregnancy BMI in an effort to optimize birth weight between 3.0-3.9 kg (Table 2.1).73 Research supports that gaining weight within the recommended ranges is associated with lower risk of delivering a SGA or LGA infant.74 Observational data also indicate that achieving a healthy weight gain within recommended amounts isassociated with decreased risk of preterm delivery, pregnancy complications such as pre-eclampsia, and adverse neonatal outcomes such as hypoglycemia and meconium aspiration.74 Recent research indicates that pregnancy outcomes for obese women may be optimized with less than recommended or even no weight gain. Results from a population-based cohort study in Missouri found that risk for pre-eclampsia, cesarean delivery, SGA, and LGA were lowest when class I obese women (BMI 30-<35 kg/mz) gained 10-25 lb, class II obese women (BMI 35-<4O kg/mz) gained 0-10 lbs, and class III obese women (BMI 40+ kg/mz) lost 2-10 lbs during pregnancy.75 Table 2.1: 1990 IOM Recommended Gestational Weight Gain Ranges According to Pre—Prflancy Weight Status 73 Pre-pregnancy BMI (kg/m2) Weight Gain (lbs) Low (<19.8) 28-40 Normal (19.8-26.0) 25-35 Overweight (>26.0-29.0) 15-25 Obese (>290) 2 15 Unfortunately, US. data indicate that only ~3 0% of pregnant women gain weight within recommended ranges, while ~46% gain excess amounts and ~23% have less than recommended weight gain during pregnancy.74 The prevalence of recommended weight gain is remarkably similar across populations, with studies involving Hispanic and/or multi-ethnic samples reporting similar rates.”78 Pre-pregnancy BMI influences 15 p? gestational weight gain with overweight women being two times more likely to gain excess weight as compared to normal weight women.” 78 Relationships Among Maternal Anthropometric Variables and Infant Birth Size Maternal height, weight, BMI prior to pregnancy, and gestational weight gain are each positively related to birth size, even with adjustment for sociodemographic factors and metabolic conditionss’ 6' 49‘ 5 " 63 Two smaller studies have also indicated that maternal overweight/obesity is associated with significantly greater neonatal fat mass and percent fat, even when overall birth weight is unaffected.” 8° Timing of gestational weight gain may also be important. Weight gain experienced from the first to second trimester is significantly related to infant length at birth, while weight gain from the second to third trimester is related to birth weights" 82 Because pre-pregnancy weight status and gestational weight gain are strongly correlated, it is difficult to determine their independent influences on birth weight. While pre-pregnancy weight status and gestational weight gain appear to have both independent and combined influences on fetal growth,6 research supports that the overweight/ obese women who gain the most weight deliver the largest infants.5' 49‘ 83 Maternal size and pregnancy weight gain may affect fetal growth by enhancing placental size and influencing placental capacity to supply nutrients to the fetus.84’ 85 Additionally, changes in lipid metabolism that accompany overweight status may also affect birth size. Clausen et al. found that high levels of leptin, a hormone involved in lipid metabolism, in maternal serum during the second trimester were associated with a birth weight greater than 4.5 kg; however, leptin concentrations were no longer significantly related after adjustment for maternal BMI.86 16 R315- Vr ilt‘ LRLI C TRIPS? inter: infinrl ‘Wuee IIPA Rsuhs 11113131 liil intl “€13 lOlin adVei The exact pathways through which maternal size and/or weight gain influences fetal growth are not yet understood; however, they are likely to be multi-factorial and involve interactions with genetic and environmental factors. Unfortunately, studies examining the contribution of maternal size and pregnancy weight gain on birth weight have rarely accounted for the influence of LTPA, which may affect both maternal and infant body size/composition. One recent study of 553 Scandinavian women found that while maternal BMI, weight gain, plasma glucose, gestational age and maternal LTPA were each independently related to risk of macrosomia, only low-level pre-pregnancy LTPA remained a significant predictor of macrosomia in adjusted analysis.ll These results indicate that maternal LTPA is a modifiable determinant of birth weight and may interact with maternal anthropometric influences. MATERNAL PHYSICAL ACTIVITY DURING PREGNANCY The American College of Obstetricians and Gynecology (ACOG) published the first US. guidelines for exercise during pregnancy in 1985.87 These original guidelines were cautious and advised women to keep their heart rate below 140 beats per minute and to limit “strenuous activity” to 15 minutes in duration.87 Specific concerns about potential adverse effects of maternal exercise included abortion, grth restriction, premature labor, fetal hypoxia, acidosis, hypertherrnia, and brain damage.88 Since that time, literature has shown that LTPA generally does not increase risk of adverse pregnancy outcomes, and may be associated with a wide range of maternal and fetal health benefits.”90 Thus, the most current ACOG guidelines (2002) recommend that all pregnant women participate in at least 30 minutes of moderate LTPA on most days of the week.9| These guidelines also state that recreational and competitive athletes may 17 1"."3‘l‘ Ed‘s-tuu.t i ‘3. "ll “rib\.']-\kk" are 3550 LS. Gt accord;- tome: strut} iilltltj proxid Elohim maintain their activities during pregnancy with medical supervision. Thus, vigorous intensity activity is not constrained and the pregnant woman is left to the discretion of her health care provider and herself to modify activity as needed. In fact, the only activities which ACOG recommends against are recreational sports with a high potential for contact or falls, scuba diving, and prolonged activity in the supine position, all of which are associated with risk for fetal trauma or impaired venous return.91 Most recently the US. Government has released evidence-based LTPA guidelines for all Americans.92 In accordance with the 2002 ACOG guidelines, these governmental guidelines state that women who are not already active should get at least 150 minutes of moderate aerobic activity per week during pregnancy, and women who habitually engage in vigorous activity may continue to do so provided they discuss their activities with their health care provider. Epidemiology of Physical Activity during Pregnancy About tvvo-thirds of pregnant women in the US. choose to engage in some amount of LTPA, but only 16% participate at recommended levels.93 Among active women, the most commonly reported LTPA during pregnancy is walking (~43-53%), followed by swimming ( ~12%) and aerobics (~10-12%).93’ 94 Few investigators have examined participation in vigorous LTPA. Using population-based data, ~6% of pregnant women report running in the past month while ~1% report playing team or racquet sports.93 A study of almost 1700 women found that 14% reported vigorous activities in the first trimester, while only 8% participated in vigorous activity in the second trimester.95 18 accent 1 decrease m Irr= IIPA 5.: I‘m 312;" Most population-based reports of LTPA participation during pregnancy do not account for gestational length; however, it is known that participation in exercise/LTPA decreases with increasing gestation.96'99 A study conducted in the Boston, Massachusetts area (n=1442) using 7—day recall measurements of LTPA, showed that mean levels of LTPA decreased from 9.6 to 6.9 hours/wk during pregnancy. '00 Additionally, the prevalence of not meeting ACOG guidelines increased from 13% pre-pregnancy to 22% during the second trimester.loo Another study of 250 women that measured several domains of physical activity found that median total energy expenditure remained fairly stable throughout pregnancy (~33 MET-hr/day), but that moderate activity energy expenditure was lower in the third trimester (0.8 MET-hr/day) compared to the first (2.3 MET-hr/day) while household/care giving energy expenditure was higher (6.8 vs. 12.5 MET-hr/day for the first vs. the third trimester, respectively).98 Thus it is possible that pregnant women replace more strenuous physical activities with lighter activities as the pregnancy progresses. Several maternal characteristics are associated with LTPA participation during pregnancy. Non-Hispanic white race/ethnicity, nulliparity, >high school education, older maternal age (>25 years), not smoking, and engaging in structured exercise pre- pregnancy are all associated with increased odds of participating in LTPA during pregnancy.96'98' 10‘ Factors associated with decreasing LTPA participation during pregnancy include parity, young maternal age, overweight pre-pregnancy BMI, low self- efficacy for LTPA, and high fiequency of pre-pregnancy LTPA.97' 100‘ '02 Reported barriers to LTPA during pregnancy include fatigue, lack of time, nausea, physical 103- discomfort, and lack of child care. '06 Studies also indicate that some women fear that 19 exerci {[15:99 111051 I H (71] Ill } . have b that pa risk for exists 5 tiabete during and eff; Who art in LIP: Pie-eels LTPA 0‘ these re: results fi lIISUffiCj tClampSi anemple. exercise might hurt their baby and perceive vigorous/ high-impact activities as unsafe. '05’ 107409 While women appear to recognize the health benefits of participating in LTPA, most rate rest and relaxation as more important than exercise during pregnancy. '09 Health Benefits of Physical Activity during Pregnancy Several maternal and fetal health benefits associated with LTPA during pregnancy have been documented. Reviews of literature published over the past two decades suggest that participation in LTPA before and/or during pregnancy is associated with decreased risk for gestational diabetes, pre-eclampsia, and preterm delivery.88'90‘ ”0" '3 Some debate exists surrounding the timing of LTPA needed for beneficial effects. For gestational diabetes, the strongest protective effect occurs with LTPA participation both before and during pregnancy.1 '3 Additionally, LTPA during pregnancy has been found to be a safe and efficacious treatment for controlling blood sugar within normal limits among women who are already diabetic or become so during pregnancy.“ ”4' ”5 Similarly, participation in LTPA both before and during pregnancy yields the strongest protective effect against pre-eclampsia.l '3 Although data are limited, it appears that participation in vigorous LTPA during pregnancy is also related to decreased risk for pre-terrn birth; however these results are based only on observational data.95’ ”6 While observational studies show clear benefits associated with maternal exercise, results from randomized trials are less conclusive. Cochrane reviews have found insufficient evidence for a protective effect of maternal exercise on the risk of either pre- eclampsia or gestational diabetes.l '7' ”8 It is important to note that very few studies have attempted to randomize exercise programs during pregnancy and that these have included 20 small samples (n=16 to 38 women per study). Thus, larger intervention studies with more power to detect results are needed before conclusions are made. Proposed mechanisms for the effect of maternal exercise on the prevention of maternal disease include enhanced placental growth and vascularity, reduced oxidative stress, reduced inflammation, improved endothelial function, and improved blood lipids.“ 90 It is also likely that differences in underlying maternal fitness, pre-pregnancy weight status, and gestational weight gain may partially account for lower rates of gestational diabetes and pre-eclampsia among women who are active during pregnancy versus those who are not. More research is needed to clarify etiological pathways for the health benefits of LTPA during pregnancy. MATERNAL PHYSICAL ACTIVITY AND INFANT SIZE AT BIRTH Evidence for the effects of LTPA during pregnancy on birth size is conflicting. While some studies show that LTPA during pregnancy decreases birth weight,1 '9' '20 121 122-124 others report increases in birth weight, some find no effect, and still others report a “U-shaped” association such that both high and low levels of LTPA decrease birth weight.‘25 These apparently conflicting results are likely due to methodological differences in assessing LTPA, a lack of control for appropriate confounders, and. variability in the choice of insufficiently active vs. completely sedentary control groups.1 '3’ '26' '27 However, inconsistent results might also reflect real differences between populations of pregnant women. While controlling for several maternal demographic characteristics, past studies have failed to account for LTPA participation prior to 122,125,128 pregnancy)” '22 have measured LTPA in only one or two trimesters, or have been based on small and/or non-diverse samples.'20"22’ ”4' '25 Despite these obstacles, the 21 mgr-3.11} . ‘ 0‘0 mtklfsUj pregrzrnt panama taxed “ 'C‘ _ 9.‘ (£1— ir. '1115 51 113763161 macroso Which la Sigrifica majority of evidence suggests that LTPA during pregnancy decreases birth weight modestly within the normal range.” '26 To date, few studies have Specifically considered the effect of LTPA during pregnancy on macrosomia/LGA. One study among 291 Coloradan women found that participation in 2 2 hrs/wk of at least moderate LTPA in the second or third trimesters reduced risk of LGA (OR=0.3, 95%CI=0.2-0.7) with no effect SGA (OR=O.8, 95%CI=0.3-2.3).8 A study on modifiable determinants of macrosomia (defined as 24.2 kg in this study) among 553 Norwegian women found that LTPA during pregnancy was unrelated to birth weight, but that low LTPA pre-pregnancy (<1 hr/wk) increased risk of macrosomia (aOR=2.9, 95%CI: 1.2-7.3).ll An imprecise definition of LTPA participation which lacked information on intensity may have contributed to an inability to find more significant results for LTPA during pregnancy in this study. A recent prospective cohort study among Danish women found no association between participation in Sports/LTPA during the second or third trimesters of pregnancy and macrosomia (defined as 24.5 kg).'29 It appears that only ~3% of women in the Danish study gave birth to macrosomic infants, as compared to the higher prevalence of LGA/macrosomia reported in the Coloradan study (12%) and in the Norwegian study (15%).8’11’]29 Furthermore, the Danish study classified LTPA by categories of hours/week spent in sport or by categories of sport type (i.e. weight bearing vs. non- weight bearing). Thus, discrepancies in results may be due to underlying differences in the birth weight distributions and/or to methodological differences in defining LTPA during pregnancy. 22 In sum, preliminary evidence from the few studies to examine LTPA during pregnancy in relation to extreme high birth weight suggest either a null relationship or a protective effect that helps to modulate birth size within the normal range. Providing some additional support, Perkins et al.7 found that total physical activity averaged over the second and third trimesters was strongly inversely correlated with fetal growth ratio, a measure of birth Size adjusted for potential confounders (r = -O.42, p<0.01), particularly among taller women. While only 51 women were involved in the Perkins et a1. study, total physical activity during pregnancy was objectively measured using accelerometry, thereby freeing their results from responder and/or recall bias.7 These results await confirmation from a larger, more diverse study with greater ability to adjust for confounding variables. Compared to LTPA, job-related physical activity during pregnancy appears to have lesser effects on birth size.130 More time spent working and in shift work contribute towards small decreases in birth weight and borderline significant increased risk of giving birth to a SGA infant.'30"33 J ob-related physical activity may result in different physiological responses than those that occur with rhythmic aerobic LTPA such as walking for exercise. Thus, it is not surprising that shifi work and long periods of standing at work have been associated with detrimental pregnancy outcomes, while the majority of evidence for LTPA during pregnancy points towards beneficial effects.88’ '30' '33 It is beyond the scope of this dissertation to fully review the literature regarding job- related physical activity and pregnancy outcomes; however, the amount of time spent working and basic job characteristics are important to assess as covariates when examining affects of pregnancy LTPA on birth size. 23 Mgr/rant products of metal grouth : promote additior cardiac growing content umbilic require Pregnar placent toward: P13111131 deliver: her We; knowle dunng] LTPA ( Mechanism of Affect of Physical Activity on Birth Size To enable fetal growth, the mother must supply precursors for placental hormone products as well as nutrients/substrates.I34 In a normal pregnancy, increased availability of metabolic substrate occurs mainly in the second and third trimesters, when fetal growth accelerates. As pregnancy advances, maternal insulin resistance and lipolysis are promoted while blood glucose levels are increased to ensure feto-placental growth.134 In addition, women experience a blood volume expansion early in pregnancy to increase cardiac output and perfuse the placenta.'35 The placenta in turn transports nutrients to the growing fetus through simple or facilitated diffusion, the promotion of transplacental concentration gradients, and the production of alternate substrates delivered via the umbilical circulation.I34 Both pregnancy in and of itself, as well as maternal exercise, require increased blood flow, substrate delivery, and waste disposal. The response of a pregnant woman to exercise must therefore balance both maternal and fetal needs. Physical activity during pregnancy is thought to affect fetal growth by reducing placental blood flow and nutrient delivery to the fetus during exercise as blood is diverted towards the working muscle.I36 Nonetheless, regular participation in LTPA throughout pregnancy results in an overall increase in blood volume thereby enhancing nutrient delivery to the fetus at rest. ”5' '36 A woman with a history of regular LTPA may also start her pregnancy with above average blood volume, promoting fetal nutrient delivery. Thus, knowledge of pre-pregnancy LTPA habits is critical for interpreting the effects of LTPA during pregnancy on the developing fetus. To complicate matters further, the effect of LTPA on fetal growth may be trimester-specific. Maintaining an exercise program late in pregnancy appears to reduce fetal growth, while reducing exercise volume in late 24 gestation enhances fetal growth, especially fat mass. '20 A review article on the effects of exercise on feto-placental growth concluded that influences are time—specific and dependent on the frequency, duration, and intensity of exercise; however, it appears that regular moderate to vigorous exercise throughout pregnancy is associated with having a lighter, leaner infant.136 The maintenance of exercise during pregnancy also improves insulin sensitivity at rest, thereby presenting the placenta with normal, rather than elevated, blood glucose levels.88 In conclusion, LTPA before and during pregnancy may decrease birth size moderately and reduce the risk of having a large infant without increasing the risk of delivering a small infant; however, these results await confirmation from larger and better designed studies.7’ 8’ '37 In particular, previous studies of LTPA during pregnancy have utilized primarily homogeneous samples of healthy Caucasian women. Additionally, the majority of research on this topic is observational and the potential for self-selection biases cannot be ignored. Finally, much of the past literature has not measured LTPA with sufficient detail and precision to calculate energy expenditure and/or examine differing health effects associated with specific modes of activity. MATERNAL SIZE, MATERNAL PHYSICAL ACTIVITY AND INFANT SIZE AT BIRTH Recently a study of 223 Swedish women (response rate 19%) demonstrated that pro-pregnancy LTPA is significantly related to gestational weight gain. '38 Among this cohort, low pre-pregnancy LTPA was associated with excess gestational weight gain; however, LTPA was not associated with pre-pregnancy BMI or birth weight. Another study of 467 Norwegian women (85% response rate) found that a significantly lower 25 prop gain less 1 111316 bnth gesta detel expet sensit mater when mater macrc intern. )ttto SUI} OUICOr unders depenc tralUma eXi‘ess proportion of women who exercised at least four times per week had a gestational weight gain of >16 kg (35.2 lbs) in the third trimester when compared to women who exercised less frequently.139 Neither of these studies investigated potential interactive effects of maternal size, weight gain, and LTPA participation on birth Size. It is reasonable to expect interactive effects between maternal size and LTPA on birth size. Research indicates that mothers with high pre—pregnancy BMI and/or high gestational weight gain deliver excess amounts of nutrients and insulin to their developing fetus, leading to higher birth weight.79 Mechanistically, LTPA would be expected to promote glucose utilization within the working muscle and enhance insulin sensitivity, thereby reducing the mother’s tendency to “overfeed” her fetus. Thus, maternal size may influence offspring birth size to a lesser extent among active women when compared to low-active/sedentary women. In fact, a recent study has shown that, maternal BMI and weight gain are no longer significantly related to odds of delivering a macrosomic infant once pre-pregnancy LTPA is included in the model. H Possible interactive affects of LTPA with maternal size on birth weight on a continuous scale have yet to be investigated. SUMMARY Appropriate fetal growth is critical to ensuring healthy pregnancy and birth outcomes, yet predisposing factors related to giving birth to a large infant are poorly understood. The incidence of LGA births varies from 6 — 12% of US. pregnancies, depending on maternal race/ethnicity, and these infants are at increased risk of birth 1.4.37 trauma and future overweight status. Thus, determining modifiable risk factors for excess fetal growth leading to macrosomia/LGA has the potential to improve pregnancy 26 outcomes as well as elucidate possible intervention efforts for the prevention of childhood obesity. The most consistently reported risk factors for macrosomia/LGA include gestational diabetes, maternal pre-pregnancy overweight status, and excess gestational weight gain.”’ 6" Leisure-time physical activity during pregnancy has also been shown to help control fetal overgrowth; however, these results are based on relatively small, select samples of women and await confirmation from larger, more diverse studies.7' 8 The mechanism of effect for LTPA to control fetal overgrth is unclear; however, there is some evidence that more active women have lower gestational weight gains. ”8’ ’39 It also seems reasonable that LTPA would promote glucose utilization within the working muscle and enhance insulin sensitivity, thereby reducing the diabetic and/or overweight mother’s tendency to “overfeed” her fetus. Recent results from 553 Norwegian pregnancies provide partial support for this hypothesis by demonstrating that maternal BMI is no longer related to odds of delivering a macrosomic infant once pre-pregnancy LTPA is also considered.ll Inter-relationships among maternal size, gestational weight gain, LTPA participation during pregnancy, and birth size requires more attention. Preliminary results suggest that LTPA during pregnancy reduces overgrth without shifting the entire birth weight distribution downward.8 If these results are confirmed, future research may seek to examine the efficacy of LTPA intervention programs during pregnancy to reduce the risk of giving birth to a macrosomic/LGA infant without fear of increasing the risk for low birth weight/SGA. 27 REFI :Jl REFERENCES 10. ll. Hediger ML, Overpeck MD, McGlynn A, Kuczmarski RJ, Maurer KR, Davis WW. Growth and fatness at three to six years of age of children born small- or large-for-gestational age. Pediatrics. Sep 1999;104(3):e33. Boney CM, Verrna A, Tucker R, Vohr BR. Metabolic syndrome in childhood: association with birth weight, maternal obesity, and gestational diabetes mellitus. Pediatrics. Mar 2005;] 15(3):e290-296. Rugholm S, Baker J L, Olsen LW, Schack-Nielsen L, Bua J, Sorensen TI. Stability of the association between birth weight and childhood overweight during the development of the obesity epidemic. Obes Res. Dec 2005;13(12):2187-2 194. Haram K, Pirhonen J, Bergsjo P. Suspected big baby: a difficult clinical problem in obstetrics. Acta Obstet Gynecol Scand. Mar 2002;81(3):185-194. Catalano PM, Drago NM, Amini SB. Factors affecting fetal growth and body composition. Am J Obstet Gynecol. May l995;172(5):1459-1463. Frederick 10, Williams MA, Sales AE, Martin DP, Killien M. Pre-pregnancy Body Mass Index, Gestational Weight Gain, and Other Maternal Characteristics in Relation to Infant Birth Weight. Matern Child Health J. Sep 2008;12(5):557- 567. Perkins CC, Pivarnik J M, Paneth N, Stein AD. Physical activity and fetal growth during pregnancy. Obstet Gynecol. Jan 2007;109(1):81-87. Alderman BW, Zhao H, Holt VL, Watts DH, Beresford SA. Maternal physical activity in pregnancy and infant size for gestational age. Ann Epidemic]. Nov 1998;8(8):513-519. Clapp J F . The morphometric and neurodevelopmental outcome at five years of the offspring of women who continued exercise throughout pregnancy. J Pediatr. Dec 1996;129(6):856-863. National Center for Health Statistics, Vital Statistics of the United States: Natality, 1990. Vol 1. Hyatssville, MD: US Dept of Health and Human Services; 1994. Voldner N, Froslie KF, Bo K, et al. Modifiable determinants of fetal macrosomia: role of lifestyle-related factors. Acta Obstet Gynecol Scand. Apr 2008;87(4):423- 429. 28 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Schwartz R, Teramo KA. What is the significance of macrosomia? Diabetes Care. Jul 1999;22(7):1201-1205. Henriksen T. The macrosomic fetus: a challenge in current obstetrics. Acta Obstet Gynecol Scand. Jan 2008;87(2): 134-145. Dunn PM. The search for perinatal definitions and standards. Acta Paediatr Scand Suppl. 1985;319:7-16. High-Risk Newborn: Large for Gestational Age. Online Source. http://www.healthsystem.virginia.edu/uvahealth/peds_hrnewborn/lga.cfin. Accessed April 2, 2006. Lawrence EJ. A Matter of Size: Part 2 Evaluating the Large-for-Gestational-Age Neonate. Adv Neonatal Care. 2007;7(4):187-197. Arbuckle TE, Wilkins R, Sherman GJ. Birth weight percentiles by gestational age in Canada. Obstet Gynecol. Jan 1993;81(l):39-48. Orskou J, Kesmodel U, Henriksen TB, Secher NJ. An increasing proportion of infants weigh more than 4000 grams at birth. Acta Obstet Gynecol Scand. Oct 2001;80(10):931-936. Chauhan SP, Grobman WA, Gherrnan RA, et al. Suspicion and treatment of the macrosomic fetus: a review. Am J Obstet Gynecol. Aug 2005;193(2):332-346. Alexander GR, Kogan MD, Himes JH. 1994-1996 US. Singleton birth weight percentiles for gestational age by race, Hispanic origin, and gender. Matern Child HealthJ. Dec 1999;3(4):225-231. Bergmann RL, Richter R, Bergmann KE, Plagemann A, Brauer M, Dudenhausen JW. Secular trends in neonatal macrosomia in Berlin: influences of potential determinants. Paediatr Perinat Epidemiol. Jul 2003 ; l 7(3):244-249. Kramer MS, Morin I, Yang H, et al. Why are babies getting bigger? Temporal trends in fetal grth and its determinants. J Pediatr. Oct 2002;141(4):538-542. Bonellie SR, Raab GM. Why are babies getting heavier? Comparison of Scottish births from 1980 to 1992. ij. Nov 8 1997;315(7117):1205. Surkan PJ, Hsieh CC, Johansson AL, Dickman PW, Cnattingius S. Reasons for increasing trends in large for gestational age births. Obstet Gynecol. Oct 2004; 104(4):720-726. Wen SW, Kramer MS, Platt R, et al. Secular trends of fetal grth in Canada, 1981 to 1997. Paediatr Perinat Epidemiol. Oct 2003;17(4):347-354. 29 18. 29. 30. 31. 32. 33. 34. (.44 (II 36. 37. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. Ananth CV, Wen SW. Trends in fetal growth among singleton gestations in the United States and Canada, 1985 through 1998. Semin Perinatal. Aug 2002;26(4):260-267. Lipscomb KR, Gregory K, Shaw K. The outcome of macrosomic infants weighing at least 4500 grams: Los Angeles County + University of Southern California experience. Obstet Gynecol. Apr 1995;85(4):558-564. Stotland NE, Caughey AB, Breed EM, Escobar GJ. Risk factors and obstetric complications associated with macrosomia. Int J Gynaecol Obstet. Dec 2004;87(3):220-226. F erber A. Maternal complications of fetal macrosomia. Clin Obstet Gynecol. Jun 2000;43(2):335-339. Boulet SL, Alexander GR, Salihu HM, Pass M. Macrosomic births in the united states: determinants, outcomes, and proposed grades of risk. Am J Obstet Gynecol. May 2003;188(5):1372-1378. Casey BM, Schaffer JI, Bloom SL, Heartwell SF, McIntire DD, Leveno KJ. Obstetric antecedents for postpartum pelvic floor dysfunction. Am J Obstet Gynecol. May 2005;192(5):1655-1662. Dietz HP, Wilson PD. Childbirth and pelvic floor trauma. Best Pract Res C [in Obstet Gynaecol. Dec 2005;19(6):913-924. Alsunnari S, Berger H, Sermer M, Seaward G, Kelly B, Farine D. Obstetric outcome of extreme macrosomia. J Obstet Gynaecol Can. Apr 2005 ;27(4):323- 328. Boulet SL, Salihu HM, Alexander GR. Mode of delivery and birth outcomes of macrosomic infants. J Obstet Gynaecol. Sep 2004;24(6):622-629. Chatfield J. ACOG issues guidelines on fetal macrosomia. American College of Obstetricians and Gynecologists. Am Fam Physician. Jul 1 2001;64(1):169-170. Lapunzina P, Camelo J S, Rittler M, Castilla EE. Risks of congenital anomalies in large for gestational age infants. J Pediatr. Feb 2002;140(2):200-204. Alexander GR, Kogan M, Bader D, Carlo W, Allen M, Mor J. US birth weight/gestational age-specific neonatal mortality: 1995-1997 rates for whites, hispanics, and blacks. Pediatrics. Jan 2003;1 1 1(1):e61-66. 30 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. Catalano PM, Thomas A, Huston-Presley L, Amini SB. Increased fetal adiposity: a very sensitive marker of abnormal in utero development. Am J Obstet Gynecol. Dec 2003; 189(6): 1698-1 704. Hammami M, Walters JC, Hockman EM, Koo WW. Disproportionate alterations in body composition of large for gestational age neonates. J Pediatr. Jun 2001;138(6):817-821. Baird J, Fisher D, Lucas P, Kleijnen J, Roberts H, Law C. Being big or growing fast: systematic review of size and growth in infancy and later obesity. ij. Oct 22 2005;331(7522):929. Oken E, Gillman MW. Fetal origins of obesity. Obes Res. Apr 2003;11(4):496- 506. Harder T, Rodekamp E, Schellong K, Dudenhausen J W, Plagemann A. Birth weight and subsequent risk of type 2 diabetes: a meta-analysis. Am J Epidemiol. Apr 2007; l 65(8):849-85 7. Schuz J, F orman MR. Birthweight by gestational age and childhood cancer. Cancer Causes Control. Aug 2007;18(6):655-663. Hemachandra AH, Howards PP, Furth SL, Klebanoff MA. Birth weight, postnatal growth, and risk for high blood pressure at 7 years of age: results from the Collaborative Perinatal Project. Pediatrics. Jun 2007;] 19(6):e1264-1270. Harder T, Plagemann A, Harder A. Birth weight and subsequent risk of childhood primary brain tumors: a meta-analysis. Am J Epidemiol. Aug 2008;168(4):366- 373. Yajnik CS. The lifecycle effects of nutrition and body size on adult adiposity, diabetes and cardiovascular disease. Obes Rev. Aug 2002;3(3):217-224. J ohannsson E, Amgrimsson SA, Thorsdottir I, Sveinsson T. Tracking of overweight from early childhood to adolescence in cohorts born 1988 and 1994: overweight in a high birth weight population. Int J Obes (Lond). Aug 2006;30(8):1265-l271. Seidman DS, Ever-Hadani P, Stevenson DK, Slater PE, Harlap S, Gale R. Birth order and birth weight reexamined. Obstet Gynecol. Aug 1988;72(2): 158-162. Dougherty CR, Jones AD. The determinants of birth weight. Am J Obstet Gynecol. Sep 1982; 144(2): 190-200. Langer 0. Fetal macrosomia: etiologic factors. Clin Obstet Gynecol. Jun 2000;43(2):283-297. 31 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. Orskou J, Henriksen TB, Kesmodel U, Secher NJ. Maternal characteristics and lifestyle factors and the risk of delivering high birth weight infants. Obstet Gynecol. Jul 2003;102(l):115-120. Kramer MS, Olivier M, McLean FH, Dougherty GE, Willis DM, Usher RH. Determinants of fetal growth and body proportionality. Pediatrics. Jul l990;86(l):l8-26. David RJ, Collins JW, Jr. Differing birth weight among infants of U.S.-bom blacks, Afiican-bom blacks, and U.S.-born whites. N Engl J Med. Oct 1997;337(l7):1209-1214. Johnston LB, Clark AJ, Savage MO. Genetic factors contributing to birth weight. Arch Dis Child Fetal Neonatal Ed. Jan 2002;86(1):F2-3. Dunger DB, Petry CJ, Ong KK. Genetics of size at birth. Diabetes Care. Jul 2007;30 Suppl 2:8150-155. Dumwald C, Huston-Presley L, Amini S, Catalano P. Evaluation of body composition of large-for-gestational-age infants of women with gestational diabetes mellitus compared with women with normal glucose tolerance levels. Am J Obstet Gynecol. Sep 2004;191(3):804-808. Kieffer EC, Nolan GH, Carman WJ, Sanbom CZ, Guzman R, Ventura A. Glucose tolerance during pregnancy and birth weight in a Hispanic population. Obstet Gynecol. Nov 1999;94(5 Pt l):74l-746. Stein Z, Susser M. The Dutch famine, 1944-1945, and the reproductive process. 11. Interrelations of caloric rations and six indices at birth. Pediatr Res. Feb 1975;9(2):76-83. Stein Z, Susser M. The Dutch famine, 1944-1945, and the reproductive process. I. Effects or six indices at birth. Pediatr Res. Feb 1975;9(2):70-76. Stephenson T, Symonds ME. Maternal nutrition as a determinant of birth weight. Arch Dis Child Fetal Neonatal Ed. Jan 2002;86(1):F4-6. Spencer N, Logan 8. Social influences on birth weight. Arch Dis Child Fetal Neonatal Ed. Jan 2002;86(1):F6-7. Dubois L, Girard M. Determinants of birthweight inequalities: population-based study. Pediatr Int. Oct 2006;48(5):470-478. 32 63. 65. 66. 67. 68. 69. '11. 73. 74. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. Ehrenberg HM, Mercer BM, Catalano PM. The influence of obesity and diabetes on the prevalence of macrosomia. Am J Obstet Gynecol. Sep 2004;191(3):964- 968. Brunskill AJ, Rossing MA, Connell FA, Daling J. Antecedents of macrosomia. Paediatr Perinat Epidemiol. Oct 1991;5(4):392-401. Schaefer-Graf UM, Kjos SL, Kilavuz O, et al. Determinants of fetal growth at different periods of pregnancies complicated by gestational diabetes mellitus or impaired glucose tolerance. Diabetes Care. Jan 2003;26(1): 193-198. Caulfield LE, Harris SB, Whalen EA, Sugamori ME. Maternal nutritional status, diabetes and risk of macrosomia among Native Canadian women. Early Hum Dev. Feb l998;50(3):293-303. Gillman MW, Rifas-Shiman S, Berkey CS, Field AE, Colditz GA. Maternal gestational diabetes, birth weight, and adolescent obesity. Pediatrics. Mar 2003;] 1 1(3):e221-226. Plagemann A, Harder T, Kohlhoff R, Rohde W, Domer G. Overweight and obesity in infants of mothers with long-term insulin-dependent diabetes or gestational diabetes. Int J Obes Relat Metab Disord. Jun 1997;21(6):451-456. Engelgau MM, Herman WH, Smith PJ, RR. G, RE. A. The epidemiology of diabetes and pregnancy in the US, 1988. Diabetes Care. Jul l995;18(7): 1029 - 1033. National Center for Health Statistics. Chartbook on Trends in the Health of Americans. Hyattsville, MD; 2006. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, F legal KM. Prevalence of overweight and obesity in the United States, 1999-2004. Jama. Apr 2006;295(13):1549-1555. Chu SY, Kim SY, Bish CL. Prepregnancy Obesity Prevalence in the United States, 2004-2005. Matern Child Health J. Sep 2008;13(5):614-620. Institute of Medicine. Nutrition during pregnancy, weight gain and nutrient supplements. Report of the Subcommittee on Nutritional Status and Weight Gain during Pregnancy, Subcommittee on Dietary Intake and Nutrient Supplements during Pregnancy, Committee on Nutritional Status during Pregnancy and Lactation, Food and Nutrition Board. Washington, DC: National Academy Press; 1990:1-233. Olson CM. Achieving a healthy weight gain during pregnancy. Annu Rev Nutr. Aug 2008;28:411-423. 33 15. 76. '11. 78. '19. 80. 81. 83. 84. 86 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. Kiel DW, Dodson EA, Artal R, Boehmer TK, Leet TL. Gestational weight gain and pregnancy outcomes in obese women: how much is enough? .Obstet Gynecol. Oct 2007;110(4):752-758. Chasan-Taber L, Schmidt MD, Pekow P, Stemfeld B, Solomon CG, Markenson G. Predictors of excessive and inadequate gestational weight gain in Hispanic women. Obesity (Silver Spring). Jul 2008;16(7):l657-1666. Stotland NE, Cheng YW, Hopkins LM, Caughey AB. Gestational weight gain and adverse neonatal outcome among term infants. Obstet Gynecol. Sep 2006;108(3 Pt 1):635-643. Wells CS, Schwalberg R, Noonan G, Gabor V. Factors influencing inadequate and excessive weight gain in pregnancy: Colorado, 2000-2002. Matern Child Health J. Jan 2006;10(1):55-62. Sewell MF, Huston-Presley L, Super DM, Catalano P. Increased neonatal fat mass, not lean body mass, is associated with maternal obesity. Am J Obstet Gynecol. Oct 2006; 195(4):l 100-1 103. Hull HR, Dinger MK, Knehans AW, Thompson DM, Fields DA. Impact of maternal body mass index on neonate birthweight and body composition. Am J Obstet Gynecol. Apr 2008;198(4): e411-416. Neufeld LM, Haas JD, Grajeda R, Martorell R. Changes in maternal weight from the first to second trimester of pregnancy are associated with fetal growth and infant length at birth. Am J Clin Nutr. Apr 2004;79(4):646-652. Abrams B, Selvin S. Maternal weight gain pattern and birth weight. Obstet Gynecol. Aug l995;86(2): 163-169. Shapiro C, Sutija VG, Bush J. Effect of maternal weight gain on infant birth weight. J Perinat Med. Nov 2000;28(6):428-431. Thame M, Osmond C, Bennett F, Wilks R, Forrester T. F eta] growth is directly related to maternal anthropometry and placental volume. Eur J Clin Nutr. Jun 2004;58(6):894-900. Hay W, Jr. The placenta. Not just a conduit for maternal fuels. Diabetes. Dec 1991;40 Supp12z44-50. Clausen T, Burski TK, Oyen N, Godang K, Bollerslev J, Henriksen T. Maternal anthropometric and metabolic factors in the first half of pregnancy and risk of neonatal macrosomia in term pregnancies. A prospective study. Eur J Endocrinol. Dec 2005;153(6):887-894. 34 87. 88. 94. 96. 97. 98. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. ACOG. Technical Bulletin: Exercise during pregnancy and the postnatal period. Washington DC: American College of Obstetricians and Gynecologists; 1985. Impact of physical activity during pregnancy and postpartum on chronic disease risk. Med Sci Sports Exerc. May 2006;38(5):989-1006. Hegaard HK, Pedersen BK, Nielsen BB, Damm P. Leisure time physical activity during pregnancy and impact on gestational diabetes mellitus, pre-eclampsia, preterm delivery and birth weight: a review. Acta Obstet Gynecol Scand. Nov 2007;86( 1 1): 1290- 1296. Weissgerber TL, Wolfe LA, Davies GA, Mottola MF. Exercise in the prevention and treatment of maternal-fetal disease: A review of the literature. Appl Physiol Nutr Metab. Dec 2006;3 1(6):661-674. ACOG committee opinion. Exercise during pregnancy and the postpartum period. Number 267. American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet. Apr 2002;77(1):79-81. U.S. Physical Activity Guidelines for Americans. US. Department of Health and Human Services. http://www.health.gov/paguidelines: ODPHP Publication No. U0036; 2008. Evenson KR, Savitz DA, Huston SL. Leisure-time physical activity among pregnant women in the US. Paediatr Perinat Epidemiol. Nov 2004;18(6):400- 407. Zhang J, Savitz DA. Exercise during pregnancy among US women. Ann Epidemiol. Jan 1996;6(1):53-59. Evenson KR, Siega-Riz AM, Savitz DA, Leiferman JA, Thorp JM, Jr. Vigorous leisure activity and pregnancy outcome. Epidemiology. Nov 2002;13(6):653-659. Fell DB, Joseph KS, Armson BA, Dodds L. The Impact of Pregnancy on Physical Activity Level. Matern Child Health J. Sep 2009;13(5):597-603. Mottola MF, Campbell MK. Activity patterns during pregnancy. Can J Appl Physiol. Aug 2003;28(4):642-653. Schmidt MD, Pekow P, Freedson PS, Markenson G, Chasan-Taber L. Physical activity patterns during pregnancy in a diverse population of women. J Womens Health (Larchmt). Oct 2006;15(8):909-918. 35 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. Rousham EK, Clarke PE, Gross H. Significant changes in physical activity among pregnant women in the UK as assessed by accelerometry and self-reported activity. Eur J Clin Nutr. 2006;60:393-400. Pereira MA, Rifas-Shiman SL, Kleinman KP, Rich-Edwards JW, Peterson KE, Gillman MW. Predictors of change in physical activity during and after pregnancy: Project Viva. Am J Prev Med. Apr 2007;32(4):312-319. Petersen AM, Leet TL, Brownson RC. Correlates of physical activity among pregnant women in the United States. Med Sci Sports Exerc. Oct 2005;37(10):1748-1753. Hinton PS, Olson CM. Predictors of pregnancy-associated change in physical activity in a rural white population. Matern Child Health J. Mar 2001;5(1):7-14. Symons Downs D, Ulbrecht J S. Understanding exercise beliefs and behaviors in women with gestational diabetes mellitus. Diabetes Care. Feb 2006;29(2):236- 240. Symons Downs D, Hausenblas HA. Women's exercise beliefs and behaviors during their pregnancy and postpartum. J Midwifery Womens Health. Mar-Apr 2004;49(2): 138-144. Evenson KR, Moos MK, Carrier K, Siega-Riz AM. Perceived barriers to physical activity among pregnant women. Matern Child Health J. May 2009;13(3):364- 375. Albright CL, Maddock J E, Nigg CR. Physical activity before pregnancy and following childbirth in a multiethnic sample of healthy women in Hawaii. Women & Health. 2005;42(3):95-110. Duncombe D, Wertheim EH, Skouteris H, Paxton SJ, Kelly L. Factors related to exercise over the course of pregnancy including women's beliefs about the safety of exercise during pregnancy. Midwifery. Aug 2009;25(4):430-438 Doran F, O'Brien AP. A brief report of attitudes towards physical activity during pregnancy. Health Promot J Aust. Aug 2007;18(2):155-158. Clarke PE, Gross H. Women's behaviour, beliefs and information soucres about physical exercise in pregnancy. Midwifery. Jun 2004(2);20: 133-141. Damm P, Breitowicz B, Hegaard H. Exercise, pregnancy, and insulin sensitivity-- what is new? App] Physio] Nutr Metab. Jun 2007;32(3):537-540. 36 111. 112. 113. 114. 115. 116. 118. 119. 120. 123. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. Dempsey JC, Butler CL, Williams MA. No need for a pregnant pause: physical activity may reduce the occurrence of gestational diabetes mellitus and preeclampsia. Exerc Sport Sci Rev. Jul 2005;33(3): 141-149. Morris SN, Johnson NR. Exercise during pregnancy: a critical appraisal of the literature. J Reprod Med. Mar 2005;50(3): 181-188. Gavard J A, Artal R. Effect of exercise on pregnancy outcome. Clin Obstet Gynecol. Jun 2008;51(2):467-480. Bung P, Artal R, Khodiguian N, Kjos S. Exercise in gestational diabetes. An optional therapeutic approach? Diabetes. Dec 1991;40(Supp12):182-185. Bung P, Artal R. Gestational diabetes and exercise: a survey. Semin Perinato]. Aug 1996;20(4):328-333. Hegaard HK, Hedegaard M, Damm P, Ottesen B, Petersson K, Henriksen TB. Leisure time physical activity is associated with a reduced risk of preterm delivery. Am J Obstet Gynecol. Feb 2008;198(2): l 80 e181-185. Meher S, Duley L. Exercise or other physical activity for preventing pre- eclampsia and its complications. Cochrane Database Syst Rev. 2006(2):CD005942. Ceysens G, Rouiller D, Boulvain M. Exercise for diabetic pregnant women. Cochrane Database Syst Rev. 2006;3 :CD004225. Bell RJ, Palma SM, Lumley J M. The effect of vigorous exercise during pregnancy on birth-weight. Aust N Z J Obstet Gynaecol. Feb 1995;35(1):46-5 1. Clapp JF, 3rd, Kim H, Burciu B, Schmidt 8, Petry K, Lopez B. Continuing regular exercise during pregnancy: effect of exercise volume on fetoplacental growth. Am J Obstet Gynecol. Jan 2002;186(1): 142-147. Hatch MC, Shu XO, McLean DE, et a1. Maternal exercise during pregnancy, physical fitness, and fetal growth. Am J Epidemiol. May l993;137(10):]105- 1114. Horns PN, Ratcliffe LP, Leggett JC, Swanson MS. Pregnancy outcomes among active and sedentary primiparous women. J Obstet Gynecol Neonatal Nurs. Jan l996;25(1):49-54. Stemfeld B, Quesenberry CP, Jr., Eskenazi B, Newman LA. Exercise during pregnancy and pregnancy outcome. Med Sci Sports Exerc. May 1995;27(5):634- 640. 37 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. Duncombe D, Skouteris H, Wertheim EH, Kelly L, Fraser V, Paxton SJ. Vigorous exercise and birth outcomes in a sample of recreational exercisers: a prospective study across pregnancy. Aust N Z J Obstet Gynaecol. Aug 2006;46(4):288-292. Campbell MK, Mottola MF. Recreational exercise and occupational activity during pregnancy and birth weight: a case-control study. Am J Obstet Gynecol. Feb 2001;184(3):403-408. Pivamik JM. Potential effects of maternal physical activity on birth weight: brief review. Med Sci Sports Exerc. Mar 1998;30(3):400-406. Khanna N. Effects of exercise on pregnancy. Am F am Physician. Apr 15 1998;57(8):1764, 1770-1762. Leiferman J A, Evenson KR. The effect of regular leisure physical activity on birth outcomes. Matern Child Health J. Mar 2003;7(1):59-64. Hegaard HK, Petersson K, Hedegaard M, et al. Sports and leisure-time physical activity in pregnancy and birth weight: a population-based study. Scand J Med Sci Sports. Apr 2009: epub DOI: 10.1111/j.1600-0838.2009.00918.x. Magann EF, Evans SF, Newnham JP. Employment, exertion, and pregnancy outcome: assessment by kilocalories expended each day. Am J Obstet Gynecol. Jul 1996;175(1):l82-187. Croteau A, Marcoux S, Brisson C. Work activity in pregnancy, preventive measures, and the risk of delivering a small-for-gestational-age infant. Am J Public Health. May 2006;96(5):846-855. Hatch M, J i BT, Shu XO, Susser M. Do standing, lifting, climbing, or long hours of work during pregnancy have an effect on fetal growth? Epidemiology. Sep 1997;8(5):530-536. Bonzini M, Coggon D, Palmer KT. Risk of prematurity, low birthweight and pre- eclampsia in relation to working hours and physical activities: a systematic review. Occup Environ Med. Apr 2007;64(4):228-243. Weissgerber TL, Wolfe LA. Physiological adaptation in early human pregnancy: adaptation to balance matemal-fetal demands. App] Physio] Nutr Metab. Feb 2006;31(1):1-11. Pivamik JM, Mauer MB, Ayres NA, Kirshon B, Dildy GA, Cotton DB. Effects of chronic exercise on blood volume expansion and hematologic indices during pregnancy. Obstet Gynecol. Feb l994;83(2):265-269. 38 136. 137. 138. 139. Clapp JF, 3rd. The effects of maternal exercise on fetal oxygenation and feto- placental growth. Eur J Obstet Gynecol Reprod Biol. Sep 2003;110(Suppl 1):S80— 85. Kramer MS, McDonald SW. Aerobic exercise for women during pregnancy. Cochrane Database Syst Rev. 2006;3:CD000180. Lof M, Hilakivi-Clarke L, Sandin S, Weiderpass E. Effects of pre-pregnancy physical activity and maternal BMI on gestational weight gain and birth weight. Acta Obstet Gynecol Scand. May 2008;87(5):524-530. Haakstad LA, Voldner N, Henriksen T, Bo K. Physical activity level and weight gain in a cohort of pregnant Norwegian women. Acta Obstet Gynecol Scand. May 2007;86(5):559-564. 39 CHAPTER 3 MATERNAL PHYSICAL ACTIVITY DURING PREGNANCY AND THE BIRTH WEIGHT DISTRIBUTION: WHERE IS THE EFFECT? ABSTRACT This study sought to characterize the relation between leisure-time physical activity (LTPA) during pregnancy and birth weight. Women enrolled in the Pregnancy Outcomes and Community Health Study (1998-2004) were followed-up in 2007. F ollow- up efforts were extensive for a subcohort and minimal for the remainder, or non- subcohort. As a result, 596 (67% of eligible) subcohort and 418 (26% of eligible) non- subcohort women who delivered at term were included in this investigation. Original data collection provided maternal demographics. Medical records provided child gender, gestational age, and birth weight. Gender and gestational age-specific birth weight 2- scores (BWz) were calculated and offspring were categorized as small-, appropriate-, or large-for-gestational-age (SGA, AGA, and LGA, respectively). At follow-up, women recalled pregnancy LTPA and were classified as inactive, insufficiently active (< 7.5 kcal/kg/wk), and meeting LTPA recommendations (2 7.5 kcal/kg/wk). Analyses were conducted separately by subcohort status. The subcohort had greater racial and socioeconomic diversity than the non-subcohort. Among the non-subcohort, meeting LTPA recommendations Significantly decreased odds of LGA (aOR=0.30, 95%CI: 0.14- 0.64) without affecting odds of SGA. In quantile regression analyses, meeting LTPA recommendations was unrelated to BWz until the 0.65 quantile, after which it significantly reduced BWz. Results for the subcohort were similar but non-significant. 4o LPTA during pregnancy may represent a significant health benefit by reducing BWz only among the upper parts of the distribution. 41 INTRODUCTION The American College of Obstetricians and Gynecologists (ACOG) recommends that all pregnant women participate in at least 30 minutes of moderate leisure-time physical activity (LTPA) on most days of the week in the absence of obstetrical complications.l Recent governmental LTPA recommendations also endorse at least 150 min/wk of at least moderate LTPA for pregnant women.2 Population-based data indicate that the majority of pregnant women (~66%) choose to engage in some amount of LTPA, but only 16% participate at recommended levels.3 While LTPA during pregnancy has been associated with several maternal health benefits including decreased risk for gestational diabetes, pre-eclampsia, and preterm delivery, effects on birth weight are still debated.4‘9 Research has shown repeatedly that LTPA during pregnancy does not increase risk of delivering a low birth weight infant.5’ 7’ '0 However, results with mean birth weight have been more varied, with some reporting no changes, others finding significant reductions in mean birth weight, and one study showing increased mean birth weight associated with vigorous exercise during pregnancy.7’ ”.13 Very few studies have considered the effect of LTPA during pregnancy on high birth weight. One small study found that participation in LTPA later in pregnancy (2nd or 3rd trimester) significantly reduced odds of giving birth to a large-for-gestational-age (LGA, 290th percentile) infant without affecting risk of delivering small-for-gestational- age (SGA, <10‘h percentile).l4 However, results from a recent cohort study showed no associations among sports/LTPA participation and odds of giving birth to a high birth weight infant (24.5 kg).15 Giving birth to a large infant is associated with several adverse 42 health outcomes including prolonged and complicated deliveries for the mother, as well as increased risk of birth trauma and future childhood obesity for the offspring.'6’ ’7 Thus, if LTPA during pregnancy could decrease the risk of giving birth to a large infant without shifting the entire birth weight distribution downwards, this could represent a Significant health benefit. The purpose of this study was to more fully characterize the effect of LTPA during pregnancy on offspring size at birth among a diverse group of women who delivered at term (2 37 weeks). Specifically, we sought to determine whether LTPA during pregnancy was associated with a mean shift in the entire birth weight distribution, whether it was associated with giving birth to an SGA or LGA infant, and whether there was an isolated effect of LTPA during pregnancy on the high end of the birth weight distribution. METHODS Study Population This study followed-up women enrolled in the Pregnancy Outcomes and Community Health (POUCH) Study. Women were originally recruited in gestational weeks 15-27 from l998-2004 from 52 clinics in five Michigan communities. Inclusion criteria were singleton pregnancy with no known congenital anomaly, maternal age 215 years, maternal serum alpha-fetoprotein screen in gestational weeks 15-22, no pre- existing diagnoses of diabetes mellitus, and proficiency in English. Women with unexplained high alpha-fetoprotein levels (2 2 multiples of the median) were oversampled due to a particular interest in this biomarker for the original POUCH study aims.18 Of the 3,038 women enrolled, 19 were lost to follow-up at birth leaving a cohort 43 of 3,019 mother-child pairs. The POUCH study was approved from institutional review boards at Michigan State University, Michigan Department of Community Health, and nine community hospitals. A subcohort of women was selected for more detailed study to maximize resources when evaluating original study aims. The subcohort included all women who delivered preterm (<37 weeks), women who delivered at term but had unexplained high maternal serum alpha-fetoprotein levels, and a race-stratified sample of women with term deliveries and normal maternal serum alpha-fetoprotein levels (i.e., 72% African American and 23% White/other women in this category). Women enrolled in POUCH but not meeting these criteria comprised the “non-subcohort”. Women in the subcohort were contacted periodically for different follow-up studies (2005-2006; 2006-2007), while the non-subcohort received minimal contact. In fall of 2007, follow-up surveys on LTPA during pregnancy and child health outcomes were sent to all POUCH participants who had not declined further contact after delivery and whose children were living with them (n=1629 non-subcohort; n=1261 subcohort). Women in the non-subcohort were sent a single mailing which asked them to complete and return an enclosed survey. No further contact was attempted for the non- subcohort. Women in the subcohort were sent the same mailing; however, phone contact was attempted to encourage participation and follow-up mailings were sent. For this investigation, women who delivered preterm (< 37 weeks of gestation) were excluded (non-subcohort n=0, subcohort n=335). A total of 1200 non-subcohort and 299 subcohort women failed to return the follow-up survey, and there was incomplete follow-up information on LTPA during pregnancy (non-subcohort n=11, subcohort n=3 1) 44 Thus, the final sample for this investigation included 418 (26% of eligible) non-subcohort and 596 (67% of eligible) subcohort women. Study Protocol The POUCH study has been described elsewhere in detail.18 Briefly, participants met with a study nurse at enrollment to sign consent forms, complete an in-person interview and a self-administered survey, and have biological samples collected. The interview and self-administered survey provided information on maternal size, race/ethnicity, education, relationship status, enrollment in Medicaid, occupational level (Low=Clerical/Sales, Service/Blue Collar, or Homemaker/Other/Unknown; High = Professional/Manager/Technica1), and smoking at mid-pregnancy. Maternal pre- pregnancy body mass index (BMI, kg/mz) was calculated from self-reported pre- pregnancy weight and height values. Birth weight, child gender, maternal age at delivery, and parity were determined through chart review. Gestational age was calculated using the last menstrual period unless it disagreed by > 2 weeks with ultrasound conducted prior to 25 weeks gestation, in which case the ultrasound value was used. Sex and gestational-age specific birth weight z-scores were calculated as the observed minus the mean birth weight divided by the population standard deviation using birth weight standards from Kramer et al.19 Offspring were classified as SGA, appropriate-for-gestational-age (AGA), or LGA if their sex and gestational age-specific birth weight was 5 101‘, between the 10th and 90‘“, or 2 90th percentile, respectively.20 The follow-up survey provided information on gestational weight gain and LTPA (Appendix). Time from delivery to follow-up was used to determine length of recall. (<4, 45 4-<6, 6-9 years). Self-reported maternal weight gain during the POUCH pregnancy was used to classify women into weight gain categories (less than, within, and greater than recommended) based on the 1990 Institute of Medicine recommendations according to their pro-pregnancy BMI.2| Finally, women were asked whether they had participated in any LTPA during the POUCH pregnancy. If so, they recalled the type, average duration (min/d), and average frequency (d/wk) of up to two activities performed most often during a typical week while pregnant. To quantify intensity, metabolic equivalent (MET) values were assigned to each reported activity using the Compendium for Physical Activities.22 MET values were converted to caloric energy expenditure (1 MET = 1 kcal/kg/hr), multiplied by reported duration and frequency values, and then summed within each woman to calculate total LTPA energy expenditure (kcal/kg/wk). Using this method, meeting ACOG recommendations for LTPA during pregnancy corresponds to expending 27.5 kcal/kg/wk, or 3 METs (i.e., the lower-bound of “moderate” intensity), times 30 min/day, times 5 days/wk, divided by 60 min/hr." 22 For analyses, women were classified as inactive (0 kcal/kg/wk), insufficiently active (< 7.5 kcal/kg/wk), or meeting/exceeding LTPA recommendations (2 7.5 kcal/kg/wk) during pregnancy. Statistical Analyses Because of the sampling scheme employed to create the original POUCH study cohort and subcohort, and the differing follow-up strategies, analyses were conducted separately for non-subcohort and subcohort participants. All analyses were conducted using SAS version 9.1. Significance was set at a two-sided alpha level of P 5 0.05. Linear, polytomous logistic, and quantile regression analyses were used to assess the 46 relation between LTPA during pregnancy (reference group = inactive) and birth size (modeled as birth weight z-score or size for gestational age categories). Stepwise regression methods were used to build adjusted models. The following variables were considered as covariates based on previous literature: maternal race (among the subcohort only), pre-pregnancy BMI, maternal height, gestational weight gain categories, parity, maternal report of smoking during pregnancy, enrollment in Medicaid, relationship status, maternal age at delivery, educational level, and occupational level. Any variable that altered parameter estimates more than 10% was examined as a potential confounder or mediator. All multivariate models included length of recall for LTPA during pregnancy. Linear regression analyses (PROC GLM) were conducted first to determine whether LTPA during pregnancy was associated with a mean shift in the birth weight 2- score distribution. Adjusted R2 values were used to assess the amount of variance explained. Next, polytomous logistic regression (PROC LOGISTIC) was used to estimate associations among LTPA during pregnancy and giving birth to an SGA or LGA infant (reference category: AGA). Likelihood ratio tests were used for significance testing in building adjusted models. Finally, quantile regression (PROC QUANTREG) was used to estimate the association between LTPA during pregnancy and the entire distribution of birth weight 2- score. As an extension of median regression, quantile regression uses all available data to examine the effect of an‘independent variable on specified quantiles of the distribution of the dependent variable, not just the mean value.23 Therefore, quantile regression can determine whether LTPA during pregnancy has a constant affect across the entire birth 47 weight z-score distribution, or has a variable affect on only certain parts. Unlike logistic regression, quantile regression does not require the transformation of birth weight into somewhat arbitrary categories of SGA, AGA, and LGA. Thus, quantile regression may provide insights into the effect of LTPA during pregnancy on birth Size that are not revealed by linear or logistic regression techniques. We assessed the association between LTPA during pregnancy and birth weight z-score for every 0.05 quantile from the 0.05 to 0.95 quantile of the distribution. Histograms of the standardized residuals were evaluated to determine goodness of fit. RESULTS Within both the non-subcohort and subcohort, follow—up survey nonresponders includedsignificantly more women who were African American, younger age, had less than high school education, single, enrolled in Medicaid and smoked during pregnancy, compared to participating women (data not shown). Non-responders within the subcohort also had a higher percentage of SGA and a lower percentage of LGA births compared to participants. Among participants, maternal characteristics differed between non-subcohort and subcohort women (Table 3.1). Some differences were expected by design (e. g. 4% non- subcohort vs. 34% subcohort women African American) and other differences may reflect self-selection. The subcohort included more women who were younger, less educated, single, enrolled in Medicaid, had higher pro-pregnancy BMI values, gained less weight than recommended, smoked during pregnancy, and worked in a “ low” level occupation. 48 Mean birth weight was 3.51 i 0.45 kg among the non-subcohort and 3.38 i 0.49 kg among the subcohort, corresponding to mean z-scores of 0.24 :t 0.97 and -0.09 i 1.04, respectively. While a similar proportion of infants were born AGA in both groups (~78%), 18% of non-subcohort infants were born LGA compared to 11% of subcohort infants. A similar proportion of women in each group were classified as meeting LTPA recommendations (31% v. 29% for non-subcohort v. subcohort); however, a lower percentage of non-subcohort women were classified as inactive (39%) as compared to subcohort women (48%). Average length of recall for LTPA during pregnancy was 5.2 d: 1.3 years for the non-subcohort and 5.4 :t 1.4 years for the subcohort. Using unadjusted linear regression, we found that meeting LTPA recommendations was associated with a mean reduction in birth weight z-score among non-subcohort women ([3: -0.23, p<0.05, Table 3.2). However, this relationship was no longer significant once adjusted for maternal height and pre-pregnancy BMI. The final model for the non-subcohort explained 16% of the variance in birth weight z-score and included LTPA during pregnancy, maternal height, pro-pregnancy BMI, gestational weight gain, parity, maternal age at delivery, and smoking during pregnancy. Within the subcohort, LTPA during pregnancy was unrelated to mean birth weight z-score. Other maternal characteristics, similar to those retained in the non-subcohort model, explained 18% of the variance in birth weight z-score in the full model (Table 3.2). We next examined whether LTPA during pregnancy might be more related to extremes in birth weight. Being insufficiently active was not related to giving birth to either an SGA or LGA infant among both the non-subcohort and the subcohort (Table 3.3). Furthermore, meeting LTPA recommendations was not related to having an SGA 49 infant in either group. However, among the non-subcohort, meeting LTPA recommendations significantly reduced the odds (by 70%) of giving birth to an LGA infant even when adjusted for adequacy of gestational weight gain, maternal height, parity, occupational level, and recall length (aOR=0.30, 95%CI: 0. 14-0.64). Among the subcohort, meeting LTPA recommendations also tended to be associated with lower odds of giving birth to an LGA infant; however, these associations were not statistically significant (OR=0.73, 95% CI: 0.39-1.36; aOR=0.78, 95%CI: 0.41-1.50). We investigated the effect of LTPA during pregnancy on the birth weight z-score distribution using quantile regression and created quantplots to display the results (Figures 3.1-3.4). In quantplots, black circles represent parameter estimates for the effect of LTPA during pregnancy at every 0.05 quantile of the birth weight z-score distribution from 0.05 to 0.95. The gray shaded region represents the 95% confidence interval. Parameter estimates located below the zero line indicate a reduction in birth weight 2- score and are considered to be significant for any region of the distribution in which the shaded 95% confidence interval does not include the zero line. A mostly flat line would indicate that LTPA during pregnancy has a constant effect on the entire distribution of birth weight z-score and would result in a mean shift of the distribution. In contrast, a curvilinear line would indicate that LTPA during pregnancy has isolated effects on specific portions of the distribution and may not result in a mean shift. Figures 3.1 and 3.2 represent the association between LTPA during pregnancy and birth weight z-score among the non-subcohort. When considering insufficient LTPA during pregnancy, the shaded area of the quantplot includes the zero line across every quantile in both unadjusted (Figure 3.13) and adjusted (Figure 3.23) models, indicating a 50 null effect. Meeting LTPA recommendations also demonstrates a null relationship with birth weight z-score until the upper quantiles where the parameter estimates drop abruptly, indicating a reduction in birth weight z-score (Figure 3.1b). The adjusted model (Figure 3.2b) shows that meeting LTPA recommendations during pregnancy is essentially unrelated to birth weight z-score until the 0.65 quantile, after which it reduces birth weight z—score by up to 0.5 units. Figures 3.3 and 3.4 represent quantile regression results among the subcohort. Results are similar to those observed for the non-subcohort but are not as statistically significant. While being insufficiently active appears to increase birth weight z-score among the mid range of the distribution in unadjusted analyses (Figure 3.3a), these results are no longer significant in the adjusted model (Figure 3.4a). Meeting LTPA recommendations during pregnancy shows the same pattern as the non-subcohort results in that parameter estimates indicate reduced birth weight z-scores among the upper quantiles (>065); however, the estimated effect is non-significant in both unadjusted (Figure 3.3b) and adjusted (Figure 3.4b) models. All analyses were re-run after removing women with high maternal serum alpha-fetoprotein levels and no substantial differences were seen (data not shown). DISCUSSION Our results indicate that meeting LTPA recommendations during pregnancy was not associated with a shift in the entire birth weight z-score distribution. Rather, LTPA at/above recommended levels decreased the odds of LGA and significantly decreased birth weight z-score above the 0.65 quantile among the non-subcohort. The subcohort demonstrated similar trends; however, their results were not statistically significant. 51 The descriptive characteristics of women in the non-subcohort and subcohort varied widely, with the subcohort generally displaying more racial/ethnic and socioeconomic diversity (Table 3.1). This is likely due to the race-stratified sampling scheme used to create the subcohort, as well as the extra efforts used to follow-up this group. Unmeasured confounding due to lifestyle factors that are not easily measured may be greater in diverse samples and could partly explain the attenuated effects of LTPA during pregnancy on birth size in the subcohort vs. non-subcohort. The subcohort had a lower percentage of LGA births (11%) compared to the non-subcohort (18%), thus differences in the statistical significance of our results may also reflect lower power within the subcohort to detect relationships between LTPA and odds of LGA. While most (67%) of the subcohort participated in follow-up, only 26% of the non-subcohort returned the follow-up survey, and they did so in response to a single mailing. The non-subcohort therefore represents a more select group of women and their results may be less generalizable to other populations. However, previous literature has demonstrated that while non-responders may significantly differ from responders in regard to prevalence of exposure and outcome variables, the exposure-outcome associations are often similar in both groups if the study hypotheses are not suspected.24’ 25 While we were unable to assess the prevalence of LTPA participation among non- responders, the prevalence of LGA was lower among non-responders in both the non- subcohort (13% vs. 18% for non-responders vs. responders, p=0.06) and the subcohort (8% vs. 11% for non-responders vs. responders, p<0.001). Non-responders also tended to display greater racial/ethnic and socioeconomic diversity compared to responders in both groups. Despite selective follow-up, both the subcohort and non-subcohort showed the 52 same trend- meeting LTPA recommendations during pregnancy decreased odds of LGA and reduced birth weight birth weight z-score above the 0.65 quantile, without affecting mean birth weight. However, the effects were more robust and statistically significant only among the non-subcohort. Our results support those previously found by Alderman et. al. who also examined LTPA during pregnancy in relation to risk of SGA or LGA. Studying a group of 291 Coloradan women, they found that participation in 22 hr/wk of at least moderate LTPA during pregnancy (i.e. the equivalent of 2 6.0 kcal/kg/wk) reduced risk of delivering an LGA infant (OR=0.3, 95%CI=0.2-0.7) with no effect on delivering an SGA infant (OR=0.8, 95%CI=0.3-2.3).'4 A study on modifiable determinants of high birth weight (24.2 kg) among 553 Norwegian women found that LTPA during pregnancy was unrelated to birth weight, but that low LTPA pro-pregnancy (<1 hr/wk) increased risk of high birth weight (aOR=2.9, 95%CI: 1.273).26 In contrast to our results, a recent prospective cohort study among Danish women found no association between participation in sports/LTPA during the second or third trimesters of pregnancy and high birth weight (24.5 kg).'5 It appears that only ~3% of women in the Danish study gave birth to high birth weight infants, as compared to the higher prevalence of LGA reported here (18% among the non-subcohort and 11% within the subcohort), in the Coloradan study (12%), and of high birth weight in the Norwegian study (15%).l4’ 15‘ 26 Furthermore, the Danish study classified LTPA by categories of hours/week spent in sport or by categories of sport type (i.e., weight bearing vs. non- weight bearing). Thus, discrepancies in results may be due to underlying differences in 53 the birth during p normali deliter} intemni adaptdt‘ placent direct e use of. 01th : drama] recom: birth 6 for b0. metab PIan; feral g the birth weight distributions and/or to methodological differences in defining LTPA during pregnancy. Physical activity during pregnancy is thought to regulate fetal growth by helping normalize maternal blood glucose and by altering placental blood flow and nutrient delivery.9’ 27 Specifically, LTPA is associated with decreased placental blood flow intermittently during the exercise bout, but increased blood flow at rest due to training adaptationsn’ 28 While LTPA during the first and second trimesters appears to improve placentation and vascularization, LTPA during the third trimester may have the most direct effect on fetal growth.27 While we lacked trimester-specific information on LTPA during pregnancy, the use of quantile regression analyses in this study allowed us to refine existing knowledge of the association between LTPA during pregnancy and birth size. Our results showed a dramatic reduction in birth weight z-score associated with meeting LTPA recommendations during pregnancy, but only among the upper quantiles (>065). Higher birth weight is associated with short- and long-term health risks including birth trauma for both the mother and infant and increased risk for obesity and components of the metabolic syndrome during childhood.”” '7’ 29 Thus, our results suggest that LTPA during pregnancy may improve matemal/child health outcomes by helping to limit excessive fetal growth without reducing normal fetal growth. Some limitations must be noted. Information on LTPA during pregnancy was recalled retrospectively by our participants 3-9 years postpartum, thus recall bias cannot be discounted. A previous study found strong correlations (r=0.57-0.85) among recalled and originally measured LTPA during pregnancy at six years postpartum.30 However, 54 they had a select sample of mainly active women, thus their results may have limited generalizability. Bias due to recall error in our study may have been reduced by using a categorical LTPA variable with broadly defined groups. In addition, differential bias is less likely because women were not told of the study aims when they were asked to recall LTPA. We only queried women about LTPA performed during a “typical week” in pregnancy, thus we were not able to examine trimester-specific effects of LTPA on birth size. There is also still potential for unmeasured confounding because LTPA is correlated with other lifestyle factors, such as diet, that were not measured in this study. Despite these limitations, our study adds significantly to the existing literature. This is the first time a quantile regression technique has been applied to an investigation of LTPA during pregnancy on birth weight, which allowed us to more precisely define the relationship. In addition, the unique follow-up strategy employed in the POUCH study allowed us to evaluate the LTPA- birth size relation in two different populations. Within the non-subcohort we were able to examine the effect of LTPA during pregnancy on birth size among a more homogeneous sample of women with potentially fewer confounding variables. In contrast, the subcohort allowed us to examine the relation in a diverse sample of women that is more generalizable to the US. pregnant population. In conclusion, our results indicate that participating in LTPA during pregnancy at or above recommended levels is associated with lower risk for delivering an LGA infant, and reduces birth weight z-score among the higher quantiles of the distribution without causing a mean shift in the distribution. Due to the adverse maternal and child health effects associated with higher birth weight, our results may indicate a substantial health benefit for LTPA during pregnancy. Future studies with prospectively ascertained, 55 detailed measures of LTPA during each trimester of pregnancy are needed to test and refine our findings. 56 Table 3.]: Participant characteristics of the POUCH study non-subcohort and subcohort* Non-Subcohort Subcohort Chi-Square N (%) N (%) p-value N 418 596 Maternal Race White/Other 401 (96.0) 394 (66.1) <0.001 African American 17 (4.1) 202 (33.9) Pro-pregnancy Maternal BMI < 25 kg/m2 232 (55.5) 289 (48.5) 25-<30 kg/m2 112 (26.8) 139 (23.3) <0.001 2 30 kg/m2 74 (17.7) 168 (28.2) Maternal Height < 65 in 195 (46.7) 288 (48.3) 0.600 Gestational Weight Gains Low 48 (11.5) 104 (17.5) Recommended 149 (35.7) 187 (31.4) 0.026 High 221 (52.9) 305 (51.2) Maternal Age at Delivery <20 yrs 23 (5.5) 60 (10.1) 20-<30 yrs 202 (48.3) 346 (58.1) <0.001 2 30 yrs 193 (46.2) 190 (31.9) Maternal Education < High School 34 (8.1) 89 (14.9) High School 77 (18.4) 154 (25.8) <0.001 > High School 307 (73.4) 353 (59.2) Medicaid Yes 98 (23.4) 277 (46.5) <0.001 Relationship Status Single 54 (12.9) 173 (29.0) <0.001 Occupational Status” Low 28] (67.2) 457 (76.7) <0.001 Smoking during Pregnancy At least some 33 (7.9) 92 (15.4) <0.001 Parity Nulliparous 186 (44.5) 245 (41.1) 0.282 Child Gender Male 200 (47.9) 297 (49.8) 0.534 57 Table 3.1: (cont) N on-Subcohort Subcohort Chi-Square N (%) N (%) p-value Size at birth AGA 324 (77.5) 463 (77.7) SGA (s 10th %tile) 19 (4.6) 65 (10.9) <0.001 LGA (2 90"1 %tile) 75 (17.9) 68 (11.4) LTPA During PregnancyI Inactive 162 (38.8) 288 (48.3) Insufficiently Active 125 (29.9) 134 (22.5) <0.005 Meeting LTPA Recs 13] (31.3) 174 (29.2) Includes only POUCH study participants who gave birth at term and were enrolled in the 2007 follow-up study 5 Based on IOM guidelines according to pre-pregnancy BMI 1‘ Low Occupational Status = Clerical/Sales, Service/Blue Collar, or Homemaker/Other/Unknown I LTPA= leisure time physical activity; “Inactive”=no LTPA during pregnancy, “Insufficiently Active”= LTPA during pregnancy less than the recommended (<7.5 kcal/kg/wk), “Meeting LTPA Recs”: LTPA during pregnancy at or above the recommended level (27.5 kcal/kg/wk) 58 Table 3.2: Linear regression analyses for the unadjusted and adjusted association between maternal LTPA during pregnancy and birth weight z-score by subcohort status. N on-Subcohort Subcohort N=418 N=596 Model 1: Adjusted R2 0.0103 0.0063 LTPA'r Insufficiently Active -0.147 (0.374, 0.080) 0.200 (0.014, 0.431) Meeting LTPA Recs 0232 (0.457, -0.008)* 0.008 (-O.188, 0.204) Model 2*: Adjusted R2 0.0741 0.0510 LTPAI Insufficiently Active -0.085 (0.310, 0.140) 0.180 (0.031, 0.391) Meeting LTPA Recs -0.l77 (-0.397, 0.043) 0.006 (-0.188, 0.200) Maternal Height (in) 0.064 (0.032, 0.097)* 0.078 (0.047, 0.107)* Pre-Pregnancy BMI (kg/m2) 0.029 (0.014, 0.045)* 0.010 (-0.001, 0.021) Model 3*: Adjusted R2 0.1648 0.1840 LTPAl Insufficiently Active Meeting LTPA Recs Maternal Height (in) Pre-Pregnancy BMI (kg/m2) Gestational Weight Gain (lbs) Black Race (ref: White) Nulliparous (ref: Parous) Education (ref: < High School) = High School > High School Age at Delivery ( Ref: <20y) 20-<30 y 2 30 y Smoking in Pregnancy -0077 (0.297, 0.144) -0.105 (0.321, 0.110) 0.050 (0.018, 0.082)* 0.034 (0.018, 0.049)* 0.015 (0.010, 0.021)* -0.269 (-0.456, -0.082)* 0.425 (0.015, 0.836)* 0.492 (0.068, 0.917)* -0.360 (0.690, -0.030)* 0.071 (0.128, 0.269) -0010 (0.193, 0.173) 0.064 (0.036, 0.092)* 0.017 (0.006, 0.028)* 0.008 (0.004, 0.013)* —0.325 (0.497, -0.153)* -0.396 (0.556, -o.237)* -0.025 (-0.287, 0.237) 0.294 (0.051, 0.537)* -0.481 (0.704, -0.257)* *significant p-value < 0.05 I LTPA=leisure-time physical activity (Ref: no LTPA during pregnancy), “Insufficiently Active”: LTPA during pregnancy less than the recommended (<7.5 kcal/kg/wk), “Meeting LTPA Recs”: LTPA during pregnancy at or above the recommended level (27.5 kcal/kg/wk) *Models 2 and 3 also adjusted for length of recall (<4, 4-<6, 26 yrs) 59 a; e m .ev4 .e vv :82 we as?» was .3583 honmamoa macaw wEonm EEoEE .Qoonom :wE 56 208 8 .8153 .55 m8: cosmosno 3822: Am=088m3059=55 bran A3 3 NE £30: 35038 .935 35:35-95 no woman nomumwnoEEooou 20“ 53 no .wcuooa .3289 55.8on waist 5% Emma? .Cofiobzfis w> x028 88 383.2: Sm coma?“ 2m PMON tozoonsm » Amen o N .mVé .v vv :82 mo E3» 38 .EEEBod $52 Ecoumasooo EEBaE Age—«8383538 bran A5 3 NE Ems; 3:5me .035 zoumqwfimoa no woman 83358888 .29 55 Ho .3588 .3208 >28:on mats“. 5mm Emma? EEBaE 8m 35:?“ 2c. PMOn tosoonsméoz m 6139588 mSNV :32 3988888 05 05% no a mommamoa macaw :o< 3856535: {352%on macaw <3; oc0.b>:om£: 53:8 Regan oE: oSmETxSHA e modvfi 8m. 7:8 88. 78.8 83:: .8 88.738 5.88 who 83 3 .53 m3 3 .85.. 62-3.8 SW. 7E8 5.82.8 5.73.8 so 888 m3 8: 3 .52 3o 3 mo . . . .. . <9— §. 73.8 8... 788 so. 73.8 63.88 5838 2: 33 3 8 .o m: 3 .20.. 8e. 7%.8 A2. 72.8 84. 78.8 33-3.8 5 {$8 93 $8 3 3o .8; 3 mo . . . , <3 +83— +u>zu< +83— (o>_~u< High School), smoking during pregnancy (yes/no), and years of recall (< 4, 4—<6, 2 6 yrs). 64 ACKNOWLEDGEMENTS This work was supported by the Perinatal Epidemiological Research Initiative Program Grant from the March of Dimes Foundation [Grants 20FY01-38 and 20-FY04-37], the National Institute of Child Health and Human Development and the National Institute of Nursing Research [Grant R01 HD34543], the Thrasher Research Foundation [Grant 02816-7], the Centers for Disease Control and Prevention [Grants U01 DP000143-01 and R36 DP001322-01], and the Blue Cross Blue Shields of Michigan Foundation [Grant 1416 Student Award Program]. 65 APPENDIX QUESTIONS FROM THE POUCH F OLLOW-UP SURVEY Section A. PHYSICAL CHARACTERISTICS: In this section of the survey we will ask about some physical characteristics, like height and weight. 1. Is your child a: Boy [1 Girl [I 2. What is your child’s height now? Feet Inches 3. What is your child’s weight now? Pounds 4. In your jugment is your child currently: (Please check one) Underweight I] Normal weight 1] Overweight C 5. What is your height now? Feet Inches 6. What is your weight now? Pounds 7. How much weight did you gain during the POUCH pregnancy? Pounds Section B. PHYSICAL ACTIVITY: This section of the survey collects information about your exercise habits both now and during the POUCH pregnancy. We also want to know about how much exercise your child gets, and what kinds of activities he/she chooses. Please answer the following questions as they apply to the time DURING your POUCH pregnancy. 1. Did you perform any physical activity in your leisure time Yes Cl No Cl during your POUCH Study pregnancy? If No, Go to 3 below If YES: What was the activity you did the most? On average, how many days a week did you perform this activity? Days per week On average, how many minutes did you perform this activity? Minutes per day 66 2. Did you perform a second physical activity in your leisure time Yes Cl No Cl during your POUCH Study pregnancy? If No, Go to 3 below If YES: What was the activity? On average, how many days a week did you perform this activity? Days per week On average, how many minutes did you perform this activity? Minutes per day Please answer the following questions as they apply to y_o_g NOW. 3. Do you currently perform any physical activity in your leisure time? Yes C] No CI If No, Go to 5 below If YES: What is the activity you do the most? On average, how many days a week do you perform this activity? Days per week On average, how many minutes do you perform this activity? Minutes per day 4. Do you currently perform a second physical activity in your leisure time? Yes CI No C] If No, Go to 5 below If YES: What is the activity? On average, how many days a week do you perform this activity? Days per week On average, how many minutes do you perform this activity? Minutes per day Please answer the following questions as they apply to your POUCH Child NOW. (Circle gag answer that best applies to your child) 5. Compared to other children of the same age and sex, how would you describe your child’s (Please circle one} level of physical activity? A. Much less than others B. Somewhat less than others C. About the same D. Somewhat more than others E. Much more than others 67 6. What does your child usually do when he or she has a choice about how to spend his or her free time? (Please Circle OI_I_e)_ 1. Almost always chooses sedentary activities, such as watching TV, playing video games, or reading Usually chooses sedentary activities, such as watching TV, playing video games, or reading Just as likely to choose physically active play as inactive recreation Usually chooses physically active play Almost always chooses physically active play 68 REFERENCES l. 10. ll. 12. ACOG committee opinion. Exercise during pregnancy and the postpartum period. Number 267. American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet. Apr 2002;77(1):79-81. U.S. Physical Activity Guidelines for Americans. US. Department of Health and Human Services. http://www.health.gov/paguidelines: ODPHP Publication No. U0036; 2008. Evenson KR, Savitz DA, Huston SL. Leisure-time physical activity among pregnant women in the US. Paediatr Perinat Epidemiol. Nov 2004;18(6):400- 407. Damm P, Breitowicz B, Hegaard H. Exercise, pregnancy, and insulin sensitivity-- what is new? Appl Physiol Nutr Metab. Jun 2007;32(3):537-540. Hegaard HK, Pedersen BK, Nielsen BB, Damm P. Leisure time physical activity during pregnancy and impact on gestational diabetes mellitus, pre-eclampsia, preterm delivery and birth weight: a review. Acta Obstet Gynecol Scand. 2007;86(11):1290-1296. Dempsey J C, Butler CL, Williams MA. No need for a pregnant pause: physical activity may reduce the occurrence of gestational diabetes mellitus and preeclampsia. Exerc Sport Sci Rev. Jul 2005;33(3): 141-149. Gavard JA, Artal R. Effect of exercise on pregnancy outcome. Clin Obstet Gynecol. Jun 2008;51(2):467-480. Impact of physical activity during pregnancy and postpartum on chronic disease risk. Med Sci Sports Exerc. May 2006;38(5):989-1006. Weissgerber TL, Wolfe LA, Davies GA, Mottola MF. Exercise in the prevention and treatment of matemal-fetal disease: A review of the literature. Appl Physiol Nutr Metab. Dec 2006;31(6):661-674. Pivamik JM. Potential effects of maternal physical activity on birth weight: brief review. Med Sci Sports Exerc. Mar 1998;30(3):400-406. Leet T, Flick L. Effect of exercise on birthweight. Clin Obstet Gynecol. Jun 2003;46(2):423-43 1. Kramer MS, McDonald SW. Aerobic exercise for women during pregnancy. Cochrane Database Syst Rev. 2006;3:CD000180. 69 13. 14. 15. l6. 17. 18. I9. 20. 21. 22. 23. Hatch MC, Shu XO, McLean DE, et al. Maternal exercise during pregnancy, physical fitness, and fetal growth. Am J Epidemiol. May 1993;137(10):1105- 1114. Alderman BW, Zhao H, Holt VL, Watts DH, Beresford SA. Maternal physical activity in pregnancy and infant size for gestational age. Ann Epidemiol. Nov 1998;8(8):513-519. Hegaard HK, Petersson K, Hedegaard M, et al. Sports and leisure-time physical activity in pregnancy and birth weight: a population-based study. Scand J Med Sci Sports. Apr 2009: epub DOI 10.1111/j.1600-0838.2009.00918.x. Boulet SL, Alexander GR, Salihu HM, Pass M. Macrosomic births in the united states: determinants, outcomes, and proposed grades of risk. Am J Obstet Gynecol. May 2003; 188(5): 1372-1378. Hediger ML, Overpeck MD, McGlynn A, Kuczmarski RJ, Maurer KR, Davis WW. Growth and fatness at three to six years of age of children born small- or large-for-gestational age. Pediatrics. Sep 1999;104(3):e33. Holzman C, Bullen B, Fisher R, Paneth N, Reuss L. Pregnancy outcomes and community health: the POUCH study of preterm delivery. Paediatr Perinat Epidemiol. Jul 2001;15 Suppl 2:136-158. Kramer MS, Platt RW, Wen SW, et al. A new and improved population-based Canadian reference for birth weight for gestational age. Pediatrics. Aug 2001;108(2):E35. Lawrence EJ. A Matter of Size: Part 2 Evaluating the Large-for-Gestational-Age Neonate. Adv Neonatal Care. Aug 2007;7(4): 187-197. Institute of Medicine. Nutrition during pregnancy, weight gain and nutrient supplements. Report of the Subcommittee on Nutritional Status and Weight Gain during Pregnancy, Subcommittee on Dietary Intake and Nutrient Supplements during Pregnancy, Committee on Nutritional Status during Pregnancy and Lactation, Food and Nutrition Board. Washington, DC: National Academy Press; 1990:1-233. Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc. Sep 2000;32(9 Suppl):S498-504. Koenker R, Hallock K. Quantile Regression. J Econ Perspect. 2001 ;15(4): 143- 156. 70 24. 25. 26. 27. 28. 29. 30. Osler M, Kriegbaum M, Christensen U, Holstein B, Nybo Andersen AM. Rapid report on methodology: does loss to follow-up in a cohort study bias associations between early life factors and lifestyle-related health outcomes? Ann Epidemiol. May 2008;18(5):422-424. Martikainen P, Laaksonen M, Piha K, Lallukka T. Does survey non-response bias the association between occupational social class and health? Scand J Public Health. 2007;35(2):212-215. Voldner N, Froslie KF, Bo K, et al. Modifiable determinants of fetal macrosomia: role of lifestyle-related factors. Acta Obstet Gynecol Scand. Apr 2008;87(4):423- 429. Clapp J F. Influence of Endurance Exercise and Diet on Human Placental Development and Fetal Growth. Placenta. Jun-Jul 2006;27(6-7):527-534. Pivamik J M, Mauer MB, Ayres NA, Kirshon B, Dildy GA, Cotton DB. Effects of chronic exercise on blood volume expansion and hematologic indices during pregnancy. Obstet Gynecol. Feb l994;83(2):265—269. Boney CM, Verrna A, Tucker R, Vohr BR. Metabolic syndrome in childhood: association with birth weight, maternal obesity, and gestational diabetes mellitus. Pediatrics. Mar 2005;1 15(3):e290-296. Bauer PW, Pivamik JM, F eltz DL, Paneth N, Womack CJ. Validation of a historical physical activity recall tool on postpartum women. Med Sci Sports Exerc. 2006;38(Suppl):SS61. 71 CHAPTER 4 REVIEW OF LITERATURE PART TWO: CHILD SIZE INTRODUCTION Recent results suggest that LTPA during prior to/pregnancy may reduce the odds of giving birth to a LGA infant; however, further effects on child growth and development have not been adequately investigated. 1’ 2 Birth size has been directly related to size during childhood and adolescence.3 Since active women are less likely to have large infants, it is therefore possible that children of more active mothers may also have a reduced risk for being overweight later in life. To date only one study has considered possible influences of maternal LTPA during pregnancy on child size.4 The authors found that offspring of more active women were lighter and leaner at birth, and had reduced body fatness at five years compared to offspring of less active mothers.4 However, these results are based on a small sample (n=40) of healthy, middle-class, White women. It has yet to be seen whether LTPA during pregnancy may affect child body size in a more diverse sample with greater power to assess confounding and/or mediation by other perinatal characteristics. Further investigation of the relationships among LTPA during pregnancy, birth size, and child body size has the potential to elucidate a practical intervention to promote healthy pregnancies, birth outcomes, and children with healthy weights. TERMINOLOGY OF CHILD WEIGHT STATUS The majority of studies evaluating body size among children have relied on body mass index (BMI) to describe appropriate size rather than body fat measures, which are more difficult to obtain and have no accepted standards for under- or over-fatness. Since 72 BMI varies considerably with age and maturation, values are usually compared to age- and sex-specific growth charts of a reference population to determine weight status via 2- scores or percentile rankings. In the US, the Centers for Disease Control and Prevention (CDC) growth charts from 2000 are most commonly used for children ages 2-19 years.5 While earlier studies used “at risk for overweight” and “overweight” to designate weight categories, in 2007 an expert committee recommended that the terms “overweight” and “obese” be universally adopted to minimize confusion, provide continuity with adult terminology, and underline the health risk associated with high weight for height status in childrenf’ Thus, for the remainder of this dissertation, the term “overweight” will be applied to children with BMI values 2 85th and < 95th percentile and “obese” will designate BMI values 2 95‘h percentile.“ 7 It should be noted that while BMI classifications among adults are based on risk of disease, risk-related cut-off values for children are not available given that adverse events do not typically occur until adulthood.7 However, recent evidence on health risks among overweight and obese children suggests that the current cutpoints of the 85th and 95th percentile in BMI are clinically relevant.8 F urtherrnore, evidence suggests that the 95th percentile cutpoint for obesity correctly identifies the fattest children with high specificity and moderate sensitivity.9 Validity of Parentally Reported Child Body Size Values The majority of studies on child overweight/obesity rely on parentally and/or self reported height and weight to calculate BMI and/or z-scores. Studies on the accuracy of parental reports of weight, height, and calculated BMI values have shown variable lO-l4 results. A study of 864 Dutch children 4 yrs of age found small mean differences 73 between measured and parentally reported values for weight (-0.02 i 1.0 and 0.02 :I: 1.0 kg for girls and boys, respectively), height (-0.5 :l: 1.7 and -O.4 i 1.7), and calculated BMI (0.1 i 1.0 for either gender).'4 In total, only 9.7% of children in this study were misclassified on the basis of parental reports into inappropriate weight categories; however, parents were asked to report weight and height values from a recent doctor visit or directly measure the children themselves. '4 A greater degree of misclassification is evident when mothers are asked to report weight and height of their children without measuring or weighing them. In a different study of 4 yr old weight status, mothers reported weight and height during a face-to-face interview without knowing that measurements would be taken a few weeks later. '0 While no significant height differences were observed, mothers overestimated boys’ weights, but not girls’, which led to greater overestimation of BMI in boys compared to girls. Despite these errors, 70% of BMI values calculated fiom maternally reported values were within one standard deviation of measured BMI and kappa statistics indicated fair to moderate agreement for weight status classification by reported and measured values for girls (k=0.337) and boys (k=0.408).'° Child age may play a role in the direction and/or magnitude of body size misreport. One study providing information on validity of maternally reported weight and height of Mexican- American children ages 6 mo — 11 yrs demonstrated that mothers tend to overestimate weight in young children (<4 yrs) and underestimate weight at older ages, while height was underestimated for all age groups. '2 Correlations between reported and measured values were high for heights and weights (FOSS-0.96) but were low to moderate for BMI (r=0.15-0.44). 74 The bulk of evidence on the validity of maternal report of child body size indicates moderate to good agreement for individual measures of height or weight, but poorer agreement for calculated BMI values and/or weight classification by BMI.'°' ‘2‘ '4 There is some evidence that the accuracy of report is improved when parents are asked to report recent measurements from a health care visit and/or directly measure their child.”’ '4 It appears that maternal misclassification of their children’s weight status may be influenced by child gender, with women being more likely to perceive their male children as belonging to a lower weight status category than they really do. '1 Maternal overweight status, child dissatisfaction with their weights, and dieting are also associated with misclassifying overweight children as normal weight at age 14 yrs.” In light of these reports, results of studies using parental reports of child height and weight and/or perceptions of child weight classifications should be interpreted cautiously. EPIDEMIOLOGY OF CHILDHOOD OVERWEIGHT STATUS Studies from several developed countries have shown marked increases in the prevalence of childhood overweight and obesity in recent years. ”'2' Within the United States, data from national surveys indicate that the prevalence of obesity in children ages 6-11 yrs has increased dramatically from ~6% in the 1970’s to ~11% in the 1990’s and has continued to rise since then.22 The most current prevalence values of high BMI 1.23 These values are based on data from the among US. children are shown in Table 4. National Health and Nutrition Examination Surveys (N HANES) conducted in 2003-04 and 2005-06 and indicate that prevalence of being overweight and/or obese is higher for older children. No significant differences by gender were seen; however, marked race/ethnic differences in weight status were observed. Specifically, non-Hispanic white 75 chfld Supp power offar H y EVEI'B hens. Inuhi Obesi 3550c Obesi their 10 be children have the lowest prevalence of obesity (14.6%) as compared to either non- Hispanic black (20.7%) or Hispanic children (20.9%).23 Table 4.1: Prevalence of high BMI for age among US Children, 2003-2006 (Adapted from Ogden et al. 2008)23 Age Group BM] 2 85fll Percentile BMI 2 95th Percentile (yrs) N % (SE) % (SE) 2-19 8165 31.9 (1.2) 16.3 (0.9) 2-5 1770 24.4 (1.6) 12.4 (1.0) 6-11 2096 33.3 (2.0) 17.0 (1.3) 12-19 4300 34.1 (1.5) 17.6 (1.2) Income/socioeconomic status (SES) appears to be weakly and inversely related to child weight status.” 22’ 24‘ 25 Data collected from participants in New York’s Special Supplemental Nutrition Program for Women, Infants, and Children (WIC) indicate that poverty may be related to obesity rates in young children.25 Within this state-wide sample of families with income levels S 185% of the federal poverty line, 16.1% of children ages 2-4 years were classified as obese, which is considerably greater than the national average of 12.4% for this age group.23’ 25 Nationwide data indicate weak associations between SES and childhood obesity that are gender and race specific.26 Using data from multiple waves of NHANES, Wang and Zhang 26 found that SES was unrelated to obesity rates among 2-9 yr old children, while among 10-18 yr olds, high SES was associated with decreased risk of obesity in white boys and girls, but increased risk of obesity among black girls and Mexican American boys. Overall, their results indicate that the increased risk of childhood obesity associated with SES/income is small and appears to be weakening over time.26 Unfortunately, not only has the prevalence of obesity among children increased, obese children have also been getting appreciably heavier. When data from the National 76 eahhl Xlikld titties: being! \‘ girl I ofover :uqaan liEti $1,11ng risks. incre child Pics dysl in p. ”121E Health Examination Survey (NHES) II or 111 (1963-1970) are compared to data from the NHANES 111 (1988-1994), little to no differences in quantile-specific BMI values between the surveys are seen in the lower quantiles, but large differences are present in the higher quantiles, indicating that the heaviest children in NHAN ES 111 were significantly heavier than in NHES II or III.22 Additionally, using estimates of the “extent of overweight” (e. g., the amount by which each obese child exceeds his/her specific BMI cutpoint), obese children were on average 12% heavier than their threshold value in 1971-74, and 14% heavier by 1999-2000.27 Thus while the prevalence of obesity increased by 182% from 1971-74 to 1999-2000, the extent of overweight increased by 247% according to national data.27 HEALTH RISKS OF OVERWEIGHT IN EARLY CHILDHOOD Short- Term Health Risks Childhood overweight/obesity is associated with both short- and long-term health risks. In fact, many obesity-related health conditions once noted among adults only are increasingly being diagnosed among children and adolescents.28 Specifically, obese children have been shown to be at increased risk for concomitant hypertension/high blood pressure, abnormal enthothelial function, insulin resistance and/or type II diabetes, dyslipidemia, and even atherosclerosisg’ 28 The emergence of childhood type II diabetes in particular appears to have paralleled the increasing prevalence of obesity and the two may be causally linked.28‘ 29 Authors have also noted a clustering of risk factors among overweight children, commonly termed the metabolic syndromeg’ 28 The metabolic syndrome is broadly defined by the presence of abdominal obesity, elevated blood pressure, glucose, and 77 triglycerides, and lowered high-density lipoprotein-cholesterol (HDL-C).30 Specific cutpoint criteria for each risk factor in children are not standardized, thus true prevalence rates are difficult to estimate. However, it appears that between 4-10% of adolescents in the US. have the metabolic syndrome and the prevalence increases to 30-50% among obese children/adolescents.3O It is important to note that children do not have to be obese to be at increased risk for metabolic syndrome components. Bell et al.3 I have reported continuous relationships between increasing BMI z-score and blood pressure and fasting insulin, as well as curvilinear relationships with HDL-C and triglycerides such that changes in BMI z-score at the high end of the spectrum had a greater impact on unfavorable lipid profiles. These results, based on a sample of only 177 children, indicate that children’s risks of cardiovascular/metabolic complications increase across the range of BMI and are not threshold-dependent.3 ' In addition to cardiovascular/metabolic co- morbidities of overweight/obesity, reviews of past literature have also noted increased prevalence of psychological problems such as depression and anxiety (especially among girls), asthma, and gastroesophageal reflux associated with overweight status.8’ 28 Long-Term Health Risks Perhaps even more troubling than the short-term health consequences, childhood overweight/obesity appears to track into adulthood, thereby placing these children at higher risk for adult obesity and obesity-related health conditions. Two recent reviews of both US. and European studies found strong evidence for the persistence of childhood overweight/obesity into adulthood, which was significantly more likely among children with at least one overweight parent, with more severe overweight status, and/or with overweight/obesity present at older ages.8’ 32 While risk estimates varied considerably 78 among studies, overweight children were at least twice as likely as their normal weight peers to become overweight adults (OR/RR ranged from 1.9 to 22.3 for overweight children >2 yrs of age).32 Childhood overweight/obesity is also linked to increased risk of obesity-related health conditions in adulthood, such as cardiovascular disease and diabetesg’ 33‘ 34 Beyond the mere presence of overweight/obesity in childhood, rapid growth during “critical time periods” appears to confer greater risk for future adverse health outcomes.35 Researchers are just now beginning to explore the individual contributions of rapid growth during infancy, the timing of the adiposity rebound in childhood (i.e. the nadir of BMI or weight), and the phenomenon of changing BMI percentile growth trajectories towards risk of later obesity.” 35‘ 36 Thus far, results indicate that factors operating in infancy and early childhood may have dramatic influences on future weight status.24’ 36 In particular, rapid growth during infancy resulting in changing growth trajectories and/or reaching the nadir of weight or BMI (e. g. hitting the “adiposity rebound”) earlier in childhood have been significantly associated with increased risk of adolescent/adult obesity.24’ 36'” Given both the short-term and long-term health consequences of childhood obesity, it is important to identify early factors which predispose children towards overweight/obesity for the purposes of intervention/prevention strategies. CHILD BEHAVIORS AND CHILD BODY SIZE The causes of childhood overweight/obesity are multifaceted and complex. This dissertation focuses primarily on perinatal exposures that may influence child growth and body size; however, child behaviors and parental influences must also be considered. 79 More proximal exposures, such as child behaviors and parental characteristics, will be briefly reviewed before considering the impact of more distal perinatal factors on child overweight/obesity in greater detail. Child Physical Activity Any investigation of factors related to child weight status must consider the influence of the child’s own behaviors. Intuitively, the emergence of childhood overweight would seem to be primarily a problem of energy balance. Either the child is consuming too much energy or is not expending enough energy to experience normal growth patterns and is thus storing excess fat and becoming overweight. While this argument seems logical, it probably oversimplifies the relationships among LTPA, diet and weight status in childhood. In fact, while research supports relationships between a child’s LTPA participation and his/her weight status, results have been neither as strong, nor as consistent as expected.39 National guidelines recommend that children participate in at least 60 min/day of at least moderate LTPA.40 U.S. youth participation in LTPA varies by age with 42.0% of 6-11 yr olds meeting recommendations, but only 8.0% of 12-15 yr olds and 7.6% of l6- 19 yr olds meeting recommendations.“ Gender differences are also seen with fewer girls meeting recommendations than boys at any age group (35% vs. 48% and 3% vs. 12% for girls vs. boys at ages 6-11 and 12-15 yrs, respectively).4| These prevalence rates are based on accelerometry data from the 2003-04 NHANES, and thus provide an objective assessment of LTPA participation in a national sample of children. One of the major difficulties in synthesizing results of past research on LTPA and body size in children is the wide array of methodologies employed to measure LTPA. 80 These have ranged from subjective measures such as parental/caregiver report and direct observation to the more objective accelerometry and/or doubly labeled water techniques. Reviews on the impact of LTPA on risk of childhood obesity have noted that much inconsistency among previous studies may be attributed to variation in LTPA measurement methods.” 42’ 43 Despite these obstacles, most studies report negative correlations between LTPA and fatness among children; however, results with BMI tend to be null.4”"5 1 Among a representative sample of 3-year-old children, obesity prevalence did not vary by outdoor play time or by minutes/day of TV viewing.5 I However, Trost et al. found that overweight 3—5 yr old boys were significantly less active than non-overweight boys.46 Trost et al. employed more exact measures of LTPA, including direct observation and accelerometry, which may have resulted in greater power to detect differences by overweight status. Interestingly, no significant differences in LTPA were observed between overweight/non-overweight girls.“5 The authors speculated that this may have been due to girls choosing to participate in less vigorous activities that were not as influenced by obesity status. Results from a study of somewhat older children (6-8 yrs) showed that neither free-living activity energy expenditure nor LTPA level calculated from 7-day doubly labeled water were significantly related to BMI; however inverse relationships with fat mass index were noted (beta coefficient = -5.57, p<0.001).48 Abbott and Davies also found LTPA level from doubly labeled water to be inversely related to percent body fat in 5-10.5 yr old children (r= -0.44, p=0.004).47 The directionality of the relationship between overweight status/body fatness and LTPA participation could not be established by the previously mentioned studies due to 81 the cross-sectional nature of these data. Longitudinal data from the F ramingham Children’s Study, however, indicate that LTPA in early childhood results in smaller gains in BMI, triceps, and sum of five skinfolds.45 After controlling for child gender, baseline age and BMI, TV watching, percentage of calories from fat, and baseline BMI and education level of both parents, the most active children had significantly lower triceps and sum of skinfolds at the end of follow-up.45 A trend for lower BMI among the most active tertile was also observed; however, this did not reach statistical significance (BMI = 20.3 i 0.6, 19.8 :t 0.5, and 18.6 d: 0.6 kg/m2 for low, moderate, and high active tertiles, respectively, p=0.052).45 Unfortunately, LTPA intervention/prevention studies have not been able to demonstrate measureable improvements in weight status to date. A recent review of school-based obesity prevention programs concluded that changes in diet and/or LTPA did not lead to improvements in overweight outcomes, but emphasized that methodological quality of these trials was low.52 A systematic review of three long term (>1 yr) and four short term (3 mo — 1 yr) interventions for preventing obesity also found inconclusive evidence for child LTPA exerting a protective effect.“ However, a third review on interventions to reduce sedentary behaviors found consistent evidence for improved weight indices associated with fewer hours/wk of sedentary activity.53 It is possible that sedentary time (e. g., time spent watching TV or playing video games) is easier to measure and/or more susceptible to intervention efforts, thus greater affects on weight status are able to be demonstrated as compared to LTPA interventional efforts. It is difficult to know the mechanism through which sedentary behavior may be related to weight status. One study of ~1000 young children found that 2+ hrs/d of TV 82 time W8 L36r6j race.nl pedhnl mawe with m 0Verht efiécts limes3 associz reconr Chfld huake OVenh boner that c: \Vhole and St dauii thela incre: time was associated with increased risk of being overweight at 3 yrs (OR=2.92, 95%CI: 1.36-6.24) and 4.5 yrs (OR=1.71, 95%CI: 1.03-2.83).54 However, once adjusted for sex, race, maternal education, marital status, and age, TV exposure was no longer a significant predictor of overweight status.54 Thus, the relationship between sedentary time and child overweight status may be confounded by other family and lifestyle factors that travel with more sedentary behavior. It is also possible that sedentary behavior increases risk of overweight/obesity through displacing LTPA, increasing energy intake via snacking, effects of food advertising on TV, and/or decreased metabolic rate during sedentary time.53 Regardless of the mechanism involved, in recognition of the health risks associated with excess sedentary behavior, the American Academy of Pediatrics has recommended that screen time be restricted to 60 minutes or less per day.55 Child Dietary Patterns In addition to energy expenditure via LTPA (or the lack thereof), energy intake/dietary habits have also been examined in relation to risk of childhood overweight/obesity. Dietary recommendations vary by age and gender among children; however, the most recent guidelines published by the American Heart Association stress that children >2 yrs old should limit juice intake, eat fruits and vegetables daily, focus on whole-grain breads and cereals, reduce intake of sugar-sweetened foods and beverages, and seek to balance dietary calories with LTPA to maintain normal growth.56 National data indicate that mean energy intake among children has not changed significantly since the late 1970’s, except among adolescent females who demonstrated a significant increase in energy intake from the 1970’s to 1988-94.57 83 To date, studies on the importance of dietary factors on child weight status have been inconclusive. Cross-sectional studies have failed to demonstrate significant relationships between either global energy consumption or specific dietary elements and child BMI or fatness.58 Similarly, longitudinal studies in children have not demonstrated relationships between energy intake or percentage of macronutrient consumption and rates of weight gain or future risk of overweight.58 Studies focusing on consumption of sugar-sweetened soft drinks and child weight status have also had limited success. More than half the studies with cross-sectional data reported null associations and only 8 of 16 longitudinal studies showed any significant relationships between sweetened drinks and weight gain, which were apparent in only select subgroups and/or disappeared with adjustment for confounders.59 Inconsistencies in methodologies for assessing dietary patterns, definitions of meeting guidelines, and a predominance of cross-sectional data may be obscuring true relationships between dietary patterns and risk of childhood obesity.” 59 For example, under-reporting of food-consumption is a problem that may vary by child weight status, with more obese children underestimating dietary intake the most.60 Also children who are currently overweight/obese may be restricting calories to lose weight, thereby disguising differences in macronutrient consumption between children who maintain healthy weights and those who become obese. More research is needed with better measures of dietary quality among children of all age ranges to further investigate relationships between dietary factors and weight status. Recently, milk consumption in childhood has received marked attention as a possible risk factor for the development of overweight/obesity. Significant inverse 84 relationships between frequency of milk consumption and child size have been reported among cross-sectional samples.6| The mechanism linking milk consumption to weight status is currently unclear, although the association was significant even after adjustment for a range of confounding variables including parental weight status and education, child. LTPA, age, gender, and frequency of consumption of other foods.61 Longitudinal research on milk consumption is needed to determine the directionality of its relationship with child obesity. Summary of Child Behaviors and Weight Status In conclusion, evidence for significant relationships between LTPA or dietary patterns and child weight status is modest, although studies utilizing objective measures of LTPA have shown significant inverse relationships with body fatness measures.39 Excessive sedentary behavior has also been shown to be a consistent risk factor for overweight/obesity.53 While the majority of dietary evidence has been weak, inverse associations between BMI z-scores and frequency of milk consumption among children have been observed.“ Further research on child behaviors is needed to determine viable avenues for prevention of overweight status. When examining the influences of child LTPA and dietary behaviors on child weight status it will also be important to consider parental characteristics that may modify the child’s susceptibility for overweight status and/or influence the child’s probability of participating in healthy behaviors. PARENTAL CHARACTERISTICS AND CHILD BODY SIZE Studies of parental influences on child weight status have focused mainly on the concurrent presence of parental overweight/obesity and parental influences on child dietary and LTPA behaviors. Both maternal and paternal size are positively related to 85 Obes “6ft (OR \rcg chflt over Resr hent nudi 500 o. meas “fig, Thus genel fluoul 0PPOl Pamer heFr child size 50’ 62 and parental overweight/obesity significantly increases risk for child obesity.”64 A study of over 2000 children found that odds of obesity at 4.5 yrs of age were significantly higher among children with one (OR=2.1, 95%CI: 1.3-3.6) or two (OR=3.2, 95% CI: 1.7-5.8) currently overweight parents compared to children of normal weight parents.63 These results have been replicated in other samples with a variety of child age groups.” 62’ 64 The mechanism for the relationship between parental and offspring weight status has not been established. Genetic influences are undoubtedly at play; however, parental overweight status may also affect child weight through alternate environmental pathways. Results from classic studies conducted by Claude Bouchard in the 1980’s indicate that heritability of fat mass in adults reaches ~25% of the phenotypic variance, while family studies indicate that the maximum heritability of obesity phenotypes ranges from 30- 50%.65’ 66 A more recent study of Hispanic children found heritability of anthropometric measures to range from 31-71%, with height having the strongest heritability (71%) and weight and fat mass having comparable heritabilities of 36% and 33%, respectively.67 Thus children of obese parents are at increased risk of overweight/obesity due to their genetic make-ups. In addition to genetic influences, parental obesity may affect child weight status through the creation of an “obesogenic” environment at home that de-emphasizes LTPA opportunities and increases access to energy-dense foods. Parental dietary and LTPA patterns are known to influence children’s participation in healthy behaviors.68’ 69 Within the Framingham Children’s Study, 7 yr olds with two active parents were 5.8 times more likely (95%CI: 1.9-17.4) to be active themselves compared to children with two inactive 86 parent: food 31 been 5. recent predie of foo. feedin prevail While amoul PETCEI $11sz Child , prESte parents.69 Young children are also dependent upon parents to provide access to nutritious food and to set appropriate serving sizes. Parental monitoring and family cohesion have been significantly related to youth healthy eating behaviors and LTPA participation.68 A recent review of child feeding behaviors concluded that parental monitoring of fat intake predicted lower child BMI, but that both maternal pressure to eat and maternal restriction of foods were associated with higher weight gain throughout childhood.7o It should be noted that ethnic and cultural differences in adult obesity rates, child feeding practices, and LTPA promotion may be partly responsible for the higher prevalence of obesity among African American and Hispanic children, as compared to White children. It is known that overweight/obesity is more prevalent among African American and Hispanic adults compared to Whites, thus minority children have a higher probability of having at least one overweight/obese parent.7l A recent review demonstrated that African American and Hispanic children also reported higher levels of obesity promoting behaviors such as skipping breakfast, eating fast food, ingesting low amounts of fruits and vegetables, and frequent TV watching.72 Additionally, parental perceptions/concem about child weight status may differ by race/ethnicity due to alternate cultural norms which may be more permissive of obesity.72 Summary of Parental Influences and Child Weight In conclusion, children with an overweight parent are more likely to be overweight themselves.63 The mechanism through which parental obesity confers risk for child obesity is multifaceted, including both genetic and environmental influences.24 The presence of maternal obesity has sometimes been shown to have a greater affect on child obesity than paternal obesity.73 It is possible that maternal obesity may confer additional 87 risks to the offspring via changes in the prenatal environment which influence fetal development.3’ 74 PERINATAL EXPOSURES AND CHILD BODY SIZE Given the increased prevalence of obesity among young children, researchers have begun to examine early life and prenatal exposures in relation to child grth and body size.“ 75 Due to our inability to directly measure fetal exposures, birth weight has often been used as an indicator of prenatal environmental conditions.76 Other perinatal factors that have received attention as potential risk factors for childhood obesity include maternal size during pregnancy and gestational weight gain, the presence of health conditions during pregnancy such as diabetes, maternal behaviors during pregnancy such as smoking, nutrition, and LTPA, and factors operating during infancy including infant growth rates and breastfeeding practices. Each of these factors will now be reviewed for relation to child size and body composition. Birth Weight and Child Size Past research has demonstrated either a linear or J- / U-shaped relationship between birth weight and subsequent child BMI.3’ 74’ 76’ 77 Overall, macrosomic/LGA infants have approximately twice the risk of becoming overweight children as compared to normal weight infants.“ 74‘ 76 Low birth weight combined with rapid growth during infancy has also been consistently associated with risk of overweight and central adiposity in childhood.3' 77 When considering body composition specifically, birth weight has shown strong direct relationships with fat-free mass, but correlations with future fat mass/percent fat have been less consistently demonstrated.76 LGA/macrosomic infants have been found to 88 have increased body fatness at birth when compared to normal weight infants. 78'80 Some have also shown that excess fat mass persists into early childhood, placing these children at higher risk of overweight/overfatness and the metabolic syndromes"83 However, while absolute amounts of body fat may be higher among children born macrosomic/LGA, lean body mass is also higher when compared to normal weight children, thus relative percent fatness may be unchanged.76’ 84 To date, results among past literature are mixed, with 81,85 some studies reporting direct relationships and others finding either no relationship or inverse relationships between birth weight and measures of fatness in childhood.“ 86'” Differing methodologies for assessing body composition, the inclusion of variable child age ranges, and racial/ethnic variations in maturation and body fat distribution make it difficult to draw conclusions based on the small number of studies that have been conducted thus far. Therefore, while birth weight has been shown to be directly related to child weight and BMI, effects on individual components of body size are less certain.” 76 The mechanism linking birth weight to child body size is not clearly understood. It is possible that birth weight mediates prenatal exposure effects on future child size. In this scenario, size at birth merely marks an adverse prenatal environment that sets the stage for later obesity. Hypothesized prenatal exposures have included maternal obesity and/or inappropriate weight gain during pregnancy (with or without the presence of diabetes), as well as maternal behaviors during pregnancy such as smoking, nutrition, and LTPA. Maternal Size / Weight Gain during Pregnancy and Child Size In Chapter Two it was shown that maternal size and weight gain during pregnancy were both positively related to birth weight, with the largest infants born to 89 0\ ED“ ill nanl' hn0“' onofl” exces: pregn. chfldr “rigt l‘ 9.) '-JJ dz posnl Contrl nhner Em C rekne mass overweight women who gained excess weight.78‘ 90 Additionally, maternal overweight status is associated with greater neonatal fat mass, even when overall BW is unaffected.” It now appears that maternal size and gestational weight gain may have lingering effects on offspring size and fatness during childhood as well.” 92’ 93 Studies have consistently demonstrated increased risk for overweight/obesity and excess body fatness among offspring of women who were overweight/obese during pregnancy.” 77’ ”’94 Relative risk of offspring obesity at 2-4 yrs is twice as high among children whose mothers were obese during pregnancy, even when controlling for birth weight, gender, maternal demographics, smoking, and gestational weight gain (RR= 2.0- 2.3 depending on child age).94 Among older children, pre-pregnancy BMI has also been positively associated with child adiposity and weight on a continuous scale.92 After controlling for child age, birth weight, infant weight gain, breastfeeding duration, maternal height, gestational weight gain, and smoking during pregnancy, maternal mid- arm circumference in late pregnancy and pre-pregnant BMI were each significantly 4.49 related to higher offspring fat mass index (fat mass (kg)/ height (m) ) , but not lean mass index (lean mass (kg)/ height (m)2'2) at 9 yrs of age.92 There are several pathways through which maternal obesity during pregnancy may influence later child weight status. These include genetic contributions and the encouragement of unhealthy dietary and LTPA patterns as discussed previously. In addition, the presence of maternal obesity may alter the prenatal environment, thereby signaling changes in fetal development that promotes excess weight gain/fat storage later in life.35 Animal models have shown that offspring of overweight rodents overfed a high- fat diet during pregnancy experienced altered pancreatic development and displayed 90 increas pregnal sheep< among oil‘spri influer ovens aherel lflfldEi exces risk c increased adiposity in adulthood.35 Additionally, increased nutrient supply during pregnancy has been shown to enhance adipogenesis and increased fat deposits in pig and sheep offspring.3 5 These results indicate that fetal over-nutrition, frequently present among obese women, may alter regulatory pathways for energy homeostasis in the offspring and increase risk of obesity independently from postnatal environmental influences. Excessive gestational weight gain has also been related to child overweight/obesity; however, the significance of this relationship has sometimes been altered by controlling for maternal pre-pregnant size.92' 93’ 95‘ 96 When compared to inadequate weight gain (based on IOM recommendations and pre-pregnancy BMI), excessive weight gain is associated with higher BMI z-scores and more than 4 times the risk of overweight at 3 yrs of age (OR=4.35, 95%CI: 1.69-1 1.24).95 In contrast, another study found no evidence of a significant main effect for gestational weight gain on offspring obesity at 3 yrs; however a significant interaction between maternal BMI and gestational weight gain was noted such that the risk of obesity was highest among offspring of obese women who gained excess weight.96 In general, it appears that the presence of maternal obesity is more important than excess weight gain for the development of childhood overweight/obesity.” 93 Maternal Diabetes in Pregnancy and Child Size Diabetes present during pregnancy is known to increase size at birth and is also associated with increased risk of child obesity.83 In a study of discordant sibships among a sample of Pima Indians, Dabelea et al.97 showed that mean BMI was 2.6 kg/m2 higher among offspring of diabetic vs non-diabetic pregnancies. This study involved 183 91 siblings <9ym. snnfidr lllll'fltli: nudies found anenu obesn apan possfl “eigl there we; anni uh DTOr bee OVe siblings from 52 families and compared offspring BMI within a 3-yr time interval (i.e., 6- <9 yrs, 9-<12 yrs). Because siblings shared similar postnatal environments and received similar genetic influences on body size, these results suggest that exposure to the intrauterine diabetic environment specifically increases risk of childhood obesity.97 Other studies involving larger, nationally representative samples of children and families have found that the relationship between gestational diabetes and risk of offspring obesity is attenuated after controlling for maternal BMI.83’ 98 Complex interactions among maternal obesity, gestational diabetes, and size at birth (i.e., LGA) are likely to exist, and teasing apart the individual contributions of each factor towards child body size is difficult.82 It is possible that shared matemal/child environments which contribute towards maternal weight and gestational diabetes also promote excess weight gain in the offspring and therefore cloud relationships between fetal exposure to diabetes and future obesity. Mechanistically, both maternal obesity and/or diabetes during pregnancy (either pre-gestational or of gestational onset) would increase the delivery of glucose, lipids, and amino acids to the fetus, leading to increased fetal growth and larger infant fat mass.” 99‘ '00 Animal studies further support fetal over-nutrition altering offspring metabolic processes leading to increased fat storage.35 In utero exposure to diabetes among rats has been shown to alter appetite control among offspring as well leading to hyperphagia and overweight later in life.3 Overall, studies support an association between gestational diabetes and offspring obesity; however, separating the individual effects of diabetes, maternal obesity, and excess gestational weight gain is difficult.77 Maternal Smoking during Pregnancy and Child Size 92 men on chi LTPA child! 811th 01638 RICE 31110" lmEr borr infa; 5010 in in hm Beyond maternal anthropometrics and health conditions during pregnancy, maternal behaviors may also alter the fetal environment, leading to downstream effects on child body size. Specifically, the impacts of maternal smoking, energy intake, and LTPA during pregnancy have been considered in relation to size at birth and in chfldhood. Studies have consistently reported a significant association between prenatal smoking and risk of childhood overweight.77’ '0' Regardless of the methods employed to measure prenatal smoking, effect sizes for the odds of child overweight/obesity have ranged from ~13 to 3.0, with most studies showing modest relationships (e.g. OR’s = 1.5-2.0).64’93’10"102 Data from the Pregnancy Nutrition Surveillance System and the Pediatric Nutrition Surveillance System covering 9 US states and 2 tribal nations indicate that the effect of maternal smoking on offspring obesity may be modified by race/ethnicity.103 In this study, smoking during pregnancy was associated with increased risk of child obesity at 2-4 yrs of age among Non-Hispanic White (OR=1.42, 95%CI: 1.34-1.50) and Non-Hispanic Black (OR=1.19, 95%CI: 1.05-1.35) children, but not among Hispanics, Native Americans, or Asians.103 Dubois et al63 also noted a significant interaction between smoking during pregnancy and infant weight gain such that children born to smoking mothers who were macrosomic and experienced the highest rate of infant growth were at similarly high risk of obesity at 4.5 yrs as were children born to smoking mothers with normal birth weight who also experienced the highest rate of infant weight gain. The biological mechanism linking smoking to later obesity is unclear. It has been hypothesized that maternal smoking leads to fetal undemutrition due to uteroplacental 93 \BSOCOII Alternat oxidant patterns expiana disadta meme logical bemee Maren; Progra- size Cc materr alIEI‘a' risk fl 11mm. vasoconstriction, which is followed by excessive “catch-up” growth postnatally.” '0' Alternatively, carbon monoxide associated with maternal smoking may affect fetal oxidative metabolism, thereby imprinting metabolic pathways and leading to differential patterns of fat metabolism/storage in the offspring.” '0‘ Aside from these biological explanations, it is also possible that maternal smoking simply marks the presence of other disadvantageous maternal behaviors/characteristics such as maternal under— or overweight and/or poor maternal nutrition. However, studies which have controlled for logical confounding factors have demonstrated a persistent, if modest, relationship between smoking during pregnancy and larger offspring size in childhood. 10' Maternal Nutrition during Pregnancy and Child Size Fetal filel supply is a key determinant of birth weight and appears to alter programming of metabolic pathways, thereby influencing future offspring body size/composition. The consequences of fetal over-nutrition, as evident in the presence of maternal diabetes and/or obesity during pregnancy, have been outlined above. The strongest evidence for a lasting impact of fetal under-nutrition comes from follow-up studies on the Dutch Famine cohort. Among 19-yr old men, risk of obesity was significantly higher among those exposed to famine conditions in the 1st or 2"‘1 trimester (OR=1.94, 95%CI: 1.54-2.44), but significantly lower among those exposed only in the 3rd trimester (OR=0.62, 95%CI: 0.44-0.stO4 These results indicate that the timing of alterations in maternal nutrition may have important consequences in terms of offspring risk for obesity. Past research has focused on the extremes of maternal nutrition (under- or over- nutrition) and offspring outcomes.“ 38’ '0' Studies on maternal dietary 94 quah resul qaah nuni exur cons .lltlll mat: nze had (n=; int] Smt fitn and par Hes sen san fine Sub quality/composition during pregnancy and child health have been few with mixed results.105 More research with detailed measures of energy consumption and dietary quality throughout pregnancy are needed to more fully understand the effects of maternal nutrition on offspring health throughout the normal range of dietary habits, as well as the extremes. From the available data, it appears that variations to maternal energy consumption within the normal range do not influence future child weight status.38 Maternal Physical Activity during Pregnancy and Child Size To our knowledge, only one investigator has examined the relationship between maternal LTPA during pregnancy and child body size. James Clapp, MD examined body size and neurodevelopmental outcomes at 5 yrs of age among offspring of women who had either vigorously exercised both before and throughout pregnancy (n=20) or had regularly exercised pre-pregnancy but voluntarily stopped activity during pregnancy (n=20).4 In an attempt to isolate the effects of LTPA on offspring outcomes, participants in this case-control study were matched based on prenatal factors including maternal smoking status, income, education, marital status, parity, occupation, preconceptional fitness and LTPA participation, and maternal/patemal weight, height, and body fatness; and postnatal factors including breastfeeding duration, type of child care, parental LTPA participation post-pregnancy, postnatal maternal change in weight, child gender, gestational age at delivery, clinically normal growth patterns in the first year, absence of serious illness, and number of new siblings. This approach resulted in a highly select sample of women, but eliminated the need for adjusted analyses. Maternal exercise was measured prospectively throughout pregnancy, birth weight and fatness (tricep and subscapular skinfold thicknesses) were assessed by a trained investigator within 2 hours 95 ofbir and r pre :17 l, O o la Offs FPO. as it lasti W35 trir Chi ass lar U1] Ca of birth, and weight, height, and skinfolds (triceps, subscapular, suprailliac, abdominal, and mid-thigh) were re-assessed within 1 month of each child’s 5 yr birthday. Results demonstrated that offspring of women who exercised throughout pregnancy were significantly lighter (BW= 3.4 :1: 8.0 v. 3.6 i: 7.0 kg, p=0.01) and leaner (%fatr— 10.5 i 0.9 vs. 15.1 :t 0.6, p=0.01) at birth compared to the control group.4 Offspring in the exercise group remained lighter (weight=18.0 i0.5 vs. 19.5 $0.6 kg, p=0.01)) and leaner (Sum of 5 skinfolds = 37 i 1 vs. 44 i 2 mm, p=0.01) at 5 yrs of age as well. These results provide preliminary support that LTPA during pregnancy may have lasting effects on child body size; however, the mechanism of effect is unclear. Our lab is currently conducting a pilot study following up women whose LTPA was carefully measured during pregnancy to determine possible relationships with child health outcomes at 8-10 yrs.106 Preliminary data analyses on 20 mother/child pairs indicate that LTPA later in pregnancy tended to be inversely associated with child BMI (r,: -0.40, p=0.08; rs= -0.41, p=0.07 for the 2nd and 3rd trimesters, respectively) and with child waist circumference (rs: -0.45, p=0.05; rs= -0.43, p=0.06 for the 2nd and 3rd trimesters, respectively). No significant associations among LTPA during pregnancy and child percent body fat were noted; however, pre-pregnancy LTPA was inversely associated with child percent body fat (rs= -0.60, p=0.005). Future studies including larger and more diverse populations are needed to confirm these findings and begin to unravel the biological pathway involved. Investigating the effect of LTPA during pregnancy on child body size requires careful consideration of potentially confounding or modifying factors. As we have seen, maternal obesity, high pregnancy weight gain, and diabetes during pregnancy all increase 96 nsk fio kxeh acnte indent shouh unorr 8&de been] inond “Ink heave SeVer chddl quant loner Mdl“ of bre “mite Obesn' 118k 01 risk for high BW and childhood overweight status.“ 74' 94 Additionally, current LTPA levels of children are inversely associated with body fatness (r = -0.52, p<0.01)49 and active children gain significantly less body fat between 3-5 years of age compared to inactive children.44 Thus, the possible impact of maternal LTPA on child body size should be evaluated in light of the current activity level of the child as well as descriptive information on maternal body size and pregnancy conditions. Breastfeeding and Child Body Size In addition to the prenatal exposures outlines above, much research interest has been paid to the possible effects of infant feeding practices on future child weight status. The American Academy of Pediatrics recommend exclusive breastfeeding for the first six months of life and support breastfeeding for the first year and beyond when possible. '07 While breastfeeding is known to confer several health benefits to the infant, associations between breastfeeding and future risk of obesity have shown contradictory results. Several cohort studies have noted small protective effects of breastfeeding on the risk of 64, 77,111-113 Recently ’ childhood obesity;'°8" ‘0 however, others have reported null findings. quantile regression analyses has demonstrated that breastfeeding had a larger effect of lowering BMI above the 90‘h percentile of the BMI distribution among children 5-7 years old.109 While these findings require confirmation, they suggest that the protective effect of breastfeeding on childhood obesity may not always be demonstrated because it is limited to the upper parts of the BMI distribution. Duration of breastfeeding appears to show dose-response relationships with obesity risk, with studies showing that infants breastfed more than 4 or 6 mo had lower risk of obesity than those breastfed for a shorter time or not at 311.24‘ '08 Unfortunately, 97 previous studies have rarely accounted for other maternal factors that may influence associations between breastfeeding and childhood obesity, such as smoking, obesity, and socioeconomic status, and none have considered the impact of interim dietary and LTPA behaviors of the child.24 Those studies which have included maternal size along with breastfeeding have noted that maternal BMI and/or overweight status is a stronger risk factor for childhood obesity than is breastfeeding.I I l’ ”4 It has been hypothesized that breastfeeding improves the child’s ability to self- regulate dietary intake, while the rapid weight gain associated with formula feeding in early infancy confers added risk of future obesity. It also appears that breastfeeding may alter future maternal feeding patterns. In particular, women who breastfed 12 months or longer demonstrated lower levels of control over food choices and frequency of eating and were more responsive to hunger/satiation cues from their toddlers when compared to women whose infants were formula fed.1 '5' “6 Thus, breastfeeding may have indirect influences on child obesity via changes to maternal and child food intake behaviors. Overall, it appears that breastfeeding offers some protection against later child obesity; however other perinatal factors including maternal size and infant growth rates appear to have larger influences on child growth and body size. SUMMARY Converging lines of epidemiological and animal study evidence support the roles of both genetic and postnatal environmental factors on the risk of childhood obesity, as well as prenatal exposures. Causal links between perinatal exposures and future risk of childhood overweight/obesity are beginning to be established. Prospective cohort studies on exposures occurring prenatally and in early infancy have demonstrated clear temporal 98 rdauo expos wendr aheral oflspr nnok‘ causa popul (habe pregr dune dust denn assoc nuns Find With amen aCor QUICl fOl’ r. relationships without reliance on recalled data which may be biased. '01 For some exposures such as maternal diabetes, obesity during pregnancy, and excessive gestational weight gain, animal studies have shown biologically plausible pathways between alterations to fetal nutrition and changes to metabolic pathways which set the stage for offspring obesity.35 Biological mechanisms are less clear for exposures such as maternal smoking and/or LTPA during pregnancy to directly impact child weight status. The third causal criteria of consistency of findings has been demonstrated across age-ranges and populations for child weight status to be associated with maternal overweight and/or diabetes during pregnancy, excessive gestational weight gain, smoking during pregnancy, and breastfeeding with few exceptions.3' 74' 77 Consistency has not been demonstrated for maternal LTPA during pregnancy as only one study has yet examined this topic.4 Some factors, such as maternal smoking and breastfeeding have also demonstrated dose-response relationships with future child size. However, the strength of association for perinatal factors has varied, with most showing modest associations and maternal overweight/obesity demonstrating the strongest effect on child obesity. ’0' Findings for these perinatal exposures appear to demonstrate a high degree of coherence with existing theory and knowledge; however, unfortunately studies have varied in their attention to confounding factors. Thus alternative explanations, including the existence of a common environment that led to both the exposure (e.g., matemallobesity) and the outcome of child obesity cannot be i gnored.IOI In total, the perinatal factors reviewed here fulfill the majority of causal criteria for relationship to child overweight/obesity status; however, more work is needed to 99 reflnt Paru nnpr beeo refine biological pathways while controlling for important confounding variables. Particularly, LTPA during pregnancy deserves more attention as a possible avenue to improve the health of the mother while also possibly decreasing risk of the child becoming overweight later in life. 100 RE 11] ll 12 REFERENCES 10. 11. 12. Alderman BW, Zhao H, Holt VL, Watts DH, Beresford SA. Maternal physical activity in pregnancy and infant size for gestational age. Ann Epidemiol. Nov 1998;8(8):513-519. Voldner N, Froslie KF, Bo K, et al. Modifiable determinants of fetal macrosomia: role of lifestyle-related factors. A cta Obstet Gynecol Scand. Apr 2008;87(4):423- 429. Simmons R. Perinatal programming of obesity. Semin Perinatol. Oct 2008;32(5):371-374. Clapp J F . The morphometric and neurodevelopmental outcome at five years of the offspring of women who continued exercise throughout pregnancy. J Pediatr. Dec 1996;129(6):856-863. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat 11. May 2002(246):1- 190. Barlow SE. Expert committee recommendations regarding the prevention, assessment, and treatment of child and adolescent overweight and obesity: summary report. Pediatrics. Dec 2007;120(Suppl 4):S164-192. F legal KM, Tabak CJ, Ogden CL. Overweight in children: definitions and interpretation. Health Educ Res. Dec 2006;21(6):755-760. Reilly JJ, Methven E, McDowell ZC, et al. Health consequences of obesity. Arch Dis Child. Sep 2003;88(9):748-752. Reilly JJ. Descriptive epidemiology and health consequences of childhood obesity. Best Pract Res Clin Endocrinol Metab. Sep 2005; 19(3):327-341. Dubois L, Girad M. Accuracy of maternal reports of pre-schoolers' weights and heights as estimates of BMI values. Int J Epidemiol. Feb 2007;36(1):l32-138. Mamun AA, McDermott BM, O'Callaghan MJ, Najman J M, Williams GM. Predictors of maternal misclassifications of their offspring's weight status: a longitudinal study. Int J Obes (Land). Jan 2008;32(1):48-54. Davis H, Gergen PJ. Mexican-American mothers' reports of the weights and heights of children 6 months through 11 years old. J Am Diet Assoc. May 1994;94(5):512-516. 101 l3. 14. 15. l6. l7. l8. 19. 20. 21. 22. 23. 24. Huybrechts 1, De Bacquer D, Van Trimpont 1, De Backer G, De Henauw S. Validity of parentally reported weight and height for preschool-aged children in Belgium and its impact on classification into body mass index categories. Pediatrics. Nov 2006;1 18(5):2109-21 l8. Scholtens S, Brunekreef B, Visscher TL, et al. Reported versus measured body weight and height of 4-year-old children and the prevalence of overweight. Eur J Public Health. Aug 2007;17(4):369-374. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, F legal KM. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002. Jama. Jun 16 2004;291(23):2847-2850. Chinn S, Rona RJ. Prevalence and trends in overweight and obesity in three cross sectional studies of British Children, 1974-94. ij. Jan 2001;322(7277):24-26. Romon M, Duhamel A, Collinet N, Weill J. Influence of social class on time trends in BMI distribution in 5-year-old French children from 1989 to 1999. Int J Obes (Land). Jan 2005;29(1):54-59. Kautiainen S, Rimpela A, Vikat A, Virtanen SM. Secular trends in overweight and obesity among Finnish adolescents in 1977-1999. Int J Obes Relat Metab Disord. Apr 2002;26(4):544-552. Kalies H, Lenz J, von Kries R. Prevalence of overweight and obesity and trends in body mass index in German pre-school children, 1982-1997. Int J Obes Relat Metab Disord. Sep 2002;26(9): 121 1-1217. Wang Y, Monteiro C, Popkin BM. Trends of obesity and underweight in older children and adolescents in the United States, Brazil, China, and Russia. Am J Clin Nutr. Jun 2002;75(6):971-977. Magarey AM, Daniels LA, Boulton TJ. Prevalence of overweight and obesity in Australian children and adolescents: reassessment of 1985 and 1995 data against new standard international definitions. Med J Aust. Jun 2001;174(11):561-564. Troiano RP, F legal KM. Overweight children and adolescents: description, epidemiology, and demographics. Pediatrics. Mar 1998;101(3 Pt 2):497-504. Ogden CL, Carroll MD, F legal KM. High body mass index for age among US children and adolescents, 2003-2006. Jama. May 2008;299(20):2401-2405. Adair LS. Child and adolescent obesity: epidemiology and developmental perspectives. Physiol Behav. Apr 2008;94(1):8-16. 102 wt. 28. 30. 31. 33. 34. 3S. 36. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. Edmunds LS, Woelfel ML, Dennison BA, Stratton H, Pruzek RM, Abusabha R. Overweight trends among children enrolled in the New York State special supplemental nutrition program for women, infants, and children. J Am Diet Assoc. Jan 2006;106(1):1 13-1 17. Wang Y, Zhang Q. Are American children and adolescents of low socioeconomic status at increased risk of obesity? Changes in the association between overweight and family income between 1971 and 2002. Am J Clin Nutr. Oct 2006;84(4):707- 716. Jolliffe D. Extent of overweight among US children and adolescents from 1971 to 2000. Int J Obes Relat Metab Disord. Jan 2004;28(1):4-9. Daniels SR. The consequences of childhood overweight and obesity. Future Child. Spring 2006;16(1):47-67. Eisenmann J C. Secular trends in variables associated with the metabolic syndrome of North American children and adolescents: a review and synthesis. Am J Hum Biol. Nov-Dec 2003;15(6):786-794. Eisenmann J C. On the use of a continuous metabolic syndrome score in pediatric research. Cardiovasc Diabetol. Jun 2008;7:17. Bell LM, Byme S, Thompson A, et al. Increasing body mass index z-score is continuously associated with complications of overweight in children, even in the healthy weight range. J Clin Endocrinol Metab. Feb 2007;92(2):517-522. Singh AS, Mulder C, Twisk JW, van Mechelen W, Chinapaw MJ. Tracking of childhood overweight into adulthood: a systematic review of the literature. Obes Rev. Sep 2008;9(5):474-488. Freedman DS, Khan LK, Dietz WH, Srinivasan SR, Berenson GS. Relationship of childhood obesity to coronary heart disease risk factors in adulthood: the Bogalusa Heart Study. Pediatrics. Sep 2001;108(3):712-718. Freedman DS, Patel DA, Srinivasan SR, et al. The contribution of childhood obesity to adult carotid intima-media thickness: the Bogalusa Heart Study. Int J Obes (Lond). May 2008;32(5):749-756. Cottrell EC, Ozanne SE. Early life programming of obesity and metabolic disease. Physiol Behav. Apr 2008;94(1):17-28. Baird J, Fisher D, Lucas P, Kleijnen J, Roberts H, Law C. Being big or growing fast: systematic review of size and growth in infancy and later obesity. ij. Oct 2005;331(7522):929. 103 38. 39. 40. 41. ~15. 46. 47. 48. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. Parsons TJ, Power C, Manor 0. Fetal and early life grth and body mass index from birth to early adulthood in 1958 British cohort: longitudinal study. ij. Dec 2001;323(7325):1331-l335. Wells J C, Chomtho S, F ewtrell MS. Programming of body composition by early growth and nutrition. Proc Nutr Soc. Aug 2007;66(3):423-434. Rennie KL, Wells JC, McCaffrey TA, Livingstone MB. The effect of physical activity on body fatness in children and adolescents. Proc Nutr Soc. Nov 2006;65(4):393—402. U.S. Physical Activity Guidelines for Americans. US. Department of Health and Human Services. http://www.health.gov/paguidelines: ODPHP Publication No. U0036; 2008. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. Jan 2008;40(1):181-188. Hills AP, King NA, Armstrong TP. The contribution of physical activity and sedentary behaviours to the growth and development of children and adolescents: implications for overweight and obesity. Sports Med. 2007;37(6):533-545. Campbell K, Waters E, O'Meara S, Summerbell C. Interventions for preventing obesity in childhood. A systematic review. Obes Rev. Aug 2001 ;2(3): 149-157. Moore LL, Nguyen US, Rothman KJ, Cupples LA, Ellison RC. Preschool physical activity level and change in body fatness in young children. The Framingham Children's Study. Am J Epidemiol. Nov 1995;142(9):982-988. Moore LL, Gao D, Bradlee ML, et al. Does early physical activity predict body fat change throughout childhood? Prev Med. Jul 2003;37(1):10-17. Trost SG, Sirard JR, Dowda M, Pfeiffer KA, Pate RR. Physical activity in overweight and nonoverweight preschool children. Int J Obes Relat Metab Disord. Jul 2003;27(7):834-839. Abbott RA, Davies PS. Habitual physical activity and physical activity intensity: their relation to body composition in 5.0-10.5-y-old children. Eur J Clin Nutr. Feb 2004;58(2):285-291. Rennie KL, Livingstone MB, Wells JC, et al. Association of physical activity with body-composition indexes in children aged 6-8 y at varied risk of obesity. Am J Clin Nutr. Jul 2005;82(1):13-20. 104 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. Davies PS, Gregory J, White A. Physical activity and body fatness in pre-school children. IntJ Obes Relat Metab Disord. Jan 1995;l9(l):6-10. Maffeis C, Talamini G, Tato L. Influence of diet, physical activity and parents' obesity on children's adiposity: a four-year longitudinal study. Int J Obes Relat Metab Disord. Aug l998;22(8):758-764. Burdette HL, Whitaker RC. A national study of neighborhood safety, outdoor play, television viewing, and obesity in preschool children. Pediatrics. Sep 2005;1 16(3):657-662. Kropski JA, Keckley PH, Jensen GL. School-based obesity prevention programs: an evidence-based review. Obesity (Silver Spring). May 2008;16(5): 1009-1018. DeMattia L, Lemont L, Meurer L. Do interventions to limit sedentary behaviours change behaviour and reduce childhood obesity? A critical review of the literature. Obes Rev. Jan 2007;8(1):69-81. Lumeng J C, Rahnama S, Appugliese D, Kaciroti N, Bradley RH. Television exposure and overweight risk in preschoolers. Arch Pediatr Adolesc Med. Apr 2006;160(4):417-422. American Academy of Pediatrics: Children, adolescents, and television. Pediatrics. Feb 2001;107(2):423-426. Gidding SS, Dennison BA, Birch LL, et al. Dietary recommendations for children and adolescents: a guide for practitioners. Pediatrics. Feb 2006;1l7(2):544-559. Troiano RP, Briefel RR, Carroll MD, Bialostosky K. Energy and fat intakes of children and adolescents in the United States: data from the National Health and Nutrition Examination Surveys. Am J Clin Nutr. Nov 2000;72(5 Suppl):l343S- 1353s. Rodriguez G, Moreno LA. Is dietary intake able to expalin differences in body fatness in children and adolescents? Nutr Metab Cardiovasc Dis. May 2006;16(4):294-301. Gibson S. Sugar-sweetened soft drinks and obesity: a systematic review of the evidence from observational studies and interventions. Nutr Res Rev. Dec 2008;21(2):134-147. Bandini LG, Schoeller DA, Cyr HN, Dietz WH. Validity of reported energy intake in obese and nonobese adolescents. Am J Clin Nutr. Sep 1990;52(3):421- 425. 105 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. Barba G, Troiano E, Russo P, Venezia A, Siani A. Inverse association between body mass and frequency of milk consumption in children. Br J Nutr. Jan ' 2005;93(1):15-19. Davey Smith G, Steer C, Leary S, Ness A. Is there an intrauterine influence on obesity? Evidence from parent child associations in the Avon Longitudinal Study of Parents and Children (ALSPAC). Arch Dis Child. Oct 2007;92(lO):876-880. Dubois L, Girard M. Early determinants of overweight at 4.5 years in a population-based longitudinal study. Int J Obes (Lond). Apr 2006;30(4):610-617. Moschonis G, Grammatikaki E, Manios Y. Perinatal predictors of overweight at infancy and preschool childhood: the GENESIS study. Int J Obes (Lond). Jan 2008;32(1):39-47. Bouchard C. Current understanding of the etiology of obesity: genetic and nongenetic factors. Am J Clin Nutr. Jun 1991;53(6 Suppl): 1561S-156SS. Perusse L, Bouchard C. Role of genetic factors in childhood obesity and in susceptibility to dietary variations. Ann Med. Apr 1999;31(Suppl 1):19-25. Butte NF, Cai G, Cole SA, Comuzzie AG. Viva la Familia Study: genetic and. environmental contributions to childhood obesity and its comorbidities in the Hispanic population. Am J Clin Nutr. Sep 2006;84(3):646-654. Moore J, Harre N. Eating and activity: the importance of family and environment. Health Promot J Austr. Aug 2007;18(2): 143-148. Moore LL, Lombardi DA, White MJ, Campbell JL, Oliveria SA, Ellison RC. Influence of parents' physical activity levels on activity levels of young children. J Pediatr. Feb 1991;118(2):215-219. Clark HR, Goyder E, Bissell P, Blank L, Peters J. How do parents' child-feeding behaviours influence child weight? Implications for childhood obesity policy. J Public Health (Oxf). Jun 2007;29(2): 132-141. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of overweight and obesity in the United States, 1999-2004. Jama. Apr 2006;295(13):1549-1555. Kumanyika SK. Environmental influences on childhood obesity: ethnic and cultural influences in context. Physiol Behav. Apr 2008;94(1):61-70. Griffiths LJ, Dezateux C, Cole TJ. Differential parental weight and height contributions to offspring birthweight and weight gain in infancy. Int J Epidemiol. Feb 2007;36(1): 104-107. 106 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. Whitaker RC, Dietz WH. Role of the prenatal environment in the development of obesity. J Pediatr. May 1998;132(5):768-776. Oken E, Gillman MW. Fetal origins of obesity. Obes Res. Apr 2003;11(4):496- 506. Rogers 1. The influence of birthweight and intrauterine environment on adiposity and fat distribution in later life. Int J Obes Relat Metab Disord. Jul 2003;27(7):755-777. Snethen J A, Hewitt J B, Goretzke M. Childhood obesity: the infancy connection. J Obstet Gynecol Neonatal Nurs. Sep-Oct 2007;36(5):501-510. Catalano PM, Drago NM, Amini SB. Factors affecting fetal growth and body composition. Am J Obstet Gynecol. May 1995;172(5):1459—1463. Catalano PM, Thomas A, Huston-Presley L, Amini SB. Increased fetal adiposity: a very sensitive marker of abnormal in utero development. Am J Obstet Gynecol. Dec 2003;189(6):1698-1704. Hammami M, Walters JC, Hockman EM, Koo WW. Disproportionate alterations in body composition of large for gestational age neonates. J Pediatr. Jun 2001;138(6):817-821. Hediger ML, Overpeck MD, McGlynn A, Kuczmarski RJ, Maurer KR, Davis WW. Growth and fatness at three to six years of age of children born small- or large-for-gestational age. Pediatrics. Sep 1999;104(3):e33. Boney CM, Verma A, Tucker R, Vohr BR. Metabolic syndrome in childhood: association with birth weight, maternal obesity, and gestational diabetes mellitus. Pediatrics. Mar 2005;1 15(3):e290-296. Gillman MW, Rifas-Shiman S, Berkey CS, Field AE, Colditz GA. Maternal gestational diabetes, birth weight, and adolescent obesity. Pediatrics. Mar 2003;111(3):e22l-226. Hediger ML, Overpeck MD, Kuczmarski RJ, McGlynn A, Maurer KR, Davis WW. Muscularity and fatness of infants and young children born small- or large- for-gestational-age. Pediatrics. Nov 1998; 1 02(5):E60. Rogers IS, Ness AR, Steer CD, et al. Associations of size at birth and dual-energy X-ray absorptiometry measures of lean and fat mass at 9 to 10 y of age. Am J Clin Nutr. Oct 2006;84(4):739-747. 107 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. Elia M, Betts P, Jackson DM, Mulligan J. Fetal programming of body dimensions and percentage body fat measured in prepubertal children with a 4-component model of body composition, dual-energy X-ray absorptiometry, deuterium dilution, densitometry, and skinfold thicknesses. Am J Clin Nutr. Sep 2007;86(3):618-624. Evagelidou EN, Kiortsis DN, Bairaktari ET, et al. Lipid profile, glucose homeostasis, blood pressure, and obesity-anthropometric markers in macrosomic offspring of nondiabetic mothers. Diabetes Care. Jun 2006;29(6):1197-1201. Chomtho S, Wells JC, Williams J E, Lucas A, F ewtrell MS. Associations between birth weight and later body composition: evidence from the 4-component model. Am J Clin Nutr. Oct 2008;88(4):1040-1048. Labayen I, Moreno LA, Blay MG, et al. Early programming of body composition and fat distribution in adolescents. J Nutr. Jan 2006;136(1):147-152. Frederick 10, Williams MA, Sales AE, Martin DP, Killien M. Pre-pregnancy Body Mass Index, Gestational Weight Gain, and Other Maternal Characteristics in Relation to Infant Birth Weight. Matern Child Health J. Sep 2008;12(5):557- 567. Hull HR, Dinger MK, Knehans AW, Thompson DM, Fields DA. Impact of maternal body mass index on neonate birthweight and body composition. Am J Obstet Gynecol. Apr 2008;198(4):416 e41 1-416. Gale CR, Javaid MK, Robinson SM, Law CM, Godfrey KM, Cooper C. Maternal size in pregnancy and body composition in children. J Clin Endocrinol Metab. Oct 2007;92(10):3904-391 1 . Gillman MW, Rifas-Shiman SL, Kleinman K, Oken E, Rich-Edwards JW, Taveras EM. Developmental origins of childhood overweight: potential public health impact. Obesity (Silver Spring). Jul 2008;16(7): 1651-1656. Whitaker RC. Predicting preschooler obesity at birth: the role of maternal obesity in early pregnancy. Pediatrics. Jul 2004;114(1):e29-36. Oken E, Taveras EM, Kleinman KP, Rich-Edwards JW, Gillman MW. Gestational weight gain and child adiposity at age 3 years. Am J Obstet Gynecol. Apr 2007;196(4):322 e321-328. Olson CM, Strawderrnan MS, Dennison BA. Maternal Weight Gain During Pregnancy and Child Weight at Age 3 Years. Matern Child Health J. Sep 2008; epub DOI: 10.1007/S10995-008-0413-6. 108 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. Dabelea D, Hanson RL, Lindsay RS, et al. Intrauterine exposure to diabetes conveys risks for type 2 diabetes and obesity: a study of discordant sibships. Diabetes. Dec 2000;49(12):2208-221 1 . Whitaker RC, Pepe MS, Seidel KD, Wright JA, Knopp RH. Gestational diabetes and the risk of offspring obesity. Pediatrics. Feb 1998;101(2):E9. Catalano PM, Ehrenberg HM. The short- and long-term implications of maternal obesity on the mother and her offspring. Bjog. Oct 2006;113(lO):1126-1133. Catalano PM, Kirwan JP, Haugel-de Mouzon S, King J. Gestational diabetes and insulin resistance: role in short- and long-term implications for mother and fetus. J Nutr. May 2003;133(5 Suppl 2):1674S-l683S. Huang J S, Lee TA, Lu MC. Prenatal programming of childhood overweight and obesity. Matern Child Health J. Sep 2007;11(5):461-473. Chen A, Pennell ML, Klebanoff MA, Rogan WJ, Longnecker MP. Maternal smoking during pregnancy in relation to child overweight: follow-up to age 8 years. Int J Epidemiol. Feb 2006;35(l):121-130. Shanna AJ, Cogswell ME, Li R. Dose-response associations between maternal smoking during pregnancy and subsequent childhood obesity: effect modification by maternal race/ethnicity in a low-income US cohort. Am J Epidemiol. Nov 2008;168(9):995-1007. Ravelli GP, Stein ZA, Susser MW. Obesity in young men after famine exposure in utero and early infancy. N Engl J Med. Aug 1976;295(7):349—353. Moore VM, Davies MJ. Diet during pregnancy, neonatal outcomes and later health. Reprod F ertil Dev. 2005;l7(3):341-348. Pivamik JM, Stein AD, Rivera J M. Effect of pregnancy on heart rate/oxygen consumption calibration curves. Med Sci Sports Exerc. May 2002;34(5):750-755. Gartner LM, Morton J, Lawrence RA, et al. Breastfeeding and the use of human milk. Pediatrics. Feb 2005;115(2):496-506. Grummer-Strawn LM, Mei Z. Does breastfeeding protect against pediatric overweight? Analysis of longitudinal data from the Centers for Disease Control and Prevention Pediatric Nutrition Surveillance System. Pediatrics. Feb 2004;113(2):e81-86. Beyerlein A, Toschke AM, von Kries R. Breastfeeding and Childhood Obesity: Shift of the Entire BMI Distribution or Only the Upper Parts? Obesity (Silver Spring). Dec 2008;16(12):2730-2733. 109 110. 111. 112. 113. 114. 115. 116. Bergmann KE, Bergmann RL, Von Kries R, et al. Early determinants of childhood overweight and adiposity in a birth cohort study: role of breast-feeding. Int J Obes Relat Metab Disord. Feb 2003;27(2): 162-172. Buyken AE, Karaolis-Danckert N, Gunther A, Kersting M. Effects of breastfeeding on health outcomes in childhood: beyond dose-response relations. Am J Clin Nutr. Jun 2008;87(6): 1964-1965; author reply 1965-1966. Hediger ML, Overpeck MD, Kuczmarski RJ, Ruan WJ. Association between infant breastfeeding and overweight in young children. Jama. May 2001 ;285(19):2453-2460. Hediger ML, Overpeck MD, Ruan WJ, Troendle JF. Early infant feeding and growth status of US-born infants and children aged 4-71 mo: analyses from the third National Health and Nutrition Examination Survey, 1988-1994. Am J Clin Nutr. Jul 2000;72(1):159-167. Li C, Kaur H, Choi WS, Huang TT, Lee RE, Ahluwalia J S. Additive interactions of maternal prepregnancy BMI and breast-feeding on childhood overweight. Obes Res. Feb 2005;13(2):362-371. Taveras EM, Scanlon KS, Birch L, Rifas-Shiman SL, Rich-Edwards JW, Gillman MW. Association of breastfeeding with maternal control of infant feeding at age 1 year. Pediatrics. Nov 2004;114(5):e577-583. Fisher JO, Birch LL, Smiciklas-Wright H, Picciano MF. Breast-feeding through the first year predicts maternal control in feeding and subsequent toddler energy intakes. J Am Diet Assoc. Jun 2000;100(6):641-646. llO CHAPTER 5: MATERNAL PHYSICAL ACTIVITY DURING PREGNANCY, CHILD SEDENTARY BEHAVIOR, AND CHILD WEIGHT STATUS ABSTRACT Purpose: To evaluate the effects of maternal leisure-time physical activity (LTPA) during pregnancy and child sedentary behavior on child weight status at 3-9 years. Methods: Women enrolled in the Pregnancy Outcomes and Community Health Study (1998-2004) were studied again in 2007. F ollow-up efforts were extensive for a subcohort and minimal for the remainder, or non-subcohort. Original data collection provided maternal demographics and birth outcomes. At follow-up women reported child height and weight (used to calculate body mass index (BMI)), child leisure-time behavior (sedentary, some sedentary/some active, or active), and recalled pregnancy LTPA. Women were classified as inactive (reference category), insufficiently active, or meeting LTPA recommendations. A four-category matemal/child LTPA variable was created (reference: any LTPA during pregnancy and non-sedentary child leisure-time behavior). Children were classified as healthy weight (5-<85th), overweight (85-<95th), or obese (Z95th) based on age- and sex-specific percentiles. Results: Among the non-subcohort, LTPA during pregnancy were associated with borderline significant reduced odds of child overweight and obese status by approximately 50%. Among the subcohort, LTPA during pregnancy significantly increased odds of child overweight (aOR= 1.9 and 2.1 for insufficient activity and meeting LTPA recommendations) but non-significantly decreased odds of obesity (aOR=0.7 for both insufficient activity and meeting LTPA recommendations). Maternal inactivity during pregnancy combined with sedentary child 111 behavior tended to increase odds of obesity within both the non—subcohort (aOR=2.5, 95%CI: 0.8-7.4) and subcohort (aOR=l.8, 95%CI: 0.8, 3.7). Conclusion: Maternal LTPA during pregnancy may have a lasting effect of reducing risk of child obesity, even when considering their children’s own sedentary behavior. 112 INTRODUCTION The prevalence of childhood obesity in the US. has increased steadily since the 1970’s, with current values indicating that 16% of US. children ages 2-19 yrs are obese (i.e., age and sex-specific body mass index (BMI) 2 the 95th percentile)" 2 Childhood obesity is associated with increased risk of concomitant high blood pressure, insulin resistance, and dyslipidemia, as well as risk for adult obesity and cardiovascular disease.3' 5 Thus for the current and future health of youth, it is imperative to determine modifiable risk factors for childhood obesity to aid prevention and/or intervention efforts. While the causes of childhood obesity are multifaceted and complex, prenatal exposures are increasingly being examined in relation to future child size. Maternal overweight/obese status during pregnancy, excess gestational weight gain, gestational diabetes, smoking during pregnancy, and extreme birth weight (low or high) have each been associated with an increased risk for child obesity or excess fatness.6'IO One case- control study has also demonstrated that offspring of women who remained sedentary throughout pregnancy had increased body fat at five years of age when compared to offspring of women who exercised throughout pregnancy. 1' These results suggest that maternal leisure-time physical activity (LTPA) during pregnancy may reduce offspring susceptibility for excess fat storage. However, this study was based on a small select sample and did not consider the impact of the children’s LTPA levels. Despite differing measurement methods, past studies have shown child LTPA participation to be inversely associated with body size, including body fatness.'2"4 In addition, longitudinal data support the hypothesis that LTPA in early childhood results in lower BMI and body fatness later in childhood.'5 However, studies examining the 113 relationship between child LTPA participation and body size have rarely considered prenatal exposures, and none have evaluated possible influences of maternal LTPA during pregnancy. The purpose of this study, therefore, was to evaluate the separate and combined effects of maternal participation in LTPA during pregnancy and children’s leisure-time behavior on child weight status at 3-9 years of age. We hypothesized that LTPA during pregnancy would be associated with reduced odds of child overweight and obese status. We also hypothesized that children who engage in mostly sedentary leisure-time activities and whose mothers were inactive during pregnancy would have the highest risk for being classified overweight or obese when compared to children who choose non- sedentary leisure-time activities and whose mothers were active during pregnancy. METHODS Study Population Participants in this study were enrolled in the Pregnancy Outcomes and Community Health (POUCH) Study. The POUCH study originally recruited women in gestational weeks 15-27 from l998-2004 from 52 clinics in five Michigan communities. Inclusion criteria were singleton pregnancy with no known congenital anomaly, maternal age 215 years, maternal serum alpha-fetoprotein screen in gestational weeks 15-22, no pre-existing diagnoses of diabetes mellitus, and proficiency in English. Women with unexplained high alpha-fetoprotein levels (2 2 multiples of the median) were oversampled due to a particular interest in this biomarker for the original study aims. '6 Of the 3,038 women enrolled, 19 were lost to follow-up at birth leaving a cohort of 3,019 mother-child pairs. The POUCH study was approved from institutional review boards at 114 Michigan State University, Michigan Department of Community Health, and nine community hospitals. A subcohort of women was selected for more detailed study to maximize resources when evaluating original study aims. The subcohort included all women who delivered preterm (<37 weeks), women who delivered at term but had unexplained high maternal serum alpha-fetoprotein levels, and a race-stratified sample of women with term deliveries and normal maternal serum alpha-fetoprotein levels (i.e., 72% Afiican American and 23% White/other women in this category). Women enrolled in the POUCH study but not meeting these criteria comprised the “non-subcohort”. Women in the subcohort were contacted periodically for different follow-up studies (2005-2006; 2006-2007), while the non-subcohort received minimal contact. In fall of 2007, follow-up surveys on LTPA during pregnancy and child health outcomes were sent to all POUCH study participants who had not declined filrther contact after delivery and whose children were living with them (n=1629 non-subcohort; n=1261 subcohort). Women in the non-subcohort were sent a single mailing which asked them to complete and return an enclosed survey. No further contact was attempted for the non-subcohort. Women in the subcohort were sent the same mailing; however, phone contact was attempted to encourage participation, and follow-up mailings were sent. For this investigation, women who delivered preterm < 37 weeks of gestation were excluded (non-subcohort n=0, subcohort n=335). A total of 1200 non-subcohort and 299 subcohort women failed to return the follow-up survey, and there was incomplete follow-up information on LTPA during pregnancy (non-subcohort n=11, subcohort n=31) or current child height and weight (non-subcohort n=0, subcohort n=4) among the 115 me int: inte pres self thro unle in \r Welt the r Offs 0r Ll 10““ (lbs), calcu CalCul respondents. Thus the sample for this investigation included 418 (26% of eligible) non- subcohort and 592 (67% of eligible) subcohort women. Study Protocol The POUCH study has been described elsewhere in detail. '6 Briefly, participants met with a study nurse at enrollment to sign consent forms, completed an in-person interview and a self-administered survey, and had biological samples collected. The interview and self-administered survey provided information on maternal size, race/ethnicity, education, marital status, enrollment in Medicaid, and smoking at mid- pregnancy. Maternal pre-pregnancy body mass index (BMI, kg/mz) was calculated from self-reported pre-pregnancy weight and height values. Birth weight, child gender, maternal age at delivery, and parity were determined through chart review. Gestational age was calculated using the last menstrual period unless it disagreed by > 2 weeks with ultrasound conducted prior to 25 weeks gestation, in which case the ultrasound value was used. Sex and gestational-age specific birth weight z-scores were calculated as the observed minus the mean birth weight divided by the population standard deviation using birth weight standards from Kramer et al.17 Offspring were classified as small-, appropriate, or large-for-gestational-age (SGA, AGA, or LGA) if their sex and gestational age-specific birth weights were 5 10'”, between the 10th and 90”, or 2 90’h percentile, respectively.’8 At follow-up, women reported their children’s current age, height (in), weight (lbs), and leisure-time behavior. Reported height and weight values were used to calculate child BMI (kg/m2). Age- and sex-specific BMI percentile values were calculated using the 2000 Centers for Disease Control and Prevention (CDC) grth 116 than pert sede acu P()‘ rnar can of} rvh ask int usi oil ()I pre charts.l9 Child BMI values that were biologically implausible according to CDC criteria (i.e., any BMI z-score value calculated as less than -4.0 or greater than +5.0 units) were excluded from further analyses (non-subcohort n=14; subcohort n=56). '9 Children were classified into weight status categories of “underweight” (<5th percentile), “healthy weight” (5th -< 85th percentile), “overweight” (85th -< 95th percentile), and “obese” (E: 95th percentile). Women reported their child’s choice of leisure-time behavior as “mostly 9, ‘6 sedentary , some sedentary and some physically active”, or “mostly physically active”.20 Women also reported gestational weight gain and LTPA participation for the POUCH pregnancy, as well as current height and weight (used to calculate current maternal BMI) and LTPA at follow-up. Women were classified into weight gain categories (less than, within, and greater than recommended) based on the 1990 Institute of Medicine recommendations according to their pre-pregnancy BMI.21 Women recalled whether they participated in any LTPA during their POUCH pregnancy. If so, they were asked to recall the type, average duration (min/d), and average frequency (dek) of up to two activities performed most often during a typical week while pregnant. To quantify intensity, metabolic equivalent (MET) values were assigned to each reported activity using the Compendium for Physical Activities.22 Women who reported at least 150 min/wk of at least moderate intensity LTPA (i.e., 2 3 METs) were classified as “meeting/exceeding LTPA recommendations” according to the American College of Obstetricians and Gynecologists (ACOG) recommendation for LTPA during pregnancy.23 Women reporting LTPA below this threshold were classified as “insufficiently active” while women reporting no LTPA were classified as “inactive”. 117 . ——.a‘ __ \Vome precr “Mal Cl'Q'd‘ thil. SCdl 8.5' dur 110 be S Women also reported their current (within the past month) LTPA levels and were classified in the same manner. Information on other domains of physical activity (i.e., occupational, household, etc) was not collected. In order to consider simultaneously the effects of maternal LTPA during pregnancy and child leisure-time behavior on child weight status, a categorical “Maternal/Child LTPA” variable was created. First, two dichotomous variables were created to classify women as participating in any vs. no LTPA during pregnancy and children as having sedentary vs. non-sedentary (includes “some active and some sedentary” and “mostly active”) leisure-time behavior. Participants were then classified as “any LTPA during pregnancy and non-sedentary child behavior” (+ / +), “any LTPA during pregnancy and sedentary child behavior” (+ / -), “inactivity during pregnancy and non-sedentary child behavior” (- / +), or “inactivity during pregnancy and sedentary child behavior” (- / -). Statistical Analyses Because of the sampling schemes employed to create the original POUCH study cohort and subcohort, and the differing follow-up strategies, primary analyses were conducted separately for non-subcohort and subcohort participants. All analyses were conducted using SAS version 9.1. Significance was set as a two-tailed Alpha level of S 0.05. Polytomous logistic regression analyses were used to assess the main effects of maternal LTPA during pregnancy (reference group = inactive) and child leisure-time behavior (reference group: sedentary) on child weight status (reference group = Healthy Weight). Separate polytomous logistic regression models then evaluated the association 118 between between maternal/child LTPA (reference group = + / +) and odds of child weight status. Stepwise regression methods were used to build adjusted models. The following variables were considered as covariates based on previous literature: maternal race (among the subcohort only), pre-pregnancy and current maternal BMI, gestational weight gain, maternal report of smoking during pregnancy, enrollment in Medicaid, relationship status, educational level, occupational status, child gender, and birth weight for gestational age group. Likelihood ratio tests were used for significance testing in building adjusted models. Any variable that altered parameter estimates more than 10% was examined as a potential confounder or mediator. RESULTS Within both the non-subcohort and subcohort, follow-up survey nonresponders included significantly more women who were Afiican American, younger age, had less than high school education, were single, enrolled in Medicaid and smoked during pregnancy, compared to participating women (data not shown). Among participating women, maternal characteristics differed substantially between the non-subcohort and subcohort (Table 5.1). While some differences were expected due to the sampling strategy used to create the subcohort (i.e., 4.5% vs. 30.6% African American women in the non-subcohort vs. subcohort in this sample), other differences may reflect self- selection in relation to the intensity of follow-up efforts. The subcohort included more women who were younger, less educated, single, enrolled in Medicaid, had higher BMI values pre-pregnancy and currently, smoked during pregnancy, and were inactive during pregnancy and currently. 119 ,M.w;_‘.—_, --. In Some child characteristics also differed by subcohort status (Table 5.2). The non- subcohort included fewer children who were born SGA and more born LGA compared to the subcohort. Mean child BMI was 16.3 i: 2.4 kg/m2 among the non-subcohort and 16.7 :t 2.8 kg/m2 among the subcohort, corresponding to BMI z-score values of 0.27 3: 1.45 and 0.39 i 1.57, respectively. Significantly more children in the subcohort were classified as obese (21.6%) compared to the non-subcohort (14.9%); however, the percent of children classified as underweight were equally represented in both the non-subcohort (10.6%) and subcohort (10.5%). Underweight children were excluded from further analyses due to the small number in this category and a primary interest in odds of overweight or obese status vs, healthy weight status. Child leisure-time behavior was similar between the non-subcohort and subcohort; however, combined maternal/child LTPA was different between groups. The subcohort had a higher percentage of “inactivity during pregnancy and sedentary child behavior” (- / -) and a lower percentage of “any LTPA during pregnancy and non- sedentary child behavior” (+ / +) compared to the non-subcohort (Table 5.2). Among the non-subcohort, being insufficiently active or meeting LTPA recommendations during pregnancy were each associated with ~50% reduced odds of child overweight and obese status (Table 5.3). These estimates were of borderline statistical significance, even after adjusting for child sedentary behavior and Medicaid enrollment. Being insufficiently active or meeting LTPA recommendations during pregnancy were also associated with non-significantly reduced odds of child obesity among the subcohort; however, LTPA during pregnancy significantly increased odds of child overweight status by about two-fold in this group. 120 Child leisure-time behavior was not significantly associated with child weight status in either group. Among the non-subcohort, being classified as “some sedentary/some active” or “mostly active” were associated with lower odds of child obesity, but non-significantly higher odds of child overweight. Within the subcohort, “some sedentary/some active” and “,mostly active" child behavior were related to lower odds of both child overweight and obese status, but these estimates were not significant in adjusted analyses. Similar to the main effect results, associations among the combined maternal/child LTPA variable and child overweight status varied between non-subcohort and subcohort participants (Table 5.4). Among the non-subcohort, maternal inactivity during pregnancy combined with either child non-sedentary (-/+) or sedentary behavior (-/-) was associated with non-significant increased odds of child overweight status (p > 0.05). In contrast, among the subcohort, maternal inactivity during pregnancy combined with non-sedentary child behavior was associated with significantly lower odds of child overweight (aOR=0.49, 95%CI: 0.27-0.89). The association between maternal inactivity during pregnancy combined with child sedentary behavior and child overweight status among the subcohort approached unity and was non-significant. Results for child obesity were more consistent among the non-subcohort and subcohort, with both groups displaying increased odds associated with maternal inactivity during pregnancy combined with either non-sedentary or sedentary child behavior (Table 5.4). In unadjusted analyses, odds of obesity were ~2.5-3.0 times higher when both the mother and child were classified as inactive/sedentary as compared to both being active/non-sedentary. Adjustment for maternal covariates attenuated results within both 121 the non-subcohort and subcohort to statistical non-significance, yet point estimates still indicated ~2.0-2.5 times the odds of child obesity when both the mother and child were classified as inactive/sedentary. Estimates for the effect of any maternal LTPA during pregnancy combined with child sedentary behavior (+/-) on child obesity were non- significant in both the non-subcohort and subcohort, but tended to be associated with increased odds among the non-subcohort. Since associations for the main effects of LTPA during pregnancy and child leisure-time behavior as well as the interactive effect of maternal/child LTPA on odds of child obesity were similar for the non-subcohort and subcohort, we considered combining these groups in secondary analyses. Formal tests for interaction between subcohort status and LTPA during pregnancy or maternal/child LTPA on odds of child obesity was non- significant. Thus the non-subcohort and subcohort were combined to increase power and improve ability to consider covariates. Overweight children were excluded from these analyses due to the contrasting effects observed for the separate and combined effects of LTPA during pregnancy and child leisure-time behavior by subcohort status. Once children fiom the two groups were combined and overweight children were excluded, 25.0% (176/703) were classified as obese, with the remainder classified as healthy weight (reference group). Unadjusted results from the combined analyses showed ~50% reduced odds of child obesity associated with LTPA during pregnancy or with some sedentary/some active child leisure-time behavior and ~2.0-3.0 times the odds of child obesity associated with maternal inactivity during pregnancy combined with either non-sedentary or sedentary child behavior (Unadjusted Models, Table 5.5). Adjustment for maternal 122 race/ethnicity, Medicaid enrollment, and pre-pregnancy BMI attenuated these relationships to borderline statistical significance (Models 1 and 2, Table 5.5). Maternal activity during pregnancy combined with sedentary child behavior was also associated with increased odds of obesity, however, the confidence intervals were wide and included unity in both unadjusted and adjusted models. Appropriateness of birth weight for gestational age category (i.e., SGA, AGA, or LGA) was subsequently added to Model 2 and examined as a possible mediator of the relationship between LT PA during pregnancy and child obesity. The addition of this variable was not significant and did not change parameter estimates appreciably, suggesting that birth weight did not mediate the effect of maternal/child LTPA on child obesity (data not shown). No other covariate significantly entered the main effects or interaction models. Both separate (i.e., non-subcohort and subcohort) and combined analyses were repeated after removing women with unexplained high maternal serum alpha-fetoprotein levels, which is associated with increased risk of preterm delivery (n=0 non-subcohort, n=95 subcohort). The direction of relationships remained unchanged; however, parameter estimates for the effects of maternal/child LTPA on odds of child obesity became more extreme and remained statistically significant even when adjusted for maternal race/ethnicity, Medicaid enrollment, and pre-pregnancy BMI. Specifically, within combined adjusted analyses, increased odds of child obesity were observed for maternal inactivity during pregnancy combined with non-sedentary (aOR=1.52, 95% CI: 0.98- 2.33) or with sedentary child behavior (aOR=2.20, 95% CI: 1.11-4.35). 123 DISCUSSION Our purpose was to evaluate the separate and combined effects of maternal LTPA during pregnancy and child LTPA on child weight status at 3-9 years of age. We hypothesized that LTPA during pregnancy would be associated with reduced odds of child overweight and obesity and that odds of overweight and/or obesity would be highest when both the mother and child were classified as inactive/sedentary as compared to when both were active/non-sedentary. While results for child overweight status were mixed, being insufficiently active or meeting LTPA recommendations during pregnancy tended to reduce the odds of child obesity by 20-40% (p>0.05) within both the non- subcohort and the subcohort. Maternal inactivity during pregnancy combined with either child non-sedentary or sedentary behavior also tended to be associated with higher odds of child obesity in both groups. Although not statistically significant in adjusted models, odds of child obesity were highest when both the mother and child were classified as inactive/sedentary, whether evaluated separately within the non-subcohort and subcohort, or within the combined sample. Combining the non-subcohort and subcohort allowed us to examine the separate and combined effects of maternal LTPA and child leisure-time behavior on odds of obesity within a diverse sample of women while also considering important covariates. Adjustment for maternal race/ethnicity, Medicaid enrollment, and pre-pregnancy BMI status attenuated our results to borderline statistical significance. However, practically speaking, they are still suggestive of a protective main effect for maternal LTPA during pregnancy and for a protective interactive effect of maternal LTPA during pregnancy combined with non-sedentary child behavior on childhood obesity risk. Other covariates 124 which have been associated previously with child obesity, including gestational weight gain, maternal educational level, smoking during pregnancy, current maternal BMI, and parity, did not significantly alter our results and were not included in the final model. Past researchers have considered the effect of maternal or child LTPA participation on child weight status separately, but never combined as we have done. To our knowledge, James Clapp, MD conducted the only study to consider lasting effects of LTPA during pregnancy on child size.ll He found that offspring of women who exercised throughout pregnancy (n=20) had lower body fatness measures at 5 years of age compared to offspring of women who had exercised pre-pregnancy but voluntarily stopped activity during pregnancy (n=20). This study was conducted using a small, select group of women and demonstrated that LTPA during pregnancy could affect child body size/composition in the absence of confounding factors such as maternal overweight. Unfortunately no information on child LTPA or diet behaviors was collected. Our results for the main effect of LTPA during pregnancy corroborate Clapp’s findings. Although we did not have measures of body fatness, we found that participating in insufficient amounts of LTPA or meeting LTPA recommendations were associated with ~50% significantly reduced odds of child obesity within the non-subcohort and the subcohort. Adjustment for child leisure-time behavior and maternal characteristics attenuated these results to borderline statistical significance. The mechanism for the effect of LTPA during pregnancy on child body size is unclear. Clapp found the offspring of the more active women were also lighter and leaner at birth compared to the control group.II Other studies have reported that LTPA before and/or during pregnancy is associated with lower odds of giving birth to an LGA 125 infant.” 25 Thus, it is possible that maternal LTPA alters fetal growth to reduce fat stores at birth and may modify metabolic mechanisms throughout childhood, leading to less excess fat storage. It is also possible that women who choose to engage in LTPA during pregnancy provide a more physically active environment for their children, thereby influencing child growth and weight status. Neither hypothesis has been tested. Regardless of the mechanism involved, our results support a role for LTPA during pregnancy to affect child weight status, even when controlling for children’s current sedentary behavior. Additionally, birth size did not appear to mediate the relationship between maternalLTPA during pregnancy and child size in our study. In contrast to the paucity of data on LTPA during pregnancy and child weight status, child LTPA and/or sedentary behavior has been studied extensively in relation to weight status. Most studies report negative correlations between LTPA and fatness among children, while relationships between LTPA and BMI weight categories tend to be null.'2"5’ 26'” Studies on sedentary behavior among children have consistently reported improved weight indices associated with fewer hrs/wk spent in sedentary activity.30 It is possible that sedentary behavior increases risk of child overweight/obesity through displacing LTPA, increasing energy intake via snacking, effects of food advertising on TV, and/or decreased metabolic rate during sedentary time.30 Our results showed that children choosing some sedentary/some active or mostly active leisure-time behaviors had non-significantly reduced odds of child obesity within both the non-subcohort and the subcohort. Sedentary children with mothers who were active during pregnancy had increased odds of obesity in combined analyses; however these estimates had wide confidence intervals that included unity (Table 5.4). Children who were sedentary and 126 also had mothers who were inactive during pregnancy had almost three times the odds of obesity in combined, unadjusted analyses. While adjustment for maternal factors attenuated this relationship, maternal inactivity combined with sedentary child activity still showed borderline significance indicating almost twice the odds of obesity. Unlike odds of obesity, results for the relationships between maternal LTPA during pregnancy or maternal/child PA and odds of child overweight status were quite different between the non—subcohort and subcohort. It is unclear why LTPA during pregnancy would be associated with increased odds of child overweight while maternal inactivity during pregnancy combined with non—sedentary child activity would be associated with significantly decreased odds of overweight within the subcohort. One possibility is that the overweight category included a greater number of children who were misclassified by maternal reports of their height and weight than did the obese category, resulting in variable associations. Past literature suggests that parentally- reported height and weight values correctly classify obese children more often than overweight children.3 1'33 Thus, the results seen for odds of obesity may be more consistent between the non-subcohort and subcohort due to the more accurate maternal identification of obese children. This study is limited by the reliance on maternally reported values for child height, weight and leisure-time behavior, as well as the women’s recalled information for LTPA during pregnancy. Previous evidence on the validity of maternal report of child body size indicates moderate to good agreement for individual measures of height or weight, but poorer agreement for weight classification by BMI (kappa=0.34 for girls and 0.41 for boys).33'35 We excluded data from children with BMI values flagged as 127 biologically implausible by the CDC criteria to help improve the validity of our outcome variable. We used a proxy measure of child leisure-time behavior, asking mothers to characterize their children as mostly sedentary, some sedentary/some active, or mostly active in their leisure-time. It would have been more ideal to have direct, objective measures of child LTPA. However, LTPA intervention studies have not been able to demonstrate measureable improvements in child weight status, while interventions to reduce sedentary behavior have more consistently shown reductions in overweight status.” 36’ 37 It is possible that time spent in sedentary behavior is easier for parents/children to report, thus greater effects on weight status are able to be demonstrated in relation to sedentary behavior. We also lacked information on children's dietary patterns which may have affected their weight status. Finally, information on LTPA during pregnancy was recalled retrospectively at 3- 9 years postpartum, thus recall bias cannot be discounted. Bias due to recall error in our study may have been reduced by using a categorical variable to indicate any vs. no LTPA during pregnancy. Differential bias due to either maternal report of child body size or recall error is less likely because women were not told of the study aims and responded to these questions as part of a larger questionnaire on maternal and child behaviors. Despite these limitations, our results add significantly to the existing literature on risk factors for child obesity. While past research provides evidence for the role of both maternal and child LTPA participation on child body size, no previous study has considered the simultaneous effect of these two factors as we have done. Insufficient activity and/or meeting LTPA recommendations during pregnancy were each associated with ~30-50% reduced odds of child obesity in the non-subcohort and the subcohort. 128 These effects were borderline significant even after controlling for child sedentary behavior and maternal covariates. Furthermore, although statistically non-significant in adjusted analyses, odds of obesity were highest when both the mother and child were classified as inactive/sedentary compared to active/non-sedentary. Future work should further explore the inter-relationships among maternal and child LTPA and child weight status, and seek to determine plausible mechanisms through which maternal LTPA during pregnancy may affect child body size. Prospective, trimester-specific measures of LTPA during pregnancy combined with more objective measures of child LTPA levels are needed to test and refine our results. Beneficial maternal health effects of LTPA are already recognized.38 Our results suggest further that engaging in LTPA during pregnancy may improve offspring health by decreasing the odds of obesity. Overall, our results indicate that both maternal LTPA during pregnancy and child leisure-time behavior are important factors in the development of child obesity. 129 Table 5.1: Maternal characteristics of the POUCH study non-subcohort and subcohort* Chi- N on-Subcohort Subcohort Square N (%) N (%) p-value N 404 536 Maternal Race White/Other 386 (95.5) 372 (69.4) <0.001 African American 18 (4.5) 164 (30.6) Maternal Age at Delivery <20 yrs 21 (5.2) 50 (9.3) 20-<30 yrs 194 (48.0) 306 (57.1) <0.001 2 30 yrs 189 (46.8) 180 (33.6) Maternal Education < High School 28 (6.9) 70 (13.1) High School 73 (18.1) 134 (25.0) <0.001 > High School 303 (75.0) 332 (61.9) Medicaid Yes 94 (23.3) 231 (43.1) <0.001 Relationship Status Single 47 (11.6) 146 (27.2) <0.001 Smoking during Pregnancy At least some 28 (6.9) 80 (14.9) <0.001 Pre-pregnancy Maternal BMI < 25 kg/m2 221 (54.7) 265 (49.4) 25-<30 kg/m2 111 (27.5) 118 (22.0) <0.001 2 30 kg/m2 72 (17.25) 153 (28.5) Weight Gain During Pregnancy Low 45 (11.3) 87 (16.7) Recommended 143 (35.9) 169 (32.5) 0.062 High 210 (52.8) 264 (50.8) Maternal BMI at Follow-Up < 25 kg/m2 190 (47.0) 187 (34.9) 25-<30 kg/m2 105 (26.0) 146 (27.2) <0.001 2 30 l(g/m2 109 (26.7) 203 (37.9) LTPA During PregnancyS Inactive 161 (39.8) 258 (48.1) Insufficiently Active 121 (29-9) 124 (23-1) 0-020 Meeting Recommendations 122 (302) 154 (28-7) LTPA at Follow-Ups Inactive 100 (24.7) 178 (33.2) Insufficiently Active 100 (24-7) 107 (19-9) 0014 Meetifl Recommendations 204 (50-5) 251 (46-8) *lncludes only POUCH study participants who gave birth at term and were enrolled in the 2007 follow-up study sLTPA= leisure-time physical activity 130 iL l|1L-( Table 5.2: Child characteristics of the POUCH study non-subcohort and subcohort* Non-Subcohort Subcohort Chi- N (%) N (%) Square p-value N 404 536 Child Gender Male 194 (48.0) 259 (48.3) 0.927 Size at birth AGA 313 (77.5) 416 (77.6) SGA (S 10'h %tile) 18 (4.5) 57 (10.6) <0.001 LGA (2 90Lh %tile) 73 (18.1) 63 (11.8) Parity Nulliparous 179 (44.3) 219 (40.9) 0.289 Child Age at Follow-Up <4years 75 (18.6) 74(13.8) 4-<6 years 177 (43.8) 266 (49.6) 0.082 2 6 years 152 (37.6) 196 (36.6) Child BMI at Follow-Up Underweight (<5th %tile) 43 (10.6) 56 (10.5) Healthy Weight (5‘h -< 85th %tile) 244 (60.4) 283 (52.8) 0.042 Overweight (85‘h -< 95th %tile) 57 (14.1) 81 (15.1) Obese (2 95'h %tile) 60 (14.9) 116 (21.6) Child Leisure-Time Behavior Mostly Sedentary 47 (11.6) 83 (15.5) Some Sedentary/ Some Active 181 (44.8) 209 (39.0) 0.103 Mostly Active 176 (43.6) 244 (45.5) Maternal/Child LTPA W + / + 219 (54.2) 247 (46.1) + / - 24 (5.9) 31 (5.8) 0.031 _ / + 138 (34.2) 206 (38.4) _ / - 23 (5.7) 52 (9.7) * Includes only POUCH study participants who gave birth at term and were enrolled in the 2007 follow—up study 3 LTPA= leisure-time physical activity V Maternal/Child LTPA “+ / +” represents any maternal LTPA during pregnancy and child non-sedentary activity; “+ / -“ is any maternal LTPA during pregnancy and child sedentary activity; “- / +” is maternal inactivity during pregnancy and child non- ‘6 ‘( sedentary activity; activity 131 is maternal inactivity during pregnancy and child sedentary Table 5.3: Associations among maternal LTPA during pregnancy, child leisure-time behavior, and child weight status at 3-9 years (Reference: Healthy Weight) within the POUCH study non-subcohort and subcohort Overweight Obese Univariate Adjusted Univariate Adjusted OR’r on‘ OR“ on1 i (95% CI) (95% CI) (95% CI) (95% CI) ' Non-suhgghort (n=361)§ LTPA during Pregnancy (ref: Inactive) W Insufficiently Active 0.53 0.52 0.47* 0.59 (0.29, 1.07) (0.25, 1.07) (0.24, 0.96) (0.28, 1.22) Meeting LTPA Recs 0.54 0.52 0.49* 0.52 (0.27, 1.09) (0.26, 1.06) (0.24, 0.99) (0.26, 1.06) Child Behavior (ref: Mostly Sedentary) Some Sedentary/ Some 1.27 1.34 0.44 0.53 Active (0.45, 3.59) (0.47, 3.84) (0.19, 1.03) (0.22, 1.27) Mostly Active 1.44 1.54 0.73 0.82 (0.51, 4.10) (0.54, 4.41) (0.33, 1.64) (0.36, 1.90) Subcohort game)§ LTPA during Pregnancy (ref: Inactive) W lnsufficiently Active 1.69 1.97* 0.55* 0.73 (0.90, 3.17) (1.03, 3.78) (0.30, 1.00) (0.39, 1.36) Meeting LTPA Recs 1.72 2.15* 0.57* 0.80 (0.96, 3.07) (1.17, 3.94) (0.34, 0.97) (0.46, 1.42) Child Behavior (ref: Mostly Sedentary) Some Sedentary/ Some 0.64 0.61 0.51 * 0.70 Active (0.32, 1.29) (0.30, 1.27) (0.27, 0.97) (0.35, 1.40L Mostly Active 0.66 0.54 0.86 0.80 (0.33, 1.32) (0.26, 1.11) (0.47, 1.58) (0.42, 1.53) §Etcludes underweight children (n=43 non-subcohort, n=56 subcohort) l Univariate OR’s are unadjusted for any other variable iAdjusted OR’s are mutually adjusted for variables shown in table as well as Medicaid enrollment for the Non-subcohort and maternal race (White/Other vs. African American), pre-pregnancy BMI, and Medicaid enrollment for the Subcohort. W LTPA during pregnancy: “Inactive” includes women reporting no LTPA during pregnancy, “Insufficiently Active” includes women reporting less than recommended amounts of LTPA during pregnancy (<150 min/wk of moderate LTPA), “Meeting LTPA Recs” includes women reporting 2150 min/wk of at least moderate LTPA during pregnancy * p<0.05 132 Table 5.4: Associations among maternal/child LTPA and child weight status at 3-9 years (Reference: Healthy Weight) within the POUCH study non-subcohort and subcohort Overweight Obese OR aOR OR aOR (95% CI) (95% CI)’r (95% CI) (95% C1)" * Non3Sub§9hort (n=361)§ Maternal / Child LTPA (Ref: + / + )"’ + / - 0.32 0.33 1.89 1.72 (0.04, 2.54) (0.04, 2.55) (0.63, 5.63) (0.57, 5.63) - / + 1.75 1.76 213* 1.86 (0.95, 3.24) (0.94, 3.28) (1.14, 4.00) (0.98, 3.54) - / - 1.73 1.74 3.02* 2.47 (0.52, 5.74) (0.52, 5.84) (1.04, 8.82) (0.83, 7.39) l Subgohort (n=480)§ _... J, Maternal / Child LTPA (Ref: + / + )"’ + / - 1.70 1.75 0.96 1.06 (0.67, 4.33) (0.68, 4.51) (0.30, 3.08) (0.32, 3.55) - / + 053* 0.49* 1.64* 1.25 (0.31, 1.00) (0.27, 0.89) (1.01, 2.64) (0.75, 2.01) - / - 0.95 0.83 242* 1.76 (0.40, 2.27) (0.35, 2.01) (1.20, 4.89) (0.84, 3.71) § excludes underweight children (n=43 non-subcohort, n=56 subcohort) l Non-subcohort model is adjusted for Medicaid. Subcohort model is adjusted for maternal race (White/Other vs. African American), Medicaid enrollment, and pre- pregnancy BMI (<25, 225 kg/mz) ‘V Maternal/Child LTPA “+ / +” represents any maternal LTPA during pregnancy and child non-sedentary activity; “+ / -“ is any maternal LTPA during pregnancy and child sedentary activity; “- / +” is maternal inactivity during pregnancy and child non- sedentary activity; “- / -“ is maternal inactivity during pregnancy and child sedentary activity. * p<0.05 133 Table 5.5: The effects of LTPA during pregnancy and child leisure-time behavior on odds of obesity among children 3-9 years in the POUCH cohort (n=703§). Obesity (n=l76) (Ref: Healthy Weight) Main Effects Models Unadjusted Model 1* Model 2’r OR (95% CI) aOR (95% CI) aOR (95% CI) LTPA during Pregnancy (ref: Inactive) S Insufficiently Active 0.48* 0.51* 0.68 (0.31, 0.76) (0.32, 0.80) (0.42, 1.09) Meeting LTPA Recs 0.52* 053* 0.74 (0.34, 0.78) (0.35, 0.81) (0.47, 1.15) Child Behavior (ref: Mostly Sedentary) Some Sedentary/ Some 047* 053* 0.67 Active (0.28, 0.78) (0.31, 0.88) (0.39, 1.17) Mostly Active 0.80 0.88 0.84 (0.49, 1.30) (0.54, 1.44) (0.50, 1.42) Interaction Models Maternal / Child LTPA Unadjusted Model 11 Model 21 (Ref: + / + )‘" OR (95% CI) aOR (95% CI) aOR (95% C1) +/ - 1.31 1.34 1.40 (0.59, 2.91) (0.59, 3.04) (0.61, 3.20) - / + 1.91* 1.50* 1.42 (1.30, 2.78) (1.01, 2.24) @95, 2.12) - / - 284* 1.94* 1.78 (1.59, 5.09) (1.05, 3.60) (0.95, 3.34) § Excludes underweight (n=1 19) and overweight (n=l38) children 1 Main Effects Model 1 is mutually adjusted for both main effects variable. Model 2 is adjusted for fiirther adjusted for for maternal race (White/Other, African American), Medicaid enrollment, and pre-pregnancy BMI (<25, 225 kg/mz). 1 Interaction Model 1 is adjusted for maternal race (White/Other, African American) and Medicaid enrollment. Model 2 is adjusted for covariates in Model 1 + pre-pregnancy BMI (<25, 225 kg/mz). s LTPA during pregnancy: “Inactive” includes women reporting no LTPA during pregnancy, “Insufficiently Active” includes women reporting less than recommended amounts of LTPA during pregnancy (<150 min/wk of moderate LTPA), “Meeting LTPA Recs” includes women reporting 2150 min/wk of at least moderate LTPA. “' Maternal / Child LTPA“+ / +” represents any maternal LTPA during pregnancy and child non-sedentary activity; “+ / -“ is any maternal LTPA during pregnancy and child sedentary activity; “- / +” is maternal inactivity during pregnancy and child non- sedentary activity; “- / -“ is maternal inactivity during pregnancy and child sedentary activity. "‘ p<0.05 134 ACKNOWLEDGEMENTS This work was supported by the Perinatal Epidemiological Research Initiative Program Grant from the March of Dimes Foundation [Grants 20F Y01-38 and 20-FY04-37], the National Institute of Child Health and Human Development and the National Institute of Nursing Research [Grant R01 HD34543], the Thrasher Research Foundation [Grant 02816-7], the Centers for Disease Control and Prevention [Grants U01 DP000143-01 and R36 DP001322-01], and the Blue Cross Blue Shields of Michigan Foundation [Grant 1416 Student Award Program]. 135 REFERENCES 10. ll. 12. Troiano RP, Flegal KM. Overweight children and adolescents: description, epidemiology, and demographics. Pediatrics. Mar 1998;101(3 Pt 2):497-504. Ogden CL, Carroll MD, Flegal KM. High body mass index for age among US children and adolescents, 2003-2006. Jama. May 28 2008;299(20):2401-2405. Daniels SR. The consequences of childhood overweight and obesity. Future Child. Spring 2006;16(1):47-67. Reilly JJ, Methven E, McDowell ZC, et al. Health consequences of obesity. Arch Dis Child. Sep 2003;88(9):748-752. Freedman DS, Khan LK, Dietz WH, Srinivasan SR, Berenson GS. Relationship of childhood obesity to coronary heart disease risk factors in adulthood: the Bogalusa Heart Study. Pediatrics. Sep 2001;108(3):712-718. Gale CR, Javaid MK, Robinson SM, Law CM, Godfrey KM, Cooper C. Maternal size in pregnancy and body composition in children. J Clin Endocrinol Metab. Oct 2007;92(10):3904-391 l. Gillman MW, Rifas-Shiman SL, Kleinman K, Oken E, Rich-Edwards JW, Taveras EM. Developmental origins of childhood overweight: potential public health impact. Obesity (Silver Spring). Jul 2008;16(7): 1651-1656. Gillman MW, Rifas-Shiman S, Berkey CS, Field AE, Colditz GA. Maternal gestational diabetes, birth weight, and adolescent obesity. Pediatrics. Mar 2003;111(3):e221-226. Huang J S, Lee TA, Lu MC. Prenatal programming of childhood overweight and obesity. Matern Child Health J. Sep 2007;11(5):461-473. Simmons R. Perinatal programming of obesity. Semin Perinatal. Oct 2008;32(5):371-374. Clapp J F . The morphometric and neurodevelopmental outcome at five years of the offspring of women who continued exercise throughout pregnancy. J Pediatr. Dec 1996;129(6):856-863. Rennie KL, Livingstone MB, Wells JC, et al. Association of physical activity with body-composition indexes in children aged 6-8 y at varied risk of obesity. Am J Clin Nutr. Jul 2005;82(1):13-20. 136 13. 14. 15. 16. l7. 18. 19. 20. 21. 22. 23. Trost SG, Sirard JR, Dowda M, Pfeiffer KA, Pate RR. Physical activity in overweight and nonoverweight preschool children. Int J Obes Relat Metab Disord. Jul 2003;27(7):834-839. Abbott RA, Davies PS. Habitual physical activity and physical activity intensity: their relation to body composition in 5.0-10.5-y-old children. Eur J Clin Nutr. Feb 2004;58(2):285-291. Moore LL, Nguyen US, Rothman KJ, Cupples LA, Ellison RC. Preschool physical activity level and change in body fatness in young children. The Framingham Children's Study. Am J Epidemiol. Nov 1 1995;142(9):982-988. Holzman C, Bullen B, Fisher R, Paneth N, Reuss L. Pregnancy outcomes and community health: the POUCH study of preterm delivery. Paediatr Perinat Epidemiol. Jul 2001;15(Supp1 2):136-158. Kramer MS, Platt RW, Wen SW, et al. A new and improved population-based Canadian reference for birth weight for gestational age. Pediatrics. Aug 2001;108(2):E35. Lawrence EJ. A Matter of Size: Part 2 Evaluating the Large-for-Gestational-Age Neonate. Adv Neonatal Care. Aug 2007;7(4):187-197. Kuczmarski RJ, Ogden CL, Guo SS, et al. 2000 CDC Growth Charts for the United States: methods and development. Vital Health Stat I I. May 2002(246):1- 190. Noland M, Danner F, DeWalt K, McFadden M, Kotchen J M. The measurement of physical activity in young children. Res Q Exerc Sport. Jun 1990;61(2):146-153. Institute of Medicine. Nutrition during pregnancy, weight gain and nutrient supplements. Report of the Subcommittee on Nutritional Status and Weight Gain during Pregnancy, Subcommittee on Dietary Intake and Nutrient Supplements during Pregnancy, Committee on Nutritional Status during Pregnancy and Lactation, Food and Nutrition Board. Washington, DC: National Academy Press; 1990:1-233. Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc. Sep 2000;32(9 Suppl):S498-504. ACOG committee opinion. Exercise during pregnancy and the postpartum period. Number 267. American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet. Apr 2002;77(1):79-81. 137 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. Alderman BW, Zhao H, Holt VL, Watts DH, Beresford SA. Maternal physical activity in pregnancy and infant size for gestational age. Ann Epidemiol. Nov 1998;8(8):513-519. Voldner N, Froslie KF, Bo K, et al. Modifiable determinants of fetal macrosomia: role of lifestyle-related factors. Acta Obstet Gynecol Scand. Apr 2008;87(4):423- 429. Moore LL, Gao D, Bradlee ML, et al. Does early physical activity predict body fat change throughout childhood? Prev Med. Jul 2003;37(1):10-17. Davies PS, Gregory J, White A. Physical activity and body fatness in pre-school children. Int J Obes Relat Metab Disord. Jan 1995;19(1):6-10. Maffeis C, Talamini G, Tato L. Influence of diet, physical activity and parents' obesity on children's adiposity: a four-year longitudinal study. Int J Obes Relat Metab Disord. Aug l998;22(8):758-764. Burdette HL, Whitaker RC. A national study of neighborhood safety, outdoor play, television viewing, and obesity in preschool children. Pediatrics. Sep 2005;1 16(3):657-662. DeMattia L, Lemont L, Meurer L. Do interventions to limit sedentary behaviours change behaviour and reduce childhood obesity? A critical review of the literature. Obes Rev. Jan 2007;8(1):69—8 1. Banach A, Wade TJ, Caimey J, Hay JA, F aught BE, O'Leary DD. Comparison of anthropometry and parent-reported height and weight among nine year olds. Can J Public Health. Jul-Aug 2007;98(4):251-253. Goodman E, Hinden BR, Khandelwal S. Accuracy of teen and parental reports of obesity and body mass index. Pediatrics. Jul 2000;106(1 Pt 1):52-58. Dubois L, Girad M. Accuracy of maternal reports of pre-schoolers' weights and heights as estimates of BMI values. Int J Epidemiol. Feb 2007;36(1):l32-138. Davis H, Gergen PJ. Mexican-American mothers' reports of the weights and heights of children 6 months through 11 years old. J Am Diet Assoc. May 1994;94(5):512-516. Scholtens S, Brunekreef B, Visscher TL, et al. Reported versus measured body weight and height of 4-year-old children and the prevalence of overweight. Eur J Public Health. Aug 2007;17(4):369-374. Kropski JA, Keckley PH, Jensen GL. School-based obesity prevention programs: an evidence-based review. Obesity (Silver Spring). May 2008;16(5): 1009-1018. 138 37. 38. Campbell K, Waters E, O'Meara S, Summerbell C. Interventions for preventing obesity in childhood. A systematic review. Obes Rev. Aug 2001;2(3):149-157. Impact of physical activity during pregnancy and postpartum on chronic disease risk. Med Sci Sports Exerc. May 2006;38(5):989-1006. 139 CHAPTER SIX RELIABILITY OF RECALLING LEISURE-TIME PHYSICAL ACTIVITY PERFORMED DURING PREGNANCY ABSTRACT Our purpose was to evaluate the reliability of recalling leisure-time physical activity (LTPA) performed during pregnancy among postpartum women. Methods: Women (n=311) reported min/wk spent in moderate and vigorous LTPA at their first prenatal care visit. Total min/wk of LTPA was calculated. At 15-30 months postpartum, 82 women recalled the type, duration, and frequency of LTPA during the trimester of their original interview. Recalled min/wk of moderate, vigorous, and total LTPA were calculated. Results: Mean differences in observed-recalled LTPA were small but standard deviations were large (-6.0 i 175.7, 3.6 i 71.5, and -2.4 i 193.6 min/wk for moderate, vigorous, and total LTPA, respectively). Recalled vigorous (rs=0.34, p=0.002) and total (rs=0.28, p=0.016) LTPA values were significantly correlated with original reports, but this was not true for moderate LTPA (rs=0.10, p=0.350). Older age, higher income, having a healthy weight, and participating in recommended amounts of LTPA at follow-up were associated with better total LTPA recall ability. There was high percent agreement for original vs. recalled report of any moderate, vigorous, or total LTPA (70-79%), but kappa values were low (0.015-0.183). Conclusions: Correlations between original and recalled LTPA are low to moderate, percent agreement for participation in LTPA during pregnancy is high, but kappa values are low. Participant characteristics may influence reliability of LTPA recall. 140 INTRODUCTION Leisure-time physical activity (LTPA) during pregnancy has been associated with a wide range of health benefits, including decreased risk of gestational diabetes, preeclampsia, and maternal obesity."3 Thus, the American College of Obstetrics and Gynecology (ACOG) recommended in 2002 that all pregnant women participate in at least 30 minutes of moderate LTPA on most days of the week.4 More recently in 2008, the US. Government released evidenced-based LTPA guidelines for Americans which also stated that previously sedentary pregnant women should engage in at least 150 minutes of moderate LTPA per week, while habitual exercisers may continue more vigorous activities under the supervision of their health care providers.5 While past research has focused on the effects of LTPA during pregnancy on birth outcomes, investigators are beginning to evaluate possible relationships with later maternal/child health outcomes. Several questionnaires have been validated for measuring multiple domains of physical activity, including LTPA, among women who are currently pregnant;6‘ 7 however, prospective measurements of LTPA are often difficult to obtain. To facilitate epidemiological research on short- and long-term outcomes associated with LTPA during pregnancy, it is important to examine the reliability of historical recall of LTPA among postpartum women. The Modifiable Activity Questionnaire (MAQ) was originally validated for historical recall of LTPA among Pima Indians; however, it was created specifically to be easily adapted to other populations and varying lengths of recall.8’ 9 Bauer et al. recently used the MAQ to assess the validity of recalled LTPA during pregnancy among women 141 six years postpartum.10 Moderate to high agreement was found for recalled LTPA with that originally recorded at 20 and 32 weeks gestation, and 12 weeks postpartum.lo Participants in the Bauer et al. study were Caucasian, relatively lean at follow-up, and many had participated in high amounts of LTPA during pregnancy. It is not known whether LTPA can be recalled reliably by women with more diverse socioeconomic status and LTPA patterns. The purpose of this study was to evaluate the reliability of recalled LTPA during pregnancy compared to amounts reported at the first prenatal care visit among women who were 15-30 months postpartum. Additionally, we sought to determine whether participant characteristics influenced the reliability of the women's recall. METHODS Participants Women were originally recruited from the waiting rooms of nine prenatal care clinics in Kent County, MI and asked to participate in a 15-20 minute face-to-face interview as part of the Michigan Alliance for the National Children’s Study (MANCS) pilot study.I " l2Eligible women were currently pregnant, at their first prenatal care visit, 18-50 years old, and proficient in English or Spanish. From April-October 2006, 311 of 342 (91%) eligible women completed the interview. Of these, 299 provided complete LTPA information at the time of enrollment and agreed to be contacted for future studies. From July 2008 - February 2009 women were contacted via mail and telephone and asked to complete a follow-up survey on pregnancy behaviors, birth outcomes and their toddler’s health. One-hundred-ten women (37%) could not be located, despite paid intemet searches. Of the remaining 188 women, 23 refused participation, 3 had 142 miscarried during their 2006 pregnancies, and 44 never responded to phone calls or mailings, leaving 118 who agreed to participate. Of these, 82 (28% of original cohort) provided complete LTPA recall information. The original study was approved by the Institutional Review Boards of Michigan State University and the three hospitals with labor and delivery services in Kent County, MI — Spectrum Health, Metro Health, and St. Mary’s Health Care Center. The Biomedical Institutional Review Board at Michigan State University approved the follow-up study. Women completed informed consent at the original recruitment and again for participation in the follow-up survey. Original Data Collection The original interview consisted of demographic questions, maternal LTPA, participation and safety perceptions, and attitudes toward participating in hypothetical research procedures.l " '2 Questions on LTPA were adapted from the 2001 Behavioral Risk Factor Survey.l3 Specifically, women reported whether or not they had participated in any moderate of vigorous physical activities during their free time in the past month (yes/no). If so, average durations (min/d) and frequencies (d/wk) were reported for a typical week. Interviewers described moderate activities as those causing light sweating or slight to moderate increases in breathing or heart rate, and vigorous activities as those that cause heavy sweating or large increases in breathing or heart rate. Scripted examples of moderate (e. g., brisk walking, gardening) and vigorous activities (e.g., running, aerobics) were given; however, specific activities were not recorded. Minutes per week spent in moderate and vigorous LTPA were calculated. Total LTPA was recorded as the sum of moderate and vigorous LTPA. Women were 143 C2 \1 categorized as reporting any vs. no participation in moderate, vigorous, or total LTPA. Additionally, women reporting at least 150 minutes per week of total LTPA were classified as meeting ACOG recommendations.4 Women also indicated whether or not they intended to be active throughout the remainder of their pregnancies as “yes”, “no”, or “it depends”. Maternal age, educational level, race/ethnicity, relationship status, family income level, and parity were recorded. Finally, women reported their current gestational week of pregnancy or the trimester if they were unsure about the week. F allow-up Data Collection Women had the choice of completing the follow-up survey through the mail or over the intemet. The MAQ was used to assess past and current LTPA. Participants recalled their LTPA for a typical week during five time-periods: the month before they knew they were pregnant, the first, second, and third trimesters, and the previous month. The time-period of recalled LTPA was matched to the trimester in which the woman completed the original interview. Women specified the type, duration (average min/d), and frequency (average d/wk) of up to five activities for each time-point. A standard list of 40 activities was provided for women to choose from, and they could write in other activities not on the list. To designate intensity, metabolic equivalent (MET) values were assigned to each reported activity using the compendium for physical activities. '4 Minutes per week spent in each reported activity were calculated. To maintain comparability with previously collected data, amount of time (min/wk) recalled for moderate activity was calculated as the sum of all activities with a MET value of 3.0 to 5.9, while min/wk recalled for vigorous activity was calculated as the sum of all activities with a MET value 2 6.0. Total recalled min/wk of LTPA excluded any light intensity 144 01c activities (i.e., < 3.0 METs). Women were classified as recalling participation in any vs. no moderate, vigorous, or total LTPA. Women recalling participation in at least 150 min/wk of total LTPA were classified as meeting ACOG recommendations.4 During follow-up, women also self-reported pre-pregnancy weight and height, gestational weight gain, and current weight. Both pre-pregnancy and current body mass index (BMI) values were calculated (kg/m2). Statistical Analyses Chi-square analyses compared descriptive characteristics between women who did and did not participate in the follow-up study. Descriptive statistics were calculated for originally reported and recalled min/wk of moderate, vigorous, and total LTPA and mean differences were assessed. Physical activity data were non-normally distributed, thus Spearman rank-order correlation coefficients were used to assess relationships between originally reported and recalled min/wk of moderate, vigorous, and total LTPA for the sample as a whole and by subgroups of participant characteristics. The standard error of measurement was also calculated for recalled min/wk of moderate, vigorous, and total LTPA. Categorical agreement among originally reported and recalled LTPA (any vs. none and meeting vs. not meeting recommendations) was evaluated using percent agreement and Cohen’s kappa, which corrects for chance agreement. All analyses were performed using SAS version 9.1 and significance was set at a two-tailed Alpha level of p<0.05. RESULTS Women who participated in the follow-up were significantly more likely to be older, more educated, of Non-Hispanic White race/ethnicity, married, have a higher 145 ([5 I'd re: income, and to have completed their original interview in the lSt trimester (Table 6.1). According to the original report, participation in vigorous and total LTPA was more frequent among those who completed the follow-up; however, participation in moderate LTPA, likelihood of meeting ACOG guidelines, and intention to be active throughout pregnancy were not significantly different between those who did and did not complete the follow-up (Table 6.1). Among women who completed the follow-up survey, average pre-pregnancy BMI was 23.7 :t 4.9 kg/mz, reported gestational weight gain was 35 3:13 lbs, and BMI at follow-up was 25.1 i 6.0 kg/mz. Mean values of originally reported and recalled LTPA are shown in Table 6.2. Recalled LTPA for a typical week in the first trimester was used for all but six women whose first prenatal care visit fell in the second (n=5) or third (n=1) trimesters. Mean differences between originally reported and recalled LTPA values were small (-6.0 to 3.6 min/wk); however, wide ranges in recall ability were present. Values for the standard error of measurement for moderate (123 min/wk), vigorous (51 min/wk), and total (138 min/wk) LTPA indicate that women were more reliable at recalling vigorous than moderate or total LTPA. Spearman correlations between originally reported and recalled moderate LTPA were not statistically significant for the group as a whole (rs=0.10), or within any of the participant characteristic subgroups (Table 6.3). However, significant relationships were observed between original and recalled reports of vigorous (rs=0.34) and total LTPA (rs=0.28). Older women, those with higher income, and those with a BMI in the healthy range also had significant, moderately strong correlations between originally reported and recalled vigorous and total LTPA (rs=0.36-0.56). Additionally, nulliparous women 146 IL‘ displayed a significant correlation for original and recalled report of vigorous LTPA while parous women did not. Interestingly, women classified as meeting LTPA recommendations according to their current activity levels at follow-up showed a significant correlation between original and recalled total LTPA, while women classified as not meeting LTPA recommendations had a significant correlation for vigorous activity (Table 6.3). We were unable to evaluate differences in recall ability by education, race/ethnicity, and relationship status due to the homogeneity of our participants regarding these characteristics. Percent agreement for women reporting any vs. no participation in moderate, vigorous, and total LTPA was high (74-79%, Table 6.4). However, using standard interpretation of the kappa statistic (i.e. kappa <0.2= poor, 0.2-0.39= fair, 0.4- 0.59=moderate, 0.6-0.79=good, 0.8+= strong agreement)15 , results indicated poor agreement between original and recalled report of participation in LTPA during pregnancy. When categorizing women on meeting ACOG LTPA recommendations, there was moderate (57%) agreement between original and recalled reports, but the kappa value was 0.117 indicating poor agreement. Cell sizes were too small to determine whether categorical LTPA agreement varied by participant characteristics. DISCUSSION Our purposes were to evaluate the reliability of recalling LTPA during pregnancy among women 15-30 months postpartum, and to determine whether participant characteristics influenced reliability of recall. Our results indicate that recall of moderate LTPA was poor, but recalled amounts of vigorous and total LTPA were moderately correlated with original values. Maternal age, parity, income, maternal body size at time 147 of recall, and participation in recommended amounts of LTPA at time of recall each appeared to influence the reliability of recall on a continuous scale. When considered categorically, percent agreement for any vs. no participation in LTPA was high and percent agreement for meeting vs. not meeting ACOG recommendations was moderate; however kappa values were quite low. Recalled amounts of moderate activity were non-significantly correlated with original values (rs=0.10), whereas moderate and significant correlations were found for recalled vigorous (rs= 0.34) and total LTPA (rs=0.28). Similar correlations have been demonstrated for middle-aged women recalling LTPA occurring 1—5 years previously.l6’ '7 Using a slightly different form of the MAQ than that used here, Chasan-Tabor et al. found poor one-year recall ability for moderate activity (r=0. 15) and stronger recall ability for vigorous (F052) and total (r=0.26) LTPA among a group of 131 women aged 39-65 yrs.’6 Similar correlations were demonstrated among 78 middle-aged women for the 3-5 year recall of moderate (r=0.16), vigorous (r=0.26), and total (r=0.20) physical activities. '7 Other studies involving mixed gender samples and/or longer recall lengths (10 to 30 yrs) have reported similar results for their female sub-cohorts as well, with correlations ranging from 0.25-0.47 for recall of moderate, 0.36-0.41 for vigorous, and 029-057 for total LTPA.'8'20 Thus, our results are in line with those reported previously for recalling LTPA and it appears that vigorous activities are more easily recalled than moderate intensity activities. The one previous study to examine historical recall of LTPA performed during pregnancy found that values recalled six years postpartum were strongly correlated with originally measured LTPA at 20 and 32 weeks of gestation (r = 0.57 and 0.85, 148 FE ca respectively).'0 There are several reasons why our results are not as strong as those found by Bauer et al. Participants in the Bauer et a1. study had much higher LTPA levels than our sample, which may have been more easily recalled. They had also been monitored carefully during pregnancy to record LTPA data via heart rate telemetry, accelerometry, and LTPA diarieslo’ 21‘ 22 Thus, it is possible that their apparent ability to recall LTPA was influenced by greater attention to their activity during pregnancy. Regardless, although our findings are not as strong, they do corroborate those of Bauer et al. using a larger and more diverse sample. We found that correlations between originally reported and recalled LTPA were stronger for both vigorous and total LTPA among women 2 30 years old, those earning 2 $50,000, and those with a healthy BMI (< 25 kg/mz). Nulliparous women also displayed a stronger correlation for recall of vigorous LTPA, compared to parous women. While women meeting LTPA recommendations at the time of recall displayed a stronger correlation for total LTPA, women not meeting recommendations had a stronger correlation for recalling vigorous activity. Falkner et al. assessed the influence of participant characteristics on the recall of LTPA >30 years in the past.23 The authors found that “underestimators” were younger while “overestimators” had higher levels of current LTPA compared to “good recallers”.23 All other participant characteristics (gender, education, marital status, BMI, pulse rate, and blood pressure) were unrelated to recall ability in their sample. Other investigators have reported that adjustment for age, length of recall interval, or current BMI did not significantly alter correlations between recalled and originally reported activity values; however, previous studies have not calculated stratum-specific correlation coefficients as we have done. '8‘ 24 It is possible 149 that recall of pregnancy-related LTPA is more susceptible to maternal characteristic influences, but more work is needed to confirm our findings. Finally, our categorical analyses showed high percent agreement between original and recalled reports of participation in any moderate, vigorous, or total LTPA (70-79%), while percent agreement for meeting ACOG LTPA recommendations was moderate (57%). Unfortunately, kappa values were quite low for these comparisons and indicated poor agreement after controlling for chance. The condition of having high percent agreement and low kappa scores has been termed the “prevalence paradox”.25 When the prevalence of the outcome is very high or very low, agreement expected by chance increases and the magnitude of kappa decreases. In this situation, the interpretation of kappa is more ambiguous than standard guidelines, and it is recommended that both percent agreement and kappa scores be taken into consideration.26 In our data 95% of women reported at least some moderate or vigorous LTPA at original data collection, thus it is not surprising that kappa values were low. Most previous studies on reliability of LTPA recall have not reported categorical agreement; however, one study on 10 year LTPA recall ability showed 70% agreement for reporting low vs. high amounts of vigorous LTPA and 66% agreement for total LTPA among 322 women.18 Thus, our results are in line with those found previously among non-pregnant samples. Some limitations of the present study must be noted. First, validation of originally reported LTPA levels was not conducted, so the possibility of bias and/or error in the original values cannot be discounted. However, the LTPA questions used in the original interview have been shown to be moderately to highly reliable for moderate (kappa 0.35- 0.53) and vigorous (kappa 0.80-0.86) activities among US. adults, and moderately valid 150 for id (kapl Add disc diff the mi [)8 for identifying recommended amounts of LTPA when compared to daily LTPA logs (kappa = 0.40-0.52).13 Thus, we are fairly confident in our original report of LTPA. Additionally, a different survey of reporting LTPA was used at follow-up, thus some discrepancies between original and recalled amounts of LTPA may be due to survey differences. Women participating in the follow-up study were substantially different from those who did not (Table 6.1), which limits the generalizability of our results. While original report of vigorous and total LTPA was more frequent among those who participated in follow-up, report of moderate LTPA, the likelihood of meeting ACOG LTPA recommendations, and intention to be/not be active throughout pregnancy did not significantly differ between those who did and did not participate in the follow-up. Finally, while small mean differences between recalled and originally reported amounts of LTPA were noted, individual differences were wide. Thus, some women grossly overestimated and others underestimated their original activities. Despite limitations, our results corroborate previous reports on the reliability of historical recall of LTPA.'°’ '6'20‘ 27 We found moderate reliability for recalling vigorous and total LTPA during pregnancy among women 15-30 months postpartum using rank- order spearman correlations. Older age, nulliparity, higher family income, and having a healthy BMI and participating in recommended amounts of LTPA at the time of recall were each associated with improved total LTPA recall ability. While kappa values for categorical analyses were low, percent agreement was high for recalling participation in moderate, vigorous and total LTPA and moderate for identifying women with recommended amounts of LTPA during pregnancy. Based on the results reported here 151 and those found by Bauer et al., we believe that future studies on the long-term outcomes associated with LTPA during pregnancy can consider use of historical recall of LTPA when prospective measurement is impractical.10 Future research is needed to establish the reliability of historically recalling other domains of physical activity during pregnancy, such as occupational and household-related physical activity. 152 Table 6.]: Participant characteristics during pregnancy according to information reported during the original interview. Did Not Original Complete Completed Chi- Study F ollow-Up F ollow-Up Square N (%) N (%) N (%) p-value* N 299 217 82 Age 18-29 yrs 208 (69.6) 164 (75.6) 44 (53.7) 30+ yrs 91 (30.4) 53 (24.4) 38 (46.3) <0.001 Education < High School 68 (22.7) 65 (30.0) 3 (3.7) <0.001 Race/ethnicity Non-Hispanic White 172 (57.5) 106 (48.9) 66 (80.5) Non-Hispanic Black 48 (16.1) 43 (19.8) 5 (6.1) <0.001 Hispanic/Other 79 (26.4) 68 (31.3) 11 (13.4) Relationship Status Single 66 (22.2) 59 (27.4) 7 (8.5) <0.001 Family Income5 < $25,000 129 (43.9) 114 (53.8) 15 (18.3) $25 _ 50,000 69 (23.5) 47 (22.2) 22 (26.8) <0.001 > $50,000 96 (32.7) 51 (24.1) 45 (54.9) Parity Nulliparous 124 (41.5) 84 (38.7L 40 (48.8) 0.115 Trimester of Interview 15' 248 (82.9) 172 (79.3) 76 (92.7) 2nd 42 (14.1) 37 (17.1) 5 (6.1) 0.023 3rd 9 (3.0) 8 (3.7) 1 (1.2) Moderate LTPA1 Yes 258 (86.6) 183 (84.7) 75 (91.5) 0.127 Vigorous LTPA Yes 68 (22.7) 41 (18.9) 27 (32.9) 0.010 Total LTPA Yes 264 (88.3) 186 (85.7) 78 (95.1) 0.024 ACOG Guidelines+ Meets 111 (37.1) 78 (35.9) 33 (40.2) 0.492 Pregnancy LTPA Intention Yes 238 (79.9) 168 (77.8) 70 (85.4) No 48 (16.1) 40 (18.5) 8 (9.8) 0.177 Depends 12 (4.0) 8 (3.7) 4 (4.9) *Chi-square p-values compare women who did and did not complete follow—up n=5 women did not report family income in the original interview lLTPA= Leisure-time physical activity; ACOG=American College of Obstetrics and Gynecology; “Meets” ACOG Guidelines = original report of total LTPA 2150 min/wk 153 1'10. Table 6.2: Average LTPA* values and mean differences in originally reported and recalled values for moderate, vigorous, and total LTPA among women who completed follow-up (n=82) Original LTPA Recalled LTPA Original — Recalled LTPA (min/wk) (min/wk) (min/wk) Moderate 131.1 :t 108.5 137.1 a: 152.1 -6.0 at 175.5 LTPA @620 to 420) Vigorous 30.5 2’: 59.6 26.9 :1: 65.8 3.6 i 71.5 LTPA (-420 to 210) Total 161.6 :t 133.6 164.0 i 179.1 -2.4 d: 193.6 LTPA (-605 to 435) Value are means :t standard deviations (ranges) * LTPA = Leisure-time physical activity Table 6.3: Spearman correlations (rs, p-value) between originally reported and recalled min/wk of moderate, vigorous, and total LTPA for the group as a whole and by participant subgroups Moderate Vigorous Total LTPA” LTPA“ LTPA“ Total Group n=82 0.10 (0.35) 0.34 (<0.001)* 0.28 (0.02)* Age at Interview 18-<30 yrs (n=44) 0.16 (0.29) 0.17 (0.27) 0.20 (0.20) 30-<50 yrs (n=38) 0.02 (0.97) 0.55 (<0.001)* 0.32 (0.05)* Parity Nulliparous (n=40) 0.19 (0.25) 0.39 (0.01)* 0.30 (0.06) Parous (n=42) 0.02 (0.89) 0.23 (0.15) 0.22 (0.16) Family Income < $50,000 (n=37) -0.02 (0.91) 0.08 (0.63) 0.08 (0.64) 2 $50,000 (n=45) 0.17 (0.27) 0.53 (<0.001)* 0.39 (0.01)* BMI at Recall < 25 kg/m2 (n=46) 0.21 (0.15) 0.37 (0.01)* 0.33 (0.03)* 2 25 kg/m2 (n=36) -0.03(0.88) 0.14 (0.48) 0.11 (0.57) ACOG Recs at Recalls Meets (n=43) 0.18 (0.24) 0.29 (0.08) 0.38 (0.01)* Does not Meet (n=39) -0.02 (0.93) 0.42 (0.01)* 0.07 (0.66) # LTPA = Leisure-time physical activity 5 “ACOG Recs at Recall” classifies women into meeting (2150 min/wk) or not meeting (< 150 min/wk) ACOG recommendations based on their report of LTPA during the previous month at the time of recall *p<0.05 154 Ta pa 1'81 E1___E1__El 3| [_ b—-* .__. Table 6.4: Categorical agreement between originally reported and recalled participation in moderate, vigorous, and total LTPA, and participation in ACOG recommended amounts of LTPA Percent A_reement Kat -a Value Recall Yes 74% 0.015 Recall No Recall Yes 1 1 16 70% 0.183 Recall No 9 46 Recall Yes 79% 0. 031 Recall No 14 Recall Yes Recall No *LTPA= leisure-time physical activity; “Meets ACOG Recommendations”= reports 2 150 min/wk of total LTPA 155 ACKNOWLEDGEMENTS This work was supported by the Centers for Disease Control and Prevention [Grant R36 DP001322-01], and the Blue Cross Blue Shields of Michigan Foundation [Grant 1416 Student Award Program]. 156 REFERENCES 10. ll. Weissgerber TL, Wolfe LA, Davies GA, Mottola MF. Exercise in the prevention and treatment of matemal-fetal disease: A review of the literature. Appl Physiol Nutr Metab. Dec 2006;31(6):661-674. Hegaard HK, Pedersen BK, Nielsen BB, Damm P. Leisure time physical activity during pregnancy and impact on gestational diabetes mellitus, pre-eclampsia, preterm delivery and birth weight: a review. A cta Obstet Gynecol Scand. Nov 2007;86(1 1): 1290-1296. . Impact of physical activity during pregnancy and postpartum on chronic disease risk. Med Sci Sports Exerc. May 2006;38(5):989-1006. ACOG committee opinion. Exercise during pregnancy and the postpartum period. Number 267. American College of Obstetricians and Gynecologists. Int J Gynaecol Obstet. Apr 2002;77(1):79-81. U.S. Physical Activity Guidelines for Americans. US. Department of Health and Human Services. http://www.health.gov/paguidelines: ODPHP Publication No. U0036; 2008. Chasan-Taber L, Schmidt MD, Roberts DE, Hosmer D, Markenson G, Freedson PS. Development and validation of a Pregnancy Physical Activity Questionnaire. Med Sci Sports Exerc. Oct 2004;36(10):l750-1760. Schmidt MD, Freedson PS, Pekow P, Roberts D, Stemfeld B, Chasan-Taber L. Validation of the Kaiser Physical Activity Survey in pregnant women. Med Sci Sports Exerc. Jan 2006;38(1):42-50. Kriska AM, Knowler WC, LaPorte RB, et al. Development of questionnaire to examine relationship of physical activity and diabetes in Pima Indians. Diabetes Care. Apr 1990;13(4):401-411. Kriska AM. Modifiable Activity Questionnaire. Med Sci Sports Exerc. 1997 1997;29(Supp1):S73-S78. Bauer PW, Pivamik JM, Feltz DL, Paneth N, Womack CJ. Validation of a historical physical activity recall tool on postpartum women. Med Sci Sports Exerc. 2006;38(Suppl):S561. Nechuta S, Mudd LM, Biery L, et al. Attitudes of pregnant women towards participation in perinatal epidemiologic research. Paediatr Perinat Epidemiol. Jul 2009; epub DOI: 10.111l/j.1365-3016.2009.01058.x. 157 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. Mudd LM, Nechuta S, Pivamik JM, Paneth N. Factors associated with women's perceptions of physical activity safety during pregnancy. Prev Med. Jun 2009; epub DOI:10.1016/j.ypmed.2009.06.004. Yore MM, Ham SA, Ainsworth BE, et a1. Reliability and validity of the instrument used in BRF SS to assess physical activity. Med Sci Sports Exerc. Aug 2007;39(8): 1267-1274. Ainsworth BE, Haskell WL, Whitt MC, et al. Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc. Sep 2000;32(9 Suppl):S498-504. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. Mar l977;33(1):159-174. Chasan-Taber L, Erickson J B, Nasca PC, Chasan-Taber S, Freedson PS. Validity and reproducibility of a physical activity questionnaire in women. Med Sci Sports Exerc. Jun 2002;34(6):987-992. DuBose KD, Edwards S, Ainsworth BE, Reis JP, Slattery ML. Validation of a historical physical activity questionnaire in middle-aged women. J Phys Act Health. Jul 2007;4(3):343-355. Blair SN, Dowda M, Pate RR, et al. Reliability of long-term recall of participation in physical activity by middle-aged men and women. Am J Epidemiol. Feb 1991;133(3):266-275. Falkner KL, Trevisan M, McCann SE. Reliability of recall of physical activity in the distant past. Am J Epidemiol. Jul 1999;150(2): 195-205. Winters-Hart CS, Brach J S, Storti KL, Trauth JM, Kriska AM. Validity of a questionnaire to assess historical physical activity in older women. Med Sci Sports Exerc. Dec 2004;36(12):2082-2087. Pivamik JM, Stein AD, Rivera J M. Effect of pregnancy on heart rate/oxygen consumption calibration curves. Med Sci Sports Exerc. May 2002;34(5):750-755. Stein AD, Rivera J M, Pivamik J M. Measuring energy expenditure in habitually active and sedentary pregnant women. Med Sci Sports Exerc. Aug 2003;35(8):1441-1446. Falkner KL, McCann SE, Trevisan M. Participant characteristics and quality of recall of physical activity in the distant past. Am J Epidemiol. Nov 2001;154(9):865-872. 158 24. 25. 26. 27. Wolf AM, Hunter DJ, Colditz GA, et al. Reproducibility and validity of a self- administered physical activity questionnaire. Int J Epidemiol. Oct 1994;23(5):991-999. Lantz CA, Nebenzahl E. Behavior and interpretation of the kappa statistic: resolution of the two paradoxes. J Clin Epidemiol. Apr 1996;49(4):43 1-434. Hoehler FK. Bias and prevalence effects on kappa viewed in terms of sensitivity and specificity. J Clin Epidemiol. May 2000;53(5):499-503. Slattery ML, Jacobs DR, Jr. Assessment of ability to recall physical activity of several years ago. Ann Epidemiol. Jul 1995;5(4):292-296. 159 CHAPTER SEVEN SUMMARY AND RECOMMENDATIONS SUMMARY OF RESULTS The purposes of this dissertation were three-fold: to determine the effects of LTPA during pregnancy on offspring size at birth, to determine the effects of LTPA during pregnancy and child LTPA on child weight status at 3-9 years of age, and to determine the reliability of postpartum recall of LTPA during pregnancy. Using data from the POUCH study we found that maternal LTPA during pregnancy at or above recommended levels was associated with decreased odds of giving birth to an LGA infant (Chapter 3). Furthermore, using quantile regression, we were able to demonstrate that the effect of LTPA during pregnancy was limited to the upper quantiles of birth weight 2- score such that participating at or above recommended levels reduced birth weight 2- scores above the 0.65 quantile without causing a mean shift in the distribution. Due to the adverse maternal and child health effects associated with higher birth weight, these results may indicate a substantial health benefit for LTPA during pregnancy, without increased risk of SGA. Beyond effects on offspring size at birth, we also sought to determine whether participation in LTPA during pregnancy was associated with weight status during childhood, while also considering child LTPA level (Chapter 5). Using a combined maternal/child LTPA variable within the POUCH cohort, we found that inactivity during pregnancy combined with either non-sedentary or sedentary child activity were associated with borderline significant increased odds of child obesity as compared to any LTPA during pregnancy combined with non-sedentary child activity. The highest odds of 160 child obesity were observed when both the mother and child were classified as inactive/sedentary. Thus it appears that both maternal LTPA during pregnancy and children’s own LTPA level may be important in the development of child obesity. Finally we used the MANCS pilot study cohort to examine reliability of recalling LTPA during pregnancy at 15-30 months postpartum (Chapter 6). Correlational analyses indicated moderate relationships between original and recalled reports of min/wk of vigorous and total LTPA, but relationships for moderate activity were non-significant. Older age, higher income, having a healthy weight, and participating in recommended amounts of LTPA at follow-up were associated with better recall ability for total LTPA. There were high percent agreements for original vs. recalled report of any moderate, vigorous, or total LTPA, and moderate percent agreement for meeting LTPA guidelines, but kappa values for all of these relationships were low. Overall, these results support moderate reliability for recalling LTPA during pregnancy. Thus, researchers may consider the use of historical recall of LTPA during pregnancy when prospective measurements are not feasible. However, larger sample sizes may be needed to counter variability due to misclassification and participant characteristics known to influence recall ability should be included in multivariate models. Taken as a whole, our results indicate a role for maternal LTPA during pregnancy to affect not only offspring size at birth, but also weight status in early childhood when combined with child activity level. Our results provide preliminary evidence that maternal participation in LTPA may help to improve offspring health by normalizing size at birth and reducing odds of child obesity. While we had to rely on recalled LTPA data 161 within the POUCH study cohort, our results from the MANCS pilot study cohort indicate that postpartum recall of LTPA during pregnancy is moderately reliable. RECOMMENDATIONS Future research is needed to further test and refine our findings. Prospective studies with detailed measures of LTPA throughout pregnancy are needed to determine trimester-specific effects of LTPA on offspring size. This will help to refine recommendations for LTPA during pregnancy to maximize maternal and offspring health benefits. Future studies should also seek to incorporate objective and direct measures of child body size and LTPA participation, such as body fatness and acceleromehy, so that combined influences of maternal LTPA during pregnancy and child LTPA participation on child body size can be more fully explored. One of the advantages of using the POUCH cohort to evaluate our aims was that a wide range of recalled LTPA participation was present. Future prospective studies should seek to recruit samples of women with diverse LTPA participation so that dose-response relationships with offspring size can be evaluated. It is yet to be determined whether a minimal threshold amount of LTPA is needed to effect offspring size. Our results for Aim One indicated that LTPA participation at or above the ACOG recommended amount was associated with reduced BWz among the upper quantiles of the distribution, while participation in LTPA below this level was not associated with BWz. In contrast, our results for Aim Two indicated that participation in any LTPA during pregnancy combined with non-sedentary child activity reduced the odds of child obesity. Future studies with more exact measures of LTPA throughout pregnancy and of child body size may be able to delineate a more specific minimum threshold level of LTPA during 162 pregnancy needed to effect offspring size. As yet no study has determined a maximum threshold of LTPA participation that would begin to increase maternal/child health risks rather than benefits. This is likely due to the fact that few women choose to exercise vigorously throughout pregnancy, thus it is hard to gather generalizable data on the effects of vigorous LTPA. Future studies should begin to examine the relationship of maternal LTPA during pregnancy with matemal/fetal biomarkers to test for biologically plausible pathways for the effects of LTPA during pregnancy on offspring size and other health outcomes. Our results provide evidence that maternal inactivity during pregnancy combined with either non-sedentary or sedentary child activity increases risk for child obesity. However, whether this relationship is the result of biological influences of LTPA during pregnancy on fetal development, or environmental influences on child behavior is still debatable. Possible biological effects could be evaluated by examining relationships between LTPA during pregnancy and fetal metabolic parameters. Examining relationships among LTPA during pregnancy, maternal postnatal behaviors, and child LTPA and dietary behaviors would help to determine how environmental influences also effects offspring size. This dissertation focused on LTPA during pregnancy, and the vast majority of women in both cohorts reported participation in aerobic LTPA (i.e., walking). Future studies are needed to evaluate the influence of other types of LTPA, such as strength training or yoga, on birth outcomes and offspring size. It is also important to consider the contribution of other domains of physical activity, such as occupational and household- related physical activity. Due to the intermittent nature and varying intensity level of occupational and household-related activities, the effects of participating in these 163 domains of physical activity may be very different from the effects associated with LTPA participation. Finally, we found moderate reliability for postpartum recall of LTPA during pregnancy using the MAQ compared to original LTPA questions based on the BRFS. It would be preferable to establish validity of recalling LTPA during pregnancy by using objective measures, such as accelerometry, throughout pregnancy. In the absence of accelerometry data, researchers should procure more detailed LTPA questionnaire data during pregnancy so that total volume of LTPA reported originally and at recall can be compared. The reliability of recalling other domains of physical activity during pregnancy also needs to be established to facilitate epidemiological studies examining outcomes associated with these variables. The results of this dissertation set the stage for a line of inquiry further evaluating the effects of maternal LTPA during pregnancy on offspring size and health. Future study incorporating prospective, detailed measures of LTPA during pregnancy, child LTPA, and child size at birth through childhood will help to more specifically define the independent effect of LTPA during pregnancy on offspring size. If our results are confirmed, then future research may also seek to evaluate the efficacy of LTPA intervention programs during pregnancy to reduce risk for child obesity. 164 111111111111 11111] 1 3 1293 03062 9970 ill1111111111111