THE  RELATIONS  BETWEEN  FEEDING  PRACTICES,  BODY  MASS  INDEX,  DEPRESSION,  AND   AGGRESSION  IN  A  LOW-­‐INCOME  SAMPLE  OF  MOTHERS  AND  CHILDREN     By     Heather  Whitty                       A  THESIS   Submitted  to   Michigan  State  University   in  partial  fulfillment  of  the  requirements  for  the  degree  of     MASTER  OF  ARTS     Human  Development  and  Family  Studies     2011               ABSTRACT     THE  RELATIONS  BETWEEN  FEEDING  PRACTICES,  BODY  MASS  INDEX,  DEPRESSION,  AND   AGGRESSION  IN  A  LOW-­‐INCOME  SAMPLE  OF  MOTHERS  AND  CHILDREN     By     Heather  Whitty       A  controlling  feeding  style  is  associated  with  difficulty  in  children’s  food  regulation,   which  is  linked  to  later  weight  gain  (Johnson  &  Birch,  1994).  Survey  data  from  Early  Head   Start  (n  =  119)  families  of  36-­‐month  old  children  was  used  to  investigate  the  relations   between  maternal  feeding  practices,  maternal  depressive  symptoms,  and  child  aggression   and  BMI.  Bivariate  correlations  revealed  a  negative  correlation  between  maternal   depressive  symptoms  and  meal  patterns,  and  a  positive  correlation  between  maternal   depressive  symptoms  and  child  aggression.  This  is  the  first  known  study  to  report  an   association  between  symptoms  of  depression  and  meal  patterns,  which  is  suggestive  of   depressed  mothers’  difficulty  in  healthy  meal  planning  for  themselves  and  their  children.   Future  work  should  confirm  this  association  with  older  children  and  identify  whether  or   not  it  is  linked  to  child  overweight.             AKNOWLEDGEMENTS       I  would  like  to  formally  acknowledge  and  thank  all  who  assisted  in  making  this   thesis  possible;  my  advisor,  Hope  Gerde,  my  committee  members,  Holly-­‐Brophy  Herb  and   Robert  Griffore,  and  the  researcher  who  headed  the  project  from  which  this  data  came,   Rachel  Schiffman.  My  husband,  friends  and  family  also  provided  me  with  invaluable   emotional  support  during  this  process.  Thank  you  all!!!                                     iii     TABLE  OF  CONTENTS   LIST  OF  TABLES........................................................................................................................................................vi     LIST  OF  FIGURES.................................................................................................................................................... vii     INTRODUCTION.........................................................................................................................................................1   Current  Study..............................................................................................................................................................1   Theory............................................................................................................................................................................2   Hypothesis ...................................................................................................................................................................3   Research  Question  ...................................................................................................................................................4     CHAPTER  1   LITERATURE  REVIEW ............................................................................................................................................5   Child  Obesity  Overview ..........................................................................................................................................5   Health  Consequences  of  Obesity ........................................................................................................................5   Economic  Consequences  of  Obesity   .................................................................................................................6   Need  for  Obesity  Prevention................................................................................................................................7   Family  Influences  on  Food ....................................................................................................................................8            Family  Impact  on  Young  Children’s  Eating ..............................................................................................8            Foods  Consumed...............................................................................................................................................10            Food  Preferences ..............................................................................................................................................14            Feeding  Styles  and  Practices........................................................................................................................16   BMI ...............................................................................................................................................................................21            Measurement  of  Body  Weight.....................................................................................................................21            Relation  of  Child  and  Maternal  BMI..........................................................................................................22   Depression ................................................................................................................................................................23            Depression  Overview  in  Adult  Females..................................................................................................23            Depression  and  Obesity  in  Adult  Females .............................................................................................24   Aggression.................................................................................................................................................................25            Child  Aggression  Overview ..........................................................................................................................25            Child  Aggression  and  Maternal  Depressive  Symptoms ...................................................................26            Child  Aggression,  Body  Weight,  and  Parental  Feeding ....................................................................28   Summary   ...................................................................................................................................................................30     CHAPTER  2   METHODS ..................................................................................................................................................................31   Early  Head  Start  Research  and  Evaluation  Project  Data  Set ...............................................................31   Participants...............................................................................................................................................................32   Procedures ................................................................................................................................................................32   Measures....................................................................................................................................................................33            Food  Control .......................................................................................................................................................33            Meal  Patterns......................................................................................................................................................33            Child  Aggression ...............................................................................................................................................34     iv              Maternal  Depressive  Symptoms   ...............................................................................................................34            Child  BMI ..............................................................................................................................................................35            Maternal  BMI ......................................................................................................................................................35   Analysis ......................................................................................................................................................................35     CHAPTER  3   RESULTS.....................................................................................................................................................................37   Descriptive  Statistics ............................................................................................................................................37   Bivariate  Correlations..........................................................................................................................................37   Regression.................................................................................................................................................................37     CHAPTER  4   DISCUSSION..............................................................................................................................................................39   Maternal  Depressive  Symptoms  and  Child  Aggression.........................................................................39   Maternal  Depressive  Symptoms  and  Meal  Patterns ...............................................................................40   Sample  versus  National  Composition ...........................................................................................................44   Limitations ................................................................................................................................................................46            Controlling  Feeding  Practices  and  BMI...................................................................................................46            Mother  and  Child  BMI.....................................................................................................................................48   Future  Research......................................................................................................................................................49   Implications..............................................................................................................................................................51   Conclusion.................................................................................................................................................................53     APPENDICES ............................................................................................................................................................55   Theoretical  Model ..................................................................................................................................................56   Health  Section  of  Pathways  Project  Survey................................................................................................57   Fruits  and  Veggies  Section  of  Pathways  Project  Survey .......................................................................59   Factor  Analysis........................................................................................................................................................60   Subscale  of  Child  Behavior  Checklist  for  Ages  1  ½  -­‐  5 ...........................................................................61   Center  for  Epidemiologic  Studies  Depression  Scale ...............................................................................62   Frequencies  of  Continuous  Variables............................................................................................................63   Crosstabulation  of  BMI  and  Depression  for  Mothers .............................................................................64   Crosstabulation  of  BMI  and  Aggression  for  Children.............................................................................65   Bivariate  Correlations..........................................................................................................................................66     REFERENCES ...........................................................................................................................................................67             v     LIST  OF  TABLES     TABLE  1   Factor  Analysis........................................................................................................................................................60     TABLE  2   Frequencies  of  Continuous  Variables............................................................................................................63     TABLE  3   Crosstabulation  of  BMI  and  Depressive  Symptoms  for  Mothers ......................................................64   TABLE  4   Crosstabulation  of  BMI  and  Aggression  for  Children.............................................................................65     TABLE  5   Bivariate  Correlations..........................................................................................................................................66                                     vi     LIST  OF  FIGURES     FIGURE  1   Theoretical  Model ..................................................................................................................................................56                                               vii     THE  RELATIONS  BETWEEN  FEEDING  PRACTICES,  BODY  MASS  IMDEX,  DEPRESSION,  AND   AGGRESSION  IN  A  LOW-­‐INCOME  SAMPLE  OF  MOTHERS  AND  CHILDREN     INTRODUCTION   Early  childhood  is  a  crucial  period  of  growth  for  physical,  cognitive  and  psychosocial   development,  during  which  food  preferences  are  formed  (Koivisto  Hursti,  1999).  Dietary   patterns,  including  daily  intakes  of  energy,  fat  and  calcium,  tend  to  track  longitudinally   from  early  childhood  and  have  implications  for  long-­‐term  health  (Boulton,  Magarey,  &   Cockington,  1995;  Klesges,  Stein,  Eck,  Isbell,  &  Klesges,  1991).  Thus,  an  affinity  for   unhealthy,  energy-­‐dense  foods  during  childhood  would  will  likely  yield  similar  long-­‐term   preferences  in  adulthood  (Wardle,  1995).    Current  evidence  supports  the  claim  that  obese   children  tend  to  become  obese  adults  (Butte,  2009;  Freedman  et  al.,  2005;  Magarey,   Daniels,  Boulton,  &  Cockington,  2003;  Serdula  et  al.,  1993;  Whitaker,  Wright,  Pepe,  Seidel,  &   Dietz,  1997).  Therefore,  it  is  a  commonly  held  view  that  the  early  childhood  period  is  an   opportune  time  during  which  healthy  food  preferences  and  behaviors  can  develop  (Wardle,   1995).    To  study  obesity  prevention,  many  investigators  have  explored  maternal  feeding   practices  and  whether  they  influence  the  type  and  amount  of  food  children  eat,  and   ultimately,  a  child’s  adiposity  (Baughcum,  Burklow,  Deeks,  Powers,  &  Whitaker,  1998;   Birch  &  Fisher,  2000;  Faith,  Scanlon,  Birch,  Francis,  &  Sherry,  2004;  Lumeng  &  Burke,   2006).     Current  Study   While  child  obesity  has  not  been  ignored  in  research,  there  is  much  yet  to  unpack   about  the  ways  in  which  obesogenic  behaviors  develop  during  childhood.    The  ways  in   which  a  mother  feeds  her  young  child  impacts  her  child’s  ability  to  regulate  food     1     consumption  (Costanzo  &  Woody,  1985;  Satter,  1990).  In  particular,  the  degree  of  control  a   mother  exercises  when  feeding  her  child  is  of  interest  due  to  the  speculative  role  that   external  cues  have  in  limiting  the  child’s  ability  to  self-­‐regulate  his  own  food  intake  (Birch   &  Fisher,  1998;  Johnson  &  Birch,  1994).  In  addition,  maternal  depressive  symptoms  has   been  linked  to  feeding  practices  (Francis,  Hofer,  &  Birch,  2001)  and  overweight  (Surkan,   Kawachi,  &  Peterson,  2008),  and  child  overweight  is  related  to  child  behavior  (Lewinsohn   et  al.,  2005;  Young  Hyman,  Schlundt,  Herman-­‐Wenderoth,  &  Bozylinski,  2003);  therefore,   the  mental  health  status  of  mothers  and  children  are  important  dimensions  to  consider  in   feeding  practice  studies.    The  present  study  measured  maternal  control  of  child  feeding  and   maternal  depressive  symptoms  with  child  body  mass  index  and  child  aggression  in  a   sample  of  low-­‐income  mothers  and  children.     A  low  SES  sample  was  used  is  this  study  given  the  association  of  low  cost,   nutritionally  comprised  diets  and  obesity  in  low-­‐income  families  (Drewnowski,  2004;  Wigg   Dammann  &  Smith,  2009).  The  present  study  aimed  to  add  to  the  body  of  knowledge   regarding  low-­‐income  status  and  controlling  feeding  practices.  Evidence  suggests  that  the   degree  of  control  a  mother  exerts  in  the  feeding  scenario  will  affect  her  child’s  body  weight   differently  in  low-­‐income  samples  versus  middle-­‐  to  high-­‐income  samples  (Hoerr  et  al.,   2009;  Robinson,  Kiernan,  Matheson,  &  Haydel,  2001).     Theory   Based  upon  the  interest  in  the  interaction  between  mothers  and  children  in  general,   and  their  interactions  in  food-­‐related  contexts,  Bronfenbrenners’s  (1994)  human  ecological   theory  is  an  appropriate  theory  by  which  to  consider  these  relationships.  This  theory   prompts  one  to  view  an  individual  in  context.  One  could  posit  that  child  and  mother  risk     2     factors  for  obesity  or  emotional  problems  stem  in  part  from  the  interactions  of  the  five   levels  of  the  environment  (microsystem,  mesosystem,  exosystem,  macrosystem,   chronosystem).  These  levels  can  influence  each  other  bidirectionally  to  produce  an   environment  that  indirectly  or  directly  promotes  obesity  or  healthy  lifestyle,  and  mental   illness  or  mental  health.  Typically,  the  majority  of  a  young  child’s  interactions  will  be  with   family  members,  and  thus  he  will  learn  about  what  foods  to  eat  and  how  to  eat  them  from   these  people  that  comprise  his  microsystem.  While  the  microsystem  is  the  emphasis  of  the   current  study,  children  are  also  influenced  by  their  parents’  interactions  with  other   systems  and  their  society’s  emphasis  on  lifestyle  factors.     Hypothesis     Based  on  the  literature  of  maternal  feeding  practices,  maternal  control  was   hypothesized  to  be  positively  associated  with  child  BMI.  That  is,  the  more  controlling  a   mother,  as  indicated  by  her  practice  of  limiting  her  child’s  ability  to  make  his  own  food   decisions,  the  higher  the  child’s  BMI  would  be  at  36  months.  Echoing  the  work  of  Birch  and   Fisher  (2000,  2003),  Fisher  and  Birch  (1999a,  1999b,  2002),  and  Johnson  and  Birch  (1994),   maternal  control  in  child  feeding  is  negatively  related  to  a  child’s  sensitivity  to  internal   cues  of  hunger  and  satiety,  which  is  linked  to  a  child’s  ability  to  regulate  his  energy  intake.   If  a  child’s  ability  to  monitor  his  own  hunger  and  decide  how  much  to  eat  is  replaced  by  an   external  source  (in  this  case,  the  mother),  a  child  may  learn  to  pay  less  attention  to  internal   cues  to  regulate  his  energy  intake  and  be  at  a  greater  risk  of  elevated  weight  status   (Johnson  &  Birch,  1994).  However,  due  to  the  fact  that  these  general  findings  have  not  been   frequently  explored  in  a  low-­‐income  sample  and  in  this  young  age  group,  the  outcomes  may   not  generalize  across  these  populations.       3     Additionally,  maternal  BMI  and  maternal  depressive  symptoms  was  hypothesized  to   be  significantly  associated  with  child  BMI  and  child  aggression.  Mothers  who  are   overweight  or  depressed  are  speculated  to  have  poorer  child  outcomes;  that  is,  their  child   will  have  elevated  body  weight  and/or  child  behavior  problems.  This  hypothesis  targets   Bronfenbrenner’s  (1994)  assertion  that  a  child’s  development  is  affected  by  what  is   present  in  his  microsystem;  in  this  study,  a  mother’s  physical  and  mental  health  status  is   hypothesized  to  be  linked  to  her  child’s  physical  and  mental  health.       Research  Question   The  principal  research  question  for  the  present  study  is,  “What  is  the  relation   between  a  controlling  feeding  style,  maternal  depressive  symptoms,  maternal  BMI,  child   aggression  and  child  BMI?”  The  hypothesized  relations  between  these  variables  are   depicted  using  a  theoretical  model  (Appendix  A).                             4     LITERATURE  REVIEW   Child  Obesity  Overview     The  prevalence  rates  of  obesity  in  childhood  in  the  United  States  have  risen   dramatically  in  the  past  several  decades  (Ogden,  Carroll,  &  Flegal,  2008).  Data  from  the   National  Health  and  Nutrition  Examination  Survey  (NHANES)  reveal  that  from  1976–1980   to  2007–2008,  the  prevalence  rates  of  obesity  in  children  aged  two  to  five  years  grew  from   5%  to  10.4%;  rates  of  obesity  in  children  from  the  ages  of  six  to  eleven  years,  similarly,   increased  from  6.5%  to  19.6%,  over  triple  the  percentage  of  children  (Ogden,  Carroll,   Curtin,  Lamb,  &  Flegal,  2010;  Ogden  et  al.,  2008).     Additionally,  this  trend  of  increasing  body  fat  has  been  documented  in  many   countries  in  the  world,  with  10%  of  school  aged  children  in  the  world  meeting   international  criteria  for  overweight  or  obese  (Lobstein,  Baur,  &  Uauy,  2004).  Economically   developed  countries  in  the  Americas,  Europe  and  the  Middle  East  lead  the  way  in  obesity   rates,  but  urbanized  areas  in  developing  countries  have  also  reported  rising  incidence  rates   of  obesity  (Prentice,  2006).  To  illustrate  its  global  significance,  the  World  Health   Organization  has  recently  recognized  obesity  and  overweight  as  a  condition  of  epidemic   proportion  affecting  all  ages  and  socioeconomic  groups  (World  Health  Organization,  2003).   Health  Consequences  of  Obesity     Obesity  is  a  major  concern  due  to  its  impact  on  health.  Physical  ailments  attributed   to  an  overweight  health  status  in  adults  are  type  2  diabetes,  hypertension,  atherosclerosis,   sleep  disorders,  nonalcoholic  steatohepatitis,  arthritis,  depression  and  certain  types  of   cancer  (Cossrow  &  Falkner,  2004).  Formerly,  chronic  conditions  resulting  from  obesity   were  thought  to  be  reserved  for  adults,  yet  this  notion  is  becoming  increasingly  challenged.     5     Type  2  diabetes,  for  instance,  is  being  more  commonly  diagnosed  in  the  pediatric   population,  comprising  as  many  as  50%  of  newly  diagnosed  diabetes  cases  (Rosenbloom,   2002).  Another  example  that  challenges  the  preconception  that  obesity-­‐related  physical   ailments  apply  only  to  adults  arose  from  a  population  based  sample  of  children  and   adolescents,  which  revealed  that  obesity  accounted  for  the  highest  rate  of  nonalcoholic   steatohepatitis  diagnoses  amongst  those  surveyed  (Schwimmer  et  al.,  2006).  Therefore,   when  considering  these  medical  complications,  it  is  prudent  to  say  that  overweight  and   obesity  could  affect  individuals  across  the  lifespan.   Economic  Consequences  of  Obesity     The  many  health  complications  of  obesity  related  conditions  pose  economic   consequences  on  the  health  care  system.  Based  on  data  from  the  1998  Medical  Expenditure   Panel  Survey  and  the  1996  and  1997  National  Health  Interview  Surveys,  estimates  of   medical  spending  attributable  to  obesity  and  overweight  related  conditions  in  1998  arrive   at  the  sum  of  78.5  billion,  or  9.1%  of  total  annual  medical  expenditures  (Finkelstein,   Fiebelkorn,  &  Wang,  2003).  Though  the  current  costs  of  obesity  are  not  trifling,  future   projections  of  health  care  costs  are  even  more  staggering.  If  existing  prevalence  rates  of   obesity  and  overweight  remain  unchecked,  spending  would  be  expected  to  double  every  10   years  until  2030,  reaching  as  high  as  860.7  to  956.9  billion  dollars,  or  16%  to  18%  of  total   health  care  costs  (Wang,  Beydoun,  Liang,  Caballero,  &  Kumanyika,  2008).      Beyond  the  realm  of  systemized  health  care,  there  are  additional  costs  that  burden   those  with  obesity.  In  reference  to  diabetes  and  obesity,  Yach,  Stuckler  and  Brownell   (2006)  claim  that  the  microeconomic  costs  to  sick  individuals,  families  and  communities   could  be  more  taxing  than  the  costs  of  the  actual  conditions.  Indirect  costs  include  lower     6     returns  on  education,  decreased  income,  earlier  retirement  and  unemployment,  and   greater  dependency  on  the  welfare  system  (Yach  et  al.,  2006).  In  short,  obesity  imposes   immediate  and  long-­‐term  economic  consequences  that  affect  the  individual  with  obesity  at   a  societal  and  personal  level.     Need  for  Obesity  Prevention     The  health  and  economic  issues  remain  central  to  obesity  in  childhood  due  to  the   finding  that  obese  children  have  a  statistically  greater  risk  of  becoming  obese  adults  than   do  normal  weight  children  (Butte,  2009;  Freedman  et  al.,  2005;  Magarey  et  al.,  2003;   Serdula  et  al.,  1993;  Whitaker  et  al.,  1997).  In  addition,  obesity  is  difficult  to  treat  once  it  is   present  (Flodmark,  Marcus,  &  Britton,  2006).  Some  predictors  of  obesity  at  the  age  of  seven   years  are  based  on  child  characteristics  before  the  age  of  three;  these  include  birth  weight,   parental  obesity,  more  than  eight  hours  of  watching  television  per  week,  rapid  weight  gain   in  the  first  12  months,  and  short  sleep  duration  (Reilly,  et  al.,  2005;  Butte,  2009).  The   majority  of  these  factors  remain  significant  in  predicting  adult  adiposity  from  child  factors,   along  with  the  addition  of  low  socioeconomic  status  during  middle  childhood  (Parsons,   Power,  Logan,  &  Summerbell,  1999).  Whitaker  et  al.  (1997)  found  that  for  children  under   three  years  of  age,  the  strongest  predictor  of  obesity  during  adulthood  was  parental   obesity.  All  of  these  factors  support  the  idea  that  the  environment  of  a  child’s  early  years   can  influence  the  risk  of  obesity  later  in  life  (Butte,  2009).         Taking  into  consideration  the  health  and  economic  effects  of  obesity,  efforts  to   prevent  the  onset  of  overweight  have  gained  momentum  (Kranz,  Hartman,  Siega-­‐Riz,  &   Herring,  2006;  Hill  &  Peters,  1998).  These  efforts  exist  in  part  to  prevent  harm  to  the   person  at  the  level  of  the  individual,  but  also  to  limit  the  negative  societal  impacts  of     7     obesity  (Flodmark  et  al.,  2006).  In  some  industrialized  nations  with  high  rates  of  obesity,   national  policies  have  been  enacted  which  specify  primary  healthcare  providers  as  the   sentry  for  identification  and  counseling  for  childhood  overweight  (Wake,  Gold,  McCallum,   Gerner,  &  Waters,  2008).  However,  some  experts  suggest  that  the  most  logical  places  to   begin  prevention  efforts  are  the  home  and  school  environments  (Lobstein  et  al.,  2004).   Support  for  school-­‐based  prevention  efforts  is  reflected  in  the  results  of  a  recent  survey  of   parents  in  the  USA,  who  report  the  conviction  that  schools  should  have  more  responsibility   in  preventing  childhood  obesity  than  health  care  providers  or  the  government  (Evans,   Finkelstein,  Kamerow,  &  Renaud,  2005).  Evaluators  of  school  obesity  prevention  programs   have  traced  numerous  peer-­‐reviewed  studies  with  BMI  as  a  chief  measure  of  the  health  of   children  and  youth  from  as  early  as  1985  (Budd  &  Volpe,  2006).  Systematic  reviews  have   yielded  vastly  mixed  results  regarding  the  effectiveness  of  obesity  prevention  programs,   with  only  several  studies  producing  modest  statistically  significant  findings  (Budd  &  Volpe,   2006;  Kropski,  Keckley,  &  Jensen,  2008;  Thomas,  2006).  The  conflicted  findings  and  lack  of   a  strong  consensus  surrounding  ‘what  works’  in  schools  to  prevent  obesity  leads  one  to  the   conclusion  that  prevention  efforts  might  best  be  targeted  to  a  child’s  initial  social   environment;  in  most  cases,  this  would  be  the  family.     Family  Influences  on  Food       Family  Impact  on  Young  Children’s  Eating:  Childhood  obesity  is  a  complex  disease   arising  from  the  interaction  of  a  genetic  predisposition  toward  obesity  and  an  unregulated   environment  (Butte,  2009).  However,  given  the  rising  accounts  of  obesity  within  a   relatively  stable  gene  pool,  the  main  source  of  the  increase  in  body  mass  likely  stems  from   changing  factors  within  the  environment,  such  as  food  availability  and  portion  size     8     (Prentice,  1997;  Hill  &  Peters,  1998).  How  decisions  are  made  regarding  these  food   considerations,  including  what  food  is  suitable  for  consumption  and  methods  of   preparation,  is  founded  on  one’s  culture  (Birch  &  Fisher,  1995;  Rozin  &  Vollmecke,  1986;   Nestle  et  al.,  1998).  The  media,  peers  and  family  are  constituent  parts  of  culture,  and   therefore,  have  an  influence  on  what  and  how  a  child  eats  (Nestle  et  al.,  1998).  This   environmental  influence  is  also  shaped  by  the  media’s  messages  regarding  ideal  body  size,   with  thinness  for  females  and  musculature  for  males  typically  considered  the  pinnacle  of   attractiveness  (Agliata  &  Tantleff-­‐Dunn,  2004;  Brownell,  2001;  Grabe,  Ward,  &  Hyde,   2008).  These  body  stereotypes  are  readily  evident  in  some  forms  of  children’s  media   (Herbozo,  Tantleff-­‐Dunn,  Gokee-­‐Larose,  &  Thompson,  2004).     Though  children’s  food  consumption  is  influenced  by  the  food  decisions  of  their   peers,  the  strongest  influence  for  a  young  child  is  said  to  be  the  family  (Lau,  Quadrel,  &   Hartman,  1990;  Birch  &  Fisher,  1998).  Generally  speaking,  parents  exert  influence  on  their   child’s  behavior,  and  weight-­‐related  behaviors  such  as  physical  activity  and  eating  style  are   impacted  accordingly  (Carnell  &  Wardle,  2007).  This  is  related  to  the  results  of  a   representative  survey  of  American  adults,  which  revealed  that  approximately  91%  of   respondents  believed  that  parents  played  the  strongest  role  in  reducing  child  overweight   (Evans  et  al.,  2005).  The  strength  of  a  parent’s  influence  is  especially  significant  when  one   considers  the  multiple  roles  that  parents  fulfill.  McCaffree  (2003)  conceptualized  parents  as   providers,  enforcers,  role  models,  protectors  and  advocates  with  respect  to  a  child’s  eating   and  physical  activity  and  thus,  weight.  The  ubiquitous  influence  of  parents  in  the  lives  of   their  children  has  lead  researchers  to  the  assumption  that  parental  behaviors  are  a     9     modifiable  element  of  a  child’s  environment  for  the  purposes  of  preventing  or  reducing   child  obesity  (Carnell  &  Wardle,  2007).     However,  research  suggests  that  mothers  exert  particular  influence  on  the   determinants  of  children’s  food  intake,  such  as  how,  when  and  why  children  are  fed   (Baughcum,  et  al.,  2001).  Mothers  more  typically  feed  their  children,  purchase  food,   prepare  meals  and  are  concerned  about  the  foods  they  feed  their  families  (Baranowski,   1997;  Brown  &  Miller,  2002;  Devine,  Connors,  &  Sobal,  2003).  Indeed,  a  mothers’  diet  has   been  found  to  be  the  strongest  predictor  of  her  child’s  diet  (Horodynski,  Stommel,  Brophy-­‐ Herb,  Xie,  Weatherspoon,  2010).  As  a  result,  most  studies  investigating  the  familial   influence  on  children’s  eating  have  focused  on  the  dyadic  relationship  of  the  mother  and   child.  Yet  despite  the  developmental  importance  of  early  childhood,  Lee,  Hoerr  and   Schiffmann  (2005)  state  that  there  have  been  infrequent  investigations  of  the  feeding   practices  of  mothers  of  young  children,  with  most  studies  focusing  on  older  children.  The   present  study  fills  a  gap  in  the  literature  as  it  focuses  on  mothers’  feeding  practices  of  their   36-­‐month-­‐old  children.             Foods  Consumed:  As  previously  stated,  an  overweight  or  obese  weight  status  can   arise  due  to  genetics  and  environmental  influences  (Butte,  2009).  Of  the  environmental   influences,  lifestyle  choices  such  as  physical  inactivity  and  poor  diet  are  regarded  as  the   most  important  contributors  to  an  overweight  status  (US  Department  of  Health  and  Human   Services  and  US  Department  of  Agriculture,  2005).  Interestingly,  some  researchers  have   recently  re-­‐evaluated  the  benefit  of  physical  activity  to  reduce  the  energy  gap,  as  some   experts  consider  food  intake  to  be  a  more  significant  variable  in  determining  body  weight   (Sonneville  &  Gortmaker,  2008).  Yet  however  much  weight  is  assigned  to  each  factor,  it  is     10     generally  agreed  that  an  overweight  or  obese  status  arises  from  “long-­‐term  net  energy   imbalances”  (Faith  et  al.,  2004,  p.  1719).       Assessment  of  dietary  quality  is  not  known  for  its  use  of  ease  in  measurement  (Kranz,   Hartman,  Siega-­‐Riz,  &  Herring,  2006).  Analysis  of  dietary  quality  has  undergone  revision  in   how  it’s  measured  over  the  past  several  decades,  with  a  shift  from  time-­‐intensive  nutrient   adequacy  towards  a  food  measurement  approach  (Hoerr,  Horodynski,  Lee,  &  Henry,  2006).   Consumption  of  food  is  typically  assessed  according  to  the  guidelines  recommended  in   MyPyramid,  the  USDA  food  guide  (Dixon,  Cronin,  &  Krebs-­‐Smith,  2001),  the  Dietary   Reference  Intakes  of  the  National  Academy  of  Science,  the  American  Dietetic  Association,   and/or  in  the  case  of  preschoolers,  the  American  Academy  of  Pediatrics  (Kranz  et  al.,  2006;   Kranz,  Siega-­‐Riz,  &  Herring,  2004).  Typical  recommendations  of  a  healthy  diet  include  an   emphasis  on  eating  a  variety  of  fruits  and  vegetables,  whole  grains  and  low-­‐fat  dairy   products,  including  a  variety  of  lean  protein  products,  and  limiting  saturated  and  trans  fats,   sodium,  and  added  sugars  (US  Department  of  Health  and  Human  Services  and  US   Department  of  Agriculture,  2005).       This  is  pertinent  information,  as  tracking  the  food  consumption  of  individuals  in  high-­‐ risk  groups,  such  as  those  in  low  SES  populations,  can  help  identify  those  in  danger  of   developing  health  problems  (Hoerr  et  al.,  2006).  The  health  benefits  of  a  high-­‐quality  diet   are  typically  related  to  chronic  disease  prevention  (US  Department  of  Health  and  Human   Services  and  US  Department  of  Agriculture,  2005).  For  instance,  epidemiological  evidence   has  revealed  that  a  diet  high  in  fruits  and  vegetables  can  be  considered  a  protective   mechanism  against  certain  cancers,  heart  disease  and  stroke,  in  addition  to  emerging   evidence  for  the  prevention  of  cataracts,  chronic  obstructive  pulmonary  disease,     11     diverticulosis  and  hypertension  (Van  Duyn  &  Pivonka,  2000).  Since  fruits  and  vegetables   are  a  significant  source  of  fiber,  many  of  the  health  benefits  of  this  food  group  extend  to   those  who  consume  optimal  quantities  of  fiber,  in  addition  to  a  reduced  risk  of  developing   type  2  diabetes  and  the  management  of  gastrointestinal  disorders  (Anderson,  Smith,  &   Gustafson,  1994;  Van  Duyn  &  Pivonka,  2000).  Fiber  can  also  be  consumed  through  whole   grains,  with  greater  intakes  of  whole  grains  being  linked  to  reduced  incidence  of   cardiovascular  disease  (Mellen,  Walsh,  &  Herrington,  2008).  Those  who  consume  the   recommended  amounts  of  calcium,  commonly  consumed  through  dairy  products  in  the   USA,  have  the  benefit  of  increasing  bone  mass  and  reducing  the  risk  of  developing   osteoporosis  later  in  life  (Weaver,  2000).  Additionally,  adhering  to  these  eating   recommendations  is  said  to  be  beneficial  for  weight  loss  or  weight  maintenance  based  on   the  healthy  properties  of  these  foods  (Anderson  et  al.,  1994).     Relevant  to  the  current  discussion  is  the  assertion  that  diet  is  a  key  part  in  the   etiology  of  childhood  overweight  and  obesity  (Fisher  &  Birch,  2001).  Kranz  et  al.,  (2004)   compared  the  diet  quality  of  preschool  children  between  1978  and  1998  and  found  that   dietary  quality  had  small  but  significant  improvements  over  time.  Yet  the  authors  pointed   out  that,  “Children  consuming  more  food,  and  thus  more  energy,  are  more  likely  to  meet   the  intake  recommendations  for  food  groups  and  nutrients  compared  with  children  who   eat  less”  (Kranz  et  al.,  2006,  p.  1596).  Thus,  a  higher  dietary  quality,  which  included  an   increase  in  the  consumption  of  fruit  juices  and  sugar,  yielded  an  overall  energy  increase.   This  change  was  suggested  as  a  partial  explanation  for  the  increased  prevalence  rates  of   child  obesity  (Kranz  et  al.,  2004).  The  finding  that  children  currently  consume  surplus   energy  was  echoed  by  Wang,  Gortmaker,  Sobol,  and  Kuntz  (2006).  These  authors  cited  the     12     energy  difference  gap  of  110-­‐167  extra  calories  a  day  in  children  aged  2-­‐7  over  a  ten  year   period  as  a  contributor  to  the  increase  in  overall  body  weight  among  US  children.         As  previously  established,  mothers  typically  provide  what  is  available  for  children  to   consume  (Barankowski,  1997).  Low-­‐income  mothers  tend  to  have  less  knowledge  of  the   nutritional  content  of  foods  as  compared  to  their  middle-­‐income  and  high-­‐income   counterparts,  and  some  purportedly  lack  the  knowledge  that  diet  is  a  contributor  in  the   development  of  some  chronic  diseases  (Barankowski,  1997;  Wiig  Dammann  &  Smith,   2009).  By  analyzing  compliance  with  the  recommended  daily  consumption  guideline  of  at   least  one  serving  of  the  five  foods  groups  from  MyPyramid,  Lee  et  al.  (2005)  found  high   agreement  between  poor  quality  diets  of  low-­‐income  mothers  and  their  children.  This   general  finding  was  consistent  in  a  study  of  beverage  intake  in  low-­‐income  families  with   children  aged  2-­‐3  years,  which  revealed  a  high  concordance  between  sweetened  beverages   and  soft  drinks  in  mother-­‐child  dyads  (Hoerr,  Lee,  Schiffman,  Horodynski  &  McKelvey,   2006).  Focus  groups  of  low-­‐income  mothers,  of  which  over  75%  were  either  overweight  or   obese,  revealed  that  these  women  wanted  to  feed  their  families  more  nutritious  foods  but   cited  the  expense  of  healthy  foods  as  the  main  deterrent  from  doing  so;  some  also  espoused   quantity  of  food  over  quality  of  food  (Wiig  Dammann  &  Smith,  2009).  This  revelation  can   be  implicated  in  the  development  of  an  energy  imbalance,  as  many  low-­‐cost  food  items   consumed  by  low-­‐income  families  tend  to  be  high  in  calories  and  composed  of  refined   grains,  added  sugars,  and  added  fat  (Drewnowski,  2004).  Therefore,  children  of  low-­‐ income  mothers  could  be  more  at  risk  of  inadequate  nutrient  consumption  and  excessive   energy  intake  and  less  likely  to  receive  the  preventative  benefits  of  good  nutrition  than   children  in  other  income  categories.         13     Food  Preferences:  Food  preferences  are  largely  based  on  familiarity  and  taste  of   foods  (Birch,  1992).  Food  preferences  are  relevant  to  consider  in  what  children  eat  due  to   the  finding  that  food  preferences  predict  consumption  (Birch  &  Fisher,  1998;  Skinner,   Carruth,  Bounds,  &  Zeigler,  2002)  and  these  preferences  develop  through  repeated   exposure  to  new  foods  when  infants  transition  away  from  a  milk-­‐based  diet  (Birch,  1998).   Studies  have  shown  that  early  exposure  to  fruits  and  vegetables,  and  to  energy  dense,   sugary  and  fatty  foods  play  a  part  in  establishing  a  hierarchy  of  food  preferences  (Hearn,   Baranowski,  &  Baranowski,  1998).  However,  children  generally  learn  to  prefer  energy   dense  foods  versus  energy  dilute  foods  (Birch,  1992;  Birch,  1999).  An  evolutionary   perspective  could  help  explain  why  this  is;  it  could  be  considered  an  adaptation  to  prefer   foods  that  contain  higher  concentrations  of  energy  during  times  when  food  is  scarce  (Birch,   1998)  or  to  maintain  the  energy  requirements  of  growing  children  (Birch,  1999).   Some  innate  preferences  are  present  at  birth,  such  as  an  affinity  for  sweet  foods  and   an  aversion  for  bitter  and  sour  foods;  others,  such  as  salty  foods,  arise  early  in  the  infant’s   life  (Birch,  1999).  Food  preferences  are  also  learned  through  associative  conditioning,   wherein  foods  offered  in  positive  conditions  may  increase  affinity,  while  foods  offered  in   negative  contexts  may  decrease  affinity  (Birch,  1998).  Associative  learning  also  occurs  via   the  physiological  responses  to  food.  Aversion  can  be  learned  through  negative   physiological  responses,  such  as  gastrointestinal  illness  (Garcia,  Hankins,  &  Rusinak,  1974),   while  preference  can  be  established  through  the  association  of  particular  flavors  with  “the   pleasant  postingestive  signals  generated  by  normal  satiety”  (Birch,  1999,  p.  54).  These   signals  involve  the  activation  of  such  neurotransmitter  systems  as  dopamine,  opioid  and   benzodiazepine/GABA  (Gibson,  2006).       14     Children  eat  what  they  like  and  what  is  familiar  to  them;  typically,  this  is  what  is   available  and  normative  in  their  every  day  environments  (Birch  &  Fisher,  1998;  Cooke,   2007).  Mothers  tend  to  stock  their  homes  with  food  that  they  like  and  eat,  and  this  food   environment  shapes  children’s  preferences  (Birch  &  Fisher,  1998;  Wardle,  1995).  Mothers   with  limited  economic  resources  have  fewer  options  regarding  food  they  provide  for  their   families,  and  report  choosing  food  that  ultimately  contributes  to  unhealthy  diets  due  to   their  low  cost  (Wiig  Dammann  &  Smith,  2009),  proximity  (Cummins  &  Macintyre,  2006)  or   fast  and  convenient  preparation  (Engler-­‐Stringer,  2010).  Low-­‐income  mothers  are  also   likely  to  prepare  meals  that  are  of  less  complexity  than  middle-­‐income  mothers,  which  is   accounted  for  by  either  a  reduced  selection  from  which  to  choose  foods  to  cook  or  reduced   food  preparation  skills  (McLaughlin,  Tarasuk,  &  Kreiger,  2003).  Low-­‐income  women  also   report  consuming  unhealthy  foods  due  to  their  convenience  and  ease  of  preparation,  as   well  a  lack  of  their  own  personal  time  and  feelings  of  stress  (Chang,  Nitzke,  Guilford,  Adair,   &  Hazard,  2008).   Regardless  of  the  varying  degrees  of  food  availability,  the  food  preferences  of   children  are  demonstrated  to  be  associated  with  the  food  preferences  of  their  mothers   (Horodynski  et  al.,  2010).  In  a  longitudinal  analysis,  Skinner  et  al.  (2002)  found  significant   correlations  between  foods  liked,  disliked  and  never  tried  between  mothers  and  their   children.  Brown  and  Ogden  (2004)  examined  snack  food  consumption  and  found  that   parents  and  children  ate  many  unhealthy  snack  foods,  and  that  there  was  a  strong   association  between  these  groups  for  snacks  in  general  and  in  the  foods  consumed  during   the  last  24  hours.  Even  though  there  are  some  physiological  components  involved  in  food     15     preferences,  children’s  learned  food  preferences  are  largely  shaped  through  what  foods  are   present  in  the  environment  and  mothers’  own  food  preferences.       Feeding  Styles  and  Practices:  Another  way  in  which  a  mother  influences  her  child’s   eating  is  the  method  or  pattern  of  behavior  by  which  she  feeds  her  child  (Moore,  Tapper  &   Murphy,  2007).  Two  paradigms  in  the  study  of  mothers’  behaviors  in  relation  to  child   eating  and  weight  have  emerged  (Hoerr  et  al.,  2009).  Some  researchers  align  an  overall   parenting  style  with  a  feeding  style,  while  others  focus  on  a  prescribed  set  of  feeding   practices.  Unfortunately,  these  different  approaches  have  created  some  confusion  in  the   research  community  (Hoerr  et  al.,  2009),  as  researchers  themselves  often  fail  to  make   distinctions  about  which  lens  they  are  using  to  study  the  general  theme  of  how  mothers   affect  the  weight  of  their  child  (Ventura  &  Birch,  2008).    Parenting  styles  are  understood  as  the  general  type  of  attitudes  and  behaviors   parents  have  that  characterize  their  interactions  with  their  child  across  domains  (Darling  &   Steinberg,  1993).  In  general  it  is  understood  that,  “Parenting  styles  have  an  indirect  effect   on  children's  outcomes:  parenting  styles  moderate  the  effect  of  parenting  practices  because   they  influence  the  effectiveness  of  specific  parenting  practices”  (Ventura  &  Birch,  2008,  p.   3).  Using  Baumrind’s  (1966)  parenting  styles,  several  investigators  have  related  these   profiles  to  feeding  styles  and  have  used  these  classifications  in  their  studies  (Hughes,   Power,  Fisher,  Mueller,  &  Nicklas,  2005;  Patrick  &  Nicklas,  2005;  Rhee,  Lumeng,   Appugliese,  Kaciroti,  &  Bradley,  2006).  Authoritarian  feeding  (high  demandingness/low   responsiveness)  is  characterized  by  forcing  a  child  to  eat  certain  foods  and  to  restrict  other   foods;  this  controlling  style  of  feeding  does  not  take  into  account  a  child’s  ability  to  choose   or  their  specific  preferences.  A  permissive  feeding  style  (low  demandingness/high     16     responsiveness)  is  distinguished  by  an  unstructured  feeding  setting  in  which  a  child  is   allowed  to  eat  whatever  they  desire  while  parents  maintain  sensitivity  to  the  child.  The   authoritative  style  of  feeding  (high  demandingness/high  responsiveness)  is  typified  as  a   compromise  between  parent  boundaries  and  child  choice,  wherein  a  child  is  given  choice  to   select  foods  from  a  variety  of  foods  made  available  by  parents  (Birch  &  Fisher,  1995).  Some   have  also  included  an  uninvolved  or  neglectful  feeding  style  (low  demandingness/low   responsiveness),  which  is  characterized  by  a  lack  of  control  in  the  feeding  scenario  and   limited  involvement  with  the  child  (Birch  &  Fisher,  1995;  Hughes  et  al.,  2005;  Rhee  et  al.,   2006).     While  studies  exploring  feeding  styles  are  generally  limited  (Hoerr  et  al.,  2009;   Hughes  et  al.,  2005),  some  results,  albeit  slightly  inconsistent,  have  been  obtained.  As   would  be  expected,  Rhee  et  al.  (2006)  found  that  children  of  authoritarian  mothers  had  the   highest  risk  of  overweight  compared  to  children  whose  parents  had  different  parenting   styles.  However,  children  of  permissive  and  neglectful  mothers  were  also  at  risk  of  being   overweight,  as  it  was  found  that  these  children  were  twice  as  likely  to  be  overweight  as   children  from  authoritative  mothers  (Rhee  et  al.,  2006).  Contrary  to  what  was  expected,   Hughes  et  al.  (2005)  found  that  children  had  significantly  higher  ratings  of  BMI  when  their   parents  were  permissive  in  their  feeding  styles  rather  than  authoritarian.  Similarly  at  odds   with  what  is  typically  found,  Hoerr  et  al.  (2009)  found  that  children  from  a  low-­‐income   population  had  better  eating  behaviors  when  their  mothers  demonstrated  an  authoritarian   feeding  style  rather  than  an  authoritative  feeding  style;  also  found  was  that  children  with   the  highest  BMI  scores  tended  to  have  parents  who  were  permissive  in  style  and  children   with  the  lowest  BMI  scores  had  parents  who  were  authoritarian  in  style.  Recent  work  by     17     Topham  et  al.  (2009)  included  an  examination  of  the  moderating  effects  of  maternal   depressive  symptoms  and  SES  on  the  association  of  authoritarian  and  permissive  feeding   styles  and  child  overweight.  This  cross-­‐sectional  examination  of  mothers  of  first-­‐grade   children  found  maternal  depressive  symptoms  and  income  status  moderated  a  permissive,   but  not  authoritative,  parenting  style  on  child  obesity.  These  differences  could  be  due  in   part  to  the  diversity  of  income  and  racial  backgrounds  in  the  families  who  participated  in   these  studies.  This  sample  of  feeding  style  studies  serves  to  indicate  that  highly  controlling   feeding  styles  are  not  always  linked  to  higher  child  weight  status.     In  addition,  there  has  been  challenge  in  merely  linking  the  four  parenting  styles  to   child  weight  status  given  the  insignificant  results  found  by  some  researchers.  Blisset  and   Haycraft  (2008)  found  no  association  between  authoritarian  parenting  style  and   controlling  feeding  practices,  but  did  find  that  permissive  mothers  tended  to  use  a   restrictive  strategy  and  permissive  fathers  pressured  their  children  to  eat.  Interestingly,   they  also  found  that  parenting  styles  were  not  associated  with  child  BMI.  As  further   support  to  the  idea  that  feeding  styles  do  not  always  map  onto  Baumrind’s  (1966)   taxonomy  of  parenting  styles,  Hughes  et  al.  (2005)  reported  many  inconsistencies  when   applying  these  categories  to  African  American  and  Hispanic  samples.     The  more  common  way  that  researchers  have  studied  how  mothers  influence  their   child’s  eating  is  through  an  examination  of  feeding  practices  (Hughes  et  al.,  2005).  Whereas   feeding  styles  are  more  consistent,  feeding  practices  are  less  apt  to  be  characterized  and   can  change  depending  on  the  context;  this  includes  potential  change  according  to  different   children  in  the  same  family,  as  well  as  a  child’s  gender,  age,  weight  status  and  eating   behavior  (Ventura  &  Birch,  2008).  Parenting  practices  are  regarded  as  behaviors  parents     18     use  to  get  their  child  to  do  something  specific;  in  the  context  of  the  feeding  scenario,   feeding  practices  are  used  to  encourage  or  restrict  food  consumption  (Ventura  &  Birch,   2008).  Specific  practices  to  aid  in  these  endeavors  include  modeling,  repeating  a  child’s   exposure  to  a  food,  offering  rewards,  monitoring,  pressure  to  eat,  and  food  restriction   (Baughcum  et  al.,  2001;  Birch  &  Fisher,  1998;  Hoerr  et  al.,  2009;  Moore  et  al.,  2007).       Restricting  access  to  certain  foods  and  pressuring  children  to  finish  their  food  (i.e.   “clean  your  plate”),  have  been  linked  to  eating  in  the  absence  of  hunger,  disinhibited  eating,   and  overweight  (Birch  &  Fisher,  2000;  Birch,  Fisher,  &  Davison,  2003;  Faith  et  al.,  2004;   Fisher  &  Birch,  1999a;  Fisher  &  Birch,  1999b;  Fisher  &  Birch,  2002;  Johnson  &  Birch,  1994).   Some  researchers  speculate  that  the  strongest  link  to  weight  gain  in  childhood  is  through   the  practice  of  dietary  restraint  (Fisher  &  Birch,  1999a;  Faith  et  al.,  2004).  This  practice   involves  restricting  a  child’s  access  to  a  highly  desirable  food,  which  tends  to  increase   consumption  of  the  food  after  the  restraint  is  removed.  In  relation  to  this  is  the  idea  that  a   high  degree  of  maternal  control  during  feeding  translates  to  a  child  learning  to  pay   attention  to  the  amount  of  food  on  their  plate  and  to  disregard  internal  cues  of  hunger  and   satiety;  a  child’s  responsiveness  to  the  caloric  density  of  food  is  disrupted  and  can  result  in   energy  imbalances  (Costanzo  &  Woody,  1985;  Johnson  &  Birch,  1994;  Birch  &  Fisher,   1998).  This  theory  was  explored  in  an  intervention  study  by  Johnson  (2000),  wherein   children  who  had  greater  adiposity  struggled  with  being  able  to  improve  in  the  regulation   of  their  energy  intake.  However,  the  field  is  complicated  by  the  differences  in   measurement,  as  some  researchers  tend  to  investigate  feeding  practices  by  focusing  on   specific  domains  (e.g.  restraint)  or  through  a  general  conceptualization  of  feeding  control   (eating  all  food  on  plate,  eating  only  at  mealtimes,  encouraging  child  to  eat  foods  good  for     19     him/her,  not  allowing  child  to  play  with  food,  pushing  the  child  to  eat  more,  letting  child   control  feeding  interaction,  etc.;  Baughcum  et  al.,  2001;  Faith  et  al.,  2004,  Robinson  et  al.,   2001).     As  previously  mentioned,  there  has  been  some  disagreement  regarding  the  finding   that  highly  controlling  feeding  practices  have  detrimental  outcomes  on  a  child’s  weight   status.  The  studies  upon  which  the  basic  assumption  that  a  mother’s  highly  controlling   feeding  practices  leads  to  a  child’s  inability  to  regulate  his  energy  intake  and  subsequent   increase  in  adiposity  has  been  concluded  based  on  samples  of  mostly  middle-­‐  to  high-­‐ income  Caucasian  mothers  (Birch  &  Fisher,  2000;  Johnson  &  Birch,  1994).  Therefore,  there   has  been  some  speculation  about  the  generalizability  of  these  findings  to  other  income  or   racial  groups  (Clarke  et  al.,  2007;  Faith  et  al.,  2004;  Hoerr  et  al.,  2009;  Robinson  et  al.,   2001).  For  instance,  Baughcum  et  al.  (2001)  utilized  a  sample  of  low-­‐income  and  middle-­‐ income  normal  weight  and  overweight  mothers  and  found  that  no  particular  feeding   practice  was  associated  with  overweight  in  young  children.  Robinson  et  al.  (2001)  explored   ethnicity  and  income  by  utilizing  a  diverse  sample  of  third  grade  children  and  found  no   evidence  to  support  the  finding  that  controlling  feeding  practices  were  positively  related  to   child  BMI.  Similarly,  high  levels  of  maternal  control  in  feeding  practices  were  not  found  to   relate  to  child  adiposity  in  several  low-­‐income  samples  of  Hispanic  children  (Contento,   Zybert,  &  Williams,  2005;  Melgar-­‐Quinonez  &  Kaiser,  2004).  In  a  sample  containing   Caucasian  and  American  Indian  families,  Topham  et  al.  (2009)  found  permissive  but  not   authoritarian  parenting  moderated  child  obesity  among  high  SES  mothers.  Hughes  et  al.   (2005)  reported  significant  findings  in  their  low-­‐income  sample  of  African  American  and   Hispanic  families,  but  it  was  the  opposite  of  what  had  been  found  previously;  a  high  degree     20     of  feeding  control  resulted  in  children  with  lower  BMI  scores.  Yet  despite  these  exceptions,   the  predominance  of  the  extant  literature  remains  largely  based  on  middle-­‐class  Caucasian   samples  and  the  consensus  of  several  literature  reviews  is  that  of  support  for  the   association  of  maternal  food  restriction  with  higher  caloric  food  intake  and  body  weight  in   children  (Clark  et  al.,  2007;  Faith  et  al.,  2004).  Nevertheless,  more  work  is  still  needed  to   explore  the  connection  between  a  controlling  feeding  practices  and  child  obesity  in   economically  and  racially  diverse  samples.   BMI   Measurement  of  Body  Weight:  The  measurement  of  adiposity,  or  body  fat,  can  be   assessed  through  several  different  methods.  These  methods  vary  in  their  complexity,   accuracy  and  ease  of  administration.  Direct  measurement  methods  of  body  fat  include   underwater  weighing,  magnetic  resonance  imaging,  computed  tomography,  and  dual   energy  x-­‐ray  absorptiometry  (DXA),  bioelectrical  impedus  analysis  and  air  displacement   plethysmography.  These  methods  are  regarded  as  highly  accurate  estimates  of  body  fat  and   are  typically  performed  in  laboratory  settings  (Lobstein,  Baur,  &  Uauy,  2004).  However,   these  methods  are  costly,  complicated  and  potentially  unsafe  to  children  (Pietrobelli  et  al.,   1998).  Anthropometric  data  provide  an  indirect  assessment  of  body  fat.  These  methods   include  weight  and  weight  for  height,  waist  circumference  and  waist-­‐to-­‐hip  ratio,  skinfold   thickness  and  BMI  (Lobstein  et  al.,  2004).     Of  the  anthropometric  methods,  the  most  commonly  used  way  to  approximate  one’s   body  fat  in  non-­‐laboratory  settings  is  through  BMI  (Whitaker  et  al.,  1997).  This  is  a  ratio  of   weight  in  kilograms  to  the  square  of  height  in  meters,  and  correlates  with  other  measures   of  adiposity  (Krebs,  et  al.,  2003).  For  children,  growth  charts  available  through  the  Centers     21     for  Disease  Control  and  Prevention  (CDC;  2009a)  indicate  that  BMI  between  the  85th  and   95th  percentile  for  age  and  sex  is  considered  overweight,  and  BMI  at  or  above  the  95th   percentile  is  considered  obese.  Some  advantages  of  using  BMI  to  approximate  body  fatness   are  that  it  is  affordable,  safe,  easy  to  obtain,  and  suitable  for  field  studies  (Pietrobelli  et  al.,   1998).  However,  these  methods  are  not  a  precise  measure  of  fatness  because  they  do  not   incorporate  other  component  of  a  person’s  weight,  such  as  muscle  mass  and  bone  weight,   into  the  calculation  (Krebs,  et  al.,  2003).  Despite  this  limitation,  BMI  has  been  validated  as  a   fatness  measure  in  children  and  adolescents  through  its  strong  association  with  DXA,  the   more  stringent  adiposity  measure  (Pietrobelli  et  al.,  1998).   Relation  of  Child  and  Maternal  BMI:  Based  on  the  shared  food  environment  and   genetics,  one  might  suppose  that  the  weight  status  of  mothers  and  their  offspring  would  be   associated  with  one  another.  Studies  investigating  the  association  between  the  weight   status  of  mothers  and  young  children  have  been  infrequent  thus  far,  but  of  the  studies  that   have  been  conducted,  some  inconsistencies  have  been  found  (Stunkard,  Berkowitz,   Stallings,  &  Cater,  1999).  Stunkard  et  al.  (1999)  conducted  a  brief  review  of  the  literature   and  reported  that  nine  out  of  12  studies  found  a  positive  association  between  the  body   weights  of  newborns  and  mothers,  yet  during  the  first  two  years  of  life  this  became  less   clear  as  only  three  studies  supported  this  association  and  five  found  no  such  association.   Other  investigators  have  also  reported  only  insignificant  to  weak  relations  between  a   mother’s  weight  and  her  young  child’s  weight  (Danielzik,  Langnäse,  Mast,  Spethmann,  &   Müller,  2002;  Stunkard,  Burkowitz,  Schoeller,  Maislin,  &  Stallings,  2004;  Whitaker,  Deeks,   Baughcum,  &  Specker,  2000).       22     Stunkard  et  al.  (1999)  noted  that  many  colleagues  were  surprised  by  their  finding   that  there  was  no  association  between  the  BMIs  of  parents  and  their  two-­‐year-­‐old  children,   as  the  influence  of  genetic  factors  on  obesity  is  hypothesized  to  be  present  throughout  the   life  cycle.  One  explanation  is  that  the  genetic  influences  on  the  body  weight  of  young   children  are  different  from  the  genetic  influences  on  adult  body  weight;  or,  that  the  genes   for  obesity  are  expressed  at  different  periods  of  human  development  (Stunkard  et  al.,   1999).  Lake,  Power,  and  Cole  (1997)  found  evidence  that  this  weak  relationship  in  early   childhood  regains  strength  in  children  above  the  age  seven,  though  Cardon  (1995)  has   suggested  that  the  heritabilities  of  body  fat  becomes  evident  beginning  at  the  age  of  three   and  stabilizes  at  the  age  of  four.       Depression   Depression  Overview  in  Adult  Females:    Similar  to  the  etiology  of  obesity,  depression   is  thought  to  arise  from  both  environmental  and  genetic  factors  (Chazan-­‐Cohen  et  al.,   2007).  Epidemiological  studies  define  depression  as  a  cluster  of  symptoms  such  as  sadness,   irritability,  loss  of  interest  in  normal  activities,  difficulty  concentrating,  and  changes  in   appetite,  sleep,  and  energy  levels  in  any  two  week  period  that  cause  impairment  in   functioning;  national  prevalence  rates  of  depression  for  the  US  population  in  2005-­‐2006   are  cited  at  5.4%,  with  6.7%  of  adult  women  reporting  symptoms  of  depression  (Pratt  &   Brody,  2008).  Also  reported  was  the  finding  that  those  who  met  the  federal  poverty  level   had  higher  rates  of  depression  than  high-­‐income  persons;  the  total  rate  of  depression  was   13.1%  for  people  living  in  poverty  versus  4.4%  for  people  living  at  or  above  the  poverty   line  (Pratt  &  Brody,  2008).  Indeed,  low-­‐income  mothers  have  been  found  to  be  at  greater   risk  for  depression  than  their  middle-­‐  to  high-­‐income  counterparts  due  to  the  stressors     23     associated  with  living  in  poverty  (Petterson  &  Friel,  2001),  and  mothers  eligible  for  EHS   services  may  be  more  prone  to  dysfunction  due  to  the  challenges  of  raising  young  children   in  poverty  (Aber,  Jones,  &  Cohen,  2000).  Chazan-­‐Cohen  et  al.  (2007)  examined  17   geographically  distributed  EHS  programs  and  found  that  32%  of  mothers  scored  above  the   clinical  cutoff  for  depression  on  the  short  form  of  the  Center  for  Epidemiologic  Studies   Depression  Scale  (CES-­‐D)  when  the  target  child  was  36  months.   Maternal  Depressive  Symptoms  and  Food-­Related  Child  Outcomes:  Research   associating  depression  and  obesity  in  adults  is  ongoing  (Heo,  Pietrobelli,  Fontaine,  Sirey,  &   Faith,  2006;  Simon  et  al.,  2006;  Strine  et  al.,  2008).  A  population-­‐based  study  revealed  that   young,  overweight  women  were  44-­‐80%  more  likely  to  have  experienced  depressed  mood   in  the  preceding  two  weeks  than  those  who  were  not  obese  (Heo  et  al.,  2006).  Obese   women  also  report  additional  kinds  of  distress.  Overweight  Caucasian  women,  but  not   African  American  women,  reported  feeling  stigmatized  and  self-­‐conscious  on  account  of   their  weight  and  as  a  consequence,  indicated  some  impairments  in  their  social   relationships  in  the  workplace  and  in  their  personal  lives  (Blixen,  Singh,  &  Thacker,  2006).   These  feelings  seem  to  be  ubiquitous  in  low-­‐income  samples,  as  overweight  and  obese  low-­‐ income  Caucasian  and  African  American  mothers  also  reported  feelings  of  unhappiness,   self-­‐consciousness,  anger,  and  depression  when  considering  their  body  and  weight  (Chang   et  al.,  2008).         Studies  linking  maternal  depressive  symptoms  and  child  overweight  have  been   relatively  infrequent  thus  far  (Surkan  et  al.,  2008).  In  one  of  the  few  studies  to  date  on  this   combined  topic,  Surkan  et  al.  (2008)  found  in  their  low-­‐income  sample  of  mother-­‐child   dyads  that  depressed  mothers  were  more  likely  to  have  6-­‐24  month  old  children  who  were     24     overweight.  Ertel,  Koenen,  Rich-­‐Edwards,  and  Gillman  (2010)  revealed  that  mothers  who   experienced  postpartum  depression  had  children  with  higher  levels  of  adiposity  when   there  children  were  three  years  of  age.       Maternal  depressive  symptoms  and  feeding  style  has  also  been  examined.  Francis,   Hofer  and  Birch  (2001)  found  that  mothers  who  reported  higher  levels  of  depression  were   associated  with  using  higher  levels  of  pressure  in  feeding  their  daughters  and  conveyed  a   more  authoritarian  parenting  style.  However,  Topham  et  al.  (2009)  found  that  it  was   permissive,  not  authoritative,  parenting  that  was  a  predictor  of  child  obesity  when  mothers   were  depressed.    Haycraft  and  Blisset  (2008)  explicated  the  role  of  depression  in  feeding   such  that  for  some  parents,  “Depression  may  relate  to  hostility  in  responding  to  children’s   signals  and  interference,  such  as  overt  pressure  to  eat,  while  for  others  it  may  be   characterized  by  a  withdrawal  from  interactions,  characterized  by  parents’  reduced   involvement  in  feeding  situation”  (p.  485).   Aggression     Child  Aggression  Overview:  Aggression  in  early  childhood  is  of  particular  concern   because  it  has  been  found  to  be  stable  throughout  childhood  and  rarely  arises   spontaneously  (Tremblay  et  al.,  2004;  Zahn-­‐Waxler  et  al.,  1990).  High  levels  of  physical   aggression  during  infancy  and  toddlerhood  track  throughout  childhood  and  can  be  linked   to  a  number  of  problems  in  adolescence  and  adulthood  (Tremblay  et  al.,  2004).  In  general,   there  is  support  for  purporting  early  aggression  as  an  antecedent  to  antisocial  disorders,   drug  abuse,  violent  crimes,  depression,  suicidal  behavior,  spousal  abuse,  neglectful  and   abusive  parenting  later  in  life  (Elgar  et  al.,  2004;  Fergusson  &  Horwood,  1998;  Hofstra  et   al.,  2000;  Loeber  &  Hay,  1997;  Nagin  &  Tremblay,  1999).       25       Evidence  of  the  stability  of  aggression  has  been  found  in  early  childhood.  For  instance,   maladaptive  aggression  in  two  year  olds  was  found  to  later  predict  externalizing  problems   when  the  child  was  five  and  six;  interestingly,  this  pattern  was  more  frequent  and  stronger   in  children  of  depressed  than  nondepressed  mothers  (Zahn-­‐Waxler  et  al.,  1990).  Tremblay   et  al.  (2004)  tracked  aggression  in  17-­‐month  infants  to  42-­‐month  children,  with  those  most   at  risk  of  not  learning  to  regulate  their  aggression  from  homes  in  which  mothers  had  a   history  of  antisocial  behavior  during  their  school  years,  became  pregnant  early,  smoked   during  pregnancy,  and  came  from  low-­‐income  families.     Child  Aggression  and  Maternal  Depressive  Symptoms:  Maternal  depressive  symptoms   and  child  emotional  and  behavioral  problems  commonly  co-­‐occur  and,  as  such,  these   associations  have  been  frequently  investigated  (Goodman  &  Gotlib,  1999;  Elgar,  McGrath,   Waschbusch,  Stewart,  &  Curtis,  2004).  There  are  numerous  detrimental  effects  upon  the   offspring  of  mothers  who  have  depression  (Goodman  &  Gotlib,  1999).  Dickstein  et  al.   (1998)  found  that  depressed  mothers  had  poorer  quality  of  interactions  with  their   children,  and  families  with  maternal  mental  illness  were  distinguished  by  overall  unhealthy   family-­‐unit  functioning.  Additionally,  children  of  depressed  mothers  were  found  to  be  more   likely  to  suffer  from  an  affective  disorder  than  other  children,  and  be  at  greater  risk  of   cognitive  and  language  difficulties,  insecure  attachments,  problems  with  emotional   regulation,  low  self-­‐esteem,  social  skills,  and  behavioral  problems  (Gladstone  &  Beardslee,   2002;  Goodman  &  Gotlib,  1999;  Pilowsky  et  al.,  2006;  Zahn-­‐Waxler,  Iannotti,  Cummings,  &   Denham,  1990).  Many  of  these  psychiatric  problems  persist  into  adulthood;  the  risks  for   psychopathology  such  as  anxiety  disorder,  major  depression,  and  substance  abuse  were     26     greater  for  children  of  depressed  parents  than  children  of  nondepressed  parents  (Hofstra,   Van  der  Ende,  &  Verhulst,  2000;  Weissman  et  al.,  2006).       Using  the  CBCL,  Black  et  al.  (2002)  found  that  maternal  depressive  symptoms  was   significantly  correlated  with  child  externalizing  (r  =  0.38)  and  internalizing  problems  (r  =   0.32)  in  a  sample  of  young,  low-­‐income  mothers  with  preschool  children.  According  to   Elgar  et  al.,  (2004)  depressed  mothers  are  thought  to  engage  in  “Parenting  behavior  that  is   too  intrusive  or  withdrawn,  which  may  trigger  a  disruptive  outburst  in  the  child,  which  the   depressed  mothers  have  difficulty  managing,  thereby  exacerbating  the  child’s  behavior,   and  so  on”  (p.  445).  Some  investigators  found  that  maternal  depressive  symptoms   preceded  the  onset  of  child  adjustment  problems  (Forehand  &  McCombs,  1988;  Elgar,   Curtis,  McGrath,  Waschbusch,  &  Stewart,  2003),  which  can  include  child  aggression,  but   others  speculate  whether  or  not  maternal  depressive  symptoms  and  child  aggression  are   bidirectionally  related  and  what,  if  any,  other  mediating  influences  exist  (Elgar  et  al.,  2004;   Malik  et  al.,  2007).  Malik  et  al.  (2007)  investigated  this  issue  and  conducted  a  path  analysis   of  maternal  depressive  symptoms,  child  aggression,  and  other  ecological  factors  among   EHS  families;  their  models  were  found  to  be  an  adequate  fit  to  the  data  and  thus  supported   both  direct  and  indirect  pathways  linking  these  variables  bidirectionally.  Regardless  of  the   directionality  of  these  variables,  Elgar  et  al.  (2004)  suggested  in  their  review  of  the   literature  that  higher  rates  of  maternal  depressive  symptoms  and  child  aggression  in  low-­‐ income  samples  result  in  stronger  associations  of  these  symptoms  in  these  families.       One  plausible  explanation  for  the  relation  between  impaired  maternal  mental  health   and  child  behavior  problems  arises  from  the  study  of  food  insecurity.  Findings  from  non-­‐ human  primate  studies  suggest  that  food  insecurity  in  humans  leads  to  maternal  emotional     27     distress,  such  as  symptoms  of  anxiety  and  depression  (Andrews  &  Rosenblum,  1994).  This   stress  impairs  mother-­‐child  interactions  and  contributes  to  behavior  problems  (Andrews  &   Rosenblum,  1994;  Lindberg,  Bohlin,  Hagekull,  &  Palmerus,  1996;  Rosenblum  &  Paully,   1984).  One  such  examination  of  food  security  in  which  maternal  mental  health  and  child   mental  health  was  investigated  revealed  a  prevalence  of  behavior  problems  among  three-­‐ year-­‐old  children.  These  behavior  problems  increased  with  the  level  of  maternal  food   insecurity,  as  did  the  prevalence  of  depression  and  anxiety  in  mothers  after  one  year   (Whitaker,  Phillips,  &  Orzol,  2006).  Thus,  mental  health  problems  in  both  women  and   children  seem  to  be  more  common  when  food  is  insecure.       Child  Aggression,  Body  Weight  and  Parental  Feeding:  Limited  information  is  available   investigating  the  relation  between  weight  status  and  aggression  in  children.  Bradley  et  al.   (2008)  conducted  one  of  the  few  studies  that  included  data  on  young  children  and  found   that  from  age  two  to  grade  six,  there  was  no  association  between  BMI  and  externalizing   problems  as  measured  by  the  Child  Behavior  Checklist  (CBCL;  Achenbach  &  Rescorla,   2000).  However,  in  another  examination  using  the  CBCL  problem  scales,  5-­‐10  year  old   overweight  African  American  children  were  reported  by  their  parents  to  be  increasingly   more  aggressive;  that  is,  aggression  scores  increased  between  the  overweight,  obese,  and   super  obese  groups  (Young-­‐Hyman,  Schlundt,  Herman-­‐Wenderoth,  &  Bozylinski,  2003).   Utilizing  an  even  older  sample  of  overweight  and  normal  weight  10-­‐15  year  olds,  parents   and  teachers  reported  more  problem  behaviors  in  overweight  children,  but  no  information   regarding  the  CBCL  problem  scales  was  communicated  (Stradmeijer,  Bosch,  Koops,  &   Seidell,  1999).  Similarly,  an  examination  of  9-­‐12  year  old  clinical  obese,  nonclinical  obese,   and  normal  weight  children  found  that  only  the  clinically  obese  children  had  significantly     28     higher  scores  than  the  other  two  groups  on  the  total  and  externalizing  scores  of  the  CBCL   (Braet  et  al.,  1997).  One  interesting  finding  revealed  through  a  longitudinal  research  design   was  the  finding  that  large  body  size  at  age  three,  in  addition  to  stimulation-­‐seeking  and   fearlessness,  was  significant  in  the  prediction  of  parent-­‐rated  aggression  at  eleven  years.  In   this  sample,  aggressive  children  at  age  three  were  found  to  be  significantly  taller,  heavier,   and  bulkier  than  nonaggressive  children,  but  this  finding  was  not  replicated  when  these   children  were  eleven  years  old  (Raine,  Reynolds,  Venables,  Mednick,  &  Farrington,  1998).     There  is  some  evidence  linking  aggression  to  issues  pertaining  to  child  feeding.  In  a   2005  study,  Lewinsohn  et  al.  linked  problematic  child  behaviors  to  feeding  and  eating   problems.  It  was  found  that  the  struggle  for  control  factor  in  feeding  was  significantly   associated  with  the  externalizing  behavior  problems  described  in  the  CBCL,  with  children   who  had  higher  BMI  being  associated  with  increased  conflict.  One  of  the  few  studies  in   which  child  behavior  and  diet  were  examined  revealed  some  evidence  linking  4  ½  year  old   children  who  ate  more  junk  food  per  week  as  being  at  greater  risk  for  behavioral  problems   at  seven  years.  Unfortunately,  the  authors  did  not  include  measures  of  child  adiposity   (Wiles,  Northstone,  Emmett,  &  Lewis,  2009).  A  study  comprised  of  low-­‐income  mothers   depicted  a  link  between  food  and  child  problem  behavior  as  a  result  of  mothers’  use  of  food   to  shape  children’s  behavior  (Baughcum  et  al.,  1998).  In  this  study,  mothers  tended  to  use   food  to  quiet  crying  babies  or  to  calm  a  child’s  temper  tantrum,  not  to  satisfy  hunger.  These   mothers  also  tended  to  set  few  behavioral  limits  for  their  children;  one  of  the  results  of  this   was  that  children  were  able  to  eat  what,  as  much  and  whenever  they  wanted.  These   authors  did  not  gather  data  pertaining  to  adiposity  or  child  behavior,  which  could  have   allowed  for  statistical  inference  about  the  relations  between  these  factors.  Lewinsohn  et  al.     29     (2005)  recognized  this  gap  in  the  literature  and  surmised  that  it  was  important  to   determine  whether  or  not  the  symptoms  of  children  with  problematic  eating  behaviors  was   embedded  in  the  context  of  other  concurrent  psychopathology.     Summary   When  considering  a  mother’s  influence  on  her  child’s  eating,  most  obesity   prevention  programs  tend  to  focus  on  what  children  are  being  fed  instead  of  how  they  are   being  fed.  In  addition,  experts  focus  on  feeding  problem  behaviors  as  risk  factors  for  eating   and  weight  disorders  (Lewinsohn  et  al.,  2005).  Targeting  mothers’  feeding  practices   directly  is  a  less  common  approach  (Clark  et  al.,  2007).  Given  the  various  links  between   controlling  feeding  practices,  maternal  depressive  symptoms,  overweight,  and  child   aggression  (Birch  &  Fisher,  2000;  Birch  et  al.,  2003;  Fisher  &  Birch,  1999a;  Fisher  &  Birch,   1999b;  Fisher  &  Birch,  2002;  Goodman  &  Gotlib,  1999;  Elgar  et  al.,  2004;  Johnson  &  Birch,   1994;  Surkan  et  al.,  2008;  Young-­‐Hyman  et  al.,  2003),  these  variables  warrant  examination   when  considering  the  context  of  a  child’s  home  food  environment.  The  study  presented   here  examined  the  relation  between  maternal  feeding  practices,  maternal  depressive   symptoms,  child  aggression,  and  adiposity  in  a  low-­‐income  sample.  I  hypothesized  that   controlling  feeding  practices  were  linked  to  child  overweight  and  aggression  in  low-­‐income   families.    This  study  expanded  on  previous  literature  by  including  measures  of  child   aggression,  maternal  depressive  symptoms,  and  maternal  weight  status  to  a  model  of   maternal  feeding  practices  and  child  BMI  (Appendix  A).  These  findings  may  be  applicable   to  the  goal  of  forestalling  escalating  rates  of  obesity  and  inform  treatment  programs  for  a   low-­‐income  population.         30     METHODS   Early  Head  Start  Research  and  Evaluation  Project  Data  Set       The  data  used  for  the  present  study  was  a  subset  of  the  local  component  of  the  Early   Head  Start  Research  and  Evaluation  Project,  a  large,  evaluation  of  a  nationally   representative  sample  of  EHS  centers  throughout  the  United  States  (Schiffman  et  al.,  1996).   This  national  undertaking  was  funded  in  three  phases.  The  Birth  to  Three  Phase  (1996-­‐ 2001),  included  three  components:  an  evaluation  of  the  impacts  of  EHS  on  specific  child   and  family  outcomes,  a  measurement  of  how  EHS  centers  implemented  program  standards,   and  various  research  initiatives  administered  by  university-­‐based  investigators.  The  Pre-­‐ Kindergarten  Follow-­‐Up  Phase  (2001-­‐2005)  was  designed  to  answer  questions  raised  by   participation  in  EHS;  the  same  university  researchers  built  on  their  earlier  research  and   tracked  children  and  families  from  the  time  they  left  EHS  until  they  entered  kindergarten.   The  Elementary  School  Follow-­‐Up  (2005-­‐2010)  was  administered  by  the  testing   organizations  from  the  first  phase  and  measured  children,  families  and  teachers  on  various   measures  as  the  former  EHS  participants  entered  their  fifth  or  sixth  year  of  elementary   school  (Schiffman  et  al.,  1996).  The  data  presented  in  this  study  are  from  the  first  phase  of   the  project  when  children  were  birth  to  three  years.   Michigan  State  University  partnered  with  the  Early  Head  Start  Program  of  the   Community  Action  Agency  in  Jackson,  Michigan,  resulting  in  the  longitudinal  studies   Pathways  Project:  Research  into  Directions  for  Family  Health  and  Service  Use  (1996-­‐2001)   and  Pathways  Project  II:  Research  into  Directions  for  Family  Health  (2001-­‐2005;  Schiffman   et  al.,  1996).  The  data  from  which  the  current  analysis  was  based  upon  was  drawn  from  the   initial  investigation.  The  project  was  designed  to  examine  family  health  status  and  program     31     and  service  utilization,  which  was  measured  using  qualitative  and  quantitative  instruments   assessing  various  aspects  of  family  health  (Schiffman  et  al.,  1996).     Participants   Families  in  contact  with  local  family  health  centers  and  other  community  agencies,   and  who  were  eligible  for  EHS  services,  were  recruited  for  participation  in  the  study   (Schiffman  et  al.,  1996).  Complete  data  was  available  for  119  participants.     At  enrollment,  mothers  (n  =  119)  were  an  average  age  of  22.2  ±  4.7  years  (15.2-­‐36.9   years)  and  at  36  months  children  were  an  average  age  of  38.4  ±  2.1  months  (35.2-­‐45.4   months).  There  were  slightly  more  male  children  in  the  sample  (52.9%).  Mothers  reported   their  child’s  ethnicity  as  Caucasian  (67.5%),  African  American  (15%)  or  other  ethnicity   (17.5%).  Mothers  reported  that  they  were  Caucasian  (75.2%),  African  American  (15.6%),   or  other  ethnicity  (9.2%).  For  education,  44.9%  of  mothers  reported  having  less  than  a  high   school  education,  33.6%  were  high  school  graduates,  and  21.5%  had  some  college   education.  Most  mothers  were  single  and  never  married  (47.2%);  the  remainder  was   married  (14.8%),  cohabiting  (20.4%),  or  separated,  divorced  or  widowed  (17.6%).  As   previously  stated,  participants  were  required  to  be  eligible  for  Head  Start  services,  and   were  therefore  all  low-­‐income.  All  participants  were  included  in  the  present  study.     Procedures   Data  was  collected  through  parent  interviews  and  reviews  of  medical  records  of   both  parent  and  child.  The  survey  was  conducted  when  the  target  child  was  36  months  of   age.  Interviews  were  conducted  by  trained  data  collectors  in  the  homes  of  the  participating   families,  which  was  an  appropriate  methodology  for  gathering  such  data  (Love  et  al.,  2005).       32     Measures     To  measure  mothers’  behaviors  about  food  delivery,  appropriate  questions  from  the   Health  (Appendix  B)  and  the  Are  You  Ready  For  Fruits  and  Veggies?  (Appendix  C)  sections   of  the  Pathways  Project  survey  were  analyzed.  A  confirmatory  factor  analysis  was   conducted  with  the  goal  of  identifying  factors  in  this  survey  that  summarized  the  questions   about  mothers’  feeding  practices.  A  maximum  likelihood  extraction  method  and  a  promax   with  Kaiser  normalization  rotation  method  was  utilized.  Results  of  the  factor  analysis   indicated  that  two  components  should  be  extracted:  Food  Control  and  Meal  Patterns  (both   with  Eigenvalues  above  1.5).  The  correlations  of  variables  with  each  of  the  components  are   presented  in  Table  1  (Appendix  D).       Food  Control:  This  factor  was  derived  from  several  questions  from  the  Health  section   of  the  Pathways  Project  survey.  These  questions  were  based  on  the  different  types  of   decisions  mothers  make  regarding  their  child’s  food  intake,  which  included  serving  size,   allowing  for  hunger  based  eating,  trying  all  foods  at  least  once,  having  a  child  eat  all  their   food,  and  allowing  a  child  to  decide  on  the  quantity  of  food  consumed.  Responses  were   scored  on  a  five  point  Likert-­‐type  scale,  with  responses  ranging  from  never  to  always.       Meal  Patterns:  This  variable  was  derived  from  several  questions  from  the  Health  and   Are  you  Ready  for  Fruits  &  Veggies?  section  of  the  Pathways  Project  survey.  These   questions  tapped  the  various  types  of  practices  mothers  could  use  to  structure  meals,   which  included  the  time  meals  take  place,  variety  of  food  available,  seating  requirements,   and  the  mother’s  own  eating  practices.    Some  responses  were  scored  on  a  five  point  Likert-­‐ type  scale,  with  responses  ranging  from  never  to  always,  while  questions  having  to  do  with   frequency  of  intake  were  scored  either  0-­‐5+  or  0-­‐7.     33       Child  Aggression:    Child  aggression  was  measured  using  the  aggression  subscale  from   the  CBCL  for  ages  1  ½  to  5  years  (Achenbach  &  Rescorla,  2000)  (Appendix  E).  This  subscale   contained  19  questions  and  included  such  items  as,  “Has  temper  tantrums”,  “Easily   frustrated”,  and,  “Wants  a  lot  of  attention.”  These  and  the  other  questions  pertinent  to   aggression  are  integrated  into  the  larger  100-­‐item  instrument,  which  is  intended  to   measure  a  child’s  behavior  and  emotional  problems  (Achenbach  &  Ruffle,  2000).  This   measure  is  completed  by  a  child’s  caregiver  and  is  based  on  their  child’s  behavior  during   the  preceding  two  months.  Respondents  are  prompted  to  assign  a  number  from  0-­‐2  for   each  item  based  on  how  accurately  the  item  describes  their  child’s  behavior.  The  sum  of   each  of  the  syndrome  scales  is  interpreted  alongside  the  percentiles  based  on  the  national   normative  sample.  Scores  below  the  95th  percentile  are  in  the  normal  range,  scores  above   the  98th  percentile  are  in  the  clinical  range,  and  scores  in  between  this  range  are  regarded   as  an  area  of  concern  (Achenbach  &  Rescorla,  2000).  Reliability  for  the  CBCL  1  ½  -­‐  5  is  high,   with  a  reported  mean  r  of  .85  across  all  scales,  and  the  construct  validity  of  the  problem   scales  is  well  established  (Achenbach  &  Rescorla,  2000).         Maternal  Depressive  Symptoms:  Maternal  depressive  symptoms  was  measured  using   the  CES-­‐D  (Appendix  F).  The  CES-­‐D  is  a  self-­‐report  scale  that  measures  the  degree  of   depressive  symptoms  experienced  over  the  past  two  weeks  (Radloff,  1977).  The  scale   focuses  on  affective  components  like  depressed  mood,  feelings  of  guilt  and  worthlessness,   feelings  of  helplessness  and  hopelessness,  psychomotor  retardation,  loss  of  appetite  and   sleep  disorders.  Answers  to  20  questions  are  provided  along  a  five  point  Likert-­‐type  scale,   with  responses  ranging  from  none  of  the  time  to  most  of  the  time  (Radloff,  1977).  The  CES-­‐ D  is  reported  to  have  high  internal  consistency  in  the  general  population  with  Cronbach’s     34     alpha  coefficient  of  .85,  while  concurrent  validity  with  other  self-­‐report  measures  of   depressive  symptoms  has  been  adequately  demonstrated  (Radloff,  1977).       Child  BMI:  This  variable  was  calculated  using  child  height  and  weight  data  available  in   the  Pathways  Project  dataset;  this  data  was  obtained  through  parent  report  or  through   medical  record  examinations  when  parents  did  not  provide  these  figures.  Additional   information  needed  to  calculate  BMIs  for  children  included  the  child’s  birth  date,  the  date   of  measurement,  and  the  child’s  gender.  This  information  was  inputted  into  the  CDC’s  Child   BMI  Tool  for  Schools,  which  is  an  excel  spreadsheet  that  generates  the  corresponding  BMI   and  BMI  percentiles  (Centers  for  Disease  Control  and  Prevention,  2009a).  Child  BMI   percentiles  were  used  in  the  current  analysis.       Maternal  BMI:  This  variable  was  pre-­‐calculated  and  directly  available  in  the  Pathways   Project  dataset.  Information  that  is  necessary  to  compute  scores  for  adults  over  20  years   includes  height,  weight,  age,  and  gender;  the  current  project  utilized  self-­‐reported  figures   of  height  and  weight.  BMI  can  be  easily  calculated  through  specific  computer  programs,   such  as  the  CDC  sanctioned  Epi  Info  program,  or  by  using  the  Adult  BMI  Calculator  on  the   CDC  website  (Centers  for  Disease  Control  and  Prevention,  2008).   Analysis   Out  of  198  participants,  cases  were  excluded  if  there  were  data  missing  from  the   food  control  and  meal  patterns  variables.  Cases  were  further  reduced  if  they  contained   missing  data  from  the  BMI  variables  (maternal  BMI  =  12  missing;  child  BMI  =  18  missing).   Accurate  data  imputation  was  not  possible  given  the  confines  of  the  data  set  (i.e.,  no   previous  or  similar  data  to  base  an  imputation),  and  thus,  cases  missing  these  key  variables   were  not  included  in  the  final  analyses.  No  data  was  missing  from  the  depressive  symptoms     35     and  aggression  variables.  The  total  number  of  cases  that  contained  complete  data  for  all  of   the  variables  was  119.     The  research  question  in  the  present  study  was  examined  using  archival  data  from   the  Pathways  Project.  A  power  analysis  using  NIESEM  software  was  conducted  to  examine   the  appropriateness  of  the  theoretical  model  (Appendix  A)  linking  mother’s  controlling   feeding  practices  and  meal  patterns,  maternal  depressive  symptoms,  maternal  BMI,  child   BMI,  and  child  aggression  for  path  analysis.  Based  on  119  participants,  the  estimated   power  was  0.1363.  Thus,  even  for  a  small  effect  size  (0.20)  the  model  was  underpowered   for  utilizing  this  analytical  technique.  Indeed,  the  low  number  of  participants  in  the  current   study  was  implicated  in  the  relative  lack  of  power  in  this  model.  For  a  small  effect  size  of   0.2  the  required  number  of  participants  would  have  been  192.  Thus,  the  theoretical  model   of  the  relations  between  maternal  feeding  practices,  maternal  depressive  symptoms,   maternal  BMI,  child  BMI  and  child  aggression  were  investigated  using  bivariate   correlations  and  regression.   Given  the  sample  size,  the  relations  between  the  variables  of  interest  were   investigated  using  bivariate  correlations  and  regression.  Descriptive  statistics  were   calculated  for  all  variables  and  were  examined  for  skewness  and  kurtosis.  A  distribution   was  considered  nonnormal  if  the  ratio  of  skewness  or  kurtosis  to  its  standard  error  was   less  than  -­‐2.00  or  greater  than  2.00  (Tabachnick  &  Fidell,  1996).  All  analyses  were   conducted  using  SPSS  version  17.0               36     RESULTS   Descriptive  Statistics     Descriptive  statistics  of  all  the  variables  were  tabulated  and  are  presented  in  Table  2   (Appendix  G).     To  gain  greater  clarity  of  the  sample  composition,  the  crosstabs  function  was  used   to  obtain  counts  and  percentages  for  each  level  of  the  variable  for  mother  and  child  data.   The  information  pertaining  to  mothers,  BMI,  and  depression  are  presented  in  Table  3   (Appendix  H).  The  information  pertaining  to  children,  BMI,  and  aggression  are  presented  in   Table  4  (Appendix  I).  Percentages  were  calculated  based  on  the  total  sample  size  of  119.     Bivariate  Correlations      The  associations  between  each  variable  were  obtained  using  Zero-­‐order  Pearson’s   correlation  analyses.  The  results  of  these  calculations  are  presented  in  Table  5  (Appendix   J).  The  variables  that  were  significantly  correlated  with  one  another  were  maternal   depressive  symptoms  and  child  aggression  (r  =  .21,  p  =  .025)  and  maternal  depressive   symptoms  and  meal  patterns  (r  =  -­‐.21,  p  =  .020).     Regression   Regression  analysis  was  used  to  test  the  relation  between  several  maternal   variables  on  child  health  outcomes.  Maternal  depressive  symptoms  were  a  significant   predictor  of  child  aggression  (t(1)  =  2.28,  p  =  .03,  R2  =  .04,)  and  maternal  depressive   symptoms  significantly  predicted  meal  patterns  (t(1)  =  -­‐2.36,  p  =  .02,  R2  =  .05).  However,   the  low  R2  values  indicated  that  these  models  only  accounted  for  4%  and  5%  of  the   variability  in  the  outcome  variables;  a  very  small  amount.  Multiple  regression  models  were   tested  including  maternal  depressive  symptoms,  food  control,  and  meal  patterns  as     37     independent  variables  to  predict  child  BMI  and  child  aggression  separately;  neither  model   was  significant  (.77(3)  =2.31,  p  =  .51,  R2  =  .02;  2.25(3)  =  6.75,  p  =  .09,  R2  =  .06).  Maternal   BMI  was  not  a  significant  predictor  of  child  BMI  or  child  aggression.                                             38     DISCUSSION     Results  of  the  present  study  replicated  previous  work  by  identifying  a  significant   relation  between  maternal  depressive  symptoms  and  child  aggression  and  provided  an   important  new  addition  to  previous  work  by  identifying  a  significant  relation  between   maternal  depression  and  maternal  meal  planning  filling  an  important  gap  in  the  literature.   A  comparison  between  the  present  sample  and  a  national  sample  on  BMI,  maternal   depression  and  child  aggression  is  provided  to  highlight  the  frequencies  of  this  sample  of   low-­‐income  mothers  and  children  against  national  prevalence  rates.     Maternal  Depressive  Symptoms  and  Child  Aggression  risk     This  study  provides  a  replication  of  previous  work  by  identifying  a  positive  relation   between  maternal  depressive  symptoms  and  child  aggression  among  low-­‐income  mothers   and  their  three  year  olds  (Black  et  al.,  2002;  Elgar  et  al.,  204;  Goodman  &  Gotlib,  1999;   Malik  et  al.,  2007;  Zahn-­‐Waxler  et  al.,  1990).  In  the  current  study  maternal  depression   predicted  children’s  aggression  indicating  an  important  influence  from  maternal  mental   health  to  child  mental  health.  As  previous  work  has  noted,  this  is  important  because   aggression  tends  to  extend  into  adulthood  (Tremblay  et  al.,  2004)  and  is  associated  with   numerous  negative  outcomes,  like  violent  crimes,  substance  abuse,  and  psychological  and   social  impairments  (Elgar  et  al.,  2004;  Fergusson  &  Horwood,  1998;  Hofstra  et  al.,  2000;   Loeber  &  Hay,  1997;  Nagin  &  Tremblay,  1999).  It  is  important  to  note  that  in  the  current   study,  this  relation  was  found  in  a  non-­‐clinical  sample  of  low-­‐income  mothers  and  children   which  reinforces  a  smaller  body  of  work  (e.g.,  Black  et  al.,  2002;  Malik  et  al.,  2007).         39     Maternal  Depressive  Symptoms  and  Meal  Patterns     To  our  knowledge  this  is  the  first  study  to  investigate  the  relation  between  maternal   depression  and  meal  patterns.  According  to  our  data,  maternal  depressive  symptoms   predicted  the  ways  in  which  mothers  prepare  and  organize  meals  for  themselves  and  for   their  young  children.  Specifically,  as  maternal  depressive  symptoms  increased,  a  mother’s   ability  to  organize  and  structure  meals  for  herself  and  her  child  decreased.  Behaviors   included  in  the  meal  patterns  measure  involved  meal  planning  for  both  the  mother  herself   (e.g.  servings  of  vegetables  consumed  per  day,  consistent  time  at  which  food  is  eaten)  and   for  her  child  (e.g.  offering  child  a  variety  of  healthy  foods  to  eat).  Therefore,  according  to   these  data  depressive  symptoms  inhibited  how  mothers  engaged  in  the  feeding  of  their   young  children,  both  directly  (by  how  they  fed  their  child)  and  indirectly  (by  how  they  fed   themselves  and  modeled  eating  behaviors).  Based  on  these  data,  it  is  clear  that  maternal   depressive  symptoms  has  important  implications  for  children’s  development  of   appropriate  mealtime  habits.  Indeed,  maternal  modeling  of  food  consumed  is  a  strong   predictor  of  children’s  own  food  consumption  (Cooke  et  al.,  2008;  Horodynski  et  al.,  2010),   as  foods  consumed  by  parents  and  children  tend  to  be  correlated  (Brown  &  Ogden,  2004;   Skinner  et  al.,  2002)  and  mothers  tend  to  purchase  foods  and  prepare  meals  that  they  find   appealing  (Birch  &  Fisher,  1998;  Wardle,  1995).  Therefore,  when  mothers  are  depressed   they  are  less  likely  to  model  the  consumption  of  healthy  foods  or  plan  and  prepare   organized  family  meals.  Further,  as  depression  increased,  a  mother’s  capacity  to  engage  in   appropriate  meal  pattern  activities  decreased.  For  example,  when  mothers  are  depressed,   they  are  less  likely  to  prepare  healthy  meals  with  multiple  food  choices  for  their  children  or     40     follow  generally  accepted  healthy  food  guidelines  (e.g.  eating  breakfast,  eating  fruits  and   vegetables,  sitting  while  eating).     Previous  work  has  identified  several  negative  consequences  of  maternal  depressive   symptoms  on  a  child’s  psychological  development  (Gladstone  &  Beardslee,  2002;  Goodman   &  Gotlib,  1999;  Pilowsky  et  al.,  2006;  Zahn-­‐Waxler  et  al.,  1990).  Certainly,  depression  also   impacts  a  parent’s  ability  to  fulfill  other  parenting  responsibilities,  including  general   household  management.  For  instance,  depressed  parents  have  been  found  to  be  less  likely   than  nondepressed  parents  to  optimally  interact  with  their  child,  which  included  such   activities  as  reading,  playing,  teaching,  and  setting  behavioral  limits  (LaRosa,  Glascoe,  &   Macias,  2009).  Taking  into  account  children’s  medical  management,  it  was  found  that   depressed  mothers  were  more  likely  to  report  less  diligence  in  adhering  to  their  child’s   asthma  therapy  and  difficulties  communicating  with  their  doctor  more  than  nondepressed   mothers  (Bartlett,  Krishnan,  Riekert,  &  Butz,  2004).  Depressed  parents  also  have  difficulty   managing  the  development  of  their  child’s  extended  social  networks,  as  these  parents  may   not  lend  support  to  their  child’s  involvement  with  events  or  others  beyond  the  immediate   family  (Zahn-­‐Waxler  et  al.,  1990).  In  a  study  assessing  quality  of  life  in  depressed  adults,   those  who  were  depressed  reported  lower  levels  of  quality  of  life  than  nondepressed   adults,  and  reported  that  they  were  more  likely  to  have  daily  periods  in  which  they  did   nothing  and  be  less  engaged  in  their  work  activities  (Barge-­‐Schaapveld,  Nicolson,  Berkhof,   &  deVries,  1999).  Considered  all  together,  these  findings  suggest  that  mothers  with   depression  tend  to  be  less  engaged  in  their  daily  activities.  Findings  from  the  present  study   add  to  this  wealth  of  knowledge  regarding  family  and  household  management  by  indicating   that  mothers  with  depression  are  less  proficient  than  nondepressed  mothers  at  family  food     41     management,  an  essential  component  of  children’s  health  and  wellbeing.  This  important   finding  extends  our  knowledge  of  the  impact  of  maternal  depressive  symptoms  beyond   influencing  children’s  socioemotional  and  cognitive  development  to  inhibiting  children’s   physical  development  as  well.         In  the  current  study,  maternal  depressive  symptoms  was  negatively  associated  with   meal  patterns.  Therefore,  mothers  who  were  depressed  were  less  likely  to  eat  at  the  same   time  every  day,  less  likely  to  offer  a  child  a  variety  of  foods  to  eat  such  as  fruits,  vegetables,   cereals  and  dairy,  less  likely  to  be  seated  at  a  table  when  they  ate  meals,  and  less  likely  to   consume  recommended  intakes  of  fruit  or  fruit  juice,  vegetables  and  breakfast  provisions.   This  finding  has  important  implications  for  children’s  development  because  low  scores  on   meal  patterns  resembles  a  neglectful  feeding  style.  A  neglectful  feeding  style  is   characterized  by  unstructured  feeding  sessions  and  limited  involvement  with  the  child   (Hughes  et  al.,  2005;  Rhee  et  al.,  2006).  A  mother’s  neglectful  feeding  style  may  result  in  the   maintenance  or  development  of  poor  eating  habits  for  herself  and  her  children,  such  as   frequent  snacking,  eating  and  serving  quick  pre-­‐processed  meals,  and  permitting  children   to  prepare  or  retrieve  these  convenience  foods  themselves  (e.g.,  Easy  Mac,  Lunchables,   etc.).  This  finding  is  of  great  concern  because  children  of  neglectful  or  permissive  mothers   are  at  risk  of  overweight  (Rhee  et  al.,  2006)  or  lower  intakes  of  nutrient  dense  foods  (Hoerr   et  al.,  2009).  Although  the  results  from  this  study  found  no  association  between  maternal   depressive  symptoms  and  child  obesity  at  age  three,  previous  work  has  established  this   connection  with  slightly  older  samples  of  children.  Topham  et  al.  (2009)  found  that  first   grade  children  whose  mothers  were  depressed  and  employed  a  permissive  feeding  style   were  more  likely  to  be  obese  than  children  from  non-­‐depressed  mothers.  Therefore,  one     42     could  hypothesize  that  having  a  depressed  mother  could  be  a  risk  factor  for  obesity  in   middle  childhood  (Topham  et  al.,  2009).  The  optimum  age  at  which  to  intervene  in  a   depressed  mother’s  food  management  system,  with  the  goal  of  preventing  child  obesity  and   the  complications  that  arise  from  living  overweight,  is  at  least  during  early  childhood,  if  not   earlier.       The  findings  of  this  study  are  even  more  concerning  due  to  the  low-­‐income  status  of   the  mothers  and  children  who  participated  in  this  study.  Low-­‐income  mothers,  already  at   greater  risk  of  depression  than  their  middle-­‐  to  high-­‐income  counterparts  (Chazan-­‐Cohen   et  al.,  2007;  Petterson  &  Friel,  2001;  Pratt  &  Brody,  2008)  may  be  in  further  peril  of   developing  poor  self  and  child  feeding  strategies  given  their  increased  likelihood  of   depression.  It  is  important  to  note,  that  although  these  mothers  were  of  low-­‐income  status,   they  were  not  a  clinically  depressed  sample.  So  even  small  variations  in  depression  within   a  non-­‐clinical  sample  impeded  mothers’  ability  to  provide  planned  meals  for  themselves   and  their  child.       Previous  work  has  identified  a  relation  between  maternal  mental  health  and   compromised  feeding  processes.  For  example,  Bronte-­‐Tinkew,  Zaslow,  Capps,  Horowitz,   and  McNamara  (2007)  found  that  food  insecurity  influenced  parents’  depression,  which  in   turn  negatively  affected  parenting  practices  and  decisions  regarding  infant  feeding.  Cooper,   Whelan,  Woolgar,  Morrell,  and  Murray  (2004)  conducted  a  study  composed  of  mothers   with  eating  disorders  and  their  approximately  4  ½  year  old  children  and  found  that   mealtime  disorganization  was  strongly  associated  with  child  feeding  problems.  In  an   investigation  of  mother,  child  and  family  factors  in  childhood  obesity,  De  Sousa  (2009)   compared  obese  and  normal  weight  boys  and  found  that  in  the  obese  group,  mothers  were     43     more  depressed  and  anxious,  scored  higher  on  the  laxness  and  verbosity  in  parenting  style   scale,  and  allowed  their  children  to  eat  more  meals  per  day,  eat  more  meals  away  from   home,  watch  more  television  and  consume  more  soft  drinks.  Clearly  many  differences  exist   in  the  design  of  these  studies,  but  in  all  of  these  instances  the  common  factor  appears  to  be   that  mothers  with  poorer  mental  health  have  compromised  family  feeding  management   systems.     Sample  versus  National  Composition     National  prevalence  rates  of  obesity,  depression  and  aggression  were  compared  to   the  prevalence  rates  of  the  participants  from  the  present  study.  The  composition  of  this   sample  was  examined  through  the  crosstabs  function  in  SPSS,  which  produced  a   breakdown  of  the  frequencies  and  percentages  of  mothers  and  children  according  to  their   weight  status  and  respective  measure  of  mental  health.  These  figures  are  presented  in   Table  3  and  Table  4.     The  percentage  of  obese  children  in  this  sample  was  greater  than  the  latest  national   estimate  of  children  with  a  BMI  greater  than  the  95th  percentile,  but  the  percentage  of   overweight  children  was  lower  than  the  national  estimate  of  children  with  a  BMI  between   the  85th  and  95th  percentile  (Ogden  et  al.,  2010).  These  figures  were  30.3%  versus  10.4%,   and  15.1%  versus  21.2%.  The  finding  that  there  were  more  obese  but  fewer  overweight   children  than  the  national  averages  may  be  explained  by  the  fact  that  the  figures  for  the   current  study  were  based  on  the  BMI  percentiles  for  three  years  olds,  unlike  the  national   estimates  which  were  based  on  data  for  two  to  five  year  olds.    The  discrepancy  in  ages   limits  comparison  between  these  two  groups.     44     The  percentage  of  obese  mothers  in  the  current  sample  was  slightly  below  the   national  average  (Flegal,  Carroll,  Ogden,  &  Curtin,  2010).  The  national  average  of  obesity   (BMI  equal  or  greater  to  30)  in  2007-­‐2008  for  women  over  the  age  of  twenty  was  reported   to  be  35.5%;  for  this  sample,  this  rate  was  31.1%.  Rates  of  overweight  adults  were  reported   in  combination  with  rates  of  obesity;  national  estimates  of  the  percentage  of  adults  with  a   BMI  greater  than  25  and  less  than  30  was  not  individually  reported  (Flegal  et  al.,  2010),   which  prevented  comparison  with  the  current  study’s  estimate  of  overweight  women.     National  rates  of  clinical  scores  of  aggression  in  young  children  were  not  located  to   facilitate  comparison  with  the  findings  from  the  current  analysis.  The  most  analogous   statistics  available  were  from  archival  data  including  over  30,000  children  visiting   pediatricians  from  1979-­‐1996;  final  estimates  of  psychosocial  problems  revealed  that  7.5%   of  four  to  fifteen  year  old  children  were  reported  to  have  behavioral/conduct  problems   when  presenting  to  a  physician  (Kelleher,  McInerny,  Gardner,  Childs,  &  Wasserman,  2000).   In  the  current  study,  1.5%  of  children  were  reported  to  have  clinical  scores  of  aggression;   this  was  well  below  the  rate  of  7.5%,  yet  a  few  key  issues  prevent  these  rates  from  being   truly  comparable.  One  such  issue  is  that  children  in  the  Kelleher  et  al.  (2000)  study  were   older  and  the  age  range  was  much  wider  than  the  three  year  olds  in  the  current  study.  In   addition,  the  use  of  a  different  assessment  tool  to  measure  child  internalizing  and   externalizing  problems  was  different  than  what  used  in  the  present  study.     Consistent  with  the  literature  (Petterson  &  Friel,  2001),  the  low-­‐income  women  in   this  study  were  similarly  depressed  compared  to  the  national  estimates  of  women  who   were  depressed.  Approximately  8.4%  of  mothers  in  the  current  sample  scored  in  the  upper   range  of  the  CES-­‐D,  while  the  latest  NHANES  data  available  involving  depression  found     45     6.7%  of  women  reported  symptoms  of  depression  when  measured  by  the  Patient  Health   Questionnaire  (PHQ-­‐9;  Pratt  &  Brody,  2008).  However,  this  figure  of  8.4%  is  well  below  the   rate  of  32%  found  by  Chazan-­‐Cohen  et  al.  (2007)  for  EHS  mothers.       Limitations   There  are  several  reasons  as  to  why  the  expected  associations  between  certain   variables  were  not  found.  These  possibilities  are  discussed  in  variable  specific  sections.     Controlling  Feeding  Practices  and  BMI:  One  of  the  main  goals  of  this  study  was  to   shed  light  on  the  issue  of  whether  or  not  mothers’  controlling  feeding  practices  were  linked   to  higher  BMI  in  children  in  a  low-­‐income  sample.  In  general,  there  was  no  support  for  this   association  through  the  results  of  the  bivariate  correlation  analysis  or  regression  analysis.   However,  despite  the  absence  of  significant  findings,  it  would  be  remiss  to  say  that   controlling  feeding  practices  were  unequivocally  unrelated  to  adiposity  in  this  low-­‐income   sample  of  mothers  and  their  three-­‐year-­‐old  children.  Though  controlling  feeding  practices   and  child  adiposity  could  simply  be  unassociated  in  low-­‐income  samples  in  general,  the   association  might  only  become  evident  as  a  child  increases  in  age,  which  the  current  study   would  be  unable  to  assess  given  its  36-­‐month  time  point.  Children  might  be  affected  by   controlling  feeding  practices  at  three  years  of  age,  but  because  decreased  sensitivity  to   internal  signs  of  hunger  and  satiety  is  proposed  to  occur  over  time  (Johnson  &  Birch,  1994;   Birch  &  Ventura,  2009)  disinhibited  eating  and  overweight  may  only  be  evident  in  the  later   years  of  childhood.  Though  children  from  the  age  of  three  years  have  been  found  to  have  an   increased  desire  to  consume  a  food  that  had  been  restricted  or  if  they  had  mothers  who   were  restrictive  in  their  feeding  practices  (Fisher  &  Birch,  1999a;  Fisher  &  Birch,  1999b;   Johnson  &  Birch,  1994),  previous  work  has  confirmed  the  link  between  controlling  feeding     46     practices  (which  include  restriction)  and  overeating  and  overweight  when  children  are  five   through  nine  years  of  age  (Birch  &  Fisher,  2000;  Birch,  Fisher,  &  Davison,  2003;  Fisher  &   Birch,  2002).     Food  control  did  not  yield  any  significant  findings  in  the  bivariate  correlation   analysis  or  the  regression  analyses.  This  lack  of  significance  may  be  in  part  due  to  issues   surrounding  the  measurement  of  controlling  feeding  practices.  The  current  study   measured  a  generalized  conceptualization  of  controlling  feeding  practices,  which  included   such  questions  as  “How  often  does  your  child  decide  how  much  to  eat  at  meals?”  and,  “How   often  do  you  give  your  child  the  same  amounts  of  food  you  give  yourself?”  The  survey   questions  from  which  food  control  was  based  seemed  to  tap  the  concept  of  feeding   practices  identifying  high  face  validity  however,  evidence  of  content  validity  and  reliability   was  unavailable  for  the  measure.  Further,  this  measure  was  a  general  measure  of  feeding   practices.  A  meta-­‐analysis  of  feeding  behaviors  and  weight  by  Faith  et  al.  (2004)  found  that   studies  that  employed  a  general  measure  of  controlling  feeding  practices  were  significantly   less  likely  to  be  positively  associated  with  child  adiposity  or  energy  intake  than  studies  that   tapped  the  specific  practice  of  feeding  restriction.  Previous  work  has  linked  the   measurement  of  restriction  of  access  to  palatable  foods  to  elevated  levels  of  BMI  (Birch  &   Fisher,  2000;  Birch  et  al.,  2003;  Fisher  &  Birch,  2002);  the  lack  of  questions  pertaining  to   restriction  thus  may  have  been  part  of  the  reason  why  there  was  no  association  between   food  control  and  BMI.  If  the  dimension  of  food  restriction  had  been  included,  preliminary   findings  could  have  laid  the  groundwork  for  further  longitudinal  study  of  the  predictive   effect  of  restrictive  feeding  practices  when  these  low-­‐income  children  matured  in  age.   Frequently  employed  in  the  study  of  feeding  style  and  feeding  practice  is  the  Child  Feeding     47     Questionnaire  (CFQ;  Birch  et  al.,  2001),  which  contains  questions  that  measure  restriction,   including  “I  have  to  be  sure  that  my  child  does  not  eat  too  many  sweets”  and,  “I   intentionally  keep  some  foods  out  of  my  child’s  reach.”  None  of  the  questions  in  the   Pathways  Project  survey  resembled  these  kinds  of  questions,  which  may  have  accounted   for  the  lack  of  association  of  control  with  child  BMI.     Maternal  and  Child  BMI:  There  was  no  significant  association  between  maternal  and   child  BMI  in  the  bivariate  correlational  analysis.  This  finding  supported  the  assertion  by   Stunkard  et  al.  (1999)  that  mother  and  child  BMI  are  unrelated  when  a  child  is  two  years   and  extends  this  finding  by  one  year  to  three  years.  This  finding  is  contrary  to  the   suggestion  by  Cardon  (1995),  who  held  that  these  variables  should  be  significantly  linked   beginning  at  the  age  of  three  years.  As  previously  stated,  this  claim  has  not  been  found  to   be  corroborated  by  others,  and  Lake  et  al.  (1997)  found  this  connection  beginning  when   children  were  seven  years  old.  However,  the  low-­‐income  status  of  mothers  and  their   children  might  provide  some  rationale  as  to  why  no  relation  was  found.  Given  that  this   sample  contained  fewer  overweight  and  obese  mothers  and  greater  obese  children  than  is   reported  in  the  national  estimates  (Ogden  et  al.,  2010;  Pratt  &  Brody,  2008),  it  might  be   plausible  that  mothers  ensured  that  their  children  were  sufficiently  fed  before  feeding   themselves,  thus  creating  a  wider  than  expected  disparity  between  their  measures  of   adiposity.  Perhaps  children’s  participation  in  EHS,  a  program  which  targets  children’s   health  and  nutrition,  provided  them  with  additional  calories  that  were  unavailable  to   mothers.     Secondly,  the  height  and  weight  figures  used  to  tabulate  BMI  of  mothers  and   children  were  based  on  self-­‐report,  and  thus,  presented  concern  regarding  validity.  While     48     self-­‐report  of  height  and  weight  is  deemed  acceptable  for  use  in  epidemiological  studies,  it   remains  that  height  tends  to  be  overestimated  and  weight  tends  to  be  underestimated  in   adults  (Engstrom,  Paterson,  Doherty,  Trabulsi,  &  Speer,  2003;  Gorber,  Tremblay,  &  Moher,   2007).  This  schism  impacts  BMI  calculations;  in  one  study  using  self-­‐report  data,  22.4%  of   men  and  18%  of  women  were  grouped  into  the  wrong  BMI  categories  (Spencer,  Appleby,   Davey,  &  Key,  2001),  which  may  be  why  the  frequencies  in  the  current  study  did  not  reflect   the  national  data  set.  Mothers  might  underreport  their  weight  either  due  to  lack  of  accurate   knowledge  or  because  of  the  social  desirability  for  women  to  be  thin  in  America  (Gorber  et   al.,  2007;  Grabe,  Ward,  &  Hyde,  2008).  The  anthropometric  data  obtained  for  children  was   also  provided  by  mothers  on  an  ad  hoc  basis.  In  the  cases  where  these  figures  were   unattained  from  mothers,  researchers  gathered  this  data  from  medical  records.  This   strategy  improved  the  completion  and  accuracy  of  the  BMI  variable,  yet  those  whose   anthropometric  data  were  provided  from  medical  records  were  not  differentiated  in  the   dataset.  Therefore,  it  was  not  possible  to  sort  child  BMI  by  the  source  of  the  height  and   weight  figures.  These  concerns  call  to  question  the  veracity  of  the  findings  in  which  the  BMI   measure  was  used.       Future  Research   Previous  work  on  feeding  practices  and  adiposity  has  infrequently  included  mental   health  variables  like  maternal  depression.  However,  as  the  results  of  the  present  study   suggest,  these  seem  to  be  important  variables  related  to  maternal  feeding  practices.  The   few  feeding  practice  studies  that  have  measured  maternal  depressive  symptoms  (Blissett   et  al.,  2007;  Haycraft  &  Blissett,  2008)  have  not  included  child  internalizing  and   externalizing  symptoms  in  their  analyses.  Findings  of  the  study  presented  here  provide     49     new  evidence  that  maternal  mental  health  status  is  an  important  variable  to  consider  in   studies  of  feeding  practices,  particularly  in  light  of  the  previously  established  link  between   maternal  feeding  practices  and  later  weight  gain  in  children  (Birch  &  Fisher,  2000;  Birch  et   al.,  2003;  Fisher  &  Birch,  2002).  Due  to  the  complex  interworkings  of  the  multiple  variables   implicated  in  children’s  physical  health  outcomes  future  work  should  investigate  the   relations  among  these  variables  using  an  analytical  strategy  capable  of  assessing  the   relation  between  multiple  variables  simultaneously.  These  strategies  (e.g.,  Structural   Equation  Modeling)  remain  uncommon  statistical  approaches  used  in  feeding  practice   studies.   Future  research  should  build  on  the  theoretical  model  proposed  here  and  include   more  than  one  standardized  measure  for  each  of  the  variables  of  interest.  Adiposity  could   be  additionally  measured  through  the  inclusion  of  skinfold  tests  or  exclusively  through  the   more  sophisticated  methods  such  as  DXA  or  via  the  direct  measurement  (vs.  self-­‐report)  of   weight  and  height.  Feeding  practices  could  be  better  evaluated  by  employing  the  frequently   used  CFQ  (Birch  et  al.,  2001)  and  observational  tools  of  actual  mealtime  practices,  such  as   Bob  and  Tom’s  Method  of  Assessing  Nutrition  (BATMAN;  Klesges  et  al.,  1983)  or  the   Mealtime  Family  Interactions  Coding  System  (MICS;  Dickstein,  Hayden,  Schiller,  Seifer,  &   San  Antonio,  1994).  As  noted  before,  measurement  of  controlling  feeding  practices  should   include  survey  items  related  to  restriction  of  palatable  foods  because  of  the  significance  of   this  practice  to  body  weight  findings  (Fisher  &  Birch,  1999a;  Faith  et  al.,  2004).  In  addition,   observational  tools  could  note  the  presence  or  absence  of  particular  feeding  behaviors  in   real  time.  Future  work  should  broaden  the  scope  of  maternal  mental  health  and  include   measurement  of  anxiety,  such  as  the  Beck  Anxiety  Inventory  (BAI;  Beck,  Epstein,  Brown,  &     50     Steer,  1988)  and  eating  disorders,  such  as  the  Eating  Disorders  Inventory-­‐2  (EDI-­‐2;  Garner,   1991);  child  outcome  variables  could  be  expanded  to  include  measurement  of   temperament  and  self-­‐regulation  to  be  included  as  mediators  between  maternal  feeding   practices  and  child  BMI.  Theoretically,  it  may  be  that  self-­‐regulatory  skills  are  influenced  by   maternal  feeding  practices  or  that  high  self-­‐regulation  might  act  as  a  protective  factor   against  overeating.     Additional  concern  stems  from  the  cross-­‐sectional  nature  of  the  majority  of  studies   that  have  investigated  feeding  practices,  with  few  investigators  exploring  the  topic  through   other  methodologies  (Moens,  Braet,  &  Soetens,  2007;  Ventura  &  Birch,  2008).  Some   researchers  have  questioned  the  generalizability  of  the  finding  that  excessive  maternal   control  in  child  feeding  is  linked  to  greater  adiposity  when  employing  low-­‐income  or   racially  diverse  samples  (Clarke  et  al.,  2007;  Faith  et  al.,  2004;  Hoerr  et  al.,  2009;  Hughes  et   al.,  2005;  Robinson  et  al.,  2001).  While  the  findings  of  this  study  do  not  allow  one  to   conclude  that  there  is  no  association  between  controlling  feeding  practices  and  weight   status  in  low-­‐income  samples,  long-­‐term  follow  up  is  nevertheless  needed  to  investigate   this  link  when  these  low-­‐income  children  are  older.  Such  a  study  would  wisely  utilize   standardized  medical  devices  to  measure  height  and  weight  to  derive  adult  and  child  BMI,   such  as  the  commonly  accepted  wall  stadiometer  and  balance  beam  scale  (Malina,  1995).     Implications     One  of  the  most  compelling  findings  of  the  study  was  the  association  between   maternal  depressive  symptoms  and  meal  planning.  Programs  linked  to  positive  nutrition   outcomes  are  well  established,  but  few  include  considerations  for  mother  and  child  mental   health  in  their  curriculum.  An  example  of  an  existing  well-­‐known  intervention  program     51     involving  nutrition  education  is  the  Special  Supplemental  Nutrition  Program  for  Women,   Infants,  and  Children  (WIC).  WIC  is  a  federal  initiative  implemented  by  state  agencies   aimed  at  providing  low-­‐income  mothers  of  young  children  with  supplemental  nutritious   food,  health  education,  and  high-­‐risk  nutrition  counseling  (Food  &  Nutrition  Service,  2010).   While  WIC  is  also  mandated  to  provide  referrals  to  other  health  and  social  services,  what  is   unknown  is  if  and  how  frequently  participants  are  referred  onwards  for  mental  health   services  and  if  participants  follow-­‐up  with  these  services.  Conversely,  there  is  no  known   federal  initiative  that  seeks  to  prevent  the  onset  of  depression;  a  rationale  for  doing  so   would  be  well-­‐founded  given  the  high  percentages  of  depression  in  low-­‐income  mothers   (Chazan-­‐Cohen  et  al.,  2007;  Petterson  &  Friel,  2001;  Pratt  &  Brody,  2008).     Therefore,  future  intervention  programs  might  employ  a  holistic  model  in  which   both  physical  and  mental  facets  of  health  in  mothers  and  children  are  targeted.  To  achieve   positive  outcomes  for  mothers  and  their  families,  nutrition  education  could  combine   traditional  content  (e.g.  nutrient  properties  of  foods,  healthy  meals,  mealtime  planning   strategies)  with  content  specifically  related  to  child  feeding.  Such  content  could  include   guidance  to  parents  on  how  to  continue  or  instill  children’s  ability  to  listen  to  internal  cues   of  hunger  and  satiety  to  guide  eating  (Birch  et  al.,  2003)  and  to  limit  the  use  of  food  as  a   way  of  managing  difficult  behaviors  unrelated  to  hunger  (Baughcum  et  al.,  1998).   Empowering  mothers  who  are  depressed  with  the  knowledge  and  skills  to  manage  healthy   child  feeding  could  serve  to  support  their  own  healthy  eating,  as  well  as  increase  their  self-­‐ efficacy  and  confidence.       Moreover,  the  positive  correlation  between  maternal  depressive  symptoms  and   child  aggression  supports  the  need  for  intervention  efforts  before  the  child  is  three  years     52     old.  To  assist  with  early  detection  and  treatment,  screenings  for  mothers  and  their  children   for  internalizing  and  externalizing  problems  could  occur  during  regularly  scheduled  well-­‐ baby  and  well-­‐child  visits  and  involvement  with  community  service  agencies,  such  as  WIC.   Again,  this  would  be  especially  important  for  low-­‐income  mothers,  given  previous  evidence   confirming  that  they  have  higher  incidence  rates  of  depression  than  their  middle-­‐  and  high-­‐ income  counterparts  (Chazan-­‐Cohen  et  al.,  2007;  Petterson  &  Friel,  2001;  Pratt  &  Brody,   2008).  Attending  to  psychological  facets  of  health  could  forestall  the  suffering  of  the   mother  and  the  many  negative  impacts  on  the  child  attributable  at  least  in  part  to  their   mother’s  depression  (Gladstone  &  Beardslee,  2002;  Goodman  &  Gotlib,  1999;  Pilowsky  et   al.,  2006;  Zahn-­‐Waxler  et  al.,  1990).  Based  on  some  researchers’  assertion  that  maternal   depressive  symptoms  leads  to  child  aggression  (Forehand  &  McCombs,  1988;  Elgar  et  al.,   2003),  treatment  of  maternal  depressive  symptoms  could  theoretically  help  to  ward  off  the   development  of  child  aggression,  other  psychosocial  problem  behaviors  (Black  et  al.,  2002;   Elgar  et  al.,  2004;  Hofstra  et  al.,  2002;  Weissman  et  al.,  2006)  and  elevated  weight  status   (Surkan  et  al.,  2008).  Therefore,  prevention  and  treatment  of  maternal  depressive   symptoms  is  key.     Conclusion     The  assertion  that  the  way  in  which  a  mother  feeds  her  young  child  affects  her   child’s  own  relation  to  food  has  been  largely  supported  in  the  literature  (Costanzo  &   Woody,  1985;  Satter,  1990).  Prior  work  has  linked  controlling  maternal  feeding  practices  to   a  child’s  inability  to  regulate  their  intake  of  food,  which  is  related  to  increased  body  weight   in  later  childhood  (Birch  &  Fisher,  2000;  Birch  et  al.,  2003;  Fisher  &  Birch,  2002)  however,   there  was  no  association  found  between  food  control  and  BMI  in  the  present  study.  What     53     was  revealed  in  the  current  study  were  associations  between  maternal  depressive   symptoms,  child  aggression,  and  meal  planning,  which  provides  empirical  evidence   implicating  maternal  depressive  symptoms  as  one  variable  which  influences  the  mental   health  of  her  young  child  and  her  ability  to  plan  meals  for  herself  and  her  child.  This  finding   attests  to  the  powerful  effect  that  depression  has  in  parenting  duties.  A  mother’s   depression  may  confer  a  more  potent  influence  over  her  young  child’s  development  than   feeding  practices  do  because  it  impedes  a  mother’s  ability  to  plan  and  prepare  healthy   meals  for  her  family.  Chronic  poor  meal  planning  could  theoretically  limit  the  development   of  healthy  eating  patterns,  which  in  turn  could  foster  deregulated  eating  and  later  weight   gain.  The  ubiquitous  presence  depression  exerts  in  a  mother’s  life  extends  beyond  her  own   mental  health  to  impact  her  child’s  behavior  and  her  food  planning  for  herself  and  her   child,  thereby  rendering  her  offspring  at  risk  for  the  myriad  of  psychological  and  physical   complications  linked  to  aggression  and  overweight.                         54                         APPENDICES                             55     APPENDIX  A:  Theoretical  Model           Maternal  BMI               Food  Control   Child  BMI                   Meal  Patterns   Child  Aggression               Maternal  Depression           Figure  1:  Theoretical  model  at  36  months               56     APPENDIX  B:  Health  Section  of  Pathways  Project  Survey     8.5   Is   (CHILD)   currently   covered   by   any   kind   of   health   insurance,   such   as   Medicaid   or   private   insurance   plan,   or   by   a   Health   Maintenance   Organization   (HMO)   that   covers   hospital  or  doctor  bills?           YES……………………………………..01       NO……………………………………....00       DON’T  KNOW……………………….-­‐1       REFUSED………………………..……-­‐3     For  each  question  I  read,  please  answer  whether  you  (or  FOCUS  CHILD)  never,  hardly  ever,   sometimes,  most  times  or  always  do  the  task  in  question.       (READ  STATEMENT).  Do  you  do  this  never,  hardly  ever,  sometimes,  most  times,  or  always   do  the  following?     Most     Never   Hardly  Ever   Sometimes   Always   Times   1.  How  often  do  you   serve  new  foods  to   1   2   3   4   5   your  child?   2.  How  often  does   your  child  decide   how  much  to  eat  at   1   2   3   4   5   meals?  (parent   does  not  decide)   3.  How  often  do  you   give  your  child  the   same  amounts  of   1   2   3   4   5   food  you  give   yourself?   4.  How  often  is   your  child  seated   1   2   3   4   5   when  eating?   5.  How  often  does   your  child  decide  if   1   2   3   4   5   he/she  is  hungry  at   meals?   6.  How  often  do  you   have  your  child   1   2   3   4   5   taste  everything  on   the  plate?       57     Next   I   am   going   to   read   some   statements   and   thoughts   on   feeding   children.   Tell   me   whether   you   strongly   agree,   agree,   disagree,   or   strongly   disagree   with   each   of   the   statements  as  they  apply  to  your  thoughts  about  feeding  children.       Strongly   Strongly     Disagree   Not  Sure   Agree   Disagree   Agree   1.  It  is  important   for  my  child  to  eat   1   2   3   4   5   or  taste  vegetables   every  day.   2.  It  is  important   for  my  child  to  eat   1   2   3   4   5   or  taste  fruit  every   day.   3.  It  is  important   for  my  child  to  eat   1   2   3   4   5   or  taste  meat  every   day.   4.  It  is  important   for  my  child  to  eat   1   2   3   4   5   or  taste  cereal   every  day.     5.  It  is  important   for  my  child  to   1   2   3   4   5   finish  food  on   his/her  plate.   6.  When  my  child   refuses  a  new  food,   1   2   3   4   5   I  don’t  offer  it   again.     7.  How  many  times  a  day  should  your  child  drink  a  glass  of  milk?  Would  you  say…   a. 0   b. 1   c. 2   d. 3   e. 4   f. 5+     8.  To  help  children  develop  a  liking  for  a  new  food,  over  a  period  of  time,  parents  should   offer  the  new  food:     a. Once  or  twice   b. Several  times   c. Many  times,  up  to  20       58       APPENDIX  C:  Fruits  and  Veggies  Section  of  Pathways  Project  Survey     1.  Usually  how  many  servings  a  day  of  fruit  or  fruit  juice  do  you  eat?  (1  serving  =  ½    cup,  1   piece,  ¾  cup  juice;  Not  juice  drinks,  Kool-­‐aid,  pop,  Hi-­‐C,  Sunny  Delight,  or  Tropical  Breeze)     2.  Usually,  how  many  servings  of  vegetables  a  day  do  you  eat?  (1  serving  =  ½  cup  cooked,  1   cup  lettuce;  Include  potatoes,  fries,  tomatoes,  and  tomato  sauce  in  pizza  and  spaghetti)     3.  About  how  long  have  you  eaten  this  many  servings  of  fruits  &  veggies?   a. Less  than  1  month   b. About  1-­‐5  months   c. About  6  or  more     4.  How  important  do  you  think  eating  fruits  &  veggies  is  to  your  health?   a. Very   b. Somewhat   c. Not  very     5.  During  the  last  six  (6)  months,  do  you  plan  to  eat  fruits  &  veggies?   a. No     b. Yes,  I  have     6.  In  the  next  six  (6)  months,  do  you  plan  to  plan  to  eat  fruits  &  veggies?   a. No   b. Probably  not   c. Yes,  I  plan  to   d. I  already  have  enough     7.  How  many  times  a  week  do  you  usually  eat  breakfast?         8.  How  many  times  a  week  do  you  eat  fast  food  or  eat  out?  (Include  pizza,  subs,  Wendys,   McDonalds,  etc.)   _________________     9.  Have  you  participated  in  Project  FRESH?   a. No   b. Yes,  this  summer   c. Yes,  in  ____________  [year(s)]       59     APPENDIX  D:  Factor  Analysis   Table  1   Pathways  Project  Survey  Factor  Analysis         Food  Control   -­‐.01   Meal  Patterns   .43   Do  you  offer  a  child  a  variety  of  foods  to  eat   such  as  fruits,  vegetables,  cereals  and  dairy?   -­‐.15   .43   Are  you  seated  at  a  table  when  you  eat?   .07   .45   Usually,  how  many  servings  a  day  of  fruit  or   fruit  juice  do  you  eat?   -­‐.05   .46   -­‐.04   .51   .02   .25   .39   .03   -­‐.19   .07   .32   .29   .99   .07   -­‐.51   .25   Do  you  plan  to  eat  at  the  same  time  every  day?   Usually,  how  many  servings  a  day  of   vegetables  do  you  eat?   How  many  times  a  week  do  you  usually  eat   breakfast?   Do  you  give  a  child  the  same  amounts  of  food   you  give  yourself?   Do  you  let  a  child  decide  if  he/she  is  hungry  at   meals?   Do  you  have  a  child  at  least  taste  everything   on  the  plate?   Do  you  have  a  child  eat  everything  on  the   plate?   Does  a  child  decide  how  much  he/she  eats  at   meals?                   60     APPENDIX  E:  Subscale  of  Child  Behavior  Checklist  for  Ages  1  ½  -­‐  5     Aggressive  Behavior  Subscale   0  =  Not  True  (as  far  as  you  know)   1  =  Somewhat  or  Sometimes  True   2  =  Very  True  or   Often  True   ___    8.  Can’t  stand  waiting;  wants  everything  now   ___  15.  Defiant   ___  16.  Demands  must  be  met  immediately   ___  18.  Destroys  property  belonging  to  others   ___  20.  Disobedient   ___  27.  Doesn’t  seem  to  feel  guilty  after  misbehaving   ___  29.  Easily  frustrated   ___  35.  Gets  in  many  fights   ___  40.  Hits  others   ___  42.  Hurts  animals  or  people  without  meaning  to   ___  44.  Angry  moods   ___  53.  Physically  attacks  people   ___  58.  Punishment  doesn’t  change  his/her  behavior   ___  66.  Screams  a  lot   ___  69.  Selfish  or  won’t  share   ___  81.  Stubborn,  sullen,  or  irritable   ___  85.  Temper  tantrums  or  hot  temper   ___  88.  Uncooperative   ___  96.  Wants  a  lot  of  attention     61     APPENDIX  F:  Center  for  Epidemiologic  Studies  Depression  Scale   How  often  during  the  paste  week  have  you  felt  (READ  STATEMENT)—would  you  say:   rarely  or  never,  some  or  a  little  of  the  time,  occasionally  or  a  moderate  amount  of  time,  or   most  or  all  of  the  time?     Rarely  or   Some  or  a   Occasionally  or   Most  or  all  (5-­‐   never  (less   little  (1-­‐2   moderate  (3-­‐4   7  days)   than  1  day)   days)   days)   A.  Bothered  by  things  that   0   1   2   3   usually  don’t  bother  you   B.  You  did  not  feel  like   0   1   2   3   eating;  your  appetite  was   poor   C.  That  you  could  not   shake  off  the  blues,  even   0   1   2   3   with  help  from  family  and   friends   D.  That  you  were  as  good   0   1   2   3   as  other  people   E.  You  had  trouble   keeping  your  mind  on   0   1   2   3   what  you  were  doing   F.  Depressed   0   1   2   3   G.  That  everything  you  did   0   1   2   3   was  an  effort   H.  Hopeful  about  the   0   1   2   3   future   I.  Your  life  has  been  a   0   1   2   3   failure   J.  Fearful   0   1   2   3   K.  Your  sleep  was  restless   0   1   2   3   L.  You  were  happy   0   1   2   3   M.  You  talked  less  than   0   1   2   3   usual   N.  You  felt  lonely   0   1   2   3   O.  People  were  unfriendly   0   1   2   3   P.  You  enjoyed  life   0   1   2   3   Q.  You  had  crying  spells   0   1   2   3   R.  You  felt  sad   0   1   2   3   S.  You  felt  that  people   0   1   2   3   dislike  you   T.  You  could  not  get   0   1   2   3   “going”       62     APPENDIX  G:  Frequencies  of  Continuous  Variables   Table  2   Frequencies  of  Continuous  Variables     N   Valid     Missing   Depression   Maternal   Aggression   Child  BMI   Food   Meal   BMI   percentile   Control   Patterns   119   119   119   119   119   119   0   0   0   0   0   0   10.60   27.17   11.9   65.06   19.31   19.71   9   26.61   11   79.9   20   20   11   31.89   8   99.90   19   21   7.94   6.54   6.23   35.1   2.96   4.84   63.08   42.8   39.1   1231.86   8.76   23.27   Range   34   39.30   32   99.90   15   23   Minimum   .00   16.46   .00   .00   10   7   Maximum   34   55.77   32   99.90   25   30   Skewness   .99   .92   .52   -­‐.62   -­‐.47   .01   Kurtosis   .65   1.92   .03   -­‐1.17   .46   -­‐.36   Mean   Median   Mode   Std.   Deviation   Variance                   63     APPENDIX  H:  Crosstabulation  of  BMI  and  Depressive  Symptoms  for  Mothers   Table  3   Crosstabulation  of  BMI  and  Depressive  Symptoms  for  Mothers       Underweight   Normal  weight   Overweight   Obese   Total       Not   Depressed   N   Percent   6   5%   35   29.4%   25   21%   26   21.8%   92   77.3%   Possibly   Depressed   N   Percent   -­‐   -­‐   8   6.7%   3   2.5%   6   5%   17   14.3%                                 64   Probably   Depressed   N   Percent   -­‐   -­‐   3   2.5%   2   1.7%   5   4.2%   10   8.4%   Total   N   6   46   30   37   119   Percent   5%   38.7%   25.2%   31.1%   100%     APPENDIX  I:  Crosstabulation  of  BMI  and  Aggression  for  Children   Table  4     Crosstabulation  of  BMI  and  Aggression  for  Children       Underweight   Normal  weight   Overweight   Obese   Total     Aggression  -­‐ Aggression  -­‐  At   Normal     Risk   N   Percent   N   Percent   8   6.7%   -­‐   -­‐   52   43.7%   4   3.4%   14   11.8%   4   3.4%   33   27.7%   2   1.7%   107   89.9%   10   8.4%                                   65   Aggression  -­‐ Clinical   N   Percent   -­‐   -­‐   1   0.8%   -­‐   -­‐   1   0.8%   2   1.7%   Total   N   8   57   18   36   119   Percent   6.7%   47.9%   15.1%   30.3%   100%     APPENDIX  J:  Bivariate  Correlations   Table  5   Bivariate  Correlations     1.  Maternal  BMI   2.  Maternal   depressive   symptoms   3.  Child  BMI   4.  Child  Aggression   5.  Food  Control   6.  Meal  Patterns   1   -­‐-­‐   2   .15   -­‐-­‐                     3   -­‐.01   -­‐.08   4   -­‐.10   .21*   5   .06   .00   6   .00   -­‐.21*   -­‐-­‐   .09   -­‐-­‐   -­‐.11   -­‐.11   -­‐-­‐   .06   -­‐.06   .05   -­‐-­‐           *  Correlation  is  significant  at  the  0.01  level  (2-­‐tailed).                                   66                                 REFERENCES                         67     REFERENCES       Aber,  J.  L.,  Jones,  S.,  &  Cohen,  J.  (2000).  The  impact  of  poverty  on  the  mental  health     and  development  of  very  young  children.  In  C.  Zeanah  (Ed.),  Handbook  of  infant   mental  health  (2nd  ed.,  pp.  113-­‐128).  New  York:  Guilford  Press.       Achenbach,  T.,  &  Rescorla,  L.  (2000).  Manual  for  the  ASEBA  preschool  forms  &     profiles.  Burlington,  VT:  University  of  Vermont,  Research  Center  for  Children,  Youth,   &  Families.     Achenbach,  T.  M.,  &  Ruffle,  T.  M.  (2000).  The  child  behavior  checklist  and  related     forms  for  assessing  behavioral/emotional  problems  and  competencies.  Pediatrics  in   Review,  21,  265-­‐271.     Agliata,  D.,  &  Tantleff-­‐Dunn,  S.  (2004).  The  impact  of  media  exposure  on  males'  body     image.  Journal  of  Social  and  Clinical  Psychology,  23,  7-­‐22.       Anderson,  J.  W.,  Smith,  B.  M.,  &  Gustafson,  N.  J.  (1994).  Health  benefits  and  practical     aspects  of  high-­‐fiber  diets.  American  Journal  of  Clinical  Nutrition,  59,  S1242-­‐S1247.     Andrews,  M.  W.,  &  Rosenblum,  L.  A.  (1994).  The  development  of  affiliative  and     agonistic  social  patterns  in  differentially  reared  monkeys.  Child  Development,  65,   1398-­‐1404.     Baranowski,  T.  (1997).  Families  and  health  factors:  Personal  and  social     determinants.  In  D.  S.  Gochmen  (Ed.),  Handbook  of  Health  Behavior  Research  I     (pp.  179-­‐206).  New  York:  Kluwer  U.  S.     Barge-­‐Schaapveld,  D.,  Nicolson,  N.,  Berkhof,  J.,  &  deVries,  M.  (1999).  Quality  of  life  in     depression:  Daily  life  determinants  and  variability.  Psychiatry  Research,  88,  173-­‐189.     Bartlett,  S.,  Krishnan,  J.,  Riekert,  K.,  &  Butz,  A.  (2004).    Children  with  asthma     maternal  depressive  symptoms  and  adherence  to  therapy  in  inner-­‐city.  Pediatrics,   113,  229-­‐237.       Baughcum,  A.  E.,  Burklow,  K.  A.,  Deeks,  C.  M.,  Powers,  S.  W.,  &  Whitaker,  R.  C.  (1998).     Maternal  feeding  practices  and  childhood  obesity:  A  focus  group  study  of  low-­‐ income  mothers.  Archives  of  Pediatrics  &  Adolescent  Medicine,  152,  1010-­‐1014.     Baughcum,  A.  E.,  Powers,  S.  W.,  Johnson,  S.  B.,  Chamberlin,  L.  A.,  Deeks,  C.  M.,     Jain,  A.,  et  al.  (2001).  Maternal  feeding  practices  and  beliefs  and  their  relationships   to  overweight  in  early  childhood.  Journal  of  Developmental  and  Behavioral   Pediatrics,  22,  391-­‐408.         68     Baumrind,  D.  (1966).  Effects  of  authoritative  parental  control  on  child  behavior.     Child  Development,  37,  887-­‐907.     Beck,  A.  T.,  Epstein,  N.,  Brown,  G.,  &  Steer,  R.  A.  (1988).  An  inventory  for  measuring     clinical  anxiety:  Psychometric  properties.  Journal  of  Consulting  and  Clinical   Psychology,  56,  893-­‐897.     Birch,  L.  L.  (1992).  Children's  preferences  for  high-­‐fat  foods.  Nutrition  Reviews,  50,     249-­‐255.     Birch,  L.  L.  (1999).  Development  of  food  preferences.  Annual  Review  of  Nutrition,     19,  41-­‐62.     Birch,  L.  L.,  &  Fisher,  J.  O.  (1995).  Appetite  and  eating  behavior  in  children.  Pediatric     Clinics  of  North  America,  42,  931-­‐953.     Birch,  L.  L.,  &  Fisher,  J.  O.  (1998).  Development  of  eating  behaviors  among  children     and  adolescents.  Pediatrics,  101,  539-­‐549.     Birch,  L.  L.,  &  Fisher,  J.  O.  (2000).  Mothers'  child-­‐feeding  practices  influence     daughters'  eating  and  weight.  American  Journal  of  Clinical  Nutrition,  71,  1054-­‐1061.     Birch,  L.,  Fisher,  J.,  &  Davison,  K.  (2003).  Learning  to  overeat:  Maternal  use  of     restrictive  feeding  practices  promotes  girls'  eating  in  the  absence  of  hunger.   American  Journal  of  Clinical  Nutrition,  78,  215-­‐220.     Birch,  L.,  Fisher,  J.,  Grimm-­‐Thomas,  K.,  Markey,  C.,  Sawyer,  R.,  &  Johnson,  S.  (2001).     Confirmatory  factory  analysis  of  the  Child  Feeding  Questionnaire:  A  measure  of   parental  attitudes,  beliefs  and  practices  about  child  feeding  and  obesity  proneness.   Appetite,  36,  201-­‐210.       Birch,  L.,  &  Ventura,  A.,  (2009).  Preventing  child  obesity:  What  works?  International     Journal  of  Obesity,  33,  S74-­‐S81.       Black,  M.,  Papas,  M.,  Hussey,  J.,  Dubowitz,  H.,  Kotch,  J.,  &  Starr,  R.  (2002).   Behavior  problems  among  preschool  children  born  to  adolescent  mothers:  Effects  of   maternal  depressive  symptoms  and  perceptions  of  partner  relationships.  Journal  of   Clinical  Child  and  Adolescent  Psychology,  31,  16-­‐26.       Blissett,  J.,  &  Haycraft,  E.  (2008).  Are  parenting  style  and  controlling  feeding     practices  related?  Appetite,  50,  477-­‐485.       Blixen,  C.  E.,  Singh,  A.,  &  Thacker,  H.  (2006).  Values  and  beliefs  about  obesity  and     weight  reduction  among  African  American  and  Caucasian  women.  Journal  of   Transcultural  Nursing,  17,  290-­‐297.       69     Boulton,  T.  J.,  Magarey,  A.  M.,  &  Cockington,  R.  A.  (1995).  Tracking  of  serum  lipids     and  dietary  energy,  fat  and  calcium  intake  from  1  to  15  years.  Acta  Paediatrica,  84,   1050-­‐1055.     Bradley,  R.,  Houts,  R.,  Nader,  P.,  O'Brien,  M.,  Belsky,  J.,  &  Crosnoe,  R.,  (2008).  The     relationship  of  body  mass  index  and  behavior  in  children.  The  Journal  of  Pediatrics,   153,  629-­‐634.       Braet,  C.,  Mervielde,  I.,  &  Vandereycken,  W.  (1997).  Psychological  aspects  of     childhood  obesity:  A  controlled  study  in  a  clinical  and  nonclinical  sample.  Journal  of   Pediatric  Psychology,  22,  59-­‐71.     Bronfenbrenner,  U.  (1994).  Ecological  models  of  human  development.  In  T.  Husen,  &     T.  N.  Postlethwaite  (Eds.),  The  international  encyclopedia  of  education  (2nd  ed.).  New   York:  Elsevier  Science.     Bronte-­‐Tinkew,  J.,  Zaslow,  M.,  Capps,  R.,  Horowitz,  A.,  &  McNamara,  M.  (2007).  Food     insecurity  works  through  depression,  parenting,  and  infant  feeding  to  influence   overweight  and  health  in  toddlers.  The  Journal  of  Nutrition,  137,  2160-­‐2165.     Brown,  J.  L.,  &  Miller,  D.  (2002).  Gender  role  preference  and  family  food  chores.     Journal  of  Nutrition  Education  and  Behavior,  34,  100-­‐108.     Brown,  R.,  &  Ogden,  J.  (2004).  Children's  eating  attitudes  and  behaviour:  A  study  of     the  modelling  and  control  theories  of  parental  influence.  Health  Education     Research,  19,  261-­‐271.     Budd,  G.  M.,  &  Volpe,  S.  L.  (2006).  School-­‐based  obesity  prevention:  Research,     challenges,  and  recommendations.  Journal  of  School  Health,  76,  485-­‐495.     Butte,  N.  F.  (2009).  Impact  of  feeding  practices  on  childhood  obesity.  Journal  of     Nutrition,  139,  412S-­‐416S.     Cardon,  L.  (1995).  Genetic  influences  on  body  mass  index  in  early  childhood.  In     Turner,  J.,  Cardon,  L.,  &  Hewitt,  J  (Eds.),  Behavior  Genetic  Approaches  in  Behavioral   Medicine  (pp.  133-­‐143).  New  York:  Plenum  Press.       Carnell,  S.,  &  Wardle,  J.  (2007).  Associations  between  mulitple  measures  of  parental     feeding  and  children's  adiposity  in  United  Kingdom  preschoolers.  Obesity,  15,  137-­‐ 144.     Centers  for  Disease  Control  and  Prevention.  (2008).  Adult  BMI  calculator:  English.     Retrieved  January  7,  2010  from  Healthy  weight:  It's  not  a  diet,  it's  a  lifestyle!:     http://www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/english_bmi_calculat or/bmi_calculator.html       70     Centers  for  Disease  Control  and  Prevention.  (2009a).  About  BMI  for  children  and     teens.  Retrieved  January  7,  2010  from  Healthy  weight:  It's  not  a  diet,  it's  a  lifestyle!:     http://www.cdc.gov/healthyweight/assessing/bmi/childrens_BMI   /about_childrens_BMI.html     Centers  for  Disease  Control  and  Prevention.  (2009b).  Children's  BMI  tool  for     schools.  Retrieved  March  10,  2010  from  Healthy  weight:  It's  not  a  diet,  it's  a   lifestyle!:  http://www.cdc.gov/healthyweight/assessing/bmi/childrens_bmi   /tool_for_schools.html     Chang,  M.,  Nitzke,  S.,  Guilford,  E.,  Adair,  C.  H.,  &  Hazard,  D.  L.  (2008).  Motivators  and     barriers  to  healthful  eating  and  physical  activity  among  low-­‐income     overweight  and  obese  mothers.  Journal  of  the  American  Dietetic  Association,  108,   1023-­‐1028.     Chazan-­‐Cohen,  R.,  Ayoub,  C.,  Alexander  Pan,  A.,  Roggman,  L.,  Raikes,  H.,  McKelvey,  L.,     Whiteside-­‐Mansell,  L.,  &    Hart,  A.,  (2007).  It  takes  time:  Impacts  of  early  head  start   that  lead  to  reductions  in  maternal  depressive  symptoms  two  years  later.  Infant   Mental  Health  Journal,  28,  151-­‐170.       Clarke,  H.  R.,  Goyder,  E.,  Bissell,  P.,  Blank,  L.,  &  Peters,  J.  (2007).  How  do  parents'     child-­‐feeding  behaviours  influence  child  weight?  Implications  for  childhood  obesity   policy.  Journal  of  Public  Health,  29,  132-­‐141.     Contento,  I.  R.,  Zybert,  P.,  &  Williams,  S.  S.  (2005).  Relationship  of  cognitive  restraint     of  eating  and  disinhibition  to  the  quality  of  food  choices  of  Latina  women  and  their   young  children.  Preventative  Medicine,  40,  326-­‐336.     Cooke,  L.  (2007).  The  importance  of  exposure  for  healthy  eating  in  childhood:  A     review.  Journal  of  Human  Nutrition  and  Dietetics,  20,  294-­‐301.     Cooper,  P.,  Whelan,  E.,  Woolgar,  M.,  Morrell,  J.,  &  Murray,  L.  (2004).  Association     between  childhood  feeding  problems  and  maternal  eating  disorder;  role  of  the   family  environment.  British  Journal  of  Psychiatry,  184,  210-­‐215.     Cossrow,  N.,  &  Falkner,  B.  (2004).  Race/ethnic  issues  in  obesity  and  obesity-­‐related     comorbidities.  The  Journal  of  Clinical  Endocrinology  &  Metabolism,  89,  2590-­‐2594.     Costanzo,  P.  R.,  &  Woody,  E.  Z.  (1985).  Domain-­‐specific  parenting  styles  and  their     impact  on  the  child's  development  of  particular  deviance:  The  example  of     obesity  proneness.  Journal  of  Social  &  Clinical  Psychology,  3,  425-­‐430.     Cummins,  S.,  &  Macintyre,  S.  (2006).  Food  environments  and  obesity:     Neighbourhood  or  nation?  International  Journal  of  Epidemiology,  35,  100-­‐104.         71     Danielzik,  S.,  Langnäse,  K.,  Mast,  M.,  &  Spethmann,  C.,  &  Müller,  M.  (2002).  Impact  of     parental  BMI  on  the  manifestation  of  overweight  5–7  year  old  children.  European   Journal  of  Nutrition,  41,  132-­‐138.       Darling,  N.,  &  Steinberg,  L.  (1993).  Parenting  style  as  context:  An  integrative  model.     Psychological  Bulletin,  113,  487-­‐496.     De  Sousa,  A.  (2009).  Maternal,  child  and  family  factors  in  childhood  obesity.     International  Journal  of  Diabetes    &  Metabolism,  17,  111-­‐114.       Devine,  C.  M.,  Connors,  M.  M.,  &  Sobal,  J.  (2003).  Sandwiching  it  in:  Spillover  of  work     onto  food  choices  and  family  roles  in  low-­‐  and  moderate-­‐income  urban  households.   Social  Sciences  &  Medicine,  56,  617-­‐630.     Dickstein,  S.,  Hayden,  L.,  Schiller,  M.,  Seifer,  R.,  &  San  Antonio,  W.  (1994).  The  Family     Mealtime  Interaction  Coding  System.  Unpublished  coding  manual.  Brown  University   School  of  Medicine:  Bradley  Hospital.       Dickstein,  S.,  Seifer,  R.,  Hayden,  L.,  Schiller,  M.,  Sameroff,  A.,  Keitner,  G.,  Miller,  I.,     Rasmussen,  S.,  Matzko,  M.,  &  Magee,  K.  (1998).  Levels  of  family  assesment:  II.  Impact   of  maternal  psychopathology  on  family  functioning.  Journal  of  Family  Psychology,  12,   23-­‐40.       Dixon,  L.  B.,  Cronin,  F.  J.,  &  Krebs-­‐Smith,  S.  M.  (2001).  Let  the  pyramid  guide  your     food  choices:  Capturing  the  total  diet  concept.  Journal  of  Nutrition,  131,  S461-­‐S472.     Drewnowski,  A.  (2004).  Obesity  and  the  food  environmnet:  Dietary  energy  density     and  diet  costs.  American  Journal  of  Preventive  Medicine,  27,  154-­‐162.     Elgar,  F.  J.,  Curtis,  L.  J.,  MGrath,  P.  J.,  Waschbusch,  D.  A.,  &  Stewart,  S.  H.  (2003).     Antecedent  –  consequence  conditions  in  maternal  mood  and  child  adjustment:  A  four-­‐ year  cross-­‐lagged  study.  Journal  of  Clinical  Child  and  Adolescent  Psychology,  32,  362-­‐ 374.       Elgar,  F.  J.,  McGrath,  P.  J.,  Waschbusch,  D.  A.,  Stewart,  S.  H.,  &  Curtis,  L.  J.  (2004).     Mutual  influences  on  maternal  depressive  symptoms  and  child  adjustment  problems.   Clinical  Psychology  Review,  24,  441-­‐459.       Engler-­‐Stringer,  R.  (2010).  The  domestic  foodscapes  of  young  low-­‐Income  women  in     Montreal:  Cooking  practices  in  the  context  of  an  increasingly  processed  food  supply.   Health  Education  &  Behavior,  37,  211-­‐226.       Engstrom,  J.,  Paterson,  S.,  Doherty,  A.,  Trabulsi,  M.,  &  Speer,  K.  (2003).  Accuracy  of     self-­‐reported  height  and  weight  in  women:  An  integrative  review  of  the  literature.   Journal  of  Midwifery  &  Women's  Health,  48,  338-­‐345.         72     Ertel,  K.,  Koenen,  K.,  Rich-­‐Edwards,  J.,  &  Gillman,  M.  (2010).  Antenatal  and     postpartum  depressive  symptoms  are  differentially  associated  with  early  childhood   weight  and  adiposity.  Paediatric  and  Perinatal  Epidemiology,  24,  179-­‐189.       Evans,  W.  D.,  Finkelstein,  E.  A.,  Kamerow,  D.  B.,  &  Renaud,  J.  M.  (2005).  Public     perceptions  of  childhood  obesity.  American  Journal  of  Preventive  Medicine,  28,  26-­‐32.     Faith,  W.  D.,  Scanlon,  K.  S.,  Birch,  L.  L.,  Francis,  L.  A.,  &  Sherry,  B.  (2004).  Parent-­‐child     feeding  strategies  and  their  relationships  to  child  eating  and  weight  status.     Obesity  Research,  12,  1711-­‐1722.     Fergusson  D.  M.,  &  Horwood  L.  J.  (1998).  Early  conduct  problems  and  later  life     opportunities.  Journal  of  Child  Psychology  and  Psychiatry,  39,  1097-­‐1108.       Finkelstein,  E.  A.,  Fiebelkorn,  I.  C.,  &  Wang,  G.  (2003).  National  medical  spending     attributable  to  overweight  and  obesity:  How  much,  and  who's  paying?  Health  Affairs,   W3,  219-­‐226.     Fisher,  J.  O.,  &  Birch,  L.  L.  (2001).  Early  experience  with  food  and  eating:     Implications  for  the  development  of  eating  problems.  In  J.  K.  Thompson,  &  L.  Smolak   (Eds.),  Body  image,  eating  disorders,  and  obesity  in  children  and  adolescents:  Theory,   assessment,  treatment,  and  prevention.  Washington,  DC:  American  Psychological   Association.     Fisher,  J.  O.,  &  Birch,  L.  L.  (2002).  Eating  in  the  absence  of  hunger  and  overweight  in     girls  from  5  to  7  y  of  age.  American  Journal  of  Clinical  Nutrition,  76,  226-­‐231.     Fisher,  J.  O.,  &  Birch,  L.  L.  (1999a).  Restricting  access  to  foods  and  children's  eating.     Appetite,  32,  405-­‐419.     Fisher,  J.  O.,  &  Birch,  L.  L.  (1999b).  Restricting  access  to  palatable  foods  affects     children's  behavioral  response,  food  selection,  and  intake.  American  Journal  of   Clinical  Nutrition,  69,  1264-­‐1272.     Flegal,  K.,  Carroll,  M.,  Ogden,  C.,  &  Curtin,  L.  (2010).  Prevalence  and  trends  in  obesity     among  US  adults,  1999-­‐2008.  Journal  of  the  American  Medical  Association,  303,  235-­‐ 241.       Flodmark,  C.  E.,  Marcus,  C.,  &  Britton,  M.  (2006).  Interventions  to  prevent  obesity  in     children  and  adolescents:  A  systematics  literature  review.  International     Journal  of  Obesity,  30,  579-­‐589.     Food  &  Nutrition  Service.  (2010).  WIC  at  a  glance.  Retrieved  from  About  WIC  on     May  18th,  2010  from:  http://www.fns.usda.gov/wic/aboutwic/   wicataglance.htm       73     Freedman,  D.  S.,  Kettle  Khan,  L.,  Serdula,  M.  K.,  Dietz,  W.  H.,  Srinivasan,  S.  R.,  &     Berenson,  G.  S.  (2005).  The  relation  of  childhood  BMI  to  adult  adiposity:  The   bogalusa  heart  study.  Pediatrics,  115,  22-­‐27.     Garcia,  J.,  Hankins,  W.  G.,  &  Rusinak,  K.  W.  (1974).  Behavioral  regulation  of  the     milieu  interne  in  man  and  rat.  Science,  185,  824-­‐831.     Garner,  D.  M.  (1991).  Eating  disorder  inventory-­‐2  professional  manual.  Odessa,     Florida:  Psychological  Assessment  Resources.     Gibson,  E.  L.  (2006).  Emotional  influences  on  food  choice:  Sensory,  physiological  and     psychological  pathways.  Physiology  and  Behavior,  89,  53-­‐61.     Gladstone,  T.  R.  G.,  &  Beardslee,  W.  R. (2002).  Treatment,  intervention,  and     prevention  with  children  of  depressed  parents:  A  developmental  perspective.  In  S.H.   Goodman  &  I.  H.  Gotlib  (Eds.),  Children  of  depressed  parents:  Mechanisms  of  risk  and   implications  for  treatment  (pp.  277–305).  Washington,  DC:  American  Psychological   Association.     Goodman,  S.  J.,  &  Gotlib,  I.  H.  (1999).  Risk  for  psychopathology  in  the  children  of     depressed  mothers:  A  developmental  model  for  understanding  mechanisms  of   transmission.  Psychological  Review,  106,  458-­‐490.       Gorber,  S.,  Tremblay,  M.,  &  Moher,  D.  (2007).  A  comparison  of  direct  vs.  self-­‐report     measures  for  assessing  height,  weight  and  body  mass  index:  A  systematic  review.   Obesity  Reviews,  8,  307-­‐326.           Grabe,  S.,  Ward,  L.,  &  Hyde,  J.  (2008).  The  role  of  the  media  in  body  image  concerns     among  women:  A  meta-­‐analysis  of  experimental  and  correlational  studies.   Psychological  Bulletin,  134,  460-­‐476.       Haycraft,  E.  &  Blissett,  J.  (2008).  Controlling  feeding  practices  and  psychopathology     in  a  non-­‐clinical  sample  of  mothers  and  fathers.  Eating  Behaviors,  9,  484-­‐492.     Hearn,  M.  D.,  Baranowski,  T.,  &  Baranowski,  J.  (1998).  Environmental  influences  on     dietary  behavior  among  children:  Availability  and  accessibility  of  fruits  and     vegetables  enable  consumption.  Journal  of  Health  Education,  29,  26-­‐32.     Herbozo,  S.,  Tantleff-­‐Dunn,  S.,  Gokee-­‐Larose,  J.,  Thompson,  J.  (2004).  Beauty  and     thinness  messages  in  children's  media:  A  content  analysis.  Eating  Disorders,  12,  21-­‐ 34.       Heo,  M.,  Pietrobelli,  A.,  Fontaine,  K.  R.,  Sirey,  J.  A.,  &  Faith,  M.  S.  (2006).  Depressive     mood  and  obesity  in  US  adults:  Comparison  and  moderation  by  sex,  age,  and  race.   International  Journal  of  Obesity,  30,  513-­‐519.       74     Hill,  J.  O.,  &  Peters,  J.  C.  (1998).  Environmental  contributions  to  the  obesity  epidemic.     Science,  280,  1371-­‐1374.     Hoerr,  S.  L.,  Horodynski,  M.  A.,  Lee,  S.  Y.,  &  Henry,  M.  (2006).  Predictors  of     nutritional  adequacy  in  mother-­‐toddler  dyads  from  rural  families  with     limited  incomes.  Journal  of  American  Dietetic  Association,  106,  1766-­‐1773.     Hoerr,  S.  L.,  Hughes,  S.  O.,  Fisher,  J.  O.,  Nicklas,  T.  A.,  Liu,  Y.,  &  Shewchuk,  R.  M.     (2009).  Associations  among  parental  feeding  styles  and  children's  food  intake  in   families  with  limited  incomes.  International  Journal  of  Behavioral  Nutrition  and   Physical  Activity,  6,  55-­‐61.     Hoerr,  S.  L.,  Lee,  S.,  Schiffman,  R.  F.,  Horodynski,  M.,  &  McKelvey,  L.  (2006).  Beverage     consumption  of  mother-­‐toddler  dyads  in  families  with  limited  incomes.  Journal  of   Pediatric  Nursing  ,  21,  403-­‐411.     Hofstra,  M.  B.,  Van  der  Ende,  J.,  &  Verhulst,  F.  C.  (2000).  Continuity  and  change  of     psychopathology  from  childhood  into  adulthood:  A  14-­‐year  follow-­‐up  study.  Journal   of  the  American  Academy  of  Child  and  Adolescent  Psychiatry,  39,  850-­‐858.     Horodynski,  M.,  Stommel,  M.,  Brophy-­‐Herb,  H.,  Xie,  Y.,  &  Weatherspoon,  L.  (2010).     Low-­‐income  African  American  and  Non-­‐Hispanic  White  mothers'  self  efficacy,  "picky   eater"  perception,  and  toddler  fruit  and  vegetable  intake.  Public  Health  Nursing,  27,   408-­‐417.     Hughes,  S.  O.,  Power,  T.  G.,  Fisher,  J.  O.,  Mueller,  S.,  &  Nicklas,  T.  A.  (2005).  Revisiting     a  neglected  construct:  Parenting  styles  in  a  child-­‐feeding  context.  Appetite,  44,  83-­‐ 92.     Johnson,  S.  L.  (2000).  Improving  preschoolers'  self-­‐regulation  of  energy  intake.     Pediatrics,  106,  1429-­‐1435.     Johnson,  S.  L.,  &  Birch,  L.  L.  (1994).  Parents'  and  children's  adiposity  and  eating     style.  Pediatrics,  94,  653-­‐661.     Kelleher,  K.,  McInerny,  T.,  Gardner,  W.,  Childs,  G.,  &  Wasserman,  R.  (2000).     Increasing  identification  of  psychosocial  problems:  1979-­‐1996.  Pediatrics,  105,   1313-­‐1321.       Klesges,  R.,  Coates,  T.,  Guendoline,  B.,  Sturgeon-­‐Tillisch,  J.,  Moldenhauer-­‐Klesges,  L.,     Holizer,  B.,  Woolfrey,  J.,  &  Vollmer,  J.  (1983).  Parental  influences  on  children's  eating   behavior  and  relative  weight.  Journal  of  Applied  Behavior  Analysis,  16,  371-­‐378.       Klesges,  R.  C.,  Stein,  R.  J.,  Eck,  L.  H.,  Isbell,  T.  R.,  &  Klesges,  L.  M.  (1991).  Parental     influence  on  food  selection  in  young  children  and  its  relationships  to  chidhood   obesity.  American  Journal  of  Clinical  Nutrition,  53,  859-­‐864.     75     Koivisto  Hursti,  U.  K.  (1999).  Factors  influencing  children's  food  choice.  Annals  of     Medicine,  31,  26-­‐32.     Kranz,  S.,  Hartman,  T.,  Siega-­‐Riz,  A.  M.,  &  Herring,  A.  H.  (2006).  A  diet  quality  index     for  American  preschoolers  based  on  current  dietary  intake  recommendations  and   an  indicator  of  energy  balance.  Journal  of  the  American  Dietetic  Association,  106,   1594-­‐1604.     Kranz,  S.,  Siega-­‐Riz,  A.  M.,  &  Herring,  A.  H.  (2004).  Changes  in  diet  quality  of     American  preschoolers  between  1977  and  1998.  American  Journal  of  Public  Health,   94,  1525-­‐1530.     Krebs,  N.  F.,  Baker,  R.  D.,  Greer,  F.  R.,  Heyman,  M.  B.,  Jaksic,  T.,  Lifshitz,  F.,  et  al.     (2003).  Prevention  of  pediatric  overweight  and  obesity.  Pediatrics,  112,  424-­‐430.     Kropski,  J.  A.,  Keckley,  P.  H.,  &  Jensen,  G.  L.  (2008).  School-­‐based  obesity  prevention     programs?  An  evidence-­‐based  review.  Obesity,  16,  1009-­‐1018.     Lake,  J.,  Power,  C.,  &  Cole,  T.  (1997).  Child  to  adult  body  mass  index  in  the  1958     British  birth  cohort:  Associations  with  parental  obesity.  Archives  of  Diseases  in   Childhood,  77,  376-­‐380.       LaRosa,  A.,  Glascoe,  F.,  &  Macias,  M.  (2009).  Parental  depressive  symptoms:     Relationship  to  child  development,  parenting,  health,  and  results  on  parent-­‐reported   screening  tools.  The  Journal  of  Pediatrics,  155,  124-­‐128.       Lau,  R.  R.,  Quadrel,  M.  J.,  &  Hartman,  K.  A.  (1990).  Development  and  change  of  young     adults'  preventative  health  beliefs  and  behavior:  Influence  from  parents  and  peers.   Journal  of  Health  and  Social  Behavior,  31,  240-­‐259.     Lee,  S.,  Hoerr,  S.  L.,  &  Schiffman,  R.  F.  (2005).  Screening  for  infants'  and  toddlers'     dietary  quality  through  maternal  diet.  The  American  Journal  of  Maternal  Child   Nursing,  30,  60-­‐66.     Lobstein,  T.,  Baur,  L.,  &  Uauy,  R.  (2004).  Obesity  in  children  and  young  people:  A     crisis  in  public  health.  Obesity  Reviews,  5,  S4-­‐S85.     Loeber,  R.,  &  Hay,  D.  (1997).  Key  issues  in  the  development  of  aggression  and     violence  from  childhood  to  adulthood.  Annual  Reviews  of  Psychology,  48,  371-­‐410.       Love,  J.  M.,  Kisker,  E.  E.,  Ross,  C.,  Raikes,  H.,  ,  Constantine,  J.,  Boller,  K.,  Brooks-­‐Gunn,     J.,  et  al.  (2005).  The  effectiveness  of  Early  Head  Start  for  3  year  old  children  and   their  parents:  Lessons  for  policy  and  programs.  Developmental  Psychology,  41,  885-­‐ 901.         76     Lumeng,  J.  C.,  &  Burke,  L.  B.  (2006).  Maternal  prompts  to  eat,  child  compliance,  and     mother  and  child  weight  status.  The  Journal  of  Pediatrics,  149,  330-­‐335.     Magarey,  A.  M.,  Daniels,  L.  A.,  Boulton,  T.  J.,  &  Cockington,  R.  A.  (2003).  Predicting     obesity  in  early  adulthood  from  childhood  and  parental  obesity.  International   Journal  of  Obesity,  27,  505-­‐513.     Malik,  N.  M.,  Boris,  N.  W.,  Heller,  S.  S.,  Harden,  B.  J.,  Squires,  J.,  Chazan-­‐Cohen,  R.,     Beeber,  L,  S.,  &  Kaczynski,  K.  J.  (2007).  Risk  for  maternal  depressive  symptoms  and   child  aggression  in  early  head  start  families:  A  test  of  ecological  models.  Infant   Mental  Health  Journal,  28,  171-­‐191.     Malina,  R.  (1995).  Anthropometry.  In  P.  Maud,  &  C.  Foster  (Eds.),  Physiological     assessment  of  human  fitness  (pp.  205-­‐219).  Champaign,  IL:  Human  Kinetics.       McLaughlin,  C.,  Tarasuk,  V.,  &  Kreiger,  N.  (2003).  An  examination  of  at-­‐home  food     preparation  activity  among  low-­‐income,  food-­‐insecure  women.  Journal  of  the   American  Dietetic  Association,  103,  1506-­‐1512.     Melgar-­‐Quinonez,  H.  R.,  &  Kaiser,  L.  L.  (2004).  Relationship  of  child-­‐feeding  practices     to  overweight  in  low-­‐income  Mexican-­‐American  preschool-­‐aged  children.  Journal  of   the  American  Dietetic  Association,  104,  1110-­‐1119.     Mellen,  P.  B.,  Walsh,  T.  F.,  &  Herrington,  D.  M.  (2008).  Whole  grain  intake  and     cardiovascular  disease:  A  meta-­‐analysis.  Nutrition,  Metabolism  &  Cardiovascular   Diseases,  18,  283-­‐290.     Moore,  S.  N.,  Tapper,  K.,  &  Murphy,  S.  (2007).  Feeding  strategies  used  by  mothers  of     35  year  old  children.  Appetite,  49,  704-­‐707.     Nagin,  D.,  &  Tremblay  R.  E.  (1999).  Trajectories  of  boys’  physical  aggression,     opposition,  and  hyperactivity  on  the  path  to  physically  violent  and  nonviolent   juvenile  delinquency.  Child  Development,  70,  1181-­‐1196.     Nestle,  M.,  Wing,  R.,  Birch,  L.  L.,  DiSogra,  L.,  Drewnowski,  A.,  &  Middleton,  S.  (1998).     Behavioral  and  social  influences  on  food  choices.  Nutrition  Reviews,  56,  S50-­‐S74.     Ogden,  C.,  Carroll,  M.,  Curtin,  L.,  Lamb,  M.,  &  Flegal,  K.  (2010).  Prevalence  of  high     body  mass  index  in  US  children  and  adolescents,  2007-­‐2008.  Journal  of  the  American   Medical  Association,  303,  242-­‐249.       Ogden,  C.  L.,  Carroll,  M.  D.,  &  Flegal,  K.  M.  (2008).  High  body  mass  index  for  age     among  US  children  and  adolescents,  2003-­‐2006.  Journal  of  the  American  Medical   Association,  299,  2401-­‐2405.         77     Parsons,  T.  J.,  Power,  C.,  Logan,  S.,  &  Summerbell,  C.  D.  (1999).  Childhood  predictors     of  adult  obesity:  A  systematic  review.  International  Journal  of  Obesity,  23,  S1-­‐S107.     Patrick,  H.,  &  Nicklas,  T.  A.  (2005).  A  review  of  family  and  social  determinants  of     children's  eating  patterns  and  diet  quality.  Journal  of  the  American  College  of   Nutrition,  24,  83-­‐92.     Petterson,  S.  M.,  &  Friel,  L.  V. (2001).  Psychological  distress,  hopelessness  and       welfare.  Women  and  Health,  32,  79-­‐99.     Pietrobelli,  A.,  Faith,  M.  S.,  Allison,  D.  B.,  Gallagher,  D.,  Chiumello,  G.,  &  Heymsfield,  S.     B.  (1998).  Body  mass  index  as  a  measure  of  adiposity  among  children  and   adolescents:  A  validation  study.  The  Journal  of  Pediatrics,  132,  204-­‐210.     Pilowsky,  D.  J.,  Wickramaratne,  P.,  Rush,  A.  J.,  Hughes,  C,  W.,  Garger,  J.,  Malloy,  E.,     King,  C.  A.,  Cerdda,  G.,  Sood,  A.  B.,  Alpert,  J.  E.,  Wisniewski,  S.  R.,  Trivedi,  M.  H.,  Talati,   A.,  Carlson,  M.  M.,  Liu,  H.  H.,  Fava,  M.,  &  Weissman,  M.  M.  (2006).  Children  of   currently  depressed  mothers:  A  STAR*D  ancillary  study.  Journal  of  Clinical   Psychiatry,  67,  126-­‐136.       Pratt,  L.  A.,  &  Brody,  D.  J.  (2008).  Depression  in  the  United  States  Household     Population,  2005-­‐2006.  Retrieved  April  28,  2010  from  NCHS  Data  Brief:   http://www.cdc.gov/nchs/data/databriefs/db07.htm.       Prentice,  A.  M.  (1997).  Obesity-­‐the  inevitable  penalty  of  civilization?  British  Medical     Bulletin,  53,  229-­‐237.     Prentice,  A.  M.  (2006).  The  emerging  epidemic  of  obesity  in  developing  countries.     International  Journal  of  Epidemiology,  35,  93-­‐99.       Radloff,  L.  S.  (1977).  The  CES-­‐D  scale:  A  self-­‐report  depression  scale  for  research  in     the  general  population.  Applied  Psychological  Measures,  1,  385-­‐401.     Raine,  A.,  Reynolds,  C.,  Venables,  P,  Mednick,  S.,  &  Farrington,  D.  (1998).     Fearlessness,  stimulation-­‐seeking,  and  large  body  size  at  age  3  years  as  early   predispositions  to  childhood  aggression  at  age  11  years.  Archives  of  General   Psychiatry,  55,  745-­‐751.       Reilly,  J.  J.,  Armstrong,  J.,  Dorosty,  A.  R.,  Emmett,  P.,  Ness,  A.,  Rogers,  I.,  et  al.  (2005).     Early  life  risk  factors  for  obesity  in  childhood:  Cohort  study.  British  Medical  Journal,   330,  1357-­‐1363.     Rhee,  K.  E.,  Lumeng,  J.  C.,  Appugliese,  D.  P.,  Kaciroti,  N.,  &  Bradley,  R.  H.  (2006).     Parenting  styles  and  overweight  status  in  first  grade.  Pediatrics,  117,  2047-­‐2054.         78     Robinson,  T.  N.,  Kiernan,  M.,  Matheson,  D.  M.,  &  Haydel,  K.  F.  (2001).  Is  parental     control  over  children's  eating  associated  with  childhood  obesity?  Results  from  a   population-­‐based  sample  of  third  graders.  Obesity,  9,  306-­‐312.     Rosenbloom,  A.  L.  (2002).  Increasing  incidence  of  type  2  diabetes  in  children  and     adolescents:  Treatment  considerations.  Pediatric  Drugs,  4,  209-­‐221.     Rosenblum,  L.  A.,  &  Paully,  G.  S.  (1984).  The  effects  of  varying  environmental     demands  on  maternal  and  infant  behavior.  Child  Development,  55,  305-­‐314.     Rozin,  P.,  &  Vollmecke,  T.  A.  (1986).  Food  likes  and  dislikes.  Annual  Review  of     Nutrition,  6,  433-­‐456.     Satter,  E.  (1990).  The  feeding  relationship:  Problems  and  interventions.  The  Journal     of  Pediatrics,  117,  S181-­‐S189.     Schiffman,  R.,  Gibbons,  C.,  Fitzgerald,  H.,  Omar  (Horodynski),  M.,  Keefe,  D.,  Reischl,  T.,     et  al.  (1996).  MSU  Pathways  Project:  Research  into  Directions  for  Family  Health  and   Service  Use:  Administration  on  Children,  Youth  and  Families.     Schwimmer,  J.  B.,  Deutsch,  R.,  Kahen,  T.,  Lavine,  J.  E.,  Stanley,  C.,  &  Behling,  C.     (2006).  Prevalence  of  fatty  liver  in  children  and  adolescents.  Pediatrics,  118,  1388-­‐ 1393.     Serdula,  M.  K.,  Ivery,  D.,  Coates,  R.  J.,  Freedman,  D.  S.,  Williamson,  D.  F.,  &  Byers,  T.     (1993).  Do  obese  children  become  obese  adults?  A  review  of  the  literature.   Preventative  Medicine,  22,  167-­‐177.     Simon,  G.  E.,  Von  Korff,  M.,  Saunders,  K.,  Miglioretti,  D.  L.,  Krane,  P.  K.,  van  Belle,  G.,  et     al.  (2006).  Association  between  obesity  and  psychiatric  disorders  in  the  US  adult   population.  Archives  of  General  Psychiatry,  63,  824-­‐830.     Skinner,  J.  D.,  Carruth,  B.  R.,  Bounds,  W.,  &  Zeigler,  P.  J.  (2002).  Do  food-­‐related     experiences  in  the  first  two  years  of  life  predict  dietary  variety  in  school-­‐aged   children?  Journal  of  Nutrition  Education  Behavior,  34,  310-­‐315.     Sonneville,  K.  R.,  &  Gortmaker,  S.  L.  (2008).  Total  energy  intake,  adolescent     discretionary  behaviors  and  the  energy  gap.  International  Journal  of  Obesity,  32,  S19-­‐ S27.     Spencer,  E.,  Appleby,  P.,  Davey,  G.,  Key,  T.  (2002).  Validity  of  self-­‐reported  height  and     weight  in  4808  EPIC-­‐Oxford  participants.  Public  Health  Nutrition,  5,  561-­‐565.       Stradmeijer,  M.,  Bosch,  J.,  Koops,  W.,  &  Seidell,  J.  (1999).  Family  functioning  and     psychosocial  adjustment  in  overweight  youngsters.  International  Journal  of  Eating   Disorders,  27,  110-­‐114.       79       Strine,  T.  W.,  Mokdad,  A.  H.,  Dube,  S.  R.,  Balluz,  L.  S.,  Gonzalez,  O.,  Berry,  J.  T.,  et  al.     (2008).  The  association  of  depression  and  anxiety  with  obesity  and  unhealthy   behaviors  among  community-­‐dwelling  US  adults.  General  Hospital  Psychiatry,  30,   127-­‐137.     Stunkard,  A.,  Berkowitz,  R.,  Schoeller,  D.,  Maislin,  G.,  &  Stallings,  V.  (2004).   Predictors  of  body  size  in  the  first  2  y  of  life:  A  high-­‐risk  study  of  human  obesity   International  Journal  of  Obesity,  28,  503-­‐513.       Stunkard,  A.,  Berkowitz,  R.,  Stallings,  V.,  &  Cater,  J.  (1999).  Weights  of  parents  and     infants:  Is  there  a  relationship?  International  Journal  of  Obesity  Related  Metabolic   Disorders,  23,  159-­‐162.     Surkan,  P.  J.,  Kawachi,  I.,  &  Peterson,  K.  E.  (2008).  Childhood  overweight  and     maternal  depressive  symptoms.  Journal  of  Epidemiology  and  Community  Health,  62,   e11.     Tabachnick,  B.  G.,  &  Fidell,  L.  S.  (1996).  Using  multivariate  statistics  (3rd  ed.).  New     York:  Harper  Collins.     Thomas,  H.  (2006).  Obesity  prevention  programs  for  children  and  youth:  Why  are     their  results  so  modest?  Health  Education  Research,  21,  783-­‐795.     Topham,  G.  L.,  Page,  M.  C.,  Hubbs-­‐Tait,  L.,  Rutledge,  J.  M.,  Kennedy,  T.  S.,  Shriver,  L.,  &  Harris,     A.  W.  (2009).  Maternal  depressive  symptoms  and  socio-­‐economic  status  moderate   the  parenting  style/child  obesity  association.  Public  Health  Nutrition,  13,  1237-­‐1244.     US  Department  of  Health  and  Human  Services  and  US  Department  of  Agriculture.     (2005).  Dietary  guildelines  for  Americans,  2005.  (6th  Edition).  Washington,  DC:  US   Government  Printing  Office.     Van  Duyn,  M.  S.,  &  Pivonka,  E.  (2000).  Overview  of  health  benefits  of  fruits  and     vegetable  consumption  for  the  dietetics.  Journal  of  the  American  Dietetic  Association,   100,  1511-­‐1521.     Ventura,  A.  K.,  &  Birch,  L.  L.  (2008).  Does  parenting  affect  children's  eating  and     weight  status?  International  Journal  of  Behavioral  Nutrition  and  Physical  Activity,   5:15.     Wake,  M.,  Gold,  L.,  McCallum,  Z.,  Gerner,  B.,  &  Waters,  E.  (2008).  Economic     evaluation  of  a  primary  care  trial  to  reduce  weight  gain  in  overweight/obese   children:  The  LEAP  trial.  Ambulatory  Pediatrics,  8,  336-­‐341.           80     Wang,  Y.  C.,  Gortmaker,  S.  L.,  Sobol,  A.  M.,  &  Kuntz,  K.  M.  (2006).  Estimating  the     energy  gap  among  US  children:  A  counterfactual  approach.  Pediatrics,  118,  1721-­‐ 1733.     Wang,  Y.,  Beydoun,  M.  A.,  Liang,  L.,  Caballero,  B.,  &  Kumanyika,  S.  K.  (2008).  Will  all     Americans  become  overweight  or  obese?  Estimating  the  progression  and  cost  of  the   US  obesity  epidemic.  Obesity,  16,  2323-­‐2330.     Wardle,  J.  (1995).  Parental  influences  on  children's  diets.  Proceedings  of  the     Nutrition  Society,  54,  747-­‐758.     Weaver,  C.  M.  (2000).  The  growing  years  and  prevention  of  osteoporosis  in  later  life.     Proceedings  of  the  Nutrition  Society,  59,  303-­‐306.     Weissman,  M.  M.,  Wickramaratne,  P.,  Nomura,  Y.,  Warner,  V.,  Pilowsky,  D.,  &  Verdeli,     H.  (2006).  Offspring  of  depressed  parents:  20  years  later.  American  Journal  of   Psychiatry,  163,  1001-­‐1008.       Whitaker,  R.  C.,  Phillips,  S.  M.,  &  Orzol,  S.  M.  (2006).  Food  insecurity  and  the  risks  of     depression  and  anxiety  in  mothers  and  behavior  problems  in  their  pre-­‐school  aged   children.  Pediatrics,  118,  859-­‐868.     Whitaker,  R.  C.,  Wright,  J.  A.,  Pepe,  M.  S.,  Seidel,  K.  D.,  &  Dietz,  W.  H.  (1997).  Prediting     obesity  in  young  adulthood  from  childhood  and  parental  obesity.  The  New  England   Journal  of  Medicine,  337,  869-­‐873.     Whitaker,  R.,  Deeks,  C.,  Baughcum,  A.,  &  Specker,  B.  (2000).  The  relationship  of     childhood  adiposity  to  parent  body  mass  index  and  eating  behaviour.  Obesity   Research,  8,  234-­‐240.     Wiig  Dammann,  K.,  &  Smith,  C.  (2009).  Factors  affecting  low-­‐income  women's  food     choices  and  the  perceived  impact  of  dietary  intake  and  socioeconomic  status  on   their  health  and  weight.  Journal  of  Nutrition  Education  and  Behavior,  41,  242-­‐253.     Wiles,  N.,  Northstone,  K.,  Emmett,  P.,  &  Lewis,  G.  (2009).  'Junk  food'  diet  and     childhood  behavioral  problems:  Results  from  the  ALSPAC  cohort.  European  Journal   of  Clinical  Nutrition,  63,  491-­‐498.     World  Health  Organization.  (2003).  Obesity  and  overweight.  Retrieved  August  12,     2009  from   http://www.who.int/dietphysicalactivity/publications/facts/obesity/en/     Yach,  D.,  Stuckler,  D.,  &  Brownell,  K.  D.  (2006).  Epidemiologic  and  economic     consequences  of  the  global  epidemics  of  obesity  and  diabetes.  Nature  Medicine,  12,   65-­‐66.       81     Young-­‐Hyman,  D.,  Schlundt,  D.,  Herman-­‐Wenderoth,  L.,  &  Bozylinski,  K.  (2003).     Obesity,  appearance,  and  psychosocial  adaptation  in  young  African  American   children.  Journal  of  Pediatric  Psychology,  28,  463-­‐472.       Zahn-­‐Waxler,  C.,  Iannotti,  R.  J.,  Cummings,  E.  M.,  &  Denham,  S.  (1990).  Antecedents  of     problem  behaviors  in  children  of  depressed  mothers.  Development  and   Psychopathology,  2,  271-­‐291.         82