SangNam Ahn, PhD, MPSA1,2, Hongmei Zhang, PhD1, Marian Levy, DrPH, RD1

1The University of Memphis School of Public Health; 2Texas A&M University School of Public Health

Abstract Objectives. Obesity is a serious public health concern affecting an increasing number of children and adolescents in the U.S. It has been suggested that adverse childhood experience (ACE)(i.e., domestic violence, parental divorce) could worsen mental health among children, in turn leading to weakened coping skills and eventual obesity. It is unclear, however, how mental health mediates the relationship between ACE and childhood obesity and how neighborhood built environment (NBE) might moderate the effects of ACE on developing mental health problems.

Methods. Data were analyzed using a sample of 42,193 children aged 10 to 17 (mean age=13.7±2.3 years) from the 2011-2012 National Survey of Children’s Health. Using the structural equation modeling (SEM), the current study applies the moderated mediation model to examine moderating effects of NBE on the associations of ACE with mental health problems.

Results. Approximately 3 out of 10 children were overweight or obese. More than half of these children had at least one ACE, and about 15% suffered from depression, anxiety, or behavioral problems. SEM model displayed a significant mediating effect of mental health between experiencing ACE and body mass index (BMI) (p < 0.001). Additionally, as a significant moderator, NBE weakened the negative effects of ACE on mental health problems that may lead to childhood obesity (p < 0.001).

Discussion. Identifying and developing social support mechanisms can improve mental health and obesity risk among children with ACE.

Key words: Childhood obesity, Adverse childhood experience (ACE), Social support, Neighborhood built environment (NBE), Mental health

Conceptual Frameworka Full Modelb Best Fit Modelc Sampling Weighted Modeld

Background

Research Objectives

Methods Table 2. Distribution of body mass index by respondent characteristics

Study Limitations

•  28.9% were overweight or obese. •  50.4% had at least one ACE. •  35.8% had a 20-min PA 3 days per week. •  15% suffered from anxiety, depression, etc. (MHP). •  Number of MHP was a significant mediator. •  NBE potentially weakened effects of ACE on MHP.

ü Obesity severely affects our children. • Childhood obesity is defined as a BMI > the 95th percentile (CDC). • Childhood obesity rates in the U.S. have tripled over the past 3 decades: 16.9%-obese; 14.8%-overweight (NCLS, 2013).

ü Children’s PA & Diet are affected by parents. • Parental modeling relates to child’s diets (Fisher, 2002). • Nearly 7.5 million children had a parent either incarcerated or on probation or parole (Roettger, 2012).

ü NBE could weaken the deleterious effect of ACE on mental health and ACE on obesity. • Buffering Model suggests that social support networks can impact the psychological, physical and social wellbeing of individuals by changing the way a person perceives and reacts to stress (Bowlby, 1971).

ü To assess how ACE relates to mental health problems and obesity. ü To examine the mediating role of mental health problems and PA between ACE and obesity. ü To examine the moderating role of NBE between ACE and mental health status/obesity.

ü Data used in this analysis were collected as part of the 2011-2012 National Survey of Children’s Health (NSCH). ü Study sample included children aged 10 through 17 (n=42,193). ü Adverse childhood experience (9); mental health problems (4), neighborhood built environment (4), physical activity. ü Multiple indices of goodness of fit were used: chi-squared fit statistic, Comparative Fit Index (CFI), BIC, and the root mean square error of approximation (RMSEA). ü SEM was estimated using Stata, version 14/MP. ü Sample weights were further included to the best model (Model D) to test the sensitivity of sample weights to parameter estimates and inferences.

ü A more holistic approach is needed to deal with childhood obesity problems especially among children with ACE. ü The holistic approach will include: 1) identifying social supports (neighborhood built environment); 2) promoting school PA education; and 3) improving mental health. ü Policymakers should invest resources to equip a community with an anti-obesogenic environment (e.g., sidewalks, walking paths, recreational parks, community centers, libraries). ü Childhood obesity should be understood using Social Ecological Model as well as Social Cognitive Theory.

ü We need to further investigate a sampling weighted model (as shown below) using M-plus or other SEM software.

Variables   Mean (±SD) or (%) Age (10-17) 13.7 (±2.3) Sex Male 52.2%

Female 47.8% Household poverty level 0-99 %FPL 12.6%

100-199% FPL 16.9% 200-399% FPL 31.4% 400% FPL or greater 39.1%

BMI Underweight 5.7% Normal 65.2% Overweight 14.8%

  Obese 14.3%

Variables Underweight

(n=2,412) Normal

(n=27,568) Overweight (n=6,217)

Obese (n=5,996)   Overall p-value

Number of ACE (0-9) 0.9 (±1.5) 1.0 (±1.4) 1.2 (±1.6) 1.4 (±1.6)   1.1 (±1.5) <0.001 Experienced extreme economic hardship

(somewhat or very often) 19.0% 18.2% 24.1% 31.1%   21.0%   Parents divorced or separated 20.9% 22.8% 27.7% 29.8% 24.4%

Lived with someone with an alcohol or drug problem 12.6% 13.3% 15.6% 17.2%   14.1%  

Witnessed or was victim of neighborhood violence 10.4% 10.5% 13.8% 16.1% 11.7% Lived with someone who was mentally ill or

suicidal 11.5% 10.9% 12.8% 14.2%   11.6%   Witnessed domestic violence 6.9% 7.1% 9.5% 11.6% 8.0%

Parent served time in jail 6.2% 5.8% 8.7% 10.3%   6.9%   Treated or judged unfairly due to race/ethnicity 4.8% 5.2% 6.3% 6.8% 5.5%

Death of parent 3.3% 4.2% 5.5% 6.5%   4.6%   Number of mental health conditions** 0.25 (±0.6) 0.21 (±0.6) 0.26 (±0.7) 0.33 (±0.8)     <0.001

No condition 82.90% 85.9% 83.9% 80.8%   84.7%   1 condition 11.80% 9.5% 9.8% 10.5% 9.9%

2 conditions 3.4% 3.0% 3.8% 5.0%   3.4%   3 conditions 1.3% 1.2% 1.7% 2.5% 1.4% 4 conditions 0.6% 0.4% 0.8% 1.2%   0.6%  

Number of days of 20 minutes PA/week 4.2 (±2.3) 4.4 (±2.2) 4.2 (±2.2) 3.9 (±2.3)   4.3 (±2.2) <0.001 0 9.9% 8.4% 9.5% 13.1%   9.3%   1 4.9% 8.4% 3.6% 4.5% 3.8% 2 9.4% 8.6% 9.6% 10.5%   9.1%   3 13.1% 13.4% 14.4% 15.6% 13.8% 4 11.5% 12.2% 13.0% 11.6%   12.2%   5 20.8% 21.4% 21.4% 20.4% 21.2% 6 6.4% 9.2% 6.9% 5.4%   8.2%   7 24.0% 23.3% 21.7% 19.0%   22.5%  

Number of NBE (0-4) 3.19 (±1.1) 3.17 (±1.1) 3.10 (±1.1) 2.98 (±1.2)   3.13 (±1.1) <0.001 Having sidewalks/walking paths 75.3% 74.4% 73.5% 70.3% 73.7%

Having a park or playground 82.2% 82.1% 81.6% 78.4%   81.5%   Having a recreation center, community center, or

boys’ or girls’ club 70.8% 70.2% 67.1% 62.9% 68.7% Having a library or bookmobile 91.2% 90.7% 88.4% 86.8%   89.9%  

Note: *BMI=body mass index; **Mental health conditions=depression, anxiety, behavioral disorders, ADHD

Next Step ü We will conduct key informant interviews and parental surveys to identify local NBE and social support systems to help children with ACE improve their PA, mental health, and weight status.

Results

ACE BMI

PhysicalAc0vity

MentalHealthProblems

NeighborhoodBuiltEnvironments

HouseholdIncome

NeighborhoodBuiltEnvironments

NeighborhoodBuiltEnvironments

ACE BMI

PhysicalAc0vity

MentalHealthProblems

NBE*ACEΧ2=4898.0(df=15)RMSEA=0.027[90%CI:0.022,0.033]CFI:0.987BIC:924823.3

NBE

HouseholdIncome

0.002

-0.01

-0.03*

0.05* -0.0002

0.13*-0.002

-0.01*

0.03*

-0.24*

-0.02* -0.11*

0.03*

bDirecteffect:Standardizedpathcoefficients(β)areindicated(*p<0.01).Do^edlineindicatesnon-significantpath(p>0.005).

aNBEasamoderatorofthemediatedpathwaysfromACEtophysicalac0vity,mentalhealthproblems,andBMI.Physicalac0vityandmentalhealthproblemsasmediatorsfromACEtoBMI.

ACE BMI

PhysicalAc0vity

MentalHealthProblems

NBE*ACEΧ2=4898.0(df=15)RMSEA=0.019[90%CI:0.016,0.023]CFI:0.984BIC:924809.0

NBE

HouseholdIncome

-0.03*

0.04*

0.13*-0.002

-0.01*

0.03*

-0.26*

-0.02* -0.11*

0.03*

cDirecteffect:Standardizedpathcoefficients(β)areindicated(*p<0.01).Do^edlineindicatesnon-significantpath(p>0.005).

ACE BMI

PhysicalAc0vity

MentalHealthProblems

NBE*ACE

NBE

HouseholdIncome

-0.02*

0.10*

0.10*-0.01

-0.001

0.02*

-0.18*

-0.01* -0.13*

0.02

*Thisposterpresenta.onissupportedbytheMemphisResearchConsor.um(MRC)PilotGrant.

Table 1. Descriptive statistics, NSCH (2011-2012)(n=42,193)

Conclusions

Adverse Childhood Experience (ACE), Mental Health, and Childhood Obesity: A Moderated Mediation Model of Neighborhood Built Environment (NBE)*