Poster #16 - Growth in Psychopathology Across the Transition to Adolescence: Associations with Body Mass, Puberty, and Genetics

• In Event: Poster Session 15

Abstract

Objective: Prior research suggests a positive association between obesity and internalizing and externalizing psychopathology in adolescence (Tubbs et al., 2020). However, findings are limited to small patient samples, wide age ranges, heterogeneity in measurement, and lack of consideration of sex, pubertal stage, direction of effects, and genetic contributions to the association. The present study fills these gaps by testing relations between the full-range of body mass index (BMI) with growth in depression, anxiety, and externalizing psychopathology across the transition to adolescence in a large, diverse, statewide sample of twins.

Methods: Participants included 1188 twin children (51.2% female; 59.4% non-Hispanic White, 28.2% Hispanic; 30.8% identical, 37.8% same-sex fraternal, 31.4% opposite-sex fraternal; 46.1% below middle class) and their parents participating in the longitudinal Arizona Twin Project (Lemery-Chalfant et al., 2019). Twins were assessed annually at mean ages 8.42, 9.71, 10.88, and 11.76. BMI was computed from height and weight measurements obtained at home visits at each wave, and primary caregivers (95% mothers) reported Depression, Anxiety, and Externalizing problems on the reliable and valid MacArthur Health and Behavior Questionnaire (Essex et al., 2002). Mothers and youth reported pubertal stage on the Pubertal Development Scale (Petersen et al., 1988) at each wave.

Results: Using a variable centered approach, latent growth models were fit to Depression (-2LL: 767.27, BIC: -1472.84, with a positive slope and significant variance around the intercept and slope), Anxiety (-2LL: 844.19, BIC: -1626.67, with a negative slope and significant variance around the intercept and slope), and Externalizing (-2LL: 883.55, BIC: -1705.40, with a non-significant slope and significant variance around the intercept and slope).
BMI at age 8 was highly heritable, with 93% of the variance due to genetic differences. BMI predicted higher anxiety intercepts (b=.011, p=.001) and a more negative slope (b=-.004, p<.001), but was unrelated to depression and externalizing growth parameters. Puberty was positively related to depression at age 8 and externalizing at age 10, and female sex predicted higher anxiety and lower externalizing intercepts.
A person-centered approach added nuance to the variable-centered findings. Table 1 provides the fit statistics (selected models are highlighted), and Figure 1 depicts the classes uncovered for psychopathology using latent class growth analysis and growth mixture modeling. BMI positively predicted membership in the moderate-increasing depression class compared to the low-stable class (logit=.096, O.R.=1.10, p=.022), and membership in the externalizing high-decreasing class compared to the low-stable (logit=.119, O.R.=1.13, p=.015) or moderate-increasing (logit=.301, O.R.=1.35, p=.007) groups. Puberty was unrelated to class membership, but males were more likely to be in the high-decreasing externalizing group than females.

Conclusions: Associations between BMI and developmental psychopathology are complex, with variable-centered approaches revealing a positive association between BMI and a higher initial level and a steeper decrease in anxiety, but person-centered approaches uncovering latent relations with increasing depression and decreasing externalizing. Findings on direction of effects and genetic influences on BMI-psychopathology associations will also be discussed. With these more precise associations, future research should consider the role of sleep and inflammation as genetically-influenced mechanisms linking body mass to developmental psychopathology.

Authors

• Kathryn Lemery-Chalfant, Arizona State University
  Presenting Author

• Jeri Sasser, Arizona State University

• Sierra C. Clifford

• Leah D Doane, Arizona State University

• Mary Davis, Arizona State University