BackHome
Refresh: OffView Personal Schedule
Individual Submission Summary
Share...

Direct link:

Association of Child Body Mass Index and the Obesogenic Environment in an Adoption study

Thu, April 8, 3:15 to 4:15pm EDT (3:15 to 4:15pm EDT), Virtual

Abstract

Introduction. Childhood obesity is determined by heritable and environmental factors, including the family environment (Silventoinen et al., 2010). To date, most studies of obesogenic environments for childhood obesity have been conducted in families where parents raise their biological children, thus confounding environment with child genotype or gene-environment correlation (rGE). We address this limitation by using an adopted-at-birth design (Leve et al., 2019) to examine the obesogenic family environment that adoptive parents (AP1, AP2) create for their children, independent of common heritable influences. The goals of this study were to 1) employ a novel application of latent profile analysis (LPA) to measure two key aspects of obesogenic family environments: parent characteristics and behavior management; 2) examine associations between AP1 and AP2 profiles; and 3) examine associations between APs’ profiles and childhood body mass index (BMI).

Study Population. Data came from a longitudinal, US-based sample of children adopted at birth (57% male, 55.6% Caucasian, 19.3% multiracial, 13% African American, 10.9% Latino, and < 1% Other) and their birth and non-relative adoptive parents (Leve et al., 2019).

Methods. Latent constructs of the family environment were estimated using LPAs in Mplus, using measures derived from assessments when children were 4-8 years old. Separate models were run for AP1 (97% mothers) and AP2 (96% fathers), and for Parent Characteristics and Behavioral Management. The Parent Characteristics LPA included BMI, physical activity (Krahnstoever, et al., 2005; Krahnstoever, 2004), screen time (Davison et al., 2005), permissive parenting (Arnold et al., 1993) and the Three Factor Eating Questionnaire (Karlsson et al., 2004 ). The Behavior Management LPA included subscales of the Birch Child Feeding Questionnaire (Spruijt-Metz et al., 2002), the Family Nutrition & Physical Activity scales (Ihmels et al., 2009) and AP support of child activity (Krahnstoever, 2004). Analyses adjusted for child sex, age, earlier BMI (age 2), and birth parent BMI (heritable influences).

Results. For AP1 Characteristics, LPA identified 3 classes: average activity, high activity, and low activity (Table 1 for fit statistics, Figure 1 for class descriptions). For AP2 Characteristics, LPA identified 3 classes: low activity, high activity, and low activity with high eating distress class. Neither AP1 nor AP2 Characteristics class membership predicted child BMI. The AP1 Behavior Management LPA identified 3 classes: average (values for all subscales were around the sample mean), weight concerned, and low support of physical activity. The weight concerned class was associated with higher child BMI, compared to the average class. The AP2 Behavior Management LPA yielded 3 classes: average, weight concerned, and a very low management class. AP2 Behavioral Management classes were not associated with child BMI. In terms of the relationship between AP classes, 44% of the sample for AP characteristics, and 55% for behavior management were in the largest class for both AP1 and AP2 characteristics. No AP1-AP2 correlations emerged as indicated by chi-squares across AP1 and AP2 classes.

Conclusion. Self-reported primary caregiver behavioral management was the only aspect of the obesogenic environment measured in this study that may be environmentally liked to child BMI in middle childhood.

Authors