Modeling Child Body Mass Index Z-score Variation from Age 2 to 12 years Associated with Heritable and Rearing Environmental Factors Using Multilevel Model with Heterogeneous Variance

Many children follow predictable growth curves for BMI, showing consistent z-scores over time. However, some children show high BMI instability. The current study explored whether variability in children’s BMI-z scores over... [ view full abstract ]

Many children follow predictable growth curves for BMI, showing consistent z-scores over time. However, some children show high BMI instability. The current study explored whether variability in children’s BMI-z scores over time reflects the impact of heritable and environmental factors that may constrain or predict instability in BMI z-scores over time. Consistent with current obesity theories, we hypothesized that children with higher heritable risk for obesity may demonstrate higher and more stable BMIs relative to peers who are not at risk. We also hypothesized that parent behaviors that encourage children to use external cues to regulate eating would lead to less stable BMI z-scores over time: children who are not taught to rely on internal cues of hunger and satiety may have more difficulty maintaining a stable BMI z-score.

Participants were drawn data from the Early Growth and Development Study, a prospective adoption study that includes 561 adoptees who were placed with nongenetically related adoptive families close to birth. Child BMI was obtained via adoptive parents’ (AP) reports and extracted from medical records. Child BMI z-score adjusted for sex and age was calculated based on the US CDC 2000 norms (Kuczmarski et al., 2002). On average, there were 9 BMI assessments per child (range = 1 to 51), and assessments spanned from age 2 to 12 years. Birth mother and birth father (birth parents: BP) BMIs were self-reported 6 times from post-partum 5 months to 9 years, and used to compute birth mother and birth father’s average BMIs. BP average BMIs were used to index heritable influences on BMI. At child age 4.5 years, adoptive mothers (AMs) reported on how much they controlled their children’s eating (food restriction, pressure to eat) vs. permitted them to regulate their own eating. Analyses were conducted using a multilevel model with heterogeneous variance, in which within-person variances were modeled as a function of heritable and environmental factors. Both heritable and rearing environmental factors were associated with mean levels and variability in child BMI z-score. Consistent with hypotheses, children whose BP have higher BMIs also have higher BMI z-scores (birth mother: ß = .03, p < .01, and birth father: ß = .03, p = .04) and more stable BMI z-scores from early childhood to early adolescence (birth mother: χ² = 13.18, df = 1, p < .01; birth father: χ² = 3.12, df = 1, p = .07). AMs’ external regulation of eating was related to more variability in child BMI z-scores over time. Although food restriction was not associated with mean levels of BMI z-scores, it was positively associated with variability in child BMI z-scores (χ² = 17.11, df = 1, p < .01); pressure to eat was negatively associated with mean levels of BMI z-scores (ß = -.28, p < .01) and positively associated with variability in BMI z-scores (χ² = 11.98, df = 1, p < .01). Therefore, more stable BMIs were observed for children at heritable risk for higher BMIs and for parents who do not externally regulate children’s eating.