Nucleus Accumbens Fractional Anisotropy and Children’s Body Mass Index: Moderating Role of Race and Family Income

Abstract

Background and aims: The nucleus accumbens (NAcc) functional and morphometric features mayinfluence children’s body mass index (BMI). Recent evidence, however, suggests that the functionand structure of the NAcc may have different predictive abilities for the BMI for the sub-groups ofchildren from different racial and socioeconomic status (SES) backgrounds. Using the AdolescentBrain Cognitive Development data, this study investigated racial and SES differences in the associationbetween NAcc microstructure (i.e., fractional anisotropy) and childhood BMI. Methods: This cross-sectional study included 9497 children aged 9 and 10. Data were collected from21 sites across 15 states in the United States. Then, the mixed-effects regression model was appliedfor data analysis. The predictor variable of interest was NAcc fractional anisotropy measured usingdiffusion magnetic resonance imaging (dMRI). The main outcome of interest was children’s BMI values,which were treated as a continuous variable. Covariates included gender, age, and family structure.Race (White, Black, Asian, and Other/mixed) and family income (<USD 50,000, USD 50,000-100,000,and USD100,000+) were the effect modifiers (moderators). Results: Higher average NAcc fractional anisotropy in dMRI was predictive of lower levels of theBMI, and net of covariates. However, this inverse association between the average intensity of thenormalized T2-weighted image and the BMI was stronger in children from Hispanic, low income, andlow-educated backgrounds compared to non-Hispanic, high-income, and high-educated backgrounds. Conclusion: Our findings suggested that although NAcc fractional anisotropy is linked to children’sBMI, this link is not invariant across racial and SES groups. The issue of whether or not obesogenicenvironments alter the implications of NAcc for childhood BMI needs further investigation. For diversegroups, NAcc microstructures may have different magnitudes of associations with childhood BMI.