Longitudinal associations between weight indices, cognition, and mental health from childhood to early adolescence

Abstract

Childhood obesity has been associated with lower cognitive performance and worse mental health in cross-sectional studies. However, it is unclear whether these findings extend longitudinally and in what causal direction. Using data from the Adolescent Brain Cognitive Development (ABCD) Study (maximum analytical n = 6671, 48.3% girls, 42.8% non-White), we examined how body mass index (BMI) at baseline (ages 9-11) relate prospectively to changes in cognition or psychopathology across the 2 years thereafter, and vice versa. Cognitive tests included the National Institutes of Health Toolbox Cognition Battery, Little Man Task of mental rotation, and Rey Auditory Verbal Learning Test. Psychopathology was assessed using caregiver-reported Child Behavior Checklist. Linear mixed models adjusted for sociodemographic and developmental covariates indicated that lower baseline performance on most cognitive tests was associated with greater longitudinal BMI gain (eg, 1 point lower than median on Picture Vocabulary corresponded to 0.012 kg/m² [1.6%; 95% CI, 0.008 to 0.016 kg/m²] more annual BMI gain, P_FDR < .001), whereas baseline BMI was unrelated to longitudinal changes in cognition (P_FDR ≥ .12; including after considering practice effects). Greater broad-spectrum psychopathology at baseline was associated with increased BMI gain (eg, each endorsement of externalizing problems than none corresponded to 0.015 kg/m² [2.2%; 95% CI, 0.009 to 0.021 kg/m²] more annual BMI gain, P_FDR < .001) and, reciprocally, greater baseline BMI was linked specifically to more longitudinal withdrawn/depressed and depression problems (0.010 [22%; 95% CI, 0.004 to 0.016] and 0.011 [15%; 95% CI, 0.004 to 0.017] more problems annually per 1 kg/m² above median BMI, P_FDR = .003 and .008). The associations did not differ in boys vs. girls (P_FDR ≥ .40), and remained stable with waist circumference as the weight index and in subgroups of participants without weight-altering medications or common baseline psychiatric diagnoses. Our longitudinal findings expand previous cross-sectional works and highlight the importance of cognitive and mental health to children's weight development and links between weight and depression.