Kavli Affiliate: Anders Dale
| Authors: Alison Rigby, Diliana Pecheva, Pravesh Parekh, Diana M. Smith, Ashley Becker, Janosch Linkersdoerfer, Richard Watts, Robert Loughnan, Donald J. Hagler, Kyung E. Rhee, Carolina Makowski, Anders M. Dale and Terry L. Jernigan
| Summary:
Abstract
Importance Adolescent body mass index (BMI) is linked to brain microstructural alterations, but it is not known if this association is uniform across the BMI range.
Objective To characterize the shape of the association between adolescent BMI and brain microstructure across the BMI range.
Design Baseline, and 2-, 4-, and 6-year follow-up brain imaging and BMI from the Adolescent Brain Cognitive Development℠ Study, analyzed between June 2025 and June 2026.
Setting Multicenter, population-based study including 21 sites distributed across the US.
Participants 10,479 adolescents (5,078 [48.5%] female; aged 9-18 years) from this population-based cohort study were included after quality control of complete data.
Exposures Centers for Disease Control BMI extended percentiles and percent from median BMI.
Main Outcome and Measure The primary outcome was the restricted normalized isotropic (RNI) signal fraction derived from diffusion imaging which indexes water confined within spherical cellular structures and is sensitive to microstructural properties including cellularity.
Results A total of 10,479 participants (5,078 [48.5%] female; aged 9-18 years) contributed 22,049 imaging observations across four timepoints. Higher BMI percentile was associated with greater RNI in bilateral reward- and appetite-related subcortical gray matter and white matter tracts, with effects varying spatially within structures, and with generally greater magnitude in males. Across these structures, the association increased steeply above the 80th percentile. This acceleration reflected compression of the percentile scale at high BMI; when BMI was reparameterized as percent from median, the association with RNI was approximately linear.
Conclusions and Relevance Our results suggest that BMI is associated with brain microstructure in a graded manner within white matter tracts and subcortical gray matter related to reward, executive function, and appetitive control. This indicates that the effect of BMI on microstructure is not confined to weight extremes and should be accounted for when modeling brain-body-behavior relationships during adolescence.
Question Does the association between adolescent body mass index (BMI) and brain microstructure vary across the BMI ranges?
Findings In this longitudinal study, adolescent BMI percentile and percent from median BMI were significantly associated with brain microstructure in subcortical structures and white matter tracts. The rate of change above the 80th BMI percentile was steep, although this reflected a compression of effects in the upper percentile range; percent from median revealed an approximately linear relationship of BMI on brain microstructure.
Meaning BMI is linearly associated with brain microstructure in adolescence, irrespective of clinically defined weight classes.