Body Mass or Belly Size? Comparing Measures of Body Fat Composition as Predictors of Executive Functioning Among Racially Diverse Middle- to Older-Aged Adults

Alexandria Bartlett, Georgia State University, Atlanta, United States
Kharine Jean, Georgia State University, Atlanta, United States
Michael Lee, Georgia State University, Atlanta, United States
Asia Hemphill, Georgia State University, Atlanta, United States
Jannah Lherisson, University of Central Florida, Orlando, United States
Vonetta Dotson, Georgia State University, Atlanta, United States


Body mass index (BMI) has become the most widely used metric of adiposity, often used to indicate disease risk. Higher BMIs are associated with greater cardiovascular burden and reduced cognitive performance, particularly in executive functioning. However, BMI may not generalize as the best measure of fat distribution for all given the documented differences in lean mass, height, extremity length, and bone density across ethnoracial groups. Abdominal circumference, an alternative adiposity measure reflecting visceral fat, may be a more universal signal of overall health than BMI given its similar cognitive and vascular associations that are less impacted by the ethnically variable factors calculated into BMI. However, the literature evaluating the interrelationships between adiposity measures and executive functioning across race is sparse. The goal of this study, therefore, was to compare BMI and waist circumference as predictors of executive functioning in middle-aged and older adults, and to determine if these relationships differ between ethnoracial groups. 

Participants and Methods:

1,164 Non-Hispanic White (NHW), 1,164 Hispanic, and 700 Black adults aged 50 and older (mean age = 65.21 ± 8.67; 62.4% female) from the Health and Aging Brain Study: Health Disparities (HABS-HD) underwent neuropsychological evaluation and physical exams including waist measurements and BMI. A principal components analysis was used to derive linear composite scores representing executive performance from four executive measures: Trail Making Test Part B completion time, Digit Span Backwards total score, FAS total score, and Animal Naming total score. Moderation analyses were performed using BMI as the predictor and the newly derived executive functioning scores as the outcome variable with racial identity as a moderator and age and education as covariates. Moderation analyses were then repeated with abdominal circumference as the predictor. 


The effects of BMI and abdominal circumference on executive functioning were significant (p = .045; p < .001, respectively), demonstrating worse executive performance with increasing BMI and waist size overall. With age as a covariate, race significantly moderated each relationship, interestingly revealing inverse associations between racial categories. In NHWs, greater BMI was associated with worse executive performance (b = -.028, p < .001). However, in Black and Hispanic adults, higher BMI was associated with better performance (b = .032, p < .001; b = .022, p < .001, respectively). Similarly, larger waist sizes were associated with poorer executive performance in NHWs (b = -.025, p < .001) but better performance for Blacks (b = .020, p = .006) and Hispanics (b = .023, p < .001). Race no longer moderated the relationships after controlling for education.


While these data support previous findings that BMI remains a useful marker of cognitive risk, our results reveal that the direction of this relationship may differ across ethnoracial lines. This difference may partly account for the mixed literature on BMI and cognitive decline in older adults, suggesting that higher BMI may be protective in Black and Hispanic adults but not in NHWs. Future studies are needed to investigate the associations between fat distribution and cognitive functioning in racially diverse samples.

Category: Executive Functions/Frontal Lobes

Keyword 1: cognitive functioning
Keyword 2: diversity
Keyword 3: aging (normal)