Background Although substantial evidence has demonstrated the impact of obesity on brain structure and cognition, the heterogeneity of adiposity-particularly in terms of fat distribution patterns-and its differential neurologic effects remain poorly understood. Purpose To identify body fat distribution patterns with MRI and latent profile analysis (LPA) and their associations with brain structure measurements, cognition, and neurologic diseases. Materials and Methods This secondary analysis used prospective data from the UK Biobank, including health records and MRI scans of the brain, heart, and abdomen. Fat distribution profiles were classified using LPA based on eight body mass index (BMI)-adjusted MRI-derived fat quantification metrics. Differences in brain volume, white matter properties, cognition, and the risk of neurologic disorders were analyzed across profiles and relative to a benchmark lean profile; analyses were stratified by sex. Group differences were examined using analysis of covariance (ANCOVA) or rank-based ANCOVA. Results Among 25 997 participants (mean age, 55 years ± 7.4 SD; 13 536 female participants), LPA identified six profiles of body fat distribution in both sexes. Four high-adiposity patterns were identified, including the pancreatic-predominant profile (profile 1), with elevated proton density fat fraction (mean BMI-adjusted z score, 2.38 ± 0.74 for male participants and 3.01 ± 1.08 for female participants; P P d, -0.63 for male participants and -0.58 for female participants; profile 3: Cohen d, -0.56 for male participants and -0.12 for female participants; P d, 0.47 for male participants and 0.42 for female participants; profile 3: Cohen d, 0.42 for male participants and 0.20 for female participants; P d, 0.25 for male participants with profile 1 and 0.32 for male participants with profile 3; P Supplemental material is available for this article.
Yu et al. (Thu,) studied this question.