Abstract
Background/ObjectivesBody mass index (BMI) is widely used to assess obesity-related disease risk but does not capture important aspects of fat distribution and ectopic fat deposition. Here we assess age-related differences in magnetic resonance imaging (MRI)-derived body composition measures and their relevance for cardiometabolic disease risk.Subjects/MethodsWe analysed 36,831 UK Biobank participants (45-82 years), with MRI-derived measures, including 1,688 participants with longitudinal measures. Nine MRI-derived traits and five cardiometabolic outcomes were analysed: type 2 diabetes (T2DM), hypertension, hyperlipidaemia, myocardial infarction, and cerebral infarction. Cox regression was used to estimate the association between adiposity traits and clinical outcomes and stepwise Cox regression to determine the relative importance of the variables. Significant associations were validated using univariable and multivariable Mendelian randomization (MR).ResultsAt the initial imaging session, several MRI-derived traits, including visceral adipose tissue (VAT) volume and pancreatic and thigh muscle fat infiltration, were higher with age at assessment and increased within participants between imaging timepoints in the longitudinal sub-cohort, despite similar BMI. In contrast, liver and pancreas volume were lower in older participants. All MRI-derived traits were associated with at least one clinical outcome. However, after adjustment for multiple testing and for BMI, only T2DM and hyperlipidemia remained significant. Including MRI-derived traits alongside BMI improved risk prediction, particularly in males. VAT showed the strongest association and accounted for a substantial proportion of the T2DM risk traditionally attributed to BMI in both females and males, as supported by multivariable MR.ConclusionsAgeing is associated with progressive, depot-specific redistribution of body fat with increasing accumulation in metabolically harmful depots later in life. MRI-derived adiposity measures provide important insight into these risk-relevant patterns. Future clinical translation may depend on more scalable tools or biological markers that capture visceral and ectopic fat accumulation and improve cardiometabolic risk assessment beyond BMI.</p>