Leisure-Time Physical Activity on Age-Modelled Trajectories of Body Mass Index and Obesity Risk Throughout Life: Multivariable Regression and Mendelian Randomization Analyses Using Electronic Health Record Data From the CORDELIA-Catalunya Study.

内容摘要

To assess whether self-reported leisure-time physical activity, moderate-to-vigorous leisure-time physical activity, and genetically determined cardiorespiratory fitness are associated with age-related BMI trajectories and incident obesity. We pooled 14&#x2009;993 adults (30-90&#x2009;years; 52.7% women; cohorts: REGICOR-ACRISC, ILERVAS, ARTPER) with estimated LTPA (moderate-to-vigorous LTPA [MVLTPA] in REGICOR-ACRISC), genotype, and 120&#x2009;352 repeated BMI measurements from electronic health records (1990-2024). LTPA was categorised into cohort-specific quartiles; MVLTPA in 0, <&#x2009;100, <&#x2009;200, and &#x2265;&#x2009;200 METs-min/day. In one-sample Mendelian randomisation analyses, we categorised participants in quartiles of a cardiorespiratory fitness polygenic risk score derived from a large GWAS in UK Biobank. Group-dependent BMI trajectories were modelled using spline mixed-effects models. Obesity onset (first BMI &#x2265;&#x2009;30&#x2009;kg/m2) was analysed with IPW-weighted Kaplan-Meier curves and Cox models. Higher LTPA was associated with slower BMI increases in ages 30-60 (Q1: +0.120 vs. Q4: +0.075&#x2009;kg/m2&#xb7;year), slower declines in ages 70-90 (Q1: -0.143 vs. Q4: -0.123&#x2009;kg/m2&#xb7;year), and lower obesity risk (Q4 vs. Q1: HR 0.83, 95% CI 0.72-0.96). Similar trends were observed for MVLTPA. Higher genetically determined cardiorespiratory fitness showed parallel gradients (ages 30-60, Q1: +0.109 vs. Q4: +0.101&#x2009;kg/m2&#xb7;year; ages 70-90, Q1: -0.130 vs. Q4: -0.102&#x2009;kg/m2&#xb7;year) and lower obesity risk (Q4 vs. Q1: HR 0.66, 0.56-0.78). Associations were present for women and men separately, but were stronger in men. Higher LTPA and MVLTPA were associated with more favourable lifelong BMI trajectories, delayed obesity risk, and convergent support from Mendelian randomisation analyses, supporting a causal protective role of physical activity (in both sexes but stronger in men).