Physical activity and genetically driven cardiorespiratory fitness on lifelong BMI trajectories and obesity risk: multivariable regression and mendelian randomization in the CORDELIA-Catalunya Study

Background High body mass index (BMI) is a growing global health problem, yet evidence on the lifelong impact of leisure-time physical activity (LTPA) on weight is limited because most available data are observational, short term and prone to confounding. To strengthen causal inference, genetically determined cardiorespiratory fitness (CRF, a physiological proxy for habitual activity) can be used as an instrumental variable in Mendelian randomization (MR) studies. Purpose To compare lifelong BMI trajectories and incident obesity across levels of LTPA and genetically determined CRF. Methods In participants of three cohorts in the CORDELIA Study (N = 8,828), we assessed BMI trajectories from ages 30 to 90 in quartiles according to LTPA (self-reported, Minnesota LTPA questionnaire; multivariable regression) and genetically determined CRF (polygenic risk scores derived from the sex-specific GWAS ). Longitudinal BMI changes across groups were modeled using smoothed cubic spline mixed-effects models, and incident obesity (BMI ≥30 kg/m2) among initially non-obese participants was assessed using inverse probability-weighted Kaplan-Meier curves and multivariable Cox models. LTPA-based analyses were adjusted for age, sex, education, baseline hypertension, diabetes, hypercholesterolemia and smoking, and MR analyses were adjusted for the first ten ancestry-informative principal components. Results Higher LTPA was associated with slower BMI increases between ages 30-65 (+0.12, +0.094, +0.093 and +0.06 kg/m2 per year in quartile Q1, 2, 3 and 4, respectively) and a delayed onset of obesity (median age at first obesity, Q1: 68.8 years; Q4: 78.5 years; hazard ratios ranging from 0.75-0.78 in Q2-4 versus Q1). After age 65, BMI declined faster in the more inactive groups. In MR analyses, higher genetically determined CRF was also associated with slower BMI increases in ages 30-65 (+0.10, +0.10, +0.092 and +0.083 kg/m2 per year in quartile Q1, 2, 3 and 4, respectively) and a delayed onset of obesity (median age at first obesity, Q1: 75.2 years; Q4: 85.5 years; HR 0.87, 95% CI 0.75-1.00 in Q4 versus Q1). Conclusions Higher LTPA was linked to more favourable BMI trajectories across adulthood, with the most active individuals showing markedly BMI gain until age 65, later obesity onset, and slower BMI loss from age 65 than inactive peers. MR analyses yielded similar patterns, supporting a causal role of sustained physical activity in maintaining healthier BMI profiles.Lifelong and incidence LTPA curvesFor image description, please refer to the figure legend and surrounding text. Lifelong and incidence CRF-PRSs curvesFor image description, please refer to the figure legend and surrounding text.