Joint non-linear dose–response associations of device-measured physical activity and cardiorespiratory fitness with cardiovascular disease: a cohort and Mendelian randomisation study

Abstract

Objectives: To characterise the non-linear joint dose–response relationship of accelerometer-measured moderate-to-vigorous physical activity (MVPA) and cardiorespiratory fitness (CRF, estimated as maximal oxygen uptake (VO₂max)) with incident cardiovascular disease (CVD), and to assess causal consistency using Mendelian randomisation (MR). Methods: We conducted a cohort study in the UK Biobank using accelerometer data linked to hospital and death registries. A Cox generalised additive model characterised the joint MVPA–CRF association with incident CVD (atrial fibrillation, myocardial infarction, heart failure (HF) and stroke), adjusting for confounders. We derived a fitness-stratified matrix quantifying the weekly MVPA minutes associated with prespecified relative hazard reductions. Complementary two-sample MR analyses leveraged genome-wide association study summary statistics for device-measured physical activity (PA) traits and CRF to assess potential causal effects on cardiovascular outcomes. Results: Among 17 088 participants, 1233 incident CVD events occurred over a median follow-up of 7.85 years (IQR, 7.39–8.27). A significant non-linear interaction between MVPA and CRF was observed (p<0.001). Meeting the 150 min/week guideline yielded a modest ~8%–9% risk reduction across fitness levels, whereas achieving a >30% risk reduction required threefold to fourfold higher volumes (~560–610 min/week). Residual analysis indicated that fitness beyond what MVPA and covariates predicted retained a modest protective association with CVD risk (HR, 0.98 per 1 mL/kg/min; 95% CI 0.97 to 0.99; p<0.001). In MR analyses, genetically proxied higher CRF was associated with lower HF risk (OR, 0.79; 95% CI 0.63 to 0.99), whereas genetic evidence for PA traits was weaker and less consistent. Conclusion: Current MVPA guidelines provide a universal but modest safety margin, whereas optimal cardiovascular protection may require substantially higher activity volumes. The fitness-stratified prescription matrix offers quantitative behavioural targets, and genetic findings reinforce the independent importance of CRF in cardiovascular risk reduction.