Interaction of accelerometer-measured physical activity and genetic risk on cardiovascular diseases: a prospective cohort study from UK Biobank

Figure 1. Dose–response relations between physical activity and specific cardiovascular diseases. The solid line denotes HRs and the shaded area denotes 95% CIs. Sex, race, education, TDI, smoking status, alcohol consumption, diet quality, LPA (for MVPA analyses) and MVPA (for LPA analyses) were adjusted. AF, atrial fibrillation; CHD, coronary heart disease; IS, ischaemic stroke; LPA, light physical activity; MVPA, moderate to vigorous physical activity; TDI, Townsend Deprivation Index.

Figure 2. Polygenic risk score (PRS)-stratified dose–response relations. (a) PRS-stratified dose–response relations between light intensity physical activity and atrial fibrillation. (b) PRS-stratified dose–response relations between total volume of physical activity and hypertension. The solid line denotes HRs and the shaded area denotes 95% CIs. Sex, race, education, TDI, smoking status, alcohol consumption and diet quality were adjusted. TDI, Townsend Deprivation Index.

Figure 3. Attributable risks of physical activity for specific cardiovascular diseases stratified by polygenic risk score. Cumulative incidence at 70 years old was estimated by Cox regression model, fixing covariates as female, white, without a college or university degree, in the second quartile group of Townsend Deprivation Index, never smoking, drinking once or twice a week, with intermediate diet quality, and with 1800–2400 min LPA (for MVPA analyses) or 150–300 min MVPA (for LPA analyses) per week. The * denotes a statistically significant additive interaction. The error bars represent the 95% CIs. AF, atrial fibrillation; AR, attributable risk; CHD, coronary heart disease; IS, ischaemic stroke; LPA, light physical activity; MVPA, moderate to vigorous physical activity.