Objectives We examined the dose–response relationship between physical activity (PA) and incidence of cardiovascular disease (CVD) risk factors in adults in Taiwan.

Methods This study included 1 98 919 participants, aged 18–97 years, free of CVD, cancer and diabetes at baseline (1997–2013), who were followed until 2016. At baseline, participants were classified into five PA levels: inactive’ (0 metabolic equivalent of task (MET)-h/week), ‘lower insufficiently active’ (0.1–3.75 MET-h/week), ‘upper insufficiently active’ (3.75–7.49 MET-h/week), ‘active’ (7.5–14.99 MET-h/week) and ‘highly active’ (≥15 MET-h/week]. CVD risk factors were assessed at baseline and at follow-up by physical examination and laboratory tests. Analyses were performed with Cox regression and adjusted for the main confounders.

Results During a mean follow-up of 6.0±4.5 years (range 0.5–19 years), 20 447 individuals developed obesity, 19 619 hypertension, 21 592 hypercholesterolaemia, 14 164 atherogenic dyslipidaemia, 24 275 metabolic syndrome and 8548 type 2 diabetes. Compared with inactive participants, those in the upper insufficiently active (but not active) category had a lower risk of obesity (HR 0.92; 95% CI 0.88 to 0.95), atherogenic dyslipidaemia (0.96; 0.90 to 0.99), metabolic syndrome (0.95; 0.92 to 0.99) and type 2 diabetes (0.91; 0.86 to 0.97). Only highly active individuals showed a lower incidence of CVD risk factors than their upper insufficiently active counterparts.

Conclusion Compared with being inactive, doing half the recommended amount of PA is associated with a lower incidence of several common biological CVD risk factors. Given these benefits, half the recommended amount of PA is an evidence based target for inactive adults.