Fig. 2.

Reallocating sedentary time into physical activity for coronary heart disease. Notes: Cox regression models using age (at the start of the accelerometry sub‐study) as the underlying timescale (not as a confounder) were adjusted for sex, smoking status, alcohol consumption, diet, Townsend Deprivation Index, use of antihypertensive medication, use of blood‐glucose lowering medication, use of cholesterol‐lowering medication, genotype array type, and first 10 principal components of genetic ancestry, with mutual adjustment for the 3 ILR terms. Light PA, light physical activity; MVPA, moderate‐to‐vigorous physical activity; ST, sedentary time.