Hazard ratios (95%CI) for any cardiovascular, non-fatal cardiovascular and coronary heart disease events per hour/day increase in television viewing in 12,608 men and women in EPIC Norfolk, 1998–2007.

Participants with self-reported or diagnosed history of stroke, myocardial infarction or cancer at baseline were excluded.Model A: adjusted for age and gender.Model B: Model A additionally adjusted for education level, smoking status, alcohol consumption, medication for hypertension, medication for dyslipidaemia, medication for depression, baseline diabetes status, family history of cardiovascular disease and sleep duration.Model C: Model B additionally adjusted for total physical activity energy expenditure (MET*hours/day).Examination of the Schoenfeld residuals and the Kaplan-Meier plots indicated that the proportional hazards assumption was reasonable for these data.