Cross-sectional study of diet, physical activity, television viewing and sleep duration in 233 110 adults from the UK Biobank; the behavioural phenotype of cardiovascular disease and type 2 diabetes

OBJECTIVES: Simultaneously define diet, physical activity, television (TV) viewing, and sleep duration across cardiometabolic disease groups, and investigate clustering of non-diet lifestyle behaviours. DESIGN: Cross-sectional observational study. SETTING: 22 UK Biobank assessment centres across the UK. PARTICIPANTS: 502 664 adults aged 37–63 years old, 54% women. 4 groups were defined based on disease status; ‘No disease’ (n=103 993), ‘cardiovascular disease’ (CVD n=113 469), ‘Type 2 diabetes without CVD’ (n=4074) and ‘Type 2 diabetes + CVD’ (n=11 574). MAIN OUTCOMES: Diet, physical activity, TV viewing and sleep duration. RESULTS: People with ‘CVD’ report low levels of physical activity (<918 MET min/week, OR (95% CI) 1.23 (1.20 to 1.25)), high levels of TV viewing (>3 h/day; 1.42 (1.39 to 1.45)), and poor sleep duration (<7, >8 h/night; 1.37 (1.34 to 1.39)) relative to people without disease. People with ‘Type 2 diabetes + CVD’ were more likely to report low physical activity (1.71 (1.64 to 1.78)), high levels of TV viewing (1.92 (1.85 to 1.99)) and poor sleep duration (1.52 (1.46 to1.58)) relative to people without disease. Non-diet behaviours were clustered, with people with ‘CVD’ or ‘Type 2 diabetes + CVD’ more likely to report simultaneous low physical activity, high TV viewing and poor sleep duration than those without disease (2.15 (2.03 to 2.28) and 3.29 (3.02 to 3.58), respectively). By contrast, 3 in 4 adults with ‘Type 2 diabetes’, and 2 in 4 adults with ‘CVD’ have changed their diet in the past 5 years, compared with only 1 in 4 in the ‘No disease’ group. Models were adjusted for gender, age, body mass index, Townsend Deprivation Index, ethnicity, alcohol intake, smoking and meeting fruit/vegetable guidelines. CONCLUSIONS: Low physical activity, high TV and poor sleep duration are prominent unaddressed high-risk characteristics of both CVD and type 2 diabetes, and are likely to be clustered together.