Abstract
AIMS: Sedentary behaviour (particularly television (TV) viewing) is thought to be a risk factor for cardiovascular disease. We employed a negative control outcome to explore whether the association between TV viewing and heart disease mortality is explained by confounding.
METHODS: The sample was drawn from the UK Biobank study and comprised 479 658 participants (aged 56.5±8.0 years; 45.7% men) followed up over a mean of 10.4 years. TV viewing was measured from self-report.
RESULTS: There were 1437 ischaemic heart disease (IHD) deaths, and 214 accidental deaths (employed as the negative control outcome). TV viewing was related to the following confounding variables: age, smoking, alcohol, diet, obesity, physical inactivity, cardiovascular disease and education. The confounding structures were similar for both outcomes. TV viewing (per hour/d) was associated with IHD (hazard ratio (HR)=1.30, 95% CI, 1.27 to 1.33) and accidental death (HR=1.15, 95% CI, 1.07 to 1.24) in unadjusted models. Associations were attenuated for both outcomes and were considerably converged after adjustment for confounders; IHD (HR=1.09, 95% CI, 1.06 to 1.12) and accidental death (HR=1.06, 95% CI, 0.98 to 1.15).
CONCLUSION: The pattern of results for TV with an implausible outcome mirrored that of IHD, suggesting that observed associations between TV and heart disease are likely to be driven by confounding.