Prolonged television viewing is associated with an increased mortality risk, including from cardiovascular disease (CVD), that may be at least partially mediated by inflammatory markers, suggest results from a large UK cohort study[1]. The findings, which were published online on June 9, 2017 inAtherosclerosis, demonstrate for the first time that over 15% of the association between mortality and increased TV viewing may be explained by C-reactive protein and fibrinogen levels. Dr Mark Hamer (National Centre for Sport and Exercise Medicine, East Midlands, Loughborough, UK) and colleagues note that previous research has shown that prolonged sitting is linked to the expression of genes associated with inflammatory responses. "Fibrinogen, the strongest predictor of mortality in the present study, may also have relevance in terms of elevated risk of vascular conditions, particularly venous thrombosis," they write. "Recent experimental data demonstrated increases in plasma fibrinogen with prolonged uninterrupted sitting that was attenuated with active breaks." While the researchers note that the study design does not allow the temporal relationship between TV viewing and the biomarkers to be determined, previous analyses of the same cohort revealed a prospective association between TV viewing and changes in inflammatory markers. "Our analysis supports the main criteria for mediation, that is, the key exposure (TV viewing) was associated with biomarkers (potential mediators), the potential mediators predicted outcome (mortality), and exposure predicted outcome; this association was attenuated after adjustment for the mediators (inflammatory markers)," they write.

            As the results of previous examinations of the interaction between TV viewing and inflammatory and metabolic risk factors have been inconsistent, the researchers studied 8451 participants in the English Longitudinal Study of Ageing (ELSA) cohort study, who had an average age of 64.8 years. The participants self-reported television viewing in wave 4 of the ELSA study in 2008–2009 and were asked about their levels of physical activity. In addition, they were asked about cigarette smoking, chronic illnesses, and depressive symptoms.

            Anthropometric data was collected, and all participants had their systolic and diastolic blood pressure measured. Blood samples were analyzed for a range of inflammatory and metabolic biomarkers. Finally, the participants were linked to mortality records from the UK National Health Service registries up to February 2012. The team reports that high levels of TV viewing, defined as over 6 hours per day, were reported by 29.1% of participants. Compared with people who watched less TV, they were more likely to be female and smokers; have low levels of physical activity; and report chronic illness, depression, and disability. Individuals who reported high levels of TV viewing were also more likely to be obese than those who reported less TV watching and were more likely to have a less favorable biomarker profile.

Over a mean of 4 years of follow-up, there were 370 deaths, including 78 from CVD and 91 from cancer. ach standard-deviation (SD) increase in TV viewing was associated with a significant increase in mortality, at a fully adjusted hazard ratio of 1.17 (95% CI 1.06–1.28), which was unaffected by removing the 72 deaths that occurred in the first year of follow-up. For CVD death, the fully adjusted hazard ratio for each SD increase in TV viewing was 1.22 (85% CI 1.00–1.49), while that for cancer death was 1.16 (95% CI 0.96–1.39). When the researchers focused on biomarkers, they found that, in models mutually adjusted for all biomarkers and body-mass index, fibrinogen was the strongest predictor of mortality (P<0.001), while HDL cholesterol (P=0.087) and C-reactive protein (P=0.061) achieved marginal significance. urther analysis revealed that adding C-reactive protein and fibrinogen to the models accounted for approximately 15.7% of the association between TV viewing and mortality. This was not seen when HDL cholesterol, triglycerides, and glycated hemoglobin were added to the models.

 The data were made available through the UK Data Archive. Hamer acknowledges support from the National Institute for Health Research (NIHR) Leicester Biomedical Research Centre, which is a partnership between University Hospitals of Leicester NHS Trust, Loughborough University and the University of Leicester. Disclosures for the coauthors are listed in the paper.