Associations of discretionary screen time with mortality, cardiovascular disease and cancer are attenuated by strength, fitness and physical activity: findings from the UK Biobank study.pdf (1.82 MB)
Download file

Associations of discretionary screen time with mortality, cardiovascular disease and cancer are attenuated by strength, fitness and physical activity: findings from the UK Biobank study.

Download (1.82 MB)
journal contribution
posted on 31.07.2018, 08:36 by Carlos A. Celis-Morales, Donald M. Lyall, Lewis Steell, Stuart R. Gray, Stamatina Iliodromiti, Jana Anderson, Daniel F. Mackay, Paul Welsh, TThomas Yates, Jill P. Pell, Naveed Sattar, Jason M. R. Gill
BACKGROUND: Discretionary screen time (time spent viewing a television or computer screen during leisure time) is an important contributor to total sedentary behaviour, which is associated with increased risk of mortality and cardiovascular disease (CVD). The aim of this study was to determine whether the associations of screen time with cardiovascular disease and all-cause mortality were modified by levels of cardiorespiratory fitness, grip strength or physical activity. METHODS: In total, 390,089 participants (54% women) from the UK Biobank were included in this study. All-cause mortality, CVD and cancer incidence and mortality were the main outcomes. Discretionary television (TV) viewing, personal computer (PC) screen time and overall screen time (TV + PC time) were the exposure variables. Grip strength, fitness and physical activity were treated as potential effect modifiers. RESULTS: Altogether, 7420 participants died, and there were 22,210 CVD events, over a median of 5.0 years follow-up (interquartile range 4.3 to 5.7; after exclusion of the first 2 years from baseline in the landmark analysis). All discretionary screen-time exposures were significantly associated with all health outcomes. The associations of overall discretionary screen time with all-cause mortality and incidence of CVD and cancer were strongest amongst participants in the lowest tertile for grip strength (all-cause mortality hazard ratio per 2-h increase in screen time (1.31 [95% confidence interval: 1.22-1.43], p < 0.0001; CVD 1.21 [1.13-1.30], p = 0.0001; cancer incidence 1.14 [1.10-1.19], p < 0.0001) and weakest amongst those in the highest grip-strength tertile (all-cause mortality 1.04 [0.95-1.14], p = 0.198; CVD 1.05 [0.99-1.11], p = 0.070; cancer 0.98 [0.93-1.05], p = 0.771). Similar trends were found for fitness (lowest fitness tertile: all-cause mortality 1.23 [1.13-1.34], p = 0.002 and CVD 1.10 [1.02-1.22], p = 0.010; highest fitness tertile: all-cause mortality 1.12 [0.96-1.28], p = 0.848 and CVD 1.01 [0.96-1.07], p = 0.570). Similar findings were found for physical activity for all-cause mortality and cancer incidence. CONCLUSIONS: The associations between discretionary screen time and adverse health outcomes were strongest in those with low grip strength, fitness and physical activity and markedly attenuated in those with the highest levels of grip strength, fitness and physical activity. Thus, if these associations are causal, the greatest benefits from health promotion interventions to reduce discretionary screen time may be seen in those with low levels of strength, fitness and physical activity.


The UK Biobank was supported by the Wellcome Trust, Medical Research Council, Department of Health, Scottish Government and the Northwest Regional Development Agency. It has also had funding from the Welsh Assembly Government and the British Heart Foundation.



BMC Medicine, 2018, 16:77

Author affiliation

/Organisation/COLLEGE OF LIFE SCIENCES/School of Medicine/Diabetes Research Centre


VoR (Version of Record)

Published in

BMC Medicine


BioMed Central



Acceptance date


Copyright date


Available date


Publisher version