Article type
Year
Abstract
Background:
Prevalent sedentary behavior is a major global public health issue. Previous meta-analyses have suggested an inverse association between certain sedentary lifestyles and health outcomes; however, their potential causality has not been sufficiently examined. In this meta-analysis we aim to investigate the independent effect of prolonged sitting and TV viewing on all-cause mortality.
Methods:
We searched five databases for eligible cohort studies on such sedentary behaviors and all-cause mortality before April 2017. We employed a robust-error meta-regression method to establish the potential dose-response relationship between sedentary behaviors and all-cause mortality. The absolute effect was computed as the number needed to harm (NNTH) at four years. We conducted a series of sensitivity analyses for different adjustment patterns to see whether predefined confounders influenced the results. (Registration: CRD42017062439).
Results:
We included 22 studies in the meta-analysis. All but one study were of moderate or high quality. Compared to four hours total sitting (reference), sitting for around eight hours was the threshold for increased risk of death (8 h: relative risk (RR) 1.03, 95% confidence interval (CI) 1.00 to 1.07). For prolonged TV viewing, the risk increase started from the reference of two hours (3 h: RR 1.05, 95% CI 1.01 to 1.10). Measured by one-hourh increases, a 2.6% and 4.1% increase in RR was observed beyond the reference, corresponding to NNTH of 1123 (95% CI 730 to 2653) and 865 (95% CI 806 to 934) for total sitting and TV viewing at four-year follow-up. The absolute effect increases with age, and becomes appreciable at the age of 60 years or older. No clear mortality association was observed for occupational sitting. Sensitivity analyses showed robust estimates.
Conclusions:
Sitting for eight hours or longer, or viewing TV for more than two hours per day are independent risk factors for all-cause mortality. The absolute effect, however, is small on average, but substantial for the elderly at high baseline risk of mortality.
Patient or healthcare consumer involvement:
None.
Prevalent sedentary behavior is a major global public health issue. Previous meta-analyses have suggested an inverse association between certain sedentary lifestyles and health outcomes; however, their potential causality has not been sufficiently examined. In this meta-analysis we aim to investigate the independent effect of prolonged sitting and TV viewing on all-cause mortality.
Methods:
We searched five databases for eligible cohort studies on such sedentary behaviors and all-cause mortality before April 2017. We employed a robust-error meta-regression method to establish the potential dose-response relationship between sedentary behaviors and all-cause mortality. The absolute effect was computed as the number needed to harm (NNTH) at four years. We conducted a series of sensitivity analyses for different adjustment patterns to see whether predefined confounders influenced the results. (Registration: CRD42017062439).
Results:
We included 22 studies in the meta-analysis. All but one study were of moderate or high quality. Compared to four hours total sitting (reference), sitting for around eight hours was the threshold for increased risk of death (8 h: relative risk (RR) 1.03, 95% confidence interval (CI) 1.00 to 1.07). For prolonged TV viewing, the risk increase started from the reference of two hours (3 h: RR 1.05, 95% CI 1.01 to 1.10). Measured by one-hourh increases, a 2.6% and 4.1% increase in RR was observed beyond the reference, corresponding to NNTH of 1123 (95% CI 730 to 2653) and 865 (95% CI 806 to 934) for total sitting and TV viewing at four-year follow-up. The absolute effect increases with age, and becomes appreciable at the age of 60 years or older. No clear mortality association was observed for occupational sitting. Sensitivity analyses showed robust estimates.
Conclusions:
Sitting for eight hours or longer, or viewing TV for more than two hours per day are independent risk factors for all-cause mortality. The absolute effect, however, is small on average, but substantial for the elderly at high baseline risk of mortality.
Patient or healthcare consumer involvement:
None.