Sedentary Behavior, Physical Activity, and All-Cause Mortality: Dose-Response and Intensity Weighted Time-Use Meta-analysis

OBJECTIVES: Previous studies have placed those with excessive sedentary behavior at increased risk of all-cause mortality. There is evidence of interdependency of sedentary behavior with physical activity, and its elucidation will have implications for guidelines and practice. This study investigated if sedentary behavior-related mortality risk can be offset by moderate- to vigorous-intensity physical activity (MVPA) considered in a time-use fashion.

DESIGN: PubMed was searched (from its inception till May 2018) for studies or meta-analyses that used data harmonized for MVPA. Of the 17 data-custodians located, 7 provided data on sitting time or TV viewing time, or both. A dose-response meta-analysis modeling log relative risks of all-cause mortality against uncompensated sedentary behavior metabolic equivalent hours (USMh) was run using the robust error meta-regression method. (Registration: CRD42017062439) SETTING: Individual subject data held by data custodians on this topic.

PARTICIPANTS: General adults.

MEASUREMENTS: Sedentary time, MVPA.

RESULTS: Five harmonized cohorts of sitting time (258,688 participants) and 4 of TV viewing time (156,593 participants) demonstrated that sedentary behavior was significantly associated with mortality, but this risk was attenuated with increasing energy expenditure through MVPA modeled in a time-use fashion. The average increment in mortality per USMh spent on sitting was 1% [relative risk (RR) 1.01, 95% confidence interval (CI) 1.00, 1.02; P = .01] and that per USMh spent on TV viewing was 7% (RR 1.07, 95% CI 1.04, 1.10; P < .001). The thresholds for risk started at 7 USMh for sitting and 3 USMh for TV viewing.

CONCLUSIONS/IMPLICATIONS: Our findings suggest that overall daily sitting time energy expenditure of 7 MET-hours (or TV viewing of 3 MET-hours) in excess of that expended on MVPA is independently related to all-cause mortality. These findings support the view that sitting is strongly influenced by consideration of concurrent MVPA in its impact on adverse health consequences and that the USMh is a more practical metric of sedentary behavior.