Files
Abstract
Substantial evidence suggests that physical inactivity is associated with cardiovascular disease, certain types of cancer, and premature death. Distinct from physical inactivity, sedentary time (characterized by very low energy expenditure while sitting/lying) is independently associated with these adverse outcomes. Measuring physical activity (PA) and sedentary behavior (SED) remains a challenge, particularly in large epidemiologic cohorts where surveys are the most common measurement method. This study aimed to: 1) evaluate the test-retest reliability and criterion validity of the Cancer Prevention Study-3 (CPS-3) PA and SED surveys and 2) estimate the mortality risk reductions associated with replacing 30 minday-1 SED with an equivalent amount of light intensity PA (LPA) or moderate-to-vigorous intensity PA (MVPA) in the Cancer Prevention Study-II Nutritional Cohort (CPS-II). The validation studies included 713 participants aged 31-72 years. Reliability was assessed by computing Spearman correlation coefficients between pre- and post-study survey responses. Validity was assessed by comparing PA and SED estimated from the CPS-3 survey with accelerometry and seven-day diaries. Reliability was acceptable or strong for all CPS-3 items and validity estimates were comparable to studies of PA/SED questionnaires with similar survey characteristics. Together, these findings suggest that the CPS-3 survey is suitable for ranking or categorizing participants according to PA or SED time. The mortality study included 101,757 participants aged 69.06.2 years. An isotemporal substitution approach to Cox proportional hazards regression was used to estimate adjusted hazard ratios and 95% confidence intervals (HR, 95% CI) for mortality associated with the substitution of 30 minday-1 SED for LPA or MVPA. Overall, 31,801 participants died during 13 years of follow-up. Among the least active participants, the replacement of 30 minday-1 SED with LPA was associated with a 14% mortality risk reduction (HR=0.86, 0.83-0.89) and replacement with MVPA was associated with a 50% mortality risk reduction (HR=0.50, 0.44-0.58). Similar associations were seen among moderately active participants (HR=0.91, 0.89-0.96 LPA replacement, HR=0.65, 0.56-0.79 MVPA replacement), but were not significant for the most active participants (HR=1.00, 0.97-1.02 LPA, HR=0.97, 0.95-1.01 MVPA). These findings suggest that replacing modest amounts of SED with even light intensity PA may improve health among less active people.