Dose-response relationship between sedentary behavior and CVD in older women

Introduction and methods

The health benefits of moderate to vigorous physical activity (MVPA) are well-established. Conversely, evidence is now emerging that shows that high levels of sedentary behavior, independently of MVPA, is a modest risk factor for CVD [1]. Studies examining sedentary time have nearly all used self-reported data. This is biased and may lead to underestimated associations [2,3]. The American Heart Association, among others, therefore called for the use of objective measures of sedentary time when evaluating relations with cardiometabolic health [4].

Use of accelerometers to quantify sedentary behavior allows characterization of accumulation patterns of sedentary time: is total sedentary time highly interrupted, or highly prolonged, thus in long, continuous sedentary bouts? The latter have been shown to be associated with impaired glucose control and with other cardiometabolic risk factors, both in experimental [5,6] and in epidemiological studies [7,8]. No prospective studies have been conducted in later life, when sedentary time is higher.

This is a prospective study evaluating accelerometer-measured sedentary time and sedentary accumulation patterns in relation to CVD events in an ethnically diverse cohort of older women with no prior history of myocardial infarction (MI) or stroke. The OPACH Study (Objectively Measured Physical Activity and Cardiovascular Health) enrolled 7058 ambulatory women of 63 years or older between 2012 and 2014. Participants wore accelerometers over their right hip 24 hours a day for 7 consecutive days. Participants also self-reported in-bed and out-of-bed times on the days they wore the accelerometer, and these periods were removed from the accelerometer readings. 6133 Women met data processing criteria for estimating average daily sedentary time (≥10 waking wear hours on ≥4 days per week). The primary outcome for this study was CVD events: first occurrence of MI, revascularization, hospitalized angina, heart failure, stroke or death due to CVD in a women without that CVD.

Main results

• 545 CVD and 137 CHD events occurred during 19350 person-years (PY) of follow-up.
• Less favorable sociodemographic and health-related characteristics were seen in women in the highest quartile of total sedentary time. Mean age, for instance, ranged from 76.3 to 80.9 years in quartiles 1 to 4.
• After adjusting for potential confounders, women in quartile 4 had 69% higher risk for CVD than women in quartile 1 (HR: 1.69, 95%CI: 1.27-2.26, P-trend= 0.001). Adjustment for potential mediators, and separately, level of MVPA, attenuated HRs, but they remained statistically significant.
• After adjusting for potential confounders, women with the most prolonged accumulation patterns had 54% higher risk for CVD (HR: 1.54, 95%CI: 1.17-2.02, P-trend=0.003) than those in quartile 1, with the most interrupted accumulation pattern. Correction for CVD risk biomarkers gave an HR of 1:36 (95%CI: 1.01-1.83, P-trend=0.10).
• Strong correlations were seen between mean bout duration and other sedentary accumulation pattern metrics, such as prolonged sedentary time: r=0.92, breaks in sedentary time: r=-0.62, usual bout duration: r=0.95 and alpha (a measure of the shape of the distribution of bout durations computed over all adherent days) r=-0.83.
• HR trajectories for sedentary time and mean bout duration were dose-responsive and linear. Plotting them suggested that 1 hour of sedentary time was associated with higher CVD risk (HR: 1.12, 95%CI: 1.05-1.19) and for 1 minute of mean bout duration HR was 1.04 (95%CI: 1.01-1.07). Associations remained statistically significant after adjusting for MVPA.
• Women with both high sedentary time and high bout duration showed a significantly elevated CVD risk after adjustment for potential confounders, than women with low sedentary time and low bout duration (HR: 1.34, 95%CI: 1.08-1.65). Adjustment took away the suggestion that the effects of two measures were additive.
• Confounder-adjusted associations of sedentary time and incident CHD yielded stronger HRs, often about two-fold of that seen for CVD.

Conclusion

References

Find this article online at Circulation