Katherine E. DeVivo, Scott Jamieson, Chih-Hsiang Yang, Christine Pellegrini. University of South Carolina, Columbia, SC.

Background: Excess sedentary behavior, defined as waking behavior in a sitting/reclining/lying posture ≤1.5 metabolic equivalents, is associated with risk for types of cancers, cardiovascular disease, and type 2 diabetes. Due to the habitual nature of sedentary behavior, it is difficult to obtain an accurate estimate of sedentary time (ST), particularly among those who spend the majority of waking hours sitting. The purpose of this study was to compare differences and level of agreement between adults self-reported ST and objectively measured ST. Methods: As part of a sedentary reduction intervention for adults following knee replacement (TKR), participants ST was self-reported (screening item and SIT-Q) and objectively measured (activPAL). The screening item asked hours per day (h/d) that participants spend sitting. The SIT-Q-7d is a self-administered questionnaire that quantifies ST in the last 7 days by assessing behavior across five domains of daily life activities and calculates weekday and weekend SIT-Q scores (h/d). For 7 days, participants wore an activPAL on their thigh, the gold standard for assessing sedentary behavior, to measure ST (h/d). Paired samples t-test compared mean differences in ST between the self-reported measures and activPAL. Interclass correlation coefficient (ICC) measured the extent of agreement between self-reported and objectively measured ST.Results: Participants (n=32) were 62.5% female, 90.6% White, 64.2±9.5 yrs, 32.8±6.3 kg/m2 BMI. Self-reported ST via screener (8.9±2.1 h/d) was significantly lower (p<0.001) than activPAL ST (12.1±2.3 h/d). There were no differences (p=0.976, p=0.083) in ST between SIT-Q weekday (12.1±3.4 h/d) or weekend (10.7±3.4 h/d) compared to activPAL. ICC between the activPAL and screener was 0.355, 95% CI 0.008 - 0.626, between the activPAL and SIT-Q weekday was 0.075, 95% CI -0.281 - 0.413, and between the activPAL and SIT-Q weekend was -0.068, 95% CI -0.407 - 0.287. Conclusion: Overall, the level of agreement in ST was low between the self-reported and objective measures in adults after TKR. The one-item screener question resulted in significantly lower estimates of ST than that measured by the activPAL. Caution is warranted on future use of this one item to obtain an accurate measure of ST in this population. If objective measures are not available, the SIT-Q would provide a closer estimate of ST. Funding: This research was supported by NIAMS R21AR074780.

This document is currently not available here.