Validity and Reliability of Two Self-report Measures to Assess Sedentary Behavior in Older Adults

Purpose To examine the test-retest reliability and criterion-related validity of the Yale Physical Activity Survey for Older Adults (YPAS) and Community Health Activities Model Program for Seniors (CHAMPS) survey to assess sedentary behavior (SB) in older adults.

Methods Adults aged 65 years and older were recruited to participate in a larger study comparing the validity of various objective and self-report measures of physical activity. Participants completed the YPAS and CHAMPS before and after a 10 day period during which they wore an ActiGraph accelerometer (ACC). ACC derived average daily SB time (defined as <100 counts per min) was used as the criterion measure against which to validate the questionnaires. Time spent sitting was reported using a single question on the YPAS (<3, 3-5.9, 6-7.9, 8+ hrs/day), and by a combination of nine questions from the CHAMPS for activities ≤1.5 METs. Test-retest reliability was examined by calculating Kappa statistics for the YPAS and intraclass correlation coefficients (ICC) for the CHAMPS. Criterion-related validity was assessed by: 1) calculating percent agreement between YPAS response categories and similarly derived categories from the ACC, and by examining the trend for ACC sitting time across YPAS category, and 2) with Pearson correlations and Bland-Altman analysis for the CHAMPS with ACC.

Results Data were analyzed for 58 adults (75 ± 7 years; 79% female) with complete YPAS, CMAMPS, and ACC data. Average daily SB time (ACC) was 9.2 ± 1.3 hrs/day (mean ± SD). YPAS reliability was fair with Kappa = 0.27 (p <0.001). Reliability of the CHAMPS was also deemed fair with an ICC of 0.47 (p <0.001). YPAS validity was poor with 8.3% agreement with the ACC. However, there was a linear trend (p <0.01) in ACC-derived SB time across YPAS response categories. CHAMPS validity was also poor as evidenced by a small, non-significant correlation with the ACC (r = 0.12, p = 0.36) and a mean underreporting of 5.21 hrs/day, 95% CI (2.2, 8.3).

Conclusions Neither the YPAS nor CHAMPS proved to be acceptably valid or reliable to estimate SB time in this group of older adults. SB time was drastically lower on both measures compared to objectively measured SB; however, the YPAS displays the potential to accurately rank individuals according to time spent in SB.