Validity of self-report methods for measuring sedentary behaviour in older adults.

Abstract

Sedentary behaviour (SB) is associated with a range of negative health outcomes, but little is known about the validity of self-report methods for measuring SB in older adults. Thus, the aim was to assess the reliability and validity of two instruments for measuring SB in older adults. Cross-sectional study. 41 community-dwelling older adults (14/27 male/female, 74.5 ± 7.6 years) wore an ActivPAL(3)™ (AP) for 7 consecutive days, then completed (1) a single question (SQ) to assess sitting time on a usual weekday, weekend day and yesterday (i.e. the last day of monitoring), and (2) a computer-delivered 24-h recall (MARCA) for the last two days. Intraclass correlation (ICC) and standard error of measurement (SEM) were used to assess test-retest reliability; validity was examined using Spearman's correlation, mean bias and limits of agreement, and kappa for classifying tertiles of time in SB, with AP as the reference standard. For the SQ, the ICC ranged from 0.64 to 0.79, with SEM 1.03-1.42 h/day. ICC for the MARCA ranged from 0.72 to 0.96, with SEM 0.47-1.18 h/day. The SQ showed modest correlation with AP (r = 0.13-0.33), with mean biases of about -3.5 h/day. The MARCA showed moderate correlation with AP (r = 0.49-0.67), with mean biases of about 1.4 h/day. When categorised into tertiles, agreement was significant but fair for the SQ, and moderate for the MARCA. Both measures have acceptable reliability, but the MARCA provides more valid estimates of SB than the SQ, which underestimates SB in this group of older adults.