Agreement between self-reported and device-measured sedentary behaviour varies with cognitive function

Sedentary behavior is a modifiable lifestyle factor associated with health and well-being among older adults. Self-report and device-based measures are commonly used to assess this behavior, but the agreement between the two approaches varies considerably. To better understand the role of cognitive function in discrepancies between measures of sedentary behavior among older adults, we conducted a secondary analysis to examine the cross-sectional association between cognitive function and the agreement between self-report and device-measured sedentary behavior in 257 participants (mean age = 78.9 years) from the Wave 4 of Lothian Birth Cohort 1936. Participants completed 13 cognitive tests assessing general cognitive function, crystallized ability, visuospatial ability, verbal memory, and processing speed. They also completed a paper-and-pencil questionnaire to self-report sedentary behavior and wore activPAL accelerometers on their thighs to record total sedentary time. We found that the agreement between the two measures was weak. Poorer cognitive function (general and all four cognitive domains) was significantly associated with poorer agreement, evidenced by larger difference scores in regression models adjusted for activPAL-measured sedentary behavior, sex, age, body mass index, and self-rated health. Our results add to the previous literature by suggesting that cognitive function may affect the accuracy of self-report sedentary behavior among older adults compared with activPAL, highlighting the importance of considering cognitive function when selecting sedentary behavior measures and interpreting findings in aging populations.