Development and testing of a past year measure of sedentary behavior: the SIT-Q

Abstract

BackgroundMost sedentary behavior measures focus on occupational or leisure-time sitting. Our aim was to develop a comprehensive measure of adult sedentary behavior and establish its measurement properties.MethodThe SIT-Q was developed through expert review (n = 7), cognitive interviewing (n = 11) and pilot testing (n = 34). A convenience sample of 82 adults from Calgary, Alberta, Canada, participated in the measurement property study. Test-retest reliability was assessed by intraclass correlation coefficients (ICCs) comparing two administrations of the SIT-Q conducted one month apart. Convergent validity was established using Spearman’s rho, by comparing the SIT-Q estimates of sedentary behaviour with values derived from a 7-Day Activity Diary.ResultsThe SIT-Q exhibited good face validity and acceptability during pilot testing. Within the measurement property study, the ICCs for test-retest reliability ranged from 0.31 for leisure-time computer use to 0.86 for occupational sitting. Total daily sitting demonstrated substantial correlation (ICC = 0.65, 95% CI: 0.49, 0.78). In terms of convergent validity, correlations varied from 0.19 for sitting during meals to 0.76 for occupational sitting. For total daily sitting, estimates derived from the SIT-Q and 7 Day Activity Diaries were moderately correlated (ρ = 0.53, p < 0.01).ConclusionThe SIT-Q has acceptable measurement properties for use in epidemiologic studies.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2458-14-899) contains supplementary material, which is available to authorized users.

Weekly recall of sedentary time: Validity of 2 weekly self‐reported measures in undergraduate students

“…Remembering sedentary time across multiple sedentary behaviors appears to be quite difficult after just 1 day, and Study 1 demonstrated that some participants were unsure up to ±5 hours over the past‐week. In previous studies, the inaccuracy of other weekly sedentary questionnaires has been demonstrated using the Bland‐Altman analysis showing very wide limits of agreement . The results of these previous studies in combination with the results of Study 2 suggest that the most accurate measures of weekly sedentary time may come from a questionnaire that has participants report daily sedentary time across multiple sedentary behaviors sequentially, in more of a “log” format, rather than retrospectively, such as the NIGHTLY‐WEEK‐U.…”

“…It is possible that continuous reporting of past‐day sedentary time over a week could make a participant aware of how much sitting they are participating in, and as a result, begin to limit their daily sedentary time. The downside of this limitation will need to be weighed against the demonstrated inaccuracies of past‐week or typical week sedentary time questionnaires . A further limitation of the NIGHTLY‐WEEK‐U is the increased burden displaced on both the participant and the researcher.…”

“…In previous studies, the inaccuracy of other weekly sedentary questionnaires has been demonstrated using the Bland-Altman analysis showing very wide limits of agreement. 7,10,20,21 The results of these previous studies in combination with the results of Study 2 suggest that the most accurate measures of weekly sedentary time may come from a questionnaire that has participants report daily sedentary time across multiple sedentary behaviors sequentially, in more of a "log" format, rather than retrospectively, such as the NIGHTLY-WEEK-U. Furthermore, to maximally reduce recall bias, the "past-day" reporting of sedentary time should be done at the end of the day in question, and not on the following day after a participant has fallen asleep.…”

“…The downside of this limitation will need to be weighed against the demonstrated inaccuracies of pastweek or typical week sedentary time questionnaires. 7,10,20,21 A further limitation of the NIGHTLY-WEEK-U is the increased burden displaced on both the participant and the researcher. More data and time are required from the participant, and as a result, there are more data to be analyzed by the researcher.…”

“…However, revisions were made to the PAST‐U questions to reflect a past‐week format to better account for the intra‐individual variability of undergraduate students' daily activities, which may differ significantly based on their lecture and work schedules throughout the week. Unlike other past‐week sedentary time questionnaires that separate weekdays and weekend days, the PAST‐WEEK‐U asks respondents to report sedentary time in each domain on each specific day of the week. While this increases the burden on participants, it better accounts for the heterogeneous nature of an undergraduate's typical week (L. Hardy, personal communication, February 12, 2018) and it is more in line with the past‐day format of the original PAST‐U …”