RESUMO
This study evaluated the test-retest reliability of a questionnaire assessing five domains of sedentary behaviour, and compared an overall indicator of sitting time with measures obtained by a multi-domain questionnaire, among adults living in Brazil. Data from two different studies carried out in Pelotas, Brazil, were used. Study A: evaluated 7-day reliability of a multi-domain sedentary behaviour questionnaire among 78 adults (≥18 years). Study B was a population-based study (n = 2,871 adults aged 20+ years) where a score of sedentary behaviour based on time spent sitting in five domains was compared with overall sitting time. Lin's Concordance Correlation coefficients (CCC) and Bland-Altman plots were used to assess agreement between measures. The 7-day reliability of multi-domain questionnaire did not show marked differences in the median of minutes spent in each domain (Study A). For the total score, the Lin's CCC was 0.87. When sitting time was measured by a single question, participants reported on average less sedentary time compared to the multiple-domain questionnaire (Study B). There was a slightly greater difference between measures among participants who reported higher sedentary time. Assessing sedentary behaviour in different domains would appear to be preferable to employing a general question on the total time spent sitting per day.