Reliability and validity of time-use surveys in assessing 24-hour movement behaviors in adults

Abstract

Background: Time-use surveys are considered a valid alternative for assessing physical activity (PA) and sedentary behavior (SB). The International Classification of Activities for Time-Use Statistics (ICATUS) has been widely adopted, as a standardized framework for categorizing time-use data. A classification system has been developed to classify ICATUS-based activities into sleep, SB, light PA (LPA), and moderate-to-vigorous PA (MVPA). This study aimed to evaluate the reliability and validity of ICATUS-based time-use classifications. Methods: Participants aged 18–59 years were recruited from five organizations located in the Bangkok metropolitian area (n = 220). The study was conducted from September to October 2022. Participants wore an accelerometer for ten consecutive days and completed two-day time-use diaries. The intraclass correlation coefficient (ICC) was calculated to assess test-retest reliability between the first and second entries of time-use surveys, as well as for accelerometer data. Validity was assessed by comparing the two time-use surveys with corresponding accelerometer data using Spearman correlations. Results: Test-retest reliability showed strong absolute agreement in the average time-use estimates for sleep and SB, with ICCs of 0.80 and 0.83, respectively. Moderate agreement was observed for LPA (ICC = 0.71) and MVPA (ICC = 0.51). Moderate validity correlations were found for SB, while LPA showed weak correlations, and MVPA results were inconsistent. Conclusion: ICATUS-based time-use data demonstrated strong reliability and moderate validity for SB, and moderate reliability and low validity for PA in working adults. The classification system appears to be a verified tool, supporting its applicability of time-use data, particularly in developing countries.