Criterion validity of a single-item question for assessment of daily breaks in sedentary time in adults.

Abstract

As the implementation of sensor-based assessment for sedentary time (ST) and physical activity (PA) has practical limitations when applied on a large-scale, most studies rely on subjective data. We aimed to examine the criterion validity of a single-item question to assess daily breaks in ST and other PA-related outcomes for the first time using sensor-based data as the criterion. In a sample of 858 adults, breaks in ST and other PA-related parameters were assessed through sensor-based accelerometry and subjective data, which included a comprehensive questionnaire with a specific question ('During the day, do you usually sit for a long time in a row or interrupt frequently?') with a three-level closed answer. The Spearman's rank correlation coefficient was used to determine the agreement between the single-item question and sensor-based data. Positive correlations were found for self-reported breaks in ST with sensor-based breaks in ST in both women (ρ=0.37; 95% CI=0.29-0.44) and men (ρ=0.15; 95% CI=0.04-0.26). Self-reported breaks in ST were inversely correlated with ST in women (ρ=-0.33; 95% CI=-0.40 to 0.25). For both sexes, self-reported breaks in ST showed a positive correlation with light-intensity PA (ρ=0.39; 95% CI=0.31-0.46 women; ρ=0.13; 95% CI=0.02-0.24 men), however, positive correlations between self-reported breaks in ST and moderate-to-vigorous PA (ρ=0.13; 95% CI=0.02-0.24) were found only in men. Our single-item question can be used as an indication for ranking people's breaks in ST during the waking day, although acknowledging that some misclassification will occur, especially in men. There must be an effort to include this question in future national and international surveys to replicate these findings.