Reliability and validity of self-reported sitting and breaks from sitting in the workplace

Prolonged sitting is a health risk factor which is ubiquitous to the workplace, and breaking up prolonged sitting is widely recommended. This study evaluated the test-retest reliability and concurrent validity of a self-report measure of duration of sitting and breaks from sitting in the workplace.

Cross-sectional study.

Fifty-nine workers who reported spending most of their work time sitting wore an activPAL inclinometer and the ActiGraph accelerometer for eight consecutive days, and completed single-item measures of duration of sitting (min/work hour) and breaks from sitting (frequency/per work hour), twice, seven days apart.

Participants reported sitting at work for a median of 420min/day (Interquartile Range=360-450min/day) and taking one break (Interquartile Range=1.0-2.0) from sitting per work hour. For reported duration of workplace sitting, test-retest reliability was adequate (Intra-Class Correlations=0.78, 95% Confidence Intervals [CI]=0.65, 0.86), and concurrent validity fair against the activPAL (Spearman's Rho=0.24, CI-1.0,0.47) and the ActiGraph (Rho=0.39, CI=0.15, 0.68). For reported breaks from sitting (frequency/per work hour), test-retest reliability was adequate (Intra-Class Correlations=0.65, CI=0.48, 0.78) and concurrent validity fair against the activPAL (Spearman's Rho=0.39, CI=0.25, 0.74) and the ActiGraph (Spearman's Rho=0.30, CI=0.15, 0.69). Self-reported duration of sitting was biased toward over-reporting compared to the activPAL (median=45.4min) and under-reporting compared to the ActiGraph (median=21.7min).

This study found adequate reliability and fair validity for self-reported duration of sitting (min/work day) and breaks from sitting (frequency/per work hour). Further validity research is needed using the inclinometer.