Examining The Accuracy Of Self-reported Sitting Time Questionnaires Compared To An Objective Measurement

Abstract

INTRODUCTION: Sedentary behavior (SB) is associated with an increased risk for adverse health outcomes such as metabolic syndrome and type 2 diabetes. Self-report instruments are commonly used to measure sitting time (ST), a surrogate measure of SB. However, these questionnaires have not been validated using a criterion measure of ST. PURPOSE: To examine the accuracy of self-reported ST compared to objectively measured ST. METHODS: Nineteen overweight and obese office workers (BMI = 33.4+5.94 kg·m-2; age = 45.6 +11.28 years) wore an activPAL (AP) activity monitor for two weeks. This monitor recorded time spent sitting/lying, standing, and stepping. At the end of each week, participants reported ST using the Sitting Questionnaire (SQ) and the Focus Questionnaire (FQ). Bias was calculated to determine the accuracy of the self-reported estimates of ST compared to the AP measurements for weekdays and weekend days. Correlation coefficients and 95% confidence intervals (CI) were calculated as measures of precision. RESULTS: For weekdays the FQ was not significantly different from the AP, but it underestimated ST by.91 hours (CI -2.1 to 0.30 hrs). In contrast, the SQ overestimated ST by 2.8 hours (CI 1.9 to 3.6). For weekend days the FQ underestimated ST by 1.8 hours (CI -3.0 to -0.58) whereas the SQ overestimated ST by 3.0 hours (CI 1.1 to 4.9). The correlations for weekdays and weekend days between the SQ and the AP ranged from r = 0.22 to r = 0.34 as compared to r=0.10 to r=0.42 for the FQ and the AP. Discussion: The self-report measures examined in this study did not accurately measure ST as compared to an objective measure. Although the correlations between self-report measures and objective measures are low to moderate, the confidence intervals of the bias reflect the high variability in the accuracy of the self-report measures. Furthermore, since the SQ overestimated ST and the FQ underestimated ST, researchers should use caution when comparing ST across studies that use different self-report measures. When possible objective monitors should be used to assess ST. Funded by ACSM Doctoral Student Research Grant