Reliability and Validity of Self-Reported Questionnaires Assessing Physical Activity and Sedentary Behavior in Finland.

Abstract

Reliable and valid data on physical activity (PA) and sedentary behavior (SB) are needed for implementing evidence-based interventions and policies. Monitoring of these behaviors is based on PA questionnaires (PAQs) and device-based measurements, but their comparability is challenging. The present study aimed to investigate the test-retest reliability and concurrent validity of Finnish versions of the widely used PAQs (IPAQ-SF, EHIS-PAQ, GPAQ, Eurobarometer) and to compare their data with accelerometer data. This study is based on the Finnish data of the European Union Physical Activity and Sport Monitoring project (EUPASMOS). Participants (n = 62 adults, 62% women) answered the PAQs twice, one week apart, and wore an accelerometer for these seven consecutive days. Intraclass correlations, Spearman's rank correlations, t-tests, and Cohen's kappa with bootstrap confidence intervals were used to analyze the data. The PAQs had typically moderate-to-good test-retest reliability (ICC 0.22-0.78), GPAQ, EHIS-PAQ, and Eurobarometer showing the highest reliability. The PAQs correlated with each other when assessing sitting and vigorous PA (R = 0.70-0.97) and had a fair-to-substantial agreement when analyzing adherence to the PA recommendations (74-97%, Cohen's kappa 0.25-0.73). All the PAQs had a poor criterion validity against the accelerometry data. The Finnish versions of the PAQs are moderately reliable and valid for assessing PA, adherence to PA recommendations and sitting among adult participants. However, the poor criterion validity against accelerometer data indicates that PAQs assess different aspects of PA constructs compared to accelerometry.