A brief questionnaire is able to measure population physical activity levels accurately: A comparative validation study

Abstract

BackgroundCategorizing individuals on the basis of physical activity (PA) level has both epidemiological utility and public health implications. Questionnaires provide a combination of accuracy and practicality when assessing population PA, but their characteristics, including summarization of raw data and scaling methods, can affect utility in large-scale studies. The aim of this study was to ascertain the concurrent validity of a brief six-point scale for the categorical scaling of PA levels. The effect of participant characteristics on the scales was also assessed. MethodsA comparative observational study was used. One hundred participants aged 18–80 years completed the valid and reliable International Physical Activity Questionnaire (IPAQ) and the Human Activity Profile (HAP) and a six-point scale. Quadratic weighted kappa, Spearman’s rho, and Fisher’s exact tests were applied to compare the questionnaire results for concurrent validity. The effect of participant characteristics on PA level was analyzed using Fisher’s exact test or analysis of variance. ResultsKappa values comparing the six-point scale with the IPAQ (κ=0.46) and HAP (κ=0.57) showed fair to good agreement. Analysis using Fisher’s exact test illustrated a significant association between the six-point scale and both the HAP (p<0.000) and IPAQ (p<0.000). Spearman’s rho correlation values between the six-point scale and the IPAQ and the HAP were 0.49 and 0.69, respectively. Scores on all questionnaires were significantly related to age (p<0.05). Marital status and number of comorbidities were related to HAP and six-point scale scores (p<0.05). Occupation was significantly related to HAP but not the other two scales (p<0.000). ConclusionThe six-point scale provides a quick measurement of usual amount and intensity of PA that a person performs. This validation study found significant associations and a moderate strength of agreement when compared with the IPAQ and HAP. The six-point scale has the potential for utility in large-scale epidemiological studies as a quick method for scaling individuals by PA level.