Gender bias in the assessment of physical activity in population studies

Objectives: Despite their generally more health promoting behaviours, women are found to participate less in physical activity than men. This study explores possible gender bias in measurement of physical activity in population studies. Methods: Data collected by telephone (CATI) from the Berne Lifestyle Panel in 1996 is utilised. A representative sample of the population of the city of Berne comprised N = 1119 cases. Gender differences are assessed for the weekly frequency of three measurements of physical activities. Results: An indicator of sport and exercise showed higher physical activity among men, while the indicator of habitual physical activity showed higher rates of daily walking and biking among women. A combined indicator of general physical activity showed no significant gender differences. Conclusions: The results provide empirical evidence on potential risk of underestimation of physical activity among women and of misclassification with respect to high or low risk behaviour patterns.