An evaluation of physical activity surveys in a multi-ethnic sample of mid-life women

The health enhancing effects of physical activity are well documented in the literature. However, women continue to report lower participation in exercise and physical activity (PA) compared to men. As women age an overall trend in decreased activity is observed. The primary place of activity reported in women is the home and one of the most commonly reported reasons for lack of physical activity is the lack of time. Few instruments have been developed that focus on the activity patterns of women. The Cross Cultural Physical Activity Study that this study was based on targeted women of color to assess the types of activity and constraints to activity experienced by African American and Native American women over 40 years old. This secondary data analysis focused on the psychometric properties of two scales used in the above study, The Physical Recreation Questionnaire (PRQ) and The Typical Week Physical Activity Survey (TWPAS). An exploratory factor analysis (EFA) was conducted on the 18 items from the Physical Recreation Questionnaire (PRQ) which focused on constraints to PA. The results of the EFA were a poor fit of a two factor model. The three factor model had a favorable fit in the EFA. Confirmatory factor analysis (CFA) was then conducted on the 18 items in the PRQ. Results of the CFA supported the presence of three latent variables: enjoyment of PA, constraints to PA, and negotiation of constraints to PA. The Typical Week Physical Activity Survey (TWPAS) is a 35 item measure of moderate PA that includes the activities most often reported by women. The purpose of the TWPAS was to capture habitual PA that might not be recorded in other PA questionnaires. The TWPAS was correlated with criterion measures of PA records, treadmill, accelerometer, and BMI. Although correlations were small, they were in the expected direction with the criterion measures. The evaluation of the instruments supported the presence of the construct of constraints to PA in the PRQ and the measurement of moderate intensity PA in the TWPAS.