Validation and modification of the Scottish Physical Activity Questionnaire for use in a female student population

Abstract

Objectives Two trials were undertaken to establish and improve the validity of seven-day physical activity (PA) recall using the Scottish Physical Activity Questionnaire (SPAQ) in a female student population. Fifty-five female students were recruited to Trials One and Two (n=30; n=25). Methods The SPAQ was completed following a monitoring period. Heart rate (HR) data was collected for three days, following which the SPAQ was completed. HR data were converted to minutes spent in moderate and vigorous PA. This required the monitoring of HR during rest and three activities selected to reflect light, moderate and vigorous activity. Questionnaire modification was informed by data analysis and the responses of 16 Trial One participants who agreed to attend group interviews. In Trial Two the method was replicated using the modified SPAQ. Results The mean of three days of HR data were compared with the equivalent SPAQ data. Parametric statistical analysis was carried out on log-transformed data owing to skewed data distributions. In all but six cases the SPAQ estimate exceeded objective estimates of time in activity, with a median difference of 40.70 minutes, found to be statistically significant on t-test (95% CI: −1.39, 0.23) and poorly correlated (r=0.03). Limits of Agreement analysis demonstrated poor agreement between estimates. Following questionnaire modification the difference between median estimates of time in PA reduced to 10.60 minutes, no longer significantly different (95% CI: −0.75,9.18E−02). The correlation improved to 0.59 and Limits of Agreement also improved. Conclusions Students undergo substantial lifestyle changes that may have a detrimental effect on their activity participation. As women are at particular risk, the measurement of PA in female students is important to allow epidemiological and interventional research. The SPAQ was found to require modification for use in this population. Alterations included reordering of questions and the provision of examples to aid the estimation of time spent in routine and intermittent activities. The modified SPAQ for students demonstrated reduced bias and improved agreement with an objective measure. It was therefore felt to be appropriate for the measurement of PA in female students.