Assessment of physical activity and inactivity in multiple domains of daily life: a comparison between a computerized questionnaire and the SenseWear Armband complemented with an electronic diary

BACKGROUND Although differences between paper-and-pencil questionnaires and accelerometers have been reported for overall physical activity and time spent in moderate and vigorous activity, few studies have looked at domain-specific behavior. This study compared estimates of domain-specific physical (in)activity obtained with the Flemish physical activity computerized questionnaire (FPACQ) with those obtained from a combination of the SenseWear Armband and an electronic diary. Furthermore, it was investigated whether the correspondence between the two methods varied with gender and age. METHODS Data were obtained from 442 Flemish adults (41.4 ± 9.8 years). Physical activity was questioned with the FPACQ and measured for seven consecutive days using the SenseWear Armband together with an electronic activity diary (SWD). Analogous variables were calculated from the FPACQ and SWD. Mean differences and associations between FPACQ and SWD outcomes were examined with paired t-tests and Pearson correlations. The Bland-Altman method was used to assess the level of agreement between the two methods. Main effects and interaction of gender and age groups (20-34; 35-49; 50-64 years) on differences between FPACQ and SWD outcomes were analyzed using two-way ANOVAs. RESULTS All parameters of the FPACQ were significantly correlated with SWD assessments (r = 0.21 to 0.65). Reported activity was significantly different from SWD-obtained values for all parameters, except screen time. Physical activity level, total energy expenditure and time spent in vigorous activities were significantly higher (+0.14 MET, +25.09 METhours·week(-1) and +1.66 hours·week(-1), respectively), and moderate activities and sedentary behavior significantly lower (-5.20 and -25.01 hours·week(-1), respectively) with the FPACQ compared to SWD. Time and energy expenditure of job activities and active transport were significantly higher, while household chores, motorized transport, eating and sleeping were significantly lower with the FPACQ. Time spent in sports was lower (-0.54 hours·week(-1)), but energy expenditure higher (+4.18 METhours·week(-1)) with the FPACQ. The correspondence between methods varied with gender and age, but results differed according to the intensity and domain of activity. CONCLUSIONS Despite the moderate correlations, significant differences between the two methods were found. In general, physical activity was higher and sedentary behavior lower as calculated from the FPACQ compared to SWD.