Reliability of pedometer data in samples of youth and older women

Lisa A Strycker, Susan C Duncan, Nigel R Chaumeton, Terry E Duncan, Deborah J Toobert

Pedometers offer researchers a convenient and inexpensive tool for objective measurement of physical activity. However, many unanswered questions remain about expected values for steps/day for different populations, sources of variation in the data, and reliability of pedometer measurements. This study documented and compared mean steps/day, demographic predictors of steps/day, and pedometer reliability in two longitudinal investigations, one involving a population-based youth sample (N = 367) and the other targeting postmenopausal women with type 2 diabetes (N = 270). Individuals were asked to wear pedometers (Yamax model SW-701) at the waist for 7 days and record steps/per day. They were also asked to record daily physical activities, duration, and perceived intensity (1 = low/light, 2 = medium/moderate, 3 = high/hard) for the same 7 days. In addition, survey data regarding usual physical activity was collected. Analyses of variance (ANOVA) were conducted to determine whether there were significant differences in pedometer results according to sex, age, and body mass index. Repeated measures ANOVAs were used to examine potential differences in results among differing numbers of days. Mean steps/day were 10,365 steps in the youth sample and 4,352 steps in the sample of older women. Girls took significantly fewer steps than boys, older women took fewer steps than younger women, and both youth and women with greater body mass took fewer steps than those with lower body mass. Reliability coefficients of .80 or greater were obtained with 5 or more days of data collection in the youth sample and 2 or more days in the sample of older women. Youth and older women were more active on weekdays than on weekends. Low but significant associations were found between step counts and self-report measures of physical activity in both samples. Mean steps/day and reliability estimates in the two samples were generally consistent with previously published studies of pedometer use. Based on these two studies, unsealed pedometers were found to offer an easy-to-use and cost-effective objective measure of physical activity in both youth and older adult populations.