Validation of moderate-to-vigorous physical activity pedometers

The objective of this study was to determine the validity of the Walk4Life model MVP 4 Function Digital Pedometer for measuring moderate to vigorous physical activity (MVPA). Sixty-five (38 female, 27 male) college students, age 18 years and older (M age = 20.17 years, SD = 4.08), volunteered for the study. Preliminary data collection included age, resting heart rate, height, body weight, body mass index (BMI), and gender. Each participant wore a pedometer and a Polar model T31 heart rate monitor to determine their MVPA steps per minute (SPM) threshold and to measure their heart rates during the treadmill tests. Manual and pedometer step counts were repeated and compared for 40 (17 male, 23 female) participants to establish step count accuracy and reliability in determining SPM. MVPA pedometer recordings were compared to the amount of time spent in their MVPA heart rate zone. The System for Observing Fitness Instruction Time (SOFIT) (McKenzie, Sallis, & Nader, 1991) instrument was used to code MVPA time during a Disc Lacrosse activity lesson and compared to MVPA pedometer readings during the direct observation portion of the study. Manually counted step counts did not significantly differ from pedometer recorded counts (p = .30), and pedometer recorded counts from the repeated trials were strongly and positively correlated (r = .99). Pedometers tended to overestimate MVPA at lower speeds (M Absolute Error (AE) = - 26.55) and underestimate MVPA at higher speeds (M AE= 23.53) during the treadmill tests. Neither gender (p= .721) nor BMI (p= .664) had a significant effect on absolute error. Pedometer estimates of MVPA significantly differed from that of the SOFIT (p= .012). Results yielded a significant effect for gender (p= .020) and BMI (p= .041) during direct observation. Walk4Life MVP pedometers are accurate and reliable for measuring steps, but may not be recommended for use by researchers to measure MVPA time due to estimation errors. Practitioners may find these pedometers useful for estimating MVPA because they are less arduous than direct observation methods such as the SOFIT