Influence of pedometer tilt angle on step counting validity during controlled treadmill walking trials
Pedometers are tools frequently used to monitor walking-related physical activity patterns of overweight and obese populations. However, there is a known association between increasing body mass index (BMI) and decreasing pedometer accuracy. The decrease in pedometer accuracy has been attributed to tilt angle, but the specific tilt angle where pedometers fail to maintain accuracy is unclear. PURPOSE: The purpose of this study was to test two popular pedometer mechanisms, accelerometry- and pendulum-based, to determine the specific tilt angle where the pedometers fail to maintain step counts accurately. METHODS: Twenty subjects (10 men, Mean±SD: 25.4±4.2 yrs, 184.4±5.3 cm, 76.7±6.3 kg; 10 women: 21.6±3 yrs, 166.1±6.3 cm, 56.7±6.7 kg) walked two sets of 21 trials at a combination of treadmill speed (67.0, 80.4, 93.8 m/min) and tilt angle (-30,-20,-10, 0,+10,+20,+30°) while wearing two pedometers; the experimental pedometer attached to a custom-built pedometer gimbal for altering tilt angle, and the control pedometer in a neutral position (±2° of 0°). The pedometers were worn on each hip just anterior to the iliac crest in line with the mid-axillary line of the thigh. The first set of trials tested one pedometer mechanism, while the second set of trials tested the second pedometer mechanism in a counterbalanced order. Mean steps from the experimental pedometer were compared to the control pedometer for each trial (ÄStep=Experimental-Control) using a 3-factor RMANOVA at an alpha level of 0.05. RESULTS: Mean Ästeps for the accelerometry-based pedometer and the pendulum-based pedometer differed significantly (P<0.05) at tilt angles >±10° for treadmill speeds <or equal to 67.0 and 80.4 m/min, respectively. However, mean Ästeps across treadmill speeds was much lower for the accelerometry-based pedometer (Mean±SE: 4.9±1.9, 2±1.9, 4±1.9, 5±1.9, 4±1.9, 9±1.9 steps) than the pendulum-based pedometer (14.1±1.9, 4.2±1.9, 8.9±1.9, 6.8±1.9, 6±1.9, 38.1±1.9 steps) across conditions > 0° (-30, -20, -10, +10, +20, +30°, respectively). CONCLUSIONS: Increasing tilt angle caused a decrease in pedometer accuracy for both pedometer mechanisms, although the accelerometry-based pedometers were less affected. Negative tilt angles had less effect on pedometer accuracy than positive tilt angles, and the combination of increasing positive tilt angle and slower speeds had the greatest impact on step count accuracy.