Evaluation of a commercially available pedometer used to promote physical activity as part of a national programme

Objective To assess the accuracy of a pedometer (manufactured by Silva) currently being used as part of a national programme to promote physical activity in the UK. Methods Laboratory study: 68 participants (aged 19.2 years (SD 2.7), body mass index (BMI) 22.5 kg/m2 (SD 3.3)) wore two Silva pedometers (over the right and left hips) while walking on a motorised treadmill at 2, 2.5, 3, 3.5 and 4 mph. Pedometer step counts were compared with actual steps counted. Free-living study: 134 participants (aged 36.4 years (SD 18.1), BMI 26.3 kg/m2 (SD 5.1)) wore one Silva pedometer, one New-Lifestyles NL-1000 pedometer and an ActiGraph GT1M accelerometer (the criterion) during waking hours for one day. Step counts registered by the Silva and NL-1000 pedometers were compared with ActiGraph step counts. Percentage error of the pedometers were compared across normal-weight (n = 58), overweight (n = 45) and obese (n = 31) participants. Results Laboratory study: Across the speeds tested percentage error in steps ranged from 6.7 (4 mph) to 46.9% (2 mph). Free-living study: Overall percentage errors of the Silva and NL-1000 pedometers relative to the criterion were 36.3% and 9%, respectively. Significant differences in percentage error of the Silva pedometer were observed across BMI groups (normal-weight 21%, overweight 40.2%, obese 59.2%, p