Reactivity: an issue for short-term pedometer studies?

Objectives: To investigate the influence of wearing a pedometer and recording daily step counts on ambulatory activity. Methods: During the first visit to the laboratory, 50 healthy volunteers, blinded to the study aim, were provided with a sealed pedometer (New-Lifestyles NL-2000) and informed that it was a "body posture monitor'' (sealed condition). Participants wore the pedometer throughout waking hours for one week. Upon returning to the laboratory step counts were downloaded and participants were informed that the device was actually a pedometer. They were requested to wear it unsealed for a second one-week period, and to record their step counts in an activity log (unsealed condition). Results: Mean (SD) daily step counts reported in the sealed condition (9541 (3186) steps/day) were significantly lower than step counts reported in the unsealed condition (11385 (3763) steps/day) (p < 0.001). Conclusions: Step counts increased significantly in the unsealed condition. Possible mechanisms for this include the knowledge of wearing a pedometer, the visible step count display and the completion of the activity log. This has validity implications for short-term studies (duration <= 1 week) investigating habitual pedometer-determined activity levels.