Step counter use and sedentary time in adults: A meta-analysis

Abstract

OBJECTIVE: Although step counters are increasingly being used in walking programmes to promote sedentary behavior changes in adults, their effectiveness remains unknown. The aim of this meta-analysis of randomized controlled trials (RCTs) was to assess the effectiveness of step counter use in reducing sedentary time among adults.
METHODS: English-language RCTs from 3 databases were searched up to December 2014. Studies were included if they evaluated the effects of step counter use in adult populations and reported outcomes in sedentary time. Summary estimates (Cohen d with 95% confidence intervals [CIs]) were pooled using a random-effects model. Subgroup analyses and random-effects meta-regression analyses based on the characteristics of participants or interventions were conducted to explore their associations with sedentary time changes.
RESULTS: Fifteen RCTs with a total sample size of 3262 adults were included. Step counter use was associated with a small but significant overall effect in reducing sedentary time (d = -0.20, 95% CI -0.33 to -0.07), equating to a reduction in sedentary time of ~23 min/d compared with controls. Subgroup analyses showed that step counter use with a step goal was associated with significantly reduced sedentary time (d =- 0.32, 95% CI -0.53 to -0.11), whereas without, it had only a trend. A greater reduction in sedentary time was observed among step counter users employing objective methods than those employing subjective methods for measurement (P = 0.03). Effects of covariates on sedentary time changes were generally unclear.
CONCLUSIONS: Step counter use is associated with reduced sedentary time among adults. Future studies are required to specify the step goal use and to employ objective as well as subjective methods for measuring both total and domain-specific sedentary time.