Workplace health promotion: a meta-analysis of effectiveness

Context: An unhealthy lifestyle may contribute to ill health, absence due to sickness, productivity loss at work, and reduced ability to work. Workplace health promotion programs (WHPPs) aim to improve lifestyle and consequently improve health, work ability, and work productivity. However, systematic reviews on intervention studies have reported small effects, and the overall evaluation of effectiveness of WHPPs is hampered by a large heterogeneity in interventions and study populations. This systematic review aims to investigate the influence of population, study and intervention characteristics, and study quality on the effectiveness of workplace health promotion programs.

Evidence acquisition: A systematic literature search was conducted identifying RCTs, published before June 2012, evaluating the effect of a WHPP aimed at smoking cessation, physical activity, healthy nutrition, and/or obesity on self-perceived health, work absence due to sickness, work productivity, or work ability. Studies were included in the meta-analyses if quantitative information was present to calculate an effect size (ES). A meta-analysis, stratified meta-analyses, and meta-regression analyses were performed in Spring 2012 using Comprehensive Meta-analysis software 2.0 and PAWS 17.0.2.

Evidence synthesis: In 18 studies describing 21 interventions, the overall effect of a WHPP was small (ES=0.24, 95% CI=0.14, 0.34). The effectiveness of a WHPP was larger in younger populations, in interventions with weekly contacts, and in studies in which the control group received no health promotion. A 2.6-fold lower effectiveness was observed for studies performing an intention-to-treat analysis and a 1.7-fold lower effectiveness for studies controlling for confounders. Studies of poor methodologic quality reported a 2.9-fold higher effect size of the WHPP.

Conclusions: The effectiveness of a WHPP is partly determined by intervention characteristics and statistical analysis. High-quality RCTs reported lower effect sizes. It is important to determine the effectiveness of WHPPs in RCTs of high quality.