Address physical inactivity, but avoid stressing harms.

In their excellent Comment (July 21, p. 192), Wen and Wu make a strong case for investing in behaviour change interventions to promote more physical activity in the inactive population. They list several helpful approaches, such as optimizing the available infrastucture and building skills of the inactive. However, their main recommendation is to stress the harms of physical inactivity, and recent evidence indicates that this is a dangerous suggestion. Recently, our meta-analysis showed that threatening communication is ineffective, and can even cause health-defeating behaviour, unless the receiver is high in efficacy. This means that when doctors stress the harms of physical activity, patients will do nothing or become less active unless they are convinced of their ability to successfully engage in physical activity. In addition, this meta-analysis showed that even when threatening communications are effective (ie, under high efficacy), the effect size is disappointing (Cohen's d = 0·31). This finding is in line with other recent meta-analyses. More effective and safer behaviour change methods exist. For example, targeting patients' efficacy has no danger of backfiring and enables safe use of threatening communications. Note that to manipulate efficacy, a simple behavioural recommendation does not suffice. Although ideally, the optimum method is determined specifically for the population and behavior at hand, an increasing number of meta-analyses also provides evidence as to effective methods in specific contexts. We advise practitioners to avoid stressing the harms of physical inactivity and focus on the other valuable suggestions of Wen and Wu.