Impact of primary care exercise referral schemes on the health of patients with obesity.

Primary care exercise referral schemes (ERSs) are a potentially useful setting to promote physical activity (PA). It is not established, however, whether interventions to increase PA, such as ERSs, have differing health outcomes according to the participants' body mass index (BMI). This paper summarizes evidence for the impact of primary care ERSs on the health of people with obesity and reports findings of a reanalysis of the EMPOWER study, providing the first data to report differential outcomes of ERSs by BMI category. Our literature review revealed a paucity of published data. A 2011 Health Technology Assessment review and 2015 update were identified, but normal-weight participants were neither excluded nor were results stratified by weight in the included studies. A study of the effect of exercise referral in overweight women reported a significantly greater increase in PA levels in the ERS group than the control group at 3 months. Reanalysis of the EMPOWER study data showed a significant improvement in PA at 3 months in both obese and overweight/normal BMI groups, with the effect size attenuated to 6 months. There was no significant difference from baseline to 6 months in blood pressure for either BMI category. At 6 months, there was a significant decrease in weight from baseline for the obese category. Comparison of crude mean differences between BMI groups revealed a significant mean difference in PA at 3 months favoring the overweight/normal BMI group, but not at 6 months. There were no further significant differences in unadjusted or adjusted mean differences for other outcomes at follow-up. We report some evidence of a differential impact of ERS on PA by BMI category. However, the effect of ERSs in primary care for patients with obesity remains unclear due to the small number of published studies that have reported outcomes by BMI category. Further research is needed.