Comparative efficacy of lifestyle intervention strategies targeting weight outcomes in people with psychosis: a systematic review and network meta-analysis.

OBJECTIVES The objective of this review was to pool and rank the efficacy of lifestyle intervention strategies targeting weight, body mass index, waist circumference and waist-to-hip ratio in people with psychosis by comparing the effect size of these weight outcomes. Secondary to this, the objective was to stratify the lifestyle interventions according to their inclusion of dietary information that adheres to Australian Dietary Guidelines. INTRODUCTION People living with psychosis have a significantly increased risk of all-cause mortality, with cardiovascular disease a considerable contributor to this risk. Controlling lifestyle risk factors, which include smoking, poor diet and inadequate physical activity, leads to significant weight reduction and decreases cardiovascular disease risk. Previous reviews on this topic have not clearly identified essential components of lifestyle interventions in people with psychosis, mainly due to statistical limitations of analyses. This review employed the network meta-analysis, which compares more than two groups of interventions and ranks them according to efficacy, thus providing a global estimate of effect. Additionally, available reviews have not assessed compliance of dietary information offered in lifestyle interventions to established guidelines. INCLUSION CRITERIA This review considered randomized controlled studies that delivered lifestyle interventions to community-dwelling adults with psychotic disorders. Outcomes of interest were weight, body mass index, waist circumference and waist-to-hip ratio. METHODS The Cochrane Library, MEDLINE/PreMEDLINE, Embase, CINAHL, Scopus and PsycINFO were searched for studies published in English from 1985 to June 2018. Data were qualitatively summarized, during which lifestyle intervention subgroups were created (based on key similarities) and then compared in direct meta-analyses and network meta-analyses. Assessments of study adherence to Australian Dietary Guidelines was conducted in a narrative format. RESULTS Thirty-two randomized control studies were included, and the overall quality of these studies ranged from what appeared to be low to moderate. Lifestyle intervention studies contained both a dietary and physical activity component, with the exception of two studies that focused solely on physical activity. Delivery of dietary and physical activity information was mainly through education; however, some studies provided additional structure to the intervention by offering tailored advice or helping participants to set goals, and providing regular review of progress for diet, physical activity or both. Results from network-meta-analyses showed that only studies with a structured approach for both diet and physical activity demonstrated significant decreases in weight (effect size = -4.12, 95% confidence interval = -7.772 to -2.760, P = 0.000) and body mass index (effect size = -2.94, 95% confidence interval = -1.78 to -0.357, P = 0.003). Waist circumference subgroup comparisons mainly comprised single studies; therefore, findings are inconclusive. Dietary information provided in studies generally complied with Australian Dietary Guidelines; however, none of the studies complied with all guidelines. CONCLUSIONS Lifestyle interventions incorporating both dietary and physical activity components led to the greatest decreases in weight (4.1 kg) and body mass index (2.9 points) among people with psychosis. Important intervention strategies for both components are the personalization of education through tailored advice or goal setting, and a corresponding progress review. Dietary information in the included studies appeared to comply with the Australian Dietary Guidelines. However, these findings were weakened by an increased risk of bias, complex and multicomponent study designs, and lack of clarity in reporting of study methodology.