BACKGROUND: Motivational interviewing (MI) is potentially useful in management of overweight and obesity, but staff training and increased delivery time are barriers, and its effectiveness independent of other behavioral components is unclear. PURPOSE: To assess the independent contribution of MI as part of a behavioral weight management program (BWMP) in controlling weight and improving psychological well-being. DATA SOURCES: 6 electronic databases and 2 trial registries, searched from database inception through 24 September 2021. STUDY SELECTION: Randomized controlled trials in adults or adolescents aimed at weight loss or maintenance and comparing programs incorporating MI versus interventions without MI. DATA EXTRACTION: Two reviewers independently screened studies, extracted data, and assessed risk of bias. Outcomes included weight, anxiety, depression, quality of life, and other aspects of psychological well-being. Pooled mean differences or standardized mean differences were obtained using random- and fixed-effects meta-analyses. DATA SYNTHESIS: Forty-six studies involving 11 077 participants, predominantly with obesity, were included. At 6 months, BWMPs using MI were more effective than no/minimal intervention (-0.88 [95% CI, -1.27 to -0.48] kg; I 2 = 0%) but were not statistically significantly more effective than lower-intensity (-0.88 [CI, -2.39 to 0.62] kg; I 2 = 55.8%) or similar-intensity (-1.36 [CI, -2.80 to 0.07] kg; I 2 = 18.8%) BWMPs. At 1 year, data were too sparse to pool comparisons with no/minimal intervention, but MI did not produce statistically significantly greater weight change compared with lower-intensity (-1.16 [CI, -2.49 to 0.17] kg; I 2 = 88.7%) or similar-intensity (-0.18 [CI, -2.40 to 2.04] kg; I 2 = 72.7%) BWMPs without MI. Studies with 18-month follow-up were also sparse; MI did not produce statistically significant benefit in any of the comparator categories. There was no evidence of subgroup differences based on study, participant, or intervention characteristics. Too few studies assessed effects on psychological well-being to pool, but data did not suggest that MI was independently effective. LIMITATIONS: High statistical heterogeneity among studies, largely unexplained by sensitivity and subgroup analyses; stratification by comparator intensity and follow-up duration resulted in pooling of few studies. CONCLUSION: There is no evidence that MI increases effectiveness of BWMPs in controlling weight. Given the intensive training required for its delivery, MI may not be a worthwhile addition to BWMPs. PRIMARY FUNDING SOURCE: National Institute for Health Research Biomedical Research Centre. (PROSPERO: CRD42020177259).
Ann Intern Med