Vagus nerve stimulation in obesity: a systematic review and meta-analysis

Background Vagus nerve stimulation (VNS) has been investigated as a potential therapy for obesity because of the vagus nerve’s role in satiety, gastrointestinal signaling, autonomic regulation, and metabolic homeostasis. However, studies vary substantially in stimulation site, interface, parameters, species, diet condition, and reported outcomes. Methods PubMed, Embase, Cochrane Library, Scopus, and Web of Science were searched from inception through May 6, 2026. Eligible studies evaluated chronic electrical stimulation of the vagus nerve or a defined vagal branch, included a comparator, and reported quantitative body weight data. Random-effects meta-analyses using Hedges’ g evaluated final body weight and weekly body weight change, while food intake, adiposity, metabolic outcomes, and safety were summarized narratively. Results Of 629 records identified, 23 publications met the inclusion criteria, yielding 25 experimental comparisons. Across 22 comparisons, VNS was associated with lower final body weight than control conditions (Hedges’ g = −1.07, 95% CI −1.43 to −0.72; I 2 = 55.5%; P < 0.001). Across 14 comparisons, VNS was associated with lower weekly body weight gain or greater weekly body weight reduction (Hedges’ g = −2.82, 95% CI −4.23 to −1.41; I 2 = 93.3%; P < 0.001). In animal studies, modality-specific final body weight estimates favored conventional implanted electrical VNS ( g = −0.84), transcutaneous auricular VNS ( g = −1.54), and self-powered triboelectric VNS ( g = −8.36), although the triboelectric subgroup comprised two comparisons from one publication. Excluding the triboelectric comparisons, VNS remained associated with lower final body weight ( g = −0.93, 95% CI −1.23 to −0.63, I 2 = 39.4%). Food intake was commonly reduced after VNS, and adiposity outcomes generally favored stimulation, whereas metabolic and safety outcomes were reported inconsistently. Conclusion Chronic, anatomically specific VNS was associated with lower body weight across predominantly preclinical studies. The final body weight estimate showed moderate heterogeneity, whereas the weekly body weight change estimate was highly heterogeneous and should be interpreted as supportive rather than as a precise estimate of effect magnitude. Translation to obesity treatment remains limited by sparse human evidence, potential small-study effects, and inconsistent reporting of metabolic and safety outcomes.