Efficacy and safety of deep brain stimulation for treatment-refractory anorexia nervosa: a systematic review and meta-analysis

Abstract Background Several pioneering studies investigated deep brain stimulation (DBS) in treatment-refractory anorexia nervosa (AN) patients, but overall effects remain yet unclear. Aim of this study was to obtain estimates of efficacy of DBS in AN-patients using meta-analysis. Methods We searched three electronic databases until 1st of November 2021, using terms related to DBS and AN. We included trials that investigated the clinical effects of DBS in AN-patients. We obtained data including psychiatric comorbidities, medication use, DBS target, and study duration. Primary outcome was Body Mass Index (BMI), secondary outcome was quality of life, and the severity of psychiatric symptoms, including eating disorder, obsessive-compulsive, depressive, and anxiety symptoms. We assessed the risk of bias using the ROBINS-I tool. Results Four studies were included for meta-analysis, with a total of 56 patients with treatment-refractory AN. Follow-up ranged from 6–24 months. Random effects meta-analysis showed a significant increase in BMI following DBS, with a large effect size (Hedges’s g = 1 ∙ 13; 95% CI = 0 ∙ 80 to 1 ∙ 46; Z-value = 6 ∙ 75; P < 0 ∙ 001), without heterogeneity (I 2 = 0 ∙ 00, P = 0 ∙ 901). Random effects meta-analysis also showed a significant increase in quality of life (Hedges’s g = 0 ∙ 86; 95% CI = 0 ∙ 44 to 1 ∙ 28; Z-value = 4 ∙ 01, P < 0 ∙ 001). Furthermore, DBS decreased the severity of psychiatric symptoms (Hedges’s g = 0 ∙ 89; 95% CI = 0 ∙ 57 to 1 ∙ 21; Z-value = 5 ∙ 47; P < 0 ∙ 001, I 2 = 4 ∙ 29, P = 0 ∙ 371). Discussion In this first meta-analysis, DBS showed statistically large beneficial effects on weight restoration, quality of life, and reduction of psychiatric symptoms in patients with treatment-refractory AN. These outcomes call for more extensive naturalistic studies to determine the clinical relevance for functional recovery. This study is preregistered in PROSPERO,CRD42022295712.

Abstract licence: CC BY-NC-ND

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.