Meta-analysis of placebo-arm dropouts in osteoporosis randomized-controlled trials and implications for nocebo-associated discontinuation of anti-osteoporotic drugs in clinical practice

Summary Dropout from placebo arms in randomized-controlled trials is a surrogate for nocebo responses, resulting from patients’ negative expectations to treatment. Among 16,460 placebo-treated patients in oral anti-osteoporotic drug trials, nocebo dropouts were 8% on average, being higher in older patients. This implies that nocebo may contribute to the osteoporosis treatment gap in clinical practice. Purpose Osteoporosis is a common disease requiring long-term treatment. Despite the availability of effective anti-osteoporotic drugs, adherence to treatment is low. Nocebo, a behavior mostly related to the negative expectations to a certain treatment, decreases adherence and negatively affects treatment outcomes and health-related care costs in chronic diseases. Since in double-blind placebo-controlled randomized trials any unfavorable outcome leading to discontinuation in placebo arms is considered as nocebo, we aimed to investigate the size of nocebo response in patients participating in osteoporosis trials. Methods We searched MEDLINE, EMBASE, SCOPUS, and Cochrane databases for dropouts due to reported adverse events in the placebo arms (nocebo dropouts) in all double-blind trials investigating anti-osteoporotic drugs published between January 1993 and March 2022. Only data on bisphosphonates and selective estrogen receptor modulators (SERMs) were analyzed (Prospero registration number CRD42020212843). Results Data from 44 trials were extracted. In 16,460 placebo-treated patients, the pooled nocebo-dropout was 8% both for bisphosphonates (average: 0.08; range 0.01–0.27; 95%CI 0.06–0.10) and SERMs (average: 0.08; range 0.03–0.15; 95%CI 0.05–0.13). Nocebo-dropouts were higher in bisphosphonate trials enrolling individuals ≥ 65 years (11%) ( n = 18) compared to trials enrolling younger individuals (6%) ( n = 18) (average: 0.11; 95%CI 0.08–0.13 vs. average: 0.06; 95%CI 0.05–0.08, respectively, p = 0.001). Participants’ sex, dosing-intervals, publication year, or severity of osteoporosis had no impact on the nocebo-dropouts. Conclusion Almost 1 in 10 osteoporosis patients receiving placebo in trials of bisphosphonates and SERMs experiences AEs leading to dropout, implying that nocebo contributes to treatment-discontinuation in clinical practice. Efforts to identify and minimize nocebo, especially in older patients, are warranted.