Systematic review and meta-analysis synthesizes acupuncture effects on sleep in Parkinson's disease

BackgroundParkinson’s disease (PD) causes multiple non-motor symptoms (NMSs), such as insomnia, anxiety, and fatigue, that worsen the quality of life. Pharmacological options offer limited relief, prompting interest in acupuncture as an adjunctive therapy.MethodsFollowing PRISMA 2020 guidelines (PROSPERO CRD420251172700), PubMed, Embase, Web of Science, and Cochrane Library were searched from 1 Jan 2015 to 28 Oct 2025 for randomized and controlled observational studies comparing manual or electroacupuncture with sham or usual care in idiopathic PD. Outcomes were sleep quality (PDSS and PDSS-2), mood (HAM-A, HAM-D, HADS, and BDI), and fatigue (MFIS, FSS, and FACIT-F). Two reviewers independently extracted data, assessed bias with RoB 2 and ROBINS-I, and rated certainty using GRADE. Only sleep (PDSS) outcomes from two RCTs were quantitatively pooled; mood and fatigue outcomes were narratively summarized because only single trials were available.ObjectivesUnlike prior reviews that pooled diverse non-motor outcomes and comparator conditions, this review focuses on sleep disturbance, anxiety/depression, and fatigue in idiopathic PD and includes sham-controlled evidence up to 28 October 2025. It differs from recent network meta-analytic approaches by emphasizing symptom-specific, sham/usual-care controlled effect estimates and aligning certainty judgments with RoB 2/ROBINS-I and GRADE. Where data permitted, we assessed robustness using leave-one-out and fixed- versus random-effects sensitivity checks; meta-regression was not feasible due to the small number of trials.ResultsTwenty-five full-text reports were assessed, and eight primary studies met the inclusion criteria (4 RCTs and 4 observational). Two sham-controlled sleep RCTs (total n = 138) reported PDSS changes; pooled analysis suggested improved sleep with acupuncture (MD 14.52, 95% CI 7.27–21.78) with moderate heterogeneity (I2 = 68%), potentially related to differences in treatment duration (4 vs. 16 weeks) and protocol. One RCT reported a greater reduction in anxiety at follow-up (HAM-A difference 7.03 points), while fatigue showed no difference from sham (SMD 0.10, 95% CI − 0.20 to 0.40). No serious adverse events were reported. Since fewer than 10 studies contributed to any outcome, publication bias could not be formally assessed.ConclusionAcupuncture shows a moderate-certainty signal for improving sleep quality in PD. Evidence for anxiety is preliminary, based on one RCT with the main signal observed at follow-up, and requires replication. Evidence for fatigue is very uncertain and does not show superiority over sham. Larger multicenter RCTs with standardized protocols and ≥6-month follow-up are needed.