<![CDATA[Introduction: Neuropathic pain frequently coexists with sleep disturbance, and current pharmacotherapy is constrained by limited efficacy and adverse effects. Melatonin, a pleiotropic neurohormone with analgesic, antioxidant, anti-inflammatory and chronobiotic properties, has been proposed as an adjunctive treatment, but its magnitude of clinical benefit on pain and sleep across heterogeneous neuropathic-pain populations had not previously been quantitatively synthesised in a meta-analysis specific to neurology practice. Methods: PubMed, Embase, Web of Science and the Cochrane Central Register of Controlled Trials were searched for randomised controlled trials of oral melatonin versus placebo or active control in adults with neuropathic pain, in accordance with PRISMA 2020. Two reviewers independently screened, extracted and assessed risk of bias using Cochrane RoB 2.0 with the crossover extension where applicable. The primary outcome was change in pain intensity (NRS or VAS); secondary outcomes were sleep quality (PSQI, ESS, sleep-interference NRS), responder rate (≥ 50 % pain reduction) and adverse events. Standardised mean differences (Hedges’ g) were pooled under a random-effects model with restricted maximum likelihood and Knapp–Hartung adjustment, with parallel raw mean differences on the NRS where instruments matched. Heterogeneity was quantified with I² and τ². Pre-specified subgroups examined etiology, dose and duration; sensitivity analyses used leave-one-out resampling and exclusion of the open-label and the mechanism-only trials. Results: Ten randomised controlled trials enrolling 491 participants were included. The exploratory primary pooled SMD across the two trials with extractable continuous pain data was −0.43 (95 % CI −4.86 to 4.00; I² = 77.8 %; P = 0.436); the wide interval reflects the Knapp–Hartung correction with k = 2 and is a structural feature of the methodology. The corresponding raw mean difference on the NRS 0–10 scale was −0.70 (95 % CI −1.99 to 0.59). Substantial heterogeneity was driven by divergent results in painful diabetic neuropathy (SMD −0.76, 95 % CI −1.16 to −0.36; raw NRS difference 1.3 points; P < 0.001) and mixed neuropathic pain (SMD −0.06, 95 % CI −0.57 to 0.45; P = 0.8). Sleep favoured melatonin in painful diabetic neuropathy (sleep-interference SMD −0.81, 95 % CI −1.21 to −0.41) and in multiple-sclerosis-related neuropathic pain (PSQI ↓ 55.9 %, P < 0.001). The ≥ 50 % responder rate in painful diabetic neuropathy was higher with melatonin (RR 1.47, 95 % CI 1.01 to 2.14). Risk of bias was low for six trials, raised some concerns for three trials and was high for one open-label trial. Adverse events were infrequent and did not differ from placebo. Conclusion: Current evidence indicates that melatonin reduces pain and improves sleep in painful diabetic peripheral neuropathy with concomitant sleep disturbance and may benefit selected post-surgical neurosensory deficits and multiple-sclerosis-related neuropathic pain, but is unlikely to benefit unselected mixed neuropathic pain in tertiary chronic-pain settings. Substantial clinical heterogeneity and the small number of pooled trials preclude routine recommendation. Adequately powered, etiology-stratified trials with harmonised pain and sleep endpoints are required.]]>
