<![CDATA[Background: Chronic low-grade inflammation is increasingly implicated in diabetic peripheral neuropathy (DPN). Tumour necrosis factor-alpha (TNF-α) and interleukin-6 (IL-6) have been proposed as circulating correlates of nerve injury, yet the evidence is inconsistent, a previous synthesis addressed TNF-α alone, and no pooled estimate has ever been reported for IL-6. Objective: To determine whether circulating TNF-α and IL-6 concentrations differ between diabetic patients with and without peripheral neuropathy, and to quantify the magnitude, precision, and consistency of any difference. Methods: Four databases were searched for observational studies reporting serum or plasma TNF-α and/or IL-6 in diabetic patients with versus without neuropathy. Standardised mean differences (SMD, Hedges g) were pooled using a random-effects model, with Hartung–Knapp adjustment, restricted maximum-likelihood re-estimation, cluster-collapsed and leave-one-out analyses, and a prediction interval. Risk of bias used the Newcastle–Ottawa Scale and certainty the GRADE framework. Results: Four studies (476 participants; 249 with DPN, 227 without) provided seven cytokine comparisons. The overall SMD was 0.70 (95% CI 0.22–1.18; p = 0.005; I² = 91%), robust across all sensitivity analyses. The effect was driven by IL-6, which was significantly and homogeneously elevated (SMD 0.65, 95% CI 0.42–0.87; I² = 2%). Pooled TNF-α was higher but non-significant and heterogeneous (SMD 0.71, 95% CI −0.18–1.60; I² = 95%), owing to one statin-treated cohort; excluding it raised the estimate to 0.88. Certainty was low for IL-6 and very low for TNF-α. Conclusion: DPN is associated with higher circulating pro-inflammatory cytokines, with IL-6 the most consistent correlate and the first for which a pooled estimate is available. IL-6 merits prospective evaluation as a biomarker of neuropathy risk. Findings are hypothesis-generating.]]>
