<![CDATA[Background: Chemotherapy-induced peripheral neuropathy is the commonest neurological complication of cancer treatment and has no established preventive therapy. Vitamin D deficiency is a plausible, cheaply correctable risk factor, but the only previous pooled estimate rested on two studies. Objective: To estimate the association between vitamin D deficiency or insufficiency and the risk and severity of chemotherapy-induced peripheral neuropathy, and to determine how much of the between-study variation is attributable to how neuropathy was ascertained. Methods: PubMed/MEDLINE and Europe PMC were screened, with four further sources and citation tracking. Cohort, cross-sectional and trial-embedded studies reporting serum 25-hydroxyvitamin D against neuropathy in adults were eligible. ROBINS-E was used. Random-effects models pooled odds ratios and, separately, Hedges g for 25-hydroxyvitamin D concentration. Ascertainment method was a pre-specified moderator. Results: Fourteen studies (3,654 participants) were included, seven poolable (2,118). The unstratified pooled odds ratio was 3.45 (95% confidence interval 1.97 to 6.03), I² 74.6%. Ascertainment explained 93.5% of between-study variance: patient-reported instruments gave 5.27 (3.41 to 8.13, I² 0%) and clinician-graded NCI-CTCAE gave 1.72 (1.29 to 2.29, I² 8.0%); ratio of odds ratios 3.01 (1.69 to 5.35). 25-Hydroxyvitamin D was lower in affected patients (g 0.58, 0.28 to 0.88). Bias-adjusted estimates ranged from 1.44 to 2.40. Certainty was very low, rising to low for the clinician-graded estimate. Conclusion: Vitamin D deficiency was associated with chemotherapy-induced peripheral neuropathy, but the magnitude depended almost entirely on how neuropathy was measured. The defensible estimate is a 1.7- to 2.4-fold increase in odds, not threefold.]]>
