Abstract
Abstract Background Epidural steroid injections are widely used for the management of lumbar radiculopathy and degenerative spinal disorders. Although these injections provide short-term pain relief, concerns have been raised regarding potential systemic skeletal effects, particularly the risk of vertebral fractures. Existing studies have reported inconsistent findings, and recent cohort data warrant an updated systematic evaluation. Main body This systematic review and meta-analysis aimed to assess the association between epidural steroid injections and vertebral fracture risk. A comprehensive literature search of Ovid MEDLINE, PubMed/MEDLINE, and the Cochrane Library was conducted from database inception to July 2025 in accordance with PRISMA 2020 guidelines. Eligible studies were peer-reviewed cohort studies comparing vertebral fracture outcomes between patients receiving epidural steroid injections and non-injection controls. Data on study characteristics, patient demographics, follow-up duration, and adjusted hazard ratios were extracted. Risk of bias was assessed using the Newcastle–Ottawa Scale, and certainty of evidence was evaluated using the GRADE framework. A random-effects model was applied to pool hazard ratios, and statistical heterogeneity was quantified using the I 2 statistic. Three retrospective cohort studies conducted in Taiwan and the United States were included, comprising 320–8901 participants with follow-up periods ranging from 2 to 5 years. Random-effects pooling (DerSimonian–Laird) yielded a hazard ratio of 1.41 (95% CI 1.10–1.80; I 2 = 89%); however, application of the more conservative Hartung–Knapp–Sidik–Jonkman (HKSJ) adjustment rendered the association statistically non-significant (HR 1.41, 95% CI 0.72–2.76). Leave-one-out sensitivity analyses further demonstrated that omission of either U.S.-based cohort eliminated statistical significance. The 95% prediction interval (τ2 = 0.0301; df = 1) ranged from HR 0.09 to 21.5, indicating that the direction and magnitude of effect in any future study cannot be reliably anticipated from the current evidence base. Based on the control event rate from the largest cohort, this corresponds to approximately eight additional vertebral fractures per 1000 person-years (95% CI 2–16 more). Given the limited number of studies and variability in exposure definitions and study designs, further quantitative exploration of heterogeneity and dose–response relationships was not feasible in a statistically reliable way. Conclusion Current evidence is insufficient to establish a causal association between epidural steroid injections and vertebral fracture risk. Given very low certainty of evidence, substantial heterogeneity, and unresolved confounding by indication, the pooled estimate should be regarded as a directional signal warranting clinical attention rather than definitive evidence of harm. Prospective studies with standardized protocols are needed. PROSPERO Registration: This review was registered with the International Prospective Register of Systematic Reviews (PROSPERO, registration number: CRD420251160896).