Effectiveness of Metformin in Preventing Herpes Zoster Among Japanese Patients With Type 2 Diabetes: A Target Trial Emulation

Patients with type 2 diabetes have a higher risk of herpes zoster (HZ). Although metformin may reduce this risk via immunomodulatory effects, prior evidence is limited by confounding by indication and an unclear causal estimand. We aimed to estimate the per-protocol effect of sustained metformin use on the risk of HZ in comparison with dipeptidyl peptidase-4 (DPP-4) inhibitors, for which no evidence suggests an elevated HZ risk. We emulated a target trial using a Japanese claims database (April 2012-December 2023) to assess the 4-year risk of HZ among new users of either metformin or a DPP-4 inhibitor. Risks were estimated using an inverse probability weighted pooled logistic regression model. Among 9691 metformin users, 154 developed HZ (weighted 4-year risk: 4.18%), while among 17 202 DPP-4 inhibitor users, 345 developed HZ (4.92%). The risk difference was -0.74% (95% CI, -1.80 to 0.39), and the risk ratio was 0.85 (95% CI, 0.66 to 1.08). In conclusion, metformin use did not reduce the 4-year risk of HZ compared with DPP-4 inhibitor use. This finding contradicts a prior observational report of a marked risk reduction, which was likely inflated by confounding by indication, and highlights the need to specify a clear causal estimand.