SGLT-2 inhibitors and the risk of migraine in patients with type 2 diabetes mellitus: a population-based target trial emulation study

Preclinical studies in animals have suggested potential neuroprotective effects of sodium-glucose cotransporter 2 inhibitors (SGLT-2i), but no epidemiological study has investigated the potential effects of SGLT-2i on migraine risk. We aimed to assess the association between use of SGLT-2i and migraine incidence. Using a large population-based electronic health database from Xiamen, China, we conducted a retrospective cohort study emulating a target trial among adult patients with type 2 diabetes mellitus (T2DM) initiating SGLT-2i versus dipeptidyl peptidase-4 inhibitors (DPP-4i) between January 1, 2019, and September 1, 2025. The primary endpoint was incident migraine. Inverse probability of treatment weighting (IPTW) with stabilized weights was applied to achieve covariate balance. Weighted Cox proportional hazards models were used to estimate adjusted hazard ratios (HRs). Prespecified subgroup and sensitivity analyses were performed to assess robustness. The final cohort included 48,276 initiators of SGLT-2i and 50,441 initiators of DPP-4i. A total of 801 incident migraine cases were identified during a mean (SD) follow-up of 3.19 (1.75) years. After adjustment using stabilized IPTW, SGLT-2i initiation was associated with a lower hazard of incident migraine compared with DPP-4i initiation (HR 0.78; 95% CI 0.66–0.93; P = 0.006). The weighted incidence rate difference was − 0.49 (95% CI -0.84 to -0.14) per 1,000 person-years, and the 5-year risk difference was − 0.30 (95% CI -0.51 to -0.09) percentage points. No statistically significant effect modification was observed across the prespecified subgroups. The findings were generally consistent across sensitivity analyses using alternative weighting approaches, washout periods, landmark analyses, active comparators, and more stringent migraine definitions. SGLT-2i initiation was associated with a lower risk of incident migraine compared with DPP-4i initiation among patients with T2DM. These findings support the neuroprotective effects of this drug class and warrant validation in prospective trials. Not applicable.