Imeglimin is a novel oral antidiabetic agent that enhances insulin secretion and improves insulin sensitivity. Randomized trials have shown glycemic benefit but have identified gastrointestinal symptoms and hypoglycemia as notable risks. In Japan, dipeptidyl peptidase-4 (DPP-4) inhibitors such as sitagliptin remain commonly prescribed, with a low hypoglycemia and gastrointestinal adverse event profile. Comparative real-world safety data between imeglimin and DPP-4 inhibitors in routine clinical practice, particularly among older patients, were limited. This study used community pharmacy records to compare treatment-related adverse events between new users of imeglimin and sitagliptin, prespecifying a 10% absolute difference in gastrointestinal symptom incidence as the threshold for clinical relevance.
This retrospective, pharmacy-based cohort employed a new-user design using community pharmacy questionnaires distributed by the Japan Pharmaceutical Association (JPA). Questionnaires and pharmacy records from affiliated pharmacies collected patient-level data for new prescriptions issued in September of 2022–2024. New users were those with at least one prior pharmacy visit before March and no recorded use of the target drug during the March–August washout. Patients who switched between the study drugs during follow-up or who had no subsequent pharmacy visits were excluded. The final pooled sample comprised 1,140 imeglimin new users and 990 sitagliptin new users. Observation for each patient covered a 4-month period (September 1–December 31) in the year of initiation.
Community pharmacists actively monitored and recorded adverse events during routine counseling and at each prescription visit. Reported events were assessed for potential causal relationship to the medication and coded using the Medical Dictionary for Regulatory Activities (MedDRA) version 27.0, converting reported low-level terms to preferred terms for analysis. Data collected at baseline included demographics, comorbidities, concomitant medications, hospitalization history, smoking and alcohol status, and daily dose.
Incidence proportions, absolute risk differences, and 95% confidence intervals (CIs) were calculated. Time-to-event analyses used days as the time scale and Cox proportional hazards models to estimate unadjusted, age- and sex-adjusted, and multivariable-adjusted hazard ratios (HRs). Adjustment covariates included age, sex, smoking, alcohol, selected comorbidities (hypertension, dyslipidemia, hyperuricemia, myocardial infarction, heart failure, cancer), hospitalization history, and concomitant medications (antihypertensives, non-insulin glucose-lowering agents, insulin, lipid-lowering drugs, hypouricemic agents). To address sparse-event bias, sensitivity analyses employed Firth’s penalized likelihood Cox regression. A prespecified sensitivity analysis restricted to monotherapy users (no other glucose-lowering agents at baseline) was performed.
Among included patients, mean age was 66.9 years for imeglimin users and 69.9 years for sitagliptin users. Average daily doses were 1,619.4 mg for imeglimin and 46.2 mg for sitagliptin. Mean observation durations were similar (approximately 99–101 days). Notable baseline imbalances (standardized mean differences >0.10) were present for age, dyslipidemia, concomitant use of other antidiabetic agents or insulin, lipid-lowering therapy, and hospitalization history; these factors were considered in multivariable adjustment.
Overall, at least one adverse event was reported in 31% (351 cases) of imeglimin users versus 22% (220 cases) of sitagliptin users (crude relative risk 1.39; 95% CI 1.20–1.60). The most frequently reported events among imeglimin users included gastrointestinal symptoms (12.8%), decreased therapeutic response (6.1%), and nausea (3.6%). In multivariable Cox models, imeglimin was associated with substantially higher adjusted hazard ratios for gastrointestinal events (aHR 5.44; 95% CI 3.44–8.61). Specific symptom-level associations included:
Other events reported exclusively or primarily in the imeglimin group included vomiting, soft stools, and dyspepsia. Sitagliptin users more commonly had events coded as decreased therapeutic response, improvement in pre-existing conditions, and inadequate diabetes control; the rate of improvement in pre-existing conditions was higher in the sitagliptin group (aHR 0.25 for imeglimin vs sitagliptin; 95% CI 0.09–0.70).
No statistically significant difference in hypoglycemia risk was observed between groups (aHR 0.66; 95% CI 0.20–2.22). The authors caution interpretation of hypoglycemia findings because the number of events was limited.
Kaplan–Meier curves illustrated cumulative incidence differences for gastrointestinal symptoms during follow-up, with a higher cumulative incidence among imeglimin users.
In the monotherapy sensitivity subgroup (imeglimin n=114; sitagliptin n=561), gastrointestinal events occurred in 19 imeglimin users and 15 sitagliptin users. The multivariable-adjusted hazard ratio for gastrointestinal events in this subgroup was 5.36 (95% CI 2.63–10.91), with an absolute risk difference of 13.4% (95% CI 7.6%–21.2%). Estimates from Firth’s penalized likelihood Cox regression were similar to the primary analyses (e.g., aHR 5.20; 95% CI 2.55–10.61), supporting consistency of the association despite some outcomes having small event counts.
Strengths of the study include use of a new-user design, active event ascertainment by community pharmacists, and MedDRA-based standardized coding. These methods can capture adverse events that may be underreported in claims databases. Limitations include the retrospective design, potential residual confounding (including confounding by indication), imbalance in some baseline characteristics, limited follow-up of 4 months per patient, and small numbers of events for certain outcomes (notably hypoglycemia) that limit precision. The data source is not publicly available; data access is restricted by the Japan Pharmaceutical Association.
In Japanese community practice captured by pharmacy-based surveillance, imeglimin was associated with a higher incidence of gastrointestinal adverse events than sitagliptin, a pattern that persisted in monotherapy sensitivity analyses and approached the prespecified threshold for clinical significance. No clear difference in hypoglycemia risk was demonstrated, but low event counts limit certainty. The findings support active monitoring for gastrointestinal symptoms when initiating imeglimin, particularly in older patients or those with complex regimens. Clinicians should weigh the increased gastrointestinal event risk against glycemic benefits when selecting therapy and consider counseling patients about potential gastrointestinal side effects early in treatment.