Cardiovascular Events in Individuals Treated With Sulfonylureas or Dipeptidyl Peptidase 4 Inhibitors

Alexander Turchin; Lucia C Petito; Emma Hegermiller; Ryan Carnahan; Andrea DeVries; Satyender Goel; M Cecilia Lansang; Marie E McDonnell; Vinit Nair; Elisa Priest; Vincent J Willey; Alan F Kaul; Miguel A Hernán

doi:10.1001/jamanetworkopen.2025.23067

. 2025 Jul 24;8(7):e2523067. doi: 10.1001/jamanetworkopen.2025.23067

Cardiovascular Events in Individuals Treated With Sulfonylureas or Dipeptidyl Peptidase 4 Inhibitors

Alexander Turchin ^1,^2,^✉, Lucia C Petito ³, Emma Hegermiller ⁴, Ryan Carnahan ⁵, Andrea DeVries ⁶, Satyender Goel ³, M Cecilia Lansang ⁷, Marie E McDonnell ^1,², Vinit Nair ⁴, Elisa Priest ⁸, Vincent J Willey ⁹, Alan F Kaul ¹⁰, Miguel A Hernán ¹¹, for the BESTMED Study Research Group

¹Division of Endocrinology, Brigham and Women’s Hospital, Boston, Massachusetts

²Harvard Medical School, Boston, Massachusetts

³Northwestern University Feinberg School of Medicine, Chicago, Illinois

⁴Indus Consulting, Inc, Chicago, Illinois

⁵The University of Iowa College of Public Health, Iowa City

⁶Humana Healthcare Research, Inc, Louisville, Kentucky

⁷Cleveland Clinic, Cleveland, Ohio

⁸Baylor Scott and White Research Institute, Dallas, Texas

⁹Carelon Research, Wilmington, Delaware

¹⁰Medical Outcomes Management, Inc, Sharon, Massachusetts

¹¹CAUSALab, Harvard T.H. Chan School of Public Health, Boston, Massachusetts

Accepted for Publication: April 25, 2025.

Published: July 24, 2025. doi:10.1001/jamanetworkopen.2025.23067

Correction: This article was corrected on December 12, 2025, to fix an error in the byline.

^✉

Corresponding Author: Alexander Turchin, MD, MS, Division of Endocrinology, Brigham and Women’s Hospital, 221 Longwood Ave, Boston, MA 02115 (aturchin@bwh.harvard.edu).

Author Contributions: Dr Turchin had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Turchin, Petito, Carnahan, Lansang, McDonnell, Nair, Kaul, Hernán.

Acquisition, analysis, or interpretation of data: Turchin, Petito, Hegermiller, Carnahan, DeVries, Goel, Lansang, Priest, Willey, Kaul, Hernán.

Drafting of the manuscript: Turchin, Petito, Hegermiller, McDonnell, Kaul, Hernán.

Critical review of the manuscript for important intellectual content: Petito, Carnahan, DeVries, Goel, Lansang, McDonnell, Nair, Priest, Willey, Kaul, Hernán.

Statistical analysis: Petito, Hegermiller, Hernán.

Obtained funding: Turchin, McDonnell, Nair, Priest, Kaul.

Administrative, technical, or material support: Petito, Hegermiller, Carnahan, DeVries, Goel, Nair, Priest, Kaul.

Supervision: Turchin, Petito, Goel, Lansang, Kaul, Hernán.

Conflict of Interest Disclosures: Dr Turchin reported receiving grants from Eli Lilly and personal fees from Novo Nordisk and Proteomics International outside the submitted work. Dr Petito reported receiving grants from Omron Healthcare Co, Ltd (unrelated research support) outside the submitted work. Dr Lansang reported receiving grants from Abbott (investigator-initiated study) and Dexcom (investigator-initiated study) outside the submitted work. Dr Priest reported receiving grants from Boehringer Ingelheim, AstraZeneca, CSL Vifor, and Owkin outside the submitted work. Dr Willey reported being employed by Carelon Research outside the submitted work. Dr Hernán reported receiving personal fees from Cytel and ProPublica, being a consultant to Adigens Health (a company of which he owns equity), and being a member of the ADIALab Advisory Board outside the submitted work; his interests were declared, reviewed, and approved by Harvard University in accordance with its institutional compliance policies. No other disclosures were reported.

Funding/Support: This study was funded in part by the Patient-Centered Outcomes Research Institute (contract Nos. DB-2020C2-20308 and RI-MISSOURI-01-PS1).

Role of the Funder/Sponsor: The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Nonauthor Collaborators: The members of the BESTMED Study Research Group are listed in Supplement 2.

Data Sharing Statement: See Supplement 3.

^✉

Corresponding author.

PMCID: PMC12290728 PMID: 40705329

This comparative effectiveness research study compares the risk of cardiovascular events after initiation of treatment with individual sulfonylureas or dipeptidyl peptidase 4 inhibitors (DPP4is).

Key Points

Question

Do any sulfonylureas (glimepiride, glipizide, or glyburide) associated increase the risk of cardiovascular events compared with dipeptidyl peptidase 4 inhibitors (DPP4is)?

Findings

In this comparative effectiveness research study of electronic health records and claims from 48 165 individuals starting a second T2D agent after metformin, individuals treated with glipizide had a higher risk of a composite end point of myocardial infarction, ischemic stroke, heart failure hospitalization, and cardiovascular death than those treated with a DPP4i.

Meaning

These findings suggest that sulfonylureas, glipizide in particular, may not be the optimal agents in treatment of individuals with T2D at moderate cardiovascular risk.

Abstract

Importance

Sulfonylureas are commonly used to treat type 2 diabetes (T2D). Research findings on cardiovascular risk associated with sulfonylureas have been inconsistent.

Objective

To emulate a target trial that compares the risk of cardiovascular events after initiation of treatment with individual sulfonylureas or dipeptidyl peptidase 4 inhibitors (DPP4is).

Design, Setting, and Participants

This comparative effectiveness research study included individuals with T2D and moderate cardiovascular risk treated with metformin monotherapy who received care at 1 of 10 US health systems or were insured by 1 of 2 large health insurance plans between January 1, 2014, and January 1, 2023. Data were analyzed from July 2024 to March 2025.

Exposure

Initiation of treatment with a sulfonylurea (glimepiride, glipizide, or glyburide) or a DPP4i (reference category) as a second line therapy after metformin.

Main Outcomes and Measurements

The primary outcome was a 4-point composite of major adverse cardiovascular events (MACE-4): myocardial infarction, ischemic stroke, heart failure hospitalization, or cardiovascular death (from any of these conditions). The 5-year risks of each outcome were estimated.

Results

Among 48 165 eligible individuals (median [IQR] age, 61 [52-69] years; 22 674 female [47.1%]; median [IQR] hemoglobin A_1C, 7.8% [7.3%-8.5%]; median [IQR] low-density lipoprotein cholesterol, 89 mg/dL [70-112 mg/dL]), 18 147 started glipizide, 14 282 started glimepiride, 1887 started glyburide, and 13 849 started a DPP4i. Over the median (IQR) follow-up of 37 (20-64) months, 3158 individuals (6.6%) experienced a MACE-4. The estimated 5-year risks of MACE-4 were 8.1% (95% CI, 7.5%-8.7%) for DPP4i, 8.4% (95% CI, 6.8%-9.9%) for glyburide, 8.6% (95% CI, 7.9%-9.2%) for glimepiride, and 9.1% (95% CI, 8.7%-9.7%) for glipizide. Compared with DPP4is, the 5-year risk ratio of MACE-4 was 1.13 (95% CI, 1.03-1.23) for glipizide, 1.07 (95% CI, 0.96-1.16) for glimepiride, and 1.04 (95% CI, 0.83-1.24) for glyburide.

Conclusions and Relevance

In this comparative effectiveness research study of sulfonylureas vs DPP4i in patients with T2D, the risk of MACE-4 events was highest for glipizide. These findings suggest that sulfonylureas, glipizide in particular, may not be the optimal agent in treatment of individuals with T2D at moderate cardiovascular risk.

Introduction

Type 2 diabetes (T2D) is a common chronic disease whose prevalence continues to grow in the US and worldwide.^1,2 Individuals with T2D have an increased risk of adverse cardiovascular events, including coronary ischemia, stroke, and heart failure (HF).^3,4,5,6,7,8 Mitigation of cardiovascular risk is therefore an important aspect of treatment of T2D.⁹

Sulfonylureas remain a common choice of pharmacotherapy of T2D,^10,11 due in part to their low cost and glucose-lowering potency.¹² In the past, individual sulfonylureas that are no longer available in the US (eg, tolbutamide) were found to increase the risk of cardiovascular events.¹³ It is not known whether treatment with any sulfonylureas currently in clinical use increases cardiovascular risk compared with other therapeutic alternatives.

A long-term randomized clinical trial to compare individual sulfonylureas currently in use in the US would be resource and time-intensive and may never be conducted. We therefore leveraged a multicenter observational study of individuals with T2D to emulate a target trial that compares the incidence of cardiovascular events after initiation of treatment with individual sulfonylureas and dipeptidyl peptidase 4 inhibitors (DPP4is) as a second line agent after metformin.

Methods

Target Trial Specification

This comparative effectiveness research study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline and was approved by the institutional review boards of all participating institutions. The requirement of informed consent was waived due to the use of deidentified data. The elements of the protocol of the target trial are summarized in eTable 1 in Supplement 1. Individuals are eligible for a target trial if they meet the following criteria between January 1, 2014, and January 1, 2023: (1) aged 30 years or older; (2) diagnosis of T2D and no evidence of diabetes of any other type; (3) hemoglobin A_1C (HbA_1C) between 7.0% and 11% (to convert to proportion of total hemoglobin, multiply by 0.01); (4) monotherapy with metformin for at least 3 months and no prior nonmetformin outpatient diabetes therapy; (5) moderate cardiovascular risk; (5) no metastatic cancer, end-stage lung disease, end-stage liver disease, dementia, previous history of pancreatitis, previous history of medullary thyroid cancer, previous history of pyelonephritis, or estimated glomerular filtration rate (eGFR) less than 45 mL/min/1.73 m²; (6) at least 12 months of engagement with their health system or continuous enrollment at their health insurance plan to ensure availability of historical information on baseline characteristics and to increase the probability of continued engagement; and (7) a clinician’s determination that a second-line treatment needs to be initiated. Moderate cardiovascular risk is defined as: (1) no history of diagnosis of atherosclerotic cardiovascular disease or hospitalization for HF and either (2) males aged 35 years and older and females aged 45 years or older or (3) males aged 30 to 34 years and females aged 30 to 44 years with hypertension, hyperlipidemia, retinopathy, kidney disease, or neuropathy.¹⁴

The treatment strategies of interest include immediate initiation of exactly 1 of the following medications after metformin: (1) glimepiride, (2) glipizide, (3) glyburide, or (4) any DPP4i (alogliptin, linagliptin, saxagliptin, or sitagliptin). Individuals can subsequently continue or discontinue study medications at their discretion. DPP4is are chosen as the reference category because multiple randomized clinical trials have shown no difference in all or most cardiovascular events between each of the DPP4is and placebo.^{15,16,17,18,19,20} In a variation of this target trial, saxagliptin was excluded from the DPP4i category because a previous randomized trial showed an increased risk of HF in individuals treated with saxagliptin.¹⁹

The outcome is the 4-point composite of major adverse cardiovascular events (MACE-4): (1) myocardial infarction (MI), (2) ischemic stroke, (3) HF hospitalization, or (4) death from any of these conditions. Individual components of the composite end point are also outcomes of interest. Follow-up starts at assignment and ended at the outcome, loss to follow-up (12 months without health care encounters, vital signs, laboratory measurements, or loss of membership with the health insurance plan), or at study end (July 1, 2023). For MACE-4 and cardiovascular death, noncardiovascular mortality is a competing event; for MI, stroke, and HF, the competing event is non-MI, nonstroke, and non-HF mortality, respectively. The causal contrast of interest is the intention-to-treat effect, defined as the total effect.²¹

In the intention-to-treat analysis, the 5-year cumulative incidence (referred to as risk) of each outcome is compared between treatment strategies via ratios and differences.²² Risks for each outcome are estimated nonparametrically via a Kaplan-Meier estimator or parametrically via a pooled logistic regression model²³ for the monthly probability of that outcome with the following covariates: month of follow-up (modeled as a quadratic polynomial), indicators for treatment strategy, and product (interaction) terms between linear month and the treatment indicators. Prespecified prognostic factors are also included as covariates in the model, with the risk estimates standardized to the distribution of these factors. Subgroup analyses for the MACE-4 outcome are conducted by sex, age, baseline HbA_1C, and baseline eGFR. Nonparametric bootstrap with 500 resamples are used to calculate 95% CIs for all estimates.

Target Trial Emulation

We emulated the target trial using data from the Observational Evaluation of Second Line Therapy Medications in Diabetes (BESTMED) study, which includes individuals with T2D who were treated at 1 of 10 health systems or had their health insurance at 1 of 2 national health insurance plans in the US. Data were harmonized across study sites using the PCORnet Common Data Model.²⁴ Individuals who had data at more than 1 study site were identified using a privacy-preserving linkage solution²⁵ and their records were merged for the analysis.

The target trial focused on individuals at moderate cardiovascular risk in part to reduce the confounding that could result if any of the study medications were less likely to be prescribed to individuals at high cardiovascular risk. Information on individuals’ baseline characteristics, analytical covariates, and outcomes was obtained from electronic health records of participating health systems and claims data of participating health insurance plans (combined with electronic health record and laboratory data of their data partners). Race and ethnicity information was collected via via patient registration, available through the electronic health record and health insurance plan data and was included to account for a known disparity. Race categories included Asian, Black, White, and other (ie, persons having origins in any of the original peoples of North and South America [including Central America] and who maintain tribal affiliation or community attachment, persons having origins in any of the original peoples of Hawaii, Guam, Samoa, or other Pacific Islands, and persons categorized with multiple races or any race not otherwise specified in their medical records). Ethnicity categories included Hispanic and non-Hispanic. Information on medications was obtained from prescription records (health care delivery organizations) and prescription fills (health insurance plans). Treatment adherence was assessed using both prescription and dispensing records. Prescriptions were considered discontinued 1 year after the last prescription (based on the typical refill duration in the US). Dispensing treatment windows were calculated based on the date of drug dispensing and the recorded days’ supply. Information on the cause of death was obtained from the National Death Index for individuals with evidence of death in health insurance plan or electronic health record data or a nationwide mortality database (derived from public and commercial data sources using a previously validated methodology²⁶).

Statistical Analysis

The 5-year risks of each outcome were estimated as described for the target trial. The pooled logistic regression models included the following baseline covariates: demographics (age, sex, and health insurance); community factors (proportion treated with brand name vs generic anticoagulants at the individual’s site and Social Deprivation Index²⁷ in the individual’s zip code); laboratory values (HbA_1C, eGFR, low density lipoprotein cholesterol); excess weight (body mass index [BMI; calculated as weight in kilograms divided by height in meters squared] ≥35, or International Statistical Classification of Diseases and Related Health Problems, Tenth Revision [ICD-10] code indicating class II-III obesity); proxies of comorbidity load (number of nondiabetes medications the patient was taking, number of outpatient visits, and number of distinct laboratory tests in the previous year and hospitalization in the previous 3 months); history of HF; history of metabolic dysfunction–associated steatotic liver disease and metabolic dysfunction–associated steatohepatitis; Charlson comorbidity index; and treatment with systemic glucocorticoids, statins, or antiplatelet drugs. eTable 2 in Supplement 1 describes these variables and the methods used to address missing data.

We also conducted several sensitivity analyses. First, to explore the impact of confounding adjustment on our estimates, we obtained estimates adjusted for age only. Second, to explore the impact of the handling of competing events, we censored individuals when they experience the competing event (death from any cause). Third, to explore the sensitivity of our analysis to our method of addressing missing data, we added indicators for missingness (details in eTable 3 in Supplement 1). Fourth, to explore the impact of adjustment for hypertension, a prevalent co-occurring comorbidity, we added hypertension (identified by ICD-10 codes and blood pressure, as available) as a confounder in the outcome model. Finally, to assess that the impact of including individuals with history of HF or atrial fibrillation, we conducted an analysis excluding individuals who had these conditions at baseline. All statistical analyses were performed in R version 4.4.0 from July 2024 to March 2025.

Results

Study Population

Of 48 165 eligible individuals (median [IQR] age, 61 [52-69] years; 22 674 female [47.1%]; median [IQR] HbA_1C, 7.8% [7.3%-8.5%]; median [IQR] low-density lipoprotein cholesterol, 89 mg/dL [70-112 mg/dL]), 18 147 initiated glipizide, 14 282 initiated glimepiride, 1887 initiated glyburide, and 13 849 initiated a DPP4i (Figure 1). Differences in the median values of baseline characteristics between individuals treated with any of the 3 sulfonylureas vs DPP4i (the reference category) did not exceed 2 years for age, 0.2% for HbA_1C, 7 mg/dL (to convert to millimoles per liter, multiply by 0.0259) for low-density lipoprotein cholesterol, and 3 mL/min/1.73 m² for eGFR (Table 1). The median (IQR) time these individuals continued treatment with the study drug was 13 (6-29) months for glipizide, 14 (6-32) months for glimepiride, 11 (3-20) months for glyburide, and 12 (5-23) months for DPP4is. See eTable 4 in Supplement 1 for baseline characteristics and study medication prescribing for eligible individuals by study site.

Figure 1. — ASCVD indicates atherosclerotic cardiovascular disease; DPP4i, dipeptidyl peptidase-4 inhibitors; eGFR, estimated glomerular filtration rate; HbA_1C, hemoglobin A_1C; UTI, urinary tract infection. To convert HbA_1C to proportion of total hemoglobin, multiply by 0.01; to convert glucose to millimoles per liter, multiply by 0.0555.

Table 1. Baseline Characteristics of Eligible Individuals in the Observational Evaluation of Second Line Therapy Medications in Diabetes Study, 2014-2023.

Characteristic	Participants, No. (%)					SMD (vs DPP4i)
Characteristic	Overall (N = 48 165)	DPP4i (n = 13 849)	Glimepiride (n = 14 282)	Glipizide (n = 18 147)	Glyburide (n = 1887)	Glimepiride	Glipizide	Glyburide
Age, median (IQR), y	61 (52-69)	60 (52- 68)	62 (53-69)	61 (52-69)	58 (50-66)	0.13	0.07	0.18
Sex
Female	22 674 (47.1)	6813 (49.2)	6484 (45.4)	8444 (46.5)	933 (49.4)	0.08	0.05	0.00
Male	25 491 (52.9)	7036 (50.8)	7798 (54.6)	9703 (53.5)	954 (50.6)	0.08	0.05	0.00
Race
Asian	1108 (2.3)	300 (2.2)	305 (2.1)	469 (2.6)	34 (1.8)	0.37	0.34	0.22
Black	3824 (7.9)	1053 (7.6)	1032 (7.2)	1580 (8.7)	159 (8.4)
White	26 132 (54.3)	6088 (44.0)	8613 (60.3)	10 433 (57.5)	998 (52.9)
Other^a	2107 (4.4)	535 (3.9)	613 (4.3)	868 (4.8)	91 (4.8)
No information	14 994 (31.1)	5873 (42.4)	3719 (26.0)	4797 (26.4)	605 (32.1)
Ethnicity
Hispanic	2623 (5.4)	650 (4.7)	757 (5.3)	1059 (5.8)	157 (8.3)	0.37	0.35	0.24
Non-Hispanic	30 407 (63.1)	7221 (52.1)	9812 (68.7)	12 263 (67.6)	1111 (58.9)
No information	15 135 (31.4)	5978 (43.2)	3713 (26.0)	4825 (26.6)	619 (32.8)
Private health insurance	38 487 (79.9)	11 738 (84.8)	11 025 (77.2)	14 312 (78.9)	1412 (74.8)	0.19	0.15	0.25
SDI percentile
1-30	13 821 (32.2)	3761 (32.1)	4015 (31.0)	5557 (33.4)	488 (29.4)	0.03	0.05	0.07
31-60	12 736 (29.6)	3376 (28.8)	3913 (30.3)	4979 (29.9)	468 (28.2)
61-100	16 409 (38.2)	4582 (39.1)	5004 (38.7)	6120 (36.7)	703 (42.4)
Hospitalization in previous 3 mo	1256 (2.6)	389 (2.8)	361 (2.5)	457 (2.5)	49 (2.6)	0.02	0.02	0.01
HbA_1C, median (IQR), %	7.8 (7.3-8.5)	7.7 (7.3-8.4)	7.8 (7.4-8.5)	7.8 (7.4-8.6)	7.9 (7.4-8.7)	0.06	0.11	0.21
eGFR, median (IQR), mL/min/1.73 m²	91.0 (76.5-101.8)	91.6 (77.2-102.3)	90.5 (76.2-101.1)	90.5 (76.0-101.4)	94.4 (79.1-104.9)	0.05	0.06	0.13
LDL cholesterol, median (IQR), mg/dL	89 (70-112)	88 (69-111)	89 (70-112)	90 (70-113)	95 (76-118)	0.01	0.04	0.20
BMI, median (IQR)^b	34.0 (29.8-39.0)	33.1 (29.0-38.6)	34.0 (30.0-39.0)	34.0 (30.0-39.0)	33.8 (30.0-39.0)	0.06	0.09	0.05
CCI, mean (SD)	2.13 (1.77)	2.11 (1.76)	2.12 (1.77)	2.16 (1.79)	1.99 (1.72)	0.00	0.03	0.07
CCI ≥5	1685 (3.5)	507 (3.7)	487 (3.4)	637 (3.5)	54 (2.9)	0.01	0.01	0.04
Smoking status
Nonsmoker	8424 (17.5)	1944 (14.0)	2575 (18.0)	3532 (19.5)	373 (19.8)	0.29	0.28	0.26
Current smoker	9068 (18.8)	1808 (13.1)	3162 (22.1)	3734 (20.6)	364 (19.3)
No information	30 673 (63.7)	10 097 (72.9)	8545 (59.8)	10 881 (60.0)	1150 (60.9)
Hypertension	38 516 (80.0)	11 201 (80.9)	11 456 (80.2)	14 419 (79.5)	1440 (76.3)	0.02	0.04	0.11
Heart failure	2828 (5.9)	754 (5.4)	851 (6.0)	1111 (6.1)	112 (5.9)	0.02	0.03	0.02
Atrial fibrillation	1471 (3.1)	415 (3.0)	453 (3.2)	560 (3.1)	43 (2.3)	0.01	0.00	0.04
Hypoglycemia in the previous 5 y	1488 (3.1)	444 (3.2)	417 (2.9)	570 (3.1)	57 (3.0)	0.02	0.00	0.01
Statin therapy	33 362 (69.3)	9572 (69.1)	9879 (69.2)	12 748 (70.2)	1163 (61.6)	0.00	0.03	0.16
Antiplatelet therapy	4564 (9.5)	1125 (8.1)	1306 (9.1)	1997 (11.0)	136 (7.2)	0.04	0.10	0.03

Open in a new tab

Abbreviations: BMI, body mass index; CCI, Charlson Comorbidity Index; DPP4i, dipeptidyl peptidase-4 inhibitor; eGFR, estimated glomerular filtration rate; HbA_1C, hemoglobin A_1C; LDL, low-density lipoprotein; SDI, Social Deprivation Index; SMD, standardized mean difference.

SI conversion factor: To convert HbA_1C to proportion of total hemoglobin, multiply by 0.01; LDL cholesterol to millimoles per liter, multiply by 0.0259.

^{^a}

Persons having origins in any of the original peoples of North and South America (including Central America) and who maintain tribal affiliation or community attachment, persons having origins in any of the original peoples of Hawaii, Guam, Samoa, or other Pacific Islands, and persons categorized with multiple races or any race not otherwise specified in their medical records.

^{^b}

Calculated as weight in kilograms divided by height in meters squared.

Cardiovascular Events

Over a median (IQR) follow-up of 37 (20-64) months, 3158 (6.6%) individuals had a MACE-4, 1082 (2.2%) had an MI, 1260 (2.6%) had an ischemic stroke, 1460 (3.0%) had an HF hospitalization, and 206 (0.4%) died from any of these conditions, while 1475 individuals (3.1%) died from noncardiovascular causes and 17 168 individuals (35.6%) were lost to follow-up. See eTable 5 in Supplement 1 for number of events by treatment group and cardiovascular outcome.

The estimated 5-year risks of MACE-4 were 8.1% (7.5%-8.7%) for DPP4is, 8.4% (6.8%-9.9%) for glyburide, 8.6% (7.9%-9.2%) for glimepiride, and 9.1% (8.7%-9.7%) for glipizide. Compared with DPP4is, the 5-year risk ratios of MACE-4 were 1.04 (95% CI, 0.83-1.24) glyburide, 1.07 (95% CI, 0.96-1.16) for glimepiride, and 1.13 (95% CI, 1.03-1.23) for glipizide (Figure 2 and Table 2). Point estimates were similar across subgroups defined by sex, age, HbA_1C, and eGFR, although confidence intervals were wide (Figure 3).

Figure 2. — Both fatal and nonfatal events are included in panels B, C, and D. In panel C, the blue line (glipizide) cannot be easily seen because it almost fully overlaps with the orange line (glimepiride). DPP4i indicates dipeptidyl peptidase-4 inhibitor; MACE-4, major adverse cardiovascular events (4-point composite).

Table 2. Estimated 5-Year Risks of Cardiovascular Outcomes by Treatment Group, Observational Evaluation of Second Line Therapy Medications in Diabetes Study 2014-2023^a.

Outcome by treatment group	No. eligible	No. of events at 5-y	5-y Risk, % (95% CI)	Risk difference, % (95% CI)	Risk ratio (95% CI)	Average hazard ratio (95% CI)
4-point MACE
DPP4i	13 849	663	8.1 (7.5 to 8.7)	0 [Reference]	1 [Reference]	1 [Reference]
Glimepiride	14 282	824	8.6 (7.9 to 9.2)	0.54 (−0.35 to 1.25)	1.07 (0.96 to 1.16)	1.05 (0.92 to 1.17)
Glipizide	18 147	1021	9.1 (8.7 to 9.7)	1.06 (0.23 to 1.80)	1.13 (1.03 to 1.23)	1.14 (1.01 to 1.27)
Glyburide	1887	83	8.4 (6.8 to 9.9)	0.29 (−1.37 to 1.84)	1.04 (0.83 to 1.24)	1.02 (0.77 to 1.29)
Stroke^b
DPP4i	13 879	266	3.1 (2.8 to 3.5)	0 [Reference]	1 [Reference]	1 [Reference]
Glimepiride	14 312	349	3.5 (3.2 to 3.9)	0.43 (−0.02 to 0.88)	1.14 (0.99 to 1.31)	1.13 (0.94 to 1.38)
Glipizide	18 192	388	3.5 (3.2 to 3.8)	0.42 (−0.05 to 0.88)	1.14 (0.98 to 1.31)	1.13 (0.96 to 1.36)
Glyburide	1893	27	2.7 (1.8 to 4.0)	−0.36 (−1.39 to 0.9)	0.88 (0.56 to 1.29)	0.86 (0.49 to 1.30)
Heart failure hospitalization or fatal heart failure
DPP4i	13 881	290	3.7 (3.3 to 4.1)	0 [Reference]	1 [Reference]	1 [Reference]
Glimepiride	14 317	372	4 (3.7 to 4.4)	0.3 (−0.21 to 0.82)	1.08 (0.95 to 1.24)	1.07 (0.90 to 1.28)
Glipizide	18 198	458	4.2 (3.8 to 4.6)	0.47 (−0.04 to 0.96)	1.12 (0.99 to 1.26)	1.12 (0.94 to 1.30)
Glyburide	1892	43	4.3 (3.3 to 5.6)	0.56 (−0.57 to 1.87)	1.15 (0.85 to 1.53)	1.24 (0.84 to 1.77)
Myocardial infarction^b
DPP4i	13 880	231	3.0 (2.6 to 3.3)	0 [Reference]	1 [Reference]	1 [Reference]
Glimepiride	14 316	256	2.8 (2.5 to 3.2)	−0.13 (−0.72 to 0.30)	0.96 (0.78 to 1.11)	0.97 (0.75 to 1.16)
Glipizide	18 203	348	3.2 (2.9 to 3.6)	0.27 (−0.26 to 0.80)	1.09 (0.92 to 1.29)	1.07 (0.89 to 1.31)
Glyburide	1896	31	3.1 (2.2 to 4.2)	0.11 (−0.93 to 1.10)	1.04 (0.71 to 1.34)	0.96 (0.62 to 1.44)
Death from cardiovascular causes
DPP4i	13 898	34	0.5 (0.3 to 0.6)	0 [Reference]	1 [Reference]	1 [Reference]
Glimepiride	14 332	64	0.6 (0.5 to 0.8)	0.18 (−0.05 to 0.39)	1.39 (0.92 to 2.19)	1.67 (1.03 to 2.91)
Glipizide	18 226	50	0.5 (0.3 to 0.6)	0.02 (−0.17 to 0.21)	1.05 (0.74 to 1.74)	0.88 (0.54 to 1.64)
Glyburide^c	1897	4	0.7 (0.3 to 1.1)	0.28 (−0.11 to 0.71)	1.71 (0.73 to 3.16)	1.71 (0.73 to 3.18)

Open in a new tab

Abbreviations: DPP4i, dipeptidyl peptidase-4 inhibitor; MACE, major adverse cardiovascular events.

^{^a}

Adjusted analysis for cardiovascular death could not be carried out for glyburide due to the small number of events.

^{^b}

Both fatal and nonfatal events are included.

^{^c}

Due to limited sample size, estimates are partially adjusted for age only.

Figure 3. — DPP4i indicates dipeptidyl peptidase-4 inhibitor; eGFR, estimated glomerular filtration rate; HbA_1C, hemoglobin A_1C. To convert HbA_1C to proportion of total hemoglobin, multiply by 0.01.

The estimated 5-year risk of components of the MACE-4 outcome ranged from 0.5% (95% CI, 0.3%-0.6%) for cardiovascular death for individuals treated with DPP4is or glipizide to 4.3% (95% CI, 3.3%-5.6%) for HF hospitalization for individuals treated with glyburide (Table 2). The estimated 5-year RRs for these outcomes (Figure 2) were broadly consistent with those for MACE-4. See eFigure 1 in Supplement 1 for unadjusted survival curves for MACE-4 by treatment group and eFigure 2 in Supplement 1 for subgroup analyses for risk differences.

Sensitivity Analyses

When 367 individuals taking saxagliptin were excluded from the DPP4i category, 5-year risk ratios of the primary outcome (compared with DPP4is) were 1.01 (95% CI, 0.79-1.30) for glyburide, 1.05 (95% CI, 0.92-1.18) for glimepiride, and 1.13 (95% CI, 1.01-1.26) for glipizide. The results of the analyses by individual end points were broadly similar (eTable 6 in Supplement 1). In particular, 5-year risk ratios for HF hospitalization (compared with DPP4is) were (1.07; 95% CI, 0.94-1.23) for glimepiride, 1.11 (95% CI, 0.98-1.30) for glipizide, and 1.13 (95% CI, 0.81-1.58) for glyburide. The results of other sensitivity analyses were also similar to those from the main analysis (eTable 7 in Supplement 1).

Discussion

In this nationwide comparative effectiveness research study of individuals with T2D at moderate cardiovascular risk, we estimated that the risk of cardiovascular events increased after initiation of glipizide compared with DPP4is. Glipizide—currently the most commonly used sulfonylurea in the US¹¹—was estimated to increase the risk of MACE-4 by approximately 13%, with risk increases of 13 ± 10% highly compatible with the data under standard conventional criteria. The risk ratio for glimepiride was 1.07, but the estimates for MACE-4 and its components were imprecise. The estimates for glyburide were closer to the null and more imprecise.

A strength of our study is the comparison of individual sulfonylureas with DPP4is, which are a clinical alternative to sulfonylureas that have been demonstrated to be neutral for cardiovascular events compared with placebo. Only one previous high-quality study—the Cardiovascular Outcome Study of Linagliptin vs Glimepiride in Type 2 Diabetes clinical trial—compared a sulfonylurea (glimepiride) to a DPP4i (linagliptin).²⁸ This trial, involving approximately 6000 participants, estimated a hazard ratio for cardiovascular events (not including HF) of 0.98 (95% CI, 0.84-1.14)—a finding compatible with our results.²⁸ Other investigations did not compare sulfonylureas with DPP4is; a randomized trial²⁹ found a hazard ratio of 1.31 (95% CI, 1.05-1.69) for glipizide compared with metformin and a meta-analysis³⁰ of 3 randomized trials estimated a relative risk of 0.84 (95% CI, 0.56-1.26) for glyburide compared with other sulfonylureas and insulin. Results from a meta-analysis³¹ that found mortality differences between sulfonylureas are hard to interpret because of the inclusion of observational studies with questionable methodology (eg, comparing prevalent users).

A 13% relative increase in cardiovascular risk, like the one estimated in this study, is potentially clinically impactful. Decreases in cardiovascular risk of this magnitude have generally been thought to be sufficient for approval of medications for prevention of cardiovascular events, particularly in high-risk populations.^32,33 The estimated absolute increase of 1.1% in risk of cardiovascular events was modest in this population at moderate risk but could be greater in a high-risk population. This study’s findings therefore suggest caution in prescribing glipizide to individuals at moderate cardiovascular risk or established atherosclerotic cardiovascular disease.

Treatment with sulfonylureas also has other risks, in particular hypoglycemia. The risk of hypoglycemia is thought to be higher for glyburide than for shorter acting sulfonylureas.^30,34 Choice between individual sulfonylureas should therefore be made in an individualized manner, balancing the risks and benefits based on the particular patient’s health profile.

The potentially different cardiovascular risk between individual sulfonylureas in our study underscores the importance of evaluating each agent in a particular pharmacological class on its own merits. In multiple instances, individual pharmaceuticals were found to have a different risk-benefit profile than other medications in the same class.^35,36,37 Among sulfonylureas, in particular, tolbutamide was found to have a greater risk of cardiovascular events, and glyburide was found to have a higher risk of hypoglycemia.^13,30 These findings caution against extrapolating results of studies evaluating a single agent to the entire class, and temper concerns about the use of research and development resources for the development of multiple agents in the same pharmacological class.³⁸

In addition to pancreatic β cells, sulfonylurea receptors (SURs) are also found on cardiomyocytes (SUR 2A) and vascular smooth muscle cells (SUR 2B). Decrease in cardiac ischemic preconditioning after exposure to sulfonylureas has been found in animal experiments.³⁹ This effect has been proposed as a possible mechanism for an increased cardiovascular risk from sulfonylureas.⁴⁰ However, glipizide is not thought to have a greater affinity for either SUR 2A or SUR 2B vs pancreatic SURs (SUR 1) compared with other sulfonylureas.⁴¹ Furthermore, this mechanism would not explain possible differential effects on stroke. Further research is therefore needed to establish pathophysiological mechanisms of the increased incidence of cardiovascular events with glipizide.

Strengths and Limitations

The present study had a number of strengths. Our explicit emulation of a target trial reduces the risk of biases commonly found in observational studies, including immortal time and selection biases.^42,43 The study population was socioeconomically similar to the US population, as indicated by the Social Deprivation Index distribution. Unlike investigations that rely solely on claims data, all individuals included in this analysis had baseline measurements of glycemic control and kidney function and most had baseline measurements of low-density lipoprotein cholesterol—all important cardiovascular risk factors. Incorporation of National Death Index data allowed us to include specific causes of death as outcomes. Finally, utilization of a privacy-preserving patient linkage permitted merger of records for the same individuals across study sites, minimizing duplicate and missing data.

Findings of this study should be interpreted in the context of its limitations. It is not possible to completely exclude confounding by indication if 1 or more of the study medications were less likely to be prescribed to individuals at higher cardiovascular risk. However, the distribution of most of the known cardiovascular risk factors was similar across the comparison groups, even before any adjustments, which makes it unlikely that unmeasured risk factors that are correlated with the measured ones (eg, duration of diabetes or smoking status) were imbalanced across groups. Second, the median follow-up of approximately 3 years may be too short to detect further differences between study medications. Third, the studied drugs reflect treatment patterns in the US at the time of the study; gliclazide was not in use, and few individuals were taking glyburide. The exclusion of saxagliptin, previously found to increase the risk for HF compared with placebo,¹⁹ from the DPP4i group did not affect the results. Fourth, racial and ethnic diversity was limited and the majority of study individuals had private health insurance. Although study findings may not generalize to populations with different access to treatment, all sulfonylureas evaluated in the study had low-cost generics available throughout the study period. The analysis did not evaluate different doses of sulfonylureas, and it is possible that the differences in cardiovascular risk between sulfonylureas could be dose dependent.

Conclusions

In this comparative effectiveness research study of individuals with T2D, individuals treated with glipizide as a second agent after metformin monotherapy had a higher risk of MACE-4 events than individuals treated with DPP4is. These findings suggest that glipizide may not be the optimal agent in treatment of individuals with T2D at moderate cardiovascular risk.

Supplement 1.

eTable 1. Specification of the Target Trial and Its Emulation in the BESTMED Database

eTable 2. Inclusion of Variables in the MACE-4 Model

eTable 3. Summary of Missing Baseline Covariate Data

eTable 4. Baseline Characteristics of and Study Medication Prescribing for Eligible Individuals by Study Site

eTable 5. Number of Events by Treatment Group and Type of Cardiovascular Outcome

eTable 6. Sensitivity Analysis Excluding Saxagliptin

eFigure 1. Unadjusted Survival Curves for MACE-4 by Treatment Group

eFigure 2. Subgroup Analyses for the Primary Outcome: Risk Differences

eTable 7. Estimated 5-Year Risks of 4-Point MACE by Treatment Group: Sensitivity Analyses

jamanetwopen-e2523067-s001.pdf^{(618.8KB, pdf)}

Supplement 2.

Nonauthor Collaborators

jamanetwopen-e2523067-s002.pdf^{(132KB, pdf)}

Supplement 3.

Data Sharing Statement

jamanetwopen-e2523067-s003.pdf^{(14.5KB, pdf)}

References

1.Sun H, Saeedi P, Karuranga S, et al. IDF diabetes atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract. 2022;183:109119. doi: 10.1016/j.diabres.2021.109119 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Xu G, Liu B, Sun Y, et al. Prevalence of diagnosed type 1 and type 2 diabetes among US adults in 2016 and 2017: population based study. BMJ. 2018;362:k1497. doi: 10.1136/bmj.k1497 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Mulnier HE, Seaman HE, Raleigh VS, et al. Risk of myocardial infarction in men and women with type 2 diabetes in the UK: a cohort study using the General Practice Research Database. Diabetologia. 2008;51(9):1639-1645. doi: 10.1007/s00125-008-1076-y [DOI] [PubMed] [Google Scholar]
4.Haffner SM, Lehto S, Rönnemaa T, Pyörälä K, Laakso M. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med. 1998;339(4):229-234. doi: 10.1056/NEJM199807233390404 [DOI] [PubMed] [Google Scholar]
5.Mulnier HE, Seaman HE, Raleigh VS, et al. Risk of stroke in people with type 2 diabetes in the UK: a study using the General Practice Research Database. Diabetologia. 2006;49(12):2859-2865. doi: 10.1007/s00125-006-0493-z [DOI] [PubMed] [Google Scholar]
6.Janghorbani M, Hu FB, Willett WC, et al. Prospective study of type 1 and type 2 diabetes and risk of stroke subtypes: the Nurses’ Health Study. Diabetes Care. 2007;30(7):1730-1735. doi: 10.2337/dc06-2363 [DOI] [PubMed] [Google Scholar]
7.Nichols GA, Gullion CM, Koro CE, Ephross SA, Brown JB. The incidence of congestive heart failure in type 2 diabetes: an update. Diabetes Care. 2004;27(8):1879-1884. doi: 10.2337/diacare.27.8.1879 [DOI] [PubMed] [Google Scholar]
8.Dunlay SM, Givertz MM, Aguilar D, et al. ; American Heart Association Heart Failure and Transplantation Committee of the Council on Clinical Cardiology; Council on Cardiovascular and Stroke Nursing; Heart Failure Society of America . Type 2 diabetes mellitus and heart failure, a scientific statement from the American Heart Association and Heart Failure Society of America. J Card Fail. 2019;25(8):584-619. doi: 10.1016/j.cardfail.2019.05.007 [DOI] [PubMed] [Google Scholar]
9.American Diabetes Association Professional Practice Committee . 9. Pharmacologic approaches to glycemic treatment: standards of care in diabetes—2024. Diabetes Care. 2024;47(suppl 1):S158-S178. doi: 10.2337/dc24-S009 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Neilson LM, Munshi KD, Peasah SK, et al. Changes in type 2 diabetes medication utilization and costs in the United States, 2014–2019. Med Care. 2021;59(9):789-794. doi: 10.1097/MLR.0000000000001597 [DOI] [PubMed] [Google Scholar]
11.Shin JI, Echouffo-Tcheugui JB, Fang M, Grams ME, Selvin E. Trends in use of sulfonylurea types among us adults with diabetes: NHANES 1999-2020. J Gen Intern Med. 2023;38(8):2009-2010. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Mishriky BM, Cummings DM, Tanenberg RJ. The efficacy and safety of DPP4 inhibitors compared to sulfonylureas as add-on therapy to metformin in patients with type 2 diabetes: a systematic review and meta-analysis. Diabetes Res Clin Pract. 2015;109(2):378-388. doi: 10.1016/j.diabres.2015.05.025 [DOI] [PubMed] [Google Scholar]
13.Miller M, Knatterud G, Hawkins B. A study of the effects of hypoglycemic agents on vascular complications in patients with adult on-set diabetes. Diabetes. 1976;25(12):1129-1153. doi: 10.2337/diab.25.12.1129 [DOI] [PubMed] [Google Scholar]
14.Bertoluci MC, Rocha VZ. Cardiovascular risk assessment in patients with diabetes. Diabetol Metab Syndr. 2017;9:25. doi: 10.1186/s13098-017-0225-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Green JB, Bethel MA, Armstrong PW, et al. ; TECOS Study Group . Effect of sitagliptin on cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2015;373(3):232-242. doi: 10.1056/NEJMoa1501352 [DOI] [PubMed] [Google Scholar]
16.Engel SS, Golm GT, Shapiro D, Davies MJ, Kaufman KD, Goldstein BJ. Cardiovascular safety of sitagliptin in patients with type 2 diabetes mellitus: a pooled analysis. Cardiovasc Diabetol. 2013;12:3. doi: 10.1186/1475-2840-12-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Zannad F, Cannon CP, Cushman WC, et al. ; EXAMINE Investigators . Heart failure and mortality outcomes in patients with type 2 diabetes taking alogliptin versus placebo in EXAMINE: a multicentre, randomised, double-blind trial. Lancet. 2015;385(9982):2067-2076. doi: 10.1016/S0140-6736(14)62225-X [DOI] [PubMed] [Google Scholar]
18.White WB, Cannon CP, Heller SR, et al. ; EXAMINE Investigators . Alogliptin after acute coronary syndrome in patients with type 2 diabetes. N Engl J Med. 2013;369(14):1327-1335. doi: 10.1056/NEJMoa1305889 [DOI] [PubMed] [Google Scholar]
19.Scirica BM, Bhatt DL, Braunwald E, et al. ; SAVOR-TIMI 53 Steering Committee and Investigators . Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus. N Engl J Med. 2013;369(14):1317-1326. doi: 10.1056/NEJMoa1307684 [DOI] [PubMed] [Google Scholar]
20.Rosenstock J, Perkovic V, Johansen OE, et al. ; CARMELINA Investigators . Effect of linagliptin vs placebo on major cardiovascular events in adults with type 2 diabetes and high cardiovascular and renal risk: the CARMELINA randomized clinical trial. JAMA. 2019;321(1):69-79. doi: 10.1001/jama.2018.18269 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Young JG, Stensrud MJ, Tchetgen Tchetgen EJ, Hernán MA. A causal framework for classical statistical estimands in failure-time settings with competing events. Stat Med. 2020;39(8):1199-1236. doi: 10.1002/sim.8471 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Toh S, Hernández-Díaz S, Logan R, Robins JM, Hernán MA. Estimating absolute risks in the presence of nonadherence: an application to a follow-up study with baseline randomization. Epidemiology. 2010;21(4):528-539. doi: 10.1097/EDE.0b013e3181df1b69 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Thompson WA Jr. On the treatment of grouped observations in life studies. Biometrics. 1977;33(3):463-470. doi: 10.2307/2529360 [DOI] [PubMed] [Google Scholar]
24.PCORnet . Common data model. Accessed March 26, 2025. https://pcornet.org/data/common-data-model/
25.Mirel LB, Resnick DM, Aram J, Cox CS. A methodological assessment of privacy preserving record linkage using survey and administrative data. Stat J IAOS. 2022;38(2):413-421. doi: 10.3233/SJI-210891 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Curtis MD, Griffith SD, Tucker M, et al. Development and validation of a high-quality composite real-world mortality endpoint. Health Serv Res. 2018;53(6):4460-4476. doi: 10.1111/1475-6773.12872 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Robert Graham Center. Social Deprivation Index (SDI). Accessed June 16, 2025. https://www.graham-center.org/maps-data-tools/social-deprivation-index.html
28.Rosenstock J, Kahn SE, Johansen OE, et al. ; CAROLINA Investigators . Effect of linagliptin vs glimepiride on major adverse cardiovascular outcomes in patients with type 2 diabetes: the CAROLINA randomized clinical trial. JAMA. 2019;322(12):1155-1166. doi: 10.1001/jama.2019.13772 [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Hong J, Zhang Y, Lai S, et al. ; SPREAD-DIMCAD Investigators . Effects of metformin versus glipizide on cardiovascular outcomes in patients with type 2 diabetes and coronary artery disease. Diabetes Care. 2013;36(5):1304-1311. doi: 10.2337/dc12-0719 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Gangji AS, Cukierman T, Gerstein HC, Goldsmith CH, Clase CM. A systematic review and meta-analysis of hypoglycemia and cardiovascular events: a comparison of glyburide with other secretagogues and with insulin. Diabetes Care. 2007;30(2):389-394. doi: 10.2337/dc06-1789 [DOI] [PubMed] [Google Scholar]
31.Simpson SH, Lee J, Choi S, Vandermeer B, Abdelmoneim AS, Featherstone TR. Mortality risk among sulfonylureas: a systematic review and network meta-analysis. Lancet Diabetes Endocrinol. 2015;3(1):43-51. doi: 10.1016/S2213-8587(14)70213-X [DOI] [PubMed] [Google Scholar]
32.Nissen SE, Lincoff AM, Brennan D, et al. ; CLEAR Outcomes Investigators . Bempedoic acid and cardiovascular outcomes in statin-intolerant patients. N Engl J Med. 2023;388(15):1353-1364. doi: 10.1056/NEJMoa2215024 [DOI] [PubMed] [Google Scholar]
33.Steg PG, Bhatt DL, Simon T, et al. ; THEMIS Steering Committee and Investigators . Ticagrelor in patients with stable coronary disease and diabetes. N Engl J Med. 2019;381(14):1309-1320. doi: 10.1056/NEJMoa1908077 [DOI] [PubMed] [Google Scholar]
34.Schopman JE, Simon AC, Hoefnagel SJ, Hoekstra JB, Scholten RJ, Holleman F. The incidence of mild and severe hypoglycaemia in patients with type 2 diabetes mellitus treated with sulfonylureas: a systematic review and meta-analysis. Diabetes Metab Res Rev. 2014;30(1):11-22. doi: 10.1002/dmrr.2470 [DOI] [PubMed] [Google Scholar]
35.Pfeffer MA, Claggett B, Diaz R, et al. ; ELIXA Investigators . Lixisenatide in patients with type 2 diabetes and acute coronary syndrome. N Engl J Med. 2015;373(23):2247-2257. doi: 10.1056/NEJMoa1509225 [DOI] [PubMed] [Google Scholar]
36.Cannon CP, Pratley R, Dagogo-Jack S, et al. ; VERTIS CV Investigators . Cardiovascular outcomes with ertugliflozin in type 2 diabetes. N Engl J Med. 2020;383(15):1425-1435. doi: 10.1056/NEJMoa2004967 [DOI] [PubMed] [Google Scholar]
37.Jüni P, Nartey L, Reichenbach S, Sterchi R, Dieppe PA, Egger M. Risk of cardiovascular events and rofecoxib: cumulative meta-analysis. Lancet. 2004;364(9450):2021-2029. doi: 10.1016/S0140-6736(04)17514-4 [DOI] [PubMed] [Google Scholar]
38.Hollis A. Me-too drugs: is there a problem. ResearchGate. Published December 13, 2004. Accessed June 16, 2025. https://www.researchgate.net/publication/228919661_Me-too_drugs_Is_there_a_problem
39.Meier JJ, Gallwitz B, Schmidt WE, Mügge A, Nauck MA. Is impairment of ischaemic preconditioning by sulfonylurea drugs clinically important? Heart. 2004;90(1):9-12. doi: 10.1136/heart.90.1.9 [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Klepzig H, Kober G, Matter C, et al. Sulfonylureas and ischaemic preconditioning; a double-blind, placebo-controlled evaluation of glimepiride and glibenclamide. Eur Heart J. 1999;20(6):439-446. doi: 10.1053/euhj.1998.1242 [DOI] [PubMed] [Google Scholar]
41.Abdelmoneim AS, Hasenbank SE, Seubert JM, Brocks DR, Light PE, Simpson SH. Variations in tissue selectivity amongst insulin secretagogues: a systematic review. Diabetes Obes Metab. 2012;14(2):130-138. doi: 10.1111/j.1463-1326.2011.01496.x [DOI] [PubMed] [Google Scholar]
42.Hernán MA, Robins JM. Using big data to emulate a target trial when a randomized trial is not available. Am J Epidemiol. 2016;183(8):758-764. doi: 10.1093/aje/kwv254 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Hernán MA, Sterne JAC, Higgins JPT, Shrier I, Hernández-Díaz S. A structural description of biases that generate immortal time. Epidemiology. 2025;36(1):107-114. doi: 10.1097/EDE.0000000000001808 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1.

eTable 1. Specification of the Target Trial and Its Emulation in the BESTMED Database

eTable 2. Inclusion of Variables in the MACE-4 Model

eTable 3. Summary of Missing Baseline Covariate Data

eTable 4. Baseline Characteristics of and Study Medication Prescribing for Eligible Individuals by Study Site

eTable 5. Number of Events by Treatment Group and Type of Cardiovascular Outcome

eTable 6. Sensitivity Analysis Excluding Saxagliptin

eFigure 1. Unadjusted Survival Curves for MACE-4 by Treatment Group

eFigure 2. Subgroup Analyses for the Primary Outcome: Risk Differences

eTable 7. Estimated 5-Year Risks of 4-Point MACE by Treatment Group: Sensitivity Analyses

jamanetwopen-e2523067-s001.pdf^{(618.8KB, pdf)}

Supplement 2.

Nonauthor Collaborators

jamanetwopen-e2523067-s002.pdf^{(132KB, pdf)}

Supplement 3.

Data Sharing Statement

jamanetwopen-e2523067-s003.pdf^{(14.5KB, pdf)}

[zoi250672r1] 1.Sun H, Saeedi P, Karuranga S, et al. IDF diabetes atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract. 2022;183:109119. doi: 10.1016/j.diabres.2021.109119 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r2] 2.Xu G, Liu B, Sun Y, et al. Prevalence of diagnosed type 1 and type 2 diabetes among US adults in 2016 and 2017: population based study. BMJ. 2018;362:k1497. doi: 10.1136/bmj.k1497 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r3] 3.Mulnier HE, Seaman HE, Raleigh VS, et al. Risk of myocardial infarction in men and women with type 2 diabetes in the UK: a cohort study using the General Practice Research Database. Diabetologia. 2008;51(9):1639-1645. doi: 10.1007/s00125-008-1076-y [DOI] [PubMed] [Google Scholar]

[zoi250672r4] 4.Haffner SM, Lehto S, Rönnemaa T, Pyörälä K, Laakso M. Mortality from coronary heart disease in subjects with type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. N Engl J Med. 1998;339(4):229-234. doi: 10.1056/NEJM199807233390404 [DOI] [PubMed] [Google Scholar]

[zoi250672r5] 5.Mulnier HE, Seaman HE, Raleigh VS, et al. Risk of stroke in people with type 2 diabetes in the UK: a study using the General Practice Research Database. Diabetologia. 2006;49(12):2859-2865. doi: 10.1007/s00125-006-0493-z [DOI] [PubMed] [Google Scholar]

[zoi250672r6] 6.Janghorbani M, Hu FB, Willett WC, et al. Prospective study of type 1 and type 2 diabetes and risk of stroke subtypes: the Nurses’ Health Study. Diabetes Care. 2007;30(7):1730-1735. doi: 10.2337/dc06-2363 [DOI] [PubMed] [Google Scholar]

[zoi250672r7] 7.Nichols GA, Gullion CM, Koro CE, Ephross SA, Brown JB. The incidence of congestive heart failure in type 2 diabetes: an update. Diabetes Care. 2004;27(8):1879-1884. doi: 10.2337/diacare.27.8.1879 [DOI] [PubMed] [Google Scholar]

[zoi250672r8] 8.Dunlay SM, Givertz MM, Aguilar D, et al. ; American Heart Association Heart Failure and Transplantation Committee of the Council on Clinical Cardiology; Council on Cardiovascular and Stroke Nursing; Heart Failure Society of America . Type 2 diabetes mellitus and heart failure, a scientific statement from the American Heart Association and Heart Failure Society of America. J Card Fail. 2019;25(8):584-619. doi: 10.1016/j.cardfail.2019.05.007 [DOI] [PubMed] [Google Scholar]

[zoi250672r9] 9.American Diabetes Association Professional Practice Committee . 9. Pharmacologic approaches to glycemic treatment: standards of care in diabetes—2024. Diabetes Care. 2024;47(suppl 1):S158-S178. doi: 10.2337/dc24-S009 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r10] 10.Neilson LM, Munshi KD, Peasah SK, et al. Changes in type 2 diabetes medication utilization and costs in the United States, 2014–2019. Med Care. 2021;59(9):789-794. doi: 10.1097/MLR.0000000000001597 [DOI] [PubMed] [Google Scholar]

[zoi250672r11] 11.Shin JI, Echouffo-Tcheugui JB, Fang M, Grams ME, Selvin E. Trends in use of sulfonylurea types among us adults with diabetes: NHANES 1999-2020. J Gen Intern Med. 2023;38(8):2009-2010. [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r12] 12.Mishriky BM, Cummings DM, Tanenberg RJ. The efficacy and safety of DPP4 inhibitors compared to sulfonylureas as add-on therapy to metformin in patients with type 2 diabetes: a systematic review and meta-analysis. Diabetes Res Clin Pract. 2015;109(2):378-388. doi: 10.1016/j.diabres.2015.05.025 [DOI] [PubMed] [Google Scholar]

[zoi250672r13] 13.Miller M, Knatterud G, Hawkins B. A study of the effects of hypoglycemic agents on vascular complications in patients with adult on-set diabetes. Diabetes. 1976;25(12):1129-1153. doi: 10.2337/diab.25.12.1129 [DOI] [PubMed] [Google Scholar]

[zoi250672r14] 14.Bertoluci MC, Rocha VZ. Cardiovascular risk assessment in patients with diabetes. Diabetol Metab Syndr. 2017;9:25. doi: 10.1186/s13098-017-0225-1 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r15] 15.Green JB, Bethel MA, Armstrong PW, et al. ; TECOS Study Group . Effect of sitagliptin on cardiovascular outcomes in type 2 diabetes. N Engl J Med. 2015;373(3):232-242. doi: 10.1056/NEJMoa1501352 [DOI] [PubMed] [Google Scholar]

[zoi250672r16] 16.Engel SS, Golm GT, Shapiro D, Davies MJ, Kaufman KD, Goldstein BJ. Cardiovascular safety of sitagliptin in patients with type 2 diabetes mellitus: a pooled analysis. Cardiovasc Diabetol. 2013;12:3. doi: 10.1186/1475-2840-12-3 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r17] 17.Zannad F, Cannon CP, Cushman WC, et al. ; EXAMINE Investigators . Heart failure and mortality outcomes in patients with type 2 diabetes taking alogliptin versus placebo in EXAMINE: a multicentre, randomised, double-blind trial. Lancet. 2015;385(9982):2067-2076. doi: 10.1016/S0140-6736(14)62225-X [DOI] [PubMed] [Google Scholar]

[zoi250672r18] 18.White WB, Cannon CP, Heller SR, et al. ; EXAMINE Investigators . Alogliptin after acute coronary syndrome in patients with type 2 diabetes. N Engl J Med. 2013;369(14):1327-1335. doi: 10.1056/NEJMoa1305889 [DOI] [PubMed] [Google Scholar]

[zoi250672r19] 19.Scirica BM, Bhatt DL, Braunwald E, et al. ; SAVOR-TIMI 53 Steering Committee and Investigators . Saxagliptin and cardiovascular outcomes in patients with type 2 diabetes mellitus. N Engl J Med. 2013;369(14):1317-1326. doi: 10.1056/NEJMoa1307684 [DOI] [PubMed] [Google Scholar]

[zoi250672r20] 20.Rosenstock J, Perkovic V, Johansen OE, et al. ; CARMELINA Investigators . Effect of linagliptin vs placebo on major cardiovascular events in adults with type 2 diabetes and high cardiovascular and renal risk: the CARMELINA randomized clinical trial. JAMA. 2019;321(1):69-79. doi: 10.1001/jama.2018.18269 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r21] 21.Young JG, Stensrud MJ, Tchetgen Tchetgen EJ, Hernán MA. A causal framework for classical statistical estimands in failure-time settings with competing events. Stat Med. 2020;39(8):1199-1236. doi: 10.1002/sim.8471 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r22] 22.Toh S, Hernández-Díaz S, Logan R, Robins JM, Hernán MA. Estimating absolute risks in the presence of nonadherence: an application to a follow-up study with baseline randomization. Epidemiology. 2010;21(4):528-539. doi: 10.1097/EDE.0b013e3181df1b69 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r23] 23.Thompson WA Jr. On the treatment of grouped observations in life studies. Biometrics. 1977;33(3):463-470. doi: 10.2307/2529360 [DOI] [PubMed] [Google Scholar]

[zoi250672r24] 24.PCORnet . Common data model. Accessed March 26, 2025. https://pcornet.org/data/common-data-model/

[zoi250672r25] 25.Mirel LB, Resnick DM, Aram J, Cox CS. A methodological assessment of privacy preserving record linkage using survey and administrative data. Stat J IAOS. 2022;38(2):413-421. doi: 10.3233/SJI-210891 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r26] 26.Curtis MD, Griffith SD, Tucker M, et al. Development and validation of a high-quality composite real-world mortality endpoint. Health Serv Res. 2018;53(6):4460-4476. doi: 10.1111/1475-6773.12872 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r27] 27.Robert Graham Center. Social Deprivation Index (SDI). Accessed June 16, 2025. https://www.graham-center.org/maps-data-tools/social-deprivation-index.html

[zoi250672r28] 28.Rosenstock J, Kahn SE, Johansen OE, et al. ; CAROLINA Investigators . Effect of linagliptin vs glimepiride on major adverse cardiovascular outcomes in patients with type 2 diabetes: the CAROLINA randomized clinical trial. JAMA. 2019;322(12):1155-1166. doi: 10.1001/jama.2019.13772 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r29] 29.Hong J, Zhang Y, Lai S, et al. ; SPREAD-DIMCAD Investigators . Effects of metformin versus glipizide on cardiovascular outcomes in patients with type 2 diabetes and coronary artery disease. Diabetes Care. 2013;36(5):1304-1311. doi: 10.2337/dc12-0719 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r30] 30.Gangji AS, Cukierman T, Gerstein HC, Goldsmith CH, Clase CM. A systematic review and meta-analysis of hypoglycemia and cardiovascular events: a comparison of glyburide with other secretagogues and with insulin. Diabetes Care. 2007;30(2):389-394. doi: 10.2337/dc06-1789 [DOI] [PubMed] [Google Scholar]

[zoi250672r31] 31.Simpson SH, Lee J, Choi S, Vandermeer B, Abdelmoneim AS, Featherstone TR. Mortality risk among sulfonylureas: a systematic review and network meta-analysis. Lancet Diabetes Endocrinol. 2015;3(1):43-51. doi: 10.1016/S2213-8587(14)70213-X [DOI] [PubMed] [Google Scholar]

[zoi250672r32] 32.Nissen SE, Lincoff AM, Brennan D, et al. ; CLEAR Outcomes Investigators . Bempedoic acid and cardiovascular outcomes in statin-intolerant patients. N Engl J Med. 2023;388(15):1353-1364. doi: 10.1056/NEJMoa2215024 [DOI] [PubMed] [Google Scholar]

[zoi250672r33] 33.Steg PG, Bhatt DL, Simon T, et al. ; THEMIS Steering Committee and Investigators . Ticagrelor in patients with stable coronary disease and diabetes. N Engl J Med. 2019;381(14):1309-1320. doi: 10.1056/NEJMoa1908077 [DOI] [PubMed] [Google Scholar]

[zoi250672r34] 34.Schopman JE, Simon AC, Hoefnagel SJ, Hoekstra JB, Scholten RJ, Holleman F. The incidence of mild and severe hypoglycaemia in patients with type 2 diabetes mellitus treated with sulfonylureas: a systematic review and meta-analysis. Diabetes Metab Res Rev. 2014;30(1):11-22. doi: 10.1002/dmrr.2470 [DOI] [PubMed] [Google Scholar]

[zoi250672r35] 35.Pfeffer MA, Claggett B, Diaz R, et al. ; ELIXA Investigators . Lixisenatide in patients with type 2 diabetes and acute coronary syndrome. N Engl J Med. 2015;373(23):2247-2257. doi: 10.1056/NEJMoa1509225 [DOI] [PubMed] [Google Scholar]

[zoi250672r36] 36.Cannon CP, Pratley R, Dagogo-Jack S, et al. ; VERTIS CV Investigators . Cardiovascular outcomes with ertugliflozin in type 2 diabetes. N Engl J Med. 2020;383(15):1425-1435. doi: 10.1056/NEJMoa2004967 [DOI] [PubMed] [Google Scholar]

[zoi250672r37] 37.Jüni P, Nartey L, Reichenbach S, Sterchi R, Dieppe PA, Egger M. Risk of cardiovascular events and rofecoxib: cumulative meta-analysis. Lancet. 2004;364(9450):2021-2029. doi: 10.1016/S0140-6736(04)17514-4 [DOI] [PubMed] [Google Scholar]

[zoi250672r38] 38.Hollis A. Me-too drugs: is there a problem. ResearchGate. Published December 13, 2004. Accessed June 16, 2025. https://www.researchgate.net/publication/228919661_Me-too_drugs_Is_there_a_problem

[zoi250672r39] 39.Meier JJ, Gallwitz B, Schmidt WE, Mügge A, Nauck MA. Is impairment of ischaemic preconditioning by sulfonylurea drugs clinically important? Heart. 2004;90(1):9-12. doi: 10.1136/heart.90.1.9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r40] 40.Klepzig H, Kober G, Matter C, et al. Sulfonylureas and ischaemic preconditioning; a double-blind, placebo-controlled evaluation of glimepiride and glibenclamide. Eur Heart J. 1999;20(6):439-446. doi: 10.1053/euhj.1998.1242 [DOI] [PubMed] [Google Scholar]

[zoi250672r41] 41.Abdelmoneim AS, Hasenbank SE, Seubert JM, Brocks DR, Light PE, Simpson SH. Variations in tissue selectivity amongst insulin secretagogues: a systematic review. Diabetes Obes Metab. 2012;14(2):130-138. doi: 10.1111/j.1463-1326.2011.01496.x [DOI] [PubMed] [Google Scholar]

[zoi250672r42] 42.Hernán MA, Robins JM. Using big data to emulate a target trial when a randomized trial is not available. Am J Epidemiol. 2016;183(8):758-764. doi: 10.1093/aje/kwv254 [DOI] [PMC free article] [PubMed] [Google Scholar]

[zoi250672r43] 43.Hernán MA, Sterne JAC, Higgins JPT, Shrier I, Hernández-Díaz S. A structural description of biases that generate immortal time. Epidemiology. 2025;36(1):107-114. doi: 10.1097/EDE.0000000000001808 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Cardiovascular Events in Individuals Treated With Sulfonylureas or Dipeptidyl Peptidase 4 Inhibitors

Alexander Turchin, MD, MS

Lucia C Petito, PhD

Emma Hegermiller, MS

Ryan Carnahan, PharmD, MS

Andrea DeVries, PhD

Satyender Goel, PhD

M Cecilia Lansang, MD, MPH

Marie E McDonnell, MD

Vinit Nair, MS

Elisa Priest, DrPH

Vincent J Willey, PharmD

Alan F Kaul, PharmD, MS

Miguel A Hernán, MD, DrPH

Key Points

Question

Findings

Meaning

Abstract

Importance

Objective

Design, Setting, and Participants

Exposure

Main Outcomes and Measurements

Results

Conclusions and Relevance

Introduction

Methods

Target Trial Specification

Target Trial Emulation

Statistical Analysis

Results

Study Population

Figure 1. Selection of Individuals for the Target Trial Emulation, Observational Evaluation of Second Line Therapy Medications in Diabetes Study, 2014-2023.

Table 1. Baseline Characteristics of Eligible Individuals in the Observational Evaluation of Second Line Therapy Medications in Diabetes Study, 2014-2023.

Cardiovascular Events

Figure 2. Cumulative Incidence of Cardiovascular Outcomes by Treatment Group.

Table 2. Estimated 5-Year Risks of Cardiovascular Outcomes by Treatment Group, Observational Evaluation of Second Line Therapy Medications in Diabetes Study 2014-2023a.

Figure 3. Subgroup Analyses for the Primary Outcome.

Sensitivity Analyses

Discussion

Strengths and Limitations

Conclusions

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Table 2. Estimated 5-Year Risks of Cardiovascular Outcomes by Treatment Group, Observational Evaluation of Second Line Therapy Medications in Diabetes Study 2014-2023^a.