Skip to main content
NCBI home page
JAMA Network logoLink to JAMA Network
. 2022 Dec 9;5(12):e2245854. doi: 10.1001/jamanetworkopen.2022.45854

Comparison of Mitochondrial Adenosine Triphosphate–Sensitive Potassium Channel High- vs Low-Affinity Sulfonylureas and Cardiovascular Outcomes in Patients With Type 2 Diabetes Treated With Metformin

Meng-Ting Wang 1,2,✉, Hsueh-Yi Pan 1, Ya-Ling Huang 2, Li-Wei Wu 2, Pin-Chun Wang 2, Yu-Juei Hsu 3, Tzu-Chieh Lin 1, ChenWei Lin 1, Jyun-Heng Lai 4, Chien-Hsing Lee 5
PMCID: PMC9856426  PMID: 36484988

This cohort study evaluates the association between mitochondrial adenosine triphosphate–sensitive potassium (KATP) channel high- vs low-affinity sulfonylureas and cardiovascular outcomes in patients with type 2 diabetes treated with metformin.

Key Points

Question

Is there an increased risk of major adverse cardiovascular events (MACEs) associated with use of cardiac mitochondrial adenosine triphosphate–sensitive potassium (mitoKATP) channel high-affinity sulfonylureas (glyburide and glipizide) vs low-affinity sulfonylureas (gliclazide and glimepiride) combined with metformin in type 2 diabetes (T2D)?

Findings

In this population-based, propensity score–matched cohort study, use of cardiac mitoKATP channel high-affinity sulfonylureas vs low-affinity sulfonylureas added to metformin was associated with a 1.18-fold increased MACE risk in patients with T2D.

Meaning

These findings suggest that high-affinity blockage of sulfonylureas to cardiac mitoKATP channels may act as an important potential mechanism underlying sulfonylureas-associated MACEs when combined with metformin in T2D.

Abstract

Importance

Sulfonylureas are frequently used as add-on to metformin in type 2 diabetes (T2D), and individual sulfonylurea agents carry different risks of cardiovascular disease. Sulfonylureas’ different affinities to cardiac mitochondrial adenosine triphosphate–sensitive potassium (mitoKATP) channels have been speculated to account for the intraclass difference in cardiovascular risk from in vitro and ex vivo studies; however, this hypothesis has not been assessed in a general population with diabetes receiving sulfonylureas added to metformin.

Objective

To compare the risk of myocardial infarction (MI), ischemic stroke, or cardiovascular death in patients with T2D treated with mitoKATP channel high-affinity sulfonylureas and low-affinity sulfonylureas as add-on to metformin.

Design, Setting, and Participants

This is a new-user, active-comparator, and propensity score–matched cohort study with analysis of the Taiwanese Diabetes Mellitus Health Database from 2006, to 2017. Data analysis was performed from August 2020 to July 2021.

Exposures

Cardiac mitoKATP channel high-affinity (glyburide and glipizide) and low-affinity (gliclazide and glimepiride) sulfonylureas combined with metformin.

Main Outcomes and Measures

Primary outcome was major adverse cardiovascular events (MACEs), a composite of cardiovascular death or hospitalization for either MI or ischemic stroke. Secondary outcomes included individual MACE components, heart failure, arrhythmia, all-cause mortality, and severe hypoglycemia. Cox proportional hazards models were used to estimate adjusted hazard ratios (aHRs).

Results

Each sulfonylurea group comprised 53 714 patients (mean [SD] age, 54.7 [12.1] years; 31 962 men [59.5%]). MitoKATP channel high-affinity sulfonylureas vs low-affinity sulfonylureas when combined with metformin were associated with an increased risk of MACE (aHR, 1.18; 95% CI, 1.03-1.34), MI (aHR, 1.34; 95% CI, 1.04-1.73), all-cause mortality (aHR, 1.27; 95% CI, 1.03-1.57), and severe hypoglycemia (aHR, 1.82; 95% CI, 1.58-2.10), but not with increased risks of ischemic stroke, cardiovascular death, arrhythmia, and heart failure. The duration analyses revealed the highest MACE risk during 1 to 90 days after initiation of mitoKATP channel high-affinity sulfonylureas (aHR, 6.06; 95% CI, 4.86-7.55).

Conclusions and Relevance

Use of mitoKATP channel high-affinity sulfonylureas vs low-affinity sulfonylureas was associated with an increased MACE risk in patients with T2D concomitantly receiving metformin, suggesting that high-affinity blockage of the mitoKATP channels could account for sulfonylurea-associated MACEs.

Introduction

Diabetes is a prevalent disease that imposes an enormous burden on health worldwide.1 Diabetes currently affects more than 537 million adults, leads to 6.9 million deaths, and cost $US 966 billion in health care expenditures in 2021 worldwide.1 Major cardiovascular events (MACEs) are the major cause of mortality and morbidity in diabetes; for example, approximately one-third of all deaths in patients with diabetes were cardiovascular causes from 2010 to 2015 in the US.2 This highlights the importance of preventing MACEs in patients with diabetes.

Despite the availability of newer types of antidiabetic medications,3 sulfonylureas remain one of the most frequently prescribed classes of noninsulin antihyperglycemic agents, primarily owing to their low cost, established glucose-lowering efficacy, and long experience of clinical use.4 Sulfonylureas are the most prescribed antihyperglycemic medications after metformin.5,6 However, use of sulfonylureas is associated with an elevated risk of MACEs.4 Different individual sulfonylureas have been found to be associated with varying risks of cardiovascular diseases,7,8 suggesting a possible intraclass difference in cardiovascular risk. Although it has been investigated whether the specificity of sulfonylureas for pancreatic β-cells could account for these drugs’ associated cardiovascular risks,9,10,11,12 no studies have confirmed the proposed mechanism.

Blockage of cardiac mitochondrial adenosine triphosphate–sensitive potassium (mitoKATP) by certain sulfonylureas may be an alternative mechanism underlying the intraclass difference in MACEs. Ischemic preconditioning is the most pivotal mechanism for cardiac protection and is involved in the opening of mitoKATP channels, which can be blocked by certain sulfonylureas, such as glyburide and glipizide.13,14 In vitro and ex vivo data have indicated that there were differential affinities to the mitoKATP channels among individual sulfonylureas,15,16,17,18,19 although this has not been translated to account for sulfonylurea-related adverse cardiovascular events in clinical settings, except for 1 observational study focusing on sulfonylurea monotherapy.20 As sulfonylureas are most used as add-on antidiabetic agents to metformin in management of type 2 diabetes (T2D), we aimed to assess whether use of cardiac mitoKATP channel high-affinity sulfonylureas (glyburide and glipizide) vs low-affinity sulfonylureas (gliclazide and glimepiride) when combined with metformin is associated with an increased risk of MACEs in a nationwide T2D population.

Methods

Study Design and Data Sources

This was a population-based, propensity score (PS)–matched cohort study using the Taiwan Diabetes Mellitus Health Database (DMHD) from January 1, 2006 to December 31, 2017. The DMHD comprises patient diagnoses, medical procedures, and prescription refill records of all patients with new diagnoses of diabetes under the Taiwanese universal national health insurance (NHI) program, with a coverage rate of greater than 99% of Taiwanese inhabitants. In the database, patients with new diabetes diagnoses were defined as those with at least 3 outpatient visits or 1 inpatient visit for diabetes in a given year, and they did not have previous visits with any diagnoses of diabetes preceding the first visit related to diabetes. The nationwide death registry and the Tri-Service General Hospital (TSGH) electronic medical records were linked to the DMHD to obtain the causes of death and glycated hemoglobin A1c (HbA1c) levels, respectively. This study was approved by the institutional review board of TSGH, National Defense Medical Center and completed before the lead author became affiliated with the National Yang Ming Chiao Tung University. Written informed consent was waived because the study analyzed a deidentified database. The report of this study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Study Population and Exposures

We identified a base cohort of patients with newly diagnosed diabetes who initiated metformin monotherapy, defined as those without any prescription refill records of any antidiabetic medications in the year before the monotherapy commencement, from January 1, 2007, to December 31, 2016. The base cohort was required to have no type 1 diagnosis codes and to continuously receive metformin monotherapy, allowing a 30-day grace period between refills. From the base cohort, we identified the study cohort of patients with an add-on of a cardiac mitoKATP channel high-affinity sulfonylurea (glyburide and glipizide) or an add-on low-affinity sulfonylurea (gliclazide and glimepiride) to their metformin monotherapy, and marked the first prescription date of add-on sulfonylureas as cohort entry date. Patients who initiated sulfonylurea and metformin on the same day were also included. The study cohort was required to be aged 20 or more years at cohort entry. The 4 examined individual sulfonylureas comprised more than 99% of the sulfonylureas combined with metformin in the study period. Patients were excluded if they received both types of sulfonylureas at cohort entry; had lack of 1-year continuous NHI enrollment; or had a hospitalization with a primary diagnosis of myocardial infarction (MI), ischemic stroke, or pregnancy in the year preceding cohort entry. eTable 1 in Supplement 1 details the disease and procedure codes used in defining the inclusion and exclusion criteria.

The 2 sulfonylurea groups were followed from the cohort entry date until the earliest of the following events: occurrence of a primary cardiovascular outcome (defined later in this study), NHI enrollment discontinuation, metformin-sulfonylurea treatment discontinuation, switch or add-on of other antidiabetic medications (except for switching within the same mitoKATP channel high-affinity or low-affinity sulfonylurea group), pregnancy, or December 31, 2017. Discontinuous use of sulfonylureas with metformin was determined according to the prescription refill records with more than a 30-day grace period. We extended the follow-up to 1 month after the dual therapy was discontinued to include the MACEs events that may have caused discontinuity in the dual therapy.

The PS, the probability of initiating mitoKATP channel high-affinity sulfonylureas, was estimated by fitting a multivariable logistic regression model, conditional on all factors listed in Table 1, to maintain comparability between groups. The 2 sulfonylurea groups were matched by the cohort entry date (within 90 days), adapted Diabetes Complications Severity Index (aDCSI; 0, 1, 2, and ≥3), number of metformin prescriptions between initial use of metformin and cohort entry date, and PS using the nearest neighboring matching scheme with a caliper width of 0.02 on the propensity-score scale. We additionally examined the effects of the average daily dose (<0.5, 0.5-1, >1 defended daily dose) and duration therapy (1-90, 91-180, 181-365, >365 days) of mitoKATP channel high-affinity sulfonylureas on the comparative outcomes.

Table 1. Characteristics of Users of Add-on MitoKATP Channel High-Affinity and Low-Affinity Sulfonylureas Before and After Matching in Patients With Diabetes.

Characteristicsa Before matching After matching
Patients, No. (%) SMDb Patients, No. (%) SMDb
MitoKATP channel high-affinity sulfonylureas (n = 54 411) MitoKATP channel low-affinity sulfonylureas (n = 193 333) MitoKATP channel high-affinity sulfonylureas (n = 53 714) MitoKATP channel low-affinity sulfonylureas (n = 53 714)
Age, mean (SD), y 54.8 (12.1) 54.4 (11.9) 0.035 54.7 (12.1) 54.6 (12.0) 0.006
Sex
Female 22 132 (40.7) 80 265 (41.5) 0.017 21 833 (40.6) 21 671 (40.3) 0.006
Male 32 279 (59.3) 113 068 (58.5) 0.017 31 881 (59.4) 32 043 (59.7) 0.006
Metformin prescriptions between initial use of metformin and cohort entry date, mean (SD), No. 1.11 (2.90) 1.68 (3.55) 0.176 1.02 (2.70) 1.02 (2.70) <0.001
Entry year
2007 5954 (10.9) 12 823 (6.6) 0.150 5874 (10.9) 5910 (11.0) 0.002
2008 6611 (12.2) 15 181 (7.9) 0.142 6484 (12.1) 6417 (12.0) 0.004
2009 7683 (14.1) 18 061 (9.3) 0.147 7552 (14.1) 7614 (14.2) 0.003
2010 6630 (12.2) 19 356 (10.0) 0.069 6537 (12.2) 6565 (12.2) 0.002
2011 6310 (11.6) 19 190 (9.9) 0.054 6230 (11.6) 6206 (11.6) 0.001
2012 5631 (10.4) 20 939 (10.8) 0.016 5574 (10.4) 5616 (10.5) 0.003
2013 4987 (9.2) 21 675 (11.2) 0.068 4942 (9.2) 4917 (9.2) 0.002
2014 4226 (7.8) 22 198 (11.5) 0.127 4186 (7.8) 4156 (7.7) 0.002
2015 3499 (6.4) 22 974 (11.9) 0.192 3479 (6.5) 3486 (6.5) 0.001
2016 2880 (5.3) 20 936 (10.8) 0.206 2856 (5.3) 2827 (5.3) 0.002
Diabetes severity indicators
Adapted diabetes complications severity index
0 44 748 (82.2) 158 060 (81.8) 0.013 44 572 (83.0) 44 572 (83.0) <0.001
1 6393 (11.8) 23 812 (12.3) 0.017 6214 (11.6) 6214 (11.6) <0.001
2 2684 (4.9) 9484 (4.9) 0.001 2497 (4.7) 2497 (4.7) <0.001
≥3 586 (1.1) 1977 (1.0) 0.005 431 (0.8) 431 (0.8) <0.001
Metformin dose
<1000 mg 20 634 (37.9) 59 693 (30.9) 0.145 20 237 (37.7) 19 822 (36.9) 0.016
1000-1499 mg 25 270 (46.4) 91 750 (47.5) 0.020 25 070 (46.7) 25 478 (47.4) 0.015
≥1500 mg 8507 (15.6) 41 890 (21.7) 0.158 8407 (15.7) 8414 (15.7) <0.001
Measures of health care utilization
Physician visits, No.
Diabetes related
First tertile 29 777 (54.7) 88 395 (45.7) 0.172 29 678 (55.3) 29 831 (55.5) 0.006
Second tertile 10 171 (18.7) 41 180 (21.3) 0.066 10 145 (18.9) 10 091 (18.8) 0.003
Third tertile 14 463 (26.6) 63 758 (33.0) 0.143 13 891 (25.9) 13 792 (25.7) 0.004
Non–diabetes related
First tertile 17 814 (32.7) 62 506 (32.3) 0.009 17 719 (33.0) 18 343 (34.2) 0.025
Second tertile 18 171 (33.4) 66 702 (34.5) 0.023 17 981 (33.5) 17 755 (33.1) 0.009
Third tertile 18 426 (33.9) 64 125 (33.2) 0.015 18 014 (33.5) 17 616 (32.8) 0.016
Hospital admissions, No.
Diabetes related
0 53 118 (97.6) 188 743 (97.6) <0.001 52 596 (97.9) 52 613 (98.0) 0.002
1 1091 (2.0) 3959 (2.1) 0.003 986 (1.8) 979 (1.8) 0.001
2 202 (0.4) 631 (0.3) 0.008 132 (0.3) 122 (0.2) 0.004
Non–diabetes related
0 51 174 (94.1) 182 978 (94.6) 0.026 50 614 (94.2) 50 682 (94.4) 0.005
1 2491 (4.6) 7971 (4.1) 0.022 2396 (4.5) 2367 (4.4) 0.003
2 746 (1.4) 2384 (1.2) 0.012 704 (1.3) 665 (1.2) 0.006
Emergency department visits, No.
Diabetes related
0 53 454 (98.2) 189 675 (98.1) 0.010 52 830 (98.4) 52 882 (98.5) 0.008
1 752 (1.4) 3206 (1.7) 0.023 713 (1.3) 683 (1.3) 0.005
2 205 (0.4) 452 (0.2) 0.026 171 (0.3) 149 (0.3) 0.008
Non–diabetes related
0 45 641 (83.9) 163 089 (84.4) 0.013 45 192 (84.1) 45 443 (84.6) 0.013
1 5892 (10.8) 21 316 (11.0) 0.006 5774 (10.8) 5580 (10.4) 0.012
2 2878 (5.3) 8928 (4.6) 0.031 2748 (5.1) 2691 (5.0) 0.005
Monthly income-based insurance premium, NTD
First tertile 22 061 (40.6) 66 682 (34.5) 0.122 21 693 (40.4) 21 742 (40.5) 0.002
Second tertile 15 471 (28.4) 61 652 (31.9) 0.076 15 285 (28.5) 15 380 (28.6) 0.004
Third tertile 16 879 (31.0) 64 999 (33.6) 0.056 16 736 (31.2) 16 592 (30.9) 0.006
Hospital level
No medical record 1769 (3.3) 4908 (2.5) 0.042 1769 (3.3) 1985 (3.7) 0.022
Academic medical centers 3599 (6.6) 14 758 (7.6) 0.040 3546 (6.6) 3445 (6.4) 0.008
Metropolitan hospitals 6649 (12.2) 23 782 (12.3) 0.002 6501 (12.1) 6368 (11.9) 0.008
Local community hospitals 5165 (9.5) 14 834 (7.7) 0.065 4915 (9.2) 4778 (8.9) 0.009
Physician clinics 37 229 (68.4) 135 051 (69.9) 0.031 36 983 (68.9) 37 138 (69.1) 0.006
Comorbidities
Cardiovascular diseases
Heart failure 1110 (2.0) 3510 (1.8) 0.016 1008 (1.9) 1003 (1.9) 0.001
Hypertension 21 205 (39.0) 79 077 (40.9) 0.040 20 726 (38.6) 20 070 (37.4) 0.025
Cerebrovascular disease 1328 (2.4) 4445 (2.3) 0.009 1239 (2.3) 1225 (2.3) 0.002
Ischemic heart disease 3885 (7.1) 14 049 (7.3) 0.007 3676 (6.8) 3768 (7.0) 0.007
Arrhythmia 111 (0.2) 325 (0.2) 0.008 93 (0.2) 82 (0.2) 0.005
Dyslipidemia 13 211 (24.3) 59 408 (30.7) 0.148 12 963 (24.1) 12 652 (23.6) 0.014
Peripheral arterial disease 621 (1.1) 2132 (1.1) 0.004 585 (1.1) 583 (1.1) <0.001
Coronary revascularization 81 (0.2) 318 (0.2) 0.004 74 (0.1) 76 (0.1) 0.001
Cardiomyopathy 79 (0.2) 276 (0.1) 0.001 70 (0.1) 65 (0.1) 0.003
Venous thromboembolism 74 (0.1) 238 (0.1) 0.004 66 (0.1) 60 (0.1) 0.003
Pulmonary disease
Asthma 1979 (3.6) 7180 (3.7) 0.004 1927 (3.6) 1882 (3.5) 0.005
Chronic obstructive pulmonary disease 1891 (3.5) 5582 (2.9) 0.033 1773 (3.3) 1731 (3.2) 0.040
Pneumonia 1219 (2.2) 4019 (2.1) 0.011 1145 (2.1) 1093 (2.0) 0.007
Mental disease
Depression 1249 (2.3) 4511 (2.3) 0.003 1219 (2.3) 1183 (2.2) 0.005
Anxiety 3928 (7.2) 14 331 (7.4) 0.007 3854 (7.2) 3769 (7.0) 0.006
Schizophrenia 593 (1.1) 1805 (0.9) 0.016 565 (1.1) 515 (1.0) 0.010
Neurologic disorders
Dementia 341 (0.6) 1048 (0.5) 0.011 313 (0.6) 298 (0.6) 0.004
Epilepsy 179 (0.3) 536 (0.3) 0.009 166 (0.3) 151 (0.3) 0.005
Bone and joint disorders
Fracture 1774 (3.3) 6044 (3.1) 0.008 1721 (3.2) 1702 (3.2) 0.002
Osteoporosis 749 (1.4) 2455 (1.3) 0.009 724 (1.4) 728 (1.4) 0.001
Osteoarthritis 5645 (10.4) 19 712 (10.2) 0.006 5490 (10.2) 5353 (10.0) 0.008
Anemia 855 (1.6) 3380 (1.8) 0.014 823 (1.5) 830 (1.6) 0.001
Thyroid disease 1279 (2.4) 5484 (2.8) 0.031 1255 (2.3) 1208 (2.3) 0.006
Chronic liver disease 5418 (10.0) 19 718 (10.2) 0.008 5251 (9.8) 5115 (9.5) 0.009
Chronic kidney disease 1899 (3.5) 7917 (4.1) 0.032 1749 (3.3) 1752 (3.3) <0.001
Obesity or weight gain 1471 (2.7) 4893 (2.5) 0.011 1441 (2.7) 1374 (2.6) 0.008
Tobacco-related 459 (0.8) 1910 (1.0) 0.015 452 (0.8) 461 (0.9) 0.002
Alcohol-related disorder 241 (0.4) 719 (0.4) 0.011 233 (0.4) 242 (0.5) 0.003
Hypokalemia 229 (0.4) 713 (0.4) 0.008 209 (0.4) 179 (0.3) 0.009
Hypoglycemia 149 (0.3) 663 (0.3) 0.012 142 (0.3) 143 (0.3) <0.001
Autoimmune diseases 998 (1.8) 3255 (1.7) 0.011 955 (1.8) 969 (1.8) 0.002
Cancer 1733 (3.2) 5996 (3.1) 0.005 1670 (3.1) 1643 (3.1) 0.003
Comedication
Cardiovascular medications
Angiotensin-converting enzyme inhibitors 3987 (7.3) 13 286 (6.9) 0.018 3846 (7.2) 3781 (7.0) 0.005
Angiotensin receptor blockers 7515 (13.8) 35 144 (18.2) 0.121 7332 (13.7) 7005 (13.0) 0.018
α-Blockers 938 (1.7) 3048 (1.6) 0.012 898 (1.7) 855 (1.6) 0.006
β-Blockers 10 462 (19.2) 36 425 (18.8) 0.010 10 143 (18.9) 9865 (18.4) 0.013
Calcium channel blockers
Dihydropyridines 13 706 (25.2) 49 949 (25.8) 0.015 13 370 (24.9) 12 968 (24.1) 0.017
Nondihydropyridines 1144 (2.1) 3838 (2.0) 0.008 1092 (2.0) 1066 (2.0) 0.003
Diuretics
Thiazides 10 816 (19.9) 43 148 (22.3) 0.060 10 541 (19.6) 10 086 (18.8) 0.021
Loop 1754 (3.2) 5583 (2.9) 0.019 1643 (3.1) 1666 (3.1) 0.002
Potassium-sparing agents 1316 (2.4) 3681 (1.9) 0.035 1209 (2.3) 1206 (2.3) <0.001
Antiplatelets 7037 (12.9) 24 716 (12.8) 0.004 6760 (12.6) 6579 (12.3) 0.010
Anticoagulants 391 (0.7) 1390 (0.7) <0.001 358 (0.7) 373 (0.7) 0.003
Lipid-lowering agents
Statins 5575 (10.3) 29 749 (15.4) 0.156 5417 (10.1) 5116 (9.5) 0.019
Others 857 (1.6) 3684 (1.9) 0.025 842 (1.6) 860 (1.6) 0.003
Nitrates 1163 (2.1) 3659 (1.9) 0.017 1070 (2.0) 1028 (1.9) 0.006
Antiarrhythmic agents 460 (0.9) 1456 (0.8) 0.010 420 (0.8) 406 (0.8) 0.003
Digoxin 531 (1.0) 1626 (0.8) 0.014 485 (0.9) 496 (0.9) 0.002
Anti-inflammatory agents
Nonsteroidal anti-inflammatory drugs 31 071 (57.1) 111 250 (57.5) 0.009 30 642 (57.1) 30 321 (56.5) 0.012
Steroids 8980 (16.5) 31 548 (16.3) 0.005 8779 (16.3) 8542 (15.9) 0.012
Potassium channel opener (nicorandil) 301 (0.6) 1395 (0.7) 0.021 287 (0.5) 274 (0.5) 0.003
Inhibitors of mitochondrial permeability transition pore 8872 (16.3) 31 692 (16.4) 0.002 8679 (16.2) 8495 (15.8) 0.009
Proton pump inhibitors 1537 (2.8) 5744 (3.0) 0.009 1485 (2.8) 1421 (2.7) 0.007
Anticonvulsants 1779 (3.3) 6704 (3.5) 0.011 1712 (3.2) 1643 (3.1) 0.007
Antidepressants 2941 (5.4) 10 562 (5.5) 0.003 2864 (5.3) 2813 (5.2) 0.004
Antipsychotics 3502 (6.4) 11 724 (6.1) 0.015 3388 (6.3) 3123 (5.8) 0.021
Open in a new tab

Abbreviations: mitoKATP, mitochondrial adenosine triphosphate–sensitive potassium; NTD, New Taiwan dollar; SMD, standardized mean difference.

a

All comorbidities, diabetes severity indicators, and adjusted Diabetes Complications Severity Index were measured in the year preceding the cohort entry date; comedications were measured 6 months before the cohort entry date.

b

SMD greater than 0.1 represents meaningful differences between 2 groups.

Outcome Definition

The primary outcome was 3-point composite outcome of MACEs, including hospitalization with any diagnosis of MI, ischemic stroke, or cardiovascular death. The positive predictive values of the coding algorithms (eTable 1 in Supplement 1) for identifying the MI and ischemic stroke events in the database were at least 88%.21,22 Secondary outcomes were the individual components of the 3-point MACEs, heart failure, arrhythmia, all-cause mortality, and hypoglycemia, as detailed in eTable 1 in Supplement 1.

Baseline Covariates

We assessed patient demographic and clinical characteristics in the year preceding the cohort entry, including age, sex, proxies of diabetes severity (eg, aDCSI and metformin daily dose), comorbidities (eg, chronic kidney disease), and health care utilizations. Comedications (eg, cardiovascular medications) were evaluated in the previous 6 months before cohort entry.

Statistical Analysis

Main Analysis

Baseline characteristics were compared using the standardized difference, where a magnitude greater than 0.1 indicated an imbalance between groups. The Kaplan-Meier method with a log-rank test was used to compare the cumulative incidence of primary and secondary outcomes. Cox proportional hazards models with 95% CIs were used to estimate the hazard ratios (HRs) for each outcome, and the proportionality assumption was assessed through Schoenfeld residuals, with all assumptions being met. All analyses were also adjusted for PS deciles after the matching process to mitigate any residual confounding. Statistical significance was defined as a 2-sided value of P < .05. Data cleaning and statistical analyses were performed using SAS statistical software version 9.4 (SAS Institute). Data were analyzed from August 2020 to July 2021.

Sensitivity and Subgroup Analysis

Multiple predetermined sensitivity analyses were conducted. First, we performed the PS calibration analysis,23 which combined PS and regression calibration, to address the unavailability of HbA1c in the DMHD by using HbA1c data obtained from a tertiary medical center, detailed in eMethods in Supplement 1. Second, unmeasured confounding was addressed through the rule-out approach24 and high-dimensional PS-matched analyses (detailed in eMethods in Supplement 1).25 Third, the grace period used for determining continuous treatment was shortened to 14 days to address any exposure misclassification. Fourth, we performed an intention-to-treat analysis without considering treatment discontinuation or switching, where follow-up was set as 1 year to mitigate potential informative censoring. Fifth, we considered noncardiovascular death as a competing event for the primary outcome using the Fine and Gray method. Sixth, all patients were restricted to those with a medication possession ratio of 80% or greater to maintain adequate medication adherence.26 Seventh, we performed a negative outcome of diabetic retinopathy analysis as there is no evidence suggesting differential risks of the eye disease between the 2 types of sulfonylureas. Eighth, we adjusted for hypoglycemic events measured during follow-up to rule out the possibility that the observed MACE risk was mediated through hypoglycemia. Ninth, cardiovascular death in the MACE outcome was redefined as a primary cause of MI or ischemic stroke. Tenth, an inverse probability of treatment weighting using PS was performed to avoid reduction in sample sizes.27 We additionally performed subgroup analyses according to sulfonylurea’s pancreas specificity (pancreas-specificity sulfonylureas, glipizide vs gliclazide; pancreas-nonspecific sulfonylureas, glyburide vs glimepiride), with re-estimated PS.

Results

A total of 247 744 patients with T2D met the inclusion and exclusion criteria, among whom after 1:1 matching 53 714 pairs of new users of mitoKATP channel low-affinity sulfonylureas and high-affinity sulfonylureas in combination with metformin were identified. Overall mean (SD) patient age was 54.7 (12.1) years, and 31 962 (59.5%) were men (Figure 1). The mean (SD) follow-up duration was 10.2 (16.3) months (corresponding to 44 790 person-years) and 14.0 (20.5) months (61 816 person-years) in the mitoKATP channel high-affinity and low-affinity sulfonylurea group, respectively. Most patients in both affinity groups were censored during the follow-up primarily owing to treatment discontinuation (29 054 [54.1%] and 31 671 [59.0%], respectively) or treatment switching (15 953 [29.7%] and 13 286 [24.7%], respectively), with the differences between groups across all types of censoring no greater than 5% (eTable 2 in Supplement 1). In our study, glyburide and glimepiride comprised 76.8% and 74.9% in the cardiac mitoKATP channel high-affinity sulfonylurea and in low-affinity sulfonylurea groups, respectively.

Figure 1. Flowchart Showing the Selection of the Base and Study Cohorts.

Figure 1.

Open in a new tab

aDCSI, adapted Diabetes Complications Severity Index; MI, myocardial infarction; mitoKATP, mitochondrial adenosine triphosphate–sensitive potassium.

As shown in Table 1, although most baseline characteristics were similar between groups before matching, more patients in the mitoKATP channel low-affinity sulfonylurea group entered the study cohort in later years, had diagnoses of dyslipidemia, and received angiotensin receptor blockers and statins than did the high-affinity sulfonylurea group. All characteristics were well-balanced after matching.

A total of 444 and 484 MACEs occurred in the mitoKATP channel high-affinity and in the low-affinity sulfonylurea group, respectively, during follow-up, with corresponding incidence rates per 100 person-years of 0.99 (95% CI, 0.90-1.09) and 0.78 (95% CI, 0.72-0.86) (Table 2). The cumulative incidence rates of the examined outcomes are shown in eFigures 1 and 2 in Supplement 1.

Table 2. Risk of Major Adverse Cardiovascular Events Between Add-on MitoKATP Channel High-Affinity Sulfonylurea and MitoKATP Channel Low-Affinity Sulfonylurea.

Outcomes MitoKATP channel high-affinity sulfonylureas (n = 53 714) MitoKATP channel low-affinity sulfonylureas (n = 53 714) HR (95% CI) aHR (95% CI)a
Events, No. Total No. of person-years Incidence rate per 100 person-years (95% CI) Events, No. Total No. of person-years Incidence rate per 100 person-years (95% CI)
Primary outcome, 3-point MACEb 444 44 790 0.99 (0.90-1.09) 484 61 816 0.78 (0.72-0.86) 1.17 (1.03-1.34) 1.18 (1.03-1.34)
Secondary outcomes
Myocardial infarction 118 44 909 0.26 (0.22-0.31) 119 62 009 0.19 (0.16-0.23) 1.34 (1.04-1.73) 1.34 (1.04-1.73)
Ischemic stroke 295 44 847 0.66 (0.59-0.74) 330 61 879 0.53 (0.48-0.59) 1.12 (0.96-1.31) 1.12 (0.96-1.31)
Cardiovascular deathc 38 44 964 0.08 (0.06-0.12) 39 62 069 0.06 (0.05-0.09) 1.32 (0.85-2.07) 1.32 (0.85-2.07)
Arrhythmia 81 44 901 0.18 (0.15-0.22) 89 61 983 0.14 (0.12-0.18) 1.20 (0.89-1.63) 1.21 (0.89-1.63)
Heart failure 59 44 952 0.13 (0.10-0.17) 55 62 045 0.09 (0.07-0.12) 1.35 (0.93-1.95) 1.35 (0.93-1.95)
All-cause mortality 169 44 961 0.38 (0.32-0.44) 174 62 066 0.28 (0.24-0.33) 1.27 (1.02-1.57) 1.27 (1.03-1.57)
Severe hypoglycemia 485 44 750 1.08 (0.99-1.19) 330 61 889 0.53 (0.48-0.59) 1.82 (1.58-2.09) 1.82 (1.58-2.10)
Open in a new tab

Abbreviations: aHR, adjusted hazard ratio; HR, hazard ratio; MACE, major adverse cardiovascular events; mitoKATP, mitochondrial adenosine triphosphate–sensitive potassium.

a

Adjusted for the deciles of propensity scores.

b

Three-point MACE includes myocardial infarction, ischemic stroke, and cardiovascular death.

c

Cardiovascular death was defined as death due to all cardiovascular diseases.

MitoKATP channel high-affinity sulfonylureas were associated with a 1.18-fold (95% CI, 1.03-1.34) increased risk of MACE compared with mitoKATP channel low-affinity sulfonylureas in combination with metformin (Table 2). Additionally, use of mitoKATP channel high-affinity sulfonylureas vs low-affinity sulfonylureas was associated with an increased risk of MI (adjusted HR [aHR], 1.34; 95% CI, 1.04-1.73), all-cause mortality (aHR, 1.27; 95% CI, 1.03-1.57), and severe hypoglycemia (aHR, 1.82; 95% CI, 1.58-2.10) but was not associated with higher risks of ischemic stroke, cardiovascular death, arrhythmia, and heart failure in patients with T2D concurrently receiving metformin.

The MACE risk varied by duration and dose of mitoKATP channel high-affinity sulfonylurea therapy (Table 3). The duration of mitoKATP channel high-affinity sulfonylurea use was inversely associated with the MACE risk, with the highest increased risk confined to 1 to 90 days of the sulfonylurea therapy (aHR, 6.06; 95% CI, 4.86-7.55). Among the 3 dose strata, only the use of mitoKATP channel high-affinity sulfonylureas greater than 1 defined daily dose was associated with an increased risk of MACEs (aHR, 1.43; 95% CI, 1.10-1.85). eTable 3 in Supplement 1 presents the absolute risks.

Table 3. Risk of Major Adverse Cardiovascular Events With Different Doses and Durations of Add-on MitoKATP Channel High-Affinity Sulfonylurea Therapy Compared With Any Use of MitoKATP Channel Low-Affinity Sulfonylureaa.

Dose and duration Events, No. Total No. of person-years Incidence rate per 100 person-years (95% CI) HR (95% CI) aHR (95% CI)b
MitoKATP channel low-affinity sulfonylureas 484 61 816 0.78 (0.72-0.86) 1 [Reference] 1 [Reference]
Cumulative duration of add-on mitoKATP channel high-affinity sulfonylurea
MitoKATP channel high-affinity sulfonylureas, d
1-90 202 2824 7.15 (6.23-8.21) 6.09 (4.89-7.59) 6.06 (4.86-7.55)
91-180 45 3652 1.23 (0.92-1.65) 1.39 (0.999-1.92) 1.38 (0.996-1.92)
181-365 49 5371 0.91 (0.69-1.21) 1.17 (0.86-1.59) 1.17 (0.86-1.59)
>365 148 32 942 0.45 (0.38-0.53) 0.59 (0.49-0.71) 0.59 (0.49-0.72)
Average daily dose of add-on mitoKATP channel high-affinity sulfonylurea
MitoKATP channel high-affinity sulfonylureas, defined daily dose
<0.5 182 18 372 0.99 (0.86-1.15) 1.13 (0.95-1.34) 1.12 (0.95-1.33)
0.5-1 197 20 872 0.94 (0.82-1.09) 1.15 (0.97-1.36) 1.16 (0.98-1.37)
>1 65 5546 1.17 (0.92-1.49) 1.42 (1.09-1.83) 1.43 (1.10-1.85)
Open in a new tab

Abbreviations: aHR, adjusted hazard ratio; HR, hazard ratio; MACE, major adverse cardiovascular events; mitoKATP, mitochondrial adenosine triphosphate–sensitive potassium.

a

Three-point MACE includes myocardial infarction, ischemic stroke, and cardiovascular death.

b

Adjusted for the deciles of propensity scores.

The main findings remained consistent across all sensitivity analyses, such as when considering baseline HbA1c in estimating the PS and using intention-to-treat analyses (Figure 2). Additionally, an unmeasured confounder was unlikely to fully explain the main findings with the rule-out approach (eFigure 3 in Supplement 1). Furthermore, the negative outcome analysis revealed a null association (aHR, 1.01; 95% CI, 0.90-1.14), as expected. The HRs were not materially changed after stratifying by the pancreas specificity of sulfonylureas.

Figure 2. Forest Plots of Sensitivity and Subgroup Analyses.

Figure 2.

Open in a new tab

aHR, adjusted hazard ratio; MACE, major adverse cardiovascular events; PS, propensity score.

aP value is less than .05.

bThe deciles of PS were adjusted for.

Discussion

The overall findings suggested that use of mitoKATP channel high-affinity sulfonylureas vs low-affinity sulfonylureas in combination with metformin was associated with an increased MACE risk in a nationwide T2D population. The increased MACE risk was primarily associated with hospitalization for MI and was particularly elevated with mitoKATP channel high-affinity sulfonylureas used within 90 days of initiation and at a high daily dose. The main findings were consistent across multiple sensitivity analyses. Our findings suggest that high affinity to cardiac mitoKATP channels is associated with sulfonylurea-associated cardiovascular risk when concomitantly used with metformin in patients with T2D.

In accordance with this current finding, a recently published observational study,20 reported a 21% (aHR, 1.21; 95% CI, 1.03-1.44) increased risk of 3-point MACEs associated with mitoKATP channel high-affinity sulfonylureas vs low-affinity sulfonylureas although the prior study focused on sulfonylurea monotherapy, and the generalizability of the reported data might be limited, given that sulfonylureas are most prescribed as an add-on medication to metformin in clinical settings. Conversely, the 2 studies observed discrepant secondary outcome findings, such as a substantial excess in cardiovascular death in the prior study as opposed to no statistically increased cardiovascular mortality for use of mitoKATP channel high-affinity sulfonylureas compared with low-affinity sulfonylureas in this current study. Although there may be differences in study cohort characteristics and diabetes severity between the 2 studies, use of metformin may at least partly account for the contradictory data. Despite inconsistent findings,28 metformin could be associated with reduced risk of cardiovascular-related events and cardiovascular mortality in patients with T2D,29,30 and this benefit may offset or modify the risk of the cardiovascular events associated with use of mitoKATP channel high-affinity sulfonylureas. On the other hand, the Action in Diabetes and Vascular Disease: Preterax and Diamicron MR-Controlled Evaluation (ADVANCE) and The Cardiovascular Outcome Study of Linagliptin vs Glimepiride in Type 2 Diabetes (CAROLINA) trials found that the mitoKATP channel low-affinity sulfonylurea gliclazide modified release vs other antidiabetic medication and linagliptin vs the mitoKATP channel low-affinity sulfonylureas glimepiride, respectively, had comparable outcomes on the risk of MACEs.31,32 Although not directly comparable to this current study, the data from the 2 large prospective studies31,32 indirectly support our reported data.

Overall, our findings are supported by biological plausibility. In vitro and animal data revealed that certain sulfonylureas could jeopardize or abolish ischemic preconditioning, the self-cardiac protection mechanism for minimizing a potentially lethal ischemic result, through antagonism of cardiac mitoKATP channels, which could subsequently cause larger infarct sizes or lead to death during an acute ischemic event.16,33,34 Our findings are in line with these preclinical data, offering insights into the cardiovascular safety of sulfonylureas combined with metformin in patients with T2D.

The findings on the hypoglycemia outcome are supported by the prior literature revealing that glyburide had the highest risk of hypoglycemia, followed by glimepiride and glipizide.35,36 In our study, glyburide and glimepiride comprised 76.8% and 74.9% in the cardiac mitoKATP channel high-affinity sulfonylurea and in low-affinity sulfonylurea groups, respectively. This could explain why there was a higher risk of hypoglycemia in the high-affinity sulfonylurea users compared with the low-affinity sulfonylurea users. Additionally, the main findings persisted after adjusting for occurrence of hypoglycemia during follow-up, which may rule out the possibility that hypoglycemia acts as an intermediator for our observed associations.

Our data have implications in prescribing practice and comparative safety research of antidiabetic agents in T2D. As cardiac mitoKATP channel low-affinity sulfonylureas were found to have lower risks of both 3-point MACEs and hypoglycemia, as compared with cardiac mitoKATP channel high-affinity sulfonylureas, we recommend low-affinity sulfonylureas over high-affinity sulfonylureas in cases where sulfonylurea treatment is considered for patients with T2D. This recommendation about the choice of sulfonylureas is of great clinical importance, given that sulfonylureas are one of the most used antidiabetic drugs after metformin in current clinical settings, despite the presence of newer antidiabetic medications. Additionally, some newer antidiabetic agents have been compared to sulfonylureas as a class regarding the risk of cardiovascular diseases,37,38 but the reported comparative data may have been associated with the compositions of varying mitoKATP channel affinity sulfonylurea. Future comparative studies involving sulfonylureas are suggested to take the affinity of sulfonylureas to cardiac mitoKATP channels into account.

Strengths and Limitations

This study had several strengths. To our knowledge, this is the first study that showed pharmacological differences in sulfonylureas combined with metformin regarding their differential affinity to cardiac mitoKATP channels, which could account for the sulfonylurea intraclass difference in the risk of MACEs among sulfonylurea agents. The inclusion of a nationwide diabetic population with the most used sulfonylurea regimen as add-on therapy to metformin substantially increases the generalizability of our findings. The validity of the composite outcome of 3-point MACEs is expected to be high, as the coding algorithms for identifying hospitalization for MI or ischemic stroke were validated with high accuracy,21,22 and cardiovascular deaths were determined through a nationwide death registry.

Several limitations of this study merit discussion. First, our study is potentially subjected to confounding by indication or disease severity. We attempted to mitigate this potential confounding by maintaining a balanced average daily dose of metformin and all proxies of diabetes severity between the 2 groups. Second, unmeasured confounders, such as smoking, could possibly confound the reported findings; however, we used an active-comparator design, and reached consistent findings when using the high-dimensional–PS approach. Third, the adopted as-treated exposure analyses were prone to potential informative censoring; nevertheless, the alternative intention-to-treat analysis yielded similar results as the main findings. Fourth, the long-term effect of cardiac mitoKATP channel low-affinity sulfonylureas and high-affinity sulfonylureas on cardiovascular safety could not be assessed due to the fairly short follow-up period. The observed short-term use of sulfonylureas combined with metformin, however, reflects clinical medication use among patients with T2D. Fifth, the adoption of a PS-matching approach led to reductions in sample sizes, and we alternatively adopted the inverse probability of treatment weighting approach which maintained the original sample sizes and led to the consistent results. Sixth, we may not have sufficient statistical power in secondary and subgroup analyses including duration analyses, which needs to be taken into consideration when interpreting the findings.

Conclusions

MitoKATP channel high-affinity sulfonylureas vs low-affinity sulfonylureas, when combined with metformin, were associated with an increased risk of MACE, suggesting that the high-affinity blockage of cardiac mitoKATP channels may act as an important determinant of sulfonylurea-related adverse cardiovascular events in patients with T2D.

Supplement 1.

eMethods. The Propensity Score Calibration Analysis

eReferences

eTable 1. Operational Definitions for the Adopted Exclusion Criteria, Outcomes, Comorbidities, and Comedications

eTable 2. The Mean Duration and the Reasons for Truncation During Follow-Up for Add-On Mitochondrial KATP Channel High-Affinity and Low-Affinity Sulfonylurea Groups, by Outcomes

eTable 3. Number Needed to Harm for Add-On Mitochondrial KATP Channel High-Affinity and Low-Affinity Sulfonylureas In Patients Continuously Receiving Metformin Monotherapy

eFigure 1. Kaplan-Meier Survival Curves of 3-Point Major Adverse Cardiovascular Events (A), Myocardial Infarction (B), Ischemic Stroke (C), and Cardiovascular Death (D) Between Metformin Users With Add-On Mitochondrial KATP Channel High-Affinity Sulfonylurea and Mitochondrial KATP Channel Low-Affinity Sulfonylurea

eFigure 2. Kaplan-Meier Survival Curves of Arrhythmia (A), Heart Failure (B), All-Cause Mortality (C), and Severe Hypoglycemia (D) Between Metformin Users With Add-On Mitochondrial KATP Channel High-Affinity Sulfonylurea and Mitochondrial KATP Channel Low-Affinity Sulfonylurea

eFigure 3. Employment of the Rule-Out Approach to Assess the Impact of Unmeasured Confounding on the Main Findings

Click here for additional data file. (808.9KB, pdf)
Supplement 2.

Data Sharing Statement

Click here for additional data file. (15.2KB, pdf)

References

  • 1.International Diabetes Foundation . IDF diabetes atlas, 10th edition 2021. Accessed July 22, 2022. http://www.diabetesatlas.org
  • 2.Gregg EW, Cheng YJ, Srinivasan M, et al. Trends in cause-specific mortality among adults with and without diagnosed diabetes in the USA: an epidemiological analysis of linked national survey and vital statistics data. Lancet. 2018;391(10138):2430-2440. doi: 10.1016/S0140-6736(18)30314-3 [DOI] [PubMed] [Google Scholar]
  • 3.Maladkar M, Patil S, Viswanathan H. Redefining the therapeutic wheel of diabetes management. J Cardiol Cardiovasc Ther. 2018;12(2):28-34. doi: 10.19080/JOCCT.2018.12.555832 [DOI] [Google Scholar]
  • 4.Scheen AJ. Sulphonylureas in the management of type 2 diabetes: to be or not to be? Diabetes Epidemiol Manag. 2021;1:100002. doi: 10.1016/j.deman.2021.100002 [DOI] [Google Scholar]
  • 5.Chu WM, Ho HE, Huang KH, et al. The prescribing trend of oral antidiabetic agents for type 2 diabetes in Taiwan: an 8-year population-based study. Medicine. 2017;96(43):e8257. doi: 10.1097/MD.0000000000008257 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Montvida O, Shaw J, Atherton JJ, Stringer F, Paul SK. Long-term trends in antidiabetes drug usage in the US: real-world evidence in patients newly diagnosed with type 2 diabetes. Diabetes Care. 2018;41(1):69-78. doi: 10.2337/dc17-1414 [DOI] [PubMed] [Google Scholar]
  • 7.Pop LM, Lingvay I. The infamous, famous sulfonylureas and cardiovascular safety: much ado about nothing? Curr Diab Rep. 2017;17(12):124. doi: 10.1007/s11892-017-0954-4 [DOI] [PubMed] [Google Scholar]
  • 8.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]
  • 9.Abdelmoneim AS, Eurich DT, Gamble JM, et al. Risk of acute coronary events associated with glyburide compared with gliclazide use in patients with type 2 diabetes: a nested case-control study. Diabetes Obes Metab. 2014;16(1):22-29. doi: 10.1111/dom.12173 [DOI] [PubMed] [Google Scholar]
  • 10.Abdelmoneim AS, Eurich DT, Senthilselvan A, Qiu W, Simpson SH. Dose-response relationship between sulfonylureas and major adverse cardiovascular events in elderly patients with type 2 diabetes. Pharmacoepidemiol Drug Saf. 2016;25(10):1186-1195. doi: 10.1002/pds.4014 [DOI] [PubMed] [Google Scholar]
  • 11.Douros A, Dell’Aniello S, Yu OHY, Filion KB, Azoulay L, Suissa S. Sulfonylureas as second line drugs in type 2 diabetes and the risk of cardiovascular and hypoglycaemic events: population based cohort study. BMJ. 2018;362:k2693. doi: 10.1136/bmj.k2693 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Douros A, Yin H, Yu OHY, Filion KB, Azoulay L, Suissa S. Pharmacologic differences of sulfonylureas and the risk of adverse cardiovascular and hypoglycemic events. Diabetes Care. 2017;40(11):1506-1513. doi: 10.2337/dc17-0595 [DOI] [PubMed] [Google Scholar]
  • 13.Foster MN, Coetzee WA. KATP channels in the cardiovascular system. Physiol Rev. 2016;96(1):177-252. doi: 10.1152/physrev.00003.2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.O’Rourke B. Myocardial KATP channels in preconditioning. Circ Res. 2000;87(10):845-855. doi: 10.1161/01.RES.87.10.845 [DOI] [PubMed] [Google Scholar]
  • 15.Ardehali H, O’Rourke B. Mitochondrial KATP channels in cell survival and death. J Mol Cell Cardiol. 2005;39(1):7-16. doi: 10.1016/j.yjmcc.2004.12.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Argaud L, Garrier O, Loufouat J, et al. Second-generation sulfonylureas preserve inhibition of mitochondrial permeability transition by the mitochondrial K+ATP opener nicorandil in experimental myocardial infarction. Shock. 2009;32(3):247-252. doi: 10.1097/SHK.0b013e31819c3794 [DOI] [PubMed] [Google Scholar]
  • 17.O’Rourke B. Evidence for mitochondrial K+ channels and their role in cardioprotection. Circ Res. 2004;94(4):420-432. doi: 10.1161/01.RES.0000117583.66950.43 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Sato T, Nishida H, Miyazaki M, Nakaya H. Effects of sulfonylureas on mitochondrial ATP-sensitive K+ channels in cardiac myocytes: implications for sulfonylurea controversy. Diabetes Metab Res Rev. 2006;22(5):341-347. doi: 10.1002/dmrr.621 [DOI] [PubMed] [Google Scholar]
  • 19.Szewczyk A, Wójcik G, Lobanov NA, Nałecz MJ. The mitochondrial sulfonylurea receptor: identification and characterization. Biochem Biophys Res Commun. 1997;230(3):611-615. doi: 10.1006/bbrc.1996.6023 [DOI] [PubMed] [Google Scholar]
  • 20.Wang MT, Huang YL, Lai JH, et al. Association between specificity of sulfonylureas to cardiac mitochondrial KATP channels and the risk of major adverse cardiovascular events in type 2 diabetes. Diabetes Care. 2022;45(5):1276-1287. doi: 10.2337/dc21-1779 [DOI] [PubMed] [Google Scholar]
  • 21.Cheng CL, Lee CH, Chen PS, Li YH, Lin SJ, Yang YH. Validation of acute myocardial infarction cases in the National Health Insurance Research Database in Taiwan. J Epidemiol. 2014;24(6):500-507. doi: 10.2188/jea.JE20140076 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Hsieh CY, Chen CH, Li CY, Lai ML. Validating the diagnosis of acute ischemic stroke in a National Health Insurance claims database. J Formos Med Assoc. 2015;114(3):254-259. doi: 10.1016/j.jfma.2013.09.009 [DOI] [PubMed] [Google Scholar]
  • 23.Stürmer T, Schneeweiss S, Avorn J, Glynn RJ. Adjusting effect estimates for unmeasured confounding with validation data using propensity score calibration. Am J Epidemiol. 2005;162(3):279-289. doi: 10.1093/aje/kwi192 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Schneeweiss S. Sensitivity analysis and external adjustment for unmeasured confounders in epidemiologic database studies of therapeutics. Pharmacoepidemiol Drug Saf. 2006;15(5):291-303. doi: 10.1002/pds.1200 [DOI] [PubMed] [Google Scholar]
  • 25.Schneeweiss S, Rassen JA, Glynn RJ, Avorn J, Mogun H, Brookhart MA. High-dimensional propensity score adjustment in studies of treatment effects using health care claims data. Epidemiology. 2009;20(4):512-522. doi: 10.1097/EDE.0b013e3181a663cc [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Hess LM, Raebel MA, Conner DA, Malone DC. Measurement of adherence in pharmacy administrative databases: a proposal for standard definitions and preferred measures. Ann Pharmacother. 2006;40(7-8):1280-1288. doi: 10.1345/aph.1H018 [DOI] [PubMed] [Google Scholar]
  • 27.Robins JM, Hernán MA, Brumback B. Marginal structural models and causal inference in epidemiology. Epidemiology. 2000;11(5):550-560. doi: 10.1097/00001648-200009000-00011 [DOI] [PubMed] [Google Scholar]
  • 28.Griffin SJ, Leaver JK, Irving GJ. Impact of metformin on cardiovascular disease: a meta-analysis of randomised trials among people with type 2 diabetes. Diabetologia. 2017;60(9):1620-1629. doi: 10.1007/s00125-017-4337-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Han Y, Xie H, Liu Y, Gao P, Yang X, Shen Z. Effect of metformin on all-cause and cardiovascular mortality in patients with coronary artery diseases: a systematic review and an updated meta-analysis. Cardiovasc Diabetol. 2019;18(1):96. doi: 10.1186/s12933-019-0900-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Zhang K, Yang W, Dai H, Deng Z. Cardiovascular risk following metformin treatment in patients with type 2 diabetes mellitus: results from meta-analysis. Diabetes Res Clin Pract. 2020;160:108001. doi: 10.1016/j.diabres.2020.108001 [DOI] [PubMed] [Google Scholar]
  • 31.Zoungas S. ADVANCE in context: the benefits, risks and feasibility of providing intensive glycaemic control based on gliclazide modified release. Diabetes Obes Metab. 2020;22(suppl 2):5-11. doi: 10.1111/dom.13968 [DOI] [PubMed] [Google Scholar]
  • 32.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]
  • 33.Skalska J, Debska G, Kunz WS, Szewczyk A. Antidiabetic sulphonylureas activate mitochondrial permeability transition in rat skeletal muscle. Br J Pharmacol. 2005;145(6):785-791. doi: 10.1038/sj.bjp.0706214 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Szewczyk A. Intracellular targets for antidiabetic sulfonylureas and potassium channel openers. Biochem Pharmacol. 1997;54(9):961-965. doi: 10.1016/S0006-2952(97)00136-6 [DOI] [PubMed] [Google Scholar]
  • 35.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]
  • 36.Leonard CE, Han X, Brensinger CM, et al. Comparative risk of serious hypoglycemia with oral antidiabetic monotherapy: a retrospective cohort study. Pharmacoepidemiol Drug Saf. 2018;27(1):9-18. doi: 10.1002/pds.4337 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Dawwas GK, Smith SM, Park H. Cardiovascular outcomes of sodium glucose cotransporter-2 inhibitors in patients with type 2 diabetes. Diabetes Obes Metab. 2019;21(1):28-36. doi: 10.1111/dom.13477 [DOI] [PubMed] [Google Scholar]
  • 38.Patorno E, Goldfine AB, Schneeweiss S, et al. Cardiovascular outcomes associated with canagliflozin versus other non-gliflozin antidiabetic drugs: population based cohort study. BMJ. 2018;360:k119. doi: 10.1136/bmj.k119 [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.

eMethods. The Propensity Score Calibration Analysis

eReferences

eTable 1. Operational Definitions for the Adopted Exclusion Criteria, Outcomes, Comorbidities, and Comedications

eTable 2. The Mean Duration and the Reasons for Truncation During Follow-Up for Add-On Mitochondrial KATP Channel High-Affinity and Low-Affinity Sulfonylurea Groups, by Outcomes

eTable 3. Number Needed to Harm for Add-On Mitochondrial KATP Channel High-Affinity and Low-Affinity Sulfonylureas In Patients Continuously Receiving Metformin Monotherapy

eFigure 1. Kaplan-Meier Survival Curves of 3-Point Major Adverse Cardiovascular Events (A), Myocardial Infarction (B), Ischemic Stroke (C), and Cardiovascular Death (D) Between Metformin Users With Add-On Mitochondrial KATP Channel High-Affinity Sulfonylurea and Mitochondrial KATP Channel Low-Affinity Sulfonylurea

eFigure 2. Kaplan-Meier Survival Curves of Arrhythmia (A), Heart Failure (B), All-Cause Mortality (C), and Severe Hypoglycemia (D) Between Metformin Users With Add-On Mitochondrial KATP Channel High-Affinity Sulfonylurea and Mitochondrial KATP Channel Low-Affinity Sulfonylurea

eFigure 3. Employment of the Rule-Out Approach to Assess the Impact of Unmeasured Confounding on the Main Findings

Click here for additional data file. (808.9KB, pdf)
Supplement 2.

Data Sharing Statement

Click here for additional data file. (15.2KB, pdf)

Articles from JAMA Network Open are provided here courtesy of American Medical Association

RESOURCES