Abstract
Background
Type 2 diabetes mellitus (T2DM) and gout are associated with an increased risk of cardiovascular events. Despite the approval for the secondary prevention of cardiovascular diseases by the United States Food and Drug Administration in 2023, evidence regarding the effectiveness of colchicine among T2DM population remains limited.
Objectives
We aimed to evaluate the association between the use of colchicine and the risk of major adverse cardiovascular events (MACE) among patients with gout and T2DM.
Methods
We conducted a nationwide, population‐based cohort study with active comparator, new‐user design using nationwide claims data of South Korea (2010–2022). Patients with T2DM and gout who initiated colchicine or non‐steroidal anti‐inflammatory drugs (NSAIDs) from 2011 to 2022 were included. The primary outcome was MACE (myocardial infarction, ischemic stroke, and cardiovascular death). Secondary outcomes were each individual components of primary outcome and hospitalization due to heart failure. As‐treated approach with 30‐day grace period was applied.
Results
Before propensity score (PS) matching, 13,019 colchicine users and 111,594 NSAIDs users were included in the study cohort (mean age, 65.5 vs. 62.9; 35.0% vs. 29.8% female). After 1:2 PS matching, 12,908 colchicine users and 25,816 NSAIDs users remained (mean age, 65.7 vs. 65.7 years; 35.2% vs. 35.1% female). The PS‐matched hazard ratio for MACE was 0.94 (95% confidence interval 0.65–1.36), and all secondary outcomes also resulted in null findings.
Conclusions
Use of colchicine does not significantly reduce the risk of MACE compared with NSAIDs in a real‐world population with T2DM and gout in South Korea between 2011 and 2022.
Keywords: cardiovascular events, colchicine, gout, Type 2 diabetes
Abbreviations
- BMI
body mass index
- CI
confidence interval
- HRs
hazard ratios
- ICD‐10
International Classification of Diseases, 10th revision
- IPCW
inverse probability censoring weights
- ITT
intention‐to‐treat
- MACE
major adverse cardiovascular events
- NHIS
National Health Insurance Service
- NSAIDs
non‐steroidal anti‐inflammatory drugs
- PS
propensity score
- T2DM
Type 2 diabetes mellitus
Introduction
Type 2 diabetes mellitus (T2DM) and gout are both chronic conditions with a steadily rising disease burden [1, 2]. Several studies suggested that there might be interactions between these two diseases [3, 4]. Gout increases the risk of cardiovascular diseases and may exacerbate T2DM by increasing insulin resistance, which also heightens cardiovascular risk by contributing to vascular stiffness [5, 6, 7]. As T2DM is a known risk factor of cardiovascular diseases, patients with both T2DM and gout have higher risk of adverse cardiovascular events, a major contributor of premature death, than those without T2DM or gout [8, 9]. Thus, effective management of cardiovascular disease in patients with T2DM, especially those with comorbid gout, is crucial.
Colchicine is an anti‐inflammatory drug widely used to manage and prevent gout flares by inhibiting neutrophil mobility [10]. Due to its anti‐inflammatory properties, colchicine was hypothesized to have a preventative effect on major adverse cardiovascular events (MACE). Clinical trials such as COLCOT and LoDoCo2 have explored the association between colchicine use and MACE, leading the United States Food and Drug Administration to approve colchicine as the first anti‐inflammatory drug for secondary prevention of cardiovascular disease in June 2023 [11]. However, these trials included participants with a history of myocardial infarction or chronic coronary disease, leaving the primary preventive effect of colchicine on individuals without cardiovascular comorbidities unclear. Moreover, the results varied between trials: the COLCOT trial found a greater effect of colchicine on preventing MACE in patients with diabetes, whereas the LoDoCo2 trial did not find difference in effect based on the presence of diabetes [12, 13, 14, 15]. Considering the effect of T2DM itself as a major risk factor of MACE, the current clinical evidence for those with T2DM remains insufficient.
Given the scarcity of evidence from those with both T2DM and gout and the different results observed in clinical trials, further research among those population is warranted on the cardiovascular effectiveness of colchicine. To bridge these research gaps, we aimed to investigate the association between colchicine use and the risk of MACE among patients with T2DM and gout by conducting a population‐based cohort study using the nationwide claims data from South Korea.
Methods
Data source
This population‐based, active comparator, new‐user, nationwide cohort study used claims data of the National Health Insurance Service (NHIS) database from South Korea, January 2010 to December 2022. As enrollment to the national insurance program is mandatory in South Korea, the NHIS database encompasses virtually the entire Korean population, totaling over 50 million residents. The database includes sociodemographic information, medical diagnoses, history of medical facility admissions, inpatient and outpatient prescriptions, biennial health examination results, and death records. Diagnoses were documented according to the International Statistical Classification of Diseases and Related Health Problems, 10th Revision (International Classification of Diseases, 10th revision [ICD‐10]) coding system. Medications are documented using domestic National Health Insurance coding systems, which contain information about ingredients, dosage, formulation, prescription dose, and administration period. Biennial health examinations include anthropometric data, laboratory measurements, and self‐reported questionnaires. Death records, obtained from both in‐hospital and out‐of‐hospital settings, were linked to cause‐of‐death data from Statistics Korea. The study protocol was reviewed and approved by the institutional review board of Sungkyunkwan University (SKKU 2024‐05‐013).
Study population
We constructed an active comparator cohort to compare cardiovascular safety of colchicine and non‐steroidal anti‐inflammatory drugs (NSAIDs), which are considered the same indication as colchicine for gout flare management and prophylaxis [16, 17, 18]. Patients registered in NHIS, who diagnosed with T2DM (ICD‐10 codes E11–E14) and gout (M10) any time before the cohort entry date and prescribed with colchicine or NSAIDs between January 1, 2011 and December 31, 2022 were included. Table S2 lists the generic names of drugs used to define the study cohort. Then, we excluded patients aged under 18, as this population has low prevalence of T2DM and gout. Patients prescribed with both colchicine and NSAIDs at the cohort entry date were also excluded to minimize exposure misclassification. We excluded patients having a record of colchicine or NSAIDs prescription for gout during the 1‐year period before cohort entry date to define new use of the study drug and to prevent prevalent user bias. Patients with a record of MACE, coronary artery disease, or heart failure within 90 days before cohort entry date were also excluded to mitigate potential outcome misclassification. Lastly, we excluded patients with a record of end‐stage renal disease or dialysis, as colchicine and NSAIDs are contraindicated for those conditions.
Outcomes and follow‐up
The primary outcome of our study was MACE, defined as a composite of inpatient‐recorded myocardial infarction or ischemic stroke and cardiovascular death. Secondary outcomes were individual components of MACE and hospitalization for heart failure. Primary and secondary outcomes were identified using the ICD‐10 codes—myocardial infarction (I21, I22), ischemic stroke (I63, I64), cardiovascular death (death records with cause of death “I”), and hospitalization for heart failure (I110, I971, I50). The definitions of outcomes have been validated to have a positive predictive value of 92.0% for myocardial infarction, 90.5% for hemorrhagic stroke, 87.5% for ischemic stroke, and 82.1% for hospitalization for heart failure [19, 20].
Patients were considered exposed from the cohort entry date, defined as the date of first prescription of the study drug alongside a gout diagnosis, until the earliest of: outcome occurrence, treatment discontinuation, switching or addition of study drugs, all‐cause death, or the end of the study period (December 31, 2022; Figure S1). We allowed a 30‐day grace period between non‐overlapping prescriptions to define a continuous treatment episode.
Covariates
In order to minimize confounding, we used more than 50 covariates for this study. Demographic characteristics (age and sex) were assessed at the cohort entry date. Antidiabetic medication use (alpha‐glucosidase inhibitors, glucagon‐like peptide‐1 receptor agonists, dipeptidyl peptidase‐4 inhibitors, insulin, meglitinides, metformin, sodium‐glucose co‐transporter‐2 inhibitors, sulfonylureas, and thiazolidinediones), number of antidiabetic medications (0–1, 2–3, and ≥4), level of diabetes treatment (level 1: none or a single class of antidiabetic medication use without insulin; level 2: two or more classes of antidiabetic medication use without insulin; level 3: insulin use with or without other antidiabetic medications), and diabetes‐related conditions (diabetic ketoacidosis, diabetic nephropathy, diabetic neuropathy, diabetic retinopathy, diabetes with peripheral circulatory disorders, and hypoglycemia) were used as proxies of diabetes severity and were assessed 1 year prior to the cohort entry date. Comorbid medical conditions (cancer, cerebrovascular disease, heart failure, peripheral vascular disease, dyslipidemia, hypertension, acute kidney injury, chronic kidney disease, chronic obstructive pulmonary disease, pneumonia, obstructive sleep apnea, and venous thromboembolism), Charlson Comorbidity Index (0, 1, 2, and ≥3), gout medications (xanthine oxidase inhibitor, and corticosteroids), and other concurrent medication use (angiotensin‐converting enzyme inhibitors/angiotensin receptor blockers, angiotensin receptor/neprilysin inhibitors, beta‐blockers, calcium channel blockers, diuretics, nitrates, statins, other lipid‐lowering agents, oral antiplatelets, and oral anticoagulants) were evaluated 1 year prior to the cohort entry date. As proxies for health‐seeking behavior, we included the number of outpatient and inpatient visits 1 year prior to the cohort entry date. Body mass index (BMI), drinking and smoking status, and lipid levels were included from the national health screening program, which were assessed on the date closest to the cohort entry date. The codes used to define the covariates were listed in Table S1.
Statistical analysis
To adjust for potential confounding, 1:2 propensity scores (PS) matching with nearest‐neighbor algorithm was applied. Baseline characteristics of patients with both T2DM and gout taking colchicine or NSAIDs were compared before and after matching. We presented continuous variables as means and standard deviations and categorial variables as numbers and percentages. We considered absolute standardized mean differences less than 0.1 well balanced for each individual covariate between study groups. We computed PS by fitting a multivariable logistic regression model on the covariates to estimate the probability of being exposed to colchicine as opposed to being exposed to NSAIDs. All covariates were included in the model except for BMI, drinking and smoking status, and lipid levels due to their high missing rates.
For each outcome, incidence rates per 100 person‐years and 95% confidence intervals (CIs) were presented based on a Poisson distribution. We visualized the cumulative incidence of MACE using Kaplan‐Meier plots. Cox proportional hazards models were used to estimate both crude and PS‐matched hazard ratios (HRs) and their 95% CIs for the risk of cardiovascular outcomes. Crude HRs were estimated in unmatched population, whereas PS‐matched HRs were estimated among 1:2 PS‐matched population. We defined statistical significance as a 95% CI that did not include 1. All analyses were performed using SAS version 9.4 (SAS Institute).
Subgroup analyses
We conducted several subgroup analyses to assess the potential of effect modifications. Subgroups were stratified by demographic features including sex, age groups (18–44, 45–64, and ≥65), and relevant comorbidities (coronary artery disease, hypertension, and heart failure). History of coronary artery disease and heart failure were accessed between 91 and 365 days prior to the cohort entry date. We also stratified subgroups by the level of diabetic treatments to evaluate effect modification by the severity of diabetes. To assess any duration‐response relationship, we conducted a piecewise Cox regression analysis with a change‐point at 120 days. The HR before 120 days was estimated from the cohort entry date until the earliest of Day 120, the occurrence of the outcome, or the censoring criteria, whereas the HR past 120 days was estimated from Day 120 until the earliest of the outcome occurrence or censoring criteria [21].
Sensitivity analyses
We performed various sensitivity analyses to confirm the robustness of this study. First, we repeated the main analysis using an intention‐to‐treat (ITT) approach to define follow‐up. In ITT approach, patients were followed from the cohort entry date until the earliest of outcome occurrence, 365 days after the cohort entry date, death, or end of the study period (December 31, 2022). Second, we used an expanded definition of MACE to capture a broader spectrum of cardiovascular outcomes by adding coronary revascularization to MACE. Coronary revascularization was defined using procedure codes for percutaneous coronary intervention and coronary artery bypass graft. Third, to assess any exposure misclassification, we used a 15‐day grace period instead of a 30‐day grace period between non‐overlapping prescriptions. Fourth, to account for potential informative censoring, we applied an inverse probability of censoring weight (IPCW) method. Covariates were updated by a 30‐day interval, and weights were truncated at 0.5% and 99.5%. Fifth, we excluded patients with a history of insulin use to rule out patients with severe diabetes. Sixth, we conducted an analysis excluding patients without a record of smoking, drinking, or BMI was also implemented to assess the possible effect of health examination participation. Finally, we adopted herpes zoster virus infection as a negative control outcome to assess the possibility of residual confounding, expecting null association.
Results
Patient characteristics
A total of 214,880 patients with T2DM diagnosed with gout were identified between 2011 and 2022. After applying the exclusion criteria, 124,613 patients newly prescribed with colchicine or NSAIDs were selected, with 13,019 (10.4%) classified as colchicine exposed and 111,594 (89.6%) as NSAIDs exposed. Following 1:2 PS matching, 12,908 colchicine users and 25,816 NSAIDs users remained.
Patient characteristics at baseline, both before and after PS matching, are presented in Table 1. Despite achieving excellent balance across all covariates after PS matching, colchicine users differed from NSAIDs users in several respects at baseline. Colchicine users were older (average age: 65.5 vs. 62.9), had a higher Charlson Comorbidity Index (≥3: 22.8% vs. 15.7%), and experienced more frequent hospitalizations (≥3 hospitalizations: 13.0% vs. 6.8%). Although colchicine users had a higher proportion of smokers (14.7% vs. 12.4%), they had a less proportion of drinkers (31.9% vs. 41.3%). Lipid levels such as total cholesterol, high density lipoprotein, low density lipoprotein, and triglycerides were well‐balanced before and after matching.
Table 1.
Baseline characteristics of patients who received colchicine or non‐steroidal anti‐inflammatory drugs (NSAIDs) before and after propensity score (PS)‐matching.
| Characteristics | Before PS‐matching | After PS‐matching | ||||
|---|---|---|---|---|---|---|
| Colchicine | NSAIDs | aSD a | Colchicine | NSAIDs | aSD a | |
| Total no. | 13,019 | 111,594 | 12,908 | 25,816 | ||
| Age, mean (SD), years | 65.5 (12.5) | 62.9 (12.4) | 0.21 | 65.7 (12.4) | 65.7 (12.1) | 0.00 |
| Age group | ||||||
| 18–44 | 4905 (37.7) | 49,817 (45.2) | 0.15 | 4900 (38.0) | 9793 (37.9) | 0.00 |
| 45–64 | 7285 (56.0) | 51,889 (47.0) | 0.19 | 7285 (56.4) | 14,591 (56.5) | 0.00 |
| ≥65 | 829 (6.4) | 9888 (8.9) | 0.09 | 723 (5.6) | 1432 (5.5) | 0.00 |
| Sex, n (%) | ||||||
| Male | 8459 (65.0) | 78,382 (70.2) | 0.11 | 8369 (64.8) | 16,759 (64.9) | 0.00 |
| Female | 4556 (35.0) | 33,195 (29.8) | 0.11 | 4539 (35.2) | 9057 (35.1) | 0.00 |
| Cohort entry year, n (%) | ||||||
| 2011 | 1001 (7.7) | 9091 (8.1) | 0.02 | 986 (7.6) | 2017 (7.8) | 0.01 |
| 2012 | 1044 (8.0) | 9607 (8.6) | 0.02 | 1033 (8.0) | 2087 (8.1) | 0.00 |
| 2013 | 1061 (8.1) | 9521 (8.5) | 0.01 | 1054 (8.2) | 2120 (8.2) | 0.00 |
| 2014 | 1053 (8.1) | 9122 (8.2) | 0.00 | 1045 (8.1) | 2072 (8.0) | 0.00 |
| 2015 | 1031 (7.9) | 9171 (8.2) | 0.01 | 1021 (7.9) | 2052 (7.9) | 0.00 |
| 2016 | 1107 (8.5) | 9636 (8.6) | 0.01 | 1098 (8.5) | 2227 (8.6) | 0.00 |
| 2017 | 1141 (8.8) | 9680 (8.7) | 0.00 | 1134 (8.8) | 2332 (9.0) | 0.01 |
| 2018 | 1180 (9.1) | 10,088 (9.0) | 0.00 | 1174 (9.1) | 2346 (9.1) | 0.00 |
| 2019 | 1135 (8.7) | 9937 (8.9) | 0.01 | 1123 (8.7) | 2194 (8.5) | 0.01 |
| 2020 | 1060 (8.1) | 8740 (7.8) | 0.01 | 1055 (8.2) | 2071 (8.0) | 0.01 |
| 2021 | 1160 (8.9) | 8740 (7.8) | 0.04 | 1150 (8.9) | 2229 (8.6) | 0.01 |
| 2022 | 1046 (8.0) | 8261 (7.4) | 0.02 | 1035 (8.0) | 2069 (8.0) | 0.00 |
| Antidiabetic medication use, n (%) | ||||||
| Alpha‐glucosidase inhibitors | 731 (5.6) | 6519 (5.8) | 0.01 | 727 (5.6) | 1425 (5.5) | 0.01 |
| GLP1RA | 98 (0.8) | 665 (0.6) | 0.02 | 96 (0.7) | 196 (0.8) | 0.00 |
| DPP4I | 6329 (48.6) | 52,081 (46.7) | 0.04 | 6282 (48.7) | 12,471 (48.3) | 0.01 |
| Insulin | 3084 (23.7) | 18,085 (16.2) | 0.19 | 3070 (23.8) | 6196 (24.0) | 0.01 |
| Meglitinides | 253 (1.9) | 1473 (1.3) | 0.05 | 252 (2.0) | 523 (2.0) | 0.01 |
| Metformin | 8915 (68.5) | 80,098 (71.8) | 0.07 | 8840 (68.5) | 17,673 (68.5) | 0.00 |
| SGLT2I | 655 (5.0) | 6221 (5.6) | 0.02 | 645 (5.0) | 1245 (4.8) | 0.01 |
| Sulfonylureas | 5984 (46.0) | 52,229 (46.8) | 0.02 | 5948 (46.1) | 11,859 (45.9) | 0.00 |
| Thiazolidinediones | 1203 (9.2) | 10,686 (9.6) | 0.01 | 1190 (9.2) | 2405 (9.3) | 0.00 |
| No. of diabetic medications being taken | ||||||
| 0–1 | 3932 (30.2) | 33,271 (29.8) | 0.01 | 3899 (30.2) | 7813 (30.3) | 0.00 |
| 2–3 | 7532 (57.9) | 67,603 (60.6) | 0.06 | 7470 (57.9) | 14,983 (58.0) | 0.00 |
| ≥4 | 1550 (11.9) | 10,695 (9.6) | 0.08 | 1539 (11.9) | 3020 (11.7) | 0.01 |
| Level of antidiabetic treatment b | ||||||
| 1 | 3577 (27.5) | 31,143 (27.9) | 0.01 | 3544 (27.5) | 7057 (27.3) | 0.00 |
| 2 | 6353 (48.8) | 62,341 (55.9) | 0.14 | 6294 (48.8) | 12,563 (48.7) | 0.00 |
| 3 | 3084 (23.7) | 18,085 (16.2) | 0.19 | 3070 (23.8) | 6196 (24.0) | 0.01 |
| Diabetes‐related conditions, n (%) | ||||||
| Diabetic ketoacidosis | 61 (0.5) | 375 (0.3) | 0.02 | 60 (0.5) | 119 (0.5) | 0.00 |
| Diabetic nephropathy | 1406 (10.8) | 9152 (8.2) | 0.09 | 1398 (10.8) | 2903 (11.2) | 0.01 |
| Diabetic neuropathy | 2602 (20.0) | 19,390 (17.4) | 0.07 | 2589 (20.1) | 5200 (20.1) | 0.00 |
| Diabetic retinopathy | 2308 (17.7) | 17,533 (15.7) | 0.05 | 2300 (17.8) | 4516 (17.5) | 0.01 |
| Diabetes with peripheral circulatory disorders | 1668 (12.8) | 12,898 (11.6) | 0.04 | 1663 (12.9) | 3308 (12.8) | 0.00 |
| Hypoglycemia | 650 (5.0) | 4447 (4.0) | 0.05 | 646 (5.0) | 1249 (4.8) | 0.01 |
| Comorbid medical conditions, n (%) | ||||||
| Cancer | 2161 (16.6) | 14,830 (13.3) | 0.09 | 2146 (16.6) | 4320 (16.7) | 0.00 |
| Cerebrovascular disease | 1764 (13.6) | 11,274 (10.1) | 0.11 | 1761 (13.6) | 3567 (13.8) | 0.01 |
| Coronary artery disease | 1823 (14.0) | 11,837 (10.6) | 0.10 | 1821 (14.1) | 3597 (13.9) | 0.01 |
| Heart failure | 1241 (9.5) | 6407 (5.7) | 0.14 | 1236 (9.6) | 2413 (9.3) | 0.01 |
| Peripheral vascular disease | 1342 (10.3) | 9773 (8.8) | 0.05 | 1337 (10.4) | 2648 (10.3) | 0.00 |
| Dyslipidemia | 5102 (39.2) | 41,592 (37.3) | 0.04 | 5056 (39.2) | 10,027 (38.8) | 0.01 |
| Hypertension | 8022 (61.6) | 66,431 (59.5) | 0.04 | 7967 (61.7) | 15,948 (61.8) | 0.00 |
| Acute kidney injury | 243 (1.9) | 961 (0.9) | 0.09 | 243 (1.9) | 500 (1.9) | 0.00 |
| Chronic kidney disease | 1409 (10.8) | 6525 (5.8) | 0.18 | 1404 (10.9) | 2826 (10.9) | 0.00 |
| Chronic obstructive pulmonary disease | 526 (4.0) | 3160 (2.8) | 0.07 | 524 (4.1) | 1051 (4.1) | 0.00 |
| Pneumonia | 812 (6.2) | 4761 (4.3) | 0.09 | 811 (6.3) | 1617 (6.3) | 0.00 |
| Obstructive sleep apnea | 323 (2.5) | 1942 (1.7) | 0.05 | 320 (2.5) | 616 (2.4) | 0.01 |
| Venous thromboembolism | 79 (0.6) | 527 (0.5) | 0.02 | 79 (0.6) | 170 (0.7) | 0.01 |
| Charlson Comorbid Index, n (%) | ||||||
| 0 | 3635 (27.9) | 41,889 (37.5) | 0.21 | 3593 (27.8) | 7142 (27.7) | 0.00 |
| 1 | 4796 (36.9) | 38,769 (34.7) | 0.04 | 4764 (36.9) | 9614 (37.2) | 0.01 |
| 2 | 1621 (12.5) | 13,343 (12.0) | 0.02 | 1599 (12.4) | 3258 (12.6) | 0.01 |
| ≥3 | 2962 (22.8) | 17,568 (15.7) | 0.18 | 2952 (22.9) | 5802 (22.5) | 0.01 |
| Other gout medications, n (%) | ||||||
| Xanthine oxidase inhibitor | 1984 (15.2) | 16,456 (14.7) | 0.01 | 1970 (15.3) | 4117 (15.9) | 0.02 |
| Corticosteroids | 8387 (64.4) | 62,921 (56.4) | 0.17 | 8324 (64.5) | 16,693 (64.7) | 0.00 |
| Concurrent medication use, n (%) | ||||||
| ACEI/ARB | 8213 (63.1) | 66,357 (59.5) | 0.07 | 8157 (63.2) | 16,414 (63.6) | 0.01 |
| ARNi | 47 (0.4) | 148 (0.1) | 0.05 | 47 (0.4) | 79 (0.3) | 0.04 |
| Beta‐blocker | 4308 (33.1) | 28,283 (25.4) | 0.17 | 4289 (33.2) | 8620 (33.4) | 0.00 |
| Calcium channel blocker | 6546 (50.3) | 51,319 (46.0) | 0.09 | 6499 (50.3) | 13,003 (50.4) | 0.00 |
| Diuretics | 6492 (49.9) | 45,947 (41.2) | 0.18 | 6454 (50.0) | 12,990 (50.3) | 0.01 |
| Nitrates | 1384 (10.6) | 7931 (7.1) | 0.12 | 1382 (10.7) | 2739 (10.6) | 0.00 |
| Statin | 7923 (60.9) | 64,350 (57.7) | 0.07 | 7870 (61.0) | 15,737 (61.0) | 0.00 |
| Other lipid‐lowering agents | 1354 (10.4) | 10,957 (9.8) | 0.02 | 1341 (10.4) | 2688 (10.4) | 0.00 |
| Oral antiplatelets | 9697 (74.5) | 80,150 (71.8) | 0.06 | 9624 (74.6) | 19,287 (74.7) | 0.00 |
| Oral anticoagulants | 1567 (12.0) | 7462 (6.7) | 0.19 | 1565 (12.1) | 3135 (12.1) | 0.00 |
| Measures of healthcare utilization | ||||||
| No. of hospitalizations, n (%) | ||||||
| 0 | 7722 (59.3) | 80,228 (71.9) | 0.27 | 7635 (59.1) | 15,212 (58.9) | 0.01 |
| 1–2 | 3610 (27.7) | 23,778 (21.3) | 0.15 | 3589 (27.8) | 7255 (28.1) | 0.01 |
| ≥3 | 1687 (13.0) | 7588 (6.8) | 0.21 | 1684 (13.0) | 3349 (13.0) | 0.00 |
| No. of outpatient visits, n (%) | ||||||
| 0–2 | 98 (0.8) | 1263 (1.1) | 0.04 | 94 (0.7) | 177 (0.7) | 0.01 |
| 3–5 | 191 (1.5) | 2612 (2.3) | 0.06 | 185 (1.4) | 384 (1.5) | 0.01 |
| ≥6 | 12,730 (97.8) | 107,719 (96.5) | 0.08 | 12,629 (97.8) | 25,255 (97.8) | 0.00 |
| BMI, kg/m2, n (%) c , d | ||||||
| <18.5 (underweight) | 144 (1.1) | 997 (0.9) | 0.02 | 142 (1.1) | 264 (1.0) | 0.01 |
| 18.5–22.9 (normal) | 2296 (17.6) | 19,159 (17.2) | 0.01 | 2280 (17.7) | 4561 (17.7) | 0.00 |
| 23.0–24.9 (overweight) | 2583 (19.8) | 22,176 (19.9) | 0.00 | 2570 (19.9) | 5204 (20.2) | 0.01 |
| ≥25.0 (obese) | 6072 (46.6) | 55,227 (49.5) | 0.06 | 6010 (46.6) | 12,342 (47.8) | 0.03 |
| Missing | 1924 (14.8) | 14,035 (12.6) | 0.06 | 1906 (14.8) | 3445 (13.3) | 0.04 |
| Smoking, n (%) d | ||||||
| Ever | 1913 (14.7) | 13,889 (12.4) | 0.07 | 1895 (14.7) | 3410 (13.2) | 0.04 |
| Never | 7914 (60.8) | 64,520 (57.8) | 0.06 | 7867 (60.9) | 15,691 (60.8) | 0.00 |
| Missing | 3192 (24.5) | 33,185 (29.7) | 0.12 | 3146 (24.4) | 6715 (26.0) | 0.04 |
| Drinking, n (%) d | ||||||
| Ever | 4154 (31.9) | 46,116 (41.3) | 0.20 | 4092 (31.7) | 8972 (34.8) | 0.07 |
| Never | 6952 (53.4) | 51,589 (46.2) | 0.14 | 6921 (53.6) | 13,434 (52.0) | 0.03 |
| Missing | 1913 (14.7) | 13,889 (12.4) | 0.07 | 1895 (14.7) | 3410 (13.2) | 0.04 |
| Lipid levels d | ||||||
| Total cholesterol, mean (SD), mg/dL | 180.2 (44.3) | 183.1 (49.7) | 0.06 | 180.2 (44.3) | 181.0 (50.2) | 0.02 |
| Missing, n (%) | 6261 (48.1) | 50,253 (45.0) | 0.06 | 6217 (48.2) | 11,947 (46.3) | 0.04 |
| HDL, mean (SD), mg/dL | 49.0 (15.4) | 49.5 (20.2) | 0.03 | 49.0 (15.4) | 49.4 (16.0) | 0.03 |
| Missing, n (%) | 6261 (48.1) | 50,253 (45.0) | 0.06 | 6217 (48.2) | 11,947 (46.3) | 0.04 |
| LDL, mean (SD), mg/dL | 97.8 (39.6) | 99.5 (46.1) | 0.04 | 97.8 (39.6) | 98.5 (46.4) | 0.02 |
| Missing, n (%) | 6370 (48.9) | 51,481 (46.1) | 0.06 | 6326 (49.0) | 12,180 (47.2) | 0.04 |
| Triglyceride, mean (SD), mg/dL | 175.5 (140.7) | 182.7 (154.2) | 0.05 | 175.5 (141.1) | 175.6 (138.6) | 0.00 |
| Missing, n (%) | 6261 (48.1) | 50,254 (45.0) | 0.06 | 6217 (48.2) | 11,947 (46.3) | 0.04 |
Abbreviations: ACEI, angiotensin convert enzyme inhibitor; ARB, angiotensin‐2 receptor blocker; ARNi, angiotensin receptor neprilysin inhibitor; aSD, absolute standardized difference; BMI, body mass index; DPP4I, dipeptidyl peptidase‐4 inhibitor; GLP1RA, glycogen like peptide‐1 receptor agonist; HDL, high density lipoprotein; LDL, low density lipoprotein; SD, standard deviation; SGLT2I, sodium glucose cotransporter 2 inhibitors.
The value >0.10 indicates a significant imbalance between the exposed and unexposed groups.
Level 1: none or a single class of antidiabetic medication use without insulin; level 2: two or more classes of antidiabetic medication use without insulin; level 3: insulin use with or without other antidiabetic medications.
Asian BMI classification.
Not included in the PS model due to a high prevalence of missing data.
Major adverse cardiovascular events
Overall, 119 (0.31%) patients experienced MACE during 4747 person‐years of follow‐up. The incidence rate per 100 person‐years for MACE was 2.35 (95% CI 1.80–3.08) for those treated with colchicine and 2.65 (95% CI 2.08–3.37) for those treated with NSAIDs. Although the crude HR indicated that colchicine use was associated with a higher risk of MACE (HR 1.48, 95% CI 1.09–2.02), this association was not retained after PS matching (HR 0.94, 95% CI 0.65–1.36).
For all secondary outcomes, 95% CI of the HRs included 1, indicating no significant association. The non‐significant decrease in risks were observed for myocardial infarction (HR 0.85, 95% CI 0.41–1.75) and cardiovascular death (HR 0.89, 95% CI 0.39–2.04), whereas the risks of ischemic stroke (HR 1.04, 95% CI 0.65–1.67) and hospitalization due to heart failure (HR 1.01, 95% CI 0.72–1.41) showed null associations.
Subgroup and sensitivity analyses
When stratified by baseline cardiovascular comorbidities, such as coronary artery disease, hypertension, and heart failure, as well as the level of antidiabetic treatment, sex, age, and the treatment duration using the 120‐day mark, did not significantly modify the effect of colchicine on MACE (Figures 1, 2, 3, 4).
Fig. 1.
Flow diagram of study cohort identification. MACE, major adverse cardiovascular events; NSAIDs, non‐steroidal anti‐inflammatory drugs; PS, propensity score.
Fig. 2.
Crude and PS‐matched hazard ratios for association between colchicine and NSAIDs. CI, confidence interval; HR, hazard ratio; MACE, major adverse cardiovascular events; NSAIDs, non‐steroidal anti‐inflammatory drug; PS, propensity score.
Fig. 3.
PS‐matched hazard ratio of the association between colchicine versus NSAIDs and risk of cardiovascular events by subgroups of interest.*Piecewise Cox regression analysis was conducted to assess duration response of colchicine versus NSAIDs on cardiovascular safety. CI, confidence interval; NSAIDs, non‐steroidal anti‐inflammatory drug; PS, propensity score.
Fig. 4.
Kaplan‐Meier curve of MACE among users of colchicine and matched users of NSAIDs. MACE, major adverse cardiovascular events; NSAIDs, non‐steroidal anti‐inflammatory drugs.
The sensitivity analyses consistently supported the primary findings, with the CIs for the HRs, including the point estimate observed in the primary analysis, as shown in Table S3. The IPCW analysis produced non‐significant results (HR 0.98, 95% CI 0.33–2.93), indicating minimal influence from informative censoring. Comparable point estimates were observed across other sensitivity analyses, confirming the robustness of the study population and exposure definitions. The negative control analysis indicated no association (HR 0.94, 95% CI 0.73–1.21), demonstrating minimal influence from shared unmeasured confounders on our results.
Discussion
Considering the increasing prevalence of T2DM and the elevated cardiovascular risk in this population, we conducted a nationwide, population‐based cohort study assessing the cardiovascular effect of colchicine used for gout among a diabetic population. The use of colchicine for gout in T2DM patients was not associated with a decreased risk of MACE, myocardial infarction, stroke, cardiovascular death, and hospitalization of heart failure, compared with the use of NSAIDs. Several sensitivity analyses resulted in estimates similar with the main analysis, and the negative control outcome analysis showed null association.
Several randomized controlled trials have investigated the cardiovascular effect of colchicine among patients with the history of cardiovascular events and found that colchicine use reduced the risk of subsequent events. However, subgroup analyses by diabetes status have shown inconsistent results across trials. Although the COLCOT trial found that colchicine reduced the risk of cardiovascular events in patients with diabetes, the LoDoCo2 trial reported no such reduction, consistent with our findings [12, 13, 14, 22]. The CHANCE‐3 trial reported that short‐term (90 days) use of colchicine did not significantly reduce cardiovascular risk in the diabetic subgroup [23]. The most recent CLEAR trial reported that colchicine did not reduce cardiovascular risk, including in patients with diabetes [24]. These conflicting results from randomized clinical trials may stem from limited statistical power, as most study designs were not specifically intended to assess effects in patients with diabetes.
Several observational studies have also investigated the association between colchicine use and cardiovascular risks, resulting in conflicting outcomes, similar to those seen in randomized clinical trials [25, 26, 27, 28]. For example, studies from the United States and the United Kingdom have shown both decreased and increased cardiovascular risk with colchicine use among gout patients, whereas a study from Taiwan suggested a lower risk of cardiovascular events in diabetic patients using colchicine, though it was limited by potential biases such as immortal time bias and confounding by indication bias [25, 26, 27, 28]. Our study addressed these gaps by using a nationwide database to provide more comprehensive evidence on colchicine's cardiovascular effects, particularly among individuals with T2DM.
Colchicine has been widely used as a treatment for gout flare and prophylaxis, and its anti‐inflammatory effect also extends to reducing the risk of cardiovascular events. Colchicine exhibits its anti‐inflammatory effects by inhibiting the migration and adhesion of neutrophils to inflammatory sites. This action, which involves both cytokine and transcriptional pathways, suggests that a longer treatment duration may be needed for optimal cardiovascular protection, as observed in the CHANCE‐3 trial, which showed no benefit for shorter colchicine use [29].
According to treatment guidelines for gout, colchicine, NSAIDs, corticosteroids, and their combinations are both recommended for gout flares. For gout prophylaxis, urate‐lowering therapy with xanthine oxidase inhibitors is recommended, followed by the use of colchicine, NSAIDs, or corticosteroids [16, 17, 18]. Although the guidelines specify that the prevalence of cardiovascular diseases is high in gout population and lifestyle modifications are needed, they do not recommend specific choices of drugs for high cardiovascular risk patients. Our findings also corroborate current guidelines for gout treatment by showing non‐differential risk of cardiovascular outcomes between the two treatment options.
There are several strengths of this study. First, using a nationwide claims data that covers the entire population of South Korea, we were able to form a large T2DM patient cohort for study with sufficient representativity. Second, by using validated algorithms to define study outcomes, we were able to minimize outcome misclassification. Third, we employed an active‐comparator, new‐user design, which minimizes bias from including prevalent users and enhances comparability between treatment groups, as both colchicine and NSAIDs are used as first‐line therapies for gout. Fourth, various sensitivity analyses were conducted to show that the possibility of exposure misclassification and informative censoring are marginal.
This study also has limitations. First, we could not rule out residual confounding due to the nature of the claims database. To minimize the effects of residual confounding, we used more than 50 covariates to estimate PS, and each selected covariate was well‐balanced after PS matching. Second, we could not completely distinguish whether colchicine or NSAIDs were prescribed for gout flare or prophylaxis using ICD‐10 codes, making it difficult to evaluate cardiovascular effect of colchicine versus NSAIDs. A previous study reported that gout flares increase the risk of cardiovascular diseases [30]. Although PS matching method was used to balance the groups, the complete balance of gout flare or prophylaxis was hard to achieve. Considering the higher prescription rate of NSAIDs compared to colchicine and the gastrointestinal adverse effects of colchicine, this may imply a preference for NSAIDs in gout prophylaxis. Thus, these results should be interpreted cautiously and conservatively, as there is a possibility of overestimating cardiovascular risk of colchicine. Third, there exists a possibility of exposure misclassification. Because we used claims records for the study, the adherence of each patient to the study drug and over‐the‐counter drug use could not be captured. However, as gout is usually managed in a clinical setting, it is unlikely that patients would use over‐the‐counter NSAIDs for this condition. Additionally, sensitivity analysis with 15‐day grace period resulted in an estimate similar to that of the main analysis, implying that exposure misclassification by differential grace period is minimal. Fourth, there was a relatively low number of outcome events, which might limit the power of the study. To address this concern and assess the robustness of our findings, we conducted sensitivity analyses using an ITT approach with a maximum of 365 days of follow‐up and an expanded definition of MACE. The estimates from these sensitivity analyses were similar to the main findings. Fifth, the generalizability of our findings may be limited, as most of the study participants were ethnic Koreans. Further studies incorporating participants of different race and ethnicities are needed.
Conclusion
This nationwide, population‐based cohort study conducted from 2011 to 2022 found that colchicine use does not significantly reduce the risk of MACE in a real‐world population from South Korea with T2DM and gout. The findings from this study implies that cardiovascular risk may not be a main consideration in colchicine prescription for gout in patients with T2DM.
Conflict of interest statement
Dr. Shin received grants from the Ministry of Food and Drug Safety, the National Research Foundation of Korea, and pharmaceutical companies, including Pfizer, Celltrion, SK Bioscience, LG Chemical, Union Chimique Belge, and GlaxoSmithKline. No other relationships or activities have influenced the submitted work.
Funding information
Ministry of Food and Drug Safety, Korea, in 2024–2028, Grant: RS‐2024‐00332632; the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT), Grant: RS‐2024‐00405650
Disclosure
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Ethics statement
This study was approved by the Institutional Review Board of Sungkyunkwan University, South Korea (No. 2024‐05‐013).
Supporting information
Table S1: Codes used to define the covariates.
Table S2: Generic names of drugs used to define the study cohort.
Table S3: Sensitivity analyses: Crude and propensity score adjusted hazard ratio for association between colchicine versus non‐steroidal anti‐inflammatory drugs and risk of cardiovascular events.
Figure S1: Graphical depiction of the study design.
Jeon M, Cho Y, Bea S, You W, Cho SK, Ryu S, et al. Colchicine and the risk of major adverse cardiovascular events in patients with gout and Type 2 diabetes: A nationwide cohort study. J Intern Med. 2025;298:604–616.
Minjeong Jeon and Yongtai Cho contributed equally to this work as co‐first authors.
Contributor Information
Yoosoo Chang, Email: yoosoo@skku.edu.
Ju‐Young Shin, Email: shin.jy@skku.edu.
Data availability statement
Data generated and/or analyzed during the current study cannot be shared publicly due to the data‐sharing policy of the National Health Insurance Service (NHIS) of Korea, governed by Article 18 of the Personal Information Protection Act (“Limitation to Out‐of‐Purpose Use and Provision of Personal Information” available at https://elaw.klri.re.kr/kor_service/lawView.do?hseq=53044&lang=ENG). However, the data are available from the NHIS (study identifier: NHIS‐2023‐1‐410) on reasonable request for researchers who meet the criteria for access to confidential data (https://www.data.go.kr/en/tcs/eds/selectCoreDataView.do?coreDataInsttCode=B551182&coreDataSn=1&searchCondition2=coreDataNmEn&searchKeyword2=).
References
- 1. Tinajero MG, Malik VS. An update on the epidemiology of type 2 diabetes: a global perspective. Endocrinol Metab Clin North Am. 2021;50:337–355. [DOI] [PubMed] [Google Scholar]
- 2. Dehlin M, Jacobsson L, Roddy E. Global epidemiology of gout: prevalence, incidence, treatment patterns and risk factors. Nat Rev Rheumatol. 2020;16:380–390. [DOI] [PubMed] [Google Scholar]
- 3. Collier A, Stirling A, Cameron L, Hair M, Crosbie D. Gout and diabetes: a common combination. Postgrad Med J. 2016;92:372–378. [DOI] [PubMed] [Google Scholar]
- 4. Cho J, Dalbeth N, Petrov MS. Relationship between gout and diabetes mellitus after acute pancreatitis: a nationwide cohort study. J Rheumatol. 2020;47:917–923. [DOI] [PubMed] [Google Scholar]
- 5. Borghi C, Agabiti‐Rosei E, Johnson RJ, Kielstein JT, Lurbe E, Mancia G, et al. Hyperuricaemia and gout in cardiovascular, metabolic and kidney disease. Eur J Intern Med. 2020;80:1–11. [DOI] [PubMed] [Google Scholar]
- 6. Kim SC, Liu J, Solomon DH. Risk of incident diabetes in patients with gout: a cohort study. Arthritis Rheumatol. 2015;67:273–280. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Hill MA, Yang Y, Zhang L, Sun Z, Jia G, Parrish AR, et al. Insulin resistance, cardiovascular stiffening and cardiovascular disease. Metabolism. 2021;119:154766. [DOI] [PubMed] [Google Scholar]
- 8. Singh JA, Ramachandaran R, Yu S, Yang S, Xie F, Yun H, et al. Is gout a risk equivalent to diabetes for stroke and myocardial infarction? A retrospective claims database study. Arthritis Res Ther. 2017;19:228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Glovaci D, Fan W, Wong ND. Epidemiology of diabetes mellitus and cardiovascular disease. Curr Cardiol Rep. 2019;21:21. [DOI] [PubMed] [Google Scholar]
- 10. Lee JJ, Lee JiS, Chung MK, Ahn JK, Choi H‐J, Hong S‐J, et al. Korean guidelines for the management of gout. J Rheum Dis. 2023;30:141–150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Agepha Pharma. LODOCO (colchicine) tablets, for oral use, NDA 215727. Agepha Pharma. 2023; Available at https://www.accessdata.fda.gov/drugsatfda_docs/label/2023/215727s000lbl.pdf. [Google Scholar]
- 12. Nidorf SM, Fiolet ATL, Eikelboom JW, Schut A, Opstal TSJ, Bax WA, et al. The effect of low‐dose colchicine in patients with stable coronary artery disease: the LoDoCo2 trial rationale, design, and baseline characteristics. Am Heart J. 2019;218:46–56. [DOI] [PubMed] [Google Scholar]
- 13. Nidorf SM, Fiolet ATL, Mosterd A, Eikelboom JW, Schut A, Opstal TSJ, et al. Colchicine in patients with chronic coronary disease. N Engl J Med. 2020;383:1838–1847. [DOI] [PubMed] [Google Scholar]
- 14. Roubille F, Bouabdallaoui N, Kouz S, Waters DD, Diaz R, Maggioni AP, et al. Low‐dose colchicine in patients with type 2 diabetes and recent myocardial infarction in the Colchicine Cardiovascular Outcomes Trial (COLCOT). Diabetes Care. 2024;47:467–470. [DOI] [PubMed] [Google Scholar]
- 15. Tardif J‐C, Kouz S, Waters DD, Bertrand OF, Diaz R, Maggioni AP, et al. Efficacy and safety of low‐dose colchicine after myocardial infarction. N Engl J Med. 2019;381:2497–2505. [DOI] [PubMed] [Google Scholar]
- 16. FitzGerald JD, Dalbeth N, Mikuls T, Brignardello‐Petersen R, Guyatt G, Abeles AM, et al. 2020 American College of Rheumatology Guideline for the management of gout. Arthritis Rheumatol. 2020;72:879–895. [DOI] [PubMed] [Google Scholar]
- 17. Qaseem A, Harris RP, Forciea MA, Clinical Guidelines Committee of the American College of Physicians , Denberg TD, Barry MJ, et al. Management of acute and recurrent gout: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2017;166:58–68. [DOI] [PubMed] [Google Scholar]
- 18. Richette P, Doherty M, Pascual E, Barskova V, Becce F, Castañeda‐Sanabria J, et al. 2016 updated EULAR evidence‐based recommendations for the management of gout. Ann Rheum Dis. 2017;76:29–42. [DOI] [PubMed] [Google Scholar]
- 19. Park J, Kwon S, Choi E‐K, Choi Y‐J, Lee E, Choe W, et al. Validation of diagnostic codes of major clinical outcomes in a National Health Insurance database. International Journal of Arrhythmia. 2019;20:5. [Google Scholar]
- 20. Kim D, Yang P‐S, You SC, Sung J‐H, Jang E, Yu HT, et al. Treatment timing and the effects of rhythm control strategy in patients with atrial fibrillation: nationwide cohort study. BMJ. 2021;373:n991. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Curtis JR, Yamaoka K, Chen Yi‐H, Bhatt DL, Gunay LM, Sugiyama N, et al. Malignancy risk with tofacitinib versus TNF inhibitors in rheumatoid arthritis: results from the open‐label, randomised controlled ORAL Surveillance trial. Ann Rheum Dis. 2023;82:331–343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Bouabdallaoui N, Tardif J‐C, Waters DD, Pinto FJ, Maggioni AP, Diaz R, et al. Time‐to‐treatment initiation of colchicine and cardiovascular outcomes after myocardial infarction in the Colchicine Cardiovascular Outcomes Trial (COLCOT). Eur Heart J. 2020;41:4092–4099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Li J, Meng X, Shi F‐D, Jing J, Gu H‐Q, Jin A, et al. Colchicine in patients with acute ischaemic stroke or transient ischaemic attack (CHANCE‐3): multicentre, double blind, randomised, placebo controlled trial. BMJ. 2024;385:e079061. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Jolly SS, d'Entremont M‐A, Lee SF, Mian R, Tyrwhitt J, Kedev S, et al. Colchicine in acute myocardial infarction. N Engl J Med. 2025;392:633–642. [DOI] [PubMed] [Google Scholar]
- 25. Solomon DH, Liu CC, Kuo IH, Zak A, Kim SC. Effects of colchicine on risk of cardiovascular events and mortality among patients with gout: a cohort study using electronic medical records linked with Medicare claims. Ann Rheum Dis. 2016;75:1674–1679. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Roddy E, Bajpai R, Forrester H, Partington RJ, Mallen CD, Clarson LE, et al. Safety of colchicine and NSAID prophylaxis when initiating urate‐lowering therapy for gout: propensity score‐matched cohort studies in the UK Clinical Practice Research Datalink. Ann Rheum Dis. 2023;82:1618–1625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Yeh JJ, Kuo IL, Yip HT, Hsueh MY, Hsu CY, Kao CH. Effects of colchicine use on ischemic and hemorrhagic stroke risk in diabetic patients with and without gout. Sci Rep. 2022;12:9195. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Shah B, Toprover M, Crittenden DB, Jeurling S, Pike VC, Krasnokutsky S, et al. Colchicine use and incident coronary artery disease in male patients with gout. Can J Cardiol. 2020;36:1722–1728. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Deftereos SG, Beerkens FJ, Shah B, Giannopoulos G, Vrachatis DA, Giotaki SG, et al. Colchicine in cardiovascular disease: in‐depth review. Circulation. 2022;145:61–78. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Cipolletta E, Tata LJ, Nakafero G, Avery AJ, Mamas MA, Abhishek A. Association between gout flare and subsequent cardiovascular events among patients with gout. JAMA. 2022;328:440–450. [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
Table S1: Codes used to define the covariates.
Table S2: Generic names of drugs used to define the study cohort.
Table S3: Sensitivity analyses: Crude and propensity score adjusted hazard ratio for association between colchicine versus non‐steroidal anti‐inflammatory drugs and risk of cardiovascular events.
Figure S1: Graphical depiction of the study design.
Data Availability Statement
Data generated and/or analyzed during the current study cannot be shared publicly due to the data‐sharing policy of the National Health Insurance Service (NHIS) of Korea, governed by Article 18 of the Personal Information Protection Act (“Limitation to Out‐of‐Purpose Use and Provision of Personal Information” available at https://elaw.klri.re.kr/kor_service/lawView.do?hseq=53044&lang=ENG). However, the data are available from the NHIS (study identifier: NHIS‐2023‐1‐410) on reasonable request for researchers who meet the criteria for access to confidential data (https://www.data.go.kr/en/tcs/eds/selectCoreDataView.do?coreDataInsttCode=B551182&coreDataSn=1&searchCondition2=coreDataNmEn&searchKeyword2=).