Real‐world usage of sulphonylureas in Asian patients with type 2 diabetes using the Joint Asia Diabetes Evaluation (JADE) register

Abstract

Aims

To explore the patterns of use of oral glucose‐lowering drugs (OGLDs) in Asian patients with type 2 diabetes (T2D), focusing on sulphonylureas (SUs), and to describe patient profiles according to treatment regimen.

Methods

We conducted a cross‐sectional analysis of data from adults with T2D from 11 Asian countries/regions with structured assessment enrolled in the prospective Joint Asia Diabetes Evaluation (JADE) register between November 2007 and December 2019. Patients receiving insulin and/or injectable glucagon‐like peptide‐1 receptor agonists were excluded.

Results

Amongst 62 512 patients (mean ± standard deviation age: 57.3 ± 11.8 years; 53.6% men), 54 783 (87.6%) were treated with OGLDs at enrolment. Most received one (37.5%) or two (44.2%) OGLDs. In the entire cohort, 59.4% of treated patients received SU‐based therapy with variations amongst countries/regions. Overall, 79.5% of SU regimens were based on SUs plus metformin, and 22.1% on SUs plus dipeptidyl peptidase‐4 inhibitors. Among SU users, gliclazide was most commonly prescribed (46.7%), followed by glimepiride (40.0%) and glibenclamide (8.1%). More gliclazide users entered the cohort with glycated haemoglobin levels <53 mmol/mol (7%) than non‐gliclazide SU users (odds ratio [OR] 1.09, 95% CI 1.02‐1.17), with less frequent self‐reported hypoglycaemia in the 3 months before registration (OR 0.81, 95% CI 0.72‐0.92; adjusted for sociodemographic factors, cardiometabolic risk factors, complications, use of other OGLDs, country/region and year of registration).

Conclusion

In Asia, SUs are a popular OGLD class, often combined with metformin. Good glycaemic control and safety profiles associated with the use of SUs, including gliclazide, support their position as a key treatment option in patients with T2D.

1. INTRODUCTION

In 2021, over half a billion people were living with diabetes worldwide, with 90% having type 2 diabetes (T2D). ¹ In the Asia‐Pacific region, adults aged 20 to 79 years accounted for 50% of the global population with diabetes. ¹ Optimal glycaemic control is essential for preventing microvascular and macrovascular complications in diabetes. ² Glucose‐lowering strategies should aim to minimize the risk of hypoglycaemia which may compromise quality of life, treatment adherence and patient satisfaction. ³ Practice guidelines recommended a patient‐centred approach to individualize selection of glucose‐lowering drugs (GLDs) in people with T2D. ⁴ , ⁵ , ⁶ Metformin is often the first‐line therapy in T2D and forms the base therapy in many clinical trials that evaluated new GLDs. ⁷ Alternative GLDs, including sodium‐glucose cotransporter 2 (SGLT2) inhibitors, glucagon‐like peptide‐1 receptor agonists (GLP‐1RAs), dipeptidyl peptidase‐4 (DPP‐4) inhibitors, sulphonylureas (SUs) and thiazolidinediones, may be considered if metformin is not tolerated or if glycaemic control is not achieved, taking into consideration the individual's risk of cardiorenal disease or hypoglycaemia. ⁴ , ⁵ , ⁶ Compared with White populations, Asian patients with T2D had reduced insulin‐secreting capacity to overcome insulin resistance. ⁸ Thus, improvement of insulin secretion via insulin secretagogues such as SUs is a popular treatment strategy amongst Asian patients with T2D. ⁹

Given that 80% of patients with diabetes live in low‐ and middle‐income countries, ¹ metformin and SUs included in the World Health Organization Essential Medicine List will remain the mainstay of treatment. ¹⁰ SUs are a low‐cost, safe and efficacious class of oral GLD (OGLD) ¹¹ albeit with increased risk of hypoglycaemia, especially in patients with chronic kidney disease (CKD). ¹² In the Cardiovascular Outcome Study of Linagliptin vs Glimepiride in Type 2 Diabetes (CAROLINA) trial, glimepiride, an SU, was associated with a higher rate of symptomatic hypoglycaemia than linagliptin, a DPP‐4 inhibitor (8.4 vs. 1.4 events per 100 patient‐years), although the cardiovascular event rates were similar in the two groups. ¹³ In randomized clinical trials (RCTs), gliclazide had greater efficacy and safety than other SUs, especially in high‐risk patients, including the elderly. ¹⁴ , ¹⁵ , ¹⁶ In the head‐to‐head GUIDE study (GlUcose control in type 2 diabetes: Diamicron MR vs glimEpiride), hypoglycaemia episodes were 50% lower in the gliclazide modified‐release (MR) group than in the glimepiride group. ¹⁷ Meta‐analyses of RCTs also reported a lower risk of hypoglycaemia with gliclazide compared with other SUs. ¹¹ , ¹⁸ , ¹⁹ Of note, gliclazide had been used safely during prolonged fasting such as during Ramadan. ²⁰ Gliclazide is primarily metabolized by cytochrome P450 2C19 (CYP2C19) rather than CYP2C9, which is the main cytochrome 450 enzyme for the metabolism of other SUs. ²¹ These findings suggest that gliclazide may have different patterns of efficacy and safety versus other SUs.

In most surveys, 70% of patients with T2D did not have cardiorenal disease ²² and thus, optimal glycaemic control remains an important strategy to reduce glycaemic deterioration for organ protection. ²³ While RCTs can inform on the clinical efficacy and safety of a drug, the generalizability of the results can be limited by strict eligibility criteria and confounding due to care setting. Thus, real‐world data collected in routine practice can complement RCTs to provide additional insights regarding the acceptability and effectiveness of a drug in optimizing care. ²⁴ There are several real‐world databases reporting treatment patterns in Asian people with T2D, although few of them had detailed documentation of patient profiles essential for evaluating treatment effectiveness and safety. ²⁵ , ²⁶ , ²⁷ Using data from the Joint Asia Diabetes Evaluation (JADE) register, a data‐driven quality improvement programme with structured assessment, ²⁸ we evaluated treatment responses and tolerability/safety issues associated with SU‐based therapy in a real‐world setting.

2. METHODS

2.1. JADE register/study design

We performed a retrospective cross‐sectional analysis of data from the prospective JADE register, established in 2007 and aimed at promoting quality improvement and collaborative research in Asia. ²⁸ The well‐defined JADE cohort serves as a rich data resource to identify unmet needs and provide insights into real‐world practice. ²⁸ , ²⁹ The web‐based JADE technology consists of a portal with built‐in protocols to guide data collection during structured assessment performed at baseline and every 12 to 24 months. ²⁸ Over 300 physicians, mainly internists or endocrinologists, from a wide range of care settings in 11 Asian countries/regions, enrolled patients with diabetes into the JADE register. During a clinic visit and supervised by attending physicians, trained nurses performed structured assessment including documentation of sociodemographic data, lifestyle and self‐management, history of comorbidities, current medications, self‐reported hypoglycaemia events, clinical measurements and examination of eye and feet for diabetes‐related complications. ²⁶ Data on treatment duration or adherence were not collected. Additionally, blood and urine samples were collected after 8 to 10 hours of fasting. ²⁶

All items were filled by trained nurses and attending physicians using a standardized case record form and through direct patient enquiry. Upon entry to the JADE portal by trained nurses, data were automatically deidentified with creation of a unique random number for identification purposes. The case record form includes itemized responses to improve data validity while the JADE portal contains value ranges to prompt outliers during data entry. Due to funding limitations and given that this was not an RCT, independent data validation was not performed. Of note, the utility of the JADE portal had been demonstrated in several RCTs, including the Asia‐Pacific JADE, ³⁰ Diabetic Kidney Disease – Treating to Multiple Targets (DKD‐TMT) ³¹ and Peer Support, Empowerment and Remote Communication Linked by Information Technology (PEARL) ³² projects. All participants gave written informed consent for analysis and reporting of the deidentified data.

The study was approved by the Joint Chinese University of Hong Kong‐New Territories East Cluster Clinical Research Ethics Committee and the local institutional boards.

2.2. Study population

We included data from Asian patients with T2D aged ≥18 years, enrolled in the JADE register between January 2007 and December 2019 from 11 countries/regions (China, Hong Kong, India, Indonesia, Malaysia, the Philippines, Singapore, South Korea, Taiwan, Thailand and Vietnam). All patients had physician‐diagnosed T2D based on the American Diabetes Association (ADA) criteria or local practice. ³³ Patients with type 1 diabetes were excluded from this analysis. ²⁶ Most of the study sites did not have electronic medical records or routinely measure insulin autoantibodies or C‐peptide assays. Thus, we defined type 1 diabetes based on presentation with diabetic ketoacidosis, unprovoked ketosis, or continuous insulin requirement within 12 months of diagnosis. ²⁶ Since this study focused on the treatment patterns of OGLDs, patients with T2D treated with injectable insulin and/or GLP‐1RAs were also excluded.

2.3. Study objectives

The primary objective was to describe patterns of OGLD use in patients with T2D in Asia, including SU‐ and non‐SU‐based therapy, in the whole cohort as well as by country/region. We compared the clinical characteristics including the control of cardiometabolic risk factors and comorbidities (cardiovascular disease [CVD], CKD, diabetic retinopathy and peripheral neuropathy) at registration between (a) the OGLD and the non‐OGLD group, (b) the SU and the non‐SU group in the OGLD group, and (c) the gliclazide and the non‐gliclazide group in SU‐treated patients. CVD was defined by self‐reported history of ischaemic heart disease (myocardial infarction or unstable angina requiring hospitalization or chest pain confirmed by cardiac investigations including stress test, imaging or intervention), stroke (hospitalization with permanent neurological damage or brain imaging) or peripheral vascular disease (history of revascularization or diminished/absent foot pulse and/or ankle‐brachial index <0.9 using Doppler scan, as recorded). CKD was defined as an estimated glomerular filtration rate (eGFR) <60 mL/min/1.73 m² calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD‐EPI) equation at registration. ²⁶

Secondary objectives were to describe, amongst SU users: (a) treatment patterns of different SUs in monotherapy or combination therapy in the whole cohort and by country/region, (b) the proportions of patients (gliclazide vs. non‐gliclazide SU) with at least one episode of self‐reported hypoglycaemia in the 3 months before registration and (c) the proportions of patients (gliclazide vs. non‐gliclazide SU) with glycated haemoglobin (HbA1c) levels <53 mmol/mol (7%) at registration.

During comprehensive assessment, self‐reported hypoglycaemia in the past 3 months was assessed by the trained nurses using a structured case record form. This was based on patients' experience of typical symptoms (eg, hunger, dizziness, perioral numbness, tremor, especially if corrected by carbohydrate intake), frequency of hypoglycaemia (at least daily, at least once weekly, at least once monthly, or less than once monthly), intensity of hypoglycaemia episode (mild, moderate, or severe) with or without confirmed low values on self‐monitoring of blood glucose (SMBG). Severe hypoglycaemia episodes were also self‐reported, and defined as any hypoglycaemia episode that required third‐party assistance, hospitalization or medical assistance.

2.4. Statistical analysis

Data are presented as mean ± standard deviation (SD) or median (interquartile range [IQR]) for continuous variables with normal or skewed distribution, respectively. Categorical variables are expressed as number and percentages. For between‐group comparisons of continuous variables, an independent Student's test was used for variables with normal distribution and Wilcoxon's rank‐sum test for variables with skewed distribution. For between‐group comparisons of categorical variables, the chi‐squared test was used. Binary logistic regression analysis was performed to estimate the probability of having HbA1c <53 mmol/mol (7%) at registration and self‐reported hypoglycaemia (at least monthly, including mild, moderate and severe episodes) comparing gliclazide and non‐gliclazide SU groups, adjusted for age, gender, duration of T2D, body mass index (BMI), smoking status, eGFR, SMBG (yes/no), prior CVD (yes/no), use of metformin (yes/no), DPP‐4 inhibitors (yes/no), alpha‐glucosidase inhibitors (AGIs; yes/no), SGLT2 inhibitors (yes/no), meglitinide (yes/no) and thiazolidinediones (yes/no), country/region and year of registration. When hypoglycaemia was the dependent variable, the logistic regression models were additionally adjusted for HbA1c at baseline. Probability was expressed as odds ratio (OR) with 95% confidence interval (CI).

All statistical tests were conducted using R (www.r-project.org). Two‐sided P values <0.05 were taken to indicate statistical significance.

3. RESULTS

3.1. Patient distribution

Between January 2007 and December 2019, 114 614 patients were recruited into the JADE register (Figure 1). Of 104 478 patients with T2D, patients treated with insulin (n = 28 239) or GLP‐1RAs (n = 277) were excluded. There were 13 645 patients (13%) with missing information on OGLDs. A total of 62 512 patients who received either OGLDs at baseline (n = 54 783) or no treatment (n = 7729) from 11 countries/regions were included in the final analysis.

FIGURE 1.

Selection of patients from the Joint Asia Diabetes Evaluation (JADE) register for analysis of usage of oral glucose‐lowering drugs (OGLDs) in Asian patients with type 2 diabetes (2007‐2019). Abbreviations: GLP‐1, glucagon‐like peptide 1; GLP‐1RA, glucagon‐like peptide 1 receptor agonist

3.2. Patients treated with OGLDs

Table 1 summarizes the clinical characteristics of patients with T2D treated with OGLDs at baseline or with lifestyle modification (LSM) only. The OGLD users were older than the LSM group (mean ± SD age: 57.5 ± 11.6 vs. 56.2 ± 12.6 years; P < 0.001) and had longer T2D duration (5 [IQR 2‐10] vs 2 [IQR 0‐6] years), higher BMI and systolic blood pressure and lower eGFR (P < 0.001, respectively). Both groups had similar HbA1c levels at registration (60 mmol/mol [7.6%], respectively). Among the patients treated with LSM only, 53.9% had HbA1c <53 mmol/mol (7%) and 28.7% had HbA1c ≥64 mmol/mol (8%). Patients receiving OGLDs were more likely to have hypertension, dyslipidaemia and complications (including sensory neuropathy, retinopathy, CKD, CVD) than the LSM group (13.9% vs. 11.5%; P < 0.001). They were also more likely to report hypoglycaemia events (6.1% vs. 3.9%; P < 0.001). Amongst patients receiving OGLDs, the mean number of OGLDs was 1.8. Metformin was used by 84.1% of patients, whilst 59.4% received an SU, and 23.0% a DPP‐4 inhibitor, alone or in combination with another OGLD at registration. Table S1 compares the clinical characteristics of patients with T2D, stratified by data availability on OGLDs. Compared with those included in the present analysis, patients with missing data were younger, had shorter disease duration and were less likely to have cardiorenal diseases or treatment with renin‐angiotensin system inhibitors and statin. These patients had worse glycaemic control, with a higher proportion having self‐reported hypoglycaemia and SMBG.

TABLE 1.

Clinical characteristics of patients with type 2 diabetes at time of enrolment in the JADE register, stratified by use of oral glucose‐lowering drugs

	Treated with LSM and OGLDs (n = 54 783)		Treated with LSM only (n = 7729)
	n		n		P value
Age at registration, years	54 783	57.5 ± 11.6	7729	56.2 ± 12.6	<0.001
>65 years, n (%)	54 783	13 994 (25.5)	7729	1845 (23.9)	0.002
Men, n (%)	54 779	29 473 (53.8)	7728	4036 (52.2)	0.010
Smoking status, n (%)	54 140		7241		0.269
Non‐smoker		39 526 (73.0)		5334 (73.7)
Ex‐smoker		8048 (14.9)		1076 (14.9)
Current smoker		6566 (12.1)		831 (11.5)
Duration of T2D, years	54 269	5 (2‐10)	6951	2 (0‐6)	<0.001
≤1 year, n (%)	54 269	11 902 (21.9)	6951	3435 (49.4)	<0.001
Young‐onset diabetes, a n (%)	54 273	8603 (15.9)	6974	1100 (15.8)	0.880
BMI, kg/m2	52 690	26.2 ± 4.5	6274	25.9 ± 4.7	<0.001
≥25 kg/m2, n (%)	52 690	29 921 (56.8)	6274	3331 (53.1)	<0.001
HbA1c, mmol/mol (%)	48 292	60 ± 17.5 (7.6 ± 1.6)	6219	60 ± 24.0 (7.6 ± 2.2)	0.366
<48 mmol/mol (6.5%), n (%)		12 795 (26.5)		2478 (39.8)	<0.001
<53 mmol/mol (7.0%), n (%)		20 024 (41.5)		3353 (53.9)	<0.001
≥64 mmol/mol (8.0%), n (%)		14 706 (30.5)		1787 (28.7)	0.006
Systolic blood pressure, mmHg	53 933	131.0 ± 17.0	7292	129.0 ± 16.9	<0.001
Diastolic blood pressure, mmHg	53 825	78.8 ± 9.7	7278	78.5 ± 10.0	0.008
Total cholesterol, mmol/L	45 412	4.6 ± 1.1	5942	5.1 ± 1.5	<0.001
LDL cholesterol, mmol/L	46 731	2.6 ± 0.9	5889	3.0 ± 1.1	<0.001
Triglycerides, mmol/L	47 792	1.5 (1.1‐2.1)	6179	1.5 (1.1‐2.3)	0.002
HDL cholesterol, mmol/L	47 015	1.2 ± 0.4	5997	1.5 ± 1.0	<0.001
eGFR, mL/min/1.73m2	45 742	81.9 ± 22.1	6005	84.3 ± 24.8	<0.001
eGFR <60 mL/min/1.73m2, n (%)	45 742	7529 (16.5)	6005	882 (14.7)	<0.001
Urinary ACR, mg/mmoL	37 458	1.7 (0.6‐5.7)	4622	1.5 (0.6‐7.4)	0.910
SMBG, n (%)	49 949	32 832 (65.7)	5940	3227 (54.3)	<0.001
Self‐reported hypoglycaemia at least monthly, n (%)	49 328	2987 (6.1)	6543	255 (3.9)	<0.001
Severe hypoglycaemia	7058	85 (1.2)	752	12 (1.6)	0.454
Class of OGLD, n (%)	54 783		7729		<0.001
Metformin		46 080 (84.1)
Sulphonylureas		32 558 (59.4)
DPP‐4 inhibitors		12 591 (23.0)
Thiazolidinediones		4776 (8.7)
Alpha glucosidase inhibitors		3509 (6.4)
Glinides		748 (1.4)
SGLT2 inhibitors		284 (0.5)
Number of OGLDs	54 783	1.8 ± 0.8	7729		<0.001
1, n (%)		20 553 (37.5)
2, n (%)		24 224 (44.2)
3, n (%)		8573 (15.6)
>3, n (%)		1433 (2.6)
RAS inhibitors, n (%)	54 783	22 434 (41.0)	7729	1078 (13.9)	<0.001
Statin, n (%)	49 567	24 636 (49.7)	7204	1252 (17.4)	<0.001
Hypertension, n (%)	54 395	46 587 (85.6)	7371	5715 (77.5)	<0.001
Dyslipidaemia, n (%)	49 894	41 447 (83.1)	6026	4702 (78.0)	<0.001
Peripheral sensory neuropathy, n (%)	54 446	5263 (9.7)	7198	267 (3.7)	<0.001
Diabetic retinopathy, n (%)	50 043	5890 (11.8)	6878	458 (6.7)	<0.001
Chronic kidney disease, n (%)	54 783	7592 (13.9)	7729	887 (11.5)	<0.001
Coronary heart disease, n (%)	54 783	4845 (8.8)	7729	463 (6.0)	<0.001
Stroke, n (%)	54 783	2125 (3.9)	7729	237 (3.1)	0.001
Peripheral arterial disease, n (%)	54 783	2118 (3.9)	7729	270 (3.5)	0.117

Note: Results are presented as mean ± standard deviation, median (interquartile range) or number (percentage), as appropriate.

Abbreviations: ACR, albumin:creatinine ratio; BMI, body mass index; DPP‐4, dipeptidyl peptidase‐4; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; LSM, lifestyle modification; OGLD, oral glucose‐lowering drug; RAS, renin‐angiotensin system; SMBG, self‐monitoring of blood glucose; SGLT2, sodium‐glucose cotransporter‐2; T2D, type 2 diabetes.

^a Young‐onset diabetes diagnosed before the age of 40 years. Self‐reported hypoglycaemia was reported in the last 3 months before enrolment in the JADE register.

Figure 2 shows the distribution of OGLD usage in different Asian countries/regions. Figure 2A illustrates the proportion of SU‐based versus non‐SU‐based therapy. In the entire cohort, 59.4% of treated patients received an SU‐based therapy. In China, South Korea and the Philippines, 32.6% to 47.4% of OGLD‐treated patients received SUs, whereas in Vietnam, 72.3% of T2D patients received SUs. Figure 2B shows the distribution of OGLD regimens. The most common therapies were metformin only (25.2%), metformin plus an SU (30.6%) or metformin plus an SU plus a DPP‐4 inhibitor (9.2%), with variation among countries/regions. For example, in India and Indonesia, metformin monotherapy was used less often than in other countries. In China, AGIs only or AGIs plus metformin were more commonly prescribed than in other Asian countries.

FIGURE 2.

Patterns of usage of oral glucose‐lowering drugs (OGLDs) in patients with type 2 diabetes (T2D) at enrolment in the Joint Asia Diabetes Evaluation (JADE) register. (A) Sulphonylureas (SUs) versus non‐SUs (n = 54 783), (B) type of OGLD (n = 54 783), (C) type of regimen in patients treated with SUs (n = 32 558), (D) type of SU used (n = 32 558). Abbreviations: AGI, alpha‐glucosidase inhibitor; DPP‐4, dipeptidyl peptidase‐4; TZD, thiazolidinedione

3.3. Patients treated with SUs

Table 2 summarizes the clinical characteristics of patients treated with an SU (n = 32 558) or other OGLDs (n = 22 225) plus LSM at baseline. The SU group was older (58.0 ± 11.5 vs. 56.8 ± 11.8 years; P < 0.001), with longer disease duration (6 [IQR 3‐12] vs. 3 [IQR 1‐7] years; P < 0.001) than non‐SU users. They had higher HbA1c level and systolic blood pressure and lower eGFR (P < 0.001, respectively). Users of SUs had lower BMI than patients receiving other OGLDs (P < 0.001) and were more likely to have hypertension (P < 0.001), complications (eg, CKD 15.9% vs. 10.8%; P < 0.001) and self‐reported hypoglycaemia (P < 0.001), with similar rates of severe hypoglycaemia between the two groups.

TABLE 2.

Clinical characteristics of patients with type 2 diabetes stratified by (a) use and non‐use of sulphonylureas (SUs) amongst patients treated with oral glucose‐lowering drugs and lifestyle management and (b) use and non‐use of gliclazide amongst patients treated with SUs at enrolment in the JADE register

	Treated with OGLDs					Treated with SUs
	Treated with SU (N = 32 558)		Not treated with SU (N = 22 225)		P value	Treated with gliclazide (N = 15 312)		Not treated with gliclazide (N = 17 246)
	n		n			n		n		P value
Age at registration, years	32 558	58.0 ± 11.5	22 225	56.8 ± 11.8	<0.001	15 312	59.7 ± 11.5	17 246	56.5 ± 11.3	<0.001
Age >65 years, n (%)	32 558	8675 (26.6)	22 225	5319 (23.9)	<0.001	15 312	4834 (31.6)	17 246	3841 (22.3)	<0.001
Men, n (%)	32 556	17 349 (53.3)	22 223	12 124 (54.6)	0.004	15 311	7900 (51.6)	17 245	9449 (54.8)	<0.001
Smoking status, n (%)	32 207		21 933		0.025	15 182		17 025		<0.001
Non‐smoker		23 567 (73.2)		15 959 (72.8)			10 787 (71.1)		12 780 (75.1)
Ex‐smoker		4832 (15.0)		3216 (14.7)			2639 (17.4)		2193 (12.9)
Current smoker		3808 (11.8)		2758 (12.6)			1756 (11.6)		2052 (12.1)
Duration of T2D, years	32 271	6 (3‐12)	21 998	3 (1‐7)	<0.001	15 133	7 (3‐12)	17 138	6 (3‐11)	0.002
≤1 year, n (%)	32 271	4965 (15.4)	21 998	6937 (31.5)	<0.001	15 133	2421 (16.0)	17 138	2544 (14.8)	0.004
Young‐onset diabetes, a n (%)	32 272	5422 (16.8)	22 001	3181 (14.5)	<0.001	15 133	2121 (14.0)	17 139	3301 (19.3)	<0.001
BMI, kg/m2	31 485	26.1 ± 4.4	21 205	26.3 ± 4.6	<0.001	14 841	25.8 ± 4.3	16 644	26.4 ± 4.4	<0.001
BMI ≥25 kg/m2, n (%)	31 485	17 668 (56.1)	21 205	12 253 (57.8)	<0.001	14 841	7728 (52.1)	16 644	9940 (59.7)	<0.001
HbA1c, mmol/mol (%)	28 596	62 ± 18.6 (7.8 ± 1.7)	19 696	56 ± 16.4 (7.3 ± 1.5)	<0.001	13 915	60 ± 17.5 (7.6 ± 1.6)	14 681	64 ± 18.6 (8.0 ± 1.7)	<0.001
<48 mmol/mol (6.5%), n (%)		5830 (20.4)		6965 (35.4)	<0.001		3360 (24.1)		2470 (16.8)	<0.001
<53 mmol/mol (7.0%), n (%)		9583 (33.5)		10 441 (53.0)	<0.001		5376 (38.6)		4207 (28.7)	<0.001
≥64 mmol/mol (8.0%), n (%)		10 443 (36.5)		4263 (21.6)	<0.001		4311 (31.0)		6132 (41.8)	<0.001
Systolic blood pressure, mmHg	32 131	132.0 ± 17.2	21 802	130.0 ± 16.7	<0.001	15 124	133.0 ± 18.0	17 007	132.0 ± 16.5	<0.001
Diastolic blood pressure, mmHg	32 067	78.9 ± 9.6	21 758	78.8 ± 9.7	0.737	15 113	78.0 ± 10.0	16 954	79.6 ± 9.1	<0.001
Total cholesterol, mmol/L	27 053	4.6 ± 1.1	18 359	4.6 ± 1.1	0.124	13 251	4.6 ± 1.0	13 802	4.7 ± 1.1	<0.001
LDL cholesterol, mmol/L	27 883	2.6 ± 0.9	18 848	2.6 ± 0.9	<0.001	13 579	2.6 ± 0.9	14 304	2.7 ± 0.9	<0.001
Triglycerides, mmol/L	28 447	1.6 (1.1‐2.2)	19 345	1.5 (1.0‐2.1)	<0.001	13 857	1.5 (1.0‐2.1)	14 590	1.7 (1.2‐2.2)	<0.001
HDL cholesterol, mmol/L	28 026	1.2 ± 0.4	18 989	1.2 ± 0.4	<0.001	13 624	1.2 ± 0.4	14 402	1.2 ± 0.4	<0.001
eGFR, mL/min/1.73m2	27 337	80.2 ± 22.5	18 405	84.5 ± 21.2	<0.001	13 554	79.7 ± 22.1	13 783	80.6 ± 22.8	<0.001
eGFR <60 mL/min/1.73m2, n (%)	27 337	5155 (18.8)	18 405	2374 (12.9)	<0.001	13 554	2684 (19.8)	13 783	2471 (18.0)	<0.001
Urinary ACR, mg/mmol	22 296	2 (0.7‐7.0)	15 162	1.4 (0.6‐4.2)	<0.001	11 585	2 (0.7‐7.5)	10 711	2.0 (0.7‐6.3)	0.023
SMBG, n (%)	29 735	19 749 (66.4)	20 214	13 083 (64.7)	<0.001	14 070	9402 (66.8)	15 665	10 347 (66.1)	0.163
Self‐reported hypoglycaemia at least monthly, n (%)	28 972	2117 (7.3)	20 356	870 (4.3)	<0.001	14 369	1062 (7.4)	14 603	1055 (7.2)	0.602
Severe hypoglycaemia	4987	56 (1.1)	2071	29 (1.4)	0.394	2572	31 (1.2)	2415	25 (1.0)	0.663
Class of OGLD, n (%)	32 558		22 225		<0.001	15 312		17 246		<0.001
Metformin		25 888 (79.5)		20 192 (90.9)			11 981 (78.2)		13 907 (80.6)
DPP‐4 inhibitors		7189 (22.1)		5402 (24.3)			2631 (17.2)		4558 (26.4)
Thiazolidinediones		2884 (8.9)		1892 (8.5)			733 (4.8)		2151 (12.5)
Alpha glucosidase inhibitors		2217 (6.8)		1292 (5.8)			684 (4.5)		1533 (8.9)
Glinides		72 (0.2)		676 (3.0)			25 (0.2)		47 (0.3)
SGLT2 inhibitors		153 (0.5)		131 (0.6)			104 (0.7)		49 (0.3)
Number of OGLDs	32 558	2.2 ± 0.7	7729	1.3 ± 0.5	<0.001	15 312	2.1 ± 0.7	17 246	2.3 ± 0.8	<0.001
1, n (%)		4939 (15.2)		15 614 (70.3)			2623 (17.1)		2316 (13.4)
2, n (%)		18 333 (56.3)		5891 (26.5)			9539 (62.3)		8794 (51.0)
3, n (%)		7881 (24.2)		692 (3.1)			2858 (18.7)		5023 (29.1)
>3, n (%)		1405 (4.3)		28 (0.1)			292 (1.9)		1113 (6.5)
RAS inhibitors, n (%)		14 494 (44.5)		7940 (35.7)		15 312	7043 (46.0)	17 246	7451 (43.2)	<0.001
Statin, n (%)		14 933 (50.4)		9703 (48.6)		14 146	6949 (49.1)	15 471	7984 (51.6)	<0.001
Hypertension, n (%)	32 370	28 147 (87.0)	22 025	18 440 (83.7)	<0.001	15 233	13 042 (85.6)	17 137	15 105 (88.1)	<0.001
Dyslipidaemia, n (%)	29 790	24 647 (82.7)	20 104	16 800 (83.6)	0.016	14 337	11 731 (81.8)	15 453	12 916 (83.6)	<0.001
Peripheral sensory neuropathy, n (%)	32 355	3555 (11.0)	22 091	1708 (7.7)	<0.001	15 247	1214 (8.0)	17 108	2341 (13.7)	<0.001
Diabetic retinopathy, n (%)	29 436	3952 (13.4)	20 607	1938 (9.4)	<0.001	13 971	2385 (17.1)	15 465	1567 (10.1)	<0.001
Chronic kidney disease, n (%)	32 558	5189 (15.9)	22 225	2403 (10.8)	<0.001	15 312	2707 (17.7)	17 246	2482 (14.4)	<0.001
Coronary heart disease, n (%)	32 558	2988 (9.2)	22 225	1857 (8.4)	0.001	15 312	1575 (10.3)	17 246	1413 (8.2)	<0.001
Stroke, n (%)	32 558	1278 (3.9)	22 225	847 (3.8)	0.511	15 312	880 (5.8)	17 246	398 (2.3)	<0.001
Peripheral arterial disease, n (%)	32 558	1341 (4.1)	22 225	777 (3.5)	<0.001	15 312	506 (3.3)	17 246	835 (4.8)	<0.001

Note : Results are presented as mean ± standard deviation, median (interquartile range) or number (percentage), as appropriate.

Abbreviations: ACR, albumin:creatinine ratio; DPP4, dipeptidyl peptidase 4; eGFR, estimated glomerular filtration rate; HDL, high‐density lipoprotein; LDL, low‐density lipoprotein; LSM, lifestyle modification; OGLD, oral glucose‐lowering drug; RAS, renin‐angiotensin system; SGLT2i, sodium‐glucose co‐transporter 2 inhibitor; SMBG, self‐monitoring of blood glucose; SU, sulphonylurea; T2D, type 2 diabetes.

^a

Young‐onset diabetes diagnosed before the age of 40 years. Self‐reported hypoglycaemia was reported in the last 3 months before enrolment in the JADE register.

3.4. Patients treated with gliclazide

Table 2 summarizes the clinical characteristics of patients treated with gliclazide (n = 15 312) or other SUs (n = 17 246) in the SU‐treated group. Gliclazide users were older (59.7 ± 11.5 vs. 56.5 ± 11.3 years; P < 0.001) and had longer disease duration (7 [IQR 3‐12] vs. 6 [IQR 3‐11] years; P < 0.001) than patients receiving other SUs (glimepiride, glibenclamide, glipizide or other). Gliclazide users had lower BMI (25.8 vs. 26.4 kg/m²; P < 0.001) and were more likely to have HbA1c <53 mmol/mol (7%; 38.6% vs. 28.7%; P < 0.001) than patients receiving other SUs.

Gliclazide users had lower eGFR (79.7 vs. 80.6 mL/min/1.73 m²; P < 0.001) with a higher proportion having CKD (17.7% vs. 14.4%; P < 0.001) and complications (sensory neuropathy, retinopathy, CVD; all P < 0.001). Gliclazide users were more likely to be treated with renin‐angiotensin system inhibitors, but less likely to have dyslipidaemia or hypertension or to be treated with statins (P < 0.001, respectively) than non‐gliclazide users. Overall, 78.2% of gliclazide‐containing regimens included metformin and 0.7% included SGLT2 inhibitors. Gliclazide users were less likely to use DPP‐4 inhibitors (17.2% vs. 26.4%), with a lower proportion receiving three or more OGLDs than other SU users (20.6% vs. 35.6%; P < 0.001).

3.5. Effectiveness and safety

Table 3 shows the OR of having HbA1c <53 mmol/mol (7%) and hypoglycaemia associated with gliclazide use (n = 15 312) versus non‐gliclazide use (n = 17 246) in the SU group. After adjustment for predefined covariables, gliclazide use was associated with a higher likelihood of achieving HbA1c <53 mmol/mol (7%) with an OR of 1.09 (95% CI 1.02‐1.17; P = 0.014) than use of other SUs. The rate of self‐reported hypoglycaemia was 19% lower (adjusted OR 0.81, 95% CI 0.72‐0.92; P = 0.001) in the gliclazide group than the non‐gliclazide SU group.

TABLE 3.

Binary logistic regression analysis for odds ratio of having glycated haemoglobin <53 mmol/mol (7%) at registration and self‐reported hypoglycaemia in the 3 months before registration (sulphonylurea‐treated patients)

	SU group (treated with gliclazide vs. not treated with gliclazide)
	HbA1c <53 mmol/mol (7%)		Hypoglycaemia
	OR (95% CI)	P value	OR (95% CI)	P value
Model 1	1.57 (1.49‐1.65)	<0.001	1.02 (0.94‐1.12)	0.586
Model 2	1.46 (1.39‐1.54)	<0.001	1.05 (0.96‐1.14)	0.326
Model 3	1.47 (1.39‐1.55)	<0.001	0.95 (0.86‐1.05)	0.345
Model 4	1.35 (1.27‐1.43)	<0.001	0.95 (0.86‐1.06)	0.382
Model 5	1.09 (1.02‐1.17)	0.014	0.81 (0.72‐0.92)	0.001

Note: Hypoglycaemia was self‐reported, at least once monthly and included mild, moderate and severe episodes. Model 1: unadjusted. Model 2: adjusted for age at registration, gender and duration of type 2 diabetes. Model 3: Model 2 plus body mass index, smoking status, estimated glomerular filtration rate, self‐monitoring of blood glucose (yes/no), prior history of cardiovascular disease (yes/no), HbA1c (when hypoglycaemia was the dependent variable). Model 4: Model 3 plus use of metformin (yes/no), dipeptidyl peptidase‐4 inhibitors (yes/no), alpha glucosidase inhibitors (yes/no), sodium‐glucose cotransporter‐2 inhibitors (yes/no), meglitinide (yes/no) and thiazolidinediones (yes/no). Model 5: Model 4 plus country/region and year of registration.

Abbreviations: CI, confidence interval; HbA1c, glycated haemoglobin; OR, odds ratio; SU, sulphonylurea.

Tables S2 to S5 show the rates of hypoglycaemia, stratified by use of SUs, gliclazide and DPP‐4 inhibitors. Amongst patients treated with OGLDs, one‐third were neither treated with SUs nor DPP‐4 inhibitors, whilst 13.1% were using SUs plus DPP‐4 inhibitors (Table S2). Compared with patients treated neither with SUs nor with DPP‐4 inhibitors, use of SUs, DPP‐4 inhibitors or SUs plus DPP‐4 inhibitors was associated with adjusted ORs of 2.37 (95% CI 2.09‐2.69), 1.87 (95% CI 1.56‐2.22) and 2.53 (95% CI 2.15‐2.98), respectively, for hypoglycaemia (Table S3). Amongst patients treated with SUs, 39% were treated neither with gliclazide nor with DPP‐4 inhibitors, whilst 8.1% were treated with gliclazide plus DPP‐4 inhibitors (Table S4). Compared with patients treated neither with gliclazide nor with DPP‐4 inhibitors, use of gliclazide was associated with reduced adjusted odds of hypoglycaemia (OR 0.81, 95% CI 0.71‐0.93), while gliclazide plus a DPP‐4 inhibitor was associated with neutral risk of hypoglycaemia (adjusted OR 0.92, 95% CI 0.75‐1.12; Table S5).

3.6. Patterns of SU use

Figure 2C illustrates the types of SU regimen used at baseline in different Asian countries/regions. SU monotherapy was used by 15.2% of patients, with SUs being part of combination therapy in most patients. Overall, 79.5% of SU regimens were based on an SU plus metformin, 22.1% on an SU plus a DPP‐4 inhibitor and 0.5% on an SU plus an SGLT2 inhibitor. Amongst all countries/regions, SU monotherapy was common in China (24.7%), the Philippines (23.6%) and Indonesia (21.2%). Amongst different types of SU (Figure 2D), gliclazide was the most popular (46.7%), followed by glimepiride (40.0%) and glibenclamide (8.1%). High usage of gliclazide was observed in Malaysia, Singapore and Hong Kong. Glimepiride was commonly used in India and Taiwan. In Thailand, glibenclamide and glipizide were used more frequently than in other Asian countries.

4. DISCUSSION

In this retrospective analysis based on a large prospective register including over 100 000 Asian patients with T2D, 30% were treated with injectables, with the majority receiving OGLDs (88%) or LSM only (12%). Amongst patients treated with OGLDs, approximately 60% were treated with SU, mainly in combination with metformin, with 1% to 2% prevalence of severe hypoglycaemia. Amongst the SU group, gliclazide use was associated with an increased likelihood of achieving HbA1c level <53 mmol/mol (7%) and lower occurrence of hypoglycaemia after adjusting for multiple covariables.

4.1. Patterns of use of OGLDs

In this report focusing on pattern of OGLDs, the most common therapies were metformin only, metformin plus an SU or metformin plus an SU plus a DPP‐4 inhibitor, with considerable inter‐country/region variations that were probably due to differences in reimbursement guidelines, prescribing habits and patient characteristics. In India and Indonesia, combination therapies including metformin were more often prescribed than metformin monotherapy. While this pattern might be influenced by prevailing HbA1c and patient profiles, international practice guidelines support early use of combination therapy in some patients to achieve glycaemic goals to increase glycaemic durability. ⁴ In China, the popularity of AGI‐based therapies might be driven by results of local clinical trials in both diabetes and prediabetes, with increased user experiences as well as national insurance coverage. ³⁴ , ³⁵

In this study, nearly half of the patients in Asia were treated with SUs, with one in four registrants from China, South Korea and the Philippines receiving SU monotherapy. In Vietnam, approximately 70% of patients with T2D received SUs with or without other OGLDs. The popularity of SUs in Asian patients may be related to the preponderance of beta‐cell dysfunction, closely associated with low BMI, compared with White populations ³⁶ , ³⁷ ; SUs have been shown to be particularly effective in patients with low BMI. ³⁸ Similarly, we also observed inter‐country/region differences in the types of SU used, probably due to variations in reimbursement policies, accessibility and affordability. Overall, 46.7% of SU users received gliclazide and 40.0% glimepiride. While India and Taiwan had high glimepiride use due to its low cost in these countries, glibenclamide and glipizide were used most frequently in Thailand, where they were listed in the Thailand National List of Essential Medicines and reimbursed by the national universal health coverage. ³⁹

4.2. Effectiveness of gliclazide and other SUs

After adjusting for confounders including country/region, gliclazide use was associated with higher odds of achieving HbA1c goals at study entry compared with non‐gliclazide. Ethnic differences in pharmacogenetics may be relevant. As discussed earlier, gliclazide is primarily metabolized by CYP2C19 rather than CYP2C9, which metabolizes other SUs. ²¹ A CYP2C19 poor metabolizer phenotype occurs in approximately 15% to 30% of Asians. ⁴⁰ In a retrospective pharmacogenetic analysis of Chinese patients with T2D, CYP2C19 poor metabolizers had greater glycaemic durability than fast metabolizers, probably due to reduced oral clearance of gliclazide. ⁴¹ These potential differences in gliclazide metabolism compared with other SUs, particularly in Asian populations, support our approach of separating assessment of gliclazide from other SUs.

In the large‐scale intensive glucose‐lowering Action in Diabetes and Vascular Disease: PreterAx and DiamicroN‐MR Controlled Evaluation (ADVANCE) trial, involving high‐risk patients with T2D, 30% of whom came from China, ⁴² the intensive gliclazide MR‐based strategy resulted in greater HbA1c reduction (48 mmol/mol [6.5%] vs. 56 mmol/mol [7.3%) than usual care. This difference in HbA1c was translated to a reduction of microvascular events, particularly nephropathy. ⁴²

Of note, treatment with gliclazide was associated with a better lipid profile and reduced likelihood of statin treatment than treatment with other SUs. Since hyperglycaemia might reduce the clearance of LDL cholesterol, the lower HbA1c in the gliclazide users might contribute to this favourable lipid pattern. ³⁰ Given the higher‐risk profiles of patients treated with gliclazide, there is a possibility that they might be managed by specialists or experienced physicians with better quality of care, which could confound the efficacy and safety of drug treatments. ⁴³ However, attributes of participating doctors were not captured.

4.3. Self‐reported hypoglycaemia with SUs and other OGLDs

Our data indicated that, while self‐reported hypoglycaemia was higher amongst SU users, severe hypoglycaemia was uncommon (1%‐2%), with similar rates between SU and non‐SU users. SUs, DPP‐4 inhibitors and glinides are insulin secretagogues. ⁴⁴ Amongst the non‐SU users, the latter two drugs were used more frequently, which might explain the similar rates of severe hypoglycaemia in the two groups. Researchers had reported that incretin and SUs, notably glibenclamide and glimepiride, both targeted the Epac2A‐Rap signalling pathway with increased insulin secretion. This effect was less evident with gliclazide. ⁴⁵ In support of this hypothesis, we found the highest rate of hypoglycaemia in patients treated with SUs plus DPP‐4 inhibitors (Table S3). However, amongst the SU users, gliclazide was associated with a lower rate of hypoglycaemia compared with other SUs. Interestingly, in the SU group, the frequency of hypoglycaemia in patients treated with gliclazide plus DPP‐4 inhibitor combination therapy was comparable with that in patients treated with neither drug (Table S5).

Several other studies have investigated the difference between gliclazide efficacy and that of other SUs. In the head‐to‐head GUIDE comparison trial, the gliclazide MR group had 50% fewer hypoglycaemia episodes than the glimepiride group. ¹⁷ Meta‐analyses of RCTs also reported lower risk of hypoglycaemia with gliclazide compared with other SUs, ¹¹ , ¹⁸ , ¹⁹ possibly due to its reversible binding to SU receptors with inactive metabolites. ¹¹ , ¹⁹ Our findings from the diverse setting of the JADE register therefore provide an opportunity to study differences between gliclazide and other SUs in a real‐world setting.

In a real‐world survey including patients with T2D observing a prolonged fast during Ramadan, gliclazide MR was associated with low risk of hypoglycaemia, with no episodes of severe hypoglycaemia. ²⁰ In our study, we did not have information distinguishing between gliclazide and the gliclazide MR formulation.

With availability of newer GLDs with low risk of hypoglycaemia, the position of SUs is being re‐evaluated. ⁴⁶ International guidelines are often based on evidence from RCTs with good internal validity, although real‐world data can provide valuable insights for external validity. ²⁴ Although over 50% of patients with diabetes come from the Asia‐Pacific region, ¹ Asians are generally under‐represented in RCTs. ⁴⁷ Consistent with the ADA/European Association for the Study of Diabetes (EASD) consensus statement, a patient‐centred approach is key in the treatment of patients with T2D. ⁴⁸ Given the large number of patients requiring GLDs to control glycaemia and prevent micro‐ and macrovascular complications, access and costs are important considerations. ⁴⁸ , ⁴⁹ The efficacy and safety of gliclazide in a large‐scale outcome trial, ⁴² supported by observational real‐world data including the current survey in a large Asian population, support the use of gliclazide as an important OGLD, especially in resource‐restricted settings. ⁴⁹

4.4. Importance of early glycaemic intervention

In our study, approximately 10% of patients were treated with LSM only, with 28.7% of these having HbA1c ≥64 mmol/mol (8%). Although these patients generally had fewer comorbidities than those treated with OGLDs, 11.5% already had CKD. Due to natural disease progression, ³⁰ , ⁵⁰ international practice guidelines recommended initiation of metformin along with LSM upon diagnosis. ⁴ In observational surveys, therapeutic inertia or insufficient glycaemic control during the early stage of diabetes was associated with increased risk of long‐term complications. These findings are supported by RCTs, where early combination therapy improved glycaemic durability and delayed treatment escalation including insulin compared with metformin monotherapy, especially in patients with young‐onset diabetes who face long disease duration. ⁵¹ , ⁵² , ⁵³ To avoid therapeutic inertia, the ADA/EASD recommended regular assessment of risk factor control and complications. In this regard, the JADE programme was designed to guide structured assessment (eye, feet, blood, urine) and systematically collect data for stratifying risk, empowering self‐management and reducing therapeutic inertia. ²⁸ In our previous reports, we consistently demonstrated that enrolment in the JADE register was associated with further reduction in risk factors and better self‐management on top of usual care in high‐, middle‐ and low‐income countries/regions. ³⁰ , ³¹ , ³² , ⁵⁴

4.5. Strengths and limitations

Our analysis has several limitations. First, due to volunteer/recruitment bias of participating physicians and patients, the JADE register might not fully represent the distribution of OGLDs and patient characteristics in the respective countries. Instead, the data were used to identify care gaps and drive quality improvement. This may limit the external validity of our findings. Second, the analysis was limited to patients treated solely with OGLDs. Due to the pragmatic nature of the study design, we did not collect data on treatment duration or adherence. Third, data in the JADE register were collected in real‐world practice and not from RCTs, therefore, missing data are inevitable. The JADE register was not automatically linked to pharmacy databases and 13% of patients had missing data for OGLDs. Compared with those included in the present analysis, patients with missing data were younger and had fewer complications but worse glycaemic control. Fourth, self‐reported hypoglycaemia events by patients might not be confirmed by SMBG or laboratory or medical records and were subject to recall bias. Rates of enrolment varied among countries/regions, which limited subgroup comparisons. Last, cross‐sectional analysis precluded causal inference, although our real‐world data accorded with what had been reported in RCTs, observational surveys and meta‐analyses, which are complementary.

Our study also has some strengths. The JADE register enrolled a prospective cohort of Asian patients who are generally under‐represented in clinical trials. The register has a well‐defined structure, procedures and objectives, with accrual of real‐world data gathered from 11 countries/regions in Asia, including 300 sites since 2007. These sites included private, public and subsidized centres attended by specialists and non‐specialists. As the JADE registry was introduced to the various countries over several years, the data were pooled from these different settings and adjusted for year of enrolment to control for any time‐related confounders (eg, change of practice, care standards or socioeconomic status). This approach was used to increase the generalizability of the results. This broad base of recruitment increased the utility of the register to inform real‐world practice regarding the heterogeneity of patient risk profiles, patterns of care and usage of medications to motivate continuing improvement and discover unmet needs.

In conclusion, in Asia, the majority of patients with T2D were treated with OGLDs, with approximately 60% of them treated with SUs, often in combination with metformin. Good glycaemic control, safety profiles and affordability associated with the use of SUs (including gliclazide) support their position as a key treatment option in patients with T2D.

AUTHOR CONTRIBUTIONS

Lee‐Ling Lim, Juliana C.N. Chan and Andrea O.Y. Luk developed the study design. Lee‐Ling Lim, Eric S.H. Lau, Siew Pheng Chan, Linong Ji, Soo Lim, Sirinart Sirinvaravong, A.G. Unnikrishnan, Andrea O.Y. Luk and Juliana C.N. Chan were involved in study conduct, data collection and plan of data analysis. Lee‐Ling Lim, Eric S.H. Lau and Johnny T.K. Cheung performed data analysis. All authors were involved in writing the manuscript.

CONFLICT OF INTEREST

Lee‐Ling Lim reports receiving grants and/or honoraria for consultancy or giving lectures from Abbott, AstraZeneca, Boehringer Ingelheim, Merck Sharp & Dohme, Novo Nordisk, Roche, Sanofi, Servier and Zuellig Pharma. Eric S.H. Lau and Johnny T.K. Cheung report no conflicts of interest. Siew Pheng Chan reports receiving grants and/or honoraria for consultancy or giving lectures from Abbott, AstraZeneca, Boehringer Ingelheim, Merck Serono, Merck Sharp & Dohme, Novo Nordisk, Sanofi, Servier and Zuellig Pharma. Linong Ji reports receiving consulting and lecture fees from Eli Lilly, Novo Nordisk, Merck, Sanofi‐Aventis, MSD, Servier and Boehringer Ingelheim. Soo Lim has been a member of advisory boards or has consulted for Merck, Sharp & Dohme, NovoNordisk and Servier, has received grant support from AstraZeneca, Merck, Sharp & Dohme and Astellas, and has served on the speakers' bureau of AstraZeneca, Boehringer Ingelheim, Eli Lilly & Co., Merck, Sharp & Dohme, CKD Pharmaceutical and NovoNordisk. Sirinart Sirinvaravong reports receiving honoraria for giving lectures from Boehringer Ingelheim, Merck, Novo Nordisk, Sanofi, Servier and Zuellig Pharma. A. G. Unnikrishnan reports receiving grants via affiliated institutions, advisory board and speaker contracts from AstraZeneca, Boehringer Ingelheim, Merck Sharp & Dohme, Sanofi, Serdia, Servier, Novo Nordisk and Roche. Andrea O.Y. Luk reports receiving grants and/or honoraria for consultancy or giving lectures from Amgen, AstraZeneca, Bayer, Boehringer Ingelheim, Merck Sharp & Dohme, Lee's Pharmaceutical, Roche and Sanofi. The proceeds have been donated to CUHK, the ADA and other charity organizations to support diabetes research and education. Viviana Cortese and Alexandra Durocher are employees of Servier. Juliana C.N. Chan reports received research grants through her affiliated institutions, honoraria and speakers' fees from Applied Therapeutics, Astra Zeneca, Bayer, Boehringer Ingelheim, Celltrion, Hua Medicine, Lee Powder, Lilly, Merck Sharpe Dohme, Merck Serono, Pfizer, Sanofi, Servier and Viatris Pharmaceutical, holds patents of genetic markers for predicting diabetes and its complications, and is a cofounder of a biotechnology start‐up company, GemVCare, with partial support from the Hong Kong Government Innovation and Technology Commission for providing precision diabetes care and Chief Executive Officer of Asia Diabetes Foundation on a pro bono basis.

PEER REVIEW

The peer review history for this article is available at https://publons.com/publon/10.1111/dom.14865.

Supporting information

Data S1. Plain Language Summary

Table S1. Clinical characteristics of patients with T2D at registration into the JADE register, stratified by the data availability on OGLD.

Table S2. Proportion of patients with T2D treated with OGLD at registration into the JADE Register, stratified by the use of SU and DPP‐4 inhibitors.

Table S3. Binary logistic regression analysis for the risk of hypoglycaemia in the 3 months before registration in SU‐ and/or DPP‐4 inhibitor‐treated patients with T2D, compared with patients receiving neither drug amongst patients treated with OGLD.

Table S4. Proportion of patients with T2D and treated with SU at registration into the JADE Register, stratified by the use of gliclazide and DPP‐4 inhibitors.

Table S5. Binary logistic regression analysis for the risk of hypoglycaemia in the 3 months before registration in gliclazide‐ and/or DPP‐4 inhibitor‐treated patients with T2D, compared with patients receiving neither drug in the SU‐treated group.

Lim L‐L, Lau ESH, Cheung JTK, et al. Real‐world usage of sulphonylureas in Asian patients with type 2 diabetes using the Joint Asia Diabetes Evaluation (JADE) register. Diabetes Obes Metab. 2023;25(1):208‐221. doi: 10.1111/dom.14865

DATA AVAILABILITY STATEMENT

Individual‐level data cannot be shared but data supporting the findings of this study may be available from the corresponding author upon reasonable request.