Sodium-glucose cotransporter-2 inhibitors (SGLT2is) are now currently the first-line treatment endorsed by worldwide and national guidelines for patients with Type 2 Diabetes Mellitus (T2DM) who have heart failure, Atherosclerotic Cardiovascular Disease (ASCVD), a high-risk of ASCVD, and Chronic Kidney Disease (CKD). ^{1– 4} Apart from that, SGLT2is are also recommended for T2DM patients who are prone to hypoglycemia and are overweight or obese. ⁵ At present, the SGLT2is available in Indonesia are dapagliflozin and empagliflozin, which were launched in 2016 and 2017, respectively. ^{6– 9}

Numerous prior extensive retrospective cohort studies have compared the efficacy of empagliflozin and dapagliflozin regarding cardiovascular occurrences, although the findings are inconsistent. ^{10– 12} These studies do not adequately represent the T2DM population in Southeast Asia, especially in Indonesia, which has the fifth-greatest number of T2DM patients globally. ¹³ Diabetes elevates the risk of CKD, coronary artery disease, and heart failure. Cardiorenal complications elevate the risk of mortality, hospital readmission, and deteriorate health-related quality of life. ^{14– 16} Consequently, efficient pharmaceutical approaches to achieve glycemic control and safeguard against cardiorenal complications are essential for patients with diabetes mellitus. Therefore, it is necessary to investigate the actual context of regular clinical practice in Indonesia, acknowledging potential differences in genetics, socioeconomic status, lifestyle, and healthcare systems that may affect cardiorenal outcomes. ^{17– 20} A meta-analysis examining the impact of race or ethnicity on cardiovascular outcomes from antidiabetic treatments indicates that varying racial or ethnic backgrounds may result in disparate outcomes. ²¹ A retrospective cohort study in the United Kingdom showed that white, black, and Asian ethnicities had different glycemic control. Asian patients with T2DM exhibit lower levels of obesity and greater insulin sensitivity compared to Caucasians, ²² which may influence cardiorenal outcomes. Thus, this study seeks to address the gap by comparing the effects and safety of dapagliflozin and empagliflozin on cardiorenal risk factors in T2DM Indonesian patients over a 12-month treatment period. The findings will provide valuable insights into these agents’ real-world effectiveness and safety profiles, guiding personalized diabetes management in Indonesia.

A total of 502 eligible patient records were collected from three data-collecting sites, resulting in the inclusion of 319 patients following the screening process, with 154 and 165 patients incorporated and evaluated in the dapagliflozin and empagliflozin groups, respectively, as seen in Figure 1. The baseline demographics exhibited similarities between the groups, except age (55 years in the dapagliflozin group versus 58 years in the empagliflozin group), history of sulfonylurea usage (62 patients in the dapagliflozin group versus 93 patients in the empagliflozin group), history of thiazolidinedione usage (23 patients in the dapagliflozin group versus seven patients in the empagliflozin group), history of acarbose usage (5 patients in the dapagliflozin group versus zero patients in the empagliflozin group), history of beta-blocker usage (10 patients in the dapagliflozin group versus 30 patients in the empagliflozin group), weight (80.7 kg in the dapagliflozin group versus 74.175 kg in the empagliflozin group), and eGFR (90.560 mL/min/1.73 m ² in the dapagliflozin group versus 81.667 mL/min/1.73 m ² in the empagliflozin group), as seen in Table 1. The average duration of SGLT2 is use observed in this study was 14.18 months.

Paired data analysis of dapagliflozin and empagliflozin demonstrated a significant reduction in HbA1c (-1.121% vs. -0.986%), FPG (-29.531 mg/dL vs. -25.238 mg/dL), SBP (-8.382 mmHg vs. -4.341 mmHg), and LDL-C (-12.212 mg/dL vs. -17 mg/dL) after 12 months of administration. Nonetheless, substantial decreases in weight, BMI, DBP, triglycerides, ASCVD risk, and a rise in HDL-C were observed exclusively in the dapagliflozin group. Empagliflozin markedly decreased total cholesterol by -16.586 mg/dL. Both groups had a reduction in eGFR at 12 months, which is -0.927 mL/min/1.73 m ² in the dapagliflozin group and -2.169 mL/min/1.73 m ² in the empagliflozin group, as illustrated in Table 2. Both groups had similar patterns in enhancing glycemic control, blood pressure, lipid profile, and ASCVD risk, alongside a reduction in eGFR.

The comparative analysis of effectiveness revealed that dapagliflozin significantly reduced body weight (-2.311 ± 5.173 kg vs. -0.875 ± 4.954) and BMI (-0.702 ± 1.863 vs. -0.316 ± 1.919) compared with empagliflozin. Due to a substantial disparity in baseline weight between the two groups, a supplementary analysis was conducted utilizing the percentage of weight difference, calculated as the difference between initial and final weight divided by the baseline weight, subsequently expressed as a percentage to normalize the baseline discrepancy, yielding robust results that favor dapagliflozin (-2.566% ± 5.863 vs. -1.080% ± 6.493; p = 0.001). No significant changes were seen in the mean values of HbA1c, FPG, SBP, DBP, LDL-C, HDL-C, triglycerides, total cholesterol, eGFR, and ASCVD risk (p > 0.05), as seen in Table 3. Additional analysis to overcome the eGFR difference between the two groups by utilizing the percentage of eGFR difference shows robust results (-3.410 ± 15.343 vs. -2.526 ± 17.913; p = 0.698), meaning no difference between the two groups. The sensitivity analysis, by excluding missing data, yielded a robust conclusion for each comparison. However, the multivariate linear regression results indicated that differences in sulfonylurea use history and baseline body weight between groups influenced the body weight and BMI parameters. The adjusted outcomes for confounding indicated a weight reduction of -1.969 kg in the dapagliflozin group compared to -1.233 kg in the empagliflozin group; p = 0.282 for BMI parameter, and -0.616 in the dapagliflozin group versus -0.367 in the empagliflozin group; p = 0.467.

The safety study of both groups revealed no significant differences in any adverse event components, including genital infections, UTI, hypoglycemia, diabetic ketoacidosis, fractures, lower extremity amputations, and overall adverse events (p > 0.05). These results are illustrated in Table 4.

According to paired data analysis, dapagliflozin and empagliflozin have comparable efficacy in enhancing glycemic control, specifically for reducing HbA1c and FPG levels. This aligns with the findings of a prior retrospective study including T2DM patients, which demonstrated a reduction in HbA1c in both groups after an average follow-up period of 24.5 months. ³¹ Nonetheless, a separate retrospective trial involving diabetic individuals with a history of chronic kidney disease indicated that dapagliflozin 10 mg was more effective than empagliflozin in reducing HbA1c, at both 10 mg and 25 mg dosages. ³²

The results of this trial indicate similar efficacy between groups in reducing body weight and BMI; even though before controlling for differences in sulfonylurea history and baseline body weight, the findings favor dapagliflozin. Sulfonylureas augment the visceral fat compartment, resulting in increased body weight ³³ and biased the body weight outcome. A prior meta-analysis demonstrated a similar result, showing comparable effects between dapagliflozin and empagliflozin on weight gain. ³⁴

An extensive retrospective cohort research in Taiwan showed that dapagliflozin resulted in a 31% reduction in LDL-C compared to empagliflozin, particularly in patients with baseline LDL levels below 100 mg/dL. ¹¹ The superior reduction in LDL-C within the dapagliflozin group may positively influence the prevention of cardiovascular events. A substantial retrospective cohort research conducted in Korea demonstrated that dapagliflozin administration in patients with T2DM decreased 24% the incidence of cardiovascular mortality and 16% hospitalization resulting from heart failure. ¹² The findings of the Korean study contradict those of another extensive retrospective cohort trial, which demonstrated empagliflozin’s superiority for both outcomes. ¹⁰ The disparity is likely due to the later study’s patients having a history of heart failure and being predominantly Caucasian, with only 4.6% identifying as Asian. ¹⁰ Consequently, dapagliflozin may be more effective in mitigating the risk of heart failure among Asians. The resemblance of ASCVD risk values between the two groups is attributable to the similar risk assessment components in this study, which include comparable SBP, HDL-C, and total cholesterol values. ³⁵ The congruence of SBP values is corroborated by prior meta-analyses, indicating no significant difference between dapagliflozin 10 mg and empagliflozin at both 10 mg and 25 mg dosages. ³⁴

A prior retrospective cohort study examining kidney function in individuals with T2DM and CKD stages G1-G4 over an 18-month follow-up revealed that the dapagliflozin 10 mg group exhibited a superior increase in eGFR compared to the empagliflozin dosages of 10 and 25 mg. ³² A separate retrospective investigation indicated an elevation in eGFR after an average usage length of 23.7 months, based on paired data analysis results, in both the dapagliflozin and empagliflozin groups. ³¹ However, the findings of the prior study contrast with those of the current study, likely attributable to the shorter follow-up period in the latter. Furthermore, the patient features in the present investigation diverge from those in the prior comparative analysis, which exclusively included patients with a history of CKD.

The findings of this investigation indicated comparable safety between the two groups. A retrospective cohort conducted in Turkey demonstrated analogous findings, particularly with the frequency of urinary tract infections and genital infections. ³¹ Furthermore, the utilization of SGLT2is is regarded as safe, presenting a limited risk of adverse events; however, with the increasing risk of genital infections, including in Asian. ^{36– 39} Adverse effects in the current study may remain unrecorded concerning the documentation of safety components. Thus, further evaluation is highly recommended.

This study has the advantage of including patients from various levels of healthcare facilities, from primary to tertiary levels. Nonetheless, multiple limitations exist in the study; primarily, it is a retrospective analysis characterized by missing data. In this case, researchers have implemented multiple data imputation techniques and conducted sensitivity analyses, which yielded robust outcomes. The second constraint is the significant difference between the two groups in some of the baseline characteristics; however, the adjustment has been made to the affected result. The third constraint is the potential for non-compliance, which may impact the outcomes. Last but not least, this study was conducted exclusively in an Indonesian population, which may limit the generalisability of the findings to populations in other geographic regions. Cultural factors, genetic backgrounds, disease epidemiology, and treatment access may differ in other countries. However, the observed associations may still be relevant to similar settings in Southeast Asia or other low- and middle-income countries with comparable patient profiles and healthcare infrastructures.

Dapagliflozin and empagliflozin both effectively reduced cardiorenal risks in Indonesian T2DM patients. Although dapagliflozin demonstrated additional advantages in terms of weight loss and BMI reduction, these disparities were not statistically significant following the adjustment. Both medications exhibited comparable safety profiles, supporting their use as equally effective options in personalized diabetes care. Longer-duration, larger-scale studies are needed to confirm the results of this study.

Before the commencement of the investigation, the research design underwent evaluation by the Investigation Ethics Committee of RSCM (KET-749/UN2.F1/ETIK/PPM.00.02/2024) dated September 26, 2024; and the Atma Jaya Catholic University of Indonesia (01/05/KEP-FKIKUAJ/2024) dated May 7, 2024. There is no necessity for informed assent in this study, as it is retrospective.