Benefits of Sodium-Glucose Cotransporter-2 Inhibitors with Renin-Angiotensin System Blockers in Type-2 Diabetes: A Cohort Analysis

Introduction

The prevalence of type 2 diabetes mellitus (T2DM) has increased over the past few decades in Taiwan [1]. Patients with T2DM carry a higher risk of cardiovascular disease (CVD) and nephropathy [2]. Furthermore, patients with T2DM without a history of CVD have an equal risk of cardiac events as those with a prior history of myocardial infarction [3]. This risk could be explained by the vascular damage from hyperglycemia, or one of several other existing comorbidities associated with increased risks, such as high blood pressure, dyslipidemia, and obesity. Diabetic kidney disease (DKD) is another complication of T2DM, seen in 25–30% of patients with T2DM [4,5] and accounting for one-third of end-stage renal disease cases requiring dialysis [6,7].

The renin-angiotensin system plays a key role in the renal pathophysiology of T2DM, and its pharmacological inhibition through ACEIs or ARBs has been shown to reduce cardiovascular and all-cause mortality [8,9]. Angiotensin II (Ang II), a key mediator in DKD pathogenesis, increases intraglomerular pressure through efferent arteriole constriction, leading to glomerular damage and proteinuria [10]. Ang II also drives renal inflammation and fibrosis via NADPH oxidase, contributing to oxidative stress [10]. By vasodilating efferent glomerular arterioles, renin-angiotensin system blockers (RASB) lower intraglomerular pressure [11]. RASBs have also demonstrated benefits in delaying the onset and progression of DKD, as evidenced by many clinical trials [12,13]. Thus, RASB use is the criterion standard for managing renal complications in high-risk T2DM patients.

Sodium-glucose cotransporter 2 inhibitors (SGLT2i) are new oral antidiabetic agents with a unique mechanism of action that functions independently of insulin secretion and action [14]. They lower plasma glucose by inhibiting glucose reabsorption in the renal proximal tubule, producing glucosuria. This unique mechanism of action corrects several metabolic and hemodynamic abnormalities beyond glycemic control and reduces the incidence of nephropathy [15,16].

While both SGLT2i and RASB (ACEI or ARBs) are well-established for their individual clinical benefits in managing T2DM, there is a critical lack of robust real-world evidence evaluating their combined use, especially in the high-risk patients who are most likely to benefit [17–19]. Moreover, there is debate regarding the synergistic effects of RASB combined with SGLT2i, as further reductions in intraglomerular pressure could decrease eGFR and heighten AKI risk. This gap in knowledge creates uncertainty for clinicians seeking to make evidence-based decisions for optimizing renal outcomes. By addressing this gap, this study aimed to analyze data from a nationwide Taiwanese database (real-world data) to shed light on the kidney safety and effectiveness of using SGLT2i together with RASB in treating patients with T2DM.

Material and Methods

STUDY DESIGN AND DATA SOURCES:

This research was a retrospective cohort study with a new-user and active-comparator design, drawing from a vast pool of information compiled within the National Health Insurance Research Database (NHIRD). This database is remarkably comprehensive, covering almost the entire population of Taiwan, a figure close to 99%, with records that have been meticulously collected since the year 1995. Prior to releasing the data analysis, Taiwan’s Ministry of Health, and Welfare (MOHW) took significant steps to ensure the privacy of the individuals whose data was included in our study. To this end, MOHW meticulously anonymized all the records of the beneficiaries making a claim, which effectively removed any personal identifiers from the dataset.

Because of this careful anonymization, it was determined that there was no longer a necessity to obtain informed consent from the subjects of the study, since they could no longer be directly or indirectly identified. Additionally, we sought and subsequently received a formal exemption from the need for informed consent by the Institutional Review Board of the Shin-Kong Wu Ho-Su Memorial Hospital. This relief from the usual consent requirement came after careful consideration and was officially sanctioned, as evidenced by the IRB approval number provided, which is 20220712R for reference. The data for this study were accessed on January 9, 2023.

STUDY POPULATION AND EXCLUSION CRITERIA:

This study analyzed a population of patients with T2DM, with an enrollment period from January 1, 2016 to December 31, 2016, based on the NHIRD (n=1 937 938). Figure 1 presents a flowchart outlining the process of patient selection. Patients enrolled in this study were required to have continuously used metformin and RASB medications for at least 3 months (n=595 501). Under Taiwan’s national health insurance (NHI) rules, metformin is the main treatment for diabetes and is recommended only for patients with a kidney function (eGFR) over 30 mL/min/1.73 m2. This criterion enabled us to effectively exclude patients with severely compromised renal function (eGFR <30 mL/min/1.73 m2).

The following were exclusion criteria for this study: (1) age less than 18 years or greater than 85 years or missing information (n=17 023); (2) history of kidney transplant before the index date (n=296); (3) history of stroke, pregnancy, heart failure (HF), immune system diseases, hospitalization for any reason, end-stage kidney disease (ESKD), acute kidney injury (AKI), coronary artery disease (CAD), or acute coronary syndrome (ACS) 3 months preceding the index date (n=155 449); (4) history of dialysis 3 months prior to the index date (n=153); (5) a diagnosis of cancer before the index date (n=21 171); (6) discontinuation of metformin and RASB meds within 3 months from index date (n=12 064); (7) use of either SGLT2i or DPP4i before the index date (n=193 532); (8) patients in the comparison group who did not take DPP4i at least 3 times (n=178 608). Excluding these cases ensured a more homogeneous study population and minimized potential confounding factors, as such conditions or prior treatments that could independently influence renal outcomes and bias assessment of the interventions.

The study population was divided into 2 groups. The case group included patients who used SGLT2i at least 3 times or more but had not used DPP4i (n=3705). The comparison group consisted of patients who had not used SGLT2i but had used DPP4i 3 times or more (n=13 500). DPP4i were selected as the comparison group owing to their prevalent application in the management of diabetes mellitus [20]. The index date was defined as the day after 3 months of SGLT2i use, and the comparison group shared this index date. Patients were followed from the index date until the occurrence of an event, death, or the end of the database period (December 31, 2020). To reduce confounding variables, propensity score matching (PSM) was done at a 1: 1 rate, adjusting for sex, age, and Charlson Comorbidity Index (CCI). After matching, the case group comprised 3622 individuals, mirroring the size of the comparison group.

COVARIATES:

The variables gathered included: sex, age, CCI score, and specific comorbidities (eg, hypertension, ischemic heart disease, arrhythmia, atrial fibrillation, stroke, chronic obstructive pulmonary disease, asthma, peptic ulcer, dyslipidemia, gout, and liver cirrhosis). Past medication history was also reviewed, including beta-blockers, calcium channel blockers, alpha-blockers, anticoagulants, diuretics, antithrombotic, insulin, sulfonylureas, thiazolidinediones, acarbose, lipid-lowering agents, urate-lowering agents, non-steroidal anti-inflammatory drugs, and sedative-hypnotics. The disease and drug codes are presented in Table 1.

STUDY OUTCOMES:

This study commenced on the index date. Patients were followed until the occurrence of clinical events, including acute kidney injury (AKI), CKD, initiation of erythropoietin (ESA) therapy, and initiation of long-term dialysis. Under Taiwan’s national health insurance (NHI) rules, ESA payments are approved for CKD patients with an eGFR under 15 mL/min/1.73 m2 and a hemoglobin level below 9 g/dL. Thus, CKD patients getting ESA in the NHIRD are in stage 5 of CKD (eGFR <15 mL/min/1.73 m2). This stage 5 definition is widely used in research [21–23]. If none of those clinical events occurred, this ongoing observation and data collection persisted until either the patient’s death or the end of the study period (December 31, 2020).

EXTERNAL DATASET VALIDATION:

The TriNetX dataset was chosen for external validation due to its large, diverse, and real-world population, which enhances the generalizability of the findings. TriNetX is a global federated health research network providing access to electronic medical records (diagnoses, procedures, medications, laboratory values, genomic information) across large healthcare organizations (HCOs). It allows users to search for patients meeting specific criteria in a de-identified database without the need for prior Institutional Review Board (IRB) approval [24,25]. This sensitivity test was run on the set of HCOs grouped into a network called Global Collaborative Network, which included 119 HCOs (n=156 955 589).

The validation analysis was performed on the date of April 22, 2024. A cohort consisting of individuals diagnosed with T2DM, exhibiting an eGFR of ≥30 mL/min/1.73 m², and concurrently utilizing metformin and RASB was assembled over the period from January 1, 2016 to December 31, 2016 (n=1 782 162).The criteria for exclusion encompassed any history of malignancy, status pertaining to transplanted organs, or prior dialysis treatment. From this population, 2 distinct cohorts were delineated: one receiving SGLT2i (n= 43 280) and the other administered DPP4i (n=53 146).

All the baseline characteristics cooperated in to PSM encompassed age, sex, hypertensive disorders, ischemic heart conditions, arrhythmias, atrial fibrillation, cerebrovascular accidents, chronic obstructive pulmonary disease, asthma, peptic ulcer disease, dyslipidemia, gout, liver cirrhosis, β-adrenergic blockers, calcium channel antagonists, alpha-adrenergic blockers, diuretics, lipid-lowering medications, insulin, sulfonylureas, thiazolidinediones, acarbose, cholesterol-lowering agents, urate-lowering therapies, non-steroidal anti-inflammatory medications, and sedative-hypnotic agents. To balance potential confounding variables, a 1: 1 PSM was applied, resulting in 41 411 matched users for each medication group. Patients in the study were matched on a one-to-one basis using a nearest neighbor algorithm without replacement, based on the propensity scores derived from a logistic regression model.

The targeted endpoints encompassed AKI, progression to advanced chronic kidney disease (CKD stage 5), and the terminal phase of renal function with dialysis. The outcomes of interest were within a specified timeframe of 1 days to 8 years following the index date, defined as the implantation of studied medications. If these clinical events did not occur, observation and data collection were continued until the last data entry or the end of the study period on April 22, 2024 employing the Cox proportional hazards regression model for analysis. All statistical analyses were conducted using the TriNetX platform, with a significance threshold set at a 2-sided P value of <0.05

STATISTICAL ANALYSIS:

The demographic findings are shown as percentages or averages along with their standard deviations. We compared categorical and continuous variables using the chi-square test and the t test, respectively. To minimize differences in baseline characteristics between the case and comparison groups for more accurate analyses, we applied propensity score matching at a 1: 1 ratio. This method pairs individuals from each group with similar propensity scores, calculated based on relevant covariates, to ensure that both groups are comparable. By balancing these covariates, propensity score matching helps reduce potential bias and confounding, allowing for a more reliable evaluation of the treatment effect [26]. Moreover, we used Cox proportional regression models to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) to evaluate the risk of the outcomes. The proportional hazard assumption was assessed using Schoenfeld residuals.

The initial Cox regression model was a crude analysis after propensity score matching, whereas the subsequent multivariable adjustment model included further adjustments for all baseline confounding variables to reduce residual confounding after propensity score matching. Baseline covariates encompassed comorbid conditions and pharmacological interventions pertinent to renal outcomes (Table 1). Additionally, in the subgroup analysis, we included only the group with more than 500 patients. Two-tailed P values less than 0.05 were considered statistically significant. All analytical procedures were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA).

Results

BASELINE CHARACTERISTICS:

Table 2 presents the demographic and clinical characteristics of diabetes patients using RASB in 2 groups: SGLT2i and DPP4i users, both before and after matching. Before matching, significant differences were observed between the groups in age, comorbidities, and medication use. SGLT2i users were slightly younger (mean age 59.2 vs 60.0 years) and had higher rates of comorbidities such as hypertension, ischemic heart disease, arrhythmia, atrial fibrillation, stroke, COPD, and asthma. They also had higher CCIS and were more likely to use medications like β-blockers, CCB, diuretics, antithrombotics, insulin, and TZD. After matching, these differences were largely balanced, with no significant differences in age, sex, or CCIS. However, SGLT2i users still showed higher rates of hypertension, ischemic heart disease, arrhythmia, atrial fibrillation, and stroke compared to DPP4i users. They were also more likely to use medications such as β-blockers, CCBs, diuretics, antithrombotics, and insulin.

THE OUTCOME OF ACUTE KIDNEY INJURY IN THE COMBINATION OF SGLT2I WITH RASB:

Figure 2A illustrates the Kaplan-Meier survival curves for AKI among groups using SGLT2i and DPP4i. The survival difference between these 2 groups was not statistically significant for AKI, as determined by a log-rank test with a P value of 0.125. Table 2 shows the rates of AKI in patients using SGLT2i and DPP4i after they have been matched, with figures of 0.8 and 0.6 per 100 person-years, respectively. The likelihood of experiencing acute renal failure as an adverse effect was comparable between the SGLT2i group and the DPP4i group. The risk difference was statistically insignificant in both the preliminary analysis (HR, 1.24; 95% CI, 0.94–1.63; P=0.126) and the multivariable analysis that adjusted for other factors (HR, 1.16; 95% CI, 0.87–1.55; P=0.304).

THE BENEFITS OF COMBINING SGLT2I AND RASB ON KIDNEY FAILURE:

SGLT2i combined with RASB outperforms DPP4i with RASB in improving renal outcomes in CKD, advanced kidney failure, and the initiation of long-term dialysis. Kaplan-Meier survival curves showed significant differences between SGLT2i and DPP4i groups for CKD (P<0.001), advanced kidney failure (P=0.019), and long-term dialysis initiation (P=0.035) (Figure 2B–2D).

Table 3 presents the incidence rates of several clinical events among SGLT2i users and DDP4i users in T2DM patients with RASB after matching, including CKD (3.2 vs 4.9 per 100 person-year), advanced kidney failure (0.3 vs 0.5 per 100 person-year), and initiation of long-term dialysis (0.2 vs 0.3 per 100 person-years). Moreover, the unadjusted HR was significant for CKD (0.66; 95% CI, 0.58–0.74, P<0.001), advanced kidney failure (0.64; 95% CI, 0.44–0.93, P<0.020), and initiation of long-term dialysis (0.61; 95% CI, 0.38–0.97, P<0.037). After multivariable adjusting, these 3 kidney outcomes still showed significant differences: the HR for CKD was 0.74 (95% CI, 0.65–0.84, P<0.001), for advanced kidney failure was 0.62 (95% CI, 0.42–0.92, P<0.017), and for beginning long-term dialysis was 0.53 (95% CI, 0.32–0.87, P<0.012) (Table 3).

SUBGROUP ANALYSIS:

A series of stratified analyses were conducted to evaluate the reliability of our findings in (Figure 3). The reduced HRs of CKD development among diabetic patients with RABS use in favor of SGLT2i use were consistent across almost all patient subgroups, except those with peptic ulcer and those without lipid-lowering agents (Figure 3A). In advanced kidney failure, only half of the subgroups experienced benefits from using a combination of SGLT2i and RASB. Nevertheless, the data suggests a favorable trend using this combination therapy (Figure 3B).

EXTERNAL DATA VALIDATION:

Table 4 delineates the demographic and clinical attributes of patients with RASB who are administered SGLT2i and DPP4i after the implementation of PSM (n=41 411 per cohort). The sex distribution within both cohorts was comparable, with 53% of the population being male, and the mean age 62 years. Prevalent comorbid conditions were hypertension (85%), dyslipidemia (73%), and ischemic heart disease (23%), revealing no statistically significant disparities between the groups. The use of medications such as β-blockers, diuretics, insulin, and agents for lipid-lowering also reflects similarity between the 2 cohorts. Additional medical conditions, including arrhythmia, stroke, chronic obstructive pulmonary disease (COPD), and asthma, were not significantly different. In summary, the groups were effectively matched (all P value >0.05).

Table 5 shows that SGLT2i group had significantly lower rates of AKI (15.2% vs 22.5%; HR, 0.91, 95% CI, 0.88–0.93), advanced kidney failure (0.8% vs 2.0%; HR, 0.63, 95% CI, 0.55–0.72), and initiation of long-term dialysis (4.1% vs 3.3%; HR, 0.79, 95% CI, 0.74–0.84) compared to the DPP4i group. These results are consistent with the finding in NHIRD, except for the outcome of AKI.

Discussion

This population-based cohort study demonstrated that combined use of SGLT2i plus RASB has clinical benefits in participants with early T2DM who used metformin. This combination demonstrated improvement in renal outcomes, including the prevention of CKD development, delaying the progression to severe kidney failure (identified as CKD stage 5), and averting end-stage kidney disease. The study also found no increased risk of AKI with this combination treatment compared to the use of DPP4i and RASB. Additionally, subgroup analyses indicate a positive trend in favor of combination therapy. Thus, the evidence suggests that the combination of SGLT2i and RASB may be more effective in promoting kidney health in T2DM patients.

SGLT2i have emerged as a promising drug class in clinical practice. Traditionally, SGLT2i are reserved for adjunctive therapy in T2DM if adequate control cannot be achieved by metformin. However, SGLT2i can also be used as monotherapy in cases where metformin is contraindicated [27]. Beyond their role in glucose control, SGLT2i have unique mechanisms that can benefit individuals with CVD, such as reduced mortality, lower risk of hospitalizations due to heart failure, enhanced blood pressure regulation, and improved overall cardiovascular health [15,28–30]. In a meta-analysis of several clinical randomized controlled trials (EMPAREG OUTCOME, CANVAS, and DECLARE-TIMI 58) by Zelniker et al, T2DM patients with previous atherosclerotic CVD can benefit from SGLT2i, with improved cardiovascular outcomes such as new-onset or recurrent heart failure and hospitalization [31].

Large-scale studies (EMPA-REG, DECLARE, and CANVAS) have reported the renoprotective effects of SGLT2i, including reduced risks of serum creatinine doubling, macroalbuminuria, and end-stage kidney disease [16,30,32]. In a meta-analysis of those trials, SGLT2i reduced the incidence of worsening renal function, end-stage renal disease, or renal death by 45% (HR 0.55; 95% CI 0.48–0.64) [31]. SGLT2i has renoprotective effects by blocking the reabsorption of glucose and sodium in the renal proximal tubule, leading to increased urinary glucose excretion and osmotic diuresis. This mechanism helps reduce hyperglycemia and blood pressure, which are critical factors in the progression of DKD [33,34]. Additionally, by decreasing sodium reabsorption, SGLT2i enhances sodium delivery to the distal tubule, restoring tubuloglomerular feedback. This process lowers intraglomerular pressure and hyperfiltration, thereby protecting the kidneys [34,35]. Furthermore, oxidative stress is a key factor linking hyperglycemia to vascular complications in diabetes, especially DKD. It results from an imbalance between pro-oxidants and antioxidants, causing excessive reactive oxygen species (ROS) production [36]. Mitochondrial dysfunction, NADPH oxidase activity, and enzymatic pathways drive ROS generation, activating NF-κB and TGF-β, which promote inflammation and fibrosis [36,37]. Inflammation, which is both a cause and result of oxidative stress, amplifies this cycle. ROS regulate immune pathways, driving chronic inflammation, while cytokines and stress kinases disrupt insulin signaling, worsening glucose metabolism and organ dysfunction [37]. SGLT2i lower ROS by reducing glucose levels, improving mitochondrial function, and promoting ketone production as a less oxidative energy source [38]. This protects blood vessels, the heart, and kidneys, reducing cardiovascular events and slowing kidney disease. They also reduce inflammation by lowering pro-inflammatory cytokines like IL-6 and TNF-α, encouraging an anti-inflammatory macrophage phenotype, and reducing systemic inflammation by improving glycemic control and reducing visceral fat [39]. These multifaceted benefits have led to their inclusion in clinical guidelines for DKD management.

Since SGLT2i has a unique role in treating patients with T2DM, the efficacy of its combination with RASB has been a crucial issue. Lytvyn et al conducted a prospective study to evaluate the effects of combination therapy in type 1 DM patients with renal hyperfiltration [19]. Adding SGLT2i to RASB resulted in an initial drop in GFR, as expected, but there was also a benefit of reduced oxidative stress and improved blood pressure. Recent systematic reviews and meta-analyses have also confirmed that combination therapy can significantly reduce albuminuria, but with no difference in eGFR compared to standard care [17,18]. Regarding outcomes, Zhao et al demonstrated that combination therapy had greater efficacy in lowering MACE, cardiovascular death or hospitalization heart failure, and composite kidney outcomes compared to RASB monotherapy among T2DM patients [40]. These results were consistent with our real-world database analysis.

The initiation of RASB and SGLT2i therapy can cause a temporary reduction in renal function, which can be intensified with combination therapy [41,42]. However, this reduction was not found in our investigation. This discrepancy may be attributed to the fact that mild renal impairments were overlooked in clinical settings, resulting in an underrepresentation of AKI diagnoses. Moreover, any preliminary decline in renal function may have ameliorated prior to subsequent outpatient consultations. The infrequent occurrence of AKI events could have further constrained the statistical power of the study to identify nuanced disparities among the groups. Additionally, since the research was based on real-world data, heterogeneity in clinical practices, medication compliance, and patient surveillance may have attenuated the discernible impacts of SGLT2 inhibitors on AKI risk.

The findings of this study align with the renal benefits observed in landmark trials such as CRDENCE [43] and EMPA-REG [44], but notable distinctions warrant further discussion. Similar to CREDENCE, which demonstrated a 34% reduction in the risk of renal-specific outcomes with canagliflozin (HR: 0.66; 95% CI, 0.53–0.81), and EMPA-REG, which reported a 39% reduction in incident or worsening nephropathy with empagliflozin (HR: 0.61; 95% CI, 0.53–0.70), our study highlights the renoprotective effects of combining SGLT2i with RASB. Specifically, the current study showed significant reductions in advanced kidney failure (HR: 0.62; 95% CI, 0.42–0.92) and long-term dialysis (HR: 0.61; 95% CI, 0.38–0.97) compared to DPP4i. However, unlike EMPA-REG, which included a high-cardiovascular-risk population, our study focused on a broader real-world cohort with fewer cardiovascular comorbidities, potentially explaining the absence of AKI risk observed in EMPA-REG. Additionally, while CREDENCE exclusively targeted patients with albuminuric CKD, our study excluded those with eGFR <30 mL/min/1.73 m2, limiting direct comparisons in advanced renal disease stages. Methodological differences, such as the retrospective design and reliance on real-world data in our study versus the randomized controlled designs of CREDENCE and EMPA-REG, may also account for discrepancies. Despite these differences, the consistent renal benefits across studies underscore the potential of SGLT2i, particularly when combined with RASB, to improve kidney outcomes in T2DM. Future research should explore these therapies in diverse populations to validate and expand upon these findings.

In the subgroup analysis of our cohort, combination therapy of SGLT2i and RASB reduced the incidence of CKD in most subgroups. Among patients with advanced kidney failure, those with female sex, underlying gout, asthma, and peptic ulcer disease seemed to have less benefit from combination therapy. Patients with history of NSAID use also did not benefit from combination therapy, likely due to its nephrotoxicity. NSAIDs can induce AKI by interfering with production of prostaglandins, which are key vasodilators in the kidneys. Decreased prostaglandin levels can lead to reduced renal perfusion, causing ischemia, especially in the setting of RAAS inhibition [45].

Our study had several strengths. First, a nationwide database was used, ensuring the broad relevance for our findings. This finding was also validated by global dataset (TriNetX). Second, we specifically selected T2DM individuals using metformin, ensuring similar CKD stages in both case and comparison groups (eGFR >30 mL/min/1.73 m2). Third, we applied an active-comparator and new-user design to approximate an RCT. While the number of observational studies on clinical interventions has grown significantly, they face challenges like confounding by indication compared to RCTs. Active-comparator and new-user designs help reduce these biases and improve validity by aligning more closely with RCT methods. Active-comparator studies compare the drug of interest with another commonly used for the same condition, rather than with a non-user group, ensuring similar treatment indications and reducing differences in patient characteristics [46]. However, our study also has some potential limitations. First, key confounding factors which could be clinically relevant, such as physical features, lifestyle, and laboratory data, were not included in the NHIRD, and we thus were unable to consider these in our analysis. Second, since this was a retrospective cohort analysis and not an RCT, there could be baseline disparities between groups, potentially introducing bias due to confounding factors. To minimize potential bias, we implemented a PSM stratified by sex, age, and the CCI scores. As demonstrated in Table 2, certain discrepancies remain evident between the cohorts. Consequently, we performed multivariable adjustments to address and minimize this bias. Within the TriNetX framework, baseline characteristics were thoroughly incorporated into the propensity score PSM, and Table 4 illustrates that both cohorts were comparable. Third, the retrospective design employed in this study is a limitation, as it is susceptible to biases and confounding variables arising from pre-existing data, thereby precluding the establishment of causal relationships. These concerns may significantly influence the interpretation of observed associations. Nonetheless, this design facilitates the analysis of extensive datasets and real-world outcomes, thereby providing invaluable insights and a foundational basis for subsequent prospective investigations. Fourth, the laboratory data, lifestyle factors, and the severity of disease were not accessible within the HIRD and the absence of these variables may have resulted in residual confounding. Finally, as the research concentrated on participants residing in Taiwan, it is imperative to exercise caution when extrapolating the results to international cohorts with different characteristics. Genetic predispositions and ethnic variances, such as discrepancies in drug metabolism, risks of complications, and the progression of renal diseases, can substantially influence therapeutic responses. Polymorphisms within the renin-angiotensin system or sodium-glucose cotransporter 2 pathways could potentially modulate the safety and effectiveness of SGLT2i and RASB therapies. The healthcare infrastructure is also of paramount importance, as Taiwan’s universal healthcare system guarantees access to these therapeutic modalities, in contrast to regions with resource limitations. Cultural and lifestyle determinants, encompassing dietary habits, physical activity levels, and the prevalence of obesity, may further affect clinical outcomes. Although the validation using the global TriNetX dataset corroborates the robustness of findings across diverse contexts, further investigations in multi-ethnic populations and varied healthcare systems are needed to ascertain the generalizability of these results.

Conclusions

Individuals with 2DM could achieve better kidney health outcomes by combining SGLT2i with RASB in real-world settings. To maximize its impact on preserving kidney function and slowing disease progression, incorporating this combination therapy into clinical guidelines for managing diabetic nephropathy is essential. Subsequent investigations should examine the prolonged effects of this regimen, particularly its influence on mortality and cardiovascular health, while subgroup analyses may reveal which patients derive the most benefit, facilitating more tailored treatment strategies.

Data Availability Statement

The dataset used in this study is held by the Taiwan Ministry of Health and Welfare (MOHW). Owing to the General Data Protection Regulation, the dataset is not available on request from the corresponding author. Any researcher interested in accessing this dataset can apply for access. Please visit the website of the National Health Informatics Project of the MOHW (https://dep.mohw.gov.tw/DOS/mp-113.html). TriNetX is a network connecting multiple research centers that offers instant access to anonymized data from the electronic health records of participating healthcare organizations. Researchers can access this database at https://live.trinetx.com.