Comparison of the Effects of SGLT-2i Versus GLP-1RA on Cardiovascular and Renal Outcomes in Patients With Type 2 Diabetes, Based on Baseline Renal Function Free

Diabetes is a collection of heterogeneous metabolic disorders characterized by hyperglycemia resulting from defects in insulin secretion and/or action. Over the past decade, the incidence of type 2 diabetes (T2D) has been increasing annually, with an average annual growth rate exceeding 1.5% (1). By 2045, diabetes is projected to affect a minimum of 783 million individuals globally (2). Diabetes, along with cardiovascular diseases (CVDs) such as heart failure (HF) and chronic kidney disease (CKD), are prevalent and interrelated conditions. Moreover, diabetes is a high-risk factor for atherosclerotic cardiovascular disease, HF, and CKD (3). Notably, the coexistence of diabetes and CKD exerts a synergistic effect on the risk of CVD (4). Consequently, preventing the progression of CKD and cardiovascular events is paramount in treating patients with T2D.

As innovative medications in diabetes care, sodium–glucose cotransporter 2 inhibitors (SGLT-2i) and glucagon-like peptide 1 receptor agonists (GLP-1RA) have been shown across numerous randomized controlled trials (RCTs) to have significant cardioprotective benefits (5–9). Importantly, SGLT-2i was recommended for the treatment of HF, regardless of the patient's diabetes status. Moreover, the 2024 American Diabetes Association Diabetes Guidelines endorse SGLT-2i as a comprehensive cardiovascular risk management therapy for patients with T2D with either key cardiovascular risk factors or an established CVD (10). In the realm of renal health, a meta-analysis along with multicenter RCTs, such as the Dapagliflozin and Prevention of Adverse Outcomes in CKD trial, have consolidated the evidence that both SGLT-2i and GLP-1RA effectively reduce urinary albumin excretion and renal composite end-point events in patients with T2D, including a sustained decline in estimated glomerular filtration rate (eGFR), increased in serum creatinine levels, end-stage renal disease, and death related to renal disease (11–13). Notably, GLP-1RA stood out for its capacity to slow the decline in eGFR and reduce urinary albumin excretion in patients with T2D, thereby conferring potential renal benefits (14). In summary, the cardiorenal benefits offered by SGLT-2i and GLP-1RA are of paramount importance.

Although SGLT2i and GLP-1RA provide substantial cardiovascular and renal benefits for patients with T2D, there is a notable absence of studies focusing on their differential efficacy in cardiovascular and renal outcomes across a spectrum of eGFR levels. Therefore, in this study, we investigated the efficacy differences between SGLT2i and GLP-1RA in patients with T2D across various levels of eGFR.

Our study was conducted based on the Preferred Reporting Items for Systematic Review and Meta-analysis checklists and guidelines (15,16). We prospectively registered this systematic review in the International Prospective Register of Systematic Reviews database (registration no. CRD42024553603).

We systematically searched the PubMed, Cochrane Library, Web of Science, and Embase databases, covering their inception until 7 January 2025. The following keywords were applied: (“Sodium-Glucose Transporter 2 Inhibitors” OR “SGLT-2 Inhibitors” OR “Sotagliflozin” OR “Canagliflozin” OR “Dapagliflozin” OR “Empagliflozin”) OR (“Glucagon-Like Peptide-1 Receptor” OR “GLP-1 Receptor” OR “Dulaglutide” OR “Semaglutide” OR “Liraglutide” OR “Exenatide”) AND (“Diabetes Mellitus” OR “Diabetes Mellitus, Type 2” OR “Type 2 Diabetes Mellitus”) AND (“Renal Insufficiency, Chronic” OR “Chronic Kidney Diseases” OR “Chronic Renal Diseases”).

The selection criteria for included publications were as follows: 1) the type of study should be an RCT; 2) the study included adult patients (≥18 years old) with T2D; 3) the study compared SGLT-2i or GLP-1RA with a placebo; 4) risk of cardiovascular and renal outcomes between treatment and placebo groups with different eGFR were compared; and 5) the study was published in English.

This study evaluated five outcomes: major adverse cardiovascular event (MACE), renal outcome, hospitalization for heart failure (HHF), all-cause death (ACD), and cardiovascular death (including fatal stroke and fatal myocardial infarction). MACE was defined as a composite of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke. If data on nonfatal myocardial infarction and stroke were unavailable, total myocardial infarction and stroke data were used instead. Renal outcome was defined as a comprehensive end point indicating the progression to advanced stages of CKD (e.g., the continuous decline of eGFR to <15 mL/min/1.73 m², the doubling of serum creatinine level, renal death).

We used EndNote software to manage the retrieved literature. After screening the title and abstract, the article meeting the inclusion criteria was obtained for evaluation and data extraction. In addition, two reviewers (Y.W. and C.X.) extracted data independently through Microsoft Excel. Different opinions on the data extraction process were solved by the other two reviewers (M.L. and G.X.). The data we extracted covered a range of aspects, including geographic location, patient characteristics, study design characteristics, details of intervention and placebo arms, outcomes of interest, and hazard ratios with their 95% CIs. We made great efforts to contact corresponding authors via email for information that could not be obtained directly. Three reviewers (Y.W., M.L., and C.X.) assessed the risk of bias for all studies independently according to the Cochrane Risk of Bias tool (17). Risk of bias for each domain was evaluated as a high, low, or unclear for included studies. Disagreements were resolved by the fourth reviewer (G.X.).

We performed a network meta-analysis using Stata, version MP 18.0. Risk ratios (RRs) and 95% CIs were used to present the efficacy of treatments. If the 95% CI of an RR did not include 1, the difference between the two groups was considered statistically significant; otherwise, it was not. We installed R, version 4.3.0 (ggplot2 and robvis packages), to evaluate publication bias and installed STATA 18.0 (mvmeta and network packages) to draw forest plots and to conduct heterogeneity tests and sensitivity analysis. In this network meta-analysis, none of the five outcomes formed closed loops, meaning there was only indirect evidence between SGLT-2i and GLP-1RA. Therefore, it was unnecessary to test the inconsistency of this network meta-analysis. Heterogeneity was evaluated using the I² statistic, which was categorized as follows: unimportant (0% < I² <40%), moderate heterogeneity (30% < I² <60%), substantial heterogeneity (50% < I² <90%), and considerable heterogeneity (75% < I² <100%) (18). Finally, we conducted a sensitivity analysis of the outcomes with P < 0.05 in the heterogeneity test to evaluate the robustness of the results.

The original contributions presented in the study are included in the article or the Supplementary Material. Additional inquiries can be directed to the corresponding author.

We retrieved a total of 7,892 articles from PubMed (n = 1,500), Cochrane Library (n = 676), Web of Science (n = 1,652), and Embase (n = 4,064) in our primary search. During the process, another 8 articles were identified through references. After exclusion of duplicates and screening of titles and abstracts, 81 records remained for full-text evaluation. After detailed evaluation and citation searching, 17 articles comprising 10 nonoverlapping RCTs met the inclusion criteria and were included in the network meta-analysis. Of the 10 studies, 5 (19–28) compared SGLT2i with placebo across different eGFR levels, and the other 5 studies (29–35) compared GLP-1RA with placebo under varying eGFR levels. The detailed study-filtering process is shown in Fig. 1.

The characteristics of the included studies are presented in Table 1 and Supplementary Table 1. The pooled population consisted of 87,334 patients with T2D: 47,411 patients in SGLT-2i studies (n = 26,262 in the intervention group and n = 21,149 in the control group), and 39,923 patients in GLP-1RA studies (n = 19,939 in the group treated with GLP-1RA and n = 19,984 in the control group). Supplementary Table 2 lists the number of participants contributed by the source studies and the definitions for MACE outcomes and renal outcomes used in these contributing studies. Although the definitions of MACE and renal outcomes differed slightly among the source studies included in the analysis, they were sufficiently similar to be used for analysis.

We assessed the risk of bias in those trials using the Revised Cochrane Risk of Bias Tool (RoB 2.0) (36). The quality evaluation of the included studies is shown in Supplementary Fig. 1. All trials were evaluated as low risk in 5 outcomes (28).

Within the range of eGFR of 30–60, 60–90, or >90 mL/min/1.73 m², SGLT2 and GLP-1RA did not show statistical differences in reducing MACE when compared with each other (RR 0.98 [95% CI 0.79, 1.21]; 0.95 [95% CI 0.78, 1.16]; and 0.90 [95% CI 0.76, 1.07], respectively; Supplementary Fig. 2).

Compared with GLP-1RA, SGLT-2i therapy correlated with a lower risk of HHF in the eGFR 30–60 mL/min/1.73 m² level (RR 1.87; 95% CI 1.15, 3.04) and the 60–90 mL/min/1.73 m² level (RR 1.37; 95% CI 1.05, 1.78). There was substantial heterogeneity within the eGFR 30–60 mL/min/1.73 m² level (I² = 85.1%; P < 0.001) (Fig. 2).

Compared with SGLT-2i, GLP-1RA demonstrated a lower correlation with ACD risk in patients with an eGFR >90 mL/min/1.73 m² (RR 0.75; 95% CI 0.58, 0.97). Although there was substantial heterogeneity within the eGFR 60–90 mL/min/1.73 m² level (I² = 75.6%; P = 0.006), the sensitivity analysis results indicated that this study's conclusion was robust (Fig. 3).

When considering different levels of renal function (i.e., 30–60, 60–90, or >90 mL/min/1.73 m²), the effect of SGLT-2i treatment was not statistically different compared with GLP-1RA treatment in reducing the risk of cardiovascular death (RR 1.08 [95% CI 0.81, 1.45]; 1.07 [95% CI 0.70, 1.63]; and 0.72 [95% CI 0.51, 1.00], respectively). Despite substantial heterogeneity within the eGFR 60–90 mL/min/1.73 m² level (I² = 56.7%; P = 0.042), the results of sensitivity analysis indicated the conclusion of this study was robust (Supplementary Fig. 2).

In terms of renal outcome, when the eGFR was >90 mL/min/1.73 m², SGLT-2i therapy demonstrated an effect in reducing the risk of kidney outcome compared with GLP-1RA treatment (RR 1.80; 95% CI 1.15, 2.84). There was substantial heterogeneity within the eGFR >90 mL/min/1.73 m² level (I² = 75.8%; P = 0.016) (Fig. 4).

We conducted heterogeneity tests to check the robustness of the results, which are detailed in Supplementary Fig. 3. When the result was statistically heterogeneous at P < 0.05, we performed sensitivity analyses (detailed in Supplementary Fig. 4). Specifically, when comparing the reduction of risk of HHF between SGLT-2i and GLP-1RA treatments in patients with eGFR 30–60 mL/min/1.73 m², the Exenatide Study of Cardiovascular Event Lowering (EXSCEL) and Cardiovascular Outcomes Following Ertugliflozin Treatment in Type 2 Diabetes Mellitus Participants With Vascular Disease (VERTIS CV) studies may be the potential sources of heterogeneity. Similarly, for the decreased risk of renal outcome in patients with eGFR >90 mL/min/1.73 m², the heterogeneity comes from the EXSCEL study. However, due to the limited number of studies included in the analysis, we were unable to conduct a meta-regression analysis to investigate the exact sources of heterogeneity further. The number of participants with events of interest and the number of all participants at risk contributing to the network meta-analysis in each category of baseline renal function (i.e., eGFR) and RRs for the comparison of SGLT-2is and GLP-1 RAs with placebo and against each other are presented in Table 2.

GLP-1RA and SGLT-2i have been proven to reduce the risk of cardiovascular and renal outcomes in patients with T2D, and subgroup analyses of multiple RCTs have shown that their effects are not influenced by baseline eGFR levels (37,38). However, the difference in benefits between SGLT-2i and GLP1-RA in patients with different renal functions remains unclear.

Our results indicate that within the eGFR 30–90 mL/min/1.73 m² range, SGLT-2i demonstrated more pronounced efficacy in reducing the risk of HHF compared with GLP-1RA, suggesting a consistent benefit of SGLT-2i across a broader eGFR spectrum for reducing HHF risk, which could be attributed to the mild natriuretic and diuretic effects of SGLT-2i. Multiple RCTs have shown that in patients with HF with reduced ejection fraction, SGLT-2i was associated with a lower risk of cardiovascular death or HHF (39,40), and SGLT-2is now are recommended as a first-line therapy for patients with HF with reduced ejection fraction (41). Apart from providing indirect protection to the heart by lowering blood pressure and blood glucose levels, SGLT-2i also inhibited the activity of Na⁺/H⁺ exchange in the heart, decreased intracellular calcium and sodium levels, and simultaneously enhanced myocardial energy metabolism and increased myocardial oxygen supply, thereby inhibiting myocardial fibrosis and reversing myocardial remodeling. Furthermore, SGLT-2i can improve cardiovascular function by reducing cardiac load through its diuretic effect and mitigating oxidative stress and endothelial cell inflammation (42–44). However, no difference was observed in individuals with eGFR >90 mL/min/1.73 m², which likely is due to the weaker natriuretic and diuretic effects of SGLT-2i in individuals with better renal function. Insufficient data for patients with eGFR <30 mL/min/1.73 m² were available for comparison analysis, due to the limited inclusion of such patients in relevant RCTs.

Within the eGFR >90 mL/min/1.73 m² level, SGLT-2i therapy was more pronounced in reducing the risk of renal outcome. A pooled analysis comparing SGLT-2i and GLP-1RA therapies in patients with diabetes with respect to the difference in composite renal outcome and MACE found that the protective effect of SGLT-2i on composite renal events was 28 times that of GLP-1RA (45). Our research suggests the difference in renal protection between these two drugs may be more significant when eGFR >90 mL/min/1.73 m². This could be attributed to the unique mechanism of SGLT-2i, which blocks glucose reabsorption in proximal tubular cells, effectively limiting the excessive accumulation of glucose in the body and subsequently mitigating the impact of glucose toxicity on the organism. This process may contribute to reduced oxidative stress and exert an anti-inflammatory effect (46). Also, SGLT-2i blocks sodium reabsorption in the proximal tubules, leading to afferent arteriolar vasoconstriction and reduced intraglomerular pressure through physiological mechanisms such as tubuloglomerular feedback (47,48).

Within the eGFR range >90 mL/min/1.73 m², GLP-1RA is superior to SGLT-2i in reducing the risk of ACD. Research indicates that, in addition to the indirect cardiovascular protective effects through blood glucose reduction and weight loss, GLP-1RA also modulated blood pressure and lipid levels, reduced reactive oxygen species and inflammation (49,50), and inhibited myocardial fibrosis. GLP-1RAs also mediate the release of nitric oxide from endothelial cells, thereby improving endothelial function (51,52). GLP-1RA exhibited cardioprotective effects through these mechanisms, although different GLP-1RA drugs may vary in their cardiovascular benefits (53).

Our study had several limitations. First, some data of our study come from subgroup analyses of RCTs. Second, the heterogeneity of some outcomes had a P value <0.05, so the results should be interpreted cautiously. Third, the level of albuminuria is also an important indicator of baseline renal function, but our study did not explore renal outcomes with respect to albuminuria, which may lead to bias. More studies, including head-to-head comparisons, are needed to explore the impact of SGLT-2i and GLP-1RA on patients with T2D with different baseline renal functions.

SGLT-2i and GLP-1RA have different effects on the protection of cardiovascular and renal functions in patients with T2D with various baseline renal function levels. Specifically, SGLT-2i had notable advantages in decreasing the risk of HHF and renal outcome of patients with eGFR 30–90 mL/min/1.73 m². Conversely, GLP-1RA demonstrated a potential to reduce the risk of ACD in patients with eGFR >90 mL/min/1.73 m². Therefore, individualized considerations should be made based on the patient's specific kidney function status and risk factors when selecting treatment medications.

Funding. This study was supported by the Key Projects of Jiangxi Natural Science Foundation (grant 20224ACB206008), the Key Clinical Research Project of the Second Affiliated Hospital of Nanchang University (grant 2022efyB01), the “Thousand Talents Plan” project of introducing and training high-level talents of innovation and entrepreneurship in Jiangxi Province (grant JXSQ2023201030), and the Jiangxi Province Key Laboratory of Molecular Medicine (grant 2024SSY06231).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. Y.W. and G.X. conceptualized the study and its design. Y.W. and C.X. conducted the literature search and data extraction. Y.W., M.L., and C.X. analyzed the data. Y.W. wrote the manuscript. G.X. edited the manuscript. All authors read and agreed to submit the manuscript to this journal.