Comparative Effectiveness of Empagliflozin and Dapagliflozin in Chinese Patients with Type 2 Diabetes Mellitus

Abstract

Background:

Sodium-glucose cotransporter-2 (SGLT2) inhibitors have become cornerstone drugs for the treatment of T2DM. Comprehensive direct comparisons between the effects of different SGLT2 inhibitors are rare in Chinese studies. We aimed to compare the effects of empaglifllozin (EMPA) and dapaglifllozin (DAPA) in Chinese patients with T2DM.

Methods:

The study used real-world data for retrospective analysis. The hospitalized patients treated with EMPA or DAPA in our hospital between December 1, 2022 and October 31, 2024 were screened. The matching of 1:1 propensity score (PS) balanced the baseline of EMPA and DAPA groups. To compare the effects of EMPA and DAPA in T2DM patients.

Results:

The study retrospectively enrolled 5171 patients and finally screened 199 patients (EMPA group) and 179 patients (DAPA group). After 1:1 PS matching, 124 pairs were found successfully. The changes of HbAlc, FBG, HbAl, eAG, TG, TC, HDL-c, LDL-c, AST, ALT, TBiL, DBiL, eGFR and BUN, or the control rates of HbAlc < 7%, FBG < 7 mmol/L, FBG < 6.5 mmol/L, TG, TC, HDL-c, LDL-c, AST, ALT, eGFR, BUN between EMPA and DAPA groups were not significantly different (P > 0.05). There was no significant difference in the incidence of CHD, DN, DR, DPVD, DPN, thyroid nodule, osteoporosis, urinary tract infection, chronic gastritis and oral inflammatory diseases (P > 0.05). The DAPA group showed a significantly lower incidence of MAFLD compared to the EMPA group during the observation period (P < 0.01).

Conclusions:

DAPA and EMPA showed parallel effects on glycolipid metabolism, liver and kidney function.

Highlights

1. A 1:1 propensity score was applied to balance the baseline information of patients.

2. The amount of data is relatively large, and the real world treatment data of 248 patients were selected out of 5131 patients.

3. In Chinese T2DM population, it was found that empagliflozin and dapagliflozin, as additional drugs, had parallel effects in regulating glycolipid liver and kidney indexes and the incidence of diabetes-related diseases, while dapagliflozin had a significantly lower incidence of fatty liver than empagliflozin.

Introduction

Type 2 diabetes mellitus (T2DM) is a common chronic disease characterized by insulin resistance or relatively insufficient secretion of insulin, that can lead to progressive impairment of blood vessel and tissue function throughout the body, and its prevalence is increasing at an alarming rate worldwide. ¹ Due to the diversity of its complications and the severity of its consequences, effective treatment of T2DM is critical to maintaining and improving the quality of life for affected patients. However, the guidance for selecting specific agent from the categories of oral agents that can be effectively used to treat T2DM is limited, especially given the wide range and specific clinical needs, making the selection of therapeutic schedules still challenging. ²

SGLT2 inhibitors are oral hypoglycemic drugs which have attracted high attention and are at the forefront of research in recent years. They can inhibit the reabsorption of glucose by the kidney, reduce the threshold of renal glucose, and thus promote the excretion of urine sugar. ³ SGLT2 inhibitors are also believed to provide benefits due to systemic as well as direct effects, including regulation of glycosuria and natriuresis, reducing oxidative stress, inflammation and mitochondrial dysfunction, rather than relying on their direct glucose-lowering effects. ⁴ Currently, SGLT2 inhibitors listed in China include Empaglifllozin (EMPA), Dapaglifllozin (DAPA), Canagliflozin, Ertugliflozin and Henagliflozin. EMPA and DAPA are the most commonly used SGLT2 inhibitors in the world. ⁵

Systematic evaluation showed that EMPA and DAPA for treating T2DM are effective in improving glycaemic control. ⁶ A cohort study conducted using data from national registries in Sweden, Denmark and Norway, which included 141 065 new EMPA users and 58 306 new DAPA users, showed that the use of EMPA and DAPA was similarly associated with the risk of cardiovascular and renal outcomes, mortality and diabetic ketoacidosis. ⁷ However, The results of the observational study from Boston, USA showed that, when adjusted for matching potential confounding factors, individuals who started using EMPA were significantly less likely to experience death or hospitalization within 1 year after initiation of an SGLT2 inhibitor compared to individuals who started using DAPA, and the risk of these outcomes held true when assessed separately. ⁸ A prospective observational study in Republic of Korea showed that after 52 weeks of treatment with EMPA and DAPA in T2DM patients inadequately controlled by glimepiide, metformin and dipeptidyl peptide-4 inhibitors, EMPA was superior to DAPA in glycemic control and lipid profile regulation. ⁹ While Another large real-world cohort study based on the population of Republic of Korea has shown that DAPA may be associated with a lower risk of hospitalization or death from heart failure and cardiovascular death than EMPA in Asian patients with T2DM, ⁴ and also have scholars believe that the reduction in the incidence of atrial fibrillation shown in the meta-analysis is mainly due to the use of DAPA. ¹⁰ Clinical trials of the two drugs have shown inconsistencies in disease outcomes among different populations, and it is not entirely clear whether there are drug-specific differences in this class. ⁴ Therefore, there are still many uncertainties about the effectiveness and safety of EMPA and DAPA, with implications for drug use and guideline development. A network meta-analysis of randomized controlled trials also have revealed that the differences among SGLT2 inhibitors in terms of reducing cardiovascular and renal outcomes are not consistent. Further studies are needed to better understand the mechanism and clinical efficacy of each SGLT2 inhibitor in specific populations. ¹¹

EMPA was approved by the Food and Drug Administration in June 2023 for the control of blood glucose in the pediatric population and adolescents with T2DM aged 10 years and older, ¹² but has not yet been approved in China. The US Food and Drug Administration has approved EMPA to reduce cardiovascular death as an indication for T2DM, while DAPA does not. ⁶ There may also have pharmacological differences between drugs, such as receptor selectivity that can lead to different effects on results. ¹³ Although China has a large population of T2DM patients, comprehensive direct comparisons between the effects of different SGLT2 inhibitors are rare in Chinese studies. Our study aims to compare whether EMPA and DAPA in Chinese patients with T2DM, have differences in glycolipid metabolism, liver and kidney function related indicators and safety.

Materials and Methods

Patients

Inpatients treated with EMPA or DAPA in Shijiazhuang Second Hospital (Shijiazhuang Diabetes Hospital, a comprehensive hospital integrating medical treatment, teaching and research, with an annual outpatient volume of 400 000 and 1000 beds) from December 1, 2022 to October 31, 2024 were retrospectively selected. Inclusion criteria: (1) Patients diagnosed with T2DM, regardless of the primary reason for hospital admission; (2) Complete clinical index data; (3) Age ≥18 years old; (4) At least two visits were recorded during the study period, and the interval between the two visits was ≥3 months. It is not limited whether other hypoglycemic drugs are used at the same time when EMPA or DAPA is used. Exclusion criteria: (1) lack of glycosylated hemoglobin test information; (2) lack of blood lipid test information; (3) lack of complete medication information; (4) patients with malignant tumors; (5) patients with liver insufficiency. Patients with EMPA or DAPA were divided into two groups, and the recommended starting doses for EMPA and DAPA were 10 mg once daily and 5 mg once daily, respectively. This study was approved by the medical ethics committee of our hospital, and informed consent of the patients and their families.

Data Collection

The case number and the time of patients using EMPA and DAPA were found through the prescribing system of our hospital pharmacy. General data of enrolled patients, including gender, age, diagnosis, diabetes course, systolic blood pressure (SBP), Diastolic blood pressure (DBP) and complete medication information, were collected through a unique case number search on MandalaT Software Corporation. Laboratory test information was searched and collected by the inpatient’s unique case number in the hospital information system (rmLis and Lis6.0)

Statistical Analysis

The SPSS 22.0 statistical software were used to analyze all the data. Numeric data values were expressed as mean ± SD and were compared by the t test. Categorical variables were expressed as cases and percentages and were compared by the Chi-square test. To adjust for confounding, we set the caliper value to 0.01 according to the baseline of 1:1 caliper matching to balance different groups using SPSS 22.0 software. Bilateral P < 0.05 indicates statistical significance. Bilateral P < 0.01 were considered to have high statistical significance. The outcomes: (1)The changes of HbAlc, FBG, HbAl, eAG, TG, TC, HDL-c, LDL-c, AST, ALT, TBiL, DBiL, eGFR and BUN; (2) The control rates of HbAlc < 7%, FBG < 7 mmol/L, FBG < 6.5 mmol/L, TG, TC, HDL-c, LDL-c, AST, ALT, eGFR, BUN; (3) The incidence rates of Coronary Heart Disease (CHD), Diabetic Nephropathy (DN), Diabetic Retinopathy (DR), Diabetic Peripheral Vascular Disease (DPVD), Diabetic Peripheral Neuropathy (DPN), thyroid nodule, osteoporosis, urinary tract infection, chronic gastritis and oral inflammatory diseases.

Results

Baseline Characteristics of the Included Population

In total, 5171 patients were treated with EMPA or DAPA in our hospital from December 1, 2022 and October 31, 2024. A total of 709 patients were hospitalized ≥2 times, including 372 patients treated with EMPA and 337 patients treated with DAPA. Patients treated with EMPA excluded five patients with malignant tumors and postoperative malignant tumor, 39 patients with less than 3 months in hospital and 129 patients with lack of complete medication and examination information, and 199 patients who met the criteria were finally included. The patients treated with DAPA excluded six patients with malignant tumor and postoperative malignant tumor, one patient with liver dysfunction, 393 patients with less than 3 months in hospital and 118 patients lacking complete medication and examination information, finally, 179 patients who met the criteria were included. The flow chart in Figure 1 illustrates the screening process for case data. After 1:1 PS matching, there were 124 PS matched pairs were successfully found between EMPA and DAPA. The baseline characteristic data of the included patients are summarized in Table 1.

Figure 1.

Case screening process and results.

Table 1.

Comparison of Patients’ Baseline Before and After 1:1 Propensity Score.

PS	Group	n	Gender		Age (year)	During (month)	SBP (mmHg)	DBP (mmHg)	DM (year)
			Male	Female
Before PS	EMPA	199	138	61	60.23 ± 12.34	9.45 ± 3.67	127.69 ± 10.84	76.62 ± 6.24	9.84 ± 6.90
	DAPA	179	134	45	60.46 ± 13.01	8.78 ± 3.88	130.28 ± 15.13	78.06 ± 9.85	11.69 ± 7.64*
After PS	EMPA	124	87	37	60.34 ± 11.96	9.08 ± 3.57	128.18 ± 11.06	76.49 ± 6.04	10.33 ± 6.79
	DAPA	124	89	35	61.46 ± 13.28	9.17 ± 3.94	127.66 ± 13.73	76.41 ± 8.71	10.62 ± 6.83

Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BUN, blood urea nitrogen; DAPA, dapaglifllozin; DBiL, direct bilirubin; DBP, diastolic blood pressure; DM, diabetes mellitus; eGFR, estimated glomerular filtration rate; EMPA, empaglifllozin; FBG, fasting blood glucose; HbAlc, hemoglobin Alc; PLT, platelet count; PS, propensity score; SBP, systolic blood pressure; TBiL, total bilirubin; TC, total cholesterol; TG, triglycerides; WBC, white blood cell count.

^*P < 0.05, **P < 0.01 versus EMPA group.

Comparison of Blood Glucose Related Indicators Between Groups

There were no significant differences in terms of the changes in HbAlc, FBG, HbAl and eAG, or in the control rates for HbAlc < 7%, FBG < 7 mmol/L and FBG < 6.5 mmol/L between the two groups (P > 0.05) (Table 2).

Table 2.

Comparison of the Change and the Control Rate of Blood Glucose Related Indexes Between Groups.

Group	ΔHbAlc (%)	ΔFBG (mmol/L)	ΔHbAl (%)	ΔeAG (mmol/L)	HbAlc < 7 (%) control rate	FBG < 7 (mmol/L) control rate	FBG < 6.5 (mmol/L) control rate
EMPA	−0.20 ± 1.86	−0.66 ± 4.51	−0.24 ± 1.98	−0.41 ± 3.05	20.00 (17/85)	33.33 (30/90)	17.82 (18/101)
DAPA	−0.42 ± 1.76	−1.40 ± 3.98	−0.46 ± 1.92	−0.67 ± 2.85	17.20 (16/93)	36.17 (34/94)	27.72 (28/101)
χ2/t	0.985	1.358	0.879	0.699	0.158	0.079	1.774
P	0.326	0.176	0.380	0.485	0.691	0.779	0.183

Abbreviations: DAPA, dapaglifllozin; eAG, estimated average glucose; EMPA, empaglifllozin; FBG, fasting blood glucose; HbAl, hemoglobin Al; HbAlc, hemoglobin Alc.

Δ refers to the change after and before treatment; control rate refers to the percentage of patients who met the standard after treatment compared to those who did not meet the standard before treatment.

Comparison of Lipid-Related Indexes Between Groups

The differences in the changes of TG, TC, HDL-c and LDL-c between the two groups were not statistically significant, and there was no statistically significant difference in the control rates of TG, TC, HDL-c and LDL-c between the two groups (P > 0.05) (Table 3).

Table 3.

Comparison of the Change and the Control Rate of Blood Lipid Related Indexes Between Groups.

Group	ΔTG (mmol/L)	ΔTC (mmol/L)	ΔHDL-c (mmol/L)	ΔLDL-c (mmol/L)	TG (0.5-1.7) control rate	TC (2.6-5.7) control rate	HDL-c (0.8-2.8) control rate	LDL-c (0-3.36) control rate
EMPA	0.18 ± 2.32	−0.36 ± 1.44	0.03 ± 0.29	−0.24 ± 0.94	24.07 (13/54)	72.41 (21/29)	75.00 (12/16)	52.17 (12/23)
DAPA	0.13 ± 1.77	−0.22 ± 1.24	0.05 ± 0.29	−0.25 ± 0.85	37.29 (22/59)	76.19 (16/21)	64.29 (9/14)	82.35 (14/17)
χ2/t	0.164	−0.815	−0.487	0.071	1.222	0.013	0.072	0.814
P	0.870	0.416	0.627	0.943	0.269	0.908	0.788	0.367

Abbreviations: DAPA, dapaglifllozin; EMPA, empaglifllozin; HDL-c, highdensity lipoprotein cholesterol; LDL-c, lowdensity lipoprotein cholesterol; TC, total cholesterol; TG, triglycerides.

Δ refers to the change after and before treatment; control rate refers to the percentage of patients who met the standard after treatment compared to those who did not meet the standard before treatment.

Comparison of Liver and Kidney Function Indexes Between Groups

The changes of aspartate aminotransferase (AST), alanine transaminase (ALT), total bilirubin (TBiL), DBiL, eGFR and BUN showed no significant difference between the two groups. The control rates of AST, ALT, eGFR and BUN did not differ significantly between the two groups (P > 0.05) (Table 4).

Table 4.

Comparison of the Change and the Control Rate of Liver and Kidney Function Indexes Between Groups.

Group	ΔAST (U/L)	ΔALT (U/L)	ΔTBiL (μmol/L)	ΔDBiL (μmol/L)	ΔeGFR (mL/min/1.73 m2)	ΔBUN (mmol/L)	AST (13-40)control rate	ALT (7-50)control rate	eGFR≥90 control rate	BUN (3.1-9.5) control rate
EMPA	−1.78 ± 10.67	−2.31 ± 15.81	−0.36 ± 4.79	−0.57 ± 2.46	−5.42 ± 31.62	0.20 ± 2.05	88.24 (15/17)	87.50 (7/8)	38.89 (7/18)	69.23 (9/13)
DAPA	−1.39 ± 28.06	−3.18 ± 22.67	−0.69 ± 5.17	−0.75 ± 2.67	0.25 ± 42.48	0.32 ± 2.53	66.67 (16/24)	88.89 (8/9)	23.08 (6/26)	85.71 (6/7)
χ2/t	−0.145	0.350	0.520	0.558	−1.193	−0.429	0.343	0.000	0.682	0.092
P	0.885	0.727	0.603	0.577	0.234	0.668	0.558	1.000	0.409	1.000

Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BUN, blood urea nitrogen; DAPA, dapaglifllozin; DBiL, direct bilirubin; eGFR, estimated glomerular filtration rate; EMPA, empaglifllozin; TBiL, total bilirubin.

Δ refers to the change after and before treatment; control rate refers to the percentage of patients who met the standard after treatment compared to those who did not meet the standard before treatment.

The Incidence of Diabetes Complications and Other Diseases

The incidences of Coronary Heart Disease (CHD), Diabetic Nephropathy (DN), Diabetic Retinopathy (DR), Diabetic Peripheral Vascular Disease (DPVD), Diabetic Peripheral Neuropathy (DPN), Thyroid Nodule, Osteoporosis, Urinary Tract Infection, Chronic Gastritis and Oral inflammatory diseases were not significantly different between the two groups (all P > 0.05). However, the DAPA group showed a significantly lower incidence of MAFLD compared to the EMPA group during the observation period (P < 0.01) (Table 5).

Table 5.

Comparison Between Groups Incorporating Other Disease Conditions.

Group	CHD (%)	DN (%)	DR (%)	DPVD (%)	DPN (%)	MAFLD (%)	TN (%)	OP (%)	UTI (%)	CG (%)	OID (%)
EMPA	13.43 (9/67)	8.45 (6/71)	10.14 (7/69)	26.00 (13/50)	30.00 (9/30)	19.54 (17/87)	17.86 (15/84)	5.66 (6/106)	2.50 (3/120)	9.10 (10/110)	2.50 (3/120)
DAPA	10.00 (6/60)	9.52 (8/84)	11.69 (9/77)	11.69 (9/77)	22.22 (4/18)	5.6** (5/89)	12.12 (12/99)	0.91 (1/110)	2.46 (3/122)	6.06 (6/99)	1.72 (2/116)

Abbreviations: CG, chronic gastritis; CHD, coronary heart disease; DAPA, dapaglifllozin; DN, diabetic nephropathy; DPN, diabetic peripheral neuropathy; DPVD, diabetic peripheral vascular disease; DR, Diabetic retinopathy; EMPA, empaglifllozin; MAFLD, metabolic associated fatty liver disease; OID, oral inflammatory diseases; OP, osteoporosis; TN, thyroid nodule; UTI, urinary tract infection.

Incidence rate refers to the percentage of patients with disease after treatment as opposed to those without disease before treatment.

^**P < 0.01 versus EMPA group.

Discussion

This study was conducted to determine whether there are differences in the efficacy and safety of EMPA versus DAPA in Chinese patients with T2DM. We used real-world data from 5171 patients to compare glycolipid and liver and kidney markers in 124 Chinese EMPA and 124 Chinese DAPA users based on baseline characteristics balanced by a 1:1 PS with a median follow-up of 9 months. The results demonstrated that EMPA and DAPA had similar change and control rate in glycolipid, liver and kidney function. Similarly, there was no significant difference in the morbidity of most diabetes-related diseases during treatment, while the incidence rate of MAFLD was lower in DAPA than in EMPA in Chinese patients with T2DM.

Diabetes urgently needs prevention, diagnosis and treatment, as an incurable and heterogeneous disease, it has different manifestations and related complications in different individuals, and even different responses to anti-sugar drugs.^14,15 Thus there is widespread agreement that medicine needs to move to a platform for individualized clinical management. At present, It is clinical common to encounter patients who insist on using alternative oral antidiabetic agents instead of injection due to various reasons such as acupuncture pain, fear of hypoglycemia, sudden weight change, ⁹ and cumbersome procedures. Multiple clinical studies have shown that the glucose-promoting effect of SGLT2 inhibitors, as an oral hypoglycemic agent, can significantly reduce diabetic burden, subsequently approved for the treatment of T2DM, ¹⁶ and can be used alone or in combination with other antidiabetic drugs. Therefore, this study tended to compare the effects of individual SGLT2 inhibitors, EMPA versus DAPA specifically in Chinese patients, instead of expending much effort to broadly confirm the overall role of these drugs in diabetes management.

A number of studies have shown that EMPA reduced blood sugar and HbA1c in T2DM patients when comparison with DAPA. ¹⁷ However, our study showed that EMPA has the same hypoglycemic effect as DAPA. This may be due to that the time of our experiment was chosen at any time during the administration, not necessarily the first administration. Currently, most patients will not change medication if there is no clinical adverse reaction or unsatisfactory blood glucose control, so observation at any time may be more representative. The first meta-analysis among 7578 patients with T2DM to investigate the relationship between SGLT2 inhibitors and dyslipidemia showed that SGLT2 inhibitors significantly increased HDL-c, TC and LDL-c, while reducing TG, compared with placebo or other oral hypoglycemic agents. ¹⁸ Our study complements that there was no significant difference in the effects of EMPA and DAPA on lipid levels. The results of meta-analysis in Asian patients with T2DM showed that SGLT2 inhibitors significantly decreased AST and ALT, while total bilirubin (TBiL) and eGFR levels were largely unchanged. ¹⁹ Our study complements that the change of AST, ALT, TBiL, DBiL, eGFR and BUN and the control rate of AST, ALT, eGFR and BUN were no statistical difference between EMPA and DAPA. In addition, The literature reports that the safety of SGLT2 inhibitors is similar to that of placebo or other glucose-lowering agents in T2DM patients. ²⁰ Both EMPA and DAPA has good efficacy and safety. ¹⁷ Our study complemented the evidence that the incidence of CHD, DN, DR, DPVD, DPN, Thyroid Nodule, Osteoporosis, Urinary Tract Infection, Chronic Gastritis and Oral inflammatory diseases had no significant difference between EMPA and DAPA. In summary, EMP and DAPA may have a parallel class of effects on glycolipid metabolism and the incidence of most diabetes-related diseases. Indeed, it has been reported that EMPA and DAPA have the same benefits demonstrated for DM, heart failure, myocardial infarction and renal failure, Therefore, they believe that SGLT2 inhibitors have a clearly favorable class effect, and the differences between individual drugs mainly apply to personalized care and minimize side effects. ²¹

To date, there are no specific drug treatment guidelines for MAFLD in T2DM patients. Previous studies have shown that SGLT2 inhibitors can improve hepatic steatosis, but the focus has mainly been on canagliflozin, while the effects of EMPA and DAPA remain largely speculative as there are few positive comparisons. ²² A meta-analysis of Asian patients with T2DM has shown that SGLT2 inhibitors, including EMPA, DAPA, Canagliflozin, Ipragliflozin and Luseogliflozin, are an effective treatment for improving MAFLD. ¹⁹ In T2DM patients with MAFLD, initiating SGLT2 inhibitors (EMPA and DAPA) to reduce mean ALT levels and may also reduce liver-related morbidity and mortality from MAFLD. Study have shown that DAPA administration can improve liver gluconeogenesis in normal mice. ²³ Another study reported that EMPA can reduce liver fat in T2DM patients with MAFLD. ²⁴ Pioglitazone has been proven to improve hepatic hyphology in diabetes patients with MAFLD and NASH, while compared with pioglitazone, DAPA significantly reduced the fatty liver index in MAFLD patients, and the changes of fatty liver index were positively correlated with the changes of insulin and HbA1c levels.^19,25 Another study have shown that EMPA also has a better ameliorative effect on liver steatosis than pioglitazone. ²⁶ Our study indicates that the incidence of MAFLD among Chinese T2DM patients in the DAPA group is significantly lower than that in the EMPA group. DAPA but not EMPA significantly increased hippocampal mitochondria-associated protein expression, interfering with mitochondrial respiration. ¹⁶ Compared with EMPA, DAPA had a stronger effect on the inflammatory response, DAPA also altered the energy metabolism of Human umbilical vein endothelial cells, and showed an anti-fibrotic effect by inhibiting migration and proliferation of human cardiac fibroblasts. ¹⁶ EMPA but not DAPA has been described as having specific effects on mitochondrial respiratory chain complex I and II in a mouse model. ²⁷ In conclusion, DAPA may have a lower incidence of MAFLD due to its lower inflammatory, additional regulation of mitochondrial respiratory and anti-fibrosis effects compared with EMPA. An Italian clinical study also provided evidence that the treatment of SGLT2 inhibitors in T2DM patients with MAFLD was associated with improvements in markers of liver steatosis and fibrosis, as well as redox status and circulatory pro-inflammatory, rather than blood glucose control. ²⁸ Although, the bioavailability of EMPA and DAPA is 90% and 78%, and the selectivity is 1200× and 2500×, respectively. ²¹

A comparative evaluation of the long-term efficacy and safety of EMPA and DAPA as adjuvant hypoglycemic agents in T2DM patients, also from a Chinese population, showed that after 36 months of use of EMPA or DAPA, there was a synergistic effect in hypoglycemic effect, stabilizing blood pressure and reducing body weight. However, No significant difference in efficacy of two agents was found even at different doses. ²⁹ As complementary data for Chinese patients with T2DM taking EMPA and DAPA, Our study indicates that EMPA and DAPA showed similar effects on glycolipid metabolism, liver and kidney function and the incidence of most diabetes-related diseases, however, DAPA significantly reduced the incidence of MAFLD compared to EMPA. Due to EMPA and DAPA have a high incidence of diabetes-related diseases in clinical use, lifestyle education should also be paid attention to during treatment, and the clinical use of DAPA may be considered in people with T2DM who are at high risk for MAFLD. However, the limitations of our study are as follows: (1) The retrospective design of the study did not include records of patients’ lifestyle habits, and valid evaluation indicators may have been missed, and also cannot assess adherence or actual medication exposure. (2) Adverse effects were not evaluated in this study. (3) The source of study cases was limited to limited numbers of a medical center. (4) The duration of the patient’s medication was unknown. (5) It was unknown whether any patients have taken the maximum dose of both medications. (6) The impact of other antidiabetic drugs taken concurrently by the patients on the results was unknown. Meanwhile, more evidence from multicenter, large-scale randomized controlled trials and basic biochemical experiments is needed.

Conclusions

Our study is the first retrospective study to evaluate the effects of EMPA and DAPA on glycolipid liver and kidney markers, and the incidence rate of diabetes-related disease over any period of time based on 1:1 PS matching. According to our study, EMPA and DAPA showed similar effects on marks of glycolipid metabolism, liver and kidney function, and the incidence of MAFLD in the DAPA group was significantly lower than that in the EMPA group, suggesting that SGLT2 inhibitors may should be distinguished in the individualized treatment of Chinese T2DM patients with a tendency to fatty liver.