GRAPHICAL ABSTRACT

Highlights

• This NMA assessed HbA1c-lowering efficacies of non-insulin therapies in early T2DM.

• Metformin plus GLP-1RA or DPP-4i showed the greatest glycemic control efficacy.

• Initial combination therapies outperformed monotherapies for HbA1c reduction.

• These findings support early dual therapy to optimize glycemic control in T2DM.

INTRODUCTION

Type 2 diabetes mellitus (T2DM) is a prevalent chronic metabolic disorder characterized by progressive insulin resistance and β-cell dysfunction, leading to sustained hyperglycemia [1]. Early and effective glycemic control is critical for preventing long-term complications, including cardiovascular disease and microvascular complications, all of which significantly impair the quality of life and increase mortality rates [2]. As the pharmacological landscape for managing T2DM has expanded, the optimal initial therapy remains a topic of considerable debate. Traditionally, metformin has been widely recommended as a first-line therapy for patients with newly diagnosed T2DM [3]. The development of more effective and safer glucose-lowering agents with diverse characteristics has allowed for a more flexible, person-centered approach to drug selection [4,5]. As such, treatment decisions can consider various demographic, medical, and socioeconomic factors specific to each patient. In particular, thiazolidinedione (TZD), dipeptidyl peptidase-4 inhibitor (DPP4i), sodium-glucose cotransporter-2 inhibitor (SGLT2i), and glucagon-like peptide-1 receptor agonist (GLP-1RA) have demonstrated a low risk of hypoglycemia [4,5]. These agents have also been shown to reduce treatment failure rates when used in combination while maintaining a favorable safety profile [6]. Consequently, recent guidelines advocate for the early initiation of combination therapy for mild hyperglycemia, reflecting a shift toward more aggressive, tailored treatment strategies aimed at preventing disease progression and improving patient outcomes [4,5].

Although multiple network meta-analyses (NMA) have evaluated diabetes medications [7,8], few have focused specifically on early T2DM or comprehensively compared monotherapy and combination therapy, leaving an important knowledge gap for this population. Our study addresses this gap by being the first to conduct an NMA comparing the efficacy and safety of various therapeutic regimens, including monotherapy and combination therapy, in people with early T2DM. This study aimed to compare the glucose-lowering efficacy of various non-insulin diabetes therapies and their impact on body weight, blood pressure, lipid profiles, and incidence of adverse events through an NMA of randomized controlled trials (RCTs). Our findings will guide clinicians in selecting the most appropriate initial treatment strategy for people with early T2DM.

METHODS

Study design

This study adhered to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines and was registered with PROSPERO (CRD42024511058), ensuring transparency and conformance with established standards for systematic reviews.

The eligibility criteria for selecting studies were specified using the Population, Intervention, Comparison, Outcomes, Time, Study design framework, as detailed in Supplementary Table 1. This framework was designed to determine the most effective pharmacological regimen for adults with early T2DM. The studies included adults newly diagnosed with T2DM or those with previously diagnosed T2DM who had never received pharmacological treatment. Studies were excluded if the mean duration of diabetes exceeded 5 years or if the mean glycosylated hemoglobin (HbA1c) level was >8.5%. These criteria were selected to ensure a focus on patients in the early period of diabetes, excluding cases where severe hyperglycemia might require insulin treatment [4,5]. The selected interventions were non-insulin antidiabetic agents currently used in Korea as of 2024 or expected to be available soon, including metformin, sulfonylurea (SU), TZD, DPP4i, SGLT2i, and GLP-1RA. Only RCTs with an intervention period of at least 24 weeks, during which the intervention was sustained without any crossover, were included. The primary outcomes were changes in HbA1c or odds of achieving target HbA1c (<7.0% or <6.5%) after 24 weeks (6 months). Secondary outcomes included changes in fasting blood glucose (FBG), body weight, blood pressure, total cholesterol, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), triglycerides, and odds for the incidence of adverse events.

Search strategy and literature selection

The search strategy was refined through multiple iterations, incorporating both controlled vocabulary (MeSH terms) and free text to ensure comprehensive coverage of relevant studies. The final search strategy (Supplementary Table 2) was developed in consultation with a professional librarian and was executed on January 31, 2024, across PubMed, Embase, Cochrane, and KoreaMed databases. A total of 69,706 documents were retrieved from the four databases. After removing duplicates by screening the title twice, 44,721 documents were included. From February 1 to February 28, 2024, six diabetes experts (J.H.C., Y.S.Y., S.H.M., B.K.K., J.S.P., and S.Y.R.) from the Committee of Clinical Practice Guidelines of the Korean Diabetes Association screened the titles of these documents, resulting in the selection of 21,214 documents. Abstract screening further reduced the number to 529 documents, which were selected for full-text review. Two independent diabetes experts (J.H.C. and Y.S.Y.) thoroughly read the full texts and selected the final documents. Discrepancies were resolved through discussions between the two experts, and if needed, a consensus meeting with all the six experts and the corresponding author (M.K.M.) was held to resolve any disagreements.

Data extraction and risk of bias assessment for individual studies

Data extraction was initially performed by S.H.M., B.K.K., and Y.S.Y. Then, J.H.C. conducted a secondary review to verify the accuracy and completeness of the extracted data. The extracted data included both continuous and dichotomous variables relevant to the study outcomes. Continuous variables included HbA1c (%), FBG (mg/dL), body weight (kg), blood pressure (mm Hg), and lipid profiles (total cholesterol, HDL-C, LDL-C, and triglycerides) (mg/dL) after 24 weeks. For continuous variables, data were extracted for the mean difference with standard error (SE) of changes from baseline to 6 months. Dichotomous variables included the proportion of patients achieving target HbA1c (<7.0% or <6.5%) at 24 weeks and the incidence of adverse events. The data for dichotomous variables were reported as odds ratio (OR) with SE for each outcome. Data discrepancies were resolved by discussion among the initial extractors or through consultation with another reviewer, if necessary. The risk of bias in the included studies was assessed independently by two reviewers (J.H.C. and Y.S.Y.) using the Cochrane risk of bias 2.0 tool. This tool was selected for its ability to thoroughly evaluate key areas, including the randomization process, deviations from intended interventions, missing outcome data, measurement of outcomes, and selection of reported results. Any discrepancies in the risk of bias assessments were resolved through discussion or, when necessary, by consulting a third reviewer (M.K.M.) to reach consensus.

Data synthesis and analysis

The NMA was conducted by H.J.K. using STATA version 16.1 (StataCorp LP, College Station, TX, USA) to compare the efficacy and safety of the included interventions. A random-effects model was applied to account for potential heterogeneity across studies, incorporating both direct and indirect comparisons. Continuous outcomes are summarized as weighted mean difference (WMD) with 95% confidence interval (CI) and dichotomous outcomes as OR with 95% CI. Consistency between direct and indirect evidence was assessed using the node-splitting method, with a P value threshold of 0.1 used to evaluate statistical inconsistency. This threshold was chosen to account for potential complexities in the network structure and indirect comparisons in the analysis. Treatments were ranked based on the surface under the cumulative ranking (SUCRA) probabilities, estimating the likelihood of each intervention being the most effective. Additionally, a pairwise meta-analysis (PMA) was conducted using the same random-effects model in STATA to validate the NMA findings through direct comparisons between interventions. The integration of NMA and PMA allowed for cross-validation and comprehensive evaluation of therapeutic options for early-stage T2DM.

RESULTS

Sixty RCTs were included in the final analysis after a thorough full-text review based on predefined inclusion and exclusion criteria. The selection process, including the number of articles at each stage and reasons for exclusion, is detailed in Fig. 1. The key characteristics of the included studies are summarized in Supplementary Table 3. Among these, 47 studies (78.3%) involved intervention periods of 24 weeks, while the longest intervention period extended to 60 months. The mean age of participants ranged from 41.0 to 70.9 years, and the mean body mass index ranged from 22.5 to 39.8 kg/m². The mean HbA1c ranged from 6.6% to 8.5%, and the duration of diabetes varied from new diagnoses to a maximum of 5.0 years. The risk of bias assessment results for the primary outcome, change in HbA1c after 24 weeks, are provided in Supplementary Fig. 1. Supplementary Table 4 outlines the types of interventions, number of studies, participants, and baseline mean HbA1c values for each intervention. The number of studies, direct comparisons, and participants for each outcome variable are detailed in Supplementary Table 5.

Glucose-lowering efficacy

The network geometry for the change in HbA1c after six months is illustrated in Fig. 2. The global test for inconsistency showed a P value of 0.0724, suggesting a potential inconsistency in the network (Supplementary Table 6). The outcomes of the NMA and PMA are summarized in Table 1, and the ranking of treatments based on the network rank test is provided in Table 2. Data for glucose-lowering efficacy were available in 58 out of the 60 included studies, comprising 91 direct comparisons involving 20,161 participants. All dual combination therapies were ranked higher than monotherapies. Metformin+GLP-1RA was the most effective in lowering HbA1c (SUCRA: 85.8 vs. placebo WMD: –1.50%; 95% CI, –2.04 to –0.96), followed by metformin+DPP4i (SUCRA: 85.2 vs. placebo WMD: –1.46%; 95% CI, –1.96 to –0.95). Among the monotherapies, GLP-1RA (SUCRA: 47.4 vs. placebo WMD: –1.10%; 95% CI, –1.65 to –0.55) was the most effective. Metformin ranked second in the rank test among monotherapies, but it was the only treatment among all therapies that did not show a statistically significant reduction in HbA1c compared to placebo (SUCRA: 45.4 vs. placebo WMD: 0.10%; 95% CI, –0.46 to 0.66). In contrast, on PMA, metformin significantly reduced HbA1c (vs. placebo WMD: –0.53%; 95% CI, –0.64 to –0.43).

The network geometry for the ORs of achieving target HbA1c at 24 weeks is illustrated in Supplementary Fig. 2. The global test indicated no evidence of inconsistency (P=0.9128). Data for this outcome were available from 30 of the 60 included studies, encompassing 48 direct comparisons with 15,018 participants. The detailed NMA and PMA results for this outcome are presented in Table 3, and the network rank test results are shown in Supplementary Fig. 2. Metformin+DPP4i emerged as the most effective combination (SUCRA: 89.9 vs. placebo OR: 34.31; 95% CI, 7.69 to 153.00), followed by DPP4i+TZD (SUCRA: 76.7 vs. placebo OR: 24.49; 95% CI, 5.20 to 115.44). Among monotherapies, GLP-1RA ranked the highest (SUCRA: 57.0 vs. placebo OR: 15.90; 95% CI, 3.48 to 72.69), followed by SU (SUCRA: 53.5 vs. placebo OR: 15.08; 95% CI, 3.16 to 71.97) and TZD (SUCRA: 52.7 vs. placebo OR: 14.63; 95% CI, 3.15 to 67.86). Despite minor changes in the ranking order among metformin, SU, and TZD, the overall rank test results were consistent with those observed for HbA1c reduction. Similar to the HbA1c reduction results, metformin was the only therapy with no statistically significant superiority over placebo in achieving target HbA1c (SUCRA: 38.4, OR: 0.58; 95% CI, 0.17 to 1.96). However, in PMA, metformin showed a significant improvement in achieving the HbA1c target compared to placebo (OR: 11.49; 95% CI, 4.07 to 32.26), indicating that metformin is also an effective therapy in direct comparisons.

The NMA results for FBG are presented in Supplementary Fig. 3. The global test results indicated a potential for inconsistency (P=0.0545). Overall, combination therapies generally demonstrated better efficacy than monotherapies, although the results varied across treatments. Among the combination therapies, metformin+DPP4i combination was the most effective (SUCRA: 87.7 vs. placebo WMD: –48.0 mg/dL; 95% CI, –71.7 to –24.4). Among monotherapies, SGLT2i ranked the highest (SUCRA: 60.3 vs. placebo WMD: –36.9 mg/dL; 95% CI, –54.8 to –19.1). Similar to the HbA1c reduction findings, metformin showed the smallest effect size compared to placebo (WMD: –13.0 mg/dL; 95% CI, –25.4 to –0.7).

Body weight

The NMA and PMA results for body weight are presented in Supplementary Fig. 4. The global test results showed no evidence of inconsistency (P=0.1806). The rank test results indicated that GLP-1RA (SUCRA: 89.0), SGLT2i (SUCRA: 62.7), metformin (SUCRA: 55.6), and their combination therapies ranked higher, while TZD (SUCRA: 3.0), SU (SUCRA: 18.6), DPP4i (SUCRA: 29.0), and their combinations ranked lower in terms of weight reduction. GLP-1RA showed the best weight reduction result (SUCRA: 89.0 vs. TZD WMD: –3.50 kg; 95% CI, –4.60 to –2.40), while the combination of metformin+ GLP-1RA had a similar SUCRA but showed the largest effect size (SUCRA: 88.5 vs. TZD WMD: –5.50 kg; 95% CI, –8.13 to –2.87). In the NMA, none of the drug regimens showed statistically significant weight reduction effects compared to placebo. However, in the PMA, both SGLT2i (WMD: –2.08 kg; 95% CI, –2.34 to –1.81) and GLP-1RA (WMD: –1.71 kg; 95% CI, –2.08 to –1.33) demonstrated significant weight reduction compared to placebo. In contrast, TZD (WMD: 3.34 kg; 95% CI, 1.79 to 4.89) and SU (WMD: 1.75 kg; 95% CI, 0.11 to 3.39) showed significant weight gain compared to placebo in the NMA.

Adverse events and hypoglycemia

The NMA and PMA results for any and serious adverse events are presented in Supplementary Figs. 5 and 6, respectively. The global test results showed no evidence of inconsistency for any adverse event (P=0.8471) and serious adverse events (P=0.9773). In the rank test for any adverse events, placebo (SUCRA: 91.9), GLP-1RA (SUCRA: 77.9), and DPP4i (SUCRA: 51.1) ranked the highest, indicating a lower likelihood of adverse events. Conversely, SGLT2i (SUCRA: 34.2) and SU (SUCRA: 28.3) ranked the lowest, indicating a higher likelihood of adverse events. For serious adverse events, DPP4i (SUCRA: 63.6), SU (SUCRA: 60.7), and metformin+DPP4i (SUCRA: 58.3) ranked the highest, whereas DPP4i+SGLT2i (SUCRA: 33.8) and SGLT2i (SUCRA: 21.4) ranked the lowest, indicating more events. However, no statistically significant differences were observed in the occurrence of any or serious adverse events across treatment combinations in both NMA and PMA.

The NMA and PMA results for hypoglycemia are presented in Supplementary Fig. 7. The global test results showed no evidence of inconsistency (P=0.9385). In the NMA rank test, DPP4i+SGLT2i ranked higher than placebo, suggesting a lower likelihood of hypoglycemia, but the difference was not statistically significant (SUCRA: 92.4, OR: 0.20; 95% CI, 0.01 to 5.50), and no direct comparisons were available. Following DPP4i+SGLT2i, placebo (SUCRA: 77.8), DPP4i (SUCRA: 70.9), TZD (SUCRA: 55.9), and SGLT2i (SUCRA: 52.8) ranked higher, indicating a lower hypoglycemia risk. Conversely, SU (SUCRA: 10.1) ranked the lowest, indicating the highest risk of hypoglycemia. SU was significantly associated with a higher risk of hypoglycemia compared to placebo (OR: 17.2; 95% CI, 1.4 to 220.3). In the PMA, although SU did not have direct comparisons with placebo, it showed a significantly higher risk of hypoglycemia compared to TZD (OR: 5.13; 95% CI, 3.21 to 8.18), GLP-1RA (OR: 9.10; 95% CI, 5.69 to 14.55), and metformin (OR: 4.87; 95% CI, 4.03 to 5.89). No other treatments were associated with a statistically significant increase in the risk of hypoglycemia.

Other secondary outcomes

For systolic and diastolic blood pressure, data were available from 21 and 20 studies, respectively, out of the 60 included studies. The number of direct comparisons was 33 for systolic blood pressure and 32 for diastolic blood pressure, involving 6,350 and 6,313 participants, respectively. However, owing to the limited number of included studies and the fragmented network structure, a coherent network could not be established for these outcomes. As a result, NMA could not be performed.

Regarding lipid profiles, data for LDL-C, HDL-C, and triglycerides were available from 27, 28, and 27 studies, respectively, of the 60 included studies. The NMA and PMA results for these outcomes are presented in Supplementary Figs. 8-10. The global test results indicated significant inconsistencies, with P<0.0001 for LDL-C, P<0.0001 for HDL-C, and P=0.0038 for triglycerides. While some treatment combinations showed statistically significant differences in either NMA or PMA, the results varied across drug classes and were inconsistent. Notably, the PMA for SGLT2i revealed a slight but statistically significant increase in LDL-C (WMD: 4.5 mg/dL; 95% CI, 1.9 to 7.0), a significant increase in HDL-C (WMD: 2.7 mg/dL; 95% CI, 1.9 to 3.4), and a significant decrease in triglycerides (WMD: –11.7 mg/dL; 95% CI, –18.7 to –4.8) compared to placebo.

DISCUSSION

This study represents the first systematic review and NMA to compare not only monotherapies but also various combination therapies in people with early T2DM and non-severe hyperglycemia (HbA1c ≤8.5%), who have minimal prior exposure to pharmacotherapy. The findings offer a comprehensive comparison of glycemic control efficacies and other outcomes across diverse therapeutic strategies in this specific patient population.

Regarding glycemic control, all dual combination therapies consistently demonstrated superior glycemic control compared to monotherapies. Among them, metformin+GLP-1RA (HbA1c reduction: vs. placebo WMD: –1.50%; 95% CI, –2.04 to –0.96) and metformin+DPP4i (HbA1c reduction: vs. placebo WMD: –1.46%; 95% CI, –1.96 to –0.95; HbA1c target achievement: vs. placebo OR: 34.33; 95% CI, 7.70 to 153.16) were the most effective combinations. In terms of the rank test results for HbA1c reduction and target achievement, the order was consistent across therapies, except for the minimal reversal in the ranking of metformin, SU, and TZD in monotherapies. Apart from the therapies not included in the analysis, all other therapies showed concordance between the two outcomes. Interestingly, metformin was the only therapy that failed to show statistically significant superiority over the placebo in both HbA1c reduction and target achievement (HbA1c reduction: vs. placebo WMD: 0.10%; 95% CI, –0.46 to 0.66; HbA1c target achievement: vs. placebo OR: 1.08; 95% CI, 0.19 to 6.08), which contradicts common clinical expectations. This discrepancy is likely due to the relatively small proportion of placebo-controlled comparisons involving metformin (14.7%, five of 34 comparisons), which may have underestimated its effect on NMA. In contrast, DPP4i had a higher proportion of placebo comparisons (51.5%, 17 of 33 comparisons), which may have provided a more accurate estimate of its efficacy. In the PMA, metformin showed statistically significant effects compared to placebo (HbA1c reduction: WMD: –0.53%; 95% CI, –0.64 to –0.43; HbA1c target achievement: OR: 11.49; 95% CI, 4.07 to 32.26). This finding reinforces the existing knowledge that metformin is an effective glucose-lowering therapy when analyzed using direct evidence, despite the NMA results. The glycemic control effects of various drug regimens in adults with early T2DM observed in this study are consistent with findings from general studies conducted in overall T2DM populations [8], further supporting the consistency of these therapeutic approaches.

In terms of weight reduction, GLP-1RA, SGLT2i, metformin, and their combination therapies were more effective than the placebo, whereas TZD, SU, and DPP4i ranked lower. Regarding any and serious adverse events, no significant differences were observed among the various agents and their combinations. Notably, and as expected, SU was the only therapy that significantly increased the risk of hypoglycemia compared to other treatments (vs. placebo OR: 17.2; 95% CI, 1.34 to 220.3). Regarding lipid profiles, including LDL-C, HDL-C, and triglycerides, the results were characterized by substantial heterogeneity and limited study comparisons, leading to inconsistent findings (global test P<0.01). In the PMA, despite the superior cardiovascular outcomes of SGLT2i, they slightly increased the LDL-C levels (vs. placebo WMD: 4.5 mg/dL; 95% CI, 1.9 to 7.0). SGLT2i also increased HDL-C (vs. placebo WMD: 2.7 mg/dL; 95% CI, 1.9 to 3.4) and decreased triglycerides (vs. placebo WMD: –11.7 mg/dL; 95% CI, –18.7 to –4.8). These results are consistent with recent reports [9], reflecting their complex effects on lipid metabolism. This pattern suggests that while SGLT2i improve certain cardiovascular outcomes, their impact on the lipid profile, particularly the increase in LDL-C, should be considered in the context of overall patient management.

This study has several limitations that should be considered when interpreting the results. First, our findings are only applicable to people with early T2DM. However, our thorough analysis confirmed that the effects of drug regimens in patients with early T2DM are consistent with those observed in the overall T2DM population. Second, the selection of studies focusing on early T2DM led to a high level of inconsistencies in each outcome, particularly when certain drug combinations were overly represented in placebo-controlled or active-controlled trials. This could still result in the overestimation or underestimation of drug efficacy, despite our mitigation efforts through PMA. As recent guidelines increasingly emphasize early combination therapy, the accumulation of results from RCTs focusing on various combination therapies in people with early T2DM who meet our study criteria, followed by NMA, could help overcome heterogeneity and further reduce the potential for interpretative errors. In our study, GLP-1RA showed a lower-than-expected effect on glycemic control and weight reduction. This outcome may be attributed to the varied efficacy of GLP-1RA agents, as the study included only one trial with subcutaneous semaglutide, two with oral semaglutide, one with liraglutide, two with dulaglutide, one with once-weekly exenatide, and three with once-daily exenatide. The relatively small number of studies involving more potent GLP-1RA agents could have contributed to the observed results. Therefore, in the future, additional RCTs involving more potent GLP-1RAs, such as subcutaneous semaglutide, in early T2DM will be necessary, and further integrated analyses of these studies may be required.

In conclusion, our NMA results demonstrated superior glycemic control with initial combination therapies compared to monotherapies without an increased risk of adverse events in early T2DM. However, the heterogeneity of the included studies and the specific focus on early T2DM may have over- or underestimated the effects. Therefore, future RCTs involving potent GLP-1RAs, novel antidiabetic agents, and various combination therapies, even in people with early T2DM, are crucial for establishing more effective and individualized treatment strategies within clinical guidelines.

SUPPLEMENTARY MATERIALS

NOTES

CONFLICTS OF INTEREST

Sang Youl Rhee has been an associate editor of the Diabetes & Metabolism Journal since 2022. He was not involved in the review process of this article. Otherwise, there was no conflict of interest.

AUTHOR CONTRIBUTIONS

Conception or design: M.K.M.

Acquisition, analysis, or interpretation of data: all authors.

Drafting the work or revising: J.H.C., M.K.M.

Final approval of the manuscript: J.H.C., Y.S.Y., M.K.M.

FUNDING

This research was funded by the Korea Centers for Disease Control and Prevention Agency (No. 2022-ER1105-00).

ACKNOWLEDGMENTS

None

Fig. 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) study flow for literature selection and exclusion process.

Fig. 2.

Network geometry for glycosylated hemoglobin (HbA1c) reduction efficacy of anti-diabetic treatments after 24 weeks (6 months) in the network meta-analysis. Each circle represents an antidiabetic treatment that was included in the analysis, with the size of the circle proportional to the number of trials involving the treatment. The width of each connecting line is proportional to the number of individuals who participated in each pair of trials. The numbers inside or next to the circles indicate the baseline HbA1c levels for each treatment group. The global test for inconsistency yielded a P value of 0.0724, which suggests potential inconsistencies in the network of treatments. TZD, thiazolidinedione; SGLT2i, sodium-glucose cotransporter-2 inhibitor; GLP-1RA, glucagon-like peptide-1 receptor agonist; DPP4i, dipeptidyl peptidase-4 inhibitor.

REFERENCES

• 1. Lee SH, Park SY, Choi CS. Insulin resistance: from mechanisms to therapeutic strategies. Diabetes Metab J 2022;46:15-37.ArticlePubMedPDF
• 2. GBD 2021 Diabetes Collaborators. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 2023;402:203-34.PubMedPMC
• 3. Sanchez-Rangel E, Inzucchi SE. Metformin: clinical use in type 2 diabetes. Diabetologia 2017;60:1586-93.ArticlePubMedPDF
• 4. Choi JH, Lee KA, Moon JH, Chon S, Kim DJ, Kim HJ, et al. 2023 Clinical practice guidelines for diabetes mellitus of the Korean Diabetes Association. Diabetes Metab J 2023;47:575-94.PubMedPMC
• 5. American Diabetes Association Professional Practice Committee. 9. Pharmacologic approaches to glycemic treatment: standards of care in diabetes-2024. Diabetes Care 2024;47(Suppl 1):S158-78.PubMed
• 6. Matthews DR, Paldanius PM, Proot P, Chiang Y, Stumvoll M, Del Prato S, et al. Glycaemic durability of an early combination therapy with vildagliptin and metformin versus sequential metformin monotherapy in newly diagnosed type 2 diabetes (VERIFY): a 5-year, multicentre, randomised, double-blind trial. Lancet 2019;394:1519-29.ArticlePubMed
• 7. Shi Q, Nong K, Vandvik PO, Guyatt GH, Schnell O, Ryden L, et al. Benefits and harms of drug treatment for type 2 diabetes: systematic review and network meta-analysis of randomised controlled trials. BMJ 2023;381:e074068.PubMedPMC
• 8. Maloney A, Rosenstock J, Fonseca V. A model-based meta-analysis of 24 antihyperglycemic drugs for type 2 diabetes: comparison of treatment effects at therapeutic doses. Clin Pharmacol Ther 2019;105:1213-23.ArticlePubMedPDF
• 9. Bechmann LE, Emanuelsson F, Nordestgaard BG, Benn M. SGLT2-inhibition increases total, LDL, and HDL cholesterol and lowers triglycerides: meta-analyses of 60 randomized trials, overall and by dose, ethnicity, and drug type. Atherosclerosis 2024;394:117236.ArticlePubMed

Figure & Data

References

Citations

Citations to this article as recorded by  

About this article

Choi JH, Koo BK, Yang YS, Min SH, Park JS, Rhee SY, Kim HJ, Moon MK. Initial Pharmacological Strategies in People with Early Type 2 Diabetes Mellitus: A Systematic Review and Network Meta-Analysis. Diabetes Metab J. 2025 Apr 29. doi: 10.4093/dmj.2024.0660. Epub ahead of print.

Received: Oct 24, 2024; Accepted: Jan 16, 2025

DOI: https://doi.org/10.4093/dmj.2024.0660.

Download citation ↓