INTRODUCTION

Type 2 diabetes (T2D) is a significant risk factor for cardiovascular disease (CVD). Coronary artery disease and stroke are the main causes of death in T2D patients [1–4]. Dyslipidemia is a major modifiable factor for CVD prevention [3,4]. Therefore, the management of dyslipidemia is necessary to reduce mortality due to CVD in patients with T2D [5,6].

Low levels of high-density lipoprotein cholesterol (HDL-C) are a key feature of diabetic dyslipidemia [6–9], as well as a major independent risk factor for CVD [10]. While statins are the mainstay of dyslipidemia management, they have limited effects on HDL-C [11]. A previous study found that HDL-C levels increased when sodium-glucose cotransporter-2 inhibitors (SGLT2-i) and thiazolidinediones (TZDs) were administered alone or in combination [4,7,12–19]. Therefore, it was expected that a combination of TZD and SGLT2-i would increase HDL-C levels more than treatment with only SGLT2-i.

Some clinical trials have added SGLT2-i to background TZD treatment [7,17,18,20], but to our knowledge, there have been no clinical trials in which TZD was added to SGLT2-i treatment. In this study, using national health insurance data, we aimed to investigate whether treatment with a combination of TZD and SGLT2-i resulted in a greater improvement in HDL-C levels than treatment with SGLT2-i alone.

METHODS

Ethical statement

This study was approved by the Institutional Review Board of Sungkyunkwan University (No. 2021-05-017). Participants who underwent national health checkup examinations provided written informed consent for the use of their data for research purposes. All personal information was deleted, and only de-identified data were included in the analysis.

Study setting and data source

Korea has a mandatory social health insurance system, called the National Health Insurance Service (NHIS), which is run by the government (i.e., the Ministry of Health and Welfare). The entire Korean population is covered by the NHIS, except for the lowest-income segment of the population (approximately 3%), who are covered by medical aid. In Korea, all medical providers are compulsorily designated by the NHIS, and the providers must register all disease diagnoses, medical procedures, and drug prescription information for reimbursement from the NHIS. All administrative procedures for NHIS subscribers and the medical aid population are performed by the NHIS, which is a public corporation.

The Korean government also operates the National Health Screening Program (NHSP). All insured adults are eligible for a regular health checkup every 2 years. In 2014, the target population's participation rate in the program's general health examination was 74.8% [21]. Currently, Korea's NHSP is the world's most extensive health screening program.

This retrospective study used the NHIS database. The NHIS maintains five databases: the qualification database, the national health checkup database, the medical use database, the long-term care insurance database (since 2014), and the healthcare provider database [22]. The NHIS operates the National Health Information Sharing Service to support policies and academic research by providing information (Wonju, NHIS; https://nhiss.nhis.or.kr). The NHIS databases have been extensively used for medical and health policy research.

For this research, we used a custom NHIS database, which was designed according to the researchers’ needs [22]. Data extraction was performed by medical record technicians at the NHIS Center, who had no conflicts of interest related to this study.

Study population

We extracted 129,666 T2D patients (International Classification of Diseases 10th revision [ICD-10] codes, E10–E14) who were first prescribed SGLT2-i from the NHIS database from October 1, 2014 to December 31, 2016. Of these, 127,066 adults aged 30 years or older were selected.

The index date was the first prescription date for T2D patients who first received SGTL2-i between 2014 and 2016. As of the index date, T2D patients who had not received a medical examination during the previous 2 years and within 2 years thereafter were excluded. We also excluded patients with type 1 diabetes mellitus (ICD-10, E10) and gestational diabetes mellitus (ICD-10, O244 and O249) from January 1, 2002 to the index date.

The group with TZD prescriptions was defined as patients who were also prescribed TZD at the time of the first prescription of SGTL2-i, and it excluded patients who received only SGTL2-i. Subsequently, we performed propensity score matching (PSM), which is known to reduce confounding in observational studies [23]. All possible covariates, including demographic and medical information (Table 1) were included in PSM, and logistic regression analysis was performed to calculate the propensity scores in a logistic model. We set the caliper for nearest-neighbor matching within the first 4 to 8 digits, allowing PSM at a 1:1 ratio. Finally, 700 patients were obtained for each group: group A (SGTL2i and TZD, regardless of other antidiabetic medications) and group B (SGTL2i only without TZD, regardless of other antidiabetic medications) (Fig. 1).

Study endpoint

The primary endpoint was the posttreatment HDL-C level. The secondary endpoints were posttreatment triglyceride (TG), low-density lipoprotein cholesterol (LDL-C), and total cholesterol (TC) levels.

Statistical analysis

The absolute value of the standardized mean difference (ASD) was used to determine the balance before and after PSM, and an ASD <0.1 was used to determine whether all covariates were sufficiently balanced through PSM. Since PSM was performed and baseline characteristics of groups A and B were comparable, differences in the study endpoints between groups A and B were compared with the Student t-test without consideration of the baseline characteristics. In addition, the paired t-test was used to analyze differences between the pre- and posttreatment values within each group. All statistical analyses were performed using SAS ver. 9.4 (SAS Institute Inc., Cary, NC, USA), and P-values less than 0.05 were considered statistically significant.

RESULTS

Patient disposition and baseline characteristics

After PSM, 1,400 patients (700 patients in each group) were included in the analysis. The characteristics of groups A and B before and after PSM are compared in Table 1. The ASDs of all covariates were below 0.1, indicating that the two groups were sufficiently balanced after PSM (Table 1).

Comparison of the primary outcome HDL-C

In the posttreatment comparison conducted using the Student t-test, group A showed significantly higher HDL-C levels (51.4±12.92 vs. 49.1±12.15 mg/dL, P<0.001) than group B. When comparing the pretreatment and posttreatment HDL-C levels within each group, a significant increase was found in group A (49.54±20.03 to 51.6±12.92 mg/dL, P=0.007), whereas no significant change was observed in group B (49.14±13.52 to 49.1±2.15 mg/dL, P=0.937) (Table 2).

Changes in the secondary outcomes

In the posttreatment comparison conducted using the Student t-test, group A showed significantly lower TG levels than group B (120.51 mg/dL, 95% confidence interval [CI] 115.6–126.62 mg/dL vs. 135.59 mg/dL, 95% CI 129.98–141.43 mg/dL; P<0.001). When comparing the pretreatment and posttreatment TG levels within each group, a significant decrease was observed in group A (141.53 mg/dL [95% CI, 135.71–147.6 mg/dL] to 120.51 mg/dL [95% CI, 115.6–125.62 mg/dL]; P<0.001), but not in group B (141.15 mg/dL [95% CI, 135.7–147.6 mg/dL] to 135.59 mg/dL [95% CI, 129.98–141.43 mg/dL]; P=0.056).

No statistically significant differences were found between groups A and B in the posttreatment comparisons of TC levels (167.4±41.49 mg/dL vs. 165.85±40.71 mg/dL, P=0.479) and LDL-C levels (88.23±36.05 mg/dL vs. 86.13±33.76 mg/dL, P=0.261) using the Student t-test. However, both groups showed significant decreases in these parameters in comparisons between pretreatment and posttreatment values. TC levels decreased from 176.01±44.08 to 167.4±41.49 mg/dL (P<0.001) in group A and from 176.48±41.25 to 165.85±40.71 mg/dL (P<0.001) in group B. LDL-C levels decreased from 94.9±36.15 to 88.23±36.05 mg/dL (P<0.001) in group A and from 95.62±37.01 to 86.13±33.76 mg/dL (P<0.001) in group B (Table 2).

DISCUSSION

In this study, we tried to determine whether dyslipidemia could be improved by using TZD in combination with SGTL2-i. Although our study was an observational study, we tried to reduce the differences between the groups by using PSM, and our real-world data might be useful considering that no clinical trial data regarding this question are available yet.

An increase in HDL-C levels was observed only in patients who were prescribed SLGT2-i and TZD (group A), who had higher posttreatment HDL-C levels than patients who were prescribed SGLT2-i only (group B). These results suggest that the combination of TZD and SGLT2-i might be helpful for the improvement of HDL-C. Many previous studies have confirmed that TZD significantly increases HDL-C levels [12,15,16,24]. Although it has been accepted that the increase in HDL-C levels with TZD treatment is entirely due to an increase in the HDL3 subfraction, the mechanism by which TZD induces alterations in the HDL3 subfraction remains unclear [25].

SGTL2-i medications are generally considered to improve HDL-C, but unexpectedly, the present study using NHIS data found no increase in HDL-C levels with SGTL2-i treatment. A possible explanation might be that previous studies inconsistent results were reported regarding the effects of different SGLT2-i drugs on HDL-C levels [22–24], which is relevant since group B in our real-world study used various types of SGTL2-i (dapagliflozin, empagliflozin, and ipragliflozin). In addition, treatment persistence and/or adherence might not have been optimal in this study population compared to a clinical trial. Limitations of insurance benefits for SGTL2-i and TZD may also be a factor (Fig. 2).

TG is also an important aspect of diabetic dyslipidemia and a substantial risk factor for atherosclerotic CVD in patients with T2D [26]. In patients who were prescribed SLGT2-i and TZD (group A), the posttreatment TG levels were lower than in those who were prescribed SGLT2-i only (group B). This is in line with the results of previous studies showing a decrease in TG with TZD treatment [15,27,28].

LDL-C and TC levels decreased after treatment within groups A and B, but the posttreatment comparison between groups A and B did not yield significant results. TZD showed a consistent LDL-C reduction effect [15,28], while SGTL2-i slightly increased LDL-C levels [14], although differences were observed depending on the specific drugs or doses. Variations in the types and doses of SGTL2-i were not considered in this study, but should be investigated in future research.

There are several limitations of this study. First, the National Health Insurance Research Database (NHIRD) was created for entitlement and reimbursement management; therefore, it lacks clinical information, such as why patients were prescribed specific antidiabetic drugs. While we tried to balance the two groups with PSM, there could have still been some unmeasured confounders. Second, the number of patients in each group was small, at 700. Currently, the reimbursement of oral antidiabetic drugs is limited to two drugs, and the patient must pay for a third drug. As most people are prescribed metformin as the first-line drug, the number of patients who are prescribed both SGLT2-i and TZD is limited. Therefore, although this was a nationwide study, the sample size was only 700 patients and a more detailed analysis was not possible. Third, our endpoint of the lipid profile is only a proxy marker of CVD. Fourth, information about the concomitant use of anti-hyperlipidemic drugs was lacking. Further research is needed to clarify whether improving the lipid profile could lead to a reduced risk of coronary artery disease.

In conclusion, our study results suggest that a combination of SGTL2-i and TZD might be more effective for improving dyslipidemia than SGLT2-i alone in T2D patients. Our findings are consistent with the recommendations for using TZD in the American Diabetes Association guidelines for T2D patients with a higher CVD risk, but further studies are needed to confirm these findings.

ARTICLE INFORMATION

Ethical statements

This study was approved by the Institutional Review Board of Sungkyunkwan University (No. 2021-05-017). Participants who underwent national health checkup examinations provided written informed consent for the use of their data for research purposes. All personal information was deleted, and only de-identified data were included in the analysis.

Conflicts of interest

The authors have no conflicts of interest to declare.

Funding

None.

Author contributions

Conceptualization: TP, KH, DS; Data curation: KH, TP; Formal analysis: KH, TP; Methodology: DS, TP, KH; Project administration: DS, TG, HK; Supervision: DS; Validation: DS, TP, KH, JP; Writing–original draft: TP, KH, DS, JP; Writing–review & editing: TP, JP, KH, DS.

All authors read and approved the final manuscript.

Fig. 1.

Patient disposition. SGLT2-I, sodium-glucose cotransporter-2 inhibitors; TZD, thiazolidinedione.

Fig. 2.

Insurance benefits for antidiabetic agents in Korea based on the Health Insurance Review and Assessment Service. Pioglitazone is 100% covered by the patient. MET, metformin; SU, sulfonylurea; a-GI, alpha-glucosidase inhibitors; TZD, thiazolidinedione; DPP4-i, dipeptidyl peptidase-4 inhibitors; SGLT2-i, sodium-glucose cotransporter 2 inhibitors.

REFERENCES

• 1. Schramm TK, Gislason GH, Kober L, Rasmussen S, Rasmussen JN, Abildstrom SZ, et al. Diabetes patients requiring glucose-lowering therapy and nondiabetics with a prior myocardial infarction carry the same cardiovascular risk: a population study of 3.3 million people. Circulation 2008;117:1945-54.ArticlePubMed
• 2. Einarson TR, Acs A, Ludwig C, Panton UH. Prevalence of cardiovascular disease in type 2 diabetes: a systematic literature review of scientific evidence from across the world in 2007-2017. Cardiovasc Diabetol 2018;17:83. ArticlePubMedPMCPDF
• 3. Rhee EJ, Kim HC, Kim JH, Lee EY, Kim BJ, Kim EM, et al. 2018 guidelines for the management of dyslipidemia in Korea. Korean J Intern Med 2019;34:1171. ArticlePubMedPMCPDF
• 4. Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK, Blumenthal RS, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines. Circulation 2019;139:e1082-143.ArticlePubMedPMC
• 5. Cholesterol Treatment Trialists' (CTT) Collaborators, Kearney PM, Blackwell L, Collins R, Keech A, Simes J, et al. Efficacy of cholesterol-lowering therapy in 18,686 people with diabetes in 14 randomised trials of statins: a meta-analysis. Lancet 2008;371:117-25.ArticlePubMed
• 6. Schofield JD, Liu Y, Rao-Balakrishna P, Malik RA, Soran H. Diabetes dyslipidemia. Diabetes Ther 2016;7:203-19.ArticlePubMedPMCPDF
• 7. Kashiwagi A, Shiga T, Akiyama N, Kazuta K, Utsuno A, Yoshida S, et al. Efficacy and safety of ipragliflozin as an add-on to pioglitazone in Japanese patients with inadequately controlled type 2 diabetes: a randomized, double-blind, placebo-controlled study (the SPOTLIGHT study). Diabetol Int 2015;6:104-116.ArticlePDF
• 8. Mooradian AD. Dyslipidemia in type 2 diabetes mellitus. Nat Clin Pract Endocrinol Metab 2009;5:150-9.ArticlePubMedPDF
• 9. Farbstein D, Levy AP. HDL dysfunction in diabetes: causes and possible treatments. Expert Rev Cardiovasc Ther 2012;10:353-61.ArticlePubMedPMC
• 10. Mahmood SS, Levy D, Vasan RS, Wang TJ. The Framingham Heart Study and the epidemiology of cardiovascular disease: a historical perspective. Lancet 2014;383:999-1008.ArticlePubMed
• 11. Yamashita S, Tsubakio-Yamamoto K, Ohama T, Nakagawa-Toyama Y, Nishida M. Molecular mechanisms of HDL-cholesterol elevation by statins and its effects on HDL functions. J Atheroscler Thromb 2010;17:436-51.ArticlePubMed
• 12. Filippas-Ntekouan S, Tsimihodimos V, Filippatos T, Dimitriou T, Elisaf M. SGLT-2 inhibitors: pharmacokinetics characteristics and effects on lipids. Expert Opin Drug Metab Toxicol 2018;14:1113-21.ArticlePubMed
• 13. Chen MB, Wang H, Cui WY, Xu HL, Zheng QH. Effect of SGLT inhibitors on weight and lipid metabolism at 24 weeks of treatment in patients with diabetes mellitus: a systematic review and network meta-analysis. Medicine (Baltimore) 2021;100:e24593.ArticlePubMedPMC
• 14. Cha SA, Park YM, Yun JS, Lim TS, Song KH, Yoo KD, et al. A comparison of effects of DPP-4 inhibitor and SGLT2 inhibitor on lipid profile in patients with type 2 diabetes. Lipids Health Dis 2017;16:58. ArticlePubMedPMCPDF
• 15. Bae J, Park T, Kim H, Lee M, Cha BS. Lobeglitazone: a novel thiazolidinedione for the management of type 2 diabetes mellitus. Diabetes Metab J 2021;45:326-36.ArticlePubMedPMCPDF
• 16. Rizzo M, Christ ER, Rini GB, Spinas GA, Berneis K. The differential effects of thiazolidindiones on atherogenic dyslipidemia in type 2 diabetes: what is the clinical significance? Expert Opin Pharmacother 2008;9:2295-303.ArticlePubMed
• 17. Rosenstock J, Vico M, Wei L, Salsali A, List JF. Effects of dapagliflozin, an SGLT2 inhibitor, on HbA(1c), body weight, and hypoglycemia risk in patients with type 2 diabetes inadequately controlled on pioglitazone monotherapy. Diabetes Care 2012;35:1473-8.ArticlePubMedPMCPDF
• 18. Forst T, Guthrie R, Goldenberg R, Yee J, Vijapurkar U, Meininger G, et al. Efficacy and safety of canagliflozin over 52 weeks in patients with type 2 diabetes on background metformin and pioglitazone. Diabetes Obes Metab 2014;16:467-77.ArticlePubMedPMCPDF
• 19. Filipova E, Uzunova K, Kalinov K, Vekov T. Effects of pioglitazone therapy on blood parameters, weight and BMI: a meta-analysis. Diabetol Metab Syndr 2017;9:90. ArticlePubMedPMCPDF
• 20. Kovacs CS, Seshiah V, Merker L, Christiansen AV, Roux F, Salsali A, et al. Empagliflozin as add-on therapy to pioglitazone with or without metformin in patients with type 2 diabetes mellitus. Clin Ther 2015;37:1773. -88. e1.ArticlePubMed
• 21. Seong SC, Kim YY, Park SK, Khang YH, Kim HC, Park JH, et al. Cohort profile: the National Health Insurance Service-National Health Screening Cohort (NHIS-HEALS) in Korea. BMJ Open 2017;7:e016640.ArticlePubMedPMC
• 22. Shin DW, Cho J, Park JH, Cho B. National General Health Screening Program in Korea: history, current status, and future direction. Precis Future Med 2022;6:9-31.ArticlePDF
• 23. Stuart EA, Lee BK, Leacy FP. Prognostic score-based balance measures can be a useful diagnostic for propensity score methods in comparative effectiveness research. J Clin Epidemiol 2013;66(8 Suppl):S84-90.e1.ArticlePubMedPMC
• 24. Kim SG, Kim DM, Woo JT, Jang HC, Chung CH, Ko KS, et al. Efficacy and safety of lobeglitazone monotherapy in patients with type 2 diabetes mellitus over 24-weeks: a multicenter, randomized, double-blind, parallel-group, placebo controlled trial. PLoS One 2014;9:e92843.ArticlePubMedPMC
• 25. Ovalle F, Bell DS. Effect of thiazolidinediones on high-density lipoprotein subfractions. Endocr Pract 2002;8:102-4.ArticlePubMed
• 26. Hirano T. Pathophysiology of diabetic dyslipidemia. J Atheroscler Thromb 2018;25:771-82.ArticlePubMedPMC
• 27. Kim SH, Kim SG, Kim DM, Woo JT, Jang HC, Chung CH, et al. Safety and efficacy of lobeglitazone monotherapy in patients with type 2 diabetes mellitus over 52 weeks: an open-label extension study. Diabetes Res Clin Pract 2015;110:e27-30.ArticlePubMed
• 28. Jin SM, Park CY, Cho YM, Ku BJ, Ahn CW, Cha BS, et al. Lobeglitazone and pioglitazone as add-ons to metformin for patients with type 2 diabetes: a 24-week, multicentre, randomized, double-blind, parallel-group, active-controlled, phase III clinical trial with a 28-week extension. Diabetes Obes Metab 2015;17:599-602.ArticlePubMedPMCPDF

Figure & Data

References

Citations

Citations to this article as recorded by