Abstract
We conducted high dimensional propensity score-adjusted cohort studies to examine whether thiazolidinedione use with a statin or fibrate was associated with an increased risk of severe hypoglycemia. We found that concomitant therapy with a thiazolidinedione + fibrate was associated with a generally delayed increased risk of severe hypoglycemia.
Keywords: cohort studies, comparative effectiveness research, diabetes mellitus, drug interactions, hypoglycemia, hypolipidemic agents, Medicaid, pharmacoepidemiology, propensity score, thiazolidinediones
1. INTRODUCTION
Dyslipidemia is a major, yet modifiable, risk factor for cardiovascular disease. While glycemic control improves the lipid profile of persons with diabetes, treatment with antihyperlipidemics is often indicated. Co-prescribing of antidiabetic and antihyperlipidemic agents, though, may not be without risks. In particular, thiazolidinediones (TZDs)—peroxisome proliferator-activated receptor (PPAR) γ agonists which increase insulin sensitivity—are metabolized primarily by hepatic cytochrome P450 (CYP) 2C8.1 This isozyme can be inhibited by some antihyperlipidemics, most notably fibrates,2 leading to higher concentrations of TZDs. In addition, the PPAR α activity of fibrates may itself have effects on glucose homeostasis.3 Some statins may also affect glucose metabolism.4 These mechanisms might result in enhanced glucose lowering effects in concomitant users of TZDs and certain antihyperlipidemics. While these effects may be desirable for some patients, drug interactions might also increase the risk of severe hypoglycemia—a major clinical and public health problem. We therefore examined severe hypoglycemia risk among concomitant users of TZDs and antihyperlipidemics.
2. METHODS
We conducted two high-dimensional propensity score-adjusted cohort studies of adult users of pioglitazone and rosiglitazone, respectively, using Medicaid data from five large states. Each cohort consisted exclusively of person-time concomitantly-exposed to the TZD plus one of eight antihyperlipidemics: atorvastatin; fenofibrate; fluvastatin; gemfibrozil; lovastatin; pravastatin; rosuvastatin; or simvastatin. The day on which the subject was first co-exposed served as the cohort entry date. Exposure was defined by the antihyperlipidemic active upon cohort entry. The primary outcome was a validated diagnosis-based algorithm for severe hypoglycemia within 30 days of cohort entry. Please see Supplemental Materials for details on: the data source; defining the study cohorts; exposure, covariate, and outcome ascertainment; and statistical analyses.
3. RESULTS
Pioglitazone
Characteristics of pioglitazone users are presented in Table 1. Unadjusted and adjusted hazard ratios (HRs) for severe hypoglycemia within 30 days are presented in Table 2 and Fig. 1, respectively. Unadjusted and adjusted HRs for severe hypoglycemia within 180 days are presented in Table 2. Time-specific association measures for concomitant use of pioglitazone and fibrates are presented in Fig. 2. No time-course effects were evident for concomitant use with statins. See Supplemental Materials for results from sensitivity analyses.
Table 1.
Characteristics of pioglitazone users, by antihyperlipidemic exposure group
| Analyses examining 30-day time period post-cohort entry |
Statins | Fibrates | |||||||
|---|---|---|---|---|---|---|---|---|---|
| pravastatin | atorvastatin | fluvastatin | lovastatin | rosuvastatin | simvastatin | fenofibrate | gemfibrozil | ||
| Ns (unless otherwise noted) | |||||||||
| Users, concomitant with pioglitazone | 21,066 | 109,371 | 4,757 | 15,818 | 13,014 | 69,847 | 10,969 | 11,531 | |
| Person-years of follow-up | 1,709 | 8,909 | 385 | 1,282 | 1,048 | 5,662 | 829 | 891 | |
| Severe hypoglycemia events within 30 days of cohort entry | 124 | 595 | 20 | 68 | 43 | 408 | 53 | 65 | |
| Cumulative incidence of severe hypoglycemia within 30 days of cohort entry (95% CI) | 0.59% (0.49–0.70) | 0.54% (0.50–0.59) | 0.42% (0.26–0.65) | 0.43% (0.33–0.54) | 0.33% (0.24–0.44) | 0.58% (0.53–0.64) | 0.48% (0.36–0.63) | 0.56% (0.44–0.72) | |
| Demographics | Group | % (unless otherwise noted) | |||||||
| Age in years at cohort entry (continuous) | Median (Q1-Q3) | 66.4 (56.0–74.1) | 64.7 (54.3–73.0) | 63.1 (52.4–72.5) | 63.3 (51.0–72.8) | 65.3 (54.5–73.1) | 65.5 (55.1–73.6) | 59.8 (48.5–70.4) | 57.8 (47.7–68.9) |
| Sex | female | 66.5 | 65.0 | 68.4 | 64.2 | 64.4 | 64.8 | 56.8 | 53.7 |
| Race | white | 35.1 | 37.1 | 35.1 | 28.9 | 32.8 | 36.2 | 49.8 | 40.0 |
| black | 13.6 | 13.9 | 14.0 | 12.1 | 11.6 | 15.0 | 6.5 | 7.0 | |
| other / unknown | 51.3 | 49.0 | 51.0 | 59.0 | 55.5 | 48.8 | 43.7 | 52.9 | |
| State of residence | CA | 58.3 | 52.0 | 59.7 | 69.6 | 40.4 | 41.2 | 41.6 | 61.5 |
| FL | 8.4 | 5.9 | 10.0 | 8.0 | 19.4 | 11.1 | 12.5 | 8.2 | |
| NY | 21.8 | 27.8 | 16.4 | 12.2 | 33.3 | 32.8 | 26.5 | 18.9 | |
| OH | 5.6 | 9.1 | 6.3 | 4.5 | 4.0 | 8.5 | 12.4 | 6.9 | |
| PA | 5.9 | 5.3 | 7.7 | 5.8 | 2.8 | 6.4 | 7.0 | 4.5 | |
| Calendar year of cohort entry | 2000–2003 | 54.7 | 36.5 | 52.5 | 10.7 | 1.4 | 22.3 | 28.0 | 39.8 |
| 2004 | 10.3 | 11.0 | 13.2 | 9.0 | 9.9 | 6.3 | 9.1 | 10.3 | |
| 2005 | 10.1 | 13.8 | 13.5 | 13.3 | 16.5 | 11.0 | 14.3 | 12.3 | |
| 2006 | 8.8 | 13.8 | 11.0 | 25.3 | 23.3 | 14.9 | 16.0 | 11.7 | |
| 2007 | 9.0 | 16.4 | 7.3 | 26.6 | 30.6 | 26.1 | 19.1 | 15.0 | |
| 2008 | 7.0 | 8.5 | 2.5 | 15.1 | 18.4 | 19.3 | 13.5 | 11.0 | |
| Medicare enrolled | Yes | 66.2 | 62.2 | 58.5 | 60.5 | 61.5 | 64.5 | 59.6 | 52.8 |
| Nursing home residence, ever during baseline | Yes | 4.7 | 6.6 | 4.0 | 5.5 | 2.8 | 6.8 | 5.1 | 5.6 |
| Healthcare utilization covariates, in baseline period* | Group | measures of central tendency | |||||||
| # prescriptions dispensed | Median (Q1-Q3) | 57.0 (33.0–90.0) | 58.0 (33.0–93.0) | 52.0 (28.0–83.0) | 44.0 (22.0–76.0) | 58.0 (32.0–93.0) | 58.0 (31.0–94.0) | 65.0 (36.0–106) | 57.0 (31.0–92.0) |
| # unique drugs dispensed | Median (Q1-Q3) | 15.0 (10.0–22.0) | 15.0 (9.0–21.0) | 13.0 (8.0–20.0) | 12.0 (7.0–18.0) | 15.0 (9.0–22.0) | 14.0 (9.0–21.0) | 16.0 (10.0–23.0) | 14.0 (9.0–21.0) |
| # outpatient diagnosis codes | Median (Q1-Q3) | 47.0 (23.0–90.0) | 47.0 (23.0–95.0) | 36.0 (16.0–71.0) | 27.0 (10.0–61.0) | 43.0 (20.0–89.0) | 44.0 (19.0–95.0) | 50.0 (24.0–98.0) | 40.0 (18.0–82.0) |
| # unique outpatient diagnosis codes | Median (Q1-Q3) | 17.0 (10.0–26.0) | 16.0 (9.0–26.0) | 13.0 (7.0–23.0) | 11.0 (5.0–20.0) | 15.0 (9.0–25.0) | 15.0 (8.0–26.0) | 17.0 (10.0–27.0) | 14.0 (8.0–23.0) |
| # outpatient CPT-4 / HCPCS procedure codes | Median (Q1-Q3) | 55.0 (28.0–103) | 55.0 (27.0–104) | 45.0 (22.0–85.0) | 38.0 (16.0–75.0) | 53.0 (26.0–102) | 51.0 (22.0–103) | 58.0 (30.0–107) | 48.0 (24.0–93.0) |
| # unique outpatient CPT-4 / HCPCS procedure codes | Median (Q1-Q3) | 30.0 (17.0–48.0) | 29.0 (16.0–49.0) | 26.0 (14.0–41.0) | 22.0 (10.0–39.0) | 29.0 (16.0–48.0) | 28.0 (14.0–48.0) | 31.0 (18.0–49.0) | 27.0 (15.0–45.0) |
| Other investigator pre- defined covariates, in baseline period | Group | % | |||||||
| Prior severe hypoglycemia | Yes | 3.4 | 3.4 | 2.7 | 2.4 | 2.0 | 3.5 | 2.7 | 3.4 |
| Alpha-glucosidase inhibitor exposure | Yes | 2.5 | 2.2 | 2.5 | 1.7 | 1.6 | 1.9 | 1.9 | 2.1 |
| DPP-4 inhibitor exposure | Yes | 0.7 | 1.1 | 0.4 | 1.1 | 3.2 | 2.1 | 1.9 | 0.7 |
| GLP-1 inhibitor exposure | Yes | 0.3 | 0.5 | ** | 0.7 | 1.5 | 0.9 | 1.1 | 0.4 |
| Insulin exposure | Yes | 25.4 | 26.1 | 23.1 | 20.4 | 19.1 | 24.7 | 23.5 | 25.3 |
| Meglitinide exposure | Yes | 6.7 | 5.3 | 5.0 | 3.0 | 5.4 | 4.7 | 6.6 | 4.5 |
| Metformin exposure | Yes | 59.4 | 60.9 | 58.8 | 65.8 | 60.7 | 61.1 | 59.6 | 65.6 |
| Sulfonylurea exposure: glipizide | Yes | 22.7 | 23.9 | 25.5 | 26.6 | 18.9 | 22.8 | 19.3 | 26.0 |
| Sulfonylurea exposure: glyburide | Yes | 29.4 | 27.0 | 29.2 | 28.0 | 22.8 | 27.0 | 23.5 | 30.2 |
| Sulfonylurea exposure: other agent | Yes | 11.4 | 11.4 | 9.5 | 10.8 | 12.7 | 10.8 | 13.5 | 10.1 |
CI = confidence interval; CPT-4 = Current Procedural Terminology-4; DPP-4 = dipeptidyl peptidase-4; GLP-1 = glucagon-like peptide-1; HCPCS = Healthcare Common Procedure Coding System; Q = quartile
the following healthcare utilization covariates were excluded from the table, as the median values were zero: # inpatient International Classification of Diseases, 9th Revision (ICD-9) diagnosis codes; # unique inpatient ICD-9 diagnosis codes; # inpatient ICD-9 procedure codes; # unique inpatient ICD-9 procedure codes; # inpatient CPT-4/HCPCS procedure codes; # unique inpatient CPT-4/HCPCS procedure codes; # outpatient ICD-9 procedure codes; # unique outpatient ICD-9 procedure codes; # other setting diagnosis codes; # unique other setting diagnosis codes; # other setting ICD-9 procedure codes; # unique other setting ICD-9 procedure codes
omitted in compliance with Centers for Medicare & Medicaid Services privacy policy (i.e., underlying cell count <11 persons)
Table 2.
Summary of findings: pioglitazone + antihyperlipidemic and the risk of severe hypoglycemia
| Analyses examining 30-day time period post-cohort entry | Statins | Fibrates | ||||||
|---|---|---|---|---|---|---|---|---|
| pravastatin | atorvastatin | fluvastatin | lovastatin | rosuvastatin | simvastatin | fenofibrate | gemfibrozil | |
| Point estimates (95% CI) from primary analyses | ||||||||
| Unadjusted HR | reference | 0.92 (0.76–1.12) | 0.72 (0.45–1.15) | 0.73 (0.54–0.98) | 0.57 (0.40–0.80) | 0.99 (0.81–1.21) | 0.88 (0.64–1.21) | 1.00 (0.74–1.35) |
| Adjusted HR [see Figure 1] | 0.94 (0.77–1.14) | 0.85 (0.53–1.37) | 0.99 (0.73–1.35) | 0.91 (0.63–1.30) | 1.07 (0.87–1.32) | 1.08 (0.78–1.49) | 1.15 (0.85–1.56) | |
| Point estimates (95% CI) from sensitivity analyses | ||||||||
| Adjusted HR, excluding SU users* | reference | 1.17 (0.87–1.58) | 1.20 (0.62–2.31) | 1.15 (0.71–1.84) | 1.05 (0.61–1.78) | 1.23 (0.90–1.69) | 0.91 (0.53–1.55) | 1.23 (0.77–1.96) |
| Adjusted HR, excluding SU or insulin users** | 1.62 (0.83–3.15) | 2.09 (0.65–6.73) | 2.05 (0.84–5.03) | 1.91 (0.73–5.01) | 1.64 (0.81–3.31) | 1.32 (0.47–3.68) | 1.62 (0.63–4.18) | |
| Adjusted HR, excluding covariates from the PS strongly related to exposure but not outcome | 0.94 (0.77–1.14) | 0.84 (0.52–1.35) | 1.00 (0.74–1.37) | 0.92 (0.64–1.31) | 1.07 (0.87–1.32) | 1.07 (0.77–1.49) | 1.14 (0.84–1.55) | |
| Adjusted HR, excluding managed care enrollees | 0.97 (0.77–1.22) | 1.03 (0.60–1.75) | 1.25 (0.84–1.87) | 0.86 (0.56–1.31) | 1.12 (0.89–1.43) | 1.08 (0.74–1.58) | 1.26 (0.89–1.79) | |
| Analyses examining 180-day time period post-cohort entry | Point estimates (95% CI) from primary analyses | |||||||
| Unadjusted HR | reference | 1.00 (0.88–1.13) | 0.88 (0.67–1.17) | 0.74 (0.61–0.89) | 0.62 (0.49–0.77) | 1.01 (0.89–1.16) | 1.05 (0.86–1.29) | 1.34 (1.11–1.62) |
| Adjusted HR | 1.02 (0.90–1.15) | 1.03 (0.78–1.37) | 1.02 (0.83–1.25) | 1.02 (0.81–1.28) | 1.11 (0.97–1.27) | 1.33 (1.08–1.63) | 1.60 (1.32–1.93) | |
| Point estimates (95% CI) from sensitivity analyses | ||||||||
| Adjusted HR, excluding SU users* | reference | 1.10 (0.91–1.34) | 1.20 (0.78–1.84) | 0.98 (0.71–1.36) | 1.14 (0.80–1.62) | 1.13 (0.92–1.39) | 1.05 (0.75–1.47) | 1.39 (1.03–1.88) |
| Adjusted HR, excluding SU or insulin users** | 1.20 (0.77–1.86) | 2.25 (1.07–4.76) | 1.12 (0.57–2.20) | 1.30 (0.64–2.61) | 1.19 (0.74–1.90) | 0.96 (0.45–2.04) | 1.52 (0.80–2.90) | |
| Adjusted HR, excluding covariates from the PS strongly related to exposure but not outcome | 1.02 (0.90–1.16) | 1.03 (0.77–1.36) | 1.02 (0.83–1.24) | 1.01 (0.81–1.27) | 1.11 (0.97–1.27) | 1.31 (1.07–1.61) | 1.59 (1.31–1.92) | |
| Adjusted HR, excluding managed care enrollees | 1.07 (0.92–1.24) | 1.11 (0.80–1.55) | 1.09 (0.83–1.43) | 1.04 (0.80–1.36) | 1.15 (0.99–1.34) | 1.36 (1.07–1.72) | 1.57 (1.25–1.96) | |
CI = confidence interval; HR = hazard ratio; PS = propensity score; SU = sulfonylurea
Bolded values met the traditional threshold for statistical significance
if co-exposed within 60 days prior to cohort entry and censoring follow-up time if subsequently exposed to a SU
if co-exposed within 60 days prior to cohort entry and censoring follow-up time if subsequently exposed to a SU or insulin
Figure 1.
Adjusted hazard ratios (HRs) for the rate of severe hypoglycemia within 30 days of cohort entry among pioglitazone users, by antihyperlipidemic of interest (vs. pravastatin)
Figure 2.
Adjusted hazard ratios (HRs) for the rate of severe hypoglycemia within 180 days of cohort entry among pioglitazone users, by fibrate of interest (vs. pravastatin), by time window
Rosiglitazone
4. DISCUSSION
We examined potential drug-drug interactions between TZDs and antihyperlipidemics. While we found no increased risk of severe hypoglycemia during the first month of concomitant pioglitazone and antihyperlipidemic therapy, the risk was elevated and increased monotonically with time during later months of concomitant therapy with a fibrate. Pioglitazone+fenofibrate was associated with an increased risk of severe hypoglycemia as much as 2.3-fold during 60–180d post-initiation of concomitant therapy, and pioglitazone+gemfibrozil as much as 2.6-fold during 30–180d. For rosiglitazone, we found no increased risk of severe hypoglycemia during the first 30d of concomitant use with a statin, but use of rosiglitazone+gemfibrozil was associated with a 1.6-fold increased risk. Subsequently, the risk of severe hypoglycemia peaked during 30–59d—1.8-fold for rosiglitazone+fenofibrate and 2.5-fold for rosiglitazone+gemfibrozil—and returned to the null by 180d.
This is first pharmacoepidemiologic investigation of these potential drug interactions. The presumptive mechanism underlying prior pharmacokinetic- and laboratory science-based work was that fibrates inhibited CYP2C8, the major metabolic pathway for TZDs. Yet, even if inhibition by fibrates significantly raises serum concentrations of TZDs, it is not generally thought that TZDs cause severe hypoglycemia.5 However, we found that severe hypoglycemia occurs at a rate of ~2.5 per 100 p-y among TZD users even in the absence of concomitant insulin or sulfonylureas. Further, Bron et al reported that TZDs are associated with a small but significant increased risk of moderate or severe hypoglycemia, especially within the first year of therapy.6 This raises the possibility that TZDs, while clearly less associated with hypoglycemia than insulin or sulfonylureas, may cause severe hypoglycemia in certain circumstances. That being said, the mechanism seems more complex than elevated TZD serum concentrations caused by CYP2C8 inhibition. Arguments against a lone, major role for CYP2C8 inhibition include: some statins also inhibit CYP2C8,7 yet we did not find elevated HRs for statins; and the inhibition and inactivation of CYP enzymes occurs rapidly,8 yet we generally found delayed rather than rapid-onset increases in the risk of severe hypoglycemia. This latter point could also be explained by the delayed onset of action of TZDs, whose effects may peak at one month.9,10
A more plausible explanation for our findings may be driven by expected actions of fibrates. The PPAR α agonist effects of fibrates beneficially impact lipid and lipoprotein metabolism. Lipid and glucose homeostasis is interrelated11 and lowering free fatty acids ameliorates insulin resistance12 and13 via protection of pancreatic islets14. Alternatively, or in addition, fibrates may induce fatty acid-binding protein and stimulate β-oxidation in skeletal muscles15. Regardless of potential mechanism, improvements in insulin resistance and glycemic control have been reported in users of gemfibrozil16 and fenofibrate13 and 17. Further, some fibrates also act at PPAR γ18, the site of action of TZDs. Of further interest are differences between pioglitazone and rosiglitazone in the time-course of their interaction with fibrates—the former increasing monotonically with time (Fig. 2) and the latter having an inverted U-shape (Supplementary Figure 2). The sustained risk of severe hypoglycemia observed with pioglitazone + gemfibrozil or fenofibrate may be mediated by pioglitazone's more favorable effect on lipids compared to rosiglitazone19; pioglitazone significantly reduces fatty acids and triglycerides20. As discussed above, this may lead to less insulin resistance. The lack of a sustained risk with rosiglitazone + fibrate may be due to a sufficiently weaker interaction to which patients can develop compensatory behaviors or endocrine adaptations over time. Future studies should investigate the relative contributions of these and other potential mechanisms.
See Supplemental Material for a discussion of this work’s strengths and limitations.
5. CONCLUSION
We found that concomitant therapy with a TZD and fibrate is associated with an increased risk of severe hypoglycemia. The mechanism underlying this apparent drug-drug interaction needs further elucidation, but may involve fibrates ‘impact on glucose (i.e., a pharmacodynamic interaction mediated by PPAR α ± γ effects). Clinicians should be attuned to both immediate- and delayed-onset hypoglycemia in their patients on this drug combination.
Supplementary Material
HIGHLIGHTS.
Concomitant use of an antidiabetic thiazolidinedione (i.e., pioglitazone or rosiglitazone) with a fibrate (i.e., gemfibrozil or fenofibrate) was associated with an approximate 1.5- to 2.5-fold increased risk of severe hypoglycemia.
This increased risk generally manifested after the first month of concomitant use.
Severe hypoglycemia occurs in some patients receiving a thiazolidinedione, even without concomitant exposure to insulin or a sulfonylurea.
The apparent drug-drug interaction between thiazolidinediones and fibrates is an underappreciated cause of severe hypoglycemia—a health outcome of interest that is the largest barrier to glycemic control, reduces quality of life, and can lead to major adverse cardiovascular events, dementia or death.
Clinicians should be attuned to both immediate- and delayed-onset hypoglycemia in their patients treated concomitantly with thiazolidinediones and fibrates.
Acknowledgments
Funding / Support: This work was supported by the National Institute on Aging (R01AG025152), the National Institute of Diabetes and Digestive and Kidney Diseases (R01DK102694), and the National Center for Advancing Translational Sciences (UL1TR000003).
This work was supported by the National Institute on Aging (R01AG025152), the National Institute of Diabetes and Digestive and Kidney Diseases (R01DK102694), and the National Center for Advancing Translational Sciences (UL1TR000003). The authors with to thank: Dr. David Margolis (University of Pennsylvania) for his clinical input in the treatment of persons with diabetes; Ms. Cristin Freeman (employed by the University of Pennsylvania at the time of her contribution) and Ms. Margaret Mangaali (University of Pennsylvania) for their project management support; Ms. Min Du (University of Pennsylvania), Mr. Maximilian Herlim (University of Pennsylvania), and Ms. Liu Qing (University of Pennsylvania) for their statistical programming support; and Ms. Geralyn Barosso (University of Minnesota) for her Centers for Medicare and Medicaid Services data support.
Footnotes
CONFLICTS OF INTEREST:
Charles Leonard: none
Xu Han: none
Warren Bilker: none
James Flory: none
Colleen Brensinger: none
David Flockhart: none
Joshua Gagne: JJG was Principal Investigator of a previous investigator-initiated grant from Novartis Pharmaceuticals Corporation to the Brigham and Women’s Hospital for work unrelated to this study.
Serena Cardillo: SC receives research funding from AstraZeneca and Bristol-Myers Squibb.
Sean Hennessy: SH has consulted for AbbVie, AstraZeneca, Bayer Healthcare, LLC and Merck Sharpe & Dohme Corp. He receives research funding from AstraZeneca and Bristol-Myers Squibb. The Center for Pharmacoepidemiology Research & Training, which he directs, receives support for pharmacoepidemiology training from Pfizer and Sanofi.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Jaakkola T, Laitila J, Neuvonen PJ, Backman JT. Pioglitazone is metabolised by CYP2C8 and CYP3A4 in vitro: potential for interactions with CYP2C8 inhibitors. Basic Clin Pharmacol Toxicol. 2006;99(1):44–51. doi: 10.1111/j.1742-7843.2006.pto_437.x. PTOpto_437 [pii] [DOI] [PubMed] [Google Scholar]
- 2.Prueksaritanont T, Richards KM, Qiu Y, et al. Comparative effects of fibrates on drug metabolizing enzymes in human hepatocytes. Pharm Res. 2005;22(1):71–78. doi: 10.1007/s11095-004-9011-5. [DOI] [PubMed] [Google Scholar]
- 3.Gross B, Staels B. PPAR agonists: Multimodal drugs for the treatment of type-2 diabetes. Best Pract Res Clin Endocrinol Metab. 2007;21(4):687–710. doi: 10.1016/j.beem.2007.09.004. [DOI] [PubMed] [Google Scholar]
- 4.Kostapanos MS, Liamis GL, Milionis HJ, Elisaf MS. Do statins beneficially or adversely affect glucose homeostasis? Curr Vasc Pharmacol. 2010;8(5):612–631. doi: 10.2174/157016110792006879. BSP/CVP/E-Pub/000084 [pii] [DOI] [PubMed] [Google Scholar]
- 5.Holstein A, Egberts EH. Risk of hypoglycaemia with oral antidiabetic agents in patients with type 2 diabetes. Exp Clin Endocrinol Diabetes. 2003;111(7):405–414. doi: 10.1055/s-2003-44287. [DOI] [PubMed] [Google Scholar]
- 6.Bron M, Marynchenko M, Yang H, Yu AP, Wu EQ. Hypoglycemia, treatment discontinuation, and costs in patients with type 2 diabetes mellitus on oral antidiabetic drugs. Postgrad Med. 2012;124(1):124–132. doi: 10.3810/pgm.2012.01.2525. [DOI] [PubMed] [Google Scholar]
- 7.Tornio A, Pasanen MK, Laitila J, Neuvonen PJ, Backman JT. Comparison of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors (statins) as inhibitors of cytochrome P450 2C8. Basic Clin Pharmacol Toxicol. 2005;97(2):104–108. doi: 10.1111/j.1742-7843.2005.pto_134.x. [DOI] [PubMed] [Google Scholar]
- 8.Tornio A, Niemi M, Neuvonen PJ, Backman JT. Drug interactions with oral antidiabetic agents: pharmacokinetic mechanisms and clinical implications. Trends Pharmacol Sci. 2012;33(6):312–322. doi: 10.1016/j.tips.2012.03.001. [DOI] [PubMed] [Google Scholar]
- 9.Schwartz S, Raskin P, Fonseca V, Graveline JF. Effect of troglitazone in insulin-treated patients with type II diabetes mellitus. Troglitazone and Exogenous Insulin Study Group. N Engl J Med. 1998;338(13):861–866. doi: 10.1056/NEJM199803263381302. [DOI] [PubMed] [Google Scholar]
- 10.Horton ES, Whitehouse F, Ghazzi MN, Venable TC, Whitcomb RW. Troglitazone in combination with sulfonylurea restores glycemic control in patients with type 2 diabetes. The Troglitazone Study Group. Diabetes Care. 1998;21(9):1462–1469. doi: 10.2337/diacare.21.9.1462. [DOI] [PubMed] [Google Scholar]
- 11.Reasner CA. Reducing cardiovascular complications of type 2 diabetes by targeting multiple risk factors. J Cardiovasc Pharmacol. 2008;52(2):136–144. doi: 10.1097/FJC.0b013e31817ffe5a. [DOI] [PubMed] [Google Scholar]
- 12.Boden G. Interaction between free fatty acids and glucose metabolism. Curr Opin Clin Nutr Metab Care. 2002;5(5):545–549. doi: 10.1097/00075197-200209000-00014. [DOI] [PubMed] [Google Scholar]
- 13.Damci T, Tatliagac S, Osar Z, Ilkova H. Fenofibrate treatment is associated with better glycemic control and lower serum leptin and insulin levels in type 2 diabetic patients with hypertriglyceridemia. Eur J Intern Med. 2003;14(6):357–360. doi: 10.1016/s0953-6205(03)90001-x. S095362050390001X [pii] [DOI] [PubMed] [Google Scholar]
- 14.Lalloyer F, Vandewalle B, Percevault F, et al. Peroxisome proliferator-activated receptor alpha improves pancreatic adaptation to insulin resistance in obese mice and reduces lipotoxicity in human islets. Diabetes. 2006;55(6):1605–1613. doi: 10.2337/db06-0016. 55/6/1605 [pii] [DOI] [PubMed] [Google Scholar]
- 15.Furuhashi M, Ura N, Murakami H, et al. Fenofibrate improves insulin sensitivity in connection with intramuscular lipid content, muscle fatty acid-binding protein, and beta-oxidation in skeletal muscle. J Endocrinol. 2002;174(2):321–329. doi: 10.1677/joe.0.1740321. JOE04709 [pii] [DOI] [PubMed] [Google Scholar]
- 16.Mussoni L, Mannucci L, Sirtori C, et al. Effects of gemfibrozil on insulin sensitivity and on haemostatic variables in hypertriglyceridemic patients. Atherosclerosis. 2000;148(2):397–406. doi: 10.1016/s0021-9150(99)00283-x. S002191509900283X [pii] [DOI] [PubMed] [Google Scholar]
- 17.Wysocki J, Belowski D, Kalina M, Kochanski L, Okopien B, Kalina Z. Effects of micronized fenofibrate on insulin resistance in patients with metabolic syndrome. Int J Clin Pharmacol Ther. 2004;42(4):212–217. doi: 10.5414/cpp42212. [DOI] [PubMed] [Google Scholar]
- 18.Willson TM, Brown PJ, Sternbach DD, Henke BR. The PPARs: from orphan receptors to drug discovery. J Med Chem. 2000;43(4):527–550. doi: 10.1021/jm990554g. [DOI] [PubMed] [Google Scholar]
- 19.Goldberg RB, Kendall DM, Deeg MA, et al. A comparison of lipid and glycemic effects of pioglitazone and rosiglitazone in patients with type 2 diabetes and dyslipidemia. Diabetes Care. 2005;28(7):1547–1554. doi: 10.2337/diacare.28.7.1547. 28/7/1547 [pii] [DOI] [PubMed] [Google Scholar]
- 20.Rajagopalan R, Perez A, Ye Z, Khan M, Murray FT. Pioglitazone is effective therapy for elderly patients with type 2 diabetes mellitus. Drugs Aging. 2004;21(4):259–271. doi: 10.2165/00002512-200421040-00004. 2144 [pii] [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.