Treatment of Type 2 Diabetes with a Combination of Two Insulin Sensitizers — Increased Efficacy and Fewer Side-effects

Register or Login to View PDF Permissions
Permissions× For commercial reprint enquiries please contact Springer Healthcare:

For permissions and non-commercial reprint enquiries, please visit to start a request.

For author reprints, please email
Average (ratings)
No ratings
Your rating
Copyright Statement:

The copyright in this work belongs to Radcliffe Medical Media. Only articles clearly marked with the CC BY-NC logo are published with the Creative Commons by Attribution Licence. The CC BY-NC option was not available for Radcliffe journals before 1 January 2019. Articles marked ‘Open Access’ but not marked ‘CC BY-NC’ are made freely accessible at the time of publication but are subject to standard copyright law regarding reproduction and distribution. Permission is required for reuse of this content.

Thiazolidinediones (TZD) and metformin lower cardiac risk factors, as do serum glucoses, and are therefore the best choice for initial therapy of type 2 diabetes. To choose between metformin and a thiazolidinedione is difficult because of the many benefits of both drugs. In most cases, a combination of metformin and a TZD is chosen, and this article describes the rationale for this decision.

The basic principle of combination therapy is that with smaller doses of two drugs there is greater efficacy and fewer side-effects than with a large dose of either drug used as monotherapy. In the case of the addition of metformin to a TZD or vice versa, because of their different sites of action (liver with metformin and muscle with the TZD), it will result in a decrease in the hemoglobal (Hb) A1c that will be greater than that achieved by monotherapy with a larger dose of either drug. Perhaps of more importance is that, at lower doses, the side-effects of metformin (anorexia, nausea, and diarrhea) and TZDs (weight gain, edema, and dilutional anemia) are much less. Around 30% of patients initiated on metformin will develop gastrointestinal symptoms,and 3% will have to discontinue the drug. Of the remaining 27%, the majority will only tolerate metformin at less than the maximum dose of 2,500mg daily (usually one gram or less).

Both metformin and TZDs lower cardiac events. This is important with type 2 diabetes, which has been described as a cardiac condition characterized by hyperglycemia and is a cardiac risk equivalent i.e., a 20% chance of a cardiac event over 10 years. Metformin lowers total low-density lipoproteins (LDL) and triglycerides, and TZDs variably lower triglycerides and raise total high-density lipoprotein (HDL). Of more importance is that by suppressing the activity of hepatic lipase, TZDs have been shown to increase the size of both the HDL and LDL particles.1The larger LDL particle is less atherogenic and the larger HDL particle (HDL2) is the only HDL particle, which is cardioprotective. Furthermore, the smaller HDL3 particle in addition to being less cardioprotective, is also more easily broken down by the liver, accounting for the low total HDL seen with insulin resistance. Therefore, insulin sensitizers used in combination can, even at lower doses by their effects on the lipid profile, provide a further lowering in cardiac risk and are complementary to the statins which most type 2 diabetic patients will be or should be utilizing.2

Endothelial dysfunction is improved by both TZDs and metformin. When the endothelium is healthy, nitric oxide is produced, vasodilatation occurs, and the endothelium is resistant to clot formation and penetration by monocytes to start the process of atherogenesis. When insulin resistance is present, superoxides are produced, nitric oxide activity is effectively quenched, and endothelial dysfunction occurs with vasoconstriction, clotting, and atheroma formation.3 Because of its association with endothelial dysfunction in the Paris Protective and the Quebec Heart Studies, the presence of insulin resistance has been shown to be an independent risk factor for cardiac events and increases the risk of a cardiac event in the non-diabetic insulin-resistant subject by over five times.4,5

With lowering of insulin resistance, vasodilatation occurs accompanied by a small but significant decrease in blood pressure.6 Furthermore, with less endothelial damage, there are reductions in the levels of both Von Willebrand factor and homocystein. Albuminuria is an easily assessed manifestation of endothelial dysfunction since the glomerulus is an arteriole, and with the decrease in permeability of that arteriole, which occurs with insulin sensitization and improved endothelial function, albuminuria decreases beyond that which would be expected from lowering of the glucose alone.6,7

With both insulin resistance and poor glycemic control, Plasminogen Activator Inhibitor (PAI1) levels are increased. Elevation of PAI1 is associated with a decrease in the conversion of plasminogen to plasmin, a slowing of fibrinogen and clot breakdown and higher serum fibrinogen levels and increased thromboembolic events. However, of more importance is that with high PAI1 levels in the vessel wall, the ensuing low plasmin levels are associated with decreased removal of collagen from the atheromatous plaque, which is needed to facilitate the entry of vascular smooth muscle cells into the plaque so that the plaque can be stabilized.8 Therefore, insulin resistance due to high PAI1 levels is associated not only with an increased frequency of thromboembolism, but with unstable atheromatous plaques and increased cardiovascular events. By reducing insulin resistance, both metformin and TZDs have been shown to lower PAI1 levels, which can be further lowered by improved glycemic control and this combination should lead to less cardiac events.9,10

The presence of a high serum C-reactive protein level is associated with both insulin resistance and unstable atheromatous plaques. An autopsy study of coronary arteries showed that C-reactive protein was only present in the segments of the artery that had atherosclerotic lesions and was present even at the very earliest stages of atherogenesis. A specific receptor for C-reactive protein (CRP) resides on the monocyte, and activation of this receptor promotes migration of the monocytes into the arterial wall with activation of complement. Therefore, C-reactive protein is not only a marker of inflammation within a plaque but also plays a very active role in the formation of the atheromatous plaque.11 While lowering of the C-reactive protein occurs with statins, aspirin, non-steroidal anti-inflammatory drugs (NSAIDs), selective estrogen receptor modulators (SERMs) and TZDs, the lowering of the C-reactive protein that occurs with TZDs is two- to three-fold greater than that which occurs with statins, and therefore could be associated with a greater decrease in the frequency of rupture of the atheromatous plaque and a decrease in cardiac events.12

In addition to being anti-inflammatory, TZDs are also anti-proliferative. An example of this activity can be seen with the vascular smooth muscle cells (VSMCS), which by their proliferation and migration are responsible not only for the atheromatous plaque, but for the restenosis that, especially in diabetic patients, often follows the endothelial injury caused by angioplasty and stent placement. Studies in animals have shown a decrease in the proliferation and migration for the VSMCs with TZDs, and studies in diabetic humans have shown a decrease in intimal-medial thickening in the carotid arteries (the first stage of atherosclerosis), and diabetic patients placed on a TZD following angioplasty and placement of a 'bareÔÇÖ stent have been shown to have less restenosis.13-15 Therefore, TZDs by their anti-proliferative action have the potential to lower cardiac events directly in addition to their ability to lower cardiac events through the lowering of the risk factors associated with the insulin resistance syndrome.

Another reason to utilize TZDs from the onset of diabetes is that they have the potential to rejuvenate the pancreatic beta cell. Destruction of the pancreatic beta cell in type 2 diabetes starts as long as 12 years before the diagnosis of diabetes is made and is due to accelerated beta cell apoptosis (programmed cell death), caused by an increased fat content in the beta cell, which increases with age more in the beta cell than in the alpha cells of the islets or the pancreatic duct cells.16

An increase in beta cell free fatty acids (FFAs) from this intracellular triglyceride reservoir in which they are stored results through the metabolite ceramide in stimulation of nitric oxide synthase activity, toxic nitric oxide levels, which combine with superoxide to form peroxinitrite. This accelerates beta cell apoptosis.17 TZDs remove FFAs from the beta cell and in animal studies TZDs have been shown to rejuvenate and regranulate the pancreatic beta cell and increase endogenous insulin production.18,19 In humans there is evidence that accelerated beta cell apoptosis is also reversed and beta cell rejuvenation and an increase in endogenous insulin production that does not occur with metformin or sulfonylureas or placebo occurs with the TZDs. In humans this is assessed indirectly utilizing the HOMAB method. A more accurate method is to assess the pro-insulin: insulin ratio. The failing beta cell secretes more of the insulin precursor pro-insulin so that the pro-insulin:insulin ratio elevates. This elevation is reversed with TZD therapy, but not with placebo or sulfonylureas, where the pro-insulin:insulin ratio continues to increase. This indicates that an improvement in beta cell function has occurred with TZDs, but not with other agents.20 A semi-prospective case control study has shown a 62% increase in stimulated endogenous insulin production with the addition of troglitazone and later rosiglitazone to the regimen of patients failing the combination of a sulfonylurea and metformin. In the comparison group, there was no change in endogenous insulin production when metformin was added to the regimen of patients carefully matched for age, body mass index (BMI), and A1cs before and after this addition who were failing a sulfonylurea.21 Furthermore, after three years, 74% of these triple therapy patients had maintained excellent glycemic control (average A1c 6.9%) due to a significant increase in endogenous insulin production whereas the 26% who failed to maintain glycemic control with triple oral therapy failed to increase their endogenous insulin production and needed the addition of insulin to their regimen.22 After five years, 62% of these subjects were well controlled with an average A1c of 7.1%, and after six years 51% were controlled with an A1c of 6.9%.23 This prospective study illustrates the longevity of glycemic control, which is undoubtedly due to beta cell preservation, rather than lowering of insulin resistance or glycemic control with the TZDs, and is the best data available on longevity of TZD action at this time. The probability is that in the group who failed triple oral therapy, TZDs were not started early enough to be able to rejuvenate the pancreatic beta cell.

This suggests that TZDs should be utilized from the time of diagnosis of type 2 diabetes. A prospective randomized study of the addition of either a TZD or insulin to the regimens of similar patients failing a combination of metformin and a sulfonylurea showed that with the addition of insulin there was no change in endogenous insulin production, but with the addition of a TZD endogenous insulin, production increased. Furthermore, the first phase insulin response (the first defect in insulin secretion to occur in both type 1 and type 2 diabetes) returned.24

It is now known from the UK Prospective Diabetes Study (UKPDS) that there is no glycemic threshold for diabetic macrovascular complications and that the lower the HbA1c, the lower the prevalence of cardiovascular events.25 It is also known from a population study that, in non-diabetic subjects whose HgbA1cs are within the normal range, the lower the A1c, even within the normal range, the lower the incidence of cardiac events.26 Therefore, in type 2 diabetic patients, an A1c that is within the normal range should be aimed for. Unfortunately, obtaining this level of glycemic control with insulin secretagogues or insulin will result in both an increased frequency and severity of hypoglycemic events. Fortunately, with a combination of metformin and a TZD, hypoglycemia is rare, severe hypoglycemia extremely rare, and lower A1cs are more easily obtained.

With a TZD, weight gain due to new adipocyte formation, and occasionally due to edema, occurs. However, with metformin, weight stabilization or even weight loss due to appetite suppression will often occur even when sulfonylureas or insulin in addition to a TZD are being utilized.27,28 Therefore, the weight gain that occurs with a TZD can be minimized by utilizing the TZD in combination with metformin.

The combination of a TZD and metformin is the ideal combination to be utilized in the type 2 diabetic patient, and the most advantageous time to utilize this combination is at the time of diagnosis of type 2 diabetes, when the availability of endogenous insulin is at its peak, so that insulin replacement or stimulation of the pancreatic beta cells to release more insulin with a secretagogue is not necessary. Smaller doses of metformin and TZDs used in combination are more efficacious and have fewer side-effects. In addition, due to the minimal risk of hypoglycemia, lower A1cs that are associated with less cardiac events can be reached, and when metformin is on board, weight gain with a TZD is minimized. Furthermore, if two compounds are combined in the same tablet, there is often an improvement in patient compliance and a decrease in cost.


  1. Brunzell J D and Hokarson J E,“Dyslipidemia of central obesity and insulin resistance”, Diabetes Care (1999), 22 (suppl 3) pp. C10–13.
  2. Ovalle F and Bell D S H,“Differing effects of thiazolidinediones on LDL and HDL subfractions and Lp(a)”, Diabetes (2001), 50 (suppl 2) A4530454 and A461–462.
  3. Steinberg H O, Chaker H, Leaming R, et al.,“Obesity/insulin resistance is associated with endothelial dysfunction: implications for the syndrome of insulin resistance”, J. Clin. Ivest. (1996), 97: pp. 2,601–2,610.
    Crossref | PubMed
  4. Fontbonne A M and Eschwege E M,“Insulin and cardiosvascular disease. Paris Prospective Study”, Diabetes Care (1991),14: pp. 461–469.
    Crossref | PubMed
  5. Despres J P, Lamarche B, Maurcege P, et al.,“Hyperinsulinemia as an independent risk factor for ischemic heart disease”, N. Eng. J. Med. (1996), 334: pp. 952–957 .
    Crossref | PubMed
  6. DeFronzo R A and Ferannini E, “Insulin resistance: a multifaceted syndrome responsible for NIDDM, obesity, hypertension, dyslipidemia, and atherosclerotic cardiovascular disease”, Diabetes Care (1991), 14: pp. 173–194
    Crossref | PubMed
  7. Weston W M, Heise M A, Porter L E, Bakris G,Viberti G and Freed M I,“Rosiglitazone mediated reductions in urinary albumin excretion are associated with changes in ambulatory blood pressure in type 2 diabetes patients”, Diabetes (2001), 50 (Suppl2) p. A134
  8. Sobel B E, “The potential influence of insulin and plasminogen activator type 1 on the formation of vulnerable atherosclerotic plaques associated with type 2 diabetes”, Proc.Assoc.Am. Physicians (1999), 11: pp. 313–318.
    Crossref | PubMed
  9. Chu W V, Kim D D, Kong A P S, et al.,“Differential effects of metformin and troglitazone on cardiovascular risk factors in patients with type 2 diabetes mellitus”, Diabetes (2000), 49 (suppl 1) p.A1-1.
  10. McGill J B, Schneider D J,Arfken C L, Lacore C L and Sobel B E,“Factors responsible for impaired fibrinolysis in obese subjects and NIDDM patients”, Diabetes (1994), 43: pp. 104–109. .
    Crossref | PubMed
  11. Torzewski M, Rist C, Mortensen R F, Zwaka T P, Bienek M,Waltenberger J, Konig W, Schmitz G, Hombach V and Torzewski J, “C-reactive protein in arterial intima: role of C-reactive protein receptor-dependent monocyte recruitment in atherogenesis”, Arteriosclerosis,Thrombosis, and Vascular Biology (2000), 10: pp. 2,094–2,099.
    Crossref | PubMed
  12. Moharty P, Aljada A, Ghanim H,Tripathy D, Syed T, Hofmeyer D and Dandora P, ÔÇ£Rosiglitazone improves vascular reactivity, inhibits oxygen species (ROS) generation, reduces P47 Phox subunit expression in mononuclear cells (MNC) and reduces Creactive protein (CRP) and monocyte chemotactic protein 1 (MCP-1): evidence of a potent anti-inflammatory effectÔÇØ, Diabetes (2001), 50 (suppl2) p.A68.
  13. Law R E, Goetze S, Xi V P et al., “Expression and function of PPAR gamma in rat and human smooth muscle cells”, Circulation (2000), 101: pp. 1,311–1,318.
  14. Minamikawa J,Tanaka S,Yamauchi M, et al.,“Potent inhibitory effect of troglitazone on carotid arterial wall thickness in type 2 diabetes”, J. Clin. Endocrinol. Metab. (1998), 83: pp. 1,818–1,820.
    Crossref | PubMed
  15. Takagi T,Akasaka T,Yamamuro A, et al.,“Troglitazone reduces neointimal tissue proliferation after coronary stent implantation in patients with non-insulin dependent diabetes mellitus: a serial intravascular ultrasound study”, J.Am. Coll. Cardiol. (2000), 36: pp. 1,529–1,535.
    Crossref | PubMed
  16. Bell D S H, “Tissue triglyceride levels in type 2 diabetes and the role of thiazolidinediones in reversing the effects of tissue hypertriglyceridemia: a review of evidence in animals and humans”, Endocrine Practice (2001), 7: pp. 135–138.
    Crossref | PubMed
  17. Shimabukuro M, Ohneda M, Lee Y, Unger R H, “Role of nitric oxide in obesity-induced beta cell disease”, J. Clin. Invest. (1997), 100: pp. 290–295.
    Crossref | PubMed
  18. Shimabukuro M, Zhou Y T, Lee Y and Unger R H,“Troglitazone lowers islet fat and restores beta cell function of Zucker diabetic fatty rats”, J. Biol. Chem. (1998), 273: pp. 3, 547–3,550.
    Crossref | PubMed
  19. Finegood D T, McArthur M D, Kojwang D,Thomas M J,Topp B B, Leonard T and Buckingham R B,“Beta cell mass dynamic in Zucker diabetic fatty rats: rosiglitazone prevents the rising net cell death”, Diabetes (2001), 50: pp. 1,021–1,029.
    Crossref | PubMed
  20. Porter L E, Freed M I, Jones N P and Biswas N,“Rosiglitazone improves beta cell function as measured by proinsulin/insulin ratio in patients with type 2 diabetes”, Diabetes (2000), 49 (suppl 1) p.A122.
  21. Ovalle F and Bell D S H, “Clinical evidence of thiazolidinedione induced improvement of pancreatic beta cell function”, Diabetologia (2000), 43 (suppl 1) p.A120.
  22. Bell D S H and Ovalle F,“Long-term efficacy of triple oral therapy for type 2 diabetes mellitus”, Diabetes (2001), 50 (suppl 2) p.A106.
  23. Ovalle F and Bell D S H, “Glycemic control and weight after six years of triple oral therapy for type 2 diabetes mellitus”, Diabetes (2004), 53 (suppl2) pp.A478–A479.
  24. Ovalle F and Bell D S H,“Rosiglitazone or insulin added to subjects failing double oral therapy”, Diabetes (2003), 52 (Suppl 1) p.A130.
  25. Stratton I M,Adler A I,Weil H A, et al.,“Association of glycemia with macrovascular and microvascular complications of type 2 diabetes (UKPDS35): prospective observational study”, BMJ (2000), 321: pp. 405–412.
    Crossref | PubMed
  26. Khaw Kay-Tee,Wareham N, Luber R, et al.,“Glycated hemoglobin diabetes and mortality in men in Norfolk cohort of European Prospective Investigation of Cancer and Nutrition”, (EPIC Norfolk) BMJ (2001), 322: pp. 15–18.
    Crossref | PubMed
  27. Bell D S H and Mayo M S,“Weight loss inpatients treated with a metformin sulfonylurea combination in comparison with twice daily mixed insulin”, Endocrine Practice (1998), 4: pp. 360–364.
    Crossref | PubMed
  28. Strowig S M, Aviles-Santa M L and Raskin P, “Improved glycemic control without weigh gain using triple therapy in type 2 diabetes”, Diabetes Care (2004), 27 (7) pp. 1,577–1,583.
    Crossref | PubMed