Prevention of Type 2 Diabetes Role of Metformin

M. Aline Charles and Eveline EschwègeNational Institute of Health and Medical Research (INSERM) Unit 21, Villejuif, France

Abstract Metformin lowers moderate (nondiabetic) fasting hyperglycaemia in individ-uals at risk for type 2 diabetes without causing hypoglycaemia. In addition, it hasdemonstrated favourable action on several cardiovascular risk factors that are oftenpresent in these individuals: it favours the maintenance of diet-induced weightloss and its associated improvement in fibrinolysis; and it lowers plasma concen-trations of fasting insulin, total and low density lipoprotein-cholesterol, free fattyacids, and of two markers of endothelial damage − tissue plasminogen activatorantigen and von Willebrand factor. These effects together with the good toler-ability profile of the drug position metformin as a first-line agent for theprevention of type 2 diabetes.

REVIEW ARTICLE Drugs 1999; 58 Suppl. 1: 71-730012-6667/99/0001-0071/$01.50/0

© Adis International Limited. All rights reserved.

The prevention of type 2 (non–insulin-depend-ent) diabetes mellitus entails the attenuation ofdiabetic hyperglycaemia and its associated compli-cations. There is epidemiological[1] and thera-peutic[2] evidence that the prevention of diabetichyperglycaemia in turn prevents the developmentof microangiopathy. The situation is quite differentfor macrovascular complications, however.

Patients with type 2 diabetes often present witha cluster of cardiovascular risk factors (visceralobesity, hypertension, high triglyceride and lowhigh density lipoprotein-cholesterol levels, andhypofibrinolysis), all of which form the insulin re-sistance or metabolic syndrome and potentiallycontribute to increased cardiovascular risk. In ad-dition, the insulin resistance syndrome precedes di-abetic hyperglycaemia and cannot be prevented ortreated solely by the lowering of blood glucose lev-els. Metformin, a well known and well toleratedantihyperglycaemic drug with effects on severalmetabolic anomalies associated with the insulin re-sistance syndrome, is a good candidate drug for theprevention of type 2 diabetes.

1. Prevention of Diabetic Hyperglycaemia

Even in nondiabetic hyperglycaemia, met-formin reduces fasting blood glucose levelswithout causing hypoglycaemia. In the BIGPRO(Biguanides and the Prevention of the Risk ofObesity) 1 trial in 324 nondiabetic men and womenwith visceral obesity who were treated for 1 yearwith metformin 850mg twice daily or placebo inaddition to lifestyle recommendations, fastingblood glucose levels decreased significantly in pa-tients who were glucose intolerant at study entrybut not in those who were normoglycaemic.[3]

Any agent used for the prevention of type 2diabetes must maintain a long term glucose-lower-ing effect. This has been shown clearly in theUnited Kingdom Prospective Diabetes Study(UKPDS). Overall, after an initial decrease, fastingblood glucose levels returned to their baseline val-ues after 4 to 6 years, and subsequently increasedsteadily regardless of drug treatment.[4] TheUKPDS showed that, once blood glucose levels havereached 8 mmol/L, the average level at inclusion,

diabetes progresses because of deteriorating β-cellfunction. It is possible, however, that initial treat-ment at a lower fasting blood glucose level willlead to a greater level of β-cell protection.

There are 2 preliminary reports of the effect ofmetformin on the incidence of type 2 diabetes inindividuals with moderate hyperglycaemia. In arandomised trial in 70 individuals with impairedglucose tolerance, a 21% conversion rate was notedin the placebo group versus 3% in the metformin-treated (250mg 3 times daily) group over 1 year(p < 0.05).[5] In a subgroup of individuals withmoderate fasting hyperglycaemia (6 to 7.7mmol/L) in the UKPDS, metformin treatment (at amaximum dosage of 2550 mg/day) was reported toconfer a 50% reduction relative to diet only, in therisk of progression to diabetic hyperglycaemia(≥7.8 mmol/L) over 6 years in obese patients (p <0.004).[6] These results do not allow firm conclu-sions to be drawn, however.

Metformin has been selected for one of the in-tervention arms in the US Diabetes Prevention Pro-gram, an ongoing 3- to 6-year clinical trial to studythe prevention of type 2 diabetes in individualswith impaired glucose tolerance. This trial will pro-vide a more accurate assessment of the protectiveeffect of metformin treatment and may address thequestion of the effect of early treatment of bloodglucose elevation on the rate of degradation of β-cell function.

2. Prevention of CardiovascularDisease in Individuals at Risk of Type 2 Diabetes

Recent epidemiological studies in patients withdiabetes have confirmed that chronic hyperglycae-

mia is an independent risk factor for cardiovasculardisease.[7,8] Because of its antihyperglycaemic ef-fect, metformin may therefore have a potential ef-fect on the prevalence of cardiovascular disease (asmay other antihyperglycaemic agents). However,the insulin resistance syndrome which precedes thedevelopment of hyperglycaemia may be the maincause of high cardiovascular risk in individualswith type 2 diabetes.[9] Metformin is an insulinsensitiser[10] and has shown beneficial effects onseveral cardiovascular risk factors in individuals atrisk of type 2 diabetes.

In the BIGPRO 1 trial in individuals with vis-ceral obesity,[3,11] those treated with metforminshowed greater weight loss (p < 0.06), a greater de-crease in fasting insulin levels (p < 0.06) and asmaller increase in low density lipoprotein (LDL)-cholesterol concentrations (p < 0.07) than thosewho received placebo. In this trial, the effect ofmetformin was most notable on levels of 2 markersof endothelial damage, tissue plasminogen activa-tor (t-PA) antigen (p < 0.02) and von Willebrandfactor (p < 0.02).[11] Bodyweight loss was associ-ated with a significant decrease in plasminogen ac-tivator inhibitor-1 (PAI-1) activity. Metformin byitself did not improve fibrinolysis in addition towhat could be expected from its weight loss-pro-moting effect,[10] and did not show any significanteffect on fasting triglyceride levels or blood pres-sure. The results of BIGPRO 1 were confirmed bythe BIGPRO 1.2 trial in men with visceral obesity,hypertension and high fasting triglyceride levels(unpublished data). A subgroup analysis of the twoBIGPRO trials suggests that metformin may lowerblood pressure only in hyperglycaemic individuals(unpublished data).

Table I. Effect of metformin in individuals at risk of type 2 diabetes

Metformin effect High risk condition

hypertension/dyslipidaemia visceral obesity hyperglycaemia

Specific Increase in diet-induced weightloss and decrease in PAI-1

Decrease in blood glucoselevels and (?) blood pressure

Common to all conditions Decrease in total cholesterol,LDL-cholesterol, free fatty acids,t-PA ag and insulin levels

LDL = low density lipoprotein; PAI-1 = plasminogen activator inhibitor-1; t-PA ag = tissue plasminogen activator antigen.

72 Charles & Eschwège

© Adis International Limited. All rights reserved. Drugs 1999; 58 Suppl. 1

In 3 recent placebo-controlled clinical trialsin nondiabetic patients with hypertension,[12-14]

metformin significantly decreased fasting insulinor C-peptide levels,[12-14] total cholesterol, LDL-cholesterol or apolipoprotein B levels,[12,13] andfasting free fatty acid[13] and t-PA antigen[13] levels.Metformin had a significant effect on blood pressureand fasting triglyceride levels in one study only.[12]

In combined hyperlipidaemia, metformin sig-nificantly reduced total and LDL-cholesterol butnot fasting triglyceride levels.[15] The effect ofmetformin on LDL-cholesterol levels was found tobe additive to that of lovastatin plus diet in patientswith hypercholesterolaemia and coronary heartdisease.[16]

In obese women with polycystic ovary syn-drome (who are frequently insulin resistant[17] andat risk of developing type 2 diabetes[18]), metfor-min decreased fasting levels of insulin and the areaunder the curve of plasma insulin level versus timeduring an oral glucose tolerance test. In contrast, therewas no significant change in the placebo group.[17]

The potentially preventive effects of met-formin in type 2 diabetes are summarised in tableI. Decreases in total and LDL-cholesterol, freefatty acid, t-PA antigen and insulin levels may beobtained with metformin whatever the condition(hypertension, dyslipidaemia, visceral obesity orhyperglycaemia). Conversely, weight loss and ac-companying decreases in PAI-1 levels should beexpected only in obese patients and will be en-hanced by concomitant dietary therapy. Effects onblood glucose levels and possibly blood pressureare seen in hyperglycaemic patients only.

Metformin responds to the pre-diabetic and pre-macroangiopathic anomalies that make up the in-sulin resistance syndrome through a variety of syn-ergistic beneficial effects. Because of theseproperties and the good tolerability profile of thedrug, metformin can be considered a first-lineagent for the prevention of type 2 diabetes.

References1. Klein R. Hyperglycemia and micro and macrovascular disease

in diabetes. Diabetes Care 1995; 18: 258-682. UK Prospective Diabetes Study (UKPDS) Group. Intensive

blood-glucose lowering control with sulphonylureas or insu-

lin compared with conventional treatment and risk of compli-cations in patients with type 2 diabetes (UKPDS 33). Lancet1998; 352: 837-53

3. Fontbonne A, Charles MA, Juhan-Vague I, et al. The effect ofmetformin on the metabolic abnormalities associated withupper body fat distribution. Results of the BIGPRO 1 trial.Diabetes Care 1996; 19: 920-6

4. Turner R, Cull C, Holman R, et al. United Kingdom ProspectiveDiabetes Study 17: a 9-year update of a randomized, control-led trial on the effect of improved metabolic control on com-plications in non-insulin-dependent diabetes mellitus. AnnIntern Med 1996; 124 (1 Pt 2): 136-45

5. Chun-Lin L, Ju-Min L, Chang-Yu P, et al. Effect of metforminon impaired glucose tolerant patients. Diabetologia 1997; 40Suppl 1: A309

6. Holman RR, Cull CA, Turner RC. Six-year randomised con-trolled trial of diet, sulphonylurea, insulin and metformintherapy in 270 subjects at risk for diabetes. Diabetes 1995; 44Suppl. 1: 57A

7. Turner RC, Millns H, Neil HAW, et al. Risk factors for coronaryartery disease in non-insulin dependent diabetes mellitus:United Kingdom prospective diabetes study (UKPDS: 23).BMJ 1998; 316: 823-8

8. Laakso M. Glycemic control and the risk for coronary heartdisease in patients with non-insulin-dependent diabetes mel-litus. Ann Intern Med 1996; 124 (1 Pt 2): 127-30

9. Haffner SM, Stern MP, Hazuda HP, et al. Cardiovascular riskfactors in confirmed prediabetic individuals: does the clockfor coronary heart disease start ticking before the onset ofclinical diabetes? JAMA 1990; 263: 2893-8

10. Bailey C, Turner RC. Metformin. N Engl J Med 1996; 334:574-9

11. Charles MA, Morange P, Eschwège E, et al. The effect of weightchange and of metformin on fibrinolysis and von Willebrandfactor in obese non diabetic subjects. Diabetes Care. In press

12. Giugliano D, De Rosa N, Di Maro G, et al. Metformin improvesglucose, lipid metabolism, and reduces blood pressure in hy-pertensive, obese women. Diabetes Care 1993; 16: 1387-90

13. Landin K, Tengborn L, Smith U. Metformin and metoprolol CRtreatment in non-obese men. J Intern Med 1994; 235: 335-41

14. Dorella M, Giusto M, Da Tos V, et al. Improvement of insulinsensitivity by metformin treatment does not lower blood pres-sure of nonobese insulin-resistant hypertensive patients withnormal glucose tolerance. J Clin Endocrinol Metab 1996; 81:1568-74

15. Pentikäinen PJ, Voutilainen E, Aro A, et al. Cholesterol lower-ing effect of metformin in combined hyperlipidemia: placebocontrolled double blind trial. Ann Med 1990; 22: 307-12

16. Carlsen SM, Rossvoll O, Bjerve KS, et al. Metformin improvesblood lipid pattern in nondiabetic patients with coronary heartdisease. J Intern Med 1996; 239: 227-31

17. Nestler JE, Jakubowicz DJ. Decreases in ovarian cytochromeP450c17α activity and serum free testosterone after reductionof insulin secretion in polycystic ovary syndrome. N Engl JMed 1996; 335: 617-23

18. Dahlgen E, Janson PO, Johansson S, et al. Women with poly-cystic ovary syndrome wedge resected in 1956 to 1965: a longterm follow-up focusing on natural history and circulatinghormones. Fertil Steril 1992; 57: 505-13

Correspondence and reprints: Dr M. Aline Charles, NationalInstitute of Health and Medical Research (INSERM) Unit 21,16, avenue Paul Vaillent-Couturier 94807, Villejuif, France.E-mail: charles@vjf.inserm.fr

Metformin in Prevention of Type 2 Diabetes 73

© Adis International Limited. All rights reserved. Drugs 1999; 58 Suppl. 1