The long-awaited results from the United Kingdom Prospective Diabetes Study (UKPDS) were published in 1998 (UKPDS Study Group 1998a, 1998b, 1998c, 1998d). This randomised, controlled trial of people newly diagnosed with type 2 diabetes refuted the suggestion that treatment with either sulphonylureas or insulin therapy was associated with increased cardiovascular risk (UKPDS Study Group 1998a). Subjects in the main study were randomised either to conventional treatment with diet or to intensive treatment with either sulphonylurea or insulin therapy. There was a significant reduction in HbA1c in the intensive treatment group and a 25% risk reduction in microvascular endpoints (neuropathy, nephropathy and retinopathy and need for laser treatment). However, despite small reductions in macrovascular endpoints, such as myocardial infarction (MI), these results did not reach statistical significance. These people have been followed for a few years since the end of the UKPDS and the reduction in MI with intensive treatment based on sulphonylureas or insulin therapy is now statistically significant.

Interestingly, a substudy from the UKPDS group in overweight individuals indicated that treatment with metformin led to a reduction in cardiovascular events, including MIs and diabetes-related deaths (UKPDS Study Group 1998b). Because MI was such a common cause of death in these individuals, the reduction in MIs was also mirrored by a reduction in total mortality. This unexpected finding was not fully anticipated by the UKPDS investigators and the results remain difficult to explain. The number of people receiving metformin was small (n = 342) and this might have exaggerated the degree of benefit. The reductions in events could not be explained on the basis of reductions in HbA1c, suggesting that metformin might have extended benefits that were not directly related to glucose-lowering effects of the drug. Metformin has a relatively minor effect in reducing insulin resistance at the level of the liver, and this is one possible explanation.

The results of the Diabetes Control and Complications Trial (DCCT), a multicentre trial, which randomised 1141 people with type 1 diabetes to usual care or intensive insulin therapy, mirrored to some extent the findings from the UKPDS Study. There was a significant reduction in microvascular endpoints in the intensive treatment group compared to the conventionally treated group (DCCT Research Group 1995). Although there was a trend towards a reduction in cardiovascular disease in the intensive treated group, again this did not reach statistical significance. This might be due to the small number of events that actually occurred throughout the period of follow-up (40 in the control group; 23 in the treatment group), which is attributable to the young age of the population studied (13–39 years). A subsequent meta-analysis, including data from DCCT and several smaller studies, demonstrated a significant reduction in the number of macrovascular events, but no significant effect on the number of people developing macrovascular disease, or on macrovascular mortality (Lawson et al 1999).

Therefore, although the results from the UKPDS and DCCT trials were beneficial in so much as they proved that treatment with insulin or sulphonylureas reduced microvascular events and did not lead to an increase in cardiovascular disease, to date, no study has demonstrated that intensive management of hyperglycaemia reduces the risk of CHD. However, in addition to the evidence that metformin might reduce the incidence of macrovascular complications, the thiazolidinediones, a new group of oral agents, have also been shown to have extended benefits beyond glucose lowering, including favourable affects on blood pressure, lipids and microalbuminuria, as well as reductions in novel cardiovascular risk markers such as C-reactive protein (CRP) (Haffner et al 2002). Whether this translates into clinical benefit in terms of a reduction in cardiovascular events is currently being tested in large studies with rosiglitazone and pioglitazone, respectively.

Three major studies published in the 1990s, in which more than 15 000 people participated, looked at the issue of secondary prevention of atherosclerotic CVD using statin therapy: the Scandinavian Simvastatin Survival Study (4S) (Scandinavian Simvastatin Survival Study Group 1994), the Cholesterol and Recurrent Events (CARE) Trial (Sacks et al 1996) and the Long-term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study (LIPID Study Group 1998).

Data from the 4S and CARE, which were conducted in the US and Canada, and from the LIPID Study, which was carried out in Australia and New Zealand, were consistent. Reducing LDL and total cholesterol is associated with reductions in all-cause mortality, CHD-related deaths and major coronary events (first MI, unstable angina). As a result, statin therapy in people with a history of coronary heart disease is normal practice.

Also in the 1990s, two large studies also examined the issue of the role of statins in the primary prevention of coronary heart disease. Data from the West of Scotland Coronary Prevention Study (WOSCOPS) and the Airforce/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS) suggested that statin therapy was also of benefit in people with no history of CVD (Shepherd et al 1995, Downs et al 1998). People with a wide range of cardiovascular risk were included in these studies and all people appeared to benefit. Few people with diabetes were included in these studies so no meaningful analysis could be made of the data from these small subgroups.

However, despite the large numbers of people in these five studies, relatively few individuals with diabetes were included. A total of 2000 people with diabetes were included in these five cohorts and, until recently, clinical practice was based on data from these small diabetic subgroups. Over the last few years, however, further studies have been published that have greatly enhanced our knowledge of the effects of lipid-lowering in the population of people with diabetes.

Some believe that everyone with type 2 diabetes should be treated as if they had pre-existing CHD, and one large study based on data from Finland suggested that a person with diabetes was at greater risk of suffering from a coronary event than a person without diabetes who has previously had a similar event (Haffner et al 1998). A similar study carried out in Tayside, however, has suggested that, while certainly at higher risk of CHD than the general population, people with diabetes without known cardiovascular disease are not more likely to suffer an event than a person who has pre-existing cardiovascular disease (Evans et al 2002). Several recent guidelines suggest that all people with diabetes are treated as a CHD equivalent, and so should be given a statin. Although all people with diabetes would gain from the widespread use of statins, the issues of cost and compliance need to be considered carefully before adopting such a ‘blanket’ prescribing policy.

A further issue that remains unresolved is the use of statins in people with type 1 diabetes. This group has a much higher incidence of CHD than their counterparts who do not have diabetes. Despite this, however, diabetes is usually diagnosed at a young age when the absolute incidence of CHD is negligible. The benefit of statin therapy in type 1 diabetes has not been proven given the very small numbers of these people included in statin trials. The timing of initiation of therapy also remains unresolved in people with type 1 diabetes, and there is still a need for further information regarding the benefits of statins in people with diabetes but who do not have CHD, particularly people with type 1 diabetes and younger people with type 2 diabetes.

A further cloudy issue in the management of dyslipidaemia is how far cholesterol should be lowered. Although the Heart Protection Study provided some indication as to target levels of cholesterol it was not specifically designed to identify this. A recent novel analysis of the reduction of CHD events with all forms of lipid-lowering therapy (Brady & Betteridge 2003) has shown that the lower the cholesterol the greater the reduction in CHD events, and the same is true for an analysis of the reduction of CHD events in people with diabetes. The Pravastatin or Atorvastatin Evaluation and Infection Therapy (PROVE-IT) Study (Cannon et al 2004) has demonstrated that high-dose atorvastatin is better than a less efficacious dose of pravastatin in reducing major cardiovascular events or death in patients with acute coronary syndromes; 18% of the cohort had diabetes but the reduction in diabetic subjects did not reach statistical significance.

Fibrates are also available as lipid-lowering therapy. They cause a reduction in serum triglycerides and elevations in HDL cholesterol with little impact on LDL cholesterol. Given that the major lipid abnormalities seen in people with diabetes are a reduction in HDL cholesterol and elevated serum triglycerides, it would seem likely that fibrates would be a more reasonable choice of lipid-lowering therapy in the person with diabetes. A randomised double-blind placebo-controlled trial has studied the effect of fibrates on progression of coronary artery disease in people with type 2 diabetes (Diabetes Atherosclerosis Intervention Study Investigators 2001) and demonstrated a 28% reduction in triglycerides and 6% increase in HDL cholesterol, with no significant change in LDL cholesterol. There was a significant reduction in coronary artery plaque progression in the fenofibrate-treated group compared to the placebo group. There was also a reduction in clinical endpoints in the treated group, although the study did not have statistical power to analyse these data. This is currently being examined in a larger randomised, prospective study of fenofibrate in people with diabetes examining harder cardiovascular end points.

The Veterans Affairs High-density lipoprotein cholesterol Intervention Trial (VA-HIT) Study Group carried out a randomised double-blind placebo controlled trial looking at the effect of treatment with gemfibrozil in 2500 men with established CHD and reduced HDL cholesterol and normal levels of LDL cholesterol (Rubins et al 1999). This cohort had a large diabetes subgroup. The study demonstrated that HDL cholesterol increased and triglycerides were significantly reduced in the gemfibrozil treated group. There was also a 22% relative risk reduction in primary endpoints in the treatment group and a 41% relative risk reduction in CHD related death in the subgroup of people with diabetes.

Although these studies look promising, larger studies looking at the effects of fibrates on clinical endpoints are necessary. No large study has looked at the combined effect of a fibrate and a statin and, although this might seem like the ideal combination in the person with diabetes, given the typical lipid profile, there are concerns regarding the safety of these agents in combination in terms of rhabdomyolysis.

Ezetimibe is a novel lipid-lowering agent that acts by selectively inhibiting cholesterol absorption without disrupting the absorption of fat soluble vitamins. It reduces LDL cholesterol by 15–20% and is mainly used in people who are intolerant of statins (especially at higher doses), or as combination therapy in people who are failing to reach target serum cholesterol levels despite high dose statins.