May be high risk, low benefit

The problems with a glucocentric approach to managing patients with type 2 diabetes^{1 2 3} can be seen using data from the UKPDS.⁴ In the 10 years of follow-up of newly diagnosed patients with a mean age of 53, macrovascular events (myocardial infarction and stroke) were five times more common than serious microvascular events (blindness in one eye and renal failure) but, unlike these, were not significantly reduced by intensive glucose lowering. Moreover the observational data from the study² showed a substantially less steep relation of mean concentrations of HbA_1c levels over 10 years with macrovascular risk than with microvascular risk. These data imply that if glycaemia per se has a role in the aetiology of macroangiopathy, the maximal potential benefit from a 1% reduction in HbA_1c is 14% for myocardial infarction and 12% for stroke. Intervention studies with statins and antihypertensives have shown benefits of around twice these amounts.⁵ A qualitative difference for the patient also exists between regimens based on tablets and those based on injections (perhaps multiple) and blood glucose monitoring, as well as the additional risk of hypoglycaemia.

The other value of UKPDS data is the possibility of calculating numbers needed to treat. In that study, the 10 year risk of macrovascular disease was 22%, around four times that of the control group in the ACCORD study.¹ If this figure is combined with the epidemiological data,² the maximal potential benefit of lowering HbA_1c by 1% in a 53 year old patient with type 2 diabetes would be a reduction of 3.1% over 10 years (14% of 22%). This implies that for every 32 people treated for 10 years with an intensive glucose lowering regimen, at least 31 would have exactly the same outcome, whether or not they were using the regimen. While absolute cardiovascular risk, and so potential benefit, increases with age, the same is true for the potential risks of adverse consequences of hypoglycaemia.

One hopes that the ACCORD study will inject a note of caution before conflating blood glucose with cholesterol and blood pressure as cardiovascular risk factors worthy of aggressive intervention. Even if ADVANCE and other intensive glucose lowering trials prove positive, informed choice should require that patients be provided with full explanations of the likely level of benefit expressed as absolute, and not relative, risk reduction.