What is all this about?
In a recent letter to The Lancet, three correspondents express their concern that in several recent Lancet publications “the reaction against “glucocentricity” in the field of diabetes has gone too far”. They refer to the data-focussed commentary relating to self-monitoring of diabetes on the type 2 diabetes section of NPC, which discusses the UK Prospective Diabetes Study (UKPDS) and states “Compared with ‘conventional control’ there was no benefit from tight control of blood glucose with sulphonylureas or insulin with regard to total mortality, diabetes-related death, macrovascular outcomes or microvascular outcomes, including all the most serious ones such as blindness or kidney failure”.
Their letter continues: “In type 2 diabetes, the UKPDS reported that a composite microvascular endpoint (retinopathy requiring photocoagulation, vitreous haemorrhage, and fatal or non-fatal renal failure) was reduced by 25% in patients randomised to intensive glucose control (p=0·0099). To imply that these are not patient-relevant outcomes is to distort the evidence.”
They conclude “Many studies have also found that improved glycaemic control reduces macrovascular complications (ref). Do not be misled: glycaemic control remains a crucial component in the care of people with diabetes”
Is the criticism fair?
The results of UKPDS quoted need further scrutiny. The authors of that paper wrote “Patients assigned intensive treatment had a significant 25% [relative] risk reduction in microvascular endpoints (p=0·0099) compared with conventional treatment— most of which was due to fewer cases of retinal photocoagulation (figure 4): the reduction in risk was of borderline significance for myocardial infarction (p=0·052) and cataract extraction (p=0·046)”. (emphasis ours).
As the Lancet correspondents say, there was a reduction the rate of the composite microvascular endpoint. Composite endpoints are used to improve the power of a clinical study (its ability to detect a difference that really exists), by grouping together several similar endpoints (such as death from coronary heart disease and fatal and non-fatal myocardial infarction in many statin trials). The UKPDS authors say very clearly that most of the 25% relative risk reduction in all microvascular endpoints was due to fewer cases of retinal photocoagulation.
It is important that composite endpoints are composed of things of similar importance to patients. It is questionable whether process measures such as retinal photocoagulation and albuminuria should be grouped with outcomes of direct important to patients such as visual acuity and renal failure. Visual acuity and blindness was not reduced by the intervention to control blood glucose intensively. Perhaps nowadays we would be less ready to accept the microvascular composite outcome used in UKPDS.
Above all, the UKPDS showed conclusively that despite intensive therapeutic efforts there was an inexorable rise in HbA1c over time. Failing to recognise this occurs in many people with type 2 diabetes mellitus may have the unfortunate consequence of distorting management priorities by patients and their health care advisers.
Writing recently in the BMJ, about Quality and Outcomes Framework measures, Iona Heath commented:
“In 1993, the results of the Diabetes Control and Complications Trial (DCCT) documented the potential and substantial benefits of tight control of blood glucose levels in patients with type 1 diabetes. The authors warned explicitly against using their findings in the care of those with type 2 diabetes, but that is precisely what has happened in the intervening years…
The results of the United Kingdom Prospective Diabetes Study (UKPDS), published in 1998, have been used to support the emphasis on tight control of blood glucose levels in type 2 diabetes. However, as McCormack and Greenhalgh have so carefully argued, the study in fact showed no clinically important benefit from the control of blood glucose with sulphonylureas and insulin over more than 10 years…. … The only effective agent was metformin, which is undoubtedly beneficial through mechanisms that seem poorly understood and are not completely explained by the effect on blood glucose levels.
The Lancet correspondents state in their letter “Many studies have also found that improved glycaemic control reduces macrovascular complications”. The reference they cite to support this statement was a meta-analysis of observational studies of the association between and cardiovascular disease in people with diabetes. This found that for every 1-percentage point increase in HbA1c, the relative risk for any cardiovascular event increased by 18% in people with type 2 diabetes mellitus. However, this is very different from showing either that an intervention aimed at delivering a lower HbA1c actually reduced HbA1c in the long term, or , more importantly, that lowering HbA1c delivered the outcome of fewer macrovascular and microvascular events, which is of much more direct importance to patients. Researchers need to demonstrate that patients live longer or better, not just that some biological parameter is improved.
Perhaps there is more evidence from more recent trials using more modern drugs that this laudable objective of improving mortality and morbidity is actually achievable? The PRO-ACTIVE study attempted to obtain good, randomised controlled trial (RCT) evidence of benefit from pioglitazone compared with placebo, in terms of mortality and vascular morbidity. Patients with type 2 diabetes and existing macrovascular disease (but not heart failure) were recruited and the study lasted nearly three years. No benefit was seen for pioglitazone in regard to the primary outcome. A benefit was seen with regard to a secondary outcome not defined in the original methodology (NNT=48). However, the P-value indicates that the possibility that this was a chance finding with approximately the same probability as throwing a double six in a game like Monopoly®, and we all know that happens quite often. There are good statistical reasons for being very cautious before accepting this result. In any event, the patients in PRO-ACTIVE were not receiving optimum care to reduce other cardiovascular risk factors (blood pressure was poorly controlled, many patients were not taking a statin and a significant proportion were smokers). As we have blogged previously, there are serious concerns regarding the cardiovascular safety of rosiglitazone. The benefit versus risk balance of widely using agonists of peroxisome proliferator-activated receptors to intensively control blood glucose looks very doubtful.
The NPC has never suggested that control of blood glucose is not important for people with type 2 diabetes: this can be seen in the material on the type 2 diabetes section of NPC. Such people’s most immediate problem is the mix of diabetes symptoms – tiredness, lethargy, polyuria, polydypsia, nausea, etc – which are related to high blood sugar levels and which adversely affect their quality of life. As the case study on the diabetes floor of NPC states, the first priority is to control blood glucose to the extent sufficient to control symptoms. This could be with diet, but may well need medication, ideally metformin. After that, encouraging smokers to stop smoking, controlling blood pressure, adding a statin (ideally simvastatin 40 mg/day) for those at more than 20% risk of cardiovascular events over 10 years and adding aspirin once their blood pressure is controlled, is likely to achieve far more with regard to reducing their risk of macrovascular and microvascular complications than prioritising blood glucose control. Above all, the best quality evidence needs to be particularised to individual patients -a fundamental tenet of good medical practice is to take the wishes and priorities of individual patients into account.
Do not be misled: controlling blood glucose well for people with type 2 diabetes mellitus can be very important. But medical or lay preoccupation with glycaemic control to the neglect of smoking cessation, control of blood pressure, control of hyperlipidaemia, addition of aspirin, and use of metformin (for its broader metabolic effects as well as its hypoglycaemic effects) is, on the basis of the currently available evidence, harmful to people with type 2 diabetes.