This observational data comes from follow-up of patients who participated in the UKPDS randomised controlled trial (RCT). It suggests that early intensive blood glucose control has lasting benefits for patients for up to 10 years. However, the original data from the UKPDS RCT (a higher level of evidence and less prone to bias) showed only limited microvascular, not macrovascular, benefits from tight glucose control. Recent data from other RCTs indicates no benefit and possible harm from tight blood glucose control.
As part of the observational follow up of UKPDS the same researchers report that good blood pressure (BP) control must be continued if the benefits on microvascular and macrovascular outcomes are to be maintained. This is not surprising.
This new UKPDS observational data should not change the view that we should be advising active management of all cardiovascular risk factors in people with type 2 diabetes on an individualised basis, rather than focussing disproportionately on tight control of blood glucose.
Controlling blood glucose well for people with type 2 diabetes is important, but after controlling the symptoms associated with having high blood glucose levels (with diet/lifestyle and metformin), clinicians should give priority to reducing cardiovascular risk. Interventions such as smoking cessation, diet, exercise, weight reduction, BP control and the use of metformin/aspirin/simvastatin should take priority. We should not become over-focussed on intensive strategies to achieve HbA1c targets.
What is the background to this?
UKPDS was a 20-year RCT which recruited over 5,000 patients with newly diagnosed type 2 diabetes. It consisted of three main arms: one looking at the effect on tight blood glucose control (using sulphonylureas/insulin) vs. conventional control (primarily with diet); the second comparing tight blood pressure control with less tight control; and the third, in a subset of overweight patients looking at the effect of tight control of blood glucose using metformin vs. conventional control.
The results, as discussed in these MeReC Briefings (Type 2 diabetes: the management of blood glucose and cardiovascular risk factors) were that while tight BP control and using metformin produced many patient oriented benefits, reducing both microvascular and macrovascular outcomes, the effect of tight blood glucose control using sulphonylureas/insulin was less convincing, reducing microvascular outcomes only.
The two papers discussed here report 10-year follow-up of the UKPDS survivor cohort, once the study interventions were stopped and patients returned to normal care. During follow-up, patients were assessed with annual clinic visits up to year 5 then questionnaires; no attempt was made to maintain previously assigned therapies.
What do these studies claim?
The follow-up of the blood glucose part of the study claims that, despite an early loss of glycaemic differences between the intensive and conventional arms of the study (baseline differences in mean HbA1c levels were lost by one year) a continued reduction in microvascular risk and emergent risk reductions for macrovascular outcomes were seen.
- Any diabetes-related endpoint occurred in 57.6% of the intensive group vs. 60.3% of the conventional group (Risk Ratio [RR] 0.91; 95% confidence interval [CI] 0.83–0.99; P=0.04)
- Microvascular endpoints occurred in 15.7% vs. 19.5% (RR 0.76; 95%CI 0.64–0.89; P=0.001)
- Diabetes-related death in 22.6% vs. 26.1% (RR 0.83; 95%CI 0.73–0.96; P=0.01)
- All-cause mortality in 42.6% vs. 47.2% (RR 0.87; 95%CI 0.79–0.96; P=0.007)
- Myocardial infarction in 24.8% vs. 28.0% (RR 0.85; 95%CI 0.74–0.97; P=0.01).
Significant risk reductions also persisted with metformin, in the sub-study of overweight patients.
- Any diabetes-related endpoint occurred in 61.1% in the metformin group vs. 63.7% in the conventional group (RR 0.79; 95%CI 0.66–0.95; P=0.01)
- Diabetes-related death in 23.7% vs. 29.2% (RR 0.70; 95%CI 0.53–0.92; P=0.01)
- All-cause mortality in 44.4% vs. 52.8% (RR 0.73; 95%CI 0.59–0.89; P=0.002)
- Myocardial infarction in 23.7% vs. 30.7% (RR 0.67; 95%CI 0.51–0.89; P=0.005).
Details of individual endpoints from aggregate outcomes are not reported.
In the follow-up of the blood pressure part of the study, differences in blood pressure between the two groups were lost by two years and risk reductions which were significant during the interventional trial e.g. for any diabetes-related endpoint and diabetes-related death were not sustained.
How does this relate to other studies?
As we have previously blogged, two recently published randomised controlled trials (RCTs), ACCORD and ADVANCE found no significant improvements in macrovascular events with intensive glucose control versus standard glucose control. Indeed, in ACCORD, intensive therapy was associated with an increased risk of death.
A further study, the VA Diabetes Trial (VADT), only reported so far at a conference supports these two papers as intensive glucose control did not significantly reduce cardiovascular events compared with standard control, once other cardiovascular risk factors had been addressed.
The follow-up data from UKPDS are interesting. They suggest early intensive glucose control with either sulphonylureas/insulin or metformin has a lasting effect on microvascular outcomes and emerging, so-called ‘legacy effects’ on macrovascular events and death. However, these results need to be viewed with caution. They are derived from observational data which, in the hierarchy of evidence, do not carry as much weight as data from RCTs due to them being more prone to bias. Data were collected by questionnaires not clinic visits after 5 years, and questionnaires may not have accurately documented all outcomes. Also, we have no data on how these follow-up patients were managed with regard to blood pressure, lipids, smoking cessation etc. In the original UKPDS RCT, only metformin, not sulphonylureas/insulin reduced macrovascular complications.
These follow-up data from UKPDS should not be used to promote early, very intensive glucose lowering treatment for all patients with type 2 diabetes. Optimal glucose control early on in type 2 diabetes may have benefits beyond reducing microvascular complications, but this needs to be shown in high-quality RCTs. Also, it is important to define ‘intensive control’. UKPDS was started in 1977 when targets for glucose control were not as aggressive. During the intervention phase of UKPDS, median HbA1c over 10 years was 7.0% in the intensive control group. At the end of the intervention phase and start of the follow-up, this was 7.9%. In ACCORD and ADVANCE, no significant macrovascular benefits were seen despite HbA1c levels of 6.4% and 6.5%, respectively, being reached.
The recent NICE clinical guideline on diabetes provides advice on setting a target for HbA1c. Although generally advocating the setting of a 6.5% target, it cautions against the use of highly intensive management strategies to achieve levels of less than 6.5%, and recognises the importance of involving the patient in the setting of their own target level, which may be above that of the 6.5%.
The bottom line remains that metformin is the only oral hypoglycaemic drug that has been shown to reduce macrovascular complications in high-quality RCTs, and it remains the drug of choice for first-line use in the majority of patients with type-2 diabetes. Control of blood glucose is important for people with type 2 diabetes, but this follow-up of UKPDS should not be used to potentiate a preoccupation with glycaemic control to the neglect of other measures. For people with type 2 diabetes who have symptoms associated with high blood glucose, the first priority is to control blood glucose to the extent sufficient to control these symptoms. This could be with diet, exercise and weight reduction, but may well require 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 BP 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.
With regard to BP, the follow-up from UKPDS shows that good BP control must be continued if the benefits on microvascular and macrovascular outcomes are to be maintained. This is not surprising.
You can find more information on the type 2 diabetes floor of NPCi.
These papers report 10-year post-interventional follow-up of the UKPDS survivor cohort. Of 5,102 original patients with newly diagnosed type 2 diabetes, 4,209 were randomised in the blood glucose part of UKPDS and 1,148 in the blood pressure part of the study. 3,277 of the blood glucose cohort entered post-trial monitoring follow-up and 884 of the blood pressure cohort. During follow-up, patients were assessed with annual clinic visits (for 5 years only) or annual questionnaires, no attempt was made to maintain their previously assigned therapies. Outcome events were adjudicated as during the original trial by members of the UKPDS endpoint committee, blinded to study groups.
Results – blood glucose follow-up: baseline differences in mean HbA1c levels between the intensive therapy group and the conventional therapy group were lost by one year. During a median of 8.5 years post-trial follow-up, risk reductions which were significant during the interventional trial remained so. Any diabetes-related endpoint occurred in 57.6% of the intensive group vs. 60.3% of the conventional group (RR 0.91; 95%CI 0.83–0.99; P=0.04). Microvascular endpoints occurred in 15.7% vs. 19.5% (RR 0.76; 95%CI 0.64–0.89; P=0.001). Post-trial risk reductions also emerged for diabetes-related death (22.6% vs. 26.1%; RR 0.83; 95% CI 0.73–0.96; P=0.01); all-cause mortality (42.6% vs. 47.2%; RR 0.87; 95%CI 0.79–0.96; P=0.007) and myocardial infarction (24.8% vs. 28.0%; RR 0.85; 95%CI 0.74–0.97; P=0.01). In the sub-study of overweight patients, again, risk reductions which were significant during the interventional trial remained so. This included, any diabetes-related endpoint (61.1% in the metformin group vs. 63.7% in the conventional group; RR 0.79; 95%CI 0.66–0.95; P=0.01); diabetes-related death (23.7% vs. 29.2%; RR 0.70; 95%CI 0.53–0.92; P=0.01); all-cause mortality (44.4% vs. 52.8%; RR 0.73; 95%CI 0.59–0.89; P=0.002); and myocardial infarction (23.7% vs. 30.7%; RR 0.67; 95%CI 0.51–0.89; P=0.005).
Results – blood pressure follow-up: baseline differences between the groups in mean systolic and diastolic blood pressures were lost by year 1 and year 2, respectively. During a median of 8.0 years post-trial follow-up, risk reductions which were significant during the interventional trial were not sustained. Any diabetes-related