 |
|
 |
 |
 |
 |
 |
|
|
 |
|||||||||||||||||||||||||||||||||||||
 |
||||||||||||||||||||||||||||||||||||||||
| Home > Evidence of the Month > Archive | |
 |
|
Education |
| February 2006 Evidence of the Month |
 |
Reduced Cardiovascular Disease With Intensive Blood Glucose Management: A Long-term Follow-up of the DCCT
Philip Home
Newcastle Diabetes Centre
University of Newcastle upon Tyne
Newcastle, UK
Background
The major cause of premature death in people with Type 1 diabetes has been cardiovascular disease (CVD), generally in middle age and often in association with diabetic kidney damage. Age-specific risk rates are around 10 times the background population.
In December 2005, the New England Journal of Medicine published a paper from the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group which examined cardiovascular (CV) events for a further 12 years after completion of the original trial (totalling 17 years in all). A 42% reduction in CV events (95% confidence interval [CI], 9% to 63%; P=0.02) was reported in the group previously managed intensively during the 5-year DCCT period.
In the original DCCT (as published in the New England Journal of Medicine in 1993), reduction in CV damage (−41%) approached statistical significance (P=0.06), but total number of events was very small and was dominated by the development of peripheral vascular disease.
Methods
The DCCT was not a study of good glucose control nor of intensified insulin therapy. It was a study of 1441 people with Type 1 diabetes randomised to intensive management or conventional management (1983-style). It is not clear what combination of aspects of intensive management led to protection against microvascular disease in the original trial. Subsequently, 93% of the cohort have been followed up for the 2005 study in regular care, but by 1997 HbA1c was very similar in the two management groups (as published in the New England Journal of Medicine in 2000).
CV events were ascertained up to early 2005 (blind to the original management group); events included myocardial infarction, onset of confirmed angina, stroke, cardiac revascularisation, and likely CV death. Assessments were made of number of events, number of people affected, time to first event, association with original randomised management group, updated mean blood glucose control (HbA1c), and the development of renal damage.
Key Results
Conventional arterial risk factors (lipids, blood pressure, cigarette smoking, body mass index [BMI]) differed little between the management groups, whether at randomisation, end of the DCCT study, or 11 years later in 2004; however, BMI was 6% higher with intensive management at the end of the DCCT, but this difference did not persist. HbA1c was much lower with intensive management at the end of the 5-year DCCT but identical with long-term follow-up.
The CV risk factor microalbuminuria was markedly reduced by intensive management not only at 5 years but also at the end of long-term follow-up. Accordingly, progression to albuminuria and to raised serum creatinine was also reduced at long-term follow-up.
Overall CV event rate was reduced in the previously intensively managed cohort after 17 years: 0.38 vs 0.80 per 100 patient-years (P=0.007). Rates for individual CV events were also consistently about 50% reduced. Time to event analysis gave a 42% reduction in risk (95% CI, 9% to 63%; P=0.02). Absolute risk reductions can be calculated as 52 events over about 17 years in around 590 people - about 10% in 20 years, with a number needed to treat of 10 to prevent 1 event in 2 decades.
Microalbuminuria was associated with increased risk of CV events. However, when the effect of lesser microalbuminuria in the intensive group was stripped out, there was still a statistically significant CV risk reduction in that group.
HbA1c during the 5-year DCCT was related to CV outcome over the whole follow-up period (17 years). A 10% reduction of HbA1c (eg, from 8.0% to 7.2%) would predict a 20% (9% to 30%) reduction in CV risk. Over half of the CV risk reduction disappeared with removal of the effect of the 5-year HbA1c; indeed the 16% risk reduction that remained was not statistically significant. No combined adjustment for renal disease and HbA1c was reported.
Clinical Relevance
The same study group previously reported that long-term follow-up of the DCCT cohort showed a ‘memory effect’ of the intensive management during the study on development of microvascular complications after 4 years, despite convergence of HbA1c when study participants returned to regular diabetes care. The current study extends that observation to much longer term follow-up but for CV outcomes. The effects are important because CVD strikes people with Type 1 diabetes relatively early in life, often when they are at their most productive and with greatest family responsibilities. Thus, while the benefits in absolute terms might seem on the low side (around 0.5% absolute per year), they have to be added to the microvascular benefits and multiplied by the long period of future health lost when disability or death occurs in middle age.
The benefits would also have presumably been even greater had the separation in management intensity been maintained for more than 5 of the 17 study years. It is not possible to estimate this reliably but a minimum would be a doubled effect, and a maximum (using the DCCT separation of HbA1c and multiplying by 17/5 for duration) would appear to give a risk reduction of most of the excess risk in the population.
Is the effect due to blood glucose control or some other aspect of intensive management (such as the dietary education given in the DCCT)? We do not know, but the pointers are all towards blood glucose control. Thus, other CV risk factors including lipids were substantially unchanged (and BMI actually higher in the intensively managed group), while the regression analysis suggests that HbA1c could account for much of the gain. It is true that microalbuminuria prevention was also important in protecting against CV events, but glucose control seems to be the major primary factor (and blood pressure secondary) in preventing the diabetic kidney damage in the first place.
Why the seeming memory effect of good blood glucose control? Vascular damage is a prolonged pathogenetic process, whether microvascular or atherosclerotic, so perhaps it is not surprising that the clock can be suspended by returning blood glucose control much closer to non-diabetic levels. But what level should be our target? The 2005 paper does not help much with that question - it would be necessary to know the glucose control levels of the people who did develop CVD in both randomised populations to answer that.
Does the study have relevance for Type 2 diabetes? Yes, the process of atherosclerosis is similar in people developing arterial damage in Type 1 and Type 2 diabetes. Meanwhile the lessons for clinical practice appear to be to intensify management and to use HbA1c to monitor the treatment effects, perhaps down to a target of 7.0% (DCCT-aligned) in the absence of troublesome hypoglycaemia or other major life restrictions. Additional CV protective management should be reserved for those patients with microalbuminuria (in particular), poorer glucose control, or other CV risk factors.
This Website Feature is funded by an unrestricted educational grant from sanofi-aventis and Pfizer Inc.
|
 |
|
Print this page |
|
||
 |
Add to Favourites |
|
|