In brief, 1,441 type-1 diabetic individuals were recruited into two cohorts: 726 in a primary prevention cohort with no retinopathy and an albumin excretion rate (AER) of <40mg per 24 hours; and 715 in a secondary intervention cohort with mild to moderate retinopathy and AER 40–200mg per 24 hours. Both genders were equally represented and the age range was 13–39 years. Each cohort was randomly assigned to either intensive treatment (?3 insulin injections per day or use of a subcutaneous insulin-infusion pump) or conventional treatment (up to two insulin injections per day).
In brief, 1,441 type-1 diabetic individuals were recruited into two cohorts: 726 in a primary prevention cohort with no retinopathy and an albumin excretion rate (AER) of <40mg per 24 hours; and 715 in a secondary intervention cohort with mild to moderate retinopathy and AER 40–200mg per 24 hours. Both genders were equally represented and the age range was 13–39 years. Each cohort was randomly assigned to either intensive treatment (?3 insulin injections per day or use of a subcutaneous insulin-infusion pump) or conventional treatment (up to two insulin injections per day). Any commercially available insulin could be used. The intensive treatment target was hemoglobin A1c (HbA1c) <6.0% (upper limit of nondiabetic range); the goal for conventional treatment was day-to-day clinical wellbeing and freedom from symptoms of glycosuria, ketoacidosis, and hypoglycemia.
Benefits of Intensive Treatment
In both the primary and secondary cohorts, over a mean follow-up of 6.5 years, intensive treatment significantly reduced the risks of development or progression of diabetic retinopathy and diabetic nephropathy. Risk reductions ranged from 34% to 76% and the treatment effect was almost entirely explained by the difference in average HbA1c over the trial (intensive 7.2% versus conventional 8.9%). It is noteworthy as a harbinger of later results that, although the difference in HbA1c levels was established by six months, it took three to four years for this to be reflected in a separation of the cumulative incidence of the complications. Moreover, the initial HbA1c level prior to any intervention was itself a significant predictor of progression of retinopathy through the trial period. Intensive treatment with a sharp, rapid lowering of blood glucose did not immediately benefit the complications.When the prevalence of neuropathy was assessed at five years into the trial by clinical examination and by peripheral nerve conduction and autonomic nervous-system testing, it was reduced 60% by intensive treatment. By the end of the DCCT, the intensive-treatment group had also experienced fewer CVD events than the conventional-treatment group, but the numbers of events were too small in this relatively young and healthy type-1-diabetic population to show statistically significant differences.
Carryover Benefits of Intensive Treatment (Metabolic Memory )
Approximately one year after the end of the DCCT, 96% of the whole cohort of research volunteers were recruited into an observational follow-up study (EDIC). After all conventional treatment participants had been offered training in intensive treatment, diabetes management was transferred to the participant’s physician of choice.There were two main purposes to EDIC. First, annual and biennial examinations were conducted to learn whether the beneficial effects of intensive treatment on early stages of retinopathy, nephropathy, and neuropathy during the DCCT would ultimately be reflected in reduced incidences of the later, clinically harmful, stages. Second, the evolution of CVD complications could be assessed and the risk factors for all the complications- environmental and genetic-could be determined.
The status of retinopathy and nephropathy at the DCCT closeout examination was used as the baseline for determining further progression of these complications during EDIC. A remarkable picture has emerged. The two former randomly assigned DCCT treatment groups have continued to show divergent patterns in their cumulative incidences of retinopathy and nephropathy, and the risk reductions with intensive treatment remain similar in magnitude for up to 10 years later. This is despite the fact that, over this observational period, the mean HbA1c of the two former treatment groups has been nearly identical (8.1% versus 8.2%). Modeling analyses of the EDIC data show that almost all of the treatment effects are explained by the mean HbA1c levels that existed during the DCCT. Thus, one may either conclude that the beneficial effect of an HbA1c around 7.0% has persisted despite a subsequent rise of 1.0%, or that the noxious effect of an HbA1c around 9.0% has persisted despite a subsequent decrease of 1.0%. By either interpretation, exposure of tissues to a particular level of hyperglycemia leads to a degree of damage that endures for years.The tissue exhibits a ‘metabolic memory’; one may say that hyperglycemia ‘casts a long shadow’. A similar metabolic memory effect appears to operate for diabetic neuropathy.
Intensive treatment during the DCCT, combined with the metabolic memory effect, has led to decidedly significant clinical benefits for retinopathy. At the end of 10 years of EDIC, of those randomized to DCCT as adults (approximately 90% of the whole cohort), 25% of the DCCT conventional-treatment group had developed proliferative retinopathy and 19% had received laser treatment. This compares to 9% and 6%, respectively, of the DCCT intensivetreatment group (p<0.001). With regard to nephropathy, although renal insufficiency is still very uncommon, 19 DCCT conventional-treatment group versus five DCCT intensive-treatment group patients have had serum creatinine levels reach ≥2.0mg per decilitre (mg/dL) and seven conventional versus four intensive have needed dialysis or renal transplant by eight years of EDIC (p=0.004). By anyone’s estimation, these are major clinical benefits of intensive treatment.Furthermore, a sufficient number of CVD events had occurred by 11 years of EDIC that a significant beneficial effect of intensive treatment on a composite of nonfatal myocardial infarction (MI) and stroke, CVD death, subclinical (‘silent’) MI, confirmed angina or revascularization by angioplasty, stent, or bypass surgery could be discerned. When only the first three ‘hard events’ were counted, a similar significant risk reduction was demonstrated. As was the case for microvascular complications, over 90% of the intensive treatment effect, could be accounted for by the decrease in HbA1c, such treatment produced during the DCCT. Since most of the CVD events occurred during EDIC, this may be yet another manifestation of metabolic memory.
Risks of Intensive Treatment
Intensive treatment certainly requires more time and effort from the patient and costs more, but the two proven clinical risks are episodes of severe hypoglycemia and excessive weight gain. Hypoglycemia was self-reported and defined in two ways: any episodes requiring assistance in treatment by a third party; and, more stringently, episodes characterized by coma or seizure. To be counted, an episode either had to be supported by a bloodglucose level of <50mg/dl or rapid improvement with intravenous glucose, subcutaneous glucagon, or oral carbohydrate. During the DCCT, 65% of intensive treatment and 35% of conventional treatment participants had at least one event. As the former were far more likely to have multiple events than the latter, the respective event rates were 62 versus 19 events per 100 person years. By five years, the cumulative incidence of a fifth episode was 22% versus 4%. Over one-quarter of all events were accompanied by coma or seizure. The event rates remained about the same throughout the DCCT period, despite concerted efforts to reduce them. There were no deaths proven to be caused by hypoglycemia, but two deaths were considered to be possibly the result of hypoglycemia by the adjudicating morbidity and mortality committee.
From the outset, there was concern that, in addition to the acute risk of serious neurological-even irreversible-sequelae, there might be more subtle adverse effects on the central nervous system, particularly from repeated hypoglycemic events, including some that were unreported and unappreciated that occurred during sleep. An epidemiological study early in the DCCT showed that 55% of events occurred while the patient was asleep.Therefore, a large battery of psychometric tests was performed at entry and conclusion of the DCCT. The domains tested were problem solving, learning, immediate memory, delayed recall, spatial information processing, attention psychomotor efficiency; and motor speed. No difference in test scores was observed when comparing intensively treated to conventionally treated patients, nor between those who had suffered multiple events (more than five) and those who had suffered no events. Despite these reassuring results, a concern lingered that damage to cognitive function might only become apparent later. Hence, the entire identical battery of tests was repeated in the 12th year of EDIC. Again, in none of the eight domains was there a difference between former intensively treated and conventionally treated patients or between those with multiple and those with no hypoglycemic events (including those that occurred during EDIC). All results were within normal population ranges.Intensive treatment led to a mean gain of an additional 4.8 kilograms of body weight than conventional treatment by the end of the DCCT, with women gaining more weight than men. Bodycomposition studies showed this excess weight comprised both fat and lean body mass. Waist circumference increased by 3.2cm in men and 1.5cm in women.When divided into quartiles of increase in body mass index (BMI), the highest-quartile BMI in the intensive group had significantly increased systolic blood pressure, serum triglycerides, low-density lipoprotein cholesterol, apoproteins B, and waist circumference.This pattern suggests the development of the insulin resistance or metabolic syndrome. While it is reassuring that the DCCT intensivetreatment group as a whole has a decreased risk of CVD events thus far, there are still insufficient numbers of events to determine whether BMI might be a risk factor for CVD in this type-1-diabetes cohort. Other possible clinical consequences of the insulin resistance or metabolic syndrome such as polycystic ovaries, infertility, uterine cancer, or breast cancer have not been apparent, but the study remains vigilant for these possibilities.
Summary
There is a balance between the benefits and risks of intensive treatment of type-1 diabetes, with normal or near-normal glycemia as the target of treatment. The benefits outweigh the risks, given the long-term reductions in microvascular and CVD complications, by one-half to two-thirds (even if glycemic levels rise a bit above HbA1c of 7.0%), compared with the threefoldgreater short-term risk of severe hypoglycemia. However, it must be emphasized that this is a general balance for the population of type-1-diabetic individuals in the DCCT/EDIC cohort.
In comparison with two population-based samples, it was calculated that about 120,000 individuals (17%) of the type-1-diabetic population would have met the DCCT eligibility criteria. Furthermore, compared with a sub-cohort of the Wisconsin Epidemiology of Diabetic Retinopathy study that would have met DCCT eligibility criteria, the relationship of diabetic retinopathy to HbA1c levels and the progression of retinopathy over time was similar to those observed in the DCCT conventional-treatment group.
The DCCT/EDIC benefit-to-risk balance may not apply to children or the elderly-neither of which group was included in the DCCT/EDIC study and may be more vulnerable to the effects of hypoglycemia -or to those with hypoglycemia unawareness and a history of repeated events or those with life-shortening comorbidities or lifestyles and conditions, such as alcoholism, that make prevention of hypoglycemia more difficult. Moreover, some patients would not consider it a good trade-off to accept even one episode of hypoglycemic coma or seizure in exchange for a lesser risk of developing nonproliferative retinopathy. Other patients might be willing to accept multiple episodes of severe hypoglycemia in exchange for a lesser risk of stroke or MI at a relatively young age.Thus, the balance between benefits and risks should be reviewed individually with each patient before setting treatment targets.
Nonetheless, the overall results of the DCCT/EDIC study support the recommendation that intensive treatment should be implemented as early as possible in the course of type-1 diabetes, with due regard for patient safety. These results also encourage the hope and belief that, when blood glucose can be normalized by artificial closed-loop insulin-delivery systems or by restoring a sufficient number of normally functioning beta cells, the microvascular, neuropathic, and CVD complications will be eliminated from type-1 diabetes.
