the sheet blog
DCCT Study

December 7th, 2009

“As you know I think tight control is a good idea. The clinical study that is going to prove that tight control prevents vascular disease is about halfway done and is looking promising. So let’s stick with your tight control. I don’t want your kidneys to fail.” This was the first I heard of the DCCT study, from Dr. Andrew Drexler, when I lived in New York City and was his patient. They were going to prove tight control is good. It seems obvious to us now, and tight control is the keystone of diabetes therapy in the early 21st century. But it was not always obvious.

We must remember that kids died of type 1 diabetes before insulin was available. They wasted away. So when the Canadians Banting and Best announced that they had found the anti-diabetic hormone there was celebration: when administered to diabetics they seemed to come back to life. They gained weight. They lived.

As therapy developed opinion was divided on on the dosing needed to manage the disease. My grandfather’s diabetes management book from the 1940′s shows that therapy was counting calories and injecting insulin on a predetermined schedule. The only way to determine that the insulin dose was right was to measure urine sugar (which required multiple test tubes and hot water on the stove). Clearly the average blood sugar would have been high by tight control standards.

When I was diagnosed in the 1970′s we were still in the same era: only urine sugars were available at home. (But it no longer required a chemistry set but rather a convenient strip.) I would get a blood sugar measurement every three months — when I visited my doctor! Then home blood glucose testing sets were launched and the new era of management began.

Once it was possible to self-measure blood sugars, interest in the relationship between blood sugar and vascular disease grew. Many thought that blood sugar abnormalities were the cause of diabetic vascular disease which meant that better control would reduce vascular disease. I remember that at the time the most compelling anecdotal evidence came from some of longest lived diabetics. When long acting insulin was introduced some diabetics did not switch but continued to inject pure insulin (what we now call regular insulin). These old-timers tended to have less vascular disease than the more up-to-date diabetics taking long acting insulin. Could it be that a bolus of insulin at meals was better than steady insulin all the time?

The DCCT study was designed to answer that question, and it did in spades. It is rare that a major study in a major disease has such a dramatic outcome. Look at figures 2 through 4. Vascular complications drop dramatically with improved hemoglobin A1c, a measure of average blood sugar. It was clear to all that tight control works. And tight control has been the quest for type 1 diabetics ever since.

Type 1 Diabetes Classic.
Follow the responses to this post through the RSS 2.0 feed.
You can leave a response.

One Response to “DCCT Study”

  1. JP Marat says:

    An alternative explanation for the better results obtained by patients who continued on the old regimen of repeated daily doses of regular insulin was that they inevitably suffered repeated daily bouts of hypoglycemia with that regimen, and these episodes interrupted the formation of cross-linkages from advanced glycation endproducts.

    A priori, it seems implausible that by correcting the first and most prominent side-effect of the autoimmune disease that causes type 1 diabetes — that is, the hyperglycemia that follows the beta cell destruction — you can also expect to remove all the effects on the body of the continuing autoimmune condition, which Faustman and others have shown never ceases to operate. While the DCCT showed a population association between lower HbA1c values and fewer complications, it also demonstrated that there was often a lack of pairwise correspondence between patients, so that some with high blood sugar values had few complications, while others with good blood sugar control did poorly. The failure of a specific HbA1c value to produce a predictable result over all patients needs to be explained.

    Epidemiological data casts some doubt on the centrality of blood sugar role in the fate of diabetic patients. While from 1968 to 1984, prior to the development of home blood sugar meters, there was a dramatic, three-fold decline in diabetes-related deaths among type 1 children up to age 18, from the introduction of the home blood sugar meter in 1984 to 1998 there was no decline in diabetes-related deaths. (J. Di Liberti, et al, “Long-Term Trends in Childhood Diabetes Mortality,” Diabetes Care, 24(2) 1348, 2001)) Similarly, there was no improvement in microalbuminuria in type 1 diabetics from 1986 to 1996. (R. Amin, et al, “Unchanged Incidence of Microalbuminuria in Children with Type 1 Diabetes Since 1986,” Archive of Diseases of the Child, 94(4) 251 (2009)) At 25 years post onset of type 1 diabetes, there was also no improvement in proliferative retinopathy and overt nephropathy from the period 1950-1959, when strict blood sugar control was often not even recommended, and 1975-1980, when blood sugar control was generally accepted as the goal of therapy. (G. Pambianco, et al, “The 30-Year Natural History of Type 1 Diabetes Complications,” Diabetes, 55(5) 1463 (2006); cf. E. Joslin, et al, Joslin’s Diabetes Mellitus (Philadelphia: Lippincott, 2004, 798)) Half of all diabetics developed nephropathy before 1950, but today, with all the improvements in blood sugar control, the percentage developing nephropathy is variously estimated at between 30% and 40%, which seems a very small improvement for so much effort. (J. Ekoe, et al, The Epidemiology of Diabetes Mellitus (London: John Wiley, 2001, 341))

    While I recognize that there is a general consensus that strict blood sugar control will eliminate or sharply reduce the incidence of diabetic complications, some stubborn data seem to stand in the way of that hypothesis. The serum of patients with type 1 diabetes has been demonstrated to be toxic to nerve growth factors even when its sugar content is normal. (G. Pittenger, “The Toxic Effects of Serum from Patients with Type 1 Diabetes,” Diabetes Medicine, 10(10) 925 (1993)) It is now well-established that there is little relation between blood sugar control in diabetics and their elevated risk for coronary and lower extremity disease, so this forces us to complicate the theory that hyperglycemia causes vascular damage by positing that it is linked only to micro- but not to macrovascular disease. (C. Lloyd, et al, “Incidence of Complications in Diabetes Mellitus,” American Journal of Epidemiology, 143(5) 431 (1996); S. Mudalier, “Serum Glucose Control in Diabetic Patients with Cardiovascular Disease,” Current Atherosclerosis Reports, 11(5) 384 (2009)) This then raises the question, what is causing such closely related damage to large blood vessels in diabetics if it results from some mechanism different from that leading to microvascular disease? It would seem odd to have two blood vessel diseases working in tandem in diabetics, one from hyperglycemia and one not, rather than both having their origin in a common underlying mechanism as yet undiscovered. Diabetic nephropathy seems particularly resistant to the effects of improved blood sugar control, and while only half of type 1 diabetics with poor blood sugar control ever develop renal failure, some of those with good control suffer renal failure. (S. Rich, “Genetics of Diabetes and its Complications,” Journal of the American Society of Nephropathy, 17, 353 (2006)) At the one extreme we have type 1 diabetics who have survived 50 years or more with their disease with minimal complications, but who when surveyed were found to have an average HbA1c of more than 10%, while at the other, we have type 1 diabetics with functioning pancreatic transplants whose improvement in complications remains a matter of controversy, and is at best only moderate. (G. Gill, et al, “Insulin-Dependent Diabetes of Over 50 Years’ Duration,” Practical Diabetes International, 10(2) 60 (2005); J. de Sa, et al, “The Evolution of Diabetic Chronic Complications after Pancreas Transplantation,” Diabetology and Metabolic Syndrome, 1(1) 11 (2009))

    These results, surprising for the conventional theory, suggest some other factors might be involved in the etiology of diabetic complications. Even if it is argued that hyperglycemia still plays a role in the genesis of complications, the fact that other causes are operative must diminish the significance of blood sugar control as a therapeutic strategy in diabetes treatment. Genetic factors are certainly operative in at least some of the classic complications of the disease.The enzyme responsible for the degradation of fibrosis from tissue injury is elevated in both diabetics with renal disease and in their first-degree relatives. (C. Bau and S. Twigg, “Fibrosis in Diabetes Complications,” Vascular Health Risk Management, 4(3) 575 (2008)) There is now an increasing acceptance of the view that diabetic nephropathy is probably a genetic trait inherited along with the complex of genes causing diabetes in a distinct subgroup of that population. (L. Tarnow, “European Rational Approach for the Genetics of Diabetic Complications,” Nephrology, Dialysis, Transplantation, 23(1) 161 (2008); L. Thorn, et al, “Clustering of Risk Factors in Parents of Patients with Type 1 Diabetes and Nephropathy,” Diabetes Care, 30(5) 1162 (2007)) The cause of retinopathy in diabetics is also now recognized to have a strong genetic component. (G. Liew, “The Rose of Genetics in Susceptibility to Diabetic Retinopathy, International Ophthalmology Clinics, 49(2) 35 (2009); S. Abhary, et al, “A Systematic Meta-Analysis of Genetic Assocation Studies for Diabetic Retinopathy,” Diabetes, 58(9) 2137 (2009)) It is now being suggested that diabetic autonomic neuropathy may also have a genetic cause. (C. Haverslev-Foss, et al, “Autonomic Neuroopathy in Nondiabetic Offspring of Type 2 Diabetic Subjects,” Diabetes, 50(3) 630 (2001)) Familial clustering of the severity and type of diabetic complications has been noted as well. (M. Moriti, et al, “Familial Risk Factors for Microvascular Complications,” Journal of Clinical Endocrinology and Metabolism, 92(12) 4650 (2007)) Of course it could be argued that while there may be genetic susceptibilities to damage from hyperglycemia, given that genetic engineering is many decades away as a practical intervention, perhaps the best place to intervene to stop the damage is by controlling hyperglycemia rather than the genetic factors, such as surplus enzymes for degrading surplus glucose to advanced glycation endproducts, which make excess blood sugar toxic in some people. But given that some changes consistent with diabetic complications can be discerned even in diabetics’ close relatives who themselves have perfectly normal blood sugar levels, perhaps these genes are always able to cause damage and are simply intensified in their effect in the presence of the genes which also cause the initial autoimmune attack which first brings the disease to clinical attention. (Cf. C. Giannattasio, et al, “Increased Arterial Stiffness in Normoglycemic, Normotensive Offspring of Type 2 Diabetic Patients,” Hypertension, 51(2) 182 (2008))

    Most autoimmune diseases attack not one but several organs, so it is natural to suspect that the continuing autoimmunity of type 1 diabetes does not just attack the pancreatic beta cells and then have no further affects. It is well-established that type 1 diabetes is associated with a higher incidence of other diseases which are known to have an autoimmune cause, such as autoimmune thyroid disease and multiple sclerosis. What if some or all of the diabetic complications were not due to hyperglycemia but instead just to the autoimmunity? (T. Staeva-Vieira, et al, “Translational Mini-Review Series on Type 1 Diabetes: Immune-Based Approaches for Type 1 Diabetes,” Clinical and Experimental Immunology, 148, 17 (2007)) This seems quite likely, since autoimmunity is known to produce an inflammatory state in the body, and this oxidative stress has well-known consequences for the health of the vascular system. (M. Zimmerman and S. Flores, “Autoimmune-Mediated Oxidative Stress and Endothilial Dysfunction,” Journal of Surgical Research, 155(1) 173 (2009); S. Devarai, et al, “Evidence of Increased Inflammation and Microcirculatory Abnormalities in Patients with Type 1 Diabetes,” Diabetes, 56(11) 2790 (2007)) It has been suggested that pro-inflammatory cytokines linked to diabetic autoimmunity may cause diabetic nephropathy. (K. Ichinose, “Recent Advancement of Understanding of Pathogenesis of Type 1 Diabetes and Potential Relevance to Diabetic Nephropathy,” American Journal of Nephrology, 27(6) 554 (2007)) Some evidence exists to suppor the theory that diabetic neuropathy may be caused by autoantibodies. (V. Granberg, “Autoantibodies to Autonomic Nerves Associated with Cardiac and Peripheral Autonomic Neuropathy,” Diabetes, 28(8) 1959 (2005); K. Kles, et al, “Pathophysiology and Treatment of Diabetic Peripheral Neuropathy,” Current Diabetes Review, 2(2) 131 (2006); C. Herder, et al, “Subclinical Inflammation and Diabetic Polyneuropathy,” Diabetes Care, 32(4) 680 (2009)) An excellent study has recently argued that diabetic retinopathy should in fact be regarded as an autoimmune rather than a hyperglycemic disease, based on a close histological examination of the nature of the retinopathic damage, which is clearly autoimmune in nature. (D.Adams, “Autoimmune Destruction of Pericytes as the Cause of Diabetic Reitnopathy,” Clinical Ophthalmology, 2(2) 295 (2008))

    The explanatory principle of Occam’s Razor warns scientists to try to minimize their resort to hypotheses to account for phenomena, and the theory that both the initial destruction of the beta cells in diabetes and the further complications that follow later are both due to one single cause — the autoimmunity and its production of antibodies, inflammation, oxidative stress, and cytokine release — rather than to two causes — autoimmunity and then hyperglycemia — has a certain elegance. Thus it has been theorized that the autoimmunity of diabetes produces elevated reactive oxygen species long before the patient is diagnosed as diabetic, and that this inflammatory state causes both the destruction of the beta cells and — as it continues its operation over the decades following — on the vascular system. (M. Brownlee, “Biochemical and Molecular Cell Biology of Diabetic Complications,” Nature, 414, 813 (2001))

    The obvious question which remains then is why diabetic complications have been found to be correlated with blood sugar control if the two are not linked as effect and cause? The answer is that blood sugar levels, or the ability of the patient to control blood sugar, may be merely the marker for some underlying factor which is the actual mechanism causing the diabetic complications. The inheritance of the disposition to type 1 diabetes is known to be quite complex and to involve many genes, so it may be the case that the inheritance of a more virulent genetic complex not only makes blood sugar control more difficult, but also causes more damage to the vascular and neurological systems of the patient independently of the blood sugar level. It seems probable to me that a more severe form of autoimmunity more thoroughly destroys the pancreatic beta cells, with the result that patients with a more severe form of diabetic autoimmuity also have more difficulty controlling their blood sugar levels, making blood sugar control seem to ’cause’ the complications which are actually caused by the more serious, lifelong autoimmune attack. Autoimmunity causes toxic cytokine expression, cytokines attack the pancreatic beta cells, and then continue to attack the rest of the body throughout the life of the patient. (O. Pankewycz, et al, “Cytokines Are Mediators of Autoimmune Diabetes and Diabetic Complications,” Endocrinology Review, 16(2) 169 (1995))

    Diabetics with any residual endogenous insulin production can achieve much better blood glucose levels than those who lack residual production, since natural insulin production is carefully tailored to present requirements and thus provides a buffer against both hyper- and hypoglycemia. But c-peptide is also produced along with insulin, although it has long been regarded as a useless metabolic by-product of insulin manufacture in the body. Now, however, there is increasing evidence that it may be the lack of c-peptide which is the actual mechanism of diabetic complications rather than the poor blood sugar control which is closely associated with it. (B. Johansson, et al, “Beneficial Effects of C-Peptide on Incipient Nephropathy in Patients with Type 1 Diabetes Mellitus,” Diabetes Medicine, 17(3) 181 (2000); J. Warren, “C-Peptide: New Findings and Therpeutic Implications in Diabetes,” Clinical Physiology and Functional Imaging, 24(4) 180 (2004); Y. Ide, et al, “Prevention of Vascular and Neuronal Dysfunction in Diabetic Rats by C-Peptide,” Science, 277(5325) 563 (1997)) Thus it may well be the case that the complications of diabetes are in fact due to the autoimmune disease destroying the beta cells of the pancreas, which in turn causes a lack of c-peptide, which is the actual cause of the complications. The association between a lack of c-peptide and poor blood sugar control is inevitable, since the residual insulin production which allows patients to achieve better blood sugar control is also matched molecule for molecule with the c-peptide production. Complications in type 2 diabetics who do not lack native insulin production may then be due to the demonstrated inability of their cell walls to permit c-peptide to cross into them.

    All of these possible causes of diabetic complications remain speculative at this point, but questions about whether the continuing effects of autoimmunity, a genetic predetermination, a lack of c-peptide, or just plain hyperglycemia cause the complications in whole or in part must discount some of the enthusiasm for efforts to control blood sugar levels as the best route to ‘curing’ this difficult disease.

Leave a Reply