Do We Need More Research to Tell Us That Diabetes Mellitus is Bad?

The relationship between diabetes mellitus, either type 1 or type 2, and future cardiovascular disease (CVD) is well known. The rise in prevalence of both types of diabetes in younger individuals suggests that the trend towards a reduction in cardiovascular morbidity and mortality in younger individuals observed over the last several decades will be attenuated by this change in the epidemiology of diabetes.¹ The Pathobiological Determinants of Atherosclerosis in Youth Study has shown a strong relationship between dysglycemia and advanced atherosclerosis (American Heart Association grade 4 or higher) in individuals as young as 15-19 years. In fact, dysglycemia is the only CVD risk factor associated with advanced atherosclerosis in this age group and removes the protection for women from early atherogenesis.² Regardless of diabetes type, current knowledge of atherosclerosis pathology suggests that dysglycemia may initiate a similar cascade of events in the vessel wall as LDL cholesterol.³ Diabetes also exerts adverse influences on cardiac function and is a risk factor for heart failure.⁴ A hallmark of diabetes is small vessel disease leading to renal, cerebrovascular, and peripheral arterial disease.

Given recent advances in understanding and assessment of antecedents of CVD, not only in measurement of risk factors but in assessment of cardiovascular target organ injury and mediators of CVD, a substantial literature has arisen documenting the early adverse consequences of diabetes. Maahs et al, in the Scientific Statement “Cardiovascular Disease Risk Factors in Youth with Diabetes”, summarize this literature.⁵ For type 1 diabetes mellitus, conventional risk factors have a similar prevalence as in the general population, placing a substantial number of affected children in a multiple risk setting. Many inflammatory factors associated with CVD are elevated compared to those without diabetes. For type 2 diabetes mellitus, the well-known associations with obesity (and obesity associated inflammatory/metabolic disturbances), hypertension, and dyslipidemia are redocumented. For both type 1 and type 2 diabetes mellitus, evidence of early cardiovascular target organ damage is apparent, with associations somewhat stronger for type 2 diabetes, probably because of longer exposure to obesity related risk. Where available, longitudinal data on children with both types of diabetes suggests CVD risk worsens over time (often despite treatment) and this deterioration is associated with poorer glucose control.

Maahs et al point out that insufficient information exists about the differences with regard to early cardiovascular injury between type 1 and type 2 diabetes mellitus at the onset of disease. While most research in adults suggest both types lead to similar complications, at disease onset, those with type 2 diabetes already have had chronic exposure to the metabolic complications of obesity. Thus, timing of specific interventions (type of glycemic control, CVD risk control) during the early phase of the illness may need to be different based on diabetes type.⁵

Is anyone actually surprised by these findings? I would hope not. This authoritative report with over 400 references should answer any lingering questions concerning the future medical issues confronted by children with diabetes. If this overview is insufficient, we should consider the limited impact of interventions, other than statin treatment for dyslipidemia in adults, with well-established diabetes.⁶ The likely diminished impact of a range of interventions, including tighter glycemic control, aspirin therapy, and anti-hypertensive treatment, suggests significant established target organ injury that cannot be overcome by treatment late in the pathologic process.

Our agenda should no longer be proving to ourselves by cross-sectional studies that diabetes mellitus is bad for people, but determining what we can do about it. I would like to suggest a more ambitious and differently focused research agenda than discussed by Maahs and colleagues.

The first step will be acquiring longitudinal data on both CVD risk and target organ damage. The results of these studies are needed primarily to inform the design of future intervention studies including randomized trials. Currently documented rates of progression of CVD risk, particularly in those with type 2 diabetes mellitus are unacceptable. Fortunately, there are ongoing studies, including the SEARCH for Diabetes in Youth Study (subjects with type 1 and type 2 diabetes) and the TODAY study (subjects with type 2 diabetes mellitus) that can accomplish this goal.^{7, 8} These studies must identify intermediate outcomes, including measures of arterial stiffness, cardiac function, and subclinical atherosclerosis that are reliably measured and can be used in trials initiated in youth.

A second step will be considering non-cardiovascular interventions to better accomplish prevention goals. Health disparities identified by Maahs et al for both type 1 and type 2 diabetes must be considered.⁵ For type 2 diabetes, race/ethnic and socioeconomic status specific efforts to prevent obesity and delay onset must be undertaken. Consideration of screening programs for high risk individuals, based on race/ethnicity specific criteria (as opposed to Caucasian) criteria, must be developed. For both type 1 and type 2 diabetes, understanding factors contributing to poor treatment compliance in specific at risk groups is critical. Patient centered research that better understands patient perspectives and barriers to compliance with care are needed.^{9, 10} Maahs et al document a high rate of depressive symptoms in diabetics with associated poorer glucose control and potential adverse effects of sleep disturbances.⁵ Interventions to control depression, improve self esteem, and improve sleep habits may lead to beneficial effects on diabetes control. Efforts to address ophthalmologic, renal, and peripheral vascular outcomes will improve quality of life and may also address small vessel disease injury and downstream large vessel disease.

A third step, and difficult to study, will be addressing the primary problem in diabetes-- lack of insulin. While insulin resistance is a consequence of type 1 diabetes mellitus treatment, helps precipitate type 2 diabetes mellitus, and creates the intermediate phenotype of double diabetes, only insulin treatment can help control the adverse impact of dysglycemia.¹¹ Almost 50% of patients of TODAY study patients progressed to need for insulin during the first 2 to 6 years of the study.⁷ Dysglycemia by itself and by an association with worsening of CVD risk is the primary culprit in target organ injury.³ The complications of excessive insulin therapy, including progressive obesity and insulin resistance in type 1 diabetes, are well documented. Nonetheless, the contributions of dysglycemia to the initiation of CVD injury, the presence of a legacy effect of early rigorous glucose control on CVD outcomes in the DCCT/EDIC studies, and the lack of benefit of many aggressive risk reduction treatments late in the disease cannot be ignored.¹²

A fourth step is too include truly effective interventions, though more aggressive, in clinical trial design. Here, bariatric surgery in those with type 2 diabetes mellitus leaps to mind.^{13, 14} Though this treatment might not be a cure, it certainly holds promise for reduction of insulin resistance and preserving beta cell function for a longer time interval. Trials comparing bariatric surgery to conventional treatment should be considered.

Finally, at least some trials, particularly those with target organ damage outcomes, must provide for some long term clinical follow up, perhaps for 10-15 years, sufficiently long to assess hard CVD outcomes. Without these studies, the true clinical value of risk intervention and intermediate outcomes will remain unknown.

We no longer need research to tell us diabetes mellitus is bad for us. To improve outcomes for children with diabetes, we need well-designed clinical trials. We need clinical trials. We need clinical trials. We need clinical trials. Do I need to repeat myself?