We thank Dr. Finucane for his interest in diabetes translation and his recognition of its importance. He questioned the effectiveness of interventions to improve glycemic control and whether such interventions are ready for translation. We agree that no clinical trial has shown that improved glycemic control prevents single end-stage microvascular complications, such as blindness and kidney disease (1). However, such end-stage events are rare, and researchers would need to conduct decades of follow-up to detect enough cases for meaningful analyses. More often, clinical trials have either examined conditions antecedent to end-stage complications (for example, retinopathy rather than blindness) or combined antecedent and end-stage complications into an aggregate measure.

For example, results of the Diabetes Control and Complications Trial among persons with type 1 diabetes mellitus showed that compared with patients receiving conventional therapy, those receiving intensive therapy had 76% less risk for developing new retinopathy and 54% less risk for established retinopathy (2). In the United Kingdom Prospective Diabetes Study of conventional and intensive glycemic control among persons with newly diagnosed type 2 diabetes (3), researchers developed aggregate outcomes that included several diabetes-related end points. "Any diabetes-related end point" included sudden death, death from hyperglycemia or hypoglycemia, fatal and nonfatal myocardial infraction, angina, heart failure, stroke, renal failure, amputation, vitreous hemorrhage, retinal photocoagulation, blindness in one eye, or cataract extraction. The researchers used this aggregate outcome to compare the results of conventional treatment with results of intensive treatment and found that intensive treatment significantly reduced any diabetes-related end point (40.9 versus 46.0 events per 1000 patient-years; P = 0.03).

The outcomes of interventions depend on the duration of a patient's diabetes, on the duration of the study's follow-up, and on the quality of care provided outside the clinical trial. Observational studies may help sort these issues out. For example, the Wisconsin Epidemiologic Study of Diabetic Retinopathy had a large enough study sample to give us some insight into the risk for end-stage disease associated with a given glycemic exposure (4).

For translation, we are interested in many end points in aggregate, not necessarily single outcomes, because aggregate outcomes can greatly affect quality of life. As such, the evidence for reducing microvascular complications is sound, but we need more direct evidence of effectiveness in terms of long-term end-stage complications.