Intensive Glycemic Control Improves Long-term Renal Outcomes in Type 2 Diabetes in the Veterans Affairs Diabetes Trial (VADT).

Diabetes accounts for almost 50% of cases of end-stage renal disease (ESRD) in high-income countries (1). Although the impact of intensive glycemic control on macrovascular complications is relatively modest, evidence suggests that tight glucose control may reduce kidney disease progression. The role of glucose control in preventing chronic kidney disease and ESRD continues to be of significant interest.

The Veterans Affairs Diabetes Trial (VADT) was a randomized, prospective, multicenter study in 1,791 veterans with poorly controlled type 2 diabetes, primarily to determine the impact of 5.6 years of tight versus standard glycemic control on cardiovascular events; secondary outcomes included microvascular disease. After the 5.6-year interventional period, the cohort was observationally followed for an additional 10 years (median follow-up was 15 years, ending in 2017) and provided a unique view of the impact of glycemic control on important clinical outcomes. As previously reported, baseline demographics, clinical characteristics, statin and blood pressure medicine use, and mean HbA1c (9.4 ± 2% [79.2 mmol/mol]) were balanced between the two groups (2). Participants received either intensive (INT) or standard (STD) glucose control for a median of 5.6 years. The median HbA1c achieved was 1.5% lower in INT versus STD during the intervention phase (6.9% vs. 8.4% [52 vs. 68 mmol/mol]). Subsequent diabetes care was provided by their primary health care team, and participants were followed using the Department of Veterans Affairs (VA) medical records system and three national data registries.

Reported here is the long-term effect of intensive glycemic control on an important renal composite outcome, namely, sustained estimated glomerular filtration rate (eGFR) <45 mL/min/1.73 m2 or urine albumin-to-creatinine ratio >300 mg/g creatinine. We selected this composite since the relative risk for additional complications, including ESRD, cardiovascular disease, and all-cause mortality further increases once eGFR drops below 45 mL/min/1.73 m2, and macroalbuminuria is a better risk marker for eGFR loss (3). For the outcome to be considered sustained, the last recorded value at the end of a study period had to meet the composite definition, and there should have been a previous value of eGFR <45 mL/min/1.73 m2 or urine albumin-to-creatinine ratio >300 mg/g. An event time was assigned when either of the events of the composite outcome first occurred. Baseline eGFR or baseline albuminuria was not used to adjust the model. Kaplan-Meier survival curves generated by the product-limit method using intention-to-treat analysis are shown in Fig. 1. At the end of the VADT interventional period, there was a nonsignificant 13% reduction in time to occurrence of the composite renal outcome in the INT group (95% CI 0.70–1.074, P = 0.19). Three years after the intervention phase ended, the difference in HbA1c between INT and STD declined to 0.2–0.3%. At the 10-year interim analysis, 445 people had reached the composite renal outcome: 22.7% in INT and 27.0% in STD groups, with a 20% reduction in the time to composite renal outcome in the INT group (95% CI 0.66–0.96, P = 0.02). By the end of the final VADT follow-up in 2017, 496 people had reached the composite outcome: 25.7% in INT and 29.7% in STD groups, for a 19% reduction in the INT group (95% CI 0.68–0.97, P = 0.02).

Figure 1

Kaplan-Meier analysis of time to renal composite from randomization in the VADT to end of 15 years (15 November 2000–30 June 2017). The x-axis shows “subjects at risk” at each year since randomization in the INT and STD cohorts. HR 0.81, 95% CI 0.68, 0.97; P = 0.02.

In this extended follow-up of the VADT cohort for 15 years, we found that 5.6 years of intensive glucose lowering reduced a composite renal outcome of development of persistent stage 3b chronic kidney disease or severe albuminuria. Consistent with other trials, there was a lag time for the effects of intensive glucose control to manifest (4). Of great interest is that the beneficial renal effect was sustained even after the separation of 1.5% in HbA1c had disappeared. This benefit from glucose lowering does not appear to be explained by simultaneous improvements in other key risk factors, since use of medicines for treating diabetes, blood pressure, and lipid control were similar in both groups. Recent studies with glucagon-like peptide 1 receptor agonists and sodium–glucose cotransporter 2 inhibitors have shown lower rates of development and progression of diabetic kidney disease despite relatively modest decline in HbA1c (5). These agents were not frequently used in the VADT, and it will be interesting to see if their use, on a background of good blood glucose control, could further reduce progression of kidney disease.

These VADT findings are important from a clinical perspective since they demonstrate that about 6 years of intensive glycemic control provided sustained reductions in renal outcomes at 15-year follow-up, even when using older and less expensive medicines, suggesting a possible beneficial legacy effect. Whether the reduction in the end points reported here translates into reduced dialysis, transplant, and renal death remains an important clinical question.