Identifying subgroups benefiting from intensive glycaemic treatment to improve renal outcomes in type 2 diabetes: Insights from the ACCORD trial.

Abstract

The ACCORD trial showed that intensive glucose-lowering therapy has a limited impact on renal function decline. We aimed to identify subgroups in the ACCORD population that might derive renal benefits from intensive glucose-lowering therapy. The primary renal outcome included a ≥50% decline in baseline estimated glomerular filtration rate or end-stage renal disease (ESRD). Using the causal tree model, we employed internal cross-validation to identify five pivotal variables influencing the renal efficacy of intensive glycaemic control. These variables were integrated into the model-based recursive partitioning approach, yielding a visualizable tree model that depicted benefitting subgroups. Node 4, characterized by no cardiovascular history, systolic blood pressure (SBP) ≤142.67 mm Hg, and triglycerides ≤172 mg/dL, showed significantly reduced hazards of the composite renal outcome (fully adjusted hazard ratio [HR] 0.66, 95% confidence interval [CI] 0.49-0.89; p = 0.006) and doubling of serum creatinine (fully adjusted HR 0.59, 95% CI 0.36-0.98; p = 0.041). Node 7 (no cardiovascular history and SBP 142.67-154 mm Hg) showed reduced hazards of the primary renal outcome (fully adjusted HR 0.67, 95% CI 0.49-0.93; p = 0.016) and ESRD (fully adjusted HR 0.35, 95% CI 0.17-0.74; p = 0.0057). Encouragingly, neither node 4 nor node 7 displayed elevated cardiovascular risk or hypoglycaemic events. Through innovative machine learning, we identified ACCORD subgroups benefitting significantly from intensive glycaemic therapy for renal outcomes, without increased cardiovascular or hypoglycaemic risks.