Measurable urinary albumin predicts cardiovascular risk among normoalbuminuric patients with type 2 diabetes

Abstract

Micro- or macroalbuminuria is associated with increased cardiovascular risk factors among patients with type 2 diabetes, but whether albuminuria within the normal range predicts long-term cardiovascular risk is unknown. We evaluated the relationships between albuminuria and cardiovascular events in 1208 hypertensive, normoalbuminuric patients with type 2 diabetes from the BErgamo NEphrologic Diabetes Complication Trial (BENEDICT), all of whom received angiotensin-converting enzyme inhibitor (ACEI) therapy at the end of the trial and were followed for a median of 9.2 years. The main outcome was time to the first of fatal or nonfatal myocardial infarction; stroke; coronary, carotid, or peripheral artery revascularization; or hospitalization for heart failure. Overall, 189 (15.6%) of the patients experienced a main outcome event (2.14 events/100 patient-years); 24 events were fatal. Albuminuria independently predicted events (hazard ratio [HR], 1.05; 95% confidence interval [CI], 1.02-1.08). Second-degree polynomial multivariable analysis showed a continuous nonlinear relationship between albuminuria and events without thresholds. Considering the entire study population, even albuminuria at 1-2 μg/min was significantly associated with increased risk compared with albuminuria <1 μg/min (HR, 1.04; 95% CI, 1.02-1.07). This relationship was similar in the subgroup originally randomly assigned to non-ACEI therapy. Among those originally receiving ACEI therapy, however, the event rate was uniformly low and was not significantly associated with albuminuria. Taken together, among normoalbuminuric patients with type 2 diabetes, any degree of measurable albuminuria bears significant cardiovascular risk. The association with risk is continuous but is lost with early ACEI therapy.

Figure 1.

Adjusted HRs for major cardiovascular events according to baseline albuminuria. Solid line shows estimated relation when logarithmic hazard is modeled as linear function of log (UAE). The reference value (HR = 1) is set at UAE = 1 μg/min. The shaded area includes the 95% CIs for more general functional relation, as estimated by P-splines. The HRs are adjusted for age, sex, duration of diabetes, history of cardiovascular events, smoking habit, body mass index, mean BP, logarithms of HbA_1c, LDL-to-HDL cholesterol ratio, triglyceride and serum creatinine levels, and whether patient was allocated to ACEI therapy. CVD, cardiovascular disease.

Figure 2.

Adjusted HRs for major cardiovascular events according to baseline albuminuria in patients who during the BENEDICT trial had been randomly allocated to ACEI therapy with trandolapril alone or combined to verapamil (solid line) or to non-ACEI therapy with verapamil alone or placebo (dashed line). The lines show estimated relation when logarithmic hazard is modeled as linear function of log (UAE). The reference value (HR = 1) is set at UAE = 1 μg/min. The HRs are adjusted for age, sex, duration of diabetes, history of cardiovascular events, smoking habit, body mass index, mean BP, logarithms of HbA_1c, LDL-to-HDL cholesterol ratio, triglyceride and serum creatinine levels, and whether patient was allocated to ACEI therapy. CVD, cardiovascular disease.