Long-term predictive power of simple function tests for cardiovascular autonomic neuropathy in diabetes: a population-based study.

No consensus exists on which test is optimal in screening for cardiovascular autonomic neuropathy in diabetic patients. The long-term predictive power of five simple function tests with regard to all-cause mortality in individuals with diabetes from the general population was assessed. The diabetic population in the municipality of Horsens, Denmark, was delineated by the prescription method, and an age- and gender-stratified sample of 240 diabetic persons was randomly selected. Five function tests were conducted: Valsalva ratio, heart rate response to standing (30:15 ratio), expiration/inspiration ratio (E/I ratio), ortostatic blood pressure response (Orto BP) and increase in diastolic blood pressure during sustained handgrip. After a 15½-year follow-up period, vital statistics were obtained from the Danish Civil Registration System. One hundred and seventy eight persons participated, 44% women, 40% type 1 diabetes, mean age 58.9 (SD 10.2) years. In Cox regression analysis, adjusting for age and gender, each test had a significant predictive effect on mortality (all P < 0.05). In the multivariate analysis, only Valsalva ratio (P = 0.0017), 30:15 ratio (P = 0.037) and handgrip (P = 0.037) were predictors of survival. As a binary variable, the presence of CAN was associated with a relative risk of all-cause death of 2.85 (95%CI: 1.75-4.65, P = 0.00003). In the period following the first 5 years, CAN's predictive value remained significant. This study indicates that Valsalva, 30:15 ratio and handgrip, but not E/I ratio and Orto BP, have an independent predictive value with regard to long-term all-cause mortality. Baseline function tests hold prognostic information after at least 5 years.