EVALUATION OF DEGREE OF DIABETIC DYSAUTON OMY WITH THE ANALYSIS OF HEART-RATE-VARIABILITY MEASURED IN REST PERIOD WITH FINAPRES SYSTEM

In this work, we developped the ERK (Eckmann, Ruelle and Khamporst) me thod of recurrent plots to analyse Heart Rate Variability (HRV), measu red by the FINAPRES system in diabetic subjects. Our aim was to search some indices that might characterize the degree of dysautonomy detect ed in diabetic subjects, using the Ewing tests. The idea was to analyz e the recurrences of the HR to previously observed values; When a valu e of HR, x(j) measured at the j-th beat come back to an already observ ed value, we compare the distances between the following measures (x(j +1) to (x(j+1) x(j+2)), (x(j+2) to x(j+2))... If the distances are und er a given criteria during k beats, x(i) is deterministic of order k. Let n(1) be the number of recurrent points and n(k) the number of k-or der deterministic points, with k = 2, 3, 4... We defined the index of determinism n(k+1)/n(k) and the Shannon entropy of the N-k=n(k)-n(k+1) . These indices will be correlated to the total score of the 5 Ewing t ests, which represent the standard measure to evaluate the diabetic dy sautonomy. Blood pressure (BP) and HR were measured during about 30 mi nutes using the FINAPRES system in 44 healthy subjects and 60 non-hype rtensive diabetic subjects. In the diabetic subjects, the age, the bod y weight and systolic and diastolic BP were 56+/-13 years (mean+/-stan dard-deviation), 79+/-14 kg, 80+/-10 mmHg and 137+/-20 mmHg. HR was we alky correlated to age (r=0.29; p=0.02) and the Ewing score (r=0.31; p =0.01), Its standard-deviation is also wealky correlated to age (r=0.3 2; p=0.01) and to the Ewing score (r=0.34; p=0.01). Using the ERK meth od, we obtained indices which are much more correlated to the Ewing sc ore. In particular, the ratio n(2)/n(1) and Shannon entropy were corre lated to the Ewing score with r=0.51 and r=0.53, respectively (p <0.00 01 in both cases). These indices are also correlated to age (r=0.40; p =0.003) in both cases, The ERK method give some indices which are easy to obtain (measurement of HR during a rest period) and easy to interp ret. These indices are strongly linked to the dysautonomy score which required a good cooperation of the patient and a great vigilance durin g its execution. This method could be applied to BP to explore BP regu lation in hypertensive subjects.