Time and frequency domain estimates of spontaneous baroreflex sensitivity provide early detection of autonomic dysfunction in diabetes mellitus

Summary Diabetic autonomic dysfunction is associated with a high risk of mortality which makes its early identification clinically important. The aim of our study was to compare the detection of autonomic dysfunction provided by classical laboratory autonomic function tests with that obtained through computer assessment of the spontaneous sensitivity of the baroreceptor-heart rate reflex (BRS) by time domain and frequency domain techniques. In 20 normotensive diabetic patients (mean age ± SD 41.9 ± 8.1 years) with no evidence of autonomic dysfunction on laboratory autonomic testing (D0) blood pressure (BP) and ECG were continuously monitored over 15 min in the supine position. BRS was assessed as the slope of the regression line between spontaneous increases or reductions in systolic BP and linearly related lengthening or shortening in RR interval over sequences of at least 4 consecutive beats (sequence method), or as the squared ratio between RR interval and systolic BP spectral powers around 0.1 Hz. We compared the results with those of 32 age-matched normotensive diabetic patients with abnormal autonomic function tests (D1) and with those of 24 healthy age-matched control subjects with normal autonomic function tests (C). Compared to C, BRS was markedly less in D1 when assessed by both the slope of the two types of sequences (data pooled) and by the spectral method (–71.3 % and –60.2 % respectively, both p < 0.01). However, BRS was consistently although somewhat less markedly reduced in D0, the reduction being clearly evident for all the estimates (–57.0 % and –43.5 %, both p < 0.01). The effects were more evident than those obtained by the simple quantification of the RR interval variability. These data suggest that time and frequency domain estimates of spontaneous BRS allow earlier detection of diabetic autonomic dysfunction than classical laboratory autonomic tests. The estimates can be obtained by short non-invasive recording of the BP and RR interval signals in the supine patient, i. e. under conditions suitable for routine outpatient evaluation. Diabetologia (1997) 40: 1470–1475] Received: 30 April 1997 and in revised form: 31 July 1997