The aim of the study was to select parameters of power spectral analysis appropriate for clinical evaluation of various degrees of diabetic cardiovascular autonomic neuropathy. The diagnosis of autonomic neuropathy was based on Ewing´s battery of cardiovascular autonomic function tests and on short-term power spectral analysis of heart rate variability during a modified orthostatic load (in positions supine-standing-supine). 52 diabetic patients and 24 age-matched controls were included. Diabetic patients were divided into two groups according to the total Ewing score. Groups of 14 subjects with early autonomic neuropathy and 38 subjects with severe AN were comparable in age and diabetes duration. Results: Cumulative spectral power of the total frequency band in all three positions (Total Power LFHF 1 + 2 + 3) proved to be the most selective and discriminating parameter of power spectral analysis between the control group and diabetic patients with severe autonomic neuropathy in stepwise discriminant analysis. Spectral power of low-frequency band in positions 1 + 2 + 3 (Power LF 1 + 2 + 3) proved to be the most selective and discriminating parameter between diabetic groups with early and severe autonomic neuropathy. Conclusion: The most discriminating parameters of power spectral analysis of heart rate variability for the determination of autonomic function seem to be Total Power LFHF 1 + 2 + 3 and Power LF 1 + 2 + 3.

        Key words: Diabetes mellitus - Cardiovascular autonomic neuropathy - Heart rate variability - Power spectral analysis