Respiratory rate is an important factor influencing heart rate variability (HRV), but is usually ignored in the studies in patients with ischaemic heart disease (IHD). This factor may significantly contribute to the changes in HRV in ischaemic left ventricular impairment. We examined 186 patients with IHD (46 females, 140 males), including 54 pts with left ventricular ejection fraction < 40% and 132 pts with preserved LVEF (> or = 40%). HRV was evaluated in supine and standing position from short-term ECG (512 beats, 1 kHz, 12 bit) in time- (mean RR interval--RRI, and standard deviation of RRI-SD-RR) and frequency-domain (FFT). Power spectral density of high and low frequency (HF 0.14-0.5 and LF 0.04-0.15 Hz) was estimated and expressed as log-values. The frequency of respiration (fResp, Hz) was defined as the frequency of maximum HF power. RESULTS: Patients with LVEF < 40% had shorter RRI in supine (879 + 157 vs 942 +/- 147 ms, p < 0.05), but similar in standing. FResp was greater both in supine (0.308 +/- 0.05 vs 0.274 +/- 0.047 Hz) and in standing (0.329 +/- 0.055 vs 0.274 +/- 0.047, all p < 0.001) in low, as compared to preserved LVEF. As expected, HRV parameters were more reduced in LVEF < 40%. The differences of raw values were as follows: SD-RR 29 +/- 16 vs 34 +/- 15 i 27 +/- 13 vs 31 +/- 12 ms, both p < 0.05, HF 3.21 +/- 0.36 vs 3.33 +/- 0.32, p < 0.05 and 3.19 +/- 0.32 vs 3.21 +/- 0.29 p > 0.1, LF 3.41 +/- 0.33 vs 31.49 +/- 0.30 and 3.52 +/- 0.30 vs 3.61 +/- 0.32, p < 0.05, ANOVA). The differences were observed despite adjustment for age, sex and RRI. When fResp was added as covariate they became not significant. CONCLUSION: Patients with left ventricular dysfunction had faster rate of respiration. This finding may explain, in part, reduced HRV in these patients. The rate of respiration should be taken into account for correct interpretation of HRV analysis.