Time and frequency analysis of beat-to-beat R-T interval variability in patients with ischaemic left ventricular dysfunction providing evidence for non-neural control of ventricular repolarisation.

BACKGROUND

Determinants of temporal lability in ventricular repolarisation are not fully recognised. We aimed to analyse the sources of RT variability by comparing normal subjects and patients after myocardial infarction (MI) with either depressed or preserved left ventricular (LV) function.

METHODS

One hundred and nine patients (27 women, 82 men, aged 51 +/- 9 years) were divided into three groups: 24 patients (pts) with an uncomplicated angiographically proven coronary heart disease (CHD-group), 59 post-MI pts with preserved LV function (LVEF > 40%, PMI-N-group) and 26 post-MI pts with depressed LV function (LVEF < 40%, PMI-L-group). An ECG signal of low-noise 512 heartbeats was recorded using a computer-assisted amplifier (16 bit, 2 kHz). The onset and offset of the R-wave and T-wave were determined automatically. The magnitude of R-R and R-T variability was measured as the standard deviation of all intervals (SD-RR and SD-RT, ms, respectively). Their relationship was quantified by the correlation coefficient r(RT/RR). Power spectral density of RR or RT variability was estimated with the FFT (Welch's averaged periodogram, Hanning window) and frequency relation was quantified using a squared coherence spectrum (SCS). For all spectral and cross-spectral measurements two frequency ranges were considered: high (0.15-0.50 Hz, HF) and low (0.04-0.15 Hz, LF). Spectral power and SCS of RR and RT variability for both ranges (HF(RR), LF(RR), HF(RT), LF(RT), SCS(HF), SCS(LF)), and the ratios LF/HF(RR) and LF/HF(RT) were drawn for comparisons. The central frequency of HF(RR) was considered as the frequency of respiration (f(resp), Hz).

RESULTS

In the PMI-L group the SD-RT was significantly greater compared to the remaining groups and accounted for almost 10% of the SDRR. Also, the coefficient r(RT/RR) was weakest in this group. The spectral indices of RR variability were similar in all groups, while the greatest value of the HFRT was observed in the PMI-L group. The SCS(LF) was insignificant in this group, contrary to the CHD and PMI-N groups. Additionally, there were significant negative relationships between f(resp) and spectral indices of RT variability in PMI-patients with depressed LV function.

CONCLUSION

A greater beat-to-beat variation in RT interval duration along with increased power of its HF component indicates an important role of respiration in ventricular repolarisation control, while reduced time- and frequency RT-RR relationships seem to relate to an impaired process of ventricular duration adaptation.