Is the T-wave alternans magnitude in apparently healthy subjects and in different subsets of patients with ischaemic heart disease T-wave amplitude dependent? Reply to the letter by J.E. Madias

We thank Dr Madias for his interest in our study on the distribution of T-wave alternans (TWA) magnitude in apparently healthy subjects and in patients with ischaemic heart disease.¹In his letter, Dr Madias suggests that TWA magnitude might depend on T-wave amplitude, a very interesting hypothesis that he has been proposing for some years, raising the possibility that adjusting TWA magnitude for T-wave amplitude might result in a TWA index able to improve the diagnostic and prognostic yield of the method.^2–4

To address the issue raised by Dr Madias, we have analysed the exercise stress test electrocardiograms (ECGs) of a random sample of 175 subjects enrolled in our study,¹including 106 apparently healthy individuals and 69 patients with documented ischaemic heart disease (IHD). T-wave amplitude in the lead where maximal TWA magnitude was detected has manually been measured as the amplitude from isoelectric line (PR point) to T-wave peak.

Spearman correlation analyses showed, in fact, an association between T-wave amplitude and TWA magnitude, which achieved statistical significance for TWA measured in precordial leads only; the correlation coefficients, however, were low (Table 1). Of note, there seemed to be no major differences in the TWA-T-wave relation when considering separately normal subjects and IHD patients, although higher correlation coefficients in the two subgroups were differently shown for TWA measured in all ECG leads and in precordial leads only, respectively (Table 1). Thus, the analyses in a random subset of individuals enrolled in our study¹actually show some influence of T-wave amplitude on TWA magnitude, but the entity of the relation seems weak.

As far as the second point raised by Dr Madias, we would just like to observe that, as TWA magnitude by modified moving average is measured as the highest value recorded in any ECG lead, increasing the number of leads (all 12 leads vs. 6 precordial leads) used for the measurement likely simply increases the probability to detect a higher TWA magnitude in individual patients, which eventually results in a TWA distribution moved towards higher values.

Conflict of interest:none declared.

References