Electrical Activity of the Heart Under Pressure

Willem Einthoven’s classic description of the normal and abnormal electrocardiogram (ECG) published as “Le Télécardiogramme” in the Archives Internationales de Physiologie in 1906 is one of the most significant milestones in medicine.1 The advances made by Einthoven on the modifications of the string galvanometer that could register deflections due to small electric currents in a wire in a strong magnetic field, enabled the first accurate recordings of ECG signals from patients in the “Academisch Ziekenhuis” in Leiden; signals which were transmitted over telephone wires from the hospital to his physiology laboratory at the University, a considerable distance across the city. [“The wires between the laboratory and the hospital are approximately 1 mile long. Their self-induction is negligible while their resistance is 106 ohms and their capacitance 0.075 pf. From this it may be calculated that wires of much greater length could be used successfully and that it would be possible to connect Leiden with The Hague or Haarlem: I believe that a practically usable connection could even be established between Leiden and Rotterdam or between Leiden and Amsterdam”].1 Although the first reported recordings of the ECG in humans was by the British physiologist A.D. Waller at Mary’s Medical School in London in 1887,2 it was the transtelephonic transmission of the ECG in clinical cases by Einthoven in Leiden that provided seminal information on the relationship between the morphology of the ECG signal and cardiac properties related to conduction disturbances and structural abnormalities. Descriptions included left and right ventricular and atrial hypertrophy, the first description of the U wave, notching of the QRS complex, ventricular premature beats, ventricular bigeminy, atrial flutter, and complete heart block.1 Einthoven’s standardization of the ECG recording using the 3 limb leads (Einthoven’s “triangle,” Leads I, II, III) and the ECG feature descriptors (P,Q,R,S,T,U) propelled numerous studies expanding the recording configurations to include augmented unipolar leads (aVR, aVL, aVF) and the chest (precordial) leads (V1-V6), leading to the now standard 12-lead ECG recording. Since this set of leads gives a fiducial description of the rotating 3-dimensional electric field vector projected on 2-dimenional planes, combinations of voltages of some or all the 12 leads can describe relationships between the electrical activity and co-localized alteration of myocardial structure and function. Using the 12-lead configuration, comprehensive diagnostic criteria for left ventricular hypertrophy (LVH) were first proposed by Sokolow and Lyon in 1949,3 where LVH is present based on a set threshold of the sum of the S wave in V1 and the R wave in V6. Due to specific limitations (e.g., the Sokolow–Lyon criteria are not gender specific), these criteria were refined and expanded by many other investigators, and with the advent of echocardiography in recent decades, it has been possible to obtain greater specificity and sensitivity. The most notable has been the work emanating from the Cornell investigators,4–6 where LVH is diagnosed based on the “Cornell Voltage ” (sum of S wave in precordial lead V3 and R wave in limb aVL, [mV]) and the “Cornell Index” (Cornell Voltage multiplied by QRS duration [mV.ms]). Comparison of the Sokolow–Lyon and Cornell criteria show notable differences.7 The routine clinical use of the ECG has greatly facilitated large-scale studies in assessing the effect of blood pressure on LVH, and particularly the efficacy of antihypertensive agents in regression of LVH.8 Due to the association of ECG voltages with LVH and given that most LVH is due to ventricular remodelling in response to the arterial pressure load,8,9 recent studies have attempted to provide evidence of a possible association of ECG voltages with arterial pressure per se in terms of transvalvular pressure gradients10 or conventional measurements of blood pressure.11 In the presence of LVH, seeking a relationship between ECG voltages and blood pressure would seem to be a tautological exercise, given the interdependence of LVH and blood pressure. However, since the propagation of the electrical signal throughout the myocardium is influenced by local ionic concentrations and potential mechanical effects due to myocardial stress affecting the mass potential,12 it is conceivable that during systole, when there is continuity between arterial and ventricular compartments, it might be plausible to expect direct effects of arterial pressure on membrane depolarization during the systolic phase of the cardiac cycle, and thus affect the 3-dimensional electric field vector detected as voltage changes in the 12-lead ECG. The study by Yang et al.13 in this issue of the American Journal of Hypertension assesses the association of ECG voltage and arterial blood pressure in a cohort (n = 177) of young male workers (mean age 29.1 years) engaged in process work related to lead recycling. The screening involved ECG recording and measurement of 24-hour ambulatory blood pressure with a device (Mobil-O-Graph) that estimates central aortic pressure from conventional brachial cuff measurements.14 Due to the relatively high sensitivity and specificity, the Cornell Voltage and Cornell Index4 were used to quantify the association of ECG voltages and central and peripheral blood pressures. The finding that systolic, diastolic, and pulse pressure at the aorta and brachial artery show similar diurnal rhythm is not unexpected due to the high correlation between central and peripheral pressure. However, due to the type of process work performed by the workers, the study suffered from uncontrolled pressure recordings during working hours, producing a greater variability of the daytime recordings. The relatively lower reduction in central compared to brachial systolic pressure at night (and so an increase in central pulse pressure) was clearly due to the reduction in heart rate, hence causing a lower pulse amplification,15 although this was not explicitly addressed in the study. The Cornell Voltage and Cornell Index had a stronger association with nighttime compared to daytime blood pressures due to the reduced variability in the nighttime pressure registrations. According to the Cornell criteria for LVH, the prevalence of LVH in the study cohort was in the range 0.6–3.4%, hence there was no categorical analysis. Notwithstanding such a low prevalence of LVH, predominantly due to the low mean age of the cohort, robust associations were found for both Cornell Voltage and Cornell Index for central and brachial systolic pressure. The intriguing finding is that a large range of the Cornell Voltage (5th to 95th percentile interval, 0.29–2.34 mV) and Cornell Index (25.5–230.4 mV·ms) was found for a relatively small systolic pressure range of <10 mm Hg. The sensitivities of the ECG voltages were similar for central and peripheral systolic pressure changes. The premise of the study of Yang et al.13 was based on the fact that since aortic systolic pressure is closer to the mechanical load on the left ventricle, there would be a closer association of ECG voltages with central compared to brachial systolic pressure. This was not found in this cohort. This is in contrast to the variable and different associations of central and peripheral pressure with LVH in a number of studies.16,17 Other studies also found variable associations with ECG voltage indices related to LVH and central and peripheral systolic pressure.18,19 Whereas these studies showed the link with ECG voltages, LVH, and aortic and peripheral systolic pressure, the study by Yang et al.13 shows the potential to obtain associations of ECG voltages and blood pressure before the development of LVH, that is, before structural alterations of the myocardium that can alter the ECG waveform morphology. The intimate association between regional myocardial electrical activity with pressure-related indices has also been shown with the relation between pacing modality and hemodynamic parameters.20 Given the study limitations, Yang et al.13 demonstrate a solid proof of concept that relationships of ECG voltages can be obtained with systolic blood pressure in normotensive subjects with little or no prevalence of LVH. The implication being that stress on the myocardial fibers determined by the peak pressure in the chamber during systole can affect membrane depolarization causing changes in the propagated electrical field to the body surface, and so detected as voltage changes in the 12 leads of the ECG. This study used the conventional Cornell indices that have been developed for classification of presence of LVH. However, other representations of the ECG waveform have recently been developed that are based on fitting of Gaussian functions, capable of producing robust parametric descriptions of the underlying mass potential and particle populations that constitute the relationship between cell membrane depolarization and the surface ECG signal.12 These models could also be applied to uncover potential associations of ECG intrinsic parameters and blood pressure indices, and compared with existing ECG voltage indices in asymptomatic individuals for early screening. While the early seminal foundations of the ECG were described by Einthoven to diagnose pathological changes of the heart,1 the novel study by Yang et al.,13 together with more recent approaches to quantify ECG morphology12 and to obtain associated information on electrical and hemodynamic parameters,20 can contribute to avenues for improved means of assessing the normal cardiovascular condition before it progresses to irreversible pathology. The authors declared no conflict of interest.