Because global T wave inversion has not been specifically characterized, 100 electrocardiograms (ECGs) with this pattern (frontal plane T vector-100-degrees to -170-degrees with precordial T inversion) were prospectively collected from approximately 30,000 consecutively interpreted ECGs and analyzed blindly. There was a striking female predominance (82 women vs. 18 men; p < 0.0005) despite an essentially equal number of female and male hospital admissions. There was a single statistically ECG correlate: a more vertical QRS axis in women (+ 14.1-degrees +/- 45.3-degrees vs. -5.6-degrees +/- 31.3-degrees; p = 0.034). The T waves were basically symmetric (68%), the influence of this factor usually altering the characteristically asymmetric T wave inversions of right bundle branch block (4 of 5) and left ventricular hypertrophy (21 of 36). Asymmetry was mainly associated with digoxin therapy (21 of 32 patients taking digoxin; p less-than-or-equal-to 0.0005) and a corrected QT (QTc) interval (0.433 +/- 0.095) shorter than with symmetric T wave inversions (0.507 +/- 0.074; p less-than-or-equal-to 0.0005) though not reaching the degree of shortening expected for digitalization. Twenty-eight patients admitted for acute myocardial infarction and 23 for a central nervous system disorder accounted for the majority of patients with symmetric T wave inversion. Fifteen of 18 patients who had coronary angiography hd some degree of coronary artery disease; 3 had angiographically normal coronary arteries. The 8% death rate in this series (eight deaths, all in women) did not differ statistically from the 7.02% death rate for all medical admissions (p = NS) or the total adult hospital death rate of 4.18% (p = 0.098), suggesting no prognostic implication of the ECG pattern itself. Although such global T wave inversion could not be attributed to any specific anatomic diagnosis, catecholamine excess was demonstrated in case studies of individuals with this pattern. Global T wave inversion is a nonspecific primary T wave change, possibly catecholamine related, of no intrinsic prognostic implication for hospital patients and with an unexplained, highly significant female preponderance.
