Handgrip-Induced Negative U-Wave in Electrocardiogram of Hypertensive Subjects

Long Tai Fu, M.D., Nobumitsu TAKAHASHI, M.D., Machiko YAMAMOTO, B.S., Masao KUBOKI, M.D., and Shintaro KOYAMA, M.D.

SUMMARY

To study clinical significance and possible mechanism of negative U- wave observed in hypertensive patients, we applied handgrip test (HG) to 35 patients with hypertension and 20 age-matched normal subjects. HG was sustained for 3min at 50% of maximum voluntary contraction. ECG recording and measurement of were made before, during, and after HG. During HG, systolic blood pressure was 188.7•}3.4mmHg (mean•} SE) and diastolic blood pressure was 114.8•}2.1mmHg in hypertensives, while 145.3•}2.6mmHg and 93.8•}1.1mmHg in normals; the difference between hypertensives and normals was statistically significant (p<0.001). Incidence of HG-induced negative U-wave was 0% in normals, while 68.6% in hypertensives. In hypertensive subjects, HG-induced nega- tive U-wave appeared in 24 of the 28 patients (85.7%) with minimum to moderate left ventricular hypertrophy, while it was not seen in all 7 pa- tients (0%) with marked left ventricular hypertrophy shown by ECG and/or UCG findings (QRS-T angle>130•‹ with left high voltage: left ven- tricular wall thickness>12mm). Further we recorded UCG during HG in 14 subjects. In 4 patients, negative U waves appeared simultaneously with increase in left ventricular dimension during . In the re- maining 10 cases that included both normals and hypertensives, however, neither increase in left ventricular end diastolic dimension nor negative U waves were observed. Thus we conclude that HG-induced negative U waves may be caus- ed by stretch of the ventricular wall resulting from acute left ventricular .

Additional Indexing Words:

Orthogonal leads LV wall thickness Acute LV dilatation

T is generally said that negative U-wave appears in electrocardiogram in 20-30% of hypertensive subjects.1),2) However, with regard to the clinical

From the Cardiovascular Institute, 3-10, 7-chome, Roppongi, Minato-ku, Tokyo, Japan. Address for reprint: Long Tai Fu, M.D., the Cardiovascular Institute, 3-10, 7-chome, Rop- pongi, Minato-ku, Tokyo 106, Japan. Received for publication April 15, 1980. 59 Jpn. J. J 60 FU, TAKAHASHI, YAMAMOTO, KUBOKI, AND KOYAMA anuary, 1981 setting in appearance of negative U-wave, there are some controversial results between statistical studies and clinical experiments. Kemp et al,1) as well as Watanabe,2) emphasized that the group of patients with negative U-wave showed a higher incidence of marked cardiomegaly and more advanced con- gestive heart failure, in addition to higher systolic and diastolic pressures than the group of patients without negative U-wave. Meanwhile, Georgopoulos et al3) observed that lowering of arterial pressure with sodium nitroprusside resulted in conversion of negative to positive U-wave in the majority of hy- pertensive patients, and thus they concluded that negativity of U-wave in hy- pertensives correlated only to arterial blood pressure and was not necessarily related to organic myocardial changes. To test the hypothesis proposed by Georgopoulos et al,3) and also to study relationship between negative U-wave and left ventricular hypertrophy we applied isometric handgrip test to hypertensive subjects with and without left ventricular hypertropy, in addition to normal volunteers. Our observa- tions may contribute to the clinical significance and possible mechanism of negative U-wave.

MATERIALS AND METHODS

Subjects: The material for this study was composed of 35 patients with hypertension

(Group H), and 20 age-matched normal subjects (Group C). Hypertension was de- fi ned as blood pressure measured at resting state equal to or greater than 160/95 mmHg at present or in the past history. Patients of Group H were further divided into 3 subgroups based on electro- and vectorcardiographic findings as follows: 10

patients with neither increased voltage of the QRS complexes nor ST•ET changes (Group H0); 18 patients with increased voltage of the QRS complexes (RV5 or V6+ SV1>3.5mV) alone, or coupled either with depression of the ST segment or with lowered or inverted , but with a QRS-T angle in transverse plane vector- cardiogram<130•‹ (Group H1); 7 patients with increased voltage of the QRS com-

plexes with a marked ST depression and T inversion, and with a QRS-T angle in transverse plane vectorcardiogram•†130•‹ (Group H2). Patients with asymmetric septal hypertrophy (ventricular septal/left ventricular posterior wall thickness ra- tio>1.3 in eehocardiogram4)) were excluded from this study. There were no patients with complication of acute myocardial infarction, con-

gestive heart failure, or cerebral vascular accident. Ischemic heart disease was com- plicated in 2 patients of Group H2, and none of other hypertensive groups. There was no patient with bundle branch block, ventricular preexcitation, or atrial fibril lation in studied cases. Handgrip exercise test: The maximum voluntary contraction (MVC) for the left hand of each subject was determined in Kg's as described by Lind et al5),6) and others.7),8) Each subject was then asked to sustain a constant effort by keeping the indicator needle on the Vol.22 No.1 HANDGRIP-INDUCED NEGATIVE U 61 dynamometer at 50% of his or her MVC for a period of 3min. A normal venti- latory pattern was observed in each subject throughout the test period. Before, dur- ing, and after handgrip corrected orthogonal leads X, Y, and Z (Frank lead system) were recorded simultaneously at a paperspeed of 50mm/sec and a gain of 20-30 mm/mV. In addition, in most cases unipolar chest leads V1 or V2 and V5 were also recorded. Blood pressure was measured with a mercury sphygmomanometer. Handgrip-induced NegativeU-wave: Handgrip-induced negative U-wave generally appeared in left-sided chest leads and lead X, and it was judged present when preexisted positive or biphasic, or neg- ative U-wave became absolute negative, or more negative in the same leads dur- ing handgrip. The negative U-wave was usually identified as a distinct negative deflection relative to the base line (P-P) between T and P waves. To avoid misac- cepting a negative terminal T-wave as a negative U-wave, we confirmed the timing of at least 2 out of the starting, peak, and end points of the negative U wave to be almost coincident with those of the positive U wave on the simultaneously recorded vertical lead (Y), where the morphology of U-wave was little affected by handgrip. We discarded the cases with of 100 beats per minute or more from this study, in which the terminal portion of U-wave often fused with the initial portion of P-wave and the correct observation of change in U-wave morphology was dis- turbed. Echocardiography: M-mode echocardiographic studies were performed in the supine or left

Table I. , Blood Pressure Before and During Handgrip, and Incidence of Handgrip-Induced Negative U-Wave in Hypertensive Patients and Age-Matched Control Subjects (mean•}SE)

BP=blood pressure; p=probability; *p<0.001, before vs during handgrip; NS=not significant. Jpn. Heart J. J 62 FU, TAKAHASHI, YAMAMOTO, KUBOKI, AND KOYAMA anuary, 1981 Vol.22 No.1 HANDGRIP-INDUCED NEGATIVE U 63 64 FU, TAKAHASHI, YAMAMOTO, KUBOKI, AND KOYAMA Jpn. Heart J. January, 1981 lateral decubitus position using an Aloka ultrasonoscope with a 2.25 megaherz me- dium (7.5cm) internally focused transducer. Permanent recordings were obtained with a Honeywell Recorder. Statistical analysis: Results were expressed as mean•}SE. A paired-samples t-test was used to as- sess changes for each group, and for comparison between groups an independent- samples t-test, or chi square test with Yates' correction was applied.

RESULTS

Summarized data of Groups C and H, detailed tabulation of data of Groups H0, H1, and H2 and statistical significance of differences in measure- ments are listed in Tables I, II, and III. Heart rate and blood pressure: With the handgrip, all of Groups C, H0, H1, and H2 showed a significant increase in heart rate with no difference between groups. There was a signi- fi cant rise in both systolic and diastolic blood pressures in each group during the third minute of sustained handgrip. Between Group C and Group H there was significant difference in degree of pressure rise induced by handgrip, and the attained pressure levels in Group H were significantly higher than those in Group C. Among subgroups of hypertension mean values of systolic

Table III. Statistical Significance of Measurements in Table II

Abbreviations: HG=handgrip; HR=heart rate; SBP=systolic blood pressure; DBP=diastolic blood pressure; HG-NU=incidence of HG-induced negative U-wave; NS=no statistically signi- fi cant difference. Vol.22 No.1 HANDGRIP-INDUCED NEGATIVE U 65 and diastolic blood pressures at rest, and also during handgrip were the hi- ghest in Group H2, intermediate in Group H1 and the lowest in Group H0, and the differences were significant in a great majority (Table III). How- ever, there was no difference in degree of pressure rise induced by handgrip in each hypertensive group. NegativeU-wave: In about two thirds of hypertensive subjects electrocardiographic U-wave became negative or increased its negativity during handgrip. As mentioned earlier, this change of U-wave morphology was most prominent in the left- sided chest leads, especially V4 and V5, and orthogonal lead X, while there were only minimal changes of U waves in other leads. Sequential changes during the grip test are illustrated in Fig. 1. A preexisted tiny negative U- wave increased in its negativity with the progression of time during handgrip, and returned to the control within 60sec after release of the handgrip. There were no subjects to show handgrip induced U-wave change among 20 subjects of Group C, who had had positive, flat or biphasic U-wave in lead X (or V5) before the handgrip. In Fig. 2 a case of Group C is shown. Dur- ing the handgrip blood pressure rose from 124/60mmHg to 150/98mmHg, together with an increase in heart rate, but there was no change in electro- cardiographic U-wave. On the contrary, in all of the 10 subjects of Group H0 the electrocardio- graphic U-wave in lead X (or V5) was inverted to negative from either posi- tive or biphasic polarity. In Group H1, 14 of the 18 subjects also showed the handgrip indued U-wave inversion or increase in U-wave negativity. In Fig. 3, one (Pt. No. 22) of the representative cases who showed handgrip induced negative U-wave is illustrated. This patient showed 190/90mmHg in blood

Fig. 1. Changes in U-wave before, during, and after sustained handgrip

(HG). Recordings of orthogonal lead X (Frank) with a gain of 1mV= 30mm (•~3). Preexisted negative U-wave (U) became enlarged during the

HG, and returned to control one after release of HG. With increasing of U-

wave negativity amplitude of positive T wave (T) increased slightly in this case, but there was no significant change in ST-Segment. Jpn. Heart J. J 66 FU, TAKAHASHI, YAMAMOTO, KUBOKI, AND KOYAMA anuary, 1981

Fig. 2. Changes in blood pressure (BP), heart rate (HR), and U wave

in leads V1, V5, and X during the handgrip (HG) in a normal subject. Dur- ing the HG BP rose from 124/60 to 150/98mmHg, together with an increase

in HR, however there was no change in U-wave. Lead X was recorded with

a gain of 1mV=20mm (•~2).

Fig. 3. An individual response to the grip test in a patient (No.22) of Group H1. During the handgrip (HG) blood pressure (BP) rose from 190/90 to 230/130mmHg, and negativity of U wave in leads V5 and X became more prominent. After release of the HG U-wave negativity returned to control with lowering of BP. pressure and a small negative U-wave in leads X and V5 before the hand- grip test. During sustained handgrip, blood pressure was elevated to 230/ 130mmHg and simultaneously negative U-wave increased its negativity, to- Vol.22 No.1 HANDGRIP-INDUCED NEGATIVE U 67 gether with an increase in heart rate. After release of handgrip U-wave nega- tivity returned almost to the control together with lowering of blood pressure and heart rate. The contrasted result in incidence of handgrip induced negative U-wave between Group C and Group H0 or Group H1 led us initially to presume that appearance of negative U-wave depended on high blood pressure, because at- tained pressure levels were significantly higher in both Group H0 and Group H1 compared to Group C. However, in Group H2, there were again no sub- jects who showed handgrip induced U-wave change though systolic and dias- tolic blood pressures rose to higher levels than those in Group H0 and Group H1. In Fig. 4 effect of handgrip on blood pressure, heart rate, and electro- cardiographic U-wave in a case (Pt. No.29) of Group H2 is illustrated. Dur- ing handgrip blood pressure rose from 180/120 to 240/140mmHg, together with an increase in heart rate. However preexisted negative U-wave did not increase its negativity at all. QRS complexesand ST, T waves: With regard to QRS complexes 7 of the 35 hypertensive subjects showed increase and 10 showed decrease in R-wave amplitude in lead X or other leads during handgrip. There were no cases to show increase in QRS duration. The change in R-wave amplitude was found to be unrelated to handgrip in- duced U-wave change. ST-segment in lead X or V5 did not change during handgrip in all subjects studied. Of the 35 hypertensives amplitude of T-

Fig. 4. An individual response to the handgrip test in a patient (No.29) of Group H2. During the handgrip (HG) blood pressure (BP) rose from 180/ 120 to 240/140mmHg, nevertheless there was no measurable change in U-wave negativity. Jpn. Heart J. J 68 FU, TAKAHASHI, YAMAMOTO, KUBOKI, AND KOYAMA anuary, 1981 wave in lead X or Z, or in both leads increased in 6 subjects and decreased in 3 subjects during handgrip. However there was no relationship between handgrip-induced changes in U and T waves. Echocardiography: Left ventricular end-diastolic wall thickness and dimension were observed in echocardiogram and compared among the hypertensive subgroups. As summarized in Table II, mean value of posterior wall thickness at rest was within normal limits9),10)in Group H0, increased slightly in Group H1, and increased markedly in Group H2. There was no significant difference in left ventricular end-diastolic dimension among Groups H0, H1, and H2. As shown in Fig. 5, handgrip-induced negative U-wave tended to appear in hyperten- sive subjects with less hypertrophied heart. Further we studied the relationship between handgrip-induced negative U-wave and change in left ventricular end-diastolic dimension in 14 sub- jects. In 4 hypertensive patients negative U-wave appeared simultaneously with increase in left ventricular end-diastolic dimension during handgrip. There were no observed abnormalities in local wall motion during . One of these cases is illustrated in Fig. 6. In the remaining 10 cases included both normal and hypertensive subjects, however, neither increase in left ven-

Fig. 5. The echocardiographically determined thickness of the left ven- tricular posterior wall (LVPW) in 7 patients of Group H0, 15 patients of Group

H1, and 7 patients of Group H2. Closed (•œ), and open (•›) circles indicate

patients with, and without handgrip induced U-wave change (either U-wave inversion or increase in U-wave negativity). Horizontal bar in each column shows mean value of the thickness in each group. The ventricular wall is

significantly thicker in Group H2 than in other groups (p<0.01). Handgrip induced U-wave change tends to be observed in patients with less hypertro-

phied left . Vol.22 No.1 HANDGRIP-INDUCED NEGATIVE U 69

Fig. 6. Simultaneous recordings of echo- and electrocardiogram (lead V5). During handgrip (HG) left ventricular end-diastolic dimension in- creased from 46 to 49mm, together with U-wave inversion. IVS=inter- ventricular septum; LVPW=left ventricular posterior wall; PCG=phonocar- diogram.

tricular end-diastolic dimension nor negative U-wave was observed.

DISCUSSION It is well known that negative U-wave appears in electrocardiogram of patients with hypertension, or ischemic, valvular, and congenital heart di- sease.2) Present studies were focused, however, on clinical significance and possible mechanism of negative U-wave related to hypertension. Clinicalsignificance of negativeU-wave in hypertensives: Kemp et al (1957),1) and later Watanabe (1966)2) showed that negative U waves were found in 20-30% of hypertensive patients in their statistical studies. These investigators also emphasized that the group of patients with negative U-wave had a higher incidence of marked cardiomegaly and more advanced congestive heart failure in addition to higher systolic and diastolic blood pressures than the group with normal positive U-wave. Surawicz et al (1957)11)studied the electrocardiograms of 297 cases with hypertension, reveal- ed that a negative U-wave in left-sided chest leads was accompanied with a negative T-wave in the majority of electrocardiograms, and an inverted U- wave in the presence of an upright T-wave was found in only 2.8% of the cases. These findings might suggest importance of organic myocardial 70 FU, TAKAHASHI, YAMAMOTO, KUBOKI, AND KOYAMA Jpn. Heart J. January, 1981 changes such as left ventricular hypertrophy, dilatation or subendocardial ischemia as a clinical basis of negative U waves. However, Surawicz et al11)also observed that in 3 cases negative U waves became positive or isoelectric following the lowering of blood pressures with antihypertensive agent or nitroglycerine without any changes in ST and T waves (Fig. 7 and 8 of the article by Surawicz and co-workers11)). The evi- dence clearly suggested that negative U-wave was not a fixed but a labile sign, and its appearance was possibly related to high blood pressure. Geor- gopoulos et al (1961)3) reported that following satisfactory lowering of arterial pressure with institution of antihypertensive treatment previously inverted U waves became upright in the majority of hypertensive patients regardless of organic myocardial changes. In addition, they also studied the relationship between level of arterial pressure and degree of U-wave negativity in 12 hy- pertensive patients, in whom blood pressure was lowered and kept at various levels with intravenous sodium nitroprusside, and they found a gradual de- crease in U-wave negativity proportional to arterial pressure fall. Thus, they concluded that presence of negative U-wave might be based on high blood pressure and did not necessarily indicate organic myocardial changes. We agree in part with the observations of Georgopoulos et al, in that handgrip induced negative U-wave was never observed in normal subjects, in whom blood pressure did not reach levels sufficient to make the U wave invert in hypertensive subjects. However, among the hypertensive subjects whose

Fig. 7. Negative U-wave during congestive heart failure in a patient (No.34) of Group H2 (A). Inverted U wave became flat after improvement of cardiomegaly (B). BP=blood pressure; CTR=cardiothoracic ratio on chest X-ray film. Vol.22 No.1 HANDGRIP-INDUCED NEGATIVE U 71 arterial pressure was elevated equally to the levels around 190mmHg in sys- tole and 115mmHg in diastole during handgrip, negative U-wave was observ- ed most frequently in the patients without left ventricular (LV) hypertrophy, less frequently in the patients with moderate hypertrophy, and was not ob- served at all in the patients with marked hypertrophy. This finding indicates that high arterial pressure is an important but not a crucial factor in the deve- lopment of U-wave inversion in hypertensive subjects. As demonstrated by the echocardiographic study, handgrip induced U- wave inversion (or increase in U-wave negativity) was found to depend sim- ply on whether LV wall could be stretched by a given high blood pressure. With regard to the reason why handgrip induced negative U-wave did not appear in hypertensive subjects with marked hypertrophied left ventricle we presume that left ventricular wall was hardly to be stretched against a given high blood pressure, because left ventricular wall stress remained less increas- ed in more hypertrophied heart. In 2 patients of Group H2, in whom negative U-wave had been observed spontaneously during heart failure and disappeared after treatment with digi- talis and diuretics (Fig. 7), handgrip failed to reproduce U-wave inversion, although it made arterial pressure rise equally to or more than the level before treatment. The evidence shows that increase in is an additional im- portant condition to form negative U-wave, and it is theoretically suggested that in the presence of a higher preload, U-wave negativity could be increas- ed by a smaller degree of arterial pressure rise. It is well known that nitro- prusside decreases not only but also preload,12) therefore nitroprus- side-induced decrease in U-wave negativity that Georgopoulos et al observed may be not only due to lowering arterial pressure, but also due to decreasing of preload. Four of the 7 patients in Group H2 had negative U-wave before hand- grip test. For the clinical background of these negative U waves, because their negativity was not increased by pressure rise, and also because left ven- tricular end-diastolic dimension of these patients was not enlarged strikingly except 1 case (Pt. No.33), a local myocardial ischemia, or other unknown factors may have to be taken into consideration.13)-15) In summary, negative U-wave in electrocardiogram of hypertensive sub- jects seems to be related to multiple factors. Among them, we have clarified the causal relation between the negative U-wave and the stretched left ven- tricle based on unproportional increase in either preload or afterload, or both against left ventricular wall. Genesisof handgripinduced negative U-wave: At least 4 categories of theory have been raised for the genesis of U-wave 72 FU, TAKAHASHI, YAMAMOTO, KUBOKI, AND KOYAMA J pn. Heart J. January, 1981 formation, i.e., (1) Purkinje fiber repolarization,16),17)(2) delayed repolariza- tion in some regions of the ventricle (septum,18),19)and basal portion of the ventricle14)), (3) delayed excitation sequence,14) and (4) stretch potentials (after-potential) of the ventricular myocardium.20) Most of these theories may have convincing powers to explain the genesis of normal positive U-wave or its accentuation, but may not give a responsible explanation for the mechanism how the normel positive U-wave turns to be negative by handgrip. From the discussion in the previous section it is clear that the interrelations between U-wave and stretching the left ventricular wall should be of our main interest. Lepeschkin20)proposed that the U wave resulted from afterpotentials in the ventricular transmembrane action potential, and that the normal posi- tive U-wave might be caused by stretching of predominantly the subendocar- dial layers and the papillary muscles during the repolarization process. Ac- cording to him, normally the 2 papillary muscles wind themselves around each other during systole and thus reinforce the thin ventricular apex. When acute dilatation occurs, such a reinforcing mechanism for the apex may become im- possible, and stretching will spread to the entire wall of the apex, resulting in a marked negative after-potential in the entire apical region but not in the thicker basal portion, leading to a reversal of the U wave. At present, we do not know exactly what the negative afterpotential is in nature. But the thought that such a potential is developed in stretch-imposed muscles of the heart, is very suggestive to postulate the genesis of handgrip in- duced U-wave change. Acute dilatation of the left ventricle due to hand- grip is likely to impose the same degree of stretch on the subepicardial layers as the subendocardial in apex or some localized area where the ventricular wall is still relatively thin, and to develop a potential there causing the nega- tive U wave.

REFERENCES

1. Kemp RL, Surawicz B, Bettinger JC, Gottlieb H, Bellet S: Prognostic significance of nega- tive U waves in the electrocardiogram in hypertension. Circulation 15: 98, 1957 2. Watanabe N: Clinical significance of the negative U wave in electrocardiogram. Jpn Circulat J 31: 9, 1967 3. Georgopoulos AJ, Proudfit WL, Page IH: Relationship between arterial presseure and nega- tive U waves in electrocardiograms. Circulation 23: 675, 1961 4. Henry WL, Clark CE, Epstein SE: Asymmetric septal hypertrophy. Echocardiographic identification of pathognomonic and anatomic abnormality of IHSS. Circulation 47: 225, 1973 5. Lind AR, Taylor SH, Humphreys PW, Kennelly BM, Donald KW: The circulatory effects of sustained voluntary muscle contraction. Clin Sci 27: 229, 1964 6. Donald KW, Lind AR, McNicol GW, Humphreys PW, Taylor SH, Staunton HP: Cardio- Vol.22 HANDGRIP-INDUCED NEGATIVE U 73 No.1

vascular responses to sustained (static) contractions. Circulat Res 20 (Suppl I): I-15, 1967 7. Kivowitz C, Parmley WW, Donoso R, Marcus H, Ganz W, Swan HJ: Effects of isometric exercise on cardiac performance. The grip test. Circulation 44: 994, 1971 8. Helfant RH, Devilla MA, Meister SG: Effect of sustained isometric handgrip exercise on left ventricular performance. Circulation 44: 982, 1971 9. Feigenbaum H: Echocardiography. 2nd Ed, Lea & Febiger, Philadelphia, p464, 1976 10. Fujii J, Watanabe H, Koyama S, Kato K: Echocardiographic study on diastolic posterior wall movement and left ventricular filling by disease category. Am Heart J 98: 144, 1979 11. Surawicz B, Kemp RL, Bellet S: Polarity and amplitude of the U wave of the electrocardio- gram in relation to that of the T wave. Circulation 15: 90, 1957 12. Miller RR, Vismara LA, Zelis R, Amsterdam EA, Mason DT: Clinical use of sodium nitro- prusside in chronic ischemic heart disease. Effects on peripheral and venous tone and on ventricular volume, pump and mechanical performance. Circulation 51: 328, 1975 13. Kishida H, Cole JS, Surawicz B: Negative U wave. Clinical significance and possible mechanism (abstr). Circulation 58 (Suppl II): 239, 1978 14. Sakamoto Y, Sano T, Suzuki F, Yamamoto M: The U vector loop in normal and various abnormal states as studied by the summation method. J Electrocardiology 11: 331, 1978 15. Gerson MC, Phillips JF, Morris SN, McHenry PL: Exercise-induced U-wave inversion as a marker of stenosis of the left anterior descending coronary artery. Circulation 60: 1014, 1979 16. Hoffman BF, Cranefield PF: Electrophysiology of the Heart. McGraw-Hill, New York, p201, 1960 17. Watanabe Y: Purkinje repolarization as a possible cause of the U wave in the electrocardio- gram. Circulation 51: 1030, 1975 18. Omura Y: Relationship between transmembrane action potentials of single cardiac cells and their corresponding surface electrograms in viro and in vitro, and related electromechanical phenomena. Trans NY Acad Sci (series II) 32: 871, 1970 19. Furbetta D, Bulfalari A, Santucci F, Solinas P: Abnormality of the U wave and the T-U segment of the electrocardiogram. The syndrome of the papillary muscles. Circulation 14: 1129, 1956 20. Lepeschkin E: Physiological basis of the U wave. In Advances in ed by Schlant RC, Hurst JW, Grune and Stratton, New York, London, p431, 1972