Electrocardiographic abnormalities in acute cerebro-vascular accidents and their correlation with cerebro-spinal fluid pressure and serum electrolytesMS Gambhir, AH Hakim, BS Bomb, RC Gupta, DK Agarwal
Department of Medicine, R.N.T. Medical College, Udaipur, Rajasthan, India
Electrocardiographic studies have been carried out in 50 patients with acute cerebrovascular accidents. These cases included 31 cases from non-haemorrhagic group and 19 cases from haemorrhagic group. `T' or T-U wave abnormalities were seen in 63.1% of cases in haemorrhagic group and in 35.4% of cases from non-haemorrhagic group. Abnormal prolongation of QTc interval (105% of normal i.e. 0.45 Secs.) was observed in 73.6% cases from haemorrhagic group and in 35.4% of cases from non-haemorrhagic group. On comparing these two groups it was found that QTc was significantly more prolonged in patiens of haemorrhagic group than non-haemorrhagic group. The CSF pressure was found to be significantly more in patients with ECG abnormalities. However, there teas no correlation between serum electrolyte levels and ECG abnormalities. After 2 weeks, the ECG abnormalities disappeared, except in two cases from haemorrhagic group. There was no significant difference in the incidence of mortality between patient having ECG abnormalities and the patients without ECG abnormalities. In the light of these findings the probable mechanism of production of ECG changes in acute cerebrovascular accidents is discussed.
Acute cerebrovascular accidents (C.V.A.) are quite often accompanied with characteristic electrocardiographic (E.C.G.) abnormalities. ,,,,, The findings may be summarised as: Prolongation of the Q-Tc interval and a large and wide T or T-U wave of the same general configuration as in myocardial ischaemia. 
Similar ECG changes have been described in nonvascular intracranial lesions like meningitis and intracranial tumours by Hersch.  The mechanism by which these changes are produced is not clear. The various mechanisms which have been suggested by different workers ,, include actual damage to the myocardium in the form of sub-endocardial haemorrhage, electrolyte imbalance, and raised intracranial tension and vagotonicity.
In view of these varied explanations for the ECG abnormalities in acute CVA the present study was undertaken to review the pattern of ECG changes and its correlation with serum electrolytes (Sodium and Potassium) and the cerebrospina fluid (CSF) pressure if any. Other aims, of this study were to find out any correlation between ECG changes and prognosis and reversibility of these changes.
The present study was carried out on 50 clinically diagnosed patients of acute cerebrovascular accidents admitted within 48 hours after the onset of stroke to the general hospital, attached to R.N.T. Medical College, Udaipur. The diagnosis depended on focal neurological signs and temporal profile. These patients were further divided into two groups (1) haemorrhagic vascular lesions and (2) nonhaemorrhagic vascular lesions, depending upon the presence or absence of grossly apparent blood (non traumatic) or xanthochromia or both in the CSF.
Immediately after hospitalisation a detailed neurological examination was carried out to assess the neurological deficit. In addition to routine investigations, following special investigation were also carried out:
In all ECG's the QTc interval was calculated according to the method described by Kissin. 
The QTc interval was compared between the haemorrhagic and nonhaemorrhagic group by the Students `t' test. The CSF pressure and serum electrolyte levels were also compared between patients showing ECG abnormalities and patients without any ECG abnormalities by the same statistical method.
The results of the present study are summarised in [Table 1] and[Table 2]. As evident from the tables, 19 patients belonged to haemorrhagic group and 31 patients, to non-haemorrhagic group. QTc was found to be prolonged (more than 105% of normal) in 14 out of 19 patients from the haemorrhagic group and 11 out of 31 patients from the non-haemorrhagic group. The average QT( was 123.3% of normal in the haemorrhagic group, and 114.6% of normal in the non-haemorrhagic group. Or statistical analysis the QTc interval in haemorrhagic group was significantly longer than the non-haemorrhagic group (p < 0.05). After two weeks the Q-Tc duration returned to normal (within 105% of normal) in all the cases who survived except two cases from haemorrhagic group in whom the Q-Tc's were 107 and 108% of normal.
The abnormal T or T-U waves were observed in 12 out of 19 patients (63.1%) from the haemorrhagic group and 11 out of 31 patients (35.4%) from the non-haemorrhagic group. The T or T-U wave abnormalities disappeared in the subsequent ECGs.
The mean serum sodium levels in patients of acute cerebrovascular accidents with and without ECG abnormalities were 137.82 and 136.12 mEq. /LI respectively, while serum potassium levels were 4.37 and 4.42 mEq/1 respectively. There was no correlation between the ECG abnormalities and serum sodium and potassium levels.
The C.S.F. pressure in patient of acute CVA with and without ECG abnormalities were 204.80 and 162.40 mm of water respectively. On statistical analysis the difference was found to be significant.
During the two weeks' follow up, in all, ten patients expired, out of which six belonged to the haemorrhagic group and four to the non-haemorrhagic group. Out of the six patients, who expired from the haemorrhagic group two had no ECG abnormalities, while four had ECG abnormalities. All the four patients from the non-haemorrhagic group had no ECG abnormalities.
The above observations on the ECGs in 50 cases of acute CVA have brought out two types of the ECG abnormalities (i) increased Q-Tc interval and (ii) abnormal T or T-U wave.
In 46% of the cases, both the above abnormalities were present, while in additional 4% cases, only prolonged ,Q-Tc was there.
The incidence of such abnormalities is significantly higher in the haemorrhagic group than in the non-haemorrhagic group Furthermore, the mean Q-Tc was mar; prolonged in the haemorrhagic group than in the non-haemorrhagic group. This observation is in confirmation with that of Hansson and Larsson. 
On follow up it was seen that in most of the cases the ECG changes reverted to normal within a period of two week; except in two cases from the haemorrhagic group, in whom slight prolongation of the Q-Tc duration persisted, although in these the T or T-U wave abnormality disappeared. This indicates that ECG abnormalities in acute CVA is produced by some pathophysiology which does not produce permanent damage to the heart.
There was no significant difference in the incidence of mortal ty between patients of acute CVA with ECG abnormalities and those without ECG abnormalities. This further indicates that there is no significant structural damage to heart in majority of the patients of acute CVA with ECG abnormalities as suggested by Koskelo et al  otherwise one would expect higher mortality in these patients than those who are not having any ECG abnormality. In agreement with our view while contrary to the observation of Koskelo et al  other workers ,, were unable to find any histopathological changes in the myocardium of the patients dying from subarachnoid haemorrhage.
From the present study it is clear that there is no correlation between the serum sodium and potassium levels and ECG abnormalities encountered in patients of acute CVA suggesting that ECG abnormalities encountered in patients of acute CVA are not because of alteration in the serum sodium and potassium levels. This observation is in confirmation with those of Huganholetz  and Shuster,  while contrary to the suggestion of Burch et al. 
Another point which is clear from the present study is that the mean CSF pressure was significantly higher in patients of acute CVA who had ECG abnormalities than those who had no ECG abnormalities; this indicates that raised CSF pressure by some mechanism may be producing the ECG abnormalities. This possibility is further supported by the observation of Hersch  who has reported similar ECG abnormalities in patients of raised intracranial tension because of nonvascular causes. How the raised intracranial tension produces the ECG abnormalities, is yet speculative. Cushing 5 demonstrated the dysfunction of autonomic nervous system to be due to increased intracranial pressure in dogs, which could be eliminated by vagotomy. This suggests that raised intracranial tension does have an effect on vagol tone. This view is supported by Chetri and De,  who correlated intracranial hypertension with vagal hyperactivity and have shown that the administration of intravenous atropine tends to reverse the non-specific ST and T or TU wave abnormalities of acute CVA to normal, while similar alteration due to coronary artery disease remained unaffected. Smith and Ray  have also implicated autonomic dysfunction as a probable mechanism responsible for E.C.G. changes and arrhythmia in patients having raised intracranial tension.
The authors are thankful to the Principal and Controller and Superintendent R.N.T. Medical College and attached groups of hospitals for their kind permission to publish this work.
[Table 1], [Table 2]