Blood Pressure Influences the Occurrence of Complex Ventricular Arrhythmia in Hemodialysis Patients
Abstract We investigated the relationship between blood pressure and the occurrence of complex ventricular arrhythmias (multiform, couplets, or runs) as assessed by 48-hour Holter monitoring in 74 stable long-term hemodialysis patients (44.5±12 years old; 54% men; 74% whites; dialysis duration, 51.3±36.1 months; systolic pressure, 146.6±19.3 mm Hg; diastolic pressure, 89.2±12.1 mm Hg; prevalence of arterial hypertension, 33.8%). Systolic and diastolic pressures represented the average of all predialysis determinations during the 3 months preceding the tests. Hemodialysis was performed midway through the Holter monitoring period. M-mode and bidimensional echocardiograms and myocardial perfusion tests were also obtained from all patients. Complex arrhythmias were observed in 37 individuals (50%). Univariate analysis showed that systolic pressure (P<.001), diastolic pressure (P<.05), age (P<.001), left ventricular posterior wall thickness (P<.01), left ventricular mass index (P<.05), and ischemic alterations on myocardial perfusion tests (P<.005) were significantly associated with complex arrhythmias. With the use of a multivariate model (stepwise logistic regression analysis) only systolic pressure (P<.01) and age (P<.05) were independently associated with complex arrhythmias. Sex; angina; dialysis duration; New York Heart Association functional class; use of digitalis; plasma levels of creatinine, sodium, potassium, calcium, and phosphate; hematocrit; left ventricular fractional shortening; left ventricular diastolic diameter; and ST segment deviation were not correlated with complex arrhythmias. The severity and frequency of complex arrhythmias were not influenced by hemodialysis. At follow-up (5 to 80 months) 5 patients had died of sudden death, 4 of whom were hypertensive and older than 45 years. We conclude that complex ventricular arrhythmias are highly prevalent in long-term hemodialysis patients, their occurrence is independently influenced by systolic pressure and age, and they are not aggravated by dialysis itself in this stable dialysis population. The data also suggest that Holter monitoring should be performed in elderly hypertensive dialysis patients.
Ventricular arrhythmias are prevalent in chronic renal failure patients on long-term hemodialysis,1 2 3 4 5 6 7 8 9 10 11 a population known to sustain an elevated frequency of sudden cardiac death.12 The role of arterial hypertension, a well-known risk factor for complex arrhythmias in patients with essential hypertension13 and an alteration exceedingly frequent in dialysis populations, has not yet been fully investigated. We designed the present study to evaluate the relationship between arterial BP and complex ventricular arrhythmias, as assessed by 48-hour Holter monitoring, in a group of stable end-stage renal failure patients treated by hemodialysis. All patients were submitted to extensive clinical and cardiological evaluation, and multivariate analysis was performed to control for the factors most likely to independently influence cardiac rhythm.
All adult hemodialysis patients on dialysis for at least 12 months and treated at three dialysis centers located in the São Paulo metropolitan area were contacted and asked to participate in the study. We limited to 12 months the minimum duration of dialysis treatment to favor the inclusion of individuals whose clinical conditions were stable under dialysis. The choice of centers was dictated by the use of similar treatment routines in all. Seventy-four individuals out of a population of 136 were included in the study. Sixty-two patients were excluded because of the following reasons: age of less than 20 years (12 patients), dialysis duration of less than 12 months (18 patients), refusal to participate in the study (22 patients), technical failure in the Holter recordings (4 patients), and substandard quality of echocardiographic records (6 patients). Hemodialysis was performed three times a week through an arteriovenous fistula with the use of a hollow-fiber dialyzer with 1.0 to 1.4 m2 of surface area and a bicarbonate bath containing 0 to 2.0 mEq/L potassium. Dialysis duration was adjusted to keep midweek predialysis serum urea less than 150 mg/100 mL (25 mmol/100 mL). Minimum flow rates were 300 mL/min, and dialysate flow rates were 400 mL/min. The causes of end-stage renal diseases were as follows: chronic glomerulonephritis (n=17), hypertensive kidney disease (n=15), tubule-interstitial diseases (n=22), diabetes (n=3), lupus nephritis (n=3), Alport’s syndrome (n=1), and gout nephropathy (n=1). In 15 patients the etiology of renal disease could not be determined. SBP and DBP represented the average of all predialysis determinations taken with patients in the upright position during the 3 months preceding the tests. Hypertension was defined as SBP higher than 160 mm Hg and/or DBP higher than 94 mm Hg. Forty-five patients (61%) were taking antihypertensive medication consisting of sympatholytics, angiotensin-converting enzyme inhibitors, calcium channel blockers, and β-blocker agents, alone or in combination. In addition, 4 patients were taking digitalis, 2 were taking antiarrhythmics drugs, and 2 were taking long-acting nitrates. Medications were not withdrawn during the tests. Patients gave informed consent to participate in the study.
Clinical and Laboratory Evaluations
Patients underwent medical and laboratory investigations that included resting electrocardiograms (to exclude previous myocardial infarction) and plasma creatinine, electrolyte, and hematocrit determinations. All tests were performed in the morning of the day preceding dialysis. NYHA functional class was established for each patient.
Holter electrocardiographic monitoring was performed for a period of 48 hours while the individuals engaged in their ordinary daily activities and included a 4-hour hemodialysis session carried out midway through the recording period. We used a two-channel Marquette 8500 device, and the records were analyzed with a Marquette Laser 8000 T system. Two leads were recorded: MV1 (modified V1) and MV5 (modified V5). The Holter recordings were evaluated regarding the presence and frequency of complex ventricular arrhythmias before, during, and after the dialysis session and the occurrence of ST segment deviation. Complex ventricular arrhythmias were defined according to the classification of Lown and Wolf14 : multiform ventricular premature beats (grade III), couplets (grade IVa), and runs (grade IVb). Significant ST segment deviations were defined according to Kennedy and Wiens.15
M-mode and two-dimensional echocardiograms were obtained according to the criteria set forth by the American Society of Echocardiography.16 LV mass (LVM) was calculated by the equation LVM=(IVST+PWT+EDD)3−EDV×1.04, where IVST is diastolic thickness of the interventricular septum; PWT is diastolic thickness of the posterior LV wall; EDD is end-diastolic dimension; EDV is end-diastolic volume; and 1.04 is the cardiac tissue density. LV mass index was calculated by the ratio of LVM to body surface area. LV hypertrophy was defined as LV mass index greater than 134 g/m2 in men and 110 g/m2 in women.
Myocardial Perfusion Studies With Dipyridamole Infusion (Thallium-Dipyridamole Tests)
Cardiac images were obtained after intravenous injection of 2 mCi thallium-201 preceded by dipyridamole infusion (0.5 mg/kg during 4 minutes) with an Ohio Nuclear gamma camera. Areas of thallium redistribution (transient filling defects) were interpreted as indicative of myocardial ischemia, and persistent defect areas corresponded to scarring.17
Patients were followed for 5 to 80 months.
Values are expressed as mean±SD. Student’s t test for unpaired samples was used to compare continuous variables, and the χ2 or Fisher’s exact test was used to compare categorical variables between patients with and without complex arrhythmias. The variables found to be significantly associated with complex arrhythmias were entered in a multivariate stepwise logistic regression analysis. ANOVA and the Tukey test were applied to assess the relationship between the severity of systolic hypertension and the degree of complexity of ventricular arrhythmias. To take into account the nonuniform distribution of arrhythmia frequency and severity and the great variability of these parameters from hour to hour, we used nonparametric tests (McNemar18 and Friedman19 tests, respectively) to assess changes in the severity and frequency of complex arrhythmias during and after dialysis compared with the predialysis period. A value of P<.05 was considered statistically significant.
All 74 patients completed the study. Patient age was 44.5±12 years (range, 21 to 71); there were 40 men (54%) and 55 whites (74%), and mean dialysis duration was 51.3±36.1 months (range, 12 to 132). Mean SBP and DBP values for the entire group were 146.6±19.3 and 89.2±12.1 mm Hg, respectively. Hypertension was present in 25 patients (33.8%), and 36 patients (48.6%) had a history of previous arterial hypertension. Patients had a negative history of previous myocardial infarction, and old myocardial infarcts were not observed in resting electrocardiograms. No individual was in NYHA class IV.
Complex ventricular arrhythmias occurred in 37 (50%) of the individuals. Multiform premature complexes were observed in 16 patients, couplets in 15, and runs in 6. The severity of complex arrhythmias was not changed by dialysis (before versus during dialysis, P=1.0; before versus after dialysis, P=.06, P=NS) (Fig 1⇓): 17 patients exhibited complex arrhythmias before, during, and after dialysis; 12 developed complex arrhythmias only during and after dialysis; and in 8 the alteration disappeared during or after dialysis. Also, the frequency of complex arrhythmias (number of events per hour) was not significantly altered by dialysis (before dialysis versus during and after dialysis, P=.113 and P=NS, respectively). Compared with the predialysis period, 22 patients did not exhibit an increase in complex arrhythmia frequency during or after dialysis; in 4 patients the frequency was increased during dialysis; in 5, only after dialysis; and in 6, during and after dialysis. ST segment deviations were observed in only 12 patients (16.2%), occurring indifferently before, during, and after dialysis. Univariate analysis (Table 1⇓) showed that the variables significantly associated with complex arrhythmias were age (P<.001), SBP (P<.001), DBP (P<.05), LV posterior wall thickness (P<.01), LV mass index (P<.05), and altered myocardial perfusion on thallium-dipyridamole test (P<.005). Sex; angina; use of digitalis; NYHA classes II and III; dialysis duration; hematocrit; plasma levels of creatinine, sodium, potassium, calcium, and phosphate; LV fractional shortening; LV end-diastolic dimension; and significant ST segment deviation were not related to the occurrence of complex arrhythmias. The variables significantly associated with complex ventricular arrhythmias were entered in a multivariate stepwise analysis which showed that only age (P<.05) and SBP (P<.01) were correlated with the occurrence of complex arrhythmias. ANOVA also showed that the severity of systolic hypertension correlated with the degree of ventricular arrhythmias (Table 2⇓). The probability of developing complex arrhythmias may be calculated with a specificity of 76% and sensitivity of 78% with the use of the following equation, which takes into account the combined effect of age and BP on the occurrence of complex arrhythmias: log(p/1−p)=−9.7489+0.0661×Age+0.0467×SBP (Fig 2⇓).
Three patients were lost to follow-up. Five patients died of sudden death. Among those who died of sudden death, four were older than 45 years and were also hypertensive.
The frequency and severity of ventricular arrhythmias in dialysis patients have been reported to be primarily dependent on the presence of preexisting ventricular dysfunction and coronary artery disease.9 11 Other researchers have pointed out the use of digitalis,1 diabetes,3 hyperparathyroidism,3 6 age,4 and LV hypertrophy7 as the most relevant factors. It should be emphasized that our patients were relatively young and in stable clinical condition and that only five presented reduced systolic function. Moreover, the prevalence of diabetes and documented previous myocardial infarction was low or absent. Therefore, our results are likely to be more representative of the true causes of risk associated with dialysis and uremia rather than with concomitant cardiac diseases. We found that only SBP and age were determinant of complex ventricular arrhythmias in hemodialysis patients after adjustment for other associated factors. DBP, LV posterior wall thickness, LV mass index, and ischemic alterations in myocardial perfusion studies, although found to be significantly associated with complex arrhythmias by univariate analysis, were no longer significant when known confounding factors were taken into account. The importance of age as a promoter of ventricular arrhythmias has been demonstrated in healthy individuals.20 Our results agree with those of Cocchi et al,4 who showed a correlation between ventricular ectopy and age in dialysis patients. We are not aware of other reports linking BP to ventricular arrhythmias in this population. However, arterial hypertension is known to be involved in the occurrence of serious cardiac arrhythmias in nonuremic patients, especially when associated with LV hypertrophy.13 21 The arrhythmogenic effect of hypertension possibly involves the promotion of myocardial ischemia aggravated by concomitant LV hypertrophy.13 21 In our patients myocardial ischemia and LV hypertrophy were significantly more frequent among those with complex arrhythmias, suggesting that hypertensive myocardiopathy was involved in the genesis of the arrhythmias.
Most patients dying of sudden death were old and hypertensive, suggesting that this group could benefit from routine Holter monitoring as a means of establishing a prognosis. However, the small number of terminal events does not allow further elaboration.
In agreement with the majority of researchers,8 9 10 11 we failed to show that hemodialysis influences the severity or frequency of complex arrhythmias, even when performed with a dialysate without potassium or with a low concentration of potassium. This finding does not exclude the possibility that patients with more serious cardiac involvement may present an increased severity of arrhythmias during dialysis.
In conclusion, in this low-risk, stable dialysis population we found a high prevalence of complex arrhythmias that was independently influenced by SBP and age and was not aggravated by the dialysis procedure. Elderly hypertensive dialysis patients may benefit from routine Holter monitoring.
Selected Abbreviations and Acronyms
|DBP||=||diastolic blood pressure|
|NYHA||=||New York Heart Association|
|SBP||=||systolic blood pressure|
This investigation was supported by the E.J. Zerbini Foundation, São Paulo, Brazil. We thank Drs João Américo da Fonseca, Luis Sérgio F. de Azevedo, Anibal D. Godoy, and Vanda Jorgetti for permission to study their patients; Júlia T. Fukushima for the statistical analysis; and Elettra Greene for editorial assistance.
- Received June 18, 1995.
- Revision received September 16, 1995.
- Accepted October 9, 1995.
Saragoça MA, Canziani ME, Cassiolato JL, Gil MA, Andrade JL, Draibe SA, Martinez EE. Left ventricular hypertrophy as a risk factor for arrhythmias in hemodialysis patients. J Cardiovasc Pharmacol. 1991;17(suppl 2):S136-S138.
Chazan A, Pono LM. Sudden death in patients with chronic renal failure on hemodialysis. Dial Transplant. 1987;16:447-448.
Lown B, Wolf M. Approaches to sudden death from coronary heart disease. Circulation. 1971;44:130-142.
Sahn DJ, De Maria A, Kisslo J, Weyman A. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic measurements. Circulation. 1978;58:1072-1083.
Eagle KA, Boucher CA. Nonexercise thallium imaging using dipyridamole coronary vasodilation. Am J Noninvas Cardiol. 1990;4:1-7.
Fleiss JL. Statistical Methods for Rates and Proportions. 2nd ed. New York, NY: JW Levy & Sons; 1981:114-119.
Hollander M, Wolfe DA. Nonparametric Statistical Methods. New York, NY: John Wiley & Sons; 1973:139-141.