Kaplan-Meier curves for free of sudden cardiac death for current, past, and never smokers. Survival is the Kaplan-Meier estimate of survival from sudden coronary death during the follow-up period. The survival curves for never and past smokers are significantly different from the survival curve for current smokers (P = .006).
Goldenberg I, Jonas M, Tenenbaum A, Boyko V, Matetzky S, Shotan A, Behar S, Reicher-Reiss H, for the Bezafibrate Infarction Prevention Study Group. Current Smoking, Smoking Cessation, and the Risk of Sudden Cardiac Death in Patients With Coronary Artery Disease. Arch Intern Med. 2003;163(19):2301-2305. doi:10.1001/archinte.163.19.2301
Cigarette smoking is a known risk factor for sudden cardiac death (SCD). However, the effect of continued cigarette smoking and smoking cessation on SCD risk in patients with established coronary artery disease (CAD) is subject to controversy. We, therefore, evaluated the effect of cigarette smoking on SCD risk in a large cohort of patients with established CAD.
The study population was composed of 3122 patients with a previous myocardial infarction or stable angina who participated in the Bezafibrate Infarction Prevention Trial. Patients were prospectively followed up for a mean of 8.2 years. The primary end point was the incidence of SCD according to smoking status.
Among the 370 patients who were current smokers, 30 (8.1%) experienced SCD; 83 (4.6%) of the 1821 patients who had quit smoking and 43 (4.6%) of the 931 patients who had never smoked experienced SCD (P = .01). In multivariate analyses, current smoking was associated with a significant increase in the risk of SCD (hazard ratio, 2.47; 95% confidence interval, 1.46-4.19). Patients who had stopped smoking had no significant increase in the risk of SCD compared with patients who had never smoked (hazard ratio, 1.06; 95% confidence interval, 0.70-1.62).
Current cigarette smoking is a powerful independent predictor of SCD risk in patients with CAD. Patients who quit smoking experienced a significant reduction in SCD risk. Thus, efforts to reduce mortality from SCD in patients with CAD should include vigorous smoking cessation strategies.
SUDDEN CARDIAC death (SCD) constitutes about one quarter of all coronary artery disease (CAD)–related major events and more than half of all CAD-related deaths. Ventricular arrhythmia is believed to be the most common direct cause.1,2 Nicotine produces a marked elevation in serum catecholamine concentration that is potentially arrhythmogenic.3 Smoking cessation is universally recommended in patients with CAD, and some of the adverse cardiovascular effects of cigarette smoking have been shown to dissipate with smoking cessation.4- 8 However, in studies that have examined the risk factors for SCD separately in patients with and without preexisting CAD, the effect of continued cigarette smoking and smoking cessation has been well established only for subjects without known CAD. In contrast, once CAD becomes overt, the effect of smoking status on SCD risk is controversial.9- 15
We, therefore, investigated the data collected in the framework of the Bezafibrate Infarction Prevention (BIP) Trial to evaluate the long-term effect of smoking status on SCD risk in a large cohort of patients with known CAD who were prospectively followed up for a mean of 8.2 years.
The BIP study was designed and initiated in 1990. The main objective of the trial was to determine whether bezafibrate would reduce CAD-related mortality and nonfatal myocardial infarction (MI) in patients with CAD and a moderately elevated cholesterol level. A detailed description of the BIP Trial methods and results has been provided elsewhere.16,17 The following is a short review of the important aspects of the trial and information relevant to the present analysis.
Between February 1, 1990, and October 31, 1992, 15 524 male and female patients with CAD were screened for inclusion in the BIP study in 18 of the 25 cardiology departments in Israel. After an initial 2 months of eating a lipid-lowering diet, 3122 patients who fulfilled the inclusion criteria were randomized to the study. Inclusion criteria were as follows: aged 45 to 74 years, history of MI for 6 months or longer but for less than 5 years before enrollment into the study and/or stable angina pectoris (AP) confirmed by coronary angiography, and/or radionuclear studies or standard exercise tests. In addition, the following lipid profile was required: a serum total cholesterol level between 180 and 250 mg/dL (4.65-6.46 mmol/L), a low-density lipoprotein cholesterol level of 180 mg/dL or less (≤4.65 mmol/L) or 160 mg/dL or less (≤4.14 mmol/L) for patients younger than 50 years, a high-density lipoprotein cholesterol level of 45 mg/dL or less (≤1.16 mmol/L), and a triglyceride level of 300 mg/dL or less (≤3.39 mmol/L). The main exclusion criteria were type 1 diabetes mellitus, severe heart failure, unstable AP, hepatic or renal failure, known sensitivity to bezafibrate, or current use of lipid-modifying drugs.
After giving written informed consent, the patients were allocated to receive either 400 mg of bezafibrate retard or placebo once a day. Patients were allowed to take prescribed medications for cardiac and other conditions. Routine visits to the clinics were scheduled every 4 months for clinical evaluation. During each 4-month follow-up visit, data on any adverse events (as defined in the study protocol), hospitalizations, and study outcomes were obtained. All study participants, regardless of whether they continued to take the trial medication, were followed up until the last patient had completed at least 5 years of follow-up (mean, 6.2 years). After discontinuation of the study medication, patients were observed for an additional mean follow-up of 2 years. An independent International Review and Advisory Board regularly monitored the progress of the study and the incidence of adverse events. All deaths were examined in depth, and their cause was established by an independent committee composed of 3 experienced cardiologists.
Sudden cardiac deaths were defined as deaths that occurred within 1 hour of the onset of cardiovascular symptoms, witnessed instantaneous death, or unwitnessed death with no preceding change in symptoms for which no other cause could be ascribed.
Smoking status was determined by the physician at the baseline examination and at each of the 4-month follow-up visits. Information included the number of cigarettes smoked, duration of smoking, and the time of quitting. Never smokers were persons who had never smoked more than 100 cigarettes in their lifetime; past smokers, persons who had smoked more than 100 cigarettes, but had quit before their baseline examination or before the initial 4-month follow-up visit; and current smokers, persons who continued to smoke after the initial 4-month follow-up visit.
Baseline characteristics according to smoking status were compared using the χ2 test for dichotomous variables and an analysis of variance for continuous variables.
A multivariate analysis was performed using Cox proportional hazards regression models. The following baseline variables were included in the model: age; sex; smoking status (present or past smokers vs never smokers); New York Heart Association functional class of 2 or greater; AP functional class of 2 or greater (according to the Canadian classification); history of diabetes mellitus, MI, cerebrovascular accident, chronic obstructive pulmonary disease, peripheral vascular disease, or hypertension; systolic blood pressure; total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, triglyceride, and fibrinogen levels; body mass index; heart rate; and treatment with β-blockers, antiplatelet agents, angiotensin-converting enzyme inhibitors, and diuretics. Kaplan-Meier survival curves with log-rank statistics were used to compare the rate of SCD in the 3 groups during the BIP study.
Of the 3122 patients who were included in the final analysis, 370 (11.9%) were active smokers (current smokers), 1821 (58.3%) had a history of smoking but had stopped before the initial follow-up visit (past smokers), and 931 (29.8%) had no history of smoking (never smokers). Table 1 and Table 2 show their baseline clinical and laboratory characteristics and medical therapy, respectively. The groups are comparable in a wide variety of characteristics, but show some differences. Active and past smokers were more likely to be male and to have a history of chronic obstructive pulmonary disease than those who never smoked. The laboratory evaluation showed that current smokers had higher baseline triglyceride and fibrinogen levels and lower high-density lipoprotein cholesterol levels than past and never smokers (Table 1). In addition, medical management at randomization with antiplatelet agents, β-blockers, angiotensin-converting enzyme inhibitors, and diuretics was used less frequently in patients who were active smokers (Table 2).
As assessed by the 8.2-year follow-up, 30 SCDs occurred in the 370 patients who were current smokers, whereas 83 SCDs occurred in the 1821 patients who were past smokers and 43 SCDs occurred in the 931 patients who had never smoked. Notably, the rate of SCD was not different between patients who quit smoking and patients who had never smoked. Current smokers also had a substantially greater cardiac-related and all-cause mortality rate compared with never smokers and past smokers (Table 3).
The relationship between SCD rate and amount and duration of smoking is shown in Table 4. Compared with past smokers, present smokers smoked fewer cigarettes per day (90.7% of current smokers smoked ≤20 cigarettes per day vs 52.6% of past smokers; P<.001). However, the risk of SCD did not increase in patients who smoked more cigarettes per day, but did increase with increased duration of smoking for current and past smokers.
There was no correlation between the risk of SCD and the number of years without smoking in patients who had quit smoking before randomization (Table 4). Thus, the reduced SCD rate of past smokers compared with present smokers was not related to the time of smoking cessation before study enrollment.
The adjusted hazard ratios for SCD associated with significant variables in the Cox proportional hazards regression models are as follows:
Current smoking was independently associated with a significant increase in the risk of SCD, while patients who stopped smoking were not at a greater risk of SCD than patients who had never smoked.
Continued cigarette smoking was associated with an SCD risk that was greater than the risk for other important variables included in the model, such as age, lipid levels, New York Heart Association and AP functional class, history of hypertension or drug therapy with β-blockers, angiotensin-converting enzyme inhibitors, or antiplatelet agents, and treatment with bezafibrate.
Kaplan-Meier curves for survival free of SCD comparing current smokers, past smokers, and never smokers demonstrate similar survival for past and never smokers (Figure 1). After 8.2 years, the estimated rate of survival free of SCD was 95% for past smokers and never smokers and 89% for patients who continued to smoke (P = .006) (Figure 1).
The findings of this long-term prospective follow-up of 3122 patients with preexisting CAD suggest the following: (1) continued cigarette smoking is associated with a significantly increased risk of SCD; (2) in patients who quit smoking, the risk of SCD is significantly lower and comparable to the risk of patients who had never smoked; and (3) the decline in the risk of SCD with smoking cessation is immediate and not time dependent.
Prior reports5,9 have shown that, for people without a history of CAD, cigarette smoking is associated with an increased risk of SCD. However, the effect of this modifiable risk factor on SCD risk in patients with established CAD is subject to controversy. In the Framingham population, positive predictors of SCD in patients with prior CAD did not include smoking.10 More recently, in a prospective 8-year follow-up11 of 7735 middle-aged men, current smoking was independently associated with an increased risk of SCD only in men without preexisting CAD. In a retrospective case-control study12 of unselected patients with established CAD who had experienced SCD, smoking was not independently associated with risk of SCD. In a prospective evaluation13 of the effect of smoking cessation on the incidence of SCD in the subjects of the Cardiac Arrhythmia Suppression Trial cohort, there was only a marginal statistical difference between present and past smokers. However, in a high-risk subgroup of the cohort (patients who did not undergo thrombolysis or revascularization after their qualifying MI), smoking cessation greatly reduced the incidence of SCD and was associated with a statistically significant benefit on survival.13 Similarly, in 310 patients with aborted SCD, the recurrence rate during a 4-year follow-up was significantly increased in current smokers.14 In our study, compared with prior reports, many of the subjects were survivors of a prior MI. In this population, the effect of continued cigarette smoking on the occurrence of SCD was even more pronounced than that of other risk factors, such as age, sex, New York Heart Association functional class, blood pressure, and dyslipidemia.
We also demonstrated that the increased risk of SCD in patients who continue to smoke is unrelated to smoking burden. In the present study population, 90.7% of present smokers smoked less than one pack of cigarettes per day compared with only 52.6% of past smokers; however, the rate of SCD was significantly higher among patients who continued to smoke. Moreover, in patients who quit smoking, the reduction of SCD risk was immediate and unrelated to the time of smoking cessation. These findings support the notion that cigarette smoking may lead to SCD mainly through direct toxic effects and not through long-term atherosclerotic additive damage. This may occur by several potential mechanisms: the nicotine in cigarette smoke stimulates the sympathetic nerves and increases the propensity toward arrhythmias and the vulnerability to cardiac arrest18- 21; cigarette smoking has been found to predispose to acute thrombosis in patients who died suddenly22,23; and an intracellular mechanism may be related to the ability of nicotine to prolong action potentials and to depolarize membrane potential, possibly through direct inhibition of cardiac potassium channels.24
Several limitations to our study should be considered. First, the baseline characteristics, mainly medical therapy, were not similar among patients with different smoking status. Thus, patients who were active smokers had a more frequent abnormal lipid profile and received less intensive secondary prevention with β-blockers, platelet antagonists, and angiotensin-converting enzyme inhibitors. However, the prospective long-term examination of a large cohort of patients, as in the present study, provides strong evidence of the effect of smoking on the risk of SCD in patients with CAD after multivariate adjustment. Second, a few patients (13%) who were defined as current smokers had quit smoking during the follow-up period. Because of the small number, the effect of smoking cessation during the follow-up period could not be demonstrated. However, even after the exclusion of these patients from the multivariate analysis, continued cigarette smoking remained a significant independent predictor of SCD. Third, because the left ventricular ejection fraction was not documented for those patients enrolled in the BIP study, this important predictor of survival in patients with CAD was not incorporated in our multivariate analysis for long-term mortality.
Despite the advances in the prediction of SCD in survivors of MI, the effect on the cumulative incidence of SCD in patients with CAD or previous MI has been relatively small, because most SCDs occur among patients without the characteristics that would have led to their inclusion in trials of implantable defibrillators.25- 28 Given the expense of therapies for the prevention of SCD in patients with a prior MI and their narrow applicability, smoking cessation counseling with its follow-up may be more cost-effective and should be emphasized as an integral part of the management of patients with CAD, with the aim of preventing an excess of SCDs.
In conclusion, our data indicate that continued cigarette smoking significantly elevates the risk of SCD in patients with CAD, and stress the importance of complete smoking cessation in risk factor modification of patients with healed MI and/or stable AP.
Corresponding author: Ilan Goldenberg, MD, Heart Institute, The Chaim Sheba Medical Center, Tel Hashomer 52621, Israel (e-mail: firstname.lastname@example.org).
Accepted for publication December 18, 2002.