Keywords
INTRODUCTION
A healthy lifestyle is the foundation for primary prevention, hence changes in diet and physical exercise and the cessation of smoking form the basis of therapy in patients with established heart disease.1 Most studies of measures recommended for secondary prevention of coronary heart disease analyze the efficacy of programs aimed at favoring changes in lifestyle or at increasing the percentage of patients treated with drugs shown to be effective at delaying progression of coronary heart disease.2
Smoking has been related with physiopathologic mechanisms of cardiovascular disease, such as raised concentrations of carboxyhemoglobin and increases in fibrinogen and platelet aggregability, as well as changes in vascular reactivity and a reduction in plasma levels of high density lipoprotein (HDL-C).3-6 These changes, either alone or in combination, favor the appearance of arterial lesions, which lead to progression of arteriosclerosis and an increased risk of thrombosis.
Most of the evidence regarding the cardiovascular benefits of smoking cessation comes from randomized studies which included patients with prior cardiovascular disease. A meta-analysis has recently been published of cohort studies analyzing the benefit of smoking cessation in patients with coronary heart disease. The main finding was a combined odds ratio (OR) of 0.54 (95% confidence interval [CI], 0.46-0.62) for death, regardless of sex, study period, observation period and geographic zone, although the risk of non-fatal reinfarction was not studied.7
Only one clinical trial examining the efficacy of multiple risk factor management has shown significant results for smoking cessation.8 The duration of hospital stay and intensive counseling during recovery from the acute episode are important for achieving higher rates of sustained smoking cessation.9
This study was designed to analyze the effectiveness of our usual program of secondary prevention counseling in reducing the risk of coronary disease, especially relating to smoking cessation after a first AMI, which is still considered to be one of the main cardiovascular risk factors.10,11
PATIENTS AND METHODS
This 1:1 case-control study included 274 coronary patients from a cohort of 985 with a firm diagnosis of AMI from 1980 to 2000 in the Hospital de Navarra. Patients were included if they were younger than 76 years of age, they survived more than 6 months after the acute event, and they had received no invasive therapeutic procedure or aorto-coronary bypass. Diagnosis of all infarctions was made according to criteria of the MONICA project12 (two or more ECGs with definite changes; an ECG with suggestive changes and abnormal enzymes; or characteristic symptoms and abnormal enzymes). Cases were considered those patients with fatal o nonfatal reinfarction from January 1997 to December 2000 (n=142; men, 90.5%). Data collection was made at the time of the second infarction. The medical charts were reviewed to ensure compliance with all the study criteria. The two patients who refused to participate in the study and the three with inadequate control data (see below) were excluded from the analysis, giving a final total of 137 patients interviewed (96.5% of the initial sample).
One control patient was paired with each case for the same age (±5 years) and sex. The controls were from the same cohort of patients; i.e., all had had just one myocardial infarction (based on the criteria as mentioned above) and after the same secondary prevention period as their corresponding case (±10%) presented no symptoms of coronary disease progression. Data from controls were obtained when they were matched with their corresponding case. The data base was ordered by age and the first control who matched the reference case was chosen. If a control refused to participate (9 patients), the next suitable control in the data base was chosen.
The interviews were carried out by two doctors (EMZ and IEI) and one nurse (BSA), who all received prior training in a pilot study. Each case-control pair was interviewed by the same researcher. The interview was conducted after explaining the reasons for the study and obtaining informed consent. The questionnaire contained information concerning the following aspects: previous diseases (1 question), family history of cardiovascular disease (1 question), work after the first infarction (3 questions), family and social integration (3 questions), personal attitudes (2 questions, about feelings of stress and perfectionism) and drug therapy after the first myocardial infarction (8 questions). Questions about lifestyle before and after the first infarction included: smoking (4 questions), physical exercise (10 questions), dietary habits (5 questions) and sexual activity (2 questions). The patients were asked about compliance with the recommended changes in self-care and about the type and frequency of visits to their cardiologist, family doctor, and primary care nurse (13 questions).
If the patient had died (29 cases and 6 controls), or when the patient was unable to maintain a telephone conversation due to deafness or cognitive deterioration (6 cases and 5 controls), the interview was conducted with a close family member. In these cases, the matching control patient was also interviewed through a close family member.
Data were recorded in a Microsoft Access data base, and data for the possible control patients were obtained from a Microsoft Excel data base. Statistical analysis was made using SPSS, version 10.0. The odds ratios were calculated by conditional logistic regression analysis, with the case/control as the dependent variable. Binary logistic regression models were adjusted to study mortality, with death as the dependent variable. Percentage differences were calculated with EpiInfo, version 6.04b.
RESULTS
Mean age was 69.8 years for cases and 69.7 years for controls. The mean time to reinfarction in the cases was 7.15±0.51 years. Only 10 cases had a reinfarction 14 years or more after the first AMI. The study concentrated on lifestyle, drug therapy, and medical and nursing care during the period between the first infarction and entry to the study, i.e. the time of the most recent infarction in the cases or the time of pairing in the controls, to determine the duration of secondary prevention.
For the variable "smoker" in the initial analysis, the cases and controls were classified as smokers or nonsmokers after their first infarction. For the second conditional logistic regression model they were classified as "nonsmokers before and after first infarction", "ceased smoking after first infarction" and "«continued smoking after first infarction"». The number of patients included in each category is shown in Table 1.
Table 2 shows possible confounding variables for cases and controls, as well as percentages differences. Stopping work after the first infarction was associated with a lower frequency of smoking cessation, as well as with greater difficulty stopping smoking in the controls, who reported a subjective feeling of stress, which was not reported by the cases. Moreover, the proportion of patients who failed to follow treatment advice adequately was greater among those who continued smoking after their first infarction. This is consistent with the responses to the survey.
Smoking after the first infarction was associated with a reinfarction, with an OR of 2.83 (95% CI, 1.47-5.47). Interestingly, after adjusting the statistical model for variables related with walking and dietary modification after the first infarction, as well as for self-reported compliance with medical recommendations for secondary prevention, the OR rose to 2.96 (95% CI, 1.47-5.95). Likewise, inclusion in the model of drug therapy regimens (antihypertensive drugs, lipid lowering drugs and antidiabetic agents) increased the OR of a reinfarction to 3.20 (95% CI, 1.56-6.56). However, after including family history of heart disease and risk factors (hypertension, dyslipidemia and diabetes), the final OR was 2.80 (95% CI, 1.35-5.80) (Table 3).
In comparison with nonsmokers, both before and after the first infarction, the decision to quit smoking resulted in an OR for reinfarction of 0.82 (95% CI, 0.4-1.42). After adjusting for other lifestyle factors, compliance with recommendations, drug therapy and risk factors (family history, hypertension, dyslipidemia and diabetes) the OR was 0.90 (95% CI, 0.47-1.71) (Table 4). Those patients who continued smoking after the first infarction had an OR for reinfarction of 3.06 (95% CI, 1.53-6.13) compared with those who quit smoking. After adjustment for other lifestyle factors, compliance with the recommendations, drug therapy, high blood pressure, hypercholesterolemia, diabetes and family history, the OR fell to 2.90 (95% CI, 1.35-6.20) (Table 4).
In an attempt to evaluate the risk of a further infarction in relation to the number of cigarettes smoked per day, the conditional logistic regression model was adjusted, using the number of cigarettes divided by 5 as an independent variable. After inclusion of the same variables as mentioned above, those patients who continued smoking after the first infarction had an additional risk of reinfarction of 39.5% for each 5 cigarettes smoked per day (data not shown).
The OR of death was 3.07 (95% CI, 1.13-8.33) for those patients who continued smoking compared with those who quit smoking after their first infarction.
DISCUSSION
This study examined several factors influencing secondary prevention in coronary patients after a first AMI, including counseling aimed at smoking cessation. An observational study is the only possible design due to ethical difficulties arising with planned intervention studies. No differences were found in the incidence of reinfarction between patients who quit smoking and those who had never smoked. Those patients who continued smoking after their first myocardial infarction had a three-fold risk of a second infarction compared to those who quit smoking.
Previous studies have compared the effect of cardiac rehabilitation programs and lifestyle changes with usual standards of care.1,2,13 Although programmed interventions resulted in a better prognosis for these patients, none of the seven clinical trials of secondary prevention which analyzed the effect of smoking cessation and other preventive measures on the incidence of reinfarction detected any significant differences in coronary events, and only one study showed an increased rate of smoking cessation in the experimental group.2
Preventive intervention programs do not always improve health outcome,14 and it is not easy to apply proven preventive measures to daily practice.15,16 Other randomized trials have promoted interventions to improve social support, quality of life and care, but with uncertain results.2,17
This study was designed to determine the effectiveness of secondary prevention measures in patients with coronary heart disease. Treatment and counseling were provided by family doctors and primary care nurses, with occasional visits by the cardiologist. Standards of clinical management of coronary patients are constantly being improved.18 The matched design of this case-control study, with one of the criteria being the duration of the secondary prevention period, was aimed at eliminating a possible cohort effect resulting from varying access to the increasing evidence supporting protective therapy. Our results, therefore, reflect the permanent updating of preventive measures as they become incorporated into patient care.
Although we included a variable to attempt to adjust for variations in patients´ adhesion to counseling, it had the logical limitations associated with subjective self-referral. The reliability of this variable can be seen from the fact that patients who continued smoking reported a lower percentage of adaptation to the recommended treatment. Of note, too, was the fact that the control group patients who were unable to quit smoking reported a greater percentage of stress. This association has been seen by others.19 For other lifestyles, the smokers walked less often and introduced fewer changes into their diet, although these differences were not statistically significant. The least healthy lifestyles, therefore, were generally found in the same patients.
Because patients who die shortly after an acute coronary event do not have the opportunity to implement measures of secondary prevention or to benefit from health education activities, we excluded those who survived less than six months after a first infarction.
The beneficial effect of smoking cessation has been studied extensively in relation to the reduction in mortality after an infarction.20-22 Several cohort studies of coronary patients, collected in a meta-analysis, found different benefits regarding mortality after cessation of smoking. Four of these studies reported no significant individual differences, though the combined risk showed a reduction in mortality of 46%.7
Few studies analyze the incidence of fatal or non-fatal infarction in coronary patients in relation with smoking. Hedback et al23 found a statistically significant reduction in the percentage of patients having a reinfarction among those who ceased smoking, after both five- and ten-year follow-up. However, no multivariate risk analysis was done. In our study, the dependent variable was the occurrence of a further infarction, either fatal or non-fatal. This event was clearly related with smoking, independently of other risk factors; with the analysis showing protection for those who ceased smoking and an increased risk for those who continued. A similar effect was reported by Rea et al,24 who demonstrated that the longer the cigarette-free period the greater the protection.
Multivariate OR in our study (smokers vs nonsmokers) revealed a possible protective role of the independent variables related with other habits (exercise and diet) and with drug therapy, as the OR rose from 2.83 to 3.39. Thus, any change towards a beneficial lifestyle or the incorporation of protective treatment would reduce the risk in those patients who were unable to quit smoking. Although this protection has been shown with specific drugs, such as beta blockers among the group of antihypertensive agents,25 our patients were questioned about the use of any antihypertensive drug in order to facilitate their response over the telephone. The same applies to statins,26 which are the drugs most often prescribed for hypercholesterolemia during the period of secondary prevention, although few patients were able to identify their generic name.
Hermanson et al27 found a significant OR of 1.7 for fatal infarction in patients who continued smoking. Using mortality as a dependent variable in our study showed that patients who continued smoking had a significantly increased risk of death, with an OR of 3.07 compared to those who quit smoking. However, our study was not designed to analyze mortality in patients with coronary heart disease.
The most important limitation in our study is probably the credibility of the responses given by the patients over the telephone, especially those related to smoking. Previous studies have relied on the cotinine assay for validation,28 but this was not possible in our study due to its design. Nevertheless, telephone interviews are not incompatible with obtaining exact responses, as the interview is conducted voluntarily with a stranger. The same can be said for the other responses concerning lifestyle and treatment.
Another factor could be the validity of responses given by close relatives. An attempt to control for this bias was made by interviewing the same type of relative in the control patient, so that the comparative analysis should still be valid. To avoid interviewer bias, both the case and the control were interviewed by the same person. The characteristics of this study, nested in the same cohort of patients, and the pairing between cases and controls according to age, sex, hospital, interviewer and duration of the secondary prevention period, strengthens the validity of our results. Patients with a prior bypass or angioplasty were excluded from the study in order to eliminate bias resulting from therapy which could interfere with the natural course of the disease or with the benefits of secondary prevention recommendations. Patients with reinfarction within six months of the initial event were also excluded, as a shorter period of secondary prevention can be considered insufficient.
Previous studies have either examined the effects of intervention programs or analyzed mortality in cohorts of patients with coronary heart disease according to smoking habits.2,7 Our observational study of current secondary prevention also provides a multivariate analysis of the association between preventive measures and the incidence of reinfarction in patients with coronary heart disease.
CONCLUSIONS
Patients who have had a first myocardial infarction are liable to have another. Those who quit smoking have a three-fold reduction in risk of a second coronary event compared to those who continue smoking. Secondary prevention of coronary heart disease requires responding to the difficulty encountered by some patients with smoking cessation. Any effort aimed at this will reduce morbidity and mortality in patients with coronary heart disease. Counseling about smoking cessation should take place in different areas, although primary care nurses, after adequate training, will probably be able to achieve high rates of smoking cessation.29
ACKNOWLEDGEMENTS
The authors thank Dr. Enrique de los Arcos, chief of the Servicio de Cardiología del Hospital de Navarra, for encouraging the development of this study and for chart examination. We should also like to thank María José Urbina (primary care nurse) for her help.
Financed by a grant of the Departamento de Salud del Gobierno de Navarra, 1998.
Correspondence: Prof. M. Serrano Martínez.
Departamento de Epidemiología y Salud Pública, Facultad de Medicina, Universidad de Navarra,
Irunlarrea, 1, 31080 Pamplona, Spain.
E-mail: mserranm@cfnavarra.es