Although the incidence and mortality in acute myocardial infarction (AMI) are rapidly decreasing in the Western world,1^–3 cardiovascular disease remains the leading cause of death, regardless of gender. Coronary thrombus occluding a central vessel is the predominant cause in patients with ST-elevation, usually progressing to transmural myocardial infarction.4 5 ST-elevation myocardial infarction is the most serious and damaging form of AMI, resulting in large infarctions, more complications and higher short-term case death.6 7

In addition to the observed changes in coronary heart disease incidence and mortality, changes in the severity of the disease with smaller AMIs, more unstable angina pectoris (UAP) and lower case death have been reported.7^–9 These changes are probably due to improved medical and interventional management of coronary heart disease (CHD). At the same time, the risk factor pattern in the population is also changing, with decreasing cholesterol levels, less tobacco smoking and lower levels of blood pressure but increasing overweight and obesity.10^–13

Tobacco smoking is one of the major risk factors of AMI.14 In a study of patients with acute coronary syndromes (ACS), in the Euroheart Survey smoking was strongly associated with presenting with ST-elevation.15 However, in the Euroheart Survey, a large proportion of the patients without ST-elevation were discharged with a diagnosis of UAP. Accordingly, it remains unclear to what extent the differences between smokers and non-smokers reflect differences between AMI and UAP. The Register of Information and Knowledge about Swedish Heart Intensive Care Admissions (RIKS-HIA) collects information on a large number of admissions for suspected ACS in Sweden. We used this large database to explore the determinants of STEMI compared with non-ST-elevation myocardial infarction (non-STEMI) exclusively among patients with diagnosed AMI.

METHODS

In Sweden, the RIKS-HIA continuously registers all patients admitted to hospitals with coronary care units (CCUs). The full protocol has been published previously16 (detailed information and a complete protocol are also available at http://www.riks-hia.se).

On admission, patients receive written information about RIKS-HIA and other quality registries, and have the right to deny participation, but few of the participating patients exercise this right. According to Swedish law, written consent is not necessary. The register is approved by an ethics committee and the National Board of Health and Welfare. All personal identifiers are removed from the RIKS-HIA data file when used for research purposes.17 RIKS-HIA started in 1995 with 19 participating hospitals, increasing gradually during the study period from about 50% in 1996 to 95% out of 78 Swedish hospitals in 2004. The present study is based on all consecutive patients aged 25–84 years without a history of prior AMI that were admitted to hospital between 1 January 1996 and 31 December 2004 and discharged with a diagnosis of STEMI or non-STEMI. Of 105 365 patients, we excluded 5989 with missing data on smoking (6%), 5325 (5.4%) with left bundle-branch block (LBBB) and 635 (0.6%) with a pacemaker ECG. Of those patients (mean age 71.4 years) with missing data on smoking, 34.2% were diagnosed as having STEMI, and 39.9% were women. After these exclusions, the final sample included 93 416 patients admitted to hospital with their first AMI.

Case report forms filled in by nurses and used to record information on age, gender, tobacco-smoking status (never smoking, ex-smoker (defined as not smoking for more than 1 month before admission to hospital) and current smoker), known hypertension, diabetes mellitus (history or medication), known hyperlipidaemia, previous MI, previous coronary revascularisation, previous medication (ACE inhibitors, beta-blockers, aspirin and lipid-lowering drugs, ie, statins), presenting symptoms, electrocardiography, reperfusion treatment, other pharmacological treatment, intervention procedures, major complications and outcomes during hospital stay.

Definitions

The criteria for a diagnosis of acute MI were standardised and identical for all participating hospitals using the World Health Organization and Joint European Society of Cardiology (ESC) and American College of Cardiology Committee criteria.18

The ST segment was recorded as the first choice of which of the following alternatives accurately described the ST segment on the electrocardiogram at entry: 1 = normal, 2 = left bundle-branch block or pacemaker, 3 = ST segment elevation, 4 = ST segment depression, 5 = T-wave inversion and 6 = other changes. ST-elevation was defined as significant ST-segment elevation in at least two adjacent leads but was not considered in cases of left bundle-branch blocks or pacemaker ECG. Biochemical criteria were revised during the study period according to the ESC/ACC consensus document18 (eg, the limit for MB fraction of creatine kinase (CK-MB) changed in 2001 from >10 to >5 μg/l).

Statistical methods

To describe differences between STEMI and non-STEMI, baseline characteristics were summarised as means or percentages. The independent association between smoking status and STEMI was assessed by means of logistic regression, where STEMI was entered as the dependent variable, and the following variables were used as covariates (possible confounders): age, gender, history of tobacco smoking (never smoking, current smoking and former smoking), history of coronary by-pass surgery (CABG), history of percutaneous coronary intervention (PCI), history of diabetes mellitus, history of hypertension, medications used before entry into the study (eg, ACE inhibitors, beta-blockers, aspirin and lipid-lowering drugs) and year of admission. To investigate potential interactions between gender and smoking, as well as between age and smoking interaction terms (gender×current smoking and age×current smoking) were defined and introduced into the models. All statistical analyses were performed using SPSS version 15.0 (SPSS, Chicago). Odds ratios (OR) were calculated from the logistic regression model. Because of the large size of the population, 99% CIs were used.

RESULTS

Of the 93 416 patients (35.6% women and 64.4% men) with a first admission for AMI, approximately 38% presented with STEMI and 62% with non-STEMI. The proportion of patients with STEMI decreased from about 43% in 1996 to 35% in 2004. During the same period, smoking at the time of admission decreased with a few per cent in patients with or without STEMI (table 1).

Table 2 shows the baseline characteristics for all patients by presentation with or without STEMI. Smoking was more common in patients with STEMI than non-STEMI: 31.0% vs 22.9%. Patients with STEMI were slightly younger and more often men. Patients with non-STEMI had more prior diseases and accordingly more often medication with aspirin, beta-blocker, ACE inhibitor and/or statin. Chest pain was a more common symptom in patients presenting with STEMI, 87.4% compared with 75.7% in non-STEMI patients. In patients with non-STEMI, 56.9% presented with STT-depression or pathological T-wave, but 43.1% had normal or unstated ECG. Thrombolysis was common in STEMI patients, about 55% compared with 4.3% in non-STEMI; in addition, almost 14% were treated with PCI or CABG in the STEMI group compared with 3.5% in the non-STEMI group.

Table 3 shows the demographic and clinical characteristics by presentation in patients with a first myocardial infarction (STEMI or non-STEMI) and odds ratios for presenting with STEMI. The proportion of women was somewhat lower among STEMI patients: age-adjusted OR 0.90 (99% CI 0.87 to 0.94). In addition, patients with STEMI were slightly younger, with a lower prevalence of hypertension and diabetes, and less prior revascularisation. Overall, 31% of the patients with STEMI and 22.9% of the non-STEMI patients were current smokers (age-adjusted OR 1.40, 99% CI 1.34 to 1.47). After further adjustment for age, hypertension, diabetes, prior revascularisation (PCI or CABG), medication on admission (aspirin, ACE inhibitor, beta-blocker or statin) and year of admission, the adjusted OR was 1.33 (99% CI 1.26 to 1.39).

Among men with STEMI, 58% of the youngest men (<50 years) were current smokers as compared with 47% for non-STEMI (fig 1). Sixty-seven per cent of the women <50 years old with STEMI were current smokers; the figure was slightly lower (ie, 50%) among women with non-STEMI (fig 2). Among men and women, the prevalence of tobacco smoking and the difference between STEMI and non-STEMI decreased with age: in the oldest age group, about 10% were current smokers.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Smoking by age among men admitted to hospital with a first acute myocardial infarction, ST-elevation myocardial infarction (STEMI) or non-ST-elevation myocardial infarction (Non-STEMI).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Smoking by age among women admitted to hospital with a first acute myocardial infarction, ST-elevation myocardial infarction (STEMI) or non-ST-elevation myocardial infarction (Non-STEMI).

In subsequent analyses we investigated the prevalence of current smoking and prior risk factors in patients with STEMI and non-STEMI. These analyses were done separately for men and women aged <65 years and ⩾65 years (tables 4, 5). Among patients <65 years with STEMI, there were a higher proportion of current smokers than among patients with non-STEMI, with a more pronounced difference among the women: multiple-adjusted OR in men 1.33 (99% CI 1.22 to 1.43) and in women 2.01 (99% CI 1.75 to 2.30) (test for interaction p<0.0001).

Prior revascularisation among men and women <65 years was associated with less risk of presenting with STEMI, whereas hypertension and diabetes were associated with increased risk though only among women. Tobacco smoking was less common in older patients: about 18% of both men and women ⩾65 years old presenting with STEMI were current smokers as compared with 14.5% of the men and 13.5% of the women ⩾65 years old without STEMI (tables 4, 5). Smoking was related to a higher risk of STEMI in older patients: multiple-adjusted OR in men 1.14 (99% CI 1.04 to 1.25) and in women 1.33 (99% CI 1.20 to 1.48). Diabetes (multiple-adjusted OR 0.90, 99% CI 0.82 to 0.98) carried less risk of presenting with STEMI in men in this age category (ie, aged ⩾65 years) but not in women. Prior PCI or CABG was associated with less risk of presenting with STEMI in both men (multiple-adjusted OR 0.48, 99% CI 0.40 to 0.58) and women (multiple-adjusted OR 0.41, 99% CI 0.31 to 0.56).

DISCUSSION

In this large survey, which included more than 90 000 individuals with a first AMI, we found that tobacco smoking was strongly associated with STEMI, particularly among younger patients (ie, <65 years of age) and among women patients. Prior research from our group has demonstrated that, while there are demonstrable differences in clinical presentation between men and women below the age of 65, these differences have, by and large, disappeared in patients over that age. Accordingly, in order to study whether the influence of smoking differed in men and women at different ages, we selected the age of 65 as a cut-off.15 The proportion of active smokers decreased with age, as did the difference in smoking prevalence between patients with STEMI and patients without STEMI. In the oldest age groups, only a small number of the patients were regular smokers, with little or no difference in clinical presentation between smokers and non-smokers.

From the early 1980s, both incidence and mortality from CHD have decreased markedly in Sweden. The decline is partly due to risk-factor modification at the population level, with not only less tobacco smoking but also decreasing levels of total cholesterol.10 11 The percentage of regular smokers in the general population in Sweden is currently among the lowest in the world, being less than 20%.11 Reduction in coronary mortality is likely due to improved treatment and changes in risk factor pattern.7 13 However, in addition to decreasing incidence and mortality, reduced case death has been observed, along with a change in clinical presentation with milder symptoms, smaller infarctions and less severity of the condition.7^–9 Although this trend might be a consequence of increased awareness and admission of less severe cases, changes in risk factors might also contribute.

Potentially, lower smoking rates in the population could explain decreasing case severity and reduced coronary mortality. Some studies have, however, found that smokers with AMI have lower mortalities compared with non-smokers, a phenomenon often referred to as the “smoker’s paradox,”19^–21 but this reduced mortality could equally well be due to the fact that smokers with AMI are generally younger, with fewer other risk factors such as diabetes, or hypertension, and they also have less extensive coronary disease.21 In addition, in the present study we found that patients with STEMI actually had a higher mortality than non-STEMI patients, 11.8% compared with 9.3%, despite the fact that patients with non-STEMI were older and had more diabetes. After adjusting for age and diabetes, patients presenting with STEMI still had a higher mortality, adjusted RR 1.58 (99% CI 1.52 to 1.63) for 30-day mortality and 1.22 (99% CI 1.19 to 1.26) for 1-year mortality. These findings are consistent with results from the CREATE study in which Indian patients with STEMI more often were men, current smokers and had a higher 30-days mortality.22

Smoking is the most important of all avoidable risk factors, causing a substantial proportion of all CHD cases. The risk associated with smoking decreases rapidly within a few years of quitting. Still, former smokers retain an excess risk of AMI over some years.23 A recent review showed that quitting smoking is associated with a substantial reduction in risk of all-cause mortality with a more pronounced effect in patients with CHD: pooled crude relative risk (RR) 0.64 (95% CI 0.58 to 0.71). In patients with non-fatal myocardial infarctions, the RR was 0.68 (95% CI 0.57 to 0.82). Furthermore, the risk reduction occurs relatively early, within 2 years, after stopping smoking.24

These findings imply that the short-term effects on the haemodynamic system are more important than the chronic exposure in the development of coronary thrombosis. Potentially, short-term effects on coagulation and fibrinolysis could be more important than long-term effects on atherosclerosis, with respect to ST-elevation ACS.

This could partly explain why we found no difference in the risk of presenting with STEMI between former smokers and those who never smoked. Former smoking was defined as >1 month since quitting, but more exact information about the time elapsed since giving up is not recorded.

Only a few studies have examined the shift in case severity as a potential cause of the decreasing case death in AMI. Data from Finland indicate decreasing incidence and mortality, which were attributed to improved primary prevention and better acute coronary care. During the same period, the proportion of milder infarctions increased.7 The US-based ARIC study (The Atherosclerosis Risk in Communities), however, provided mixed support for a decrease in the severity of myocardial infarction. The proportion of patients with STEMI increased by 10% per year from 1987 to 1994 while the mean peak CK-MB level decreased 5% per year during the same period. Furthermore, during the same period, there was an increase in hospitalised infarctions and a decline in case death.9 It should be noted that neither the ARIC study nor the FINMONICA MI Register study included information on smoking, which, according to these data, could be important when exploring factors determining severity of myocardial infarction.

A stronger association was observed between tobacco smoking and STEMI in women, with two-thirds of the women <55 years with STEMI still active smokers on admission. Cigarette smoking is a well-known risk factor for AMI and sudden cardiac death.25 Moreover, cigarette smoking has been shown to cause endothelial dysfunction and coronary vasospasm, and can increase the risk of atherotrombosis, influencing fibrinolytic and antithrombotic factors.26 27 Women with chest pain more often than men have normal angiograms28 and a higher prevalence of plaque erosion.29 Furthermore, elevated levels of procoagulant factors have been shown to increase the risk of myocardial infarction in young women.30 The mechanisms involved in tobacco-smoking-related cardiovascular disease are not entirely clear; however, there is evidence for an increased propensity toward thrombosis in smokers.31 Why this should affect men and women differently is not clear.

More than 95% of all CCU patients in Sweden are included in the RIKS-HIA register. This provides a unique opportunity to investigate in detail the effects of smoking on clinical presentation in a large population of AMI patients.

Limitations

Despite the decline in tobacco smoking at the population level in the past few decades, smoking is still common in patients with AMI. The decreasing smoking rates in the Swedish population could contribute to the temporal trends regarding disease severity and decreasing mortality from coronary disease.

A limitation of this study was that the diagnostic criteria for myocardial infarction were changed in 2001, with a lower limit for CK-MB from >10 to >5 μg/l.18 This lower limit has probably led to a transition from UAP to non-STEMI diagnoses. This will have contributed to a larger proportion of patients with non-STEMI. However, we did control for year of hospitalisation in the multivariable analyses, with no material change in the estimates for current tobacco smoking.

Another limitation is that a validation of the RIKS-HIA shows a misclassification of tobacco smoking in a small proportion (<5%) of the patients. To what extent this classification error influenced the differences between patients with or without STEMI is unknown, but the effect is likely to have been minor. Missing data on smoking habits may underestimate the proportion of smokers, especially in older age groups. Moreover, in former smokers no data on when the patients gave up tobacco smoking are recorded, other than the fact that it was more than 1 month before the onset of the AMI. In addition, we have no information about the amount of tobacco consumed, or pack-years. The extent to which this potential misclassification may have influenced our findings is unknown. However, because smoking is less common in the older age groups, this problem is mainly limited to the younger age groups.

Conclusion

Overall, one in three AMI patients in Sweden is currently a regular smoker on admission to hospital but with a large variation in relation to age, gender and clinical presentation. Noteworthy was the fact that two-thirds of the women with STEMI and <50 years were smokers. This figure can be compared with the oldest patients of which one in 10, or less, were smokers, with small differences between genders or in clinical presentation. Overall, tobacco smoking was associated with increased risk of presenting with ST-elevation, particularly among younger patients and women patients. These results underline the importance of smoking as a major risk factor for presenting with more severe AMIs.