Gender perspective on risk factors, coronary lesions and long-term outcome in young patients with ST-elevation myocardial infarction
- 1Department of Medical and Health Sciences, Division of Cardiovascular Medicine, Linköping University Hospital, Linköping, Sweden
- 2Department of Cardiology, Uppsala University Hospital, Uppsala, Sweden
- 3Uppsala Clinical Research Center, Uppsala University Hospital, Uppsala, Sweden
- Correspondence to Sofia Sederholm Lawesson, Department of Medical and Health Sciences, Division of Cardiovascular Medicine, Linköping University Hospital, SE-581 85 Linköping, Sweden;
- Accepted 17 November 2009
Objective Previous data on young patients with acute coronary syndrome (ACS) have indicated higher rates of normal coronary angiograms but higher mortality in women than men. However, ST-elevation myocardial infarction (STEMI) differs from non-ST-elevation ACS in many aspects. We elucidated sex differences in risk factors, angiographic findings and outcome in consecutive STEMI patients below 46 years of age.
Design Retrospective cohort study.
Setting The Swedish registers for CCU care and coronary angioplasty; RIKS-HIA and SCAAR.
Patients 2132 STEMI patients below 46 years of age admitted to intensive coronary care units in Sweden between 1995 and 2006 and followed for at least 1 year.
Main outcome measures Angiographic findings and short-term and long-term mortality.
Results Risk factors were more common in women. Significant coronary lesions were equally common (92.1% vs 93.1%, p=0.64) while single vessel disease was more common (72.9% vs 59.3%; p<0.001) in women. Women had higher multivariable adjusted in-hospital mortality, OR 2.85 (95% CI 1.31 to 6.19) while long-term mortality was the same, HR 0.93 (95% CI 0.60 to 1.45). The catch-up of mortality in men might be related to a higher occurrence of re-infarctions, HR 1.82 (95% CI 1.25 to 2.65).
Conclusions STEMI below age 46 is a more rare condition in women than in men and more often related to cardiovascular risk factors. More than 90% of both men and women had coronary lesions, in women more often single vessel lesions. Female sex is associated with higher in-hospital mortality, while long-term mortality is low without difference between genders.
Myocardial infarction (MI) among adults aged less than 46 is unusual, comprising 2–10% of all cases.1 Women are relatively spared from coronary artery disease (CAD) before menopause and in order for it to occur, it is mostly accompanied by the presence of cardiovascular risk factors, especially smoking.2 Previous studies with mixed MI populations including both ST-elevation MI (STEMI) and non-ST-elevation MI (NSTEMI) have shown higher prevalence of normal coronary arteries on coronary angiography in young patients compared to old patients with suspected or diagnosed acute coronary syndrome (ACS), especially in females.2 3 This could indicate a higher prevalence of non-atherosclerotic causes of MI such as coronary artery dissection,4 hypercoagulable states5 or coronary spasm in women. A sex-age interaction with higher mortality in younger but not in older, women compared to men has been shown in two registries of mixed MI patients,6 7 but the opposite was found in a population of Q-wave infarctions.8 In older populations, previous studies have shown higher multivariable adjusted mortality in women compared to men with STEMI9 but the opposite in NSTEMI patients.10 These contradictory findings emphasise the need for additional studies of young MI patients separating these two conditions, which also differ regarding treatment recommendations.11 12
Thus we undertook the present observational multicentre study on sex differences in aetiology, management and prognosis of young patients with STEMI. The objectives of the study was to elucidate coronary risk factors, coronary lesions at angiography, treatment and short-term and long-term outcomes, in patients below 46 years of age admitted to coronary care units (CCUs) in Sweden between 1995 and 2006 with STEMI and to compare the genders.
Material and methods
Data for this study came from the prospective observational Register of Information and Knowledge about Swedish Heart Intensive care Admissions (RIKS-HIA) and from the Swedish Coronary Angiography and Angioplasty Register (SCAAR), more thoroughly described in previous reports.13 14 Both registries are sponsored by the Swedish health authorities and are independent of commercial funding. Further information about the registries and the full RIKS-HIA protocol is available at http://www.ucr.uu.se/rikshia and http://www.ucr.uu.se/scaar.
The RIKS-HIA register
RIKS-HIA started as a national register in 1995 and covers 95% of all CCUs in Sweden from the year 2000. Variables including age, sex, smoking habits, co-morbidity, delay times, symptoms, biochemical markers, cardiac procedures, complications, therapies and diagnoses were continuously recorded online over the internet for all CCU patients. The criteria for the diagnosis of acute MI were standardised and identical for all hospitals using the WHO and Joint European Society of Cardiology/American College of Cardiology Committee criteria.15 16 Source data verification has continuously been performed yearly showing an overall agreement of 94–97%.14 17 Missing data for the variables analysed in the main study ranged between 0% and 6%.
The SCAAR register
Angiographic data were recorded locally at some hospitals in 1995 and in 1998 the national register SCAAR was formed. Available data before 1998 were gathered from local databases and uploaded into SCAAR. Since 2004 all consecutive patients from all Swedish PCI centres were recorded. Variables such as previous coronary procedures, indication for angiography/PCI, findings, number, diameter and length of stents, restenotic lesions and overall success were registered. A coronary lesion was defined as significant if at least 50% of lumen diameter was reduced. Since 2001, the SCAAR register was internet-based with recording of data online through a web interface in the catheterisation laboratory. Monitoring and data verification has been performed in all hospitals since 2001 and the overall agreement has been about 95% each year.13 Data from performed angiograms were available for 84% of the patients, who, according to the RIKS-HIA register, had undergone coronary angiography and/or PCI within 7 days from admission to hospital with STEMI.
From the Swedish Heart Surgery Register, http://www.ucr.uu.se/hjartkirurgi, information was gathered about previous heart surgery. From the National Board of Health and Welfare, http://www.socialstyrelsen.se, the National Cause of Death and National Patient Registries were available with information about cause of death until 31 December 2005, vital status of all Swedish citizens until 31 December 2006 and about previous history of myocardial infarction, heart failure, stroke, dementia, cancer, chronic obstructive pulmonary disease or chronic kidney disease.
In all, 5088 patients below 46 years of age (4018 men and 1070 women) were diagnosed with an MI between 1 January 1995 until 31 December 2006. Of those, 1790 men and 401 women had a first register recorded diagnosis of STEMI, defined as ST-elevation on admission ECG, and a diagnosis of acute MI at discharge. Fifty-nine patients with other heart rhythm than sinus rhythm or atrial fibrillation/flutter were excluded and the final study population consisted of 1748 men and 384 women.
The substudy of coronary angiographic findings was performed in the 1257 patients where angiograms were performed within 7 days of admission. This cohort was also divided into three subcohorts depending on admission year since as the SCAAR register was not complete in the beginning of the study period. Two further subcohorts were created depending on timing of the angiography within the first week since the dominating reperfusion therapy has changed during this period from fibrinolytic therapy to primary PCI in Sweden.
The registers were approved by the National Board of Health and Welfare and the process of merging registers were approved by the Swedish Data Inspection Board. The study was approved by the local ethics committee and complies with the Declaration of Helsinki.
Continuous variables were summarised by their mean and SD or median and IQR as appropriate. Categorical variables were summarised by counts and percentages. Comparisons between different strata were performed by χ2 tests for categorical variables and by Student t tests or Mann-Whitney tests for continuous variables. Crude and multivariable-adjusted ORs with 95% CIs were calculated from logistic regression analyses in order to identify independent predictors of mortality and to compare men and women regarding in-hospital and long-term mortality adjusting for possible confounders. In addition to sex, reperfusion therapy and cardiovascular risk factors with differences in prevalence between genders were included as covariates in the analyses. All variables were coded as dichotomous dummy variables (yes or no). Comparison of 10-year cumulative mortality or new incidence of myocardial infarction was performed by the Mantel-Cox log rank test and by multivariable Cox regression models using the same covariates as in the logistic regression models. All statistical analyses were performed with the SPSS Version 15.0 software.
A total of 384 women and 1748 men were included, with mean ages of 40.4 and 40.8 years respectively (p=0.14). The age distribution is given in figure 1. The women had a higher cardiovascular risk factor burden with higher incidence of hypertension, diabetes mellitus and smoking than their male counterparts and more often multiple risk factors. Smoking was the most prevalent risk factor; 78% of the women and 75% of the men were former or active smokers, 64% and 58% were active smokers (table 1).
On admission, 6.3% of the woman compared to 3.5% of the men were in cardiogenic shock, p=0.01. Women had higher Killip class on admission and more symptoms of heart failure during the hospital stay. However, there was no statistically significant difference in systolic function in patients that had been investigated with echocardiography. There were no differences in symptom-to-door time or in acute reperfusion therapy frequency between the sexes. Bleeding complications were rare in this young cohort. At discharge there were significant differences between the sexes in therapy. Men were more often discharged with β-blockers and statins while women more often were prescribed calcium-channel inhibitors. There were no sex differences in platelet inhibitor or ACE inhibitor therapies (table 2).
During the whole study period, coronary angiograms were equally common among the genders and performed within one day in 46% of the women and in 45% of the men (p=0.94) and within 1 week of admission in 60% of the women and 59% of the men (p=0.77) (table 2). Data from the angiograms were available for 84% of these patients.
There was no difference between genders in normal coronary angiograms as 92% of the women and 93% of the men had significant coronary artery disease. On the other hand there was a difference in the extent of coronary lesions where men more often had multi-vessel or left main disease (33.6% vs 19.2 %; p<0.001), whereas one-vessel disease was more common in women (59.3 % vs 72.9 %; p<0.001). The validity of these findings was verified by comparisons of subcohorts based on admission year or day of investigation, respectively (table 3).
In-hospital mortality was low, 3% in women versus 1% in men (table 2), crude OR for women 2.83 (95% CI 1.32 to 6.03). Female sex appeared as an independent predictor in the multivariable model of in-hospital mortality, OR 2.85 (95% CI 1.31 to 6.19). Diabetes mellitus more than doubled the in-hospital mortality risk while acute reperfusion therapy was associated to halved mortality risk. Also after one year, multivariable adjusted mortality was higher in women, OR 2.00 (95% CI 1.03 to 3.89) (table 4). When the cohort was followed up to 10 years (mean 5.4 years) the mortality rates were similar between genders (p=0.40) (figure 1). The same result was seen in the multivariable Cox model (HR 0.93, 95% CI 0.60 to 1.45; p=0.75) (table 4). In the Mantel-Cox log rank test men had significantly (p=0.01) higher incidence of a second new MI during the following 10 years (figures 2 and 3). This was also true after adjusting in the Cox multivariable model where male gender was associated with new MI, HR 1.82 (95% CI 1.25 to 2.65; p=0.002).
The majority of studies published on young MI patients have been single-centre studies with mixed kinds of ACS patients, or populations from randomised trials. Very few studies have done gender comparisons. Our study includes all consecutive patients below 46 years of age treated for STEMI in Sweden 1995–2006 and thus gives a solid base for conclusions about young STEMI patients in a gender perspective.
The proportion of women was 18%, which is in accordance with previous reports of MI patients in this age group18 19 but considerably lower than in older patients.17 A higher-risk factor burden was seen in the young women compared to men, consistent with previous data.20 A majority of the young patients had at least one risk factor for CAD, and a quarter of the women compared to less than a fifth of the men had two or more risk factors. The largest gender difference was seen in hypertension and diabetes, with a 50% higher rate in women than in men. As in almost all previous studies on young MI patients, smoking had high prevalence also in our study population and was the most common cardiovascular risk factor in both genders.3 19 21 In contrast to previous reports,3 we found that smoking was more common in women and as many as two-thirds of the women were current smokers, which reflect its importance as a risk factor for MI in Swedish women, who currently smoke more than the men do.22 This is particularly alarming as smoking is known to be more dangerous to women than to men,23 and epidemiological data show that it is the most important risk factor for developing an MI at a young age.21
Several studies have found a sex difference in normal angiograms in young populations with mixed MI, indicating a higher incidence of non-atherosclerotic causes of MI in women compared to men.2 3 This phenomenon has been associated with vasospastic syndromes and hypercoagulable states such as oral contraceptive use24 or hereditary coagulation disorders.5 It has been shown that young women have more active platelets after MI,25 and, more often, plaque erosions instead of plaque ruptures such as in men and older women.26 In a study from the CASS register, a third of the women with documented MI had zero-vessel disease compared to a fifth of the men.2 An Italian multicentre study on mixed MI patients who underwent angiography showed similar results with normal or no significant lesions in 29% of the women compared to 15% of the men.3 In our material more than 90% of both women and men showed some type of coronary lesion. However, single-vessel lesions were most common in both genders and even more frequent in women, whereas multivessel disease was considerably rarer than in men. These findings support the suggestion that the thrombotic occlusion in STEMI in women is less often related to severe coronary atherosclerosis than it is in men. Although the proportion of early coronary angiographies increased over time the results in the male and female cohorts remained consistent when comparing cohorts at different times, which supports the concept of a gender difference in underlying pathophysiology.
In contrast to several previous reports it is noteworthy that the acute phase treatment was strikingly similar between genders—for example, in delay times, reperfusion treatment, primary PCI, coronary angiography, revascularisation procedures, antithrombotic medications and β-blockade. There were slight differences in medication at discharge with somewhat fewer women prescribed β-blockade and statins, which might be related to differences in patient characteristics and lipid levels. However, even when patients included between the years 2004 and 2006 were analysed in our study, we still found a significant sex difference in statin therapy in spite of the higher risk burden with more diabetes, smoking and hypertension in women than in men in this age group. The significant 5% less β-blockers and 10% less lipid-lowering drugs in women did, however, not translate into any difference in the long-term mortality. On the contrary men had a higher incidence of a second myocardial infarction. Unfortunately, we have no knowledge in this study about possible changes in medication after discharge.
Multivessel or left main disease was thus shown to be much more common in men than in women. In spite of this more advanced disease in men, the young women had three times higher mortality compared to the men. Actually, after adjusting for diabetes and other risk factors, the sex difference in mortality was even more pronounced. Female sex was in fact the strongest independent predictor of in-hospital death in the multivariable adjusted model. A unique aspect of the present study is the possibility of completing long-term follow-up of all patients up to 10 years. Although myocardial infarction at a very young age might be perceived as a catastrophe it is then very comforting to report that the long-term outcome is very favourable in both women and men. Using the modern treatment the in-hospital mortality was very low, 1–3%, and thereafter there was around 1% mortality and 1–1.5% re-infarction yearly. There was no difference in long-term mortality between genders. In contrast to studies on older MI populations,27 women did not have longer symptom-to-door time compared to men in this young population. Several older studies on fibrinolytic therapy have more bleeding and intracranial haemorrhages in women.28 29 This could, however, not be confirmed in our or another newer study.30 Thus, neither longer delay times nor more complications could explain the difference in early mortality in our young female STEMI population. The women had, however, higher Killip class and cardiogenic shock on admission and a tendency for more heart failure symptoms during hospital care. As women had more single-vessel disease the higher in-hospital mortality might be related to more cases of sudden thrombotic occlusion of a major coronary artery without any preformed collaterals leading to more negative haemodynamic effects than in men with a more chronic coronary artery disease and available collateral circulation. In men there was, on the other hand, a substantially higher risk of re-infarction at long-term follow-up, probably explained by their more severe coronary artery disease and probably also explaining the catch-up in long-term mortality.
The young population below 46 years old with STEMI is luckily uncommon. Therefore we had to use more than a decade of consecutive patients from a whole country to accumulate a large enough patient cohort, which is to our knowledge the largest consecutively gathered young STEMI group studied so far in the world from a gender perspective. STEMI below age 46 is a rarer condition in women than in men and more often related to cardiovascular risk factors. More than 90% of young STEMI patients have identifiable coronary lesions, which in women more often are single-vessel and less often multi-vessel lesions. Secondary prevention is thus necessary according to current guidelines. Female sex is associated with higher in-hospital mortality, while long-term mortality is low without difference between genders. The catch-up of mortality in men seems related to the higher rate of re-infarction. These findings might have implications for the selection of management strategies in patients with MI at a young age.
Some variables important for development of CAD are optional to register in the RIKSHIA registry—for example, heredity and lipid levels. Thus it was not possible to adjust for those variables in the present study.
Selection bias regarding coronary angiograms cannot be excluded during the first years when all hospitals did not participate in SCAAR. However, analyses of cohorts divided into the different time periods do not indicate significantly different data in the early compared to the recent years. Stenosis severity was determined by each individual interventional cardiologist and no core laboratory was used. It is however unlikely that the rough division into significant and not significant stenosis that we used in the two sexes would have been influenced by the absence of a core laboratory.
We acknowledge all participating hospitals for their help and cooperation to contribute with data to RIKS-HIA and SCAAR. We also acknowledge Max Köster at the National Board of Health and Welfare in Stockholm for help with merging the databases.
Funding The registers are supported by the Swedish Heart-Lung Foundation, the Swedish Association of Local Authorities and Regions, the National Board of Health and Welfare and the Swedish Society of Cardiology.
Competing interests None.
Ethics approval The registers were approved by the National Board of Health and Welfare and the process of merging registers were approved by the Swedish Data Inspection Board. The study was approved by the local ethical committee and complies with the Declaration of Helsinki.
Provenance and peer review Not commissioned; externally peer reviewed.