Author + information
- James Palmer, BMedSci, MBChBa,
- Amelia Lloyd, BMedSci, MBChBa,
- Lloyd Steele, BMedSci, MBChBa,
- James Fotheringham, PhDb@DrFothers,
- Dawn Teare, PhDb,
- Javaid Iqbal, PhDc and
- Ever D. Grech, MDc,∗ (, )@edgrech
- aSheffield Medical School, The University of Sheffield, Sheffield, United KingdomSheffield Medical School, The University of Sheffield, Sheffield, United Kingdom
- bSchool of Health and Related Research (ScHARR), The University of Sheffield, Sheffield, United KingdomSchool of Health and Related Research (ScHARR), The University of Sheffield, Sheffield, United Kingdom
- cSouth Yorkshire Cardiothoracic Centre, Northern General Hospital, Sheffield, United KingdomSouth Yorkshire Cardiothoracic Centre, Northern General Hospital, Sheffield, United Kingdom
- ↵∗Address for correspondence:
Dr. Ever D. Grech, South Yorkshire Cardiothoracic Centre, Northern General Hospital, Herries Road, Sheffield S5 7AU, United Kingdom.
Background Smoking is a well-documented risk for acute ST-segment elevation myocardial infarction (STEMI). The differential effect between sexes has yet to be quantified.
Objectives The purpose of this study was to differentiate the effect of smoking on increased risk of STEMI between sexes.
Methods For this retrospective ecological cohort study, all patients at a U.K. tertiary cardiothoracic center who presented between 2009 and 2014 with acute STEMI were combined with population data to generate incidence rates of STEMI. Age-standardized incidence rate ratios (IRRs) using the Poisson distribution were calculated comparing STEMI rates between smokers and nonsmokers stratified by sex and 3 age groups (18 to 49, 50 to 64, and >65 years).
Results A total of 3,343 patients presented over 5,639,328 person-years. Peak STEMI rate for current smokers was in the 70 to 79 years age range for women (235 per 100,000 patient-years) and 50 to 59 years (425 per 100,000 patient-years) in men. Smoking was associated with a significantly greater increase in STEMI rate for women than men (IRR: 6.62; 95% confidence interval [CI]: 5.98 to 7.31, vs. 4.40; 95% CI: 4.15 to 4.67). The greatest increased risk was in women age 18 to 49 (IRR: 13.22; 95% CI: 10.33 to 16.66, vs. 8.60; 95% CI: 7.70 to 9.59 in men). The greatest risk difference was in the age 50 to 64 years group, with IRR of 9.66 (95% CI: 8.30 to 11.18) in women and 4.47 (95% CI: 4.10 to 4.86) in men.
Conclusions This study quantifies the differential effect of smoking between sexes, with women having a significantly increased risk of STEMI than men. This information encourages continued efforts to prevent smoking uptake and promote cessation.
Cardiovascular disease remains the leading cause of mortality worldwide, with an increasing prevalence. In 2000, this accounted for 28% of deaths, increasing to 31% in 2015 (1,2). Heart disease also accounts for over one-quarter of all deaths in the United States (3). Acute ST-segment elevation myocardial infarction (STEMI) is amongst the most life-threatening manifestations of cardiovascular disease, affecting all age groups, resulting in death within 30 days in 5.7-11.0% of cases (4–6).
In 2015, there were an estimated 933.1 million smokers worldwide, with 82.3% of these being men (7). In contrast, there is a much narrower gender divide in the United Kingdom’s 7.6 million smokers (8). In 2016, 17.7% of the male population and 14.1% of the female population were smoking, the lowest recorded values since 2010 (8). The comparative similarity of smoking prevalence between sexes in the United Kingdom makes the effects of smoking easier to compare. Studies assessing sex-differentiated risk of ischemic heart disease attributable to smoking have yielded varying results. A Danish study found that smoking was a greater risk factor for all acute coronary syndromes in women compared with men (9). A study of a large Swedish registry has identified smoking as a more significant risk for STEMI for women than men age <65 years (10). Older studies have either suggested an increased risk of myocardial infarction for women compared with men (11,12), no difference in risk (Framingham) (13), or less risk than men (14).
Although the magnitude of risk of acute STEMI attributable to smoking has been studied (15), none have quantified and compared the incidence of STEMI associated with smoking between sexes and within different age groups. This study aims to assess smoking as an independent risk factor for STEMI, and determine differences in risk between age and sex groups.
In this retrospective ecological cohort study, data were compiled for all patients presenting with acute STEMI managed by primary percutaneous coronary intervention in the South Yorkshire region over a 5-year period, between January 4, 2009, and July 31, 2014.
Using a departmental database mandated by national audit processes, patients were identified and their individual patient case notes were examined. Data collected included patient age, sex, smoking status, other key cardiovascular risk factors, and cardioprotective drugs taken prior to STEMI onset. Culprit artery of STEMI was also recorded. Ex-smokers were described as being abstinent for a minimum of 28 days prior to STEMI, although duration of smoking cessation was omitted from patient case notes in 38% of ex-smokers.
Differences in the normally distributed continuous variables of ages within sexes were compared using independent samples Student's t-tests. Differences between sexes in categorical variables, other STEMI risk factors, and pre-STEMI cardiac medications and the percentage of cases attributed to culprit arteries were compared using chi-square tests.
STEMI incidence rates by age and sex were calculated, stratified by smoking status. These were derived by comparing the raw STEMI numbers to the entire population served by the South Yorkshire Cardiothoracic Centre, with data obtained from the Integrated Household Survey from the United Kingdom Office for National Statistics (ONS-IHS).
This second dataset was derived from responses from South Yorkshire residents age ≥18 years participating in the ONS-IHS between April 2009 and March 2012. The ONS-IHS asks randomly sampled respondents a range of topics using telephone or face-to-face interviews, and is the biggest pool of U.K. social data after the census. It does not contain medical diagnoses due to the self-reported nature of the questionnaire. Respondents were asked both if they had ever smoked and if they currently smoked. Regional and/or national estimates of population responses are generated using a multistage population weighting procedure that accounts for probability of selection and adjusts for non-response, and have reported similar estimates to other social surveys. The weighted responses from the local authorities served by the cardiac center were used to estimate population age and smoking status strata:
95% confidence intervals (CIs) were obtained from the Poisson distribution.
Incidence rate ratios (IRRs) and their CIs, indirectly standardized to account for age differences between the sexes were calculated, and were directly compared to assess differences in the impact of smoking on STEMI risk between sexes. IRRs were calculated by age group by the formula:
Relative risk of smoking-associated STEMI was then compared between the sexes across the age groups by dividing the IRRs through each other.
Permission to undertake this study was obtained from the Sheffield Teaching Hospitals Research and Development Department. It used previously collected data from hospital records, which was subsequently anonymized to preserve patient confidentiality. At the study’s conception, approval was not required from the NHS Research Ethics Committee for retrospective studies of this sort.
A total of 3,343 STEMIs were recorded within our 5-year study period. Of these, 27.3% occurred in women whose mean presentation of age was 5.8 years older than men (66.6 years vs. 60.8 years; p = 0.011). The prevalence of most risk factors and all pre-STEMI cardiac medications were similar (Table 1). However, hypertension, diabetes, and history of cerebrovascular accident were significantly more common in women, whereas a previous history of myocardial infarction was more prevalent in men (Table 1). The proportion of STEMI patients who were current smokers were similar between sexes (46.8% of female patients vs. 47.6% of male patients).
The distribution of culprit artery for STEMI was affected by sex, age group, and smoking status (Table 2). STEMI in women was more likely to involve the right coronary artery (RCA), and less likely to involve the circumflex. The RCA was involved less often in the 18- to 49-year-old age group. Current smokers were more likely to have the RCA as their culprit lesion, and less likely to involve the left anterior descending artery.
In smokers, the highest rate of STEMI was in the Q5 50- to 59-year age group, at 286.3 per 100,000 patient-years (95% CI: 262.1 to 312.2 per 100,000 patient years), whereas in nonsmokers, the highest rate was in the 70- to 79-year-old group, at 95.1 per 100,000 patient-years (95% CI: 82.7 to 108.9 per 100,000 patient years). An incidence rate graph displaying the raw STEMI figures versus the general population demonstrated a similar risk between ex- and never-smokers for STEMI incidence, with CIs overlapping between the groups at every 10-year age group (Figure 1). Figures 2 and 3 demonstrate the incidence rates of acute STEMI for different smoking groups at each 10-year age group, differentiated for sex (female and male, respectively), with similarity between nonsmoking groups confirmed. These figures justified the combination of the ex- and never-smoker groups for further comparison of the effect of smoking versus not smoking, by looking at age-standardized IRRs.
Compared with their nonsmoking counterparts, the STEMI risk was 6.62 times higher (95% CI: 5.98 to 7.31) in female and 4.40 (95% CI: 4.15 to 4.67) in male smokers. Female smokers age <50 years ran the highest relative risk of acute STEMI: 13.22 (95% CI: 10.35 to 16.66) times greater than their nonsmoking counterparts. This was significantly higher than men of the same age group, with a smoking associated risk of 8.60 (95% CI: 7.70 to 9.59).
The largest relative risk difference in smoking-associated STEMI was found in the middle-aged group (age 50 to 64 years), with female smokers at 9.66 (95% CI: 8.30 to 11.18) times increased risk of STEMI versus 4.47 (95% CI: 4.10 to 4.86) for men, indicating that smoking is a more severe risk factor for women of this age group compared with men by a factor of 2.16. Figure 4 demonstrates the relative risk of smoking associated STEMI for female smokers, compared to male smokers.
Smoking is an established reversible risk factor for coronary heart disease. Our group has previously identified smoking as the causative agent for STEMI in nearly 50% of all cases (16), and has highlighted that smoking poses the greatest risk in the young (age <50 years) (15). Despite this, there was no significantly increased mortality risk post-STEMI according to smoking status (17). Expanding from these previous studies, the established prevalence of risk factors at baseline has remained very similar.
We have found that smoking increases STEMI risk in all patients, regardless of age or sex (Central Illustration). However, this is the first study to quantify this risk, its differential effect according to sex, and its variation within age groups. Smoking increases STEMI risk in women more than men by a significant degree at all ages, with the largest risk difference present in the middle-aged (50 to 64 years) group. However, the highest risk increase for both sexes was in the youngest patients (age 18 to 49 years). This study emphasizes the differential effect of smoking on STEMI risk between younger and older smokers. This is most striking in young women, in whom smoking increases STEMI risk by >13×, with young male smokers at 8.6× increased risk.
The protective effects of endogenous estrogens have long been known, due to their effects on serum lipid concentrations and on vessel walls, notably vasodilation and inhibiting response to injury, thus preventing the development of atherosclerosis (18). It has been observed that estrogen activity or production is inhibited by cigarette smoke (19). This has been supported by studies finding statistically lower levels of serum estrogen and increased failure rates of in vitro fertilization in smoking women compared with their age-matched counterparts (19,20). Therefore, smoking poses a double risk to premenopausal women, with its causality to atherosclerosis as well as its disturbance on estrogen levels.
Men have also been found to have larger-caliber coronary arteries than women in a large study, regardless of body habitus or left ventricular mass (21). The pathological effect of smoking on STEMI is multifactorial and is attributable to thrombosis, endothelial dysfunction, and inflammation. Whereas atherogenic change and a hypercoagulable state will have similar effects regardless of arterial lumen diameter, chronic inflammation may lead to a greater degree of arterial narrowing in women than men, due to the already reduced-caliber vessel (22–24). This could result in female smokers having further reduced coronary artery diameter than men, compared with nonsmokers. Furthermore, it has been suggested that the etiology of female STEMI may differ from men. Although atherosclerotic change is the main culprit for both sexes, other mechanisms have been recognized as more prevalent in female STEMI patients: vasospasm, vasculitis, fibromuscular dysplasia, spontaneous coronary artery dissection, and plaque erosion (as opposed to plaque rupture) (25). It is highly likely that cigarette smoking perpetuates some of these extra-atherosclerotic events, thus imposing a greater increase in STEMI risk for women. For example, nicotine, a key component of cigarette smoke, induces vasospasm, a mechanism of STEMI known to be more common in women (22,25).
STEMI patients present later and receive less standardized treatment if they are women, resulting in greater mortality (25). This discrepancy in outcomes is mirrored at all stages of risk prevention for coronary heart disease. In a study of 172 physicians, the “attitude study” found that physicians perceived coronary artery disease in men as being more important than in women. Despite identical demographics, laboratory results, and degree of atherosclerotic disease, preventative therapy was prescribed significantly more often in male than female patients (26). Although this study did not evaluate differences in smoking cessation advice between sexes, others have, and we may speculate that smoking is considered less of a cardiac risk for women than for men, and therefore advice may differ between sexes (27,28).
The similarity in STEMI risk for ex- and never-smokers has been identified in this study. This suggests marked reversibility in STEMI risk by cessation, possibly in as little as a few weeks or months. Although chronic atherosclerotic change is unlikely to be affected, it is plausible that risk of acute events, including thrombosis and vasospasm, quickly regress with cessation (29,30). This interesting observation carries a strong public health message to encourage smoking abstinence. Exploring the reversibility of harm from smoking for cardiovascular risk is a potential avenue for future research.
As a cohort study, this ensured all patient groups were equally and fully represented from the South Yorkshire region, with a reduction in the risk of group under-representation. The population data gained from the Office for National Statistics provided accurate denominators to create genuine STEMI rates that could be applied both to the statistics of the study, as well as to clinical advice within the study region.
This study categorized patients as current, ex-, or never-smokers. However, this study did not provide information about volume of smoking or length of smoking cessation in ex-smokers.
The retrospective nature of data collection may have resulted in variable accuracy of detail within some patients’ past medical history from the case notes. However, attempts were also made to complete missing data from other database sources. This has potential to create error if a certain patient group was more at risk of not having paper records still on file, but was an unfortunate necessity of the study.
This study includes only those presenting with STEMI as a candidate for percutaneous coronary intervention. It does not include those who died in the community prior to admission. Smoking may contribute to death before hospital admission, and this is therefore a potential source of bias. This study does not include other subtypes of acute coronary syndrome.
This study is the first to quantify the differential effect of cigarette smoking between sexes on STEMI risk. It has provided strong evidence that smoking incurs a greater STEMI risk to all female patients, compared with male. The differential effect of smoking was most significant in the middle-aged, by 2-fold. However, the highest increased STEMI risk attributable to smoking was in young women (age 18 to 49 years), who had >13 times greater risk than equivalent nonsmokers.
This study also demonstrates that smoking cessation, regardless of age or sex, reduces STEMI risk to that of a never smoker, possibly within a month. Patients who smoke merit encouragement to give up their habit, and this study adds quantitative evidence to the benefits of doing so. The results shown here may also be used to demonstrate the negative effects of smoking to those who may otherwise seek to start, in particular, young, otherwise healthy adults.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: Cigarette smoking increases the risk of acute myocardial infarction to a greater degree in young women than in men. The risk of STEMI in ex-smokers is similar to that in people who never smoked.
TRANSLATIONAL OUTLOOK: Further investigation into the reversibility of the cardiovascular risk associated with cigarette smoking could lead to more effective smoking cessation strategies.
The team acknowledges the U.K. Office for National Statistics for providing Integrated Household Survey and census data for the incidence analysis.
All authors have reported that they have no relationships relevant to the contents of this paper to disclose.
Listen to this manuscript's audio summary by Editor-in-Chief Dr. Valentin Fuster on JACC.org.
- Abbreviations and Acronyms
- confidence interval
- incidence rate ratio
- Office for National Statistics Integrated Household Survey
- right coronary artery
- ST-segment elevation myocardial infarction
- Received March 7, 2019.
- Accepted March 26, 2019.
- McAloon C.J.,
- Boylan L.M.,
- Hamborg T.,
- et al.
- Martínez-García M.,
- Salinas-Ortega M.,
- Estrada-Arriaga I.,
- Hernández-Lemus E.,
- García-Herrera R.,
- Vallejo M.
- Margolis G.,
- Barkagan M.,
- Flint N.,
- Ben-Shoshan J.,
- Keren G.,
- Shacham Y.
- Widimský P.,
- Budesínský T.,
- Vorác D.,
- et al.
- Antoniucci D.,
- Rodriguez A.,
- Hempel A.,
- et al.
- Collaborators G.T.
- ↵ONS SSD. Opinions and Lifestyle Survey, Tobacco Consumption and E-Cigarettes Modules, January - February, April - May, July - August and October - November 2016 [data collection]. 2nd Edition. UK Data Service, 2017.
- Prescott E.,
- Hippe M.,
- Schnohr P.,
- Hein H.O.,
- Vestbo J.
- Björck L.,
- Rosengren A.,
- Wallentin L.,
- Stenestrand U.
- Njølstad I.,
- Arnesen E.,
- Lund-Larsen P.G.
- Janghorbani M.,
- Hedley A.J.,
- Jones R.B.,
- Zhianpour M.,
- Gilmour W.H.
- Seltzer C.C.
- LaCroix A.Z.,
- Lang J.,
- Scherr P.,
- et al.
- Lloyd A.,
- Steele L.,
- Fotheringham J.,
- et al.
- Steele L.,
- Lloyd A.,
- Fotheringham J.,
- Sultan A.,
- Iqbal J.,
- Grech E.D.
- Steele L.,
- Lloyd A.,
- Fotheringham J.,
- Sultan A.,
- Iqbal J.,
- Grech E.D.
- Tweed J.O.,
- Hsia S.H.,
- Lutfy K.,
- Friedman T.C.
- Hiteshi A.K.,
- Li D.,
- Gao Y.,
- et al.
- Benowitz N.L.
- Tamis-Holland J.E.
- Abuful A.,
- Gidron Y.,
- Henkin Y.
- Meijer E.,
- Verbiest M.E.A.,
- Chavannes N.H.,
- et al.
- De Smedt D.,
- De Bacquer D.,
- De Sutter J.,
- et al.
- Rallidis L.S.,
- Pavlakis G.
- Girard D.,
- Delgado-Eckert E.,
- Schaffner E.,
- et al.