Alcohol use was implicated as a cause in >2.5 million

Transcription

1 Coronary Heart Disease Patterns of Alcohol Consumption and Myocardial Infarction Risk Observations From 52 Countries in the INTERHEART Case Control Study Darryl P. Leong, MBBS, MPH, PhD; Andrew Smyth, MB, BCh; Koon K. Teo, MB, BCh, PhD; Martin McKee, MD, DSc; Sumathy Rangarajan, MSc; Prem Pais, MD; Lisheng Liu, MD; Sonia S. Anand, MD, PhD; Salim Yusuf, MBBS, DPhil; on behalf of the INTERHEART Investigators* Background Although moderate alcohol use is associated with protection against myocardial infarction (MI), it is not known whether this effect is generalizable to populations worldwide. It is also uncertain whether differences in the pattern of alcohol use (and in particular heavy episodic consumption) between different regions negate any beneficial effect. Methods and Results We included cases of first MI and age- and sex-matched controls from 52 countries. Current alcohol use was associated with a reduced risk of MI (compared with nonusers: adjusted odds ratio, 0.87; 95% confidence interval, ; P=0.001); however, the strength of this association was not uniform across different regions (region-alcohol interaction P<0.001). Heavy episodic drinking ( 6 drinks) within the preceding 24 hours was associated with an increased risk of MI (odds ratio, 1.4; 95% confidence interval, ; P=0.01). This risk was particularly elevated in older individuals (for age >65 years: odds ratio, 5.3; 95% confidence interval, ; P=0.008). Conclusions In most participants, low levels of alcohol use are associated with a moderate reduction in the risk of MI; however, the strength of this association may not be uniform across different countries. An episode of heavy drinking is associated with an increased risk of acute MI in the subsequent 24 hours, particularly in older individuals. (Circulation. 2014;130: ) Key Words: alcohol drinking epidemiology myocardial infarction risk factors Alcohol use was implicated as a cause in >2.5 million deaths worldwide in 2010, primarily from injuries, cirrhosis, cardiovascular disease, and cancer, and was ranked the fifth most important risk factor contributing to the global burden of disease. 1 The relationship between alcohol consumption and myocardial infarction (MI) is complex. Several studies have reported that low or moderate alcohol use is protective against MI, 2 5 but these studies have been conducted mostly in high-income countries, so it is not known whether such a protective association is generalizable worldwide. In addition, the pattern (ie, frequency and amount) of alcohol use may be an important modifier of the association between alcohol and MI. Regular moderate consumption and episodic heavy drinking have differing physiological effects on lipids and on platelet and endothelial function. 6 There is now considerable evidence that episodic heavy drinking is associated with myocardial damage and is a risk factor for sudden cardiac death. 7 However, the evidence linking episodes of heavy drinking to confirmed MIs (as opposed to the constellation of conditions that include sudden cardiac deaths) is contradictory. One prior case control study reported that consuming >4 alcoholic beverages was associated with a reduced risk of MI in the subsequent 24 hours, 8 whereas a case-crossover study found that any alcohol use (including heavy episodic use) increased risk of MI in the following 12 hours. 9 It is not known whether regular alcohol use protects against MI across different geographic regions, given that the prevalence of episodic heavy drinking varies. Editorial see p 383 Clinical Perspective on p 398 INTERHEART was a case control study of MI undertaken in individuals with a first MI and age- and sexmatched controls from 52 countries in Asia, Europe, the Middle East, Africa, Australia, and North and South America. This study provides an opportunity to assess whether the protective association of alcohol consumption with MI is consistent in different regions of the world, to assess whether the protective association between alcohol consumption and MI differs Continuing medical education (CME) credit is available for this article. Go to to take the quiz. Received November 19, 2013; accepted May 6, From the Population Health Research Institute, McMaster University and Hamilton Health Sciences, Hamilton, ON, Canada (D.P.L., A.S., K.K.T., S.R., S.S.A., S.Y.); London School of Hygiene and Tropical Medicine, London, UK (M.M.); St. John s Medical College, Bangalore, India (P.P.); and Fu Wai Hospital, Beijing, China (L.L.). *A complete list of the INTERHEART investigators can be found in the online-only Data Supplement. The online-only Data Supplement is available with this article at /-/DC1. Correspondence to Darryl P. Leong, MBBS, MPH, PhD, C2-5B DBRI, Hamilton General Hospital, 237 Barton St E, Hamilton, ON, L8L 2X2, Canada. leongd@phri.ca 2014 American Heart Association, Inc. Circulation is available at DOI: /CIRCULATIONAHA

2 Leong et al Alcohol Use Pattern and Myocardial Infarction Risk 391 in people with varying demography, and to determine whether episodic heavy consumption of alcohol may be an immediate trigger for MI. We hypothesized that regular alcohol use is protective but that an acute episode of heavy drinking may increase the risk of MI within the 24 hours after consumption. Methods Participants The INTERHEART study was a case control study of first acute MI undertaken in 262 centers across 52 countries. The details of this study have been published previously. 10,11 In brief, consecutive cases of first MI presenting within 24 hours of symptom onset were eligible. MI was defined by the presence of characteristic symptoms accompanied by ischemic ECG changes (webappendix 1, com/extras/04art8001webappendix1.pdf). For each case, at least 1 sex- and age-matched (±5 years) control with no history of heart disease or exertional chest pain was simultaneously recruited from the hospital or community. Hospital controls (58%) were identified from patients at the same center with illnesses not obviously related to coronary heart disease or its risk factors. Community controls (36%) were visitors or relatives of a patient from a noncardiac ward or an unrelated visitor of another cardiac patient. Another 3% of controls were from the World Health Organization s Monitoring of Trends and Determinants in Cardiovascular Disease (MONICA) study, 12 and the remaining 3% were from an undocumented source. 10 Data on alcohol consumption were not recorded for 54 controls and 266 cases. The following analysis is based on the remaining controls and cases. Procedures Trained staff administered a structured questionnaire and performed a standardized physical examination on participants. Alcohol exposure was characterized by asking all participants whether they consumed any alcoholic beverage within the past 12 months and, if so, how frequently (<1 time per month, <1 time per week, 1 2 times per week, 3 4 times per week, 5 6 times per week, and daily). Alcohol use was defined as the consumption of 1 alcoholic beverage within the previous 12 months. 13 To evaluate the role of alcohol consumption as a trigger for MI, cases were asked how many alcoholic beverages were consumed in the 24 hours before the onset of MI symptoms (referred to as the hazard period) and how many alcoholic beverages were consumed in the period 24 to 48 hours before the onset of MI symptoms (referred to as the control period). Heavy episodic drinking was defined as consumption of 6 alcoholic drinks within 24 hours before MI. Information on age, ethnicity, dietary patterns, 14 physical activity, tobacco use, marital status, education, employment, psychosocial factors, 15 and cardiovascular risk factors was obtained. Height, weight, waist, and hip circumference were measured in a standardized manner. Serum total cholesterol, high-density lipoprotein cholesterol, triglyceride, and apolipoprotein B and A1 concentrations were measured in a core laboratory as previously described 16 ; low-density lipoprotein cholesterol concentration was calculated from these measurements. Smoking was classified as current, former (no smoking within the previous year), or never. Marital status was considered single, married/ common-law partner, separated/divorced, and widowed. Participants highest level of education was categorized as less than grade 9, grades 9 to 12, or university/college/trade school. Employment status was categorized as used, retired, unemployed, and home duties. All data were transferred to the Population Health Research Institute, McMaster University and Hamilton Health Sciences (Hamilton, ON, Canada) for data checking and analysis. The study protocol was approved by relevant regulatory and ethics committees in participating countries and centers. All participants gave informed consent before study involvement. Statistical Analysis Categorical variables are presented as percentage, and continuous variables are presented as mean±sd or median (25th 75th percentile) as appropriate for their distribution. To examine the relationship between alcohol consumption and MI, the following approaches were adopted. First, logistic regression was used to examine exposure to any alcohol use in the previous year, both unadjusted and adjusted for potential confounders. Two adjusted models were generated: Model 1 was adjusted for age (categorized as <45, 45 65, and >65 years), sex, geographic region, Dietary Risk score, 14 exercise, smoking, marital status, employment, education level, depression, stress at work or at home, financial stress, body mass index, and waist-to-hip ratio. Model 2 was adjusted for the same covariates as model 1 plus further adjustment for serum ratio of apolipoprotein B to apolipoprotein A1 ratio; total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, and triglyceride concentrations; and history of hypertension or diabetes mellitus. 17 In an attempt to account for selection bias from the use of hospitalized controls, sensitivity analyses were performed comparing cases first with community controls and second with hospital controls. This analysis, which yielded similar results, is presented in Tables I and II in the online-only Data Supplement. Unadjusted and adjusted analyses were performed in subgroups defined by age, sex, and geographic region. Differential exposure effects between subgroups were considered to be present only if the interaction term between the subgroup and the exposure was significant. Second, logistic regression (unadjusted and adjusted) was repeated with frequency of alcohol use as the exposure. Alcohol frequency was classified as none in the last 12 months, <1 time per week, 1 to 4 times per week, and >4 times per week. As a supplementary analysis, conditional logistic regression was used to evaluate the relationship between MI and alcohol use to account for the paired recruitment of cases and controls within ±5 years of age of each other. The effect of alcohol exposure was adjusted for Dietary Risk score, 14 exercise, smoking, marital status, employment, education level, depression, stress at work or at home, and financial stress. In this analysis, geographic region did not vary between matched participants, so the analysis was stratified by geographic region, and the estimates for each region were meta-analyzed. The pooled random-effects estimate and the Cochran Q and I 2 statistics for heterogeneity of the effect estimate are reported. To assess whether an episode of acute heavy drinking might act as an acute trigger for MI, a case-crossover approach was adopted. 9 Using this technique, we defined a hazard period of alcohol consumption within the 24 hours before MI in cases. The period of 24 to 48 hours before the MI was considered a control period. Conditional logistic regression was used to estimate odds ratios (ORs) for alcohol consumption within the hazard period compared with the control period. For this analysis, only MI cases were used, and each participant served as his or her own control. We also performed the analysis by the following subgroups: age, sex, frequency of alcohol use, and geographic region. Differential exposure effects between subgroups were considered to be present only if the interaction term between the subgroup and the exposure was significant. We sought to characterize alcohol use by comparing drinkers and nondrinkers among controls. Continuous variables were evaluated by ANOVA if normally distributed or the Kruskal-Wallis test otherwise. Categorical variables were assessed by the χ 2 test. This analysis is presented in Table III in the online-only Data Supplement. A 2-sided value of α=0.05 was used for all analyses, which were performed with STATA 12.1 (StataCorp, College Station, TX). Results The characteristics of the cases and controls included in this analysis are presented in Table 1. A slightly higher proportion of controls consumed alcohol in the previous 12 months compared with cases; however, the controls who had consumed alcohol in the previous 12 months used alcohol less frequently than cases (>4 times per week in 19% versus 21% of each respective group). Current Alcohol Use and the Risk of MI Alcohol consumption within the previous year was associated with a significantly lower risk of MI (Table 2). The unadjusted

3 392 Circulation July 29, 2014 Table 1. Characteristics of Cases and Controls Table 1. Continued Characteristic Controls (n=14 583) Cases (n=12 195) Age, y 57±12 58±12 Male sex, n (%) (74) 9277 (76) Geographic region, n (%) Western Europe 763 (5) 655 (5) Eastern and Central Europe 1917 (13) 1693 (14) Middle East 1776 (12) 1605 (13) Africa 783 (5) 560 (4) South Asia 2195 (15) 1670 (14) China and Hong Kong 3055 (21) 3013 (25) Southeast Asia and Japan 1198 (8) 936 (8) Australia and New Zealand 679 (5) 587 (5) South America and Mexico 1878 (13) 1188 (10) North America 339 (3) 288 (2) Consumed alcohol in 6875 (47) 5484 (45) previous year, n (%) Current smoker, n (%) 3860 (26) 5453 (45) No. of cigarettes smoked per day among ever smokers, n (%) < (57) 3347 (43) (43) 4426 (57) Diabetes mellitus, n (%) 1092 (7) 2230 (18) Hypertension, n (%) 3195 (22) 4742 (39) Daily fruit or vegetable (85) 9697 (80) consumption, n (%) Dietary Risk score* 5.3± ±5.4 Undertakes leisure-time 3289 (23) 1845 (15) exercise, n (%) Home or work stress, n (%) None 3859 (27) 3010 (25) Some periods 7553 (53) 5708 (48) Several periods 2293 (16) 2302 (19) Permanent 608 (4) 917 (8) Financial stress, n (%) Little or none 7101 (49) 5398 (44) Moderate 5695 (39) 4981 (41) Severe 1758 (12) 1771 (15) Depressed, n (%) 977 (7) 991 (8) Marital status, n (%) Never 746 (5) 405 (3) Married/common-law partner (82) 9949 (82) Separated/divorced 610 (4) 510 (4) Widowed 1232 (9) 1314 (11) Education, n (%) <Grade (38) 5474 (45) Grade (25) 3160 (26) >Grade (37) 3548 (29) Employment, n (%) Employed 7684 (55) 5800 (50) Retired 4279 (31) 4066 (35) Unemployed 732 (5) 654 (6) (Continued) Characteristic Controls (n=14 583) Cases (n=12 195) Home duties 1213 (9) 1129 (9) Body mass index, kg/m ± ±4.15 Waist-to-hip ratio 0.91± ±0.084 ApoB, mmol/l 0.90 ( ) 0.95 ( ) ApoA1, mmol/l 1.2 ( ) 1.1 ( ) ApoB/ApoA1 ratio 0.75 ( ) 0.86 ( ) Total cholesterol, mmol/l 5.1 ( ) 5.2 ( ) HDL, mmol/l 1.0 ( ) 0.99 ( ) LDL, mmol/l 3.1 ( ) 3.3 ( ) Triglycerides, mmol/l 1.6 ( ) 1.6 ( ) Numbers represent the mean±sd or median (25th 75th percentile) as appropriate. Otherwise, numbers in parentheses refer to the percentage of the column total. Apo indicates apolipoprotein; HDL, high-density lipoprotein; and LDL, low-density lipoprotein. *Refer to the work by Iqbal et al 14 for score derivation. OR was 0.92 (95% confidence interval [CI], ; P<0.001), and the adjusted ORs were 0.81 (95% CI, ; P<0.001) and 0.87 (95% CI, ; P=0.001) for models 1 and 2, respectively. When smoking was quantified as number of cigarettes per day (as well as never smokers or ever smokers), the respective ORs were 0.85 (95% CI, ; P<0.001) and 0.91 (95% CI, ; P=0.03) for models 1 and 2. When smoking was quantified as number of cigarettes per day (as well as never smokers or current smokers, ie, with former smokers censored), the respective ORs were 0.83 (95% CI, ; P<0.001) and 0.90 (95% CI, ; P=0.02) for models 1 and 2. The conditional logistic regression yielded a pooled OR of 0.84 (95% CI, ; P=0.04; see the online-only Data Supplement). There were significant differences in the association of alcohol intake with MI risk by sex, age, and geographic Table 2. Pattern of Alcohol Use Patterns of Alcohol Use and OR for MI OR (95% CI) for MI P Value Any within 12 mo Unadjusted 0.92 ( ) <0.001 Adjusted (model 1) 0.81 ( ) <0.001 Adjusted (model 2) 0.87 ( ) Pooled* 0.84 ( ) 0.04 Any alcohol and risk of MI in the subsequent 24 h 1.0 ( ) Drinks and risk of MI in the subsequent 24 h 1.4 ( ) 0.01 Model 1 was adjusted for age (categorized as <45, 45 65, and >65 years), sex, geographic region, Dietary Risk score, 14 exercise, smoking, marital status, employment, education level, depression, stress at work or at home, financial stress, body mass index, and waist-to-hip ratio. Model 2 was adjusted as for model 1 and for serum ratio of apolipoprotein B to apolipoprotein; total cholesterol, high-density lipoprotein, low-density lipoprotein, and triglyceride concentrations; and history of hypertension or diabetes mellitus.ci indicates confidence interval; MI, myocardial infarction; and OR, odds ratio. *Pooled effect estimates from conditional logistic regression were stratified by geographic region and adjusted for Dietary Risk score, 14 exercise, smoking, marital status, employment, education level, depression, stress at work or at home, financial stress, body mass index, and waist-to-hip ratio.

4 Leong et al Alcohol Use Pattern and Myocardial Infarction Risk 393 region (Figure 1). This analysis suggested that the protective association of alcohol against MI was greater in women and in individuals 45 years of age. This subgroup analysis also demonstrated differences in the risk of MI with alcohol use between different geographic regions. Geographic heterogeneity was confirmed in the meta-analysis of effect estimates from the conditional logistic regression (Cochran Q=15.9; P=0.007; I 2 =69%; Figure I in the online-only Data Supplement). Given the unexpected finding of an increased risk of MI among South Asians in model 2, we performed a post hoc exploratory analysis of alcohol use and MI risk by country within the South Asian region. The estimates for this analysis were consistent (although CIs were wide) with the exception of Nepal, which contributed relatively few cases (n=239) and controls (n=239). The respective ORs for Sri Lanka, Pakistan, Nepal, India, and Bangladesh were 1.4 (95% CI, ), 1.2 (95% CI, ), 0.85 (95% CI, ), 1.3 (95% CI, ), and 1.0 (95% CI, ). We also performed an exploratory analysis of participants of South Asian ethnicity who were living outside the South Asian countries. There were 805 such individuals (436 cases and 369 controls), among whom the adjusted ORs of MI with alcohol use were 0.66 (95% CI, ; P=0.03) for model 1 and 0.80 (95% CI, ; P=0.3) for model 2. In contrast, among participants of South Asian ethnicity living in South Asia, the adjusted ORs of MI with alcohol use were 1.1 (95% CI, ; P=0.2) for model 1 and 1.4 (95% CI, ; P=0.007) for model 2. Another post hoc subgroup analysis was performed in which alcohol use was stratified by ethnicity. This analysis yielded findings similar to the analysis when alcohol use was stratified by geographic region (Table IV in the online-only Data Supplement). Figure 1. Subgroup analysis for the relationship between alcohol use in the previous year and myocardial infarction (MI). A, Model 1 adjusted for age (categorized as <45, 45 65, and >65 years), sex, geographic region, Dietary Risk score, 14 exercise, smoking, marital status, employment, education level, depression, stress at work or at home, financial stress, body mass index, and waist-to-hip ratio. B, Model 2 adjusted as for model 1 and for serum ratio of apolipoprotein B to apolipoprotein A1; total cholesterol, high-density lipoprotein, low-density lipoprotein, and triglyceride concentrations; and history of hypertension or diabetes mellitus. P values refer to the group-alcohol interaction. Aus indicates Australia; CI, confidence interval; Nth, North; NZ, New Zealand; and SE, Southeast.

5 394 Circulation July 29, 2014 We found that the protective association between alcohol use and MI was no longer significant when alcohol was consumed >4 times per week (Figure 2). This relationship was evident in the unadjusted analysis and persisted after adjustment (Figure 2). Alcohol as an Immediate Trigger of MI Data on alcohol consumption in the 48 hours before MI were available for a total of MI cases. Of these, 9177 (78.8%) consumed no alcohol during either the hazard or the control period, 527 (4.5%) consumed alcohol during the hazard period but not the control period, 514 (4.4%) consumed alcohol during the control period but not the hazard period, and 1434 (12.3%) consumed alcohol during both periods. There was no excess risk of MI if any alcohol was consumed within the hazard period (OR, 1.0; 95% CI, ; P=0.7). In contrast to the neutral effect of any alcohol consumption during the hazard period, heavy drinking ( 6 drinks) during the hazard period was associated with a significant elevation in the risk of MI (OR, 1.4; 95% CI, ; P=0.01). The harmful association of heavy episodic drinking was also evident when heavy drinking was defined using an alternative definition of 5 drinks for men or 4 drinks for women within a 24-hour period 18 ; in this case, the OR was 1.4 (95% CI, ; P=0.001). This harmful effect was notable particularly among older individuals compared with younger individuals (respective ORs for those <45, 45 65, and >65 years of age were 0.84 [95% CI, ; P=0.5], 1.6 [95% CI, ; P=0.01], and 5.3 [95% CI, ; P=0.008]; age episodic heavy drinking interaction P<0.001; Figure 3A). We also observed that the harmful association for triggering MI was more pronounced for those who drank more frequently (respective ORs for those who drank <1, 1 4, and >4 times a week were 0.75 [95% CI, ; P=0.4], 1.3 [95% CI, ; P=0.2], and 1.9 [95% CI, ; P=0.003; interaction P=0.02; Figure 3B). There was a modest relationship between geographic region, heavy episodic drinking, and a triggered MI (interaction P=0.049; Figure 3C). Discussion This study has examined the relationship between alcohol consumption and the long- and short-term risk of MI in all inhabited continents of the world. The most important findings are that moderate (but not high levels of) alcohol use is associated with a lower risk of MI in most populations, that this protective effect was not uniform across geographic regions, and that episodic heavy drinking precipitates acute MI, especially among older individuals. Long-Term Pattern of Alcohol Use and Risk of MI This study confirms the findings from previous research indicating that alcohol consumption in general is associated with a Figure 2. Adjusted odds ratios with 95% confidence intervals (CIs) for myocardial infarction (MI) in participants with different frequencies of alcohol use. A, Model 1 adjusted for age (categorized as <45, 45 65, and >65 years), sex, geographic region, Dietary Risk score, 14 exercise, smoking, marital status, employment, education level, depression, stress at work or at home, financial stress, body mass index, and waist-to-hip ratio. B, Model 2 adjusted as for model 1 and for serum ratio of apolipoprotein B to apolipoprotein A1; total cholesterol, high-density lipoprotein, low-density lipoprotein, and triglyceride concentrations; and history of hypertension or diabetes mellitus. The reference group is nondrinkers, defined as participants who did not consume alcohol in the previous year.

6 Leong et al Alcohol Use Pattern and Myocardial Infarction Risk 395 Figure 3. Subgroup analyses for the association between episodic heavy drinking and acute myocardial infarction (MI). A, Age; (B) frequency of usual drinking; and (C) region. P values refer to the group-alcohol interaction. Aus indicates Australia; CI, confidence interval; NZ, New Zealand; and SE, Southeast. reduced risk of MI. A systematic review and meta-analysis of studies mostly performed in high-income countries concluded that alcohol consumption was associated with a reduced incidence of coronary heart disease compared with nondrinking. 5 In this meta-analysis and in other studies, 5,19 22 a U-shaped relationship was seen between quantity or frequency of alcohol consumed per day and the risk of MI. In INTERHEART, we observed a similar pattern of association in countries that were not previously studied. Importantly, the protective effect of alcohol use against MI did not appear to be uniform across different geographic regions. Although moderate or low alcohol consumption was protective in most countries, such an effect was not seen in South Asia (India, Bangladesh, Nepal, Pakistan, and Sri Lanka). This new observation forces us to temper the generalization that alcohol use (even in moderation) is protective against MI (although it is often forgotten that even in those populations where it has been shown to be protective against MI, it has this effect only in those at significant risk such as those at older ages). 4 Furthermore, alcohol may have other harmful effects to mitigate any protective effect against MI. Evidence of the role of genetic polymorphisms in the enzymes involved in alcohol metabolism in ethnic variations in cardiovascular disease is evolving rapidly. 23 However, the situation is complex. A recent meta-analysis found that Chinese and Korean carriers of the ALDH2*504lys allele had an almost 30% increased risk of cardiovascular disease but that Japanese carriers seemed to be at no increased risk. 24 Another

7 396 Circulation July 29, 2014 meta-analysis has found an increased risk of similar magnitude with carriage of the ALDH2 rs671 allele among East Asian populations. 25 Evidence for a gene-environment interaction comes from a study in a Chinese Han population, in which those carrying the ALDH2 rs671 allele were at greater risk of hypertension and lipid disturbances but only if they drank alcohol. 26 However, to the best of our knowledge, this issue has not been examined in any detail in South Asian populations. Interestingly, our exploratory analysis of South Asians living outside South Asia, although including too few participants to be conclusive, suggested that these individuals may have a reduced risk of MI with alcohol use, in contrast with South Asians living in South Asia. This finding both raises the hypothesis that contextual factors such as the type of alcohol consumed and the pattern of alcohol use might be more important modifiers of the relationship between alcohol and MI than genetic factors and highlights the need for further studies of the effects of genetic variants related to alcohol metabolism in different settings. Alcohol Intake as an Immediate Trigger for MI This study has demonstrated little immediate hazard associated with low or moderate alcohol intake but found an increased risk of MI in the 24 hours after episodic heavy drinking. There is a paucity of existing evidence on the role of alcohol as a trigger for MI. Two smaller case control studies of a total of cases and 6801 controls found a protective association between alcohol use and the risk of MI in the subsequent 24 hours. 8,27 In contrast, in a case-crossover study of 250 Swiss patients with MI, Gerlich et al 9 reported that alcohol use within 12 hours increased the risk of MI. Thus, there remains uncertainty about the risk of MI in the period immediately after alcohol intake. Close comparison of the present study with the previous research is limited by differences in methodologies, but the present study is the largest to investigate the role of alcohol as a trigger specifically for MI, as opposed to a combination of MI and other forms of sudden cardiac death. Our findings suggest that episodic heavy drinking may be an important precipitant of MI. The potential biological mechanisms by which alcohol may predispose to MI acutely after ingestion have previously been reviewed. 6 When alcohol is consumed heavily in an episodic pattern, the withdrawal of the inhibitory effects of alcohol on platelet aggregation may create a rebound prothrombotic state. 28 Episodic heavy drinking may also be associated with adverse changes in lipid profiles such as an elevation in lowdensity lipoprotein cholesterol without the protective increase in high-density lipoprotein cholesterol observed with lower and more regular alcohol intake. 6 In addition, although alcohol ingestion lowers blood pressure within the first 4 hours after consumption, a significant elevation in blood pressure is observed at 20 to 24 hours, which may be mediated by vasoactive cytochrome P450 eicosanoids, the concentration of which is affected immediately by alcohol consumption. 29 Strengths and Limitations The major strength of the present study is that it is a large study performed across many countries, geographic regions, and ethnicities. A larger number of possible confounding covariates were included in the study than in previously reported work, enabling us to minimize their effects. These features provide broad generalizability and robustness of the study findings. The most important limitation is that this study is confined to those who survived to reach hospital. The proportion of patients who die of an acute cardiac event before admission varies considerably among countries. This reflects differing rates of sudden cardiac death and different access to care (including prehospital defibrillation). Although it has long been recognized that some sudden cardiac deaths are due to arrhythmias not linked to coronary occlusion, it is now apparent that they may make an important contribution to the burden of cardiovascular disease in certain settings characterized by high levels of hazardous drinking. 30 In these populations, hazardous consumption (measured in a variety of ways, including consumption of nonbeverage products and innovative biomarkers of alcohol consumption) is associated with protective increases in high-density lipoprotein cholesterol and apolipoprotein A1 but increases in brain natriuretic peptide, a marker of myocardial strain, that are only minimally attenuated by adjustment for hypertension. 31 Consequently, these findings can be applied only to those who have a confirmed MI, not to patients with the entire spectrum of cardiovascular events, which will include a proportion of arrhythmia-induced sudden deaths in individuals without coronary atherosclerosis. As with most observational studies, causality cannot be firmly established. The associations observed between alcohol use and MI may be accounted for by unmeasured confounders such as genetic differences between populations, variation in alcohol type or preparation, and heterogeneity of social context. The present analysis should therefore prompt further research to clarify the nature of the association between alcohol use and MI. Another limitation is that the quantity and type of alcohol consumed by participants were not recorded. In addition, it is uncertain what proportion of nondrinkers in this study were former drinkers. As noted above, some of those who stopped drinking may have done so for health reasons, which may lead to underestimation of any of the effects associated with alcohol. In addition, as with any survey of alcohol use, there is likely to be a social desirability bias; respondents in countries where alcohol use is taboo may be more likely to underestimate their consumption compared with countries where drinking is socially acceptable. Recall bias, in which there is differential recollection of alcohol use depending on case or control status, may also have influenced our findings. Selection bias may have resulted from the approach used to identify controls because these participants represent a population with contact with a hospital. Finally, despite the large sample size in INTERHEART, this study had limited power to quantify alcohol effects within some of the countries, except for China, which had >5818 participants. Findings in particular geographic regions such as South Asia, which included Bangladesh, India, Sri Lanka, Pakistan, and Nepal, require further exploration within the respective countries. Conclusions In most countries, low levels of alcohol use appear to be associated with a lower risk of MI, but such a protective effect was

8 Leong et al Alcohol Use Pattern and Myocardial Infarction Risk 397 not observed in South Asian countries. Episodic heavy drinking can precipitate acute MI, particularly in older individuals. Sources of Funding Dr Leong is supported by the Michael G. DeGroote Fellowship Award, McMaster University. Dr Anand is a recipient of the Heart and Stroke Michael G. DeGroote Chair in Population Health Research and a Canada Research Chair in Ethnicity and Cardiovascular Disease. Dr Yusuf is funded by the Marion Burke Chair of the Heart and Stroke foundation of Canada. The INTERHEART study is funded by the Canadian Institutes of Health Research; by the Heart and Stroke Foundation of Ontario; by the International Clinical Epidemiology Network; through unrestricted grants from AstraZeneca, Novartis, Hoechst Marion Roussel (now Aventis), Knoll Pharmaceuticals (now Abbott), Bristol Myers Squibb, King Pharma, and Sanofi-Sythelabo; and through various national bodies from different participating countries. None. Disclosures References 1. Lim SS, Vos T, Flaxman AD, Danaei G, Shibuya K, Adair-Rohani H, Amann M, Anderson HR, Andrews KG, Aryee M, Atkinson C, Bacchus LJ, Bahalim AN, Balakrishnan K, Balmes J, Barker-Collo S, Baxter A, Bell ML, Blore JD, Blyth F, Bonner C, Borges G, Bourne R, Boussinesq M, Brauer M, Brooks P, Bruce NG, Brunekreef B, Bryan-Hancock C, Bucello C, Buchbinder R, Bull F, Burnett RT, Byers TE, Calabria B, Carapetis J, Carnahan E, Chafe Z, Charlson F, Chen H, Chen JS, Cheng AT, Child JC, Cohen A, Colson KE, Cowie BC, Darby S, Darling S, Davis A, Degenhardt L, Dentener F, Des Jarlais DC, Devries K, Dherani M, Ding EL, Dorsey ER, Driscoll T, Edmond K, Ali SE, Engell RE, Erwin PJ, Fahimi S, Falder G, Farzadfar F, Ferrari A, Finucane MM, Flaxman S, Fowkes FG, Freedman G, Freeman MK, Gakidou E, Ghosh S, Giovannucci E, Gmel G, Graham K, Grainger R, Grant B, Gunnell D, Gutierrez HR, Hall W, Hoek HW, Hogan A, Hosgood HD 3rd, Hoy D, Hu H, Hubbell BJ, Hutchings SJ, Ibeanusi SE, Jacklyn GL, Jasrasaria R, Jonas JB, Kan H, Kanis JA, Kassebaum N, Kawakami N, Khang YH, Khatibzadeh S, Khoo JP, Kok C, Laden F, Lalloo R, Lan Q, Lathlean T, Leasher JL, Leigh J, Li Y, Lin JK, Lipshultz SE, London S, Lozano R, Lu Y, Mak J, Malekzadeh R, Mallinger L, Marcenes W, March L, Marks R, Martin R, McGale P, McGrath J, Mehta S, Mensah GA, Merriman TR, Micha R, Michaud C, Mishra V, Mohd Hanafiah K, Mokdad AA, Morawska L, Mozaffarian D, Murphy T, Naghavi M, Neal B, Nelson PK, Nolla JM, Norman R, Olives C, Omer SB, Orchard J, Osborne R, Ostro B, Page A, Pandey KD, Parry CD, Passmore E, Patra J, Pearce N, Pelizzari PM, Petzold M, Phillips MR, Pope D, Pope CA 3rd, Powles J, Rao M, Razavi H, Rehfuess EA, Rehm JT, Ritz B, Rivara FP, Roberts T, Robinson C, Rodriguez-Portales JA, Romieu I, Room R, Rosenfeld LC, Roy A, Rushton L, Salomon JA, Sampson U, Sanchez-Riera L, Sanman E, Sapkota A, Seedat S, Shi P, Shield K, Shivakoti R, Singh GM, Sleet DA, Smith E, Smith KR, Stapelberg NJ, Steenland K, Stöckl H, Stovner LJ, Straif K, Straney L, Thurston GD, Tran JH, Van Dingenen R, van Donkelaar A, Veerman JL, Vijayakumar L, Weintraub R, Weissman MM, White RA, Whiteford H, Wiersma ST, Wilkinson JD, Williams HC, Williams W, Wilson N, Woolf AD, Yip P, Zielinski JM, Lopez AD, Murray CJ, Ezzati M, AlMazroa MA, Memish ZA. A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, : a systematic analysis for the Global Burden of Disease Study Lancet. 2012;380: Mukamal KJ, Chiuve SE, Rimm EB. Alcohol consumption and risk for coronary heart disease in men with healthy lifestyles. Arch Intern Med. 2006;166: Mukamal KJ, Conigrave KM, Mittleman MA, Camargo CA Jr, Stampfer MJ, Willett WC, Rimm EB. Roles of drinking pattern and type of alcohol consumed in coronary heart disease in men. N Engl J Med. 2003;348: Romelsjö A, Allebeck P, Andréasson S, Leifman A. Alcohol, mortality and cardiovascular events in a 35 year follow-up of a nationwide representative cohort of 50,000 Swedish conscripts up to age 55. Alcohol Alcohol. 2012;47: Ronksley PE, Brien SE, Turner BJ, Mukamal KJ, Ghali WA. Association of alcohol consumption with selected cardiovascular disease outcomes: a systematic review and meta-analysis. BMJ. 2011;342:d McKee M, Britton A. The positive relationship between alcohol and heart disease in eastern Europe: potential physiological mechanisms. J R Soc Med. 1998;91: Leon DA, Saburova L, Tomkins S, Andreev E, Kiryanov N, McKee M, Shkolnikov VM. Hazardous alcohol drinking and premature mortality in Russia: a population based case-control study. Lancet. 2007;369: Jackson R, Scragg R, Beaglehole R. Does recent alcohol consumption reduce the risk of acute myocardial infarction and coronary death in regular drinkers? Am J Epidemiol. 1992;136: Gerlich MG, Krämer A, Gmel G, Maggiorini M, Lüscher TF, Rickli H, Kleger GR, Rehm J. Patterns of alcohol consumption and acute myocardial infarction: a case-crossover analysis. Eur Addict Res. 2009;15: Yusuf S, Hawken S, Ounpuu S, Dans T, Avezum A, Lanas F, McQueen M, Budaj A, Pais P, Varigos J, Lisheng L; INTERHEART Study Investigators. Effect of potentially modifiable risk factors associated with myocardial infarction in 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364: Ounpuu S, Negassa A, Yusuf S. INTER-HEART: a global study of risk factors for acute myocardial infarction. Am Heart J. 2001;141: Wilhelmsen L, Rosengren A, Johansson S, Lappas G. Coronary heart disease attack rate, incidence and mortality in Göteborg, Sweden. Eur Heart J. 1997;18: Alcohol use and alcohol use disorders in the United States, a 3-year follow-up, Sep U.S. Alcohol Epidemiologic Data Reference Manual. Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health. 2010:Vol 8;No Iqbal R, Anand S, Ounpuu S, Islam S, Zhang X, Rangarajan S, Chifamba J, Al-Hinai A, Keltai M, Yusuf S; INTERHEART Study Investigators. Dietary patterns and the risk of acute myocardial infarction in 52 countries: results of the INTERHEART study. Circulation. 2008;118: Rosengren A, Hawken S, Ounpuu S, Sliwa K, Zubaid M, Almahmeed WA, Blackett KN, Sitthi-amorn C, Sato H, Yusuf S; INTERHEART Investigators. Association of psychosocial risk factors with risk of acute myocardial infarction in cases and controls from 52 countries (the INTERHEART study): case-control study. Lancet. 2004;364: McQueen MJ, Hawken S, Wang X, Ounpuu S, Sniderman A, Probstfield J, Steyn K, Sanderson JE, Hasani M, Volkova E, Kazmi K, Yusuf S; INTERHEART Study Investigators. Lipids, lipoproteins, and apolipoproteins as risk markers of myocardial infarction in 52 countries (the INTERHEART study): a case-control study. Lancet. 2008; 372: Brien SE, Ronksley PE, Turner BJ, Mukamal KJ, Ghali WA. Effect of alcohol consumption on biological markers associated with risk of coronary heart disease: systematic review and meta-analysis of interventional studies. BMJ. 2011;342:d Wechsler H, Nelson TF. Binge drinking and the American college student: what s five drinks? Psychol Addict Behav. 2001;15: Schröder H, Masabeu A, Marti MJ, Cols M, Lisbona JM, Romagosa C, Carión T, Vilert E, Marrugat J; REGICOR Investigators. Myocardial infarction and alcohol consumption: a population-based case-control study. Nutr Metab Cardiovasc Dis. 2007;17: Britton A, Marmot M. Different measures of alcohol consumption and risk of coronary heart disease and all-cause mortality: 11-year follow-up of the Whitehall II Cohort Study. Addiction. 2004;99: Keil U, Chambless LE, Döring A, Filipiak B, Stieber J. The relation of alcohol intake to coronary heart disease and all-cause mortality in a beerdrinking population. Epidemiology. 1997;8: Mukamal KJ, Jensen MK, Grønbaek M, Stampfer MJ, Manson JE, Pischon T, Rimm EB. Drinking frequency, mediating biomarkers, and risk of myocardial infarction in women and men. Circulation. 2005;112: Han H, Wang H, Yin Z, Jiang H, Fang M, Han J. Association of genetic polymorphisms in ADH and ALDH2 with risk of coronary artery disease and myocardial infarction: a meta-analysis. Gene. 2013;526: Gu JY, Li LW. ALDH2 Glu504Lys polymorphism and susceptibility to coronary artery disease and myocardial infarction in East Asians: a metaanalysis. Arch Med Res. 2014;45: Wang Q, Zhou S, Wang L, Lei M, Wang Y, Miao C, Jin Y. ALDH2 rs671 Polymorphism and coronary heart disease risk among Asian populations: a meta-analysis and meta-regression. DNA Cell Biol. 2013; 32:

9 398 Circulation July 29, Wang Y, Zhang Y, Zhang J, Tang X, Qian Y, Gao P, Zhu D. Association of a functional single-nucleotide polymorphism in the ALDH2 gene with essential hypertension depends on drinking behavior in a Chinese Han population. J Hum Hypertens. 2013;27: McElduff P, Dobson AJ. How much alcohol and how often? Population based case-control study of alcohol consumption and risk of a major coronary event. BMJ. 1997;314: Hillbom M, Kangasaho M, Löwbeer C, Kaste M, Muuronen A, Numminen H. Effects of ethanol on platelet function. Alcohol. 1985;2: Barden AE, Croft KD, Beilin LJ, Phillips M, Ledowski T, Puddey IB. Acute effects of red wine on cytochrome P450 eicosanoids and blood pressure in men. J Hypertens. 2013;31: ; discussion Tomkins S, Collier T, Oralov A, Saburova L, McKee M, Shkolnikov V, Kiryanov N, Leon DA. Hazardous alcohol consumption is a major factor in male premature mortality in a typical Russian city: prospective cohort study PLoS One. 2012;7:e Leon DA, Shkolnikov VM, Borinskaya S, Casas JP, Evans A, Gil A, Kee F, Kiryanov N, McKee M, O Doherty MG, Ploubidis GB, Polikina O, Vassiliev M, Blankenberg S, Watkins H. Hazardous alcohol consumption is associated with increased levels of B-type natriuretic peptide: evidence from two population-based studies. Eur J Epidemiol. 2013;28: Clinical Perspective Previous evidence suggests that alcohol use is associated with a reduced risk of coronary heart disease. Although the present study also found that any alcohol consumption within the previous 12 months was associated with a reduced risk of myocardial infarction (MI) in most geographic regions, this reduced risk was not seen in all regions studied. Furthermore, an episode of heavy alcohol use ( 6 drinks) may act as a trigger for acute MI, particularly in individuals >65 years of age. Therefore, the key message from this study is that alcohol use does not have a uniformly protective association against MI. In fact, alcohol consumption may have a harmful association with MI in particular social contexts and when consumed heavily. The implications for clinicians are that caution should be exercised before promoting even low to moderate alcohol use as a strategy to reduce MI risk and that heavy alcohol consumption should be discouraged. Go to to take the CME quiz for this article.