Alcohol consumption and the risk of sudden cardiac death in women: An evaluation from the Nurses' Health Study

Abstract

The impact of alcohol consumption on total and cardiovascular mortality is well studied and often discussed in lay publications. Alcohol consumption is associated with decreased all-cause mortality and cardiovascular disease in a J-shaped or U-shaped curve. 1–3 Patients who do not consume alcohol or consume excessive amounts have higher mortality and prevalence of cardiovascular disease compared with those who drink alcohol in moderate amounts. As with many other cardiovascular risk factors and therapeutic interventions, gender differences are noted. Decreased hypertension, coronary heart disease (CHD), stroke, and mortality have been associated with moderate alcohol use in women. 4–6 Studies have also demonstrated a high risk of sudden cardiac death (SCD) and cardiac arrhythmias with heavy drinking but less SCD with moderate alcohol use 7–10 ; these studies had limited inclusion of women. Women also can develop alcohol-related medical problems at lower levels of consumption than men, 11 and although this may reflect lower body weight, it could also represent gender differences in alcohol metabolism or effects on postmenopausal estrogen levels. The study by Chiuve et al 12 in this issue of Heart Rhythm addresses the association of alcohol consumption and SCD in 121,700 women ages 30 to 55 years from the Nurses’ Health Study. In this ambitious, prospective cohort study participants were surveyed every 4 years to document the amount and type of alcohol consumption. Because other medical comorbidities or illness can cause patients to cease alcohol consumption and introduce bias, nondrinkers were separated into lifetime abstainers and former drinkers. After excluding women with a baseline diagnosis of cancer or cardiovascular disease or incomplete information on alcohol intake and dietary data, the study population comprised 85,067 subjects. Studied end points included SCD, other fatal CHD, and nonfatal myocardial infarction. Extensive efforts to increase the accuracy of SCD classification included obtaining death certificates, interviewing family members, and obtaining hospital records. These efforts at accurate classification were appropriate, as proper diagnosis of SCD remains a major challenge in studies. The authors hypothesized that the greatest association of CHD to alcohol consumption would be the most recent intake, and this intake amount was used to estimate risk in the subsequent 4-year period. Results were adjusted for age, diet, cardiovascular disease risk factors, and caloric intake. Women who consumed 5.0 to 14.9 g alcohol daily (approximately ½ to 1 drink/day) accounted for 50% of the study population and had the lowest risk of SCD in a U-shaped association (P .0003) over 26 years of follow-up. There was no significant difference in risk of SCD between abstainers and women who had higher amounts of alcohol intake. No differences in results were observed when probable but not definite SCD cases were excluded. Moreover, there were no differences in light to moderate consumption of different types of alcohol (wine, beer, or liquor) on the observed association with less SCD (RR 0.78 for beer, 0.84 for wine, 0.77 for liquor). Alcohol consumption was also associated with lower risk of nonfatal myocardial infarction or fatal CHD, but this was a linear and inverse relationship. A study of this size evaluating the association of alcohol consumption and SCD in women has not been performed previously. This study has a high number of study participants followed up over an extended time as well as a significant number of SCD and CHD events. Such an analysis generally has several significant challenges. First, correct classification of SCD events can be difficult. SCD is defined as death that occurs within 1 hour or less of symptom onset or during sleep in a person without a condition that would seem fatal, 13 but some methods of SCD identification used in previous studies are not accurate. Every et al 14 demonstrated that SCD classification on death certificates can be a sensitive but not specific method of identifying SCD events. In addition, the general public often confuses SCD and myocardial infarction. Although the authors of the present study made extensive efforts to properly classify SCD, accurate identification of such events remains a weakness in most studies of SCD. Addressing multiple comorbidities in a high number of patients can be quite complicated. Risk of SCD in this study was controlled for over 15 potentially confounding vari