Current issues of ACP Journal Club are published in Annals of Internal Medicine


Fish consumption in men decreased nonsudden death from MI

ACP J Club. 1997 Nov-Dec;127:80. doi:10.7326/ACPJC-1997-127-3-080

Source Citation

Daviglus ML, Stamler J, Orencia AJ, et al. Fish consumption and the 30-year risk of fatal myocardial infarction. N Engl J Med. 1997 Apr 10; 336:1046-53.



To determine whether men who consume fish have a reduced 30-year risk for death from cardiovascular disease (CVD).


Cohort study of men in the Chicago Western Electric Study.


Chicago, United States.


1822 men who were 40 to 55 years of age in 1957, worked in occupations related to the manufacture of telephones, and were free of coronary heart disease (CHD) at baseline. 68% of the men were in blue-collar jobs, 42% were Catholic, 58% smoked (mean 18 cigarettes/d), and 85% consumed alcohol (mean 18.8 mL/d). Follow-up was 99% after 31 years.

Assessment of risk factors

Dietary information was collected at baseline (1957 to 1959) by using an extensive dietary history questionnaire (frequency and quantity of 195 foods) and was used to determine fish consumption in grams per day (none, 1 to 17, 18 to 34, and ≥ 35), macronutrients, micronutrients, total energy, and saturated and monounsaturated fatty acids. Age, education, body mass index, blood pressure, serum cholesterol levels, smoking status, heart rate, history of diabetes, electocardiographic abnormalities, and religion were ascertained at baseline.

Main outcome measures

Death from myocardial infarction (sudden and nonsudden), CHD, and CVD.

Main results

During follow-up, 293 deaths from myocardial infarction (196 sudden deaths), 430 from CHD, 573 from CVD, and 1042 from all causes occurred. Multivariate Cox-proportional hazards analysis using 1.0 as the relative risk (RR) for the lowest quartile of fish consumption (none) showed that only the highest quartile of consumption (≥ 35 mg/d) was associated with a decreased risk for death from other causes. The risk for death from nonsudden myocardial infarction was decreased (RR 0.33, 95% CI 0.12 to 0.91, P for the linear trend across quartiles = 0.007) as was death from CHD (RR 0.62, CI 0.40 to 0.94, P for the linear trend across quartiles = 0.04) and all myocardial in-farction (nonsudden and sudden deaths) (RR 0.56, CI 0.33 to 0.93, P for the linear trend across quartiles = 0.017). For death from CVD, the linear trend for the deceased mortality rate across quartiles was significant (P = 0.01).


Fish consumption in men was associated with a decreased long-term risk for death from CHD, especially nonsudden death from myocardial infarction.

Sources of funding: In part, American Heart Association; National Heart, Lung, and Blood Institute; Chicago Health Research Foundation; Otho S. Sprague Foundation; Presbyterian-St. Luke's Hospital; Illini Foundation.

For article reprint: Dr. M.L. Daviglus, Department of Preventive Medicine, Northwestern University Medical School, 680 North Lake Shore Drive, Suite 1102, Chicago, IL 60611, USA. FAX 312-908-9588.


Some epidemiologic studies (but not all) have shown that increased fish consumption is inversely associated with death from CHD (1, 2). Unfortunately, the interpretation of the strength of this association is difficult given the correlation between fish consumption and other protective lifestyle habits.

In this analysis from the Western Electric Study, the type of dietary assessment and the repeat assessment at 1 year are methodologically sound. This reassessment increases our confidence in the classification of men by levels of fish consumption. However, given that the baseline dietary assessment was completed in 1957 and the ability to account for changes in diet over time is impossible, we must assume that no major changes occurred in consumption patterns.

The protective advantage of increased fish consumption against CHD is believed to be mediated through a reduction in triglyceride levels and a reduction in platelet aggregation caused by eicosapentaenoic acid. Unfortunately, in this study, the serum triglyceride measurements were not provided, thereby precluding a comparison between what is biochemically expected and what is clinically observed. Also, the type of fish consumed is not reported, which is of interest because fatty fish contains higher levels of eicosapentaenoic acid than does white fish.

In this analysis, persons who consumed greater amounts of fish also consumed greater amounts of fruits and vegetables and alcohol; they were also more likely to be Catholics and white-collar workers. It is reassuring that, after careful statistical adjustment for these known correlates, a strong inverse association persists between fish consumption and mortality from CHD. However, it is not possible in cohort studies to adjust for unknown confounders or other variables, such as physical activity, which were not recorded. Therefore, although it is likely that the reported association exists, consideration of other potential confounding factors remains important in our interpretation of the strength of the inverse association.

Sonia S. Anand, MD, MSc
McMaster UniversityHamilton, Ontario, Canada

Sonia S. Anand, MD, MSc
McMaster University
Hamilton, Ontario, Canada


1. Kromhout D, Bosschieter EB, de Lezenne Coulander C. The inverse relation between fish consumption and 20-year mortality from coronary heart disease. N Engl J Med. 1985;312:1205-9.

2. Ascherio A, Rimm EB, Stampfer MJ, Giovannucci EL, Willett WC. Dietary intake of n-3 fatty acids, fish intake, and the risk of coronary artery disease. N Engl J Med. 1995;332:977-82.