Background and purpose: Evidence of a relationship of fish intake to stroke incidence or mortality is weak. This report examines the association of fish consumption with stroke.
Methods: A cohort of 2107 men aged 40 to 55 years from the Chicago Western Electric Study who were free of coronary heart disease and stroke through their first annual reexamination was investigated in relation to baseline fish intake and 30-year risk of fatal and nonfatal stroke. Data on baseline fish intake, categorized into four levels (> or = 35 g/d, 18 to 34 g/d, 1 to 17 g/d, and 0 g/d), were available for 1847 men. Average values of macronutrients and micronutrients from the first two examinations and major coronary and stroke risk factors were assessed in relation to fish consumption. Stroke mortality was ascertained from death certificates and nonfatal stroke from records of the Health Care Financing Administration.
Results: During 46,426 person-years of follow-up, 76 stroke deaths occurred. Men consuming > or = 35 g/d of fish (highest level) had a higher age-adjusted death rate from stroke (23.5 per 10,000 person-years) than men in the three other categories of fish consumption. Based on a Cox proportional hazards regression model with adjustment for age, systolic blood pressure, cigarette smoking, serum cholesterol level, diabetes, electrocardiographic abnormalities, and table salt use, hazards ratios (and 95% confidence intervals) for fish consumers compared with nonconsumers were 1.34 (0.53 to 3.41) for > or = 35 g/d, 0.96 (0.41 to 2.21) for 18 to 34 g/d, and 1.00 (0.43 to 2.33) for 1 to 17 g/d. Age-adjusted and multivariate analyses for fatal and nonfatal strokes (n = 222) yielded similar results.
Conclusions: With stroke rates highest in the subgroup reporting highest fish intake, these data do not support the hypothesis of an inverse association of fish consumption with strokes.