See Related:
Authors' Articles

Context  Some prospective studies have shown an inverse association between fish intake and risk of stroke, but none has examined the relationship of fish and omega-3 polyunsaturated fatty acid intake with risk of specific stroke subtypes.

Objective  To examine the association between fish and omega-3 polyunsaturated fatty acid intake and risk of stroke subtypes in women.

Design, Setting, and Subjects  Prospective cohort study of women in the Nurses' Health Study cohort, aged 34 to 59 years in 1980, who were free from prior diagnosed cardiovascular disease, cancer, and history of diabetes and hypercholesterolemia and who completed a food frequency questionnaire including consumption of fish and other frequently eaten foods. The 79 839 women who met our eligibility criteria were followed up for 14 years.

Main Outcome Measure  Relative risk of stroke in 1980-1994 compared by category of fish intake and quintile of omega-3 polyunsaturated fatty acid intake.

Results  After 1 086 261 person-years of follow-up, 574 incident strokes were documented, including 119 subarachnoid hemorrhages, 62 intraparenchymal hemorrhages, 303 ischemic strokes (264 thrombotic and 39 embolic infarctions), and 90 strokes of undetermined type. Among thrombotic infarctions, 90 large-artery occlusive infarctions and 142 lacunar infarctions were identified. Compared with women who ate fish less than once per month, those with higher intake of fish had a lower risk of total stroke: the multivariate relative risks (RRs), adjusted for age, smoking, and other cardiovascular risk factors, were 0.93 (95% confidence interval [CI], 0.65-1.34) for fish consumption 1 to 3 times per month, 0.78 (95% CI, 0.55-1.12) for once per week, 0.73 (95% CI, 0.47-1.14) for 2 to 4 times per week, and 0.48 (95% CI, 0.21-1.06) for 5 or more times per week (P for trend = .06). Among stroke subtypes, a significantly reduced risk of thrombotic infarction was found among women who ate fish 2 or more times per week (multivariate RR, 0.49; 95% CI, 0.26-0.93). Women in the highest quintile of intake of long-chain omega-3 polyunsaturated fatty acids had reduced risk of total stroke and thrombotic infarction, with multivariate RRs of 0.72 (95% CI, 0.53-0.99) and 0.67 (95% CI, 0.42-1.07), respectively. When stratified by aspirin use, fish and omega-3 polyunsaturated fatty acid intakes were inversely associated with risk of thrombotic infarction, primarily among women who did not regularly take aspirin. There was no association between fish or omega-3 polyunsaturated fatty acid intake and risk of hemorrhagic stroke.

Conclusions  Our data indicate that higher consumption of fish and omega-3 polyunsaturated fatty acids is associated with a reduced risk of thrombotic infarction, primarily among women who do not take aspirin regularly, but is not related to risk of hemorrhagic stroke.

JAMA. 2001;285:304-312

Author Contributions: Dr Iso participated in the study concept and design, analysis and interpretation of data, and drafting of the manuscript and provided statistical expertise.

Dr Rexrode participated in the study concept and design, acquisition of data, analysis and interpretation of data, and drafting of the manuscript; provided critical revision of the manuscript for important intellectual content; obtained funding, and supervised conduct of the study.

Drs Stampfer and Willett participated in the study concept and design, acquisition of data, analysis and interpretation of data; provided critical revision of the manuscript for important intellectual content, statistical expertise and administrative, technical, or material support; and obtained funding.

Dr Manson participated in the study concept and design, acquisition of data, analysis and interpretation of data; provided critical revision of the manuscript for important intellectual content and administrative, technical, or material support; obtained funding; and supervised conduct of the study.

Dr Colditz participated in the study concept and design, acquisition of data, analysis and interpretation of data; provided critical revision of the manuscript for important intellectual content and statistical expertise.

Dr Speizer participated in the study concept and design and acquisition of data; provided critical revision of the manuscript for important intellectual content, and statistical expertise and administrative, technical, or material support; obtained funding, and supervised conduct of the study.

Dr Hennekens participated in the study concept and design and acquisition of data; provided critical revision of the manuscript for important intellectual content and administrative, technical, or material support; and supervised conduct of the study.

Funding/Support: This work was funded by research grants CA40346 and HL34594 from the National Institutes of Health. Dr Iso is the recipient of an overseas research fellowship from the Japan Society for the Promotion of Science.

Acknowledgment: We thank the participants in the Nurses' Health Study for their continuing cooperation and Karen Corsano, BA, MSL, Barbara Egan, and Lisa Dunn, BS, for their expert help.