Assessment: Fish Consumption and Incidence of Stroke Meta-Analysis

Purpose of review: Further research was necessary into relationship between fish consumption and risk of stroke because observational studies were inadequate for determining relationship.

Review strategy: Meta-analysis on cohort studies relating to fish consumption and risk of stroke.

Description: Researchers searched Medline and Embase databases with relative keywords (e.g. “fish) to identify 9 total relevant cohorts (note: one study counted as two) published in English language journals from 1966 to 2003.

The cohort studies chosen each provided “relative risk” and “corresponding 95%” confidence interval factors for relationship of stroke with fish consumption. Duration average of each was 12.8 years with a range of 4 to 30 years. Each study had been properly adjusted for multiple covariates. Regional variance existed: of the 9 studies, 6 were from the U.S., 1 from Europe, 1 from China, and 1 from Japan creating a regional variance. Five of the studies attained data through personal interviews presenting variance with stratified analyses.

Disadvantages: The researchers note limitations in that most studies chosen “did not separate ischemic from hemorrhagic stroke”, which are different in etiology. It was noted that Japan and China studies presented a population of which may have higher risk of hemorrhagic stroke versus most studies from the U.S., where ischemic stroke is higher risk. Confounding variables from each cohort also must be taken into consideration.

Results: Review suggests “an inverse association between fish consumption and risk of stroke, particularly ischemic stroke, although there was not a strong dose-response relationship.”

Assessment: Large populations and different ethnic groups among the 9 cohorts introduce multiple variables that challenge this meta-analysis considerably. These would include, as noted in the study, even how and what kind of fish was cooked. The review also is limited in determining what amounts of fish should be eaten for most benefit or whether or not there was a relationship of stroke risk with amount of n-3 fatty acids.

Reference: He K, Song Y, Daviglus ML et al. Fish consumption and incidence of stroke: a meta-analysis of cohort studies. Stroke 2004;35:1538-42. Available at: http://stroke.ahajournals.org/cgi/reprint/35/7/1538?ijkey=50a74da6182d65ded6f8ab521844953e5469f6a6.