Abstract
Fish consumption is an integral part of a healthy diet, as it is an excellent source of protein, essential Ω-3 fatty acids, eicosapentaenoic acid and docosahexaenoic acid, as well as certain micro-nutrients. A large number of observational studies have suggested beneficial effects of fish intake on all-cause mortality and cardiovascular disease [1–10]. However, studies examining the relationship between fish consumption and stroke show conflicting results. These differences could be attributed to methodological diversities; different populations studied, variety in the types of fish consumed, as well as in preparation methods. In addition, most studies focus on total stroke risk, meaning that the results could differ if a type-specific stroke analysis for ischemic and hemorrhagic stroke was performed [11].
Methods
The Swedish Mammography Cohort [12] was established in 1987–1990, inviting all women born between 1914 and 1948, and living in central Sweden to complete a questionnaire on diet. The questionnaire was completed by 66,651 women (74% of the source population). In 1997, a new expanded questionnaire on diet and other lifestyle factors was sent to 56,030 women, and was completed by 39,227 women (70%). After the exclusion of participants with a history of stroke, ischemic heart disease or cancer, 34,670 women aged 49–83 years were finally included in the present analysis. The study was approved by the Regional Ethical Review Board at the Karolinska Institute (Stockholm, Sweden) [13].
The following variables were assessed: education, weight, height, smoking status, physical activity, aspirin use, history of hypertension and diabetes, family history of myocardial infarction before 60 years of age, alcohol consumption and diet. Diet was assessed with a 96-item self-administered food-frequency questionnaire. Incident cases of first stroke were ascertained by linkage of the cohort to the Swedish hospital discharge registry. The International Classification of Diseases 10th revision (ICD-10) was used to identify stroke events. Strokes were classified as cerebral infarction, hemorrhagic stroke and unspecified stroke.
Results
After a mean follow-up period of 10.4 years, 1680 incident stroke cases were recorded, including 1310 cerebral infarctions, 233 hemorrhagic strokes, and 137 unspecified strokes. The mean (± standard deviation) servings of fish consumed per week were 2.1 ± 2.2 servings of total fish, 1.0 ± 1.3 servings of cod, saithe and fish fingers, 0.5 ± 0.9 servings of salmon, whitefish and char, and 0.7 ± 0.9 servings of herring and mackerel. Fish consumption was significantly inversely associated with risk of total stroke; compared with women in the lowest quintile of fish consumption (<1.0 servings of fish/week), the multivariable relative risk (RR) of total stroke for women in the highest quintile (>3.0 servings of fish/week) was 0.84 (95% CI: 0.71–0.98; p for trend = 0.049). An inverse association was also observed between fish consumption and cerebral infarction or hemorrhagic stroke; compared with women in the lowest quintile of fish consumption (<1.0 servings of fish/week), the multivariable RR of cerebral infarction or hemorrhagic stroke for women in the highest quintile (>3.0 servings of fish/week) was 0.87 (95% CI: 0.73–1.04; p for trend = 0.19) and 0.67 (95% CI: 0.42–1.08; p for trend = 0.08), respectively. Finally, the associations focusing on the type of fish consumed and the preparation method showed that consumption of lean fish but not of other fish types was inversely associated with risk of stroke (multivariable RR of total stroke 0.67 (95% CI: 0.49–0.93; p for trend = 0.07) for >3 servings of lean fish/week, compared with that for no consumption), and that fried fish consumption was not associated with a risk of stroke 0.95 (95% CI: 0.84–1.06), highest versus lowest quartile.
Significance of results
This study underlines the protective effect of fish consumption and especially lean fish consumption, regarding stroke incidence. The results of the present study are of considerable importance, since very few studies examining the relationship between fish consumption and stroke have performed type-specific stroke analysis, while the results on total stroke risk are conflicting. This prospective cohort, with a study sample of 34,670 women and a follow-up period of approximately 10 years, is one of the largest prospective studies that have examined this relationship, after the Nurses' Health Study [14] and the Health Professionals' Follow-up study [15], with a sample from the general population, and linkage with population-based Swedish registries. Most importantly, this study is the largest to date that examines the associations between different types of fish with the risk of stroke.
The results of the present study are in accordance with the results of two meta-analyses, a meta-analysis combining data from eight prospective cohort studies and one of five prospective studies and one case–control study. Results from the first meta-analysis suggest an inverse association between fish consumption and risk of stroke and particularly ischemic stroke; compared with those who consume fish less than one time/month, those who consume fish one to three times/month have 9% lower risk, one time/week 13% lower risk, two- to four-times/week 18% lower risk, and more than five-times/week 31% lower risk [2]. In addition, the second meta-analysis showed that any fish consumption was associated with a 12% decreased stroke risk, compared with no fish consumption [1].
In the study by Larsson
Findings from this prospective cohort suggest that high consumption of lean fish may reduce the risk of stroke, while insignificant associations were observed regarding consumption of salmon, whitefish, and char or herring and mackerel. This observation could be attributed to the fact that herring and salmon are commonly eaten salted in Sweden, thus affecting blood pressure levels, a risk factor for stroke development. However, it should also be mentioned that results from other studies focusing on cardiovascular risk according to the type of fish consumed, show that fatty fish intake, such as salmon, herring and sardines (i.e., species more rich in the essential Ω-3 fatty acids), is associated with lower risk, compared with lean fish [5,16]. In addition, results from the present study, indicate that no associations were observed between consumption of fried fish and stroke, in contrast with a study conducted in US men and women, which showed positive associations between fried fish and fish sandwich consumption and the risk of stroke [17].
As a limitation of the present study it should be noted that fish consumption was assessed by using a self-administered questionnaire, which could probably lead to a misclassification of fish intake. Furthermore, it has been suggested that gender may influence the relationship between stroke risk and fish intake. In fact, according to the results of two studies, fish intake has different effects on stroke risk in different genders, with more favorable effects on women [18,19]. The results of this study, having been performed only in women, are not able to give more data in this area of interest.
Future perspective
The results of the present study add to the better understanding of the role of fish consumption and different types of fish and preparation methods to the risk of stroke and stroke subtypes. However, this area merits further investigation, not only because of the few number of studies that have examined the association between the consumption of specific types of fish and stroke subtypes, but also because of the presence of methodological issues and confounding factors that affect the relationship between diet and disease [20].
Executive summary
This prospective cohort study is one of the largest prospective studies to have examined the relationship between fish consumption and stroke, and is the largest to date examining the associations between different types of fish with the risk of stroke.
Fish consumption was significantly inversely associated with risk of total stroke.
Inverse associations were observed between fish consumption and cerebral infarction or hemorrhagic stroke; however, this did not reach statistical significance.
Further investigation in this field is warranted owing to the low number of studies that have examined the association between consumption of specific types of fish and stroke subtypes.
Footnotes