Benefits and risks of a fish diet—should we be eating more or less?

Khursheed N Jeejeebhoy  About the author

The role of omega-3 fatty acids such as eicosapentaenoic acid and docosahexaenoic acid in preventing coronary artery disease has gradually been elucidated, and there is increasing evidence that fish oils have a cardioprotective effect. A controlled trial by Burr et al.¹ showed that patients who had had a myocardial infarction had a 29% reduction in mortality over 2 years by eating three fish meals a week. By contrast, patients who were randomized to high-fiber and low-fat diets did not have a significant reduction in mortality. Many trials have since shown the benefit of taking fish oils or eating a diet rich in fish, and the American Heart Association recommends an increase in the dietary intake of omega-3 fatty acids.²

Cardioprotective benefits have been observed with daily consumption of both wild and farmed fish, which are both high in omega-3 fatty acids. The consumption of as little as one fish meal weekly has been shown to be beneficial, with dose-dependent greater benefits up to about five fish meals per week.³

A large randomized trial of 18,000 patients with hypercholesterolemia (observed for 5 years) showed that adding 1800 mg/day of eicosapentaenoic acid to statin treatment resulted in a significant reduction in major coronary events compared with the controls who received statin treatment alone.⁴

The euphoria of conquering coronary artery disease by a simple diet change has been tempered by a plethora of papers that warn against eating fish because of the risk of mercury poisoning of the central nervous system. The stark choice, if one believes the mercury poisoning enthusiasts, can be summarized as between living a shorter life with the mental abilities of an Einstein or living a long life as a moron. What are the facts?

Mercury enters the atmosphere by combustion of waste and coal. The element then enters the oceans from the atmosphere where it is converted to methyl mercury by microorganisms and then taken up by marine life and concentrated in fish. As methyl mercury is not fat soluble, unlike dioxins, it does not reside in the fatty tissues. Methyl mercury is strongly neurotoxic, as shown by studies in Iraq where the consumption of bread contaminated by a fungicide containing methyl mercury resulted in mental retardation, seizures and microcephaly in infants.⁵

The concentration of methyl mercury in fish is increased by fish eating other fish for food. Fish that are not predatory, such as sardines, salmon and shrimp, therefore have very low levels of methyl mercury. By contrast, predatory fish such as shark, tuna, swordfish and orange roughy have higher levels of methyl mercury. Interestingly, the much-maligned farmed fish have the lowest levels of methyl mercury. To add another level of complexity to the debate, although methyl mercury per se is very neurotoxic, in fish methyl mercury is bound to cysteine, and this compound has a tenth of the toxicity of pure methyl mercury.⁶

What evidence is there that the intake of methyl mercury from eating fish causes neural damage in humans? In the Faroe Islands a study was conducted in a cohort of infants over a 14-year period.⁷ The study examined the development of the nervous system in children who were born to mothers who ate pilot whale meat daily in their diet. This study showed that there was a correlation between high prenatal mercury intake by the mother and neurological developmental deficits in the infant.

By contrast, in the Seychelle islands where women eat 12 fish meals a week, no effects on infant neurological development were noted despite the fact that the mean methyl mercury concentration in the hair of Seychelle island inhabitants, including infants, was 10–20 times that seen in US inhabitants.⁸ The concentration of methyl mercury in fish caught around the Seychelles, however, was similar to that found around the US—0.05–0.25 ppm. The higher levels of methyl mercury found in the Seychelle islanders were therefore due to the islanders eating more fish rather than eating highly contaminated fish. By contrast, pilot whale meat has 10 times the concentration of methyl mercury that is found in ocean fish (1.6 ppm). The difference between the data from the Faroes and the Seychelles is therefore likely to be largely because individuals in the Faroe Islands had much higher exposures to methyl mercury, because they ate marine mammals and not fish. It should be noted that a toxic level of mercury in hair is estimated to be 50 ppm, and even with a safety factor of 10, which reduces the threshold to 5 ppm, the levels of the Seychelle islanders eating 12 fish meals a week was only a mean of 6.8 ppm.

On the basis of these observations, should we eat fish and, if so, how much? It is clear that the degree of methyl mercury contamination in food determines its toxicity. Furthermore, the average individual intake of fish in North America will not approach 12 ocean fish meals per week (as in the Seychelles where no toxicity has been observed) and will certainly not include the consumption of highly contaminated marine mammals (as in the Faroes). Furthermore, by eating fish with low levels of methyl mercury, such as sardines, salmon and shrimp, the dietary intake of methyl mercury can be reduced further.

Other potential contaminants in fish such as dioxins and polychlorinated biphenyls could potentially increase the risk of cancer. An analysis of the potential harmful effects of these contaminants in fish versus the benefits of omega-3 fatty acids has, however, concluded that the "[l]evels of dioxins and polychlorinated biphenyls in fish are low, and potential carcinogenic and other effects are outweighed by potential benefits of fish intake".⁹

The US Institute of Medicine has recommended that pregnant women restrict their intake of fish with a higher methyl mercury content (e.g. shark, tuna, or swordfish) to 1 meal per 2 weeks; however, these women can eat 2–3 meals of other fish per week (e.g. sardines, salmon, or shrimp). In nonpregnant individuals the recommended dietary fish intake is 1 and 2–3 per week, respectively, for high and low methyl-mercury-containing fish. On the basis of the data given in the main body of this article, the recommendations of the Institute of Medicine are ultra-conservative, and there is little evidence that 2–3 meals of low-mercury-containing fish per week can cause harm. In North America wild and farmed salmon would provide an ideal option to reduce the risk of both heart disease and methyl mercury poisoning, as these fish are rich in omega-3 fatty acids and very low in methyl mercury. Finally, for individuals who want a diet with zero methyl mercury but would like to enjoy the benefits of omega-3 fatty acids, there are always fish oil supplements.

Competing interests

The author declared no competing interests.

ARTICLE NAVIGATION -ARTICLE TOOLS

Full text
• Download PDF
• Send to a friend
• Export citation
• Rights and permissions
• Order commercial reprints

ISSUE NAVIGATION - THIS ISSUE

• Table of Contents
• Previous article
• Next article