Article Text

Download PDFPDF

Fish consumption for cardiovascular health: benefits from long-chain omega-3 fatty acids versus potential harms due to mercury
  1. Xiong-Fei Pan1,2,
  2. Matti Marklund1,3,
  3. Jason HY Wu1
  1. 1 The George Institute for Global Health, Faculty of Medicine, University of New South Wales, Sydney, New South Wales, Australia
  2. 2 School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China
  3. 3 The Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA, USA
  1. Correspondence to Dr Jason HY Wu, The George Institute for Global Health, Faculty of Medicine, University of New South Wales, Sydney, NSW 2042, Australia; jwu1{at}

Statistics from

Cardiovascular disease (CVD) was the top cause of premature deaths globally in 2017, contributing to ~17.8 million deaths, including 8.9 million from coronary heart disease (CHD).1 There has been long-standing scientific interest in the cardioprotective effects of fish consumption. A substantial amount of research concerning cardiovascular benefits of fish has focused on long-chain omega-3 polyunsaturated fatty acids (PUFAs, including eicosapentaenoic acid (EPA; 20:5 n-3), docosapentaenoic acid (DPA; 22:5 n-3) and docosahexaenoic acid (DHA; 22:6 n-3)) found in fish, but other nutrients such as protein and vitamin D may also play a role.2 Conversely, fish is also the leading dietary source of mercury.3 Mercury in rivers, lakes and oceans is assimilated into fish by the process of bioaccumulation and its organic form (methylmercury), is found in higher levels among long-living predatory fish. In animal experimental and in vitro studies, mercury demonstrates harmful physiological effects such as causing dyslipidemia, thrombosis and cardiovascular toxicity,3 raising concerns that exposure might increase the risk of CVD. Therefore, understanding the independent cardiovascular effects of putative beneficial and harmful nutrients in fish, such as omega-3 PUFAs and mercury, and their possible interactions, is of crucial public health and clinical importance.

In their Heart paper, Tajik and colleagues examined the cross-sectional associations of serum levels of long-chain omega-3 PUFAs, as well as hair methylmercury levels with exercise-induced myocardial ischaemia.4 The assessment of serum omega-3 PUFAs and hair mercury levels is a notable strength, which are validated biomarkers that reflect their long-term dietary consumption.3 The study was conducted among 2199 men aged 42–60 years participating in the Kuopio Ischaemic Heart Disease Risk Factor Study (KIHD) in eastern Finland, with the exposure and outcomes assessed in 1984–1989. The investigators reported that total long-chain omega-3 PUFAs were significantly inversely associated with risk of exercised-induced ischaemia, with OR comparing those in the top to the bottom quartile of 0.67 (95% CI 0.51 to 0.87) across the whole cohort. In contrast, higher hair levels of methylmercury were associated with substantially higher risk of exercised-induced ischaemia, with 62% higher odds comparing the highest to the lowest quartile (OR, 1.62, 95% CI 1.22 to 2.14). The effects of omega-3 PUFAs and methylmercury appeared to be independent of each other, and there was little evidence that they modified each other’s relationship with the outcome. That is to say—magnitudes of protective associations for omega-3 PUFAs for exercise-induced ischaemia was similar whether the participants had high levels or low levels of hair methylmercury, and vice versa. Finally, the investigators reported that protective association of omega-3 PUFAs with exercise-induced ischaemia was most pronounced among participants with a history of CHD, whereas harmful associations observed between methylmercury and the outcome appeared similar among subjects with and without prior history of CHD.

The study has important limitations including its relatively modest sample size, and the cross-sectional design that precludes establishing temporal relationships between the exposures and the outcome. On the other hand, investigating exercise-induced ischaemia was a novel aspect of this study, since very few observational studies have assessed associations of both omega-3 PUFAs and methylmercury biomarkers with this outcome. The biological plausibility of the protective associations observed for omega-3 PUFAs with exercised-induced ischaemia is supported by prior experimental studies. For instance, omega-3 PUFA supplementation improved cardiac function and reduced infarction severity in animal models of myocardial ischaemia.5 Since exercise-induced myocardial ischaemia is predictive of future cardiac events, particularly sudden cardiac deaths, results of the current study are also consistent with those from meta-analysis of prospective cohort studies, in which higher biomarker levels of long-chain omega-3 PUFAs and intake of fish were significantly associated with lower risk of CHD deaths.6 In the largest study of omega-3 PUFA biomarkers and incident CHD to date that included 45 637 participants across 16 countries, both DHA (relative risk (RR), 0.90; 95% CI 0.84 to 0.96 for 1 SD increase) and DPA (RR, 0.90; 95% CI 0.85 to 0.96) were significantly associated with lower risk of fatal CHD, with EPA also demonstrating a trend towards inverse associations with fatal CHD (RR, 0.91; 95% CI 0.82 to 1.00).6 The differential associations of serum omega-3 PUFAs with exercise-induced ischaemia for participants with and without history of CHD is an intriguing finding from Tajik and his colleagues, although this result should be interpreted cautiously due to reduced power to detect associations within the subgroups.

The observed positive association of methylmercury with exercise-induced ischaemia in the current study is consistent with previous results published in the KIHD cohort, which found that higher hair mercury levels were associated with higher risk of acute coronary event, CHD and total CVD.7 Such findings are generally inconsistent with results from other population-based observational studies that have assessed the association between methylmercury biomarker levels and CVD or CHD.8 Overall, pooled estimates across six studies suggest that mercury exposure was not appreciably associated with risk of either CVD (pooled RR, 0.94; 95% CI 0.66 to 1.36) or CHD (RR, 0.99; 95% CI 0.65 to 1.49).8 However, there was significant heterogeneity in results for both CVD and CHD across the studies. Different associations across studies appear unlikely to be explained by levels of exposure, as similar levels have been observed in populations in the USA, where mercury was not associated with CVD risk,9 and the KIHD7 and another study across Europe,10 where mercury was related to higher CVD risk. Other study characteristics such as sampling design (prospective cohort, nested case–control and case–control), types of biosamples used to assess mercury levels (hair, toenail and blood) and sample size (from as small as a few hundred participants up to thousands of participants) might explain the discrepant findings, which requires further exploration. Of note, most of these included studies generally did not report significant interaction between mercury and omega-3 PUFAs for risk of CVD, which means that protective effect of omega-3 PUFAs appears to be equivalent across different levels of mercury exposure.

Overall, results in the current study are consistent with likely cardiovascular benefits of long-chain omega-3 PUFAs and fish consumption. The current totality of evidence does not support significant harmful effects of mercury on CVD outcomes—however, the study by Tajik and colleagues highlights the need to further investigate reasons that may underlie the different associations between mercury and CVD across studies and populations. On balance, the current body of evidence suggests the benefits of moderate fish consumption on cardiovascular health very likely outweigh the risks of any potential harm attributable to mercury exposure.11 Consistent with such interpretation, major international and national organisations such as the WHO, American Heart Association and the European Society of Cardiology consistently recommend 1–2 servings of fish per week as part of a healthy diet.

For those who are concerned about mercury exposure, risks can be further mitigated through consuming a diverse variety of fish, and avoidance of large predatory fish such as swordfish, shark and king mackerel. Of note, although omega-3 PUFA supplements contain no mercury, they cannot replace the role of fish as part of a healthy diet.2 Fish provides other nutrients such as high-quality protein, vitamins and minerals that have potential cardioprotective effects. Current trial evidence suggests that omega-3 PUFA supplements may be recommended for patients with prevalent CHD to prevent CHD death but their use for primary prevention of CVD is still to be substantiated.12 A recent meta-analysis of 10 randomised trails (each with at least 500 participants and treatment duration of at least 1 year) in 2018 showed omega-3 PUFA supplements reduced the risk of CHD death by 7% (RR, 0.93; 95% CI 0.85 to 1.00) among 77 917 high-risk individuals with a mean treatment duration of 4.4 years,13 which was fairly consistent with another meta-analysis of all major trials conducted until 2012 that identified a reduction in cardiac death (RR, 0.91, 95% CI 0.85 to 0.98) with omega-3 PUFA supplementation.14 In addition, in individuals who cannot consume sufficient fish and are at higher risk of CVD, omega-3 PUFA supplements can still be a safe and reasonable option to reduce the risk of CVD, given the cumulative evidence of their impact on risk factors from clinical trials, in combination with the low risk profile of omega-3 PUFAs.15 Finally, since high mercury exposure can cause other health problems (eg, detrimental neurocognitive effects in infants),3 public policy should continue to reduce mercury contamination in the environment and fish, such as through reducing reliance and lowering pollution from coal power plants which is a major anthropogenic mercury source.16


View Abstract


  • Contributors JW conceived the study. XP, MM and JW conducted the study and drafted the manuscript. JW, XP and MM agreed upon the final version of the manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

  • Patient consent for publication Not required.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles