This prospective cohort study examined the effects of fish oil supplements on the clinical course of cardiovascular disease, from a healthy state to atrial fibrillation, major adverse cardiovascular events, and subsequently death.

The analysis is based on the UK Biobank study (1 January 2006 to 31 December 2010, with follow-up to 31 March 2021 (median follow-up 11.9 years)) including 415 737 participants, aged 40-69 years. Incident cases of atrial fibrillation, major adverse cardiovascular events, and death, identified by linkage to hospital inpatient records and death registries. Role of fish oil supplements in different progressive stages of cardiovascular diseases, from healthy status (primary stage), to atrial fibrillation (secondary stage), major adverse cardiovascular events (tertiary stage), and death (end stage).

Among 415 737 participants free of cardiovascular diseases, 18 367 patients with incident atrial fibrillation, 22 636 with major adverse cardiovascular events, and 22 140 deaths during follow-up were identified. Regular use of fish oil supplements had different roles in the transitions from healthy status to atrial fibrillation, to major adverse cardiovascular events, and then to death:

  • For people without cardiovascular disease, hazard ratios were 1.13 (95% confidence interval 1.10 to 1.17) for the transition from healthy status to atrial fibrillation and 1.05 (1.00 to 1.11) from healthy status to stroke.
  • For participants with a diagnosis of a known cardiovascular disease, regular use of fish oil supplements was beneficial for transitions from atrial fibrillation to major adverse cardiovascular events (hazard ratio 0.92, 0.87 to 0.98), atrial fibrillation to myocardial infarction (0.85, 0.76 to 0.96), and heart failure to death (0.91, 0.84 to 0.99).

The authors concluded that regular use of fish oil supplements might be a risk factor for atrial fibrillation and stroke among the general population but could be beneficial for progression of cardiovascular disease from atrial fibrillation to major adverse cardiovascular events, and from atrial fibrillation to death. Further studies are needed to determine the precise mechanisms for the development and prognosis of cardiovascular disease events with regular use of fish oil supplements.

I must admit that I am slightly puzzled by this study and its findings. The authors clearly speak of the ‘role’ regular use of fish oil supplements has. This language implies a casual impact. Yet, what we have here are associations, and every 1st year medical student knows that

correlation is not causation.

Other things to note are that:

  • the associations are only very weak;
  • they go in opposite directions depending on the subpopulation that is examined,
  • there is no plausible mechanism of action to explain all this.

Collectively, these facts suggest to me that we are indeed more likely dealing here with a non-causal association and not a causal link. All the more surprising then that the (UK) press took up this paper in a major and occasionally alarmist fashion (the headline in THE TELEGRAPH was Revealed: How cod liver oil could be bad for your health). I learned of it by listening to the BBC headline news.


6 Responses to Fish oil supplements and cardiovascular diseases

  • Interesting in this respect, perhaps:

  • Biobank is an open access resource. Which is good, except it does mean lots and lots of studies can be generated that produce correlations but with no real insight to causation. We then have the perennial problem of science journalism producing headlines that are hyped up and unhelpful.

  • Dear Edzard,

    thank you very much for sharing this interesting study.
    A few comments from my side which shall not be seen as destructive critics of the study but as additional viewpoints which may help to see the study in a more realistic context.

    The study was based on data bank information containing information about British citizens. It was not done as a comparative experiment, but as a comparison of groups formed within a sampling of data not meant for this special purpose.

    My thoughts on it:

    1) Supplements may be useful, if people are missing certain elements in their daily food. We know that fish oil contains such elements, which are missing if people do not eat a lot of fish. Now, my question: Do you think that British people have a need for supplements that may compensate not eating enough fish? It is a question. I am from Germany and if I should tell you where I can get best fish & chips – I would have no answer. I would not even know where to get any fish & chips! In Germany, food is often in connection with pig meat – if you have not vegetarian food.
    I would like to see a study like that in Germany, and I would expect much more clear results supporting the value of fish oil supplements! For a nation where there is almost no need for fish oil supplement, any finding would be a surprise.

    2) Like all of these database studies, we have to deal with a bias coming with the decision of people to take supplements or not. These people may be different in terms of general health and sensitivity to signs from the body. If you never had health issues, you may be not interested in supplements. On the other side, people who fear for keeping their heath status OK – up to hypochondriac attitudes – may take more likely supplements and meet the doctor for any problem they detect. These people would get much more easily a medical finding, just from the fact that they are being controlled much more regular. The bias on the data, you get from such a setup, has to be expected.

    3) It looks like fish oil would cause atrial fibrillation. On the other hand, the number of cases where atrial fibrillation leads to death is significantly reduced by fish oil. From the hazard ratio, the adverse effects may even be overcompensated by the positive effects. This is an interesting finding and I would like to understand the reason behind it.

    4) The dose makes the poison, and it would be exciting to see, if there is a relation. Maybe some people get too much of it, which would easily happen, if the general provision by the substances added by the supplements is good from the daily food.

    In general, these database related studies give a good idea for further studies looking at the findings in an experimental approach. Now, the next step should be made …

    Best regards,

  • And yet the multitudes of purveyors of fish oil will cite this study as “evidence” or even “proven”.

    • the associations are only very weak;

    they go in opposite directions depending on the subpopulation that is examined,

    • there is no plausible mechanism of action to explain all this.

    Simpson’s paradox is a phenomenon in probability and statistics in which a trend appears in several groups of data but disappears or reverses when the groups are combined. This result is often encountered in social-science and medical-science statistics, and is particularly problematic when frequency data are unduly given causal interpretations.

    Simpson’s paradox has been used to illustrate the kind of misleading results that the misuse of statistics can generate.

    Edward H. Simpson first described this phenomenon in a technical paper in 1951, but the statisticians Karl Pearson (in 1899) and Udny Yule (in 1903) had mentioned similar effects earlier. The name Simpson’s paradox was introduced by Colin R. Blyth in 1972. It is also referred to as Simpson’s reversal, the Yule–Simpson effect, the amalgamation paradox, or the reversal paradox.

    Post hoc subgroup analysis is a pseudoscientific tactic to manufacture positive results. Often deployed in conjunction with convenience sampling.
    — Pete Attkins

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Subscribe via email

Enter your email address to receive notifications of new blog posts by email.

Recent Comments

Note that comments can be edited for up to five minutes after they are first submitted but you must tick the box: “Save my name, email, and website in this browser for the next time I comment.”

The most recent comments from all posts can be seen here.