select search filters
roundups & rapid reactions
before the headlines
Fiona fox's blog

dietary fats and heart disease

A clinical trial in the BMJ suggested that replacing saturated animal fats with omega-6 polyunsaturated vegetable fats is linked to an increased risk of death among patients with heart disease. This accompanied a roundup which can be viewed here.



Title, Date of Publication & Journal

Use of dietary linoleic acid for secondary prevention of coronary heart disease and death: evaluation of recovered data from the Sydney Diet Heart Study and updated meta-analysis. BMJ 2013


Claim supported by evidence?

This paper takes a small study published in 1978 and adds some clarification of the original findings.

The paper does suggest the claim that data from the trial tend to indicate that a specific source of unsaturated fat (n-6 fatty acids) might lead to a slightly increased cardiovascular risk.

The paper does not provide evidence for the claim that substitution of “any polyunsaturated fatty acids in general” for saturated fats leads to a different cardiovascular risk.

Overall, there is no new base of evidence. The paper does not alter our understanding of the possible relationship between diet and cardiovascular risk.



  • The Sydney DHS had been performed between 1966 and 1973. Its data (e.g. death rates) had been reported, but not up to the current scientific standard. Therefore, this study had not been incorporated in existing meta-analyses.
  • There already exists a solid based of meta-analysis on the relationship between fatty acids and health. The inclusion of the Sydney DHS is a useful exercise, but it does not turn around the overall evidence that already existed.
  • In the new study, the use of n-6 fatty acids was related to increased cardiovascular risks. Previous meta-analysis suggested:
    • A small reduction in cardiovascular risk is observed when more unsaturated fat is digested; however the result appears to be driven by so called n-3 fatty acids.
    • When using n-6 fatty acids only, there was so far no evidence of cardiovascular risk reduction or risk increase.  
      There is a meta-analysis available from 2010 by the same authors as the current paper that concluded in 2010: “Advice to specifically increase n-6 PUFA intake, based on mixed n-3/n-6 RCT data, is unlikely to provide the intended benefits, and may actually increase the risks of CHD and death.” 
      Hence the new paper raises no new claim.
  • The authors describe a potential mechanistic model of how the n-6 fatty acids may contribute to cardiovascular risk. This model cannot be generalised to other unsaturated fatty acids.
  • I had no access to the supplementary material so could not perform a thorough review of the statistical methodology; but overall the description of the statistical methods appears to be adequate for the type of the paper.
  • It is generally desirable to make results from all clinical trials publically available. However the current paper is not an example of that as the reporting of the Sydney diet heart study was not intentionally hidden, as it is the case for some drug trials.


Study Conclusions

  • For reporting this new result, specific attention should be paid to the difference of available unsaturated fats, as the combined evidence suggests different impacts of n-3 vs. n-6 fatty acids on cardiovascular risks.



  • The new data, originating from an old study contribute to the overall evidence for the different portions of fats in the diet.
  • The study was performed in men, 30-59, with coronary heart disease. It is not clear whether the results can be generalised to other populations, specifically to healthy people.
  • The general scientific standard of the paper appears to be adequate.
  • The claims in the paper are not new or at odds with existing evidence. They may be different to existing nutrition guidances.


‘Before the headlines’ is a service provided to the SMC by volunteer statisticians: members of the Royal Statistical Society (RSS), Statisticians in the Pharmaceutical Industry (PSI) and experienced statisticians in academia and research.  A list of contributors, including affiliations, is available here.


in this section

filter Headlines by year

search by tag