Lack of evidence that saturated fat causes cardiovascular diseaseBMJ 2014; 348 doi: https://doi.org/10.1136/bmj.g3205 (Published 14 May 2014) Cite this as: BMJ 2014;348:g3205
- Uffe Ravnskov, independent researcher1
From table 2 and the references of their systematic analysis Micha and colleagues seem to recommend replacing saturated fatty acids (SFAs) with polyunsaturated fatty acids (PUFAs) to lower the risk of cardiovascular disease.1 As support they refer to the joint WHO/FAO expert consultation from 2003, in which the evidence came from an American Heart Association consensus report. The only argument in that report came from the Nurses’ Health Study, but from the tables in that study it seems that, after multivariate adjustment, including various types of dietary fats, the relative risk of coronary heart disease did not differ significantly between the different levels of dietary SFA intake.
Micha and colleagues also refer to Bradford Hill’s criteria for causation, but the idea that saturated fat causes cardiovascular disease does not satisfy any of them.
For example, more than 25 cohort and case-control studies have shown that patients with coronary heart disease do not eat more SFAs than others. In seven studies, stroke patients had eaten significantly less.2 3
A meta-analysis of 10 cohort studies with more than 400 000 people found that high consumers of dairy products had a significantly lower risk of cardiovascular disease than low consumers,4 and a thorough review of the associations between dairy products and cardiovascular disease showed no evidence of harmful effects.5
A recent systematic review of 76 observational and experimental studies that looked at 530 525 patients found no association between increased intake of PUFAs or decreased intake of SFAs and cardiovascular disease.6
To recommend replacing SFAs with PUFAs without defining which type of PUFA is also questionable because the dominant PUFA in processed food is omega 6, and food with a high omega 6 to omega 3 fatty acid ratio is associated with serious effects on human health.
Cite this as: BMJ 2014;348:g3205
Competing interests: None declared.
Full response at: www.bmj.com/content/348/bmj.g2272/rr/695803.