In 2020, Cochrane published an analysis that claimed a 17% lowered risk of cardiovascular events from reduced saturated fat intake. This result can be seen in the following image (taken from reference 1):

You will notice that I've also highlighted the Houtsmuller 1979 trial. Participants in that trial were allocated to a control diet supposedly rich in saturated fat or to a diet rich in omega-6 polyunsaturated fat (linoleic-rich diet). At the end of the trial, the results allegedly showed a large cardiovascular benefit for the polyunsaturated group.

This of course sounds like a great result. But unfortunately, it is too good to be true. Many researchers warn that we should be skeptical of large effects from small studies, which is certainly the case here.

Although the Houtsmuller trial takes only 4.7% of the weight in the Cochrane meta-analysis, removing the trial changes the overall result from a 17% reduction to a statistically non-significant 12% reduction.

(Image: Cochrane's main analysis for combined CVD events after exclusion of the Houtsmuller trial)

Thus, if the Houtsmuller trial is removed, the case against saturated fat becomes weaker.

But why should the Houtsmuller trial be removed?

For one, the lead author (Dr. Houtsmuller) has been suspected of fraud. In fact, Cochrane noted this on multiple occasions. To quote from their 2018 paper (2):

We found Houtsmuller 1979 to be at high risk of other bias as concerns about fraud of the first author were raised around his later research (on cancer diets).

After reading about this possible fraud, I became curious and scoured the web for more information. Lo and behold, I discovered a piece from the ‘Newsletter of The Association for Skeptical Enquiry’ that mentioned Dr. Houtsmuller. To quote (3):

Two diets have dominated the field: the Moerman diet, originating in the 1950s, and its successor, the Houtsmuller diet, which was invented in the 1990s. Both inventors were charismatic medical doctors. Recently, Houtsmuller, who had promoted his diet by claiming that he had cured himself of a metastasised melanoma, was exposed on television by the chairman of the Dutch Society against Quackery and was forced to admit that he never had metastases of a melanoma in the first place.

Disturbing, to say the least.

More Red Flags

These concerns of fraud are further justified when we look at the trial information (or lack thereof).

Cochrane correctly stated that the trial was "extremely vague across all publications about its methods" (2). Also, given that the integrity of Dr. Houtsmuller is in question, and that no information is provided about how randomization was generated (e.g., a coin flip?), we have no reason to believe that the trial was actually randomized.

Not to mention the lack of blinding. Even the American Heart Association excluded the study because "those reading the ECGs were not blinded to treatment assignment" (4). Notably, assessment of cardiac events was not a "primary aim" of the study, and about 85% of the cardiac events were due to angina — an endpoint susceptible to major bias.

The Questionable Dietary Data

Adding to the list of problems is the dietary data. Ramsden and colleagues noted that the control group had an "unrealistic amount of SFA (35%E) exclusively from ‘saturated margarine’" (5). Furthermore, this "saturated margarine" is not butter and likely not saturated fat at all. Houtsmuller stated that only four subjects in the control group preferred butter over "saturated margarines" (6). And as Hamley observed in his 2017 analysis (7):

As the average cholesterol intake in the control group was identical to the experimental group and was 41% lower than the participants in the control group who ate butter, the ‘saturated margarines’ were not of animal origin and most likely comprised of hydrogenated vegetable oils.

Therefore, the unfortunate participants of the control group were likely subjected to a hefty dose of industrial trans-fat, not saturated fat.

Lack of Replication

If the massive cardiac benefit in the Houtsmuller trial is credible, then the results should be easy to replicate.

But this has not happened.

Other trials employing a linoleic-rich diet (without confounding by omega-3) have demonstrated no effect on cardiac events. If anything, these trials indicate a harmful effect from increased linoleic acid intake on heart disease mortality — the opposite of what the Houtsmuller trial suggests.

(Image: Results for high linoleic acid (LA) intakes on CHD mortality. Taken from reference 5)

Since the primary aim of the Houtsmuller trial was to assess progression of diabetic retinopathy (DR), I also wondered if the alleged benefit on this endpoint was replicated by any well-controlled randomized trial. But, all I found was another flawed trial (Howard-Williams et al.) where 40% of the randomized subjects were not available for examination (no intention-to-treat analysis) (8). Even so, a very limited observational "complier" analysis showed no significant results (9):

A 1979 study by Houtsmuller and associates reported a significant reduction in DR progression among participants on an unsaturated fat diet rich in linoleic acid, compared to those on a saturated fat diet. In contrast, a later study by Howard-Williams and colleagues reported no significant differences in incident DR between compliers of a modified fat diet (high PUFA to saturated fat ratio) and those on a low carbohydrate diet (low PUFA to saturated fat ratio).

In short, I could not find any credible replication of the Houtsmuller trial.

Conclusions

Cochrane's saturated fat review has many flaws (10), but the inclusion of the Houtsmuller trial ranks as a major one. It also shows that Cochrane — an organization whose mission is to "produce trusted synthesized evidence" — cannot always be trusted. Thus, systematic reviewers might be wise to follow the words of Richard Smith, former editor of the British Medical Journal (11):

We have now reached a point where those doing systematic reviews must start by assuming that a study is fraudulent until they can have some evidence to the contrary.

Regardless of whether the poor-quality Houtsmuller trial is fraudulent or not, meta-analyses should be based on the highest-quality evidence available. Anything less is not acceptable.


References

1) Hooper, L., Martin, N., Jimoh, O. F., Kirk, C., Foster, E., & Abdelhamid, A. S. (2020). Reduction in saturated fat intake for cardiovascular disease. Cochrane database of systematic reviews, (8). (https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD011737.pub3/epdf/full)

2) Hooper, L., Al‐Khudairy, L., Abdelhamid, A. S., Rees, K., Brainard, J. S., Brown, T. J., ... & Deane, K. H. (2018). Omega‐6 fats for the primary and secondary prevention of cardiovascular disease. Cochrane Database of Systematic Reviews, (7).

3) Heap, M. Skeptical Adversaria. (A similar statement can be found in https://www.tandfonline.com/doi/abs/10.1080/J157v02n02_08)

4) Sacks, F. M., Lichtenstein, A. H., Wu, J. H., Appel, L. J., Creager, M. A., Kris-Etherton, P. M., ... & Van Horn, L. V. (2017). Dietary fats and cardiovascular disease: a presidential advisory from the American Heart Association. Circulation, 136(3), e1-e23.

5) Ramsden, C. E., Zamora, D., Majchrzak-Hong, S., Faurot, K. R., Broste, S. K., Frantz, R. P., ... & Hibbeln, J. R. (2016). Re-evaluation of the traditional diet-heart hypothesis: analysis of recovered data from Minnesota Coronary Experiment (1968-73). bmj, 353.

6) Houtsmuller, A. J., van Hal-Ferwerda, J., Zahn, K. J., & Henkes, H. E. (1981). Favorable influences of linoleic acid on the progression of diabetic micro-and macroangiopathy in adult onset diabetes mellitus. Progress in lipid research, 20, 377-386.

7) Hamley, S. (2017). The effect of replacing saturated fat with mostly n-6 polyunsaturated fat on coronary heart disease: a meta-analysis of randomised controlled trials. Nutrition journal, 16(1), 1-16.

8) Howard-Williams, J. U. L. I. A., Patel, P., Jelfs, R., Carter, R. D., Awdry, P., Bron, A., ... & Hockaday, T. D. (1985). Polyunsaturated fatty acids and diabetic retinopathy. British Journal of Ophthalmology, 69(1), 15-18.

9) Wong, M. Y., Man, R. E., Fenwick, E. K., Gupta, P., Li, L. J., van Dam, R. M., ... & Lamoureux, E. L. (2018). Dietary intake and diabetic retinopathy: a systematic review. PLoS One, 13(1), e0186582.

10) https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD011737.pub3/detailed-comment/en?messageId=320249873

11) https://blogs.bmj.com/bmj/2021/07/05/time-to-assume-that-health-research-is-fraudulent-until-proved-otherwise/