Cochrane saturated fat reviews
Executive summary
* This week’s note examines the latest Cochrane report to be published on saturated fat and health outcomes (mortality and heart disease).
* The first Cochrane report on dietary fat was published in 2000 by Hooper et al. The same lead researcher, with varying colleagues, has produced the subsequent Cochrane dietary fat reports (2011, 2015 and 2020).
* The 2000 and 2011 reports were focused on dietary fat. The 2015 and 2020 reports were focused on saturated fat. This reflected changes in the 2015 US dietary guidelines, which became less concerned about total fat and more concerned about saturated fat.
* None of the four Cochrane reports has ever found anything for any dietary fat intervention (reduced or modified, total or saturated) for anything related to mortality. Not all-cause mortality, or cardiovascular disease (CVD) mortality, or coronary heart disease (CHD) mortality, or fatal heart attacks.
* None of the four Cochrane reports has ever found anything for any dietary fat intervention (reduced or modified, total or saturated) for anything related to CHD events, strokes or non-fatal heart attacks.
* The only claim that has ever been made by the Cochrane reports, in 20 years, is that long-term trials suggested that reducing dietary saturated fat reduced the risk of cardiovascular events. When this finding has been further examined – e.g. looking at trials that actually reduced saturated fat, rather than those that intended to, the findings have ceased to be significant.
* There have been no findings against total fat ever, and there have been no findings against saturated fat that have withstood scrutiny (as reported in the Cochrane publications themselves).
* There is nothing new in the 2020 report. This was admitted on P285 of 294. The other 293.75 pages were unnecessary.
* It is important to note that the two dietary fat guidelines (no more than 30% of calories in the form of total fat and no more than 10% of calories in the form of saturated fat), introduced in 1977/1983, are not evidence based and never can be evidence based. For something to be evidence based, the evidence must come first and there was no evidence for either guideline at the time they were introduced. That was the main finding from my PhD research and publications.
* Four other teams of researchers have examined fat and mortality and have also found nothing whatsoever against total fat, or saturated fat, for any combination of fat reduction or modification for any health outcome examined.
* The demonisation against saturated fat continues; the lack of evidence against saturated fat also continues. This is now a vendetta and it needs to stop.
Introduction
A report was published on May 19th called “Reduction in saturated fat intake for cardiovascular disease” (Ref 1). It’s important because it’s a Cochrane review and Cochrane publications are seen as the gold standard in academic research. My personal view is that Cochrane tainted its reputation with the treatment of Dr Peter Gøtzsche, who was a co-founder of the Cochrane Collaboration and the former leader of the Nordic Cochrane Center in Copenhagen, Denmark. Peter was ousted from the Cochrane board in 2018 and he documented the events related to this in his book called “Death of a Whistleblower and Cochrane’s Moral Collapse.”
Notwithstanding this, Cochrane reviews are still as good as we get and so this one deserves attention. The Cochrane reviews have also focused on systematic review and meta-analysis of randomised controlled trials, and not population studies, which also adds to the value of the publications. This latest one has already been touted as “The [fat] War is Over: The New Report Finds Sat Fat Causes Heart Disease” (Ref 2). That article was written by one of the vegan leaders with no sense of irony given that the single food highest in saturated fat on the planet is coconut oil – pure vegan. Anyway, let’s take a look at this latest report, but first, the background… (This note will serve as a definitive account of all the Cochrane reviews on dietary fat – total and saturated).
Background
The first thing to note is that the 2020 report is not new, and it says nothing new. This is the fourth substantive Cochrane review of dietary fat – it’s the second one to emphasise saturated fat in the title. The lead author on all of these is Dr Lee Hooper, who works at the University of East Anglia. Hooper first published a Cochrane systematic review of dietary fat studies in 2000 (Ref 3). This paper was called “Reduced or modified dietary fat for prevention of cardiovascular disease.” It was updated in two publications in 2001. First in one called “Reduced or modified dietary fat for preventing cardiovascular disease“ (Ref 4). Second in one called “Dietary fat intake and prevention of cardiovascular disease: systematic review” (Ref 5).
The 2001 paper was the definitive Cochrane report until 2011 when a revised review was conducted and published. This one was called “Reduced or modified dietary fat for preventing cardiovascular disease. Cochrane database of systematic reviews” (Ref 6). This was also updated the following year, 2012, with the same title (Ref 7).
The 2011/2012 report was then updated in 2015 with a report called “Reduction in saturated fat intake for cardiovascular disease” (Ref 8). Please note that dietary fat was dropped and replaced with saturated fat in the title and focus of this review. This reflected the 2015 US Dietary Guidelines, which went quiet on total fat, but continued the war on saturated fat.
I was particularly amused by the 2015 report because I might have had a hand in it being done. The first major paper from my PhD was published in February 2015 and it ‘went viral’, as the saying goes (Ref 9). It became the 64th most impactful paper in any discipline in 2015 (Ref 10). It concluded that there was no evidence base for the introduction of the two dietary fat guidelines introduced in the US in 1977 and in the UK in 1983 (no more than 30% of calories in the form of fat and no more than 10% of calories in the form of saturated fat).
The powers-that-be might have thought that this needed to be countered. The 2015 Cochrane report opened by referencing my paper, in the first sentence under the heading “Why it is important to do this review.”
Finally, we have the 2020 report published on May 19th. Let’s look at each of the four clusters of reports: 2000/2001; 2011/2012; 2015 and 2020…
The 2001 report
Given that all of the Cochrane dietary fat reviews focus on randomised controlled trials (RCTs), it is important to note that there were only six such trials that had reported on all-cause mortality and heart disease mortality at the time the guidelines were introduced. The six trials pooled together found nothing against fat for all-deaths or heart deaths (my first PhD paper) and none of the six trials called for any recommendations to be made against any fat. On the contrary, three issued cautions that their dietary fat intervention (reduced and/or modified fat intake) may have been harmful. It will always remain the fact, therefore, that dietary fat guidelines will never be based in evidence because, for that to happen, the evidence must come first. This point seems to escape most public health advisors who continue to promote these guidelines, as if they are evidence based.
Some famous dietary fat trials were published in the 1980s and 1990s (DART, Minnesota Coronary Survey, and STARS). All of these were available to Hooper and her team for the 2000/2001 report. I included all of these in my follow-up paper, which found that there would be no more evidence for introducing the two dietary fat guidelines today than there was at the time of their introduction (Ref 11). The only major RCT with a dietary fat intervention published this century was the Women’s Health Initiative (2006) (Ref 12). This was an examination of the low-fat/ high-carb (fruit/wholegrain) diet in post-menopausal American women.
With those RCTs available, Hooper et al 2001 reviewed RCTs lasting at least six months and reported three conclusions:
1) There was no significant effect on total mortality.
2) There was no significant effect on CVD mortality, but this was reported as “a trend towards protection from cardiovascular mortality“, which demonstrated bias. (The rate ratio and 95% confidence interval were reported as 0.91, (0.77 to 1.07) – the latter includes 1.0, the line of no effect, and thus is not significant).
3) There was significant protection from cardiovascular events (rate ratio 0.84, 95% CI 0.72 to 0.99). This was immediately followed by this sentence: “The latter became non-significant on sensitivity analysis.” (Every use of the word significant in this note means statistical significance, unless otherwise specified).
There were no findings, therefore.
The 2011 report
Hooper et al 2011 reviewed RCTs lasting at least six months, extending the investigation beyond total mortality and cardiovascular disease (CVD) mortality to include CVD morbidity and individual outcomes including myocardial infarction (heart attack), stroke, and cancer diagnoses. The review included 21 studies involving 71,790 people. There were 12 main results – 11 of these were non-significant i.e. they found nothing. This is never press-released; it should be the headline.
1-4) There were no significant results for total mortality from all RCTs; from modifying fat; from reducing fat; or from reducing and modifying fat.
5-8) There were no significant results for CVD mortality from all RCTs; from modifying fat; from reducing fat; or from reducing and modifying fat.
9-11) There were no significant results for CVD events from modifying fat; from reducing fat; or from reducing and modifying fat.
12) There was one significant result for CVD events from all RCTs. When all RCTs were examined together, the risk ratio (RR) for CVD events from meta-analysis was 0.86 (95% CI 0.77 to 0.96).
Eight sensitivity tests were undertaken – first excluding studies with systematic differences in care between the intervention and control arms and second excluding studies with systematic differences in diet other than fat. The results ceased to be significant in all eight tests. I.e. when genuine randomised controlled trials were examined, nothing was found.
There were no findings, therefore.
(There was no effect of altering dietary fat intake on cancer deaths or cancer diagnoses, for completeness).
The 2015 report
Hooper et al 2015 reviewed RCTs with a minimum of 24 months duration. The review included 15 studies involving approximately 59,000 people and was focused on the impact of saturated fat reduction on dying, heart disease and stroke. There were eight main results – seven of these were non-significant i.e. they found nothing. This is never press-released; it should be the headline.
1) There was no significant effect from reducing saturated fat on total mortality.
2) There was no significant effect from reducing saturated fat on CVD mortality.
3) There was no significant effect from reducing saturated fat on CHD (coronary heart disease) mortality.
4) There was no significant effect from reducing saturated fat on fatal heart attacks.
5) There was no significant effect from reducing saturated fat on non-fatal heart attacks.
6) There was no significant effect from reducing saturated fat on CHD events.
7) There was no significant effect from reducing saturated fat on strokes.
The one significant finding, again, was for CVD events where it was claimed that the risk ratio (RR) for CVD events from meta-analysis was 0.83 (95% CI 0.72 to 0.96). Dr Trudi Deakin alerted me to the fact that Table 8 in Hooper et al (2015) ran a sensitivity analysis for RCTs that did actually reduce saturated fat – excluding studies that aimed to reduce saturated fat but didn’t – and the one finding for CVD events ceased to be significant.
There were no findings, therefore.
The 2020 report
Hooper et al 2020 reviewed RCTs with a minimum of 24 months duration. The review included 15 studies involving approximately 59,000 people and was focused on the impact of saturated fat reduction dying, heart disease and stroke. This was essentially the same as the 2015 review therefore – same number of studies, same number of people, same length of study, same focus.
There were the same eight main results. Again, seven of these were non-significant i.e. they found nothing. Again, this is never press-released; it should be the headline.
1) There was no significant effect from reducing saturated fat on total mortality.
2) There was no significant effect from reducing saturated fat on CVD mortality.
3) There was no significant effect from reducing saturated fat on CHD mortality.
4) There was no significant effect from reducing saturated fat on fatal heart attacks.
5) There was no significant effect from reducing saturated fat on non-fatal heart attacks.
6) There was no significant effect from reducing saturated fat on CHD events.
7) There was no significant effect from reducing saturated fat on strokes.
The one significant finding, again, was for CVD events where it was claimed that the risk ratio (RR) for CVD events from meta-analysis was 0.79 (95% CI 0.66 to 0.93). Analysis 1.38 (of a number of analyses), on Page 159, ran a sensitivity analysis for RCTs that did actually reduce saturated fat – excluding studies that aimed to reduce saturated fat but didn’t – and the one finding for CVD events ceased to be significant.
There were no findings, therefore.
The 2015 vs 2020 CVD events difference
The 2015 paper claimed that the risk ratio for CVD events and reduction in saturated fat was 0.83 (95% CI 0.72 to 0.96). It was claimed that 13 studies, involving 53,300 people, were used for that finding. The 2020 paper claimed that the risk ratio for CVD events and reduction in saturated fat was 0.79 (95% CI 0.66 to 0.93). It was claimed that 12 studies, involving 53,300 people, were used for that finding.
The reason for the same investigation reaching different results in 2015 and 2020 was that the Women’s Health Initiative was treated as two studies in 2015 (women with or without CVD) and it was treated as one study in 2020. You might think that everyone with CVD already should have been omitted from a study looking at people who develop CVD – and you would be right. However, virtually every dietary fat intervention was undertaken on people who already had heart disease because the goal of the RCT was to stop a second event – what we call secondary prevention. If we left out studies on sick people and left out studies on single sexes (mostly just men, but the WHI was just women), we would have precisely one RCT left to examine! The Minnesota Coronary Survey (Ref 13). And even this was conducted on inpatients in mental institutions and so we have precisely zero RCTs available for examination that actually apply to any of us in general populations. The Minnesota Coronary Survey, by the way, was excluded by Hooper et al, for not meeting the 24-month duration criteria.
Even if anything had been found, it would have lacked generalisability.
The totality of the evidence
There were five main papers published from my PhD (Ref 14). Two examined the evidence available (from RCTs and from population studies) at the time the dietary fat guidelines were introduced (1977-1983). Two examined the evidence available (from RCTs and from population studies) at the time I was doing the PhD (up to 2016). All four found no evidence to support either guideline. The fifth paper from my PhD reviewed other systematic reviews and meta-analyses on this subject to examine the totality of the evidence (Ref 15).
From 6 different teams of researchers (including mine and Hooper’s), examining mortality, in up to 32 studies and over 500,000 people there were 38 different findings (Ref 16). The findings covered variations of reduced fat and/or modified fat on all-cause mortality, CVD mortality, CHD mortality, events and so on. There were 35 non findings. All that research, from all those team, on all those studies and people, and the vast majority of findings were non-findings. Of the three findings, one was the Chowdhury finding against trans-fats – no disagreement there. The other two were the Hooper 2011 and 2015 findings against CVD events alone, which then failed sensitivity tests.
There’s a section called “What’s New?” on p285 of 294 in the latest Cochrane report. In a couple of lines, this summarises that nothing is in fact new:
“29 December 2019 – New search has been performed.
9 January 2020 – New citation required but conclusions have not changed.”
We didn’t need the other 293.75 pages.
There’s no evidence against saturated fat. There never was. No new large-enough, long-enough dietary fat trials will be undertaken in the future (they’re too expensive and they can’t risk the cost knowing the answer already) and so there will be no new evidence to examine.
It makes no sense that there would be evidence against saturated fat per se, as the only food that doesn’t contain saturated fat (even a trace) is sucrose! Nature really didn’t put saturated fat in everything from fish to fruit and beef to broccoli to kill us. She really didn’t.
References
Ref 1: Hooper et al. “Reduction in saturated fat intake for cardiovascular disease.” Cochrane Database Syst Rev. 2020 https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD011737.pub2/full
Ref 2: https://thebeet.com/the-big-fat-war-is-over-saturated-fat-causes-heart-disease-says-dr-joel-kahn/
Ref 3: Hooper et al. “Reduced or modified dietary fat for prevention of cardiovascular disease.” Cochrane Database Syst Rev. 2000. https://pubmed.ncbi.nlm.nih.gov/10796866/?
Ref 4: Hooper et al. “Reduced or modified dietary fat for preventing cardiovascular disease.” Cochrane Database Syst Rev. 2001. https://pubmed.ncbi.nlm.nih.gov/11687015/
Ref 5: Hooper et al. “Dietary fat intake and prevention of cardiovascular disease: systematic review.” BMJ. 2001. https://pubmed.ncbi.nlm.nih.gov/11282859/?
Ref 6: Hooper et al. “Reduced or modified dietary fat for preventing cardiovascular disease.” Cochrane database of systematic reviews. 2011. https://pubmed.ncbi.nlm.nih.gov/21735388/
Ref 7: Hooper et al. “Reduced or modified dietary fat for preventing cardiovascular disease.” Cochrane database of systematic reviews. 2012. https://www.cochranelibrary.com/cdsr/doi/10.1002/
Ref 8: Hooper et al. “Reduction in saturated fat intake for cardiovascular disease.” Cochrane Database of Systematic Reviews. 2015. https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD011737/full
Ref 9: Harcombe et al. Evidence from randomised controlled trials did not support the introduction of dietary fat guidelines in 1977 and 1983: a systematic review and meta-analysis. Open Heart 2015.
Ref 10: https://www.altmetric.com/top100/2015/
Ref 11: Harcombe et al. Evidence from prospective cohort studies does not support current dietary fat guidelines: A systematic review and meta-analysis. Br J Sports Med 2016
Ref 12: Howard et al. Low-fat dietary pattern and risk of cardiovascular disease: the Women’s Health Initiative Randomized Controlled Dietary Modification Trial. JAMA 2006
Ref 13: Frantz et al. Test of effect of lipid lowering by diet on cardiovascular risk. The Minnesota Coronary Survey. Arteriosclerosis, Thrombosis, and Vascular Biology 1989.
Ref 14: https://twitter.com/zoeharcombe/status/955436360541798400
Ref 15: Harcombe Z. Dietary fat guidelines have no evidence base: where next for public health nutritional advice? Br J Sports Med 2016
Ref 16: The four research teams and their main papers (other than mine and Hooper’s) were:
– Chowdhury et al. Association of Dietary, Circulating, and Supplement Fatty Acids With Coronary Risk: A Systematic Review and Meta-analysis. Ann Intern Med. 2014.
– Skeaff & Miller. Dietary fat and coronary heart disease: summary of evidence from prospective cohort and randomised controlled trials. Ann Nutr Metab. 2009.
– Schwingshackl & Hoffmann. Dietary fatty acids in the secondary prevention of coronary heart disease: a systematic review, meta-analysis and meta-regression. BMJ Open. 2014.
– Siri-Tarino et al. Meta-analysis of prospective cohort studies evaluating the association of saturated fat with cardiovascular disease. Am J Clin Nutr. 2010.
What do you take from this report : “Effects on Coronary Heart Disease of Increasing Polyunsaturated Fat in Place of Saturated Fat: A Systematic Review and Meta-Analysis of Randomized Controlled Trials”
https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1000252
which found that “.. consuming PUFA (polyunsaturated fat) in place of SFA (saturated fat) reduces CHD (coronary heart disease) events in RCTs (randomized controlled trials)”
Hi John
I included that in my PhD with the following passage:
“Mozaffarian et al examined the RCTs that increased polyunsaturated fat (PUFA) in place of saturated fat (SFA) (Mozaffarian, Micha and Wallace, 2010). They concluded that “a shift toward greater population PUFA consumption in place of SFA would significantly reduce rates of CHD” (p.1). The authors were criticised (Ravnskov et al., 2014) for having excluded The Rose Corn Oil Trial (Rose, Thomson and Williams, 1965), and The Sydney Diet Heart Study (Woodhill et al., 1978), which observed greater mortality from CHD and all-causes in the intervention groups. This meta-analysis was also the only one to include The Finnish Mental Hospital Study (Miettinen et al., 1983; Turpeinen et al., 1979). This cross-over trial has been excluded in all other systematic reviews for not being randomised or controlled and because a cross-over trial design is inappropriate for the study of a long term condition.”
Best wishes – Zoe
Excellent info which i can use to convince my clients to eat more whole eggs with butter.
Hi Coachbart
Eggs have more monounsaturated fat than saturated remember! ;-)
Best wishes – Zoe
Zoe
This is probably off-topic, but I read a lot about how bad seed oils are for one. Since olive oil is a seed oil, and we use it a lot, is there any researched basis for this?
Incidentally I found your article above excellent
Hi Chris
Many thanks for your kind words.
I think of Olive oil as a fruit oil, rather than a seed oil. I think of sunflower seed, rapeseed, cotton seed as the seed oils. Olive oil is fine (in my opinion) to use as a salad dressing and for stir fry kind of cooking. It will mutate less at high temperatures than the seed oils, as they are higher in polyunsaturated fats.
I’m not one of the people who thinks that olive oil is anything special. It’s nutritionally poor (low in nutrients compared to energy) but it has its uses!
I hope this helps
Best wishes – Zoe
Excellent! Thank you, Zoe! You are an absolute rockstar!
Thanks Brad :-)
Bravo, Zoe. I have been sidetracked by Covid-19 and am just as frustrated in that field that no-one is looking at the right things, and indeed doing the wrong ones. If you need a break from fats have a look at https://bamjiinrye.wordpress.com…
Keep up the good work. Truth outs eventually!
Hi Andrew
Lovely to hear from you and great blog. You’ve been prolific!
I’ve tagged it to take a good look when I allow myself surf time.
Best wishes – Zoe
What really bothers me about this report is that the author’s summary seems to claim a benefit from reducing saturated fat, whereas apparently the data don’t say any such thing.
Is this just them being politically afraid to make a bold statement or do they not accept that the significance of the results fall away under scrutiny or are they simply biased?
It seems they leave the door open for the various food lobbies to read whatever they wish into this.
Richard
Hi Richard
Remember teams get paid for these reviews – handsomely no doubt – and if you go back to the same team, you know what you’re going to get – one finding (apparently) against CVD events alone, which falls away in the nearly 300 pages that no-one will read. Except me, Trudi Deakin, Nina Teicholz, George Puddleg and Nicolai Worm!
Funnily enough, we never get asked to do a review!
Best wishes – Zoe
p.s. and yes – I think the unsat is good,sat is bad message does leave the door open for the fake food industry :-(
Thanks for your reply Zoe.
One other question comes to mind. What is a CVD event that isn’t included in the list of 7 null-findings in your piece? Do possible events include elective medical procedures (stents and the like)?
Regards
Richard
Hi Richard
Hooper et al won’t define CVD events – they will just use whatever was reported as a CVD event in each RCT. Given that many of these date back to the 1960s, there weren’t stents etc, but the more recent RCTs did include these.
Taking the most recent – WHI – Howard included stents as a CVD event and other coronary interventions. You’re right that you then have a confounder. The women on the low fat high carb diet would have consumed way more plant sterols than the control group. Their cholesterol would likely be lower as a result. Two women arrive at the doctor complaining of chest pains – the one with the higher cholesterol will more be suspected of needing treatment than the one with the lower cholesterol.
Best wishes – Zoe
Ah, my Monday fix – and an excellent one at that :-).
“There is nothing new in the 2020 report. This was admitted on P285 of 294. The other 293.75 pages were unnecessary.”
Thank you
Thank you Philip (blush!) and for your support for what I do – it’s so appreciated :-)
Best wishes – Zoe