Is Saturated Fat Bad For You?

​For over 40 years, conventional wisdom has been that saturated fat causes heart disease and therefore should be avoided or reduced.

The targets for reduction have gone progressively down over that time from < 10% to the current recommendations from the American Heart Association of 5-7%.

During this time, the rate of cardiovascular disease (CVD) has decreased to approximately 1/3 of their 1960's levels.  While there are many factors (decline in smoking, better control of hypertension, use of statin drugs, and the timely use of blood thinners in acute myocardial infarctions) some cardiologists point to this decline as proof that the nutritional recommendations made in the late 1970's to reduce fat intake and specifically target saturated fat is a big part of this success.

More recently, prominent experts have begun to challenge this.  

"Is saturated fat bad for you?", remains one of the most contentious and confusing questions in medicine today.

I would like to unpack this for you, give you some background, and lay out a practical non-dogmatic approach for you to follow personally.

Background - The Diet:Heart Hypothesis"
In the 1960's several observations were combined to form the diet:heart hypothesis which stated:
Lowering cholesterol by replacing saturated fat with poly unsaturated fat (PUFA) from vegetable oil would:

• diminish the deposition of cholesterol in the arterial wall, 
• slow progression of atherosclerosis, 
• reduce cardiovascular disease and 
• improve survival

 
This hypothesis, which has been the dominant paradigm for cardiology over the past 40 years was based on:

• evidence from randomized controlled studies that showed replacement of saturated fat in the diet with PUFA from vegetable oil lowers total cholesterol and LDL cholesterol
• observational evidence linking serum cholesterol to coronary artery disease and mortality

 
The logic of the diet:heart hypothesis:

• decreasing saturated fat decreases total cholesterol and LDL (A) and lowering cholesterol and LDL (B) lowers coronary artery disease and mortality (C) (A+B ---> C)

Then:

• decreasing saturated fat (A) lowers coronary artery disease and mortality (C) (A ---> C)

 
The problem with the Diet:Heart hypothesis is that there has been no solid evidence to support the logical leap (A--->C)

​The original evidence to support the notion that decreasing saturated fat lowers coronary artery disease came from epidemiological studies in the 1960's that demonstrated a positive correlation between national levels of dietary fat consumption, and specifically saturated fat, and mortality from heart disease.  The most famous (or infamous) study was performed by the legendary Ancel Keyes and was called the "7 Country Study".   

This was an observational study based on fairly low-quality data:  food diaries and public health records on cause of death.  It also was not originally based on 7 countries - Keyes actually reviewed 22 countries - and when the whole dataset is reviewed the correlation, while present is far weaker.  This has led to the suggestion that Keyes essentially cherry-picked the data to make the correlative conclusion stronger.  He also refused to look at outlier cultures with very high saturated fat consumption and low rates of heart disease: Masai, Inuit etc.

By today's standards the 7 Country Study would be considered deeply flawed, and in fact recent observational studies have shown different results. 

Most recently the PURE study, a large 18 country epidemiological cohort study that followed 135,335 people for an average of 7.4 years, demonstrated that high carbohydrate intake was associated with higher risk of total mortality whereas total fat and saturated fat intake were related to lower total mortality.  Specifically saturated fat intake was not associated with increased risk of cardiovascular disease, heart attack, or cardiovascular disease mortality, and was actually associated with a decrease risk of stroke.

Of course, the challenge with these observational studies is that the best they can do is show correlation.

But correlation is not causation - my favorite example of this is the number of bathrooms one has correlates with net worth - the higher one's net worth the more bathrooms.  Now if we confuse correlation with causation, we could erroneously conclude that having more bathrooms leads to higher net worth and could advocate that individuals should build themselves new bathrooms to increase their wealth!

The gold standard for proving causation is the randomized controlled study.  This is where the diet:heart hypothesis really seems to fall apart.

No randomized controlled study has shown that replacement of saturated fat with vegetable oil significantly reduces coronary heart disease or mortality.

One randomized controlled study that attempted to test the causal role of saturated fat in heart disease was the Minnesota Coronary Experiment (MCE).

Conducted from 1968 - 1973, MCE was the largest (9570 people) and most rigorously executed trial of the diet:heart hypothesis, and the only one to ever include post-mortem assessment of coronary, aortic and cerebrovascular atherosclerosis grade and infarct status.

The MCE followed over 9000 institutionalized people living in state mental institutions and nursing homes and randomly assigned them to two groups: one that maintained the standard diet high in saturated fat and the other group in which half of the calories from saturated fat were replaced with vegetable oil and corn oil margarine - high in the Omega 6 PUFA linoleic acid.  Unlike observational studies, the MCE had detailed records of every meal administered to these subjects over a 56 month period. 

This study probably could never be repeated as today's ethics boards would not approve experimenting on institutionalized patients without consent, and it would be impossible to control diet to the extent that they were able to in this trial.

So what were the findings:
In keeping with the first part of the diet:heart hypothesis - replacing saturated fat with linoleic acid did lower cholesterol by an average of 14%, BUT this lowering of cholesterol DID NOT result in people living longer.

In fact, the lower the cholesterol the higher the rate of death (22% for every 0.75 mmol/L) and the vegetable oil group did not have fewer heart attacks or fewer atherosclerotic plaques.

So the MCE, the most rigorous trial ever done to test the diet:heart hypothesis essentially disproved the notion that decreasing saturated fat improves cardiovascular outcomes –it even suggested that increasing vegetable oil was associated with poor outcomes.

If this rigorous study was finished in 1973 and essentially disproved the diet:heart hypothesis, why did it not change the prevailing wisdom that saturated fat was bad?

It did not change minds because it was never published!

The investigators led by Ivan Frantz did not publish because the results were not what they expected and they felt that something must have been wrong with their data.  When part of the study was published in 1989 it only reported that the substitution of saturated fats with vegetable oils did not reduce the risk of heart disease or death. 

It was not until 2017, that the full data was analyzed and the true results were presented.  This study by Christopher Ramsden, was assisted by Ivan Frantz’s son Robert (a prominent Mayo clinic physician himself) who had found old computer tapes and documents in his father’s basement.  Ivan Frantz died in 2009.  The account of this discovery was the subject of a brilliant Malcolm Gladwell podcast “The Basement Tapes”
 
It was Ramsdem et al. who showed that when the whole data set was thoroughly reviewed that the MCE study results are counter to the diet:heart hypothesis and show that the replacement of saturated fat with vegetable oil increases coronary events and death.

While this was the first full reporting from the  MCE trial, a 2013 re-analysis of another 1960’s era landmark study – the Sydney Diet Heart Study (again by Ramsden) also showed that volunteers that replaced much of their saturated fat with polyunsaturated fats high in linoleic acid had a higher risk of death from heart disease.

In 2014, Chowdrury et al. reported a meta-analysis of 78 studies involving 650,000 people and concluded that there was no evidence that lowering saturated fat and increasing polyunsaturated fat intake decreases the risk of cardiovascular disease.

A landmark 2015 systematic review and meta-analysis of observational studies showed no association between saturated fat consumption and

1. all-cause mortality
2. coronary heart disease (CHD)
3. CHD mortality
4. ischaemic stroke or
5. type 2 diabetes in healthy adults

These studies in many ways disprove the diet:heart hypothesis as overly simplistic.  Focusing on evidence that saturated fat elevates LDL cholesterol misses the fact that increasing saturated fat in the diet also increases HDL cholesterol and lowers triglycerides. 

In recent years, the ratio of TG/HDL has been shown to be the best predictor of heart disease, superior to LDL cholesterol.  Ratios that included HDL: Total Cholesterol/HDL and LDL/HDL also are stronger predictors that LDL alone. 

Other evidence points to evidence that LDL particle size is likely more important than total LDL measurements.  Individuals with small, dense LDL particles (Phenotype B) are at greater risk than individuals with larger, more buoyant LDL particles (Phenotype A).

Interestingly  a high TG/HDL ratio correlates with the atherogenic Phenotype B – small, dense LDL particle size as well as insulin resistance.

Returning to the logic of the diet:heart hypothesis:

• decreasing saturated fat decreases total cholesterol and LDL (A) and lowering cholesterol and LDL (B) lowers coronary artery disease and mortality (C) (A+B ---> C)

Then:

• decreasing saturated fat (A) lowers coronary artery disease and mortality (C) (A ---> C)

From this we can see that the problem with the diet:heart hypothesis is that it ignores the effect of saturated fat on HDL and TG’s and assumes that the effect of LDL is all that there is to consider in terms of cardiovascular risk.

On the whole, this brings us to the answer to our question – Is saturated fat bad for you?

The overall evidence from these studies says probably not – but the real answer lies in your own individual response to increases or decreases in saturated fats, polyunsaturated fats, and carbohydrates.  Specifically, what we track at Wellness Garage are your HDL, tryiglycerides and LDL particle number – in addition to markers of inflammation and insulin resistance, and with this data we have a far better idea of what effect your diet is having and what risk you face from cardiovascular disease.

Our view is not conventional wisdom – but we believe that it does represent the emerging consensus view
At Wellness Garage our approach is personalized:

• have a clear-eyed look at all the evidence
• assess our clients fully
• understand their goals
• make a change
• measure the results
• understand what works for that individual

This is the basis of our Precision Health Tune-Up's.

If you think that you benefit from a personalized approach to cardiovascular risk - please reach out to us and book a free consultation, we can help you come up with a plan to address your health needs.