by Marion Nestle
Mar 22 2010

Saturated fat vs. heart disease: current state of the science

Despite recent publications finding no correlation between intake of saturated fat and coronary heart disease (CHD) – see, for example, the recent meta-analysis in the American Journal of Clinical Nutrition – the debates over the role of saturated fat continue.

In that same issue of the Journal, another study says that reducing saturated fat only works if you replace it with something better.  If you replace saturated fat with carbohydrates, the effects on heart disease will be worse.

The fat story is not simple (in What to Eat, I explain the biochemistry of food fats in the chapter on fats and oils and in an appendix).  The main reason for the complexity is that different kinds of fats do not occur separately in foods.

Without exception, food fats are mixtures of  three kinds of fatty acids: saturated (no double bonds and solid at room temperature), monounsaturated (one double bond), and polyunsaturated (two or more double bonds and liquid at room temperature).  Food fats just differ in proportions of the three kinds.

Meat, dairy, and egg fats generally are more saturated.  Plant fats and oils are generally more unsaturated.

How to make sense of the saturated fat story? An expert panel from WHO and FAO just produced a new review of the evidence.  The panel evaluated CHD morbidity and mortality data from epidemiological studies and controlled clinical trials.  It found:

  • Convincing evidence that replacing saturated fat with polyunsaturated decreases the risk of CHD.
  • Probable evidence that replacing saturated fat with largely refined carbohydrates (starch and sugar) has no benefit and even may increase the risk of CHD.
  • Insufficient evidence relating to the effect on the risk of CHD of replacing saturated fat with monounsaturated fats or whole grain carbohydrates, but a trend suggesting that these might decrease CHD risk.
  • Possible positive relationship between saturated fat and increased risk of diabetes.
  • Insufficient evidence for establishing any relationship of saturated fat with cancer.

The panel’s recommendations:  (1) Replace saturated fat with polyunsaturated fats (omega-3 and omega-6) in the diet, and (2) Limit saturated fat to 10% of daily calories or less.

Translation: Eat less animal fat and replace it with vegetable fats.

Historical note: These are precisely the same recommendations that have been standard in the U.S. for at least fifty years.  This was good advice in the late 1950s.  It is still good advice.

UPDATE, March 22,2011:  Another major review has just come to precisely the same conclusions, this one from an international expert panel.  It also suggests areas for future research.  See American Journal of Clinical Nutrition 2011;93:684-88.

  • JimB

    I have long wondered about the data contamination associated with fats. At one time, all fats were claimed equally detrimental. Later, there were distinctions made. For example, vegetable oils were blessed and animal fats horrible.Early in the 20th century, hydrogenated and partially hydrogenated vegetable oils were taken as healthy alternatives to or substitutes for saturated fats from animals.

    Then it was discovered that these hydrogenated vegetable oils were actually WORSE biological actors than were saturated fats.

    I suspect that numerous studies on “Saturated Fats” were actually conducted with “Hydrogenated Vegetable Oils”….. as they appeared to serve the same functions in foods.

    I know that some cholesterol feeding studies were conducted with dried eggs as a source of dietary cholesterol…. being a cheap industrial product and easily stored. Later knowledge revealed that the dried egg product actually contained oxidized cholesterol substances, and that these oxidized cholesterol products did not behave the same as more natural unoxidized food products. Some of the cholesterol feeding experimental results had to be discarded because not all Cholesterols are made equally active chemically.

    So, in the years since about 2000, have the “saturated fat” research articles and conclusions of the actions of “saturated fat” been made pure by discarding all of those “saturated fat” samples and results that were either pure or blends of “Trans Fats”?

    In other words, has the “Saturated Fat” data set for CVD been cleansed of these substitute or alternative “Trans Fats” which are evidently known to be even worse biological actors than Sat Fat for CVD?

    I remember the most difficult part of doing experiments was admitting that an unknown variable had crept into the experimental plan, and we would have to trash a lot of work. There was always the group that wanted to argue that the contamination was “small” or “ignorable” or “we can adjust for that”. Many of those adjustments were intended to mainly be able to go ahead as if nothing had really happened from the faulty experimental control.

    I have never heard of anyone doing this cleansing or re-review of data to see if the old published conclusions can still be accepted in view of the contamination of Sat Fat and Trans Fat as the Bad Fats in potentially many experiments reported in the literature.

    This might be a can of worms. Since people were confused on this issue for over 50 years, it seems likely that these things got mixed up a time or two. or three or four, or five…… or more..

  • Wait a moment Dr. Nestle, you say that replacing animal fats with plant fats is good advice? Why then did heart disease rates increase after the US government prescribed such advice? Just as the amount of saturated fat in the U.S. diet decreased over the 20th Century, what increased were industrial seed oil (canola, soybean, cottonseed, corn, etc) consumption and heart disease.

    It’s true that this is only a correlation, but it really merits further investigation.

    The establishment is oft to recommend more unsaturated plant fats: i.e. polyunsaturated “vegetable” oils. But the Weston A. Price Foundation notes that “in test animals, diets high in polyunsaturates from vegetable oils inhibit the ability to learn, especially under conditions of stress; they are toxic to the liver; they compromise the integrity of the immune system; they depress the mental and physical growth of infants; they increase levels of uric acid in the blood; they cause abnormal fatty acid profiles in the adipose tissues; they have been linked to mental decline and chromosomal damage; and they accelerate aging.”

    Section entitled “The Dangers of Polyunsaturates”

  • As an addendum, archeological and anthropological evidence supports the idea that humans evolved to thrive on animal fats. It would then readily follow that eating large amounts of polyunsaturated fats, which are needed in very small amounts by our bodies, is harmful to humans.

    The blogger Stephan Guyenet unpacks this topic with scientific rigor in this blog post:

    I strongly encourage you, Dr. Nestle, to reconsider the idea that animal fats are bad for us. The reality is likely more complicated.

  • Margeretrc

    I agree with Colin and Jim. The recommendations of the AJCN, the WHO, and others are not based on science, they are based on the usual lipophobic nonsense that has been perpetrated on the American (and now World) population for 50 years. It’s time to stop.

  • Gary

    Colin’s wrong – Fat(and meat) consumption INCREASED in the 20th century, along with heart disease. Humans haven’t evolved, and haven’t adapted to un-natural eating (obviously).

    We’re still primates and still thrive on fruit and veggies

  • Today, I went to the beachfront with my children.
    I found a sea shell and gave it to my 4 year old daughter and said “You can hear the ocean if you put this to your ear.” She placed the shell to her ear and screamed.
    There was a hermit crab inside and it pinched her ear.
    She never wants to go back! LoL I know this is completely off topic
    but I had to tell someone!