Butter is not your enemy. Decades of dogma got it wrong about saturated fat

News & Analysis

2nd November 2015

Challenging a medical or scientific paradigm can be seriously bad for your health, as any heretic, from Galileo to John Yudkin, can tell you.

The medical dogma surrounding saturated fat is a case in point. Until a few years ago it was considered beyond certainty that saturated fat — mainly found in meat and dairy — was the culprit behind heart disease. Guidelines published in 1981 (shown subsequently to be based on low-quality to no evidence) led to the golden age of low-fat high carbohydrate foods, consumption of which also coincided with a sharp increase in obesity, diabetes, cardiovascular disease and forms of cancer.

The paradigm is now being thoroughly weakened, with studies showing no correlation between saturated fat intake and risk of death, little to no correlation between cholesterol levels and mortality in otherwise healthy subjects, not to mention other studies showing that a higher fat intake in tandem with low-carb intake irrespective of the number of calories resulted in an improvement in various markers of inflammation and cardiovascular health.

Sadly, most of the medical profession have tunnel vision when it comes to assessing the impact of a particular drug, intervention or nutrient, especially when they already hold a common position on it. Unless a randomised controlled trial has been done, the thinking goes, any assertion is based on ‘low-quality or no evidence’ — something that did not seem to bother anyone when the original 1981 high-carb, low-fat nutritional guidelines were prepared.

Professor Frank Hu of Harvard University and colleagues published an interesting paper on saturated fat intake and risk of death from cardiovascular disease. The conclusion was that replacement of saturated fat with monounsaturated fat (in olive oil and nuts), polyunsaturated fat (fatty fish, flax seeds) or complex carbohydrates from an unrefined source (wholegrain bread, brown rice) was associated with fewer deaths from cardiovascular disease.

Supporters of the traditional dogma will, of course, trumpet this as further proof that saturated fat is bad for you; it may not have been a randomised trial, they will say, but it merely confirms what we already know.

Opponents will point to the fact that, as a non-randomised study, we cannot hold it up as ‘high-quality evidence’. Furthermore, they will make the point that the study merely underlines what we already know; not that saturated fat itself is bad, but that saturated fat of the type consumed by Americans is bad for you. Saturated fat in the US diet is primarily processed, derived from the consumption of processed meats, sausages, fatty cuts of meat and dairy. Why so? Because it has been demonstrated that the meat of animals fed grain as opposed to grass (standard US agricultural practice) has higher tissue levels of inflammatory mediators, something which may impact human health more than we give it credit.

The Hu study was a large one, but the question must be asked: are these findings reflected outside the US? Not according to a study from Denmark, which found that substituting carbohydrates, albeit of the high GI index kind (eg grains), resulted in higher rates of heart attack, but that saturated fat intake itself seemed to slightly decrease the risk of heart attacks.

Moreover, a recent meta-analysis (the highest level of evidence) which has generated some controversy found that there is insufficient evidence to support the supposition that saturated fat should be replaced by polyunsaturated fat for cardiovascular health.

Confused yet? Based on the available evidence, one can state with confidence that the healthiest forms of saturated fats are the ones to be found in grass-fed meat, grass-fed butter, fish, coconut oil, nuts, olive oil, cocoa butter, eggs and avocados, as well as full-fat dairy.

There is no scientific, evidence-based reason to restrict healthy fat consumption, unless of course you wish to uphold the belief that ‘fat makes you fat’ or ‘fat is bad for you’ in the way that experts 400 years ago believed that the world was flat.

Vegetable oils like sunflower or canola are high in omega-6 fatty acids which can contribute to vascular inflammation and, in my opinion, should be avoided.

If a person plans to increase the amount of healthy fat they eat, they really should think seriously about decreasing their overall carbohydrate consumption simply to avoid the synergy between higher insulin levels and dietary fat that will result in extra insulation.

Personally, I believe the key to navigating the minefield that is current science on fats, cardiovascular health, cholesterol and optimum nutrition for health is self-education. This area is too important to be left to medics or nutritionists, neither of whom have a good track record in giving accurate advice.

There are some excellent books written for a lay audience that do not dumb down the science (Good Calories, Bad Calories by Gary Taubes and Cholesterol Clarity by Jimmy Moore are an excellent place to start). The medical and nutrition professions have misled the public for the last 40 years on this subject; the day of reckoning is fast approaching.


  • trefjon1

    Very interesting article and a shot in the arm for Welsh sheep farmers and the dairy inustry. If there is a deeper message, it is that the sensationalising and celebrity led infantilisation of the diet debate masks its sheer complexity.

    • Tarek

      Thank you very much; I totally agree

    • Matthew Curry Pye

      Mate this article is the exact same thing, its just the pendulum of public opinion is swinging the other way

  • Recommended!

  • Peggy Holloway

    Excellent article! The two books you recommend are also two that I often cite when others are looking for “what to red.”

  • Mark Bailey

    “in the way that experts 400 years ago believed that the world was flat.”

    No they didn’t. Just like Bishops didn’t argue over how many angels could dance on the head of a pin, etc., etc.. We can blame Andrew Dickson Wright – the C19th answer to Richard Dawkins – for that