Butter and Heart Attacks in Sweden

butter and heart disease over the years

This graph from Andreas Eenfeldt (via Mark’s Daily Apple) shows heart attacks (left axis) and butter consumption (right axis) in Sweden over the last quarter-century.  Heart attacks have been going down, at least in men. What interests me is that when butter consumption suddenly increased, starting in 2006, heart attacks went down more quickly. If you fit a straight line to the heart attack rates for 1988-2005, you will see that the remaining rates (2006-2012) are below the extrapolation of the line, both for men and women. The Swedes made other dietary changes when they started eating more butter (the butter replaced other foods, for example). Nevertheless, these data make it more plausible that if butter has any effect on heart attacks, it reduces them, the opposite of what we’ve been told.

I eat a half stick (60 g) of butter daily. It improves my brain speed. After I gave a talk about this, a cardiologist in the audience said I was killing myself. I said I thought my experimental data was more persuasive than epidemiology, with its many questionable assumptions. The new data suggests I was right — butter does not increase heart attacks. It also supports my belief that by learning what makes my brain work best, I will improve my health in other ways (such as reduce heart attack risk).

No doubt a low-carb high-fat diet is better than what many people eat, but I believe the never-eat-sugar part of such a diet is a mistake. There are plenty of reasons to think sugars eaten at the right time of day improve sleep. Whatever you think about nutrition, don’t get too comfortable.

13 Responses to “Butter and Heart Attacks in Sweden”

  1. gwern Says:

    > If you fit a straight line to the heart attack rates for 1988-2005, you will see that the remaining rates (2006-2012) are below the extrapolation of the line, both for men and women.

    Really? It just looks like the usual meandering time-series to me. Data and source code for this claim?

  2. hertm Says:

    “Really? It just looks like the usual meandering time-series to me. Data and source code for this claim?”

    Don’t you have a worthless anime essay to write?

  3. dearieme Says:

    I’ll say it again: nobody knows why heart attack rates have been trending down in so many countries for so many decades. But it’s so marked that it can’t be to do with the decline of smoking, with the adoption of low-fat diets, or with medical changes. The infection conjecture sounds plausible, but until the relevant micro-organisms are identified, it remains conjecture.

    My personal variant is that heart attacks are indeed caused by infections, and that the decline is caused by “over use” of antibiotics by the medical trades. When I’m proved right, I shall happily share my Nobel with those who provided the evidence to back me up.

    Seth: Epidemics usually decline, in my understanding, because it becomes harder and harder to surprise a host. The virus or whatever spreads faster than the epidemic of acute illness. You may recall my theory that shaking hands is an early warning system. However, this is such a slow epidemic it is hard to know if the usual ideas apply.

  4. Eric Says:

    The most likely thing butter is replacing is margarine. It seems like the data could be accounted for by a decline in margarine consumption alone.

  5. Jim Says:

    there are people who have questions about this graph but it is dramatic

    http://www.drdavidgrimes.com/2013/07/an-epidemic-of-coronary-heart-disease.html

    Dearmieme, if you haven’t read his book, check it out

    Seth, I’m now with you on the sugar. It can’t make sense that something everybody likes is bad for you, but there does seem to be some context that makes it work. Time of day or amount or some co-factor.
    Ditto on saturated fat, meat has it because that’s what mammals are made out of, including us. So maybe there is a situation where it’s bad but it can’t be universally bad, that wouldn’t make any sense

  6. Ripken Holt Says:

    Seth,
    I’ve been following your blog and you have pretty much convinced me that sugar is important to your body’s health. However, I am curious as to what DOES cause the astronomical heart disease rates in America if it is not sugar. What do you think the likely cause(s) are?

    Seth: I don’t know what causes heart disease nor can I explain the recent decrease, which as dearieme says is substantial. I explain the fact that heart disease is very low in France and Japan by their high consumption of fermented foods (France and Japan) and omega-3 (Japan). I think Mr. Heisenbug may be on to something when he asks why smoking is a very bad risk factor for heart disease.

  7. Gina Says:

    I agree with Eric.

    I’m only convinced that soybean oil, corn oil, cottonseed oil, etc. are bad, not that butter is good. This makes sense with the Esselstyn/Ornish/Barnard stuff (i.e. studies that reverse heart disease and diabetes with ultra-low fat diets). I’m guessing that avoidance of these fats would prevent heart disease no matter what the rest of one’s diet.

    Seth: My Agatston score — a measure of circulatory system calcification and an excellent predictor of heart attacks — went down during a year in which I greatly increased my butter intake compared to all previous years of my life. I did not eat less soybean oil, etc., during that year. I just ate more butter — much more fat overall.

  8. dearieme Says:

    “astronomical heart disease rates in America”: nah, the epidemic is practically over. From Jim’s link
    “At the peak of the epidemic in 1970 there were 520 deaths per 100,000 per year in England and Wales, 700 in the USA, and … Thereafter there was an abrupt and apparently spontaneous decline, by 83% during the next 20 years to 1990. This is an important observation as during this time there was no widespread effective medical intervention.

    The decline continues. At present the death rate is about 20 per 100,000 per year and the decline appears to continue. We appear to have experienced a natural epidemic and it is now almost over. The cause of it has not been obvious, but when we realise that CHD has been an epidemic we can start to think objectively about what might have caused it.”

  9. dearieme Says:

    Which takes us to the Swedish study: if heart attack rates are declining anyway, the correlation with butter consumption is probably meaningless. There could be a similar trend for the consumption of, say, Australian Shiraz or New Zealand Sauvignon Blanc or Single Malt whiskies. Mere observational studies are feeble things, especially when there’s no demonstration that it’s among the people who eat the butter that the decline is concentrated.

  10. Gina Says:

    “My Agatston score — a measure of circulatory system calcification and an excellent predictor of heart attacks — went down during a year in which I greatly increased my butter intake compared to all previous years of my life. I did not eat less soybean oil, etc., during that year. I just ate more butter — much more fat overall.”

    I wonder if that can be attributed to the K2? That would explain why some people insist that only grass-fed butter is healthy (higher K2 content).

    Surely the butter displaced something in your diet?

    Seth: It immediately displaced pork fat — I stopped trying to eat pork fat every day in order to sleep better, which I’d been doing for about a year. When I started eating more pork fat was when my overall fat consumption went way up. I think at that time everything else went down.

  11. Crystal Says:

    Why are there 3 extra data points at the left of the graph? Does that data cover a different time period or are the points not aligned properly?
    I thought butter was demonized because it was a saturated fat, not that it, alone, was correlated to heart attacks. Publish the same graph against total sat fat consumption and you’ll have a stronger case to dissociate the two.
    Ps there’s interesting work that charts the total lack of correlation between sat fat consumption and heart attacks in coconut producing (& eating) countries, so now there’s a bit of work investigating possible heart benefits from specific medium chain saturated fats, like coconut oil, in the diet.

  12. gwern Says:

    hertm: aww, is someone butthert over my criticisms here?

    > Seth: My Agatston score — a measure of circulatory system calcification and an excellent predictor of heart attacks — went down during a year in which I greatly increased my butter intake compared to all previous years of my life.

    Interesting, but where’s the control Seth’s scores? This comment seems to assume that if your Agatston score is different at Time 2 than Time 1, it was because of the change in dietary butter. There are dozens of possible reasons a person’s Agatson score might differ at the two times (leaving aside measurement error, another possibility). Dozens of things that affect Agatson score were not kept constant. This would seem to be committing a very basic and fundamental purely logical error that anyone with a good training in philosophy or science or who has taken the trouble to consciously develop good intellectual habits would easily avoid.

  13. Joe Says:

    If you have a Kindle, you can read about the various reasons heart attack rates are plummeting, and for only $1.46!

    You Will Not Die From A Heart Attack, by Dr. David S. Grimes:
    http://www.amazon.com/WILL-HEART-ATTACK-Medical-Briefs-ebook/dp/B00C4C0XNG