Drilling Down: Saturated Fat Epidemiology

I’ve got to get this reader data out there; sooooo, the post I’ve been planning and writing to slam dunk a certain "researcher" from New Zealand will just have to wait until the first of next week. I’d planned to incorporate that with this, per my style, but I’ll just have to link it.

Whenever back when, I did a post on saturated fat epidemiology from a UK reader, Alex Thorne. He was pretty careful in that — unlike science frauds and grant whores — as an honest guy, he anticipated objections and so constructed and graphed the data in different ways. You can go read that post here.

Fast forward to a couple of weeks ago. Seth Roberts, author of The Shangri-La Diet, wasn’t satisfied. I’d dropped a comment at his place over something I can’t recall, and I guess he checked out my place and his comments begin here. So, you can read the basis for Alex’s additional work, which I’ll now display. First, Seth was kind enough to take the time to offer a method:

Here’s a better way to look at these data: make a scatterplot with X = percent saturated fat and Y = disability adjusted lost years per 100,000 population. To see if large and small countries are different, divide the countries into two groups splitting at the median of their population sizes. Then plot large countries with one color of dot (say, red) and the small countries with a different color of dot (say, blue). That will show you if the large and small countries show the same trend. If you want to see even more clearly, fit separate lines to the two sets of points and plot those lines in different colors.

Now, Alex:

The trends are still evident – the more saturated fat in the diet the fewer years lost to disease.


It seems no matter how you slice it, more real, natural, saturated fats from animals and coconuts are wonderfully healthful.

Go figure. I mean, should this be a surprise? Are predatory animals who eat other animals and prize the [saturated] fat dropping dead of strokes and heart disease? No, and why? Because they’re less arrogant than humans. They don’t, because they can’t, eat in any way other than how they evolved to eat.

Go forth and do likewise.

Since Covid killed my Cabo San Lucas vacation-rental business in 2021, this is my day job. I can't do it without you. Memberships are $10 monthly, $20 quarterly, or $65 annually. Two premium coffees per month. Every membership helps finance this work I do, and if you like what I do, please chip in. No grandiose pitches.


  1. Grok on November 6, 2009 at 15:55

    Kudos for the last paragraph!

  2. Ian Johnston on November 6, 2009 at 18:06

    Interesting blog, came here via way of a friend’s link. I haven’t had the time to go through the rest of your pieces yet, but I was wondering, is the correlation between increased saturated fat intake (as a proportion of food) and reduction in disease related to other variables? E.g., animal-based protein? That is, is an increase in saturated fat intake associated with other lifestyle, cultural or dietary/cooking practices that might also help? (e.g., do you have to be wealthy to have a high saturated fat diet? Does your culture find it easy to access animal protein, and if so, what factors of this type of culture might be beneficial?) Or do you think it’s the fat itself?

    It’d be interesting to hear your thoughts on this,

    • Alex Thorn on November 7, 2009 at 03:12

      While collating the data for the graphs I did notice a trend for lower saturated fat intakes for the ‘poorer’ countries so any correlation between personal wealth and lost years due to disease can still be correlated with lower intakes of saturated fats in this sub-group. I did do a comparison between the major macronutrients and found that protein intake was neutral, carbohydrate was positively correlated with lost years (i.e., the more you ate the greater the number of lost years for that population) and total fat was negatively correlated with lost years (i.e., the less you ate the greater the lost years). It would seem likely that poorer nations would eat less meat and thus less animal fats than wealthier nations. This anecdotally obvious from an historical viewpoint in that many of these countries are renown for high carb ‘peasant foods’. I also think poorer populations also tend to fill up more on unsaturated sources of fats especially polyunsaturated omega-6 fats. Not only do they pack more calorific punch per gram but they are also implicated in increased pattern B LDL and reduced HDL and this to simple sugars and starches and you have a recipe for health disaster! Then of course there is the link between polyunsaturated fats and sugar with cancer.

    • Richard Nikoley on November 7, 2009 at 10:10

      “is the correlation between increased saturated fat intake (as a proportion of food) and reduction in disease related to other variables?”


      Absolutely possible. This is just observational, so one can never be entirely certain what variable(s) are truly causal. This sort of observation serves only to suggest what variables we might want to isolate in controlled, random, intervention trials to see if we can squeeze out true cause.

      But, I think this sort of thing (and there’s lots of this sort of thing, if you follow the links) serves another role, and that is to essentially falsify the conventional diet-heart hypothesis that holds that sat fat raised cholesterol causes heart disease.

      And on top of all that, we have the benefit of observations of traditional people’s going back hundreds of years and they simply do not suffer our modern diseases (heart, stroke, diabetes, cancer, auto-immune, etc.) to any significant degree. Accordingly, claims that eating foods like animal fats that have been in our diet for millions of years of evolution are causing disease, while a host of industrial processed “foods” (vegetable oils, processed grains, refined sugars) are protective out to be met with incredulous skepticism.

  3. Aaron Blaisdell on November 6, 2009 at 19:49

    What’s truly amazing is that once average saturated fat intake reaches 12% or more of the diet the between-country variance in incidences of all three diseases becomes extremely small. Looks like a threshold or “magic number” to stay above to almost guarantee to cut the incidence in half. That is, if the correlation is due to causal relationships (Dietary SAT -> disease) rather than a spurious non-causal relationship.

    • Alex Thorn on November 7, 2009 at 03:16

      Well spotted – and I agree with you there!

  4. g on November 7, 2009 at 12:29

    Hey Ian,

    Like omega-3s pufa’s, saturated fatty acids come in a variety of chain lengths: short, medium and long. For fish oil — the long chain fish oils DHA EPA are related to longevity, increased lifespan, anti-inflammatory status, anti-cancer and anti-heart disease properties. The same with saturated fats.

    Mozzafarian at Harvard has shown that at sat fat >= 12% of dietary energy, women with coronary heart disease in the ERA trial demonstrated REGRESSION of the plaque in the vascular arteries. This was the ONLY GROUP to exhibit regression (they also SMOKED MORE and TOOK LESS DRUGS, eg STATINS):

    Butter is about 50/50 monounsaturated and short chain saturated fat (butryic acid, 4-carbon chain). Like omega-3s, this sat fat binds the PPAR receptor series which is a potent controller of inflammation and cancer.

    It’s all about the sat fat…

    Dr. T, a Paleo nephrologist from So Cal, articulately and eloquently explains the role of PPARs here:

    Hey Paleo KING!

    I love all your graphs and the contributions you make to understanding the value of saturated fats in our health and lifespan extensions. Keep up the strong work !!! I learn so much from you, dude.

    Someday our U.S. food guidelines will change with your work and many others. Will it be soon enough?? I’m getting fed up with the tragic, reversible damage that I see everyday…


  5. Vegan Trolls | Free The Animal on November 7, 2009 at 14:21

    […] Drilling Down: Saturated Fat Epidemiology […]

  6. Seth Roberts on November 7, 2009 at 17:50

    Thanks for doing a better analysis. It’s a lot clearer — if I say so myself. The feature of the data that Aaron points out would be easier to learn from if the Y axis was plotted with a log scale. There is an obvious floor on disability values (you can’t go below zero) so there must be a decrease in variance as the disability values go down and saturated fat values go up. (You can see the decrease.) Do the disability values really stop changing above 12% saturated fat? Or is this an illusion produced by the inevitable decrease in variance? Using a log scale for the Y axis will help choose between these two possibilities.

    • Alex Thorn on November 8, 2009 at 16:01

      I have done this now:

  7. Seth’s blog » Blog Archive » Saturated-Fat Epidemiology on November 7, 2009 at 21:20

    […] Here, at Free the Animal, are three scatterplots that show better health (less heart disease, less stroke) correlated with more saturated fat (= animal fat) in the diet. Each point is a different European country (Albania, Bulgaria, etc.). Small and large countries show the same relationship. […]

  8. Vince on November 8, 2009 at 03:36

    Alex, have you tried controlling for wealth? Richer countries tend to be healthier (for a variety of reasons), and they also tend to eat more meat. The “above 12%” countries that Aaron noted, for instance, are all among the richest countries in the world (France seems to be the poorest of the lot). And some of the countries that are outliers on your graph are countries with a mismatch between wealth and saturated fat consumption. Israel is relatively wealthy, eats relatively little saturated fat, and is very healthy. Turkmenistan is very poor, eats a moderate amount of saturated fat, and is very unhealthy. You could get GDP per capita data, and then just run a regression using that & the saturated fat % to predict years lost. Or, if you post a table of your data, then anyone could test whether other variables can explain the relationship between saturated fat consumption and health.

    • Vince on November 8, 2009 at 06:22

      I decided to run the numbers myself, seeing if this relationship between saturated fat consumption and health holds up after controlling for wealth. It doesn’t. The raw correlation between Saturated Fat Consumption as a Percentage of Total Calories (SatFat%) and Disability Adjusted Lost Years (DALY) is large, r = -.69 (p less than .0001), but after controlling for GDP Per Capita (GDP/person) it basically becomes zero (it actually reverses direction, with more saturated fat associated with worse health outcomes, but it’s nowhere close to statistically significant).

      All the numbers that I needed except for GDP per capita were in Alex’s first post. His first graph has labels with the SatFat% for 45 countries, and his last graph has labels with the DALY figures for those countries*. I copied those numbers into a spreadsheet, and then added the GDP Per Capita numbers (from the IMF, 2008). If anyone is interested, I could email them the spreadsheet.

      GDP/person correlated strongly with both SatFat% (r = .86) and DALY (r = -.84). And in a linear regression predicting DALY from both SatFat% and GDP/person, only GDP/person was a significant predictor (F(1,42) = 33.9, p less than .0001); SatFat% was nowhere near significant (F(1,42) = .49, p = .49). This model explains 71% of the variance in DALY (R^2 = .710). The regression equation says that every additional $1000 in GDP/person is associated with 98 fewer DALYs, and each 1 percentage point increase in SatFat% is associated with 54 additional DALYs (but this is not significantly different from zero).

      I tried a few variations on this analysis, seeing if transforming the variables made for a better model. Using the log of DALY instead of DALY is an improvement: the pairwise correlations are stronger (-.75 and -.89 instead of -.69 and -.84), R^2 for the regression model goes up (.787 instead of .710), and Norway is no longer predicted to have a negative number of DALY. The pattern of regression results remains the same: GDP/person is a strong predictor of log(DALY) (F = 43.7), and SatFat% has a very slight positive association with log(DALY) that is not close to being statistically significant (F = .05, p = .83).

      These analyses suggest that wealth (or something closely correlated with wealth) has a big impact on saturated fat consumption and on health, and that whatever impact saturated fat consumption has on health is too small to show up in this data set.

      *Two countries, Moldova and Macedonia, were missing from the last graph, so I used their approximate DALY from the first graph, which plots DALY on the y-axis but doesn’t label the exact number.

      • Richard Nikoley on November 8, 2009 at 13:37


        I really appreciate you going to that work.

        Listen, the las thing I — and Alex, I’m pretty sure — want is to fool ourselves or others. And, I think you can see from the first post, then Seth’s objections and Alex’s own anticipation of objections and subsequent crunching that everyone is being as objective as possible.

        I’ve also read Seth’s response to your analysis over at his place.

        I tend to agree with that, at least in the sense that I think it’s perfectly sensible to _presume_ that saturated fat, particularly in the context of a whole foods diet with minimal neolithic foods (grains, sugar, vegetable oils and derivatives), out to be perfectly healthful. As I said at the end of the post:

        “Go figure. I mean, should this be a surprise? Are predatory animals who eat other animals and prize the [saturated] fat dropping dead of strokes and heart disease? No, and why? Because they’re less arrogant than humans. They don’t, because they can’t, eat in any way other than how they evolved to eat.”

        This, I think, is what we ought to expect at a very fundamental level and ought to really demand extraordinary proof in the contrary.

        What we have is some confirmation based on the raw correlations that what actually makes perfect sense is likely the case.

  9. Alex Thorn on November 8, 2009 at 15:35

    I’m glad others are running with this data as my statistical skills are basic! I think it is important to differentiate between percentage of total calories and absolute amounts per person per day for the foods/macronutrients. Obviously poorer populations can afford less food (and less calories overall) so a modest or small amount of saturated fat (in absolute grams-per-person-per-day terms) may translate as percentage of total calories in the 12%+ range but fails to have a positive effect because the overall amount is low. As an example Turkmenistan (as Vince points out) is a poor country (ranked at 102 by the IMF) and eats, on average, just six grams of saturated fat per person per day which equates to 10.1% of total calories and has high DALYs. While Portugal (ranked at 32 by the IMF) eats, on average, eats 35g of saturated fats per person per day, which equates to 10.6% of total calories and have much lower DALYs. It may also be a function of the quantity and quality of the carbohydrate foods that are consumed along with the saturated fats – maybe the ratio of one to the other has some bearing on health outcomes?

    • Alex Thorn on November 8, 2009 at 15:45

      Forgive the grammatical/typographical errors in the above! I should just clarify from my brief example that the difference in saturated fat intake between the poorer Turkmenistan and richer Portugal, in percentage of total calorie terms, is 0.5% while, in absolute grams per person per day terms, it is 29g or 500% more saturated fat!

    • Richard Nikoley on November 8, 2009 at 15:53

      This suggests an alternate way of looking a the data. Alex, perhaps you can re-runn to look at sat fat in the absolute, and Vince can do the same, age adjusted.


      • Alex Thorn on November 9, 2009 at 09:30

        I’ll certainly give that a go and mail you the results!

  10. Vince on November 10, 2009 at 20:17

    I’ve left a couple more comments at Seth’s blog, and I figured I should add an update here. First, I’ve posted my data set as a spreadsheet on Google Docs. If someone can get me a spreadsheet with the other data that they’d like to see analyzed (e.g. by adding the absolute saturated fat #s to the google spreadsheet) then I’ll be happy to run that analysis.

    Second, to summarize my position on using these data to test the theory that saturated fat is good for your health. We know that there are many ways in which richer, more developed countries differ from poorer, less developed countries: they’re healthier in various ways, more educated, do less physical labor, have higher gender equality, eat more meat, have more political freedom, use more energy, etc. So when you’re interested in looking at some variable between countries, it’s important to check whether it’s one of the many variables that are associated with development level. If it is, then it’s a safe bet that it will be correlated any of the other variables that goes with development level, regardless of whether or not there’s any causal relationship between them, since they are both part of the same package (development). As a first step to see if there is a causal relationship between them, you can run a regression controlling for some measure of development (such as GDP per capita) to see if their relationship still holds up – are they related to each other beyond what you’d expect from them both being features of developed countries? There are limitations to that analysis (you want to think about whether you’re using the best measures of the variables, and whether there’s something more you should be controlling for), but if your predicted relationship holds up in the regression then that’s some evidence that it’s a real causal relationship. If it doesn’t then that’s some evidence that the there isn’t much of a causal relationship between them, and the correlation between them is just due to them both being part of the development package.

  11. […] Drilling Down: Saturated Fat Epidemiology […]

  12. […] Drilling Down: Saturated Fat Epidemiology […]

Leave a Comment

You must be logged in to post a comment.