What happens to your metabolism when you eat ultra-processed foods?

Daniel Lakeland writes:

Hey, you wanted examples of people doing real science for the blog!

Here’s a randomized controlled trial with a within-subjects crossover design, and completely controlled and monitored conditions, in which all food eaten by the subjects was created by the experimenters and measured carefully, and the participants spent several weeks in a metabolic monitoring facility.

They mention how expensive this was, but really, this was CHEAP, think of all the money wasted paying salaries of people doing observational regression studies with crappy measurements over the years? Like Wansink’s stuff for example.

The link is to a news article by Maria Godoy. Here’s more:

The study, conducted by researchers at the National Institutes of Health, is the first randomized, controlled trial to show that eating a diet made up of ultra-processed foods actually drives people to overeat and gain weight compared with a diet made up of whole or minimally processed foods. Study participants on the ultra-processed diet ate an average of 508 calories more per day and ended up gaining an average of 2 pounds over a two-week period. People on the unprocessed diet, meanwhile, ended up losing about 2 pounds on average over a two-week period.

“The difference in weight gain for one [group] and weight loss for the other during these two periods is phenomenal. We haven’t seen anything like this,” says Barry Popkin, a nutrition professor at the University of North Carolina who has studied the role of ultra-processed foods in the American diet but was not involved in the current research. . . .

Study participants were allowed to eat as much or as little as they wanted but ended up eating way more of the ultra-processed meals, even though they didn’t rate those meals as being tastier than the unprocessed meals. . . . 20 healthy, stable-weight adults — 10 men and 10 women . . .

Popkin says the take-home message for consumers is, “We should try to eat as much real food as we can. That can be plant food. It can be animal food. It can be [unprocessed] beef, pork, chicken, fish or vegetables and fruits. And one has to be very careful once one begins to go into other kinds of food.”

That last bit seems like pure story time, and I wish the reporter had shouted a Whoa! at this point.

But let’s set aside the hype and move on to the research paper, “Ultra-Processed Diets Cause Excess Calorie Intake and Weight Gain: An Inpatient Randomized Controlled Trial of Ad Libitum Food Intake, by Kevin Hall, Alexis Ayuketah, Robert Brychta, Peter Walter, Shanna Yang, and Megan Zho.

I think what Lakeland likes about this paper is that they get into the details. Instead of considering the treatment as a black box and using causal identification to get a reduced-form estimate of an average causal effect, they take lots of individual-level measurements and try to track what’s going on.

I was concerned about how to think about effects in a short time period. Obviously you’re not going to see people gain 2 pounds a week for a sustained period. Short term, you give people food that tastes better, they’ll will eat more, but it’s not clear what that would do to a person’s equilibrium weight and food consumption level.

Lakeland replied:

That’s a good point, but I think the useful information here is more about mechanism than about specific quantitative estimates of gain / week or whatever. For example, although the participants chose how much to eat themselves, *both* groups came close to consuming the same protein! This is consistent with a previously suggested mechanism: that people eat until they get “enough” protein. In the absence of a protein feedback mechanism. So it provides some evidence to raise our posterior probability for that feedback mechanism acting.

You can easily criticize their use of “significant” everywhere, but let’s be honest, do we expect any better? No. But on the other hand, hey the data are available online for Bayesian reanalysis: (Warning, OBNOXIOUS proprietary SAS dataset formats apparently readable using R “haven” library.)

Also, there’s the information that people eating the “processed” diets consumed their food more quickly, potentially getting more calories precisely because the satiety feedback occurs more slowly than the consumption rate for “soft” foods.

Sure, lots of these things are “story time” until you test them on followup studies, but I think this is what a quality non-degenerate research program looks like: do some controlled studies, take lots of within-person measurements, and then come up with plausible explanations… then test those explanations in followup studies…

So I think it’s worth mentioning the high quality measurements, the controlled design, the within person design, the data availability etc as generally providing us with much more information than other types of dietary research.

Fair enough. I have to get out of the trap of thinking that a single study should do it all.

P.S. I was amused by this bit from the Hall et al. article:

As an alternative to traditional approaches that focus on nutrient composition of the diet, the NOVA (not an acronym) diet classification system considers the nature, extent, and purpose of processing when categorizing foods and beverages into four groups . . .

I don’t get it. If NOVA is “not an acronym,” then where does the name come from? And why is it all caps? Here’s an article with some of the history but it doesn’t explain this particular mystery.

Hey, I have no problem with gimmicky names. I just never heard of this trick of using all caps for something that’s not an acronym. What will they think of NEXT?

51 thoughts on “What happens to your metabolism when you eat ultra-processed foods?

  1. “I was concerned about how to think about effects in a short time period. Obviously you’re not going to see people gain 2 pounds a week for a sustained period.”

    I think this was likely due to the increased sodium content in the processed food diet. This initial gain seems unsurprising in light of that (notice the fat mass change is not as great as the total weight change; also notice the initial bump and plateau in body weight; see Figure 3). I would think the weight change would continue, but at a rate like that of the fat mass change.

    • Could be, but it also fits with the general idea of weight gain/loss. It takes about 3500 calories to burn a pound of so the typical diet recommendation is to try to cut 500 calories a day from what you usually eat if you want to lose a pound per week. Flip side, if you want to gain a pound per week you should add 500 calories a day.

  2. Thanks for discussing our study! I have no idea where the NOVA (not an acronym) comes from! ;-)

    You say: “you give people food that tastes better, they will eat more”. If only it were so easy to explain our observations! The participants rated the pleasantness of the meals on several occasions and we didn’t find any differences between the diets. We are now designing a new study to both replicate the original findings in a new cohort and also test a reformulated ultra-processed diet matched for non-beverage energy density with the unprocessed diet. Six weeks of continuous inpatient testing in our metabolic ward.

    Regarding long-term extrapolation, our dynamic modeling predicts roughly 9 lbs of total weight difference between the diets within ~1 year. But I thought the modeling stuff was too speculative to include in our paper. See here for more details on our dynamic modeling:

    https://www.ncbi.nlm.nih.gov/pubmed/29896621
    https://www.ncbi.nlm.nih.gov/pubmed/29635495
    https://www.ncbi.nlm.nih.gov/pubmed/27804272

    • I’m not known for being overly solicitous of academics But Andrew covers so many domains. I’ve been impressed. The blog has been a great learning experience for me. Plus I can joke a bit more freely here.

      I’d love to see more of the ones I think are mischievously smart. lol

    • Just think of it as a technical jargon term. The definition of processed and ultra-processed is fairly precisely given in the NOVA people’s literature… I actually had a little bit of conversation with Andrew regarding the definitional issue, because it was also something my wife brought up. like she asked why not give a processed vs unprocessed version of the same diet, so she said things like “processed bread” (I assume she was thinking white flour etc) vs “unprocessed bread” (I assume she was thinking whole wheat flour etc) but if you look at the definitions in NOVA *all* bread is processed.

      The definition is worth checking out just so you can understand what the terms mean, as they are technical/jargon terms for sure.

    • A description of the NOVA classification scheme lives at:
      https://archive.wphna.org/wp-content/uploads/2016/01/WN-2016-7-1-3-28-38-Monteiro-Cannon-Levy-et-al-NOVA.pdf

      Two little clips:

      “Group 1 foods include fresh, squeezed, chilled, frozen, or dried fruits and leafy and root vegetables; grains such as brown, parboiled or white rice, corn cob or kernel, wheat berry or grain; legumes such as beans of all types, lentils, chickpeas; starchy roots and tubers such as potatoes and cassava, in bulk or packaged; fungi such as fresh or dried mushrooms; meat, poultry, fish and seafood, whole or in the form of steaks, fillets and other cuts, or chilled or frozen; . . . ”

      “Examples of typical ultra-processed products are: carbonated drinks; . . . ice-cream, chocolate, candies (confectionery); . . .infant formulas, . . . pre-prepared pies and pasta and pizza dishes; . . . ”

      So, corn on the cob and steaks are in group 1; pizza in group 4. Seems reasonable to me.

      Bob

      • White rice shouldn’t be in group 1. Sheesh. It makes Wonder Bread look good.

        (I live in Japan, where I don’t see genmai anywhere other than at home. And whole wheat bread is rarer than hen’s teeth. It ain’t the diet that makes Japanese live longer than anyone else, it’s that they don’t kill each other with cars, guns, and opioids. And that the infant mortality rates is so low. But they do live a long time. My niece is a cardiologist. “How old is your youngest patient”? I asked her. “Forty six, but the next youngest one is 93. And they all die of cancer, not circulatory problems.”)

        • The classification system is based on how the food is processed. White rice is “minimally processed” because it’s basically raw rice that’s been tumbled to remove the bran. Wonder bread is ultraprocessed because it requires lots of derived ingredients and processes to produce. The classification system is fairly consistent and objectively verifiable.

        • I could be wrong. Some rice brands use a bleaching process. Not all rice can be labeled as ‘minimally processed’. Or at least that is my recollection from reading an article a long while back. With bread, the additive potassium bromide has been alleged to be a carcinogen. I haven’t checked how it is labeled in commercial bread brands.

          My view is that curbing sugar and high glycemic indexed foods have real health benefits. Plus I value Sydney Wolfe’s advice; eat small portions. Most of us eat too much.

        • Daniel: That makes sense. But that means that whole wheat bread is way more processed than white rice, so the concept of degree of processing of the foodstuff isn’t quite the right thing.

  3. OK, as an advance of experimental design in nutritional science I see the progress. But in terms of the science (underlying causal argument) I don’t really get it. It offers a policy recommendation: x tends to be better than y (in some sense of better). But there’s a clear confounder here. The supplement shows that the two menus. IMO the unprocessed menu is just less appetizing. If the argument is that it’s the “unprocessedness” doing the work, then ‘tastiness’ needs to be controlled for. And search me if I’d know how to control for that in a rigorous manner.

    • One presumes the “ultra-processed” stuff was the stuff you called “less appetizing” or at least, I certainly looked at it and said “ugh”… And yet, it was the ultra-processed diet that led to weight gain, so evidently people were eating more calories of the less appetizing stuff…

      The nice thing about a study involving careful and multiple measurements is that you can analyze the data multiple ways.

      • It seems as if you misread my comment. I had written that, “the UNprocessed menu is just less appetizing.” To be clear, I’m not saying that my aesthetic judgement is correct, just that there’s an alternative explanation that can’t be adequately ruled about by the study’s design, and as such draws attention to the muddy causal account on offer. [Also, I never used the term “stuff”. I’m really interested in the science here, not the aesthetic values one way or another.]

        • Interesting… Did you actually look at the pictures of the plates of food or are you basing your own aesthetic judgement on textual descriptions?

          Food photos for each of the days (first all the ultra-processed, and then all the unprocessed ones) here:

          https://www.cell.com/cms/10.1016/j.cmet.2019.05.008/attachment/f7d43756-3f67-4557-8322-59a9d143d63c/mmc1.pdf

          I found that the aesthetics weren’t even close… the unprocessed foods were *wildly* more appealing.

          But your main point, that there are alternative causal explanations is absolutely correct. This is actually *always* true, and so in good scientific analysis, you put forward a variety of possible theories, and then try to find ways in which those theories break down in terms of explanatory power, and then after multiple experiments of various kinds… you are left with a more narrow range of plausible theories.

          In his comment above: https://statmodeling.stat.columbia.edu/2019/10/28/what-happens-to-your-metabolism-when-you-eat-ultra-processed-foods/#comment-1151434

          Kevin describes their effort to address this mechanism you mention (I think he intended to reply directly to your comment)

          “The participants rated the pleasantness of the meals on several occasions and we didn’t find any differences between the diets.”

          To address your proposed mechanism they are also “We are now designing a new study to both replicate the original findings in a new cohort and also test a reformulated ultra-processed diet matched for non-beverage energy density with the unprocessed diet. Six weeks of continuous inpatient testing in our metabolic ward.”

          I think this research does a reasonable job about not making definitive claims of having discovered the one true mechanism or anything like that.

          @Kevin Hall, are you a regular reader of Gelman’s blog or are you here because someone mentioned we were talking about your study? Just wondering, and thanks for commenting.

    • “If the argument is that it’s the “unprocessedness” doing the work, then ‘tastiness’ needs to be controlled for”

      Actually, I think this is a mistaken notion of the kind of causality being tested here. The claim isn’t that unprocessedness does the work independently of tastiness (so that things with equal “tastiness” will turn out lower calorie in the unprocessed group), the claim is that processedness or unprocessedness entails a whole variety of qualities, which taken together, result in people eating less unprocessed calories when left to their own appetite.

      It’s like if I said that “biology education” causes “better health outcomes as an adult” but you want me to control for how much biology knowledge a child has at age 18… If you’re trying to study the effect of “the amount of time spent in a class called biology” sure, but the real point is to find out how learning about biology results in people doing things that are healthier… The intermediate outcome difference of “amount of biology knowledge at age 18” is part of the *purpose of the treatment* and so it’s supposed to be there, not held constant.

      In the same way, tastiness is an intermediate outcome we should NOT control for here unless we are investigating just this sub-mechanism. Perhaps a main reason people create processed foods is to hyper-activate “tastiness” among some groups of people, so they’ll buy a lot of it… It’s not a claim that in the abstract “the act of processing” magically causes equally tasty equally calorie dense foods that you eat equal quantities of to make you fatter than the unprocessed things.

      There are a variety of sub-mechanisms that may be involved, such as macronutrient balance (amount of protein), salt content, spice and flavor content, texture (soft/hard etc) and others. If you want to probe specific sub-mechanisms you could try to control each of these various things, using measures of salt, of protein density, of texture etc…

      But if you use a large enough sample of the variety of things called “Unprocessed” and the variety of things called “processed” and you see a long term difference, then you can determine that evidently there are some consistent differences between the two methods of preparing food which result in people eating more ultra-processed foods and less unprocessed…

      The alternative is that there is some biased sample of each diet, which for example leaves out certain kinds of foods that other people would have naturally included more of or the like. That’s also a possibility, you could investigate it.

    • Both menus look pretty bleak to me. Where’s the KD hotdogs and bacon?

      I have a suggestion for a Processed Dinner:

      1 box KD
      4 hot dogs, sliced 1/4″ and fried
      1/2 bag frozen peas
      optional: 1/2 cup diced jalps from a jar

      Cook separate.
      Mix together while hot.
      Add 1 cup ketchup (organic OK) add jalps and/or 1 heaping teaspoon of pepper.
      Stir in Ketchup and jalps/pepper

      Serve in pot. MMMMMMMMM!!!!!! :)

  4. Study participants on the ultra-processed diet ate an average of 508 calories more per day and ended up gaining an average of 2 pounds over a two-week period. People on the unprocessed diet, meanwhile, ended up losing about 2 pounds on average over a two-week period.

    “The difference in weight gain for one [group] and weight loss for the other during these two periods is phenomenal.

    This is totally unimpressive… water weight fluctuations over the course of a couple days can be far greater than this. If you want to see impressive look at people who eat a low carb diet and lose 2 lbs per week for months straight while eating until satiation the whole time. But nutrition researchers still call low carb a “less than 50% carb diet”.

    • As I said above, the impressive thing about this research isn’t the diet and its weight loss results… the impressive thing is how the researchers took scientific study of diet seriously and did careful measurements, controlled as much as possible, made multiple kinds of measurements, made the data publicly available, and investigated a variety of mechanisms that might explain the results.

      As for the particular result on the 2 week period, if you just look at two points in time, the start and the end, then yes water weight fluctuation could explain it. On the other hand, if you see a consistent trend whose magnitude is -2 pounds per 2 weeks you can’t really explain this by water weight fluctuations unless you think these people are slowly dehydrating at a consistent rate… Water weight goes up and down throughout the day, but can’t easily trend for long periods of time.

      • On the other hand, if you see a consistent trend whose magnitude is -2 pounds per 2 weeks you can’t really explain this by water weight fluctuations

        This is the difference between individual and average we were discussing in the other thread. Your “model” is for an individual, but their figure 1 is for the average.

        • In figure 3C they show actual fat mass change, which shows that the results are not due to water weight change. (again, this is down to good choices in measurement)

          As for the Simpson’s Paradox it is a paradox of prediction, not causality. If something causes an average increase in X for all subgroups, it is not possible for each subgroup’s X to decrease if you implement the intervention.

        • This has nothing to do with simpsons paradox. The average curve could be due to different people suddenly losing water weight (ie, reducing glycogen stores) at different times after the beginning of the study.

          How does figure 3c show that it is due to fat mass change? Those are a fraction of the total weight change.

          Body composition measurements were performed at baseline and weekly using dual-energy X-ray absorptiometry (General Electric Lunar iDXA; Milwaukee, WI, USA). Changes in body energy stores were calculated using the measured changes in body fat and fat-free mass along with the corresponding energy densities of 9300 kcal/kg and 1100 kcal/kg, respectively. Liver fat measurements were performed using T1 and T2 corrected proton magnetic resonance spectroscopy with a breath-holding technique in a 3T scanner (MAGNETOM Verio; Siemens, Tarrytown, NY) (Ouwerkerk et al., 2012
          ).

          I am not familiar with this, but how does losing water affect measurements? It looks like they observe the volume of fatty tissue: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6150431/

          That volume will be affected by interstitial water.

        • For example, at time 7 in figure 3A, the average body mass change in the ultra processed diet was 1kg/person. If there are N people, then if you put all of them on a scale at once, there would be N persons * 1kg/person more mass on the scale on day 7.

          In figure 3C it shows that of the N*1kg about N*0.2kg would be fat.

          You can argue that the measurements in 3C are inaccurate of course, but if you assume they know how to measure this quantity, then they’ve shown that in aggregate across all the people in the study, after 7 days on the ultra-processed diet there was about N*0.2kg more total fat mass than on day 0.

        • In figure 3C it shows that of the N*1kg about N*0.2kg would be fat.

          Yes, only 1/5 of the weight loss was fat according to the DXA method. What was the other 4/5?

          You can argue that the measurements in 3C are inaccurate of course, but if you assume they know how to measure this quantity, then they’ve shown that in aggregate across all the people in the study, after 7 days on the ultra-processed diet there was about N*0.2kg more total fat mass than on day 0.

          I don’t see how they could have possible measured it accurately (accounting for the amount of water in the fat tissue) by looking at “fat shadows”, but that is just based on quickly skimming one paper and wikipedia.

          Looking closer I see they handwave this possibility:

          While the dual-energy X-ray
          absorptiometry (DXA) methodology used to measure body
          composition in our study tends to underestimate body fat
          changes (Pourhassan et al., 2013), the relatively large fat-free
          mass changes may be due to extracellular fluid shifts associated
          with differences in sodium intake between the diets. Indeed, in-
          dividual differences in sodium intake between the diets were
          significantly correlated with changes in fat-free mass (r = 0.63;
          p = 0.004) and body weight (r = 0.64; p = 0.002). Such fluid shifts
          may also affect the accuracy and precision of the measured
          body fat changes (Lohman et al., 2000; Mu
          €ller et al., 2012).

          https://www.cell.com/cell-metabolism/fulltext/S1550-4131(19)30248-7#secsectitle0075

        • I’m open to questioning the measurement reliability, and will admit to not having looked in depth at how that was done. I do know they had people in their caloric chamber measuring respiration as well, so they have a variety of measures: total calorie intake, respiration rate on the chamber days, xray and MRI measurements that are at least correlated with fat content. I doubt you can point to *any* nutritional study that does dramatically better than this. Certainly any observational study of people doing very low carb diets doesn’t even get into the same country much less ballpark.

        • Certainly any observational study of people doing very low carb diets doesn’t even get into the same country much less ballpark.

          Try it for a week or two. The denial of the nutrition research community is absurd. It is like someone spending billions of dollars every year to convince you hitting your finger with a hammer every day doesn’t actually hurt… No, it is t hat your finger always hurt which is why you hit it with a hammer.

        • Sure, try it for a week or two gives some information to you, but you know what would be a good scientific method to study it on a large scale?

          Do this kind of experiment described here and observe the results under controlled conditions with lots of fairly precise measures and then release the data publicly to let people do various kinds of analyses on it.

          As I keep saying, the exciting thing about this study is that it actually *is* science applied to diet, not that it alone once-and-for-all definitively determines the whole truth about diet.

  5. No, it is like you will reach a “jumping from altitude without a parachute means you die over 99% of the time” obvious conclusion that the people paying for (and paid to do) nutrition research apparently do not want. Cause and effect is totally reversed in most of these studies.

    The way it works is that you eat fewer carbs and you feel less hungry as a result so you then consume fewer calories.

    • “The way it works is that you eat fewer carbs and you feel less hungry as a result so you then consume fewer calories.”

      You do realize that “eat until reaching your protein requirements for the day” is the primary mechanism proposed in this study for why their unprocessed diet works right? Basically the unprocessed diet *is* a lower carb diet naturally. And that there are a whole bunch of measurements that they are able to bring to bear on the subject.

      Anyway, you can dismiss the typical nutrition study all you want, but this study is anything but typical, nor does it try to hide what you apparently think is the “obvious” outcome that higher protein less carbs leads to less total calories consumed and weight loss.

      • Basically the unprocessed diet *is* a lower carb diet naturally.

        No, it isn’t. And you are not going to convince me that +/- 5 lbs in 2 weeks is a meaningful amount of weight loss/gain.

        higher protein less carbs leads to less total calories consumed and weight loss.

        That isn’t how a low carb diet works. You degenerate due to “rabbit starvation” on a high protein diet. You’re body craves high fat diet, which they made a sin so you would buy more food.

        • Typo” “you’re” should be “your”.

          But what I would like to see is one study published using an actual low-carb diet. That means under 0.5% of calories from carbs, or about 100 calories per day. The way this diet works is you feel less “hungry”, but actually most of what people think of as “hunger” is an addiction to sugar. Real hunger feels different.

          You will see it blow all these interventions they have been wasting their time on out of the water. And really 20% (~ 400 cal/day from carbs) is good enough to start seeing an effect for most people.

        • You:

          “But what I would like to see is one study published using an actual low-carb diet.”

          Me: “you know what would be a good scientific method to study it on a large scale?

          Do this kind of experiment described here and observe the results under controlled conditions with lots of fairly precise measures and then release the data publicly to let people do various kinds of analyses on it.”

          Getting beyond what you would like to see be the primary subject of the next study, can we at least agree that if you’re going to do a study having precise measurements, and complete control and measurement of the actual food being consumed etc is the way to do it… I’m having a hard time figuring out whether you just are annoyed that they haven’t done experiments on your favorite pet theory of how to do weight loss, vs you actually think that the methodology they’re using here is not actually good.

          In any case, I also think conceiving of this study as a “study about weight loss” is completely mistaken, this is a study about “methods of producing meals” and what using each one regularly implies about people’s actual eating experiences.

          As Keith likes to say this tells you something about “what you would regularly expect to happen” if you started eating from the “unprocessed diet” defined in the NOVA literature.

        • My complaint is that: The effect seen in this study is of no practical significance, yet they call it phenomenal.

          It is only phenomenal compared to the other studies based on totally wrong-headed ideas about diet and weight. Anyone can prove the low-carb diet works, and intuitively understand the mechanism, for themselves within a week or two. You do not need any expensive study that tells you average results.

          This study still does not exist.

        • Maybe you are pointing more at the popular write up. The paper itself doesn’t say anything about “phenomenal results” and is very explicit about its purpose and motivation: study how presenting ultra-processed vs unprocessed foods alter people’s *choices* in what they eat, and the net result of those choices on their metabolisms over a 2 week period.

          In this study people aren’t told “eat all this food” they’re just given a plate of food and told to eat until they’re done and get up and leave… (leaving the rest to be measured I guess).

          On Avg, when people were given one kind of food they ate more, and when given another kind of food they ate less. I would have to download the individual data to plot for example individual time-series of total calories eaten, but I strongly suspect you will see in the 2 week period at the transition point, a notable change in the time-series curves.

          Anyway, I think it’s pointless to continue here, we seem to be talking in different directions, not really about the same thing.

        • It seems to be almost all water weight. A loss of >10 lb can be “achieved” via a low carb diet in ~3 days. But that is not a meaningful accomplishment.

        • You need to think of “quitting carbs” the same way as “quitting smoking”. It is an addiction you are introduced to at a very young age. Then you are told it is normal your whole life.

        • QAnoneuoid,

          Who is ‘they’? I assume you are referring to fat being demonized during the past few decades and/or a food pyramid recommendations. To my knowledge those were only suggestions made by USDA, not a forced diet camp. If you look up the US meat/fat consumption during the food pyramid ‘reign’, you’ll see it went up,together with processed carbs and everything else. But let’s blame bread.

          This is a classical mistake of ‘if there is no evidence(effect) in one direction, it must be the opposite’. No. It just means there is no sufficient evidence in one direction. You have to demonstrate the opposite. Direction and magnitude of the effects are separate.

          A host of very dangerous new fads (paleo, stone-age, or whatever they call it these days, being one of them) will only create health problems for medical establishment, just give them a couple of decades or so. Confusing weight loss and health or fitness and health can kill a person. Most sumo wrestlers probably have better blood panel results than you and I, despite being obese and many ultra-runners or whatever other ‘fitness’ fanatics we think of, are at higher risk than both of us.

          Regarding carbs, once humans cultivated grass, civilizations started and spread all over. We had fat/protein during ‘paleo’ times, but didn’t leave much cultural heritage from that period, did we? When sea-faring nations spread all over Pacific, including Hawaii, one of their main worries was to make sure they bring breadfruit, taro and other starchy (canoe) plants. Why? They had an abundance of fish/fruit/whatnot.

          Again, confusing processed carbs (carbs and fat or carbs and protein are rarely found in nature) and carbs from cereals and other sources doesn’t put them in the same bin.
          There was a fairly decent study published in Nature some years ago looking at longevity, and some tribes in Amazon part of Peru whose diet consisted mostly of tubers and fish came on top, while ‘modern’ diseases were rare there.

          The study posted in this blog is very well controlled (experimentally, not so much statistically), which is rare in ‘Nutrition science’.

          But to each his/her/its own

    • “What some call health, if purchased by perpetual anxiety about diet, isn’t much better than tedious disease.”

      –Pope Alexander

      P.S. My apologies to Eternal Sunshine of the Spotless Mind

  6. Thanks Daniel for pointing a study that you think does a good job and that did some of the important things properly! Look forward to studying and learning from this example.

Leave a Reply to Alex Cancel reply

Your email address will not be published. Required fields are marked *