Protein for Longevity: Debunking the Plant vs. Animal Protein Debate

Is plant protein better than animal protein for preventing chronic disease and promoting longevity? Many are now making this claim, but I for one am skeptical. So on popular request, I am going to dive into this question and expose nuances I think you should see to make an informed decision on this matter for yourself, as well as to reveal how I think about these sorts of claims. So I'm gonna start off with some general high-level points about dietary studies and then delve into the chronic disease and longevity data. The first thing, in my opinion, that you should assess on studies making a claim about diet X versus diet Y is what did the participants in the study actually eat? Often in feeding trials, the diet formulations can be rather shocking. Take, for example, a recent randomized controlled trial comparing a "beef diet"—that's what it was named—to a "veg" or vegetable diet. As shown, not in the main text, but in the supplement, the "beef diet" included for breakfast an English muffin with peanut butter and apple, low-fat milk, and then spaghetti and a salad with Italian dressing with lunch, bread rolls, peanuts and beans, and fruit with dinner, and chips, hummus, and almonds as a snack. Now, I'm not saying there was no meat in this diet, as there was a very lean steak with the dinner and some hamburger ground on the pasta at lunch. But the point here is that it's important to get a picture of what people are actually eating in trials to make any reasonable conclusion as to whether this might generalize to your eating style. I, for example, do eat a lot of meat and animal protein.. But my diet looks nothing like the diet described in this study. Nor does this population, which included persons with overweight or obesity eating about 280 or 300 grams of carbs daily, they don't reflect my metabolic health status and dietary pattern. Maybe they reflect yours, maybe they don't, but a key point to hammer home is that context matters. And if data on dietary component X is being collected in a population that grossly, that is grossly distinct from you in terms of their baseline metabolic health and overall dietary pattern, then it's wise to be skeptical of those data. And to be clear, this is not a one-off. Rather, it's typical to see oddly, and in my opinion poorly formulated, animal-based or animal-inclusive protein diets, such as this animal-based keto diet from a famous randomized controlled trial published in 2021 in Nature Medicine. Does this look like a diet you would eat? Doesn't look like a diet I would eat. Okay, that's one high-level point, but I've conspicuously failed to address the question that I opened this video with: is plant protein better than animal protein for preventing chronic disease and for longevity? For this, we typically need to turn to epidemiological studies, given the nature of the outcomes we're assessing. Now, the most talked about And perhaps strongest among these studies as of late was a 2024 study of 49,000 women from the Nurses' Health Study, which was published in the American Journal of Clinical Nutrition. It had an aim to evaluate the long-term role of dietary protein intake in healthy aging among participants in this Nurses' Health Study cohort. And to cut to the chase, the researchers reported that, now quoting from the study, Consumption of total and animal protein was inversely, unfavorably associated with the development of several chronic diseases, whereas consumption of dairy and plant protein was favorably associated with this domain, development of chronic diseases. However, we should assess these results with skepticism for a few reasons. First, higher total protein and animal protein intake was associated not with lower, but higher BMI. So people eating more protein tended to have higher BMI. In fact, the obesity rate in the highest quintile of protein intake was 2 times that of obesity in the lowest quintile protein intake. And this contradicts general wisdom that increased protein, up to a point at least, is a relatively anti-obesogenic macronutrient. So it was just a little odd. You can consider it a red flag of sorts. What's more, higher BMI in those with higher protein intake associated with substantially lower caloric intake by about 330 calories per day, comparing the highest to the lowest quintile in terms of protein intake and BMI, as reported by a food frequency questionnaire. And again, this is odd. I wouldn't say it's impossible, but it is what I'd say is a red flag, or at least a yellow flag, for data reliability. As across a large population, it's unlikely that the group with the highest obesity rates was eating less than 1,600 calories per day. That's not a comment on individual cases, but at a population level, it does make me at least skeptical. But moving on, we again need to critically assess what the women in this study were actually eating. And if you go to Supplement 2 and look at the breakdown for the contribution from animal protein, it actually doesn't look too bad, with a combined 81% of animal protein from beef, chicken, fish, and dairy. However, you need to dig a level deeper to get proper insights. For example, when you actually look back at the original questionnaire in 1984 used to collect these data, the quote meat category includes sandwiches, casserole, and lasagna. And I don't know about you, but in my opinion, Beef crumbled on pasta constitutes an entirely different meal than a pure steak. And this was a general problem across categories. For example, those with higher animal protein intake actually tended to have greater intake of vegetables, but quote vegetables included corn, yams, and sweet potatoes as equal in terms of vegetable count to celery, Brussels sprouts, and kale. And the same was true for fruits, where raisins and orange juice are treated as equal in terms of relative fruit quantity to servings of strawberries. Point being, these sorts of studies tend to cluster food into groups and lose resolution in terms of what people were actually eating. So you need to remain skeptical and critical. And again, we return to the theme of what were people actually eating and does it reflect your dietary pattern? In this case, these women who were eating higher animal protein were also doing so in the context of a mixed macronutrient setting with between 40-50% of calories from carbs and may have been getting their daily dose of animal protein in the form, at least partially, of lasagnas and casseroles. I don't know if that's a fair representation. Okay, well, with that said, what if we were to zoom out and look at a broader level, let's say a meta-analysis of trials on total animal and plant protein? On all-cause mortality. What do we see in the meta-analysis? Well, here, this study reported a benefit of all-cause mortality for plant protein but not for animal protein. And it's a meta-analysis, so it must represent a higher truth, right? Wrong. I disagree. There was large study heterogeneity, which is another way of saying different studies said different things. And there were even studies, such as these two, which showed trending benefit of animal protein over plant protein in terms of all-cause mortality. Now, one could try to accuse me of cherry-picking from this dataset. However, my point is broader. The fact is that there is such noise in these data with different studies showing conflicting findings, and also framed by the fact we already dissected that patterns of eating that persons in these studies are engaging in may be vastly different from yours or mine, All this calls into question the reliability of the conclusions. And finally, I have one more really important point to make: biological plausibility, or in this case, really shaky biological plausibility. You see, while findings from epidemiological studies can be good for hypothesis generation, and also it's not always possible to conduct a randomized controlled trial for outcomes like all-cause mortality, you still need to tell a compelling story with the data. In other words, you need to explain why the data makes sense from a biological perspective. And one might fairly expect, given all the effort that goes into studying the advantages of plant-based diets, this will be worked out. But it's not. In fact, when you go to the biological plausibility and mechanism sections of these papers dissecting plant versus animal protein, they're stunningly weak. And generally fall into a pattern of hand-waving and re-attributing the relative benefits of "plant proteins" to other compounds found in plants with purported benefits like antioxidants, fiber, phytonutrients, etc., etc., or unsaturated vs. saturated fats. To be clear, none of these are protein arguments, they're tangents. And where you can actually find "protein arguments," they're either weak, sneaky, or they're both. Take this excerpt from the paper we went over earlier. This was in the discussion. They said, "Animal protein intakes were positively associated with concentrations of insulin-like growth factor 1, or IGF-1, which has been implicated in the growth of cancer." Now, that's an interesting claim, but if you actually chase the reference they're using, what you'll find is they actually report no difference in IGF-1 between meat eaters and vegetarians, Albeit vegans did have lower IGF-1, but they also had lower BMI and lower overall protein intake. But what was really interesting is sneakily in this paper, soy protein was actually clustered with animal protein because, quoting from that paper, "Intake of protein rich in essential amino acids was positively associated with serum IGF-1." So they actually changed the clustering of protein types, putting soy with animal protein in order to create a particular narrative. Now, what I mean for you to take away from this is it can be really tricky to trace the validity of claims in a paper, since to do so you often have to reference dig, as a claim in a paper may not be well matched to what the paper is actually saying, or the reference paper is actually saying. And with biases in the peer-reviewed literature, which is not a meritocracy, you can end up with a narrative creep that skews common opinion potentially in an inaccurate direction. Okay, so all that said, just summarizing up my 5 high-level thoughts, the things I want you to take away from this video. When it comes to evaluating studies, first ask yourself, what was this population really eating and does it reflect my dietary pattern? Also, what is the metabolic health status of this population and does this reflect my metabolic health state? I hope it's obvious, but a fit, healthy person eating a steak and eggs on a low-carb keto diet is not the same metabolically speaking as a person with obesity and diabetes eating a Reuben sandwich at a deli with a side of french fries. I mean vegetables, because of course french fries are potatoes and potatoes are vegetables. Okay, moving on. Point 2: when it comes to protein, get enough and get it from high-quality sources rich in other nutrients. This includes beef, eggs, and fish, intake of which has actually been associated in other studies with improvements in all-cause mortality and cognitive longevity. So yes, I definitely think fish is a health food. Moving on, 3, you can also have plant-based proteins, be that legumes or tofu if you prefer. I have nothing against plant proteins. You do you. Although personally, I've never had any tofu that compared to a nice juicy ribeye or fatty filet of salmon. But that's my preference, it doesn't need to be yours. 4, track your biomarkers. Some papers claim that, for example, intake of animal proteins is associated with insulin resistance and metabolic dysfunction. I think this can be skewed by healthy user bias, but taking it to a real level and figuring out if those data are relevant to you, like if you're eating 6 eggs and a steak every single day and you have no signs of metabolic dysfunction, then clearly those data don't apply to you. So you do you because this population is clearly different. Finally, 5. This is my last point and it's actually super critical, so listen up. Always be mindful where you place the burden of proof in an argument. This video was clearly structured to assess the claim that plant protein is better than animal protein for chronic disease and longevity. I did not claim the counter. Rather, I dissected the data underlying the claims of plant over animal protein and found it to be grossly lacking with respect to details of diet design, generalizability, data consistency, and honestly, biological plausibility. Don't underestimate this approach, because if you want to win an argument against someone who is making a claim, a very bold claim, the best way is typically to keep the burden of proof on them. Don't let them shift it to you. They are making the claim, and they need to support it. And well, If their data are weak and their logic broken, I say bring on the steak and eggs.