Unfortunately this discussion devolved in to low vs high protein arguments. That’s a misguided discussion IMO.
The issue is what protein source and more specifically what amino-acid composition is associated with healthspan ?
Not a ton of studies but seems like plant protein may be more beneficial for a couple of reasons:
They are typically lower in methionine and (iso)leucine - thus less mTOR activation.
They typically come packaged with fiber, resistant starches, minerals, phytonutrients and vitamins.
Animal sourced protein comes often processed and tends to accumulate more toxins due to biomagnification. It also comes with bad saturated fats (there are good kind as well).
The executive summary is that it may be hazardous to listen to the suggestions of Peter Attia, who advises 2.2 g protein /kg/d. Excess of protein seems to be related to a higher risk of cancer, more so with animal protein, less so with plant-derived protein.
At the end of it, it is desirable to optimize the risk, optimizing protein intake.
Keep in mind that 4 bit precision with 120 b weights it’s virtually guaranteed sources will be hallucinated and there’s no good reason they should try to fix it (instead rely on search tool calls for citation). The text is more likely to be true as there’s more of it or can reasoned to than specific links.
It’s fantastic for medical text generation and reasoning though, basically anything other than its drawbacks like link urls (which more precisely is generating url links for some passage of text, not exactly easy to do from memory).
Regarding the complaints, personally I also only summarize what I’ve learned, unless we’re comparing AI responses or similar. A forum function of allowing to place certain text in maybe 15 lines of text and allowing for expansion would solve the problem of long AI responses on the forum that people just want to scroll by.
Now, I posed this question to DeepSeek R3 (gpt-oss is no longer available for today).
Q: what do you think of Peter Attia’s recommendations on high protein intake and how to reconcile it with the recommendations of Valter Longo, who suggests about one third of it?
Now, the answer is pretty extensive and it can be replicated by posing the same question tio deepseek R3. So I had it summarize it. It is a pretty balanced and pragmatic reasoning.
Here’s a concise summary preserving the key insight:
Conflict:
Peter Attia advocates higher protein (1.6–2.2 g/kg/day) for healthspan—preserving muscle/metabolic health, especially with aging.
Valter Longo recommends lower protein (~0.7–0.8 g/kg/day, mostly plants) for lifespan—reducing IGF-1 to activate cellular repair (autophagy) and lower cancer risk.
50 or active: Lean toward Attia’s range (1.6–1.8 g/kg).
Cancer risk: Lean toward Longo’s approach.
Key Insight (Unchanged):
Attia and Longo address complementary goals:
Attia prevents physical decline now (muscle/function); Longo targets cellular aging long-term. The optimal strategy integrates both—enough protein to maintain strength (primarily from plants), with periodic reduction to activate repair pathways.
This balances muscle preservation and longevity mechanisms without fixed dogma.
Now, same question to GPT-5, with a very concise summary. The extended version is interesting and captures some nuances different from deepseek, thsoe interested can pose the question directly.
Here’s the short version:
Attia: High protein (1.6–2.2 g/kg) all through adulthood to maintain muscle, strength, and metabolic health.
Longo: Lower protein (~0.7 g/kg) before 65 to reduce growth‑pathway activation linked to aging; slightly increase after 65.
Why they differ: Attia prioritizes muscle preservation; Longo prioritizes longevity via reduced IGF‑1/mTOR signaling.
Middle ground: Adjust intake by age, activity, and health — possibly cycle protein levels, use more plant sources, and track biomarkers to fine‑tune.
Last, the conclusion of Gemini Flash 2.5. I didn’t know it, but there is a video on low vs. high protein with Kaberlein and Attia. Have to watch it.
Reconciling the Two Perspectives
While their recommendations appear to be in direct opposition, they can be reconciled by considering the specific goals and life stages they address:
Age-Dependent Recommendations: Longo himself acknowledges that protein needs increase with age. He recommends a higher protein intake for people over 65 to prevent sarcopenia, bringing his advice closer to Attia’s. This suggests that the trade-off between growth and repair shifts as we get older.
Context and Goals: Attia’s recommendations are tailored to those actively engaged in strength training and focused on maximizing muscle mass and performance, particularly as they age. Longo’s advice is more for the general population seeking to minimize disease risk through dietary patterns that have been associated with extreme longevity.
Cyclical Approach: One way to integrate both philosophies is to cycle between phases of higher protein intake (e.g., to build muscle) and lower protein intake (e.g., during periods of caloric restriction or fasting mimicking) to trigger cellular repair. This approach leverages the benefits of both strategies at different times.
In essence, Attia focuses on the “hardware” (muscle) of longevity, while Longo focuses on the “software” (cellular pathways). Neither view is definitively “right,” but rather, they offer two valid, evidence-based approaches that can be applied depending on an individual’s specific health goals and stage of life.
Moderate resistance training plus ~ 1g/kg/d is working fine.
Previously posted picture taken in Feb. 2024 at age ~84. I will post a new one sometime this fall. Spoiler alert: nothing has changed except for a few more wrinkles from intentional weight loss. I don’t count calories, but I do keep track of protein. I make sure I get at least 10 grams of leucine on training days.
Yes, I too have noticed that higher amounts do not contribute much to muscle hypertrophy, unless the workouts are really grueling.
For one period I went up to 1.5 - 2.0 g/kg/d and more, applying all the rules on taking advantage of the metabolic window, ingesting high-quality whey isolate, and so on. To no avail. Supplying more raw material to the building cycle makes no sense unless the workers work double or treble shift (unless we take steroids).
So, there is a natural threshold to the translational process and ribosome’s capacity to synthetize protein and muscle protein specifically. The surplus of amino acids is simply turned into other nutrients and waste.
I don’t know why Peter Attia and others insist on huge amounts of protein, often hard to ingest and digest. Moderate amounts plus resistance exercise will usually guarantee sufficient muscle hypertrophy.
At my age, the law of diminishing returns comes into play, and very little muscle is gained, no matter how hard I try. Since I have been doing resistance training for many decades, I am now only trying to maintain muscle mass. And that is the reason I only do moderate workouts.
That chart isn’t right though unless you are pretty advanced in age. Lots of people can gain a few pounds of lean mass per year even after a decade or more of training.
I thought I saw you are 84, so you are right at that cut off from what I’ve read, so it may be true for you. But if you are maintaining, that’s definitely a win. I’m a big believer in eccentric strength for longevity. Even with heavy deadlifts I do slow eccentrics and barely touch the plates to the floor before the second rep. I don’t do singles for this reason.
It is unclear to me how much of the benefit of high protein is that your body will respond with greater muscle protein synthesis vs the satiety from prioritizing protein that results in fewer calories overall. If you count calories and have a target in mind, maybe it matters less. In my experience, if I have a protein goal and a fermented food goal and a vegetables goal, I don’t have room (time or appetite) for much else.
Yes, the genetic potential is an individual response for gaining muscle mass. This appears to follow, like many things in nature, a “Bell Curve”. I was very skinny as a youth, and an ectomorph, "characterized by a lean and slender build, with long limbs and a narrow frame. "
“They often have a hard time gaining weight (both muscle and fat) and may appear naturally thin.” At the gym I was in the lingo, a hard gainer. I put in a lot more effort for the same gains as a good gainer.
As opposed to mesomorphs, who are known for their athletic and muscular build, with a naturally strong physique. “They tend to have broad shoulders, a well-defined chest, and relatively narrow hips, making it easier for them to gain muscle and lose fat.”
So, if you enjoy doing hard exercise, go for it. I became bored with the gym many decades ago. That is why I only do exercise in moderation.
I really, really like the idea of this approach. Overnight fasting (plus a late night “bump” of sulforaphane, for instance, for extra autophagy/recycling) along with perhaps a once monthly 5-day fasting-mimicking protocol to really boost recycling/repair while minimizing loss of lean mass. Otherwise, high quality protein and progressive resistance training throughout the month for lean mass and strength gains.
That sounds too frequent on average; in myself, it would trigger an excessive catabolic signal, and I wouldn’t be able to make up for it during the other weeks.
Always in myself, I have noticed that the body tends to follow naturally cycles of higher and lower food intake (and this of course decreases and increases cyclically the intake of protein).
Diabetes Atlas further notes about 10.8 million cases in 2024, placing Japan among the top ten countries globally in terms of number of adults (20–79) with diabetes
