Creatine, AI Doctors, and Sleep Trackers: Notes from the First Don't Die Podcast
Bryan Johnson
This first episode of the Don't Die podcast is a loose weekly roundtable. Bryan Johnson is joined by Kate Tolo, his co-founder at Don't Die, and Dr. Mike Mallin, his lead physician. Johnson says he talks with Tolo roughly 150 times a day and with Mallin five or six times a day, and that the goal of the show is to share the ongoing work of refining his protocol. The conversation covers four things: what emergency medicine looks like from the inside and how patients can advocate for themselves (including with ChatGPT), a new study questioning creatine's effect on muscle, how body awareness develops once you start measuring yourself, and Andrej Karpathy's personal comparison of sleep trackers.
From the ER to "future-focused" medicine
The episode opens with some banter about plants. Tolo's leafy background turns out to be fake vines. Mallin admits his wife looks after their houseplants better than he does, because he waits until they wilt before watering them. Tolo jokes that this is the ER doctor in him: he triages the plants and only responds when death is imminent.
The joke leads into a more serious comparison. Johnson asks Mallin how a day in the emergency department compares to a day working on his protocol, and Mallin says the two "couldn't be any more different." He describes emergency medicine as "banging your head against the wall all day long." In his account, patients arrive with problems that stem from a culture that doesn't value health, and they show up at the end of a disease process asking to be fixed. His current work, by contrast, is about optimizing someone's health 20 years out rather than trying to reverse terrible trends. He says the ER "was tough on me."
Johnson asks how much of what arrives in the ER could have been prevented. Mallin's first answer is "at least 90-plus percent," which he then revises to about 80% for emergency medicine specifically. Acute events such as car crashes, appendicitis, gallbladder problems, and infections are hard to prevent. Chronic disease, though, is in his view the dominant issue across medicine, and much of it is preventable.
Routine versus judgment in emergency care
Johnson recounts injuring his hands years ago when a five-gallon glass water jug shattered while he was holding it. The glass lacerated both hands and severed a tendon in his left index finger. A nurse told him they see about one such injury a day, usually from scooter accidents. She mentioned a 17-year-old girl the day before who had badly injured her face in a scooter crash. Johnson says he now wants to tell every scooter rider to stop. While he was in the ER, he looked up the standard procedure for his injury and found it laid out as something like a 19-step process. He asked Mallin how much of ER work follows routine and how much is the physician's own call.
Mallin says a lot of medicine is routine, with established standards of practice. The "art" is in diagnosis. He compares treatment pathways to pre-planned Disney rides: the doctor's job is to put the patient in the right seat. A hand laceration like Johnson's is easy, and Mallin says he could walk him through it with his eyes closed. The hard cases are vague complaints like headache, belly pain, or feeling weak and dizzy, which could be produced by thousands of different conditions.
America versus New Zealand
Tolo says the ER in America feels like a war zone, with people who seem to be dying and not getting seen. She adds that she doesn't know whether that impression is accurate, and that it has been ten years since she lived in Australia, whose system she considers better. Mallin agrees that American ERs are chaos, to varying degrees depending on the hospital. He hasn't worked in Australia but has worked in emergency departments in New Zealand. The biggest difference he saw was the ratio of physicians and nurses to patients. On a regular US shift a doctor might see about 30 patients, compared to maybe 10 in New Zealand, roughly three times the workload in the same time. In the US, he says, you are "not as much putting out fires as you are just trying to stop the speed of the burn." Critical cases like severe trauma and cardiac arrest get full attention. Chronically ill patients are checked to make sure they aren't dying and are then sent back to their regular physicians.
"Throw your complaints into an LLM"
Tolo's mother recently went to the ER with an eye problem, and she asks Mallin how people should advocate for themselves inside a system with its own procedures. Mallin prefaces his answer by saying his ER friends will "kill" him. His advice is to enter your complaints into a large language model, ask what should be happening, and then make sure it happens. He acknowledges that doctors often dislike patients who Google their symptoms. He argues, however, that this kind of active self-advocacy matters because the system is broken. Staff are overwhelmed and likely burned out, so a passive patient will be placed on whatever path is easiest for the physician and nurses.
Tolo distinguishes this from googling. She thinks the stigma around googling is partly deserved, because people have a cognitive bias toward catastrophe and fixate on rare outcomes like cancer. In her experience, ChatGPT gives more reasonable, better-calibrated answers: probably this, and in rare cases this. Mallin adds a caveat. The quality of the output depends on the quality of the input. He knows which questions to ask and which positives and negatives to include, and he isn't sure someone without medical knowledge would get equally good answers. He notes that ChatGPT has outperformed physicians on standardized tests, so it should do well in "standard" situations, but "life is rarely standard."
Tolo describes what she actually did. With her mother on the phone, she told ChatGPT to treat her as a patient and conduct an intake. It asked about symptoms, and she relayed perhaps a hundred questions back and forth. It then recommended exams and next steps. Her mother wasn't seen by an ophthalmologist in the ER, so she went to an optometrist for exams. According to Tolo, ChatGPT had essentially correctly identified the problem as some inflammation of the eye and a scratched cornea. Her mother is now on what Tolo believes are steroid eye drops.
The creatine study
The "topic of the week" is a recent UNSW study. Tolo summarizes it: taking 5 grams of creatine daily, a standard dose, did not produce more muscle gain than training without it. Over 12 weeks, both groups gained the same amount of lean muscle, about 2 kg. Early weight gain in the creatine group likely came from water retention rather than muscle. Johnson points out that creatine's audience is broader than gym-goers. He calls it one of the most commonly taken supplements, possibly the most common, and notes that it is used for cognitive benefits as well as muscle. The study measured only muscle.
Mallin finds the study's design interesting. The classic way to start creatine is a loading phase of about 20 grams a day for five to seven days to saturate the muscles. Because creatine increases fluid in the body, lean mass readings go up during that period. The researchers tried to account for this by measuring lean body mass at day zero and day seven, which they called a "wash-in" period, and only then began the 12-week comparison, with both groups following the same training plan. Mallin's reading is that earlier studies may have found muscle gains because they didn't separate out this water effect.
He raises several caveats. First, he thinks the wash-in was too short. At 5 grams a day without a loading phase, which is what this study used, saturating the muscles takes three to four weeks, not seven days. Second, he says the evidence that creatine increases muscle mass "was never actually that good," since earlier studies showed only small changes. In his view, the stronger evidence for creatine concerns cognitive function, bone, muscle-related longevity, metabolic health, and performance measures like speed and power. Those performance effects appear mostly in already well-trained athletes, not the untrained participants in this study. So what the study shows, according to Mallin, is that untrained people on a valid 12-week program won't see significantly different muscle mass with or without creatine. It says nothing about cognition, recovery, or power and strength output. He adds that 12 weeks is short, and that trained athletes followed for longer might show a difference.
Tolo says that, as someone "on the lighter side of science," she feels like a new study debunks the old consensus every week, which can paralyze ordinary people. Mallin's recommendation is to change nothing. By his account, creatine has more data behind it than almost any other supplement. The data is overwhelmingly positive, some studies show no benefit, and none show harm. He attributes the attention this study received to a culture that finds negative findings more interesting and likes headlines about accepted beliefs being overturned. Later in the episode he also notes that the study had about 60 participants, "so tiny," set against decades of research, and that new findings tend to be assumed better simply because they are new.
How much creatine to take
Johnson notes that Blueprint's Longevity Mix contains 2.5 grams a day, while many people take much more, especially during intense resistance training. He asks how to choose a dose and measure whether it works. Mallin says efficacy is hard to measure. For general health he would dose by weight at about 0.1 g per kilogram per day, which works out to roughly 7 grams for a 70 kg person. A flat 5 grams isn't precise, he explains, because smaller people, often women, may need less and people with more lean mass may need more.
Diet matters too. Mallin estimates that a generally healthy standard American diet already supplies 1 to 2 grams a day, so such a person might need only about 2.5 additional grams. Vegans and people who eat little red meat get less and probably need a higher supplemental dose, since the goal, at least for performance, is muscle saturation. If he had to give a single number, he would say 5 grams a day is adequate for an average-sized person with a normal diet. A loading phase isn't necessary, because saturation happens naturally over three to four weeks.
For cognitive purposes the doses are different. Mallin says there is decent data that creatine improves cognitive performance under high metabolic demand, such as brain trauma, poor sleep, or mild cognitive impairment and early dementia. Those studies used around 10 to 20 grams a day. He cites a study of up to 30 grams a day for up to five years with no significant downsides. As far as he can tell, excess creatine is simply excreted.
Tolo started with 5 grams when she joined Blueprint. Now she takes the Longevity Mix's 2.5 grams plus an extra half scoop to reach 5. At around 57 kg, she concludes that this is about right for her under Mallin's formula. Asked whether blood work can guide dosing, Mallin says not really. Creatinine, a kidney-function marker, will likely rise and appear falsely elevated with supplementation, but he wouldn't use it to judge dosing because hydration and kidney function can confound it. When Johnson asks how Mallin's own protocol will change because of the study, he answers, "Absolutely nothing."
Johnson says he weighs 77 kg and currently takes 5 grams: 2.5 from the Longevity Mix, and he is experimenting with 10. Because he usually has to get up around 3 a.m. to catch flights and sleeps less on travel days, he is trying around 20 grams on those days to see whether he feels more cognitively fresh. He has done this only twice and hasn't noticed a difference yet.
Tolo proposes something like a "travel advisory" for health. When a study goes viral, a central source would explain what it means, for example that a new creatine study shouldn't change your behavior, and why. Johnson agrees. His concern is that such headlines settle into the background of people's health consciousness as shorthand ("I remember it didn't really help muscle") and eventually harden into accepted truth.
Body awareness, and its strange side effects
Johnson describes himself as intensely body-aware after several years on his protocol. He says he can estimate his heart rate at any moment with pretty good accuracy, and that the many treatments he undergoes have taught him to inspect his body for color, tone, and function. He contrasts this with how unaware he used to be, when he would simply push through a headache. His explanation is that pairing measurement with how you feel builds fine-grained intuitions, effectively turning you into your own sensor.
Mallin raises a downside. He and his wife recently walked through a Las Vegas casino on the way to a show and wondered how everyone was still upright while drinking, smoking, and eating pizza and corn dogs. If he did that, he says, he would "literally be on the ground." Once you get healthy, he says, you can't allow yourself to slip, yet these people appear to be thriving without obvious back pain. He wouldn't choose differently, but he calls the position "unique." The response in the conversation is that the body is incredibly adaptive, able to postpone big problems for a long time before they come crashing down.
This recalls a study at Kernel, Johnson's brain-measurement company, on inebriation. Johnson explains that participants were tested sober and at low, medium, and high levels of intoxication. At low and medium levels, people could behave and pass tests as if they weren't intoxicated, but the brain scans showed impairment that the brain was compensating for. At high levels the compensation failed and impairment appeared in both behavior and the scans. Johnson draws two lessons. Cognitive decline begins long before symptoms appear, which makes early detection possible. And brain measurement can reveal things that self-perception misses. Mallin summarizes the analogy: you're "more drunk than you are aware," and the casino patrons may be closer to trouble than they feel.
Johnson then describes a recent team dinner of about 25 people at a well-known health restaurant in Venice, California. Watching the food arrive at tables around the room, he says he was "beside myself," given what he has learned about food, toxins, and how food moves through restaurant systems. His team teased him with "welcome to normal society." He says it felt like traveling back in time to watch a movie of the early 21st century, because the system they have built is so far outside that norm.
When will we know the protocol works?
Tolo asks how long it will take to know whether Johnson's protocol truly works, suggesting aging may behave like a cliff around age 70. Mallin says the cliff's location differs for everyone depending on genetics and lifestyle. For now, the best available approach is tracking Johnson's biological age and his speed of aging. Mallin reports that they have slowed his speed of aging to below 0.5, which he calls "pretty phenomenal." But to truly "not die," he says, the process has to keep slowing further. More still needs to happen to avoid a cliff at some point.
Karpathy's sleep tracker experiment
Tolo brings up a post by Andrej Karpathy, who ran a two-month n-of-1 experiment comparing four sleep trackers: Oura, Whoop, Eight Sleep, and Apple. He rated Oura and Whoop top tier. What interested Tolo most was his conclusion, which she reads aloud. Karpathy wrote that he could say "with absolute certainty" that Bryan is basically right and that his sleep scores correlate strongly with the quality of his work. With low scores he lacks agency, courage, and creativity. With high scores he can work 14 hours and barely notice time passing. He added that the effect depends on accumulated sleep over the last few days rather than a single night: one bad night is usually fine, but several in a row is bad news. Tolo says this matches her experience. People who sleep badly regularly can't tell what sustained good sleep feels like.
Mallin frames the post as evidence for measurement. Objective data helps you interpret subjective signals, and combining the two builds the knowledge to make good health decisions. Tolo's standard advice is to start measuring even without any interventions, so that you build a relationship between how you feel and what the data shows, and only then begin changing things. She jokes that Johnson should challenge Karpathy to beat his eight-month sleep score.
Johnson describes Karpathy as a founding member of OpenAI and former director of AI for Tesla's autonomous driving program, and calls him "one of the most formidable intellects of our time." He connects the post to his own thinking after selling Braintree Venmo. He concluded that humanity is at a pivotal moment, evolving into a new species, and that the most important thing to do was to improve our own intelligence so we could see the moment clearly and act wisely. That was the motivation for Kernel: measure the brain, reveal what is invisible, and pair it with AI to improve intelligence faster than otherwise possible. By intelligence he means not only IQ but emotional development, correcting blind spots, and avoiding narrow, tribal worldviews. As Johnson sees it, Karpathy is building the future of intelligence through AI and, through this sleep experiment, is also building his own. That an "architect of superintelligence" would endorse investing in one's own intellect through sleep and self-care felt to him like the best outcome he could ask for, and a reward for daily efforts urging people to stop eating junk food and drinking alcohol, stop going to bed late, prioritize sleep, exercise, and measure themselves. He says he greatly appreciated Karpathy's kind words.
Open questions about the format
The hosts end unsure whether the format works and ask listeners to comment on what they would change. Johnson notes they had six or seven more topics outlined that they didn't reach, including an embryo selection technology currently on the market. Mallin says he measures a podcast's quality by whether it was fun, and by that standard this one was.
We went out to a restaurant. The team was all together. I just so rarely do this. I was just shocked at watching how the food arrived at various tables across the restaurant. I mean, it was just — I was beside myself. But I guess knowing what I've learned about food and toxins and the way that food is passed through restaurant systems, the team was kind of like, "Bryan." Yeah. Like, "Hey, welcome to normal society. This is kind of what people do." So we've built this system that's just so outside of that. It was very shocking to me. It almost felt like I was traveling back in time and viewing the early 21st century, and like, this is crazy, watching a movie or something.
Today everyone, welcome to the... what is this? The Don't Die podcast. Yeah, that sounds great. Yeah, there we go. So, Don't Die podcast. We believe we are the first generation who won't die, and we are trying to practice that philosophy and we're trying to bring it to you as well, so you and your family and your friends can also do it. So with me today, Kate Tolo is my co-founder of Don't Die, and Mike Mallin, who is my doctor. I mean, Kate and I talk like 150 times a day. Mike and I talk five or six times a day, and we're always chipping away at everything we can do to improve protocol. So we're going to go through the various things we're working on on a weekly basis. Also, we'll fill some of your questions in time as we create this dialogue.
Kate, is that your real background? It is. Prove it. You guys are killing me on the background. Are those vines? Yeah, they're fake. Oh no. What does that say about me? Is this falsely alive? Or it's just convenient. Yeah, when we put a whole bunch of plants in the Kernel office, we wanted to make it like a forest or an arboretum, but it's still a gorgeous environment. It just took us some time to find the natural stable state for nature to coexist with the mechanical world. Yeah, the ecosystem. Yeah. It's amazing how much of a difference some plants will make, which you can obviously tell based on my background.
Do you have plants in the house? I do. Yeah. In the other room, not here. I need to move them. Do you maintain them? Sort of. Not very well. Does somebody in the family have the responsibility of taking care of the plants? Yeah, I would say my wife does a much better job of taking care of the plants than I do. I wait until they wilt and then I water them. She proactively waters them. Signs of leaving life. Maybe that's because you were formerly an ER doctor, you're basically waiting for the plants to show up and be like, "I'm in urgent need of some care." Yeah. Unless you're dying, don't talk to me, right? And you just triage the plants. Which one needs attention? If you're dying at a certain rate, like if death is imminent, I'll talk to you. That sounds about right.
That's — I'd like to think that I've changed, hopefully. I was going to say a little bit more future focused these days. I mean entirely. No, you spend basically all of your time on the frontiers of how to slow down speed of aging and reverse aging damage, pretty much. Which is super fun. It's a way different experience than emergency medicine. That's for sure.
What is that like emotionally and intellectually, if you just contrast a day in the ER with a day in our life together? Oh wow. It couldn't be any more different, to be honest. I mean, emergency medicine is just like banging your head against the wall all day long, in the sense that you've got all of these issues in front of you that are secondary to a culture that doesn't value health, right? And the side effect of that is people get super sick, and then at the very end of life or at the end of their disease process, they come into the emergency department and they present to you like, "Hey, fix me." And you're stuck with all of these problems that are all ingrained in culture. And now my day-to-day is thinking so future focused, like years down the road, in terms of optimizing health rather than trying to reverse these terrible trends, or focusing on your health or someone else's health 20 years from now. And yeah, it couldn't be any more different, which is I think why I enjoy it so much. The ER was tough. It was tough on me.
What do you think percentage of people that end up in the ER, how much of it could have been prevented? Oh, at least 90-plus%. I mean, eventually most of that disease would eventually happen. But for emergency medicine, maybe 90 is a little aggressive, maybe 80%. Because a lot of it is acute, like trauma. You're driving down the road, you get in a car wreck, or you get appendicitis, you get gallbladder issues. Those are hard to prevent, right? Infections are hard to prevent. But the majority of what you see in medicine in general is chronic disease. Chronic disease is the most prominent issue in medicine these days. And so much of chronic disease is preventable. So quite a bit.
When I hurt my hand years ago, I was holding that five-gallon glass container, a water jug, and it shattered in my hands and it just lacerated both my hands. It severed the tendon on my left finger, my pointer finger. So I was trying to assess the damage, and my hand was just filleted. And I tried to close my hands and this finger wouldn't go down. I was like, "Oh boy, this is a bad situation." And so when I went into the ER, the nurse looked at me, she's like, "Oh yeah, we see one of these situations a day. It's typically from the scooters, where people come in." Yeah. And she's like, it's just like clockwork. And so now every time I see someone on a scooter rolling around, I'm like, "Please stop what you're doing right now. Don't do this." And she was saying that the day before they had this 17-year-old girl, beautiful girl, and she just destroyed her face with this terrible scooter accident.
So, but when I went in there, I was going to say on the ER, I looked up in real time what the ER procedure was for the conditions of an injury like my hand, and it laid it out in something like a 19-step process. So when you're in the ER, how much are you following routine steps, and how much are you basically making the call?
Oh, that's a good question. There is a lot of routine in medicine. In traditional medicine, there are standards of practice that are very much routine. So the art of medicine comes in the diagnosis. It's putting the person in the right seat for them to go on their Disney ride, right? So there's lots of different Disney rides that are pre-planned, that are pre-organized, and it's putting them on the appropriate seat. So once you make the diagnosis, so much of care is standard. So very much, the majority of it. The challenge — not for a laceration of the hand, like your case, Bryan, that's easy. I can close my eyes and take you through that journey, no problem. The challenge is when someone presents with a very vague complaint that could be a thousand different things, and trying to figure out which one of those things it is. So like, I have a headache or something, or belly pain. Yeah. Yeah. Or I feel weak and dizzy. These things can be produced by thousands of different ailments, and trying to figure out which ailment it is is the challenge.
Mike, I don't want to just spend the time interviewing you, but I feel like people in the ER... Yeah, it's great. Let's go there. I feel like people — what is the secret sauce? You show up in the ER, how do you get done? Because I feel like most people show up in the ER, and if you've never done it before, it's a war zone, especially in America, right? It's like, what's the protocol? You're saying it's different than Australia? Well, it's been 10 years since I've lived in Australia, but Australia has, I would say, a better medical system than America. And you're saying specifically in the ER? My experience has been, yes, the ER here, it's bad. You show up and it's like people are dying and they're not getting seen. That's how it feels. I don't know if that's actually true, but it definitely feels really disastrous in there.
It is chaos. Yeah. And depending on which ER you go to, there's a variety of chaos, but it's generally chaos everywhere. I will say, I haven't worked in Australia, but I did spend some time working in New Zealand in the emergency department there, and it was extremely different, mostly with the degree of providers, physicians and nurses, to patient ratio. It was much different. Whereas in the States, on a regular shift, doctors are seeing like 30 patients a day, whereas in New Zealand you'd see maybe 10. So, 3x the level of effort over the same time frame. So it's just chaos. You're not as much putting out fires as you are just trying to stop the speed of the burn. And then there are some fires where obviously those fires require all of your attention. The severe traumas, the sick patients, the people coming in cardiac arrest, those are the ones that require all the focus. And then the people with the chronic illnesses that you're not really going to do much with, you just try to make sure they're not dying, and then you send them off to go back and see their regular physician. So yeah, it's hard to describe, but it is chaos for sure.
So, like Mike, if a loved one showed up in the ER, what would you recommend to them? I just had this — my mom just had an eye issue and showed up in the ER, and it's like, what do you do? How do you advocate for yourself in the system when the system has its own procedures? So what recommendation would you give to people?
My ER friends are going to kill me. Yeah, exactly. We need them behind the scenes. Yeah. I would tell them, honestly, I would throw your complaints into an LLM and ask it what should be happening, and then make sure that that happens for yourself, if I'm being totally honest. This feels like we should blur Mike's face, right? And voice, and be like, we just referenced an anonymous source who's an ER doctor at some prestigious center. Doctors always talk about how much they hate it when their patients Google their problems and then come to them with questions about their Google problems. But that is such an active form of advocacy for yourself, and it's honestly so important in our system, because our system is so broken. So if you're not asking the questions, you're going to be placed on that easy path. And it's going to be the easy path for the physician, for the nurse, for everyone in the system, because they're getting destroyed. They're getting absolutely murdered on a day-to-day basis. Chances are good that they're burned out. And if you're not advocating for yourself, you put yourself in a challenging situation.
Yeah. And I feel like that stigma around googling things is valid, because we have this cognitive bias toward catastrophe. And so when we research something and it has a possibility of cancer, we go down that path. As opposed to, I feel like ChatGPT, when I've interacted with it from a healthcare position, actually gives me really reasonable advice. It's like, well, it's probably these things. In the really rare cases, it's these things. So it feels more informed than obviously just googling something.
Yeah. I think the challenge is it's going to give you good advice based on how good you are at giving it information, right? So the quality of which — like, I feel like because I know the right questions to ask, if I give ChatGPT all of the symptoms that somebody is presenting with and all of the pertinent negatives and all those things, it's going to give me great responses. I don't know if that's true for someone who doesn't understand physiology or medicine quite as well. Are you going to get as quality of responses? I can't say that's true or not, but we do know that ChatGPT, for example, has performed better than physicians now on standardized tests. So in situations where things are quote-unquote standard, we would expect it to do well. Life is rarely standard. But yeah, interesting world.
Yeah. What I ended up doing with my mom was I had her on the phone and I had Chat up, and I said, treat me like a patient and do an intake with me, like, ask me my symptoms, and I just went back and forth. And so on the phone with her, I was asking probably a hundred questions, and eventually it came back with, here is what I would do, here are the exams I would do. And then she went off on her way on her Disneyland ride. And Kate, did you get resolution to that? Yeah. So she didn't get seen in the ER by an ophthalmologist, and so she went then to an optometrist and got exams. And yeah, ChatGPT essentially correctly diagnosed that it was some sort of inflammation of the eye, scratching of the cornea. She's now on, I think, steroid eye drops to resolve it. So yeah.
All right. So we have a topic of the week. I don't know if you guys got to check out the study. It was essentially suggesting that creatine might not be as beneficial as what we thought it might be. So let me read you a little summary. The study found that taking 5 g of creatine daily, which is a standard dose, did not lead to more muscle gain than training without it. Over 12 weeks, both groups, with and without creatine, gained the same amount of lean muscle, 2 kg. And any early weight gain from creatine likely came from water retention and not real muscle. So what does this mean for the gym bros out there that have been taking creatine for 10 years?
I think it's a broader group of people than gym bros taking creatine. I think creatine is one of the most commonly taken supplements across society. Maybe the most. Yeah, maybe. Wow. Yeah. I mean, it's not just for muscle but for cognitive benefits. It has a very broad appeal. Yeah. And this study was just looking at muscle, I believe, right? It was only measuring that. Yeah. So what do you guys think? It's interesting.
Well, what I'll tell you is I think the most interesting thing about the study is they actually tested the wash-in period, which is what they call it. So if you're familiar with creatine, the classic gym-bro style to use it is to do this 20-gram ramp-up for like five or seven days, right? And during that time frame, you basically saturate the muscles with creatine. And why that's important is because when you increase creatine in your body, you increase the amount of fluid in your body as well. So what this study tried to do was account for that by measuring the lean muscle mass, or the lean body mass, at zero days and then 7 days in, and they call that the wash-in period, right? Because there's a change in those people who are on creatine, because they're holding on to more fluid, right?
And then after that is when they started their study. They did a 12-week study where everybody did the same training plan, and then one group took creatine and one group did not take creatine. So what's interesting is they accounted for the extra water weight that you take on when you start creatine. And the difference that that amounted to, at least for this study, as compared to other studies that have looked at muscle mass, is they found no change in the creatine group, as opposed to prior studies that found change. But the theory is they found change because they didn't include this wash-in period, basically. So that's
That's interesting in my mind. A couple of caveats. They did a 7-day wash-in period when it really takes longer than that. Like if you're not going to do the 20 grams a day and you only do five grams a day, which is what this study did, it really takes 3 to 4 weeks to saturate the muscles. So, they didn't really do a long enough wash-in period in my opinion.
And then the other thing I'll say is that the data that creatine actually increases muscle mass was never actually that good. Like the prior studies that we have are like very small changes in muscle mass. The data that's good for creatine is actually the things that Bryan was mentioning, the longevity aspect stuff, like the, you know, cognitive function, the bone muscle mass, metabolic health. That's the stuff to me, honestly, that's really interesting. And then from a performance standpoint, speed and power, those are the things that change, and those are mostly in already well-trained athletes, not these untrained athletes that the study was in.
So what did we learn from the study? We learned that people who are not particularly trained, who do a 12-week exercise program, that was a valid exercise program, are not going to have significant differences in muscle mass whether they take creatine or don't take creatine. But we didn't learn anything about their cognitive function. We didn't learn anything about their recovery. We didn't learn anything about their power or strength output.
Okay. So, when it's muscle mass specifically, it's still an unknown how it affects it, but there's plenty of potential upside elsewhere, sounds like.
Yeah. And potentially like longer term, you might see something. 12 weeks isn't a long time. You know, 12 weeks for training is not much. So, if you took more highly trained people that are performing at a higher level and you studied them with five grams of creatine or not, you might actually see a difference, especially if you watch them for a longer time period. 12 weeks is just not that long to perform a study.
Yeah, I always struggle with these things because, as someone who is really on the lighter side of science, it feels like every week a new study comes out that debunks what was previously known. And it kind of paralyzes the average person, 'cause like, I take creatine every day. Should I stop? Should I keep going? What does this mean? Headlines take it out of proportion. So yeah, for those who are taking creatine or are not, what would you say? What would you recommend?
I wouldn't change a thing about your creatine intake based on the study, to be honest with you. I mean, creatine has, as Bryan was mentioning earlier, more data backing it than I think almost any other supplement that we're aware of. And the data is overwhelmingly positive. There are certainly some studies that have shown no change, no benefit, but not studies that show harm. And there's a lot of studies that have shown benefit, and not just one area, but multiple areas of life. So, personally, I read the study and it doesn't change a thing for me in terms of how often I'm going to suggest that people take creatine or take it myself.
And this is a challenge. We live in a society where the negative is much more interesting than the positive. So, people like to publish headlines about common things that are accepted that are no longer accurate. And whether you can actually believe those headlines or not is TBD until you get down into the research and really look at it.
Mike, how does someone know how much creatine to take a day? So, some people, like for example in our longevity mix, it's 2.5 grams a day, and then many people do much higher doses, especially when you're trying to increase muscle mass, people doing very intense resistance training protocols. But how much should someone take, and how can they measure the efficacy of the dosing protocol?
Hard to measure the efficacy of the dosing protocol. But we do know a lot about creatine supplementation and the different doses, and like if I were just to give you a specific number for general overall health, I would probably actually do it weight-based and I would say something like 0.1 g per kilogram per day. So that means if you're 70 kilograms, which is the size of an average male generally, that would be about seven grams a day. So, probably just saying five or something like that is not really accurate because, you know, women may not need quite as much because they're smaller. If you're larger, have larger lean muscle mass, you might need more.
And it also depends a little bit on what you're getting in your diet already. So if you're eating a standard American diet that's generally overall healthy, you're probably getting 1 to 2 grams a day already. So, if you're getting that, you might only need 2.5 additional grams a day. But if you're vegan, for example, and you get very little red meat, then you're probably getting less creatine. And in that case, you probably need a higher dose of daily creatine in order to saturate your muscles, because the goal here is to saturate your muscles, at least for performance benefit.
So, I would say if I had to throw a number out there, I would say five grams a day seems adequate for the average-sized human, given what the normal diet generally includes, and that's going to get you to saturation. You don't need to do the loading phase where you're taking 20 grams a day. You can just start at five and just do it consistently, and then over 3 to 4 weeks the load will happen naturally on its own.
Now, if you're looking to take creatine for other reasons, like cognitive performance, like for example, if you are trying to offset poor sleep, there's some decent data to suggest that you improve cognitive performance, especially in high metabolic demand like brain trauma or poor sleep, or if there's mild cognitive impairment, like early dementia. Those studies, though, look at more like 10 to 20 grams of creatine a day. So, much higher doses.
And there was actually one study that looked at, like, what is the max dose of creatine you could take, and we have studied up to 30 grams a day in people for up to 5 years with no significant downsides. So as far as we can tell, you take extra creatine, you pee it out. It's not a big deal. So it doesn't appear that there is too high of a dose, as far as we can tell.
Kate, you began taking creatine yourself just when you started Blueprint. No?
Yeah, I did. I took five milligrams and then I... micrograms. Grams. What is it? What's the unit? Grams. Yeah. And then I switched to the longevity mix, which has 2.5. And so then I take an extra half scoop of creatine separately to get my five. But maybe... Yeah, I mean that sounds like a right amount for me then, because I think my weight is around 57. Yeah.
So, Mike, in summary, if someone is approximating five grams a day for an average intake, is there anything they're looking for in blood work that would help inform them about whether or not that dose is appropriate for them, whether they go up or down?
Not particularly. You could look at your creatinine, which will be a little bit falsely elevated. So, creatinine is a measure of kidney function, and that will elevate with creatine supplementation. I wouldn't necessarily use it to tell whether you're at adequate dosing or not, because I think it can be confounded by other things like hydration or kidney function. But you will likely see an increase in your creatinine if you start taking creatine.
Okay. Your creatine protocol is going to do what based upon this study and this response?
Absolutely nothing. That's what I will add. Keep pushing forward.
Actually, I was thinking I wish we had like a travel advisory equivalent for health, because these studies come out, they go viral, they influence society, but it's not contributing to some sort of updated information platform. So, it's like you have travel advisory, you know, here's what's happening in the country, here's what you should know before you travel to Japan. Same thing for creatine, like go onto the website. It's like, okay, this new study came out, shouldn't really impact because of these reasons. Here's what you should know. You know, wouldn't that be cool?
That's true. You're right. Because they do. They hit and then it somehow sits in the background noise of our health consciousness, and it's like, I thought I remember something like it didn't really help on muscle, and so why are you doing it, right? It becomes shorthand and then it somehow becomes truth.
Yep, yep. It's like the latest thing to come out too, it has this bias toward that too, because it's new, it's assumed to be better, right? When in reality this was a study in 60 people, which is like so tiny. Such a small study. Yeah. After decades of research, right.
Bryan, how much do you take? Five. Five grams. Okay. Are you going to continue? Yeah. 2.5 and longevity. Are you going to increase it based on your body weight now?
I'm 70... I'm 77 kg. I'm experimenting with 10. And then I've been playing around with, like with travel, increasing it, because where I live, I basically have to get up at like 3:00 in the morning to catch a flight. So, I always get less sleep. So on travel days, I'm experimenting with higher doses of like 20 grams to see if it makes me feel more cognitively fresh. Have you noticed it makes a difference? Not yet, but I've only tried that twice now. So we'll see.
You say general body awareness of status of your body, making sure you're aware of basically like a level of familiarity. I know after doing my own health protocol for the past couple years, I'm intensely body aware. Like I can feel my heart rate at any second. I can tell you how fast or slow my heart is beating at any second, with pretty good accuracy. And, you know, we get so many treatments that I've carefully inspected every part of my body. I know what to look for: color, tone, function. So, it's been actually really cool, and it creates a nice contrast for me of how unbody aware I was before. You know, you just kind of get up and do your thing, and if you have a headache or don't feel great, you just kind of push through.
But yeah, that's been really great, that you really can create intuitively really good sensors of your body just by pairing up measurement and that emotion, like you get to feel what the measurement shows you, and you create these really fine intuitions.
Do you guys ever feel like... Kate, I don't know if you have had the same experience with body awareness since you started doing Blueprint, but do you ever feel like there's a bad side of that as well? Like, my wife and I were recently in Vegas and we were just sort of looking around the casino, and we weren't actually playing at the casino. We were going to a show. I'm not a big casino guy. But we were going to a show and we had this conversation, like, how are these people all upright? You know, everybody's drinking, everybody's smoking, everybody's eating like pizza and corn dogs, and if I did that, I would literally be on the ground. I'd feel so bad. It's like once you get healthy, you create this body awareness that now all of a sudden it's almost like you can't even allow yourself to mess up anymore, right? But yet there are these humans who appear to be thriving. They're walking around, they don't have chronic back pain. They seem like they feel great when they're just putting, you know, trash into their bodies. It's an interesting situation that we put ourselves in, which obviously I wouldn't choose anything else, but it's unique.
Yeah. The body is incredibly adaptive. That's what that tells me. It's like we can push off big problems for a very long time and then they can come crashing down.
It also reminds me, Bryan, of the study at Colonel, the inebriation study, where for a certain period of time your brain is able to make up for the deficit of cognitive function from alcohol until it can no longer. And that's how it feels. Those people who are standing upright, their body is working hard to make up for the deficit of health that is happening in that transaction. That's at least how... is that accurate?
In that study they did no alcohol, low, medium, and then high. And so what they found is that when someone is in the low-medium state, you could behaviorally appear non-intoxicated, right? You could pass all the tests, but the brain scan showed that you were in fact impaired and the brain was just compensating for it. The brain pulled up and cured for the deficit, but in the high intoxication scale, the brain could no longer have the compensation ability to do so. And so you saw impairment both in behavior and the scan.
And so this is cool, like that's why, of course, cognitive decline begins a long time before symptoms emerge. And you can in fact do early detection of cognitive decline, because your brain has impairment, you just don't see it in your daily life. Your body awareness doesn't capture the impairment. So that kind of granularity of measurements... So that was cool to see, that these brain scans can significantly improve your awareness of how your brain is doing, just beyond your own perceptive behavior.
You're more drunk than you are aware. And so the analogy would be you're more deathly than you're aware in those moments in the casino.
Yeah. I don't know. Yeah. Actually, it reminds me, we went out last night to a restaurant. The team was all together, and I think 25 of us or so went out to a restaurant, and I just so rarely do this, you know, and I was just shocked at watching how the food arrived at various tables across the restaurant. I mean, I was beside myself. This is a health restaurant, too. This is one of the known health restaurants in Venice, in California. And yeah, I guess knowing what I've learned about food and toxins and the way that food is passed through restaurant systems... I mean, the team was kind of like, Bryan, yeah, hey, welcome to normal society. This is kind of what people do. But yeah, I guess we've built these systems that are just so outside of that that it was very shocking to me. It almost felt like I was traveling back in time and viewing the early 21st century, like, this is crazy, watching a movie or something.
I got a question. How many years until we know whether or not Bryan's... like, I know we know that Bryan's protocol is working in the moment, right? Like his fitness and all this kind of stuff, but I imagine age is kind of like a cliff where you get to 70 or something like that and things really start to deteriorate. Is there a point where, when Bryan hits that chronological age, we'll be like, "This is for sure working"?
That's a really good question. I think that point is different for just about everybody in terms of, you know, when you start falling off the cliff. That cliff is just at a different location for everyone, depending on their genetics and their lifestyle choices. I mean, I think what we're doing now, tracking Bryan's biological age, is probably the best that we can do in terms of watching, you know, how fast he is continuing to age and how much we are changing that. So both the speed of aging and then his actual biological age is probably our best metric right now, and we're continuing to slow the speed of aging. So down to less than, what, 0.5, which is pretty
pretty phenomenal. But you know, theoretically, at some point, to not die, at some point we have to continue to slow that process, right? So more still needs to happen in order to not have a cliff at some point in the future.
Okay, so on body awareness, what I loved seeing the other day was Andrej Karpathy, well-known figure in AI, did his own n-of-1 study at home, which I feel like everyone loves to on you, Bryan, for your n-of-1, and it's so fun to see other people coming out and publishing their data and results.
So I'm going to read you this. So over two months, he tested four sleep trackers. So Oura, Whoop, Eight Sleep, and Apple. And what he found was Oura and Whoop were top tier. But what I was most interested in, I think this links back to the body awareness, was he said: overall, I say with absolute certainty that Bryan is basically right and my sleep scores correlate strongly with the quality of work I'm able to do that day. When my sleep score is low, I lack agency. I lack courage. I lack creativity. I'm simply tired. When my sleep score is high, I can power through anything. On my best days, I can sit down and work through 14 hours and barely notice the passage of time. It's not subtle. The effects are not a function of a single day's sleep, but of the accumulated sleep depth across a duration of the last few days. So, in other words, a single bad night's sleep is usually okay, but a few in a row is bad news and vice versa. Listen to Bryan.
I love that. That's so accurate to how I've experienced sleep, too. And that's like the weird thing: if you're constantly getting bad nights' sleep occasionally, you can't detangle what it feels like to have this accumulative good night's sleep. So I just thought that was really cool.
It really speaks to the value of measurement, right? Like, and the associations that we can start to create in our life once we start actually measuring things. And I think that goes beyond sleep as well. Like, you know, once you start measuring things, you start having this sense of your body like Bryan was talking about earlier. And without that objective measure, it's hard to understand what all those subjective signals mean. And once you can put those two things together, now you're creating the knowledge and the wisdom to make the right decisions for your own health.
Yeah, totally. I always say to people, just start measuring, because as soon as you know your data every morning, even if you do no interventions, it starts to build that relationship of like, oh, I noticed this today. What's my data say? I've noticed this today. And then once you've got that down, you can start doing interventions. But yeah, Bryan, it made me think, like, you should challenge Andrej to beat your eight-month sleep score.
The thing I absolutely loved about this write-up. First, I mean, Andrej is one of the most formidable intellects of our time. He's a founding member of OpenAI. He was director of AI at Tesla's program for autonomous driving. I mean, he has been on the forefront of the AI revolution.
And this is the contemplation that I had after selling Braintree Venmo: we are at this incredibly interesting moment in this part of the galaxy where we're evolving into a new species. And so my observation was that the most important thing we could do would be improving our own intelligence to see this moment for what it is and act wisely. You know, not foolishly, but have wisdom beyond our capacity.
And so my effort was first around Kernel, which is: if we can measure the brain and reveal the invisible things that happen in the brain and then pair that with AI, that will give us a vector to begin improving our intelligence at speeds which we otherwise can't. And not just intelligence in terms of IQ, but all dimensions of intelligence. Like, what if you actually could pair your mind with AI as you're working on emotional development, or on correcting your blind spots, or in trying to have a more well-rounded thought process, or not be sucked into these very narrow worldviews that get you in this tribal mindset? So what I was really trying to do is say we need to augment ourselves.
And so what Andrej is doing is Andrej is building the future of intelligence in the form of AI. And in doing this study, he's actually building his own intelligence. And that's what he's saying: like, hey, I did this experiment with sleep. I see a substantial improvement in my intelligence along the realms of creativity, right? And the elements he talked about.
So to me that was the best representation. It's the best outcome I could ever ask for, which is an architect of superintelligence is leaning into what we're saying: it makes sense on all dimensions to invest in your own intellect and try to acquire abilities beyond your natural state by taking care of yourself.
So yeah, it was so fulfilling when, like, man, we just take so much every day, all day. And so to have someone incredible like him come out and say, you know what, sleep works. It dramatically improves my life. You can measure it. You can actually create life systems for it. It's just like, man, what an absolute joy to read that. So I appreciate him doing that. I appreciate his kind words. It's just such a great moment. It's such a good reward for the efforts we do every day of, like, you guys, let's stop doing stupid stuff. Let's not eat junk food. Let's not drink alcohol. Let's not go to bed late. Let's prioritize sleep. Let's exercise. Let's measure ourselves. So yeah, I was really excited about that.
That's awesome. Yeah, I think that's a good place to wrap up. How did we do, you guys? I mean, do you think people are actually going to enjoy this? I don't know. Okay. Well, if you're still listening, put in the comments below: did this work? What do you wish we did differently? Yeah, let us know.
I'm excited. I think it's a cool format to be able to talk about what's happening this week. Yeah. So if you listen to the podcast and you enjoyed it, let us know. We have, I think, six or seven more topics we didn't even get to today that we had outlined. We had some stuff about some embryo selection technology that is currently in market and a bunch of other stuff.
Mike, what was your take? How was today? That was great. Had fun. That's how I measure the quality of a podcast, is whether it was fun or not. So that was a good time. Hopefully people enjoy it. All right. Thank you guys.
Article published