You just realized your business needed to hire someone yesterday. How can you find amazing candidates fast? Easy. Just use Indeed. When it comes to hiring, Indeed is all you need. Stop struggling to get your job posts seen on other sites. Indeed Sponsored Jobs puts your posting at the top of the page, right where your ideal candidates are looking. And it works. According to Indeed data, sponsored jobs on Indeed get 45% more applications than non-sponsored posts.
That's the difference between finding your perfect hire today or still searching weeks from now. There's no need to wait any longer. Speed up your hiring right now with Indeed. And listeners of this show will get a 100 pound sponsored job credit to get your jobs more visibility at indeed.com slash Wondery UK. Just go to Indeed.com slash Wondery UK right now and support our show by saying you heard about Indeed on this podcast. Indeed.com slash Wondery UK. Terms and conditions apply.
Hiring indeed is all you need. How do I cope with waiting for results? It's all I can think about. Is the cancer going to come back? I can't sleep. How do I stop worrying? We know what goes through your head when you're living with cancer. The questions, the anxieties. We understand because we've been there too.
If you or someone you care about has been diagnosed with cancer, Macmillan Cancer Support's online community can help from day one. It's a free, safe online forum for people with cancer to connect anonymously. No question is too big or too small. With over 80,000 members, we get it. We've been there. Whatever you need to ask, we're here. To join, search Macmillan Online Community. Hello everyone, welcome to the Mindscape Podcast. I'm your host, Sean Carroll.
A couple of years ago, 2023, I did a holiday message entitled Reflections on Immortality. This was in response, or at least inspired by, a workshop we had at the Santa Fe Institute on immortality, what it would mean, how you might get there. And I mentioned there that a lot of the people at the workshop basically said they don't want.
immortality. Like, that's too much. Either personally, they couldn't imagine what to do for all that time, or socially, they think that society would stagnate if people just live forever rather than being replaced. That's fine. You can argue whether or not you'd want to live forever or not. Most people, though, would want to live a little longer, at least, than they probably will live. We're very interested in longevity and aging and not only living, but... living in healthier ways.
I just now, for this intro, looked up the graph of life expectancy in the United States. For the last 100, 150 years or so. And it's been going up. That's the good news. It went down once. In this particular graph. Shows five year. bins, right? So only once in the 20th century did the life expectancy decrease from one five-year bin to another between 1915 and 1920. I'm not exactly sure why that was. The flu? The Spanish flu maybe?
2015 to 2020, it also decreased. It only decreased a little amount, okay? In 1920, it was 53 years old, the life expectancy from birth in the U.S. In 2020, it was 78.8. So it's gone up quite a bit. That's impressive. But in 2015, it was 78.9. It was a little bit... bigger. And we can talk about why that's true, the healthcare system, nutrition, whatever. That is the kind of thing we're going to talk about on today's podcast. Eric Topol is a leading medical researcher.
Cardiovascular system is his specialty, but he's a broad thinker who's done a lot of work and has been thinking for a while now about medicine in a very broad way and been writing books about it. We will actually take a little bit more than a little bit of a detour to talk about his previous book, which was called Deep Medicine, how artificial intelligence can make health care human again.
He has this fascinating point of view that not only can AI be super duper useful in recognizing, diagnosing some medical conditions that human beings are not very good at. But in fact, it can be better than human beings in some very, very important ways. And, you know, we think about how to make an ideal system where there'd be both. humans and AI there, but one of the steps there is making the humans understand that the AI can actually be pretty good at some things. It's not about
AGI or superintelligence. It's about pattern recognition, which is what AIs are very good at. But mostly in today's podcast, we're going to talk about Eric's new book called Super Agers. an evidence-based approach to longevity. And here we're not talking about modifying the human genome to live for centuries or anything like that. We're just talking about living a healthier life and preventing disease. in ways that are very plausible, very down to earth, very tangible, and can actually add.
several years, maybe decade, to an ordinary human healthy life expectancy. This is something which most people would sign up for if they had that option. And it's a combination of understanding the diseases that tend to end our lives at some point. In particular, the fascinating thing that we learn is that many different sounding diseases are all related to the immune system and inflammation.
might be promising for sort of treating or preventing them. And then, of course, the other thing is that things that we individually can do, things that have to do with diet and exercise, as you might expect. but also mental health, mental acuity, social activity. These all turn out to be really, really important. for not only living a long time, but living a long time in a healthy way. So maybe this is a podcast which, unlike many Mindscape episodes, has a really tangible effect.
on how people live their lives. I hope it does. So let's go. Great to be with you, Sean. Thank you. So we're going to talk about a lot of things with medicine, aging, AI, things like that. But one thing, there's a little news item that struck me and I thought, I don't want to forget. asking you about it while I have you here, which is...
As a person of a certain age, I recently got the shingles vaccine. And now I'm told that this might even have a positive effect on preventing dementia in the future. That doesn't seem what it's there for. Is this something that is... completely weird and surprising or true and understandable. Well, it's a mixture of your questions about surprising and understandable. It's really fascinating because there are these two so-called natural experiments.
whereby huge numbers of people got the vaccine just because of a policy change or a date change. So it's like a randomized trial. In fact, it's even better because there's no selection criteria. And whether it was comparing the new Shingrix vaccine versus the old live attenuated virus vaccine called Zostavax or. The vaccine, the old vaccine versus no vaccine, everything pointed to this about 20% reduction of dementia.
And this is fascinating because While other studies had suggested this, they weren't like this ideal. natural experiment. So it does look like there's a twofer. It helped me decide to get the shingles vaccine because I don't do well with these vaccines in terms of acute reactions. It's unpleasant, yeah. And I was lucky because of all things, this vaccine didn't knock me out like some others do. But yeah, I think for people over 50 in the U.S.
It's wise to get this because there really isn't appears to be an added bonus. It doesn't mean for sure that the virus, herpes zoster, causes Alzheimer's, but it could be that just revving up your immune system with this vaccine helps. So it isn't unequivocal proof that there's a root cause of a virus for Alzheimer's or dementia, but it does certainly suggest the immune system plays a role. So we're still pretty unsure of what the root causes are for dementia for Alzheimer's.
Yes. I mean, the theory of a virus, like for multiple sclerosis, Epstein-Barr virus, which is... unequivocal. It's very soft here. We don't have anything like the multiple sclerosis story, but there's still people who believe the virus plays an important role and perhaps in a subset of people. And even other viruses besides the herpes zoster varicella virus may be invoked. So there's still work to be done on that.
Yeah, okay. I mean, there usually is, especially medicine. It's very hard. Okay, thank you for indulging me with that. Now we can get on to the work you've done, the books, etc. Part two of the podcast is going to be about your most recent book on aging, which is something everyone cares about. But previously, you talked a lot about the potential impact that AI can have on medicine. And I think this is perfect for this podcast because...
I love the idea of AI. I'm very skeptical about some overblown claims about artificial general intelligence. Here is an actual use case where you don't need to worry that the AI is conscious or intelligent or anything. You can just use it. So why don't you give us the big picture overview of why AI might be helpful in medicine? Yeah, I mean, in some ways, it's stunning, Sean. I mean, we're going to get into some of the concerns, of course, but some of the things that AI is bringing to us.
we couldn't even have conceived, frankly. So perhaps the one that is just kind of knocks you out in terms of wow factor is the so-called machine or digital eyes, the vision that AI has. Which is so it makes our vision as physicians, experts in certain domains look weak. OK, because it turns out that. Once you train an AI to see things, if you will.
is so much better, not just for the things that you're supposed to be looking at, but for finding things that you can't even see. So for example, you know, the retina is a perfect example because it turns out. Even ophthalmologists, the retina specialists, they can't see things that the AI can. So not just about retina disease, but about every system in our body.
not just the brain, but about risk for heart attacks and strokes and kidney disease and liver disease. And I mean, the list goes on and on. So the point being is that there's this opportunistic aspect. Because the supervision of AI is just extraordinary. And so that is something that I think we're just starting to get in touch with.
No one had anticipated the richness of ability to interpret medical images, which is a big part of practice of medicine. But then we're also seeing some things that also are surprising. That is. One of the most important parts of medicine is making the diagnosis. you know the the there were things that were listed as the three most important things diagnosis diagnosis diagnosis it turns out we have 12 million serious diagnostic errors a year in the United States, we see 800,000 Americans.
either disabled or dead from these serious diagnostic errors. So it's hard to make things work. Well, it turns out, even though the medical community doesn't like to fess up to these problems of errors, these errors. The AI is making a big difference because it is so good at coming up with a differential diagnosis from a person's data. much better than we expected. So the chance of it being as good as one of the experts for coming up with the diagnosis like we see with the...
storied case presentations at the New England Journal of Medicine. It's amazing. So this is another way to accuracy, not just medical images, but also for helping making the diagnosis. through what's really fascinating now sean is this we're not just seeing large language models we're seeing large reasoning models
And so this is really helping that reasoning and following the chain of thought. So you can just say, how did you get there? How did you come up with these diagnoses and get the details of how the AI is thinking, if you will? Now, the other thing that I think we wouldn't have expected, and again, some of these are really outgross of the transformer architecture of AI. But now when you go with a patient in a clinic visit or even at the bedside.
that interaction can not only be transcribed, that's nothing, but what it can do is do all the downstream tasks. So it would do like if you need prescriptions, follow up appointments, procedures, labs, setting up follow up appointments for. And nudges for things that were discussed during the interaction. Like, did you check your blood pressure? What are the values? And you get it back to the doctor. And even pre-authorization to deal with the darn insurance company.
So what we're seeing is average for those who are adopting these two to three hours less data clerk work for doc. And then the other thing that's, I think, very refreshing is to see the patients getting empowered to make their own diagnoses. uh through these tools so whether that's for heart rhythm or skin lesion or ear infections in a child or urinary tract infection so All around, they were seeing effects, and they're just real starting to take hold.
I love the example that you already mentioned it briefly, but I saw you give it in a talk and it really stuck with me of this retinal scan. So you're literally looking at a picture of someone's retina and you ask... Human doctors. Okay, is this a man or a woman? And they have no idea. It's 50-50. They're just guessing. And the AI gets it right 99% of the time. Yeah, yeah. It's amazing. And we still don't really know why the AI is so good.
But that is perhaps the best single example of the gap between you and I. discrimination and what an AI when it's seen train with hundreds of thousands of images it just It's actually pretty mind-boggling to me. And, you know, I've known about it for years. And there's just now so many other examples, whether it's a cardiogram or a path slide, an x-ray. I mean, who would have thought you could diagnose diabetes?
from a chest x-ray i mean these are crazy things you know so um it's it's wild but it kind of makes sense you know because you're kind of using the ai for what it seems like it should be good at. It's pattern recognition, right? And it's looking for differences. And human beings are okay at that, but they're not perfect. But you're not trying to shoehorn the AI into something that being something that is not.
That's right. And what's really interesting is that capability was just an outgrowth of deep learning without the extra. uh turbo charge capability of the newer models so yeah i mean just think of where we're going to go with the models that continue to get refined and improved on a weekly basis it's just crazy right now
And it's a reminder, I guess, probably as a medical doctor you don't need to be reminded of this, but the human body is very interrelated. I mean, the fact that there is some subtle difference, even if human beings don't know what it is, between a male retina and a female retina. Yeah, maybe makes sense if you tell it to me, but I wouldn't have been surprised if it hadn't been there. But, you know, yeah, the body is—all the different parts are talking to each other all the time.
That's right. Absolutely. And this is something where, again, is the AI able to tell us why it thinks that a certain retinal scan is man or woman? Well, this explainable AI is kind of an overarching goal. And sometimes we do get a pretty good explanation. But in the case of...
the retina, it's still elusive. There's these things called saliency maps and masking and all these techniques that you can use to try to kind of reverse engineer what is this neural network doing you know but uh it hasn't been as uh a good explanation yet provided for the retina, as it has for many other capabilities that were not anticipated.
Okay, perfectly fair. I guess on the flip side, there are people who worry that their medical care is going to be handed over to computers rather than human beings. And I suppose ideally it should be both, right? There should be both in the loop and there should be interactions. But like you said, with the paperwork, at least in principle, the existence of AI could give us more interactions with humans.
That's right. And there was something other, another thing that caught us by a big surprise recently. And my colleague Pranav Rajperkar and I wrote an op-ed in the New York Times about it. Six studies. now there's seven, where the AI was compared to a doctor with AI. for various tasks, whether it's a diagnosis or...
images, and the AI was better than the doctor with the AI. Okay? We would not have expected that. Everyone, including us, all anticipated that the AI and the doctor is the best winning combination. Now, what is the reason for that? Well, it could be because the AI right now isn't in high regard for doctors, and so they don't really take it seriously. Or they could just have a bias against it, or they're not grounded yet on how to use it.
But it's a little concerning that the AI could have this superior performance in various medical. tasks, important tasks. We're going to have to see how this plays out. But one thing for sure is I think it's going to be imperative that all physicians learn the nuances of AI, not to immediately reject it.
but to actually consider embracing it because it can help. But when the AI beats out the doctors, that's a flag right there. Yeah. Right. Okay, good. That is absolutely food for thought. But I would imagine that in some... circumstances like you know looking for polyp or nodules in some scan, if the AI could say, look here, it looks like there's something, that just sounds unambiguously helpful to the doctor. Yeah, the biggest study yet.
You touched on one, which is the second biggest, which is colonoscopy cumulatively. But the biggest single study, over 100,000 women in Sweden. had either the doctors, the radiologists with AI interpret the mammograms or doctors without the AI. And what was amazing, of course, is that there was this 20-some percent better detection of important cancers with the doctors who had the AI. So you would say now, in addition to other studies, that why are we doing mammograms in women without AF?
Because now that's missing, you know, one out of five. That's a lot of potential cancers. And we've known for years that mammography has lots of inaccuracies with false positives. false negatives, you know, there we're seeing that should probably be universally adopted. The problem in the United States. is that the company, the big radiology company that's got scans all over the country, they're charging the women.
If they want to have the AI reading, $35 or something like that, that's crazy. That should be part of a normal, and someday I guess it will, but we're in a transition zone. We are. Are the AIs that are being used, are they off-the-shelf GPT kind of things? Or do you need a specially-tuned medical... Yeah, these are commercially available.
software packages that have been FDA approved now. The ones that are here in the U.S., particularly the one that RadNet uses, is different than the one that was in the randomized trial in Sweden, but they're pretty similar. the point being is they they they the the good ones have all validated this 20 improvement of of making the diagnosis and so for polyps it's similar to the gastroenterologist miss lots of polyps you know like about 20 percent So the fact that AI can give us this booster.
accuracy, who wants to go through a colonoscopy and have a tired gastroenterologist who's been working all day doing this stuff? Particularly, that's when it happens is later in the day, human eyes start to lose some of their... And, you know, who wants to go through all that and have a missed important policy?
I got that from your talk that you should always have your colonoscopy in the morning. The doctors are fresher. I would. Plus, it avoids you having a fast, you know, after having gone through everything another day. So, yeah. And maybe other than diagnostics, are there other promising uses of AI? Yeah, I mean, I think it's across the board. I mean, it's just... You name it. There's like.
You know, for years as a cardiologist, we had the electrocardiogram had an overread or what do you call it underread by the machine. And it was stupid. The readings were terrible. The error rate was really high. Well, now, not only is the error rate really low, impressively low, but also what we're seeing is making diagnoses that the cardiologist might not have come up with accurately. So, for example, it can say
There's a low ejection fraction. And you say, what? And in fact, Mayo Clinic has already made that routine for all of their ECG AI interpretation. But also things like... uh there's a risk of developing atrial fibrillation even though there's no atrial fibrillation in this person okay and there's a risk of stroke or this or that
Kind of like the retina. So not only have we got rid of the error rate that was kind of a major deal, but we're getting all this opportunistic readings that we never had before. just a pretty, again, You know, I just don't see, don't expect this kind of thing. And it's just happening. And how prevalent is it right now for these kinds of things? Is it yet to ask special and pay or is it already there?
the ecg is not anywhere else that i know of yet except for mayo clinic i hope it will be universally used because it's a sense that sean that we're just leaving all this stuff on the table you know that all these interpretations that could be useful for a patient. This should be of no charge. It's there. It's in the data. We need the AI to extract it.
Well, tell me, as someone who's been around for a while, do you get the feeling that practicing clinical doctors are open to these kinds of changes in general? Are they looking to be on the cutting edge, or is it that they have a way that works and are kind of happy? Well, the older ones like me are usually not too receptive. Younger digital docs, you know, they kind of grew up in this environment. They're apt to be a little more willing, but overall.
There's not an embracement eager to adopt AI. And it's because of the other concerns like that. what is this going to do with errors hallucinations and privacy and security and biases which we know exist or you know there's many things that are that are deterrents, perhaps the one that I think that doesn't get enough emphasis, is you really have to have compelling data to get the medical community to move.
And the only compelling data we have that really compelling is what we're seeing with the mammography. uh colonoscopy perhaps you know certain things where you know that it's just unequivocal benefit okay but a lot of things that we're seeing evidence of they're still looking for the proof points that are just so clear cut that it would be considered below the standard of care.
which is kind of a bit nebulous term but it's changed when the needle is moved when the stuff the data evidence is just overpowered Well, it does seem to me like having better diagnoses of cancer or heart disease would help me live longer, which provides a natural segue into our next big topic, which is your new book on aging. And again, let me just start.
At the very highest level, aging and especially extending human lifespans. This is a topic that gets a lot of talk. A lot of the talk is maybe not reliable. I mean, how do you feel wading into this area? Yeah, I felt, Sean, I had to wade in because the pseudoscience and the false claims are totally out of control. And basically, it's kind of predatory, all these anti-aging supplements.
you know, stem cell clinics and, you know, peptides and I mean, you name it, it's just so much malarkey, so much hype of fall. You know, you could say some of this is fraudulent. It's a shame. And it basically is playing into the interest that people not just want to live longer, but live longer, healthier health space. And that's why I decided it was time for me to go deep on this and write a book, which is Super Agers, which is trying to set the record straight.
to really put the excitement in context. And it isn't about reversing aging, amazingly. It's about extending healthspan to quash age-related diseases. So basically what this is doing is taking the science of aging, which is formidable, big advance. And using that not to come up with some magical potion, which doesn't exist for people, but rather to use those advances so that we can take the big three. heart cancer.
neurodegenerative, mainly Alzheimer's, how we can apply these advances in science of aging to prevent these three major killers and major causes of losing one's health.
And just to be super clear, on the podcast, sometimes we talk about very science fiction-y topics. So your focus is not on... genetically engineering human beings so they live for 10 000 years you're really thinking about you know how we can just live a few more years in a more healthy way As I go through in the book, if you were to really work on this, you could eke out seven, perhaps even 10 years of healthspin. Healthspan meaning freedom from these three majors.
um diseases age-related diseases so yeah that's exciting to me it's not we're not talking about a few weeks or months we're talking about years yeah and the fact is that we don't need a magic potion to do this As we'll get into, we need lots of data on a person and we need AI. And we need to use the advances, which are basically these clocks. molecular clocks. That's perhaps the biggest thing that a shake-up in medicine that people don't realize right now, biomarkers and clots.
And that's what's given us this newfound potential. Well, let's get into that right away. Where does human lifespan come from? It seems clear from even the most obvious data that people tend to... live a certain amount, and if they don't have something terrible happen to them, then they're going to die. It's not an exponential fall-off, right? It's not like the same fraction of people die every day. Our biology gives out after a point.
Probably we can guess at evolutionary explanations for that, but maybe give us a clue as to what's going on. Well, everybody used to think it was the genes, and that was deterministic, and that turned out to be completely wrong. While it may be an issue in a very limited number of people, it's not really about our genes. And so a lot of this is...
It turns out it's a multidimensional, complex story about why we... um have this what's called compression of morbidity where you live perfectly healthy and then you just fall off a cliff you know that's just it doesn't work that way yeah the way it works is you know at some point you're gonna start to see some frailty some decline in capabilities like whether it's cognitive or or other and so the point being is how can we extend that
you know what intactness lack of this uh compromise or frailty uh and so what is the determinant of that well we studied 1400 people uh who we call the welder And they had been over 85 and never been sick. And they had none of these diseases. And what was interesting, we sequenced all their genomes, whole genomes. We found almost nothing from that. That is, the genes were some story. And the protagonist patient of mine in the book. Mrs. L.R., she's 98.
And she has never been sick. I mean, I saw her for a trivial thing as a patient. But what's fascinating is all of her relatives.
her brothers and sisters her parents they all died in their 60s 50s she's 98 okay so that's basically the story it's not the gene so like for me i have terrible family right but when i as the more i've learned about this the less worried i've been about that most of us have this overhang throughout our lives oh my my father had this my mother had that whatever We don't need to succumb to those thoughts because there's so much we can do now to find out our risk.
individual at the individual level. And that's what's a sharp contrast to the way we used to think about this in the past. So specifically in this study, you had these people who were living long and they were healthy, and I presume that you were hoping to see, oh, they all share this gene, or at least many of them do, but in fact essentially no connection whatsoever.
Just about no connection. I mean, certainly nothing that had a lot of weight. So we were surprised. We were very disappointed at the time. But now in context. It's starting to come together that we have to have a rethink about the gene story because it just doesn't really click. There may be something to it if you're over 110. There's not too many people like that. I'll get to that point, yeah. They're called super centenarians. But there's so few of those, and even that's questionable.
But is there something to the idea that, you know, once we're past our reproductive age, evolution is done with us and it sort of lets us decay? yeah we're in steady decay mode sadly yeah so what happens is for these three diseases in particular uh it takes 20 years for them to take hold incubate if you will So you don't have a heart attack without having your arteries getting clogged up with atherosclerosis for two decades or more.
cancer. You don't just all of a sudden show up with pantheatic cancer. This has been developing over more than two decades. these clones and their ability to eventually proliferate and spread throughout the body. And the same thing is for Alzheimer's. We now know it's 20, 30 years. You could predict ahead of time if that is going to be a risk.
These processes of our deterioration, the good thing is they're really slow. It gives us a great opportunity to get ahead of them. Imagine if they took place over a month. or a year and then of course we wouldn't have that anticipatory runway to work with. So we're really lucky in that respect. And it's that combination of these long lag time periods and the tools that we have today that's making this exciting opportunity going forward.
So you mentioned the big three. Let's dig into them a little bit more. Probably if you ask people on the street, what are the three big diseases that will end your life? They probably would have come pretty close to saying like cancer, heart disease, brain diseases. Are they related? Are they all completely different causes? What do we know about why these things start kicking in at a certain age?
Great question. There is a common thread besides the fact that they each take 20 plus years. And that thread is the immune system. and inflammation. So our immune cells can secrete all these what's called cytokines, chemokines that are inflammatory mediated. And so what's the problem with the three diseases is that they all have a marked untoward inflammation.
so inflammation in our artery wall or in our brain or um the immune system not responding to the appearance of a of a tumor cell or tumor clone So our immune system is the... is the real problem for the three diseases. And that's one of the big deficiencies we have right now. We don't have... away like you can go to the lab and get a blood count kind of thing you can't get an assay to tell you what is the status of your immune system.
And we really desperately need that. That's like the next frontier. And people say the brain is really complex. Well, I got to tell you, the immune system is right up there. It's very, you know, I have a whole chapter about that in the book. And what's exciting, of course, there is we're learning about how to rev it up. and tone it down suppress it like never before so if we can start to get a routine way people can get their immune system and inflammation markers assessed
That's going to give us even better ways to prevent these diseases. We did recently do a podcast with James Allison about using the immune system to fight off cancer. Fits in very well. But let's assume that some of our listeners didn't listen to that episode yet and maybe explain more of the details about how our immune system is supposed to work when everything's going well.
Well, you know, that breakthrough that Allison had for immunotherapy was just the beginning of now a chain of many discoveries. And the point is that it's pretty hard to get cancer spread around the body if your immune system is completely intact. Cancer tries to hijack cells and suppress your immune system, but... It's hard to get it rooted if your immune system is really high integrity. But the other thing we've learned is that as we keep adding more different ways.
of revving up the immune system in people with cancer that aren't responding to the current therapies, we keep getting more successful of getting cured. So it's almost like we can't do enough with personalized neoantigen vaccines and tumor infiltrating lymphocytes and conjugate drugs. All these things are basically trying to get the immune system to go into high power mode, you know.
That's a big learning. And the point being is, if I could say, well, let's say you, Sean, you have this significant risk of cancer in your life. What I'd be doing is maybe every year assessing your immune system. And when it was showing signs of...
showing a bit of what we call immunosenescence, aging. We would give you something like a vaccine, like we were talking about at the... the shingles type vaccine, but we might give you a vaccine just to give you a higher gear of your immune system so that you are in defense mode. And of course, we don't want to go too exuberant because we don't want you to go into self-attack mode either and develop an autoimmune component. But that's where this is.
The cancer story that's unraveled since the... great work of Alison and others is really built on the fact that it's really an immune system. that's allowing cancer to take hold and spread in our body. And we have got to prevail over cancer. This is a fundamental goal. We've been talking about this moonshot, you know, for years, for decades. But now there's a there's a way to see through.
So tell me more about the cellular biology here. I understand that there's like T cells involved. What is supposed to be happening in my bloodstream when my immune system is going great gun? Well, there's a lot of different components. The B cells that are known, of course, they're making the antibodies. and the T cells that are kind of a primary way that we can kill cells that are troublesome, that have foreign proteins on their surface.
And of course, there's so many other aspects of, you know, interferons and there. regular neutrophil white cells and there's natural killer cells. It's a big system. It comes down to when you're trying to deal with cancer, as an example. you really want to have the ability to recognize these foreign antigens on the surface of cells, the cancer cells. And that's where the T cells really kick in.
and so what you can do now which is really quite extraordinary we've seen like we've never seen before a small numbers of patients but refractory cancer they failed every therapy kidney cancer pancreatic cancer They had the kitchen sink of everything we got and they couldn't respond. And then they got a so-called neoantigen personalized vaccine for their cancer. That is, they were the cancer. They looked at these antigens taken out of the body and then engineered T cells to fight them.
for that person's specific cancer and guess what happened i mean that was like melted away i mean just Now, the problem, of course, is it's not something that you could scale in millions of people who have cancer each year around the world. But on the other hand, if we can get off the shelf ways to do this. That's where the prospects lie today is that we don't want to have to necessarily make T cells that are individualized, but rather if we can come up with the common antigens or protein.
that are responsible, that can stimulate our immune system. That's what we want to do. And, well, I don't know. I mean, can we imagine? Some combination of AI and DNA robots and I don't know, that will sort of fix up our immune system in a kind of bespoke way for whatever it is that we're most in danger of.
I wouldn't rule it out. It's certainly possible. I like your thinking. It's futuristic and you're starting to bring all this stuff together. I'm more kind of looking at the here and now potential, but you're right. already AI is being used to determine the matchup of a person. antigens of their tumor cells and the T cells. We're already starting to rely on AI because In any given cancer, there's hundreds and hundreds of these.
particular proteins. And so they're finding the ones that are to develop a vaccine again. You can't do it against hundreds. You've got to pick the most important ones. And AI is helping to do that. Okay, good. And I think I... do see what we're talking about here vis-a-vis cancer, especially because we did talk to Jim Allison before. So what's the analogous story for cardiovascular disease, neurodegeneration? Yeah, so...
The Alzheimer's story is, a lot of people don't realize this, how extraordinary change there is. so there's a blood biomarker called p tau 217 and there's a couple others like it and basically what they are is as good as a pet scan for the brain okay so you just have simple blood tests it's available It tells you your risk of Alzheimer's. And by the way, if that marker is high, abnormal, and you get another one in six months or a year, then you can say exactly when you might get.
mild cognitive deficit so it stages the disease no less makes the diagnosis of risk Now, it used to be we have a genetic... tests and we'd say, oh, well, you have a higher risk of Alzheimer's. And we didn't know if it was at age 98 or 68. Now we know, you know, very precise timing, not just the extent of the risk. Now what's amazing here is that you can do all these things that reduce inflammation in the brain.
so like for example diet and exercise the right kind of diet and the right kind of exercise um can really help reduce inflammation in the body and the brain lifestyle Okay. But also, what's really intriguing right now is all these GLP-1 drugs like Ozempic and Munjaro, they reduce inflammation in the brain. It's one of the most important ways that they work. So they're being tested now to reduce Alzheimer's in big trials that are due out in less than a year.
So there are other ways to reduce inflammation. The point being is that You might have some amyloid plaque in your brain, but if your brain is reacting to it and it's inflamed, that's the ticket to developing Alzheimer's. So knowing your risk. and getting all over it with the lifestyle of preventable factors and likely drugs in the future. Not the kind of drugs we have right now that are used for people who already have Alzheimer's, that are dangerous drugs and don't work very well.
But this is the ability to prevent Alzheimer's in the future, which is really quite remarkable. I don't know a lot about Ozempic, the weight loss drug, or it's usually marketed as a weight loss drug, but I keep hearing how, oh, it has this extra unanticipated good side effect, which is... Kind of the opposite of what I'm used to hearing about all the deleterious side effects. Are we learning more about why this is so awesome? Does it have something to do with the immune system?
Yeah, I mean, I think what's fascinating here, that if you take these drugs, these so-called GLP-1 drugs, they don't get into the brain very well. So people started to wonder, how do you see such great reduction of brain inflammation? No less, how do you see the reduction in appetite? desire for healthier foods and, you know, basically completely rewiring your reward circuits. Yeah. And it turns out the gut brain.
So basically, it's affecting the gut. There's a vagal nerve that connects the gut to the brain. And it's through this circuitry that basically... The brain is kind of the master regulator of our inflammation, both in the brain and throughout the body, which we only learned about that in the last couple of years.
you know we thought it was just the blood but no the brain's talking to the blood and the gut's talking to the brain and so this is what is we're learning about these GLP-1 drugs that have had more impact across more diseases. some of which have nothing to do with weight loss. The inflammation is reduced before anyone loses weight. Okay. So this is a whole new day for understanding these inter-organ connections.
Okay, and then heart disease. This is your thing, right? I mean, it does sound different than the other two. Like, heart disease seems more sudden, or maybe that's just my misimpression. Yeah, people think, oh, I had a heart attack and this was sudden. No, no, it's been brewing for at least 20 years. We have seen autopsies from 20-year-olds, 30-year-olds. They already have plaque building up in their arteries, in the heart arteries, the so-called coronary.
so no this is something the only sudden event is that finally there's a crack in the plaque but that plaque has been growing and kind of inflamed and here's another point is that if it wasn't that the plaque got inflamed And that's in part why statins work. They not only lower the LDL, but they reduce inflammation. They're not as powerful as the GLP-1 drug.
but statins and anything that reduces your ldl cholesterol will also reduce some inflammation for those for the arteries because those are those are a species of the fatty proteins in our blood that the artery does not like that having it in their wall. It just reacts to it. And so years and years of accumulation of bathing your arteries. LDL, and now there's another one that will become routinely measured and new drugs to fight against called LpA.
which we never had anything for it. But now we are seeing five different drugs that are emerging to tackle that. So this is exciting, too, that we're coming up with new ways to prevent. The toll of heart disease. People don't die of a plaque. They die of a heart attack or a stroke for the arteries that go up into the brain.
Sometimes informally people talk about dying of old age. Is that a thing or is that just like a catch-all term for your immune system not doing what it's supposed to be doing? You don't die just of old age. You got to die of something, right? Because age itself isn't the killer. So there's got to be something else like explaining it. Could it be heart rhythm problem?
Could it be, you know, something that doesn't get diagnosed, an indolent infection, you know, many, many possible things, but you don't die just of old. There is the idea that one's telomeres decay, right? As cells divide, our genetic code is sort of being eaten away from the ends. How relevant is that? Well, there's a lot of things happening as we age. Our telomeres, which at the end of our chromosomes, they get shorter and shorter.
But there isn't any good way of lengthening them or lengthening them safely because they're also connected to your propensity to develop cancer. You don't want to muck around with them. And they're not nearly as good as the new clock. for telling you about your aging. So the new clocks, they include organ clocks. So what's amazing here is I could get a blood tube of blood from you and I could check. It's kind of like going to your car for every system, right?
It would tell me about your aging of your brain, your heart, your kidneys, your liver, your immune system, eight systems of your body from a tomb of blood to say. Is that organ in you aging faster than your chronological age, your actual age? And then I say, you know what, Sean? I got to get all over this risk for your heart.
and i'm going to check these other things i'm going to you know confirm that this is it and now this is what we're going to do for you so you never first of all we're going to slow your heart aging We're going to slow it. Here's how we're going to slow it. And we're going to measure it again to prove that we slow it. That's what's so different today. We didn't have organ clocks. We didn't have epigenetic clocks that tell us our body wide.
um aging relative to our actual aging if we have all these ways to detect what someone is not just the person but the organ is off track This is the science of aging crystallized to make a difference. That's great. Yeah. the nitty-gritty of what an individual can do to live longer and more healthy. But maybe we can sweep the deck of some of the crazier ideas out there first. I mean...
If I got a blood transfusion from, you know, children's blood or something like that, would that clean out my system and help me live longer? Yeah, there's a lot of these things that are out there, certainly in mice and animals, that if you take old young blood plasma and put it in old, you can see some... improvements in metrics.
In people, it's never been established that it's safe or effective with any durable impact. And it's a very expensive procedure called plasmapheresis that these clinics... particularly out here in california california exactly yes horrible but they're around the country too they're even in other countries It's shameful because there's no data. And, you know, this is not covered by any insurance. This is really expensive. And this is for the likes of these extremists like Brian Johnson.
the guy who had the netflix special who takes hundreds of supplements a day and has you know a penile sensor and naiden i mean this guy and he had his blood transfused from his son This is crazy stuff. And he's kind of the one, this don't die. He's promoting this stuff and people believe it, which is a shame.
when a 200 year old person starts promoting that i'm going to start believing them but until then i'm going to be skeptical but I was shocked to learn recently that the supplement industry in the United States is several times larger than the entire prescription drug. That's a really sad fact. As far as I know, it's true. And a lot of it is anti-aging. The other day I was in an airport in Oakland and there was a vending machine of anti-aging supplements.
Wow. Wow. I mean, wow. What are we doing here? Right. Well, I happen to know that the anti-aging supplement industry is very fond of advertising on podcasts. So I don't do that. I don't actually, I turn those down, but we're all over the place. The sad truth about that is that these longevity, so-called longevity doctors and scientists, they're the ones that are hawking the supplement. Horrible. None of them are proven. And they're not only...
lack of efficacy, but they're also questioned on safety. And of course, they're a big dent in your finance. Everything bad about it. Okay, good. So we've gotten that off the deck a little bit. I guess there's things we can do individually and things we can do as a society. I mean, let's talk about it as a society first.
One cause of death is just a pandemic or an infectious disease or something like that. I mean, how good are we at... I know this is a politically charged question right now, but how good are we these days at protecting against future pandemics? Not good. And the problem here is, you know, obviously we didn't do well for the COVID pandemic, but we haven't learned a lesson. So instead of... taking advantage of the things we learn, we're seeing gutting of our resources in the United States.
I mean, unbelievable takedowns of the programs at CDC, NIH, FDA. So this anti-vaccine, anti-science culture. So we're actually in much worse position than we were for COVID because of this. movement against science. And it's a shame because we learned a lot of what it takes to be ready. And now... even with threats that are out there, whether it's avian flu or others that we know are potential pandemic threats.
There's reluctance, there's denialism to confront what we should be doing. And it's really sad. Do you find yourself personally involved in political discussions or strategic discussions? Or is it something where, you know, all the scientists are already on one side. We sort of don't have to talk to them. Yeah. And then the problem is that the people who are running the show now... You can't reason with them because they were now empowered to take apart the systems that we have.
And so we're kind of in this auto-destruct mode. Whether you're dealing with measles and you say you got to get everybody vaccinated. you know this is our best protection and what you're hearing is take cod liver oil and vitamin a vitamin a have children in the hospital because of liver toxicity that has nothing there's no data to support that for preventing measles
But there are data, obviously, for the toxicity of taking these at high doses. Really, this is a sad state. We're going backwards. It's almost like we're going in the dark ages here. Yeah, I figured that was true. I figured it was never less important to get it on the record there. So thanks for that update. Let's instead ask questions about what you and I can do a little bit. I mean, you have a really interesting section of the book where you talk about lifestyle plus. Explain what that is.
Yeah, so we used to think lifestyle was all about diet, exercise, and in more recent years, sleep, health. Turns out that's missing about 10 other layers of lifestyle. they extend through things like ultra processed foods as a component of diet, no less the protein component. And all the details of... not just food, but also beverages, you know, alcohol and all that, caffeine. But then as you get into other layers like exposure to air pollution, microplastics, forever chemicals.
Time in nature. I mean, it's a long list of what we would consider lifestyle plus. You could say lifestyle. environmental, socioeconomic, you know, all these other factors, but it's hard to, it's too many syllables and words. So I just call it plus, you know. So, OK, I mean, that's just I know that some of these are kind of cliched and everyone is supposed to know them. What is a healthy diet? And if for specifically aimed at the goal of being healthy and active for more.
Yeah, so this is really topical because very recently there was a study, 30 years of follow-up, of over 105,000 people. And only 9% lived to age 70 without the chronic diseases we've been talking about. age-related chronic diseases, the big three. And so they could zoom in on what did they eat? What were these 9% different than the rest of the folks? And what was fascinating is it just confirmed everything we knew. at this powerful study. vegetables, the Mediterranean-like diet.
reduction in, you know, red meat, animal fat, ultra processed foods. So it was confirming everything we've known, but now in a whole new study, and it's called the optimal diet for healthy aging. So that's one thing. The other point about it is we don't, as we get older, we don't want to be light on our protein intake, but we don't want to be too heavy either. You take too much protein.
which you've got to work at getting, but there's at least some people out there advocating these super high protein diets. Wrong. Because that promotes inflammation and promotes atherosclerosis. So you don't necessarily want to be at the recommended dietary allowance. 0.8 grams per kilogram you want to be maybe around 1.2 but you sure don't want to go to you know 2.0 or higher which is what um some of the
People out there have been without data, you know, to preserve your muscle mass as you age. But then you go into other things like, you know, what type of exercise? Resistance training is really important. and it doesn't mean there's no such thing as 10 000 steps you know it's like no one ever there's even less than that you even weekend warrior exercise is great and then sleep the sleep thing is wild because
We always think that how many hours did you sleep? No, it's about how much deep sleep you got. which is reduced with aging. And also people take drugs like Ambien, which backfire. They sleep, but it turns out that sleep is how you get rid of all your toxic waste products in your brain that promote inflammation.
So if you take Ambien, what's amazing is you sleep more, but it blocks the efflux of these toxic... chemicals from your brain it's like the worst thing you could do i probably probably extends to all the other drugs because that's the only one that's been carefully studied So I knew that sleep helped sort of rewire the connections in our brain a little bit, but I didn't know that it was literally cleaning out toxic chemicals. cleaning out. That's what it's all about.
And there's interactions with our heart rhythm and the efflux of these ways. But drugs don't help at all. They seem to make things worse, especially the one that's most commonly used. So you can learn. how do you get better deep sleep you know you can you get a smart watch or a fitness band or um you know the aura and other rings and you can you can train yourself Okay, so that's crucially important. So say more about that.
Because everyone says sleep is good. I like sleep. It sounds good to me. But the fact that the drugs are counterproductive and nevertheless, I can train myself in what sounds like a more natural way, actually. and that the smart watches are actually helpful. That sounds important. Yeah, because it turns out if you start tracking it, like I've done, what you can see is I had terrible amount of deep sleep.
And of course, it does reduce as you age and men are generally less than women. So I was kind of in a high risk group to start. But what was interesting is by trying different things. I could see, oh, wow, I'm getting more and more and more sleep. So, for example, I never was trying to go. very precise about when I was going to go to bed each night and try to get into as close to a ritual as possible. That makes a difference, your clock, your biologic clock.
Other things like, you know, attending what exercise you're doing. what you're eating, how late in the day you're eating. some you know so many things interactions that you could start to see the pattern without AI and with AI, of course, it even better. We don't have that yet, but we will. At any rate, each person can learn about getting better deep sleep and it's really important.
And I don't want to skip over something you mentioned already, but it's worth emphasizing, which is the ultra-processed food. That's something that's hard to avoid in the United States of America in the year 2025. Hard to avoid, but easy to reduce. Okay. What counts as ultra-positive? Oh, gosh. Anything that you don't have in your kitchen that's normal ingredients for good home food.
um so you can tell pretty you you kind of do a smell test or a sight test if you look at a package and you see lots of ingredients that's a bad sign okay um and there's different types of ultra process there's some of these are emulsifiers like you know xanthan gum and but if you don't recognize this stuff yeah it's bad it's bad and what again it goes back to that theme we've been talking about is promoting inflammation in our body
and putting us at risk there it's like the opposite of what we've been talking about is increasing the risk of the big three age-related diseases yeah now the average american takes in 60 of their food is ultra which is the highest in the world. I'm sure. A friend of mine who wrote a book, Ultra Processed People, Chris Vantellican, who's a UK physician scientist, he did a 30-day test of maximizing ultra processed food. and a brain scan before and all these inflammation markers.
And what happened was, you know, he gained 20 pounds. He had big time inflammation in his brain in 30 days of this. And he had all these off the charts, inflammation blood tests. So the point being is that it makes you eat more. It creates much more inflammation. And if you just are careful.
about what, you know, the food that you intake and includes beverages that they have added stuff in there, particularly then you can reduce this. It's not hard to get it down to 20 or 30% instead of 60 or 70. And the stuff that kids eat, it gets up to 70% or 80%. of their diet. Do you have strong feelings about what are marketed as nutrition bars, like these little bars that have everything good for you in them, but it seems like they're probably pretty processed?
most of them are horrible okay horrible you look at them in the greens the other day um person gave me one that i said oh wow this one's pretty good it was you know just
truly natural, nothing added. But you have to really read the labels. And some of them that are widely used, marketed, that you think, oh, these are great. Just be... open-minded because you'll find that those labels are the good thing is that labels exist yeah and the bad thing is they got lots of ingredients that you don't recognize Okay. All right. Eat a carrot, I suppose. I guess I'm not going to be surprised. Carrots are great.
So, okay, I mean, even though it is obvious, this is very, very good to be reminded. It's not obvious, but it should be well known that eating better alcohol, not actually good in anything other than the tiniest of doses, right? Yeah, I think there's a lot of debate about that. The question is, what is a moderate? a dose right you know i recently had a patient and his wife i was seeing them and just before we were wrapping up the wife said to me is it okay if he has continues his moderate drink
And I said, well, what is the moderate drinking? And she said, well, he has two tequilas and six beers each night. I thought she was joking with me. I thought she was joking with me. Well, yeah. So the question is, what is moderate? And, you know, if it's two drinks. on each of a weekend day, you know, because it's helped social interactions and it's enjoyable. There's probably little data to show that that's harmful.
On the other hand, if you're drinking every day and you're having a fair amount of intake, usually it's more than people like to admit or think. Anyway, when you start getting to high doses, as you pointed to, that's a problem. And I guess the other thing to sort of wrap things up, the other thing that struck me in the book was your talk about mental health and social interaction.
You have a very vivid graph that you reproduce that basically, if I read it correctly, says after age 60, people are either alone or they're with their partner and they don't interact with other people that much. And that's a contributor. Absolutely. And it got worse during the pandemic, people just living in a cave, especially the older, the more likely. But it's still a big problem that annotated the pandemic.
The connection of that with bad outcomes is striking. So for mental health, but obviously, as you well know, everyone knows there's a lot of interactions between our mental health. stress and our physical health. And for mental health, with the more social interaction, the less isolation. the aversion to loneliness. These are really important things to strive against that, again, going back to that feature patient of mine. She has, you know, she plays Rummy Cube for every.
every week with eight of her friends. There you go. All in their 90s, you know, and that's a good thing. And she, you know, frequently interacts with other people in her retirement.
community building so these are things that are really much more they seem like social science but the data are striking and so we really should pay attention to it is there any truth to the idea that if i just If I solve puzzles, if I do the crossword puzzle, if I keep my brain active in that way, play video games, read novels, does that help at all with aging? Well, that's an interesting question. The studies are mixed.
There's a new one coming out that says the more you work as you get older with technology like iPads and computers, it's actually better, you know, which is interesting. Brains, just like humans are meant to be interactive. as a social as we are really a social animal even though people like to not be that way the same thing is stimulating the brain it's got to be good for you uh proving
is tricky because, you know, we didn't have the right tests. Like, for example, what I'd want to know is that biomarker we had for Alzheimer's. which can be improved by exercise and diet. I want to see that biomarker go down with this brain stimulation. Yeah. No studies like that have yet been done. Interesting. Okay. It seems like an obvious thing to do. Why haven't they done that one yet? Because these are new. We didn't have a brain organ class. until like the end of 2023. I see. Okay.
so it's these are studies are going to be and then we can ice it and we can say hey you know what these things are really good for you and you know the other thing sean it might be really good for you but not for another person Of course. None of this stuff is like all one one size for tall. And we have to keep that in mind, too.
And to bring it home, that's another place where AI might be useful, right, in personalizing your regimen, whether it's diet or exercise or mental activity. That's exactly right. We would know in general the things that would help, but then by tracking these different metrics over time and people who are at high risk. The AI can say, well, this is what's working and this is what, you know, the gaps and this is what's going to change.
I call it precision medical forecasting. You may know this, but AI has radically changed weather forecasting. It hasn't been implemented. But now you can have 99.99% accuracy in weather forecasting like we never had before. I mean, most of the time, it seems like the weather forecasting is pretty far off, but that's going to be changed quickly because multiple groups.
have shown that now with these large language models, multimodal data, you can get such remarkably accurate weather forecasting. Well, we're about to do that for people, for their health. That, to me, is even more important than the web. recognizing patterns. That's what they're very, very good at. And I'm sure so we can finally all agree that listening to the right podcast...
is probably the best thing that can help you live a healthier, longer life, right? You got it. No, this has got to be the one. This is definitely stimulating the mind. So let's do it. That's all we can ask for. Eric Topol, thanks so much for being on the Mindscape podcast. Thanks for having me, Sean.