What if aging was a disease, and that disease was treatable?
Enter the “Information Theory of Aging”: A radical approach to aging by Harvard Professor David Sinclair.
Even as we grow old, Sinclair believes a “youthful” version of the epigenetic expression of our DNA is saved in our cells. If we are able to retrieve and manipulate that stored youthful version...can we extend life itself?
Geoffrey Woo has read Sinclair's new book "Lifespan", dove into the research, and spent the time formulating a compelling analysis. Continue below for Geoff’s final verdict on “Lifespan” & the most promising anti-aging practices.
"Lifespan" is structured in three main components:
David Sinclair really ties a nice personal trajectory throughout each of these three components.
First, let's go over the history of aging research. We all might remember our high school biology textbooks where it covers some hypothesis of how life started billions of years ago on this planet earth. What I think Sinclair does really well is that he ties that primordial story with some of the early, early versions of longevity pathways that we all know and love today. He describes early precursors of the sirtuins, a family of proteins that assist the longevity pathways.
From the past, we transition into the present day. What are academics and researchers looking into today? What are the main mechanisms and targets people are focused on? Also, what are some practical tips that you and I can start doing today? These are things like intermittent fasting, exercise, et cetera.
From the present day research and best practices, we build into the future...the key novel idea presented in the Lifespan book: The information theory of aging. I'll spend most of the time talking about this because this is really the meat of the book here.
The last section Sinclair talks about the morality and ethics of living substantially longer. What are the ramifications?
I'm going to focus on areas that I find particularly novel or areas that I disagree with Sinclair on. I hope my focus on disagreements doesn't imply that I don't think this was a good book. Simply, disagreements are the most interesting part and what really drives progress.
Sinclair makes a lot of analogies with computer science concepts. He's a geneticist using computer science as a metaphor. My formal training is in computer science, and that's a lens I bring into human performance and human physiology. The computer system is really a great analogy for the human system. There are subcomponents and programs within the human like there is within a computer, so to me, we're really converging on a similar analytical framework, but from very different areas of training. We should still be very humble to acknowledge that the human is orders of magnitude more complicated than the computer.
On page 21, he starts with analogy that "if the genome were a computer, the epigenome would be the software." I think this is a decent approximation as a biologist using a computer science analogy, but as a computer scientist interested in biology, I would propose a more nuanced, refined analogy. "If the genome is the lines of code in a computer program, the epigenome is a runtime environment for the code."
Now, let me break down what I mean by that. The genome is really lines of code that are turned into amino acids that construct the myriad of different proteins that make up our tissues, organs, and bodies. This means that the epigenome really controls which proteins are expressed or produced at which times. I'll be making minor suggestions and critiques of different computer science analogies throughout the review of Lifespan.
Let me cover a few key concepts that Sinclair covers in part one of Lifespan.
Waddington's landscape. The idea here is that cells start undifferentiated, or as pluripotent stem cells. They eventually specialize and differentiate into different types of cells: neurons, liver hearts, skin, nail, hair, and these are essentially grooves or paths down a mountain that go from unspecialized to specialized, and that specialization is very important. Obviously, you don't want brain cells on your skin, and you don't want hair or toenails in your brain. That is essentially tumors or cancer. As the epigenome, or the runtime environment, gets damaged, the differentiation becomes jumbled, so now you're getting cells that have partly neuron aspects and partly skin attributes. That's a tumor. The epigenetic mechanisms really truly affect the DNA core. As we talked about, the DNA is a massive string of programming code, and the epigenome is infrastructure that controls which part of the code is run at which time. As aging happens, the runtime environment gets confused, so you're printing out lines of code at the wrong time, and then the differentiation gets jumbled.
Another key idea is Horvath's clock. We have the notion of chronological aging. I am 30. You might be 35, or you might be 15, and that's based on aging by time. However, that's really just the correlation of our biological age, meaning how much aging or damage has my body accrued over my lifespan. Horvath's clock is an interesting idea that you can tell the biological age of a person through epigenetic and methylation on a cell's DNA. It's done through machine learning on thousands of samples, and there's some exciting data that you can predict the lifespan, the health span of a person just through the DNA methylation and epigenetic signatures on a cell. But I do want to caveat here is that in some private conversations of mine with researchers, Horvath clock doesn't work on all tissues. I've heard that the Horvath clock doesn't work in muscle tissues, so I'd be careful on to close the book just yet. For example, another very good predictor of health span and lifespan is the functional strength of muscle. How much can you be pushing a weight? How much VO2 max do you really have? As we know, those are really good markers for discussing health and longevity. The notion of a Horvath's clock or epigenetic clock for aging really sets up the metaphor that Sinclair uses as he sets up the information theory of aging. You have a DVD disc, and that DVD has your movie on it. That's kind of like your DNA. It has all the data on this disc. However, as you age, this DVD gets scratched, so the information gets obfuscated. This is kind of like epigenetic markers of methylation. The DNA gets marked or scratch with methyl tags. Before we dive really deep into the information theory of aging...
Let's circle back to what Sinclair believes are the current best practices for life extension.
I think generally, good best practices. Obviously, eating less than intermittent fasting is a big part of my personal lifestyle. I've been intermittent fasting for the last four years, and it's a very big topic of conversation in a lot of our content.
However, point two, the notion of amino acid or protein restriction, especially that of animal protein, I do have some questions and I want to push back on. The bulk of the science behind why red meat or animal protein might be harmful is based on epidemiology. These are observational studies with a high likelihood of confounding factors. There's no randomized controlled experiment. The study is based on food surveys that correlate that people that eat a lot of red meat have higher cardiovascular disease risk, but the main confounding factor here is that in a Western culture, in a Western civilization, what do we tend to eat with that red meat? We tend to eat red meat with alcohol, with soda, with other dangerous non-healthy behaviors. The epidemiological study that is critiquing red meat is actually just detecting a health or user bias. Over the last 20, 30 years, we've been told that red meat is bad, so health conscious people tend to avoid red meat. When you actually look at different population cohorts in different eating patterns, for example, in East Asian cultures, there's actually epidemiological data that suggest the very opposite. You get better longevity, reduced mortality with higher levels of red meat consumption. I'm not convinced that red meat needs to be avoided likes Sinclair declares in the book.
I think there truly needs to be balance here, where if you restrict too much protein and too much amino acid, you risk sarcopenia, or muscle wasting, or muscle loss. I understand why Sinclair cautions against too much protein because you want to reduce overstimulation of mTOR, which is a nutrient sensing pathway. We'll talk a little bit more about that. But on the other hand, you do need the right amount of protein so you can retain and maintain lean muscle mass. Again, lean muscle mass prevents injury. It's actually a very good metabolic dump for extra glucose or extra substrate through overconsumption. To me, there's a balance. Obviously, I'm not advocating eat a ton of red meat 24/7, eat 100% of your calories from protein, but I think it is important not to underdose protein because- ... but I think it is important not to underdose protein because sarcopenia, frailty is actually one of the leading causes of mortality.
The third guidance is exercise. Obviously I'm a big proponent of exercise, no arguments there. We should all be exercising more.
Point four, he talks about, is getting exposure to cold. And I think there's actually pretty good data that exposure to cold helps convert white adipose tissue into more metabolically active brown fat. I think it's reasonable. If you want to get some cold exposure, definitely go get some cold exposure. But I do want to caveat, this is very different from cryotherapy for sports recovery. There's really good data that shows that cold treatment bull once exercise adaptation, that is, cold treatment actually cancels out some of the metabolic benefits of exercise. So, be careful not to use cold overabundantly or using it wrong. And his last point is to avoid environmental toxins as much as possible. This is like radiation and cigarettes. I think this is pretty basic stuff. I'm just trying to agree with David Sinclair, avoid doing stupid stuff. Again, I understand that some of the stupid stuff might be fun and that's a lifestyle choice, but try to avoid it as much as possible. Now, we've covered some of the historical work and historical concepts that are important to understand aging as well as some of the practical best guides of what to do with lifestyle.
Now let's move on to current day research and what Sinclair thinks are four promising metabolic longevity pathways.
Area one are the sirtuins. These are proteins that go around the DNA that helps repair DNA breaks and also controls and mediates expression of longevity pathways. If we can activate sirtuins, can we extend health span, extend lifespan? One of the key compounds that Sinclair focuses on is resveratrol. You might have heard of resveratrol as one of the miracle compounds from red wine. Obviously, a lot of that hype has really tempered down in recent years and it's kind of a mixed bag. How actually effective is resveratrol? The most compelling scientific results with resveratrol are on fat, obese mice. That means you're extending health and longevity on mice with a fat obesogenic diet. However, the effects really start attenuating or reducing or even disappearing on healthy mice. So that is to say resveratrol might be beneficial for unhealthy diets but maybe has no benefit or added benefit on a healthy diet. But that is not to say that this pathway isn't interesting area of research.
Next up, Sinclair proposes NAD boosters. You might have heard of things like nicotinamide riboside, NR or NMN, nicotinamide mononucleotide. These are all precursors that turn into NAD. NAD is an important part of the Krebs Cycle. You remember from high school biology, the Krebs Cycle is a process that is conducted in the mitochondria that makes ATP. To make ATP, which is the energy currency of life and of cells, you need NAD to make ATP. NAD levels tend to decline with aging. So, the theory is, if it can boost up NID through things like NR or NMN, can you reverse some of the aspects of aging?
The third direction of research is AMPK and things that activate AMPK. Well, what does AMPK stand for? It's a little bit confusing. AMPK stands for AMP-Mediated protein kinase. You kind of got two AMPS nested within each other. Well, what does the first AMP stand for? Adenosine monophosphate. And that's essentially a building block or a lower energy state of ATP, adenosine triphosphate. And ATP, when it gets used up, converts down into ADP or AMP. What AMPK does in the cell is AMPK detects the AMP to ATP ratio. The more ATP you have, the more energy you have. But when AMP to ATP ratio increases, meaning more AMP versus ATP, it means that the cell detects less energy available for the cell. And that kicks off a lot of healing and metabolic processes that tries to upregulate or increase the availability of ATP. This is a beneficial response to low energy. So, things like additional mitochondria are formed and other things related to energy processing are kicked off when AMPK is activated. Now, one of the most interesting compounds that people have been looking at relating to AMPK is metformin. As we've discussed before, metformin is a diabetes drug. And why do people think this works? Well, it's an AMPK activator. So, what is actually going on here? Well, metformin is actually kind of like a mitochondrial poison. When metformin is in the cell, it actually disrupts ATP production. So metformin reduces the efficiency of AMP to ATP conversion so that AMP level increases. That kicks off an adaptive response. As AMPK increases, the cell responds by creating more mitochondria and other protective pathways. As we'll talk in a little bit, I'm skeptical about this approach.
Lastly, fourth direction of research is around mTOR and mTOR inhibition. mTOR, mechanistic target of rapamycin is the primary nutrient sensing pathway for our body. MTOR is especially sensitive to protein and amino acids, especially leucine, and also takes in insulin and glucose as inputs into mTOR. And when mTOR is activated, it means it's a time of growth. But when mTOR is inhibited or off, it's a time of autophagy. When you talk about fasting or ketogenic diet, you really inhibiting mTOR and when you're talking about heavy exercise and eating a lot of protein, you're talking about activating mTOR. I don't know. One of the more hyped compounds in the longevity space is rapamycin. The target of rapamycin refers to this compound. The way rapamycin works is that it's a very potent mTOR inhibitor. That might lead to autophagy and all the productive pathways when there's low nutrients available in the system. I had a really good podcast with UC Davis Professor Keith Barr, and we talk about rapamycin and mTOR. I think that conversation really covers it well. But to review, I think with a general inhibition of mTOR that's not tissue specific, that runs the risk again of sarcopenia and not allowing your muscle to really build. Again, you want to inhibit mTOR, but to some level. But you also don't want to activate mTOR so much. They are not actually building muscle, building protein when you actually need to. There's a trade off here. So, something like rapamycin, which is, shall I say, not very targeted, that might bring high risk to the situation.
I would agree with Sinclair that these are four very important pathways and mechanisms that we should think about when we talk about longevity research.
I would personally also add insulin pathway as well. Insulin is also a nutrient sensing pathway and it sense that sensitive, obviously the glucose, and insulin kicks off a bunch of longevity and growth factors as well. One point of feedback and one thing that I wish Sinclair did a little bit better is tie all these different pathways together. You really have to have a broad background in all of these ideas to understand how these components are not actually independent. They're all actually very related and work sort of together in a cohort.
For example, exercise increases the NAD pool by improving the NAMPT regeneration pathway. Exercise also increases and activates AMPK because as you're exercising, you're using a lot of ATP, therefore increasing AMP. And exercise also likely activates the sirtuins by increasing the NADH to NAD ratio. So, exercise hits three of those four pathways we just talked about. And we should really think about all these pathways as a system and as a cohort. So, thinking about this from a holistic, integrated system of pathways and networks.
We should also really talk about ketosis. Ketosis impacts the sirtuins, particularly SIRT1 and SIRT3, and there's actually a histone acetylation inhibition that triggers FOXO3, which is a longevity pathway. It also impacts the NAD pool by increasing the NADH to NAD ratio. Again, I think there's a concept between the availability of overall NAD, but what might be also more important, or just as important, is the NADH to NAD ratio. And that ratio is increased while the cell is in ketosis. This again has implications on sirtuin activation. Ketosis also activates AMPK. When a cell is in ketosis, that often implies energy deficit and AMPK increases fatty acid oxidation and all of the things that we want AMPK for. And then lastly, being ketosis is avoiding mTOR sugaring because when you're in ketosis, you're typically in a low carbohydrate environment. So, ketosis seems to be a very key metabolic mediation state that actually controls a lot. Key metabolic mediation state that actually controls a lot of the downstream pathway that Sinclair talks about. So now you're starting to see the puzzle pieces of why I'm so excited about ketosis. Fasting is a great way to get into ketosis, but I'm particularly excited about ketone esters. What if you could get the benefits of ketosis without having to fast or eat a ketogenic diet. I don't think ketone esters will completely cover all the different implications of fasting or a ketogenic diet, but it might cover a lot or the most important parts of fasting and the ketogenic diet. Ultimately, I think that some combination of all three, a combination of fasting, eating low carbohydrate diet or a ketogenic diet at times cyclical, and incorporating ketone esters will be the future of nutrition, especially for longevity.
It's also why I'm skeptical of drugs or supplements like metformin, rapamycin, sirtuin activators, NID precursors or boosters. It's just not clear to me if any of these molecules do anything on top of exercise or diet. So what I mean by that is that the studies that I've seen show relatively limited benefit, if not negative benefit of these things like metformin when you put them on top of good exercise and good diet habits. So that is not to say that there's no benefit on people that are not eating well or not exercising, but for healthy people for extension of longevity or extension of health span, I just haven't seen the data. I wish that could be true because I want that magic pill. But I'm not convinced by the science. I don't think Sinclair is giving wrong advice or misguided advice. I don't think that the risk benefit is there for me. But to find middle ground, I think that Sinclair and I would both agree that all of these are sensible targets to think about when we try to come up with strategies to maximize lifespan. We might disagree on the best techniques that target them, but we will both agree that these targets are reasonable and we should think about triggering them in the right way.
Now let's talk about the meat of the book.
The information period of aging. To be honest, it really wasn't what I expected. When I hear information theory, I think of Claude Shannon's theory, which is very, very mathematical and quantitative. Information theory to me as a computer science is about the theoretical mathematical limits of compression algorithms, of bandwidth, of maximal error correction. However, Sinclair is more inspired by the structural conceptual ideas that Shannon had. It's a much more qualitative use of information theory concepts.
With the information theory of aging, Sinclair proposes that there is an observer that tracks a youthful version of the epigenome. The idea that aging is simply a loss of integrity of epigenetic information is not really novel. What I believe is novel is that Sinclair believes that there's a stored version of a youthful epigenome in every single cell and that youthful epigenome can be recovered. Now, Sinclair doesn't know how exactly this youthful epigenetic state is stored, but he speculates that this might be related to methyl tags on the DNA or some mechanism yet to be uncovered like a protein, an RNA or a novel chemical attached to that DNA that marks this stored youthful epigenetic state.
Another use of the bridging of information theory from computer science to aging was Sinclair's use of TCP as a major analogy. I think TCP is an overengineered analogy here because TCP entails ideas of buffering and packet ordering and is extraneous to the point he's trying to make. Because the protocol doesn't really capture what he means. A simpler analogy is like Google Docs. If you ever use Google Docs, you can go into the history and find the revision history of every single edit made on that Google doc. You basically have stored safe states of a cell and Sinclair believes that biology also works like that. You can have an epigenetic state stored in its youth, and we can reverse to it. Again, I think the idea is super exciting. I think the analogy can be a little bit better. Hopefully my analogy makes it a little bit more understandable for people out there.
However, the actual lab benchwork that allowed Sinclair and his team to come up with the observer idea was pretty damn phenomenal science. So what he did was that he was able to reprogram optic nerves in mice using a cluster of gene edits, OSK. I thought this was actually the most interesting part of the book for me. In fact, when I had read this part of the book, I immediately looked up and read the original bio archive manuscript. The paper is still pending peer review, so it's very, very cutting edge. So what do they do? Lou and Sinclair in mice regenerated crushed optic nerves. They reversed induced glaucoma damage and they reversed age-related blindness in rodent eyes. It literally took years of work and a lot of credit to Yuanchang Lu who was the first author in the paper and Dave Sinclair to really make this work possible. And they did this through epigenetic reprogramming. I see two explanations here.
One, the epigenome is reversed to a more youthful state because Sinclair is right and there is an observed youthful state that's stored in each of these cells. Two, the cells just mimic the epigenomes of the surrounding healthier cells. The latter is what I hypothesize.
The last section talks about the ethics and morality and justification for life extension research. It really opens up the possibility of what might be possible if we cured aging. I fundamentally agree with Sinclair that dying so soon is very sad. We spend so many decades learning skills, being educated, and really learning how to be human beings. I know that you know at 30 I'm so much more of a complete human than when I was 20 and I imagine at age 40 I'll have so much more life experiences and maturity and wisdom. If we can have this wisdom compound decade over decade, we'll have so much more of a productive and hopefully more enlightened society.
One thing that I personally don't have a good answer to, and one thing that wasn't covered in the book was how does wealth and power transfer in a society where people are living longer if not forever? It's arguable that death and estate taxes really helps prevent a massive wealth inequality. Imagine if Jeff Bezos or Donald Trump gets to live forever. Someone like Bezos will literally compound forever like crazy. It's not a farfetched situation that you have capitalism run amok where he literally owns all the capital in the world because he has thousands if not millions of years to keep compounding and keep acquiring assets. And he reaches some escape velocity of wealth capture. One thing that we all face is that there some time limit. You die.
I still think that the trade off for living forever outweighs the risks, but it is an important thought that we need to resolve if we do really have a path to immortality.
In conclusion, I really enjoyed this book. David Sinclair, thank you so much for putting together this work and really sparking the conversation around longevity and a path towards living much, much longer. Obviously I have some pointed critiques and feedback, but Sinclair really sparked and inspired me to dig in and formalize my own thoughts and understanding of longevity. There's just open science to be done here.
I would like to take a bet on this. I would take cyclical ketogenic diet plus daily cardio and strength training versus a regimen of metformin, NMN and resveratrol and whatever other supplements you want. Run it over a couple of decades in humans. Hell, we can even include a third arm that includes the keto diet, cardio training and the supplements like NMN, metformin and resveratrol. I would take the side that regular exercise and a clean cyclical ketogenic diet would have better health outcomes and life span compared to the supplement group. Which bet would you make?
To me, this is a 5/5 for the lay reader and a 4/5 for people that really want to dive into the biology and the mechanism behind some of the pathways. Let me know what you think of my review by commenting below, on our Youtube video, or tweeting me @geoffreywoo.
Once a week, we'll send you the most compelling research, stories and updates from the world of human enhancement.