In this session, Alexander Fedintsev, co-founder of the Radical Life Extension biohacking group, came to discuss why biohacking is a necessary component for radical life extension, how they do biohacking in the RLE group (with a few tips for better and more scientific self-experimentation), what they already achieved (like their well-known plasmapheresis trial), and more. One of the focuses of the talk was also the importance of extracellular matrix, possibility of it being the 10th hallmark of aging or maybe even cause of aging, as well as their future plans of studying it.
Presentation: Alexander Fedintsev
- As Feynman said, there is no physical law that limits longevity. This is basically what motivates us and drives us to try reaching radical life extension (RLE).
- How do we define it? As at least 2x increase in life expectancy, with the ultimate goal of course being longevity escape velocity.
In the presentation I’ll try to answer that by answering these three questions:
- Why is biohacking a necessary component for radical life extension?
- How do we do biohacking in our group?
- What have we already achieved?
- First let’s try to analyze how far we are from achieving longevity escape velocity. I’ve gathered the data from European databases.
- Here we can see how many years of life expectancy on average we gain every year. On average in the whole Europe it’s 0.25 years of life expectancy every year. This is life expectancy at birth, and it looks pretty good. It’s not that huge of a difference if you put it in perspective though – a 30 year old would have an additional 15 years with this pace, so it’s not a dramatic increase even with this linear progress.
- Now if we look at life expectancy at 65 years, we see the slope becoming gentler – so the life extension is not uniform across age groups.
- And if we look at life expectancy at 85+ years, it flattens out almost completely. Which means that the current progress in healthcare doesn’t really translate to deaths at advanced ages – the current healthcare progress doesn’t work against aging.
- Here you can see the difference it results in, it means that people in old ages are virtually doomed with the progress of current medicine.
- We can clearly see that in this chart. You can see mortality rates at logarithmic scale divided by income brackets – poor and rich people. The difference in mortality at 40 is huge, but then it shrinks towards no difference at advanced ages. Another proof that current healthcare cannot significantly prolong life of older people, however rich they are. So this is where we need new therapies and approaches to medicine to do something with aging.
- So what is the current state of the affair in the current longevity research field? It’s also not very optimistic, at least according to this study. It has shown that out of almost 30 life extending genetic manipulations in mice, only 2 have shown signs of demographic aging retardation. What does it mean in practice?
- Let’s look at this chart – these are survivorship curves. Let’s say the yellow one is the control cohort. The green curve represents the aging rate retardation – the intervention that actually decreases the demographic rate of aging. We can see how dramatic the effect is, all cohorts already died out while a half of the green population is still alive. That’s what happens when we influence aging per se. The blue curve represents current healthcare – it extends life in the beginning, but then the curves converge without any increase in lifespan, and that’s not sufficient. Luckily we don’t need to retard the rate of aging to achieve maximum lifespan extension, as we can see in the orange curve – this is a parallel transfer of the survivorship curve which we can see with some of the mice intervention. They don’t retard aging rate, but still demonstrate quite decent improvement in both median and maximum lifespan. And this is what we want to see developed from big pharma.
- So what can we see from big pharma? We see that the process of adoption is pretty slow. And this is why we need biohacking, because it can speed up the process of adoption of different already known interventions dramatically. But only if biohacking is done right.
We identified three main pillars of correct biohacking.
- Use mathematical frameworks for evaluation of potential interventions, as we need to understand how many years of average it can add to the lifespan to see whether it is even worth testing
- Avoid typical experimentation mistakes, as self-experimentation is not an easy thing, you can fool yourself easily if you don’t do it correctly from the statistical point of view
- Try to identify and target the underlying cause of aging, because without it we’re only again solving symptoms
- We developed a mathematical model that is able to convert risks of all-cause mortality to years of life expectancy change, because many studies report relative risks of all-cause mortality, but the quantities are not intuitive. It’s hard to tell whether a risk reduction of 10% is worth it for example. People usually don’t understand it, so this model converts these risk reductions to more understandable quantity of life expectancy
- Many biohackers do not account for the statistical phenomenon of regression to the mean. Most of the biomarkers, especially blood biomarkers, have naturally occurring variability, some with relatively high amplitude. So you can accidentally take one measurement when that biomarker is at the top of the amplitude, and then the second one when it is at the bottom of the amplitude. That can lead you to an erroneous conclusion that the experiment led to decrease of this biomarker of interest.
- What to do with that? First we can try to pick markers that have relatively small variance. Markers that are robust, stable, and ideally almost monotonically increase with age. For highly variable markers, we have a history of measurement per person, so we can plot a distribution of values and see whether the after experiment value is an outlier – that it is unlikely that this value is a result of the natural variance of this marker.
- Usually biohackers measure dozens or more of biomarkers. Just by pure probability theory, you can have several biomarkers significantly change from measurement to measurement due to the number of comparisons. To avoid that we apply sophisticated statistical methods like Benjamini-Hochberg correction. It’s not a silver bullet, but it improves the chance that we discover something meaningful and not just noise.
- We believe that the current view of the aging process is cell-centric. Most of the scientists seem to focus on cells – epigenetic reprogramming, manipulating NAD+ levels, killing senescent cells. But they seem to overlook the changes that are happening in the environment of these cells, namely extracellular matrix, which is comprised of long living proteins like collagen, elistin. These proteins undergo chemical modification like glycation, which leads to cross-link formations, adducts formation, etc. And we have shown in this paper that changes in these molecules can lead via various mechanisms to virtually all hallmarks of aging – so the true underlying cause of aging might be just this damage to the extracellular matrix.
- We can see here that old mesenchymal stem cells placed on extracellular matrix of younger mice expanded like young cells – they preserved the traits of young cells. While younger cells placed on extracellular matrix from old mice displayed traits of old cells – they divide less, have specific biomarkers of old cells, etc.
- So my hypothesis that I’m advocating for is that since senescent cells contract fibrosis, then changes in extracellular matrix during aging (stiffening, matrix becoming resistant to collagenesis, etc.) mimic fibrosis and cells erroneously think that there is fibrosis and turn to senescent cells and start doing their job to come and correct the fibrosis. In a way you could therefore call aging pseudo-fibrosis. We are working on experiments that will either confirm or deny this hypothesis. We believe that therapies that can rejuvenate this extracellular matrix (cross-link breakers) could then lead us to radical life extension. Currently we don’t have any cross-link breakers, so it is difficult, but we can still do something.
- Let’s discuss what we’ve already achieved and what we plan in the future.
- These are the main things our group is focusing on.
- And these are our main focuses.
- This is the timeline of our progress
- Let’s talk about our history, in 2016 I worked on a dataset with a huge table with lots of biomarkers from one gerontology center in Russia analyzing data. I found out that several markers of arterial health were strongly correlated with chronological age. So I decided to build a model (a clock) to predict the age.
- I was surprised that the model I’ve built eventually, it likely was not only good predictor of chronological age but also a good predictor of overall health – you can see that people with diabetes or hypertension have significantly higher biological age. This biological age was mainly driven by four factors – carotid intima-media thickness, pulse wave velocity, degree of stenosis, and augmentation index. So only four biomarkers resulted in a model with an error of 6 years and likely represent biological age estimation. And remarkably the most significant biomarker in this index – carotid intima-media thickness – happened to be a very robust biomarker. Very stable, almost monotonically increasing with age, so very hard to reverse – only a few drugs can do this.
- But we have managed to find a drug combination that exceeded our expectations. Particularly valsartan + fluvastatin reduced the carotid intima-media thickness almost two-fold in both of our volunteers and co-founders Stanislav and Yuri. This reduction was pretty stable, it persisted for about a year, which is remarkable.
- Another very important biomarker in this model is pulse wave velocity, which reflects arterial stiffness. It is pretty variable biomarker compared to carotid intima-media thickness, and hence we needed a device that would allow us to measure it multiple times before and after the experiment to average our measurements and have more statistically robust conclusions. There are not many devices that can accurately measure this biomarker on the market, so we developed it ourselves.
- This is the first prototype, we used it already for our trials, we took our learnings from that and we are working on the next version that should be on the market next year.
- Our another achievement is the plasmapheresis trial, which is pretty well-known in the community. We didn’t expect to observe dramatic improvements in biomarkers that we would treat as promising, we just wanted to understand the logistics of the whole plasmapheresis process. Because you need to replace half of your plasma with the saline + albumin solution and this is not a simple and standard procedure. But we managed to calculate how many plasma you need to donate with each visit to the doctor and how many albumin you need to replace and we did this and surprisingly we have found some pretty interesting changes in the biomarkers of this gentleman. We have found, for instance and contrary to our expectations, that cholesterol goes both directions – bad LDL goes down and good HDL goes up, which is pretty interesting. Of course we have only two data points, so we cannot draw too many conclusions from that, but we have started a clinical trial aiming to compare plasmapheresis with albumin and without albumin, because the role of albumin of the whole procedure is an interesting question.
- A few smaller things our group has achieved. We have tried various senolytics in our volunteers. Created a lentiviral vector for APO-A1 Milano gene delivery. And also a microbiome replacement experiment, because we have access to samples from soviet cosmonauts (who are usually considered exceptionally healthy, so our hypothesis is that transferring the microbiome could yield interesting health improvements).
- Here are several things we are planning to deliver in the upcoming years.
- We are intrigued by the study showing muscular aging through 15-PGDH, and we want to reproduce it on ourselves.
- Another target is epigenetic rejuvenation of hematopoietic stem cell via targeting Cdc42. This type of cell is very reluctant to different approaches in reversing aging (even our extracellular matrix one), so we plan to rejuvenate them and investigate how to maintain the useful environment for these rejuvenated cells.
- The third thing is targeting elastogenesis. Elastin is now considered to be one of the longest living proteins in our body, elastogenesis is limited to early infancy and then the old synthesized elastin remains in our body, accumulates calcium, is degraded by enzymes and so on, therefore we lose elastin which leads to progressive deterioration of various tissues – blood vessels, skin, lungs, ligaments, muscles,… All tissues lose their elasticity and that is crucial not only for appearance but also functional health. We can try – and already have some methods – to increase elastin production in vivo.
- The fourth thing is the development of anti-aging gene delivery system. This is what we have tested already.
- We have some promising early results that it might work, so I hope we’ll be able to achieve some impressive increases in life expectancy with it in the future.
What convinced you that biohacking can be impactful? At what point did you decide that it is a worthwhile investment of time and money?
- My interest in the life extension topic began in 2012 and at that time I believed that we will defeat aging shortly, within a few years. But then after a few years, radical life extension was still very very far away, and I started digging deeper into the science trying to understand why the progress I anticipated isn’t there. From that a critical mass of information accumulated and the strategy of radical life extension emerged. We understood that the cause of aging might likely be in the extracellular matrix and that it will take a pretty long time to find ways to rejuvenate this vast majority of long living molecules. It’s not an easy task and therefore we need some ways to win some time before that, because it can easily take a few decades before we get to that point.
Do you have any idea how we can do better in educating potential biohackers about the common pitfalls you mentioned at the beginning? How do we prevent other people coming into biohacking from repeating the same mistakes?
- We might start an initiative where we can accumulate the best practices, a centralized hub that will communicate the learnings from previous mistakes. It’s a very risky thing and biohackers are really brave trying all the new things. It would be a good thing creating a place where all these things could be shown to the biohacking community – promoting the successes and warning about risks.
How do you guys control for compounding variables for example in lifestyle? Do you track your sleep, exercise, diet daily?
- Of course, we have wearables like Fitbit where we can see the distribution of activity before, during and after the experiment. However we think that controlling for confounding variables is especially important if the effect size is very small, so based on the effect size different confounding variables play a huge role. We are trying to pick up markers that are very stable and aren’t that easily influenced by changes in lifestyle, so it increases the robustness of our findings.
You showed a nice chart of all mortality, I wondered if you have access to some data that excludes things like accidents, injuries, wars, so we can look more focusedly on age-related diseases.
- Yes, this data is already filtered to deaths from natural causes.
What are you doing personally in terms of health extension and how do you avoid microplastics when doing plasma dilution?
- Since I am relatively young, I’m not doing anything special, just physical exercise, a more or less healthy diet, and trying to avoid accidents. Other than that I am not taking many supplements.
- We haven’t done anything about that yet, we’ll evaluate the importance of that in the future. We tried this as a quick pilot study.
You mentioned re-expressing elastin. I’m wondering what the limitations of that are if you are not removing the old damaged elastin. You could even make things worse potentially. What are your thoughts on that? And even more broadly about this type of approach for other molecules or macromolecules that accumulate stochastic covalent damage over time.
- There are studies that are promising – some types of drugs increase the expression of tropoelastin gene and the mice that took the drug had more elastic aorta throughout the time – the elastin was properly organized. So this gives us hope that simply stimulating the expression might work but of course we won’t take it first on ourselves, we’ll probably try it on some animals. And of course tropoelastin is not the only thing that might be needed to restore the elastin network, because elastin is a complex protein composed of monomers and the main structure of elastin is made of fibrillin, so we might also need to overexpress this protein. It’s difficult to say how this will turn out in advance, that’s why we need lots of experiments.
We have many individuals who are coming from the engineering or biomedical background who are 35 years old or younger, such as yourself, who are working on biohacking. But it seems to me that the correct population for this would be those where it would be clear whether a thing is actually having an impact or not. So I think biohacking would be better focused in the 65-70 age where you have remaining normal functionality, whereas later might be too late for some elements of bodily function. In your research, how many people using your protocols are in the 65-70 range, and are they actually experiencing reversal of functional limitations?
- The majority of the volunteers in our group are of age 55-60, so it’s very close to that range you mentioned. Maybe they are a bit young, but that might also be a good thing because they can withstand more negative influences if the intervention goes wrong. So maybe that’s another reason why too old people might not be a good idea. But it is probably correct that the effect will be more apparent on older people, and we should aim for that. But maybe we’ll find somebody who is older and wants to participate.
My question orients about the demonstration of regression of troubles and therefore functional improvement. Have you seen that? Not biometric improvement, but actual functional improvement like strength, gait speed, cognition rates, and so on.
- Yes the functional improvements should be more apparent for example in our experiment with the 15-PGDH, which should rejuvenate muscles – we expect to see increased muscle mass. It is a beautiful experiment because we can inject a drug into one leg quadriceps and the other will be intact or with saline. And then after doing bilateral exercises with the same load, we will see how things progress with time. So that is an experiment when the functional improvement might be apparent.
As a biohacker, how do you feel about sharing all your results but also personal biodata with the public? And also do you think other biohackers share the same sentiment or is there a diversity in opinion on privacy amongst biohackers?
- I am not that familiar with the biohacker community in general, so I cannot speak for everybody, but in our team it depends – some biomarkers are ok to publish, some probably not. I wouldn’t publish my genetic data for instance. But things like vascular age or blood glucose, I regularly publish that even on my Facebook.
For the arterial stiffness sensor – can you give more details on how that works or is there a place online that we can go and read about it or is it published anywhere?
- No it’s not published, it’s basically two photo photoplethysmographic (PPG) sensors in two locations – finger on the hand and a toe and that’s basically it. It measures the time it took the pulse to travel a certain distance and then you calculate the pulse velocity out of that. The same principle that medical devices use, but it’s portable.
What is the most impressive achievement in biohacking that you discovered?
- We have observed reduction in atherosclerotic plaque in some older people, but we are a bit skeptical about it because we need more data to confirm it – seems like a too optimistic finding.
What data did you have so far as evidence for the removal of the atherosclerotic plaque?
- Ultrasound examinations of patients, basically carotid artery scans, before and after, but we have to ensure there were no compounding variables or instrumental errors or whatever, so we’ll need to reproduce this multiple times.
Did you test for baseline fluctuation of artery health in a normal healthy person with no intervention?
- Yes, we measure markers in a longitudinal fashion – to have a distribution of values throughout the time. So we have a relatively clear picture of how markers vary and see how much they diverge from the mean. That allows us to say whether the change was expected or unexpected. For example I measure my pulse wave velocity every day basically.
How can this group help your work and help further your work?
- One thing everyone can do is to spread awareness about the overlooked hallmark of aging or maybe even a cause of aging – damage to the extracellular matrix. That can definitely speed up the progress in that direction. Otherwise if someone is willing to support us financially or by other means, you are very welcome, we need your help to speed our progress. You can reach us at [email protected].