Dr. Brian Kennedy is a longevity heavyweight: a former President of the Buck Institute, currently a Distinguished Professor in the Department of Biochemistry and Physiology at the National University of Singapore, a co-editor of Aging Cell, and one of the researchers behind a recent study of alpha-ketoglutarate that resulted in winding back chronological age in humans by 7 years on average (with some caveats, as usual). We tapped into Brian’s scientific wisdom on topics ranging from Singaporeans’ remarkable longevity to the reproducibility crisis in biology research.
You live and work in Singapore now, which brings me to my first question: Singapore ranks very high in terms of life expectancy, as well as Japan and Hong Kong. What’s their secret?
First, there’s a healthier diet in Asia than in the West. They eat much less red meat, dairy, calories overall, and they try to get fresh food every day. There are other factors as well. One is that people here still exercise or keep active as part of their daily lives. Particularly in Japan, maybe more so than in Singapore, you will see little older ladies walking to the grocery store and carrying back some groceries every day. People are just more active than they are in the West, particularly after retirement. I, on the other hand, tell people that I have to go for a run every day because I don’t walk anywhere.
So, sustainable routine exercise as part of a daily life, and then there’s also good management of aging. The healthcare systems here are generally better than in the US, for sure, and in many other parts of the world. They take better care of people as they get older, and they catch diseases earlier. Singapore, Japan, Hong Kong have great healthcare systems. So, it’s mostly environmental and societal factors, not particularly genetics.
From what you’re saying, the West and the US in particular have a lot to learn?
Yes. We eat too much, we don’t exercise enough, and we also have a very inconsistent healthcare system.
What do you mean by inconsistent?
The motivation to do healthcare in the US is somewhat perverted. I’m a free market type of person, not a socialist, but there’s just too much emphasis on profit. The way we’ve done things with the insurance industry… a lot of people still don’t have insurance or are under-insured, and the cost of healthcare is so exorbitant that they can’t afford to get healthcare.
As a result, they don’t get primary care. They avoid all the things that aren’t essential, and then they get really sick. At that point, they have stage four cancer or something, and then a lot of money is spent trying to treat their cancer, but we’re starting too late. Also, the medical system is really targeted toward what I call “sickcare”.
Even for people that have means, the system typically doesn’t focus on prevention, and then people get some chronic disease of aging or an infectious disease. Again, a lot of resources are spent after they become sick. We need to refocus our healthcare approach on how do we keep people healthy longer? Of course, people still need to be treated when they get sick, but if we can extend healthspan, it’s a huge economic benefit and a huge benefit to our quality of life as well. I’d rather stay healthy than sick.
I like the term “sickcare” that you use in your talks. Do you think the American system is so focused on sickcare because of bad incentives?
Yes. People get paid for procedures and reimbursements. They don’t necessarily get paid to keep people healthy. I think that’s a big factor. To be fair, all healthcare systems have their problems. None of them are perfect. Singapore, though, has a public health care system. Everybody gets a certain base level of care that’s provided by the government, and then if you want to get a higher level of care, often that doesn’t mean better medical care, it just means private rooms, etc. If you want, you can get it, but everyone has a certain baseline that’s effective. The last time I looked, healthcare expenditure in Singapore was about 8% of the GDP, whereas the US, it’s almost 20% of the GDP and it’s not as effective.
I think, if you go entirely to public health care, that could create problems too, but if you have some balance where a certain level of care is provided for everybody and then there are opportunities for other people to get more if they want to, that’s probably a more efficient system.
I don’t think any healthcare system focuses enough on thinking about a person over the entirety of their life course. We think about people when they’re born, when they’re at a very young age and then we ignore them mostly until they get to 55 or 60 and start getting sick. That time in the middle is where we really should be focusing on prevention.
Aging, of course, is the biggest risk factor for practically everything, including mortality due to COVID-19. We should be taking a life course approach to health, trying to measure how people are aging, what’s happening to them throughout their life and developing interventions that are designed to keep them healthy before they get sick.
Let’s dig a bit deeper into the science of aging. I really liked your idea of the interconnectedness of the hallmarks and pillars of aging. Could you elaborate on this for our readers?
As you know, there are nine hallmarks of aging, according to the Lopez paper. In the paper that I wrote with a bunch of other people, there were seven pillars of aging, and they overlap somewhat.
What made me think about this interconnectedness was that every time we look at an intervention, whether it’s rapamycin, alpha-ketoglutarate, NAD precursors, metformin, or pick your favorite intervention, it doesn’t seem to be hitting just one pillar or hallmark but all of them.
This means there must be a lot of connectivity between these pillars and hallmarks. If you can hit a node in the system that integrates all of them together, then the signaling changes that you get both cell-autonomously and cell-non-autonomously affect all the pillars of aging and all the hallmarks. That’s why you get one drug that can impact aging. We’ve always thought aging was an extremely complex process. When I started in the field, everybody was saying, oh, you’re not going to find single gene mutations or single drugs that affect aging or lifespan.
Even though we already started to find some in worms, and later in yeast and flies, people were saying, oh, those are simple organisms, it’s not going to be true in mammals. Turns out, it is true, there are drugs that extend life span in mice, maybe even humans.
There are genetic mutations that are associated with long lifespan in mice and humans. And so, even though aging is a complex process, there must be nodes or integration points that you can hit that will alter the whole system so that it stays healthy longer. There’s a lot of effort right now, especially in the private sector, to try to develop something that hits every hallmark, and then you can put it all together and have a massive effect on aging.
I doubt it’s going to be that simple. The challenge is that it’s not so easy to figure out what you can combine together to get bigger effects. That’s already something we see in our lab. We’ve combined a lot of interventions in mice, and I can’t predict what’s additive and what’s not. So, I get a little bit nervous when I hear people talking about all kinds of different supplements, because when you start adding things together, it leads to unpredictable outcomes.
Generally, I think, it’s fine if people want to be early adopters and try certain things as long as they’re safe, but when you’re taking three or four different things, especially at high doses, there’s no way to know what’s happening.
Frankly, I can’t really figure out how I should feel about this interconnectedness and complexity. Should this make me feel more optimistic or more pessimistic as to our ability to fight aging?
It’s a good question, but you can also say, did we get all the hallmarks and pillars right? Most probably, they weren’t exhaustive, and we’re now trying to go back and write another review, trying to think about what was missed, and other people have done this as well since those two original papers.
What does that network structure look like? It’s easy to say “network”. It’s like systems biology: everybody talks about it, but nobody can define it. We’re reaching that point, and that’s also why the AI stuff is starting to lead the way because AI doesn’t really care about mechanisms of action, but it can take complex data sets and distill something out of it that’s effective.
Does this black box approach scare you a bit?
I don’t know if it scares me. I mean, we’re doing it, and I think it’s the right approach, but I also find it a little bit depressing. I miss those days in yeast aging research where you could come up with an elegant experiment to test a specific hypothesis and you would find that the sirtuin protein complex re-localizes to the rDNA and that’s what affects aging.
There was a lot of satisfaction in this, and those results, even if they were a bit naive, retrospectively, you felt like you were learning something that was tangible. Now, we get AI data and it’s telling us something, and when we test it, it often validates, but we still can’t explain why. I can’t tell you why any intervention works, and that’s a bit frustrating.
This interconnectedness is probably the reason why many geroprotective drugs work more or less along the same lines, but AKG seems to be your molecule of choice. What makes it special?
I’m happy to tell you about alpha-ketoglutarate, but before I do that, I have to say that in Singapore, we’re testing many kinds of interventions and starting to combine them together in animal models. Working with Andrea Meyer, we’re starting to do human intervention studies as well, and we’re trying to be agnostic there. We will test AKG, of course, but there’s others too. We want to test five or ten things side by side and start comparing what works on which biomarkers and in which people.
Alpha-ketoglutarate is a central metabolite in the TCA cycle [Krebs cycle]. It’s involved in respiration, in amino acid metabolism and anabolism. It’s a center point of many different metabolic pathways. Its levels go down with aging. As was first shown in a paper in Nature, when you give it to worms, it extends their lifespan quite dramatically. Working with Gordon Lithgow, we were able to see that as well, and so, we started testing it. We saw that it really compresses morbidity in mice. It extends lifespan by a little bit, but it reduces frailty by up to 50%. That’s an exciting molecule in my book.
We recently published a human paper on it, but it wasn’t a controlled clinical study. Let me give you the caveats first. There’s no placebo control. It’s people buying the product and doing a baseline DNA methylation test and a follow-up around seven months later. It’s a relatively simple methylation test, but at the time it was the only one we could use, because all you need is a piece of paper. You put saliva on the paper and send it back in. It was a lot simpler than a blood draw or anything like that. What we found is that you could reduce biological age by seven or eight years according to that assay, which I think is exciting.
The company that sponsors the Rejuvant product is called PDL Health [Ponce de Leon Health]. They’re doing a placebo-controlled clinical study at Indiana University which is nearing completion, and we should have some data from that study very shortly. Then, we’re planning to do a sustained release AKG-only study in Singapore. Rejuvant, in addition to AKG, contains low-dose vitamin A for men and low-dose vitamin D for women. That’s what the published paper is about, but we want to see specifically what AKG is doing, because we think we can glean more mechanistic insights from that.
We need to be a little careful with these clinical studies that are still in their early days. We’re just beginning to use biomarkers as end points, and there’s probably a placebo effect too. I suspect that if you’re buying a product and paying for it, you probably become biologically younger by a couple of years.
It will be interesting to see how things operate against the controls and look at immune parameters, metabolic parameters, etc. We’re very excited about that data.
We at lifespan.io covered this study, and the results are exciting indeed. Now, how do you see a possible timeline for anti-aging interventions – say, small molecules versus the more radical approaches such as gene therapy?
Lifestyle changes are here; you can do them now. We don’t know everything, and we probably don’t know as much as we think we know, and when it comes to diet and longevity, there’s still a lot of debate. Everybody thinks exercise is good, but which kinds of exercise for which people is still kind of a question. Natural products are almost here. If you’re an early adopter, you can buy these things now – not just Rejuvant, but NAD precursors, and a whole range of other things that are tested to varying extents.
Repurposing drugs is really starting to be tested. We certainly want to look at a rapalog in our studies. I’m sure you know about the TAME trial with metformin. Those are not necessarily that far from the market. The only concern I have is that, like with the TAME trial, the FDA said, well, we’re excited to see if this is going to affect healthspan, but that doesn’t mean they’re going to move super quickly, like with the coronavirus. Even if you get positive results, it’s still going to take time. But we’re moving toward that.
Most companies developing new chemical entities right now are trying to target them towards specific diseases or aspects of aging. I guess that’s a necessity. I’m not sure it’s my favorite choice, but I understand why they’re doing it, so the process is going to be longer. Then, you have things like stem cell therapy, gene therapy, young blood, and a whole range of other types of interventions.
What is your opinion on senolytics?
Senolytics, I think, are intriguing, though there are some challenges. One type of senolytics are drugs that have been repurposed, and some of them have toxicity issues. The other type is natural products, safer drugs that people like Jim Kirkland are using and they’re targeting them towards specific diseases of aging.
I certainly think that cell senescence is a component of the aging process, but it’s still unclear to me how big a component. Some people want to put it at the center of the pillars or hallmarks of aging. I’m not quite ready to do that, but certainly, if we can effectively remove senescent cells, there will be many benefits to that. So, it’s an exciting pathway to look at.
What is your take on the effect of compressed morbidity? Is this the best we can hope for?
I think that in many cases, it’s not even there. I don’t see that much of it in the invertebrate models. There, if you extend median life span, you extend maximum life span, and if you can measure healthspan parameters, they’re extended as well, for the most part. We’re starting to see that with some mouse models.
I also think that’s what you get from exercise. I don’t know whether this will affect maximum lifespan that much, but exercise gives you a much better chance of getting closer to it, because by exercising, you’re avoiding some of the behaviors that lead to earlier stage breakdown of aging pathways and disease onset. In other words, there are many different things that can drive disease onset, and you can accelerate that based on your behaviors.
Even with mice, we don’t give them the healthiest diets all the time. We don’t let them exercise that much. We keep them under stress conditions. Maybe some of the interventions that we think are anti-aging are actually helping with the lifestyle of the animal rather than affecting the maximum lifespan.
I think, for now, that’s fine. If we can extend healthspan 5-10 years, even with no effect on maximum lifespan, that’s still a major victory, both in terms of quality of life and economic benefit. The things that can extend maximum lifespan are more interesting to think about, and I would argue that it’s certainly possible. We can do it in all the other animal models, but right now, the evidence suggests that maximum lifespans are not going up that much in humans, even though median lifespan is.
It’s unclear how easy it’s going to be to affect maximum lifespan. I think, ultimately, it’s possible, but whether the interventions we’re looking at now are evening out the bad things that are happening or whether they’re really slowing down the intrinsic aging process, which would lead to extending maximum lifespan, is still an open question.
Recently, you became a founding member of LBA – Longevity Biotechnology Association. We covered their launch extensively, and we are very excited about this initiative.
I’d like to hear your take on this – why do we need it, how is it different from some other initiatives in the longevity field, has something been happening since the launch?
I think there’s now a little bit of trying to figure out how best to structure the association so that it can have the most impact. There’s a growing private sector in longevity right now. It’s a good thing, but it’s kind of a Wild West. You got money flowing in all directions, people with all kinds of different ideas on both diagnostics of age and interventions, clinics are opening up. We need an association that can bring all that together and get the government involved so that they understand this revolution that’s happening. Certainly, the motivations of the association are great, and I’m just trying to help in any way I can, but I also think it’s going to take some time to figure out how to do this in the most effective way.
As a co-editor of Aging Cell, what do you think about what appears to be sort of a reproducibility crisis in biology? How severe is it, and do you think that maybe the whole publishing system needs updating, say, in terms of incentives to scientists?
Aging Cell has no problems, it’s always the other journals. I’m joking, of course. I do think there’s a problem with publishing right now. There are several things. For one, the most important papers are not necessarily getting published in the top journals because there’s too much emphasis on a person’s reputation rather than the science they’re putting out there. There’s also certainly a lot of motivation to get positive results, and I think that can bias how a person does experiments and interprets the data. I think the vast majority of scientists are trying to learn the truth, but it’s easy to convince yourself of something without all the controls in place.
I also think that the publication process is just taking too long. One change I would suggest is that we shouldn’t have multiple rounds of review of a paper. When somebody submits a paper, the reviewers can say whatever they want, suggest experiments, change things, edit, whatever. And then the author responds to that. At that point, I think the reviewer shouldn’t be suggesting more experiments. They should just say if it’s good enough or not good enough.
Particularly if you want a paper in one of the top journals, it can now take a year and a half to get it published. This is not serving anybody. We need to be disseminating scientific knowledge quickly. I’m perfectly happy with a paper that says, here’s the data, this is our favorite interpretation of it, but here are the weaknesses of the paper as well, and these are other potential interpretation. I think that’s fine. The problem is, if you’re not willing to put in those weaknesses, it can be misleading.
I also think we have to realize that biology is not a system where we can reduce the variables sufficiently to guarantee that we’re all doing the same experiment. We’ve seen this with the ITP program, which is an intervention testing program studying longevity interventions in mice through the National Institute of Aging. There are three different sites, all very good scientific sites, all do very well-controlled experiments, but they’ve had cases where rapamycin extends lifespan here but not there, and they traced it all the way back to the food source or some other issue.
Reproducibility is not always easy, even if you have great scientists involved. There are just a lot of variables, especially in a long-term study in a mammal. Not everybody’s going to get the same result, and when they don’t, it’s not the time to start having a fight. What we need to be doing is sorting out what the reasons are, why there was something that works here and doesn’t work there.
When you have disagreeing data, the answer is usually somewhere in the middle. I don’t think we should shy away from data that’s contradictory. It’s an opportunity to sort out the details and understand more.
Do you think that the system may be guiding scientists towards “torturing the data until it confesses” or not publishing negative results? Is it a genuine problem?
Not publishing negative results is a big problem, and we always try to do that, though we have negative results that we haven’t published too, and there are a couple of reasons for that. One is you get limited value for publishing negative data. I really liked when PLOS One started several years ago because it was a site where you could publish negative data. We published a bunch of papers there, and other people have too.
The other side of it… let’s say you’re testing 10 different interventions in mice. You’re often not doing it to the same extent than if you were only testing one intervention. You may have less mice, less doses of the small molecule, et cetera. Then, you don’t get it resolved, and the data is not good enough to publish. That was your pilot experiment, and if you want to make it good enough to publish, you have to go back and do it on a much larger scale.
Are you really going to put a large investment into something that you think will give you a negative result? No, you choose the most promising intervention, and this is how the negative data doesn’t get published. Some of it is intrinsic to the process. I do think we should do a better job of getting negative data out there, but again, I don’t recommend using it to say that this or that person is definitely wrong. The important point is: OK, we tried this experiment under these conditions, and it didn’t work. Somebody else tried to do this, and it worked, so let’s try to find out what’s going on.
I noticed that you did your BA in mathematics, and I’d say that this kind of shows. Do you think this clear, logical thinking, this mathematical approach is what biology is still lacking, like the famous 2002 article “Can a biologist fix a radio” postulates?
I do think it’s lacking and it’s lacking in me too – for instance, I’m not an AI expert, I have to bring in people that know what they’re doing. I don’t write code. I had a biochemistry degree too, so the math to me was something that I could get relatively easily, with a few more classes based on the program I was in. I just took classes that were fun, like the set theory, abstract algebra, a lot of logic courses. And I think it’s helped me a lot, though I probably should have taken some statistics.
Yes, a lot of biologists are not well-versed in math, and now we’re entering a new world where you need those skills. It’s a transition right now. Maybe I’m part of the way to that transition. You know, the people making the discoveries in biomedicine 10 years from now may look very different than the people doing it today.
I understand you’re an avid runner. Does this hobby have something to do with life extension? What else do you do to keep aging at bay?
I also try to eat reasonably healthy and to have one big meal a day, and, as you said, I’m a runner. I think managing stress is really important. We all live stressful lives, but it’s what you do with the stress that is important and how it affects your body. Of course, sleep quality is important, though that’s hard to manage. I always told people that I’m not going to try any interventions until I’m 50, and then I became 50, so that excuse didn’t work anymore, and now I’m now trying Rejuvant because I’m involved with the company.
Do you have any anecdotal evidence to share?
I think it helped my exercise performance, and actually, there’s a long list of people adopting AKG that are either endurance athletes or resistance trainers. They think it’s because it drives amino acid production. I’m not sure that’s the mechanism, but I definitely felt like I could do better endurance exercise when I was taking AKG.
By the way, isn’t it interesting that AKG seems to help muscle building, while for metformin we have some data suggesting that it may have some adverse effects on this?
I think that they probably work through different mechanisms, but we need to study metformin deeper to really understand its role in skeletal muscle. Some studies suggest that people that are exercising don’t gain the muscle mass if they’re taking metformin. Nir Barzilai has a lot of reasons for why that might be the case. We need to understand that better.
Certainly, building muscle mass is very good for aging. In fact, if you look at people who are overweight, as long as they have relatively high lean muscle mass, they’re pretty protected.
Which directions excite you in the longevity field today?
I think the big excitement over the last several years has been the development of biomarkers of aging. I don’t think they’re perfect, but things like methylation clocks are really allowing us to get an estimate of how fast a person is aging – something that was never feasible, especially in younger people.
This was the excitement, and now it’s about applying it, testing interventions with people and trying to validate that they really affect biological age. People have been trying to slow aging for millennia, and we’re at a point now where we’re going to be able to validate that it’s possible with interventions that directly target aging. I’m very excited to see that happening in the next few years.
We would like to ask you a small favor. We are a non-profit foundation, and unlike some other organizations, we have no shareholders and no products to sell you. All our news and educational content is free for everyone to read, but it does mean that we rely on the help of people like you. Every contribution, no matter if it’s big or small, supports independent journalism and sustains our future.
Take a close look at these so-called lifespan curves, each of which shows a surviving fraction of a cohort as a function of time, always going from 100% to zero. The precise shape of such a curve and slight differences between multiple curves tell us a lot about an intervention and its interaction with an organism.
