Bayesian Investor Blog

Book review: Lifespan: Why We Age – and Why We Don’t Have To, by David A. Sinclair.

A decade ago, the belief that aging could be cured was just barely starting to get attention from mainstream science, and the main arguments for a cure came from people with somewhat marginal formal credentials.

Now we have a book by an author who’s a co-chief editor of the scientific journal Aging. He’s the cofounder of 14 biotech companies (i.e. probably more than he’s had enough time to work for full time, so I’m guessing some companies are listing him as a cofounder more for prestige than for full-time work). He’s even respected enough by some supplement companies that they use his name, even after he sends them cease and desist letters.

I’m glad that Sinclair published a book that says aging can be cured, since there’s still a shortage of eminent scientists who are willing to take that position.

I found this book to be frustrating and hard to read. Not because it’s technical (although parts are technical enough to add some difficulty). It’s mainly because I kept feeling that something was missing. I couldn’t figure out how much of that was because I wasn’t reading carefully enough, how much I should expect Sinclair to fill in the missing pieces later in the book, and how much was due to Sinclair having avoided important pieces.

Sinclair struggles to walk a fine line between mainstream and eccentric, and doesn’t quite succeed. I think he could have written a book that I enjoyed if he were 20% more eccentric, or an influential, but dry, book by being 20% less eccentric.

I kept looking for a grand vision, yet only found bits and pieces of a vision. And that’s not because Sinclair focuses on near-term practical research. He sometimes does that, but he also covers a wide-ranging landscape that would be appropriate if he were presenting a grand vision.

Recent Progress

My favorite part of the book dealt with anti-aging strategies that are available now.

I’ve been taking nicotinamide riboside (NR) for a while, without having much confidence that it’s helping. Sinclair provides some interesting anecdotes about the similar NAD+ booster NMN, which increased my confidence in NAD+ boosters.

Sinclair also discusses life-extending benefits of exercise, metformin, saunas, cryotherapy, and calorie restriction. The evidence for exercise seems pretty solid. The rest are more speculative.

Also, he mentions some really interesting results from Yamanaka factors, but the evidence is harder to evaluate, and the treatment doesn’t sound close to being ready for general purpose use in humans.

Should I Believe?

What does that progress add up to in terms of increased lifespan? Sinclair was not helpful here.

Sinclair’s best evidence comes from research in mice. How optimistic should I be about translating successful mouse treatments to humans?

For many age-related diseases, the track record is mixed. It’s failed disastrously for Alzheimer’s, but it it’s quite plausible that that’s because Mouzheimer’s just isn’t the same disease as Alzheimer’s.

We have stronger evidence for aging being similar across species, but mice age faster than humans, and that seems like a sign that humans have evolved solutions to the some of the age-related problems that kill mice. So I expect that some mouse aging treatments will work well on humans, but that the treatments which are easiest to find for mice will do little for humans.

I tried to estimate how many treatments have been shown to slow aging in mice, but don’t work in humans. The evidence I found was mostly weak and inconclusive.

Calorie restriction (CR) is an example of a treatment that has good results in mice, and has something vaguely resembling good evidence about how well that applies to humans. Alas, CR slows aging across a wide variety of animals, but shows much weaker signs of benefits in monkeys.

Sinclair cites one study in monkeys which showed decent benefits, but neglects to mention the study which showed no benefit. I.e. the evidence is much messier than Sinclair leads us to think. Sinclair has tried CR, and has mostly given up because it’s too hard. He expresses a somewhat more favorable view of intermittent fasting, as do I.

There are a number of compounds which have been known for a while to extend lifespan in mice, but it’s hard to find evidence about how well they work in humans.

I looked in this database for the treatment that gave the best increase in average mouse lifespan (pineal gland extract). Then I looked for human trials, and found some comments by Sarah Constantin with promising results in a small human trial. Yet 17 years after that trial was published, there seems to be little interest in pineal extract. That seems to be about par for aging research.

A study published in 1979 showed that BHT increased mouse lifespans. Subsequent research shows … I don’t know, maybe there’s no interest because it’s unpatented?

Information Theory of Aging

The highly simplified version of Sinclair’s theory is that aging is just a function of epigenetics.

That seems at least partly right – there have been plenty of hints that epigenetics play an important role in aging. Yet other researchers are either cautious about how much of aging is explained by epigenetics, or are a bit eccentric and their arguments aren’t quite strong enough to command mainstream acceptance.

Sinclair isn’t cautious here. Instead, he portrays his theory as a big insight:

Scientific observations that had previously made no sense to me were falling perfectly into a larger picture.

Alas, that does not at all describe my reaction. In fact, I read that chapter twice, and forgot that claim both times, because I didn’t see what observations fell into place. It was only the act of rereading my notes, in order to organize this review, that I was able to flag this as an important incongruity.

Here’s Sinclair’s attempt to summarize the more complete version of his theory:

Broken DNA causes genome instability, I wrote, which distracts the Sir2 protein, which changes the epigenome, causing the cells to lose their identity and become sterile while they fixed the damage.

This describes a real phenomenon, which he demonstrated in yeast.

But how does he get from showing a phenomenon associated with aging to “a singular reason why we age”, which he calls the Information Theory of Aging? I read carefully, and only found vague hints.

What does Sinclair’s theory predict (or explain) that other theories don’t? Or alternatively, what’s wrong with the competing theories? Sinclair probably thinks he’s answered these questions, but I’m confused as to which of his points are supposed to qualify as answers.

I’ll guess there’s a 50% chance that Sinclair is right about epigenetics being the main cause of aging. But Sinclair did not increase my confidence in that.

Why Did Aging Evolve

Sinclair says we need (but haven’t had) “a unified explanation for why we age, not just at the evolutionary level but at the fundamental level.”

Yet he says little about the elephant in room: why doesn’t antagonistic pleiotropy (AP) qualify? It seems to be the most widely accepted theory at the evolutionary level, and it seems to explain why aging doesn’t look much like a unified phenomenon at other levels (i.e. why there appear to be 9 hallmarks of aging, or maybe seven, if you ask Aubrey de Grey). Indeed, I’m unclear why Sinclair implies there’s a “fundamental level” that’s different from evolution.

If I squint really hard, I can see evidence in Sinclair’s book that raises doubts about AP (e.g. organisms that don’t age). But it seems almost as plausible that Sinclair imagines his theory is compatible with AP.

If aging is easy to fix as Sinclair implies, why didn’t evolution fix it? I’m a little unclear on how easy he thinks it is, but he often seems to suggest that we can undo a fair amount of aging with a few molecules such as NAD+.

One possibility is that Sinclair considers evolutionary theory to be unimportant for his work. It’s certainly possible to do good research based on more immediate causes of aging, while having a faulty understanding of the original causes. I expect that a good theory will increase a researcher’s chance of finding the most cost-effective approach to aging, but I also expect success to be influenced by intuitions at a less theoretical level, and by luck.

Most explanations of why aging evolved don’t provide clear guidance for how to cure aging. Rather, they typically provide vague hints about which approaches are likely to be easiest. My main concern is that most evolutionary explanations suggest that curing aging is harder than Sinclair implies, which weakly suggests that Sinclair’s approach is unlikely to work.

The other possibility is that Sinclair believes aging has been actively selected for by evolution, but he has kept quiet about that because it’s controversial, and somewhat associated with some rather questionable advocates. I don’t see any clear evidence that this is what he’s doing. I’d have to really stretch to see any hints or dog whistles. I find it quite plausible that Sinclair ought to be following this approach, but it doesn’t mix well with his claim to have articulated a neat fundamental theory.

[Note: after writing a first draft of this post, I noticed that I’d previously read a similar complaint about Sinclair’s theory by Josh Mitteldorf, who has influenced my thinking on this subject. I completely forgot that Mitteldorf had been referring to Sinclair.]

Miscellaneous Comments

There’s enough uncertainty about which theory is right that theoretical insights only provide weak guidance. I wouldn’t want researchers to unite behind a single theory of aging without better arguments for which theory is correct.

A few days before Lifespan was published, results from the TRIIM trial showed evidence that epigenetic aging was reversed in humans by something like 2.5 years.

That result looks at least as important as the results that Sinclair describes. It increases my hope that aging is caused by an easily reprogrammable epigenetic clock. However, it also slightly increases my concern that epigenetic clocks are yet another set of biomarkers that represent symptoms rather than causes of aging.

If increases in NAD+ help reverse aging, then that should help COVID-19 patients. There is in fact some evidence of that, and a clinical trial to test it. [Disclosure: I own stock in Chromadex (which sells NR), as does an author of the paper I linked to.]

Toward the end of the book, Sinclair tackles a fairly broad set of topics. His mostly unremarkable opinions there are about what I’d expect from a Harvard professor.

His enthusiasm for quantified-self style health tracking leads him to mention the risks of pandemics, with hints of how we might avoid them, and warnings that we weren’t well prepared at the time the book was published (2019). He mentions, without useful comments, a quarantine bill of rights, which looks in hindsight like it might have been one of the many obstacles that prevented the US from deterring the initial COVID-19 spread.

Sinclair repeats a common claim of demographers that the world population will plateau within a century, and then decline. As far as I can tell, that overlooks several populations that are growing at seemingly stable exponential rates. E.g. the Amish population, whose trajectory suggests they’ll exceed a billion in about two centuries. I’m not saying we should be concerned now. But regardless of whether we cure aging, it sure looks like we should expect that some significant changes will be needed.

Conclusion

Sinclair’s research seems to be among the most valuable medical research being done today. I enjoyed the parts of the book where he focused on what his research has accomplished so far. I’m cautiously optimistic that his work will add years of healthy life for many people.

But the book doesn’t live up to the lofty expectations that it sets.