About Josh Mitteldorf

Josh Mitteldorf studies evolutionary theory of aging using computer simulations. The surprising fact that our bodies are genetically programmed to age and to die offers an enormous opportunity for medical intervention. It may be that therapies to slow the progress of aging need not repair or regenerate anything, but only need to interfere with an existing program of self-destruction. Mitteldorf has taught a weekly yoga class for thirty years. He is an advocate for vigorous self care, including exercise, meditation and caloric restriction. After earning a PhD in astrophysicist, Mitteldorf moved to evolutionary biology as a primary field in 1996. He has taught at Harvard, Berkeley, Bryn Mawr, LaSalle and Temple University. He is presently affiliated with MIT as a visiting scholar. In private life, Mitteldorf is an advocate for election integrity as well as public health. He is an avid amateur musician, playing piano in chamber groups, French horn in community orchestras. His two daughters are among the first children adopted from China in the mid-1980s. Much to the surprise of evolutionary biologists, genetic experiments indicate that aging has been selected as an adaptation for its own sake. This poses a conundrum: the impact of aging on individual fitness is wholly negative, so aging must be regarded as a kind of evolutionary altruism. Unlike other forms of evolutionary altruism, aging offers benefits to the community that are weak, and not well focussed on near kin of the altruist. This makes the mechanism challenging to understand and to model. more at http://mathforum.org/~josh

Prolonging Life with Fecal Transplants

A game-changing result this week from the laboratory of Dario Valenzano (Max Planck Inst).  A single treatment of antibiotics in middle-aged fish followed by transplant of gut bacteria from young fish resulted in extension of mean lifespan by 41%, max lifespan by 30%.  Treated fish remained active at ages where untreated fish were slowing down.  I say “game-changing” because up until now the gut microbiome has been a fascinating but peripheral discipline in the study of health.  This single study raises the possibility that understanding the microbiome as a system could be a powerful new avenue toward health and longevity.  [Preprint of Journal Article]  [News Article in Nature]

There have been intriguing hints that the ecosystem of bacteria in the intestine have major effects on mood, on wellbeing and on disease.  But there has been no way to get a handle on the causal variables involved.  The mix of gut bacteria varies widely from person to person, depending on diet, genetics, social contacts and environment.  Thousands of species of commensal bacteria form a constantly-shifting ecosystem.  

Who is working for whom?  Do we think of the microbiome as a parasitic colony that manipulates the host’s biochemistry for its own ends?  or as as managed by the host (that’s your body)?

I’ve seen articles about the former proposition, but I’m skeptical because I can’t imagine an evolutionary mechanism.  It seems that these thousands of bacterial species don’t stay together from one individual to another.  They are not readily transmitted (in nature) as a group, except perhaps from mother to infant.  And if there is natural selection on the microbiome ecosystem as a whole, it must be for something that maximmizes opportunities for transmission.  It’s easier to imagine individual species, specialized to living in the human gut, that learn to gain an advantage over other species by manipulating the human metabolism in ways that favor that particular species over its rivals.

The latter possibility — that our immune systems have a handle on who may live and who may die in our intestines — is both easier to conceptualize and more promising.  It raises the possibility that part of the way the body regulates its own metabolism is indirectly, via bacterial secretions.  I have advocated the position that aging, like development, unfolds on under central regulation.  The medium for instructing the body in age-appropriate behavior is likely to be signal molecules in the blood.  Could it be that some of those signal molecules originate not in our brains or our endocrine systems, but in the bacterial reservoirs of our guts?

Background: Fish

Eleven years ago, Valenzano introduced African Killifish (Nothobranchius furzeri) as a new lab model for study of aging.  Evolved for a life cycle in short-lived African ponds that dry up after a brief rainy season, they have one of the shortest life cycles of any vertebrate.  As a grad student, Valenzano demonstrated substantial lifespan increases adding resveratrol to the fish’s water.

Loss of diversity is one of the ways that the gut microbiome is known to change with age, both in humans and in fish.


Background: Humans

There has been a great deal of study and writing over the last decade, but so far only one clinical intervention, plus this guidance for the general public: a high fiber diet encourages beneficial bacteria.

Four years ago, Michael Pollan wrote about microbiomes for the NYTimes magazine. Mark Lyte has connected the microbiome to psychology: depression, anxiety, maybe autism (popular article in the NYTimes two years ago).  Turns out that gut bacteria produce some powerful hormonal signals that go right into our bloodstreams and are decoded by our brains.  

Gut microbiomes vary widely from one individual to the next, but, strikingly, different sets of bacteria are able to perform similar services.  The bacterial gene profiles in healthy individuals don’t vary nearly so much as the specific component bacterial do [ref].  

In hospitals and in people treated with antibiotics, a new disease has arisen in recent years characterized by intestinal infection with a bacterium called Clostridium difficile.  Symptoms include chronic diarrhea, stomach cramps ad nausea.  The most effective treatment developed to date (90% cure) is a transplant of fecal matter from a healthy individual.  This can be accomplished with enema, but there is some indication that it is more effective if the fecal matter is introduced from the other end, into the stomach, and this has inspired freezing and encapsulation technologies to get around the disgust factor.

Beyond this one clinical application, there is speculation about treating other intestinal disorders with fecal transplants, including ulcerative colitis, inflammatory bowel, and Crohn’s disease, extending to Type 2 diabetes, obesity, and even flatulence.  Having the right mix of microbes is important for triglycerides, glucose regulation, and the insulin metabolism [ref].  There have been multiple studies in rodents and one (successful!) study in humans of fecal transplant to treat diabetes.  

Many of the diseases of old age, (arthritis in particular), are connected to autoimmunity.  Intriguing, if speculative, work has been done connecting gut microbiomes to autoimmunity [review].  Maybe the ubiquity of antibiotics in the developed world has led to a hyper-sensitivity, connected to increases in asthma, lupus, type 1 diabetes, possibly autism.  Maybe the mechanism by which this has hit us is through our gut microbiomes.

Nearly two decades ago, scientists put forth a concept called the ‘hygiene hypothesis’. According to this hypothesis, an improvement in personal hygiene as observed in the developed countries has led to an increase in the risk of allergic and autoimmune disease [ref]. Increase in incidences of various inflammatory and autoimmune diseases like inflammatory bowel disease (IBD), asthma, type 1 diabetes (T1D), and rheumatoid arthritis in the developed countries support this concept.

It is suggested that gut microbiomes are connected to immune function more generally.  Both in mice and in humans, resistance to sinus and bronchial, including pneumonia, has been demonstrated with the right kind of gut microbiota.

Gut microbiomes in supercentennarians have been analyzed, and differences from average people have been distinguished as specific bacterial familes that seem to be associated with longevity [ref].


Summary of the Killifish Results

Turquoise Killifish normally live 16 weeks (black line).  At 9½  weeks, fish were treated with antibiotics to kill their gut microbiota.  Those that received no transplant at all lived a little longer (purple line), and those that received gut biota from same-age fish (9 weeks) lived insignificantly longer (red line).  But those that received transplants from younger fish lived 22 weeks (green).

Fish that received young transplants were more active and showed more exploratory behavior later in life.  The authors performed proteome analysis on the microbiome as a whole, and found gene expression that suggested a stronger resistance to infections in the young-transplanted fish.  

Young fish transplanted with the microbiota of old fish quickly recovered their youthful biodiversity and their lifespans were unaffected.

Authors note that

  • Microbiomes of killifish are comparable in complexity to mammals, including humans.
  • Although short-lived, killifish suffer many of the same declines as humans in old age, including neurodegeneration, muscle loss, and increased risk of cancer, heart disease, and diabetes.
  • The four most abundant phyla of gut bacteria in the killifish are the same four that predominate in human intestines.
  • Like humans, fish lose diversity of their gut microbiomes with age.  The bacteria lost with age in fish and in humans include those that digest complex carbohydrates.
  • Fish in the lab have comparable lifetimes and comparable gut microbiomes to fish in the wild.
  • Microbiomes transplanted at 9 weeks persisted, and were mostly intact at the end of the fishes’ lives 10-15 weeks later.

The authors were able to characterize explicitly the network of bacteria associated with youth (and also with enhanced longevity), naming the specific species that seemed most important.  Some of the most important species were able to digest carbohydrates and ferment them into short-chained fatty acids, which are known to be anti-inflammatory.

In their “discussion” section, the authors suggest that the gut microbiome may be managed by the host (fish)’s immune system, and that management becomes lax in old age, allowing some commensal bacteria to disappear and more pathogenic types to predominate.  They go on to speculate that perhaps there is a feedback loop between the immune system and the gut microbiome that is activated with age: poorer management of the gut ecosystem by the host immune system results in takeover by bacteria that further weaken the host immune system, leading to a vicious circle.


Remember that life extension percentages in short-lived species are always diluted when applied to long-lived species.  Sometimes they disappear altogether.  Resveratrol extends life of killifish by 60%, but failed to extend lifespan in most mice.

The microbiome transfer in killifish was done at 9 weeks of age, and it lasted the rest of their lives, which was another 8-15 weeks.  People live much longer, and the microbiome transplants would probably have to be repeated and maintained to have an effect.


Just in the last decade, the importance of the microbiome for many aspects of health has been uncovered.  But the microbial ecosystem has been considered too diverse, too irregular, too complex for study with the reductionist paradigms that Western science is so good at.  Transplanting entire microbiomes has proved to be quite feasible, however, if not to everyone’s taste.

If these results hold up (it looks to me like a very careful experiment, and Valenzano has an impeccable reputation), there is now strong motivation for studying microbiome transplants en masse, and this will certainly be accompanied by proteomic analysis.  It’s hard for me to imagine that life extension in humans will prove to be so simple as in killifish, but I wouldn’t be surprised if a host of benefits appear from youthifying our intestinal flora.

The intriguing possibility is that in addition to metabolic self-regulation by the rich network of hormones, RNAs and signal molecules, the body is also managing its metabolism by managing the bacterial mix in the intestine (and the chemicals they produce, many of which are bio-active).  A more disturbing possibility is that the gut’s microbial ecosystem manipulates the body for its own benefit; but I’d bet against this because it seems implausible from an evolutionary perspective.

Is fasting senolytic?

I am finishing a four-day fast today.  In the hope of synergizing senolytic modalities, I took 2.5 g of quercetin last night and another 2.5 g this morning.  I don’t take quercetin regularly, because studies in mice show that daily administration doesn’t lengthen lifespan, and may shorten it.  But for the present, quercetin is the most readily-available senolytic agent we have.  

It’s my speculation that fasting might prime senescent cells for elimination, in the same way that fasting has been shown to prime cancer cells for elimination by chemotherapy or radiation.  Valter Longo has been at the center of the latter research, and he tells me that he is testing the senolytic hypothesis now, with no results yet.

Quercetin is a flavonol found in many vegetables and fruits, especially capers, radishes, cilantro and onions.

Quercetin by itself has been found to be only a weak senolytic in mice, and it has not yet been tested in humans.  It is somewhat better when combined with dasatinib, but dasatinib is highly toxic and not something I would experiment with based on current knowledge.

People have asked about my fasting discipline.  I answer that different people have vastly different experiences, and you won’t know until you try for yourself.  I fast regularly one day a week, and when I take on a longer fast there is often a hump to get over on the second day, sometimes headaches and malaise.  I use enema to clear the colon on the second day, and often this seems to help.  I take caffeine daily while fasting, which is more than in my regular life.

The chief complaint I have after the second day is that I am too mellow, content to lie in bed.  No restlessness during sitting meditation.  Time slips by, and I feel no urgency about accomplishing anything.  I read things I might not otherwise find time for.  I go for long walks and entertain wide-ranging thoughts.

Senolytics against Aging: Snapshot of a Fast-Moving Field

Aging at the cellular level is called “cell senescence”, and it contributes profoundly to whole-body aging.  The most promising near-term prospects for a leap in human life expectancy come from drugs that eliminate senescent cells.  Programs in universities and pharmaceutical labs around the world are racing to develop “senolytic” drugs, defined as agents that can kill senescent cells with minimal harm to normal cells.

Apoptosis is cell suicide, and (from the perspective of the full organism) it’s the best thing that can happen to senescent cells.  The authors of this newest Dutch study ask how it is that senescent cells escape apoptosis.  

FOXO is a protein that controls gene expression, a master transcription factor associated with aging and development.  (It is the homolog in mammals of the pivotal life extension protein first identified in worms as DAF16 in the 1990s.)  FOXO4 activiation in a cell can block apoptosis.  P53 is the most common trigger of apoptosis, the first protein biochemists usually think of in connection with apoptosis.  P53 has multiple functions in the cell nucleus, but as a trigger for apoptosis, it works through the mitochondria.  FOXO4 binds to p53 and blocks its induction of apoptosis.

The treatment studied in this paper is an artificially modified FOXO4, a dummy that binds to p53 in place of regular FOXO4, but doesn’t block senescence.  It has been named FOXO4-DRI, and it works by crowding out the native FOXO4.

The authors note, with caution, that mice with no FOXO4 at all appear normal; but apoptosis is an important cell function throughout the lifespan.  A cell must have “good judgment” about when to eliminate itself, and that works in both directions; in older animals and people, we not only see failure of apoptosis to eliminate senescent cells, but we also see healthy muscle and nerve cells undergoing apoptosis prematurely, and we lose muscle and brain mass as a result.  Other functions of FOXO4 include DNA repair, and mice that lack FOXO4 are subject to a high burden of DNA damage.

By analogy with chemotherapy for cancer, the value of a senolytic treatment is measured by its ability to kill senescent cells without doing harm to normal cells. The index called SI50 (SI for “selectivity index – 50%”) is defined by analogy to LD50 = the “lethal dose” of a toxin, the dose at which half of all cells die.  SI50 is defined as the ratio of LD50’s for normal and senescent cells.  It is the concentration of the agent at which half the normal cells die, divided by the concentration at which half the senescent cells die.  Authors of the current paper report SI50 about 12.  My guess is that 12 is an encouraging beginning, but it is not high enough to support a useful therapy.  After a standard dose is injected in humans, the cellular concentrations vary from person to person and from tissue to tissue. 

The encouraging fact is that, at the optimal dose, more than 80% of the senescent cells have succumbed to apoptosis, while the number of eliminated normal cells is still below detection:

In other words, the vertical distance between the black and red curves is encouraging, but the horizontal distance is cause for concern.  Senolytic agent studied previously, including dasatinib, quercetin and ATTAC, did not include measurements of SI50 that we might use for comparison.


How does FOXO4-DRI perform in live mice?

Authors of this study were excited in a rush to publish.  They used a fast-aging strain of mice, and even for these, they did not wait to see survival curves.  The indicatators of rejuvenation that they do report look positive:  increased activity levels, regrowth of lost fur, and improvement of kidney function lost with age.


Comparison to Last Year’s Senolytic

Here’s how authors of the current study characterize FOXO-DRI compared to two previously reported senolytic agents:

Two classes of anti-senescence compounds have been reported so far: Quercetin/Dasatinib, either alone or in combination [ref], and the pan-BCL inhibitors ABT-263/737 [ref, ref]. Quercetin and Dasatinib have been reported to be non-specific. We found no selectivity toward senescent IMR90 and therefore this cocktail was not explored further. ABT-263 and ABT-737 target the BCL-2/W/ XL family of anti-apoptotic guardians. Indeed, ABT-737 showed selectivity for senescent IMR90. However, already at low doses, it appeared to influence control cells as well. Also in a treatment regimen where both compounds were added in consecutive rounds of lower concentrations, FOXO4-DRI proved to be selective against senescence yet safe to normal cells.

I reviewed the Quercetin/Dasanatib paper two years ago.  It was an early proof-of-principle, using medications that are already known (and FDA approved).  But the 1-2 punch is not sufficiently selective–it is toxic to normal cells.

I missed the two papers about ABT-263  [ref] and ABT-737  [ref].  BCL-2 is the founding member of another family of proteins that signal a cell to resist apoptosis.  Both ABT-263 and ABT-737 were identifed in screens for agents that block BCL-2.  These two studies published in Nature last year, one from University of Arkansas, the other from the Weizmann Institute, both use radiation exposure to create a large population of senescent cells, and then show that the senescent cells are selectively eliminted by ABT.  The ABT-263 paper included some in vivo results, indicating enhanced growth of blood stem cells after senescent cells have been removed.  In vivo testing of ABT-737 was limited.  Neither group reports the selectivity index (SI50) as calculated by Keizer in the latest study; but from graphs that they do present, it is clear that ABT-263 is more selective than ABT-737, and that neither is as selective as FOXO-DRI.

DIS and OIS are the senescent cells; G and V are the control (normal) cells.  ABT 737 (the middle bar in each of the 4 sets) kills more than half the senescent cells, but at the cost of taking out ~20% of the healthy cells.

ABT-263 appears to be more selective than ABT-737.  Normal cells (left) are not noticeably affected at a concentration where ~70% of the senescent cells are eliminated (right).

The original marker used to identify and target senescent cells by the Mayo Clinic’s 2011 study was p16Ink4a.  The selective elimination technique they used (in 2011) was limited to genetically modified mice, but a year ago, a new paper from Mayo Clinic demonstrated a similar procedure for ordinary, non-GMO mice.  Twice weekly injections of an antibody that induced apoptosis in cells that expressed p16Ink4a extended lifespan of the mice by 25% – 30% compared to controls, comparable to the results in the 2011 paper.  Caveat: the control mice received sham injections that shortened their lifespans.


The Bottom Line

The idea of removing senescent cells has a lot of appeal.  Not only does it enjoy broad empirical support in mammals; it also pulls together several ideas about the origin of aging:

  • Parabiosis experiments and their follow-ons have convinced us that circulating chemical signals form the basis of an epigenetic clock.  Some of these circulating molecules are known to come from senescent cells.
  • Aging commonly accelerates exponentially with age, as though it were driven by a positive feedback loop.  Senescent cells secrete cytokines that make more senescent cells–there’s your feedback.
  • Short telomeres initiate senescent cells.  At any given time, there is a bell-shaped curve of telomere length among the body’s cells.  The tail of the telomere distribution contains a few cells that are driven to senescenceby having very short telomeres.

There is now a world-wide effort, making rapid progress toward specifity in senolytic treatments.  In other words, FOXO-DRI is the newest agent, and it shows the best ratio yet for killing senescent cells while avoiding collateral damage to healthy cells.  (It cannot be taken orally and must be injected, but perhaps this is not such a great drawback for a treatment that is needed only intermittently, every few years.)

How will such promising mouse results translate into human health and life extension?  We have as yet no data, not even anecdotes.  But perhaps we are near the point where hope and courage will motivate the first self-experimenting volunteers.  Caloric restriction and its mimetics produce much greater percent increases in lifespan in mice (2 year lifespan) compared to dogs (10 yrs) or monkeys or humans.  Senolytics work via a completely independent pathway; we can hope that percent benefits in humans will be closer to the mice.  Since this is about upregulating elimination of cells via apoptosis, the strongest benefits are likely to be against cancer, and mice are more vulnerable to cancer than humans.

This is a fast-moving field in which researchers are in a rush to publish and (presumably) pharmaceutical companies are taking pains to keep their results hushed up.  Sharing of information and resources could push this research over the top and give us the first full decade of human life extension.

Anti-aging breakthrough? This one looks authentic

Yes, there are too many reports the last few years of people announcing the end of aging.  But this, I feel, is the first time a drug has been discovered that has the potential to extend our lives by decades, with a few injections.

Background: As we get older, some of our cells become ‘senescent’.  They don’t just fall down on the job; they send signals out to the body which make us old and (especially) promote inflammation, leading to higher rates of cancer, heart disease and AD.  There are very few senescent cells—perhaps one in 10,000.  But they do enormous damage.

A few years ago, Jan van Deursen of Mayo Clinic showed that by killing off the senescent cells in mice, he could make them live 25% longer.  This was done with a drug, but the catch was that the mice were genetically modified.  Each senescent cell had a bomb in it, and all it took was delivering a trigger.  The treatment only worked if the mice were specially prepared (before birth) with this genetic modification.

This was 2011.  University labs  and drug companies around the world took appropriate notice, and they started working on ways to kill senescent cells without harming normal cells that would work in animals (including you and me) who have not been prepared ahead of time with bombs in their senescent cells.

Yesterday, scientists from the Netherlands announced successful deployment of a magic bullet that would kill senescent cells only.  It works in cell cultures, and it works as an injection in mice that have short lifespans.  (They haven’t had time yet to test it in fully normal mice, but in theory, it should.)

It’s a large molecule that you can’t take in a pill, because your digestive system would dismember it.  But it can be injected, and one dose ought to be rejuvenating.  I expect that people will be lining up to try this within a year, and that the injections will turn out to be needed only once in every few years.

My previous article on the subject
Link to Science News article by Mitch Leslie
Research Article

New Database of Lifespan Trials

Human Ageing Genomic Resources announced last week their on-line database of animal studies that evaluated drugs and supplements for extended lifespan.  HAGR is a project of the University of Liverpool, spearheaded by João Pedro de Magalhaes, who has been an activist-scientist in aging research since his days as a grad student at Harvard.

The database is a great resource for researchers, and helps assure that we have no excuse for overlooking a substance or a perspective or a particular result.  Maintaining and updating it will continue to be an important and demanding project.

The full database covers 1316 studies, and I will review here just those on mice and rats.  My reason is that life extension in simpler animals turns out to be too easy.  There is much we can learn about universal biochemistry from studies in worms and flies, but most of the successes there fail when the (longer and costlier) studies are done in mammals.

Here is a spreadsheet extracting just the 93 studies on mice and rats.  You can view it online, and if you download it or copy it into your own GoogleDrive account, you can sort and edit and re-arrange it at will.


Old News

Rapamycin: Has the most studies and the best data.  Clearly works, but has side effects and it is not yet clear if it is appropriate for general use.  Make your own decision.  [read more]

Metformin: We have extensive experience with humans, and clear indications that it lowers cancer rates and ACM*, but there are dangers and side-effects. [read more]

Melatonin: Good evidence for modest life extension in rodents. For some people, it’s also a good night’s sleep; for others it can lead to grogginess or depression.

Aspirin:  The best evidence for lower cancer and ACM* is in humans.  Most people can tolerate a daily mini-aspirin without stomach complications.  

Epithalamin (and other short peptides):  This is work by Anisimov in St Petersburg, and it is so promising that I can’t understand why it isn’t being replicated all over the world. [read more]

Deprenyl:  Old studies, but they show consistent, if modest life extension.  It affects CNS in ways that you might feel, might like or might not.  [read more]

Vitamin E:  This is just one study, dosage equivalent to hundreds of pills a day, mice kept in shivering cold conditions.  [ref]  In a large human study, antioxidant vitamins increased mortality. [ref]

Acarbose: A diabetes drug that blocks the digestion of carbohydrates.  Side effects and toxicity make it less promising than metformin as a general recommendation.  [drug info]

C60 Fullerene:  Just one study in 6 rats, with spectacular results.  Replication has failed [private communication from Anton Kulaga].  Nevertheless, there are thousands of people experimenting on themselves. [read more]

Curcumin: There are major questions about absorption and dosage, but no question that anti-inflammatories are a good general strategy, and curcumin is a good anti-inflammatory. [read more]

Green tea:  Small but consistent life extension from polyphenols extracted from tea.  From a number of high-profile experimentalists, 2013.

Resveratrol: Works great in simpler animals, including some vertebrates, but in mammals life extension has been limited to overweight mice on a high-fat diet. [read more]


The New Part

BHT:  This is an anti-oxidant and chelating agent, which means that it is attracted to metal ions, it pulls them out of circulation and takes them out of commission.  This sounds good when it’s removing mercury or lead, but less good when it’s removing iron and dangerous if it’s removing zinc or other essential trace minerals.  BHT has long been used as a food packaging additive to preserve freshness, and it is still avoided by natural foods types. This Russian study [2003] found 17% life extension in mice. 

Creatine:  Used by body-builders, it encourages muscle growth by blocking myostatin.  It also increases nerve growth, and slows shrinking of the brain.  In one promising mouse study [2008], average lifespan increased 9%.

Icariin: This is an active ingredient in the traditional Chinese herb which in the West is known as Horny Goat Weed.  One mouse study, 6% increase in lifespan.

VI-28: Another Chinese herb.  Just one study, up to 14% increase.

Royal Jelly:  Queen bees are genetically identical to worker bees, yet they live 100 times longer.  Is it the royal jelly they are fed?  One mouse study [2003] showed a 25% increase in mean lifespan, but no increase in max lifespan.

N-Acetyl Cysteine:  Glutathione is an antioxidant associated with mitochondria.  Unquestionably, glutathione is a good thing.  Too bad we can’t just eat it.  The next best thing is to take the precursor, NAC, which seems to lead to increased glutathione throughout the body. This one study [2010] came out of the same prestigious group at Jackson Labs that brought us rapamycin.  Mean lifespan increased a stunning 25%. Two reservations: (1) they used enormous dosages, and (2) the mice on high-dose NAC ate less, so they probably benefited from caloric restriction.

Ginkgo biloba: Extract from the stinky fruit of an ancient oriental tree.  Traditionally used as a neuroprotective and concentration enhancer, for which it is mildly effective.  In 1998, a single study found 17% life extension in rats.  Who knew?


The Bottom Line

Clearly there is a great deal of promise here, but there is also much work to be done before we have it sorted out.

  • Many treatments have shown promising results in just one study, and that needs confirmation.  My top priorities would be epithalamin, NAC, and royal jelly.
  • Other treatments inspire enough confidence that we should be optimizing dosage for human use.
  • As I have written, the most important work before us now is to see how these different treatments combine.  Most combinations won’t work together, but when we find the few that synergize we will have a candidate protocol for major life extension in humans.

If you’re curious, of the substances reviewed here, I personally take metformin, aspirin, creatine and NAC.  I season with turmeric a few times a week.  I have dabbled with deprenyl and rapamycin.


* All-Cause Mortality


NF-kB Beyond Inflammation

At different times, I take about 8 different anti-inflammatory supplements, including aspirin, ibuprofen, omega 3 oils, curcumin, berberine, resveratrol, ashwaghanda, and boswellia, in addition to eating foods such as ginger, rosemary, tea and several mushroom species with anti-inflammatory effects.  There is good evidence for benefits from each of these individually, but I have no idea how they interact with one another.  Just last week, I learned that they all act (in part) through inhibition of NF-kB.

It’s certain that the separate benefits of each of these don’t just add up in combination.  It could be that all of them together are no better than just one of them individually.  It might even be that they interfere destructively with one another, competing for a common receptor, so that piling on more supplements is counter-productive.  There is no research on interactions among longevity supplements.



In this (2007) study, a multidisciplinary team performed a systematic search for blood factors that change most consistently with age over a sample of mammalian models, and organized these into modules that tend to vary in a coordinated way.  They then searched for transcription factors that can turn each module on or off.  Their most prominent finding was that NF-κB turns on the suite of factors characteristic of old age.  Out on a teleological limb, they were bold enough to call it “Enforcement of aging by continual NF-κB activity” in the title of the article.  (In this perspective, senescence is an active process, coordinated by the genome.  I agree there is good evidence for this.)

After a long path leading to NF-κB as their prime subject, the authors go on to test whether inhibiting NF-κB can have anti-aging effects.  The first obstacle that they encounter: NF-κB has important developmental functions (in young animals) and also is essential for regulating apoptosis (in older animals as well).  Mice with genes for NF-κB knocked out don’t survive gestation.  So they arranged to selectively “blockade” the binding of NF-κB to DNA in old mice, in skin cells only.  The result was a dramatic rejuvenation of the skin.



Research with rodents and humans suggest that there are factors in the blood that keep us young and, more important, factors that make us old.  Prime suspects in the latter category are the signals that dial up inflammation.  It’s my hunch that the most effective anti-aging strategy over the next 10 years will be to re-adjust signal molecules in the blood, adding what we lose with age but, more important, neutralizing or inhibiting pro-aging factors.

For various reasons, NF-κB is a good place to start.

“NF-kB has been termed the central mediator of the immune response. Gene knockout and other studies establish roles for NF-kB in the ontogeny of the immune system but also demonstrate that NF-kB participates at multiple steps during oncogenesis [ref] and the regulation of programmed cell death [ref].” [John Hiscott]

It is a complex of different molecules that acts as a master transcription factor.  It is always resident in the periphery of the cell, waiting so that it can be activated quickly when needed.  Latent, NF-κB is bound to an inhibitor molecule called IκB.  When a stimulus comes along that phosphorylates the IκB, the NF-κB is freed to enter the cell nucleus and switch on a variety of different genes, which varies from one cell type to another.  The best-known activity of NF-κB is in white blood cells (T and B cells) where it activates an inflammatory response  involving TNFa and IL-6.  Overactivity of NF-κB with age is a mediator of the systemic inflammation that contributes so much to cancer, heart disease and dementia.

NF-κB itself is not circulated in the blood, but signals in the blood can cause it to be turned on.  The Conboys early recognized NF-κB as one of the pathways that promote aging in parabiosis and transfusion experiments, where blood from older mice is introduced into younger mice.  It is a very good bet that inhibiting NF-κB would slow inflammaging, perhaps relieving arthritic and other auto-immune symptoms immediately, while reducing long-term risk of mortality and disease.


Inflammaging and Auto-immunity

Exponential amplification is a basic principle of the body’s immune response.  When the signal is received announcing an invader or an infection, there are just a few cells involved.  These send signals that trigger an immune response in other cells, triggering a chain reaction.

The beauty of such a system is that it ramps up quickly, and can mobilize a response throughout the body in short order.  This is also the danger of the system.  It requires an accurate and reliable switch to turn it off; otherwise, it can become like the Sorceror’s Apprentice, each magic broomstick producing two more to carry water until the workshop is flooded.

Note:  Dr. Katcher, in a note below, makes an important point about this positive feedback loop.

  • Senescent cells  spit out inflammatory cytokines
  • This activates NF-κB, which blocks apoptosis that could get rid of the senescent cells.
  • Inflammation from NF-κB turns more cells senescent, beginning the cycle over again.

NF-κB is a master switch that sets in motion a chain of events that is specific to a cell type and its environment.  Some auto-immune diseases (e.g., arthritis, type 1 diabetes, asthma, Crohn’s disease and irritable bowel) are associated with an excess of NF-κB [ref].  Its activation generally rises with age [in mice, in humans], but it is necessary at all ages, particularly for its contribution to the regulation of apoptosis (the selective elimination of cells that are potentially damaging).  Animals lacking NF-κB are not viable; so it will probably be necessary to strongly but selectively inhibit NF-κB, beginning in middle age.

Inflammatory responses are complex and focused on the immediate threat at hand.  NF-κB is a master switch that sets in motion a chain of events that is specific to a cell type and its environment  It’s true that without NF-κB this response doesn’t happen, but the response to NF-κB varies from cel to cell.  In this sense, inhibiting NF-κB is a kind of blunt instrument.  It works to damp the body’s inflammatory response globally, but even better would be if we could selectively shut off the body’s attack on itself.  It’s true that NF-κB activation rises with age [in mice, in humans].  But the real problem is not too much NF-κB expression, but the fact that NF-κB becomes defocused, so that the inflammatory response is not focused on a particular threat, but generalized throughout the body [ref].

Here’s an angle I learned about recently from Steve Cole:  While inflammation is an important defense against bacterial infection, it is actually counter-productive against viruses.  Inflammation can create an environment that invites viral infection, perhaps because apoptosis is suppressed.  (NF-κB suppresses apoptosis.)  Viruses aren’t so dumb, and some of them have learned the advantage of promoting NF-κB.  Some of the reason that NF-κB is upregulated with age may be a residue of chronic viral infections.  [Hiscott, again]  (Just to confuse us, NF-κB can also promote apoptosis in other contexts.)


Two pathways

All the anti-inflammatory agents that I have been able to catalog work by one or both of these two pathways:  NF-κB and COX2.  By most accounts, NF-κB is upstream of COX2, but the two are interrelated.  NF-κB regulates COX2, and also COX2 feeds back to regulate NF-κB.

NSAID drugs (aspirin, ibuprofen, naproxen, celecoxib, etc.) target cyclooxygenase-2=COX2.  Common herbal anti-inflammatories, including curcumin, resveratrol, vitamin D and omega 3 oils (the last two not exactly herbs) are active both against COX2 and NF-κB.  Inhibiting COX2 is a classic strategy for combatting arthritis.  The more potent COX2 inhibitors have a tendency to decrease cancer risk, while increasing cardiovascular risk.  This doesn’t necessarily mean, “it’s a wash”–rather the stronger NSAID’s are right for people with some genetic risk profiles and should be avoided by others.  Aspirin is the cheapest and oldest of the NSAIDs, for which there is copious data available on tens of millions of individuals.  There is reasonably good evidence that aspirin leads to lower heart risk as well, probably because of anti-clotting rather than anti-inflammatory action [read more].  Daily aspirin also lowers risk of several cancers.


Inhibiting NF-κB

Intermittent fasting or caloric restriction tends to prevent NFκB binding to chromosomes..  There’s also a long list of natural products that inhibit NFκB.

source of this chart

There are a few pharmaceutical products that inhibit NFκB, though none has been developed explicitly for this purpose.  These include emetine, fluorosalan, sunitinib malate, bithionol, narasin, tribromsalan, and lestaurtinib.  Emetine (as the name suggests) is used to induce vomiting and also to treat amoebic diseases.  It is the most potent inhibitor of NFκB among the listed drugs.  Sunitinib and Lestaurtinib are cancer drugs. Bithionol is used in de-worming animals. Narasin is an uncommon antibiotic. Tribromsalan is used externally as an antiseptic. None of these is marketed to inhibit NF-κB, and none have (to my knowledge) been tested for anti-aging properties.  The larger pool of prescription drugs that affect NF-κB are all steroids.  For example, dexamethasoneis a glucocorticoid (steroid) drug that was one of the earliest inhibitors of NF-κB to be discovered.

Many items in the list of natural products have multiple benefits. Silymarin has been reported to promote telomerase.  Rosemary and cloves protect against infection.  Berberine helps maintain insulin sensitivity, and was found to be as good as metformin in one test.  Tea polyphenols and resveratrol have been promoted as generally anti-aging.  Too much has already been written about curcumin.

New to me in this list is celastrol, an ingredient in thunder god vine (Tripterygium wilfordii).  This is a Chinese herb (leigong teng = 雷公藤), that has been prescribed for centuries in formulas to relieve arthritis, along with lupus, MS and other autoimmune disorders.  It is reported to be a powerful appetite suppressant and weight loss aid.  The trouble is that it is toxic, and the thunder god root must be prepared carefully in order to exclude triptolide, which is yet more toxic.  Experienced practitioners of traditional Chinese medicine know how to mix with other herbs and control dosage to minimize side-effects.  In the absence of this kind of expertise, I can only counsel experimenting gingerly with tiny quantities of thunder god vine in order to guage your personal response.


How important is inflammation?

It would be interesting (from a theoretical and a practical vantage) to know what is the maximum benefit available from manipulating the inflammatory pathway.  Inflammaging is linked to all the diseases of old age.  Suppose we dialed the systemic inflammation in a 80-year old back to where it was when he was 20, but we made no other change in the body.  What would be the effect on mortality and morbidity?  On vitality, resistance to infection, and stamina?  In other words, how much of the aging process is directly attributable to inflammation?

We might try to get a handle on this question via an epidemiological calculation: What is the correlation between inflammation and all-cause mortality?  If we extrapolate back to the inflammation level of a 20-year-old, how far does that go toward restoring the mortality rates of a 20-year-old?

We might think to look at genetically modified mice without NF-κB; however, they die in utero.  (There are no pure aging genes; aging is caused by re-balancing hormones and proteins, all of which have life-supporting as well as life-denying functions.)  There is a genetic variant of NF-κB that tends to be more common in centennarians than the rest of us [ref].


How much does inflammation rise with age?
Erythrocyte Sedimentation Rate and C-Reactive Protein

120 years old, the ESR test is still the most basic (and cheapest) measure of systemic inflammation.  The quantity measured is the number of red blood cells that clump together and fall out of solution in one hour.  Inflammation makes red blood cells sticky, and I’ve seen two explanations for the reason.  One is that fibrinogen, the clotting protein, rises with inflammation; the other is that the negative charge (zeta potential) naturally associated with oxygen-carrying red blood cells decreases with inflammation, so there is less mutual electrostatic repulsion.  The increase in blood’s tendency to clot that is associated with inflammation is part of the reason that inflammation is a risk for heart attacks and stroke.

C-Reactive Protein is a protein created in the liver as part of the response to inflammation.  It is easily measured with an antibody, so it has become the second most common blood test for inflammation.

ESR rises with age, but not dramatically compared to interpersonal variation:

 In fact, the difference between women and men is more than the difference between an 80-year-old and a 20-year-old of either sex.  Increase in CRP with age is even more subtle.  (This article claims it doesn’t rise at all, in a small sample of <400 patients.)

This article claims that CRP does rise with age (using a sample of 21,000):

A Glasgow study of 160,000 patients found strong correlation between CRP and near term mortality (within a year, HR=20) but not much for longer term. This is not what I was expecting.

In 26,000 patients, inflammatory markers were associated with a 1.5-fold increase in all-cause mortality (ACM) over 8 years.

A Norwegian study of 7,000 men and women found that high levels of CRP raised ACM only by a factor 1.25 (equivalent to just 2 years of aging).  For comparison, the ACM risk for an 80-year-old male is 60 times higher than a 20-year-old male.  The corresponding number for females is nearly 120.

The implication is that either inflammation is a minor (though significant) cause of mortality, or else the markers that we have for inflammation (including ESR, CRP and leucocytes) are not capturing the rise in systemic inflammation.

Hint: “Centennarians, on the other hand, manage to stave off these deleterious sequelae.Despite signs of inflammation, such as high levels of interleukin-6 (IL-6), fibrinogen, and coagulation factors, they are remarkably free of most age-related diseases that have an inflammatory component.” [ref]


NF-κB in the Brain

It is my favorite hypothesis that aging is mediated through hormonal signaling, under control of a clock in the neuroendocrine regions of the brain.  So I am interested in changing NF-κB activity in the aging brain.  This paper describes roles for NF-κB in brain development, regeneration after injury, and also evidence that NF-κB can be activated in response to nerve signals.  In the other causal direction, neural signaling (and presumably behavior) can change in response to NF-κB.  Directly on target (in my book) is this paper from Nature (2013). “By systematically controlling NF-kB activity in the hypothalamus alone, the authors are able to increase the healthspan as well as the lifespan of mice.”


The Big Picture

In the Prelude above, we found evidence that NF-κB is a master regulator that turns on a suite of genes that “enforces aging”.  But in the section, How important is inflammation?, we found evidence that, while inflammation certainly increases with age, the increase is not large compared to the scatter among individuals.  Centennarians commonly have high levels of inflammation, along with robust health.  Large increases in inflammation are common just in the last year of life, but they are not well correlated with the gradual increase in mortality with age.

The combination of these two findings suggests that NF-κB has other powerful roles in promoting senescence, in addition to its well-known role as effector of inflammation.  Maybe it is a master regulator of development and aging, akin to mTOR and FOXO.  It is a hypothesis worth testing that carefully tailored inhibition of NF-κB is a life extension strategy, so long as we can preserve apoptosis at an appropriate level.

Fasting-Mimicking Diet: Can You Make it a Habit?

I was glad to see Valter Longo’s Fasting-Mimicking diet in the news again this week.  I have been enthusiastic about Longo’s work ever since he documented altruistic suicide of yeast cells for his PhD thesis in the 1990s. Programmed death in one-celled protists was considered an affront to evolutionary theory at the time, and he had a devil of a time getting his findings into print.

Longo discovered in 2002-2005 that fasting had a powerful benefit for cancer patients, and that in conjunction with either radiation or chemo, it greatly magnified the benefits while mitigating the side effects.  Intermittent fasting had benefits, too, for the general population, independent of cancer.  It seems to be a way to get the health benefits of caloric restriction and it is easier to stick to for many people than a consistently low calorie diet.

But Longo couldn’t get either doctors or patients interested in the fasting program.  Part of the problem was the toxic mix of capitalism with medicine: the US relies on testing and promotion by profit-making companies to push medical technology forward, and fasting isn’t a product that anyone can make money on.  There was also an emotional truth: cancer patients feel scared, beleaguered, emotionally drained.  So much is dragging them way outside their comfort zone that it takes extraordinary strength not to fall back on food as one of life’s most reliable comforts.

So many medical researchers see their job as finding treatments, and leave the problem of adoption and compliance to someone else.  But Longo set himself single-mindedly to the task of bringing the benefits of fasting to a wider swath of the population.  The question that led to the Fasting-Mimicking Diet (FMD):  What dietary regimen can provide the greater part of the benefits of a water fast while inducing less hunger and minimal disruption to concentration, vitality and the pace of life?

I have done about 9 cycles of the FMD over the past 1½  years, including this week.  In my personal experience, 5 days of the FMD is eminently tolerable once I begin, though I still face resistance when I think in advance about disrupting my comforting food routines.  (Yoga, swimming and meditation have the same barriers for me–difficult in the anticipation, enjoyable once I begin.)  For me, FMD is not a weight loss program.  I gain back all the weight I’ve lost within a week after the 5-day program is finished.  Others may have different experiences.  People have used the 3-meal version of FMD (3 meals at 360 calories) as a medium-term weight-loss program, but the protein content is probably too low for those concerned about maintaining lean mass long term.

In his new study, 100 participants were randomized into two groups.  The first group did three rounds of 5 days on FMD over three months.  The second group did nothing special for three months, but were given an opportunity to try the same three rounds during the following three months.  The average weight loss was 6 pounds for those who completed all three cycles; lean body mass was lost, in the same proportion as fat.  Blood pressure, blood sugar, triglycerides all improved on average, and C-reactive protein (an inflammatory marker) went down as well.  

The study included a range of healthy people as well as people who carried more weight and had higher risk factors.  It was those with the higher inflammation and blood sugar who realized big benefits from the program, and the already-healthy were averaged in.  There is a lot of evidence to indicate that intermittent fasting works, and that the FMD delivers similar benefits.  But if you’re already lean and healthy with low blood sugar, then it’s less clear whether there are substantial additional benefits from intermittent fasting.

Notable was a reduction in the hormone IGF-1, which I don’t necessarily regard as a good thing.  On the one hand, lower IGF-1 is characteristic of all caloric restriction models, in animals and humans, in which life span is increased.  So it is an indication that the FMD was effective.  But low IGF-1 has consistently been found to increase risk for all-cause mortality, and heart disease in particular [ref, ref, ref].  Benefits of higher IGF-1 include maintenance of muscle mass and growth of nerves that diminishes with age.  

An additional benefit documented in the past is a “reset” of the immune system.  The white cell population is pruned during fasting, and the most-needed naive T-cells regrow after eating resumes.

Antidote to Obesity

It’s the (large and increasing) population of unhealthy people that Longo is targeting.  There is every reason to expect substantial benefits, but the big issue remainss: how many people can be motivated to take up the practice and stick to it?  The question was touched on only peripherally in the current study, without discussion; of 48 subjects selected for the first round, 39 stuck it out for three FMD cycles over three months (81% compliance).  That’s encouraging, but what we really want to know is: how many people will actually modify their eating rhythms for years at a time?  Will they feel the benefits and will that motivate them to stick with it month after month, five days each month?

And will this translate into long-term weight loss?  Sustaining weight loss is notoriously difficult for about 95% of the population.  People can stick to a diet for a time, but the Siren song of food is with us everywhere, and sooner or later we succumb.  Fortunately, there’s good reason to hope that the FMD discipline offers benefits even if weight loss is not sustained.


How to do it

Longo’s own company, L-Nutra offers a packaged diet, called ProLon, availble through health care professionals.  It’s pricey and may be covered by insurance.  LifeBox offers a non-prescription alternative that is not as much cheaper as it might be.  For my own experimentation, I have preferred to use fresh, whole foods approximating the same macronutrient proportions as the ProLon package.  It requires some time and attention in food prep, but it costs less than you’re currently spending on the same meals, and you can fill yourself with satisfying portions of fresh vegetables for the same 360 calories.  Here is my page of instructions and advice, with recipes designed by Enid Kassner.

For many of us, our relationship to food is central to our psychology.  Shaking up food habits disrupts everything else as well.  It’s the main reason that food habits are so hard to change, but for me, it’s also a good thing.  I enjoy the challenge and the self-awareness that come from a new frame of reference; fasting changes my perspective, my emotional baseline, and my mental state.  For me, the first day is unfocused, low energy, but often a time of creative new ideas; the second day is distracted, obsessive, sometimes headache-y or otherwise uncomfortable, and beginning with the third day there is returning energy, along with a freedom that I didn’t know I missed.


Why is life expectancy in America lagging?

Part of the answer is certainly cultural.  Advertising, parties, lunch and dinner meetings often reinforce consumption of food that is designed to be addictive for the sake of corporate profits.  In America, we are surrounded by overweight people, but France and Italy have much lower obesity rates, and you can walk all day around the cities of Japan or China without encountering anyone who is seriously overweight.  Even in America, the problem has grown way out of proportion only in the last 40 years.  This and income disparity are the main reasons that life expectancy in America is at the bottom of the developed world.  Our unaffordable, dysfunctional healthcare system provides many additional reasons.  Meanwhile, life expectancy in Asia is climbing at an exemplary pace.

Longo’s FMD is designed to address this most accessible factor in the diseases of late life for a large swath of people who find they cannot lose weight.  His research is based not just on effectiveness but equally important on tolerability.  How many people will adopt it and reap the benefits?  This will be an important question for public health going forward.  But you are a unique individual, not a statistical median.  I encourage you to experiment with FMD, see what you learn about yourself, and decide if it can be a valuable part of your health program in the long term.