What are the most effective things you can do to slow the aging process and extend your life expectancy? This is the question being asked by a clinical trial that I am organizing, and which seems to be rapidly taking shape. But before the study begins, we have to have candidates to evaluate. We should begin with hypotheses about what we are evaluating. My idea is to consult some experienced experts, and also to crowd-source this choice, and to ask for your help in selecting the supplements and life habits to be evaluated.
Details of the trial were described in two blog posts last spring [One, Two] and a more technical manuscript submitted in May. Outcome will be evaluated based on a variant of DNA PhenoAge, taken from a blood test before, amid, and after the two-year trial. We use methylation pattern differences rather than mortality or health outcomes because the latter take a long time to reveal themselves, and make anti-aging trials prohibitively expensive. Using methylation clocks as an endpoint is a new idea, and we don’t know if it will work, but if it does, it will be 100 times cheaper and 10 times faster than previous methods. We will have enough bandwidth to test a dozen different measures at once, which itself is a revolutionary step.
Many measures are known that are thought to increase life expectancy by a year or a few years each. Of course, we want to know which ones offer their greatest benefits. But even more important, we want to know how they interact, synergize, and interfere with one another. If any one of these measures offered major benefits—say 20 years of life—its effects would be so apparent that we would probably know it already. Likewise, if these measures added up to 20 years of extra life, we would all know some people who are obviously younger than their chronological age. Realistically, we must assume that most of the things we do are redundant. Combining metformin with berberine and gynostemma may offer little additional life expectancy compared to any one separately. A panoply of different anti-inflammatory strategies may be little improvement over an aspirin a day.
But we hope there are exceptions. If two different measures act via completely different metabolic pathways, we have reason to hope that their effects should compound. For example, perhaps life extension interventions based on mitochondrial health synergize with interventions based on rebooting the immune system. Then we might hope that the life extension available from these two measures together is greater than the sum of what we get from the two separately.
We will not tell the people who sign up for this study what to eat, what pills to take, or how much to exercise. We will ask people what they are doing, what they are eating, and what supplements they are taking in a detailed questionnaire. We will select subjects so as to represent a broad array of different strategies and different combinations among these strategies.
Broad, but not too broad. We will have enough statistical power to estimate the interactions among every pair of measures out of 12 that we take as our independent variables. These 12 should be chosen in advance, so the study has a clear focus. If there are more than 12, the number of interactions increases rapidly beyond what we can hope to distinguish (with multivariate statistics). I’ve decided to start with 15, and winnow the list as people sign up for the study and we see what
Here are the criteria I propose:
- Each measure, separately, should have either human mortality data or longevity data to back it up.
- The measures should be easily available to all (excluding intravenous drugs or transfusions)
- The measures should be well-enough known that they are already in common use (and we will have no trouble identifying a diverse group of subjects who use them)
I find that it’s hard to limit the list to 15. It may make sense to include a few measures that are so well established that every participant will be required to comply in order to be included in the study. In this category, we might put
- Limited alcohol consumption (or none)
- Vitamin D at least 5,000 IU daily
- Multi-mineral supplement with magnesium, zinc, chromium, and selenium
- Exercise equivalent to minimum 5 hours a week of walking or yoga
Body weight (BMI) is an important longevity factor, but difficult to account. Studies show that maximum lifespan is associated with BMI between 21 and 25, but in my interpretation, lower BMI is always beneficial for any given individual.
The reason for the apparent paradox is that individuals have genetic disposition to be overweight or underweight. Those who are genetically underweight tend to overeat, because they can do so with no social stigma. Those who are genetically overweight feel compelled to diet all the time (women more than men), and they may be restricting calories just to keep their BMI at 25. These dieters get the most benefit from Caloric Restriction, despite the fact that they don’t look thin.
List of things that lengthen your life
- Love. Men who are married or in close relationships have 7% lower mortality than singles. The number is 4% for women [ref]. These numbers correspond to less than a year of life expectancy. A different study finds loneliness increases mortality by 50%, corresponding to almost 5 years of life.
- Empowerment: Staying employed is worth up to 14 years, and I like to think this is more about being needed than making money. This study claims that the big difference is wealth.
- Anti-inflammatories: Aspirin, ibuprofen, curcumin, or fish oil. This study attributes about a year of life expectancy to daily aspirin.
- High fiber, a proxy for healthy gut flora. We know that they are important, but don’t yet know how to manage the biota for maximal life expectancy.
- Vegetable-based diet This review concluded that vegetarians live 3 years longer, but methodologies and results vary considerably.
- Meditation This is difficult to evaluate, and data is unreliable but encouraging [ref]. The only careful study linked meditation not to mortality but to telomerase activity. I confess I am including meditation in the list from my own intuition and experience.
- Intermittent fasting Extends lifespan in mice and lowers mortality in nursing home studies [ref].
- Interval training Reputed to be the most efficient path toward cardiovascular fitness [ref], and there is limited documentation of benefit for all-cause mortality [ref].
- Donating blood This is another quirky inclusion on my part, but there is data to support it, which I reviewed a few years ago.
- NAC just one study — 30% increase in lifespan of mice
- DHEA Lower blood levels of DHEA are clearly associated with greater age and higher mortality at the same age, but the direction of causality is in dispute.
- Metformin Prescribed for diabetes for decades, this drug also lowers mortality from cancer and heart disease [planned clinical trial].
- Rapamycin The most convincing data available for any supplement for life extension in mice. Early adopters are beginning to experiment on their own.
- Quercetin + Dasatinib (or other senolytics). Senolytics are the best near-term hope we have for a breakthrough in anti-aging medicine; but the combination of quercetin +dasatinib is not yet discriminating enough to be safe for humans, meaning it kills too many regular cells.
- Epithalamin Very promising data from Russia [ref], both in rodents and in people, but there is no one trying to reproduce them in the West.
- Ashwagandha [many benefits, but no mortality or rodent lifespan data]
- Selegiline (deprenyl). In classic studies from the 1980s, lifespan of rats was extended. Data in humans is contradictory [ref].
Perhaps we should begin with a guess about which combinations are likely to be highly redundant. In this way, we could cluster together different strategies and condense more strategies into a manageable list. It might look like this:
- Anti-inflammatories (aspirin, ibuprofen, statins, omega 3, curcumin, boswellia)
- Blood sugar control (metformin, berberine, gynostemma, chromium)
- Social factors (family, employment, wealth, community support, marriage, sex, communing with nature, empowerment)
- Mitochondrial supplements (NAC, CoQ10, PQQ, NR, melatonin, glutathione, carnosine, ALA)
- Immune support (reishi mushrooms, cistanche, andrographis, goldenseal, echinacea)
- Adaptogens (rhodiola, ashwagandha, bacopa, silymarin, pycnogenol)
- Telomerase activators (silymarin, astragalus, ashwagandha, horny goat weed)
- Senolytics (quercetin, dasatinib, fasting)
- Diet (everything from high-protein to high fiber to vegan to paleo in one cluster?)
- Limiting and intermittency of diet (long- and short-term fasting, CR, BMI)
- Exercise (aerobic, strength, interval, walking, competitive sports, yoga)
- Mental focus (meditation, prayer, yoga, tai chi, spiritual practice, cafeine)
- Neuroprotective (ashwagandha, rhodiola, ginkgo, melatonin, bacopa, selegiline, gotu kola, coffee, tea, blueberries, chocolate)
- Multivitamin supplements (including mega D, mega-C, B12, carotenoids and tocopherols)
- Sex and steroid hormones (DHEA, prostaglandin, progesterone, SAMe, testosterone)
- Angiotensin inhibitors (Lotensin, captopril, enelapril)
Or—the best of both worlds—we might structure the study in such a way that it can be analyzed after the fact either with individual strategies or clusters of strategies.
The best way to design a study is to start at the end. I imagine I am two years down the road, taking my first look at results from 5,000 life-extenders. The first thing I will want to do is to look for outliers. Are there a few people who stand out from the bell curve, aging much more slowly? If so, what do they have in common? The advantage of this approach is that it gives maximal flexibility in telling us exactly what we most want to know. The disadvantage is that it is easy to fool ourselves and imagine patterns in a small set of random errors. When there is no initial hypothesis, there is no objective way to calculate a probability that what we find is the result of chance.
Our null hypothesis is that there are no systematic outliers, but only a smooth tail to the probability curve. If there are a few scattered individuals in 5,000 who are aging much more slowly than the rest, we will not find any common thread in what they are doing, and so we must explain their data as anomalies or mistakes. If this is our result, it will be disappointing, sobering, but liberating as well. Those of us who are compulsive about one or another life extension strategy can ease our discipline.
But there is a chance we will find something more interesting. We may find that there are dozens of outliers, that their life extension strategies all overlap in some clear and unambiguous way. We will then have, for the first time, a solid foundation for our personal life-extension habits, and a clear hypothesis for further experiments.
What are your thoughts? Please comment.