David Sinclair on Aging as Lost Information, Fasting, Biosensors and Living Past 100

High-Level Overview

Lex Fridman talks with David Sinclair, professor of genetics at Harvard Medical School, co-director of the Paul F. Glenn Center for the Biology of Aging and author of Lifespan. Sinclair lays out his view that aging comes down to one thing: the loss of information in our cells, mostly in the epigenome that decides which genes each cell reads. From there the conversation covers his lab's "reset switch" that restored sight to old mice, the wearables and home tests he believes will turn medicine into a data-driven dashboard, and the everyday habits he practices himself: eating one meal a day, leaning on plants, short bursts of exercise, deep sleep and staying calm. It ends with a long exchange on death, meaning and whether we would enjoy life less if we lived forever.

The tone is warm, curious and often playful. Sinclair opens with the Winnie the Pooh poem his grandmother read to him and says he still feels six years old, and Fridman adds his own experiments with fasting, carnivore eating, brown noise and all-nighters. The conversation swings between hard molecular biology, near-future health technology, personal lifestyle habits and philosophy.

The Discussion Explores:

• Why Sinclair frames aging as information lost to noise, and the DVD-with-scratches metaphor for the epigenome
• How DNA breaks and cell stress scramble the epigenome, and how his lab speeds up and reverses aging in mice
• Three of the Yamanaka genes as a reset switch that restored vision in old mice, with human studies planned
• Biosensors, blood tests, cheek swabs and AI as a dashboard for the body, and why doctors are "running blind"
• Fasting and one meal a day as his top recommendation, plus when you eat versus what you eat
• Plant-heavy eating, mTOR and xenohormetic molecules from stressed plants
• Exercise, sleep quality and stress as levers on biological age
• How long humans could live, uploading minds, and whether death gives life meaning

Deep Dive into Main Ideas

Staying Six Forever: A Grandmother's Lesson

Fridman opens by asking whether Sinclair feels the tension of being the same person inside while his body ages. Sinclair says he still feels like a kid, and that this helps him as a scientist because researchers need wonder. He says he can see things for the first time and explain them as he would to a six-year-old "because I am that mentally", even though he has to pretend to be a grown-up while running a Harvard lab of 20 people and several companies.

Fridman reads A. A. Milne's poem that ends "now I am six. I am as clever as clever, so I think I'll be six now, forever and ever." Sinclair gets misty-eyed: his grandmother, who partly raised him, read it to him often. She lived through the Second World War in Budapest, tried to shelter Jewish friends in her apartment and fled after the 1956 revolution. Her message to him was to stay young and innocent, keep his wonder and do his best to make humanity better. He says that is what he gets up each morning to do. The two also muse about using AI to recreate people who have died, from Einstein and Feynman to lost loved ones, and Sinclair says he would love to bring Benjamin Franklin into the future to show him the thinking machine in our pockets.

Aging as Lost Information

Sinclair explains that most scientists list eight or nine hallmarks of aging: shortening telomeres, senescent cells that release inflammatory molecules, failing mitochondria, stem cell loss, protein problems and more. Most biologists say aging is too complex to boil down, but he believes it reduces to one equation: the preservation of information and its loss to entropy, or noise. The idea began in the 1990s at MIT, working with baker's yeast, where his team found a gene called SIR2, the silent information regulator. Adding one extra copy made yeast cells live 30% longer. Mammals have seven related genes, the sirtuins, and in mice, turning up SIRT1 slowed aging in the brain.

He describes two kinds of information in a cell. The genome, the digital ATCG code, stays largely intact into old age, which is why a friend of his could clone his dog three times. The epigenome, how DNA is wrapped or looped so that a cell reads the right genes, is what degrades. His analogy: the genome is the pits on a DVD and the epigenome is the reader. Over time brain cells start to look more like skin cells and kidney cells like liver cells, a process his lab calls ex-differentiation, like scratches that stop the reader accessing the music. He says 80 to 90% of the diseases of old age are caused by aging itself, yet medicine keeps treating them with band-aids.

What Scratches the Disc, and Proof in Mice

Aging is not programmed, Sinclair says: there is no selection for it, we simply live long enough to reproduce and then entropy takes over. But he adds a twist called antagonistic pleiotropy, where systems that keep us alive when young come back to bite us later. His lab has found two definite sources of noise. The first is broken chromosomes. When DNA breaks, proteins that normally regulate genes leave their posts to help repair the break, then return, like balls in a ping-pong game. With age some balls never come back. We have close to a trillion DNA breaks a day, he says, so over 80 years this adds up. The second is cell stress, such as damaged nerves, and he notes that smoking also speeds up biological age.

To test this, his team created a small, adjustable number of extra DNA breaks in mice, like a dial that goes to 11 in a nod to Spinal Tap. At a setting of about four, a mouse gets old in ten months, with gray hair, old skin and age-related disease. Using machine learning to read the epigenome as a clock, those mice measure 50% older than normal. That gives him a way to speed aging up in a controlled way, and then test ways to reverse it.

The Reset Switch

In a paper in the December 2020 issue of Nature, Sinclair's lab reported that three embryonic genes, drawn from the four discovered by Shinya Yamanaka, whose discovery won a Nobel Prize, can reset the age of adult tissue in four to eight weeks. Yamanaka's full set, switched on strongly, turns adult cells into stem cells. Sinclair's question was whether you could take age partly back without creating tumors or stem cells. The answer, he says, is yes: the body has a system that can wind a cell back to a certain point and no further. Old mice that had lost their vision to aging could see again, and prematurely aged mice are getting their ability to learn back.

The risk is tumors if it is overdone, but the genes are delivered by viruses so the team controls where and when they switch on, and mice given high levels throughout the body did not get cancer for over a year. He expected human studies in less than two years. He also finds that rejuvenated nerve cells seem to recover old memories, suggesting memories are not lost but the cells forget how to read them. Asked about AI, he says that if you are not using it in biology you are getting left behind: his lab uses it for aging clocks, protein folding, assembling genomes and predicting a mouse's lifespan from frailty, hearing and eyesight, and he struggles to hire enough bioinformaticians.

A Dashboard for the Body

Sinclair describes the monitoring he predicted in Lifespan as largely here already, sped up by COVID-19. He wears an Oura ring, a watch and a small BioButton biosticker on his chest that records a thousand times a second and tracks sleep position, coughing, temperature and heart signals, with the aim of telling pneumonia from a cold or predicting a heart attack before it happens. Hospitals already use such devices to send cardiac patients home early, saving at least $2,000 per patient. He has done blood tests for 12 years with InsideTracker, a company he consults for, which tracks 34 parameters and gives food and supplement advice drawn from thousands of papers; he says he co-authored a paper showing its recommendations beat the leading type 2 diabetes drug. By those numbers, at 51 he is at least as healthy as someone in their early 40s.

He tells of a video checkup where his doctor was "running blind" and refused to order an HbA1c glucose test because he had no symptoms or family history, so Sinclair pays for such tests himself. He calls the annual checkup medieval, predicts a lawsuit within a few years over a death a $20 sensor could have prevented, and argues hospital budgets will force adoption. Next he sees cheek swabs mailed from home that read hormones, stress, glucose and epigenetic age, with his lab working to cut test costs from hundreds of dollars to about a dollar. Fridman wants to be able to delete data he shares; Sinclair agrees that health data security has to come first, noting that a chest sensor could probably even tell if you had sex that night.

Fasting and When You Eat

Fridman describes his own 72-hour fast and eating one meal a day. Sinclair says fasting is one of the oldest ways to improve health, and his lab showed in the early 2000s that calorie restriction turns on longevity genes, first in yeast. Hunger, exercise, heat and cold make the body think it is under threat and switch into a defensive mode, as whales do and mice do not. If he could recommend one thing to slow aging it would be to skip a meal or two a day. He went from skipping breakfast to skipping lunch too and says he has his physique from age 20 back, feels sharper and no longer gets tired.

His practical advice: work up to it gradually, fill up on tea, coffee with milk or diet sodas, and give it at least two to three weeks to break the habit of chewing and feeling full. He admits he cannot fast beyond a day or two and hates exercise. He cites a study by his friend Rafael de Cabo at the National Institutes of Health in which 10,000 mice were fed different mixes of carbs, protein and fat; the only ones that lived longer were those that ate once a day. So when you eat matters more than what you eat, he says, though the best bet is to do both.

Plants, mTOR and Food That Has Suffered

Fridman says a meat-heavy diet makes him feel great. Sinclair recommends mostly plants: long-lived people in Mediterranean and Okinawan diets eat mostly plants with little red meat, and in animals a sensor called mTOR responds to amino acids plentiful in meat and shortens lifespan when activated. The drug rapamycin blocks it. He likens meat to testosterone and growth hormone: benefits now, possibly at the expense of long-term health. In the terms of the disposable soma theory, meat may put you in the mouse category of fast growth and reproduction rather than the long-lived whale category.

He eats lots of leafy greens, especially spinach, avoids fruit juice and most fruit, and rarely eats sugar, using stevia instead. He also explains xenohormesis, a term he coined with Konrad Howitz: stressed plants make colorful protective molecules, and eating them turns on our own defenses. His example is resveratrol, which grapevines make when dried out or under fungus, and which his team showed activates the sirtuin enzyme; he takes it as a supplement. So he prefers colorful, organic, locally grown produce with caterpillar bites over pale lettuce. Fridman calls these the David Goggins of vegetables.

Exercise, Sleep and Stress

Exercise clearly works, Sinclair says, citing studies of cyclists who ride over 80 miles a week having around a 40% reduction in diseases such as heart disease. But the benefit curve flattens quickly: about ten minutes on a treadmill a few times a week, getting out of breath, seems very beneficial, and overdoing it can wear out joints. He also lifts weights to maintain hormone levels and muscle, which is lost at about a percent a year, because he does not want to fall and break a hip. Fridman shares that he runs long and slow while listening to brown noise, which helps him focus.

On sleep, rats deprived of sleep for two weeks develop type 2 diabetes, and disrupting the body's daily cycle causes premature aging. As we age that cycle weakens, which disturbs sleep and speeds aging further, a feedback loop that is why shift workers suffer more age-related disease. Sinclair watches the quality of his sleep more than the hours, especially deep sleep, which alcohol ruins. On stress, he says the brain is the center of longevity: raising the sirtuin gene only in neurons made mice live longer, and calming inflammation in the hypothalamus does too. Once a shy, perfectionist child who cried over grades, he has worked to become someone who almost never gets stressed. His advice: "get your brain in order first before you tackle anything else."

How Long Could We Live, and Does Death Give Life Meaning?

Sinclair says a person who does nothing special lives to about 80, while eating well, not overeating, exercising a bit, sleeping well and not stressing adds an average of 14 years, to about 94. People in their 20s should aim for at least 100. With medicines such as rapamycin and metformin, which he takes, plus lifestyle, we should get well past 100. He sees no maximum limit: whales live hundreds of years and some trees thousands. Uploading a mind would probably require death and scans of which genes are on in each neuron, using methods like George Church's, so he would rather rejuvenate the brain we have, perhaps resetting the body hundreds of times.

Fridman raises Ernest Becker's The Denial of Death and Viktor Frankl's view that mortality gives life meaning. Sinclair agrees we are wired to avoid thinking about death, which is one of the hardest things he faces when he speaks publicly, but for himself he calls the idea that death makes life sweet "completely wrong": he enjoys each day because he makes it joyous. Fridman worries he might enjoy life less if he lived forever, and argues joy is a matter of perspective. Sinclair closes by describing how he compares himself to a lifeless metal cube and reminds himself that he can think, move and eat, and that we are among the most wondrous things in the universe.

Conclusion

Sinclair offers an unusually unified picture of aging: the genome stays mostly intact, but the epigenome that reads it gets scratched by DNA breaks and stress, and in mice that damage can be both caused and partly reversed. His everyday advice follows the same logic of putting the body under mild adversity: eat less often, lean on colorful plants, get out of breath a few times a week, protect deep sleep and keep stress low. He pairs this with a confident forecast that sensors, swabs and AI will soon give everyone a dashboard their doctors lack. Throughout, the six-year-old's wonder his grandmother taught him frames the ambition: live much longer, and enjoy every day of it.

This is a summary of "#189 – David Sinclair: Extending the Human Lifespan Beyond 100 Years" from Lex Fridman Podcast, giving you the GIST so you can decide if you want the REST. Listen to the full episode.