Body Count

Illustration: Viktor Koen; Michael Snyder headshot: Courtesy Stanford Medicine

You look in the bathroom mirror. There’s you—groggy, disheveled, bleary-eyed—and there’s you: numeric, quantified, displayed along the glass’s margins in rows of illuminated data. Stats from your smartwatch appear: time spent in deep sleep, REM and light sleep; skin temperature and emotional arousal as gauged by sweat gland activity; blood pressure and oxygenation; and, of course, the hustle and bustle of your heart. A chip beneath your skin transmits info on glucose, cortisol, ketones, cholesterol and inflammation levels. You prick your fingertip, put it on a sensor, and more biometrics scroll along the glass, all in blue, none in red to flag an anomaly.

This is the future that Michael Snyder, director of the Center for Genomics and Personalized Medicine and chair of the department of genetics, wants for everyone on Earth. He’s already cobbling it together for himself. Each morning, he straps on four smartwatches and an Exposometer to measure levels of airborne particles. He has a continuous glucose monitor for his blood sugar and an Oura ring to track his sleep. 

His smartphone maps his locations and quantifies his movements, and his smart scale measures his weight, body-mass index, body fat, body water, muscle mass and bone mass. He has also compiled a decade of lab data—on his genome (the sum of his genetic code), his epigenome (the markers regulating gene expression), his transcriptome (RNA transcripts of expressed genes), his blood proteome (the proteins in his blood), his urine metabolome (the molecular by-products of metabolism), his blood and urine lipidome (the fat contained therein), and his microbiome (the microbial organisms living in his mouth, sinuses, skin and gut). In essence, he has defined his Snyderome. The ever-more-exhaustive portrait he has been composing of himself includes immune profiling, hormone measurements, and 12 whole-body MRIs over the past five years, which show his physical structure in minute detail. The data from his body alone amounts to two petabytes—the storage capacity of 2,000 top-of-the-line iPhones or 31,250,000,000 Apollo 11 moon-landing computers. “And, you know, I’m only going to add to that,” Snyder says. At 66, he plans on squirreling away data to the grave and almost certainly enabling it to
be gathered afterward.

Photo of Michael Snyder in a lab.Michael Snyder (Photo: Lee Abel)

 

Of course, Snyder has set his sights on quantifying more than his own self. He is using himself as a test subject to iron out the bugs for a larger study he has been running on more than 100 people for the past 8 ½ years. In gathering more data on his biology than any other known human, he has predicted his own diabetes, detected its emergence and calibrated his lifestyle to mitigate it. On a plane from Germany to Norway to visit his wife’s family, he saw his blood oxygen drop far more than it usually does when he flies. This, combined with an accelerated heart rate and a recent stint in rural Massachusetts, made him wise to a Lyme disease infection. And for those in his study, his lab has discerned the earliest evidence of numerous ailments. Snyder believes that this big-data approach to health, built on longitudinal measurements (those taken regularly over long periods of time), will allow individuals to know the unique biometric signatures of their own health and detect changes the moment they occur. “Medicine is broken in a lot of different ways,” Snyder says. “It’s very focused on treating people when they’re ill—very reactive, very costly. We should be focused on keeping people healthy, but we have to understand what it means to be healthy.” 

What Snyder wants everyone—patient and doctor—to grasp is that the biometric signature of health can look different from person to person. The current system, which evaluates whether someone is healthy according to norms averaged from the larger population, fails to account for how much variation exists among humans. The solution he envisions will require better smart wearables, better testing and better algorithms to crunch the vast data from a variety of “omes.” It will require start-ups to innovate technologies and make them accessible. And it will demand an overhaul of health care so that people can be alerted when their biometrics change and a doctor can investigate what has gone awry. “In the future it can all be done just like driving a car,” Snyder says. “Your car has lots of sensors that make it run smoothly, and most of the time you just jump in and drive it. That’s the way health monitoring will be. It’ll happen in the background and then, when problems come up, it will warn you, long before symptoms.”

Comparing Apple-omes and Orange-omes

As a geneticist, Snyder’s early contribution to the field was as a professor working on the yeast genome at Yale between 1986 and 2009. During that period, the evolving field of genetics research was busy defining single genes. “It was kind of one gene, one PhD,” he says. “Where I first made a mark was in saying that this doesn’t make sense.” He wanted to move away from the reductionist approach and instead integrate information into the larger biological whole. To do this would require seeing how genes worked in conjunction with each other to influence biological traits. Eventually, he found a way to study all 6,000 yeast genes at once. 

Increasingly, Snyder wondered why the same systems approach wasn’t taken with health. When he joined Stanford’s faculty in 2009, he went for a medical checkup. “They gave me back my usual 10 or 15 measurements,” he recalls. “I wasn’t even sure how useful some of those things were.” He found himself thinking how ridiculous it was not to profile people more deeply. “That became the goal—to apply the same ideas from systems biology to systems medicine, if you will.” He and his team set up his lab to analyze metabolic by-products, sequence DNA and deploy RNA-Seq, a technique Snyder invented at Yale that is now used worldwide to sequence RNA and show not only which genes are being actively expressed but also to what degree. “We weren’t sure which technologies would be most powerful and what we might learn from them,” he says, “but the only way to know for sure was to get started.”

Thus began Snyder’s new vocation as a guinea pig. He sequenced his own genome in 2010 and identified variants that predisposed him to type 2 diabetes. The lab tested his biomarkers frequently, especially when he had a cold, which allowed them to detect a sudden rise in his blood glucose a year later, after a viral respiratory infection. In fact, after each subsequent viral infection (frequent, given that he had small children), he saw dramatic changes in the transcription of his genes. He then took a good look at his epigenome (the markers controlling gene expression). If individual genes were likened to book chapters, the epigenome (epi- meaning “above”) would be those sticky neon index tabs marking individual lines for quick reference. The invisible hand of each viral infection had reordered some of those Post-its. Most affected were the genes responsible for regulating metabolism.

‘I didn’t know that I’d be interesting. It turns out everybody’s interesting.’

In 2012, Snyder published “Personal Omics Profiling Reveals Dynamic Molecular and Medical Phenotypes” in Cell about the onset of his diabetes, attracting significant media attention to his self-measuring project. “I didn’t know that I’d be interesting,” he says. “It turns out everybody’s interesting.” In fact, in the trial he has been running on more than a hundred people since 2013, 49 participants have discovered a health problem—early signs of lymphoma, diabetes, heart issues—from a variety of monitoring approaches, including genome sequencing, blood panels and wearable sensors. 

Studies such as the one Snyder did on himself—known as N-of-1 trials, in which N refers to the number of participants—were rare, at least in scientific journals. The norm has long been to include numerous participants in trials to determine mean and median values for health metrics. Snyder’s view is that N-of-1 trials can explore each individual as a system of biological data points related to their genome and environment. Underpinning this idea is how much variation the human genome contains. The oft-mentioned fact that two average human genomes are 99.9 percent similar risks obscuring that 0.1 percent of the genome’s 6 billion base pairs “represents a substantial amount of DNA,” Snyder writes in his 2016 book, Genomics and Personalized Medicine. Already, hundreds of millions of genetic variants have been found in humans. “If a martian were to land on Earth and look to people, they’d say we’re all the same,” he says. “People have two arms and two legs and two eyes. At some gross level, we’re all the same, but at other levels we’re really quite different. Our behaviors are different. Our health risks are very, very different. The way we age is very different. On one hand, you might say, ‘Well, that’s just the difference between living to be 60 or 90.’ Maybe to martians, that’s not a big deal, but I’d rather live to 90 than to 60.”

While a quick glance about in a crowd of humans will reveal that we’re not a bunch of clones, the eye fails to see how many variations the human blueprints contain and how many factors influence the execution of those blueprints. “It’s not just the genes but also the environment playing on those genes,” Snyder says. Two 19th-century English figures, randomly pulled from history’s archives—say, Jake Apple, who migrated to Arkansas in 1886, and Benjamin Orange, who settled in Philadelphia in 1820—might face significantly different environmental factors: food, work, climate, stressors, air quality and pathogens. Rural or urban settings would account for vast differences in biological markers. The bacteria colonizing the two men’s bodies would differ. Their social statuses would influence not only their nutrition but also their stressors. And individual behaviors and proclivities could hardly be ignored: one a teetotaler, the other a tippler. Genetics aside, the Apple-ome would hardly resemble the Orange-ome. But unlike in 19th-century America, today we can discern the reasons they are different; the tech revolution keeps giving us ever-smaller, more powerful processors, the capacity to crunch big data, an expanding universe of smart devices and, for many, the ambitions to measure every detail of our body’s earthly existence. 

Ome Sweet Ome 

With the publication of his 2012 paper on the Snyderome, Snyder found himself courted by a growing cultural movement intent on measuring the self. Though people have been using self-tracking devices for exercise since at least the 1970s, this trend accelerated in recent decades. In 2007, Wired editors Gary Wolf and Kevin Kelly proposed the term quantified self for this expanding movement of people measuring everything from sleep, mood, posture and caloric expenditure to their genome, microbiome and metabolome. Inevitably, the QS movement has branched into the quantified baby, with parents rigging newborns up with wearables, taking helicopter parenting to a new level. QS even has a philosophical bent—the “know thyself” attributed to an array of ancient Greek sages, and the “care of the self” emphasized by French philosopher Michel Foucault, who argued that self-awareness is not an end in itself but a means to living better. In Wolf’s 2010 TED Talk, he emphasized the importance of gathering knowledge of the self, saying, “If we want to act more effectively in the world, we have to get to know ourselves better.”

We are currently far from this dream of understanding ourselves and our health, Snyder believes. “Imagine you have a 1,000-piece puzzle, and you grab 10 pieces and try and figure out what the picture is,” he says. “That’s what medicine is today.” Moreover, those 10 jigsaw tiles are evaluated according to the average measurements from other puzzles. Snyder gives the example of body temperature. “You’re told that your temperature, when you put a thermometer in your mouth, is 98.6,” he says. “It turns out, first of all, that it’s more like 97.5, but more importantly there’s a spread. The 25th percentile will be 94.6. The 75th percentile will be 99.1. That means that if your normal, healthy baseline is 94.6, and you walked into a physician’s office today and they measure you at 98.6, they’ll tell you you’re healthy. But you’re off by 4 degrees, and I guarantee you’re not healthy.” There are many such examples, Snyder points out—cases in which one person’s healthy measurements can differ dramatically from another’s, or in which a single metric can spike from the bottom to the top of the “normal” range. Since medicine views both measurements as healthy, the changes are often ignored. Absolute values, Snyder believes, are far less useful than deviations. “Understanding your healthy baseline is important for everyone, so you can detect those shifts,” he says. “That’s how you find problems.”

If measuring oneself in ever-greater detail seems increasingly possible, it’s because of the plummeting cost not only of tech but also of medical testing. Whereas sequencing the first whole human genome required more than 10 years, nearly 3,000 scientists and $2.7 billion, sequencing a genome now takes a day and a few hundred bucks. Snyder sees testing moving in the same direction, no longer demanding visits to doctors and hospital labs but done on the fly by mail or via Amazon, which already sells tests for STIs, food sensitivity, metabolism, allergies, thyroid function and more. “The Theranos concept was right, but it was obviously poorly executed,” he says, referring to the former company that claimed it could perform rapid, automated testing with tiny amounts of blood. “We’ll have home tests, like for pregnancy, and they’ll get back half a dozen measurements. That’ll happen in the not-too-distant future, and you’ll get a more detailed panel if you prick your finger and mail that off to a testing lab, like a Quest. They’ll give you back maybe 200 measurements,” Snyder says. “It’s going to be a matter of measuring people with reasonably high frequency. People’s blood measurements are done in a physician’s office right now, usually once every few years when you’re healthy. I think that can be moved up to once a week for some measurements, and wearables, of course, will measure 24/7.” Similarly, Snyder believes that with 1 in 10 Americans having diabetes and 1 in 3 being prediabetic, continuous glucose monitoring with a sensor under the skin would enable people to adjust their diets accordingly. In 2018, he published a paper showing that there are different glucotypes based on which foods cause people’s blood glucose to spike. (Caveat lector—cornflakes and milk caused glucose to surge into the prediabetic range in 80 percent of study participants.) “I think food monitoring will become more commonplace,” Snyder says. “It’s very complicated to do it now, but it will be done much more efficiently and easily in the future.” 

‘Understanding your healthy baseline is important for everyone. That’s how you find problems.’

Part of the challenge facing a big-data vision for health care is a belief common among medical professionals that patients shouldn’t know everything—for instance, that a detailed analysis of their genome could give them undue reason to fear problems that may never arise. “Usually, the general principle in medicine is not to do measurements or tests unless we really know what we’re going to do with the answer, so open-ended testing is kind of rare,” says Atul Butte, professor and director of the Bakar Computational Health Sciences Institute at UCSF and chief data scientist for the University of California Health System. “It’s amazing to see how much Mike’s work has influenced others to measure more things on more people at more time points and even longitudinally to see if we can detect signs of disease earlier.”

Just as Snyder encourages everyone to understand their genetic information, he advocates broader data gathering with magnetic resonance imaging, which doesn’t use radiation but rather polarizes the water molecules in the human body to create precise images of its makeup. “In today’s world, any physician will tell you don’t do a whole-body MRI, and that’s the wrong view in my opinion,” Snyder says. Doctors prefer not to do such scans because they will inevitably turn up evidence of nodules—abnormal growths present in virtually all bodies—and this will result in unnecessary panic and expense. “If you make multiple measurements longitudinally—one now, one in three months or six months—you can see if any nodules are growing,” he says. Snyder so deeply believes in the importance of doing such MRIs that he co-founded Q.bio, a company focused on deep profiling that has developed a whole-body MRI scanner. Its goal is to build a platform that “can comprehensively measure and analyze changes in the human body—not limited to a single type of information, but any information that could be collected about a body,” says Jeffrey Kaditz, co-founder, CEO and CTO, who approached Snyder in 2016. “When I had done research about who in the world had been studying biochemical changes in people within the context of their genetics, [Snyder] was the pioneer in that field,” Kaditz recalls. Once fully up and running, Q.bio intends to cheaply perform comprehensive physicals—combining MRI and molecular data—to facilitate in-depth health monitoring. 

For Snyder, combining omes allows people to discover their risks and monitor them carefully, as he did with diabetes. This approach is valuable even with specific diseases, such as COVID-19, he says, since DNA analysis can show which people are most likely to develop severe COVID. Snyder’s lab has also been looking at data from participants’ Fitbits and Apple Watches to determine if asymptomatic COVID could be detected, and in 10 out of 14 cases, the health data showed noticeable deviations. In fact, Snyder’s family speculated that he might deliberately contract COVID just to see how it changed his biomarkers. (He refrained.) But for a person’s broader, lifelong well-being, frequent measurements could alert them to problems long before they become serious. “We’ll have people living much better, healthier lives,” he says, “and save a lot of money in the process.” 

The Ome Stretch 

Snyder’s lab—among the biggest at Stanford, with sometimes more than 100 people at work—has broken so much new ground that it’s easy to forget that his vision for medicine is a largely unexplored frontier. For instance, only two years ago, the lab found that people aged distinctly along different pathways, such as metabolic, immune, hepatic (liver) and renal (kidney). Whereas someone who is primarily a metabolic ager might have rising blood sugar, an immune ager might have elevated inflammation. And whereas one person might primarily age in one pathway, another—say, Snyder—could be aging in all of them. (“I’m a pretty typical ager,” he says.) 

Many challenges face Snyder’s vision of longitudinal health monitoring. For instance, the microbiome is still incompletely understood and “not very clinically actionable,” he says. As for the other omes, the cost of assessment remains prohibitive for most people. “The technology is not yet there,” he points out. This is as true of wearable devices and rapid testing as it is of the information processing necessary to sieve, condense and store so much data. “We will have to distill it into the more valuable components because it’s very expensive to store two petabytes of data,” he says. 

And where there’s talk of data, there are worries about privacy. Snyder believes this concern is overblown because the data can be encrypted and safely stored. “Everybody has a credit card, and that has incredibly personal information, stuff that’s probably just as sensitive—maybe even more sensitive—than your health data.” People use their cards for pharmacy purchases and a host of other activities that they want to keep private, but do so out of convenience. “Nobody wants to walk around with bags of cash,” he says. He believes the health benefits of measuring oneself outweigh the risks. A future like that depicted in the 1997 film Gattaca, in which people’s careers and opportunities and social statuses are determined by their genetics, strikes him as unlikely. He acknowledges that in some cases people who have genetic disease risks might want to choose their careers or lifestyles accordingly, such as those with hypertrophic cardiomyopathy, a common cause of sudden death in young athletes. But Snyder isn’t concerned about the possibility that businesses or institutions would use genetic information against people: “The government should step in and block that with legislation.” 

The biggest challenge is perhaps the health care industry itself—highly regulated and set in its ways—and the question of who will pay for all the monitoring. “We don’t have good incentivized plans for keeping people healthy,” Snyder says. “The whole payment plan and reimbursement plan for health in general is really geared towards sick care, not health care.” For instance, though repeated whole-body MRIs appear prohibitively expensive, they might catch cancer before it spreads, allowing it to be treated with a relatively simple surgery and preventing a drawn-out illness that would be far more costly.

Eric Topol, a professor of molecular medicine at Scripps Research and the founder and director of Scripps Research Translational Institute, agrees that medical care is ready for a shake-up. “You get your blood pressure taken at one visit while you’re in the contrived situation of visiting a doctor at a medical facility”—and that’s after you’ve navigated the parking lot and languished in the waiting room for an hour, he points out. “The whole idea of being able to get lots of measurements in the real world passively and accurately is upending medicine, but it’s not accepted easily because it challenges the control nature of physicians and the paternalism.” Yet Topol doesn’t see mainstream health care adopting the new approach anytime soon. “There’s no real traction out there—not yet. But it’s inevitable. It’s probably a number of years off.”

Snyder has been working on what he sees as the key to larger adoption of an “omics” approach to medicine: scaling it. “Academics are really good about proof of principle and discovery,” he says. “The way you scale things for the planet is to form a company.” So far, he has been involved in launching 13—for genome sequencing, metabolic profiling, glucose monitoring, smart watches and more. One of the start-ups, Personalis (Atul Butte is among the co-founders), sequences cancer DNA to support the development of personalized cancer vaccines that will help an individual’s immune system target their specific type of cancer. Just as with health, it turns out, disease often looks different from individual to individual.

‘We’ll have people living much better, healthier lives, and save a lot of money in the process.’

On the long road to his envisioned future, Snyder keeps measuring. His Exposometer has helped him study the air he breathes—an increasing concern in an age of rampant wildfires. He has catalogued particles from thousands of plant species (he correlated his mild allergies with eucalyptus exposure, whereas he previously thought the cause was pine) and thousands of artificial compounds. “Certain carcinogens are everywhere,” he says. “Pesticides are everywhere.” For a while, he also wore a Geiger counter to detect gamma radiation, which, as expected, gave higher readings on airline flights and in the hills of Montana. Once, in a coffee shop, it began buzzing louder than ever before. “Clearly, somebody in there was undergoing some sort of radiation treatment,” he says. 

More often, though, Snyder takes his measurements of others intentionally. He participated in analyzing the data from the astronaut twin study, in which Mark Kelly stayed on Earth while his twin, Scott Kelly, spent a year on the International Space Station. Snyder is also preparing an extensive analysis of two men during an 80-day crossing of Antarctica. And he has recruited 1,000 Pac-12 athletes to detect COVID-19 with wearables. Meanwhile, others have also been busy doing their own measurements—of Snyder, given that he has made his data available to the larger public. At a conference, a scientist approached him and broke the ice by saying that he believed Snyder’s metabolic dysregulation was in part caused by defective mitochondria. 

Most days, however, the ever-expanding data set that is Michael Snyder goes about his work on campus with little fanfare. To manage his diabetes, he recently switched from running to weight lifting, since muscle mass is known to regulate glucose levels. A whole-body MRI has since confirmed that he has gained 10 pounds of muscle. He continues to track his daily metrics closely to see how he is influenced by the air he breathes, the food he eats, the pathogens he encounters and the many other elements that make up his days. “That’s the power of the quantified self,” he says. “Nobody can track you better than you.” 


Deni Ellis Béchard is a former senior writer at Stanford and the author of eight books. Email him at stanford.magazine@stanford.edu.