Wearables, AI, and the Future of Personalized Health

Michael: If you look at this guy, he's in great shape—he's 76. And then one day, he died right after a Peloton run. His wife generously donated all his data, so we had his Apple Watch data. It turns out, as we looked at this data going back over the last five years, there's clearly a very significant shift.

Michael: About four months before he died.

Brent: Welcome to Death Clock. I'm your host, Brent Franson. Today we speak with Dr. Michael Snyder about personalized medicine. Dr. Snyder is the Stanford Ascherman Professor and Chair of the Department of Genetics at the Stanford University School of Medicine. He's the Director of the Center for Genomics and Personalized Medicine at Stanford, and he previously served as the Director of the Center for Genomics and Proteomics at Yale University.

Brent: He's quite an interesting researcher. He has spent his career gathering as much data as he can about himself. He has over 2 million data points on his own health that he's tracked over the past 14 years. He used those data points to diagnose himself with type 2 diabetes and treat himself. For the last 13 years, he has also been running a study where he has collected 135,000 data points on each participant—from their blood, microbiomes, and full-body scans—to track their health while healthy and better predict disease.

Brent: He's a great guest to walk us through what we want to know about our own health, how all of this relates to understanding metabolic health and type 2 diabetes, and how AI is going to shape the way we predict disease in ourselves going forward. Great guest. Hope you enjoy!

Brent: Dr. Michael Snyder, welcome to the show.

Michael: Pleasure to be here.

Brent: I'm excited to chat with you about your work in genomics and all of the self-tracking you've done. But before we get into that, can you give us a sense of your background and day job?

Michael: Yeah. I'm on faculty at Stanford in the Department of Genetics. I run the Center for Genomics and Personalized Medicine, and our lab does a lot of big data research, which is how we got into this area.

Brent: What do you mean when you say personalized medicine? What does that mean to you?

Michael: For us, it's really more personalized health. Believe it or not, we're actually profiling people while they're healthy and trying to keep them that way. Everybody is different; we all have different baselines and profiles. I can give you a good example: you're used to thinking everyone's body temperature is 98.6°F. But if you measure people's temperature with a thermometer in their mouth, first of all, that number is often wrong.

Michael: It's more like 97.5°F. But the more important point is that there's a range; we all have different baselines. Why is that important? If you go to a physician today and they measure 98.6°F, they think you're healthy. But if your baseline is 94°F, you're up four degrees and not healthy. So we're big believers that you should know your healthy baseline.

Michael: Then you can detect shifts, catch disease early, and keep people healthy. That's the goal. We also respond to foods differently—we respond to so many things differently that knowing these things on an individual level is very important.

Brent: It seems that we're trying to transition from Medicine 2.0 to Medicine 3.0, but we group things for efficiency reasons. We say around 98 degrees is the right temperature, and if you're above that, maybe you have a fever. There's utility in that because it's difficult for everything to be caveated at the individual level.

Brent: So how practical is personalized medicine in the context of our healthcare system? How do we get from where we are today to there?

Michael: Great question. First of all, we can get there. Medicine has simply been trained to treat people on population averages rather than at the individual level. I can give you an example from our research: medicine treats diabetics as two main classes—Type 1, which comes early and involves autoimmune disease, and Type 2, which is everybody else.

Michael: Our lab has looked into Type 2 and broken it down. There are actually several subtypes of Type 2: some people are muscle insulin resistant, others have beta cell defects where they don't release enough insulin from the pancreas, and some don't produce incretin hormones properly.

Michael: There are also hepatic insulin-resistant subtypes. You might ask, "So what?" Well, the way people respond to food is very different. In the US, 11.6% of people are diabetic and 38% are pre-diabetic—meaning half the population is either pre-diabetic or diabetic.

Michael: Those numbers are getting worse, yet we treat everybody the same. Have you heard of continuous glucose monitors (CGMs)? They measure glucose every five minutes. I recommend everyone wear one—you'll never eat the same way again.

Michael: I mean that in a good way, because everybody responds to food differently. Some people spike their glucose from potatoes, others from pasta, white bread, or brown bread. We're all very different. How does that relate to diabetes subtypes? We've subtyped Type 2 diabetes into these different forms.

Michael: Depending on your subtype, specific foods cause spikes. If you're muscle insulin resistant, you'll spike from potatoes and pasta, but not if you're insulin sensitive. Likewise, if you have a beta cell defect, you'll spike from potatoes. How we respond to food—and drugs—depends on what's occurring under the hood.

Michael: I'm fairly thin, but I'm a Type 2 diabetic with a beta cell defect—I developed it after a viral infection. It turns out I don't respond to Metformin, which is the most common drug prescribed, but I do respond to a medication that promotes insulin release from my pancreas.

Michael: The point is that how you respond to food or drugs depends on your form of diabetes. That's why breaking this down matters. Today, we treat people uniformly, but if you knew what was going on, you could adjust lifestyle or prescribe medications far more precisely.

Michael: We can reach this point inexpensively. These days, you can buy an over-the-counter continuous glucose monitor for around $50. With new AI methods, we can tell you your subtype for several of these forms.

Michael: We can determine what subtype you are for many of these conditions.

Brent: What is the difference? If you take the average person in the population, they probably don't know their fasting glucose—the standard diagnostic method for prediabetes and diabetes. Fasting glucose between 100 and 125 mg/dL indicates prediabetes, and 125+ indicates diabetes.

Brent: Others might say fasting glucose isn't optimal—hemoglobin A1C gives a three-month view. An A1C above 5.7% indicates prediabetes, and 6.5% or above indicates Type 2 diabetes. Some might advocate for an oral glucose tolerance test instead.

Brent: You're taking it further by suggesting wearing a glucose monitor to identify specific subtypes within Type 2 diabetes. If I get a blood test showing an A1C of 6.0% (prediabetic), compared to a parallel scenario where I wear a CGM and Dr. Snyder determines my subtype, what is the practical difference in how I manage my health?

Brent: What what is the difference between those two worlds in terms of how I'm treating myself?

Michael: Hemoglobin A1C gives you an average glucose level, which is useful because it alerts you to prediabetes. However, a glucose monitor shows you exactly which foods cause glucose spikes so you can avoid them, while continuing to eat foods that don't spike your levels.

Michael: Using the same drugstore CGM, we can subtype you. If you're muscle insulin resistant, you should watch out for potatoes or take a brisk 15-minute walk right after eating them.

Michael: We're also researching ways to mitigate spikes through meal order. You may have heard that you should eat your salad before your French fries; eating protein and fiber before carbohydrates helps prevent sharp spikes.

Michael: Again, it depends on your subtype. Knowing your subtype guides what you should do. Wearing a CGM directly is great, but combining it with AI coaching models is even better.

Michael: I co-founded a company called January AI that builds predictive models using a 54-million-food database. They pair with a CGM—or even work without one—to predict what you should or shouldn't eat based on individual parameters.

Michael: We are moving into an information-rich world. Where people used to rely solely on search engines, they now use conversational AI for useful insights. People are becoming comfortable with data-driven health, and wearing sensors provides real-time feedback. If you drink a milkshake, you can watch your glucose rise immediately.

Michael: And I think people are getting more comfortable with this information based world. And I think, you know, putting these monitors, these sensors on yourself, which are very easy to do. And the other reason these glucose monitors are so powerful is their real time feedback. Meaning you eat that McDonald's shake and you will watch your glucose go through the ceiling.

Michael: And that is very impactful—I can tell you from personal experience.

Brent: Do you feel the average person understands the difference between a normal glucose response and an abnormal spike? Glucose spikes after a meal, even a healthy one. If I eat a plain salad, does it spike at all?

Brent: When I wore a CGM, my glucose spiked after eating and I concerned myself unnecessarily. A previous guest mentioned that spikes under 140 or 180 mg/dL may be normal physiological responses rather than a concern.

Brent: What are your thoughts on the psychology of glucose monitoring and helping people distinguish a healthy post-meal rise from a harmful spike?

Michael: Two comments on that. First, if a salad spiked your glucose, it was likely the dressing containing added sugar rather than the greens, which have very few carbohydrates.

Michael: A New York Times reporter called me while wearing a CGM and said, "Mike, I eat salmon on salad every day for lunch—no carbs! Why am I spiking through the roof?"

Michael: It turned out to be the salad dressing. Simply leaving that out gave him the healthy lunch he intended. That real-time feedback is what makes CGMs effective. As for normal levels:

Michael: When you're young, post-meal spikes are quite small. Ideally, you want to maintain a tight target range below 140 mg/dL.

Brent: So if I'm wearing a monitor and spike above 140 mg/dL, I should pay attention and consume less of whatever caused that spike?

Michael: Exactly. As glucose regulation declines with age, clinical guidelines often use 180 mg/dL as an upper target for diabetics and pre-diabetics because 140 mg/dL can be challenging to achieve consistently.

Michael: It's helpful to wear a monitor periodically to see which foods cause the largest spikes and adjust. Subtypes can shift over time, so checking periodically makes it easy to avoid trigger foods.

Michael: I once ate a pulled pork sandwich and my glucose shot over 300 mg/dL. I showed it to a friend, who said, "Mike, everybody knows barbecue sauce has tons of sugar in it!"

Michael: I didn't realize it at the time. Many of these insights seem obvious in hindsight, but CGMs serve as an effective, real-time teaching tool.

Brent: Many people view Type 2 diabetes as reversible through diet, lifestyle, and medication. However, as someone who studies this and tracks health data meticulously, you refer to yourself as having Type 2 diabetes rather than having reversed it. Is there a reason you avoid using the language of reversal?

Brent: You're somebody who studies this who's obviously, you know, you're you're spending a lot of time thinking about your own, your own health, and you're tracking everything. And you say, I have type two diabetes. Is there a reason that you're not using the reversible language or you haven't reversed that yourself? Yeah.

Michael: Great question. Early on, my genome predicted a high risk for Type 2 diabetes, though I wasn't diabetic at the time. I developed it following a viral infection—which was one of the first demonstrated cases of viral-triggered Type 2. As mentioned, I have a less common beta cell defect subtype.

Michael: Because I track myself closely with smartwatches and smart rings, I try to understand what triggers physiological shifts. After that viral infection, my glucose levels rose significantly.

Michael: We now know that 2% to 4% of people who get COVID-19 develop Type 2 diabetes afterward. We want to study which specific subtypes they develop. People should stay vigilant about post-viral metabolic changes.

Brent: So you don't view your Type 2 diabetes as reversible?

Michael: Initially, I brought my glucose levels down to normal through running. But when I stopped running and caught a second viral infection, my glucose rose again. I managed it partially, but levels gradually crept up with age.

Michael: I shifted from running to weightlifting and gained 10 pounds of muscle mass. Although I did everything right, it had no effect on my glucose because of my subtype: I have a beta cell defect, meaning my pancreas doesn't release enough insulin. Gaining muscle mass doesn't resolve that specific issue.

Michael: Knowing my subtype earlier would have been valuable. I tried Metformin first, but I was a non-responder. I kept increasing the dosage with no effect. Eventually, I switched to a secretagogue (like glipizide) that stimulates insulin release from the pancreas.

Michael: That worked well once I understood my subtype. Currently, I use GLP-1 receptor agonists, which work very well for me, though they cause minor indigestion. If I could manage it entirely through lifestyle, I would prefer that.

Brent: When you say controlling it, you mean standard measures like blood glucose or A1C, right?

Michael: Yes, A1C and CGM metrics like "time in range"—aiming to keep glucose between 70 and 180 mg/dL.

Brent: So exercise brought your A1C down initially, but Metformin didn't, and levels rose again when exercise decreased?

Michael: I lowered it as far as lifestyle allowed, but it gradually crept back up. When my A1C reached 6.3%, Metformin failed to lower it. Once it reached 6.6%, I transitioned to targeted medication.

Brent: Was the viral trigger an epigenetic mechanism? You had a genetic predisposition that was dormant, and the infection triggered gene expression changes or an autoimmune response?

Brent: What role did the viral infection play?

Michael: You're spot on—it was epigenetic. I was genetically predisposed with variants putting me at high risk, but I didn't become diabetic until the viral infection. DNA modifications alter how the genetic code is expressed, similar to how an orchestra interprets a musical score.

Michael: Many of my metabolic genes showed altered DNA methylation after that viral infection, and again after the second one. These changes altered gene expression and triggered diabetes.

Michael: Post-viral shifts are likely more common than realized. For instance, myalgic encephalomyelitis / chronic fatigue syndrome (ME/CFS) affects roughly 2% to 4% of the population and often follows an infection.

Michael: Long COVID is another example. Severe viral infections—in my case, an RSV infection with a high fever that kept me in bed for days—can leave lasting physiological changes.

Michael: And I was in bed for a few days, which is very unusual for me at a high temperature.

Brent: We recently discussed ME/CFS and long COVID on the show. Initially, I received the COVID-19 vaccine but was hesitant about boosters, feeling I wasn't in a high-risk group for severe acute illness.

Brent: However, learning about long COVID and chronic fatigue syndrome changed my perspective. I decided to stay updated on boosters not out of concern for the acute infection, but to avoid long-term post-viral complications.

Brent: I want to avoid potential long-term systemic impacts, such as epigenetic changes that could trigger metabolic disease.

Brent: I hadn't heard of that.

Michael: COVID-19 was significantly more severe than the flu, and many individuals suffered acute complications or death. The long-term effects, like long COVID, are substantial.

Michael: While rare vaccination complications exist, far more people experienced severe complications from COVID-19 itself. Playing the odds, protecting yourself against long-term outcomes makes sense.

Michael: You did the right thing.

Brent: I was never skeptical of the vaccine technology itself, but I initially questioned whether a healthy 43-year-old needed ongoing boosters. Understanding these chronic, long-tail risks convinced me otherwise.

Brent: Yeah.

Michael: Estimates suggest around 10% of COVID-19 cases lead to some degree of long COVID, with fatigue being the most common symptom. While symptoms can lessen over time, they can be persistent and severe.

Michael: I co-founded a company called Rhythmia that focuses on treatments for long COVID and chronic fatigue. These conditions affect millions of people and can be debilitating.

Brent: This brings us to your research on aging surges. I'm 43, and your published work highlights distinct molecular shifts or aging peaks occurring around ages 44 and 60. Could you explain what is happening during these periods?

Brent: What's happening. Why at those particular times.

Michael: In our study, running for nearly 13 years, we collect deep longitudinal data on healthy individuals: sequencing genomes and analyzing thousands of molecules in blood, urine, and stool micro-biomes.

Michael: We track wearable device data and take roughly 135,000 measurements per participant quarterly to define healthy baselines and track changes over time.

Michael: Profiling healthy individuals allows us to detect health conditions presymptomatically before clinical symptoms emerge.

Michael: We caught early lymphoma and serious cardiovascular issues presymptomatically. Everyone ages differently—we identified distinct "ageotypes," such as metabolic, immune, or cardiovascular agers (or combinations thereof).

Michael: Instead of aging being entirely linear, we analyzed whether molecular changes occur in non-linear crests or waves. People often spend their final decade in ill health, so our goal is to extend healthspan.

Michael: We found two major wave periods where significant biomolecular shifts concentrate, rather than occurring at a completely steady pace.

Michael: The shift in the 60s was expected: we observed changes in muscle mass, skin elasticity, immune function, kidney markers, and oxidative stress pathways.

Michael: The other distinct wave occurred around age 44. This represents a broader crest rather than an exact birthday cutoff.

Brent: My 44th birthday is in ten days!

Michael: Hit this big burst and change. I think you're probably.

Brent: Okay around that.

Michael: In our mid-40s data, we observed pronounced shifts in muscle tissue, skin, and lipid metabolism—which explains why body fat distribution often changes around this age.

Michael: We also observed shifts in caffeine and alcohol metabolism pathways. Personally, I noticed I had to stop drinking caffeine by noon in my 40s to avoid sleep disruption.

Michael: These findings are actionable: knowing lipid metabolism shifts in your mid-40s allows you to monitor cardiovascular markers closely and intervene early, such as with statins or lifestyle modifications.

Michael: The shift in the 60s is largely driven by sarcopenia (loss of muscle mass) and immunosenescence (immune decline). That is why resistance training and staying current on vaccinations are critical as we age.

Michael: For the mid-40s wave, our hypothesis is that lifestyle shifts accumulated over the 20s and 30s—such as career stress, dietary changes, or reduced physical activity—manifest molecularly around this time.

Michael: We are investigating whether maintaining an active lifestyle can delay or flatten that mid-40s molecular shift.

Brent: What are the most important biomarkers to track? Standard primary care medicine focuses on basic lipid panels (LDL, HDL) and metabolic panels (fasting glucose, A1C).

Brent: In preventative medicine, markers like Apolipoprotein B (ApoB) offer clearer insight into cardiovascular risk than LDL alone, and Lipoprotein(a) [Lp(a)] indicates genetic risk.

Brent: We evaluate genetic predispositions using ApoE testing for Alzheimer's risk. However, extensive testing in asymptomatic individuals is controversial in traditional medicine due to potential over-testing.

Brent: Where do you find the right balance between necessary testing and excessive paneling for asymptomatic individuals?

Michael: I believe comprehensive information helps people better manage their health. Standard checkups measure around 15 markers, while concierge testing services measure a few hundred for a clearer picture.

Michael: In our research using 135,000 measurements, roughly half of participants discovered an actionable health item. We co-founded Q Bio to perform whole-body MRI, which some traditional physicians advise against.

Michael: I strongly support whole-body MRI because establishing a baseline allows you to distinguish benign incidental findings from active nodule growth over time.

Michael: Tracking growth over time is key. Establishing regular baseline screening helps catch aggressive conditions early before metastasizing.

Michael: So you need to pick that up early before it's big and metastasize, because then it's too late.

Brent: Full-body screening carries a risk of false positives and unnecessary invasive follow-ups. For instance, lung nodule biopsies carry procedural risks like pneumothorax, which can increase overall morbidity in heavily screened populations.

Brent: Invasive procedures to investigate incidental findings carry inherent risks. We must ensure secondary screening does not create a false sense of security or replace gold-standard procedures like colonoscopies or mammograms.

Brent: Advanced testing, such as lipid fractionation for LDL particle size, often does not alter clinical management beyond standard ApoB or statin protocols according to preventative cardiologists.

Brent: How do you navigate these complexities—such as handling incidental findings or medical anxiety—for patients who may not know how to interpret ambiguous results?

Brent: I don't need that lipoprotein fractionation. And so I think I would directionally agree with you that more is better. But there is some complexity. There's that complexity around complacency. There's probably a lack of education for most people on. All right. What I'm not going to really know what to do. And my doctor's not going to know what to do.

If they know the particle size that's any different from how they're going to treat me based on my apob. So how do you think about all of that nuance? And you've got a patient base that's, I don't know, just not that informed.

Michael: You raise valid points. Patient and physician education is essential as medicine shifts toward an information science. While lipid particle size isn't universally used in standard practice, specialized clinical services utilize it effectively.

Michael: We co-founded a platform called Halo Health that analyzes 650 blood metabolites from a single drop to evaluate 20 health categories, such as oxidative stress and systemic inflammation.

Michael: AI can synthesize complex biological data into clear, personalized, actionable recommendations without overwhelming patients, helping them improve health markers over time.

Michael: These markers are scientifically validated. As testing costs drop, continuous tracking at home using consumer wearables will become increasingly accessible.

Michael: Continuous tracking via smartwatches and smart rings can detect early infection. Our algorithms flag early COVID-19 onset around 80% of the time prior to symptom presentation due to elevated resting heart rate.

Michael: Early critics argued that personal genomics would create widespread health anxiety or burden healthcare systems, but that hasn't materialized. Discovering mutations like BRCA allows for proactive cancer screening.

Michael: High-risk individuals benefit from enhanced surveillance or prophylactic care. Personal data can be managed constructability to guide preventative health decisions.

Michael: I track two petabytes of personal health data. AI platforms analyze this data to surface recommendations that traditional medical reviews might overlook.

Michael: AI analysis of my dataset highlighted two specific recommendations that previous reviews hadn't flagged.

Michael: One suggestion was optimizing zinc intake for specific immune cells, which matched published clinical literature well.

Michael: Similarly, an algorithm analyzed supplement timing for a colleague taking zinc and magnesium together, flagging that they share a intestinal absorption transporter and should be taken six hours apart to avoid competition.

Michael: Algorithmic models easily synthesize specific biochemical interactions that clinicians might not recall during routine visits.

Michael: Information management is similar to a vehicle dashboard: complex underlying sensor data is summarized into simple indicators, like a check engine light, that alert you when action is needed.

Michael: Future personal health dashboards will present intuitive status indicators—green, yellow, red—to alert individuals before minor issues turn into critical concerns.

Michael: As advanced imaging like whole-body MRI becomes less expensive, broad access to preventative screening will help more people stay in optimal health ranges.

Michael: That is where I hope to see the healthcare field evolve.

Michael: And I think that's where I'd like to see the world go.

Brent: I agree that data-driven health decision-making is the future. However, commercialized testing occasionally markets non-standard asymptomatic panels—such as celiac screening without clinical indication—which raises concerns.

Brent: Incidental findings on advanced imaging can also trigger complex diagnostic workups. For example, a CCTA scan I received for preventative screening showed zero coronary plaque, but revealed an anomalous right coronary artery (ARCA).

Brent: Initial evaluations suggested a high risk during intense exercise, leading to a referral to the Cleveland Clinic for invasive cardiac catheterization to determine if open-heart surgery was warranted.

Brent: The diagnostic catheterization carried a 1-in-100 procedural risk of major complications. Thankfully, the catheterization confirmed a low-risk anatomical variant requiring no surgical intervention.

Brent: While I don't regret getting the CCTA, it highlights how asymptomatic screening can lead to complex medical pathways. Proper patient counseling on incidental findings is essential before undergoing screening.

Michael: Better patient education is crucial. Incidental nodules are common on high-resolution MRI scans—I have nine stable, non-growing nodules tracked across 22 MRIs over nine years.

Michael: Longitudinal measurement prevents overreaction by distinguishing stable findings from growing ones. Let me share a contrasting example showing the value of continuous tracking:

Michael: A 76-year-old visiting scholar in our lab was in great shape and exercised regularly. He passed away suddenly following a Peloton workout.

Michael: His family donated his wearable data (Apple Watch, Oura Ring, Peloton). Retrospective analysis showed clear physiological shifts starting four months before his death—including altered heart rate variability, resting heart rate, and gait metrics.

Michael: The data captured these changes, but no connected clinical system existed to alert him. His final workout pushed his heart rate from a normal resting baseline of 50 bpm up to 150 bpm, triggering fatal cardiac failure.

Michael: An automated alert system flagging that baseline shift four months prior could have prompted a medical evaluation before undertaking extreme exertion.

Brent: What was the specific cause of death?

Michael: It was cardiac arrest / sudden heart failure. Continuous passive tracking offers actionable indicators without inducing undue health anxiety.

Michael: Systematic monitoring helps identify significant deviations early. Continuous tracking must remain unobtrusive so it enhances health management without causing perpetual anxiety.

Michael: Understanding personal risk markers helps individuals focus on high-priority preventative actions.

Brent: We agree that longitudinal data improves outcomes. As health data tracking expands, ensuring patients understand the complexity of human physiology remains essential.

Brent: It's not as simple as just we're just going to scan you and scanning is good. You know, it's just it's not it's not quite that simple. And so I.

Michael: Human physiology is a complex homeostatic system. Once out of balance, restoring equilibrium takes careful management.

Brent: What are your thoughts on GLP-1 receptor agonists in preventative health? I recently started microdosing tirzepatide and found it reduces appetite and dietary cravings, making healthy choices easier.

Brent: How do you view GLP-1 medications for preventative health beyond clinical obesity treatment?

Brent: What? What do you think about the role of GLP ones? It's obvious in obesity, but outside of the context of obesity.

Michael: For obesity and Type 2 diabetes, GLP-1 agonists are very effective. My A1C reached 8.4% before GLP-1 therapy reduced it to 5.7%.

Michael: Beyond metabolic control, GLP-1s demonstrate benefits for renal, cardiovascular, and potential cognitive outcomes, making them strong candidates for longevity therapy research.

Michael: While multi-decade longevity data is still emerging, current clinical results are promising, making microdosing an interesting strategy.

Brent: Our modern environment surrounds us with hyper-palatable foods, refined sugars, and addictive stimuli that differ from evolutionary conditions. GLP-1 medications help restore hormonal balance against hyper-palatable environmental cues.

Michael: I agree. Modern diets contain vastly more refined sugar than historical averages, and ultra-processed foods trigger strong dopamine feedback loops that encourage overconsumption.

Michael: While lifestyle choices are ideal, pharmaceutical interventions become essential when metabolic shifts persist. However, medication cannot replace foundational lifestyle habits like exercise.

Michael: Exercise provides microvascular, cardiovascular, and metabolic benefits that no single medication provides. GLP-1 medications do not preserve or build lean muscle mass on their own.

Michael: Combining foundational health behaviors with appropriate therapeutics remains essential.

Brent: 100% agreed. Dr. Michael Snyder, thank you so much for your time and for leading the charge in personalized medicine.

Michael: Thank you for having me—it was a pleasure!

Brent: Death Clock is recorded in Boulder, Colorado, and San Francisco, California. Produced by Patrick Gudino, music by Patrick Lee, and hosted by yours truly, Brent Franson, founder and CEO of Death Clock.