Blood Sugar According to a Stanford Scientist

If you had to engineer the healthiest way to eat your standard potato, what would it be?

The ultimate goal of this is personalized nutrition.

You would've heard both Professor Snyder and I at different points when going over the why behind this incredible observation.

Yes and no. I mean, grapes are sugar. That's really what it comes down to.

Is there a threshold at which you'd say you're probably insulin resistant?

Rice is worse than ice cream. Yeah.

So the number one food that spiked people was We're now back with Michael Snyder's fourth appearance on this channel. For those who missed the other three, this will be self-contained. Professor Snyder served as the chair of genetics at Stanford School of Medicine, and he's currently the director of Stanford Center for Genomic and Personalized Medicine. He's one of my all-time favorite scientists. You could call him my scientist crush. In our first interview though, which was very popular, we talked about a study where he sub-classified different types of diabetes. It was really remarkable, but then there was a follow-up study that we didn't get to get into in Nature Medicine in 2025, "Individual Variation in Glycemic Response to Carbohydrates and Underlying Metabolic Physiology." What does that title mean? Why that title? What was the goal?

Yeah, great to be here. So as you mentioned, we found different subtypes, if you will, of prediabetes and diabetes, or glucose dysregulation, we like to say. Musulin resistance, beta cell defects, incretin, these are these GLPs, and hepatic, meaning liver insulin resistance. And in this study, we wanted to match it up with people's responses to foods. Been known before that people do spike their glucose to different foods. We carried out a very precise study where we had 55 people eating the exact same amount of carbohydrates, but the carbohydrates were in different forms. There were grapes, there were bread, you know, potatoes, pasta, rice. And then we also later, we'll talk about that, I'm sure, we put mitigators where They had some prefoods, if you will, or foods they ate before they ate the spiking carbs to see what would mitigate this. So the goal was to try and match people's subphenotypes with what foods cause them to spike. So the ultimate goal of this is personalized nutrition, if you will, personalized, well, eating is another way to say it. Knowing that we all respond to food differently, we've seen this in many forms, whether it's fiber, whether it's, an Ensure shake, what have you, the people respond differently. And ultimately we'd like to understand what's causing that and then personalize this so that if you eat these things and they're pro-inflammatory or they're spiking your glucose, you know, you should avoid them and go to these other things that you like that don't do that. That's the big picture goal.

One thing I found interesting, and this is kind of like a cute aside, but Figure 1D, was that heat map, and there was one cell that just stood out. It's like blaring red among all the other cells. We'll show it on screen now, which surprised me. It was, it was grapes, and basically was showing grapes just like have this really acute rapid glucose response for most people. Did that surprise you?

Yes and no. I mean, grapes are sugar. That's really what it comes down to, and that most of their carbs are in the form of sugar. And so what you're seeing is like a glucose spike, and that's why it's so high. It's high, it spikes on pretty near everyone, but it's transient. It's very, very rapid. So the area under the curve, meaning the maximum amount of glucose, is not as bad as most other of, of the other carbs that were there.

So it's rapid, but it's not necessarily dangerous.

Yeah, not like rice, which is often very high but more prolonged. So the number one food that spiked people was rice, something like half or just under half the people, that was their top food that spiked their glucose. And interesting, one surprise to me was that Asians spiked more than non-Asians. I did not expect that. This is when you have pure white rice. Anyway, but some people spike the bread, some the potatoes, some the pasta. It's all over the map. Not so many to beans and one of the other things we had. And I think that—

Beans and berries, I think, were the other two carb sources.

Oh yeah, berries. Yeah, berries, I might have thought more, but, um, that— but both those have more fiber, especially beans. And so the carbs are in not raw glucose kind of form— well, sugar forms. So that's why they don't have the same effect. They're more spread out, and you have to break them down if they're in a fiber, the carbs are in a fiber.

So the bottom line is, but grapes, they're just like you say, solid sugar, and She asked ChatGPT, I like, I explained what the image was and said make a graphical depiction. I had this cartoon of grape rocket ships shooting into space with like little sugar stars.

By the way, another little fact there, you may know this, is that some people, uh, for many people, uh, rice is worse than ice cream.

Yeah, that came up in— it was Aaron Seagal had a TED Talk. So he did the other paper we were talking about before, the 2025 Nature Medicine papers, is building on other work showing just how different people's responses were and it was the cookie versus banana figure. They had two famous panels where two different people ate a cookie and a banana and the responses were completely opposite. But then I remember watching his TED Talk on personalized food and he was like— I think his wife's a dietitian or nutritionist or something and she was like, "I was shocked and now I'm saying look to my clients, if you want to have something between those two and not spike your blood sugar as much, ice cream—" Sorry, the other thing was ice cream. Is probably better than, than rice. So I don't know, we might have to reevaluate the healthfulness of rice. But yeah, I did know that.

Well, I don't remember if we talked about this before, but white rice only has 8% protein.

You mentioned that. You said you like this high protein rice, right? There are—

the ancient forms are supposed to be 20%. So there's this thing called Rocky Mountain barley or Rocky Mountain oats or something. I've gotten that. I'm the only one in the family who likes it because it has a pretty strong taste, but it's, it's 20% protein. And as I say, it's one of the few things I actually like. I'm usually pickier than my family. Anyway, it obviously then has lower glycemic index and therefore doesn't spike you as much.

Were there any molecular or hormonal signatures that explain why two people might respond so differently to the same food? Let's say bread as an example. Like, did anything stand out?

For the bread, we noticed that hypertension— people with higher hypertension tend to be bread spikers. Why that's the case, I don't know. That one is not so obvious to me unless it somehow relates to, you know, bread does have a lot of wheat and gluten in it, unless there are some folks in that study had a lot of inflammation or something that was triggering this further. It's not so obvious to me.

What else did you see that was a notable pattern?

That was a good one. I remember, yeah, the Asian one was a good one. You know, we did see, we could correlate various microbes with glucose spiking and some of the other molecules. I forgot what else was in that paper. You may remember, Nick. There must be something that caught your eye.

The grape to potato ratio is just brilliant. So I think we talked about that before. I don't know if we dug into it. I know we were going back and forth after our last conversation about the protocol, which I'll drop below, but can you explain what's going on there?

So one of the key findings of the paper is that depending on your subtype, determines what foods you spike to. So meaning muscle insulin resistant people will spike to potatoes and they'll spike to pasta, but not insulin sensitive folks. And likewise, people with beta cell defects, which is me, I have a beta cell defect, they'll spike to potatoes as well. Why this is all the case, not so clear. But one thing that potatoes and pasta have in common is resistant starch. So somehow maybe it's related to that. Then the nice thing is what you were getting at is this, we could set this up as an assay. And I mentioned this in the paper, that your potato to grapes ratio. So you treat grapes as though they're just essentially sugar and potatoes. If your muscle's resistant, you're gonna spike to that. So that ratio will go higher than in insulin-resistant folks. So in fact, that could be a common person's way to tell what their subphenotype is by the amounts. Now, I would prefer, being the chemist, I mean, I was trained as a chemist, say drink the glucose shot because it's a little more precise. But if you can't, can't do that, maybe you just eat this— these potatoes and grapes. It's a little harder to control the amounts on that. We're running a study now, by the way, to figure out what types of potatoes, uh, are better or worse for you, like french fries, mashed potatoes versus baked potatoes. With your prediction, that'll be fun.

If you had to— you had to engineer the healthiest way to eat your standard potato, what would it be?

French fries probably won't spike your glucose much because they're fatty, but they may not be as good for your cholesterol. So there's a trade-off. But from the glucose standpoint, I'm thinking the baked potato, keep the skin on, is not a bad thing to do. And of course, if you eat a raw potato, it won't also break down as well. So meaning the temperature of the potato actually affects how glycemic it is.

Can you eat a raw potato? I thought they contained something slightly toxic, like a raw potato.

I'm not aware of that. That could be— I've eaten raw potatoes in the past, so maybe it's a dose thing.

I'll be transparent about where I'm driving this. I took a— I did a survival camp when I was a teenager, so this is, this is not going to be a PubMed ID for people. Um, I was like 12 and they made us like start our own fires to cook our food, and we had meat, raw meat, and raw potatoes. Um, but it just rained like poured, so like a bunch of 12-year-olds trying to start a fire with wet wood. And we couldn't. And I was so hungry at this point. I wasn't fasting or anything, um, and I just got so angry. And I knew I couldn't eat the raw meat. It's like raw ground beef. So I'm like, screw it, and I ate a bunch of raw potatoes, and then I started throwing up. So that's where that's coming from. I could be making up that there's a toxin.

I'll have to look it up. My understanding is you don't break down the cellulose so well if it's not cooked. So, so probably, and that's why it's less glycemic. And I've actually seen that, uh, in me personally. That's my own personal experiment.

All right. And with or without sour cream?

If you do the baked potato, uh, yeah, I usually add butter on it just because we don't have much sour cream at home, so I get my fat added that way, I guess.

So in theory, people could put like a continuous glucose monitor on and then have a controlled amount of potatoes and a controlled amount of grapes on different days, probably under the same conditions, and then ratio this, maybe like taking the highest point on the spike of each, then potato spike to the grape spike. Is there a threshold at which you'd say you're probably insulin resistant?

I think these are continuous variables in the first place. That's true even when we classify people as muscle insulin resistant. It's a, you know, it's a continuum. I would have to translate what that number is into, you know, the gold standard clinical test for insulin resistance, but it's doable for sure. Because it is a continuous value, both the potato spiky and the grape. So, so absolutely could. And well, we are going to validate this a little bit further for— and then we're thinking to put this algorithm up for free. And my own preference would be that they do the glucose version versus— but it may be if you're diabetic, you wouldn't want to take the full 50 grams. You do something a little smaller.

Fair enough. Wise. All right, now I want to get on to these mitigators. In particular, a really interesting finding of this paper I thought was that if people had insulin resistance, they had a diminished response to fiber as a mitigator. So the idea is like, you know, we hear these ideas like that if you eat fiber before you eat, you know, more carbohydrates, the glucose spike will be less, but that blunting effect is reduced in someone that has insulin resistance, correct?

And that was the goal. So maybe to describe the experiment a little more, we gave people 10 grams of protein, which was egg white, fiber, which is chickpea, and the third was fat, which was cream, I believe, as I recall. Anyway, we gave them 10 grams of that and then they ate rice because everybody does like this white rice. And then we looked for whether it would mitigate the effects or not. And both fat tended to delay things, but I wouldn't say it really changed the overall, you know, peak or area in the curve as much. But I would say the biggest effects were definitely the protein and the fiber. Both could mitigate the spikes if you were insulin sensitive, but not if you're resistant. The whole reason for that is still not 100% clear to me. We've come up with different theories, but I don't know that we have the full answer yet. Yeah, one idea is that you're lining your gut better. You line your gut with these things and maybe that's helping. Eat some of this. Love to hear your theory on it, Nick. And then why doesn't it work for the insulin resistant? Well, maybe they're just more sensitive overall. Bottom line is that, yeah, that didn't work on these folks as well.

Do you remember what the fiber type was?

Yeah, it was a chickpea, which is usually inulin. Interesting. It's a mixture, but inulin's probably the major component of that particular one.

I mean, I don't have an answer in terms of like a, a direct model for how insulin sensitivity, uh, would impact it, not off the top of my head. But I think something I want to emphasize is the more and more I learn about the microbiome and fiber, the more I become interested in the diversities of this thing we call fiber and how it interacts with the body. I think a lot of people, including probably a lot of people listening, have this conception of this is just the thing that, say, slows gastric emptying or pulls in water to make you fuller, and it's just something like mechanical, which is not at all how it is. So you mentioned inulin. There's another paper I'll have to send you, was out recently. It was the coolest fiber paper I'd ever read, and it was actually about inulin specifically. They were following up on some other data showing that inulin improves insulin sensitivity and fatty liver in people with insulin resistance and diabetes. And then they did a follow-up study and what they found, at least in the preclinical models which associates with the human controlled trial data, was the way in which it reshapes the microbiome was actually quite fascinating. So this is gonna sound crazy, but it created like an immunization against sugar, in this case fructose. So this is over chronic use, which is why I don't think it necessarily applies and this is somewhat of a tangential comment, but basically what it did is it changed the microbiome so that the fructose sugar could be catabolized more. Right. And the carbons ended up getting then shunted of all things into protein. So amino acids that were used to synthesize glutathione. So what they were able to show remarkably was that this fiber adapted the microbiome so that sugar was shunted into protein and antioxidant defenses. The point here is not to like jargon bomb people listening, but You would have heard both Professor Snyder and I at different points when going over the why behind this incredible observation that people have insulin resistance and have this blunted response kind of pausing and thinking and maybe being a little bit stumped. And it's because these things are actually way more interesting and complex than I think most people appreciate in the way it generally gets communicated. So I don't have a good answer for you. But the simple observation, you know, like any good study, opens up the door for so many more questions that you could ask. And then actionability items based on once we understand the why, like what can we do with that? Just like with your original subphenotyping paper, this was the follow-up. So now you have another follow-up. That's how science progresses.

Maybe to add to this, so we ran a study where we ran people through two different fibers, arabidopsylan, which is very common in physium husk, metamusil. Etc., and inulin. These are purified fibers now. We wanted to see, you know, as opposed to mixtures, see if we can just see exactly what's going on. 'Cause if you read the literature at the time, it was all over the map. Some people saying these are better for glucose control, other, some saying there's no effect, some saying this is for each of the, those two fibers I mentioned, and others saying they're good for cholesterol or not. And what we discovered is that it was a small study, 18 people, but it was a crossover. So they did either arabinosylan increasing doses, then a washout. It was basically over 3 weeks. And then they did, and then the washout for 6 weeks, then they did the next one. And then they did a mixture after that, after another washout. So the point out of all this is that we did discover that in general, arabinosylan did decrease cholesterol quite significantly actually, whereas inulin had no effect. Neither had a general effect on glucose. But when you drilled into individual people, we discovered that some people actually didn't have their cholesterol reduced by rabezailan, but they did by inulin. So there are personal differences, and we could see very specific microbiome shifts with these fibers. So it, it's a, it comes back to what you're saying. It's a personal relationship. My own view on this is that the microbiome it is personal. You may know we all have very different microbiomes. It gets set early in life, the first 3 years. Not our work, work from Baylor group has shown that basically you establish your microbiome and your microbiome lives off the fibers and probably everything that was happening in those first 3 years of life. And then, you know, you're maintaining, you can shift it by the way, but it's a lot harder than it was probably when in your first 3 years. And so basically, and so your, but your microbiome's complex, it's a guild. If you will. And this is why people haven't figured it out, meaning everybody knows the microbiome is important, but people haven't been able to manipulate it the way we were hoping to. I mean, there are some things that work. Akkermansia is great for lowering your glucose. I think that one's become pretty well established. There's companies selling that. I have no stock, so it's real. That one drops your hemoglobin A1c 1%. Um, and so I, and I think it's hard, and I think it's going to come down to the actual Hydrolase is the exact community you have that determines how you react to this. And so we're actually spending a lot of time on this now. We have people running through 4, we've actually just completed this. They've run now through 4 of these fibers, including resistant starch and beta-glucan, and plus the other 2. And we're trying to see how they react. And we're looking at their microbiome. And this is, I think, gonna be perfect for AI where we use, you know, collect a, 'cause a lot of data and we see what communities, you know, get shifted, how does it affect their glucose and other control mechanisms. And I think you raise a good point. We need to know not just short-term but long-term effects of these things.

Not only the many different types of fiber and how they interact with the host microbiome and how the host microbiome changes over the course of days, weeks, months, but then there's also like, what are the long-term health impacts versus how you feel? I think that gets underemphasized. I'll draw from personal experience now because like I have a complex gastrointestinal history. You mentioned microbiome gets set up early in life. I had a fever early in life when I was a neonate, so I got these harsh antibiotics, IV antibiotics, and it probably screwed my microbiome up for life. You know, I developed ulcerative colitis. I've never been the same since. And when it comes to things like fiber, like that paper I just, you know, talked to you about, or the paper you just mentioned, I think was Lancaster et al. in Cell Host and Microbe. It's funny, actually, now total tangent, but I'm terrible with like, I can't memorize a single song, but when I have a good paper, it like imprints on my brain. So you're talking through that study and I'm like, that was Lancaster et al., 2022, I think it was Figure 3B. And I was like looking, I'm like, it was Figure 3B.

You remember better than me, what can I say?

I don't remember artwork, I don't remember song names, but when the science hits right, it just it imprints. But the reason I was bringing that up is because I have this conflict where the science is so cool and I can, you know, talk to the benefits of what this might mean for general human health long-term. And at the same time, I have my personal experience where, you know, if I have a head of cauliflower, I'll balloon up for 2 days. And it's just like, it becomes very hard on the personal level to know what to do and how to like play with these interventions because there's so much heterogeneity. I don't even really have a question for you. Because I don't think it's clear-cut. I just wanted to name that because one of the things that really excites me about the work you're doing is the specificity, the resolution, the personalized nature, I think will make it so that people like me that are struggling with these individualized aspects will have a lot more to go off of, in terms of like adapting our lifestyle in 5 to 10 years, we're Right now we're building off like very gross, rudimentary, blurry understandings of these things.

Yeah, 100% agree. And maybe to add to that, here's a good example. Uh, inulin. When we ran our inulin study, 3 people including me, when I hit the higher dose, 30 grams per day, which isn't a lot for most people, it actually caused inflammation. My case, it was sub- clinical. I didn't feel anything. But 2 people, they actually felt, you know, we could measure it in me, but in other people they felt that way. We saw their markers go up, CRP. So it can have different effects. And I think this one reason it's complex is your immune system. Again, we've talked about that before, 70% of your immune cells are in your gut. And so they have this relationship with the microbes that are there as well as the foods you're eating.. And obviously that's what's going on with you. You're having these immune reactions to the fiber that's coming in here, and it's probably, it could also relate to the microbes that are in your gut as well. And it is a complex relationship. I, I do think, you know, in general, the party line is we need these microbes. You need to make vitamin B12. You can supplement, of course, but a lot of these vitamins would be good to get naturally. And certainly there, there's a lot that's made in your gut. That you don't normally get. Now, again, you can try and get around that with supplements. I think it reflects the fact, again, we're homeostatic systems. We get shifted. You probably got shifted early in life because of this fever and the antibiotics and that shifted you in a certain direction. And you're still obviously a strong functioning, you know, homeostatic system, if you will, that works pretty well. But, you know, I would wonder whether you're missing something. And I don't know whether, like, if you do these, you can now do these deep micronutrient assays, you know, Genova and some of these deep medical tests, Boston Heart. We run our own through our, you know, the steep profiling studies and you could see if you're missing something. So if you are, hopefully you could add that back. All this stuff has not worked out perfectly. There's getting to be more and more studies like vitamin D3. I think we can now say Vitamin D3 supplementation is good for people because you can correlate with all these health benefits in general. But back to your other point, I think that we do need not only knowing what works in general, but I think getting specific for people because obviously you can't run around with an inflamed gut. By the way, we saw the same thing with an Ensure shake. People reacted to that shake very—

those things are just sugar bombs. You look, they're all over the hospital. I remember throughout med school they were just like freezer, like fridges stocked with them. You look in the back and they're just, they're loaded with sugar. Some of them like have 450 grams of added sugar.

It's crazy. Yeah. We found that for some people it was pro-inflammatory, for other people it was anti-inflammatory, same shake. Pretty wild.

It's, it's incredible. And I'm just optimistic for the future because I think we'll just have such an order of magnitude better understanding in 10 years that right now, double down on my own story, I've been on a keto diet for 6 and a half years. I don't mind the food. The food's great. Social elements are, you know, create difficulty. But like the way I'm seeing it in my mind is for the foreseeable future, this is the thing that I need therapeutically. And if I were to not eat another piece of bread for the rest of my life and die at 150, my plan, I wouldn't mind. However, there is a part in my brain that's like, I do think in a decade or so the technology and scientific understanding will have evolved to the point that I will at least have more tools to try things that I never could have imagined before. So, you know, you never know. Things are evolving so quickly.

You wanna watch out for the inflammation response. Yes. Are you missing something? For example, are there some microbes that because you're not feeding them, uh, with your diet that they're not, you know, you're missing something and that's what you wanna catch.

Yeah. It's, it's hard though. It's like, I don't know. Do you know of any really reliable, reputable microbiome assays to say you're missing this thing? Because different bugs have different functions in different people. So like, it's really hard to assess the functionality of your microbiome.

I don't think we're there yet, right? No, because it's a guild. Again, there, there are many ways to make a good product, if you will, a microbiome product I'll call it in this case. And they're, you know, they're hundreds of strains in your gut. Some dominate, obviously, but it, it's— they all have different, you know, some different microbes can bring in different hydrolases, for example, that will break down your fibers and things. And you can have, you can have multiple solutions to the same problem of being able to break down, as well as personal differences, you know. Probably if through an immune reaction, you're reacting to some of these fibers. And presumably, if you could avoid those but still have fibers that are missing those particular immunoreactive things, that they'll work fine for you. Problem is most fibers, as you buy them, they are mixtures. So that's a little hard to—

Yeah, no, by empiric, you figure out what lands and what doesn't. Weirdly, like, I don't know what's in kale, but I have a huge glycemic response to kale, like more than milk chocolate. Which is pro— I mean, it's weird. And I think it's just I'm having an inflammatory reaction and that's spiking the sugar. There's not much sugar in kale.

Yeah. I wonder what fibers are in there that are probably, you have a lot of hydrolases that are probably ready to attack those kale, right? Anyway. Yeah, it must be. I'm guessing that that's just how your microbes have tuned themselves, right? To be able to digest that very efficiently. Whereas if they were to digest it more slowly, perhaps that would—

I don't think it's a sugar release from the kale. I think it's a matter of like, like there's some sort of immune reaction and then the inflammatory is secondarily causing a glycemic.

Oh, oh, I see the cortisol. Yeah.

There's some sort of, I, point being kale, kale's not the fixation. It's that it's hard to predict in advance how different bodies are gonna respond.

I mean, that's kind of the major takeaway from this and other work, but that we are, what we'd like to do in the future, right, is profile you at baseline and say, aha, this is what your profile looks like. These are the things you'll react to. These are the things that'll be good for you. That's the ultimate goal. Absolutely.

Well, you've done n=1s and published them on yourself in Cell. Maybe I could offer myself up as a deep profiling subject at some point.

Yeah, well, you have to be in the Bay Area to get deep profiled.

I'll fly out. I'll fly out for that.

We're very careful about how we, you know, process the blood samples and all that. So, sure.

I want to be respectful of your time. Thank you for now our 4th interview. I've already extended an invitation for 5th and 6th, etc., as I get my new setup. And I really enjoyed these conversations. There's something really special for me. So thank you. I know my audience thanks you. If people want to check out your work, where can they go?

They can go to Snyder Lab. We have studies running there. Um, and as far as, uh, YouTubes and things, I'd say just search Snyder and see which one likes you. Actually, go look at your other podcasts there, Nick.

All right, well, thank you for your time, and, uh, we'll be signing off.