In this episode, I am speaking with Dr. Tim Spector – a professor of genetic epidemiology, genetics and microbiome expert, researcher, and one of the top 1% of most cited scientists in the world. We will talk about the truth about the link between your genes, your microbiome, and your health; and some easy tips to optimize your health today.
In this podcast, Dr. Spector will cover:
Listen outside iTunes
Ari Whitten: Hey, everyone! Welcome to The Energy Blueprint podcast. I’m your host, Ari Whitten. Today I have with me: Dr. Tim Spector, who is a professor of genetic epidemiology and director of the TwinsUK Registry at Kings College London, and has recently been elected to the prestigious fellowship of the Academy of Medical Sciences. He trained originally in rheumatology and epidemiology, and in 1992 he moved into genetic epidemiology and founded the UK Twins Registry of 13,000 twins, which is the richest collection of genotypic and phenotypic information worldwide. Don’t worry, if you don’t understand those terms, we’ll explain in the podcast. He is the author of The Diet Myth, his current work focuses on the microbiome, and he directs the crowdfunded British Gut Microbiome project. I also want to mention—and this is a pretty crazy and impressive stat—he’s published over 800 research articles and is ranked as being in the top 1% of the world’s most-cited scientists.
So, welcome to the show, Dr. Spector! Such a pleasure and an honor to have you.
Dr. Tim Spector: It’s a pleasure for me, too!
The PREDICT study
Ari Whitten: Great! So your work, as of late, and you have—obviously with 800 research papers—you’ve done a lot of work and I would imagine a lot of different fields and maybe related fields, but a lot of different topics that we could probably talk about in this podcast. But we’re going to talk about, mainly the PREDICT Study—which you’re doing now—and in particular, the gut microbiome and how different individuals can respond differently to different kinds of foods. I would love for you to, first of all, give me an overview of how you got into this field and then I want to dig into the PREDICT Study in particular and have you tell everyone what it’s all about.
Dr. Tim Spector: Great. So I got into it like I get into it through everything: through with my twins that I’ve been following for 25 years. Initially, I was interested in why they’re so similar. And the last 5 to 10 years, I’ve been much more interested in why, occasionally these identical twins—these clones—end up being quite different. We explored a bit about, sometimes, they get these genetics changes—epigenetic changes—but they are not very big. The biggest difference we started to notice was in their gut microbes. The reason that one twin will be overweight and the other one’s skinny, tend to be the most obvious clue was they had very different gut microbes and behave nothing like all the other traits that we’ve been interested in. Suddenly, by studying these twins, we suddenly realize there’s something inside all of us that makes us very different, very individual.
And explain why, if that’s true—that we’ve all got these very different microbes inside us that process foods differently—could explain a lot of the mysteries and the reason we’ve got so much wrong about our modern diets and about giving everyone exactly the same advice, when we’re all so completely different. So that was the sort of entry to this. About five years ago, we published a first twin study on microbes and showed there was only a very small genetic component to everyone’s microbes. That led, really, to this whole study of nutrition, and then the idea that we wanted to see how personalized nutrition was—that if our microbes are different, then maybe the way we process food is also very different because of that, or to some extent, because of that. And that really led to this whole idea of the PREDICT Study, which is the largest food intervention study of its kind ever performed.
I was giving a talk about my book, The Diet Myth, as I do to the public in London quite a lot, and two guys came up to me afterwards said, “We think it’s a fantastic idea. We’d like to try and get a company together to invest in a big study that could lead to personalizing diets based on these rough concepts.” And so anyone who wants to throw research money at me, I’m always very open-minded and it gave them a listen. And sure enough, they went away and raised the money and got a company together called Zoe that founded this series of studies, which we’ve just finished the first part of—which was to look at roughly a thousand twins in the UK and a hundred volunteers in Boston in mass general, giving them all basically identical meals throughout the day, and seeing how they responded differently or the same to those identical meals.
Dr. Tim Spector: Because everything we’ve been taught by science and nutritionists and guidelines is that really there shouldn’t be much difference between us, that the experiments were done well.
Ari Whitten: And what did you find?
Dr. Tim Spector: We found about an eight-fold difference between people in their response to the same food, in terms of their changes in their sugar level, changes in their insulin level, and changes in their blood fats—their triglyceride levels. Immediately we could see, in a repeatable way—because not only did we do a day in the hospital where they had about 10 blood draws, but also they had, two weeks later at home, where we repeated some of these experiments and they did free meals—and we saw the same thing. People just react totally differently to identical foods. And that suddenly opened up this whole area and concept of personalized nutrition and “one size doesn’t fit all”.
And the other thing was that, even within identical twin pairs, we could see that they not only had different microbes, but they were responding very differently to the same foods as well. One would react badly to fats, and one might react badly to carbs. And this isn’t something that people would guess beforehand. It was a surprise even to me, as the biggest sort of skeptic.
Ari Whitten: Now, how long did this study go on? How many days or weeks or months was it?
Dr. Tim Spector: It was a two-week study for each individual. It was a very intensive two weeks, and it was just observations. We just gave them set meals, set foods, and everybody was given a continuous glucose monitor. You may have seen these before—these little stick-on devices that you put on fairly simply, and you can read out on your phone or mobile device, and it gives you a reading about every five minutes of your blood sugar. So that lasted for two weeks. We gave them also some testing of blood spots so we could look at that fat levels—so we can now measure your fat just based on a blood spot level pretty accurately. So, tests and responses to fatty meals. We also made them log all their foods for that two weeks on an app, at the same time as working out exact timing of the foods, and how much sleep they had with a sleep monitor. They had an activity monitor so they were really intensively followed, as well as talking to a nutritionist every day to see how well they were accurately logging their foods. Because, as you know, most nutrition studies are dogged by the fact that the date is pretty rubbish, and people either forget or make up the foods they think people want to hear about.
Ari Whitten: Or selectively remember.
Dr. Tim Spector: Yeah, exactly. We all do that. How many glasses of wine did you actually have last night? We all tend to round it down. So, that was what we did with these 1100 people: put the data into big computer algorithms, and worked out that by looking at all this data together, we could predict pretty well for a first pass, about how any one individual is going to respond to any particular common food, even if they hadn’t tested it. So even if we hadn’t tested that particular food, we say, “Okay, how’d you respond to eating a beef lasagna? What sort of score we give you for that based on your basic results plus the results of everybody else?”
Dr. Tim Spector: So it’s partly an individual study, and it’s partly a crowd science study, because the more data we get from everybody, the better the overall predictions and we can separate out the noise and the errors.
Ari Whitten: Yeah. Okay, so you’ve measured blood sugar, blood triglycerides, and—was it one other thing in there?
Dr. Tim Spector: Their insulin as well.
Ari Whitten: Insulin, okay.
Dr. Tim Spector: Blood sugar was the main thing because technology has allowed us to get thousands of readings per person without any pain. Obviously, you can stab your finger many times and most people don’t want to do that too much. So it was the first study that’s looked not just at sugar, but also at fat. And we found some people respond well to sugar: they can have lots of carbs without any bad glucose, insulin spiking. Others, the opposite: they can have as much fat as they like and they have poor sugar responses. So nearly every possibility we saw in this data—which was interesting because before then, people thought that the two are very closely linked—but if you responded badly to sugars, you are going to respond badly to fats, and that those responses were just linked in the same metabolic disorder. But we’re showing very clear differences between them, so the advice we give is going to be different. There’s a big crazy at the moment for low-carb, high-fat diets, particularly for people with diabetes or prediabetes—that’s going to suit some people, but not others. And this is what we’re attempting to get at a more holistic approach to this kind of testing, then just focusing on the sugar peaks.
Sugar might be why we’re testing these peaks at all, because in a way, it’s not just the responses. What we’re trying to do long-term is give people the best advice, to give an alternative say of breakfasts—whether you’re going to have your granola, you’re going to have yogurt, you’re going to have English muffins or porridge—what’s the best one for you, given the same-portion sizes. It’s allowing people to make choices based on some real data. What we’re trying to do is to give someone the food that’s going to give them the least amount of a sugar peak or a fat peak because we know that a very recent science is telling us that people with regular sugar peaks are more likely to get diabetes and gain weight, and people with fat peaks at six hours—because much later at six hours, you get your fat peak after a meal—people that find it hard to get rid of the fat after those meals means that those little particles linger on in the body and start irritating the blood vessels which leads to heart disease. They seem to also tickle up the fat cells and make you store more of your energy as fat. The more we can just slightly, every day, reduce the number of peaks, then over a year you’re going to start losing weight and having less health problems. And that’s the basic concept behind what we’re trying to do. It’s not a rapid, instant diet. It’s not a “six weeks and change your total body”, but it’s just making small adjustments to your diet that are going to have long-term effects.
The problem with personalized health advice
Ari Whitten: Gotcha. Now I want to play devil’s advocate here for a moment, and I want you to convince me why these arguments are not very good. Within the fitness community, the exercise community—there was a theory that was kind of around for a long time before it was ever tested: the basic idea of which is that the hormonal response to the exercise—to the weight-lifting workout—was really critical in terms of the muscle-growth response. For example, it was known that if you do squats, or deadlifts, or big multi-joint movements with heavyweights, you’ll have more of a growth hormone and testosterone surge following that workout. And it was thought that that hormonal response to the workout was going to be really critical in the overall long-term, muscle-building, strength-building effects.
When it was actually put to the test, they found that there was almost no correlation at all between the hormonal response to the workout and the long-term effects. It was basically irrelevant or close to irrelevant. Another example of this is metabolic ward studies—I’m sure you’re familiar with like, for example, Kevin Hall’s recent metabolic ward studies—where they basically looked at equal-calorie diets that are either very low-carb/high-fat diets versus high-carb/low-fat diets. And even though they have equal calories, there’s going to be guaranteed big differences in terms of blood sugar responses to the individual meals or insulin responses to the individual meals. Meaning, the people eating the low-carb/high-fat diet are going to have much lower insulin and blood sugar responses to those meals. Yet in the long-term actual effect on body weight and fat loss at equal levels of calories, those two diets resulted in the same amount of fat loss despite the hormonal differences. I guess I’m asking, are you certain that those acute responses to meals are actually going to be of very real and significant predictive value in terms of long-term, real world effects on say, fat loss or health outcomes?
Dr. Tim Spector: So I was equally skeptical when we started talking about this internally, because other people have run away with this idea and said, “Yeah, it’s all about sugar peaks.” And I’m a naturally skeptical scientist—epidemiologist—and I said, “Well, we’d have to do the study for 20 years to really find out whether that’s true or not.” Yeah, it sounds like a theory and yes, if you believe this—the general Insulin Theory that actually it may not be the sugar peak, but it might be the amount of time the insulin is being stressed—well, that’s the bad thing. It sounds like a good theory, but we couldn’t prove it until two years ago. This large study of about 400,000 people came out using something called Mendelian randomization. This is a very complex way of saying, “We’re using genetic testing in a way to an unbiased study in a large population.” Because through the genetic work, which I’ve been taking in the last 20 years, we’ve actually found a lot of genes that—well, they’re not very useful [inaudible] themselves, they do that statistically associated with people that have either high peaks or low peaks. So we know that genes that distributed randomly in the population. If you take a half a million people, and half of them have the genes for the high peaks, and half for the low peaks, and everything else should be the same, on average—if you see a difference in those people’s weight, it means that their lifelong genes have had an effect through this small but statistical association. And that’s exactly what happened to this Boston group: published this data two years ago, and clearly showed that people with those set of genes had a greater weights and greater risk of diabetes than the others.
So, in a way, without doing a 20 year study, that’s about the only way at the moment we can get that sort of information for what is a relatively small effect over time. So that’s how we moved from these short-term changes which, as you rightly say, you can find any hormone that goes up and down with an activity, but linking it to something lifelong—like your genes, to say—that’s how I’ve done it. And that whole method has been proven many times in many other fields. It’s not like it’s just made up for this purpose. I think there’s been similar work with the lipids as well. It’s not as clear-cut that the triglycerides are related to a weight gain in as much evidence as the sugar, but there is probably equal evidence that having your fat hanging around longer is [as is] with heart disease. There’s some quite good data in that direction.
So I’m much more confident now than I was just two years ago, because the data has moved on. And I think you’re absolutely right, we have to be skeptical about anyone who says that these short-term changes in mice or in a study of 10 people in a gym or wherever—that that’s going to lead to these long-term changes. And that’s why these new techniques—using half a million people—give a lot of confidence in these areas when they exist. But ultimately, we want to show that by changing people’s sugar peaks and fat peaks, we can get them to lose weight whilst keeping everything else the same.
There is a group in Israel, the Weizmann Institute, that did this about three years ago: small numbers, and it was only a short-term over a few weeks. But they did show that by following this low-peak diet versus a normal diet prescribed by a nutritionist, they did improve the peaks. And in theory, if they carried them on longer, they would have lost weight. But the ultimate test will be to see if people who use these predictions can maintain weight better than those that can’t, given equal advice and equal calories. It is a science that’s evolving and, based off most of the nutrition science, it’s pretty weak. And so, everybody—all your listeners—are right to be skeptical about anything [like that].
Ari Whitten: No, I doubt, actually, that most of my listeners would be skeptical. I personally am skeptical because of some of the reasons that I brought up. But I think the trend now is in the direction of “highly-personalized everything”. So there’s lots of very shady companies kind of popping up around this field of “personalize this, personalize that”. For example, “We’ll do this test on you and determine your personal best multivitamin and mineral supplement. We’ll custom formulate one that’s unique to your genetics, or unique to your blood chemistry.” For example, like uBiome or Viome will give you specific diet recommendations based on what we’ve determined is the optimal diet for that particular microbiome. I just think the science, the actual research in those areas—and this is your area of expertise so I’m curious if you agree with me—but, I think that the science in those areas is in its infancy, and we’re not at a level yet where we should be charging money to say, “We know the optimal diet for people with your microbiome. Here it is, pay me some money and I’ll tell you,” or, “Pay me some money and I’ll formulate the optimal multivitamin and mineral supplement for people with your genes,” or something like that. I think there’s a huge gap between the claims and the actual research.
Dr. Tim Spector: I absolutely agree. And as someone who’s been doing genetics for the last 25 years, I think the microbiome is a new science, genetics is not, and we do know much more about what we can and can’t do with genetics. These claims that you can get someone’s DNA—which is the same when you’re born and when you die—and assess sporting ability, for example, whether you’re going to be a marathon runner or a tennis player, there are all kinds of claims made about how well you can build your muscles, and these things are complete nonsense because they’re changing what is the fact that we do know some of these genes are associated with say, bodybuilding or running marathons, but they explain perhaps less than 1% of the difference between people. They’re statistically significant in a population level, but for any one individual, they are completely meaningless. And it’s really important for people to understand that difference. They can be saying something that’s true in a population—yeah, it’s slightly better than tossing a coin in the population. But once you get down to an individual, it’s no different to tossing a coin.
Ari Whitten: Yeah, I’m so glad you brought this up. I think there is a real deficit in the general public ability to understand this concept that you’re talking about, and kind of effect size the difference between statistical significance versus something that actually has a large effect size. There’s a lot of focus in the natural health and alternative-health communities now. I see a lot of practitioners who are hyper-focused on genes and doing these advanced genetic tests: if you have this mutation, this variation of this particular snip, or that one, MTHFR or COMT, or this or that, then we’re going to do all these interventions and give you this supplement protocol and that supplement protocol… From my perspective, I think in most cases, where they’re doing these kinds of things, they’re dramatically overestimating the effect size from those different gene variations.
Dr. Tim Spector: Absolutely. The vast majority of these gene tests are completely worthless. I was part of the team that discovered the vitamin D genes. We know it has a genetic basis, but trying to work out while your vitamin D status is based on your genes is still complete nonsense. We’ve found 10, 15 genes, but there’s probably hundreds of little genes that will make up these factors. And that’s true for most of these traits. They’re usually thousands of genes, little bits of them that all go together to make up the trait. So, until you know everything, you’re not going to get there.
The other point is that most people are in the middle of the range. There is this one caveat that is increasing this thing called gene risk scores, and this is where someone is at a very high or very low risk of a disease in the lot type, like the top 1%. You can say something and that’s like a clinical test. So the very top or the very bottom, you can say, “Yes, you’re extremely unlikely to get breast cancer,” or, “You’re extremely highly likely to get it.” But for 95% of the people in the middle, that test is useless. For most people buying a test online, they’re going to be in that 95%, they’re not going to be helped at all. So that’s really important for people to understand. And even when they get the result back, they suggest they are at higher or lower risk. It’s a difference between 50.1% and 49.9% difference.
Ari Whitten: Yes. I was just going to say that. Basically people will read on their report this list of genes, and one will say, “slightly increased risk for Alzheimer’s disease,” or, “increased risk for obesity,” or something like that. But I think what people have a hard time wrapping their head around more broadly is exactly what you just said, that yes, it may be statistically significant that your gene variants are more likely to have that, but it might be the difference, as you said, between 49% chance versus a 50% chance. At the end of the day, it’s actually totally insignificant and is probably drowned out completely—like, let’s say, in the case of obesity or diabetes—is drowned out completely by one’s lifestyle habits.
Dr. Tim Spector: Yeah. And coming back to diet, there are companies out there that claim—based on your genes—to be able to provide you your ideal optimum diet for your genes. All the caveats apply to that. And we’ve done studies on the genetics of dietary choice and, yeah, it is slightly genetic, but these effects are small. And even if you take all the genes, you’re not going to get anywhere close to predictive ability. But in the PREDICT Study, because we have twins, we’re able to look at the genetic effects. And having twins allows you to do the perfect genetic test, because it’s like saying we’ve got every single gene test ever known in the future for the next hundred years. You can’t do better than a pair of identical twins. And when it turns out their responses to food, in terms of sugar, it’s not that great. So they only correlate about 50%, which means that only about 30% of the response—the glucose response—is genetic. Insulin is down to about 20%, and the response to fats is less than 10%. So even if you found all the genes, it would still be a pretty worthless test, compared to the environment and the ability to change your diet.
How lifestyle factors affect your health
Ari Whitten: Yeah, I’m so glad you said that because I actually know practitioners, again, who are just really enamored with genetic testing. I even have some friends who are trying to develop softwares—there’s a race to develop the most advanced gene analysis software with the idea that once we accumulate this massive amount of genetic information, then it’s going to finally arrive at a place where it has really strong predictive values. And what you just said right there completely blows that out of the water. It shows that the vast majority of the effect size are coming from non-genetic factors.
Dr. Tim Spector: Yes. And so you’ve got the basic, at rest, how much can you predict [where you are] at one point in time? And then if I say, “I’m gonna give you a burger, and based on your teams, I can say how you’re gonna respond to that burger,” I can now tell you that less than 20% of that response is due to genetics. At the moment, we can only maybe pick up 5% of that 20%, so it’s a tiny, tiny fraction. And the vast majority of it is other environmental tests and other things inside your body, like your microbiome… Our PREDICT Study has shown that other things we didn’t think were important are also there. People have different circadian rhythms. The time at which they eat seems to have a big effect on their spikes in their sugar and their fats; when they last exercised, how long they had a fasting period, how much they slept; the individuality of the microbiome, which explains between 5-10% of that.
So all these factors together—this individuality about us—it seems to be the main reason that we respond differently. We can measure these and come up with ways than predicting what the best foods are to put into that system, because everyone has a different motor, really, that’s what we’re showing. And that’s why some people prefer exercising before meals, some people prefer after, eating late in the day, skipping breakfast… All these things that we know instinctively—we’re not the same as every other guy—is something we’re showing for the first time in a big study, that nearly everything we’ve looked at seems to have this weird difference between people that are replicating.
I mean, you’re interested in exercise, and most people, when they exercise, are like, “My blood sugars tend to drop with exercise if I go off cycling after breakfast.” About a quarter of people, it goes up.
Ari Whitten: Interesting.
Dr. Tim Spector: And that repeats the [inaudible] and we say, “Well, why is that? What’s different about the way they’re responding?” And it’s exactly right. So that’s why some marathon runners prefer eating carbs, others prefer fat, others prefer to be vegan. In a way, we’ve got to find out what’s right for our bodies and not just believe some dogma because that is the theory of the moment.
Ari Whitten: Right. Now in your PREDICT Study, you used identical twins, correct?
Dr. Tim Spector: And not identical, and we had some other normal volunteers as well. A mixture.
Ari Whitten: Now I’m curious, to what extent—I’m going to phrase this in a complex way and in a blunt way. The blunt way is: I’m curious to what extent the differences that you found in responses in blood sugar and insulin and blood fats, are explained by one of the twins just being, objectively, much more overweight and having unhealthier habits for a longer period of time—let’s say, nutrition habits, circadian rhythm habits, stress, etc., sleep deprivation, and so on. And then, I guess in the more complex way of saying this is: to what extent are these differences a cause or effect or epiphenomenon?
Dr. Tim Spector: So we looked at health status and we looked at body mass index as a contributor to the differences between people. It has an effect, as you would expect, because as soon as you put on more weight, generally your metabolism will change. But it was [inaudible] less than 10% of the variation between people.
Ari Whitten: Oh, interesting.
Dr. Tim Spector: So none of the factors really dominated. It was really a combination of multiple factors that we saw, often of equal weight. We couldn’t explain these differences in response. We saw many pairs of identical twins, were one responds to fat and the other to sugar, had identical weights, but it turned out that the one that had the high response seemed to have high blood pressures or stress levels.
These are anecdotal stories, but they do help paint a picture of what’s going on here. So it wasn’t the obvious things that were causing these differences. We’re still analyzing the data because there’s so much, and we have probably about 30 scientists around the world helping us do this, including appetite experts and sleep scientists. We’re finding things like, time since the last meal to be important. We’re finding all kinds of carryover effects from the last meal, from the day before, how much fiber they had… Every time we look here, it gets more and more complex, really, the way we’re looking at the data to pry it apart.
Ari Whitten: Did you look at someone’s baseline diet? For example, if they were consuming more of a higher-fat diet, or higher-carb/lower-fat diet? Did you examine that? And I’m curious—I would expect it to predict somewhat of the responses someone may have to carbs or fats in the direction of—to the extent that the diet is more similar to their baseline diet—they probably will have better responses, as opposed to if somebody’s eating a higher-fat/lower-carb diet and then eats carbohydrates, they probably will have a worse response to those carbohydrates.
Dr. Tim Spector: We haven’t analyzed it in that way, but we did put in that habitual diet for the last year into the model, and it explains, again, a few percent of the variation—
Ari Whitten: Interesting.
Dr. Tim Spector: —but doesn’t, again, dominate it. But we haven’t tried to piece apart the extremes to divide them into high-carb groups and then go onto these… But in a way, we were giving people a combination of high-fat and high-carb stresses. In the hospital day, basically, we were giving them 50 grams of fat and 70 grams of carbs—so it was a typical U.S. feast of sugar and fat together—and seeing the response. And then the second meal—the lunchtime meal—is there to push the fats up, and that was a more carby meal. So they’re slightly artificial meals on that day, but yeah, habitual. And we’re looking whether the long-term diet has more effect than the short-term-diet-only responses. We haven’t got that data yet, we’re still looking at that in the moment.
Ari Whitten: Interesting. Okay, so, genetics, obviously, if you’re looking at identical twins, don’t explain basically any of the variation.
Dr. Tim Spector: A little bit. A little bit.
Ari Whitten: Still a little bit. I mean, they are looking at variations between pairs of identical twins, so one pair versus another pair, right?
Dr. Tim Spector: Within identical twins, yeah. You’ve taken out the genetics so any difference between them has to be non-genetic.
Ari Whitten: Right. So having said that, there’s lots of variables that are influencing it by—it sounds like between 2 to 10 or 15 percentage points. What were the top factor or the top couple of factors, as far as the things that had the biggest influence?
Dr. Tim Spector: Well, obviously, for sugar, it’s things like the amount of grams of carbohydrates in that meal…
Ari Whitten: No, but, what were the individual variables within that person that are influencing their response?
Dr. Tim Spector: Well, their gut microbiome, their health status, their BMI, their resting fat levels, their age, their gender… And then there were these [inaudible] unexplained factors we haven’t managed to get across yet—so, none of the above—but we’ll probably include things like the meal context. So, again, as I was saying, how long it was before they last ate, and how close to [inaudible] waking up it was, which we think is a proxy for their circadian rhythms. Because we gave everyone something to eat at same time, but if people’s metabolism and all is on a slightly different clock, then that’s going to be an individual factor, which we need to tease apart.
Ari Whitten: Right.
Dr. Tim Spector: We haven’t managed to—even with over a thousand people—we still haven’t managed to get all of the factors explained with this one design, and I think we’re going to have to go on and do other ones to specifically look at things like meal times. I have done some experiments on myself. For one day, I just ate these high-fat/high-carb muffins every three hours, and I saw my sugar spikes just going up and down like this, and first thing to say was it was probably the most horrible day I’ve had.
Ari Whitten: In terms of your what, subjective [energy]?
Dr. Tim Spector: Yeah, I was trying to work, I was trying to write my next book, and my brain just wouldn’t function.
Ari Whitten: Interesting.
Dr. Tim Spector: I was getting real big peaks up into the diabetic range—because I’m pre-diabetic and then I was falling down again—I didn’t go below baseline, but I still felt unwell in a way that I didn’t really boot. Other people say they’d get these sugar crashes and they feel terrible, but I never believed them—you know I’m, again, a skeptic—until it happened to me. So I was getting these big peaks, but what was interesting in me is that as the day wore on, the peaks got less. So I had less of a peak when I had my muffin at 8 and evening than when I did at 8 in the morning.
The theory suggests that we should get worse as the day goes on, because all these theoretical studies based on about 20 people, about “eat all your meals early” because our metabolism is best-suited, and our cortisol, and all our hormones—and it’s all theoretical and all based on very small numbers of very young, very fit people. But it seems that with age, we all change anyway. What might’ve been our perfect time when we were in our twenties, maybe quite different in our fifties.
Ari Whitten: Yeah. I would add, there’s another factor—and I’m very familiar with all the literature around meal-timing and circadian rhythm—I would say the other big factor there is whatever your habitual pattern is, because the circadian clock and the hormones that are influenced by the circadian clock, are entrained by meal-timing. So if somebody is used to and has been consuming their biggest meal of the day in the evening for years or decades, that’s probably very well and trained on the circadian clock, and you’re probably going to see really bad responses in the immediate short-term when somebody starts having really big meals in the early part of the day.
Dr. Tim Spector: Yeah, I agree. That’s a really good point. And that’s probably why the Spanish should never eat before 9PM. I’m really healthy and, Americans in the Midwest who eat at 5:00 PM are really not.
Ari Whitten: Yeah. My wife is Chilean and when we go to visit her family down there, they often stay up really late till midnight. Dinner happens at 11PM to 1AM. Drives me crazy because I like to be asleep like an hour and a half before they start eating dinner. And then I like to have my dinner three hours before going to sleep. So yeah, that very much disturbs my whole rhythm of everything that I’m doing, and I do not feel good eating that way.
I was watching an interview of yours prior to recording this with you, and one of the things you said that really stuck out to me is: you said that, “You can know way more and you can predict way more about a person’s health from looking at their gut microbiome as opposed to their genes.” I think that’s a really profound statement, again, given what we talked about earlier about how all this focus and popularity of people trying to look at their genes. Why do you say that? And what is the evidence to support that?
Dr. Tim Spector: The evidence really comes from, first, the fact that in our discussion about genetics, for most common diseases, we aren’t able to really give you a good indication of whether you’re likely to get it. And no genes are going to be the same when you’re aged three months or when you’re aged a hundred. So if you’re asking for a snapshot in time about your health, I can’t tell whether I’m dealing with a three-month-old baby or a centenarian. So there’s the failings of the genetic system for the common disease. I think I could tell if you’ve got a really rare one-in-a-million syndrome, but otherwise it’s not particularly useful. Whereas your microbes can give you a snapshot of how your body is at the moment. They change day to day, they change year to year; but there’s a group of them that stays constant to you over your lifetime that seems to go through thick and thin with you, and will tell me, in very crude terms, how you’re doing. It won’t give me precise information about what kind of disease you might be suffering from, but there’s a measure of—in your gut microbes—which is called diversity, which is a sort of rough way of looking at the number of different species you’ve got compared to other people.
We know that in the West, we’ve lost about 40% of our species compared to hunter-gatherers. And we know also from doing some of our citizen science projects, where we teamed up with the American [inaudible] to the 11,000 people, that the people with the most diverse gut microbes have the least diseases. And that’s true in virtually every study of any decent size. It’s a very crude measure, but it correlates very well with a number of diseases. The least diverse your gut microbes, the more likely you are to have all common diseases, and the opposite is true. That’s really where that statement came from. So if I had to look at your DNA and get the sequence, I couldn’t tell whether you’re healthy or not, or whether you’re going to be at risk in 30-years’ time of getting something. But if you’ve got a really a rough-looking microbiome—it’s non-diverse, you haven’t got the healthy microbes that keep your immune system going, I’m pretty sure that you’ve got either a number of diseases at the moment or you’re, in the next couple of years, you’re going to develop them if you don’t change your diet and your lifestyle.
Ari Whitten: Now in your PREDICT Study, did you analyze the microbiome of these individuals and did you—maybe you mentioned this in passing, I think—but did you determine that differences in microbiome were a significant explanatory factor in the variations of responses that people had? And where there any particular bacterial species that you found where were key to that?
Dr. Tim Spector: The analysis is still ongoing about the detail. In general, when we look crudely, we found that microbiomes explain the responses to sugar, insulin, and fat—but it seems to be much more related to fat than anything else. So for people that were healthy at baseline—up to 20% of their fat response—we could predict from some of the microbes they have. We’re not quite sure which way around it’s going, whether you need the microbes first to digest the fat, or having a regular fatty diet builds up the microbes that help it—we’re still exploring that. But these are quite big effects for fat metabolism, which haven’t been looked at before. But we’re still looking at the individual microbes. We haven’t presented that data yet, but it’s looking pretty exciting for fat and for the sugars. Particular microbes are floating the surface. Not particularly exciting, but we do, again, see this general trend for diversity, which is well-known from other literature because low-diverse people tend to have prediabetes, or diabetes, or metabolic problems.
So I think the question is whether these will ever be good enough to be predictive. I suspect they won’t on their own. What we’re finding with the microbiome is that there’s a lot of redundancy. So, one microbe might do one job in one person, but in another community, it’s going to be replaced by something else. I think we’re going to start changing our views that no, [it’s not one] disease that is caused by a lack of one microbe, it’s going to be how do you get these good communities all working together to provide the right chemicals.
We may be, in the future, moving from this: measuring the microbiome with genetics, to looking at the chemicals. And that’s another part of the research we’re doing at the moment—is to say, they eat lots of different microbes [but they all] produce the same chemicals, like brain chemicals, like serotonin, and things, so what you’re really interested in is the chemical, not the microbe. And you don’t care how you get there. You just want lots of nice serotonins so you can feel happy. And whatever feeds that is [good]. So there’s lots of data, it’s very exciting. We’re just scraping the surface of all this stuff; we’re collecting it. And the study is still ongoing because we finished the first part, PREDICT 1, but there’s ongoing in the U.S.—we were doing a home version called PREDICT 2, and at the moment, we’re still recruiting to get a whole other set of people who are very keen to log their own foods and find out about themselves and, again, contribute their data so each time, our prediction models get better.
Ari Whitten: Now are you looking for more people to sign up for that?
Dr. Tim Spector: Yes.
Ari Whitten: Most of my listeners—80%—are in the U.S. so if you’re specifically looking for people in the U.S. to sign up for this, is there a place where they can go to sign up?
Dr. Tim Spector: Yeah, they go to the joinzoe.com website, and the details of the study—the PREDICT Study—are on that. And if they’re quick, they can get signed up. Particularly looking, at the moment, for any ethnic minorities; we’re a bit short of. They’ll get preferential treatment, but anyone who’s keen is going to be a really good logger. We’re also keen to have those. And so far, the U.S. participants have been amazing, I’ve never seen such super keen people. Some of these health-hackers, they’re blowing our minds in how organized and excited they are about finding out about themselves.
Ari Whitten: Nice. And just to be clear, for people who might be interested, is this something that costs money for them?
Dr. Tim Spector: Not at this stage, so this is all free, and they get shipped everything free. At this stage, it’s all a science project.
Ari Whitten: And what are the benefits for them to participate?
Dr. Tim Spector: They get to know a lot about themselves. They get to test lots of the gear, of course—they’ll get to test the CGM devices, the exercise activities, sleep monitors…
Ari Whitten: CGM, for those listening, is the continuous glucose monitor.
Dr. Tim Spector: …And at the end of it, they’ll get—on their app—predictions about what we think the foods they should be favoring above others, and get to test first before anyone else. But the basic thing is we want people to be excited in the science, and feel that they’re helping the community, as well as helping themselves learn more about their own bodies.
Ari Whitten: Yeah. Now, given that these responses that you’re measuring with these different tools—blood sugar, blood fats, insulin, and even something like the microbiome—are all heavily influenced and extremely dynamic in terms of being influenced by these environmental inputs? Like what you mentioned, what are you actually eating is going to influence your microbiome, is going to influence some of these responses? What is your level of body fat? What’s your level of stress? How are your circadian rhythm habits? How often are you eating? How much are you sleeping? And so on. All these different levels of nutrition and lifestyle habits are in dynamic interplay with all of these things that you’re measuring. So, would it also makes sense to do this kind of measurement, let’s say, once every 3 months or 6 months, and then kind of get an updated list of, maybe, what foods your body is responding to best now, given the changes that you’ve incurred over the last 6 months?
Dr. Tim Spector: Absolutely, yeah. We see this as a dynamic tool going forward because people’s habits will change. They change in the seasons, and they might’ve changed diets completely. So this is evolving. We’re not saying we’re solving all the problems, even when some commercial product comes on next year. It’s going to be relatively crude at first, getting better and better all the time, and we will be going to something that says, “Yes, look at your app now. Put in your current features, including how much sleep did you have last night? How much exercise have you done in the last three days? What did you eat yesterday?” And we can adapt your profile for what ideally you should be eating now because this is a state of your engine at the moment, which might be different to what it was three months ago.
So I think as we get more data, this will be much more of a dynamic tool. And once we got these basic algorithms, we can start changing it and there’ll be a little [diet]. You say, “I only had four hours sleep last night. For me, does that matter?” For some, it may not; for others, it might be a really big impact, and say, “Okay, you should shift towards more fat or more carb”—whichever it is. So people can start playing with it, and we may even have a tool that people say, “If tomorrow I only have one hour sleep, what’s my best diet? Or my sustain night? What am I going to have on Sunday for brunch?” So yeah, I think it’s going to be really, really fun to be doing this.
And I think we’ve had a great enthusiasm from people wanting to take part and realizing that we’re all our own experiment. We want millions of people that are doing these n=1 experiments. It’s the only way I think we’re going to make progress here against the commercial interests that just want us to follow their particular pattern, that “there’s only one true way” and “there’s only one true diet” and “if you don’t succeed, you’re a loser”. And 99% of us end up being losers.
Dr. Spector’s tips for optimizing your health
Ari Whitten: Yeah, absolutely. Having said all of this—we’re pretty much out of time here so I want to wrap up, but—are there any key takeaways, having said the caveat of “there’s no one-size-fits-all”, let’s say, nutrition recommendation, for example—do you feel there’s any key takeaways from the information that you’ve gathered so far that can be translated into any sort of practical advice that is somewhat universal? Like, for example, people are neglecting their gut microbiomes and there’s a widespread deficiency of fiber intake and diversity of fiber intakes, and it’s leading to microbiome problems; or circadian rhythm is a big deal—is there any sort of big takeaway that you’ve found in your research thus far that you want to leave people with?
Dr. Tim Spector: Yeah. Well, I don’t want to give the impression that the only way is some commercial tool that’s going to cost money, because that’s rather depressing for many people who may not be at a fault. But yes, definitely some takeaways. The gut microbiome is key to our health. We need to maintain it, keep it as diverse as possible. The way everybody can do that is to eat a diverse array of plants. All studies suggest that the optimum is around 30 different plants a week. So we go to supermarkets where they have 30,000 products, and we always end up with the same things in our trolley; we all get into food ruts and go for the same variant. 30 plants sounds a lot, but actually that includes nuts, seeds, lots of herbs, and spices, so it’s actually relatively easy. Either do sprinkle things on your yogurt or whatever. So that’s the first thing: double the fiber. Fermented foods, we don’t have enough of those in the U.S. Real yogurt without any vanilla or flavorings or sugar or artificial stuff. Kefir fermented is fermented milk; it’s like super yogurt, massive probiotic, really good for your gut. Kombucha, also a well-known in California. And then polyphenol foods. We could do a whole other podcast about that—but berries and nuts and coffee…
Ari Whitten: Actually, I’ll just say right now, if you have a lot of expertise on that subject, I would love to do another podcast with you because that’s a pet interest of mine and I’ve actually been looking for someone who’s an expert on that to interview them. So maybe we’ll have to do a part 2.
Dr. Tim Spector: Yeah, I look forward to it. But gut health is really important, varying your diet; some diversity is really important. We’re far too boring in our choices of foods and we’re finding that also, we’ve not found anyone who responds badly to fats, particularly good fats. So at the moment, we’re looking at things like [olive oil], you really can’t have too much of it. At the moment, you can have too much bad fats, but I think we’re treating all fats the same and we have to start separating those out. And basically cutting out all highly-processed foods and artificial chemicals and sweeteners. If you do all that, you can’t go far wrong. Before we get to this personalization, you’d be in a very good position, then tweak your personal metabolism and health in exactly the right way and still enjoy life and enjoy the fantastic diversity of foods without suffering.
Ari Whitten: Yeah, beautifully said. Dr. Spector, thank you much for your time. It was really a pleasure, I enjoyed this conversation very much. I’m sure everybody listening did as well. Again, for people listening who want to participate in the PREDICT Study—and please, everyone listen, keep in mind, I have no vested interest here, I don’t get any kickbacks or anything from Dr. Spector if you sign up, this is purely a matter of furthering his scientific research so you can help the scientific community and human health, humankind, more broadly; and you can simultaneously have a selfish benefit of learning more about yourself and what diet is likely to be more optimal for you. So where can people go again to sign up for that?
Dr. Tim Spector: So the website is joinzoe.com
Ari Whitten: Okay. And Z-O-E, or Z-O-E-Y?
Dr. Tim Spector: No, Z-O-E, as in the ladies’ name, Zoe, which is the Greek word for life.
Ari Whitten: Beautiful. Again, Dr. Spector, thank you much for your time, I really appreciate it. I know it’s late over there in the U.K., so have a wonderful evening.
Dr. Tim Spector: Thank you.
The PREDICT study (1:21)
The problem with personalized health advice (13:39)
How lifestyle factors affect your health (29:45)
Dr. Spector’s tips for optimizing your health (57:21)