A (0:00)

People that have a good diet are potentially more likely to have better physical activity, to take more supplements, to not smoke. All of these things could comprise a healthy lifestyle set of behaviors which can positively affect their disease and so accordingly, there's strong potential for diet based intervention to improve Ms. Risk and modulate disease.

B (0:19)

Course welcome to Living well with Ms. This show comes to you from Overcoming ms, the world's leading multiple sclerosis healthy lifestyle charity which helps people live a full and healthy life. Through the Overcoming Ms. Program, we interview a range of experts and people with multiple sclerosis. Please remember, all opinions expressed are their own. Receive monthly tips and ideas about Living well with Ms. By signing up for our newsletter@overcomingms.org and now let's meet our guest. Today's episode features audio from the Diet in Ms. What the Evidence Tells Us webinar which was part of the finding hope with OMS 10th anniversary edition webinar series Overcoming Ms. Medical advisor Dr. Jonny White speaks with Neuroepidemiology Unit senior research fellow Dr. Steve Simmons. Yep. About the evidence around diet in Ms.

C (1:23)

Welcome everybody. Thank you for joining us on what is a very cold and snowy morning in Northern Ireland. Looks a little bit better in Melbourne where Steve is. I'm really, really honoured, privileged and really grateful, Steve, that you're able to join us today. It's a great area of excitement for me to be able to talk to somebody about diet and Ms. So I just want to introduce you formally to our audience. So Steve Simpson Yap is a Senior Research Fellow at the Neuroepidemiology Units at the University of Melbourne and many of you will have heard of that unit. It was founded by Professor George Jelinek, who some of you might know. Steve is originally from Indiana in the United States and he undertook his graduate training in Hunter College. His postgraduate training was at the University of California, Berkeley, and then he moved to Australia and Tasmania, specifically where he did postdoc in the Menzies Institute in Tasmania. His current work is as an epidemiologist and specifically examining the connections between lifestyle and diet and autoimmune conditions. But specifically at the moment, we're very grateful. Multiple sclerosis. So Steve, thank you for joining us. You're very welcome and I'm really excited to hear what you have to tell us.

A (2:26)

Well, thank you for that kind introduction. I'm presenting today about diet and Ms. What the Evidence Tells Us, but particularly I'm going to kind of give you guys a little bit of a primer about what it is that goes into A study of diet and may give you an idea of why it is. I say it the repeatedly actually that there is not enough studies in diet in Ms. So who am I and what do I do? I'm an epidemiologist. What an epidemiologist does is we take data sets of demographic, clinical and lifestyle and some other things like genetic and MRI and other things, and we apply statistical analyses to try to assess the relationships with a goal of finding what causes what. So I want my work to help our understanding of the factors both modifiable like diet and otherwise, like for instance, your genotype or your just clinical history and so on that affect human health and disease. I'm not a clinician. However, there are a lot of doctors that do epidemiology research, so clinicians to do EPI research. But I'm not a clinician and so unfortunately I can't give you any guidance about your particular conditions. But even though I'm not a clinician, I'm not right at the front lines applying my work. It is my hope that the work I do do will lead to information that your clinicians can use and help guide their treatment decisions in positive ways. And really, of course, it's my fervent hope that the work I do will lead to a cure for Ms. Or at the very least a treatment for Ms. And I believe that a lot of the work that we do that I'll touch on a little bit here does show, for instance, ways that diet or supplement use or stress reduction or things like that. The modifiable programs, lifestyle factors that are in the OMS and other lifestyle modification programs, they might end up being a component of clinical care. So why do we study diet, nms? Does it even matter? Well, I believe that there is relevance for diet, nms. There's significant what we call biological plausibility, which is like if I told you for instance that your shoe size may affect the number of lesions on your mri, that doesn't sound like there's any reason to that, but there's a lot of reason to think that diet may actually affect your disease. So it can obviously, as you might think, affect your body mass, index how heavy you are and affect that because that's been associated with Ms. It can modulate fat and vitamin levels in the blood. Those have each been associated with Ms. They can affect what's really extinct, the microbes in your gut because they've been found that those microbes, which helps you digest food and so on, actually kind of talk to your immune system across the gut associated lymphoid tissue or the gulp. And so it's possible that if you have diet that kills off some good microbes in your gut and they get replaced by unhealthy microbes that could potentially lead to a more inflammatory state, which could be worse for your disease. And there are other pathways, there are a lot of potential pathways by which diet can affect disease. And also just frankly diet covaries with other behaviors and environmental exposures which have been associated with Ms. So things like smoking, physical activity, supplement use, like, with like people that have a good diet are potentially more likely to have better physical activity, to take more supplements, to not smoke. All of these things could comprise a healthy lifestyle set of behaviors which can positively affect their disease. And so accordingly, there's strong potential for diet based intervention to improve Ms. Risk and modulate disease course. So diet programs, modification programs have been proposed for people with Ms. As you be well aware, you're familiar with the Overcoming Ms. Program. Due to promising evidence from the observational studies that I'll touch on and the aforementioned biological plausibility, there's actually been a number of diet programs that have been proposed that may benefit people with Ms. So these programs are idiosyncratic and varied. And importantly, they're not complementary per se. Some of them actually are quite distinct. For instance, the Overcoming Ms. Program recommends no land animal protein, no meat, no dairy, lots of seafood. But the McDougal program, for instance, recommends no seafood at all. It's all entirely plant based. So those can't both be right. And then for instance, the Terry Walls elimination diet actually goes hard the other way. You eat like a hunter gatherer did back in the caveman days. And so that's that conflict. The astronomy best bet has some alignments with things, and so they can't all be right. And so there is the potential that having a healthy diet may benefit your disease. But it's important that the evidence be solidly there. George would say this, George Jellinek would say this too. It's important that the diet program that you propose to people has the evidence underlying it. And some of the work that we're doing has aligned with some of that. In addition to the diet programs that people have put forward as saying this may benefit your ms, there are other diet programs out there. For instance, ones you've heard of, like Mediterranean diet or low fat, low carb, the ketogenic, vegetarian and vegan diets, and intermittent fasting. These diets have been found that potentially may improve your heart health, may help your Health in general. For instance, the Mediterranean diet. And so people with Ms. Attempt to either follow these diets or they take little bits of these diets are like, well, if these diets are all healthy, then some parts of it must be good. And unfortunately, there's a potential that you put all the energy into following and changing your diet to follow the Mediterranean whole diet, for instance. And if the evidence isn't there, you may be investing your effort inappropriately. And so nonetheless, diet modification in Ms. Is very common. So there's a large study in the United States that assessed over 7,000 people with Ms. And they found that almost half, 41%, reported following some diet programs. And a large proportion of those were following it to improve their Ms. People following the low carb Mediterranean Paleolithic especially large proportions were following it to improve their ms, even though the evidence is not quite there yet. In our own study, the holosom study, almost half of the participants reported following a specific diet for their Ms. We asked them specifically, unlike the US Study, we said, do you follow a diet to help your ms? Almost half said, yes, I do. As you can imagine, a large proportion of those said they were following the overcoming Ms. But even still, people were following the Terry Wahls or the Swank diet. And what these results show is that people with Ms. Are motivated, they want to take charge, they want to take action that's going to help their disease and not just wait on, you know, taking the drug, whatever the neurologist tells them. And they want to improve their health. However, it's important to ensure that they are making the right health behavior changes. And there are some that there's pretty foolproof. You know, you could go and quit smoking. If you're a smoker, quit smoking. Their evidence is so solid for that, that that is unchallengeable, that will help your lung health and your general health. But there's also some strong compelling evidence that it will improve their Ms. But there's other evidence that can really only come from solid epidemiological studies like those that I'm going to propose in here. And so some of you that know a little bit about epidemiology may say, well, geez, why don't you just go do a clinical trial? That's what they do with a drug. They get a drug out of the laboratory and they give it to a bunch of people. Some people get a sugar pill, some people get the actual drug and they see if it works. And there's indeed a strong desire to undertake such trials to assess whether Diet affects Ms. And some studies, like for instance, Terry Wahls diet, have been assessed in clinical trials. Little ones. However, I say this skips a critical step in ascertaining whether you should bother getting to the clinical trial stage, whether diets actually associate with the outcomes. And that actually comes from observational study data, which is what we do in the nau, for instance. Is diet associated with Ms. And clinical progression? Well, what kind of diet? Does it just have to be a healthy diet in general? Does it have to be a diet that excludes all dairy or excludes all meat or. And once you find out what kind of diet, what does diet affect? You can invest millions of dollars to randomize and follow people for five or 10 years because you're looking at, say, fatigue. And well, we can tell from observational studies that actually the evidence about diet and fatigue says there may not be something there. So maybe you're wasting time doing clinical trial for that. Instead you should really focus on what we actually show, which actually disability in observational studies seems to. So disability progression by EDSS or so on that may actually be worth your time. So it's important, though, to do these observational studies to inform the clinical trial development so that you can actually do the right diet with the right exposure in the right population. Because there's also potential that it may only be beneficial in subgroups of people with Ms. So you want to know that information. And the other important thing, though, is once you've found out what kind of diet you want to do and what clinical. What outcome you want to do, it's not as simple as just till a and till bio. Clinical trials are very complicated, but much more complicated than a pill. I wish there was a way I could just give everybody in clinical trial a pill. And I say, this has all the food and nutrition UD on it. You just take this every morning. And then I can count up how many pills they take and I know exactly what they've been eating because they've only been eating my pill. Fortunately, our diet is not like that. This isn't the Jetsons yet. You have to actually measure their diet. And measuring diet is hard. It's also very difficult. If you say somebody with a clinical trial, you need to follow only this diet. It's hard. I mean, I've seen. I've gone to restaurants with George and he has to follow a diet that follows the overcoming Ms. And sometimes the restaurant doesn't have something on the menu that aligns with that or it only kind of doesn't. So he Ends up, you know, let me be able to have a salad or something. And he can't actually have something negative. And it's hard to do that. And some people on a clinical trial might say, well, they don't have it, so I guess I'll just have a burger because that's all they have on the menu. And that would really mess up our ability. If they're on the intervention, it really messes up our ability to actually see if the diet that we've allocated them to actually works. And that's why it's very hard. And so what I say, it's very important to first obtain the necessary evidence from observational studies, as I said, to inform what kind of diet and what kind of outcome. And then once you do it, it's really doing clinical trial is possible, but it's a lot harder than a drug trial. So Ms. In other conditions, how do we assess diet? Well, first, do we need to measure diet in our study at all? What's our research question? What do we want to find out? If we're doing a study about physical activity, for instance, we probably need some measurement of diet. But if we're doing a study, say, of blood levels of exposure to viruses, then maybe you don't need to measure diet because again, diet is complicated. In order to get an adequate measure, as I'll show you, you have to give them a lot of questions, and that can be overwhelming and tiring for people that are doing your study. So you ideally want to make your questionnaire as short as possible. So let's say we do we need a measured diet? Well, we believe, yes, we do. Next question. What kind of study design do we need to implement? So really, I mean, there's a little more complicated than this, but it really boils down to cross sectional studies and longitudinal or prospective studies. A cross sectional study just measures your exposure and your outcome at one point in time. And there have been a lot of these because they're relatively easy to do. You could just recruit a hundred people in your clinic and ask them, you know, what are their lifestyle behaviors? And then you look at their clinical records and you know what their level of disability is. And you can say, based on this, we find the exposure is associated with less disability, which is all well and good, but you don't know if is, is it because the better exposure causes them to have a better outcome, or is it because the people that have a better outcome, less disability, whatever, they're able, they got more bandwidth, more energy to follow some healthier lifestyle and so you could find what we call reverse causality, where it's actually the outcome causing the exposure, rather than what you really hope for, which is your exposure to is causing your outcome. And the way you can better assess this is doing a longitudinal study where you measure exposures and outcomes over time. Ideally, you get your exposure at time point one, and then you follow them for some amount of time, six months or two years, or as we've done in Hollison now, seven and a half years. And you can say what they were doing at baseline, is it associated with their outcome way later? Because if you do, there's a better chance you can conclude that this actually means the exposure caused the outcome. It's much less likely if you go over a long enough timeframe that you're still seeing that reverse causality. So that is what we need to think of when we do an observational study in terms of design, in terms of once you figure that out, you really need to figure out logistical considerations. And these inform what approaches and measures you might take to assess diet. So how many people are you going to recruit for that? We actually do a thing called power calculations that says how many people you actually need to find the statistical association? How are you going to follow them and interview them? Are you going to do it, do it on the Internet? Are you going to call them up on the phone? Are you going to bring them into the clinic? Are you going to go to their house? There are different modes of this that cost more, obviously, if you send your people out to where they live, or it's potentially more burdensome on your participants if you ask them to come to the clinic, but then you know that affects the quality of the data you may get. And then also, as I said before, do you need to do it a single time point or multiple measures over time? And if it is multiple measures, how many measures? These are all things that go into it. Once doctors and scientists are putting together these studies. And then finally, if you've figured out all the design of your study and how many people you need and how many, how long you're going to follow them, how are you actually going to measure diet? Well, it kind of depends on how comprehensive the rest of your questionnaire. If you don't want to study like Hollisom, if any of your participants in Hollisom, you may know that diet is one part of a much larger question here, because we wanted to assess a lot of other factors that gives us a lot of ability to then control for those other factors. But it then sometimes Means that you have a very long questionnaire that not every participant with Ms. Is going to want to sit and complete all the way through the end. Sometimes we can link with other data sources, so instance, their medical records to reduce what we need to ask the participant. Because then we can get information about, say, what medications they're on, what kind of Ms. They have, and so on from there. Then we don't need to bother, you know, adding that to the questionnaire. Also, if you can, if you can get bloods or other biological measures, this can really help substantiate the findings that you might find with diet exposure and its association with outcomes. So if I ask you what foods you're eating, and then I take your blood, I can measure what vitamins you have, what your lipid levels, you know, your cholesterol and other things. And then if I find a beneficial association with diet, I can help substantiate that by seeing, oh, it's actually cholesterol is also associated in the same way. And you can also potentially look to what extent cholesterol or other factors is acting to mediate the association that you see with diet. And then finally, how detailed a measure of diet do you need? Sometimes you can have a fairly superficial measure. There are short surveys of diet that are only five questions long. Sometimes you can simply ask them a very basic proxy question. How many serves of fruit and vegetable they have, they have each day? And generally people that have more fruits and vegetable serves are probably have a better diet in general. Or do you need a more detailed and comprehensive measure of diet intake? If you do, there are various measures out there. So there are questionnaires. And you gotta balance the length and the detail you want to get versus really what a participant is willing to sit through. Because if I could ask the participant about every single thing that they've eaten in the last year, I would love to have data. I would love to have that data to work with. Fortunately, I probably get like two people that would give me every single food that they've eaten in the last year. And so it asks about individual foods and says, you know, how often you had it. Never less than once a month, all the way up to six or more times a day. And you ask them about different serves of different, either fruit in general or individual fruit. Strawberry, raspberry, kiwi fruit. And with that you can get a lot of data. But again, it's a matter of what the participant is able to put through. If you've already asked them a lot of questions in other parts of the survey and, and then you give them a book of a diet questionnaire, you may not get as many people actually answering all the questions. And that's problematic for us. Another way that we can assess diet is called a food diary. This is more comprehensive, but it's basically you literally ask them about every single food that they, a food beverage they had in the last couple of weekdays and last weekend day. So you generally try to get three days, a couple of weekdays, one weekend a day. And that the problem is that it's very idiosyncratic. It requires a lot of processing as opposed to that questionnaire which is just one size fits all for everybody. And then the best way really is weighing their food intake. If you could actually take your participants and put them in a laboratory and just follow them and make them live there and follow them over time, it would measure every single bit of food that they had. But as you can imagine, we're never going to do that. Nobody's going to come in and do that. And it's also unethical. It's a significant burden of participant and it would be data that we really could not use. So, uh, the ideal measure when we do diet and really any exposure is we want to get it as close in time to the actual exposure as possible because there's issues of recall bias and error. So if I asked you what you ate yesterday, a lot of you probably be able to give me a pretty reasonable answer, but if I asked you what we were eating this time last year, kind of guess, but you're not really going to know. And there's also an issue of bias, which is to say somebody who's really motivated to care about their lifestyle, they may actually keep a record of this or may just really be switched on or remember their lifestyle in a way that if I was recruiting some, you know, control, you know, random healthy person off the street and asked them what they ate a year ago, they probably look at me like I was an idiot. They wouldn't know. But that difference between your cases and your controls is something you have to consider when you're doing a study like this. And as I was saying here, so short term recall is the best, but again, it's good to get serial measures over time. But it's all. These are all the considerations that go into it. And so the conclusion of all this is diet studies are complicated. It's hard to get accurate and sufficiently detailed information about diet from participants, particularly if you then ask them to come in every year or every couple of years. Over multiple years of a study, because that's what you really need in order to get that level of detail. And when we do studies like this, we have to balance the optimal ideal of what detail we'd love to get through our participants and what they're willing and able to give us without getting so overwhelmed and tired that they either come in once and then we never see them again, or they just get halfway through the questionnaire and say this is too much and then we don't give that data. Whereas if we'd taken the time to try to get a shorter questionnaire that was actually able for them to complete it reasonably even still, a number of studies have been done to assess diet's relationships with clinical outcomes. One of the basic ones was done by Swank Ron Swank in Norway back in the 1950s, where you've actually looked the frequencies of Ms. Inland and on the coast and found that there were significant differences in the diet behavior between inland and on the coast, which suggested diet may be related to Ms. There have been some studies looking at individual foods and nutrients. Study here in Australia compared people with Ms. And controls and it found that overall dairy was not associated with case status, but particularly yogurt intake was associated with 11% less disease risk. But importantly, these studies are only cross sectional. They only measure these people at one time point, which means we can't derive causal conclusions. And several studies have been published using their own data set, the Holosung International Perspective study, which we showed that diet was prospectively associated with disability, both cross sectionally where we measure them at one point in time, but importantly prospectively. We queried their diet at one time point and then assessed it with subsequent change in hearers with disability. Higher diet quality on the X axis is associated with less disability on the Y axis. And indeed we found that those in the top two levels of diet quality had 0.3 points less disability accrual and those same people also had 41, 36% less risk of increasing their disability over time. And then when we that was the up to the 2.5 year time point, when we continue that up to the most recent, a 7.5 year time point. And now we're showing that those people that are in the top levels of diet quality have less disability progression and also almost half less risk of increasing disability over the 7.5 years. We've also assessed, in addition to disability, we've looked at other outcomes including fatigue, depression and quality of life. As I said earlier, fatigue. We haven't actually shown a prospective association, which may suggest that you may not want to invest in that, at least based on the evidence so far, if you want to design a clinical trial. But looking at quality of life, we found that diet quality is associated with better quality of life. So those people at the higher levels of diet quality have higher quality of life, whereas those people that are in the lower quality have roughly 0.5 points less, less physical and mental quality of life over time. Their needs yet exist for high quality evidence. So a number of studies have been conducted assessing diets, association with ms, risk for disability, progression with quality of life. But these studies are lacking. They're often very small, which, in terms of the number of people, they're often only cross sectional, so measured at one point in time, which means you can't infer a causal conclusion. They're also followed for relatively short times, often only up to one or two years. Whereas if you're looking at an outcome like disability, you really need five or 10 years. And also there's very heterogeneous modes of diet assessment. You can't ask, unfortunately, people about how many oranges they eat because that's, that's not really efficient. And it's also not really reflective of how we think diet's going to affect disease. It's not one factor. It's going to be a complex of all the factors you eat. And also, even the diet quality scores like we've used in holosom just give an indication of some predefined diet quality, but not necessarily about what the constituent elements may be or what kind of diet they should follow. It really just shows that healthy diet is potentially beneficial. And so I propose a holistic approach to diet, which is the index analysis method. So index analysis just sets a predefined schedule of some food or beverage intake, maybe just a general healthy diet. It may be modeled upon some existing style of diet, like the Mediterranean diet. It may be based upon some research developed for the purpose. So, for instance, you may have heard the DASH diet for improving hypertension, or of course the Overcoming ms, or the Terry Wahls diet, which is proposed to benefit Ms. It may be algorithmic. So there's something called the dietary Inflammatory Index, basically gives you a score for how inflammatory your diet is, even as you know Ms. Is an inflammatory condition. That is very relevant. Regardless of the index analysis method, though, it follows some schedule of foods, beverages and nutrient intakes as per some protocol. And then you score people up based on what they're eating with, basically how well their diet aligns with what the Model diet is Mediterranean diet, OMF's diet, anything. So an example of this application is my colleague Cindy Black assessed the Mediterranean diet in the Alzheimer's case control study and they assessed that the Mediterranean diet, so the alternative Mediterranean diet index found that higher diet index score was associated with less disease risk. So what do I conclude from diet studies in ms? There's some promising results unfortunately because they're often cross sectional, it means we can't make causal conclusions. And there's relatively a lack of studies in general assessing diet's role clinical progression. It's actually relatively easy to recruit a bunch of people with Ms. And a bunch of people that don't have Ms. And, and then compare them. But if you really want to know about disease progression, you need to follow people with Ms. Over time and that's harder to do. So what I'm proposing to do is use a beautiful study called the UK Ms. Register, which some of you may even be participants in. This is a national survey of people across England, Wales and Scotland that has been following people with Ms. Since 2011. They actually recruited over 10,000 people. It's got a core survey that people get sent in their email every six months which queries their physical, like clinical symptoms, their demographic characteristics and a small number of lifestyle characteristics. But importantly, my colleague Shelly Koh, who's based at Oxford Brookes University, did a large diet questionnaire in 201516 and we have just done that, that same survey again just last month and we got another 3,000 people on top of the 2,200 people that they got in 2015. And this will allow us to do is we're going to be able to assess diet quality in general, going to use those index analysis methods. We're going to score people up on the Mediterranean diet. I'm going to try to score them up on the OMS diet, the WAHLs diet, all these other diets. And we're going to assess whether the evidence exists for diet quality in these particular diets to be associated with Ms. Progression and particularly what kinds of progression. Because the UKMS Register has relapse as disability, as fatigue, it has depression, has anxiety, it has MRI parameters, it has quality of life, it has all these things which we can inform so much about what, what kind of diet and what does it affect. And yes, and importantly we're going to have this follow up over 10 years which is exactly the kind of long term duration that one has never been done for diet anymore. That's the longest it's been done is around Hollisom study with seven and a half years. But we're going to have this in a very large sample of potentially a couple thousand people, which gives us that stability, statistical power that I said earlier that we need. And it is my hope that the studies that we do will give guidance to people living with Ms. Whether and how diet can really predict their Ms. Progression and how they can improve their diet to improve their life. So I want to acknowledge my colleagues in the neuroepidemiology unit, so Drs. Jeanette Reese and Sandra Meat, who Ms. Maggie Yu, our research assistant and my colleagues in the uk, Shelly Co Thanatis tectonitis, Richard Nicholas and Ron Middleton. Thank you.