Voice-over: Welcome to Spotlight on Migraine: The Professional Series, a podcast hosted by the Association of Migraine Disorders. Join us as we dive deeper into migraine topics with guests from the medical field. This episode includes educational content intended for medical professionals, but may be useful or interesting for patients as well.
This episode is brought to you by our generous sponsors, Amgen and Novartis.
Margaret Slavin: Hi there. I am Margaret Slavin, and I am here to talk with you about dietary patterns in the context of migraine prevention.
I am a PhD-trained food scientist specializing in food composition as well as a registered dietician nutritionist who has taken up the mantle of research on dietary migraine.
So there have been a spectrum of dietary approaches suggested to improve migraine across various venues and expertise over the years. These are the bins that I divide up the different dietary approaches in my own mind, and I’ve highlighted here what I most want to talk about today, and that is the overall eating patterns associated with migraine.
I want to be very deliberate about this. I am not intending to ignore food triggers or supplements or weight management, but some of these are being covered by other speakers, and I can only do so much in one presentation. So I’m focusing on eating patterns today.
And when I submitted these goals originally for the March program, I said that I would summarize the state of the evidence of dietary interventions. But, meanwhile, that’s already been done for me. There’s a great systematic review that came out of Stanford and is published in Headache, and with my short time, I’ll direct you there with just a few comments here. I very much appreciated that they broke up their review into a few different categories, including dietary patterns as observed in the population; triggers arising from within the diet; and, finally, dietary interventions that have made that leap into controlled trials.
So, briefly, the conclusions probably won’t surprise anyone who’s been thinking and keeping up with the data about diet and migraine, in that there is limited high-quality RCT data on diet patterns or diet triggers. And there are a few small diet interventions assessed by RCT, but they don’t have strong results.
And my interpretation as a dietician from this is that it’s difficult to suggest that someone change their eating pattern — which we know is difficult for a whole host of reasons — on the basis of limited or weak support. Then, as a researcher, I think what’s most interesting to me is more this bottom piece, which is that most diet interventions were related to decreases in frequency of migraine attacks.
And so we know that there is a high placebo effect observed in migraine research outcomes, and more broadly, in pain research. And given that it’s often difficult to show or to demonstrate that a dietary intervention moves the needle very far on a health outcome, and especially during a short duration of some of these trials, it’s logical to conclude or at least to be suspicious of some of these results in that the placebo effect is probably at play and showing up here in some of these results.
And so then what do we do? Where do we go? How do we think about this? What comes next? So I’d like to explore this idea with you by looking at dietary interventions as a paradigm oriented to disease status. And that’s what I’ve tried to depict here. And so from this vantage, when we’re designing a dietary intervention relative to any particular disease, we can see that diets realistically can be tailored to do one of two things: either prevent the disease from happening or manage a disease once it’s happened.
And then one you’re in a disease or you have a diagnosis, there are steps you can take from diet in some cases to prevent exacerbation of the disease from advancing further, and then there are some diet aspects that we might be able to tweak to manage symptoms.
So if we reorient ourselves to the types of study designs that are necessary to inform interventions for any of these goals, we see that for chronic-disease prevention, really, the primary means of evidence for chronic-disease prevention, at least in terms of nutrition, remains longitudinal cohort studies. There are some limited long-term clinical trials in nutrition, but their very nature makes them expensive. And so, frankly, there are relatively few long-term studies with large samples. And the ones that do exist tend to be toward the more major diseases of public-health concern, like cardiovascular disease and cancer, for example.
So that’s chronic-disease-prevention types of studies. And when we look at disease management in really either subcategory over here, we get into the territory of smaller, potentially, randomized controlled trials as well as the lesser-quality open-label trials or prospective observational trials over here.
And the importance of those long-term prospective epi-studies in nutrition is that they help us to generate hypotheses for the limited RCTs that we are able to do. And so we move up the pyramid in terms of quality of evidence as we do this. Now, prospective cohorts help us to identify which dietary interventions are justified in testing them out as an independent variable in those full clinical trials.
So here is my interpretation of where we are with dietary migraine interventions based on the types of evidence that are available for migraine. And, so, to my knowledge, we have essentially no longitudinal cohort data that would help us to understand how long-term dietary interventions influence the development of migraine. We also have limited longitudinal data on migraine after a diagnosis, especially as it relates to diet, so we really don’t have evidence for preventing exacerbation of migraine through diet either.
What I will say is we do have evidence for the prevention of comorbidities of migraine through diet and lifestyle, so we have evidence for prevention of cardiovascular disease and stroke, diabetes, and obesity. Although it’s not explicitly in individuals with migraine, it’s potentially low-hanging fruit as an opportunity for the migraine community.
Then from the perspective of managing symptoms of migraine, this is where most of the trials have investigated it as a means for migraine management and managing symptoms. Yet even here there’s limited evidence, as was mentioned in the systematic review discussed earlier. I divide these sort of arbitrarily into those diet interventions that target prevention of attacks in general as opposed to those that target specific triggers, which can be a very narrow aspect of diet. And so we again are focusing here on kind of this more general approach for broad dietary pattern to prevent attacks.
This slide spells out in a lot of verbiage what was on the previous slide, so I’ll just highlight a few things. And one is that without that longitudinal epi-data that I keep mentioning, we’re left to other means to generate hypotheses to justify randomized controlled trials, which, frankly, can make it a bigger stretch, and therefore it can be harder to find support for this type of work.
One thing also that I haven’t mentioned yet and on that previous slide, because it didn’t really fit, is that there’s an increasing amount of cross-sectional data on diet quality in individuals with migraine. The one thing I’ll emphasize here is to caution against big conclusions with cross-sectional data, as is usually the custom. When you see a dietary pattern was associated with migraine odds, for example, we don’t know the directionality of the relationship. Do migraines cause a change in diet? Does a diet cause a change in migraine? It may be possible that neither is true and there happens to be something else in the data that explains the relationship. So, again, potentially hypothesis-generating data, but cross-sectional certainly has its limits here as well.
This slide shows the controlled trials that had beneficial results for migraine and those that made the cut in the systematic review that I mentioned earlier. And so the focus on dietary patterns, as we saw in those interventions from the previous slide, aligns with modern nutrition, this focus on patterns. We don’t eat single nutrients. We eat meals, we eat snacks, we eat foods, and most of us follow some semblance of a pattern of eating similar foods over a period of time.
And so the pendulum of modern nutrition is sort of swinging away from our overconfidence in maybe the eighties and the nineties about the roles of individual nutrients, and we’re going kind of back to what you see on the screen now, in a lot of cases, where it’s all about dietary patterns, which, frankly, can be tricky but possible to quantify. There are scales, for example, where we can quantify adherence to something like the Mediterranean diet, as an example of a pattern that’s been studied.
This picture — I quite like — is from Canada’s food guide, and it shows the overall recommended pattern by food groups, where half the plate is fruits and vegetables, and then the other half is filled in with whole grains and protein foods. And in all of these cases in the food groups, they’re choosing a variety of foods within each category — so not just peas within the vegetable group, but a variety — and minimizing processed foods. So that whole description makes up the pattern there.
So some patterns like this one are based on food groups. Patterns can take on a number of different forms, food groups. They can be strictly around macronutrient composition, and they can be even very individualized as something done with IgG antibody testing to make an individual pattern or some combination thereof.
Diets investigating macronutrient patterns have been of high interest to date, and the macronutrients are the energy-contributing nutrients — your carbs, fats, and proteins. So this is where it gets interesting in that almost every pattern that’s been tested here seems to have a positive impact, which doesn’t seem particularly likely, and they’ve been very dramatically different from pattern to pattern.
And one of the reasons I’m suspicious of this — and until the more recent trials, which have better designs — a lot of the trials were not designed to adequately isolate the effect of the macronutrient modifications they were making. So to decrease carbs, by necessity in the diet, you’re changing something else in the diet, which means you’re either decreasing overall calories, which could have unintended consequences elsewhere. Or when you decrease carbs, you’re concomitantly increasing one of the other macronutrients in order to create and isocaloric diet as your control. So I eye these studies a little suspiciously; but, thankfully, that is changing.
This is a good example of a well-controlled feeding trial in this space. Now, their inclusion criteria was chronic daily headache, not exclusively migraine, so they didn’t make it into the aforementioned systematic review. I do mention it here because they worked hard to design meaningful comparison groups. Individuals were randomly assigned to 12-week interventions and were provided diets that were either high omega-3 and low omega-6 or just low omega-6 intake.
And their designs were mechanistically driven. Their intention was to promote omega-3 derivatives with antinociceptive properties in the bodies while also decreasing omega-6 derivatives with pronociceptive properties. And they succeeded in showing a difference between groups. The high-omega-3 groups showed a greater reduction in migraine symptoms and positive influences elsewhere.
So this was an elegant study to evidence a dietary intervention for reducing migraine symptoms, so, yes! But the diet is very far from the typical American diet. Omega-6 is the predominant fatty acid in our food supply. It’s in high levels in just about any processed food that contains vegetable oil, which there are many different kinds: sunflower, soybean, canola, corn. So this team went to great lengths to design their own recipes to achieve this dietary intervention with low omega-6 and provided food to their participants for 12 weeks.
So for a typical migraine patient to do this at home, to realistically achieve an omega-6 diet is quite difficult to do, and it would require a lot of home cooking after education even. To their credit, the team is conducting a follow-up study to assess more practical versions of the diet, and they’re doing it specifically to migraine patients, so I eagerly await those results.
This brings me to a second example diet, the ketogenic diet, which, of course, limits carbohydrate intake in an attempt to mimic a fasting state. Ultimately, while doing this, fat becomes the main dietary source of energy, as we see depicted over here in the left in the composition of the classic ketogenic diet. There are some different approaches that vary the composition a little bit while still maintaining a state of ketosis and making it a little easier to do.
The ketogenic diet has a variety of mechanisms by which it’s believed it could influence the migraine brain. They’re discussed elsewhere but useful for designing studies around this. And there was a flurry of interest in the 1930s, if you dig back into that literature, closer to the time when the diet was first investigated for epilepsy, and then not a whole lot in the intervening years until a recent resurgence. And I’ve listed a few citations here for reviews of this current evidence, which are relying on more recent data, though it’s still pretty heavily on prospective open-label studies and in case series.
Now, there are a few new crossover trials that are becoming available, which is exciting. And the data has been pretty consistent crossing from those prospective open-label studies and into this one new crossover trial, with data at least. And it looks like there might be a there there. There seems to be something beneficial with the ketogenic diet on migraine outcomes.
The trial at the bottom is ongoing, and they are investigating whether it’s the ketones themselves or the lack of carbohydrates that is exerting the influence, so this is also exciting. And they’re trying to tease out this difference by using dietary supplements of exogenous ketones.
At this point, the two dietary patterns that are getting traction are doing so in part because they’re building mechanistic insight into their studies. But it seems that they have more mechanistic insight to build in partially because these diets are so different from the typical American diet. And there are biomarkers we can more easily point to. It’s fairly easy to say the diet intervention is producing ketones. But these are diets that are difficult to follow and would not be attainable for a lot of people, migraine or not.
So we’re currently, in my mind, in this kind of catch-22 situation where we really need those tightly controlled dietary studies for dietary patterns that are a little more approachable for the average person before we can strongly recommend a diet to prevent migraine at any stage. The problem is it’s not as straightforward to connect those mechanistic dots for a diet that’s not as dramatically different from our typical diets. It’s harder to write a sexy proposal for a diet that adheres to the US dietary guidelines, and yet most people don’t adhere to those guidelines.
So, collectively, we need to figure this out. Studies would likely need to be larger to be powered properly because we wouldn’t expect as dramatic of a difference from a diet that follows the dietary guidelines, for example. While I’m adding to the dream list here, we might also adjust the outcomes that we target, not just managing symptoms, but also turning our study designs to be longer-term to be able to assess any relationship between diet and exacerbation of symptoms over time. And then, if these work, we need behavioral research to find ways to really support people with a chronic pain condition to make these dietary changes. So we have a lot of work to do.
On the optimistic side, in May of this year, the NIH released their first agency-wide strategic plan for nutrition research, which is centered around precision nutrition. And it opens up the possibility for answering new nutrition questions with new research designs in precision nutrition so that maybe we can kind of skirt the old way of doing things, in that there’s that traditional epi-research, and offer up a more direct approach to assessing dietary patterns in the context of migraine. And the second and fourth strategic goals I think may be particularly relevant: investigating the role of dietary patterns and reducing the burden of disease in clinical settings.
So while I can’t really recommend a diet for preventing development or exacerbation of migraine — though I will mention that there’s really no evidence against the current healthy diet according to the US dietary guidelines, and adherence to that would hopefully prevent comorbidities of migraine. We’re a little better off when it comes to evidence for preventing migraine attacks or more generally controlling symptoms.
Unfortunately, it’s difficult to recommend either of those two diets I discussed because of how difficult they are to follow. But if a patient came in and really wanted to try a ketogenic diet or a high omega-3 omega-6 ratio diet, in my mind there’s not a distinct reason not to if they’re motivated, other than the usual contraindications for them.
And lastly, I’m optimistic that we’ll find new methodologies in this upcoming era of precision nutrition that will help us advance the science in this space and hopefully enable us to expand into studying more approachable dietary patterns as interventions for managing migraine.
And so, one last time, thank you.
Voice-over: Thank you for tuning in to Spotlight on Migraine. For more information on migraine disease, please visit MigraineDisorders.org.
This podcast is sponsored in part by Amgen/Novartis.
*The contents of this podcast are intended for general informational purposes only and do not constitute professional medical advice, diagnosis, or treatment. Always seek the advice of a physician or other qualified health provider with any questions you may have regarding a medical condition. The speaker does not recommend or endorse any specific course of treatment, products, procedures, opinions, or other information that may be mentioned. Reliance on any information provided by this content is solely at your own risk.