Peter A. Lio, MD

It is written that Hippocrates, father of modern medicine, once said: “Let food be thy medicine and medicine be thy food.” There is deep wisdom in this sentiment and many have suggested that diet is the key to atopic dermatitis. The truth, unfortunately, seems a little bit more complicated.
Despite heroic leaps in scientific understanding of the disease, the fundamental cause of atopic dermatitis (AD) remains obscure. That said, we certainly have some new ideas about what causes it, and have probably learned more in the past 10 years than in the past 10,000. Beyond a better understanding of the type of inflammation and immune system issues, we have discovered a great deal about the role of the skin barrier. In particular, a protein called “filaggrin” explains a lot about why patients might develop AD in the first place, but also—possibly—why they develop other types of allergies, including food allergies. [Dennin M] This protein, or more precisely the lack of it, leads to a condition that I like to refer to as “leaky skin”. Similar to the “leaky gut” phenomenon that is increasingly discussed in some alternative medicine circles, the idea here is that lacking normal filaggrin due to a gene mutation can lead to a weakened skin barrier that allows for water to leak out and allergens, irritants, and even infections to sneak in. This is not just speculation: there is good evidence that having leaky skin really does lead to AD. [Dennin M] And abnormal filaggrin is probably just one of many factors that can lead to leaky skin. And, in a way, this is compelling in its own right: if you trace your skin as it rounds the lips, it becomes the lining of the digestive tract! It does not seem so strange that leaky gut inside correlates with leaky skin outside. While there are still many details to work out, this basic understanding has led to something of a revolution in thinking about AD and has important implications for the connection between foods and AD as well.
Root Cause?
Many families in my practice have felt strongly that food is the “root cause” of the disease. And this does not appear to be an anomaly: in one study, parents had manipulated the diet in children with AD in some 75% of cases in an attempt to control it. [Johnston GA] This thinking is not new. In fact, a paper from the British Medical Journal from way back in 1830 said this about the relationship between food and atopic dermatitis: “There is probably no subject in which more deeply rooted convictions have been held…as regards the causation and treatment of the latter…”[Mackenzie, 1830]
And there are some good reasons for this. Most importantly, there is a clear association between AD and food allergies. Real, verifiable, and often dangerous food allergies are much more common in AD, and probably occur in up to a third of patients with eczema. [Schneider L] Undoubtedly, if someone with AD eats a food that he or she is allergic to, the resulting hives and itching could cause a flare up of the eczema—that makes perfect sense. And many providers still stress this idea: in order to avoid triggers for an eczema flare, we also must consider foods.
Now, here comes the hard part: the kind of food allergies we’re talking about here are fairly easy to test for and usually cause a clear clinical pattern: almost immediate hives, swelling, and itching, vomiting, diarrhea, and sometimes progressing to more severe angioedema and even anaphylaxis. So, if someone has a true food allergy, it may well cause an eczema flare, but that’s usually the least of their worries! They would have much more powerful and urgent issues almost universally. Because of this clear relationship, what we’re focusing on here are the situations where people suspect that this is some hidden food in the diet that are causing the eczema.
This situation might be called an “eczematous food reaction.” They certainly exist and there is no doubt that some patients over the years have found a food or foods driving all or almost all of their skin disease. So, while it is certainly possible, it is also important to note that there are some people that have fallen from an airplane without a parachute and survived just fine [Wikipedia]. But… I would not recommend trying that out! In other words: just because something can happen does not make it likely.
Such an extreme example is not very helpful. How common (or uncommon) is it? That is a useful question to answer as it could be worth investigating foods even if it is only a small percentage of patients: the risks and costs are fairly low but the potential benefits are very compelling: to be cured or even better just by avoiding certain foods is an almost irresistible concept!
Testing the Connection in a Controlled Environment
To get better insight, it is worth closely examining a study by Dr. Jon Hanifin’s group to address this issue. In this study, children with severe, refractory atopic dermatitis were admitted to the hospital where they could have strict dietary control. [Rowlands 2006] The patients were in air-conditioned, low-allergen environments and placed on a “rare foods” diet—foods rarely consumed and rarely allergenic—consisting of lamb or turkey, rice, rare fruits and vegetables. After 2 days of topical therapy to decrease inflamed skin, food challenges were conducted and patients were observed carefully for 5-7 days. Notably, this is very different than a prick test which looks for immediate reactions (within minutes) or a blood test that generally is looking for the same type of reactions. By observing the patients for days, this study could capture both types of reactions. The challenges consisted of foods like milk, wheat, egg, and soy (very common allergens and difficult to avoid completely), as well as other common foods such as chicken, pork, beef, fish, and various other common vegetables. Seafood and nuts which are common causes of anaphylaxis and generally not considered essential for a balanced diet were not tested.
In the 17 children hospitalized and food challenged in the study, a total of 91 food challenges were performed. One patient developed an eczematous reaction to milk that occurred within 30 minutes. A repeat challenge was also positive, but prior skin prick testing was negative to milk, suggesting that this was the real deal: a true eczematous food reaction! But that was about it. Essentially all of the other tests to eggs, wheat, soy, various meats, fish, vegetables, potatoes, and citrus were negative (with the exception that the same patient with the milk reaction also had a reaction to chicken). And this was in a very select group of patients, most prone to food allergy, and most likely to have something driving their severe disease. [Rowlands 2006]
It is important to stress that some of these patients had a known food allergy to things like peanuts or seafood—these were not tested for the reasons we discussed above: they cause a very different type of allergy and were strictly avoided by the families so were clearly not playing a role in the eczema.
So if the foods are usually not the cause of the eczema, what is the relationship between them? That remains an extremely compelling question that is still being worked out. But, perhaps the biggest breakthrough came in 2015 with the publication of the “Learning Early about Peanut Allergy (LEAP) study. [Du Toit, 2015] In this landmark study, the notion that peanut allergy is rare in countries where peanuts are introduced early, and that telling families to hold off on introducing peanuts until later might actually be causing the allergy was tested. Importantly, it worked really well: giving infants peanut-containing foods seemed to decrease the risk of peanut allergy significantly! What was perhaps even more interesting was the connection between eczema and peanut allergy. The idea put forth was somewhat revolutionary: “…early environmental exposure (through the skin) to peanut may account for early sensitization, whereas early oral exposure may lead to immune tolerance.” [Du Toit, 2015] This concept—that leaky skin could actually lead to developing food allergies—turns the whole thing upside down. Instead of looking for foods that are causing the eczema, it sounds like we need to stop the eczema (and its associated leaky skin) in order to prevent the food allergies!
Sealing the Skin
This is still somewhat speculative, but has been borne out in one other very tantalizing way: there are now studies that suggest simply moisturizing the skin from birth can actually prevent eczema and possibly even food allergy! [Horimukai 2014 and Simpson 2010] This supports the concept that leaky skin may be the doorway for all of these conditions to develop and that by “sealing up” the leaky skin moisturizers and perhaps supplements to help strengthen the skin, we can actually prevent the cascade of events that leads to atopic dermatitis and food allergies. This is incredibly exciting as the possibility of prevention has tremendous ramifications in both quality of life and cost to both patients and the healthcare system. [Xu 2017]
But why not try it? What harm can there be to at least see if it is a food, however rare that may be? There are two main issues that I worry about. The first is more practical: while we are working to find a potential food driving the disease, sometimes people are not doing any other treatments so as not to complicate things. For an adult this is fine, I suppose, but especially in a child I hate to see someone suffering with itch, poor sleep, and sometimes even infections and pain while we’re searching for a proverbial needle in a haystack… especially one that might not even be there. The second is more insidious: I have had the unfortunate experience over the years of working with colleagues who feel strongly that diet is the cause. They feel so strongly that this is the case that they keep pushing the patients to try harder, wait longer, and avoid more foods. It can be a dangerous game. To hear something like: “You’re so close, if we just go another month or two without gluten and dairy, I think you will be cured!” is almost irresistible. But for so many it seems to be false hope, and even a source of guilt for some families. I’ve heard things like: “We really tried our best to keep the diet, but there was a birthday party about 4 months in, and our provider said that the cupcake ruined everything and we would need at least 6 more months…” And there are stories about severe malnutrition and even death from extreme diets, some that seem more like nightmares than reality. [Benny-Morrison 2016]
Which brings us back to that paper from the British Medical Journal in 1830. The paper concluded that in most instances, diet plays an “exceedingly unimportant part in the production and cure of diseases of the skin.” And, nearly 200 years later, the verdict is not too different: looking broadly at all the published studies strongly suggests that elimination diets do not seem to help most patients with AD. [Lim NR] What to do in the meantime? I think we have to do what we have done thus far until we know more, and fortunately we know that getting the skin under better control not only leads to direct improvements in quality of life, but also seems to allay some of the food allergy concerns, suggesting that it is a helpful approach while we learn more about this terrible skin disease. [Thompson 2005]

Peter A. Lio, MD is an Assistant Professor of Clinical Dermatology & Pediatrics at the Northwestern University Feinberg School of Medicine, and a Diplomate of the American Board of Dermatology. Dr. Lio received his medical degree from Harvard Medical School, completed his internship at Boston Children’s Hospital and his dermatology training at Harvard. He served as a full-time faculty at Harvard (Beth Israel & Children’s Hospital Boston) from 2005-2008 before returning home to his native Chicago to join Northwestern and Children’s Memorial Hospital. He is also a trained acupuncturist and a leader in the Chicago integrative health care community.