About a third of Americans report that they are trying to reduce or avoid gluten in their diet. If Jimmy Kimmel’s funny stunt is any indication, most probably don’t know what gluten even is. The gluten-free diet has officially become a fad, and “gluten” has been tagged as something vaguely bad that should be avoided.

About 1% of people have a disease called Celiac, which is an autoimmune reaction to gluten. This is a serious disease that can make people very ill if they consume even the smallest amount of gluten. A diagnosis of Celiac can be confirmed with an antibody test (anti-gliadin antibodies), or, if necessary, a stomach biopsy.

Gluten is a composite protein composed of two parts, gliadin and glutenin. It is found in wheat, rye, barley, spelt, and related grains. It is a springy protein that gives bread its elasticity. Celiac disease is an immune reaction to the gliadin part of the protein.

Celiac is fairly well understood and is non-controversial. What is controversial is a disorder known as non-celiac gluten sensitivity (NCGS) – believed to be an intolerance to gluten that causes gastrointestinal symptoms. NCGS is controversial, and in fact may not truly exist at all.

NCGS should be considered a hypothesis, not a confirmed entity, but this has not stopped self-diagnosis and treatment from becoming popular.

It is sometimes tricky to confirm whether or not a new possible diagnosis truly exists, or if it is just a misdiagnosis of other diseases and disorders. Diseases are usually first recognized by their clinical syndrome, and then later investigation uncovers the cause or pathophysiology of the disease. Often at this stage, when we discover what is happening biologically, diseases are reclassified, and diagnoses are sometimes combined, and other times split apart.

There are some diagnoses, however, that live on the fringe, never gaining scientific support. Throughout history, it seems, there have always been faddish diagnoses used as popular labels for common symptoms. At the turn of the 19th century “neurasthenia” was a common label for vague or common symptoms. In the mid 20th century syphilis (although a real disease) was often used as a convenient diagnosis for any unexplained symptoms.

More recently we have chronic Lyme, candida hypersensitivity, multiple chemical sensitivity, electromagnetic sensitivity, and a host of other vague syndromes.

Electromagnetic sensitivity is similar to NCGS in that both are believed to be a sensitivity with symptoms resulting from a specific exposure. In both cases, therefore, we can address the core question (does the sensitivity exist) by studying blinded exposures. In the case of electromagnetic sensitivity, when properly blinded those who believe they have this condition cannot detect exposure.

What about NCGS? It has not been established that NCGS exists, or that people who believe they have this condition actually are responding physiologically to gluten. There are two possibilities that need to be carefully considered. The first is that perceived gluten sensitivity is an observational artifact, a type of nocebo effect. GI symptoms are notoriously sensitive to mood and expectation. There are also generic biases such as confirmation bias that can lead to the perception of false associations.

It is still not clear, in other words, that there is an actual association between consuming gluten and GI symptoms. Individuals may firmly believe that they have such an association, but we know from countless historical examples and experiments that such firm beliefs can form in the absence of a true association.

The second possibility that needs to be seriously considered is that in some people who are self-diagnosed with NCGS, they are reacting to something else that is common in gluten-containing foods. If this is the case, then gluten is an innocent bystander. This would be very important to discover, for obvious practical reasons.

A recent study suggests that this might be the case. Biesiekierski et. al. did a well controlled series of studies in which they challenged subjects with possible NCGS with carefully controlled diets with various amounts of gluten. They found no association between gluten consumption and reported symptoms, arguing very strongly against NCGS as a real entity.

Their study did, however, suggest another possible culprit – FODMAPs (fermentable, oligo-, di-, monosaccharides, and polyols). These are also common in breads and other foods containing gluten. In the study subjects, GI symptoms improved when FODMAPs and gluten were removed, but then reintroducing gluten had no association with return of symptoms. The authors conclude:

In a placebo-controlled, cross-over rechallenge study, we found no evidence of specific or dose-dependent effects of gluten in patients with NCGS placed diets low in FODMAPs.

They were not, however, testing whether or not FODMAPs were a cause of GI symptoms, and so cannot conclude if this is the true cause. A follow up study would need to be done to verify that (perhaps we’ll see a FODMAP-free fad before this science can be done). If true it would explain why some people do have reduction in GI symptoms when they avoid gluten, because they are also avoiding FODMAPs.

Conclusion

The best evidence we currently have suggests that NCGS is probably not a real entity. Blinded challenges do not show any correlation, and there is currently no evidence for a specific mechanism. Those who are self-diagnosed with NCGS probably fall into one of three categories:

1- Borderline true Celiac disease (a small minority that can be sorted out with diagnostic tests)

2- GI symptoms due to non-dietary reasons with a false association with gluten due to confirmation bias and nocebo effects

3 – GI symptoms due to some other food exposure. FODMAPs are one possibility, but more research needs to be done.

The real risk of the gluten-free fad is that it distracts from what is really going on. Popular diagnoses (whether real or not) do tend to attract self-diagnosis, and become an impediment to a more proper diagnosis. There is a tendency to prematurely settle on the popular diagnosis, and then fail to consider all possibilities.

In the case of NCGS, there may be something else in food to which some people are sensitive. Or, diet may not be the answer at all.