Non-celiac gluten sensitivity (NCGS): does it exist and how is it diagnosed

An international group of experts, whose proposals are known as the Salerno criteria, described NCGS as a syndrome with intestinal and extra-intestinal symptoms related to eating gluten-containing foods in people in whom neither celiac disease nor wheat allergy has been found. They stressed the lack of a biomarker for the condition and the need to standardise the procedure that confirms the diagnosis.

Typical intestinal symptoms include bloating, abdominal pain, diarrhoea or constipation, and commonly reported extra-intestinal ones include fatigue, 'brain fog', headaches, joint and muscle pain or skin changes. Note that this list overlaps with the symptoms of irritable bowel syndrome (IBS), with celiac disease, and with many problems unrelated to the gut. The list of symptoms alone therefore does not allow a diagnosis, and discomfort after eating bread does not prove that gluten is the culprit. We write about bloating after meals and its work-up separately, in the article on causes of bloating after eating.

This is the most important practical point of this article. The American College of Gastroenterology guidelines on celiac disease state that both serological tests and biopsy should be performed on a diet containing gluten. If someone eats gluten-free for several weeks or months, antibodies may fall and changes in the gut may reverse, and celiac disease may be missed. The consequence of such a mistake is often years of an unnecessarily relaxed diet, no monitoring for complications and no information for the family, in which celiac disease is more common.

In practice this means a simple sequence. If you suspect that gluten harms you and you still eat gluten-containing foods, see a doctor (a general physician or gastroenterologist) and ask for tests for celiac disease before eliminating anything. If you are already on a gluten-free diet, agree with your doctor how to carry out a challenge before testing, because its length and the amount of gluten depend on the procedures adopted and should not be decided on your own. Wheat allergy is a separate matter: it is diagnosed by an allergist on the basis of the history, specific IgE, skin tests and, in doubtful cases, a food challenge.

Only when celiac disease and wheat allergy have been excluded is further work-up for NCGS possible. It is worth knowing that a 2015 systematic review covering 17 studies and 1,561 patients with a diagnosis of NCGS found that among 189 patients in whom advanced diagnostic techniques were used (lymphocytic duodenosis combined with HLA-DQ2/DQ8 haplotypes), 39 people (20%) were reclassified as having celiac disease. This means that some people judged to be 'sensitive' in fact have a mild form of celiac disease.

A turning point for this topic was an Australian study that asked a simple question: do people who feel better on a gluten-free diet really react to gluten if the fermentable carbohydrates (FODMAPs), which wheat also contains, are reduced in the background?

What follows from this? In people with IBS and a suspected sensitivity to gluten, improvement after stopping wheat may result from reducing FODMAPs and not gluten. It is also worth noting that after whey protein (the placebo) symptoms worsened just as much as after gluten. Such a result cannot be explained by the action of gluten, and researchers and commentators often link it with the expectation of worsening (the nocebo effect) and with the presence of an additional protein in the diet itself. This interpretation is a hypothesis, not proof, but it shows why blinding of the challenge is so important.

Fructans are fermentable chains of fructose present in wheat, rye, onion, garlic and many other plants. This is why a low-fermentable (low FODMAP) diet is for people with IBS sometimes more effective, or at least easier to justify, than a gluten-free diet alone. We write about the principles of this diet in the low FODMAP diet entry of the knowledge base. Note that in the Oslo study the highest score came after placebo in 22 people and after gluten in 13, so most participants did not react to gluten, and some reacted to placebo itself.

For balance we must cite studies that gave a different result. An Italian double-blind, placebo-controlled cross-over trial included 61 adults who believed that gluten caused intestinal and extra-intestinal symptoms in them, without celiac disease or wheat allergy.

How can these results be reconciled? The differences may come from the selection of participants (the Australian and Norwegian studies were dominated by people with IBS who had first undergone FODMAP restriction), from the dose and form of gluten (capsules versus muesli), from the number of symptoms analysed and from the limited size of the studies. A reasonable conclusion is that NCGS is probably a heterogeneous group: in some patients symptoms are provoked by fructans and other components, in some by the expectation of a reaction, and in some really by gluten or other wheat proteins. The literature also considers a role for other wheat proteins, for example alpha-amylase/trypsin inhibitors (ATIs), but this remains a mechanistic hypothesis, without clinical confirmation in studies that could change practice.

The Salerno experts proposed that the full diagnostic procedure should include assessing the clinical response to a gluten-free diet and measuring the effect of a gluten challenge after a period on that diet. Symptoms are assessed with an instrument containing a modified version of the GSRS scale: the patient chooses one to three main symptoms and rates them numerically on a scale of 1 to 10.

The reality is that the full blinded challenge protocol is carried out mainly in research centres and specialised clinics, and in typical outpatient care the diagnosis rests on excluding celiac disease and allergy and on structured observation of the response to diet. The conclusion 'it is definitely gluten' based only on improvement after stopping it is therefore always somewhat uncertain.

Since the evidence indicates that in many people FODMAPs, not gluten, are the problem, for some patients a reasonable strategy is not a 'full gluten-free diet' but a low-fermentable diet conducted in phases: restriction, controlled reintroduction of food groups and establishing individual tolerance. Such a strategy is usually run by a dietitian, because without supervision it is easy to restrict for too long and unnecessarily. It is also worth remembering the role of the gut microbiome and prebiotic fibre, which we describe in separate entries.

If, after celiac disease and allergy have been excluded, a gluten-free diet really helps you and the result after a blinded challenge is positive, there is no reason to abandon it. Many people with NCGS nevertheless tolerate some amounts of gluten, and the intensity of symptoms varies, so it is worth assessing tolerance from time to time, in agreement with a doctor or dietitian. Remember that no risk of the complications typical of celiac disease has been described in NCGS, but the absence of such a risk does not excuse you from clarifying the cause of symptoms, especially when they are chronic.

What we do not know: there are no reliable data on the frequency of NCGS. The systematic review gives a range from 0.5% to 13%, which reflects different definitions and methods rather than real differences between populations. The review also indicates that the benefit of a gluten-free diet in people with NCGS is still disputed, although there may be a subgroup of patients with diarrhoea-type IBS and the HLA-DQ2 haplotype who might benefit. There is also no approved biomarker or simple test, and tests sold as a 'gluten sensitivity test' from blood or saliva have no proven diagnostic value.

Gluten sensitivity is a topic in which it is easy to fall into two extremes: the belief that gluten harms everyone, and the conviction that it is all suggestion. The evidence says something in between. The symptoms are real, but their cause often lies somewhere other than in gluten, and the only reliable route to an answer leads through excluding celiac disease, through a sensibly conducted diet and, if possible, through a blinded challenge.