Gluten intolerance symptoms, the three conditions they could mean, and the sequencing error that wastes the diagnosis

1. 01

   One phrase, three conditions, three mechanisms

   Coeliac disease is a permanent immune-mediated response to gluten in wheat, barley and rye, characterised by small bowel injury and specific antibodies (Rubio-Tapia et al. 2023). Wheat allergy is immunoglobulin E mediated. Non-coeliac gluten sensitivity is intestinal and extra-intestinal symptoms related to gluten-containing food in people affected by neither of the other two, and it has no biomarker (Catassi et al. 2015). They share a symptom list and share almost nothing else.

2. 02

   Biopsy-confirmed coeliac prevalence is about 0.7 percent

   A systematic review and meta-analysis put pooled global coeliac seroprevalence at 1.4 percent, 95 percent confidence interval 1.1 to 1.7, in 275,818 individuals, and pooled biopsy-confirmed prevalence at 0.7 percent, 95 percent confidence interval 0.5 to 0.9, in 138,792 (Singh et al. 2018). That is roughly one person in 140. Common enough that persistent unexplained symptoms warrant a test, uncommon enough that most people who feel better without bread do not have it.

3. 03

   Prevalence varies by region, sex and age

   Regional biopsy-confirmed prevalence ran 0.4 percent in South America, 0.5 percent in Africa and North America, 0.6 percent in Asia and 0.8 percent in Europe and Oceania. It was higher in women than men, 0.6 against 0.4 percent, and higher in children than adults, 0.9 against 0.5 percent, both at P less than 0.001 (Singh et al. 2018). A prior probability that varies twofold across those axes is worth carrying into a conversation with a clinician.

4. 04

   Removing gluten before testing destroys the test

   Coeliac serology detects an immune response that is currently active, and biopsy detects injury that is currently present. Remove gluten and antibodies fall while villi recover, so both tests trend toward normal in someone who genuinely has the disease. Doctors do not recommend starting a gluten-free diet before diagnostic testing because a gluten-free diet can affect test results (NIDDK). This is the single most consequential sequencing decision in the topic and it costs nothing to get right.

5. 05

   The diagnostic pathway is serology first, biopsy in most cases

   Detection of coeliac-specific antibodies such as tissue transglutaminase in the serum is very helpful for initial screening of patients with suspicion of coeliac disease, and intestinal biopsy is required in most patients to confirm the diagnosis, with a non-biopsy strategy suggested only for selected children (Rubio-Tapia et al. 2023). Neither step is something a person can run on themselves, which is precisely why the order of operations has to be decided before the diet is changed.

6. 06

   Coeliac disease frequently does not present as a stomach problem

   The guideline describes a wide spectrum of clinical manifestations resembling a multisystemic disorder rather than an isolated intestinal disease (Rubio-Tapia et al. 2023). The presentations that get missed are the ones that do not look gastrointestinal: iron deficiency anaemia that will not correct with supplementation, unexplained transaminase elevation, reduced bone density, and dermatitis herpetiformis, which is diagnosed on skin biopsy (NIDDK). A symptom list confined to bloating will therefore miss a substantial share of cases.

7. 07

   When FODMAPs were removed first, gluten stopped separating from whey

   In a double-blind crossover of 37 subjects with self-reported gluten sensitivity and irritable bowel syndrome, coeliac excluded, everyone first ran a two-week reduced FODMAP diet, then received high gluten at 16 g per day, low gluten at 2 g plus 14 g whey, or 16 g whey control. Symptoms consistently improved on reduced FODMAP intake and then worsened to a similar degree whether the diet included gluten or whey protein (Biesiekierski et al. 2013).

8. 08

   Gluten-specific effects appeared in 8 percent of participants

   In that same trial, gluten-specific effects were observed in only 8 percent of participants, there were no diet-specific changes in any biomarker of intestinal inflammation, injury or immune activation, and in a three-day rechallenge in 22 participants symptoms increased by similar levels among groups with gluten-specific gastrointestinal effects not reproduced. An order effect was also observed (Biesiekierski et al. 2013). Eight percent is not zero, and it is far below what the study population expected.

9. 09

   Fructan beat gluten in a head-to-head blinded challenge

   Fifty-nine people on self-instituted gluten-free diets, coeliac excluded, completed seven-day blinded challenges with gluten at 5.7 g, fructans at 2.1 g and placebo, concealed in muesli bars, with washouts between. Overall Gastrointestinal Symptom Rating Scale scores were 33.1 plus or minus 13.3 for gluten, 38.6 plus or minus 12.3 for fructan and 34.3 plus or minus 13.9 for placebo, P equals 0.04, with fructan significantly worse than gluten, P equals 0.049 (Skodje et al. 2018).

10. 10

    Bloating specifically tracked the fructan, not the gluten

    Mean bloating scores in the same crossover were 9.3 plus or minus 3.5 on gluten, 11.6 plus or minus 3.5 on fructan and 10.1 plus or minus 3.7 on placebo, P equals 0.004, with fructan significantly higher than gluten at P equals 0.003 (Skodje et al. 2018). Bloating is the symptom most often used as evidence of gluten intolerance, and in the one trial that separated the two components of wheat it was the symptom that most clearly belonged to the carbohydrate.

11. 11

    More people had their worst week on placebo than on gluten

    In the same 59 participants, 13 recorded their highest overall symptom score after the gluten challenge, 24 after the fructan challenge and 22 after the placebo challenge, and there was no difference in scores between gluten and placebo (Skodje et al. 2018). Twenty-two against 13 is a direct measurement of expectation effects in a group that had already concluded gluten was their problem, and it is why unblinded self-experimentation cannot settle this question.

12. 12

    The field has a published protocol, and it is blinded for a reason

    The Salerno Experts' Criteria assess clinical response to a gluten-free diet and then measure the effect of a gluten challenge. The patient identifies one to three main symptoms scored 1 to 10 on a numerical rating scale using a modified Gastrointestinal Symptom Rating Scale, then completes a double-blind placebo-controlled challenge at 8 g of gluten per day, one week on, one week of washout on a strict gluten-free diet, then crossover to a second week, with cooked, homogeneously distributed gluten in the vehicle (Catassi et al. 2015).

13. 13

    A 30 percent symptom variation is the threshold, and it is there to filter noise

    The same criteria require at least a 30 percent variation in one to three main symptoms between the gluten and placebo challenge to discriminate a positive from a negative result (Catassi et al. 2015). That threshold is the formal recognition that a real but small difference cannot be distinguished from day-to-day variation. Anyone running an informal trial should adopt both features of the protocol, the blinding and the numeric bar, or accept that the result will not mean much.

14. 14

    An unnecessary gluten-free diet has a nutritional cost, not just a price

    In the United States, enriched grain products are the fortification vehicle for folic acid, iron, thiamin, riboflavin and niacin, so removing them removes a meaningful share of several micronutrient intakes along with dietary fiber, on top of the shelf price premium on substitutes. Coeliac disease requires strict lifelong adherence and medical follow-up (Rubio-Tapia et al. 2023), which is a justified cost. Paying it for a fructan problem that a different adjustment would have solved is not.

What people ask us next.

Q01

What are the symptoms of gluten intolerance?

The reported symptom set is broad and non-specific: bloating, abdominal pain, altered bowel habit, fatigue and headache, which is why the symptoms alone cannot separate the three underlying conditions. Coeliac disease in particular presents as a multisystemic disorder rather than an isolated intestinal disease, and the missed cases are often iron deficiency anaemia that will not correct, unexplained transaminase elevation, reduced bone density, or the dermatitis herpetiformis rash (Rubio-Tapia et al. 2023; NIDDK). Symptoms indicate that testing is warranted; they do not identify which condition is present.

Q02

Should I stop eating gluten to see if I feel better?

Not before testing, because the test depends on the exposure. Coeliac serology detects an active immune response and biopsy detects current injury, so both trend toward normal once gluten is removed. Doctors do not recommend starting a gluten-free diet before diagnostic testing because a gluten-free diet can affect test results (NIDDK). The order that preserves every option is to keep eating gluten, see a clinician, complete testing, and only then run a diet trial. Reversing that order can cost a diagnosis of a lifelong condition.

Q03

How common is coeliac disease?

Pooled global seroprevalence is 1.4 percent, 95 percent confidence interval 1.1 to 1.7, across 275,818 individuals, and pooled biopsy-confirmed prevalence is 0.7 percent, 95 percent confidence interval 0.5 to 0.9, across 138,792, so roughly one person in 140 (Singh et al. 2018). Prevalence ran 0.4 percent in South America, 0.5 percent in Africa and North America, 0.6 percent in Asia and 0.8 percent in Europe and Oceania, and was higher in women than men and in children than adults, both at P less than 0.001.

Q04

Is non-coeliac gluten sensitivity real?

It is a recognised syndrome without a biomarker, and the blinded data suggests the gluten attribution is usually wrong even when the symptoms are not. In 37 people with self-reported sensitivity, gluten-specific effects appeared in only 8 percent once fermentable carbohydrates had been reduced first, with no diet-specific biomarker change (Biesiekierski et al. 2013). In 59 others, blinded challenges produced no difference between gluten and placebo while fructan produced significantly worse scores (Skodje et al. 2018). Some people do react; fewer than believe they do.

Q05

Could my symptoms be fructans rather than gluten?

Frequently, and this is the best-supported alternative explanation. Wheat carries both gluten and fructans, a fermentable oligosaccharide. In a double-blind crossover in 59 people on self-instituted gluten-free diets, overall symptom scores were 33.1 on gluten, 38.6 on fructan and 34.3 on placebo, P equals 0.04, with fructan significantly worse than gluten, P equals 0.049, and bloating scores of 9.3, 11.6 and 10.1, P equals 0.004 (Skodje et al. 2018). Bloating, the symptom most cited as evidence for gluten, tracked the carbohydrate.

Q06

How can I test my own reaction to gluten properly?

Only after a clinician has excluded coeliac disease and wheat allergy, and then using a blinded protocol. The Salerno Experts' Criteria specify identifying one to three main symptoms scored 1 to 10, then a double-blind placebo-controlled challenge at 8 g of gluten per day for one week, a one-week washout on a strict gluten-free diet, and crossover to a second one-week challenge, with cooked, homogeneously distributed gluten in the vehicle, requiring at least a 30 percent symptom variation to count as positive (Catassi et al. 2015).

Q07

Why does blinding matter so much here?

Because the placebo arm outperformed expectations in the direction nobody wants. Among 59 participants who had already concluded gluten was their problem, 13 recorded their worst overall symptom score after gluten, 24 after fructan and 22 after placebo, with no difference between gluten and placebo (Skodje et al. 2018). More people had their worst week on an inert challenge than on the substance they were avoiding. An unblinded personal trial measures expectation as much as physiology, which is why it cannot settle the question.

Q08

What does an unnecessary gluten-free diet cost?

More than the shelf premium on substitutes. In the United States, enriched grain products are the fortification vehicle for folic acid, iron, thiamin, riboflavin and niacin, so removing them removes a share of several micronutrient intakes along with dietary fiber. For coeliac disease, strict lifelong adherence plus medical follow-up is the treatment and the cost is justified (Rubio-Tapia et al. 2023). For a fructan problem misread as a gluten problem, the same cost is being paid permanently for a result a different adjustment would have produced.