Expectation, the weak point of an open test

01Dummy foods

In blinded food challenges in children, about one placebo day in eight produced a “reaction” (Vlieg-Boerstra et al., 2007). In a blinded trial of muesli bars hiding gluten, fructans or neither, 22 of 59 people scored their worst week on the bars with neither (Skodje et al., 2018).

02What people expect shows up in the scores

Expectation can change what people report even when the food is the same. In one trial, adults who said gluten upset their gut were told at random whether to expect it, and given bread with or without it, also at random (de Graaf et al., 2024). Among those who ate gluten, the ones who expected it scored more than twice as high as the ones who didn’t. The authors saw a nocebo effect at work, symptoms that come from expecting harm, though they couldn’t rule out some effect of gluten (de Graaf et al., 2024).

Other blinded studies point the same way. Across 10 gluten trials, only 16% of people with suspected gluten sensitivity had symptoms specific to gluten (Molina-Infante and Carroccio, 2017). Adults who reported severe trouble with even less than a glass of milk had minimal symptoms on either ordinary or lactose-free milk, given blind (Suarez et al., 1995).

03Open tests find more than blinded ones

When everyone knows what is being given, more “reactions” can turn up. When children with a positive open food challenge were retested blind, 8 of 11 one-day results and 20 of 35 one-week results held up (Venter et al., 2007). The authors concluded that delayed, mainly subjective symptoms may need a blinded challenge.

Self-reports run high as well: far more people report a food allergy than tests confirm (Rona et al., 2007; Nwaru et al., 2014).

04What helps, and what doesn’t

A test at home is usually open. Only the last of these habits deals with expectation itself; the others guard against chance and over-reading:

• Repeat it. One run can’t tell a real change from an ordinary bad day, and the authors of one study of blinded food challenges suggested repeating the test in selected cases (Vlieg-Boerstra et al., 2007). Repeating an open test won’t separate a real change from an expected one, though.
• Decide in advance how big a change must be, so a small shift can’t be talked up afterwards. That won’t remove expectation, but it stops the goalposts moving.
• Accept “inconclusive”. A small, mixed or one-off result is unclear, and saying so is an honest answer.
• Have a helper hide what is being tested, where possible. Someone else prepares the food so that test days and comparison days look and taste the same, as research does: one trial hid gluten in muesli bars (Skodje et al., 2018), and another sweetened milk so both kinds tasted alike (Suarez et al., 1995).

Before any test of a food, read Safety, and when to involve a clinician.

05What the studies found

• In a 2007 study, a double-blind retest of 41 children with a positive open food challenge confirmed 8 of 11 one-day open results and 20 of 35 one-week ones (Venter et al., 2007).
• In a 2007 study of 132 double-blind food challenges in 105 children, 17 placebo days (12.9%) produced a “reaction”, some quick and some late, and the authors suggested repeating the test in selected cases (Vlieg-Boerstra et al., 2007).
• In a 2007 review of 51 studies, self-reported food allergy was very common compared with objective tests, ranging from 3% to 35% for any food (Rona et al., 2007).
• In a 2014 review of European studies, about 6.0% of people reported having had a cow’s milk allergy, against 0.6% where a food challenge confirmed it (Nwaru et al., 2014).
• In a 2017 review of 10 double-blind gluten-challenge trials with 1,312 adults, only 38 of 231 people with suspected gluten sensitivity (16%) had symptoms specific to gluten (Molina-Infante and Carroccio, 2017).
• In a 2018 double-blind trial, 59 people with self-reported gluten sensitivity ate muesli bars hiding gluten, fructans or neither for a week each (Skodje et al., 2018). By person, 22 of the 59 scored highest on placebo.
• In a 2024 trial, 83 adults who reported gut symptoms within 8 hours of eating gluten were told they would get gluten or not, and separately, at random, did or didn’t get it (de Graaf et al., 2024). Scores were highest when gluten was expected and eaten; eating it unawares scored no higher than having none. Funders included a Dutch government program and bakery and wheat companies.
• In a 1995 double-blind trial, 30 adults who described severe trouble with milk, with symptoms after less than 240 mL, drank 240 mL a day of ordinary or lactose-free milk, a week of each (Suarez et al., 1995). Symptoms were minimal both weeks; the 21 who absorbed lactose poorly passed wind about 2.5 more times a day on ordinary milk.

06Sources on this page

Each source is listed once, with links to its record. The reference list says what was read for each, and when it was checked.

1. de Graaf et al., 2024de Graaf MCG, Lawton CL, Croden F, Smolinska A, Winkens B, Hesselink MAM, et al. 2024. The effect of expectancy versus actual gluten intake on gastrointestinal and extra-intestinal symptoms in non-coeliac gluten sensitivity: a randomised, double-blind, placebo-controlled, international, multicentre study. The Lancet Gastroenterology & Hepatology 9(2):110–123.DOI 10.1016/S2468-1253(23)00317-5 · PubMed 38040019
2. Molina-Infante and Carroccio, 2017Molina-Infante J, Carroccio A. 2017. Suspected Nonceliac Gluten Sensitivity Confirmed in Few Patients After Gluten Challenge in Double-Blind, Placebo-Controlled Trials. Clinical Gastroenterology and Hepatology 15(3):339–348.DOI 10.1016/j.cgh.2016.08.007 · PubMed 27523634
3. Nwaru et al., 2014Nwaru BI, Hickstein L, Panesar SS, Roberts G, Muraro A, Sheikh A; EAACI Food Allergy and Anaphylaxis Guidelines Group. 2014. Prevalence of common food allergies in Europe: a systematic review and meta-analysis. Allergy 69(8):992–1007.DOI 10.1111/all.12423 · PubMed 24816523
4. Rona et al., 2007Rona RJ, Keil T, Summers C, Gislason D, Zuidmeer L, Sodergren E, et al. 2007. The prevalence of food allergy: a meta-analysis. Journal of Allergy and Clinical Immunology 120(3):638–646.DOI 10.1016/j.jaci.2007.05.026 · PubMed 17628647
5. Skodje et al., 2018Skodje GI, Sarna VK, Minelle IH, Rolfsen KL, Muir JG, Gibson PR, et al. 2018. Fructan, Rather Than Gluten, Induces Symptoms in Patients With Self-Reported Non-Celiac Gluten Sensitivity. Gastroenterology 154(3):529–539.e2.DOI 10.1053/j.gastro.2017.10.040 · PubMed 29102613
6. Suarez et al., 1995Suarez FL, Savaiano DA, Levitt MD. 1995. A comparison of symptoms after the consumption of milk or lactose-hydrolyzed milk by people with self-reported severe lactose intolerance. New England Journal of Medicine 333(1):1–4.DOI 10.1056/NEJM199507063330101 · PubMed 7776987
7. Venter et al., 2007Venter C, Pereira B, Voigt K, Grundy J, Clayton CB, Gant C, et al. 2007. Comparison of open and double-blind placebo-controlled food challenges in diagnosis of food hypersensitivity amongst children. Journal of Human Nutrition and Dietetics 20(6):565–579.DOI 10.1111/j.1365-277X.2007.00828.x · PubMed 18001378
8. Vlieg-Boerstra et al., 2007Vlieg-Boerstra BJ, van der Heide S, Bijleveld CM, Kukler J, Duiverman EJ, Dubois AE. 2007. Placebo reactions in double-blind, placebo-controlled food challenges in children. Allergy 62(8):905–912.DOI 10.1111/j.1398-9995.2007.01430.x · PubMed 17620068