The Yale study has been replicated
One result can be chance. Two independent results in the same direction, both double-blind and both using a sham diet as the comparison, are a body of evidence. That is the whole point of this note.
What the new trial did
Design. Randomised, double-blind, sham-controlled. Sheba Medical Center, Ramat Gan. Enrolment 5 August 2020 to 11 April 2022. Population. 68 patients with IBS-D and IBS-M (44 and 24). Primary endpoint. A 50-point reduction on the IBS Symptom Severity Scale.
Intervention. An Alcat-guided diet compared with a sham-balanced diet over eight weeks. The control group did not receive no dietary change, but an equally demanding diet that was not guided by the test result. Patients could not tell which one they were following.
That is the decisive design detail. An elimination diet works partly through attention, structure and expectation. A sham diet carries all of that except the thing being tested: whether it matters which foods are removed.
| Outcome at week 8 | Alcat | Control | P |
|---|---|---|---|
| Primary endpoint (−50 pt IBS-SSS) | 85.7% (30/35) | 54.5% (18/33) | 0.005 |
| IBS-GIS improvement | 74.3% | 42.4% | 0.008 |
| Positive response (yes/no) | 85.7% | 57.6% | 0.010 |
No serious adverse events were reported.
Note the control arm: 54.5% met the primary endpoint on a sham diet. That is the size of the effect of changing your diet at all under structured conditions. It is also why a trial without a sham arm could not have demonstrated anything.
The trial's weakness, as its authors state it
The groups were not comparable at baseline. The Alcat group was sicker: median IBS-SSS 390 (305–435) versus 330 (240–390), P = 0.013.
This matters. A 50-point reduction is easier to achieve from 390 than from 330, and a sicker group has more room to improve.
The authors address it with a balanced sub-cohort — the 49 patients whose IBS-SSS fell between 250 and 450, where the medians did not differ (380 versus 355, P = 0.487). The difference held: 90% versus 59%, P = 0.022.
That is the right way to meet the objection. It is also a subset defined after data collection, and should be read as support for the main finding rather than as an independent result.
What the two trials together support
Yale 2017 (Ali et al., BMJ Open Gastroenterology): 58 patients, double-blind, controlled, with a reduction in plasma neutrophil elastase as mechanistic support for innate immune involvement. Sheba 2026 (Engel et al., Gastroenterology Report): 68 patients, double-blind, sham-controlled.
Two centres. Two countries. Two independent research groups. The same direction.
The clinical question now tested is narrow and clear: do IBS patients improve more on a diet guided by this test than on an equally demanding diet that is not? The answer has been given twice, and it is yes.
What it still does not support
- Two trials are two trials. 58 and 68 patients. Neither is large.
- IBS only. Both concern irritable bowel syndrome. Nothing in this evidence speaks to skin conditions, fatigue, joint pain or autoimmune disease — however often the test is used for them.
- Reproducibility has no independent published answer. The original and load-bearing objection to cytotoxic testing was that the same sample could yield different results. The current method is read by automated impedance flow cytometry rather than by human judgement, which is a reasonable reply to that objection. It is not the same thing as a published reproducibility study from an independent laboratory. There is none, and that is the study that would settle the question.
- It is not an allergy test. ALCAT does not measure IgE. The criticism on that point is correct and stands unanswered. The test should not be compared with, or substituted for, allergy investigation.
- Summaries have not caught up. The AAAAI and ACAAI positions rest on evidence up to around 2010. They were reasonable then. They have not been revised against trials published seventeen years later — which does not make them wrong, only out of date.
What this means for a patient here
Nothing changes in how we work. The test has always been used as one input among several, interpreted against symptoms, dietary history and clinical reassessment, and never as an answer in itself. See the ALCAT test and IBS and gut health.
What changes is how confidently we can say why. In IBS there are now two double-blind trials supporting that the diet this test produces makes a difference beyond the dietary change that happens anyway. For other conditions there are not, and there the test should continue to be described as what it is: an input to a hypothesis that is tested clinically and reassessed. That is the same evidence standard set out in Longevity and healthspan.
References
- Engel T, et al. Immune-based personalized elimination diet for the treatment of irritable bowel syndrome: a double-blind randomized sham-controlled study. Gastroenterology Report (Oxf). 2026;14. doi:10.1093/gastro/goag058. PMID 42294460.
- Ali A, Weiss TR, McKee D, et al. Efficacy of individualised diets in patients with irritable bowel syndrome: a randomised controlled trial. BMJ Open Gastroenterology. 2017;4:e000164.
This note sets out published evidence and does not constitute individual medical advice. See our editorial and medical review policy.