Cutting gluten and dairy is one of the most common first moves, and trial data shows it frequently fails for a specific reason: in blinded testing, fewer than 1 in 5 people who believe they react to gluten actually do, and self-reported dairy intolerance is wrong in both directions almost half the time. The reaction is often real. The blamed ingredient often isn’t.
You gave up bread. You gave up your coffee creamer, then milk entirely, then anything with a wheat flour in the ingredient list, checking labels standing in the aisle. Weeks of it, and the bloating that sent you looking in the first place didn’t move, or moved a little and then came back. That’s not a sign you didn’t try hard enough. It’s a sign the thing you cut may not have been the actual problem.
The gluten question almost never gets tested the way people assume
In 2013, a research team ran a tightly controlled trial on people who were already convinced gluten caused their symptoms. Before testing gluten specifically, they first put everyone on a diet low in FODMAPs, the fermentable carbohydrates found in wheat along with plenty of other foods. Symptoms improved for nearly everyone at that point, before gluten had been reintroduced at all. Then, under blinded conditions, adding real gluten back produced no measurable difference from adding a placebo.
That single trial didn’t settle the question on its own, so a later analysis pooled the evidence properly: 10 double-blind, placebo-controlled gluten challenge trials, 1,312 adults who believed they reacted to gluten. Fewer than 20% showed a gluten-specific symptom response when neither they nor the researchers knew which capsule was which. About 40% reported symptoms just as bad, or worse, on the placebo.
That 40% reacting to placebo isn’t a character flaw, and it isn’t “all in your head” in the dismissive way that phrase gets used. It’s called a nocebo effect, expecting a food to hurt you measurably changes how your gut interprets normal signals, and it’s a documented, physiological pattern, not an accusation of imagining things. Researchers see the same pattern in reverse with the placebo response generally: expectation changes measurable gut function in both directions, which is inconvenient for anyone trying to run a clean test on themselves at home without a blind.
The practical result is the same either way: gluten specifically was very likely never the actual mechanism for most people who believe it is. That doesn’t mean the wheat-free years were pointless. Cutting wheat also cuts fructans, a FODMAP, which is very plausibly why symptoms improved in the first place, for reasons that have nothing to do with the gluten protein itself.
Dairy self-diagnosis is unreliable in both directions
Lactose gets blamed constantly, and the research on how well people self-diagnose it is not reassuring in either direction. A 2025 systematic review pooling six studies and 845 people with IBS compared self-reported lactose intolerance against objective hydrogen breath testing. Fewer than 4 in 10 people matched on both self-report and the test. Sixteen percent believed they were intolerant and tested tolerant. Twenty-seven percent believed they tolerated dairy fine and tested positive for lactose malabsorption anyway.
There’s also a separate finding worth naming: some people react to whole cow’s milk specifically while tolerating lactose itself just fine in isolation, which points to a protein or fat component rather than the sugar most dairy-free marketing focuses on entirely. Cutting “dairy” as one category can miss that distinction the same way cutting “gluten” misses the FODMAP distinction.
| What you cut | What the evidence suggests is often the real trigger |
|---|---|
| Gluten (as a category) | Fructans, a FODMAP found in wheat, not gluten protein itself, for most self-reported reactors |
| All dairy | Sometimes lactose, sometimes a milk protein or fat component, not always predictable from symptoms alone |
| Both, all at once | Impossible to know which one, if either, was doing anything |
What this looks like in practice
The instinct after this information is usually to cut something else next. That’s the wrong direction. The more useful move is testing what you already cut.
If you’re strictly gluten-free with no celiac diagnosis, a supervised reintroduction is worth doing. Celiac disease itself needs to be ruled out first, ideally before going gluten-free, since testing is unreliable once gluten is already out of your diet.
Distinguish “wheat” from “gluten” when you reintroduce. If FODMAPs, not gluten, were the actual driver, a low-gluten but still wheat-based food might still cause symptoms, while a genuinely gluten-containing but low-FODMAP food might not.
For dairy, a lactose-specific product is a cleaner test than cutting dairy entirely. Lactose-free milk that still contains milk proteins can isolate whether lactose specifically, versus dairy generally, is the actual issue.
Don’t test gluten and dairy back in during the same week. The same one-variable-at-a-time logic that applies to FODMAP reintroduction applies here, for the same reason: change two things at once and a result tells you nothing specific.
Common questions
Does this mean gluten sensitivity isn’t real?
No. It means that in blinded trials, most people who believe gluten specifically causes their symptoms don’t react to gluten when they can’t tell what they’re eating. The symptoms are real. The blamed trigger, for most people, isn’t.
Should I get tested for celiac disease before assuming it’s something else?
Yes, and ideally before cutting gluten entirely, since celiac testing becomes unreliable once gluten has already been removed from your diet for a meaningful stretch of time.
If I feel better gluten-free, does the research even apply to me?
It’s worth testing rather than assuming. Feeling better while avoiding wheat is consistent with a FODMAP-driven improvement, a gluten-specific one, or a nocebo-driven one, and only a structured reintroduction distinguishes between them.
Why did cutting dairy help a little but not completely?
That’s consistent with lactose being a partial contributor rather than the whole story, or with a milk protein component playing a role instead of, or alongside, lactose itself.
Is a food sensitivity test worth taking to sort this out?
Most commercial food sensitivity panels, particularly IgG-based ones, lack validated evidence connecting their results to actual symptom improvement. A structured elimination and reintroduction, imperfect and slower as it is, has more real evidence behind it than most panel tests currently sold, which is a frustrating thing to hear when a blood draw feels like the faster answer.
What if reintroducing gluten and dairy both do nothing, and I’m still bloated?
That’s useful information, not a dead end. It points toward a mechanism unrelated to either food, which is exactly what the evidence-graded remedies elsewhere on this site are organized to address.
How long does a proper reintroduction of one food actually take?
Most structured protocols use a few days to a week per food, at a measured amount, tracking symptoms daily, rather than a single bite followed by an immediate verdict. A single meal reaction is harder to interpret than a tracked pattern over several days.
I don’t have access to a dietitian or an allergist. Can I still do this myself?
Yes, with the caveat that self-directed reintroduction without blinding can’t fully rule out a nocebo response the way a formal trial does. Tracking symptoms on paper before and after, one food at a time, is still more informative than an open-ended, all-at-once elimination with no reintroduction plan at all.
Cutting more food was never guaranteed to be the next right step, and for most people testing what’s already been cut turns out to matter more. What’s actually graded by evidence to help, once the guesswork is out of it, is covered here.
Fortylight is written by Nina Halvorsen, a food scientist, not a physician, dietitian, or licensed health professional. Nothing here is medical advice. Digestive symptoms can signal serious conditions. If you have blood in your stool, unexplained weight loss, persistent vomiting, difficulty swallowing, bleeding after menopause, or symptoms that began after age 50, see a doctor before trying anything on this site.