Two small studies followed people with HS on a yeast-free diet. In the first (12 people, after surgery), lesions regressed or stabilized and reappeared after eating beer, yeast or wheat. In a later follow-up of 37 people, 70% reported improvement on exclusion and 87% of those who reintroduced yeast reported recurrence within a week.
Neither study was randomized or controlled, and in the first one diet was combined with surgery, so the effect of yeast alone cannot be isolated. Guidelines consider the evidence insufficient to recommend exclusion for everyone.
Read "worsening" as a reported signal, not a rule (small clinical series and patient reports). It does not mean brewer's yeast works the same way for everyone with HS.
Luca's experience · one person, not a prediction. Beer was one of the first things I removed during my elimination phase, and I never properly reintroduced it: to this day it stays a rare exception I allow myself, not a food that came back.
If you test it: Yeast hides in more places than beer: bread, stock cubes, some supplements. If you test it, read labels and change one thing at a time.
A 2015 review proposed a dairy-free, low-glycemic diet based on hormonal mechanisms and an uncontrolled personal series. A 2024 case-control study found people with HS consumed dairy more often than controls.
No elimination-and-rechallenge study exists. The often-quoted "47 patients improved" series had no validated outcomes or control group. Guidelines do not recommend routine dairy exclusion, and cutting dairy without substitutes has nutritional costs. The popular casein hypothesis (digestion of A1 beta-casein yields beta-casomorphin-7, linked to digestive discomfort in general studies) has never been tested in HS.
Read "mixed" as a reported signal, not a rule (patient surveys and mechanistic hypotheses). It does not mean dairy works the same way for everyone with HS.
Gut barrier. In a two-week human randomized crossover trial, a milk-fat-globule-membrane-enriched dairy beverage did not change small-intestinal or colonic permeability compared with a soy-based comparator in adults with metabolic syndrome. This is short-term evidence about one dairy formulation, not dairy as a whole and not an HS outcome.
Luca's experience · one person, not a prediction. Dairy was part of my AIP elimination. From my own notes, my suspicion falls on casein more than lactose: and when I test dairy, I start from goat products, whose casein profile differs from cow's A1. My result is mine, not a prediction for yours.
If you test it: If you test dairy, reintroduce one product at a time (milk, cheese and whey behave differently) and give each test at least a week of notes. A common starting order, and mine, is goat or sheep dairy first: their casein profile differs from cow's A1.
A 2024 case-control study found a higher dietary glycemic index in people with HS, correlating with disease severity scores. In surveys, sweets and refined carbohydrates are among the foods patients most often report as aggravating. Separately, a 2018 case-control study found insulin resistance more often in HS than in matched controls, supporting a metabolic mechanism but not identifying sugar as the cause.
There is no randomized trial of a low-glycemic diet in HS. Insulin resistance may be influenced by inflammation, weight and the overall dietary pattern, so the metabolic association does not show that sugar alone drives flares.
Read "worsening" as a reported signal, not a rule (patient surveys and one case-control study). It does not mean sugar and high-glycemic foods works the same way for everyone with HS.
Gut barrier. In an animal study, mice given 15% glucose for nine weeks had increased paracellular permeability in the jejunum and cecum, whereas the matched fructose exposure did not alter permeability. This finding is specific to the tested sugars and doses, not a general high-glycemic-diet result and not an HS outcome.
Luca's experience · one person, not a prediction. Refined sugar is the one I felt most clearly in my own tracking. I wrote a full article about the insulin mechanism and my experience.
If you test it: Test categories, not "sugar" in general: a fruit is not a soft drink. Track amount and timing, not just yes/no.
In a 242-person survey, 36.7% of respondents had changed their tomato consumption. A 2024 case-control study found different nightshade consumption between people with HS and controls.
No study shows that removing nightshades improves HS. The claim "nightshades cause flares" is community-reported only: it describes what patients try, not what has been demonstrated.
Gut barrier. In animal studies, potato glycoalkaloids (solanine and chaconine) disrupted the intestinal barrier and aggravated gut inflammation at dietary-level doses. Mechanistic evidence from mice, not an HS outcome.
Luca's experience · one person, not a prediction. Nightshades are excluded by default in the AIP elimination phase, so I tested them on reintroduction. Tomatoes came back fine in my notes, and so did potatoes: well peeled, since glycoalkaloids concentrate in and near the skin.
If you test it: The nightshade family is broad: test tomatoes, peppers and potatoes separately before deciding the whole family is a problem. If you test potatoes, peeling them well removes most of the glycoalkaloids, which sit in and near the skin.
In a retrospective series of 40 people, two turned out to have undiagnosed celiac disease and both reported their HS improved on a gluten-free diet. In surveys, gluten is one of the foods patients most often modify. Outside HS, gliadin has been shown to activate zonulin-related tight-junction signaling and increase intestinal permeability; this is mechanistic plausibility, not clinical evidence for an HS effect.
For people without celiac disease, there is no clinical evidence that going gluten-free changes HS. The zonulin literature does not show that intestinal permeability mediates HS flares. If you suspect celiac disease, testing before cutting gluten matters because exclusion can make diagnosis harder afterwards.
Gut barrier. Gliadin triggers zonulin release, which loosens intestinal tight junctions. Documented mechanism, mostly studied in celiac disease: not an HS outcome.
Luca's experience · one person, not a prediction. Gluten went out with the AIP elimination and came back in stages. What I noticed is in my reintroduction notes, and it says nothing about your body.
If you test it: If celiac disease runs in your family or you have gut symptoms, talk to your doctor about testing before any exclusion.
In a 242-person survey, 37.1% had changed their alcohol consumption. Reviews acknowledge the absence of reliable clinical data linking ethanol itself to HS flares. Outside HS, chronic alcohol exposure can alter gut microbiota and intestinal permeability; that mechanism has not been connected to HS outcomes.
The beer relapses seen in the yeast studies cannot be attributed to alcohol because beer also contains brewer's yeast. No controlled HS study has separated ethanol from beverage type or tested whether the gut-barrier mechanism changes flare activity.
Read "mixed" as a reported signal, not a rule (patient surveys). It does not mean alcohol works the same way for everyone with HS.
Gut barrier. Chronic alcohol exposure alters the gut microbiota and increases intestinal permeability in general studies. Not HS-specific.
Luca's experience · one person, not a prediction. I mostly experienced alcohol through beer, which is also a yeast question. Wine and beer behaved differently for me, which taught me to test them separately.
If you test it: Test wine and beer as different things, not as one "alcohol" box: beer also carries yeast.
Case-control studies found that overall diet quality, including red and processed meat consumption, differs between people with HS and controls; a 2024 study compared 1,004 people with HS against 4,436 controls.
No study isolates meat's effect on HS from weight, smoking, socioeconomic status and the rest of the diet. There is no exclusion trial.
Gut barrier. In an animal study, a haem-iron-enriched diet at a concentration intended to model meat exposure increased colonic permeability in rats. Haem iron is only one component of red meat, and this was not an HS outcome.
If you test it: If you test this, distinguish fresh red meat from processed meat: they are different foods with different additives.
Three separate studies associate higher adherence to a Mediterranean-style pattern with lower HS severity. In one case-control study, dietary n-3 polyunsaturated fatty acids correlated inversely with Sartorius score after adjustment. In a 221-person study, using extra-virgin olive oil as the main culinary fat was the individual PREDIMED item associated with lower self-reported disease activity.
No randomized trial has tested the Mediterranean diet as an HS intervention. Reverse causality is possible: people who feel better may simply eat better.
Read "beneficial" as a reported signal, not a rule (three observational studies). It does not mean mediterranean diet works the same way for everyone with HS.
Gut barrier. In a human randomized trial of 260 women with baseline barrier impairment, a Mediterranean-diet intervention was associated with lower plasma LBP and fecal zonulin and higher short-chain fatty acids after three months. These are indirect barrier markers in a specific population, not an HS outcome.
Luca's experience · one person, not a prediction. My long-term way of eating after reintroduction ended up looking a lot like this pattern, more by results than by design.
In a cross-sectional study of 221 people with HS, using extra-virgin olive oil as the main culinary fat was the individual Mediterranean-diet item associated with lower self-reported disease activity.
The study did not assign olive oil, measure flare changes prospectively or isolate it from the wider dietary and physical-activity pattern. There is no HS-specific olive-oil trial.
Read "beneficial" as a reported signal, not a rule (one cross-sectional dietary association, not a tolerance or rechallenge study). It does not mean extra-virgin olive oil works the same way for everyone with HS.
Gut barrier. In an animal study, several extra-virgin olive-oil cultivars reduced measured intestinal permeability in mice with DSS-induced colitis, but the response was not uniform across cultivars. This disease-model evidence was not an HS outcome.
If you test it: No standardized HS-specific reintroduction protocol exists.
A 55-person observational within-person study followed people with HS through Ramadan fasting while medical treatment stayed stable. Mean IHS4 decreased from 11.00 to 10.15, without an association with weight change.
There was no non-fasting control group, the mean score change was modest and participants remained on varied antibiotics or biologics. Ramadan fasting does not establish the effect of other time-restricted eating schedules.
Read "beneficial" as a reported signal, not a rule (one small within-person study during Ramadan). It does not mean intermittent fasting works the same way for everyone with HS.
Gut barrier. In an animal study of ovalbumin-induced food allergy, two intermittent-fasting schedules were associated with higher ileal ZO-1 expression and better epithelial-barrier integrity. The study did not directly establish the same effect in humans and did not assess an HS outcome.
A 28-day single-arm pilot in 12 treatment-naive women with HS and overweight or obesity reported a 24.37% mean reduction in Sartorius score alongside weight loss and decreases in TMAO and oxidative-stress markers.
There was no control group and the sample was small and narrow. The study cannot separate calorie restriction, ketosis, weight loss or other dietary changes, and it does not establish durability or applicability outside women with excess weight.
Read "beneficial" as a reported signal, not a rule (one uncontrolled 28-day pilot). It does not mean very-low-calorie ketogenic diet works the same way for everyone with HS.
Gut barrier. In an uncontrolled human pilot study of 24 adults with obesity, eight weeks of a very-low-calorie ketogenic diet increased the lactulose/mannitol ratio and dysbiosis markers despite weight loss. Calorie restriction, ketosis and weight change cannot be separated, and this was not an HS outcome.
The Mediterranean pattern is a way of eating, not a single food: the association with lower severity is covered in the full Mediterranean diet article. For the method I used to test foods one at a time, start from the reintroduction phase guide.
The only test that counts
Evidence tells you where to look. Your body tells you what's true.
Everything on this page is a map of what's been studied, not a list of rules. The way I found my own triggers was an elimination and reintroduction protocol: one change at a time, a defined window, notes on everything. The full method, and the journal I used to track it, are in the book.
How I tested mine → the book The reintroduction method (free)
I am not a doctor: this page gathers published research and patient-reported information for informational purposes. It is not medical advice, and it is not a recommendation to start, stop or change any treatment, diet or product. Decisions about your health belong in a conversation with your doctor.