AIP Autoimmune Protocol Diet: Evidence Review

Disclaimer: This article is for educational and research purposes only and does not constitute medical advice. The Autoimmune Protocol is a highly restrictive dietary approach. Always consult a qualified healthcare practitioner — ideally an accredited practising dietitian and your specialist — before undertaking any elimination diet, particularly if you are managing an autoimmune condition with medication.

The Autoimmune Protocol (AIP) diet has moved from wellness circles into early clinical research. There are now published human pilot trials showing measurable outcomes in inflammatory bowel disease and Hashimoto's thyroiditis — but no controlled trials, and no guideline endorsement. This article covers AIP's origins, its two-phase structure, what the research actually shows, its legitimate limitations, and practical considerations for Australian implementation.

---

Origins: From Paleo to Autoimmune Protocol

The Paleo Foundation

The Autoimmune Protocol evolved from the broader paleo dietary framework, which itself centres on removing grains, legumes, dairy, and processed foods in favour of animal proteins, vegetables, fruits, and nuts. The paleo rationale — that modern agricultural foods introduce antigens and antinutrients that humans are not evolutionarily adapted to — is contested in the nutrition literature, but it provided the conceptual scaffolding on which AIP was built.

AIP goes considerably further than standard paleo. It removes an additional layer of foods hypothesised to have immunostimulatory or gut-permeability-altering effects: nightshade vegetables, eggs, nuts and seeds, alcohol, NSAIDs, and all dairy including ghee. The working hypothesis is that by removing these potential triggers, the immune system's inflammatory burden is reduced enough to allow intestinal healing and immune recalibration.

The Wahls Protocol Overlap

AIP shares conceptual ground with the Wahls Protocol, developed by Dr Terry Wahls for her own secondary progressive multiple sclerosis management. Both eliminate grains, legumes, eggs, and nightshades, but differ in emphasis: Wahls structures intake around nine cups of specific vegetables daily for nutrient density, while AIP focuses on systematic trigger identification via reintroduction. AIP is the more studied in IBD and Hashimoto's contexts; Wahls has its own small evidence thread for MS fatigue. They are related but separate frameworks.

---

Protocol Structure: Elimination and Reintroduction

Phase 1 — Elimination

The elimination phase of AIP removes:

• Grains (all, including gluten-free grains such as rice and oats)
• Legumes (including lentils, chickpeas, soy, and peanuts)
• Dairy (all forms, including ghee and whey protein)
• Eggs (including egg whites, which contain lysozyme hypothesised to cross gut epithelium)
• Nightshade vegetables (tomato, capsicum/pepper, eggplant, potato, goji berries)
• Nuts and seeds (including nut oils, seed-based spices such as coriander, cumin, and sesame)
• Alcohol (all forms)
• NSAIDs and most over-the-counter anti-inflammatory medications
• Added sugars and sugar alcohols

What remains is a diet built primarily on: bone broth, organ meats, fatty fish, grass-fed red meat, leafy green vegetables, non-nightshade vegetables (including cruciferous, brassica, and root vegetables excluding potato), fermented vegetables such as sauerkraut and kimchi (without nightshade additions), small amounts of fruit, coconut products, and olive oil.

The elimination phase is typically maintained for a minimum of 30 days, with many practitioners and researchers extending it to 60–90 days before reintroduction begins. The rationale is that mucosal healing and immune recalibration require sustained time without the hypothesised triggers, not just a brief washout.

Phase 2 — Systematic Reintroduction

The reintroduction phase is where AIP diverges sharply from a simple exclusion diet. Foods are reintroduced one at a time, in a structured sequence, with several days between each trial to allow assessment of symptomatic response.

The standard reintroduction sequence moves from foods with the lowest hypothesised immunoreactivity to the highest:

1. Stage 1: Egg yolks, fruit-based spices (not seed-based), legume pods (green beans, snow peas), ghee from grass-fed dairy
2. Stage 2: Seeds (seed-based spices first, then seed foods), cocoa and coffee, egg whites, grass-fed butter
3. Stage 3: Nuts, nightshade spices (paprika, cayenne), dairy (aged cheese first, then fermented dairy)
4. Stage 4: Full nightshades, whole grains, legumes, eggs in full

Each food is tested over approximately three days: consumed in a meaningful portion on day one, then not eaten for two days while symptoms are monitored. If no response occurs, the food is considered tolerated and reintroduced to the regular diet. If symptoms — digestive changes, joint discomfort, skin changes, fatigue, or mood shifts — emerge, the food is excluded and retested after the protocol has progressed further.

The goal is not permanent elimination of all these food categories. It is personalised identification of individual trigger foods, with the broader AIP elimination phase functioning as a controlled baseline from which individual tolerance can be assessed.

---

The Clinical Evidence

Konijeti 2017: IBD Pilot Trial

The most cited human trial of AIP in an inflammatory bowel disease population is the Konijeti et al. 2017 pilot study, published in Inflammatory Bowel Diseases (PMID: 28858071; full text: PMC5647120). Led by gastroenterologist Dr Gauree Konijeti at Scripps Clinic, San Diego, this was the first published clinical trial to formally evaluate AIP in IBD patients.

The study enrolled 15 adults with active Crohn's disease or ulcerative colitis, all of whom had been symptomatic on stable medication doses. The protocol comprised a six-week elimination phase followed by a five-week maintenance phase, with structured dietary support including cooking classes and health coaching.

Key findings:

• 73% of participants achieved clinical remission by week 6 of the elimination phase, as measured by validated symptom indices (the Harvey-Bradshaw Index for Crohn's, the Mayo Score for UC).
• Endoscopic remission (assessed in a subset of participants) was achieved in approximately half of those assessed.
• Faecal calprotectin — a marker of intestinal inflammation — trended downward, though did not reach statistical significance in this sample.
• The diet was well tolerated: no participants withdrew due to dietary difficulties, though the highly structured support environment likely contributed to adherence.

Critical limitations acknowledged by the authors: no control group, small sample size (n=15), single-centre design, and the confounding effect of concurrent medication. The study cannot establish whether improvements were caused by AIP, by placebo effect, by regression to the mean, or by the intensive support structure. The authors explicitly described it as a feasibility and pilot study, not a definitive efficacy trial.

A follow-up quality-of-life study (PMID: 31832627) from the same Scripps group showed sustained improvements in patient-reported outcomes in IBD participants following AIP, further supporting the case for a larger randomised controlled trial — which, as of this writing, has not yet been published.

Abbott 2019: Hashimoto's Thyroiditis Study

The second significant human trial is Abbott et al. 2019, published in Cureus (PMID: 31275780; full text: PMC6592837). This study examined 17 women aged 20–45 with confirmed Hashimoto's thyroiditis who participated in a structured 10-week online health coaching programme implementing AIP.

Key findings:

• Statistically significant improvements in health-related quality of life (HRQL) across all eight subscales of the SF-36 questionnaire, with the most marked improvements in physical role functioning, emotional role functioning, vitality, and general health.
• No statistically significant changes in thyroid function markers (TSH, free T4, free T3) or thyroid antibody titres (TPO-Ab, TG-Ab) over the 10-week period.
• A decrease in mean high-sensitivity CRP (hs-CRP) and changes in white blood cell differential consistent with modulation of the inflammatory response.

The authors interpreted these findings as suggesting AIP may reduce systemic inflammation and improve quality of life in Hashimoto's patients, even in the absence of measurable changes in thyroid-specific markers. A 10-week window is arguably too short to expect antibody titre changes; autoimmune remission is a longer-term process.

Limitations are similar to the Konijeti study: very small sample (n=17), all-female cohort, no control group, and an online coaching format that may limit generalisability. The study population also reported high baseline dietary restriction, suggesting potential selection bias toward participants already oriented toward dietary intervention.

---

Mechanistic Rationale

AIP's working hypotheses centre on three overlapping mechanisms. First, intestinal permeability: compounds in grains (gliadin, wheat germ agglutinin), legumes (lectins, saponins), and nightshades (glycoalkaloids) are proposed to loosen tight junctions in the gut epithelium, allowing luminal antigens systemic access and amplifying immune reactivity. This is supported in cell and animal models but remains contentious in human trials. For the intestinal permeability evidence in depth, see our review of leaky gut research and intestinal permeability.

Second, nutrient density: the AIP food list — organ meats, fatty fish, leafy greens, coloured vegetables, bone broth — is exceptionally micronutrient-rich. Adequate zinc, selenium, vitamin A, and vitamin D are each required for immune regulation, and AIP delivers them in quantity.

Third, antigenic load reduction: eliminating a large number of dietary proteins with immunostimulatory potential in sensitive individuals reduces the mucosal immune burden, potentially restoring regulatory T-cell balance. None of these pathways has been confirmed as the active driver of AIP's clinical effects — they are plausible hypotheses, not established mechanisms.

---

Legitimate Limitations and Concerns

Small Study Sizes

The existing evidence base for AIP consists of two small pilot studies with a combined enrolment of 32 participants. This is an insufficient foundation for clinical recommendations. The effect sizes reported are promising enough to justify larger randomised controlled trials — but those trials have not yet been conducted or published.

Nutritional Restriction Risk

The elimination phase removes whole food groups that provide dietary fibre, calcium, and B vitamins. Removal of dairy, almonds, and seeds reduces calcium intake substantially. Eliminating legumes, oats, and grains removes major prebiotic substrates — beta-glucan, arabinoxylan, resistant starch — that sustain Bifidobacterium and Akkermansia populations critical for mucosal immunity. Prolonged AIP without dietitian supervision carries real risk of fibre insufficiency, microbiome diversity loss, and nutrient deficiencies that are counterproductive to the goals of the protocol.

Disordered Eating Risk

A highly restrictive elimination diet carries risk of reinforcing disordered eating patterns, particularly in autoimmune patients who may already have an anxious relationship with food. Practitioners should screen for eating disorder history and monitor psychological wellbeing throughout.

Lack of Long-Term Data

Neither published trial followed participants with meaningful longitudinal data beyond the active intervention. Whether remission is sustained, microbiome changes are durable, or reintroduction is consistently achievable in practice remains unanswered.

---

Australian Implementation Considerations

Sourcing: Grass-fed beef and lamb are stocked as standard at most independent Australian butchers, and chicken frames or beef knuckles for bone broth are inexpensive and widely available. Organ meats — liver, kidney, heart — are reliably sourced from independent butchers and farmers' markets, less so from major supermarkets.

Nightshades: Australian cooking relies heavily on tomato, capsicum, and potato — all eliminated during AIP. Sweet potato, carrot, parsnip, and beetroot substitute effectively as starchy carbohydrate sources.

Fermented vegetables: Plain sauerkraut (without nightshade additions) is available refrigerated at Coles and Woolworths, or straightforwardly made at home from cabbage and salt.

Healthcare system: GPs and specialists will not typically supervise AIP given it lacks guideline endorsement. An Accredited Practising Dietitian (APD) with elimination diet experience is the appropriate professional. A Chronic Disease Management plan via your GP can subsidise up to five APD consultations per calendar year.

Medications: AIP does not replace prescribed autoimmune therapies. Immunosuppressants, biologics, and aminosalicylates in IBD, and levothyroxine in Hashimoto's, continue as prescribed. Never reduce medication based on dietary improvement without specialist guidance.

---

Who Might Consider AIP

AIP is not a first-line intervention. It is most worth considering for individuals with a confirmed autoimmune diagnosis (IBD, Hashimoto's, rheumatoid arthritis, psoriasis) who have ongoing symptomatic burden despite standard medical management, the time and cooking capacity to implement it properly, and access to APD supervision. It should be framed from the outset as a time-limited 6–12 week trial with a systematic reintroduction plan — not a permanent diet.

Those with eating disorder history, significant nutritional deficiencies, or limited food access are not appropriate candidates without close professional oversight.

For chronic inflammation without a formal autoimmune diagnosis, a less restrictive approach — the Mediterranean diet or a structured anti-inflammatory diet protocol — is better supported by evidence and nutritionally more robust for the long term. For a broader view of how dietary patterns interact with immune and nervous system function, see our overview of gut-brain axis nutrition strategies.

---

Frequently Asked Questions

Is AIP the same as a paleo diet?

No. Standard paleo removes grains, legumes, dairy, and processed foods. AIP removes all of those plus eggs, nightshades, nuts, seeds, seed-based spices, alcohol, and NSAIDs. AIP is a stricter, time-limited clinical protocol with a structured reintroduction phase — paleo is a broad long-term dietary pattern without a reintroduction component.

How long should the elimination phase last?

Most practitioners and the published trials use 6 weeks minimum, with many recommending 8–12 weeks for individuals with longstanding autoimmune conditions. Shorter periods are unlikely to allow adequate mucosal healing or immune recalibration to produce a meaningful reintroduction baseline.

Do I need to eat organ meats?

Organ meats are strongly emphasised because they supply vitamin A, zinc, B12, copper, and selenium in quantities that the elimination of dairy, eggs, nuts, and seeds otherwise reduces. They are not strictly mandatory, but skipping them significantly increases nutritional risk. If organ meats are not tolerable, supplementation guided by an APD is advisable.

---

Key Takeaways

The Konijeti 2017 IBD pilot trial and Abbott 2019 Hashimoto's study are the two published human trials of AIP, with a combined enrolment of 32 participants. Both showed meaningful signals — clinical remission rates, quality-of-life improvements, reduced hs-CRP — but neither was controlled, and neither is sufficient for clinical recommendations. Larger randomised trials are needed and have not yet been published.

AIP is not an indefinitely restrictive diet. The elimination phase is a controlled baseline; systematic reintroduction is the mechanism by which individual trigger foods are identified, and most foods can eventually be returned to the diet. Practitioners presenting AIP as permanent elimination are misrepresenting the protocol.

For Australians with confirmed autoimmune conditions and residual symptomatic burden despite medical management, AIP is worth discussing with an APD and treating specialist as an adjunctive time-limited intervention. The nutritional risks of prolonged restriction — particularly for fibre, calcium, and prebiotic diversity — are real and require professional oversight throughout.