THE RESEARCH

What the research shows about sublingual immunotherapy for food allergy

About twenty years of studies
Mostly peanut
Not FDA-approved
Reviewed by Hunter Smith, MD · Last reviewed October 2026
Illustration of a woman placing a dose of sublingual immunotherapy liquid under her tongue at home
IN SHORT
Sublingual immunotherapy is often called SLIT. It means putting a small, measured amount of a food protein under the tongue every day.
Studies going back about twenty years show it can raise how much of a food a person with a food allergy can be exposed to before they react.
For most people it does not make the food allergy go away. Most people keep avoiding the food. What changes is how much of an accidental exposure it takes to cause a reaction.
Side effects are usually mild. The most common one is itching in the mouth or throat.
Most of the research is on peanut. There is much less on milk, egg, hazelnut and peach, and almost none on other foods.
SLIT is not approved by the FDA for food allergy. Allergists who offer it are prescribing it off-label.
It works less well than oral immunotherapy (OIT) at building up tolerance to a full serving. It causes fewer whole-body reactions.

Words you will see in these studies

Food allergy studies measure a few specific things. Knowing what they mean makes the studies easier to read.

Desensitized

you can be exposed to more of the food without reacting, as long as you keep taking the daily dose. The protection depends on staying on treatment.

Remission

protection that lasts after treatment stops. It is tested by pausing the dose for a set time and then doing a supervised challenge.

Oral food challenge

a supervised test in a clinic where you eat increasing amounts of the food while a medical team stands by. This is how studies measure whether treatment worked.

Milligrams (mg) of protein

doses are measured in milligrams (mg) of the food's protein, not in grams (weight) of food. One average peanut contains roughly 250 mg of peanut protein, so 750 mg is about three peanuts.

Intention-to-treat vs Per-Protocol

some study results count everyone who started the study whether they finished the study or not. This is called intention-to-treat. Other results only count the people who followed the study protocol without major deviations. This is called per-protocol. Both are given when studies report both.
HOW BIG IS A MILLIGRAM DOSE?
Chart comparing peanut protein amounts: 4 milligrams is the top sublingual immunotherapy dose, 300 milligrams equals one peanut, and 5,000 milligrams equals one full serving

What SLIT can do, depending on what you want from it

People choose SLIT to treat food allergy with different goals. The evidence is stronger for some goals than others. It is worth being clear about your goals before you start.
GOAL 1

Protection against an accidental exposure

67 in 100
This is what SLIT does best, and it is what most of the research measures. In the longest study, 67 out of 100 children could eat at least 750 mg of peanut protein (about three peanuts) after three to five years of daily dosing. That is not enough to eat peanut freely. It is usually enough that a bite of the wrong cookie is no longer an emergency.
GOAL 2

A gentler way to start treatment, or something to do while you wait

35 of 50
Some people use SLIT as a first step before oral immunotherapy. In one study of 188 patients, a small group went on to a low-dose challenge, and 35 of 50 of those challenged passed cleanly enough to start OIT at a 300 mg dose without the usual slow build-up. Note: That study followed patients up to the point of starting OIT. It did not report how they did afterward.
The wait lists for OIT can be long, buildup can take many in-clinic appointments, and a number of studies show that the younger a person begins treatment, the better their outcomes. SLIT may be something you can begin now because the observed visits are much shorter and the risk of severe reactions is very low.
GOAL 3

Lasting protection in very young children

48 in 100
In a randomized, placebo-controlled trial of 50 children aged 1 to 4, 48 out of 100 were still protected three months after stopping treatment, a state doctors call remission. None of the children on placebo reached remission. Results were best in the youngest children, aged one and two.
About four months is as long as anyone has tested, so nobody knows whether this protection lasts for years. A separate study that followed older children after stopping found that about half had lost their protection by around 22 weeks (about 6 months). Desensitization, and even remission, is not the same thing as a cure.

The studies on SLIT for food allergy

A plain-language summary of what each study found. The exact figures and journal citations are in the full data table further down.

Kim 2024

50 people
Peanut, ages 1–4
60 in 100 could eat about 4,400 mg of peanut protein after 3 years of SLIT, compared with none of the children on placebo. 48 in 100 were still protected 3 months after stopping.
What to know: Randomized and placebo-controlled. No child in either group needed epinephrine.

Kim 2023

54 people
Peanut, ages 1–11
About 70 in 100 who completed treatment could eat at least 800 mg. 36 in 100 passed a 5,000 mg challenge.
What to know: Open-label, no control group. Protection faded after stopping: half had lost it by about 22 weeks.

Kim 2019

48 people
Peanut, ages 1–11
67 in 100 could eat at least 750 mg — about 3 peanuts. 25 in 100 passed a 5,000 mg challenge.
What to know: 3 to 5 years of daily dosing. Just under a quarter of families stopped before the end.

Fleischer 2013

40 people
Peanut, ages 12–37
The typical person who responded could eat about 496 mg of peanut powder — roughly 1 peanut. No one could eat the full 5 g challenge.
What to know: The authors describe the effect as modest. One person had a reaction at home that needed epinephrine.

Narisety 2015

21 people
Peanut, ages 7–13
SLIT and OIT both raised thresholds, but OIT raised them far more — about 22× versus 141×.
What to know: 9% of SLIT doses caused any reaction, compared with 43% of OIT doses. No one on SLIT needed epinephrine.

Keet 2012

30 people
Milk, ages 6–17
1 of 10 who stayed on SLIT passed a full 8 g milk challenge, compared with 14 of 20 who moved on to OIT.
What to know: OIT caused about 11× more whole-body reactions. Two children needed epinephrine after breathing in the liquid.

Soller 2024

188 people
Several foods, ages 4–18
Among a small group who reached a test challenge, 35 of 50 passed the challenge and transitioned into a low OIT dose.
What to know: Over 19,000 doses, only one serious reaction during buildup. None during maintenance.

Williams 2025

305 people
Several foods
40 of 57 test challenges passed into a low OIT dose.
What to know: About 2 in 100 had a moderate reaction during buildup. No severe reactions.

Safety of SLIT

Safety is one of the biggest concerns about immunotherapy. This is what the research does and does not show.
Illustration of a patient describing mouth discomfort to an allergist, with a dropper bottle and syringe on the table
45 of 48
children had mild symptoms at some point, on about 5 in 100 doses (Kim 2019)
0
epinephrine for a home dose, across 75,366 doses (Kim 2019)
80 vs 28
in 100 children had mouth itching, on treatment versus placebo (Kim 2024)
Most reactions to SLIT for food allergy are mild and are limited to itching or tingling in the mouth and throat. Mouth itch is quite common. In one long study, 45 of 48 children had mild symptoms at some point, on about 5 in 100 doses.
Kim 2019
Across 75,366 doses taken at home in that study, no one needed epinephrine for a home dose.
Kim 2019
In the randomized trial of 1 to 4 year olds, no child needed epinephrine in either the treatment group or the placebo group. Mouth itching was more common with treatment — about 80 in 100 children compared with 28 in 100 on placebo.
Kim 2024
Compared with OIT, SLIT causes far fewer whole-body reactions. In a head-to-head milk study, whole-body reactions were about 11 times more common with OIT. But the total number of reactions of any kind was about the same between OIT and SLIT.
Keet 2012
Eosinophilic esophagitis, a condition that causes trouble swallowing, has been reported with oral immunotherapy in about 3% of patients. SLIT is a recognized risk that should be discussed and watched for.
FULL DATA

The studies, with the numbers in full

The same eight studies as the summary above, with the clinical detail and citations.

Kim 2024

Peanut, ages 1–4. Randomized, placebo-controlled.
N = 50
DOSE AND DURATION
4 mg of peanut protein under the tongue daily for 36 months (3 years).
EFFICACY (WITH QUALIFIERS)
15 of 25 children on SLIT (60 in 100) ate the full 4,443 mg of peanut protein in the 3-year challenge, compared with 0 of 25 on placebo. The median amount tolerated was 4,443 mg on SLIT and 143 mg on placebo. 12 of 25 (48 in 100) were in remission 3 months after stopping; none on placebo reached remission. Results were best in children aged 1 and 2. Remission tested at 3 months only.
SAFETY
No child in either group needed epinephrine. Mouth itching in about 80 in 100 children on treatment compared with 28 in 100 on placebo.
CITATION
Kim EH, Bird JA, Keet CA, et al. Desensitization and remission after peanut sublingual immunotherapy in 1- to 4-year-old peanut-allergic children: a randomized, placebo-controlled trial. J Allergy Clin Immunol. 2024;153(1):173-181.e10. doi: 10.1016/j.jaci.2023.08.032. PubMed: 37815782.

Kim 2023

Peanut, ages 1–11. Open-label, no control group.
N = 54
DOSE AND DURATION
4 mg of peanut protein under the tongue daily for 48 months (4 years), then a randomly assigned break of 1 to 17 weeks before a final challenge.
EFFICACY (WITH QUALIFIERS)
47 of 54 completed treatment. The average amount eaten in the challenge rose from 48 mg to 2,723 mg of peanut protein. About 70 in 100 who completed treatment could eat at least 800 mg, and 36 in 100 passed the full 5,000 mg challenge. Protection faded after stopping: the modeled median time to losing it was about 22 weeks (about 6 months).
SAFETY
The median reaction rate was about 0.5 in 100 doses, most often brief mouth or throat itching. No dosing symptoms needed epinephrine.
CITATION
Kim EH, Keet CA, Virkud YV, et al. Open-label study of the efficacy, safety, and durability of peanut sublingual immunotherapy in peanut-allergic children. J Allergy Clin Immunol. 2023;151(6):1558-1565.e6. doi: 10.1016/j.jaci.2023.01.036. PubMed: 36828080.

Kim 2019

Peanut, ages 1–11.
N = 48
DOSE AND DURATION
2 mg of peanut protein under the tongue daily, for 3 to 5 years.
EFFICACY (WITH QUALIFIERS)
32 of 48 (67 in 100) could eat at least 750 mg (about 3 peanuts). 12 of 48 (25 in 100) passed the full 5,000 mg challenge. 37 of 48 completed 3 to 5 years; just under a quarter stopped before the end.
SAFETY
45 of 48 children had mild symptoms at some point, on about 5 in 100 doses. No epinephrine needed for any of 75,366 home doses.
CITATION
Kim EH, et al. Long-term sublingual immunotherapy for peanut allergy in children: clinical and immunologic evidence of desensitization. J Allergy Clin Immunol. 2019;144(5):1320-1326.e1. doi: 10.1016/j.jaci.2019.07.030. PubMed: 31493887.

Fleischer 2013

Peanut, ages 12–37.
N = 40
DOSE AND DURATION
Up to 1.4 mg (1,386 micrograms) of peanut protein under the tongue daily for 44 weeks, compared with placebo. The placebo group then switched to a higher dose of up to 3.7 mg daily.
EFFICACY (WITH QUALIFIERS)
14 of 20 on SLIT (70 in 100) met the study’s definition of a response, compared with 3 of 20 on placebo (15 in 100). Among responders, the median amount eaten rose from 3.5 mg to 496 mg of peanut powder (about 250 mg of peanut protein, roughly 1 peanut). No one could eat the full 5 g of peanut powder in the challenge. The authors describe the effect as modest.
SAFETY
One person had a reaction at home that needed epinephrine.
CITATION
Fleischer DM, Burks AW, Vickery BP, et al. Sublingual immunotherapy for peanut allergy: a randomized, double-blind, placebo-controlled multicenter trial. J Allergy Clin Immunol. 2013;131(1):119-127.e7. doi: 10.1016/j.jaci.2012.11.011. PubMed: 23265698.

Narisety 2015

Peanut, ages 7–13. SLIT compared with OIT.
N = 21
DOSE AND DURATION
Double-blind, double-dummy: SLIT up to 3.7 mg of peanut protein daily, or OIT up to 2,000 mg daily, for 12 months of maintenance, with an optional 6 more months after unblinding.
EFFICACY (WITH QUALIFIERS)
SLIT and OIT both raised thresholds; OIT raised them far more — about 22× (SLIT) versus 141× (OIT).
SAFETY
9% of SLIT doses caused any reaction, compared with 43% of OIT doses. No one on SLIT needed epinephrine.
CITATION
Narisety SD, Frischmeyer-Guerrerio PA, Keet CA, et al. A randomized, double-blind, placebo-controlled pilot study of sublingual versus oral immunotherapy for the treatment of peanut allergy. J Allergy Clin Immunol. 2015;135(5):1275-1282.e6. doi: 10.1016/j.jaci.2014.11.005. PubMed: 25528358.

Keet 2012

Milk, ages 6–17. SLIT compared with OIT.
N = 30
DOSE AND DURATION
SLIT up to 7 mg of milk protein daily, or SLIT followed by OIT up to 1,000 or 2,000 mg daily. Challenges after 12 and 60 weeks of maintenance.
EFFICACY (WITH QUALIFIERS)
1 of 10 who stayed on SLIT passed a full 8 g milk challenge, compared with 14 of 20 who moved on to OIT.
SAFETY
Whole-body reactions were about 11× more common with OIT. Total reactions of any kind were about the same between OIT and SLIT. Two children needed epinephrine after breathing in the liquid.
CITATION
Keet CA, Frischmeyer-Guerrerio PA, Thyagarajan A, et al. The safety and efficacy of sublingual and oral immunotherapy for milk allergy. J Allergy Clin Immunol. 2012;129(2):448-455.e5. doi: 10.1016/j.jaci.2011.10.023. PubMed: 22130425.

Soller 2024

Several foods, ages 4–18. SLIT as a step before OIT.
N = 188
DOSE AND DURATION
Over 19,000 doses. Buildup to 2 mg of food protein under the tongue daily, over 3 to 5 clinic visits, then 1 to 2 years of daily maintenance before an optional 300 mg challenge.
EFFICACY (WITH QUALIFIERS)
Among a small group who reached a test challenge, 35 of 50 passed the challenge and transitioned into a low OIT dose (300 mg) without the usual slow build-up. Patients were followed only up to the point of starting OIT; later outcomes were not reported.
SAFETY
Only one serious reaction during buildup. None during maintenance.
CITATION
Soller L, Williams BA, Mak R, et al. Safety and effectiveness of bypassing oral immunotherapy buildup with an initial phase of sublingual immunotherapy for higher-risk food allergy. J Allergy Clin Immunol Pract. 2024;12(5):1283-1296.e2. doi: 10.1016/j.jaip.2024.02.024. PubMed: 38423293.

Williams 2025

Several foods.
N = 305
DOSE AND DURATION
SLIT solutions made from grocery-store foods (protocols for 31 foods), taken daily for 1 to 2 years, then an optional 330 to 340 mg challenge.
EFFICACY (WITH QUALIFIERS)
40 of 57 test challenges passed into a low OIT dose.
SAFETY
About 2 in 100 had a moderate reaction during buildup. No severe reactions.
CITATION
Williams BA, Baaske A, Soller L, et al. The use of grocery-sourced real food solutions in sublingual immunotherapy for food allergies. Ann Allergy Asthma Immunol. 2025. doi: 10.1016/j.anai.2025.03.001. PubMed: 40064397.

What the evidence does not show

SLIT builds less tolerance than OIT.

Every study that compares SLIT with OIT has the same finding. If your goal is eating a full serving of the food, OIT is more likely to get you there.

Almost all the evidence is about peanut.

Milk and egg have essentially no published SLIT trials. The hazelnut evidence rests on 23 adults. Talk to your allergist about their internal data on SLIT for other foods.

Nobody knows how long protection lasts after stopping.

The longest period studied was about four months (17 weeks).

SLIT is not FDA approved for food allergy.

There is no approved food SLIT product in the United States. Insurance coverage is inconsistent as a result.

There are limited studies of SLIT for food allergy for adults.

Almost all of the studies were done in children.

Real-world cohorts are uncontrolled.

Soller 2024 and Williams 2025 describe patients treated in one allergy program in Vancouver, with no placebo or comparison group. They show what happened in practice, but they cannot show how much of the change was caused by SLIT itself.

Trial populations excluded the highest-risk patients.

The controlled trials excluded many of the people at highest risk. Narisety 2015 excluded children with a history of severe anaphylaxis (low oxygen, low blood pressure or neurological symptoms), poorly controlled asthma, or eosinophilic esophagitis. Fleischer 2013 excluded people with asthma needing high-dose inhaled steroids or a history of intubation. The results may not apply to these patients.
KEEP READING