Every LabelCheck every label, every time.

← Peanut allergy

Peanut cross-reactivity

If your child is allergic to peanut, the reassuring headline is that the allergy is usually narrower than the food label aisle makes it feel. Peanut is a legume, so the foods it can genuinely cross-react with are other legumes, and most of those turn out to be safe to eat even when a blood test flags them. This page is the deep version of that story: the short list of cross-reactions that really do matter, the longer list that looks alarming on a test and rarely is, the molecular reason some related foods are safe and others are not, and what to actually do about each one. It is the companion to the peanut profile; the profile gives you the summary, this page gives you the why.

The short answer: what to do with each food

If you only read one section, read this. The rest of the page is the evidence behind it.

  • Other beans, peas, lentils, chickpeas, and soy. Most peanut-allergic children tolerate these, because a positive test for them is usually a flag without a fight, not a real-world allergy. A positive panel result is a reason to ask your allergist, not a reason to pull the food on your own.
  • Lupin. This is the one legume to take seriously. Peanut and lupin genuinely cross-react, so do not assume tolerance, and read labels for lupin flour.
  • Fenugreek. A real cross-reaction, and it hides in spice blends. Worth knowing by name.
  • Tree nuts. Many peanut-allergic children are also tree-nut allergic, but that is co-occurrence, not a shared protein. The rule is “test the tree nuts, do not assume,” not “avoid all nuts.”
  • Coconut. Despite the name and the US label rule, coconut is botanically a fruit, and most tree-nut-allergic people tolerate it.

The thread running through all of it: a positive test is not the same as a reaction, and the decision to add any of these foods back is one to make with your allergist, not from a page.

Confirmed cross-reactive partners

These are the foods where the cross-reaction is real, not just a test artifact. They split into two groups by what you do about them: the ones you treat as a managed pair, and the ones you test rather than assume.

Manage as a pair: lupin and fenugreek

Lupin is the real legume exception. Peanut and lupin cross-react, and a peanut-allergic person can develop a true lupin allergy through that overlap. This is the cross-reaction that genuinely changes the plate, so do not assume a peanut-allergic child tolerates lupin. The practical problem is that lupin is easy to miss: lupin flour is a common wheat-flour substitute in European and imported breads, croissants, pasta, pizza dough, vegan meat substitutes, and gluten-free baked goods, and reactions can be severe. In the EU and UK, lupin is a mandatory declarable allergen and must appear in bold; in the US it is not a major allergen under labeling law, so a US label may not flag it at all. That labeling gap is the reason it stays hidden.

Fenugreek is the second. Peanut and fenugreek can cross-react, and fenugreek allergy most often grows out of a primary peanut allergy rather than appearing on its own. Fenugreek is a spice, so it hides in curry powders and spice blends, and it carries the sotolon note that turns up in some maple-flavored products. If your child is peanut-allergic, fenugreek is worth flagging to your allergist as a possible cross-reactant rather than assuming it is fine.

Test, do not assume: soy and chickpea

These two are the in-between case. They cross-react often enough on a test that allergists pay attention, but the test result usually overstates the real-world risk.

Soy. Peanut and soy commonly show cross-sensitization (frequent co-positive testing), but clinically relevant cross-reactivity is rare, and most peanut-allergic people tolerate soy. A positive soy result in a peanut-allergic child is usually not a reason to pull soy from the diet, but whether your child can have it is a confirm-with-your-allergist decision, not a blanket green light.

Chickpea. The same pattern. Chickpea and peanut commonly show co-sensitization, with clinically relevant cross-reactivity much lower than the testing suggests. Treat a positive chickpea result the way you treat a positive soy result: a reason to ask, not a reason to assume either safety or danger.

A note on peas: pea protein is showing up more and more in plant-based foods, and a positive pea test in a peanut-allergic child is a “test, do not assume” situation, the same as soy and chickpea. Do not treat pea as automatically safe, and do not treat it as automatically dangerous; it is an allergist conversation. Where pea hides is in the next section.

Hidden and non-obvious cross-reactive exposures

The legume family hides in places a “no peanuts” habit does not catch, mostly because the words do not look like the food. These are worth a one-time read so you can spot them.

  • Lupin flour in European, imported, and gluten-free baked goods (covered above; the single most important hidden legume to learn).
  • Pea protein isolate is rapidly replacing soy as a “soy-free, dairy-free” protein in plant-based meats, dairy substitutes, protein bars and shakes, and some infant products, often labeled only “pea protein,” “pea fibre,” or “vegetable protein.” Pea is not a US major allergen, so it can be easy to miss. If you switch to pea-based products, read labels closely.
  • Chickpea under other names. Besan, gram flour, and garbanzo flour are all ground chickpea, central to South Asian cooking (pakora, bhaji, dhokla) and increasingly common as a gluten-free flour (socca, farinata, gluten-free breads and crackers). Hummus, falafel, and channa are chickpea-based and ubiquitous; papadum is often chickpea flour but is sometimes lentil, so check.
  • Lentil in mixed dishes. Lentil is a common pediatric legume allergen, it can cause anaphylaxis, and its allergens are heat-stable, so thoroughly cooked lentils in dal, soups, and stews can still trigger a reaction. Lentils are often unlabeled in mixed dishes, so confirm with the cook.
  • Concentrated soy proteins. Textured soy or vegetable protein (TVP, TSP) and hydrolyzed soy protein are concentrated soy-protein ingredients, a higher-risk class than trace-protein soy lecithin, common in vegan and meat-substitute products and in processed and deli meats. Watch the generic term “hydrolyzed vegetable protein.”

These are exposure routes to recognize. Whether any of these foods is one your child needs to avoid at all depends on the testing and the allergist conversation in the section above; the hidden-source point is only that if a food is on your child’s avoid list, these are the names it travels under.

This is the part the profile only summarizes, and it is the key to the whole page. Peanut is not one thing to the immune system. It is a handful of different proteins, and which protein your child reacts to is what decides whether a related food is a real risk or a false alarm.

The storage proteins drive the serious reactions, and they do not reassure. Ara h 2 and its close partner Ara h 6 are 2S albumins, which means they are heat-stable and digestion-stable. That stability is exactly why roasting does not defuse them and why a reaction to them can go whole-body rather than staying in the mouth (Sicherer and Sampson 2018; Beyer 2020). Ara h 1 and Ara h 3 are the 7S and 11S storage proteins, also associated with systemic reactions but weaker individual predictors than the 2S albumins. The rule to carry from this: a positive result to a storage protein is a not-reassuring result. When cross-reactivity runs through these proteins, as it does with lupin, it tends to be the real, clinically meaningful kind.

Ara h 8 is the usually-milder pattern. Ara h 8 is a PR-10 protein, structurally similar to a birch-pollen protein. Isolated Ara h 8 positivity usually points to birch-pollen-driven oral allergy syndrome, the itchy-mouth kind that tends not to progress to anaphylaxis (Sicherer and Sampson 2018). This is a fragile protein, broken down by heat and digestion, which is the opposite of the storage proteins and the reason its reactions are usually local and mild.

Ara h 9 is the cofactor-amplified pattern. Ara h 9 is a lipid transfer protein (nsLTP), and isolated Ara h 9 positivity indicates the LTP phenotype, which is Mediterranean-dominant and amplified by cofactors such as exercise. This phenotype is much more common in southern Europe than in the US or northern Europe.

So why is soy a false alarm so often while lupin is not? Because the test can pick up a shared shape between proteins (cross-sensitization) without that shape being one the body actually reacts to (cross-reactivity). With soy and chickpea, the overlap is mostly the harmless kind, which is why the test lights up far more often than a real reaction occurs. With lupin and fenugreek, the overlap runs through the storage-protein machinery that drives real reactions, which is why those two genuinely travel with peanut. The molecule is the difference between a flag and a fight.

What is NOT cross-reactive: the over-avoidance corrections

This is the reassuring core of the page, and also where care matters most. The corrections below are well established, so they are stated plainly. What no page can give you is a blanket “go ahead and feed it” for a specific food; that decision stays with your allergist.

One legume allergy does not mean avoid all legumes. Most people with peanut allergy tolerate other legumes such as peas, soybeans, lentils, and chickpeas, because cross-sensitization on testing is usually not clinically relevant. Only a small minority of peanut-allergic people have a true allergy to other legumes. This is the correction to the instinct that one bean allergy means the whole family of beans is off the table. The instinct is understandable and usually wrong.

There is a real line here worth being honest about. “Your child can safely eat lentils” or “go ahead and add soy back” is a per-food introduction decision, and that stays with your allergist. Read the over-avoidance correction as “the door is probably not as closed as it looks, so ask,” not as “the door is open.”

Tree nuts are a co-occurrence, not a shared protein. Many peanut-allergic children are also allergic to one or more tree nuts, but through separate, independent sensitization, not because peanut and tree nuts share proteins. Peanut is a legume and tree nuts are tree seeds; they are botanically unrelated. Allergists test for both because the two co-occur in the same children, not because one causes the other. The practical version is “test the tree nuts your child has not eaten, and do not assume they are all off the table because peanut is.”

Coconut is usually fine. Coconut is botanically a fruit (a drupe), not a true tree nut, and although US labeling law treats it as a tree nut, most tree-nut-allergic people tolerate coconut and coconut allergy is rare. As with the legumes, confirm with your allergist before introducing it, but the name on the label overstates the botanical risk.

Where studies disagree

There is one genuine tension in the peanut cross-reactivity literature, and it is worth seeing as a disagreement rather than a settled fact, because it explains why two reasonable sources can sound like they contradict each other.

The disagreement is between co-sensitization and clinical cross-reactivity. One set of findings shows that peanut-allergic children very frequently test positive to other legumes: in some cohorts the large majority of children with one legume allergy test positive to lentil, chickpea, pea, and peanut together. Another set of findings shows that when those same children are actually fed the food under supervision, real reactions to the other legumes are much less common than the testing predicted. Both are true. The testing genuinely co-fires across legumes, and the clinical reactions genuinely do not follow at the same rate. The studies are not contradicting each other; they are measuring two different things, sensitization and reaction, and the gap between the two numbers is the whole point of this page.

The practical reading: when a panel comes back positive across several legumes, that is the expected, often-overstated co-sensitization pattern, not a verdict that all of them are dangerous. It is the reason a supervised challenge, not a broad blood panel, is what actually answers “can my child eat this.”

Testing and confirmation

The test that answers the cross-reactivity question is not the broad panel; it is component-resolved diagnostics and, where the answer is still unclear, a supervised challenge.

A standard skin prick or whole-extract blood test tells you only that the immune system has noticed peanut, and a broad legume panel tells you only that it co-fires across legumes, which it usually does. Component testing breaks the result down protein by protein, which is what separates the storage-protein pattern (Ara h 2 and Ara h 6, the serious, cross-reactivity-meaningful kind) from the PR-10 pattern (Ara h 8, the usually-mild kind) and the LTP pattern (Ara h 9). Component platforms in clinical use are ImmunoCAP, the ImmunoCAP ISAC multiplex, and ALEX2, the last of which includes a CCD inhibitor that cuts the carbohydrate-driven false positives common in children who test positive to many things (Beyer 2020; Santos 2023). That CCD-inhibitor detail matters specifically for cross-reactivity, because a lot of the alarming cross-positivity on a broad panel is exactly that carbohydrate-shape false signal.

When component testing and history still disagree, the supervised oral food challenge is the reference standard (Santos 2023). It is the one test that distinguishes a flag from a fight for a specific food, which is why “ask for a challenge” is often the right answer to “is this cross-reaction real for my child.”

How to act on this

Sort each food into one of three buckets, and treat each bucket differently.

  • Avoid and learn the hidden names. Lupin and fenugreek are the genuine peanut cross-reactants. Treat them as foods to avoid unless your allergist clears them, and learn where they hide (lupin flour in baked goods, fenugreek in spice blends).
  • Test, do not assume. Soy, chickpea, pea, lentil, and the other legumes that light up on a panel sit here. A positive test is a reason to talk to your allergist, not a reason to pull the food on your own and not a reason to assume it is safe. The honest default is “ask before you decide.”
  • Probably over-avoided, still ask your allergist. The over-avoidance correction is real: most peanut-allergic children tolerate most other legumes, and most tree-nut-allergic children tolerate coconut. But the decision to introduce or reintroduce any specific food is one to make with your allergist, who knows your child’s component profile and history. The page can tell you the door is probably not as closed as it looks. It cannot tell you to walk through it.

The single most useful action across all three buckets is to ask your allergist to run component testing and to interpret a broad positive panel against it, rather than treating the panel as a list of foods to fear.

Frequently asked questions

Can my peanut-allergic child eat other beans, lentils, and chickpeas?

Most peanut-allergic children tolerate other legumes, because a positive test for them is usually co-sensitization rather than a real allergy, and only a small minority have a true allergy to other legumes. That said, whether to add a specific legume back is a decision to make with your allergist, not from a blood panel alone.

Does peanut allergy mean my child is allergic to tree nuts?

No, not automatically. Peanut is a legume and tree nuts are tree seeds, so they do not share the proteins that cause cross-reactions. Many peanut-allergic children are also tree-nut allergic, but through separate sensitization, which is why allergists test the tree nuts rather than assuming they are all off the table.

What about lupin flour?

Lupin is the one legume to take seriously, because peanut and lupin genuinely cross-react and a peanut-allergic person can develop a true lupin allergy. Read labels for lupin flour, especially on European, imported, and gluten-free baked goods, and do not assume your child tolerates it.

My child tested positive to soy. Do we have to avoid it?

Probably not, but ask your allergist before deciding. Peanut and soy commonly show co-sensitization on testing, but clinically relevant cross-reactivity is rare and most peanut-allergic people tolerate soy. A positive soy result is a reason to confirm with your allergist, not a reason to pull soy on your own.

Is coconut a tree nut my child needs to avoid?

Usually not. Coconut is botanically a fruit, and although US labeling treats it as a tree nut, most tree-nut-allergic people tolerate it and coconut allergy is rare. Confirm with your allergist before introducing it, but the label overstates the botanical risk.

References

The claims on this page draw on the studies and guidance below. As new, credible research meaningfully changes the picture, we update the page to match.

← Peanut allergy