Fish parvalbumin cross-reactivity

By Symon, Every Label · June 30, 2026

If you are allergic to one fish, the single most useful thing to understand is the protein behind it. Almost all fish-to-fish cross-reactivity runs through one molecule called parvalbumin, a small muscle protein found across finned fish. Parvalbumin is heat-stable and digestion-stable, so cooking does not defuse it and the stomach does not break it down. Because the parvalbumin in cod, salmon, and many other fish looks alike to the immune system, an allergy to one fish often means an allergy to several. That is why fish allergy tends to travel broadly, and it is the reason this page is built around the protein rather than around any one fish.

This is the mechanism hub. It explains the protein first, then groups the fish by how strongly they actually cross-react, then sends you to each fish’s own page. One common fear gets corrected at the end: a fish allergy is not a shellfish allergy, because fish and shellfish run on completely different proteins. Where a claim is a verified cross-reactivity fact, it is drawn from the project’s cross-reactivity floor. None of it replaces your allergist.

The molecular why: one protein, shared across finned fish

Parvalbumin is the spine of this whole story, so it goes first.

Parvalbumin is a small, calcium-binding muscle protein that finned fish carry in abundance, and it is the major fish allergen. The cod version is the prototype, the protein that fish allergy testing is anchored to, but every finned fish carries its own parvalbumin, and across species these proteins are similar enough that an antibody trained on one can bind another. That shared recognition is cross-reactivity, and it is why “fish allergy” behaves less like an allergy to one fish and more like an allergy to a protein that turns up in many fish at once. Cod, salmon, and tuna all cross-react through it.

Two properties of parvalbumin drive everything below:

It is heat-stable and digestion-resistant. Cooking, frying, and baking do not break it down, and it survives the stomach intact. So the cross-reaction does not weaken when the fish is cooked, and the protein can even become airborne: the steam from cooking fish carries enough parvalbumin to trigger a reaction in some people without a bite ever being taken.
How much parvalbumin a fish carries tracks how strongly it cross-reacts. White-flesh fish such as cod carry a lot of parvalbumin and cross-react broadly and predictably. Some dark-flesh and large open-water fish, such as tuna and swordfish, carry less of it, and a subset of fish-allergic people turn out to react to them less, or not at all. That variation is real, but it is a reason to test specific fish under an allergist, not a reason to assume any fish is safe. More on this below, because it is the most common way people get hurt on this topic.

The practical rule that falls out of this: a fish allergy is broad by default, and the protein is what makes it broad. The rest of the page is that map.

The member fish, grouped by how strongly they cross-react

This is the spine of the cluster: the fish, sorted by how much an allergy to one tells you about the others. The grouping follows the verified parvalbumin cross-reactivity, not a guess. Strongest-predicts-first.

Broadly cross-reactive: most finned fish, treat as a group

Cod, salmon, and most other finned fish carry parvalbumin that is similar from species to species, and clinical cross-reactivity among them is high. In practice, an allergy to one of these fish means the others should be treated as high-risk until an allergist says otherwise. This is the default shape of fish allergy: broad, and predictable enough that most allergists manage finned fish as a single avoidance group rather than testing them one at a time.

What to do with that: if you react to one fish, treat the others as fish you have not cleared yet, and confirm tolerance of any specific fish only under supervision, never with a home trial. This is the part of the cluster where one allergy genuinely predicts the rest.

The lower-parvalbumin fish: a reason to test, not a green light

Some fish carry less parvalbumin than white-flesh fish like cod. Tuna and swordfish are the usual examples, and there is a real, repeatedly observed pattern of fish-allergic people who react to cod but tolerate tuna. It is tempting to read that as “tuna is the safe fish,” and that is exactly the mistake to avoid. The lower-parvalbumin fish are still members of the cluster, they still cross-react in many people, and true tuna allergy, including anaphylaxis, is documented. The same parvalbumin record that explains the broad cross-reactivity names tuna among the cross-reacting fish.

So the honest version of this is narrow and it matters: whether you personally tolerate a lower-parvalbumin fish is a question to answer with your allergist, through testing and, where appropriate, a supervised challenge. It is a reason to ask “can we test whether I tolerate tuna,” not a reason to put tuna on the plate at home. This page does not clear tuna, swordfish, or any other fish for you, in any form, including canned. If you have been told you tolerate a specific fish, that came from your allergist and applies to you, not to fish allergy in general.

Where a fish reaction is not a parvalbumin allergy at all

Two situations get mistaken for fish allergy and are worth naming, because both send people down the wrong path.

A reaction only to raw or lightly cooked fish. Some people react to raw or barely cooked fish (sushi, lox, ceviche) but tolerate the same fish well cooked. This points to minor, heat-sensitive fish proteins rather than to parvalbumin, which is the opposite of the heat-stable main story. It is still a fish reaction to take to an allergist, not a clearance to eat raw fish, but it is a different mechanism and worth flagging so the testing is read correctly.
Scombroid (histamine) poisoning. A flushing, headache, and rash reaction soon after eating tuna, mackerel, or other dark-flesh fish that was not kept cold enough is usually scombroid, a toxic reaction to histamine that built up in the spoiled fish. It is not a fish allergy, it is not caused by parvalbumin, and it can affect anyone at the table regardless of allergy. It matters here because it is routinely mistaken for a new fish allergy. If this has happened to you, your allergist can help tell the two apart.

What is NOT cross-reactive: fish is not shellfish

The cluster is broad, but it has a hard and frequently over-avoided edge, and it is the most important correction this page makes.

Fish is not shellfish. Finned fish (cod, salmon, tuna, and the rest) run on parvalbumin. Shellfish (shrimp, crab, lobster, and the molluscs) run on a completely different protein called tropomyosin. Because the proteins differ, a finned-fish allergy does not mean a shellfish allergy, and clinical cross-reactivity between the two is low. The reverse holds too: a shellfish allergy does not mean a fish allergy. The one real-world caveat is cross-contamination, fish and shellfish are prepared and fried together constantly, so shared fryers, grills, and seafood counters are still a risk even though the allergies are separate. Confirm with your allergist before introducing the other type.

Shellfish has its own cross-reactivity cluster, driven by tropomyosin, and it reaches well beyond food to dust mites and insects. That is a different protein and a different story, covered on the tropomyosin (shellfish) syndrome spoke rather than here. If “is my fish allergy also a shellfish allergy” is your question, that page is the other half of the answer.

This correction clears one specific fear, the fish-versus-shellfish one. It does not clear any fish inside the parvalbumin cluster; the lower-parvalbumin fish in particular stay in the test-do-not-assume group above.

Where studies disagree

One genuinely unsettled area is worth seeing as a disagreement rather than a settled fact.

Whether a positive test means a real reaction, especially for the lower-parvalbumin fish. A blood or skin test can light up across many fish at once because the parvalbumin is shared, but a positive test is not the same as a real-world reaction. The gap is widest exactly where it matters most: for fish like tuna that carry less parvalbumin, where the test may be positive but a meaningful share of patients tolerate the fish, and for the reverse, where a clean test is not a guarantee. The serology and the clinical reaction are measuring two different things, and that gap is the whole reason the lower-parvalbumin fish are “test, do not assume” rather than “safe because the number is lower.” A supervised challenge, not a broad fish panel, is what actually answers “can I eat this fish.”

A serology confounder worth knowing about: Anisakis. Anisakis is a parasite found in some fish, and a positive Anisakis test is sometimes turned up when fish allergy is being investigated. Anisakis carries a tropomyosin similar to the tropomyosin in shrimp, so a positive Anisakis result can reflect crustacean cross-reactivity rather than a true parasite exposure. It is named here only as a reason a fish-allergy workup can throw a confusing extra positive, to be read against your actual history with your allergist, not as a statement about whether any particular fish is safe to eat.

Testing and confirmation

Cross-reactivity questions in this cluster are answered the same way the core diagnosis is: component-resolved testing, and where needed a supervised challenge.

A standard whole-extract blood test or skin prick tells you the immune system has noticed fish, but because parvalbumin is shared, it tends to flag many fish at once and does not, on its own, tell you which fish you can actually eat. Component testing breaks the result down to the actual proteins, so a result driven by parvalbumin (the pan-allergen behind the whole cluster) can be told apart from one driven by a minor, fish-specific protein. That distinction is part of what decides whether a positive result for a lower-parvalbumin fish like tuna is worth challenging.

Where component testing and history still disagree, the supervised oral food challenge is the test that settles whether a specific fish can actually be eaten. It is done with your allergist, never at home, and it is the only thing that turns “related on a test” into “safe to eat.” This is exactly the route for the lower-parvalbumin fish: a reason to test, under supervision, rather than a reason to assume.

The members (each fish’s own page):

Salmon allergy: the full profile
Cod allergy and tuna allergy: the rest of the cluster

The neighboring hubs:

Tropomyosin and shellfish: the other seafood cross-reactivity cluster (why fish is not shellfish, the shared-protein reach to dust mites and insects)
Finned fish: the family page (the category, the fish-species label and additive detail)

Frequently asked questions

If I am allergic to one fish, am I allergic to all fish?

Often, yes, which is why fish allergy is treated as broad by default. Finned fish share a heat-stable muscle protein, parvalbumin, and cod, salmon, and most other fish cross-react through it, so an allergy to one usually means the others should be treated as high-risk until an allergist says otherwise. Most allergists manage finned fish as a single avoidance group and confirm tolerance of any specific fish only under supervision.

I react to cod but I have heard I might tolerate tuna. Is tuna safe for me?

This is a question to answer with your allergist, not an assumption to make at home. Some fish, like tuna and swordfish, carry less parvalbumin than cod, and a subset of fish-allergic people do tolerate them, but tuna still cross-reacts in many people and true tuna allergy is documented. Whether you personally tolerate tuna, in any form including canned, is something to test under supervision, not to try on your own. This page does not clear tuna for you.

Does a fish allergy mean I am also allergic to shellfish?

No. Finned fish run on parvalbumin and shellfish run on a different protein, tropomyosin, so a fish allergy does not mean a shellfish allergy, and clinical cross-reactivity between them is low. The same is true the other way around. Cross-contamination where fish and shellfish are prepared together is still a real risk, so stay alert at seafood counters and shared fryers, and confirm with your allergist before introducing the other type.

Can the steam from cooking fish cause a reaction?

It can for some people. Parvalbumin is heat-stable and can become airborne in cooking steam, so a person with fish allergy can react to the vapor from fish being cooked nearby, without eating any. How much this affects you is worth raising with your allergist, especially before situations like a seafood restaurant or a shared kitchen.

I got flushed and had a headache after eating tuna. Is that a fish allergy?

Possibly not. A flushing, headache, and rash reaction soon after eating tuna, mackerel, or other dark-flesh fish that was not kept cold enough is often scombroid, a histamine reaction to spoiled fish rather than a fish allergy, and it can affect anyone regardless of allergy. It matters because it is easily mistaken for a new allergy. Your allergist can help tell scombroid and a true fish allergy apart.

References and medical review

This page is pending independent medical review; the note at the top of the page applies until a reviewer is assigned. The verified cross-reactivity and reassurance claims resolve to the project’s conservative cross-reactivity floor, each carrying its own tier-1 source there: the parvalbumin cluster (cod, salmon, and tuna cross-react through the major fish allergen, and the named-tuna caution rests on the same record), the cleared fish-is-not-shellfish correction (parvalbumin versus tropomyosin, clinical cross-reactivity low, cross-contamination caveat retained), and the Anisakis serology confounder (a positive Anisakis test can reflect crustacean tropomyosin cross-reactivity, named only as a workup confounder). The syndrome-level clinical framing (the white-flesh versus dark-flesh parvalbumin shape, the cooking-steam and raw-versus-cooked notes) draws on consolidated finned-fish research still pending final review and is rendered qualitatively, with no abundance figure or tolerance rate stated.