Fragrance allergy and citrus photosensitivity from balms, stated in numbers

Fragrance is among the most frequent contact allergens, and oxidised limonene and linalool are the usual sensitisers. Bergapten is the sun sensitivity problem.

Two quite different things go wrong with scented balms, and they get muddled because both end in a red, sore patch of skin. One is an immune reaction to a fragrance molecule, which needs no sunlight and recurs for life. The other is a chemical burn that only happens when a specific class of citrus oil constituent meets ultraviolet light. This page separates them, gives the rates and limits that are actually published, and says what a buyer can do with a label.

• Declared above 0.001% (10 ppm) in the EU
• 80 named allergens since 2023
• Furocoumarin ceiling 15 ppm
• Oxidised terpenes are the sensitiser
• Natural is not exempt

Allergy, irritation and phototoxicity are three different faults

Getting the category right decides everything you do next, because the three behave differently in time.

An allergic contact reaction is an immune response to a particular molecule. It needs prior sensitisation, so it can appear after two years of using a product happily. Once established it is essentially permanent, it recurs on every exposure regardless of how little is applied, and it typically shows up 12 to 72 hours after contact. That delay is why people so rarely finger the right product.

Irritation needs no immune system. It is dose dependent, it appears within minutes to hours, it fades when you stop, and anyone will get it if the dose is high enough. Peppermint on cracked lips, clove in a mouth ulcer, a heavy hand with menthol: that is irritation, and the triage for the immediate version of it is in balm stings or burns.

Phototoxicity is a photochemical burn. The molecule sits in the skin doing nothing until ultraviolet A arrives, absorbs it, and damages the cells around it. No prior exposure is needed, the first time is as bad as the tenth, and the reaction is usually delayed by 12 to 24 hours after the sun rather than after the application. It heals into brown pigmentation that can take months to fade.

How common fragrance allergy actually is

Two populations get quoted interchangeably and they differ by roughly a factor of five. Patch test clinic figures come from people who already had a rash worth investigating, so they run high. General population estimates from European studies put fragrance contact allergy at roughly 1 to 3 percent of adults, women more often than men, which is still a lot of people in absolute terms.

Three things in that table matter to a buyer. Balsam of Peru scores as high as it does partly because it cross reacts with a wide set of fragrance and spice materials, so a positive there tends to mean broad fragrance avoidance rather than one molecule. The two oxidised terpene preparations are tested air exposed on purpose, which is the tell that condition matters as much as identity. And a mix result names a group of eight or six materials, not a single culprit, so the follow up test with individual constituents is the one that produces an avoidance list you can shop with.

The sensitiser is usually not what you bought

Fresh linalool and fresh limonene are weak sensitisers. Their hydroperoxides are not. Oxygen attacks the terpene at its double bonds, and the hydroperoxides formed are potent contact allergens. Most people with a positive test to these materials are reacting to the degradation products rather than to the molecule as it left the still, which is exactly why the standard patch test preparations are made from deliberately air exposed material.

That converts an ingredient question into a storage question. A bottle of sweet orange or lavender oil that has been opened weekly for two years, half empty, on a warm shelf, is not the same material as the one in the sealed bottle beside it. The same is true of the balm on your desk: a citrus scented tin opened daily through a hot summer accumulates the same chemistry, more slowly, because the oil is diluted and part of the headspace is wax. The underlying mechanism, and the antioxidants that slow it, are set out in rancidity and oxidation, and the domestic controls in storing balms at home.

Two practical consequences follow. Buying a heavily citrus scented product in a large size that will take you three years to finish is worse than buying a small one twice. And a product that has visibly changed, gone sharp or turpentine like in smell, or thin and sticky, has changed chemically as well as sensorially, and is a poor thing to keep applying to sore skin.

What the ingredient list will tell you

In the EU and Northern Ireland, a listed fragrance allergen has to be named in the ingredient list when it exceeds 0.001 percent of a leave on product, which is 10 parts per million, or 0.01 percent of a rinse off one. Almost every balm is leave on. That threshold is tight enough that anything you can smell is normally above it, which is why scented balms end with a tail of names most buyers do not recognise: linalool, limonene, geraniol, citronellol, coumarin, eugenol, citral.

The list itself changed. Regulation (EU) 2023/1545 took it from 26 entries, covering 24 named chemicals plus oakmoss and treemoss extracts, to 80 substances across 81 entries. The additions include ordinary aroma chemicals such as vanillin, menthol and camphor, hive and balsam materials including Myroxylon pereirae and propolis extract, and, for the first time, a set of whole essential oils named in their own right: lavender, peppermint, eucalyptus, clove, rose, geranium, ylang ylang and others. Products not meeting the new rules could be placed on the EU market until 31 July 2026 and may be made available until 31 July 2028, so through 2026 and 2027 you will see both label styles on the same shelf, the shorter one on older stock.

Great Britain has not adopted the amendment, so a GB label may legitimately carry only the original set for a product whose EU label lists twice as many. That is a labelling difference, not a formulation one. The full mechanics, thresholds and arithmetic are in fragrance allergens, and how to read the rest of the list is in reading a balm label.

Fragrance free, unscented, and why natural changes nothing

These are not synonyms. Fragrance free, used properly, means nothing aromatic has been added. Unscented means the finished product does not smell of much, which is sometimes achieved by adding a masking fragrance to cover the raw material odour, and that fragrance declares like any other. A product can therefore be honestly labelled unscented and still contain the exact molecule you react to.

The larger confusion is that natural fragrance, essential oils and botanical extracts sit outside the rules. They do not. The EU labelling duty attaches to the molecule, not to its origin: linalool from a small distiller's lavender declares at the same 0.001 percent as linalool from a drum of synthetic isolate, and the immune system draws no distinction either. If anything the natural route delivers a wider spread of candidate allergens, because an essential oil is a mixture of dozens to hundreds of constituents rather than a defined material. The trade offs between the two routes are set out in essential oils versus fragrance oils, and the marketing vocabulary is unpicked in chemical free and clean beauty claims.

Where a buyer actually gains is by shortening the list. On lips this matters most, because the vermilion has a thin stratum corneum and no meaningful barrier, which is why flavour, sweeteners and strong actives cause so much lip trouble. Flavour and sweeteners encourage licking on top of any irritant effect, and menthol, camphor and similar actives feel like they are working while adding irritation. If you want a genuinely bland product, a wax and oil base with nothing aromatic in it is the target, and what that looks like is set out in fragrance free balm.

Furocoumarins: the sun problem, and the number that governs it

Furocoumarins, chiefly bergapten (5-methoxypsoralen), are heavy, non volatile molecules found in citrus peel. Because they are non volatile they do not carry over into a steam distilled oil, and they do come through in a cold pressed or expressed one. That single fact explains most of the confusion in this area: two oils with the same botanical name behave completely differently depending on how they were made.

IFRA restricts total furocoumarins to 0.0015 percent of a finished product applied to skin exposed to sunshine, which is about 15 parts per million. The EU goes further in one narrow case: Annex II of Regulation 1223/2009 requires furocoumarins to be below 1 part per million in sun protection and bronzing products specifically, while permitting the normal content of natural essences elsewhere. Both figures exist and they are not in conflict, they cover different products.

The arithmetic ties the two numbers together neatly. Expressed bergamot oil typically carries bergapten at roughly 0.2 to 0.4 percent of the oil, that is 2,000 to 4,000 parts per million. Used at the 0.40 percent cap, that puts 8 to 16 ppm of bergapten into the finished balm, which lands right on the 15 ppm total furocoumarin ceiling. The material limit and the product limit are two views of the same constraint, which is a good sign that the numbers are real rather than borrowed.

Bergapten free grades, sold as bergamot FCF (furocoumarin free), have been through an extra process to strip the furocoumarins down to that ceiling or below. They are not covered by the 0.40 percent entry because there is nothing left to restrict. They smell slightly different from the expressed oil, which is why they are not a universal substitute in perfumery, and none of that has any bearing on the allergen question: bergamot FCF still carries limonene and linalool, and still declares them. The maker facing version of all this, including how the limits combine, is in essential oils in balms, and anyone blending at home should run levels through the essential oil dilution calculator rather than guessing.

Reacting now: what to do this week

This is an elimination, not a treatment, and it works because it removes variables rather than adding ingredients.

1. Stop everything aromatic on the affected area. Balm, cleanser, hand cream, and on lips also toothpaste flavouring and mouthwash. Substitute one bland product with the shortest list you can find.
2. Keep the packaging. Do not throw out the suspect product. The ingredient list, the batch code and the container are what a dermatologist needs, and the allergen tail is often the only route to a named molecule.
3. Write down the timing. Hours from application to symptoms, and whether the day involved sun. Minutes points at irritation, one to three days at allergy, and a day after sunshine at phototoxicity.
4. Reintroduce one product at a time. Leave at least four days between additions, because a delayed reaction to product one will otherwise be blamed on product two.
5. Test before you commit to a new product. Repeat application on a small patch of forearm skin for several days lowers your risk without proving safety; the method and its limits are in how to patch test a balm.
6. Get a proper patch test if it recurs. A baseline series plus the products you brought in identifies the molecule, which is the only thing that lets you shop confidently for the next twenty years.

Two patterns are worth recognising because they mimic each other. Persistent lip dryness that worsens the more balm you apply is far more often allergic contact cheilitis than any dependence on the product, which is the argument made at length in does lip balm dry your lips. And a reaction to a beeswax or propolis rich balm may be to the hive material rather than to the fragrance at all, which is a separate avoidance list; see beeswax and propolis allergy.

Who should simply skip scented balms

For most people, a scented balm used at ordinary rates is uneventful, and treating every fragrance molecule as a hazard is not supported by the numbers. Four groups have a better than average reason to default to unscented anyway: anyone with a confirmed fragrance or balsam allergy, anyone with eczema or an otherwise compromised barrier, where a damaged stratum corneum lets more of everything through, anyone using a product on already broken or sore skin, and children, whose exposure per kilogram of body weight is higher and whose products are covered in lip balm for children.

Beyond those, the useful habit is not avoidance but attention. Buy small, finish it, notice when the smell changes, and keep anything citrus off skin that is about to spend a day in the sun.

Where the evidence stops

The decision rule is short. If a reaction takes a day or more to appear and comes back every time, treat it as a possible allergy, keep the packaging and get patch tested. If it appears within minutes and settles when you stop, treat it as irritation and shorten the ingredient list. If it appears the day after sun and leaves brown marks, suspect an expressed citrus oil and keep that product for evenings.

What the published material does not support is more interesting than what it does. There is no evidence that the length of an allergen tail predicts a reaction in any individual: the declaration is a composition statement, set at a level that keeps labels usable rather than at a no effect level, and a sensitised person can react below 10 parts per million. The list is not exhaustive either, so an unlisted sensitiser can be present with no trace on the label at all. Nothing on the label tells you the oxidation state of the oil, which is often the variable that decides whether a material provokes anything. Patch test positivity rates describe clinic populations and cannot be converted into personal odds. And no ingredient list, read carefully or not, substitutes for a patch test when something has actually gone wrong.

Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.