Balms and eczema-prone skin: what the evidence supports
An evidence review of emollients and balms on eczema-prone skin: what regular emollient use is established to do, what a balm adds, and where the clinical line sits.
This page reviews what is known about emollients and anhydrous balms on eczema-prone skin. It is a summary of published evidence written by the Trebalm editorial desk, not medical advice, and it does not recommend any product for any condition. Eczema is a medical diagnosis and its management belongs with a clinician.
Regular emollient use is well established as part of routine care for dry, eczema-prone skin, mainly because it reduces water loss and softens the stratum corneum. An anhydrous balm can contribute occlusion and emolliency. It cannot rebuild barrier lipids, and it is not a treatment for a flare.
What emollients are established to do
The evidence base for emollients in atopic skin is unusual in dermatology in that the broad conclusion is stable while the detail is messy. Clinical guidance across several countries recommends leave-on emollients used regularly and generously as the maintenance foundation of care. The Cochrane review of emollients and moisturisers for eczema, covering dozens of trials, concluded that moisturisers as a class produce better outcomes than no moisturiser and that they extend the time between flares, while also finding that the trials were mostly short, heterogeneous and unable to establish that any one product beats another.
That last finding is the one to hold onto. The established effect belongs to the category, not to a formulation. Anyone claiming that a specific fat outperforms others on atopic skin is going beyond what has been shown.
The picture on prevention is different and worth stating plainly, because it is often misreported. A large randomised trial of daily emollient in infancy did not find that routine emollient use prevented eczema developing in high-risk infants. Emollients as maintenance for existing dry skin, and emollients as prophylaxis for babies, are separate questions with different answers.
The barrier lipid picture
Skin that is prone to eczema tends to differ in its stratum corneum lipids: shorter ceramide chain lengths, altered ceramide subclass ratios, and a less ordered lipid organisation. In many people there is also a genetic contribution affecting filaggrin, a protein involved in producing the skin's own natural moisturising factor. The consequence is a stratum corneum that loses water faster and admits irritants and allergens more readily.
This matters for how you read ingredient claims. The barrier is built from ceramides, cholesterol and free fatty acids in roughly equimolar proportions, arranged in ordered lamellae. It contains almost no triglycerides. Every plant butter and animal fat in a balm is a triglyceride. Spreading triglycerides on the surface does not supply the missing lipid classes or repair the lamellar structure, whatever a marketing line implies. What it does is sit on top and slow evaporation, which is a genuine and useful effect. The distinction is explained further under moisturisers and the skin barrier.
What an anhydrous balm can and cannot contribute
A balm is a water-free blend of oils, butters and wax. That definition, set out under what is a balm, determines what it can do.
It can occlude, strongly. It can soften and smooth by filling the spaces between corneocytes, which is the emollient action. It can stay put on hands, feet, lips and small patches for hours, which a lotion cannot. It can be made with very few ingredients, which reduces the number of possible sensitisers.
It cannot add water, because it contains none, so it does nothing on its own for a dehydrated stratum corneum unless there is water underneath to trap. It cannot carry humectants such as glycerin or urea in any useful way, because those are water-soluble and do not dissolve in an oil phase. It cannot deliver the ceramide-cholesterol-fatty-acid combination that some prescription and clinical-grade products attempt. And it cannot be spread economically over large areas, which is why whole-body emollient regimes use creams and ointments rather than balms.
Paraffin-containing emollients, and paraffin-free ones too, transfer to fabric. Dressings, clothing and bedding soaked with emollient residue burn faster and hotter once ignited, and regulators have issued warnings after fatal fires. Anyone using large quantities of a greasy product should keep away from naked flames and cigarettes, and wash bedding and clothing at a high temperature.
Ingredient by ingredient: what the evidence supports
| Ingredient | What the evidence supports | What it does not support |
|---|---|---|
| Petrolatum | The most complete occlusion measured of any common material, very low sensitisation rate, long clinical record | Any active repair effect, or the claim that it suffocates skin |
| Mineral oil and paraffins | Effective occlusion and emolliency, well characterised, low reactivity | Claims of barrier restoration |
| Shea butter | Good emolliency and moderate occlusion as a fat | Any demonstrated advantage over other emollients on atopic skin |
| Colloidal oatmeal | Recognised as a skin protectant in US regulation, with supporting trial data on dry and itchy skin | Being usable in an anhydrous balm, where it sits as an undissolved powder |
| Coconut oil | Small trials report improvements in measured skin hydration and water loss compared with mineral oil | Being suitable for everyone, given its lauric acid content and reports of irritation on some skin |
| Beeswax | Effective structurant, long history of cosmetic use | Freedom from allergy. Contact cheilitis and dermatitis from beeswax are documented |
| Lanolin | Very high occlusion, close to skin surface lipid in structure | A clean allergy record. It is a recognised sensitiser, especially on damaged skin |
| Essential oils | Nothing relevant to barrier function | Any risk-benefit case on compromised skin. Several are established contact allergens |
| Ceramides in a balm | Interest in clinical formulations at controlled ratios | Meaningful effect from trace amounts added to an oil blend |
| Honey, aloe, hydrosols | Nothing that survives being added to an anhydrous product | Safety without a preservative. Adding water to a balm makes it perishable |
Two of those rows deserve emphasis. Petrolatum is the best-evidenced occlusive there is and the least likely to sensitise, which is inconvenient for the assumption that plant-derived is gentler. And lanolin, which is technically excellent, is a documented sensitiser precisely in the population most likely to use it. Meanwhile shea butter's reputation in this area rests on plausibility and popularity rather than on comparative data, and the small trials involving coconut oil report favourable numbers on hydration and water loss but are too small and too short to change anyone's practice. Neither of those facts is a recommendation. Both are reasons to read the ingredient list rather than the front of the pack.
Why fragrance and essential oils are a poor bet here
Fragrance materials, whether synthetic or extracted from plants, are the most common cause of cosmetic contact allergy. Essential oils are not exempt: they are mixtures of dozens of volatile compounds, and the ones that must be declared on EU labels (limonene, linalool, geraniol, citral, eugenol and others) are declared precisely because they are recognised allergens. Several of them oxidise on storage into hydroperoxides that sensitise more strongly than the parent compound.
Compromised skin changes the calculation twice over. It absorbs more, and it is a more efficient route to sensitisation, because the immune machinery that drives allergic contact dermatitis is more accessible through a disrupted barrier. Against that, the benefit on offer is that the product smells nice. That is a poor trade on skin that is already inflamed, and it is why fragrance-free is the default in clinical guidance.
Products in open jars are dipped into with fingers. On skin that may be broken or colonised, that introduces organisms into the jar and returns them to the skin. Use a clean spatula, keep a jar to one person, and do not use a product that smells, looks or feels different from when it was new.
Food-derived proteins on broken skin
This one should be reported neutrally, because it is unresolved. Observational work published in 2003 found an association between the use of skin preparations containing peanut oil and later peanut allergy in children, and the wider hypothesis that early cutaneous exposure to food proteins through a disrupted barrier can promote sensitisation, while oral exposure promotes tolerance, has attracted substantial research since. It is a plausible and actively studied mechanism. It is not a demonstrated causal chain for cosmetic ingredients in general.
The practical reading: highly refined oils contain very little protein, and cold-pressed or unrefined materials contain more. Nut-derived oils, oat, wheat germ and milk-derived ingredients are the ones people ask about. If a food allergy exists in the household, or a child has significant eczema, this is a question for a clinician who knows the case, not one to settle from an ingredient list.
Where the line is
Weeping, crusting, yellow crusting, rapidly spreading redness, pain, fever or a sudden severe flare can indicate infection and need medical assessment the same day. Occluding an infected area with a heavy balm is not a neutral act. Balms are cosmetic products. They do not treat, cure or prevent disease, and anything sold as if it does is making a drug claim.
Within that line there is still a sensible place for a simple balm: on hands that are washed constantly, on heels, on lips, and on small stubborn patches where a cream rubs off too fast. Short ingredient lists, no fragrance, applied over damp skin, in quantities that suit the area rather than the tub, as discussed under how much balm to apply. The same logic applies to occluded areas in infants, where the reasoning behind a nappy area balm is barrier protection rather than treatment.
What to do with all this
If your skin is eczema-prone, the useful moves are unglamorous: keep to whatever routine your clinician has set, prefer short ingredient lists, avoid fragrance, introduce one new product at a time so you can attribute a reaction, and treat any worsening as a signal to stop rather than persist. Judge a balm on whether your skin is more comfortable after two or three weeks, and be suspicious of any product, natural or otherwise, that promises more than a reduction in water loss. That reduction is real, it is worth having, and it is the whole of what a balm offers.
Frequently asked questions
Is shea butter good for eczema?
Shea is a capable emollient and occlusive, and it is used widely in products aimed at dry skin. There is no substantial trial evidence that shea specifically outperforms other emollients on atopic skin. It is a reasonable base material if it suits your skin, and it is not a treatment. Reactions to shea are uncommon but reported.
Can a balm replace a prescribed emollient?
No, and the question belongs with whoever prescribed it. Prescribed emollients are chosen for a reason, are usually well characterised, and in some cases carry regulated status. A cosmetic balm is not an equivalent product and is not assessed for that use. Use one alongside routine care only if your clinician agrees.
Why avoid essential oils on eczema-prone skin?
Essential oils are complex mixtures containing known contact allergens such as limonene, linalool and geraniol, which oxidise on storage into stronger sensitisers. Compromised skin absorbs more and is more easily sensitised. There is no established benefit to offset that risk, so the rational choice on broken or inflamed skin is fragrance-free.
Does an anhydrous balm hydrate skin?
Not directly. A balm contains no water, so it adds none. It works by slowing evaporation from the skin surface so that water arriving from deeper tissue accumulates in the stratum corneum. That is why balms are often applied over damp skin or after a water-containing product rather than onto dry skin alone.
Are food-derived ingredients on broken skin a problem?
It is a genuine open question rather than a settled danger. Some observational and mechanistic work has linked early skin exposure to food proteins through a disrupted barrier with later food sensitisation. Refined oils and butters contain little protein, cold-pressed and unrefined ones more. If a food allergy is in the picture, raise it with a clinician.
How much balm should I use?
Clinical guidance for emollients generally emphasises frequent and generous use rather than thin application, with quantities measured in hundreds of grams a week for whole-body use in some cases. A balm is more concentrated and more occlusive, so it is usually used in smaller amounts on specific areas rather than as a whole-body emollient.
Can a balm make eczema worse?
It can. Contact allergy to an ingredient, irritation from fragrance, occlusion over an infected area, and contamination of a jar dipped into with wet fingers are all documented ways a topical product causes trouble. Worsening after starting something new is a reason to stop it and to seek advice, not to apply more.
Sources and further reading
- National Institute for Health and Care Excellence, Atopic eczema in under 12s: diagnosis and management (CG57), NICE.
- van Zuuren EJ, Fedorowicz Z, Christensen R, Lavrijsen A, Arents BWM, Emollients and moisturisers for eczema, Cochrane Database of Systematic Reviews, 2017.
- Chalmers JR et al., Daily emollient during infancy for prevention of eczema: the BEEP randomised controlled trial, The Lancet, 2020.
- Medicines and Healthcare products Regulatory Agency, Emollients: new information about risk of severe and fatal burns with paraffin-containing and paraffin-free emollients, Drug Safety Update, 2018.
- van Smeden J and Bouwstra JA, Stratum corneum lipids: their role for the skin barrier function in healthy subjects and patients with atopic dermatitis, Current Problems in Dermatology, 2016.
- Lack G, Fox D, Northstone K, Golding J, Factors associated with the development of peanut allergy in childhood, New England Journal of Medicine, 2003.
Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.