Chemicals in Deodorant: What You're Applying Daily

You'll buy roughly thirty of these in your adult life and rub their contents into the same patch of underarm skin every morning — skin you've often just shaved or sweated through, sitting directly above the lymphatic drainage of the breast. Across an adult lifetime that adds up to 18,000 deodorant or antiperspirant applications across an adult lifetime — once daily over fifty years applications. Six chemical families typically share that patch: aluminium salts, parabens, phthalates (under the word 'fragrance'), triclosan (where it's still permitted), propylene glycol, and a fragrance mixture that can lawfully hide up to 3,691 individual ingredients on the IFRA Transparency List that 'fragrance' or 'parfum' can lawfully obscure on a cosmetic label disclosed-as-one ingredients.

The strongest-evidence ingredient on that list — parabens, found in 99.1% of Americans tested by the CDCCenters for Disease Control and Prevention — sits next to the weakest-evidence one — aluminium chlorohydrate, where decades of epidemiology have produced one positive case-control with a confidence interval wide enough to drive a regulator's car through. This guide does both honestly: doesn't overstate aluminium, doesn't understate parabens, and points at the swaps where the alternative costs the same as the conventional product. Part of our endocrine disruptors guide. Six chemicals. One patch of skin. Eighteen thousand times.

A typical drugstore antiperspirant or deodorant stacks six chemical families on the same 8 cm² of underarm skin — roughly the area of a credit card — once or twice a day, on skin that is often abraded by shaving and warmed by a recent shower. An aluminium salt blocks sweat. Parabens preserve the formulation. Phthalates carry the fragrance. Triclosan kills the odour-causing bacteria (where it's still permitted). Propylene glycol humects and quietly enhances penetration. A fragrance mixture rounds out the dose with dozens of additional undisclosed substances. Every line on the table below is something that has crossed a regulator's desk in the last fifteen years.

Each row hides a different argument. Aluminium has the noisiest cancer hypothesis and the thinnest epidemiology. Parabens have the cleanest mechanism and the noisiest 99% biomonitoring number. The phthalate carrier is undisclosed by design. Triclosan is the chemical that survived a four-decade efficacy gap. Propylene glycol does its job but quietly increases how much of everything else gets through your skin. And the word 'fragrance' is the one place in cosmetic law where the ingredient list is allowed to lie. Six different ways to hide a chemical. One product.

An antiperspirant stops you sweating because it contains an aluminium salt — usually aluminium chlorohydrateAn aluminium salt formed from aluminium chloride and aluminium hydroxide. The most common antiperspirant active ingredient — typically 15–25% by weight in the product. (ACH) at 15–25% by weight, or an aluminium-zirconium variant at up to 20%. The salt dissolves in sweat, hydrolyses inside the duct, and forms a temporary plug. US OTC drug ≤25% 2003 EU Cosmetic permitted 2026 UK Cosmetic permitted 2026 The US FDAFood and Drug Administration classifies antiperspirants as over-the-counter drugs because blocking sweat is a physiological action. The EU and UK classify them as cosmetics. The chemistry doesn't care which paperwork applies.

The cancer hypothesis is older than the regulatory paperwork. The argument: ACH is applied directly to skin in immediate contact with the lymphatic drainage of the breast, repeatedly, over decades, often after shaving has compromised the skin barrier — so even a small absorption fraction would compound into a measurable internal dose. The recent anatomical literature has moved away from the older 'axillary tail of Spence' framing of the underarm as breast tissue extending into the armpit (Standring et al. 2022) — but the lymphatic continuity is preserved through the foramen of LangerAn opening in the deep fascia through which lymphatic vessels pass between the breast and the axilla. Confirms that lymphatic drainage from the breast routes directly through the underarm region.. The skin under your antiperspirant drains into the same nodes the breast does.

How much aluminium actually gets through? Three studies, three different answers. Flarend's 2001 pilot used radioactive ²⁶Al on two volunteers (one man, one woman) and reported 0.012% absorbed (Flarend et al. 2001). Pineau's 2012 Franz-cell experiment found stripped (shaved) skin took up roughly six times more aluminium than intact skin from the same stick formulation (Pineau et al. 2012). De Ligt's larger industry-funded follow-up (Cosmetics Europe paid for it; n=12 women) found 0.0094% mean absorption, range 0.002–0.06% (de Ligt et al. 2018) — three orders of magnitude between the cleanest and dirtiest skin in a 12-person study. The fraction is small in any single application. The denominator is 18,000.

Then there is the breast-cancer epidemiology, which is the fight everyone wants to have. Here is the cleanest summary the evidence allows.

Linhart enrolled 209 cases / 209 controls hospital case-control study, 1:1 age-matched within 3.5 years from an Austrian breast-clinic case-control, frequency-matched on age. The full-cohort odds ratios were small and crossed null. The 3.88 came from one subgroup: women applying underarm cosmetics 'several times per day' BEFORE age 30. Twenty-seven women drove that estimate. The 95% confidence interval ran from 1.03 to 14.66 — barely clearing the line of statistical significance, with an upper bound that's almost worthless. The breast-tissue aluminium difference (5.8 vs 3.8 nmol/g, p=0.0014) is more robust than the OR — it confirms the chemical reaches the tissue, which everyone already accepted. The strongest single positive signal in the literature is also the wobbliest.

Pool the case-control evidence — Linhart, Mirick, McGrath, and four others — and the picture flattens. A 2024 meta-analysis of seven studies returned a pooled odds ratio of 0.96 pooled OR for antiperspirant use and breast cancer across 7 case-control studies (95% CI 0.78–1.17), heterogeneity I²=60% (Trinh et al. 2024). Three of the seven studies were rated high quality; six relied on participant self-report. McGrath 2003, the often-cited 'earlier age of diagnosis' paper, surveyed 437 women all of whom already had breast cancer with no healthy comparison group (McGrath 2003) — a study design that cannot tell you whether antiperspirants cause cancer, only whether women who got cancer used them more often than each other. Mirick's 2002 J Natl Cancer Inst case-control of 813 cases and 793 controls returned a flat null (antiperspirant ever-use OR 0.9, P=0.23) (Mirick et al. 2002).

The mechanistic picture is more interesting than the epidemiology. Darbre's 2006 metalloestrogen review reframed aluminium as one of a class of inorganic chemicals that bind estrogen receptors at low concentrations (Darbre 2006). Gorgogietas's 2018 study in MCF-7 breast cancer cells showed ACH at 10⁻⁶ to 10⁻⁴ M increased estrogen-receptor-α protein two- to threefold over four days, activated estrogen-response-element-driven reporter genes, and induced p53, cyclin D1, and c-fos — all effects that the antagonist ICI 182780 abolished, indicating receptor-mediated activity (Gorgogietas et al. 2018). Sappino 2012 showed aluminium chloride promotes anchorage-independent growth in human mammary epithelial cells (Sappino et al. 2012); Mandriota 2016 extended the same group's work to in vivo tumorigenesis in NMuMG cells (Mandriota et al. 2016). Cells respond to ACH the way cells respond to a weak estrogen. The receptor work is real. The cancer epidemiology is null. Mechanism is not destiny.

The honest position: aluminium chlorohydrate has biological plausibility (it reaches breast tissue, it acts on estrogen receptors, the dose accumulates) and weak conclusive epidemiology (one positive case-control with a wide CI, six null studies). The EFSAEuropean Food Safety Authority 2008 tolerable weekly intake of 1 mg Al/kg bw/week (EFSA 2008) is set for dietary aluminium, not cosmetic — and is already exceeded by significant fractions of European children and adults from food alone, before deodorant adds anything. The industry-funded Willhite review concluded 'no clear evidence' the cosmetic exposure raises breast cancer risk (Willhite et al. 2014); the funding came in part from the International Aluminium Institute, which is the kind of disclosure worth saying aloud rather than burying. The decision the reader has to make is whether 18,000 applications of a chemical that demonstrably reaches breast tissue is worth taking on faith from a literature that has yet to look at the right exposure window. Watch

Parabens are the family of preservatives that almost every personal-care product on your shelf uses to stop bacteria and fungi colonising the bottle: methylparaben, propylparaben, butylparaben, ethylparaben. The CDCCenters for Disease Control and Prevention biomonitored 99.1% of Americans aged six and over had detectable methylparaben — NHANES 2005-2006, n=2,548 of Americans aged six and older for methylparaben in the NHANESNational Health and Nutrition Examination Survey — the CDC's recurring population-level health surveillance programme 2005-2006 cycle, and 92.7% for propylparaben (Calafat et al. 2010). When essentially every body in the country has detectable levels of something, that something is in almost every product.

Parabens are weak estrogen-receptor agonists in vivo. Routledge's 1998 study in immature rats found subcutaneous butylparaben to be roughly 100,000× weaker than estradiol — Routledge 1998 immature rat uterotrophic weaker than estradiol on the uterotrophic assay (Routledge et al. 1998); the 4-hydroxytamoxifen blocker confirmed receptor-mediated activity. Weak isn't nothing — at 99.1% population exposure, a weak agonist that everyone is exposed to at low concentrations is a different kind of problem from a strong agonist that almost nobody gets. Janjua's whole-body topical study showed butylparaben in serum within hours of application (Janjua et al. 2007, n=26 men, mean peak ≈135 µg/L).

Darbre's 2004 paper found parabens in all 20 breast tumour samples she analysed, mean concentration 20.6 ng/g, methylparaben dominant (Darbre et al. 2004). The honest reading: the study had no healthy-tissue controls, so it cannot show parabens cause breast cancer, only that they reach breast tissue at measurable concentrations — which the receptor work and the biomonitoring already implied. Presence is not causation. Repeated daily presence at receptors that respond to it is the kind of exposure the regulatory model was not designed to measure.

The EUEuropean Union took action where the science was firmest. EU 5 parabens banned 2014 Commission Regulation 358/2014 banned isopropyl-, isobutyl-, phenyl-, benzyl-, and pentylparaben from cosmetics from 30 October 2014, on advice from the SCCSScientific Committee on Consumer Safety — the European Commission's independent scientific advisory body for cosmetic-ingredient safety. Regulation 1004/2014 then capped propylparaben and butylparaben at 0.14% (individually or as a sum), restricted them in leave-on products for the under-3 nappy area, and applied from 16 April 2015. UK Same caps retained 2026 retained the same caps post-Brexit. US No federal cap 2026 has no equivalent federal restriction; California AB 2762 banned isobutyl- and isopropylparaben specifically (effective 1 January 2025) but left methyl/propyl/butyl untouched. Most major deodorant brands quietly reformulated to 'paraben-free' between 2014 and 2018. Read the back of any current stick — most no longer list a paraben. The replacement is usually phenoxyethanol, sodium benzoate, or a formaldehyde-releaser like DMDM hydantoinA formaldehyde-releasing preservative — slowly hydrolyses to release formaldehyde gradually over the product's shelf life. The TikTok 'TRESemme scandal' in 2021 was about exactly this ingredient.. Substitution rarely solves the problem. It changes the surname on the question.

DEPDiethyl phthalate — the fragrance solvent. Makes scent linger on skin. The metabolite MEP shows up in urine biomonitoring. (diethyl phthalate) is the phthalate the personal-care industry uses as a fragrance solvent — it makes scent stick to skin and last through the day. It almost never appears on a deodorant ingredient list under its own name. It appears on the label as one word: 'fragrance.' Or 'parfum.' That word is permitted to obscure up to 3,691 individual ingredients on the IFRAInternational Fragrance Association — the global trade body that maintains the master list of permitted fragrance ingredients 2025 Transparency List — a list longer than the combined ingredient declarations of every cosmetic in your bathroom — including DEP.

EWGEnvironmental Working Group and the Campaign for Safe Cosmetics commissioned an independent lab in 2010 to test 17 mass-market fragrances — perfumes and body sprays from major brands — for chemicals not on the label. Twelve of the seventeen contained DEP. Concentrations ran from 30 ppm at the low end up to 32,000 ppm — 3.2% by weight — in Calvin Klein Eternity for Women. Across all 17 products, the lab found 38 secret chemicals that didn't appear on any label. The average product had 14 undisclosed chemicals per fragrance — EWG/Campaign for Safe Cosmetics 2010 lab analysis of 17 mass-market products undisclosed ingredients; the worst, American Eagle Seventy Seven, had 24 (EWG 2010 Not So Sexy).

The dermal-absorption picture for DEP is unusually clean. Janjua's 2008 whole-body topical study applied DEP, dibutyl phthalate, and butylparaben to the skin of 26 men over two weeks and recovered the metabolites in urine: 5.79% of applied DEP recovered as the urinary metabolite MEP within 24 hours — Janjua 2008, n=26 men of applied DEP showed up as monoethyl phthalate (MEP) in urine, 1.82% of DBP as MBP, 0.32% of butylparaben (Janjua et al. 2008). The fragrance solvent does not stay on your skin. It goes through it. Duty's 2005 study of 338 men at the Massachusetts General Hospital fertility clinic found cologne use associated with 2.57× higher urinary MEP in men using cologne — Duty 2005, fertility-clinic cohort higher urinary MEP, and a per-additional-product-type dose-response of 33% per extra fragranced personal-care product (Duty et al. 2005).

The intervention study you want is HERMOSA. The 2016 trial recruited 100 Latina adolescent girls in Salinas, California, and gave them three days of personal-care products explicitly labelled free of phthalates, parabens, triclosan, and benzophenone-3 — replacing their usual deodorant, body wash, shampoo, and lotion. Then they re-measured urinary metabolites.

Three days. Not three weeks. (Harley et al. 2016) The kinetic story this tells is the most important fact in the article: the body clears these chemicals fast — half-lives are hours to days, not years like PFASPer- and polyfluoroalkyl substances — the 'forever chemicals,' which have half-lives measured in years and accumulate in the body. Phthalates and parabens behave very differently.. The reason population biomonitoring shows everyone has them is not bioaccumulation. It is reapplication. Stop reapplying and your levels collapse in three days.

The regulatory loophole that keeps DEP off the label is the trade-secret protection in 21 CFR 701.3(a) for the US and the corresponding fragrance-mixture treatment under EUEuropean Union 1223/2009 Article 19. CA Quaternium-15 banned, DEP NOT 2025 California's AB 2762 — the most progressive US state cosmetics law, in force from 1 January 2025 — banned formaldehyde, paraformaldehyde, methylene glycol, quaternium-15, isobutylparaben, isopropylparaben, mercury, and 13 PFAS, plus DEHP and DBP from the phthalate family. DEP is not on the list. The fragrance carrier remains permitted. EU 80-allergen list 2026 2023/1545 expands required allergen disclosure from 26 to 80+ named substances from 31 July 2026, but the underlying 'fragrance' loophole survives — the rule names which allergens must be declared, not the unnamed mixture as a whole. Our phthalates profile covers the broader family across all consumer-product routes.

Triclosan was added to deodorants in the 1970s as an antibacterial — the bacteria on your underarm skin are what convert sweat into body odour, and killing them stops the smell. The chemical was so widely used that the CDCCenters for Disease Control and Prevention found 74.6% of Americans tested in 2003-2004 had detectable urinary triclosan — Calafat 2008 NHANES, n=2,517 of Americans had detectable triclosan in urine in 2003-2004 (Calafat et al. 2008). The chemical is endocrine-active in animal models — it disrupts thyroid hormone signalling and the gut microbiome — and there's mechanistic concern about antibiotic resistance from chronic low-dose exposure.

US FDA banned consumer hand soap 2017 The FDAFood and Drug Administration pulled triclosan from over-the-counter consumer hand and body soaps in 2016 (81 FR 61106, effective 6 September 2017) — 19 ingredients in total, deemed not generally recognised as safe and effective. The agency's reasoning was framed as efficacy, not safety: 40 years on the market and the manufacturers couldn't show the products performed better than plain soap and water. Triclosan in toothpaste (Colgate Total) was outside the rule's scope and remains legal in the US. So is triclosan in deodorant — both labelled-as-deodorant products and antiperspirants — under the OTC monograph framework.

EU 0.3% cap, mouthwash banned 2025 Europe took a tighter cosmetic line later. Commission Implementing Regulation 2024/996, published 4 April 2024, capped triclosan at 0.3% in toothpastes, hand soaps, body soaps and shower gels, non-spray deodorants, face powders, blemish concealers, and nail-cleaning products — and banned its use in mouthwash entirely. Required label: 'Not to be used for children under three years of age.' No new placements after 31 December 2024; sell-through to 31 October 2025. UK Pre-2024 EU position retained 2026 The UK has not adopted the 2024/996 update — UK cosmetics law retains the older EU position pre-2024 and has not laid the secondary instruments that would mirror the new caps. If you're reading a UK deodorant label and triclosan still appears on it, that's the regulatory divergence in plain text. Our triclosan profile covers the chemistry and the toothpaste angle.

Propylene glycol shows up in roll-on and spray deodorants as a humectant — it binds water, keeps the formulation from drying out, and gives the product its smooth glide. By itself, the toxicology is mostly reassuring: it's classified as Generally Recognized as Safe by the FDA for food contact, and serious systemic toxicity from cosmetic doses is rare. The interesting property is what it does to the formulations around it.

Propylene glycol is one of the most widely used penetration enhancersExcipients added to topical formulations specifically to increase how much active ingredient gets through the stratum corneum and into systemic circulation. Their job is to make skin more permeable. in topical pharmaceutical and cosmetic formulation. Mistry and Notman's 2024 molecular-dynamics paper showed PG molecules disrupt the lipid packing of the SCstratum corneum, the outermost layer of skin, increase lipid mobility, and may extract cholesterol from skin lipid membranes (Mistry & Notman 2024). Translation: a deodorant containing 5–10% propylene glycol does not just deliver fragrance and aluminium and parabens to the skin surface; it makes the skin more permeable to all of them. PG is the chemical that quietly increases the dose of every chemical it's formulated with.

Propylene glycol was named the American Contact Dermatitis Society's Allergen of the Year in 2018 (Jacob, Scheman & McGowan 2018) — a designation given annually to under-recognised everyday exposures producing clinically meaningful allergic contact dermatitis. The patch-test positivity rate in dermatitis-clinic populations is several percent. If you've had unexplained underarm rash after a new deodorant, propylene glycol is on the short list of suspects. The skin-barrier disruption that makes it useful as a humectant is the same property that makes it sensitising for some users.

The 'aluminium-free' badge is the most-misleading single claim on the deodorant shelf. Crystal deodorantsSolid sticks marketed as 'mineral salt' or 'crystal' deodorants — typically a single block of crystallised potassium alum that you wet under running water and rub on your skin. marketed as 'mineral' or 'natural' are made from KAl(SO₄)₂.₁₂H₂O — potassium alum, CASChemical Abstracts Service registry number — the unique numerical identifier each chemical compound is assigned 7784-24-9. The first letter of the chemical formula is aluminium. Potassium alum is unambiguously an aluminium compound. Industry has avoided the FDA's antiperspirant monograph by formulating with potassium alum instead of aluminium chlorohydrate, which lets the brand label 'aluminium-free' under FDA labelling rules — but the underlying chemistry is still aluminium. The 'aluminium-free' badge on a potassium-alum stick is an exercise in punctuation, not chemistry.

The peer-reviewed dermal-absorption literature on potassium alum specifically is essentially empty — there is no head-to-head study comparing aluminium uptake from potassium alum versus aluminium chlorohydrate in matched skin. The chemistry of mineral-salt deodorants involves precipitation of an aluminium-protein complex on the skin surface that may absorb less readily than ACH formulations, but in the absence of measurement that's a hypothesis, not a finding. If your reason for switching to aluminium-free is to reduce aluminium exposure to breast tissue, a crystal deodorant is not what you bought.

Genuinely aluminium-free deodorants exist. They use combinations of: baking soda (sodium bicarbonate, raises skin pH to interfere with bacterial enzymes), magnesium hydroxide (similar mechanism, less alkaline, better for sensitive skin), zinc ricinoleate (binds odour molecules), arrowroot powder (absorbs moisture), activated charcoal (similar), and essential oils (antibacterial, depending on the oil). Trade-offs are real: these don't block sweat, only mask or reduce odour. If you sweat heavily, switching to a deodorant-only formulation will produce visible underarm wetness on a hot day or under stress. If you don't sweat heavily, you probably won't notice. The transition period — anywhere from a week to a month — sometimes produces an adjustment in the underarm microbiome and a brief uptick in odour, which most users find resolves.

The honest filter is structural, not ingredient-by-ingredient. The closer the product is to a single-purpose formulation, the fewer chemical families share your underarm skin. The phrase on the front of the package is almost never useful; the ingredient list on the back almost always is.

The eso-friendly approach here is closer to triage than chemistry. The strongest evidence sits with parabens and the fragrance carrier DEP, and most major deodorant brands now have paraben-free, fragrance-free options that cost roughly the same as the conventional product. The aluminium debate is genuine but unsettled, and the only way to resolve it for yourself is to look at the available studies — Linhart, Mirick, Trinh — and decide whether 18,000 lifetime applications of a chemical that demonstrably reaches breast tissue is worth taking on faith from a literature that hasn't tested the relevant exposure window. The crystal-deodorant shortcut isn't a shortcut. It's the same chemistry under a different headline.

Reach into the cabinet next time. The brand on the front and the chemistry on the back are two different products. The product on the back is what your skin meets every morning, the same skin that drains into the same lymph nodes the breast does, the same skin you've often just shaved. If you're going to do this 18,000 times — and you are — the smallest change with the biggest evidence base is removing the parabens and the fragrance. The aluminium argument is the loud one. The paraben and DEP arguments are the ones the evidence actually supports. The chemicals you remove first should be the ones with the cleanest case against them, not the ones the internet shouts loudest about.