Don’t take our word for it. Here’s every fact on this site, with its evidence grade, its caveats, and a link to the source, so you can check us.
We start you on the strong ones: findings that have been replicated or come from one large, well-controlled study. Moderate and weak facts are here too, labeled for what they are. Use “Link to this fact” to share a single card.
Skin allergy / Other evidence
Methylisothiazolinone allergy in North America appears to be falling from its 2017-2018 peak (15.3% to 11.5%), with lower rates at Canadian centers (9.9%) than US centers (13.5%) after Canada banned it in leave-on cosmetics in 2018.
Moderate evidence: Ecological comparison of clinic networks; Canada vs US difference could reflect other factors; the authors call the regulatory link a hypothesis consistent with the data.
Skin allergy / Other evidence
Nickel patch testing at 5% petrolatum picked up more allergy than 2.5% (24.9% vs 22.1% positive), at the cost of slightly more irritant reactions (0.9% vs 0.3%).
Moderate evidence: Positivity is not proof of clinical relevance; 59% of nickel reactions were past relevance only.
Skin allergy / Measured in people or real products
In a 40-month (1977-1980) prospective series by 11 North American contact dermatitis specialists, 487 cosmetic-related contact dermatitis cases were identified, about 6% of the 8,093 patients they patch tested.
Moderate evidence: Selected referral series of contact dermatitis specialists, 1977-1980 (40 months); not general-population rates.
Source: Eiermann 1982, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Measured in people or real products
Dermatologists diagnosed 80% of these cosmetic reactions as allergic contact dermatitis and 16% as irritant, but the authors call that skew an artifact because irritation was under-recorded and phototoxicity was not routinely tested.
Moderate evidence: Authors' own caveat; irritation diagnosis was by exclusion and subjective stinging was never recorded.
Source: Eiermann 1982, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Measured in people or real products
Skin care products caused the most cosmetic reactions (26%), then non-color hair preparations (16%) and facial makeup (12%); perfumes, colognes and toilet waters were only 7%.
Moderate evidence: Selected referral series of contact dermatitis specialists, 1977-1980 (40 months); not general-population rates.
Source: Eiermann 1982, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Measured in people or real products
Quaternium-15, a formaldehyde-releasing preservative, was the single most frequent named allergen in the series (32 reactions), ahead of p-phenylenediamine (24) and propylene glycol (22).
Moderate evidence.
Caveats
Selected referral series of contact dermatitis specialists, 1977-1980 (40 months); not general-population rates; ingredient counts are called 'unrealistically low' by the authors and reflect what was in the screening series ('we observe what we seek').
Source: Eiermann 1982, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Measured in people or real products
The authors suspected that 'some or many' of their 22 propylene glycol 'allergy' cases were really irritation, because positives at 4% or more mostly failed to reproduce on retest weeks later.
Moderate evidence: Authors' suspicion, not a measured rate; many cases predate their stricter criteria.
Source: Eiermann 1982, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Measured in people or real products
The face was the most affected site (37% of cases), then the eye area (13%), lower arm (11%) and armpit (8%); the authors say this partly reflects how often and how visibly cosmetics are used there.
Moderate evidence: Selected referral series of contact dermatitis specialists, 1977-1980 (40 months); not general-population rates.
Source: Eiermann 1982, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Other evidence
In the human 4-hour skin patch test cited in this review, 19 of 56 volunteers reacted to 7.5% benzalkonium chloride, against 32 of 56 reacting to 20% sodium lauryl sulfate, a standard irritant.
Moderate evidence: Relayed from a paper co-authored by the lead author; different concentrations, so not a head-to-head potency ranking.
Source: Basketter et al. 2004, Contact Dermatitis Industry-funded Link to this fact
Skin allergy / Other evidence
Reports of benzalkonium chloride allergy come from nurses, doctors, dentists and veterinarians and from antiseptics, antifungal creams, toothpaste, plaster of Paris and especially eye preparations; the review itself says the eye may be an exception to its skin irritant conclusion.
Moderate evidence: Case reports are not prevalence data and often lacked repeat testing or use tests, which the authors say they should have had.
Source: Basketter et al. 2004, Contact Dermatitis Industry-funded Link to this fact
Skin allergy / Other evidence
Contact allergy to parabens is relatively rare for how widely they are used: 0.6% of patch-tested US patients in 2001-02 (down from 1.7% in 1996-98) and 0.5-1.0% in Europe, against about 9% for formaldehyde and quaternium-15 in the US.
Moderate evidence: Patch-tested patients, not the general public; secondary figures; test preparations and referral differ between regions.
Skin allergy / Other evidence
North American patch-test clinics found about 9% positive reactions to formaldehyde and to quaternium-15 (1996-2002) against 2-2.5% and about 1% in Europe (1991-2000); diazolidinyl urea 2.7-3.7% vs 0.5-1.5%.
Moderate evidence: Patch-tested referral patients; the review does not separate protocol or referral differences from exposure differences; quaternium-15 limits are 1% US vs 0.2% EU.
Skin allergy / Other evidence
In seven patch-test studies run by the maker of the lotions (738 adults completing), body lotions with 1-5% behentrimonium chloride caused no allergic contact dermatitis and only mild, transient irritation; every lotion rated 'mild material' on the standard cumulative-irritation scale.
Moderate evidence: Manufacturer-sponsored, own products, healthy adults without skin disease; scores from different labs were normalized by the authors; no independent replication.
Source: Cameron 2013, Toxicology in Vitro Industry-funded Link to this fact
Skin allergy / Other evidence
In four- and two-week home-use tests, 175 adults and children with dry, eczema-prone or atopic skin used behentrimonium chloride body lotions with no clinically assessed reactions; a handful of brief itch or sting reports (five people) were the only complaints.
Moderate evidence: Short (2-4 weeks), small, specific formulations from one company, no unexposed or independent comparator in most arms; manufacturer-run.
Source: Cameron 2013, Toxicology in Vitro Industry-funded Link to this fact
Skin allergy / Other evidence
CIR's report states that behentrimonium chloride was 'a sensitizer at 16% in guinea pigs' in its Summary, but its body text describes the positive guinea pig test at a 10% solution and no reactions in the 16% Buehler test, and a rinse-off formulation with 3.4% behentrimonium chloride caused no sensitization in 104 people.
Moderate evidence.
Caveats
Reading the report's text versus its Summary; the dossier tests used different vehicles and challenge levels. The report also writes 2.4% in the text for the human test but lists 3.4% in Table 12. CIR is financed by the Personal Care Products Council, the cosmetics industry trade association.
Source: Becker et al. 2012 (CIR), Int J Toxicol, International Journal of Toxicology Industry-funded Link to this fact
Skin allergy / Other evidence
In 15 human repeated-insult patch tests of hair products containing cetrimonium chloride, behentrimonium chloride or both (one test used cetrimonium methosulfate), the industry panel found no sensitization at challenge. Induction irritation was common but mostly mild (for example 12% of 102 participants, 12 of 105, 32 of 213), and a hair treatment with 0.8% cetrimonium chloride tested at 25% and 50% dilution irritated 50% and 80% of 107 participants at induction.
Moderate evidence.
Caveats
Industry-submitted and unpublished studies at low concentrations; leave-on skin products at 1-5% are not represented in the table. CIR is financed by the Personal Care Products Council, the cosmetics industry trade association.
Source: Becker et al. 2012 (CIR), Int J Toxicol, International Journal of Toxicology Industry-funded Link to this fact
Skin allergy / Other evidence
Rabbit skin was not irritated by hydrogen peroxide up to 10%, was irritated at 35% and corroded at about 50% and above, and guinea pigs were not sensitized at up to 6%, which the CIR panel used to call dermal irritation and sensitization concern from cosmetics 'negligible'.
Moderate evidence: Animal data summarized by an industry-funded panel; not human use data.
Source: Becker et al. 2024, International Journal of Toxicology Industry-funded Link to this fact
Skin allergy / Review of other studies
Benzyl alcohol was named the American Contact Dermatitis Society's 2026 Allergen of the Year; US dermatologists describe it as an underrecognized contact allergen that is a preservative, solvent and fragrance ingredient in personal care products and skin medicines.
Moderate evidence: Narrative review; selection by a dermatology society that promotes the award; no new data.
Source: Zheng et al. 2026, ACDS Allergen of the Year (Cutis), Cutis Link to this fact
Skin allergy / Other evidence
In North American Contact Dermatitis Group patch testing in 2021-2022, 0.5% of 3,050 patients reacted to benzyl alcohol; it had been added to the NACDG series in 2019-2020 because of increasing use.
Moderate evidence: Referred dermatology clinic patients, not the general population; relayed from the primary, which has 15 positives.
Source: Zheng et al. 2026, ACDS Allergen of the Year (Cutis), Cutis Link to this fact
Skin allergy / Review of other studies
Benzyl alcohol is also an inactive ingredient in prescription and over-the-counter skin treatments, including corticosteroid creams and ointments, calcineurin inhibitors, topical JAK inhibitors, prescription shampoos and antiseptic creams, so a medicine for dermatitis can contain an allergen that worsens it.
Moderate evidence: Product lists are examples, not a survey; formulation analysis is unpublished data.
Source: Zheng et al. 2026, ACDS Allergen of the Year (Cutis), Cutis Link to this fact
Skin allergy / Other evidence
Among more than 70,000 patch-tested patients, a topical medication was the suspected allergen source in 34.9% of benzyl alcohol-positive patients versus 16.6% of benzyl alcohol-negative patients.
Moderate evidence: Suspected source is clinician judgment in a patch-test clinic population; association, not exposure measurement.
Source: Zheng et al. 2026, ACDS Allergen of the Year (Cutis), Cutis Link to this fact
Skin allergy / Other evidence
Adults with benzyl alcohol allergy were about twice as likely to have leg dermatitis as benzyl alcohol-negative adults (17.8% vs 8.6%), and hand dermatitis was the commonest site (24.0%); the authors link this to chronic dermatoses and compromised skin barriers.
Moderate evidence: Stasis-dermatitis explanation is the authors' interpretation, not tested.
Source: Zheng et al. 2026, ACDS Allergen of the Year (Cutis), Cutis Link to this fact
Skin allergy / Measured in people or real products
In a German car-industry cohort followed from apprenticeship, people with atopic skin diathesis had about 2.5 times the odds of current hand eczema (adjusted OR 2.47, 95% CI 1.71-3.55; 110 cases vs 120 controls), and it was the only significant factor measured.
Moderate evidence.
Caveats
Single-company cohort with an active skin-protection program; small sample; prevalent cases; controls deliberately enriched with people of unhealthy skin; the narrow CI looks unusually tight versus the crude prevalences (38.5% vs 10.9%).
Source: Apfelbacher et al. 2010, Contact Dermatitis, Contact Dermatitis Link to this fact
Skin allergy / Measured in people or real products
Among 57 cases of current irritant hand eczema in the same car-industry cohort, atopic skin diathesis was again the only significant factor (adjusted OR 1.84, 95% CI 1.20-2.80), somewhat lower than for all hand eczema.
Moderate evidence: Same small single-cohort dataset as the all-hand-eczema analysis, not an independent replication.
Source: Apfelbacher et al. 2010, Contact Dermatitis, Contact Dermatitis Link to this fact
Skin allergy / Self-reported survey
In a prospective cohort of 721 Dutch apprentice nurses, 18% of the 445 with no earlier hand eczema (81 people) developed it during 1-3 years of traineeships, mostly in the first traineeship.
Moderate evidence: Symptom-based self-report on diary cards (probably overestimates HE); AD, rhinitis and asthma overrepresented in the cohort; 26% lost early; nurse apprentices only.
Source: Visser et al. 2014 (Part I), Contact Dermatitis, Contact Dermatitis Link to this fact
Skin allergy / Self-reported survey
Among Dutch apprentice nurses, about 21-31% reported hand eczema in a given training year, and about 15% of those followed through two traineeships had it in both.
Moderate evidence: Abstract and Results disagree on the first-year figure (23% vs 21%); symptom-based definition; selected cohort.
Source: Visser et al. 2014 (Part I), Contact Dermatitis, Contact Dermatitis Link to this fact
Skin allergy / Self-reported survey
About 29% of 383 apprentice nurses exceeded 2 hours a day of wet work or 20 hand washes or disinfections a day (43% in hospitals, 6% in psychiatry).
Moderate evidence: Diary cards filled on self-chosen shifts; cut-offs from a German guideline and a Dutch expert group; the abstract rounds this to about one-third.
Source: Visser et al. 2014 (Part I), Contact Dermatitis, Contact Dermatitis Link to this fact
Skin allergy / Other evidence
In a Cleveland contact-dermatitis clinic, people with a history of childhood eczema reacted to weak allergens (propylene glycol, eugenol, vanillin, parabens) about 1.6 times as often as other patients (odds ratio 1.62, 95% CI 1.01-2.59).
Moderate evidence.
Caveats
Single-clinic retrospective study of referred patients; lower confidence bound is 1.01; atopy is a recalled-history proxy; weak allergens pooled, not tested one by one; association, not cause. Funded by the American Contact Dermatitis Society; no conflicts declared.
Source: Kohli & Nedorost 2016, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Other evidence
Wet workers reacted to medium-potency allergens about 1.3 times as often as dry workers (25.3% vs 20.8%; odds ratio 1.29, 95% CI 1.01-1.66), and to the rubber chemical thiuram mix 2.8 times as often (8.7% vs 3.0%).
Moderate evidence: Borderline interval; occupation assigned by one investigator from current job only; glove rubber exposure explains part of the result.
Source: Kohli & Nedorost 2016, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Other evidence
The same clinic found no extra reactions to strong sensitizers (neomycin, PPD, MCI/MI) in people with childhood eczema or in wet workers, which the authors read as strong allergens not needing a damaged barrier to sensitize.
Moderate evidence.
Caveats
Non-significant result is not proof of no effect; retrospective single-clinic data; the authors' mechanism (barrier damage supplies a danger signal that weak allergens lack) is a hypothesis, not tested.
Source: Kohli & Nedorost 2016, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Other evidence
Wet workers were far more likely than dry workers to present to the clinic with hand dermatitis (51.1% vs 30.3%; odds ratio 2.40, 95% CI 1.94-2.97).
Moderate evidence: Clinic-referred patients, so this is not a population rate; occupation classification crude.
Source: Kohli & Nedorost 2016, Journal of the American Academy of Dermatology Link to this fact
Skin allergy / Other evidence
At a large US referral clinic, 60.7% of 2,582 people patch tested in 2011-2015 had at least one allergic reaction; the most common were nickel (18.3%), methylisothiazolinone (13.6% of 964 tested), balsam of Peru / Myroxylon pereirae resin (9.2%) and neomycin (7.8%).
Moderate evidence: Referred dermatitis patients at a tertiary clinic, not the general population; methylisothiazolinone tested in only 964 patients.
Source: [[src-2018-trends-patch-testing-mayo-clinic-standard-series|Veverka 2018]], Dermatitis® Link to this fact
Skin allergy / Other evidence
Benzalkonium chloride, a disinfectant and preservative in hand sanitizers, eye products, cleaners and personal care products, gave allergic-pattern patch reactions in 6.9% of 2,546 patients tested (sixth highest), most of them weak.
Moderate evidence: Single clinic; weak reactions can be irritant, relevance not reported, and benzalkonium chloride is a known irritant.
Source: [[src-2018-trends-patch-testing-mayo-clinic-standard-series|Veverka 2018]], Dermatitis® Link to this fact
Skin allergy / Other evidence
Fragrance markers were among the commonest positives at this clinic: Myroxylon pereirae resin 9.2% and fragrance mix I 6.7% of patients tested, with fragrance mix II 2.6%.
Moderate evidence: Marker mixtures miss fragrance allergy to other ingredients; clinic-referred patients; Mayo fragrance mix I was about half the NACDG rate.
Source: [[src-2018-trends-patch-testing-mayo-clinic-standard-series|Veverka 2018]], Dermatitis® Link to this fact
Skin allergy / Measured in people or real products
Among 38,723 North American patients referred for allergy patch testing (2001-2016), the 303 who ended up with a stasis dermatitis diagnosis reacted more often to fragrance mix I (16.9% vs 9.8%) and to balsam of Peru (15.6% vs 8.3%) than other patients.
Moderate evidence.
Caveats
Referral population, stasis dermatitis labeled by physician after testing with no formal criteria; adjusted odds ratios were modest (1.45 and 1.54); NACDG centers, so not independent of the other NACDG reports.
Source: Silverberg et al. 2021, Arch Dermatol Res, Archives of Dermatological Research Link to this fact
Skin allergy / Measured in people or real products
In the same stasis dermatitis group, 16.2% reacted to bacitracin, a topical antibiotic sold over the counter in the US (odds about 2.3 times other patients), and the source was an antibiotic product in 47 of 48 recorded cases.
Moderate evidence: Referral series; source recorded for only 48 of 49 bacitracin positives; patch-test positivity is not proof the antibiotic caused the leg dermatitis.
Source: Silverberg et al. 2021, Arch Dermatol Res, Archives of Dermatological Research Link to this fact
Skin allergy / Measured in people or real products
Stasis dermatitis patients also reacted more often to several preservatives and moisturizer ingredients: quaternium-15 (11.0%), diazolidinyl urea (6.0%) and propylene glycol (5.0%), with moisturizers and other personal care products the most common recorded sources.
Moderate evidence.
Caveats
Quaternium-15 odds ratio 1.48 had P = 0.038 against the paper's own cutoff of 0.039; propylene glycol is also an irritant at the 30% test dose; source data cover only positives with a recorded source (n = 19 to 33 per allergen).
Source: Silverberg et al. 2021, Arch Dermatol Res, Archives of Dermatological Research Link to this fact
Skin allergy / Government agency
EFSA found no data on food allergy or intolerance from swallowed polysorbates, only two injection case reports (polysorbate 80, non-immunological anaphylactoid and subcutaneous hypersensitivity) and skin tests in which 2 of 1,206 eczema patients reacted and patch tests were positive in 5 of 737 (polysorbate 40) and 4 of 737 (polysorbate 80) dermatitis patients.
Moderate evidence: Case reports and clinic series; patch-test denominators are patients already seen for dermatitis; oral allergenicity untested.
Source: EFSA ANS Panel (2015), full opinion, EFSA Journal Link to this fact
Skin allergy / Other evidence
Among 5,595 people referred to North American skin-allergy clinics and patch tested in 2015-2016, 11.3% reacted to fragrance mix I, 7.0% to balsam of Peru and 5.3% to fragrance mix II, three of the commonly positive allergens in the group's 70-allergen series.
Moderate evidence.
Caveats
Referral patients, not the general public; mixes are partial markers; the series tested no oxidized limonene or linalool; one of several NACDG cycles (not independent of the 2021-2022 report, same group).
Skin allergy / Other evidence
Clinicians judged about 85% of fragrance mix I positives, 88% of fragrance mix II positives and 91% of balsam of Peru positives to be currently relevant, but mostly as 'possible' (skin contact with fragrance-containing products likely), not a verified product match.
Moderate evidence: Relevance is the investigator's judgment with minimal follow-up or use tests; the authors say so in their limitations.
Skin allergy / Other evidence
Fragrance mix I positivity in North American clinics was 11.3% in 2015-2016 against 11.9% two years earlier (not a significant change), and has swung between about 8.5% and 12.1% across 2005-2016 cycles, so a single cycle can differ from the previous one by several points.
Moderate evidence: Same referral network with changing membership; no multiple-comparison adjustment.
Skin allergy / Other evidence
Allergy to the preservative methylisothiazolinone kept rising in North American clinics: 13.4% of patients reacted in 2015-2016 against 10.9% in 2013-2014, making it the second most common allergen after nickel and the top-ranked by the clinic group's relevance-weighted index.
Moderate evidence: Referral patients; MI was tested at 2,000 ppm, far above cosmetic use levels; later NACDG data show a decline (MI 11.5% in 2021-2022).
Skin allergy / Other evidence
Parabens had the lowest patch-test positivity of any major preservative in North American clinics (paraben mix 0.6% in 2015-2016, down from 1.2% in 2005-2006), so 'paraben-free' is not an allergy-driven choice for most people.
Moderate evidence: Paraben allergy is not the only paraben question (endocrine evidence is separate); the allergy conclusion applies to patch-test sensitization only.
Skin allergy / Other evidence
About 1 in 5 (22.5%) patients in 13 North American contact-dermatitis clinics had a relevant allergic reaction to something outside the group's 70-allergen standard panel, and about half of those came from cosmetics, beauty and skin-care products (47.4%).
Moderate evidence: Referral patients; the specific non-series allergens were not recorded.
Skin allergy / Other evidence
The T.R.U.E. Test, the ready-made patch-test panel, would hypothetically have detected only 61.6% to 74.0% of the reactions found by North American clinic specialists' broader series, missing allergens such as methylisothiazolinone, fragrance mix II and propylene glycol.
Moderate evidence: A hypothetical calculation by the NACDG, which designs its own series; no patient was actually tested with the T.R.U.E. Test.
Skin allergy / Other evidence
The clinic group counted 8.7% of all its allergic patch-test reactions from doubtful (barely visible) second readings that the investigators then interpreted as allergic, including 42.9% of the doubtful fragrance mix I and balsam of Peru reads.
Moderate evidence: The upgrade rests on each investigator's judgment; whether such calls are true allergy is the open question.
Skin allergy / Other evidence
Formaldehyde at a 2% patch-test concentration found 8.4% of referred patients positive against 6.4% at 1%, and the authors will drop the 1% preparation from the 2019-2020 series because testing 1% in addition to 2% was judged to have low utility.
Moderate evidence: Their strong-reaction comparison contains an arithmetic slip (149/299 should be 149/357 = 41.7%); the 2% advantage is real in count but its irritant share is not shown.
Skin allergy / Other evidence
Parabens and benzophenones are personal-care ingredients (preservatives and UV filters), not fragrance ingredients; this pregnancy cohort did not measure fragrance compounds or product use, so it says nothing directly about fragrance and eczema.
Moderate evidence: Scope statement. Source of exposure (lotions, cosmetics, sunscreen) was not recorded.
Source: Miller 2025 (ECHO), Environment International Link to this fact
Skin allergy / Other evidence
A 2026 therapy review calls linalool and lavender oils non-toxic, relying on a 2003 industry-panel assessment, and never mentions that air-oxidized linalool is a leading contact allergen.
Moderate evidence: Absence of discussion is not a test; the review's subject is neuropharmacology and its patients are older adults, not dermatitis patients.
Source: Piccialli 2026, Phytotherapy Research Link to this fact
What you can do / Self-reported survey
About twice as many Australians would prefer fragrance-free workplaces (42.8%) as fragranced ones (22.2%), and over three times as many prefer unscented flights (57.7% vs 16.3%).
Moderate evidence.
Source: Steinemann 2017, Preventive Medicine Reports Link to this fact
What you can do / Self-reported survey
Most people would rather have fragrance-free spaces: 65% prefer fragrance-free airplanes, 61% hotels, 51% health care, and 48% workplaces, against 16-22% who prefer them fragranced.
Moderate evidence: Preference survey, adults 18-65.
Source: Steinemann 2019, Air Quality, Atmosphere & Health Link to this fact
What you can do / Self-reported survey
People with asthma back fragrance-free spaces by wide margins: 66.2% vs 16.1% for fragrance-free workplaces and 72.1% vs 14.8% for fragrance-free health care.
Moderate evidence.
Source: Steinemann 2018, Air Quality, Atmosphere & Health Link to this fact
What you can do / Measured in people or real products
Screening out products whose labels list 'fragrance' and a few other terms cut synthetic fragrance: 26 of 43 screened products had no detectable fragrance chemical, and high fragrance levels appeared 21 times in conventional samples vs 2 in screened ones.
Moderate evidence: Conventional composites vs single alternatives not directly comparable; did not remove parabens, phthalate substitutes.
Source: Dodson 2012, Environmental Health Perspectives Link to this fact
What you can do / Self-reported survey
Swedes overwhelmingly prefer scent-free spaces: 80.2% vs 6.0% for airplanes, 77.7% vs 9.8% for hotels, and 64.1% vs 14.0% for health care.
Moderate evidence.
Source: Steinemann 2018 (Sweden), Air Quality, Atmosphere & Health Link to this fact
What you can do / Measured in people or real products
Perfume was the strongest single predictor of diethyl phthalate exposure, so skipping perfume and fragranced personal care products is the most obvious way to lower it.
Moderate evidence: Association only; not tested as an intervention.
Source: Parlett 2013, Journal of Exposure Science and Environmental Epidemiology Link to this fact
What you can do / Measured in people or real products
Green, organic, or natural labels on baby products did not mean fewer released chemicals, so choosing products labeled fragrance-free (without essential oils) is the more reliable way to avoid scent chemicals.
Moderate evidence.
Caveats
Fragrance-free products were not tested directly for VOC emissions in Nematollahi 2018 (its own scope); Temkin 2023 (EWG-authored; unblinded chamber test of 30 US cleaners, 7 fragrance-free) later found fragrance-free green cleaners significantly lower in total VOC concentration and VOC count than conventional ones, while green-with-fragrance cleaners were not significantly different, which supports the advice for cleaners but not yet for baby products. Hamann 2015 shows a different but relevant caution: even 'fragrance-free' US pediatric products commonly contain non-fragrance contact allergens (80.8% of 52 had >=1 NACDG-tray allergen, mostly preservatives), so 'fragrance-free' is more reliable specifically against added-scent VOCs, not against contact allergens generally.
Source: Nematollahi 2018, Air Quality, Atmosphere & Health Link to this fact
What you can do / Measured in people or real products
Researchers recommend lower-terpene (less citrus/pine-scented) cleaners, diluting them, ventilating during and for hours after cleaning, cleaning when fewer people are home, and never running ozone generators or ionizers with these products.
Moderate evidence: Author recommendations derived from chamber chemistry, not tested as interventions.
Source: Singer 2006, Atmospheric Environment Link to this fact
What you can do / Self-reported survey
More UK adults prefer fragrance-free over fragranced workplaces (44.7% vs 23.3%), health care (43.3% vs 26.7%) and airplanes (61.9% vs 18.4%).
Moderate evidence: Pluralities, not majorities, for workplace and health care; preference, not health benefit.
Source: Steinemann 2018, Air Quality, Atmosphere & Health Link to this fact
What you can do / Review of other studies
People with suspected fragrance dermatitis should ask for patch testing that includes oxidized limonene and linalool, because standard fragrance screens miss many cases.
Moderate evidence.
Source: Sukakul 2024, Acta Dermato-Venereologica Industry-funded Link to this fact
What you can do / Measured in people or real products
Taking the air freshener out of the car removes its main added chemical: without one, in-car limonene was about 5-6 ug/m3.
Moderate evidence: Small pilot; limonene only.
Information, not medical advice. See also: myths we won’t tell you.