Every fact on this site, with its evidence grade, its caveats, and a link to the source.
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.
Health effects / Lab study (cells or chemistry)
Three fragrance and cosmetic chemicals used in bodycare products applied around the breast — benzyl salicylate, benzyl benzoate, and butylphenylmethylpropional (Lilial) — activated the human estrogen receptor and increased growth of estrogen-dependent breast cancer cells in a laboratory dish, an effect blocked by an anti-estrogen drug, suggesting the mechanism runs through the estrogen receptor.
Moderate evidence.
Caveats
Independent, non-industry-funded MCF7 cell study; concentrations required were 10,000 to 3,000,000-fold higher than the estradiol positive control, and none of the three compounds produced a full agonist response within 7 days (partial agonists at that timepoint, converging toward full over 35 days). No animal or human confirmation exists for any of the three compounds. A 2023 independent, guideline-validated retest of Lilial alone (OECD/EPA HeLa9903 ER-transactivation and MDA-kb2 AR-transactivation reporter assays, up to 100 µM, including its major Phase I S9 metabolites) found no estrogenic or androgenic agonist activity at all; this does not test benzyl salicylate or benzyl benzoate, and uses a different (receptor-transactivation vs. proliferation/pS2) endpoint, so it complicates but does not retract this finding for Lilial specifically.
Source: Charles & Darbre 2009, Journal of Applied Toxicology Link to this fact
Health effects / Lab study (cells or chemistry)
These estrogen-receptor effects only appeared at concentrations tens of thousands to millions of times higher than the natural hormone estradiol needs to produce the same size effect — this is laboratory-dish evidence that these chemicals can act like a weak estrogen, not evidence that normal skin or cosmetic use delivers enough of them to the breast to do so.
Moderate evidence.
Caveats
The exposure comparison is our own back-of-envelope calculation, not in the paper, using the fragrance industry's own worst-case 100% dermal-absorption assumption for benzyl salicylate only; benzyl benzoate and Lilial have no systemic-exposure estimate in our evidence base at all, so no comparable check exists for them.
Source: Charles & Darbre 2009, Journal of Applied Toxicology Link to this fact
Health effects / Measured in people or real products
In lab tests, basophils (a type of blood immune cell) from people with perfume-related breathing symptoms released significantly more histamine when exposed to strong perfume than basophils from healthy volunteers, but only at the strongest concentration tested; both groups' basophils started releasing histamine at the same dose.
Moderate evidence.
Caveats
Small sample (17 patients vs 20 controls) from a dermatology patch-test clinic, not a general population; significant only at the single highest (5%) perfume concentration; no chemical fractionation to identify which fragrance component(s) drove release.
Source: Elberling 2007, Clinical and Experimental Allergy Link to this fact
Health effects / Measured in people or real products
The excess basophil reactivity to perfume in symptomatic patients was not explained by allergy antibodies in the blood (a passive-serum-transfer test found no difference between groups) and did not track whether a patient also had a skin (contact) allergy, an inhalant allergy, or asthma.
Moderate evidence.
Caveats
Same research group, clinic and perfume as the 2006 eye-challenge study, so this corroborates rather than independently replicates that program's earlier 'not classic allergy' findings.
Source: Elberling 2007, Clinical and Experimental Allergy Link to this fact
Health effects / Lab study (cells or chemistry)
In a laboratory study, triclosan exposed to UVB at an intensity matched to real midday sunlight in a subtropical city generated damaging free radicals and killed human skin cells (HaCaT keratinocytes) in a dose-dependent way.
Moderate evidence.
Caveats
Single lab, single immortalized human keratinocyte cell line, not primary human skin or an in vivo model; triclosan concentration (5 ug/ml) not validated against measured human skin-surface levels; not replicated.
Source: Dubey 2019, Ecotoxicology and Environmental Safety Link to this fact
Health effects / Lab study (cells or chemistry)
Triclosan and UVB/sunlight together caused significant DNA damage (comet assay, micronuclei, cyclobutane pyrimidine dimers) in human skin cells in the laboratory, while triclosan alone in the dark and UVB/sunlight alone caused none.
Moderate evidence: Single in-vitro cell-line study; genotoxicity not shown in whole skin or in vivo.
Source: Dubey 2019, Ecotoxicology and Environmental Safety Link to this fact
Health effects / Measured in people or real products
In a 2021 audit of 50 stores across 8 Boston neighborhoods, stores in lower-income Roxbury and Mission Hill were more than twice as likely to stock a hair product with a high (7-10) Environmental Working Group hazard score than stores in higher-income, predominantly White Beacon Hill (risk ratio 2.3 for both).
Moderate evidence.
Caveats
Retail-shelf audit of what stores stock, not what people buy or use; EWG hazard score is a derived ingredient-label rating, not a measured chemical concentration; only 2 of 6 neighborhood comparisons reached statistical significance, and Chinatown was excluded for 94% missing scores.
Source: Chan 2023, Environmental Health Perspectives Link to this fact
Health effects / Measured in people or real products
Hair relaxers were found for sale only in the lower-income, more racially diverse Boston neighborhoods studied; none were cataloged in the two highest-income/highest-White or Asian-share neighborhoods (Beacon Hill, Chinatown).
Moderate evidence.
Caveats
Descriptive product-availability finding from one city in 2021, not a formal statistical test; does not measure whether residents of any neighborhood actually purchase relaxers there.
Source: Chan 2023, Environmental Health Perspectives Link to this fact
Health effects / Measured in people or real products
In the best-controlled human test available, propylene glycol methyl ether (PGME) solvent vapor at 150 ppm produced no measurable objective eye damage (redness, corneal swelling, tear-film stability, blinking rate) when tested double-blind against a matched-odor placebo in 12 volunteers.
Moderate evidence: Small study (n=12, all male), single 2.5 h exposure; does not test longer or repeated exposures, other populations, or higher occupational concentrations.
Source: Emmen et al. 2003, Toxicology Letters Industry-funded Link to this fact
Health effects / Measured in people or real products
A statistically significant 'nose irritation' rating from a scented solvent vapor disappeared once how strongly subjects could smell the vapor was accounted for statistically, showing that self-reported irritation from a scented product can really be a report of smell, not tissue injury.
Moderate evidence: One study, one chemical, one nasal-irritation question; the eye-irritation rating in the same study did not similarly vanish.
Source: Emmen et al. 2003, Toxicology Letters Industry-funded Link to this fact
Health effects / Measured in people or real products
In a 3-hour chamber test at realistic ozone and citrus-terpene (limonene) levels, people with mild asthma rated the air as less irritating to their eyes, nose, and throat than healthy people did, but reported more chest tightness, headache, and flu-like symptoms.
Moderate evidence.
Caveats
Single lab, n=33 mild asthmatics only (moderate/severe excluded by design), no unexposed/clean-air control arm, so this is a between-group comparison at one dose, not a test of the chemistry against no exposure.
Health effects / Measured in people or real products
People with mild asthma performed a concentration task significantly less accurately than healthy people during a 3-hour ozone-limonene chamber exposure, despite working faster.
Moderate evidence: Single study, no unexposed control arm, so it is not established that the chemistry (rather than general chamber occupancy or task load) caused the accuracy difference.
Health effects / Measured in people or real products
In the nasal-lavage arm of the same 130-woman chamber study, breathing a VOC-ozone mixture containing citrus and pine scent chemicals (limonene, alpha-pinene) caused no significant increase in nasal irritation symptoms or in objective inflammation markers (immune cells, IL-6, IL-8, protein) in nasal fluid.
Moderate evidence.
Caveats
Single 2-hour exposure in healthy, non-asthmatic, non-chemically-sensitive women; a different, smaller earlier study using a different lavage technique found a positive result at a similar VOC dose.
Source: Laumbach 2005, Journal of Occupational and Environmental Medicine Link to this fact
Health effects / Review of other studies
General reviews of the skin-penetration literature put the size ceiling for crossing intact skin's outer barrier (the Stratum Corneum) at around 100 nanometers; larger particles usually do not cross it.
Moderate evidence: Stated as a citation to a prior review (Revel et al. 2018), not new primary data from this paper; an order-of-magnitude rule, not a precise cutoff.
Source: Menichetti et al. 2025, Journal of Xenobiotics Link to this fact
Health effects / Lab study (cells or chemistry)
When the outer skin barrier (Stratum Corneum) is removed or already compromised, far more 100-500 nanometer plastic nanoparticles get through than when it is intact, pointing to skin condition, not just particle size, as the deciding factor.
Moderate evidence.
Caveats
Single ex vivo study (Martin et al. 2024) with the barrier surgically removed; no data on how much a naturally inflamed, eczematous, sunburned, or detergent-irritated (but not surgically stripped) human barrier changes penetration in real use.
Source: Menichetti et al. 2025, Journal of Xenobiotics Link to this fact
Health effects / Measured in people or real products
In a small Swedish clinic study, people whose breathing symptoms were set off by scents reacted to a blinded perfume challenge with more symptoms than a saline placebo, even though a nasal clamp meant they could not smell either one, and their lung function, oxygen levels, heart rate, and breathing rate never changed at all.
Moderate evidence: N=9, single clinic, self-rated symptoms, perfume-vs-saline contrast single-blind (not double-blind); no healthy control group.
Source: Millqvist & Löwhagen 1996, Allergy Link to this fact
Health effects / Measured in people or real products
Wearing a mask with an active-carbon filter, the kind designed to remove odors and organic vapors from the air, did not stop a scent-triggered breathing reaction to perfume: symptom scores with the carbon filter were no different from a dummy filter.
Moderate evidence.
Caveats
N=9, one perfume, unmeasured chamber concentration, single clinic; this arm of the study was double-blind. The paper's own hypothesis is that the reaction may enter through the eyes, which the mask does not cover, so this does not test filters that also shield the eyes.
Source: Millqvist & Löwhagen 1996, Allergy Link to this fact
Health effects / Measured in people or real products
In the clinical case series that first defined "irritable larynx syndrome," perfume was a named symptom trigger in about a quarter of patients (10 of 39), well behind reflux (72%) and other airborne irritants (51%).
Moderate evidence: Single specialist voice clinic, retrospective, uncontrolled, n=39; self-reported triggers, multiple allowed per patient; not an occupational or fragrance-specific population.
Health effects / Measured in people or real products
About 85% of patients in the founding irritable larynx syndrome case series were judged to have gastroesophageal reflux contributing to their laryngeal symptoms, making reflux the single most common named trigger and etiologic factor, ahead of any airborne trigger including perfume.
Moderate evidence.
Caveats
Single clinic, n=39, reflux mostly diagnosed by symptoms/history rather than pH monitoring or gastroscopy; the paper's own abstract rounds this to ">90%", but the Results table gives 33/39 = 84.6%.
Health effects / Measured in people or real products
In a small double-blind lab test, a drug that numbs sensory nerves (lidocaine) cut in half the number of coughs that chemically sensitive patients had after inhaling capsaicin, evidence that their airway sensitivity runs through nerve pathways rather than the kind of airway narrowing seen in asthma.
Moderate evidence.
Caveats
N=12, single Gothenburg clinic, same research lineage as most other evidence base capsaicin-challenge studies; the same lidocaine blockade also works in healthy subjects (cited within the paper), so it shows the pathway exists, not that it is disease-specific.
Health effects / Measured in people or real products
A blinded, controlled 30-minute lab exposure to two fragranced household aerosols caused no objective ocular redness or nasal congestion (acoustic rhinometry) change in mild or moderate asthmatics, and no drop in lung function (FEV1).
Moderate evidence.
Caveats
Industry-funded (SC Johnson, co-authored by an SC Johnson employee); single acute exposure to two specific products; asthmatics were unselected for fragrance sensitivity; severe asthma excluded by design.
Source: Opiekun 2003, Clinical and Experimental Allergy Industry-funded Link to this fact
Health effects / Measured in people or real products
Moderate asthmatics self-reported more nasal congestion after one of two fragranced aerosols, but objective nasal-airway measurement (acoustic rhinometry) on the same people at the same time found no change.
Moderate evidence.
Caveats
A real self-report/physiology mismatch in a single study; moderate-asthmatic subgroup sizes were small (n=15-54 in some cells), which limits power to detect a true objective effect if one exists.
Source: Opiekun 2003, Clinical and Experimental Allergy Industry-funded Link to this fact
Health effects / Review of other studies
Across 6 independent lab studies using a forced-choice 'which one is the real chemical?' test, people who report multiple chemical sensitivities were not better than other people at telling active chemical exposures apart from blank or sham ones.
Moderate evidence.
Caveats
A systematic review, not new data; the 6 studies span different research groups, chemicals, and blinding tiers but are heterogeneous (worker cohorts, clinic patients, different named subgroups pooled together).
Source: Das-Munshi 2006, Journal of Allergy and Clinical Immunology Link to this fact
Health effects / Measured in people or real products
In one lab study of 180 people, simply telling them beforehand that an odor was an 'industrial solvent' (versus a 'relaxing natural extract' or a neutral label) made them report significantly more irritation and symptoms from the identical odorant, and the effect was strongest in people who normally rate odors as more reactive/irritating.
Moderate evidence: Reported via Das-Munshi's review, not read from the Dalton 1999 primary source directly; acquire Dalton 1999 (Health Psychol 18:579) to verify before treating as primary-grade.
Source: Das-Munshi 2006, citing Dalton 1999, Journal of Allergy and Clinical Immunology Link to this fact
Health effects / Self-reported survey
In a large random sample of Swedish adults (Vasterbotten, n=3,406), 12.2% said they get symptoms from odorous or pungent chemicals such as perfumes and cleaning agents in doses that don't bother most other people, but only 3.3% had actually been given a doctor's diagnosis of a chemical-intolerance condition.
Moderate evidence: Single region (Vasterbotten county), not national; diagnosis status is itself self-reported, not verified against medical records.
Source: Palmquist 2014, International Journal of Hygiene and Environmental Health Link to this fact
Health effects / Self-reported survey
People who report being intolerant to one environmental factor (chemicals, certain buildings, electromagnetic fields, or everyday sounds) are far more likely than chance to also report intolerance to another — for example, self-reported chemically intolerant Swedes were about six times more likely than average to also report building-related symptoms (observed 96 vs. 19.5 expected under chance).
Moderate evidence: Cross-sectional single-item design; cannot separate a shared physiological mechanism from a shared symptom-reporting/attribution style or shared exposure circumstances.
Source: Palmquist 2014, International Journal of Hygiene and Environmental Health Link to this fact
Health effects / Measured in people or real products
In a small case series, an Uber driver and four Uber passengers developed facial and eyelid airborne contact dermatitis linked to fragrance diffusers used inside the cars; all five patients had positive patch tests to fragrance mix I and/or balsam of Peru, and the driver's dermatitis improved after he removed his own diffuser.
Moderate evidence.
Caveats
Only 5 patients from one clinic; published in the journal's 'Contact Points' section, which the journal states is exempt from mandatory peer review (editor review only); no chemical analysis of the diffusers, no air sampling, no control group; diffuser-removal resolution documented for the driver only, not the 4 passenger cases.
Source: Perper et al. 2017, Contact Dermatitis Link to this fact
Health effects / Measured in people or real products
In an unblinded human chamber study, 23 volunteers exposed to about 250 ppm propylene glycol monomethyl ether (PGME) vapor for 1-7 hours judged the odor objectionably unpleasant, and 20 of 23 developed eye irritation within about an hour, with some needing to leave early due to severe lacrimation; no subject at any tested concentration up to 2,050 ppm showed impaired vision, coordination, or reaction time.
Moderate evidence.
Caveats
Design is completely unblinded (no placebo or masking condition) and reports no objective ocular measurements (redness, corneal thickness, tear-film, blink counts); a later double-blind study with objective measures (Emmen 2003) found much smaller effects but only tested up to 150 ppm, so it does not directly re-test this 250 ppm finding.
Source: Stewart 1970 (Arch Environ Health), Archives of Environmental Health: An International Journal Industry-funded Link to this fact
Health effects / Measured in people or real products
At 100 ppm PGME vapor, most of 6 subjects found the odor intolerable on first exposure, but odor tolerance developed within 25 minutes and all judged it mild; only 1-2 of the 6 reported mild or slight eye irritation over the following two hours, with no measurable effect on vision, coordination, or neurological tests.
Moderate evidence: Unblinded, n=6, no objective eye measurements; a double-blind study at the same concentration (Emmen 2003) found no eye effects at all, subjective or objective.
Source: Stewart 1970 (Arch Environ Health), Archives of Environmental Health: An International Journal Industry-funded Link to this fact
Health effects / Self-reported survey
In a nationally representative German survey, one in five adults (19.9%) report at least one health effect from fragranced products, and more than half of them (55.3%) report respiratory problems, not skin problems — the focus of most prior German fragrance research.
Moderate evidence: Self-report; no clinical confirmation; n=1,102, Germany 2019; same dataset as Klaschka 2020b.
Source: Klaschka 2020a, Environmental Sciences Europe Link to this fact
Health effects / Self-reported survey
People who already blame fragrances for their own symptoms do not use less of them: 98.6% of Germany's fragrance-sensitive survey respondents are still exposed to fragrances from their own weekly use, slightly above the general population's 96.9%.
Moderate evidence: Self-report; cannot rule out that affected people simply notice or report their own exposure more accurately rather than actually using more.
Source: Klaschka 2020a, Environmental Sciences Europe Link to this fact
Health effects / Self-reported survey
Among adults with a diagnosis of autism or autism spectrum disorder, 71.4% (US) and 51.2% (Australia) report health problems from the scent of laundry products coming from a dryer vent — the highest prevalence of any group measured (general population, asthmatics, autistic adults, migraineurs) in either country.
Moderate evidence.
Caveats
These country-specific figures are restated from Steinemann's 2018 autistic-adults survey, which this evidence base has not yet independently ingested and read in full; the primary's own main-text tables report only a pooled 3-country figure (57.5%) for this specific exposure item, so the US/AU split has not been directly verified against the primary's own tables here (only checked for internal arithmetic consistency).
Source: Goodman, Nematollahi & Steinemann 2020/2021, Air Quality, Atmosphere & Health Link to this fact
Health effects / Self-reported survey
Among people who get migraines, 40.8% (US) and 29.1% (Australia) report health problems from the scent of laundry products coming from a dryer vent, with asthma attacks the most frequently reported effect in this group.
Moderate evidence.
Caveats
Restated from Steinemann & Nematollahi's 2020 migraine survey, which this evidence base has not yet independently ingested; self-report only, no clinical migraine diagnosis confirmation described in this source.
Source: Goodman, Nematollahi & Steinemann 2020/2021, Air Quality, Atmosphere & Health Link to this fact
Health effects / Review of other studies
Quaternary ammonium compounds such as benzalkonium chloride (BAC) and didecyldimethylammonium chloride (DDAC) have been linked to occupational asthma, contact dermatitis, and urticaria in healthcare workers, professional cleaners, and hairdressers with heavy, repeated exposure to disinfectants, antiseptics, and cosmetic preservatives.
Moderate evidence.
Caveats
Evidence is case reports, small cohorts, and a few specific-inhalation-challenge studies of occupational/disinfectant/cosmetic exposure, not laundry products; two cited cohort studies of QAC and healthcare-worker asthma directly disagree (see contradictions).
Source: Peyneau 2022, Frontiers in Toxicology Link to this fact
Health effects / Other evidence
In mice, repeated exposure to a specific odorant made the sense of smell more sensitive to that exact odorant, and the change arose at least partly in the smell-sensing tissue of the nose (the olfactory epithelium), not only in the brain.
Moderate evidence.
Caveats
Single-lab mouse study (though it replicates the same group's own earlier work); only two odorants tested; the boosted sensitivity faded back to baseline after about 4-7 months; 'sensitivity' means detection threshold for a smell, not symptom severity, irritation, or an allergic-type reaction. Mouse-only, single lab (Monell); human relevance to MCS not tested here.
Source: Yee & Wysocki 2001, Physiol Behav 72:705-711, Physiology & Behavior Link to this fact
Health effects / Other evidence
In mice, odor detection thresholds were not fixed: 10 days of exposure to one odorant sharpened detection of that odorant 3- to 20-fold without changing sensitivity to an unexposed odorant.
Moderate evidence.
Caveats
Animal, olfactory-only data; does not test trigeminal chemesthesis or human multiple-chemical-sensitivity symptoms, so it does not by itself settle whether normal human olfactory thresholds can rule out chemical intolerance. Mouse-only, single lab (Monell); human relevance to MCS not tested here.
Source: Yee & Wysocki 2001, Physiol Behav 72:705-711, Physiology & Behavior Link to this fact
Health effects / Measured in people or real products
In a lab EEG study, people who scored high on a validated chemical-sensitivity questionnaire showed brainwave (ERP) evidence of difficulty disengaging attention from a pungent gas (CO2) and an odorant (amyl acetate) specifically, not from a neutral tone, when told to ignore them and focus on another task.
Moderate evidence.
Caveats
Single lab study, n=21 chemically sensitive vs 17 nonsensitive Swedish adults defined by one self-report questionnaire (CSS-SHR), not a clinical MCS diagnosis; not independently replicated; ERP habituation comparison across only two session halves is a crude proxy per the authors' own caveat.
Source: Andersson 2009, Journal of Psychosomatic Research Link to this fact
Health effects / Measured in people or real products
The same chemically sensitive group did not habituate to repeated chemical exposure over 1.5 hours (perceived intensity of CO2 and an ERP amplitude marker stayed flat), while a nonsensitive comparison group habituated normally on both measures.
Moderate evidence.
Caveats
Small single-lab sample (n=21/17, fewer after ERP exclusions); the CS group is defined by a questionnaire cutoff, not a clinical diagnosis; sensitization was chemosomatosensory (CO2)-specific and did not extend to the olfactory (amyl acetate) magnitude-estimation measure.
Source: Andersson 2009, Journal of Psychosomatic Research Link to this fact
Health effects / Review of other studies
In human nasal provocation studies, hypertonic saline triggers substance P release within the first 3 minutes and dose-dependent glandular mucus secretion over the following 3-5 minutes, but produces no measurable increase in nasal-lavage albumin (a marker of vascular permeability) — a pattern reproduced in normal volunteers, acute allergic rhinitis, and acute and chronic rhinosinusitis patients.
Moderate evidence: This review summarizes the authors' own earlier provocation trials rather than presenting independent replication; sample sizes and statistics are not restated in this review.
Source: Baraniuk & Merck 2009, Annals of the New York Academy of Sciences Link to this fact
Health effects / Other evidence
The human nasal 'axon response' that follows hypertonic saline provocation drives glandular secretion but, unlike neurogenic inflammation described in rat trachea and other rodents, produces no vascular-permeability (plasma leak) component in human nasal mucosa — a stated species dissociation between rodent and human neurogenic inflammation.
Moderate evidence.
Caveats
Based on the authors' own human provocation data set against separately cited rodent literature, not a head-to-head experiment; the review does not name a specific rodent study for the vascular-leak side of the comparison.
Source: Baraniuk & Merck 2009, Annals of the New York Academy of Sciences Link to this fact
Health effects / Other evidence
Histamine nasal provocation produces itch, increased vascular permeability, and cholinergic reflex-driven glandular secretion (the classic allergic-rhinoconjunctivitis pattern), but histamine-induced 'axon responses' — direct local neuropeptide release from sensory nerve terminals — have not been clearly demonstrated in human airway mucosa.
Moderate evidence: This is an absence-of-evidence statement from a narrative review; the authors do not report having done a systematic search for such studies.
Source: Baraniuk & Merck 2009, Annals of the New York Academy of Sciences Link to this fact
Health effects / Review of other studies
A family of temperature- and chemical-gated TRP ion channels spans the range of stimuli relevant to breathing: TRPV2 detects tissue-damaging heat (>=52C), TRPV1 detects warmth/capsaicin/low pH/ethanol (>=42C), TRPV3/TRPV4 detect normal ambient temperature and osmotic/mechanical changes (roughly 22-40C), TRPM8 detects menthol and cool temperatures (8-22C), and TRPA1 detects very cold temperatures, mustard oil, garlic isothiocyanates, and THC (below 8C) but not menthol.
Moderate evidence: This is a conceptual/diagrammatic synthesis of previously published receptor characterization studies, not new data generated in this paper.
Source: Baraniuk & Merck 2009, Annals of the New York Academy of Sciences Link to this fact
Health effects / Review of other studies
TRPV1 is expressed well beyond nociceptive nerves, including on airway, skin, gastrointestinal, bladder, and rectal epithelial cells, plus the substantia nigra, hippocampus, and hypothalamus — meaning capsaicin and other TRPV1 ligands can act directly on epithelium without involving nerves at all.
Moderate evidence: Synthesized from cited prior localization studies, not measured in this paper.
Source: Baraniuk & Merck 2009, Annals of the New York Academy of Sciences Link to this fact
Health effects / Review of other studies
Cold, dry air causes dose-dependent nasal airflow obstruction in people with idiopathic nonallergic rhinitis but not in healthy controls, while separately, low-dose chlorine that produces no perceptible sensory irritation can still cause neurogenic nasal airflow obstruction with no detectable neuropeptide release — pointing to a non-classical, local mucosal mechanism distinct from typical axon-response neurogenic inflammation.
Moderate evidence.
Caveats
Both findings are drawn from single cited primary studies rather than replicated within this review; the mechanism proposed for the chlorine effect is explicitly called a hypothesis by the review's authors.
Source: Baraniuk & Merck 2009, Annals of the New York Academy of Sciences Link to this fact
Health effects / Review of other studies
A systematic review of 27 human chemical-exposure challenge studies on multiple chemical sensitivity (1998-2015) found no scientific consensus on whether the reaction is physical (toxicological) or psychological: only a minority of studies concluded toxicological causation, a similar minority concluded psychological causation, and the single largest group of studies reached no conclusion at all.
Moderate evidence.
Caveats
Narrative bibliographic review, not a PRISMA-registered systematic review; no formal quality scoring of the 27 underlying studies, which used heterogeneous designs (facemask aerosol, exposure chamber, sticks, bottle, dynamic olfactometer) and small samples (as few as 8 per arm); classification into toxicological/psychological/no-conclusion groups is the review authors' own qualitative judgment, not a quantitative synthesis.
Source: Rossi 2018, Journal of Occupational & Environmental Medicine Link to this fact
Health effects / Self-reported survey
Among 129 clinically diagnosed MCS patients in Rome, hyperosmia and asthenia were the most common symptoms (about 97% and 83% of patients) and clustered together statistically, followed by respiratory symptoms (dyspnea, cough, tachypnea; about 72%) and neurocognitive symptoms (confusion, attention/memory/decision-making deficits; about 62%).
Moderate evidence: Single-center clinical (not population) sample of physician-diagnosed patients; no control group; no exposure measurement; describes symptom co-occurrence, not cause.
Source: Del Casale 2020 Link to this fact
Health effects / Review of other studies
Across 13 independent qualitative studies from four countries, people with multiple chemical sensitivity consistently describe being kept out of everyday places by scented/chemical products, and losing jobs, income, or social relationships because of it.
Moderate evidence.
Caveats
Systematic review of qualitative (not survey) evidence; meta-aggregation of 53 findings from 13 small, mostly clinic/support-group/self-selected-website-recruited studies (sample sizes 4-203), none rated top quality by the review's own NICE checklist. Converges in direction with, but does not quantitatively confirm, Steinemann's self-report survey percentages already in our evidence base.
Source: Driesen 2020, Journal of Psychosomatic Research Link to this fact
Health effects / Measured in people or real products
In a controlled study of 7 asthmatic subjects, 0.5 ppm sulfur dioxide alone and cold, dry air alone each caused no significant airway narrowing, but breathing the two together raised specific airway resistance more than 3-fold (6.94 to 22.35 l*cmH2O*l-1*s, P<0.001), with worsened wheeze or breathlessness in 6 of 7 subjects.
Moderate evidence.
Caveats
Small single-lab study (n=7); mouthpiece breathing bypasses the nose's normal filtering/warming/humidifying function, so real-world nasal-breathing exposure may show a smaller effect; sulfur dioxide is an industrial/combustion gas, not a fragrance chemical.
Source: Bethel 1984 Link to this fact
Health effects / Measured in people or real products
In asthmatic subjects, self-rated shortness-of-breath severity correlated only loosely with objectively measured airway resistance after inhaling irritant stimuli; the categories overlapped so much that a person's own symptom report gave little information about how much bronchoconstriction had actually occurred.
Moderate evidence: Small study (n=7), single ordinal symptom scale; consistent with our broader point that self-reported symptom severity and objective physiological response are different measures.
Source: Bethel 1984 Link to this fact
Health effects / Measured in people or real products
In a controlled chamber study of 85 volunteers, people with mild asthma and people with moderate/severe (medication-dependent) asthma showed similar increases in airway narrowing and symptoms per increment of sulfur dioxide exposure (0.2-0.6 ppm); clinical asthma severity did not reliably predict who reacted most to the gas itself, though moderate/severe asthmatics had a higher baseline (clean-air) symptom and airway-resistance level.
Moderate evidence.
Caveats
Single-lab study (n=16 mild, n=24 moderate/severe asthmatics); sulfur dioxide is a combustion/industrial irritant gas, not a fragrance chemical; the severity comparison itself was underpowered and not statistically significant.
Source: Linn et al. 1987, American Review of Respiratory Disease Link to this fact
Health effects / Measured in people or real products
An individual's sensitivity to a given dose of an irritant gas (sulfur dioxide) was highly reproducible when the same person was retested 1-7 weeks later (r=0.85 across all subjects, r=0.82-0.74 within asthmatic subgroups), even though that sensitivity could not be predicted well from a clinical asthma diagnosis, allergy skin testing, or serum IgE level.
Moderate evidence: Single-lab study, n=85 total; reproducibility was much weaker for exercise responses measured in clean air alone.
Source: Linn et al. 1987, American Review of Respiratory Disease Link to this fact
Health effects / Measured in people or real products
Atopic (allergic, IgE-sensitized) volunteers without asthma were essentially as unresponsive to sulfur dioxide exposure as non-atopic normal volunteers; airway hyperreactivity (asthma), not IgE-mediated allergy/atopy, was what predicted reaction to this irritant gas.
Moderate evidence.
Caveats
Single-lab study, n=21 atopic subjects; sulfur dioxide is a non-allergenic irritant gas, so this speaks to the general atopy-vs-hyperreactivity question rather than to any allergic mechanism for SO2 itself.
Source: Linn et al. 1987, American Review of Respiratory Disease Link to this fact
Health effects / Test-chamber measurement
Telling people a real chemical exposure is a harmful industrial solvent versus a beneficial natural extract changed how irritating they rated identical 800 ppm acetone air and how many health symptoms (throat/nasal irritation, lightheadedness, nausea, drowsiness) they reported afterward, even though every group breathed the same concentration.
Moderate evidence: Single lab, single cohort, not independently replicated; co-funded by the Chemical Manufacturers' Association alongside NIH grants (coi/industry-funded).
Source: Dalton et al. 1997, Int Arch Occup Environ Health 69:407-417 Industry-funded Link to this fact
Health effects / Other evidence
Objective, blinded odor-detection thresholds for acetone and butanol did not differ between subjects given positive, negative, or neutral framing of the same acetone exposure, even though their subjective irritation and symptom ratings differed sharply by framing.
Moderate evidence: Single lab, single cohort; thresholds tested olfactory (CN I) sensitivity only, not trigeminal chemesthesis.
Source: Dalton et al. 1997, Int Arch Occup Environ Health 69:407-417 Industry-funded Link to this fact
Health effects / Measured in people or real products
In a small Swedish clinic study, people whose breathing symptoms were triggered by scents and chemical irritants — but who had normal lung function and negative allergy and asthma tests — coughed much more after inhaling capsaicin (the hot-pepper compound) than either people with verified asthma or healthy volunteers (93 vs 23 average coughs at the highest dose, P<0.001).
Moderate evidence.
Caveats
Small (n=10 per patient group, n=28 healthy controls), single-clinic study; capsaicin dosing was open-label (patients not told doses would increase) rather than placebo-blinded, so it is a correlational patient-vs-comparator design, not an intervention.
Source: Millqvist, Bende & Löwhagen 1998, Allergy Link to this fact
Health effects / Measured in people or real products
The extra coughing in these chemically sensitive patients could not be explained by asthma: their lung function, methacholine reactivity, and skin-prick allergy testing were all normal, yet they still coughed dose-dependently to inhaled capsaicin, and even to a plain saline control, while a separate group of people with verified asthma did not cough more than healthy volunteers.
Moderate evidence.
Caveats
Single small clinic sample (n=10 per patient group); normal methacholine and skin-prick tests were inclusion criteria for the chemically sensitive group, not an independent finding in an unselected population.
Source: Millqvist, Bende & Löwhagen 1998, Allergy Link to this fact
Health effects / Government agency
As of 1998, EPA's own environmental-health office stated that certain types of childhood cancer were increasing but the causes were largely unknown, and that the agency lacked a reliable basis to determine whether most untested high-production-volume industrial chemicals were safe or unsafe for children.
Moderate evidence.
Caveats
A general statement about data gaps as of 1998, not a study establishing or ruling out any specific chemical-childhood-cancer link; the paper explicitly states EPA does not know how many, if any, childhood cancers are caused by these exposures.
Source: Goldman 1998, Environmental Health Perspectives Link to this fact
Health effects / Self-reported survey
People with asthma were about 4 times more likely to report a doctor's diagnosis of environmental illness/multiple chemical sensitivity than people without asthma (19.0% vs. 4.6%), and over twice as likely to self-report unusual sensitivity to everyday chemicals (31.5% vs. 13.9%).
Moderate evidence.
Caveats
Cross-sectional association from a single population survey with no independent replication cited here; does not establish that asthma causes or is caused by chemical sensitivity, only that they co-occur more than chance in self-report.
Source: Kreutzer, Neutra & Lashuay 1999, American Journal of Epidemiology Link to this fact
Health effects / Self-reported survey
Women were about 1.6 times more likely than men to self-report sensitivity to more than one everyday chemical, after adjusting for other factors, though doctor-diagnosed MCS itself was not significantly associated with sex.
Moderate evidence.
Caveats
Single population survey; unadjusted sex ratios for this item (F:M ~2.3) sit well below the ~4:1 ratio later found in a national survey using the same question wording, so the sex effect size varies by sample.
Source: Kreutzer, Neutra & Lashuay 1999, American Journal of Epidemiology Link to this fact
Health effects / Self-reported survey
Reports of doctor-diagnosed multiple chemical sensitivity and self-reported chemical sensitivity were spread fairly evenly across race/ethnicity, income, education, marital status, and geography in a representative California sample, unlike clinic-based case series which have reported a predominance of middle- to upper-class white women.
Moderate evidence.
Caveats
Homogeneous distribution is consistent with either a real physiologic mechanism or widespread cultural apprehension about chemicals; the study explicitly does not distinguish between these explanations.
Source: Kreutzer, Neutra & Lashuay 1999, American Journal of Epidemiology Link to this fact
Information, not medical advice. See also: myths we won’t tell you.