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 / Measured in people or real products
Scaled to a realistic 90-square-meter apartment burning 5 candles for 3.5 hours, the highest-soot candle type produced an estimated fine-particle (PM2.5) concentration of 70 micrograms per cubic meter, above the WHO 24-hour air quality guideline of 25 micrograms per cubic meter, while the estimated cancer-risk marker (benzo[a]pyrene equivalent) stayed at, not above, the WHO guideline in both the highest- and lowest-emitting candle scenarios tested.
Moderate evidence.
Caveats
Modeled estimate from measured chamber emission factors scaled with assumed apartment volume, air exchange rate, and deposition loss rate, not a direct measurement in an actual apartment; represents burning 5 candles at once, a higher-than-typical single-evening scenario.
Health effects / Measured in people or real products
When candle emissions were converted to realistic household exposure (a 30 m3 room, candle burning 4 hours a day, 4 days a week), most measured pollutants (carbon monoxide, carbon dioxide, formaldehyde, PM2.5, benzene, most VOCs) stayed below international indoor air guide values, but nitrogen dioxide, acrolein, and benzo[a]pyrene exceeded some (not all) of the guide values tested, depending on which health agency's reference value was used.
Moderate evidence.
Caveats
Acrolein exceeded the US EPA reference concentration in all 8 experiments where it was quantifiable (compliance factor 1.3-6.3); benzo[a]pyrene exceeded the WHO guide value (factor 0.32-6.35) but not the more lenient US EPA reference concentration; nitrogen dioxide exceeded a precautionary German guide value but not the WHO/EU long-term value. Single chamber study; guide-value comparisons vary substantially by which agency's number is applied.
Source: Salthammer 2021, Environment International Industry-funded Link to this fact
Health effects / Review of other studies
People with chemical intolerance or multiple chemical sensitivity react more strongly than healthy controls in controlled chamber exposures to irritant chemicals, with derived sensitivity factors typically 1 to 3 times higher and, in some studies, up to about 20 times higher.
Moderate evidence.
Caveats
Objective physiological confirmation was rare across the underlying studies; most of the signal is in subjective symptom ratings, which can be influenced by odor perception and expectation.
Source: Mangelsdorf et al. 2021, Int J Hyg Environ Health, International Journal of Hygiene and Environmental Health Link to this fact
Health effects / Review of other studies
Benzalkonium chloride (BAK), the preservative used in about 70% of eye-drop formulations, is toxic to eye-surface cells at the concentrations it's commonly used (0.02-0.04%), which is 4 to 8 times its own cited cytotoxicity threshold of about 0.005%.
Moderate evidence.
Caveats
Narrative review citing decades of separate cell-culture, animal, and clinical studies rather than new data; authored entirely by employees of Ocular Therapeutix, Inc. (coi/industry-funded), a maker of BAK-free sustained-release eye therapies, which also funded the paper's editorial assistance and open-access fee.
Source: Goldstein 2022, Eye Industry-funded Link to this fact
Health effects / Review of other studies
Glaucoma patients on chronic BAK-preserved eye drops have ocular surface disease at much higher rates (30-70% across cited studies) than similarly aged adults without glaucoma (5-30%), and even a single 7-to-14-day course of BAK-preserved drops caused goblet-cell loss and other surface damage in eyes that started out healthy.
Moderate evidence.
Caveats
Observational comparisons in glaucoma patients cannot fully separate BAK toxicity from other differences between glaucoma and non-glaucoma eyes (disease itself, other medications, dosing frequency); short-exposure findings come from small clinical studies cited secondhand.
Source: Goldstein 2022, Eye Industry-funded Link to this fact
Health effects / Lab study (cells or chemistry)
BAK's ocular toxicity is proposed to work partly through mitochondria: because BAK is positively charged and mitochondria are the only negatively charged compartment inside cells, cell studies cited in this review found BAK triggers oxidative stress and can inhibit mitochondrial function by more than 90%.
Moderate evidence: Mechanism drawn from in vitro/cell-culture studies cited by the review, not measured directly in intact human eyes.
Source: Goldstein 2022, Eye Industry-funded Link to this fact
Health effects / Review of other studies
Benzalkonium chloride (BAC), used at 0.005%-0.02% as an eye-drop preservative, is used at roughly 5-100 times that concentration (0.1%-1.0%) to deliberately induce dry-eye-disease signs in rabbit, mouse, and rat models.
Moderate evidence.
Caveats
The 0.005%-0.02% preservative-use figure is this paper's own background citation (Goldstein 2022, Baudouin 2010), not its primary data; the induction-dose comparison is animal-only.
Source: Thacker 2023, Indian Journal of Ophthalmology Link to this fact
Health effects / Animal study
Topical BAC induces dry-eye-disease signs in animals through a detergent-like mechanism: it disrupts the tear film's lipid layer and causes corneal/conjunctival epithelial cell apoptosis, mitochondrial dysfunction, and DNA single-strand breaks, with elevated pro-inflammatory cytokines (IL-6, IL-8, IL-1β, TNF-α) and reduced mucin/goblet cells.
Moderate evidence.
Caveats
Animal-only evidence (rabbit, mouse, rat), but the qualitative mechanism is independently replicated across at least five separate labs, which supports the general mechanism claim more than any single lab's numbers.
Source: Thacker 2023, Indian Journal of Ophthalmology Link to this fact
Health effects / Animal study
In a rabbit dose-comparison study, 0.2% BAC (twice daily for 2 weeks) sustained dry-eye-disease signs for 3 weeks after BAC was stopped, versus only 1-2 weeks for 0.1% and 0.15% BAC.
Moderate evidence: Single-lab original data from this paper's own unpublished experiment, not yet independently replicated at these exact concentrations.
Source: Thacker 2023, Indian Journal of Ophthalmology Link to this fact
Health effects / Animal study
In the same rabbit study, both the fluorescein staining score and the corneal smoothness index rose with BAC concentration after 2 weeks: about 3 at 0.1%, about 4 at 0.15%, and about 5 at 0.2% (both scored on a 0-5 scale).
Moderate evidence: Approx., read from Fig. 3b/3c bar charts; the paper does not print these exact scores in the text.
Source: Thacker 2023, Indian Journal of Ophthalmology Link to this fact
Health effects / Lab study (cells or chemistry)
In cultured human corneal stromal fibroblasts, benzalkonium chloride (BAC) reduced cell viability and impaired mitochondrial respiration at concentrations 10-100,000 times lower than the classic ocular-surface-toxicity threshold (as low as 10^-8%-10^-6% vs. ~5×10^-3%).
Moderate evidence.
Caveats
Single 2D human cell-culture model from only 2 donors, no in vivo confirmation, no measured intraocular BAC concentration from real dosing regimens; in-vitro finding, not a demonstrated clinical harm.
Source: Umetsu et al. 2024, Graefe's Archive for Clinical and Experimental Ophthalmology Link to this fact
Health effects / Lab study (cells or chemistry)
BAC dose-dependently suppressed mRNA expression of collagen I, collagen VI, and fibronectin in human corneal stromal fibroblasts, with collagen I significantly reduced even at the lowest concentration tested (10^-8%).
Moderate evidence: In vitro, 2D monolayer model, 2 donors; extracellular-matrix gene suppression has not been linked to a measured clinical outcome (e.g., corneal thinning) in this study.
Source: Umetsu et al. 2024, Graefe's Archive for Clinical and Experimental Ophthalmology Link to this fact
Health effects / Lab study (cells or chemistry)
BAC suppressed rather than activated the ER-stress/unfolded-protein-response pathway (GRP94, IRE1, PERK, ATF4, sXBP1, CHOP) in human corneal stromal fibroblasts, in a concentration-dependent manner, which the authors interpret as failure of an adaptive stress response rather than reassurance.
Moderate evidence: Interpretation (failed adaptive response) is the authors' own; ER-stress-gene suppression alone does not establish a clinical mechanism of corneal damage.
Source: Umetsu et al. 2024, Graefe's Archive for Clinical and Experimental Ophthalmology Link to this fact
Health effects / Measured in people or real products
No toxicological testing has been done on the specific nanoparticles formed during scented wax melt use; the authors explicitly call for future studies on their toxicological properties before drawing health conclusions.
Moderate evidence: This is the paper's own stated evidence gap, not a finding of harm or safety.
Source: Patra et al. 2025, Environmental Science & Technology Letters Link to this fact
Health effects / Lab study (cells or chemistry)
Exposing primary human bronchial epithelial cells (air-liquid interface culture) to a disinfectant-concentration (0.4%) benzalkonium chloride aerosol for 24 h produced 2,810 differentially expressed genes (1,475 up, 1,335 down), mapping onto DisGeNET disease networks for asthma, COPD, and pulmonary fibrosis; IL1RL1 and IL6R (upregulated) are independently reported as elevated in human asthma patients.
Moderate evidence.
Caveats
Single cell-culture model, one lab, no animal or human validation of the transcriptomic findings; the authors themselves call the existing evidence that BKC causes respiratory disease in humans 'insufficient' and frame this as hypothesis-generating.
Source: Kim et al. 2025, Ecotoxicology and Environmental Safety Link to this fact
Health effects / Lab study (cells or chemistry)
The two genes with the largest fold-change increases in BKC-exposed bronchial cells were both pulmonary-fibrosis-associated: FGF2 (11.4-fold) and GREM1 (19.4-fold).
Moderate evidence: No documented human case of BKC-caused pulmonary fibrosis exists; this is a gene-expression signal in one in vitro model, not a demonstrated disease outcome.
Source: Kim et al. 2025, Ecotoxicology and Environmental Safety Link to this fact
Health effects / Review of other studies
A narrative research review of 96 studies (32 human, 26 in vivo, 38 in vitro) on BKC respiratory toxicity found mixed clinical-trial results (some report impaired nasal ciliary motility and bronchoconstriction, others find no significant effect), a handful of occupational-asthma case reports, and animal-inhalation evidence of decreased tidal volume, inflammation, and histological changes resembling pulmonary fibrosis/obstructive lung disease; the authors conclude evidence that BKC causes respiratory disease in humans 'remains limited.'
Moderate evidence.
Caveats
A narrative (not systematic/PRISMA-registered) review by one research group; only 96 of 930 initially identified records were retained, and screening judgment calls are not independently reproducible from the published methods alone.
Source: Kim et al. 2025, Ecotoxicology and Environmental Safety Link to this fact
Health effects / Review of other studies
Hydrolyzed wheat protein used in cosmetics (facial creams, moisturizers, shampoos, gels) has caused immediate allergic reactions, including hives and anaphylaxis, in sensitized users when applied to the skin.
Moderate evidence.
Caveats
Relayed from a single cited study (Lauriere et al. 2006); this is a different mechanism (immediate hypersensitivity) from the delayed contact allergy that is the review's main topic.
Source: Brasch 2007, Current Opinion in Allergy and Clinical Immunology Link to this fact
Health effects / Review of other studies
Underarm body odor is mostly made by skin bacteria, not by sweat itself: apocrine-gland secretions are odorless when released, and only turn into the characteristic musk/urine smell after aerobic Corynebacteria metabolize odorless steroid precursors into 16-androstenes, or after bacteria break the goat-like acid 3-methyl-2-hexenoic acid free from its carrier proteins.
Moderate evidence: Review synthesis of older primary biochemistry literature, not this paper's own new data.
Source: Kippenberger 2012, Experimental Dermatology Link to this fact
Health effects / Review of other studies
A single gene variant (ABCC11 538G->A) largely explains why most East Asians have a faint, acidic body odor instead of the strong axillary odor typical of Caucasians and Africans, and the same variant also determines dry vs. wet earwax.
Moderate evidence: Review synthesis of a single genetics study, not replicated here.
Source: Kippenberger 2012, Experimental Dermatology Link to this fact
Health effects / Review of other studies
Several metabolic and infectious diseases have documented, specific body-odor signatures: trimethylaminuria smells of rotten fish in urine/sweat/breath/saliva/semen, diabetic ketoacidosis smells of acetone, maple syrup urine disease smells of caramelized sugar, and typhoid fever produces a musty or fresh-baked-bread body odor.
Moderate evidence: Compiled from decades-old, mostly single case-series/case-report literature, not systematically re-verified by this review.
Source: Kippenberger 2012, Experimental Dermatology Link to this fact
Health effects / Review of other studies
Body odor changes measurably with age, but studies disagree on which compound tracks it: one GC study found dimethyl sulfone, benzothiazole, and nonanal correlating with age (and did not detect 2-nonenal), while a separate Japanese study found 2-nonenal rising sharply after age 39, the compound blamed for the so-called old man smell.
Moderate evidence: The review attributes the discrepancy to likely diet/cohort differences (the Japanese cohort's diet was unreported), not a resolved mechanism.
Source: Kippenberger 2012, Experimental Dermatology Link to this fact
Health effects / Review of other studies
Diet measurably changes how pleasant other people rate a person's body odor: in a controlled study cited here, eating red meat decreased how pleasant axillary odor was rated.
Moderate evidence: Single cited study; effect size and diet duration not given in this review.
Source: Kippenberger 2012, Experimental Dermatology Link to this fact
Health effects / Lab study (cells or chemistry)
Among 62 individually tested deodorant ingredients, only fragrance chemicals (plus one insect-pheromone contaminant) showed any estrogenic activity; no carrier, stabilizer, emulsifier, plasticizer, antitranspirant, or other active ingredient (including triclosan and BHT) tested positive.
Moderate evidence: Only 62 of the many ingredients found by GC/MS screening were selected for individual testing, based on frequency of occurrence; ingredients not selected were not tested at all.
Health effects / Measured in people or real products
Acetaldehyde, a probable human carcinogen, rose far more than its boiling point alone would predict once candles were lit — one strawberry candle's on/off concentration ratio for acetaldehyde was 3,688-fold, the largest ratio of any of the 34 compounds tracked in the study.
Moderate evidence.
Caveats
Acetaldehyde stayed far below the ACGIH/OSHA guideline values in absolute terms (max 640 ppb vs. 25,000/200,000 ppb guidelines); this finding is about combustion massively amplifying this one compound's relative output, not about it reaching a regulatory threshold.
Source: Ahn et al. 2015, Journal of Hazardous Materials Link to this fact
Health effects / Measured in people or real products
Kim et al. 2016 found their own original 2015 emission factors for carbonyl compounds from scented candles (formaldehyde and related aldehydes) had been computationally underestimated by exactly a factor of 250, due to a missed sample-dilution correction; the revised formaldehyde emission factor for the highest-emitting candle (SB-on) rose from 383 ng/g to 95.7 ug/g.
Moderate evidence: Single lab, self-corrected; independent replication of the corrected emission factors has not been reported.
Source: Kim et al. 2016, Journal of Hazardous Materials Link to this fact
Health effects / Measured in people or real products
A Korean government risk assessment of 64 deodorants found that at the maximum concentrations actually measured in the products, none of the 19 fragrance and biocidal ingredients carried to detailed (tiered 2) assessment exceeded its margin-of-exposure or hazard-quotient safety threshold, for either the inhalation or the dermal route.
Moderate evidence.
Caveats
A single government risk-assessment study using modeled Korean population-average exposure factors and mostly-animal (rat) toxicity reference values with standard uncertainty factors, not a clinical safety trial; not independently replicated.
Source: Kim et al. 2018, Science of The Total Environment Link to this fact
Health effects / Review of other studies
Fragrance chemicals have been reported to cause immediate-type contact urticaria, photosensitivity reactions, and rarer miscellaneous effects such as depigmentation and orofacial granulomatosis, but the review characterizes all of these as currently rare and not causing significant clinical problems.
Moderate evidence: Causal relationship for the miscellaneous effects in Table 13 is explicitly stated by the review as "not always established beyond doubt" (mostly case reports).
Source: de Groot 2020, Dermatitis 31(1):13-35, Dermatitis® Link to this fact
Health effects / Animal study
Scented candle fumes alone, at an occupational-intensity chamber dose, significantly increased inflammatory cytokines (TNF-alpha, IL-6, IL-1beta) and NF-kB activity in the lungs of mice, and caused measurable lung histological injury, without significantly raising heart inflammatory markers, heart NF-kB, cardiac injury markers (LDH, CPK, CKMB), plasma BNP, or heart HIF-1alpha above unexposed controls.
Moderate evidence.
Caveats
Single mouse study, N=8 per group, no independent replication; the candle-only lung effect is a within-study finding, but the same paper found no cardiac effect from candle fumes alone, so the finding is respiratory-specific in this model.
Source: Chandrasekaran 2021, Current Research in Toxicology Link to this fact
Health effects / Animal study
In mice, chronic social-disruption stress combined with scented candle fume exposure produced significantly worse cardiopulmonary inflammation and injury (BALF/lung/heart cytokines, NF-kB activity, serum LDH/CPK/CKMB, plasma BNP, heart HIF-1alpha, and lung histological injury) than either stress or candle fumes alone, though no group showed significant heart tissue histopathology.
Moderate evidence.
Caveats
Single lab, single mouse cohort (N=8/group), no independent replication; the stress model (repeated social defeat) and candle-fume dose are both occupational-intensity, so the additive effect size may not generalize to lower-intensity human exposures.
Source: Chandrasekaran 2021, Current Research in Toxicology Link to this fact
Health effects / Advocacy group testing
Six substances found in the tested perfumes are classed by this report as suspected endocrine disruptors: benzyl salicylate, BHT, butylphenyl methylpropional, ethylhexyl methoxycinnamate, ethylhexyl salicylate, and octocrylene.
Moderate evidence: 'suspected' means the substance appears on one or more screening/priority lists, not that endocrine disruption in humans at perfume-use doses has been demonstrated.
Source: Tegengif 2022 Link to this fact
Health effects / Advocacy group testing
Butylphenyl methylpropional, found in several of the tested perfumes, is the one substance in this report's list of 26 that is also classed as reprotoxic.
Moderate evidence.
Source: Tegengif 2022 Link to this fact
Health effects / Self-reported survey
Among 472 Saudi university students who used scented candles, 24.8% reported at least one adverse health symptom in the past 12 months, most commonly headache (15.2%), shortness of breath (8.9%) and cough (7.8%); skin irritation was rare (1.3%).
Moderate evidence.
Caveats
Cross-sectional self-report; exposure and symptoms measured simultaneously so recall bias and reverse causation cannot be excluded, and no chemical/particulate exposure was measured alongside symptoms.
Source: Al Khathlan et al. 2023, BMC Public Health Link to this fact
Health effects / Self-reported survey
Among the same scented-candle users, respiratory-specific complaints in the past 12 months included waking with chest tightness (36.0%), waking from a coughing attack (33.1%), nasal allergy/hay fever (25.0%), a diagnosed respiratory illness (16.3%), a shortness-of-breath attack on waking (17.2%), wheezing (14.6%) and an asthma attack (9.5%).
Moderate evidence: Self-reported, non-clinically-confirmed symptoms in a non-representative convenience sample; no comparison group of non-candle-users was asked the same respiratory questions.
Source: Al Khathlan et al. 2023, BMC Public Health Link to this fact
Health effects / Lab study (cells or chemistry)
An academic in vitro toxicology screen of Lilial (butylphenyl methylpropional, the fragrance chemical banned from EU cosmetics in 2022) found no cytotoxicity, no mutagenicity in a mammalian gene mutation assay, no nephrotoxicity, and no NF-κB or NRF2 stress-pathway activation, in either Lilial itself or its major liver-generated (S9) metabolites, up to 100 µM.
Moderate evidence.
Caveats
Single academic lab, one round of testing, not yet independently replicated; a genotoxicity endpoint (γH2AX) in the same paper was inconclusive, not clean negative, so 'no adverse effect' does not extend to every endpoint tested.
Source: Jablonská 2023, Sci Rep 13:18536, Scientific Reports Link to this fact
Health effects / Lab study (cells or chemistry)
Using OECD/EPA guideline-validated receptor-transactivation assays (HeLa9903 for estrogen receptor, MDA-kb2 for androgen receptor), Lilial and its major liver metabolites showed no estrogen- or androgen-receptor agonist activity from 1 nM to 100 µM; the study's authors said they were unable to confirm the European Chemicals Agency's earlier suspicion that Lilial is an endocrine disruptor.
Moderate evidence: Single academic lab; this directly conflicts with a weaker, non-guideline finding from a different assay type (MCF7 proliferation) for the same compound — see contradiction.
Source: Jablonská 2023, Sci Rep 13:18536, Scientific Reports Link to this fact
Health effects / Lab study (cells or chemistry)
The mutagenic/genotoxic evidence for Lilial is genuinely mixed: bacterial Ames tests were mostly negative (one strain's positive result did not reproduce in a confirmatory test), a mammalian CHO/HPRT gene-mutation assay found no mutagenicity up to 500 µM with metabolic activation, but a DNA-strand-break (γH2AX) assay in the same paper gave results the authors themselves call inconclusive rather than negative.
Moderate evidence: The γH2AX result depended on which of two readouts (fold-change vs. area-normalized fluorescence) was used, and these disagreed in most of the conditions tested.
Source: Jablonská 2023, Sci Rep 13:18536, Scientific Reports Link to this fact
Health effects / Other evidence
Polycyclic aromatic hydrocarbons identified as carcinogens, including naphthalene, anthracene, and pyrene, have been found in candle fumes from wax, aroma substances or combustion dyes.
Moderate evidence: Directionally consistent with our own chamber PAH data (Salthammer 2021, Andersen 2021), though the editorial's own citation (Orecchio 2011) is not itself in this evidence base.
Source: Nazir et al. 2024, Annals of Medicine & Surgery Link to this fact
Health effects / Animal study
In rats, burning a scented candle 1-6 hours/day for 8 weeks in a sealed room raised serum TNF-alpha to about 115-220 pg/mL and IL-6 to about 120-215 pg/mL (both approximate, read from Fig. 3), versus about 30 and 50 pg/mL in fresh-air controls, and caused progressively worse lung damage on histopathology, reaching severe injury (fibrosis, necrobiotic changes, dense inflammatory infiltration) after as little as 3 hours/day.
Moderate evidence.
Caveats
Single lab, one scented candle product (brand/composition undisclosed), n=6 rats/group; values for cytokines and oxidative-stress markers are approximate, read from bar charts rather than printed as numbers in the text.
Source: Mohamed et al. 2025, Frontiers in Public Health Link to this fact
Health effects / Government agency
An international expert working group convened by the World Health Organization and UN Environment Program concluded that close to 800 chemicals are known or suspected to interfere with hormone receptors, hormone synthesis, or hormone conversion, but only a small fraction of them have been tested for whether they actually cause overt endocrine effects in an intact organism.
Moderate evidence.
Caveats
This is a 2012 expert-consensus count covering every kind of suspected endocrine disruptor (industrial chemicals, pesticides, plastics, cosmetic ingredients, natural compounds), not a fragrance-specific count, and the report is now well over a decade old.
Source: WHO/UNEP 2012 Link to this fact
Health effects / Government agency
Animal studies show that mixtures of endocrine-disrupting chemicals can produce additive effects even when every chemical in the mixture is present at a level too low, on its own, to cause any observable effect, meaning many 'individually safe' exposures could add up to a harmful one.
Moderate evidence.
Caveats
Based on animal mixture studies cited in the report; the report itself states there are no data showing how real-world mixtures of hundreds of EDCs at low concentrations actually affect human or wildlife health.
Source: WHO/UNEP 2012 Link to this fact
Health effects / Government agency
Hormone-disrupting chemicals often do not follow a simple 'more dose equals more effect' pattern; some show non-monotonic dose-response curves, so standard toxicology testing at high doses, divided by a safety factor to estimate a low 'safe' dose, may miss real effects that occur only at low doses.
Moderate evidence: This is the report's own methodological critique of standard regulatory toxicology testing, not a finding about any specific chemical.
Source: WHO/UNEP 2012 Link to this fact
Health effects / Other evidence
A widely cited 2012 review (Vandenberg et al.) cataloged many examples of endocrine-disrupting chemicals producing non-monotonic (U-shaped) dose-response curves; an independent 2013 Danish government reanalysis of that same catalog found that 45% of its cited in vitro examples were artifacts of general cell toxicity rather than true hormone-mediated effects, and only 5 of 34 cited in vivo (whole-animal) examples showed clear evidence of a genuine non-monotonic response.
Moderate evidence.
Caveats
Known only secondhand through an industry-funded critique paper with a direct interest in this conclusion; the original Vandenberg et al. 2012 review and the DTU Food 2013 reanalysis should be read as primaries before this is treated as settled.
Source: Lamb et al. 2014, Regulatory Toxicology and Pharmacology Industry-funded Link to this fact
Health effects / Other evidence
The WHO-UNEP 2012 report changed its conclusion on PCBs and DDT causing adrenocortical hyperplasia (Cushing-like adrenal gland disease) in Baltic seals from 'uncertain association' in the WHO's 2002 report to 'sufficient evidence' of a causal link in 2012, without citing any new supporting data, while not addressing contrary findings that tie similar adrenal changes in other marine mammals to chronic stress rather than chemical exposure, and rat studies finding the opposite direction of adrenal effect (atrophy, not hyperplasia) from the same chemicals.
Moderate evidence: From an industry-funded critique paper; the underlying wildlife and rat studies have not yet been read as primaries by this evidence base.
Source: Lamb et al. 2014, Regulatory Toxicology and Pharmacology Industry-funded Link to this fact
Health effects / Other evidence
The WHO-UNEP 2012 report's claim that the rise in autism spectrum disorder (ASD) prevalence is at least partly caused by environmental factors, including endocrine-disrupting chemicals, rests on only two citations (from 1976 and 2007); separate, non-industry-affiliated research attributes much of the reported increase in ASD prevalence to changes in diagnostic criteria, broader case ascertainment, and diagnostic substitution (children previously diagnosed with intellectual disability now diagnosed with ASD) rather than to any confirmed environmental cause.
Moderate evidence.
Caveats
This does not establish that endocrine-disrupting chemicals play no role in ASD, only that the WHO-UNEP report's specific citations for an environmental/EDC contribution are thin; no causal chemical-to-autism claim is made or implied here.
Source: Lamb et al. 2014, Regulatory Toxicology and Pharmacology Industry-funded Link to this fact
Health effects / Other evidence
In a mouse study the WHO-UNEP 2012 report cites as evidence that DES and the plant estrogen genistein cause endometrial cancer, the doses of DES used were more than 1,000 times a typical daily estrogen dose from a low-dose oral contraceptive, and the genistein dose was more than 50 times the highest human daily dietary intake reported in a high-soy-consuming population; the study also used mice bred with a genetic defect in DNA-repair, a detail the WHO-UNEP report does not mention.
Moderate evidence: Single cited animal study; dose-comparison figures come from the critique paper's own calculations, not independently re-verified here.
Source: Lamb et al. 2014, Regulatory Toxicology and Pharmacology Industry-funded Link to this fact
Health effects / Government agency
The European Commission states that 'a share of scientists' holds that no safe threshold can be established for endocrine disruptors at all, an open scientific question the Commission says still needs to be resolved.
Moderate evidence.
Caveats
The Commission does not adopt this position itself; it presents it as one side of an ongoing scientific disagreement that predates this document (see WHO/UNEP 2012 vs. Lamb et al. 2014's industry-funded critique).
Source: EU Commission 2018 Link to this fact
Health effects / Animal study
Repeated animal studies attribute transgenerational reproductive effects of endocrine-disrupting chemicals to epigenetic changes (DNA methylation, histone modification, noncoding RNA) transmitted through the germline to descendants who were never themselves exposed to the chemical, rather than to a new genetic mutation.
Moderate evidence.
Caveats
This is the accepted mechanistic framework across many of the cited rodent studies, but no human transgenerational endocrine-disruptor reproduction study exists to confirm the mechanism operates the same way in people.
Health effects / Review of other studies
Antiperspirants do not stop sweating; astringent aluminum or aluminum-zirconium salts narrow the sweat duct and reduce underarm sweat emission by only about 50%.
Moderate evidence: Secondhand via a narrative review; the underlying efficacy studies are not individually cited by concentration here.
Source: Martini 2020, Annales de Dermatologie et de Vénéréologie Link to this fact
Health effects / Other evidence
Aluminum from antiperspirants penetrates intact skin only in tiny amounts (up to 0.07% of the applied dose in an in vitro Franz-cell study), but rose roughly 6-fold on damaged or shaved skin, reaching 11.5 micrograms per square centimeter.
Moderate evidence: Single in vitro study on excised human skin, known only secondhand via this review; primary (Pineau 2012) not yet in our evidence base.
Source: Martini 2020, Annales de Dermatologie et de Vénéréologie Link to this fact
Health effects / Review of other studies
Multiple independent expert reviews (a 2008 oncology/pharmacology literature review, France's Afssaps in 2011, and the EU's Scientific Committee for Cosmetic Products in 2014) found no conclusive evidence that antiperspirant aluminum causes breast cancer or Alzheimer's disease.
Moderate evidence.
Caveats
Relayed secondhand through a single review; concern-side biopsy and mechanistic studies (aluminum content in axillary breast tissue, proposed metalloestrogen mechanism) are also cited and not yet resolved in our evidence base; even the reassuring reviews add a precautionary <0.6% concentration cap.
Source: Martini 2020, Annales de Dermatologie et de Vénéréologie Link to this fact
Health effects / Test-chamber measurement
Scented candle fumes failed to reduce chronic-stress-induced serum corticosterone and instead increased inflammatory biomarkers, proinflammatory cytokine production, and NF-kB activation in the lungs and heart, causing cardiopulmonary injury.
Moderate evidence.
Caveats
Single-lab mouse study at an occupational-intensity chamber dose (already our own grading for this primary at C328); not a demonstration that candle scent itself fails to provide subjective psychological relief in humans.
Source: Singh 2023, Environmental Science & Technology Link to this fact
Health effects / Review of other studies
In a systematic review of 15 human studies, 11 found significantly higher bisphenol A (BPA) in the blood, urine, or follicular fluid of women with polycystic ovary syndrome (PCOS) than in women without it, with several also finding positive correlations between BPA and testosterone, androstenedione, insulin resistance, or low-grade inflammation markers.
Moderate evidence.
Caveats
Every included study is cross-sectional or case-control with a single blood/urine/follicular-fluid sample per woman; the review itself notes this cannot capture an individual's true exposure given BPA's short urinary half-life, and cannot establish which way causation runs.
Health effects / Lab study (cells or chemistry)
In a lab-grown 3D model of human skin, fragmented polystyrene particles smaller than about 2 micrometers reached the epidermis within about an hour of a single dose and the dermis within about two hours, peaking at an average of 4.7 micrograms in the dermis after 6 hours, and dose-dependently switched on inflammation genes and proteins (IL-1alpha, IL-1beta, IL-18, IL-6, IL-8, ICAM-1, FOS, JUN) at both the mRNA and protein level.
Moderate evidence.
Caveats
Single academic group, one 3D skin-equivalent model (not living human skin), no measured real-world human dermal dose to compare against; not independently replicated. The paper's own abstract rounds the dermal timing to 'within one hour', but its Results section reports the dermis specifically at 2 hours.
Source: Song et al. 2024, Journal of Hazardous Materials Link to this fact
Health effects / Animal study
In live hairless mice given fragmented polystyrene topically on shaved skin once a day for two weeks, only about 0.04% to 0.12% of the total applied plastic was detectable in the skin afterward, yet the treated skin still showed a significant, dose-dependent increase in the same inflammatory genes and proteins seen in the lab skin model, with no change in body or liver weight.
Moderate evidence.
Caveats
N=10 mice per group, single lab, unreplicated; topical dosing under an occlusive patch is not the same as ambient skin contact; the percentage penetrating was calculated by our evidence base from the paper's own mass figures, not stated as a percentage by the authors.
Source: Song et al. 2024, Journal of Hazardous Materials Link to this fact
Health effects / Lab study (cells or chemistry)
In donated human abdominal skin dosed daily with fragmented polystyrene for 10 days, the plastic reached depths of about 183 to 291 micrometers (into the dermis) and penetrated proportionally more than in mouse skin given a similar daily dose — roughly 1.7% of the applied dose in human skin versus about 0.04% in mouse skin at the lower dose level — which the authors attribute to human hair follicles being much larger than mouse hair follicles.
Moderate evidence.
Caveats
Only 3 human donor skin samples per group, ex vivo tissue with no blood flow or active immune response; the exact penetration percentages are our own calculation from the paper's mass tables, not stated by the authors; the paper does not report what size of particle penetrated the human skin (only the mouse skin was size-measured after digestion), so 'more permeable' here refers to total mass, not confirmed to be the same size cutoff as mouse skin.
Source: Song et al. 2024, Journal of Hazardous Materials Link to this fact
Health effects / Lab study (cells or chemistry)
Fragmented polystyrene was cytotoxic to human skin keratinocytes and dermal fibroblasts in culture at concentrations above 200 micrograms per milliliter over 24 hours, with visible cell membrane and cytoskeleton damage at 400 micrograms per milliliter; below that, the plastic was still taken up into the cytosol of both cell types at 100 micrograms per milliliter without killing them.
Moderate evidence: Cell-culture (2D monolayer) result, not skin tissue; concentrations are not benchmarked against any measured real-world skin exposure.
Source: Song et al. 2024, Journal of Hazardous Materials Link to this fact
Health effects / Lab study (cells or chemistry)
RNA sequencing of human keratinocytes exposed to fragmented polystyrene found nearly 4,700 genes with altered activity, with pathway analysis pointing to the p53, IL-17, and TNF inflammatory signaling pathways as significantly affected.
Moderate evidence.
Caveats
Single-dose, single-cell-line transcriptomic screen (100 micrograms/mL, 24 h); a broad gene-expression signature, not a direct measure of a disease outcome; not independently replicated.
Source: Song et al. 2024, Journal of Hazardous Materials Link to this fact
Health effects / Review of other studies
Across multiple large studies and meta-analyses, people with higher levels of the plastic chemical bisphenol A (BPA) in their urine or blood have consistently higher rates of obesity and abdominal obesity; one meta-analysis found an 11% increase in obesity risk for every 1 ng/mL rise in blood BPA.
Moderate evidence.
Caveats
Nearly all human evidence is cross-sectional or short-follow-up observational; BPA exposure correlates with eating packaged and ultra-processed food, which is itself linked to obesity, so diet and lifestyle confounding cannot be excluded.
Source: Dalamaga 2024, International Journal of Molecular Sciences Link to this fact
Health effects / Review of other studies
In mouse, rat, and zebrafish studies, bisphenol A and phthalates such as DEHP promote weight gain by switching on the master fat-cell gene PPAR-gamma, boosting appetite-driving brain chemicals (neuropeptide Y and AgRP), and triggering low-grade inflammation in fat tissue, often at doses meant to mimic real-world human exposure.
Moderate evidence.
Caveats
These are animal and cell-culture mechanisms; extrapolation to human obesity risk is not established, and effects vary by species, sex, dose, and life stage in ways not yet mapped for humans.
Source: Dalamaga 2024, International Journal of Molecular Sciences Link to this fact
Information, not medical advice. See also: myths we won’t tell you.