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.
Health effects / Other evidence
Clove, cinnamon leaf, and bay leaf oils are 63-91% eugenol, and for five of the oils the panel's own table shows the eugenol-group intake is above its screening threshold (90 micrograms/person/day, up to 1,900 for clove bud oil), so the safety case rests on a 1983 rat feeding study (no-effect level 300 mg/kg/day) instead of the threshold.
Moderate evidence: The margins computed are large; the point is that the threshold was exceeded and one industry-reviewed study carries the conclusion. Not evidence of harm at flavor levels.
Source: Gooderham 2020, Food and Chemical Toxicology Industry-funded Link to this fact
Health effects / Measured in people or real products
In a US cohort of 5,303 adults followed about 10 years, people in the top third of urinary DEHP metabolites (a phthalate from plastics, food packaging, and medical tubing) had about 42% higher all-cause mortality than the bottom third after adjusting for diet, smoking, activity, and BMI (HR 1.42, 95% CI 1.13-1.78).
Moderate evidence.
Caveats
Observational cohort with spot-urine exposure, so association not proof of cause; one cohort, with the later global model built on its single hazard ratio (not independent). The DEHP result turned null when the earliest NHANES wave (2001-2002) was excluded.
Source: Trasande, Liu & Bao 2022, Environmental Pollution Link to this fact
Health effects / Measured in people or real products
In the same study, the low-molecular-weight phthalate group, which includes diethyl phthalate (DEP), the phthalate used in fragrance and personal care, showed no all-cause mortality signal (continuous HR 1.01, 0.96-1.07; top third vs bottom third 1.11, 0.91-1.34).
Moderate evidence.
Caveats
A null result in one cohort using single spot-urine samples; the pooled group also includes other phthalates, and the top tertile of the DEP metabolite MEP was linked to cardiovascular death (HR 1.64, 1.07-2.51) but not in the continuous model.
Source: Trasande, Liu & Bao 2022, Environmental Pollution Link to this fact
Health effects / Other evidence
Hydrogen cyanide forms alongside benzaldehyde whenever bitter almond, apricot kernel, cherry, or sloe materials are soaked in water for flavor oil; makers remove it to a limit of 150 ppm (25 ppm for cherry laurel oil) and the panel calculates a resulting intake of about 0.001 to 0.003 micrograms per kg body weight a day, far under the 20 micrograms per kg reference value.
Moderate evidence: Calculated at the maximum allowed cyanide level, not measured in products; assumes a 60 kg adult; industry-panel calculation.
Source: Guengerich 2026 (FEMA Expert Panel), Food and Chemical Toxicology Industry-funded Link to this fact
Health effects / Other evidence
For 3 of the 18 flavorings (vanilla extract, cumin oil, apricot kernel oil) the panel found a child's lower body weight pushed intake above the generic safety threshold, and cleared each with a single animal no-effect level (ethyl vanillin 500 mg/kg/day in rats, D-limonene 215 mg/kg/day, apricot kernel oil 10 mg/kg/day in a one-dose 13-week feed study).
Moderate evidence: Each rescue relies on one animal study, in one case a surrogate chemical (ethyl vanillin for the vanillin group); industry panel's own judgment; no human child data.
Source: Guengerich 2026 (FEMA Expert Panel), Food and Chemical Toxicology Industry-funded Link to this fact
Health effects / Other evidence
The flavor panel judges methyl salicylate (wintergreen oil) safe as a flavoring from a 50 mg/kg/day no-effect level in dogs; intake as an added flavoring is about 1,190 micrograms a person a day (0.02 mg/kg), giving a margin of about 2,500; the same chemical also appears in cosmetics and rubs where this food-only assessment does not count it.
Moderate evidence.
Caveats
Industry panel's calculation; the margin is a ratio of two industry-chosen figures; combined food plus cosmetic exposure is not assessed in either the flavor or the fragrance industry's review.
Source: Guengerich 2026 (FEMA Expert Panel), Food and Chemical Toxicology Industry-funded Link to this fact
Health effects / Other evidence
The US National Toxicology Program found 'clear evidence' that isoeugenol caused liver tumors in male mice (combined adenoma or carcinoma in 43 of 50 at each dose versus 28 of 50 in controls) and 'equivocal evidence' in male rats and female mice; the flavor panel judged this irrelevant to people at flavor intake.
Moderate evidence.
Caveats
Numbers are the panel's quotation of an NTP draft report; the panel's dismissal (high control tumor rate, hepatotoxicity) is the industry position, and its margin arithmetic contains a 750-fold versus 'six orders' inconsistency.
Source: Smith et al. 2009 (FEMA Expert Panel), GRAS Flavoring Substances 24, Food Technology 63(6), Food Technology Industry-funded Link to this fact
Health effects / Self-reported survey
In two US national telephone surveys, about 1 in 5 adults said air fresheners or deodorizers give them headaches, breathing difficulties, or other problems: 17.5% in 2002-03 and 20.5% in 2005-06 (19.0% averaged).
Moderate evidence.
Caveats
Self-report; the two values are within each sample's stated plus or minus 3% precision, so the rise is borderline. An earlier paper by the same author quotes 17.8% for the first survey, and the sister 2003 paper prints 17.6%.
Source: Caress & Steinemann 2009, Journal of Environmental Health Link to this fact
Health effects / Self-reported survey
People with asthma or chemical sensitivity reported fragrance reactions far more often than the general public: air-freshener problems in 33.5% of asthmatics and 57.9% of the chemically sensitive vs 19.0% of all adults, and laundry-vent irritation in 21.2% and 39.8% vs 10.9% (2005-06).
Moderate evidence.
Caveats
Self-report; the chemically sensitive subgroup is small (about 11% of about 1,057, roughly 120 people per survey) and its second-survey value of 60.2% is printed identically in two different tables, so subgroup numbers may contain a copy error.
Source: Caress & Steinemann 2009, Journal of Environmental Health Link to this fact
Health effects / Self-reported survey
Women reported fragrance reactions more often than men in both national surveys: others' scent irritating 36.0% vs 22.5% (2002-03) and 31.6% vs 27.0% (2005-06); air-freshener problems 21.1% vs 11.4% and 22.0% vs 17.7%.
Moderate evidence: Self-report; reporting differences are not the same as biological differences; later surveys show a smaller sex gap on some items.
Source: Caress & Steinemann 2009, Journal of Environmental Health Link to this fact
Health effects / Self-reported survey
Among US adults with diagnosed asthma, 29.2% also described themselves as chemically sensitive and 9.6% had diagnosed multiple chemical sensitivity (two-survey average), against 11.4% and 3.2% in the whole sample.
Moderate evidence: Self-reported diagnoses without medical confirmation; simple averages of two surveys; association only, no direction of effect.
Source: Caress & Steinemann 2009, Journal of Environmental Health Link to this fact
Health effects / Measured in people or real products
The 17 fragrances together contained 24 chemicals EWG classed as skin sensitizers, about 10 per product (19 in Armani Acqua Di Gio); limonene was in 16 products, linalool in 14, and geraniol in 12.
Moderate evidence: Presence is lab- or label-based; sensitizer status is EWG's literature classification, not a potency or dose assessment; the sensitizer table omits one of the 17 products.
Source: Sarantis et al. 2010 (Not So Sexy, Canadian edition) Link to this fact
Health effects / Other evidence
The US National Toxicology Program reported 'clear evidence of carcinogenic activity' of pyridine (a flavor and food constituent) in male and female mice, with liver carcinoma in males 15/50 (control) vs 35/50, 41/49, and 40/50 across three doses and hepatoblastoma in females 1/49 vs 2/50, 9/50, and 16/50, and 'some evidence' in male rats (kidney adenoma 2/50 vs 10/50 at the top dose).
Moderate evidence.
Caveats
Rodent bioassay (2-year, drinking water), numbers quoted by the industry panel from NTP 2000, not the NTP report itself; mouse liver tumors in B6C3F1 are known to have high background rates.
Source: FEMA GRAS 25 (2011), Food Technology Industry-funded Link to this fact
Health effects / Other evidence
Pyridine reaches the diet far more from ordinary food than from added flavor: the panel says natural intake exceeds added-flavor intake 'by a factor of at least 100' (coffee 37-49 ppm, cocoa 0.5 ppm, shrimp 4.1-9.9 ppm), and estimates added intake at about 7 micrograms per person per day for 'eaters only'.
Moderate evidence: Industry intake model from poundage survey (54 kg/year, 28 million eaters, 0.8 correction); not a measured exposure.
Source: FEMA GRAS 25 (2011), Food Technology Industry-funded Link to this fact
Health effects / Other evidence
The FTC stated that for 'the vast majority of OTC homeopathic drugs' there are 'no valid studies using current scientific methods showing the product's efficacy,' so therapeutic claims for them 'lack a reasonable basis and are likely misleading.'
Moderate evidence: A regulator's characterization of the evidence; no trials are cited or reviewed in the statement itself.
Source: FTC 2016 OTC homeopathic enforcement policy statement, Federal Register 81(239):90122-90123 Link to this fact
Health effects / Other evidence
In 2013-14 an AVMA council literature review found studies of homeopathy in veterinary medicine were anecdotal or flawed and concluded there was no clinical evidence to support homeopathic preparations for treating or preventing animal disease (as reported by JAVMA News, 2017).
Moderate evidence: Secondhand news summary of a council review; the review's methods are not described, and the AVMA House of Delegates still took no modality-specific position.
Source: AVMA 2017 (JAVMA News) Link to this fact
Health effects / Review of other studies
The European national science academies' joint body (EASAC, 2017) concluded that there are no known diseases for which there is robust, reproducible evidence that homeopathy works beyond the placebo effect.
Moderate evidence: A policy synthesis of other bodies' reviews with no systematic search or new data; it takes a position rather than reanalyzing the trials, and does not seek a ban.
Source: EASAC 2017, EASAC Statement (September 2017) Link to this fact
Health effects / Other evidence
FDA formally asked Hyland's maker to recall every in-date lot of Hyland's Baby Teething Tablets and Baby Nighttime Teething Tablets in April 2017, and classed it Class I, the category for products with a reasonable probability of causing serious harm or death.
Moderate evidence: Regulator's finding in a letter; no assay data, sample counts or adverse-event numbers are given in the document.
Source: FDA requested recall letter, 7 Apr 2017, FDA letter (7 April 2017) Link to this fact
Health effects / Other evidence
FDA noted that this firm's homeopathic drugs included poisonous plant materials (Nux vomica, which contains strychnine; belladonna; Aconitum napellus; Gelsemium sempervirens) and that it had not run the required strychnine assay on one Nux vomica mother tincture lot.
Moderate evidence: The letter reports missing tests, not measured strychnine or alkaloid levels, and no injuries.
Source: FDA warning letter 521717, 7 Sep 2017, FDA warning letter (MARCS-CMS 521717), 7 September 2017 Link to this fact
Health effects / Other evidence
In June 2017 FDA warned Raritan Pharmaceuticals that its Homeopathic Infant's Teething Tablets were adulterated (CGMP violations) and misbranded, because FDA's tests found belladonna alkaloid content that varied widely and far exceeded the label.
Moderate evidence: Regulator's finding in a letter; no measured amounts, lot counts or assay method are given.
Source: FDA warning letter 17-NWJ-09, 20 Jun 2017 Link to this fact
Health effects / Other evidence
The Royal College of Veterinary Surgeons' Council stated that 'Homeopathy exists without a recognized body of evidence for its use. Furthermore, it is not based on sound scientific principles.'
Moderate evidence: Institutional judgment with no references or data in the statement itself; evidence is in other evidence base pages (Shang 2005, Ernst 2002, Mathie & Clausen 2014).
Source: RCVS 2017 complementary and alternative medicines position statement Link to this fact
Health effects / Government agency
FDA reported that a 2016 search of its adverse event database found 99 cases consistent with belladonna toxicity, including infant deaths and seizures possibly related to homeopathic teething products, and that the poisonous belladonna alkaloids in some products far exceeded the labeled amounts.
Moderate evidence.
Caveats
Reports 'consistent with' toxicity and 'possibly related'; FDA does not show case-by-case causality review or denominators. The excess-alkaloid finding is stated without lab numbers in this document (see the Hyland's recall letter for the product-specific action).
Source: FDA draft guidance, Drug Products Labeled as Homeopathic (Oct 2019) Link to this fact
Health effects / Government agency
FDA received more than 130 reports of loss of smell (anosmia) linked to Zicam homeopathic intranasal zinc products by 2009, and judged that use as labeled gave significant daily intranasal zinc exposure.
Moderate evidence: Complaint count and FDA's determination as summarized in one paragraph; the underlying case reviews and the exact product dates are not in this document.
Source: FDA draft guidance, Drug Products Labeled as Homeopathic (Oct 2019) Link to this fact
Health effects / Other evidence
FDA's July 2018 inspection of King Bio, a maker of homeopathic drugs, found recurring extremely high bacterial counts in the water used to make its products and in finished products, and FDA's own lab testing also found 'inordinately high' contamination.
Moderate evidence: Regulator's account of one firm; the letter gives no CFU values, sample counts or full product list, and the firm's side is known only through FDA.
Source: FDA warning letter to King Bio, 20 Mar 2019, FDA warning letter (20 March 2019) Link to this fact
Health effects / Other evidence
FDA found Burkholderia multivorans in multiple King Bio drug products with no investigation by the firm.
Moderate evidence: The letter does not say which products, how many, or at what level; it reports no infections.
Source: FDA warning letter to King Bio, 20 Mar 2019, FDA warning letter (20 March 2019) Link to this fact
Health effects / Other evidence
The EU committee set its benzophenone-3 safety limit from a rat study (Nakamura 2015): a maternal and lactational feeding study where spermatocyte numbers in male offspring fell at 3,000 ppm in feed, so it took 1,000 ppm (67.9 mg/kg body weight a day) as the no-effect level.
Moderate evidence.
Caveats
Single rat study chosen by the committee; testosterone was not measured in fetuses, and the study authors called the effects only at much higher doses than human exposure. Rat dose, not human outcome.
Source: SCCS 2021 (BP-3 opinion), SCCS final opinion (PDF) Link to this fact
Health effects / Other evidence
The EU committee confirmed that benzophenone-3 can cause photoallergic skin reactions, citing the European photo-patch test task force ranking it the second most frequent photo contact allergen among UV filters, plus one reported anaphylaxis case.
Moderate evidence: Secondary summary of clinical series that largely tested patients already suspected of photocontact allergy; no prevalence among all sunscreen users is given.
Source: SCCS 2021 (BP-3 opinion), SCCS final opinion (PDF) Link to this fact
Health effects / Other evidence
People who used 4-6% benzophenone-3 sunscreens under maximal-use conditions reached average peak blood levels of 169-258 ng/mL in the FDA trials the EU committee cited, and the committee noted human levels exceeded the blood level seen in rats at the no-effect dose.
Moderate evidence.
Caveats
The SCCS did not use these trials in its margin of safety because absorption percentage was not measured; the comparison to rat blood levels is the committee's qualitative remark. Maximal-use conditions, not typical use.
Source: SCCS 2021 (BP-3 opinion), SCCS final opinion (PDF) Link to this fact
Health effects / Other evidence
FDA says homeopathic products have caused or could have caused significant harm, citing 99 adverse events consistent with belladonna toxicity in 2016 (including infant deaths and seizures, possibly related to teething tablets) and more than 130 anosmia reports with Zicam intranasal zinc by 2009.
Moderate evidence: Agency case summaries without denominators or causation analysis; 'possibly related'. Shows signal, not incidence.
Source: FDA 2022 Homeopathic Drug Products guidance Link to this fact
Health effects / Other evidence
FDA says improper manufacturing of homeopathic products can cause contamination and incorrect dilutions, and that consumers who use them instead of approved treatment may delay or stop effective treatment.
Moderate evidence: Statement of concern without incidence data on this page.
Source: FDA consumer page: Homeopathic Products, FDA web page Link to this fact
Health effects / Other evidence
The panel's 2024 evaluation criterion for flavor oils containing genotoxic-carcinogen constituents (safrole, estragole, methyl eugenol, myristicin) is a margin of exposure of 10,000, with a 0.15 microgram per person per day threshold as the first screen.
Moderate evidence: The panel states the benchmark matches EFSA and JECFA; the intake inputs are industry-derived.
Source: FEMA Expert Panel 2024 (GRAS 31), Food Technology Industry-funded Link to this fact
Health effects / Other evidence
A 2025 systematic review and meta-analysis of 58 mostly clinic-based studies found that about half of people with migraine (47.8%, 95% CI 42.0-53.6) report osmophobia, an aversion to or heightened sensitivity to odors.
Moderate evidence: Clinic samples, mostly self-report survey questions, varying definitions; selected on reporting osmophobia; not a population rate.
Source: Briggs et al. 2025, World J Otorhinolaryngol Head Neck Surg, World Journal of Otorhinolaryngology – Head and Neck Surgery Link to this fact
Health effects / Other evidence
In the same review, about 40% of migraine patients (40.1%, 95% CI 32.7-47.8) said an odor can trigger their attacks; among those naming odor triggers, perfume was the most common (65.0%), then smoke (36.3%) and household cleaners, bleach, or chemicals (29.0%).
Moderate evidence: Retrospective self-report; trigger vs premonitory aversion not separated; the same 40.1% figure also appears for in-attack hypersensitivity with a different CI.
Source: Briggs et al. 2025, World J Otorhinolaryngol Head Neck Surg, World Journal of Otorhinolaryngology – Head and Neck Surgery Link to this fact
Health effects / Other evidence
People with migraine who have osmophobia are more likely to also have light sensitivity (OR 1.45), aura (OR 1.66) and nausea (OR 1.73), but not sound sensitivity (OR 1.18, p = 0.6), in a 2025 meta-analysis.
Moderate evidence: Cross-sectional associations among clinic patients; one included study recruited osmophobic patients on purpose; heterogeneity (I2) not reported.
Source: Briggs et al. 2025, World J Otorhinolaryngol Head Neck Surg, World Journal of Otorhinolaryngology – Head and Neck Surgery Link to this fact
Health effects / Other evidence
A 2006 citizen petition asked FDA to revoke diacetyl's GRAS status because of studies linking the buttery flavor chemical to severe respiratory reactions; as of GAO's 2010 report FDA had not decided it, 3 years after its 180-day interim letter.
Moderate evidence: GAO only reports the petition and FDA's inaction; it does not evaluate the diacetyl evidence. Status is as of December 2009.
Source: GAO 2010, GAO-10-246 Link to this fact
Health effects / Other evidence
In 2010 a UK parliamentary committee concluded that the systematic reviews and meta-analyses of homeopathy 'conclusively demonstrate that homeopathic products perform no better than placebos'.
Moderate evidence.
Caveats
This is a committee weighing existing reviews, not new analysis; the word 'conclusively' is the committee's, one member dissented on related votes, and the Ernst evidence it relied on came from a known critic. The underlying trial evidence is independently replicated across reviews.
Source: UK House of Commons Science and Technology Committee, Evidence Check 2: Homeopathy (HC 45), Feb 2010 Link to this fact
Health effects / Other evidence
Patient-reported improvement is not evidence of efficacy: UK homeopathic hospital surveys (over 6,500 patients, 70% improved; 500 patients, 72% reduced conventional drugs) were uncontrolled and the committee attributed them to placebo, self-limiting illness, self-selection, and long, empathetic consultations.
Moderate evidence: Committee's explanation, not tested within these surveys; placebo-response literature supports the mechanisms but not their size here.
Source: UK House of Commons Science and Technology Committee, Evidence Check 2: Homeopathy (HC 45), Feb 2010 Link to this fact
Health effects / Other evidence
The UK Department of Health's Chief Scientist told the Committee that most independent scientists see the evidence for homeopathy's efficacy as weak or absent and that there is no plausible mechanism, but the Government's reply still describes 'some controversy' and did not adopt the Committee's 'conclusively demonstrate' wording.
Moderate evidence: Opinion of a government official reported in a policy reply; no data cited, and the reply names no specific peer-reviewed reports.
Source: UK Department of Health, Government Response to the Science and Technology Committee report 'Evidence Check 2: Homeopathy', Cm 7914 (July 2010), Command Paper Cm 7914 (July 2010) Link to this fact
Health effects / Other evidence
The UK Government agreed to give weight to rigorous studies but declined to comment on the ethics of using placebos, saying there is no scientific consensus on how placebos work and that it is not for the Department of Health to judge a particular treatment in a particular setting.
Moderate evidence: Reflects the Department's position; the claim about no consensus is the Government's, not a literature review.
Source: UK Department of Health, Government Response to the Science and Technology Committee report 'Evidence Check 2: Homeopathy', Cm 7914 (July 2010), Command Paper Cm 7914 (July 2010) Link to this fact
Health effects / Other evidence
Australia's NHMRC concluded in March 2015 that 'there are no health conditions for which there is reliable evidence that homeopathy is effective', after assessing 57 systematic reviews containing 176 individual studies.
Moderate evidence.
Caveats
A government overview of published systematic reviews, not a new trial; the finding is about the quality of the evidence (generally small, poor studies), and conclusions rest on statistical significance rather than clinical importance.
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
NHMRC found homeopathy was reported to be no more effective than placebo for 13 conditions, including asthma, headache and migraine, upper respiratory tract infections, premenstrual syndrome, and warts.
Moderate evidence.
Caveats
A condition entered this group if all studies found (regardless of size and quality) or a large majority of the reliable ones were null, so some rest on small studies. Taken from systematic reviews, not re-run.
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
For 29 conditions (such as eczema, lower back pain, osteoarthritis, and stroke) NHMRC found only one placebo-controlled homeopathy study, and that study was unreliable, so it said no conclusion about effectiveness could be drawn either way.
Moderate evidence.
Caveats
'No conclusion' means insufficient evidence, not evidence of no effect. The paper's own counts have small internal inconsistencies (for example, lower back pain is also listed as lacking placebo studies).
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
For 14 conditions, including allergic rhinitis, ADHD in children, fibromyalgia, and rheumatoid arthritis, some studies reported homeopathy beat placebo, but NHMRC judged them unreliable because of poor quality, too few participants, or both.
Moderate evidence: The paper says these results would need confirming by well-designed studies with adequate numbers; it does not say they are false.
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
NHMRC's reliability rules were explicit: a study counted as 'sufficient size' only above 150 participants (fewer than 50 was 'very small'); quality was graded from Jadad, risk-of-bias, and similar scores; when ratings conflicted it used the lower one; and overall confidence started at 'high' and was marked down for poor quality and sparse data.
Moderate evidence: The size cutoffs are NHMRC's own guide for continuous outcomes, not a universal standard; the appendix says the criteria 'should not be treated as universal rules'.
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
NHMRC's test for calling a treatment effective required four things together: improvement not explained by placebo, improvement unlikely to be chance, improvement meaningful for overall health, and the result seen consistently in several studies.
Moderate evidence: The paper then notes that its conclusions mostly rest on statistical significance, because outcomes could not be ranked by clinical importance in advance.
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
NHMRC did not treat 'no reliable evidence' as proof homeopathy fails: it sorted conditions into reported-no-better-than-placebo, unreliable-positive, and no-conclusion-possible groups, and in its method assumed 'no effect' only where enough reliable evidence existed.
Moderate evidence: The null-hypothesis section says the onus of proof lies on those proposing the treatment. Readers should quote the three tiers, not one sentence.
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
Homeopathy supporters' own submitted studies did not change NHMRC's conclusion: 343 articles sent before the review led to 9 studies on 8 conditions, and 153 articles cited in 48 public submissions led to 40 studies, mostly at moderate-to-high risk of bias.
Moderate evidence.
Caveats
Counts as printed; the paper's later 'additional 42 studies' does not equal 9 + 40. Only two low-risk-of-bias studies were found (a small favorable cough study and an 86-person mental-fatigue study with no difference).
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
NHMRC says it could not separate individualized from clinical homeopathy because most reviews did not analyze them separately, and that most studies used clinical homeopathy; only for children's diarrhea was the evidence split.
Moderate evidence: A limitation the paper itself states; classical individualized homeopathy is therefore under-tested in this review.
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
NHMRC noted publication bias could not be ruled out: only 10 of the 57 reviews considered it, only two searched systematically for missing studies, and one reported significant publication bias from under-reporting of non-significant and negative results.
Moderate evidence: Clinical trial registries searched (WHO, ClinicalTrials.gov, ANZCTR) turned up no extra studies.
Source: NHMRC 2015 Information Paper on the effectiveness of homeopathy (CAM02A) Link to this fact
Health effects / Other evidence
BSAVA says health claims for many complementary and alternative therapies 'can be in excess of the available scientific data, and sometimes contradicts the scientific evidence.'
Moderate evidence: Institutional judgment with no references; it names no specific therapy and says some products within a modality do have reasonable evidence.
Source: BSAVA position statement on complementary and alternative therapies (reviewed 2024) Link to this fact
Health effects / Other evidence
BSAVA notes that owners, therapists, and vets may think a therapy works because of belief (placebo), anecdote, or cognitive bias, and that many conditions fluctuate or resolve on their own, which makes efficacy hard to judge.
Moderate evidence: Position statement without data; our veterinary evidence on this is in the Conzemius 2012 and Mathie 2014 pages.
Source: BSAVA position statement on complementary and alternative therapies (reviewed 2024) Link to this fact
Health effects / Other evidence
In a Puerto Rico pregnancy cohort, urinary benzophenone-3, triclosan, and parabens averaged two to three times higher after recent lotion use, benzophenone-3 about fivefold higher after sunscreen use, and triclosan and triclocarban two- and fourfold higher with liquid and bar soap use (EWG's relay).
Moderate evidence: Secondhand numbers; the page gives no authors or DOI; exposure biomarkers are not health outcomes.
Source: EWG 2025 (Friedman & Huang), EWG news (web page) Link to this fact
Health effects / Other evidence
EPA's 2006 ADBAC assessment classed the chemical as corrosive to rabbit skin, a category I eye irritant, and category II by inhalation (rat LC50 between 0.054 and 0.51 mg/L).
Moderate evidence: Animal acute studies on a concentrated test substance, summarized by the agency; not a measure of consumer exposure to diluted products.
Source: EPA RED 2006 Link to this fact
Health effects / Other evidence
The EU committee's 2006 review found cetrimonium chloride caused irreversible eye damage in rabbits at 25-29% active, and behentrimonium chloride caused irreversible damage at 8% active but only transient irritation at 3% active.
Moderate evidence: Rabbit eye tests with small groups (3 per test) on concentrates, from unpublished industry reports; they show hazard of the concentrate, not harm from finished products at 0.5-5%.
Source: SCCP 2006 Industry-funded Link to this fact
Health effects / Other evidence
In rabbit skin tests, cetrimonium chloride at 25-29% active and steartrimonium chloride at 79% active were irritating (scarring at day 22 in one rabbit), while a 2% solution of a steartrimonium/cetrimonium blend (about 0.56% active) was non-irritating.
Moderate evidence: Small groups (3 rabbits, one rabbit for the 79% test); the opinion's own summary table labels the 0.56% test 'irritating', contradicting its study text.
Source: SCCP 2006 Industry-funded Link to this fact
Health effects / Other evidence
Cationic trimonium surfactants such as cetrimonium chloride showed little skin penetration in an in vitro pig-skin test (about 1% of the dose in skin at 72 h, none in the receptor fluid).
Moderate evidence: Single in vitro model, short exposure, rinse-off style; says nothing about leave-on use or damaged skin. CIR is industry funded.
Source: CIR 2010 trimoniums (draft) Industry-funded Link to this fact
Health effects / Other evidence
In 2010 the EU's cosmetics committee judged powder hair dye with up to 3% sodium perborate on the head not to pose a risk to consumers, with a margin of safety of 700 for the hair dye and 440 for all cosmetic uses combined.
Moderate evidence.
Caveats
Based on a small in vitro pig-skin absorption study the SCCS itself called not up to modern standard (6 chambers) and a rat NOAEL; damaged skin would lower the margin; non-cosmetic exposure (detergents) not counted.
Source: SCCS 2010 (SCCS/1345/10), SCCS opinion SCCS/1345/10 (PDF) Link to this fact
Health effects / Animal study
In rabbits, daily skin application of polysorbate 60 for 30 days caused moderate irritation at 5% and skin necrosis at 10%, yet the same ingredient is reported in cosmetics at up to 6% in rinse-off and 4% in leave-on products.
Moderate evidence: Old study with few details (n not given); cosmetics are not applied daily at full strength, and the Panel's 'formulated to be non-irritating' condition is meant to cover this.
Source: CIR 2015 draft Industry-funded Link to this fact
Health effects / Other evidence
The only reproductive study of polysorbate 60 in this report fed pregnant rats up to 10% of diet (7,693 mg/kg/day) on gestation days 7-14 and found no malformations or fetal weight effects, so the report's NOAEL is simply the highest dose tested.
Moderate evidence: One rat feeding study over a short window (days 7-14); oral, not skin; no human pregnancy data.
Source: CIR 2015 draft Industry-funded Link to this fact
Health effects / Other evidence
The 1984 CIR report said polysorbates 'enhance the activity of known chemical carcinogens' in multiple studies; the 2015 draft says the data were 'not consistent' and the Panel 'was not concerned', and no new carcinogenicity data were found.
Moderate evidence: Positions are from the same industry-funded panel; the underlying tumor-promotion studies were not re-examined in the draft.
Source: CIR 2015 draft Industry-funded Link to this fact
Information, not medical advice. See also: myths we won’t tell you.