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
A toxicokinetic model estimates that fat-soluble fragrance/solvent chemicals from feminine hygiene products (p-cymene, hexane, n-nonane) can build up in body fat to levels 1 to 4 orders of magnitude higher than in blood or urine, even though skin levels stay low.
Moderate evidence: Modeled, not directly measured in body fat; single research group's model not yet independently replicated.
Source: Lin 2025, Environmental Health Perspectives Link to this fact
Health effects / Other evidence
In a worst-case modeled exposure scenario, n-nonane from feminine washes approached a conservative no-appreciable-risk threshold (hazard ratio 0.07 of 0.1) and benzene from moisturizers exceeded a conservative cancer-risk threshold of 1 in 10 million, reaching about 1.8 in 10 million.
Moderate evidence: Highest-exposure scenario, not typical use; best-guess scenario risks were well below any threshold.
Source: Lin 2025, Environmental Health Perspectives Link to this fact
Health effects / Other evidence
In a randomized lab study of 90 young women, being told a room scent was 'relaxing' lowered heart rate by about 4.7 beats per minute and being told it was 'stimulating' raised it by about 3.2, whether the room held lavender, neroli, or a placebo scent.
Moderate evidence: One small single-lab study in young women with one delivery method and 25 min exposure; no replication in this evidence base; cells about 10 each.
Source: Campenni 2004, Psychological Reports Link to this fact
Health effects / Measured in people or real products
In this study neither linalool form changed mood scores significantly compared with plain water; the mood difference was only between the two forms against each other, and the calming 'vigor' boost mentioned in the abstract was not seen with (R)-linalool.
Moderate evidence: Seven analyzed subjects, so non-significance may reflect low power; (R)-linalool lowered tension, depression and anger scores in the expected direction.
Source: Kuroda et al. 2005, European Journal of Applied Physiology Industry-funded Link to this fact
Health effects / Measured in people or real products
A UK study of 160 breast-tissue samples from 40 women with breast cancer found at least one paraben preservative in 99% of samples, with a median total of 85.5 ng/g tissue, mostly methylparaben and n-propylparaben.
Moderate evidence.
Caveats
All women had breast cancer and there was no cancer-free comparison group; blanks were large and no recovery correction was applied, so levels are uncertain; the lab is the same one that earlier reported parabens in breast tumors.
Source: [[src-2012-measurement-paraben-concentrations-breast-tissue-serial-locations|Barr 2012]], Journal of Applied Toxicology Link to this fact
Health effects / Self-reported survey
In a 1981 UK experiment with 835 women, the same tablets in a famous brand's pack gave better self-reported headache relief at one hour than in a plain pack, for both inert and aspirin tablets (mean relief 2.18 vs 1.78 for placebo, 2.7 vs 2.48 for aspirin on a -1 to +4 scale).
Moderate evidence.
Caveats
Self-rated six-point outcome in one study of UK women; no no-treatment arm; funder not stated; 'branding' bundles pack, label and tablet markings; the branded-vs-unbranded difference was not significant at 30 minutes.
Health effects / Self-reported survey
In the same trial, 325 mg aspirin gave more relief than the inert tablet (F = 40.96), larger than the branding effect, and branding and formulation did not interact; packaging added a similar increment on top of both.
Moderate evidence.
Caveats
One single-study ANOVA on self-reported relief; the authors' split of 'two-thirds to three-quarters' drug and 'one-quarter to one-third' branding is their own apportionment, not a tested quantity.
Health effects / Other evidence
A week-long investigation by a Nature team (editor, magician, misconduct investigator) at the lab behind the 1988 'memory of water' paper found no high-dilution effect in the three strictly blind experiments it ran, and concluded the original claims 'are not to be believed'.
Moderate evidence: Three experiments, one week, one investigating team that was not an immunology replication consortium; this is not a test of homeopathic products or clinical effects.
Source: Maddox, Randi & Stewart 1988, Nature, Nature Link to this fact
Health effects / Other evidence
In the lab's own notebook data the repeat readings of identical samples agreed more closely than counting statistics allow, and the lab re-counted control wells whenever a diluted sample read above control, which inflates apparent degranulation.
Moderate evidence: The authors' own analysis of the lab's notebooks, not independently re-analyzed here; the same artifact was absent in strictly blinded duplicate readings.
Source: Maddox, Randi & Stewart 1988, Nature, Nature Link to this fact
Health effects / Other evidence
The flavor industry's panel judged the food flavor d-limonene (also the most common fragrance terpene) still GRAS after the US National Toxicology Program found clear evidence of kidney tumors in male rats, reasoning the finding is irrelevant to human eating; this is a swallowed-route conclusion by the industry panel.
Moderate evidence: Industry panel's own reading; the male-rat alpha-2u-globulin mechanism is accepted by US EPA, but the call covers food intake, not skin or inhalation exposure to limonene.
Source: Hallagan 1995, Regulatory Toxicology and Pharmacology Industry-funded Link to this fact
Health effects / Other evidence
People expect pill color to signal strength and purpose: across four surveys reviewed in 1996, red, yellow, and orange pills were read as stimulating and blue and green pills as calming or sedative.
Moderate evidence: Perceived action only (expectation surveys), not drug effect; small, old studies; white and green were inconsistent across surveys.
Source: de Craen 1996, BMJ Link to this fact
Health effects / Other evidence
The paper's panel reports safety margins of at least 1,000 times (up to more than 1,000,000 times for isophorone) between no-effect doses in animals and estimated flavor intakes for these 119 ring-structured terpene flavors; the margins are the industry panel's calculation and depend on its 'eaters only' intake estimates.
Moderate evidence: Industry calculation using animal studies from the 1960s-1980s; at least one stated margin does not compute (menthol: 375 mg/kg against 0.22 mg/kg is about 1,700x, not '>10,000x').
Source: Adams 1996, Food and Chemical Toxicology Industry-funded Link to this fact
Health effects / Other evidence
Pulegone, a mint-family flavor used at only about 15 kg a year in the US, is the one alicyclic flavor with notable toxicity: the body converts it to menthofuran and then a reactive metabolite that damages the liver, and pennyroyal oil (35-52% pulegone) has killed people when swallowed in doses of about 24 g (about 500 mg/kg).
Moderate evidence.
Caveats
Mechanism from rodent studies at doses above 100 mg/kg; the paper says the same pathway 'may also exist in humans'; flavor-level intake is about 50 ng/kg/day, far below the damaging dose.
Source: Adams 1996, Food and Chemical Toxicology Industry-funded Link to this fact
Health effects / Other evidence
Camphor swallowed above about 50 mg/kg causes vomiting, convulsions, and hallucinations; 50 mg/kg by a pregnant woman caused fetal death and 700 mg killed children, while the flavor-use estimate is only about 3 ug/kg/day, roughly 100,000-fold below a plasma level a survivor reached.
Moderate evidence: Medicinal-camphor poisoning reports, few cases; flavor exposure is far below these doses; the 100,000x plasma comparison is the panel's estimate.
Source: Adams 1996, Food and Chemical Toxicology Industry-funded Link to this fact
Health effects / Other evidence
In a double-blind trial of 35 people with non-allergic chronic rhinitis, 10 days of three-times-daily oxymetazoline nasal spray caused no measurable rebound congestion, whether or not it contained the preservative benzalkonium chloride.
Moderate evidence: Small (n=35), single center, manufacturer (Draco)-funded; 10 days only; sample at 14-17 h after last dose.
Source: [[src-1999-ten-days-oxymetazoline-nasal-spray-benzalkonium-chloride|Graf 1999 (Arch Otolaryngol)]], Archives of Otolaryngology–Head & Neck Surgery Industry-funded Link to this fact
Health effects / Review of other studies
In lab studies on animal and human nasal tissue, benzalkonium chloride (BKC), a preservative in many nasal sprays, slows or stops the beating of nasal cilia within minutes at spray-like concentrations (0.01-0.02%), and keeps doing so at far lower ones over an hour.
Moderate evidence: Review of mostly 1980s-90s in vitro studies of tissue pieces without natural mucus; it does not show harm in people using sprays. No funding statement printed.
Source: Bernstein 2000, Journal of Allergy and Clinical Immunology Link to this fact
Health effects / Review of other studies
Human studies of BKC-containing nasal sprays on mucociliary clearance disagree: one found a single 10-minute 0.02% BKC exposure slowed saccharin clearance (about 763 vs 620 seconds), while repeated-dose human trials found no change.
Moderate evidence.
Caveats
Secondhand summary; small groups, large variation, van de Donk used BKC 100-fold more dilute than clinical products; the author suggests mucus dilution and biopsy site could explain the mismatch.
Source: Bernstein 2000, Journal of Allergy and Clinical Immunology Link to this fact
Health effects / Review of other studies
In healthy volunteers using an oxymetazoline nasal decongestant for 30 days, adding the preservative benzalkonium chloride (BKC) gave more nasal swelling (about 1.14 mm increase) and more stuffiness than oxymetazoline alone; in a separate 30-day trial BKC alone raised swelling by 0.67 mm versus 0.29 mm for oxymetazoline alone and 0.09 mm for placebo.
Moderate evidence.
Caveats
Same author and group as Graf 1999 (Clin Ther); not independent replication. Healthy volunteers, small groups, no allergic-rhinitis patients, and oxymetazoline may prolong BKC contact. AstraZeneca-funded review.
Source: Graf 2001, Respiratory Medicine Industry-funded Link to this fact
Health effects / Review of other studies
A 2002 review of 17 systematic reviews of homeopathy found no health condition and no homeopathic remedy that was convincingly better than placebo.
Moderate evidence: Review by one evaluator, a trained homeopath and longtime critic, many included reviews from his own team; search ended 2001; adds no new trial data.
Source: Ernst 2002, Br J Clin Pharmacol, British Journal of Clinical Pharmacology Link to this fact
Health effects / Other evidence
The best-known positive homeopathy meta-analysis (Linde 1997, odds ratio 2.45) lost its effect on re-analysis: the better the trial, the smaller the effect, and adjusting for publication bias moved the odds ratio in high-quality trials from 0.52 to 1.19.
Moderate evidence.
Caveats
Re-analyses overlap heavily in the same 89 trials; Sterne 2001 chapter likely not peer-reviewed; Linde's own team also said the re-analysis 'weakened' the original, but did not abandon it.
Source: Ernst 2002, Br J Clin Pharmacol, British Journal of Clinical Pharmacology Link to this fact
Health effects / Other evidence
Arnica, the most-tested homeopathic remedy, was not shown to differ from placebo in two systematic reviews (8 and 37 trials).
Moderate evidence: Two reviews with overlapping trials; one is a non-peer-reviewed company report; author's team wrote one of them.
Source: Ernst 2002, Br J Clin Pharmacol, British Journal of Clinical Pharmacology Link to this fact
Health effects / Other evidence
Peak particle counts in these chamber tests reached the analyzer's limit of 670,000 particles per cubic centimeter within minutes of using the air fresheners and the wood floor cleaner, mostly in the smallest (ultrafine) sizes.
Moderate evidence.
Caveats
The counter's ceiling means the true peak may be higher; no health effect was tested; the smallest size range (0.02-0.1 micrometers) came from subtracting two instruments, which adds uncertainty.
Source: Sarwar 2004, Journal of the Air & Waste Management Association Link to this fact
Health effects / Other evidence
In a 2005 Lancet analysis, 110 placebo-controlled homeopathy trials were matched with 110 conventional-medicine trials; when only the larger, higher-quality trials were counted, homeopathy showed no significant effect (odds ratio 0.88, 95% CI 0.65-1.19, 8 trials) while conventional drugs kept a clear one (0.58, 0.39-0.85, 6 trials).
Moderate evidence.
Caveats
Only 8 and 6 trials in the filtered sets; quality judged from published reports; unpublished trials probably missed; search ended January 2003; tests specific remedy effects only. Publicly funded (Swiss federal program), no declared conflict.
Health effects / Other evidence
Small and lower-quality trials overstated benefit in both homeopathy and conventional medicine, but the bias markers (not double-blind, unclear randomization, not in MEDLINE) significantly inflated homeopathy results while mostly not reaching significance for conventional drugs.
Moderate evidence: Ratios of odds ratios with wide intervals; conventional-medicine sample limited to one matched trial per pair; reporting quality assessed from publications.
Health effects / Other evidence
The authors concluded the clinical effects seen in homeopathy trials are compatible with placebo and context effects, and that future research should address those context effects and homeopathy's place in health care instead of more placebo-controlled trials.
Moderate evidence: Authors' interpretation; they state that proving a negative is impossible and that the study addressed only specific remedy effects.
Health effects / Other evidence
Homeopathic remedies are made by repeatedly diluting a substance (about 1 part in 100 million by the fourth dilution of Hahnemann's scheme) and shaking it; Hahnemann claimed this 'potentization' kept or increased potency as a 'dematerialized spiritual force', an idea orthodox physicians called nonsense from the start.
Moderate evidence: Describes the doctrine and its reception, not an experimental test of it; the dilution arithmetic assumes the starting substance is fully soluble.
Source: Loudon 2006, Journal of the Royal Society of Medicine Link to this fact
Health effects / Other evidence
In a 2007 study of 1,207 migraine patients at one Atlanta headache clinic, 43.7% said perfume or odor triggered their attacks at least occasionally, but only 10.7% said it did so in most attacks (more than 66%).
Moderate evidence.
Caveats
Single specialist clinic, 84% of the wider cohort women; self-reported recall with no diary or exposure measurement; 'perfume or odor' is one lumped item, not fragrance products specifically.
Source: Kelman 2007, Cephalalgia, Cephalalgia Link to this fact
Health effects / Other evidence
Migraine patients in this clinic named an average of 6.7 triggers (of 15 listed), and 61% named 4 to 9, so odor is rarely the only trigger and avoiding it alone is unlikely to prevent most attacks.
Moderate evidence: Single-clinic recall; the list had 15 items, which inflates counts compared with open questions.
Source: Kelman 2007, Cephalalgia, Cephalalgia Link to this fact
Health effects / Other evidence
In a 2006 Boston experiment, 82 healthy volunteers were told an identical placebo pill was a new opioid painkiller costing either $2.50 or $0.10; 85% of the 'regular price' group rated electric-shock pain lower afterward versus 61% of the 'discount' group (P = .02).
Moderate evidence: One small paid-volunteer lab study (n = 41 per arm), induced brief pain in healthy adults, no no-pill control; authors call for replication.
Source: Waber 2008, JAMA Link to this fact
Health effects / Other evidence
In 2009 the FEMA authors reported that eight flavor ingredients had lost FEMA GRAS status over the program's life (including calamus, brominated vegetable oil, and musk ambrette) and that a ninth, cinnamyl anthranilate, was abandoned by industry then banned by FDA; FDA also prohibits coumarin and safrole in food, and neither is FEMA GRAS.
Moderate evidence.
Caveats
Industry-reported list; reasons for each revocation are not given in this paper. Written by two FEMA employees and supported by FEMA (the flavor industry trade association); it is the industry's own account of its own program.
Source: Hallagan 2009, Food and Chemical Toxicology Industry-funded Link to this fact
Health effects / Other evidence
In 21 fragrance-allergic patients, an hour of breathing 1,000 µg/m³ isoeugenol or HICC produced no significant change in lung function (FEV1), methacholine airway sensitivity, blood eosinophils, CRP, or mast cell tryptase.
Moderate evidence.
Caveats
Small sample; methacholine thresholds could be evaluated in only 4-7 people per condition; a nonsignificant trend toward more airway sensitivity the next day; one hour, pure single compounds.
Source: Schnuch 2010, British Journal of Dermatology Link to this fact
Health effects / Other evidence
In that review, placebo effects appeared mainly on self-reported outcomes: pain (standardized mean difference -0.28) and nausea (-0.25) improved modestly, while laboratory outcomes did not improve, and the authors say a true effect cannot be told apart from biased reporting.
Moderate evidence: Continuous-outcome pooling had an asymmetrical funnel plot (small trials larger effects); effect on nausea about 7 mm on a 100 mm scale.
Source: Hróbjartsson 2010, Cochrane Database of Systematic Reviews Link to this fact
Health effects / Other evidence
Among the seven pain trials judged low risk of bias, four similar German acupuncture trials found a large sham-treatment effect (about 16 mm on a 100 mm pain scale, SMD -0.68) while the other three found about 3 mm (SMD -0.13), so the size of the placebo effect on pain varied from negligible to clinically relevant.
Moderate evidence: The four acupuncture trials had overlapping authors and near-identical design, so they are not independent replications; the review does not explain the gap.
Source: Hróbjartsson 2010, Cochrane Database of Systematic Reviews Link to this fact
Health effects / Other evidence
The review's strongest-supported single-condition results were small and uncertain: significant placebo effects on asthma (SMD -0.35, 4 trials) and phobia (SMD -0.63, 3 trials, 57 patients) rest on very small, high-risk-of-bias trials, and no significant effect was found for smoking, depression, dementia, obesity, hypertension, insomnia, or anxiety.
Moderate evidence: Wide confidence intervals; low-quality evidence for most single conditions; 'no significant effect' is not proof of no effect.
Source: Hróbjartsson 2010, Cochrane Database of Systematic Reviews Link to this fact
Health effects / Other evidence
In a randomized trial of 1,621 Australian adults, those told to apply SPF 16 sunscreen daily for about 4.5 years developed about half as many melanomas over roughly 15 years as those using sunscreen as they pleased (11 vs 22 cases; hazard ratio 0.50).
Moderate evidence.
Caveats
Melanoma was a secondary end point, the confidence interval includes no effect, one trial with 33 events, controls used sunscreen at their own discretion (no placebo), not blinded; lead author discloses L'Oreal research funding (trial itself NHMRC-funded).
Source: Green 2011, Journal of Clinical Oncology Industry-funded Link to this fact
Health effects / Other evidence
The Nambour sunscreen was SPF 16 containing 8% octinoxate and 2% avobenzone, two of the actives FDA's later absorption trials found above its 0.5 ng/mL threshold; this melanoma trial measured no absorption or toxicity outcomes.
Moderate evidence: Benefit data and absorption data come from different studies of different products; neither shows net harm or net benefit of the chemicals.
Source: Green 2011, Journal of Clinical Oncology Industry-funded Link to this fact
Health effects / Other evidence
A 2012 systematic review of published case reports and case series found 38 reports of harm from homeopathy, covering 1159 patients and including four deaths.
Moderate evidence.
Caveats
Case-report evidence cannot give how often harm occurs; about 92% of the 1159 is one poison-unit series (von Mach 2006, 1070 mostly mild, undescribed cases). Authors are known critics of homeopathy.
Source: Posadzki, Alotaibi & Ernst 2012, Int J Clin Pract, International Journal of Clinical Practice Link to this fact
Health effects / Other evidence
The documented direct harms of homeopathic products were mostly allergic reactions and poisoning from remedies containing arsenic, mercury, thallium, aconite, or other toxic substances, with Rhus toxicodendron (poison ivy) the most often implicated remedy.
Moderate evidence: Cause was largely the primary authors' judgment; several reports lacked detail; case reports only.
Source: Posadzki, Alotaibi & Ernst 2012, Int J Clin Pract, International Journal of Clinical Practice Link to this fact
Health effects / Other evidence
In 94.7% of the harm cases the remedy was below 12C, the dilution beyond which essentially no original molecule remains, so the reported poisonings and allergies came from remedies that still contained measurable material, not from ultra-dilutions.
Moderate evidence: Potency was not reported in many cases; the paper does not test any ultra-dilution for toxicity.
Source: Posadzki, Alotaibi & Ernst 2012, Int J Clin Pract, International Journal of Clinical Practice Link to this fact
Health effects / Other evidence
Homeopathy can also cause harm indirectly: in 8 reports (16 patients) people used it instead of effective treatment for melanoma, malaria prevention, sarcoidosis, atopic dermatitis, and childhood infections, and outcomes included death, permanent hypertension, and hydrocephalus.
Moderate evidence: Whether this is an adverse effect of homeopathy or of poor care is debatable (the authors say so); indirect harm is probably underreported; case reports only.
Source: Posadzki, Alotaibi & Ernst 2012, Int J Clin Pract, International Journal of Clinical Practice Link to this fact
Health effects / Other evidence
In a re-analysis of 58 arthritic dogs that received only a placebo pill for 42 days, 50% of owners said their dog had improved and blinded veterinarians graded 45-48% as improved, even though a force-plate measure of limb loading showed no change in 79% of dogs (46 of 58).
Moderate evidence: One drug-trial placebo arm (7 US sites) with no untreated arm; Novartis supplied the data. Force plate measures limb loading, not pain or quality of life.
Source: Conzemius 2012, Journal of the American Veterinary Medical Association Industry-funded Link to this fact
Health effects / Other evidence
Using the force plate as the yardstick, owners over-reported improvement 56.9% of the time and blinded veterinarians 40-45% of the time (39.7% walk, 44.8% trot, 43.1% joint palpation), and the effect grew over 42 days (significant at every visit, P < 0.001).
Moderate evidence.
Caveats
The paper's own definition of 'caregiver placebo effect' (report better than the plate); not separated from regression to the mean or natural fluctuation. The abstract misprints owners as 39.7%; Results say 56.9%.
Source: Conzemius 2012, Journal of the American Veterinary Medical Association Industry-funded Link to this fact
Health effects / Other evidence
Owner and veterinarian ratings agreed poorly with the objective measure over time (Cronbach alpha 0.18 to 0.37, versus 0.96-0.97 for the force-plate measures), so owner testimonials about an arthritis treatment for a pet are weak evidence of benefit on their own.
Moderate evidence: Single study of dogs with osteoarthritis; reliability of ratings across visits, not accuracy for pain.
Source: Conzemius 2012, Journal of the American Veterinary Medical Association Industry-funded Link to this fact
Health effects / Other evidence
Dogs' arthritis lameness changes little over 50 days without treatment on objective testing: only 20.7% changed by 5% or more and 3.4% by 10% or more (12.1% improved and 8.6% worsened at the 5% cutoff).
Moderate evidence: Placebo-arm dogs only; the 5%/10% cutoffs and the 2-of-3 visit rule were the authors' choices.
Source: Conzemius 2012, Journal of the American Veterinary Medical Association Industry-funded Link to this fact
Health effects / Other evidence
The 'water memory' explanation for homeopathy is at odds with measurements: hydrogen bonds in liquid water rearrange in under about 50 femtoseconds, far too briefly to carry the identity of a diluted substance.
Moderate evidence: Relies on one cited experiment; the paper does not review the wider water-structure literature, and proponents dispute the relevance of such time scales.
Source: Grimes 2012, Focus Altern Complement Ther, Focus on Alternative and Complementary Therapies Link to this fact
Health effects / Other evidence
The 1988 Nature experiment claiming diluted antibody still triggers allergy cells was judged an artifact of observer bias by an investigative team, and independent labs reportedly failed to replicate it.
Moderate evidence: Secondhand and compressed account; Grimes also lists Belon 1999 as a failure although its title reports an inhibition effect; see the Maddox and Davenas pages.
Source: Grimes 2012, Focus Altern Complement Ther, Focus on Alternative and Complementary Therapies Link to this fact
Health effects / Other evidence
In a Belgian referral clinic, 17 of 44 workers (39%) who had asthma symptoms they linked to cleaning or disinfecting products had a measurable asthmatic reaction (a sustained 20% or greater fall in FEV1) when exposed to their own work products under controlled conditions.
Moderate evidence.
Caveats
Retrospective single-center chart review of workers referred (mostly via compensation claims); not a prevalence among cleaners. No blinding described; products were commercial mixtures and exposure was not measured.
Health effects / Other evidence
Of the 17 workers who reacted to cleaning products, 10 reacted to products containing quaternary ammonium compounds (mainly benzalkonium and didecyldimethylammonium chlorides), 3 to glutaraldehyde, 1 to both, and 2 to ethanolamines.
Moderate evidence.
Caveats
The causal agent was inferred from product ingredients, not isolated; 9 workers challenged with QAC-containing products did not react (derived from Table 1). Authors name lack of quantitative exposure assessment as the main limitation.
Health effects / Other evidence
FDA has not recommended any validated screening test for endocrine disruption in food additives, and in 2012 rejected an ICCVAM-validated estrogen-receptor assay that EPA, the Consumer Product Safety Commission, and OECD accepted.
Moderate evidence: Pew authors' account of FDA's position as of 2013; FDA scientists argue the Redbook's multigenerational studies can reveal endocrine activity. Later FDA actions are not covered.
Source: Maffini 2013, Comprehensive Reviews in Food Science and Food Safety Link to this fact
Health effects / Other evidence
FDA's Redbook neurotoxicity guidance (2000) lists only observational signs such as seizure, tremor and activity, with no learning, or memory tests, while EPA and OECD adopted such behavioral tests years earlier; the review says FDA's test guidance for subtle behavioral effects lags those agencies.
Moderate evidence: Describes test guidance as of 2013, not evidence of any additive's effect on behavior; the paper takes no position on additives and hyperactivity.
Source: Maffini 2013, Comprehensive Reviews in Food Science and Food Safety Link to this fact
Health effects / Other evidence
A 2014 systematic review of placebo-controlled trials of homeopathy in animals found only 18 peer-reviewed trials, and only 2 of them were both methodologically reliable and free of funding from a homeopathic pharmacy.
Moderate evidence.
Caveats
Both authors are employed by homeopathy charities (British Homeopathic Association; Carstens Foundation); assessments were by the two authors; trial authors not contacted; literature search to 2013.
Source: Mathie & Clausen 2014, Vet Rec, Veterinary Record Link to this fact
Health effects / Other evidence
In that review the two reliable, interest-free trials disagreed: one found homeopathic Coli reduced diarrhea in piglets (odds ratio 3.89, 95% CI 1.19-12.68), the other found no benefit of individualized homeopathy over placebo for cow mastitis (SMD -0.31, CI -0.97 to 0.34).
Moderate evidence.
Caveats
Each result is a single small trial; the piglet trial treated sows and measured piglets and was not replicated; the mastitis CI does not exclude a benefit; the authors call the evidence 'modest' and draw no general conclusion.
Source: Mathie & Clausen 2014, Vet Rec, Veterinary Record Link to this fact
Health effects / Other evidence
The review's authors conclude that mixed findings from the only two reliable trials preclude generalizable conclusions about the efficacy of any homeopathic medicine, or of individualized homeopathy for any condition, in animals; there is no reliable (or any) research evidence for other conditions.
Moderate evidence: The authors work for homeopathy organizations and write that neither critics nor advocates should feel 'gloom nor glee'; an absence of reliable evidence is not proof of no effect.
Source: Mathie & Clausen 2014, Vet Rec, Veterinary Record Link to this fact
Health effects / Other evidence
Of 18 veterinary homeopathy trials, 11 were at high risk of bias, only 7 used adequate random sequence generation, and 16 named no main outcome; none reported a power calculation.
Moderate evidence: Judgments by the two reviewers; some papers were too poorly reported to appraise easily, so 'unclear' ratings may hide either flaws or good conduct.
Source: Mathie & Clausen 2014, Vet Rec, Veterinary Record Link to this fact
Health effects / Other evidence
Researchers who study the placebo effect say it works on how symptoms are felt and reported (pain, nausea, fatigue, hot flashes), not on the underlying disease: they report no evidence that placebos shrink tumors, and in an asthma experiment placebo left lung function (FEV1) unchanged while relieving perceived symptoms.
Moderate evidence.
Caveats
Expert opinion piece with 5 references, not a systematic estimate; the asthma result is one crossover experiment. A Cochrane review reaches the same subjective-versus-objective pattern but reads it as possible reporting bias.
Source: Kaptchuk & Miller 2015, New England Journal of Medicine Link to this fact
Health effects / Review of other studies
A 2016 review of professional clothing-cleaning methods reports that perchloroethylene (perc), the most widely used dry-cleaning solvent, is classed by the International Agency for Research on Cancer as probably carcinogenic to humans (group 2A), with kidney cancer, liver cancer, and non-Hodgkin lymphoma among the cancers linked to it in workers or residents.
Moderate evidence.
Caveats
Secondary review funded by Australian Wool Innovation (wool industry body); 'systematic' is a self-label with no study counts or appraisal; relies on agency, trade and manufacturer sources. The review restates other studies' findings; it contains no new exposure or outcome data, and the IARC class is a hazard rating, not a risk estimate for wearing dry-cleaned clothes.
Source: Troynikov 2016, Journal of Environmental Management Industry-funded Link to this fact
Health effects / Other evidence
Exercise, painkillers (NSAIDs) and alcohol can lower the threshold at which a food triggers anaphylaxis: in 74 cases of cofactor-enhanced food allergy NSAIDs were involved in 58%, exercise in 52.7%, and alcohol in 12.2%.
Moderate evidence: Clinic case series without matched controls; the review reports them and does not re-analyze them.
Source: Skypala 2019, Frontiers in Immunology Link to this fact
Health effects / Review of other studies
Some people react to tiny amounts of hidden allergens: eliciting doses of 0.05 mg for mustard and 0.5 mg for lupin in peanut-sensitized adults, with celeriac ED10 1.6 mg protein.
Moderate evidence: Secondary reporting of other groups' challenge data; thresholds vary by person and food matrix.
Source: Skypala 2019, Frontiers in Immunology Link to this fact
Health effects / Measured in people or real products
Controlled inhalation challenges have shown asthma reactions to quaternary ammonium disinfectants (quats, such as benzalkonium chloride) in individual workers: one review counted 37 reported cases, 29 confirmed by a specific inhalation challenge.
Moderate evidence.
Caveats
Case reports and clinic series, mostly without exposure levels, named compounds, or fully clear asthma history; the counting review is a non-systematic, agency-cited sample funded by a cleaning-products trade association (HCPA).
Source: LaKind 2019, Allergy, Asthma & Clinical Immunology Industry-funded Link to this fact
Health effects / Measured in people or real products
Whether quats trigger asthma by true allergic sensitization or by delayed irritation is unresolved: quats-specific IgE was negative in the reported cases and the two mechanisms can look the same clinically.
Moderate evidence: Review of small case reports; funded by a cleaning-products trade association; no mechanistic data of its own.
Source: LaKind 2019, Allergy, Asthma & Clinical Immunology Industry-funded Link to this fact
Information, not medical advice. See also: myths we won’t tell you.