How to read fragrance-safety studies: hazard, exposure and weight of evidence

Fragrance-safety research is easily flattened into “animals reacted, therefore dangerous” or “one test was negative, therefore safe”. Neither is enough. Hazard, exposure, relevance to humans and the direction of the whole evidence base have to be separated. Another way to flatten it is to swap the route of administration: an oral trial's conclusion cannot hold up an olfactory claim.

Coumarin: species use different metabolic routes

Lake’s 1999 review found marked species differences in coumarin metabolism. Most humans mainly use the detoxifying 7-hydroxylation route, while rats and mice rely more on a 3,4-epoxidation route that can generate toxic metabolites. Animal findings cannot be translated word for word into a human conclusion.

Nor does species difference erase concern. A 2010 human-centred assessment found a subgroup unusually sensitive to liver effects and arrived at a tolerable daily intake of 0.1 mg/kg body weight. Risk assessment works by reconciling animal data, human observations and realistic exposure—not by choosing the most convenient paper.

Cashmeran: weight of evidence is not a vote count

A 2024 Cashmeran review integrated ToxCast, cellular, receptor, 90-day repeat-dose and reproductive-developmental evidence. One cellular assay gave a weak signal while most related evidence was negative; the overall weight-of-evidence conclusion did not predict endocrine effects.

That conclusion did not come from counting negative papers. Researchers considered model relevance, general cytotoxicity, whether a molecular event could progress to an adverse outcome, and whether independent lines supported one another.

Iso E Super: a mechanistic signal needs a complete path

OTNE is the name commonly used for Iso E Super in scientific literature. A 2024 study placed thyroid-related findings inside an adverse-outcome-pathway network, asking whether an initiating event connects to later biological changes. Mechanistic evidence is one section of an assessment, not a finished verdict on product use.

Raw material tests, formula concentration, dermal exposure, inhalation and aquatic release are different cases. They cannot share one unqualified label of safe or toxic.

Vetiver: in vitro activity is not a product claim

A 2017 vetiver review reported antimicrobial activity under laboratory conditions while stating that the oil had not shown a particularly notable new cosmetic value beyond fragrance. Candidate activity needs further study before it becomes efficacy copy.

Four questions to keep

For every safety paper, ask what hazard was measured, at what dose and route, in which model, and where the authors’ conclusion stops. If an article says only that an effect was found without those four details, it is not yet enough for a formulation decision.

Current IFRA Standards, supplier safety documents and local product law still govern actual use. Critical reading prevents exaggeration; it does not replace compliance calculations.

References

B. G. Lake, Coumarin metabolism, toxicity and carcinogenicity: relevance for human risk assessment, Food and Chemical Toxicology, 37(4), 423–453 (1999). PMID 10418958

K. Abraham et al., Toxicology and risk assessment of coumarin: focus on human data, Molecular Nutrition & Food Research, 54(2), 228–239 (2010). PMID 20024932

S. Tayal et al., Cashmeran as a potential endocrine disrupting chemical: What does the weight-of-the-evidence indicate?, Food and Chemical Toxicology, 184, 114351 (2024). PMID 38081530

E. Hulzebos et al., Determination of OTNE-induced thyroid effects within an adverse outcome pathway network, Food and Chemical Toxicology, 190, 114784 (2024). PMID 38834167

P. Burger et al., Vetiver Essential Oil in Cosmetics: What Is New?, Medicines, 4(2), 41 (2017). PMID 28930256