Nontargeted chemical testing identified potentially harmful unlabeled chemicals across personal care and cleaning products, including products marketed as “natural,” “non-toxic,” and “fragrance-free,” according to research published in Environment & Health.
Among 113 products tested, 98% contained at least 1 unlabeled potentially harmful chemical, 85% contained at least 1 unlabeled chemical with known or suspected health hazards, and 58% failed the researchers’ contamination criteria.
Kim Schultz of the Million Marker Research Institute in Berkeley, California, and colleagues analyzed products across 12 categories using high-resolution two-dimensional gas chromatography time-of-flight mass spectrometry with machine learning–assisted identification. The analysis detected 3,141 chemical features, including 303 compounds assigned confirmed, probable, or tentative identifications. Products contained an average of 51 detected chemicals compared with 16 ingredients listed on labels and an average of 4 unlisted potentially harmful chemicals per product.
Unlabeled potentially harmful fragrance ingredients were detected in 39% of products, potentially harmful botanical constituents in 37%, and synthetic contaminants in 56%. Frequently detected compounds included citronellol in 70 products, myrcene in 46, cyclohexasiloxane in 22, phenethyl alcohol in 20, and benzyl chloride in 19. Cyclohexasiloxane reached estimated abundances of up to 2,200 ppm, while dimethyl oxazolidine exceeded 200 ppm, although these estimates were not intended to quantify exposure or clinical risk.
Label-based classifications did not reliably distinguish products that subsequently failed contamination testing. Among 27 products initially rated “Go,” 59% failed, and failure prevalence did not differ significantly among the investigators’ Go, Slow, and Stop categories. Twenty-nine products, or 26%, contained chemicals contradicting label claims, including products advertised as free of phthalates, parabens, fragrance, silicone, or oxybenzone. Some products labeled “hypoallergenic” contained allergens or sensitizers.
The products were purchased in 2021 and 2022 and were not selected to represent the broader market. The researchers noted that COVID-19–related supply-chain disruptions may have influenced product composition. The analysis evaluated chemical hazard rather than clinical risk, could not definitively establish the sources of individual contaminants, and did not account for exposure dose, route, or bioavailability.
The authors declared no competing financial interests; the study was supported by the Million Marker Research Institute, the Taber Trust, Michael Dubrovsky, and additional individual donors.
Source: Environment & Health
