Final Amended Report on the Safety Assessment of Methylparaben,Ethylparaben,Propylparaben,Isopropylparaben,Butylparaben,Isobutylparaben,and Benzylparaben as used in Cosmetic Products 1

Abstract

Parabens is the name given to a group of p-hydroxybenzoic acid (PHBA) esters used in over 22,000 cosmetics as preservatives at concentrations up to 0.8% (mixtures of parabens) or up to 0.4% (single paraben). The group includes Methylparaben, Ethylparaben, Propylparaben, Isopropylparaben, Butylparaben, Isobutylparaben, and Benzylparaben. Industry estimates of the daily use of cosmetic products that may contain parabens were 17.76 g for adults and 378 mg for infants. Parabens in cosmetic formulations applied to skin penetrate the stratum corneum in inverse relation to the ester chain length. Carboxylesterases hydrolyze parabens in the skin. Parabens do not accumulate in the body. Serum concentrations of parabens, even after intravenous administration, quickly decline and remain low. Acute toxicity studies in animals indicate that parabens are not significantly toxic by various routes of administration. Subchronic and chronic oral studies indicate that parabens are practically nontoxic. Numerous genotoxicity studies, including Ames testing, dominant lethal assay, hostmediated assay, and cytogenic assays, indicate that the Parabens are generally nonmutagenic, although Ethylparaben and Methylparaben did increase chromosomal aberrations in a Chinese Hamster ovary cell assay. Ethylparaben, Propylparaben, and Butylparaben in the diet produced cell proliferation in the forestomach of rats, with the activity directly related to chain length of the alkyl chain, but Isobutylparaben and Butylparaben were noncarcinogenic in a mouse chronic feeding study. Methylparaben was noncarcinogenic when injected subcutaneously in mice or rats, or when administered intravaginally in rats, and was not cocarcinogenic when injected subcutaneously in mice. Propylparaben was noncarcinogenic in a study of transplacental carcinogenesis. Methylparaben was nonteratogenic in rabbits, rats, mice, and hamsters, and Ethylparaben was nonteratogenic in rats. Parabens, even at levels that produce maternal toxicity, do not produce fetal anomalies in animal studies. Parabens have been extensively studied to evaluate male reproductive toxicity. In one in vitro study, sperm were not viabile at concentrations as low as 6 mg/ml Methylparaben, 8 mg/ml Ethylparaben, 3 mg/ml Propylparaben, or 1 mg/ml Butylparaben, but an in vivo study of 0.1% or 1.0% Methylparaben or Ethylparaben in the diet of mice reported no spermatotoxic effects. Propylparaben did affect sperm counts at all levels from 0.01% to 1.0%. Epididymis and seminal vesicle weight decreases were reported in rats given a 1% oral Butylparaben dose; and decreased sperm number and motile activity in F₁ offspring of rats maternally exposed to 100 mg/kg day^–1 were reported. Decreased sperm numbers and activity were reported in F₁ offspring of female rats given Butylparaben (in DMSO) by subcutaneous injection at 100 or 200 mg/kg day^–1, but there were no abnormalities in the reproductive organs. Methylparaben was studied using rats at levels in the diet up to an estimated mean dose of 1141.1 mg/kg day^–1 with no adverse testicular effects. Butylparaben was studied using rats at levels in the diet up to an estimated mean dose of 1087.6 mg/kg day^–1 in a repeat of the study noted above, but using a larger number of animals and a staging analysis of testicular effects—no adverse reproductive effects were found. Butylparaben does bind to estrogen receptors in isolated rat uteri, but with an affinity orders of magnitude less than natural estradiol. Relative binding (diethylstilbesterol binding affinity set at 100) to the human estrogen receptors α and β increases as a function of chain length from not detectable for Methylparaben to 0.267 ± 0.027 for human estrogen receptor α and 0.340 ± 0.031 for human estrogen receptor β for Isobutylparaben. In a study of androgen receptor binding, Propylparaben exhibited weak competitive binding, but Methylparaben had no binding effect at all. PHBA at 5 mg/kg day^–1 subcutaneously (s.c.) was reported to produce an estrogenic response in one uterotrophic assay using mice, but there was no response in another study using rats (s.c. up to 5 mg/kg day^–1) and mice (s.c. up to 100 mg/kg day^–1) and in a study using rats (s.c. up to 100 mg/kg day^–1). Methylparaben failed to produce any effect in uterotrophic assays in two laboratories, but did produce an effect in other studies from another laboratory. The potency of Methylparaben was at least 1000 × less when compared to natural estradiol. The same pattern was reported for Ethylparaben, Propylparaben, and Butylparaben when potency was compared to natural estradiol. In two studies, Isobutylparaben did produce an estrogenic response in the uterotrophic assay, but the potency was at least 240,000 × less than estradiol. In one study, Benzylparaben produced an estrogenic response in the uterotrophic assay, but the potency was at least 330,000 × less than estradiol. Estrogenic activity of parabens and PHBA was increased in human breast cancer cells in vitro, but the increases were around 4 orders of magnitude less than that produced by estradiol. Parabens are practically nonirritating and nonsensitizing in the population with normal skin. Paraben sensitization has occurred and continues to be reported in the case literature, but principally when exposure involves damaged or broken skin. Even when patients with chronic dermatitis are patch-tested to a parabens mix, parabens generally induce sensitization in less than 4% of such individuals. Many patients sensitized to paraben-containing medications can wear cosmetics containing these ingredients with no adverse effects. Clinical patch testing data available over the past 20 years demonstrate no significant change in the overall portion of dermatitis patients that test positive for parabens. As reviewed by the Cosmetic Ingredient Review (CIR) Expert Panel, the available acute, subchronic, and chronic toxicity tests, using a range of exposure routes, demonstrate a low order of parabens' toxicity at concentrations that would be used in cosmetics. Parabens are rarely irritating or sensitizing to normal human skin at concentrations used in cosmetics. Although parabens do penetrate the stratum corneum, metabolism of parabens takes place within viable skin, which is likely to result in only 1% unmetabolized parabens available for absorption into the body. The Expert Panel did consider data in the category of endocrine disruption, including male reproductive toxicity and various estrogenic activity studies. The CIR Expert Panel compared exposures to parabens resulting from use of cosmetic products to a no observed adverse effect level (NOAEL) of 1000 mg/kg day^–1 based on the most statistically powerful and well-conducted study of the effects of Butylparabens on the male reproductive system. The CIR Expert Panel considered exposures to cosmetic products containing a single parabens preservative (use level of 0.4%) separately from products containing multiple parabens (use level of 0.8%) and infant exposures separately from adult exposures in determining margins of safety (MOS). The MOS for infants ranged from ~6000 for single paraben products to ~3000 for multiple paraben products. The MOS for adults ranged from 1690 for single paraben products to 840 for multiple paraben products. The Expert Panel considers that these MOS determinations are conservative and likely represent an overestimate of the possibility of an adverse effect (e.g., use concentrations may be lower, penetration may be less) and support the safety of cosmetic products in which parabens preservatives are used.