Estrogen exposure in a child from hair lotion used by her mother: Clinical and hair analysis data

Abstract

Premature estrogenic effects may result from exogenous exposure to estrogenic substances. We report the case of a 36-month-old girl who presented with vaginal bleeding, uterus enlargement, and thelarche. Questioning of the parents revealed that the child's mother had used hormone-based hair lotions on her own scalp and that the child was in the habit of playing with her mother's hair while falling asleep, and that the girl played with her mother's combs and the empty lotion vials. The onset of hyperestrogenic syndrome was temporally related to the handling of lotions containing ethynylestradiol 0.5%. Analysis of long scalp hairs from the girl and her mother identified ethynylestradiol in concentrations of 10.6 and 46.6 µg/g, respectively. Six months after the mother discontinued use of the estrogen-containing hair lotion, the girl's hyperestrogenic signs resolved. This case highlights the importance of obtaining histories of possible food and non-food environmental sources of contamination, the suitability of hair sampling to identify the origin of the contamination, and the opportunity to warn parents about hazards related not only to oral contraceptives, but also custom-compounded topical hormone preparations.

Keywords:

• Estrogens
• Children
• Pseudo-puberty
• Hair lotions
• Hair analysis

Introduction

Precocious puberty, gynecomastia, and anovulatory dysfunctional uterine bleeding are sometimes observed in prepubertal children following exposure to exogenous estrogens or estrogenic substances, although cause and effect are often not proven (13). For example, prepubertal gynecomastia was thought to be caused by indirect exposure to estrogen cream in a report by Felner and White (4), although direct confirmation of absorption of an estrogenic substance was not shown. We report a child who was indirectly exposed to hair lotions used by her mother that contained high concentrations of the xenobiotic hormone ethynilestradiol. We obtained direct evidence of hormone absorption and storage in the hair of both mother and daughter.

Case report

A 36-month-old girl presented with a history of recurrent vaginal bleeding, each lasting a few days. She was healthy, with normal weight (16 kg) and height (99 cm). She achieved developmental milestones appropriately and had no recent history of illness or surgery and no past use of medications or cosmetic ointments. Clinical examination revealed breast enlargement 2.3 cm (left) and 2.5 cm (right), and nipple hyper-pigmentation. Ultrasonography showed an enlarged uterus (41×–18×–19 mm) with evidence of endometrial rime, and small ovaries. The clinical pattern suggested an exogenous hormonal exposure. A few weeks later, when queried about possible exposures to endocrine disrupters from sources other than food, her mother acknowledged that she was using a custom-compounded hair lotion. The girl used to caress and play with her mother's hair before falling asleep, and would play with her mother's combs and the empty hair lotion vials. Hair samples were gathered from the girl and both parents. The mother was asked to not use topical products. Six months later the girl had a complete regression of signs of feminization and a normalization of hormonal data and uterine size.

Laboratory analysis

Hair lotions obtained from the mother were analyzed for the presence and concentration of estrogenic hormones using a liquid chromatographic technique coupled both with UV-diode array and mass spectrometer detectors (1100 MSD Trap from Agilent Tech Italia, Milan-Italy). Briefly, galenic preparations were diluted 1:100 in methanol and a 20 µl sample was injected directly into the chromatographic system under the following conditions: mobile phases: methanol and 1% acetic acid ph 2.5; linear gradient from 5% methanol to 100% in 20 min, then 100% methanol for 10 min. Column: symmetry C18 3.0 ×– 150 mm (5 micron) (Waters Milford, MS-USA), flow rate: 300 µl/min. UV detection at 240 and 280 nm, while UV spectra acquisition from 220 to 350 nm; mass spectrometry acquisition in full scan (50–500 atom mass unit - amu), using nitrogen as dry gas and setting the nebuliser at 20 psi and the temperature at +350°C. Hair analytical procedures followed the ones described by Thieme et al. (5). Identification was carried out by high resolution gas chromatography-low resolution tandem mass spectrometry, monitoring one precursor and two products ions; respectively, for estradiol isomers, m/z 416, 326 and 285; for ethynylestradiol, m/z 425, 407 and 231. Quantification was performed with the internal standard method; to this purpose, 20 ng of deuterated 17-α estradiol (m/z 419) were added to the matrix at the beginning of the procedure.

Results

Blood and urine samples of the girl and her parents were negative for synthetic hormones (e.g., boldenone, diethylstilbestrol, ethynilestradiol, 19 nor-testosterone, stanozolol, trenbolone, zeranol). Analyses revealed low plasma levels of gonadotropin and estradiol, and no tumor markers in the girl. However, the hair lotions used by the mother contained the hormone ethynylestradiol (Table 1). Hair samples of the girl and her mother were positive for ethynylestradiol in concentrations of 10,600 and 46,600 ng/g, respectively (mean of two determinations on the same extract).

Table 1.  Drugs found in hair lotions used by mother

Compound	Dosage	Lotion
Cyproterone acetate	2%	1
Progesterone base	1%	1
Cyproterone acetate	5%	2
Ethinyl estradiol	0.5%	2

The preparations included minoxidil 5% and ethanol as a vehicle. Lotions were prepared at different times and sometimes included different dosing instructions. All were intended for dermatological use.

Discussion

Often investigators have difficulty determining the exact mechanism of drug exposure (1,47). One important source of uncertainty is the time lag between the observation of signs and the biological matrix sampling (several weeks, in the present case), which allows active metabolites to be cleared from blood and urine. We report here the demonstration of contamination with xenobiotic estrogenic substances through cutaneous absorption in a prepubertal girl, with long-term accumulation in a stable matrix such as hair. Our investigation was facilitated by the use of the xenobiotic hormone ethynylestradiol and by the length of the hair of the girl and her mother.

In the same period of time two male siblings (4 and 2 years) with gynecomastia came to our attention. However, hormone identification was unsuccessful in hair samples from these children and their mother, presumably because of recent haircuts, and possibly due to the lower lipophilicity of the hormone (17-α estradiol) contained in the lotion (instead of ethynylestradiol) that could affect the accumulation rate in scalp hair.

This case demonstrates that analytical methods performed on biological matrices which accumulate steroids for a long period of time, such as hair (8), may identify the source of contamination. Exposure to steroids via transdermal preparations enhances their incorporation as non-polar (non-conjugated) compounds into the hair. Oral exposure often does not result in drug deposition in the hair, due to extensive drug metabolism (5).

The clinical pattern in the girl is consistent with the presumed hormone exposure. Two milliliters of lotion (10 mg ethynylestradiol) were applied over the 600–800 cm² of the mother's scalp daily, resulting in an average hormone dose of 12 to 17 µg/cm². If we assume that estrogen bioavailability administered as a transdermal preparation is of the order of 20%, as suggested by Walters et al. (9), about 10–20 cm² of the girl's hands were exposed daily to hormone residues on mother's scalp. This exposure may result in an absorbed dose of ethynylestradiol of the same order of magnitude as a contraceptive dosage in adult woman weighing 60 kg. In fact, commercial contraceptive patches containing ethynylestradiol deliver 20–25 µg/day (0.35–0.40 µg/kg body weigh per day to an adult woman) (10). Given the girl's lower weight and less keratinized skin, we hypothesize that contact between her hands and the mother's hair allowed enough drug absorption to give signs of estrogen exposure.

However, our investigations have some limitations. Cutaneous exposure of the girl to the hair lotion resulted in absorption sufficient to produce the clinical features of estrogen intoxication, but we cannot exclude other possibilities. For example, children might have ingested the lotion whilst unsupervised or the lotion might have been applied to children's hair by other children.

Regression of clinical signs of estrogenization does not exclude medium- and long-term effects. Short duration exposure to estrogenic substances during developmental ages (2) may produce long-term damage: this was found both in cases in which estrogens were probably assumed as residues in the food (11) and in cases in which active substances were absorbed through the use of paraben–containing cosmetics (12).

Conclusion

The health risks from dermatologic preparations may be underestimated. Several preparations contain highly effective drugs in the form of creams or slow-release patches (anti-inflammatory agents, analgesics, hormones, etc.). These are applied “for external use” and the drugs which are released into the skin may be misperceived by consumers to be safe because of the presumption that only small amounts can be absorbed. In any case, our patient was exposed by some route that was sufficient to cause hyperestrogenism. These types of preparations, as well as empty vials, should be kept out of children's hands.

Acknowledgments

The authors would like to thank Dr. Maurizio Fiori from Istituto Superiore di Sanità and Dr. Guglielmo Dusi and Dr. Michele Curatolo from Istituto Zooprofilattico Sperimentale di Brescia for their analytical assistance. Thanks are also due to Mrs. Carla Campanella for her expert editorial work. This work was supported by a Ministry of Health Project, an Italian National Institute of Health Programme.