Sunscreens and thyroid function in humans after short-term whole-body topical application: a single-blinded study

Sir, The three sunscreens benzophenone‐3 (BP‐3), 3‐(4‐methylbenzylidene) camphor (4‐MBC) and octylmethoxycinnamate (OMC) are widely used in commercial sun protection products at a maximum use concentration of 10% for OMC and BP‐3 and 4–6% for 4‐MBC in the European Union. All three compounds have shown to have oestrogenic effects in vitro and in animal studies.1 We recently reported in the same groups included in this study a substantial systemic absorption and urinary excretion of the three sunscreens in men and postmenopausal women after whole‐body topical application; no effect was observed on the levels of reproductive hormones.2 However, recent studies in rats have demonstrated a direct effect on thyroid hormone synthesis by OMC and 4‐MBC after 12 weeks of oral treatment.3, 4 We now investigate for the first time whether the three sunscreens BP‐3, OMC and 4‐MBC have any effect on the hypothalamic–pituitary–thyroid axis in humans after whole‐body topical application. In this 2‐week single‐blinded study approved by the ethics committee of Copenhagen and Frederiksberg [No. (KF) 11–088/03], 15 young men (age range 23–29 years, mean 26) and 17 postmenopausal women (age range 54–86 years, mean 65) were assigned to daily whole‐body topical application (2 mg cm−2) of basic cream in week 1 and with a mixture of the three sunscreens at 10% (w/w) each in week 2. The daily mean ± SD amount of cream applied over 4 days for each subject was 40 ± 3 g for men and 35 ± 3 g for women. All subjects were healthy and not taking any regular medication. Thyroid and postmenopausal status was verified about 4 weeks before enrolment, with a blood test. Hormone levels were measured by commercially available automated immunoassay systems. Normal distribution of data was verified by a Kolmogorov–Smirnov test and P‐values were calculated with a paired t‐test. We observed no biologically significant effects on hormone levels, indicating that the concentrations of sunscreen compounds absorbed were not capable of disturbing the homeostasis of thyroid hormones in adult humans. Minor but statistically significant differences in hormone levels between the two weeks were observed in free thyroxine (FT4), free triiodothyronine (FT3) and total triiodothyronine (T3) for men (Fig. 1) and in thyroxine‐binding globulin (TBG), FT4, FT3, total thyroxine (T4) and T3 for women (Fig. 2). These differences did not seem to be related to the exposure to sunscreens and at least some of these statistically significant differences between the two weeks may be chance findings due to mass significance (when six different hormones are tested at seven different time‐points in both genders approximately four significant findings can be expected just by chance when using a significance level of 0·05). Normal biological variations in hormone levels may have contributed to some of the differences. This is supported by the fact that FT4 and T3 in both genders and FT3 in men and TBG and T4 in women were generally at a higher level during the control week and before the first application of the active formulation. There was no effect on the thyroid‐stimulating hormone (TSH) levels and additionally we did not observe any increase in the T4 or T3 level in either gender. Oestrogen is known to increase serum TBG concentrations, thereby increasing serum T4 concentrations. Serum FT4 concentrations, however, remain normal.5 We did not observe any of these effects in either gender. Mean ± SD serum concentrations of hormones in men during the week of treatment with control cream (open circles) and with cream containing sunscreens (filled circles). Significant differences in hormone levels at the same time points between the two weeks are indicated by *P < 0·05, **P < 0·01. Differences were compared using a paired t‐test. TBG, thyroxine‐binding globulin; TSH, thyroid‐stimulating hormone; T4, total thyroxine; FT4, free thyroxine; T3, total triiodothyronine; FT3, free triiodothyronine. Mean ± SD serum concentrations of hormones in women during the week of treatment with control cream (open squares) and with cream containing sunscreens (filled squares). Significant differences in hormone levels at the same time points between the two weeks are indicated by *P < 0·05, **P < 0·01. Differences were compared using a paired t‐test. TBG, thyroxine‐binding globulin; TSH, thyroid‐stimulating hormone; T4, total thyroxine; FT4, free thyroxine; T3, total triiodothyronine; FT3, free triiodothyronine. In terms of exposure, our short‐term study is not comparable with the rat studies,3, 4 as animals were fed for 12 weeks orally with OMC and 4‐MBC. Hence, average highest total intake of 4‐MBC can be estimated to be above 26 g per animal during this period, which is > 10% of the animals’ body weight. However, recent studies6, 7 showed that after single oral and topical administration of 4‐MBC to rats and humans a higher concentration of 4‐MBC was measured in rats and humans after topical administration as compared with rats after oral administration. This is due to an extensive first‐pass metabolism in the liver and/or gut wall. The authors also estimated that < 1% of 4‐MBC is absorbed in humans. In the light of these findings the authors concluded that the extent of hormonal effects of 4‐MBC in rodents depends on bioavailability of 4‐MBC, and the extent of formation of metabolites and oral toxicity of 4‐MBC are less relevant as a basis for human health risk assessment, given that the major pathway of 4‐MBC exposure for humans is dermal. Furthermore, under natural conditions humans use sunscreens at < 2 mg cm−2, typically between 0·5 and 1·5 mg cm−2, which gives lower exposure concentrations and lower doses in comparison with our study.8 We applied all three sunscreen compounds simultaneously and might expect an additive or possibly even a synergistic effect. The above findings may indicate that adult humans have a high threshold for regulation of thyroid hormone levels in terms of sunscreen use, which is in accordance with the fact that treatment of postmenopausal women with conjugated oestrogens caused a significant rise in TSH, while the FT4 index remained unchanged.5 Furthermore, another study showed that four oral contraceptives had only minor effects on thyroid function with a rise in T4 and T3 in young women (age range 18–35 years), indicating that thyroid function is not, or is only slightly affected by, oral contraceptives.9 Alternatively, the tested sunscreen compounds have a very low potency for hormone disruption in this short‐term study. Conflicts of interest: none declared.