Cosmetics use and age at menopause: is there a connection?

While the US Food and Drug Administration (FDA) monitors the chemicals that go into foods, drugs, and medical devices closely, cosmetics are not subjected to similar scrutiny. The FDA does not have to approve any cosmetics that go on the market unless the product claims to treat or prevent disease or alter the body in any way (in which case the product is classified as a drug) ( 10 ). There are only 11 chemicals that are outright prohibited or restricted for use in cosmetics (bithionol, chlorofluorocarbon propellants, chloroform, halogenated salicylanilides, hexachlorophene, mercury compounds, methylene chloride, cattle materials, sunscreens, vinyl chloride, and zirconium-containing complexes) ( 11 ). Of note, color additives must be approved by the FDA before use in any cosmetics ( 12 ). The exception to this rule is color additives derived from mineral, plant or animal sources, or additives derived from coal tar or petroleum ( 12 ). However, coal tar dyes, especially para-phenylenediamine, have been linked to DNA damage ( 13 – 15 ). This paper describes the chemicals that cosmetics contain and discusses the few studies that address how these chemicals can potentially affect human physiology, especially in relation to menopause.

Women are susceptible to the societal pressures of using cosmetics to beautify themselves ( 1 – 3 ). One theory behind the origins of the ♀ symbol used to denote “woman” is that it represents the hand mirror used by the Roman goddess Venus or the Greek goddess Aphrodite ( 4 ). In their efforts to look beautiful, both men and women apply cosmetics to hide their flaws and accentuate their features. Cosmetics have been a part of human history as far back as the ancient Egyptians ( 5 ). The ancient Egyptians, Romans, and Greeks used various ingredients to soften, improve, exfoliate, and detoxify skin ( 5 ). The ancient Romans and Greeks used walnut extracts as hair dye, antimony (a known toxic heavy metal) as eye shadow, white lead carbonate as a skin lightener, charcoal crocodile excrement as a skin darkener, and cinnabar as rouge ( 5 ). This article will broadly address the question of “Cosmetics use and menopause, is there a connection?” The Oxford English Dictionary defines “cosmetics” as “A preparation intended to beautify the hair, skin, or complexion” ( 6 ). The word comes from the Greek word kosmetikos (“relating to adornment”), which is taken from the Greek word kosmein (“to arrange, adorn”), which itself is taken from the Greek word kosmos (“order, adornment”) ( 6 ). For the purposes of this review, we define cosmetics as any product applied to the skin to enhance and beautify, i.e. products often labeled as “makeup.” In 2014, the revenue of the cosmetic industry in the United States alone was 56.63 billion dollars ( 7 ), compared to the global oral contraceptive pills market which was valued at 5.236 billion that same year ( 8 ). Companies sell a broad spectrum of cosmetic items – each item containing a huge variety of chemicals that all contribute to the color, texture, patina (sheen vs. matte), odor, preservation, suspension, lubrication, thermal stability, and finishing texture of the cosmetic. Given the widespread and frequent personal use of cosmetics containing classes of compounds that are endocrine disrupters, it is of great importance for women and health care providers to understand the potential harm that ingredients in cosmetics can have on women’s reproductive health and reproductive aging. In a survey administered to pregnancy planners and pregnant women regarding risk perception of cosmetic use, out of 128 respondents (68 of whom were pregnant), 39.5% felt that cosmetics outside of pregnancy were “fairly safe” and 37.7% felt that cosmetics were “not really safe” ( 9 ). Despite this fairly high level of concern, most women did not intend to/had not changed their cosmetics use during pregnancy ( 9 ).

Due to the vast number of chemical ingredients in cosmetics, we devised the following methodology to identify chemicals for which to conduct our literature review (see Figure 1 ). Once these chemicals were identified, we generated a word cloud to visualize the frequency of chemicals and undertook a literature review. To summarize our identification methodology, we began by using the Forbes list of top 10 global beauty brands in 2012 ( 16 ). We did not use the top grossing global beauty companies because many companies own several brands. From the top ten beauty brand list, we went to the websites of the top five. These brands are denoted by symbolic letters: X , A , B , C , and D (in order from largest to smallest global brand revenue) ( 16 ). On each site, we looked at lip makeup, face makeup, and eye makeup products of each company and extracted the ingredient list of the top 3–5 best-selling or top rated products in each category, depending on the company (see Table 1 for a complete list of the products assessed in this paper). Brands X and D only carry skin care products and not makeup, so they are not included in the table. We defined face makeup as any product that is applied to the skin for enhancing purposes. Eye makeup encompassed any makeup that is applied near the eye, including eye liners, mascaras, eye shadows, and brow liners. Lip makeup is any lip color or shape enhancing makeup, therefore not including lip balms. Since several ingredients are in the same chemical family but have different names, we simplified the list of ingredients by replacing some with their chemical family name (e.g. paraben in place of methylparaben and ethylparaben) in order to better isolate which chemical families most commonly appear in the ingredients. Table 2 shows the ingredients and their associated chemical family names. Since different companies named some ingredients differently (e.g. “safflower seed oil” vs. “carthamus tinctorius [safflower] seed oil”), we also standardized the names, but we did not include the standardizations in Table 2 . The simplified ingredient list was then inserted into a word cloud generator to visualize which chemical families appeared the most abundant ( Figure 2 ). There are two algorithms used in word cloud generation – one is a direct correlate of the count data and the other is a log function of the count. We used the direct correlation to best represent what cosmetic users may be most concerned about. From the word cloud, it was immediately apparent that coal tar dyes, siloxanes, and parabens were the most frequent chemical exposures from cosmetics application. The iron oxides and titanium dioxide color dyes did appear with much frequency, but as they are inorganic compounds that have little dermal penetration we did not include them in our literature search ( 17 ). One thousand three hundred twenty-two ingredients were compiled for the word cloud. Of the 1322 ingredients, we consolidated chemicals into 9 chemical families. Of the three largest chemical families, 145 ingredients were classified into the family of coal tar dyes, 95 into siloxanes, and 39 into parabens. Of the remaining 1043 individual chemicals we identified that could not be consolidated into chemical families, we cross-referenced our list of chemicals with websites detailing chemicals of concern. From that, we identified an additional 12 chemicals to assess for ovarian toxicity and age at menopause. In order to determine if there were any remaining chemicals that were not identified through the methods above in our definition of makeup, we turned to the David Suzuki Foundation (a Canadian environmental non-profit organization; www.davidsuzuki.org ) and the cosmetics branch of the FDA ( www.fda.gov/Cosmetics ). In 2010, the David Suzuki Foundation investigated a list of “dirty dozen” cosmetics ingredients that contained butylated hydroxyanisole/butylated hydroxytoluene, coal tar dyes, diethanolamine compounds, dibutyl phthalate, formaldehyde-releasing preservatives, parabens, parfum, polyethylene glycol compounds, petrolatum, siloxanes, sodium laureth sulfate, and triclosan ( 18 ). The FDA website highlights certain additional cosmetic ingredients, which are alpha hydroxy acids, beta hydroxy acids, phthalates, and talc ( 19 ). However, beta hydroxy acids were an infrequent ingredient and did not appear in the word cloud. Based on our inclusion criteria for the word cloud, some other ingredients did not appear in the word cloud: diethanolamine (only triethanolamine appeared in the ingredients; the ethanolamine compounds are mostly used in moisturizers or shampoos/soaps which were not included in this review), butylated hydroxyanisole (mostly used in lipsticks and eye shadow but did not appear as an ingredient in any of the products we assessed), sodium laureth sulfate (infrequent in cosmetics – it is mostly used in shampoos/soaps which were not included in this paper), and triclosan (mostly used in personal hygiene and deodorant products which were not included in this paper). After identification of the final list, Pubmed was used to conduct a literature review with the chemical of interest and the following other search terms: “menopause,” “reproductive senescence,” and “ovarian failure.” To assess the toxicity and endocrine disruption of each chemical, we used The US National Library of Medicine’s Toxicology Data Network ( http://toxnet.nlm.nih.gov/ ), Cosmetic Ingredient Review’s safety assessment reports ( http://www.cir-safety.org/ingredients ), The Endocrine Disruption Exchange’s List of Potential Endocrine Disruptors ( http://endocrinedisruption.org/endocrine-disruption/tedx-list-of-potential-endocrine-disruptors/ ), and Pubmed searches for the chemical of interest and the following other search terms: “genotox*,” “carcino*,” “endocrine,” “estrogen,” and “androgen.” Table 3 (“Summary of chemicals in chemical families and potential harmful mechanisms”) is a compilation of all of these chemicals, along with their primary use and their known toxicity to the human body (some ingredients have not been shown to be toxic to humans, but for the sake of completion we have listed all ingredients as well as all known information about relevant toxicity). We will focus on those chemicals that have relevant toxicities to humans.

Some studies have examined the relationship between use of personal care products and serum or urinary concentrations of the ingredients. Since very few studies specifically examined the association between cosmetics use and ingredients absorption, we will also include studies that examine topical personal care product use in order to provide a better picture of how these ingredients can be absorbed through the skin. This section will discuss the studies that have been published on siloxanes, diethanolamine, phthalates, and parabens. One study looked at use of personal care products in a cohort of 94 postmenopausal women in Norway and serum levels of estrogenic ( 20 ) and anti-androgenic ( 21 ) cyclic volatile methylsiloxanes (cVMS) ( 22 ). Sources of exposure to cVMS are generally through cVMS present in personal care products, inhalation due to the volatility of the compounds, or cVMS contained in breast implants ( 22 ). None of the women in the cohort had breast implants and 90.5% of the women creamed more than 10% of their skin area each day ( 22 ). cVMS also has a low blood:air partition coefficient and therefore inhalation exposure would have only had a minor contribution to serum levels of cVMS unless there was risk of workplace air exposure ( 22 ). Unfortunately, since different products carry varying levels of cVMS, no direct correlation was drawn between serum levels and specific personal care product use ( 22 ). Of the 94 women examined, 85% of the women exceeded the limit of quantification for octamethylcyclotetrasiloxane, 18% for decamethylcyclopentasiloxane, and 5% for dodecamethylcyclohexasiloxane ( 22 ). A study provided three premenopausal women with lotion containing 1.8 mg diethanolamine/gram lotion and instructed them to apply the lotion every day for a month ( 23 ). Diethanolamine is a compound that is anti-estrogenic, anti-androgenic, and possibly carcinogenic ( 24 – 26 ). Blood samples of the subjects revealed detectable plasma concentrations of diethanolamine after lotion application ( 23 ). An extensive study of personal care product use among pregnant women revealed significant associations between product use and urinary concentrations of phthalate metabolites and anti-androgenic ( 27 ) and estrogenic ( 28 ) parabens. Phthalates are a class of chemicals that have different hormonal activities depending on the congener’s compound structure ( 29 ). Braun et al. found that cosmetics users had 53% more urinary monoethyl phthalate (MEP) (95% CI: 9–113), 89% more butylparaben (BP) (95% CI: 21–198), 66% more methylparaben (MP) (95% CI: 17–137), and 105% more propylparaben (PP) (95% CI: 31–220) than non-cosmetics users ( 30 ). A study with 332 postmenopausal women in Norway examined the association between serum paraben levels and personal care product use ( 31 ). Median serum levels of MPs were correlated with increasing percentage of skin area creamed per day ( 31 ). The observed trend was also significant for the total combination of MPs, PPs, and ethylparabens (p < 0.001), but increases in ethylparabens and PPs were observed in participants creaming more than 100% of their skin surface per day (i.e. multiple applications) ( 31 ). A study of 337 women at follow-up visits 3–36 months after their pregnancies found that use of basic makeup (including eye makeup, foundation, and lipstick) was associated with increased urinary concentrations of MEP (β = 0.054, p = 0.029) and monomethyl phthalate (β = 0.040, p = 0.023) ( 32 ).

We present here the research on exposure to cosmetics ingredients/cosmetic use and age at menopause. Unfortunately, most of the chemicals that we mention in this review have not been evaluated in relation to age at menopause, and of the chemicals that have been studied, there is very little published data. Due to the limited amount of studies, we expanded our literature search to include studies examining topical personal care product use. The studies noted below were found with the same search terminology reported in the methods section and represent the complete set of articles in that search. One study examined phthalate levels in the personal care products of 195 women aged 45–54, urinary phthalate levels, and self-reported hot flashes history experienced ( 73 ). Phthalate concentrations in personal care products were estimated by summing the metabolite molar concentrations of monobutyl phthalate and MEP ( 73 ). The sum of personal care product urine phthalate levels was associated with: 1) ever experiencing hot flashes (odds ratio = 1.45, CI = 1.07–1.96), 2) experiencing moderate/severe hot flashes (odds ratio = 1.31, CI = 0.95–1.82), 3) experiencing hot flashes in the past 30 days (odds ratio = 1.43, CI = 1.04–1.96), and 4) experiencing daily hot flashes (odds ratio = 1.47, CI = 1.06–2.05) ( 73 ). Previous research has linked hot flash severity in the perimenopausal transition to earlier age at menopause ( 74 ). Another study of urinary paraben levels in 192 women presenting for fertility care in the Boston, Massachusetts area found MP and PP in more than 99% of urine samples and BP in more than 75% ( 75 ). The study found a trend towards decreased antral follicle counts and increased day three FSH levels with increasing PP tertiles ( 75 ). However the study did not achieve statistical significance ( 75 ). No other consistent associations were found with the other parabens or with ovarian volume ( 75 ). Exposure to polycyclic aromatic hydrocarbons (PAHs), which are a known contaminant of inadequately refined petrolatum, has been shown to increase proapoptotic gene expression in the ovary as well as induce oocyte depletion via p53 ( 76 ). There is some evidence that the ovarian toxicity of PAHs work through the aromatic hydrocarbon receptor which induces Bax gene expression that induces apoptosis ( 77 ). Benzo(a)pyrene (BaP), a PAH contaminant sometimes detected in cosmetics ( 78 ), also induced significant DNA damage in oocytes and cumulus cells ( 79 ), as well as decreased CL counts and ovarian volume in mice ( 80 ). A study of BaP and two other PAHs in mice and rats found that exposure to all three PAHs resulted in drastically lower numbers of primordial and primary follicles ( 81 ). Surprisingly, the study also found that chronic low dose exposure to the three PAHs was more toxic compared to acute high dose exposure ( 81 ). PAHs have also been found to be anti-androgenic, estrogenic, carcinogenic, and potentially genotoxic ( 79 , 82 – 84 ). A study examined how gestational exposure of Wistar rats to a mixture of proven human EDCs (italicized compounds or their metabolites have been found in cosmetics; di-n-butyl phthalate , di-(2-ethylhexyl) phthalate , vinclozolin, prochloraz, procymidone, linuron, epoxiconazole, octyl methoxycinnamate , dichlorodiphenyl-dichloroethylene (p,p′-DDE), 4-methyl-benzylidene camphor , BPA, butyl paraben , and paracetamol ) affected ovarian follicle reserves and reproductive aging ( 85 ). The study examined the effects of subjecting the rats to a mix of all the EDCs (Totalmix), to a mix of the anti-androgenic EDCs (4-methyl-benzylidene camphor, octyl methoxycinnamate, BPA, and BP; AAmix), and to a mix of the estrogenic EDCs (di-n-butyl phthalate, di-(2-ethylhexyl) phthalate, vinclozolin, prochloraz, procymidone, linuron, epoxiconazole, p,p′-DDE; Emix) ( 85 ). Rats exposed to AAmix had significantly reduced numbers of primordial follicles (78.2% of control values; p = 0.02) ( 85 ). Rats exposed to AAmix also had significantly lower percentages of primordial follicles (p = 0.005) and significantly higher percentages of secondary (p = 0.05) and tertiary follicles (p = 0.04) out of the total number of follicles ( 85 ). The number of total recruited follicles was also significantly higher in rats exposed to AAmix (p = 0.01) ( 85 ). At 12 months of age, rats exposed to Totalmix had a significant increase in irregular estrous cycles (p = 0.041), as well as significantly lower ovarian weight in rats exposed to Totalmix and AAmix ( 85 ). In addition, at 13 months of age, rats exposed to AAmix had a significant increase in incidence of complete absence of CL (p = 0.033), as well as a significant reduction in mean number of CL to 55% of controls in the Totalmix group (p = 0.04) and approached significance for the AAmix group (p = 0.056) ( 85 ).

Naturally occurring menopause is defined as amenorrhea for 12 full months ( 33 ). The current established average age at menopause in the US is 51.4 years old ( 33 ). In utero , the ovarian reserve of primordial follicles is built from ovarian germ cells. At four months postconceptional age, the ovary peaks at 6–7 million oocytes ( 34 ). Due to apoptosis, at birth only 1–2 million primordial follicles remain ( 34 ). Primordial follicle numbers continue to fall exponentially, albeit less rapidly, until menarchal onset when the ovary has 300,000 to 400,000 follicles remaining ( 34 ). At menopause, less than 1000 follicles remain ( 34 ). Natural menopause is a physiologic manifestation of a depleted follicular pool. Elevated follicle-stimulating hormone (FSH) and luteinizing hormone levels, and low anti-Mullerian hormone (AMH; a hormone used to estimate ovarian reserve ( 35 )) are hallmark biochemical signs of menopause ( 33 , 36 ). Menopause is preceded by a transition period (perimenopause) that can last several years starting from when a woman enters her fourth decade of life ( 33 ). Perimenopause is associated with changes in a woman’s reproductive hormones and is associated with waist thickening, vaginal dryness, and irregular periods ( 33 ). Other symptoms include hot flashes, irritability, and sleep disturbance ( 33 ). Age at menopause can be affected by various aspects such as race, family history, smoking -history, genetic predisposition, Fragile X syndrome, and autoimmune disorders ( 8 , 37 – 44 ). There is some evidence that chemical exposure can affect age at menopause, but little research has been done directly linking exposure to chemicals in cosmetics with age at menopause. We will therefore discuss the potential mechanisms by which cosmetics ingredients may have an effect on age at menopause: The onset of reproductive senescence is due to depletion of primordial follicles. It is believed that prenatal follicle assembly and rate of recruitment of primordial follicles are factors in the rate of depletion ( 45 – 47 ). Lower ovarian follicle reserves, increased number of follicles recruited for maturation per cycle, and absence of corpora lutea (CL) development indicate onset of reproductive senescence. Since the hypothalamic-pituitary-ovarian axis is crucial in the maintenance of reproductive organs as well as the menstrual cycle in females, any chemicals that disrupt the axis may also cause early reproductive senescence ( 48 ). Endocrine disrupting chemicals (EDCs) are a class of exogenous chemicals that disrupts some aspect of a hormone’s mechanism ( 49 ). EDCs that are either estrogenic or anti-androgenic seem to play the largest role in this process of premature reproductive senescence ( 45 ) (See Supplement 1 for a summary of known estrogenic or anti-androgenic EDCs and reproductive senescence). Failure of DNA repair mechanisms may also lead to failure in ovarian follicle reserves and earlier age at menopause. Ovarian follicle aging has been linked to lower expression of DNA double strand break repair genes BRCA1, MRE11, Rad51, and ATM ( 50 ). Women with BRCA1 mutations have shown significantly lower concentrations of AMH and earlier age at menopause ( 50 – 52 ). A meta-analysis of 22 genome-wide association studies in 38,968 women found additional associations between age at menopause and genes implicated in DNA repair (EXO1, HELQ, UIMC1, FAM175A, FANCI, TLK1, POLG and PRIM1) ( 53 ). Another study also found an association between a variant in the mismatch repair gene MSH6 with age at menopause ( 54 ). Shortened leukocyte telomere length may also be associated with age at menopause, although the mechanisms are still unclear. It should be noted that DNA repair mechanisms and telomere length are mechanistically related ( 55 ), and telomere length can be affected by many elements, such as telomerase activity, oxidative stress, antioxidant activity, inflammation, the hypothalamus-pituitary-adrenal axis (glucocorticoid levels), and mitochondria regulation ( 56 ). One study found that among non-Hispanic white women, one standard deviation in longer leukocyte telomere length was associated with a 0.43 year later age at menopause ( 57 ). Among Mexican-American women, one standard deviation in shorter leukocyte telomere length was associated with a 1.56 year earlier age at menopause; and among non-Hispanic black women, no association was found ( 57 ). Another study with a cohort of 486 white women found that for every 1 kilobase increase in leukocyte telomere length, average age at natural menopause increased by 10.2 months (95% CI: 1.3–19.0), with no association seen in women with surgical menopause ( 58 ). Oxidative stress can also affect the ovarian reserve in other ways. Reactive oxygen species (ROS) can induce primordial follicle loss and apoptosis in the ovary ( 59 ). ROS can also deplete glutathione, which then leads to atresia of antral follicles and apoptosis of granulosa cells ( 59 ). Oxidative stress also increases permeability in mitochondria ( 60 ), which exposes mitochondrial DNA to damaging elements ( 61 ). Damaged mitochondrial DNA has been linked to reproductive aging ( 55 , 62 ). Mice lacking the glutamate cysteine ligase modifier subunit, the rate-limiting enzyme in production of the most abundant intracellular antioxidant, glutathione, showed increased oxidative stress, apoptosis in follicles, and accelerated age-related decline in primordial follicles compared to wild type mice ( 63 ). Direct ovarian toxicity also accelerates reproductive senescence. While there is extensive evidence of environmental toxins and their effect on the ovaries ( 64 ), the effect of iatrogenic chemicals (chemotherapy and radiotherapy) is very well documented. Women who have undergone anticancer treatment, especially in adolescence, show accelerated depletion in ovarian reserve, decreased AMH levels, decreased ovarian volume, and advanced age at menopause ( 65 – 67 ). Some of the chemicals in cosmetics have known carcinogenic effects. Studies of age at menopause and cancer in humans examine cohorts of patients that have undergone gonadotoxic chemotherapy. There is one study that demonstrated reproductive senescence due to the presence of cancer alone in rats. In that study, only 49% of female Wistar rats with administered neoplasms had regular estrous cycles by 30 weeks of age, and at 112 weeks of age, only 24% of female rats were still cycling but the majority of the animals had major cycle abnormalities ( 68 ). The contribution of these mechanisms to early reproductive aging means that chemicals that are either anti-androgenic or estrogenic EDCs, that damage DNA repair mechanisms, that affect telomere length, that increase oxidative stress or reduce antioxidant activity, that increase ovarian toxicity, or that are carcinogenic have the potential to contribute to premature age at menopause. While this review examines the limited data on cosmetics use and age at menopause, it should be noted that there are a handful of articles that have published data indicating that there is a secular trend towards later age at menopause ( 69 – 72 ). This trend towards later age at menopause is associated with increasing physical activity and education level, as well as better childhood nutrition and health in the general population, but these studies also do not examine this secular trend stratified by cosmetics use ( 69 , 72 ).

While we have examined the ingredients in cosmetics, specifically limited to face makeup, eye makeup, and lip makeup, these products are but a subset of personal care products. This paper does not extensively cover the hazards of sunscreen, hair dye, makeup remover, lotion, skin lightener, anti-aging/anti-wrinkle creams, and chemicals associated with nail embellishment, to name a few. Another subset of products we do not address that is gaining particular traction is the “fountain of youth” agents, which often contain hydroquinone, retinol, sunscreen, antioxidants, and alpha hydroxyl acids. As such, evaluation of other personal care products and use of mixtures in relation to ovarian health is very important. Most of the reviewed chemicals have demonstrated dermal absorption. Certain chemicals have limited dermal absorption, such as formaldehyde ( 23 , 114 ). However, studies evaluating formaldehyde are limited by studying the exposure for as short as one month. Other limitations include no evaluation of body repositories of these chemicals such as visceral and subcutaneous fat compartments. Women apply these products to their face daily, sometimes even more than once daily, which can lead to several grams of exposure per product per day, over the course of a woman’s lifespan ( 115 – 118 ). As seen above, chronic low dose exposure to PAHs is more toxic to the ovaries than acute high dose exposure ( 81 ), and future studies need to explore if this is true for other chemicals. In addition, while studies have examined the impact of these chemicals individually, little work has been done on the potential interactions that these chemicals can have on each other and how these mixtures can then affect human physiology. Women generally use more than one product ( 115 – 117 ), and the ingredients that are contained in different products can interact with each other. In addition, while one product may contain a level of these ingredients that have been deemed safe, continuous use and use of several products containing these ingredients can easily expose a woman to levels that are beyond what has been defined as safe. Another consideration that must be taken into account is the packaging for these cosmetics. Packaging and plastic materials have been shown to have toxins that can leach into products and adversely affect human health ( 119 – 121 ). It would be reasonable to extrapolate that the plastic packaging of many cosmeticsmay leach additional toxins into the products, which are then applied to the skin. These are considerations that need to be taken seriously as we continue to evaluate the long term adverse impact that cosmetic use has on a woman. One last consideration is incomplete reporting of ingredients. While cosmetics companies are for the most part required to report all of the ingredients they use, there are still some exceptions. As mentioned before, all chemicals that are used as fragrance do not need to be reported and can simply be listed as “parfum” or “fragrance.” The same applies for cosmetics that have flavors, most often found in lipstick. Some of the cosmetics we looked at simply had “Flavor” written, with no indication of what ingredients this broad term referred to. In addition, a couple of the Brand A products that we examined contained “Shimmer Shades,” which was followed up by a note that said “Shimmer Shades may contain:” with potential ingredients listed. Consumers have no way of knowing the exact composition of ingredients of the Shimmer Shades due to the vague wording. Lastly, Brand B products have several ingredients in their products that are listed as FIL D*****/2, with * representing a number. Our search for what these ingredients may be reveals nothing. Currently, limited evidence has demonstrated that few patients are counseled regarding safe cosmetics use. A study found that only 23.4% of 128 women surveyed had received advice about personal care product use and only 18.9% of them had received advice about make-up product use ( 9 ). Health care providers should make attempts to make their patients aware of the developing literature around the chemicals used in cosmetics and help educate their patient population to make more informed personal care product choices. Many groups are now disseminating information in hopes of making consumers aware of the toxins in these cosmetics, such as the Environmental Working Group (EWG), Campaign for Safe Cosmetics, and the David Suzuki Foundation. In addition, the EWG has created a Skin Deep® Cosmetics Database that allows consumers to search for personal care products and determine the hazard level of each product. Limitations of this review are in large part due to our inability to assess each and every ingredient and their mixture effects, and the lack of available data regarding the vast number of chemicals that are in cosmetics with age at menopause. A more comprehensive investigation of cosmetics and their effects on menopause, reproductive health, and the health of other organ systems need to further investigated.

While we have examined the ingredients in cosmetics, specifically limited to face makeup, eye makeup, and lip makeup, these products are but a subset of personal care products. This paper does not extensively cover the hazards of sunscreen, hair dye, makeup remover, lotion, skin lightener, anti-aging/anti-wrinkle creams, and chemicals associated with nail embellishment, to name a few. Another subset of products we do not address that is gaining particular traction is the “fountain of youth” agents, which often contain hydroquinone, retinol, sunscreen, antioxidants, and alpha hydroxyl acids. Here we will give a brief overview of some of the above products in hopes that future research will also take these areas into consideration. The Braun et al. study mentioned previously also found several associations between use of different personal care products and increased urinary phthalate and paraben levels. Users of shampoo, conditioner, and nail polish had significantly higher urinary levels of phthalates ( 30 ). Users of hair gel had higher urinary levels of parabens ( 30 ). Lastly, users of lotion and cologne/perfume had higher urinary levels of both pthalates and parabens ( 30 ). Perfume, lotion, deodorant, hair spray, crème rinse, other hair products, and bar soap were all positively and significanty associated with urinary MEP levels ( 32 ). There was also a significant association between total number of personal care products used in the 24 hours before the urine spot test and median creatinine-adjusted log 10 -transformed MEP concentrations. In a study of 108 women in Mexico, increased use of body lotion, deodorant, perfume, and anti-aging facial cream use was associated with increased median urinary phthalate concentrations ( 86 ). Another study of 186 minority pregnant women in New York City revealed that women who used perfume had 2.3 times higher urinary MEP concentrations (95% CI: 1.6–3.3) ( 87 ). Suntan and sunscreen products are also particularly harmful. Polyaromatic hydrocarbons ( 78 ) as well as benzophenones are used as ingredients in these products ( 88 ). Benzophenone passes through the skin and enters the blood stream ( 89 ), along with other sunscreen chemicals such as octyl-methoxycinnamate and 3-(4-methylbenzylidene) camphor ( 90 ). Benzophenone has been shown to be endocrine disrupting in animals ( 91 , 92 ), positively associated with women with uterine leiomyomas and endometriosis ( 93 , 94 ), and increased oxidative stress and genotoxicity ( 95 , 96 ). Benzophenones are not just limited to sunscreen products. They are found in makeup, lotion, and hair products in order to protect consumers from the sun ( 97 ). Skin lighteners are commonly used in minority races/ethnicities ( 98 – 101 ), and as a result this population of women are at particular risk of exposure to toxic chemicals that are in skin lightening products such as mercury, hydroquinone, and steroids ( 102 , 103 ). Vaginal douching, which a study on the National Health and Nutrition Examination Survey (NHANES) 2001–2004 found to be more common in black women than white or Mexican American women, is significantly associated with higher urinary concentrations of MEP ( 104 ). Hair products, which contain a significant amount of phenylenediamine, a coal tar dye, also contain formaldehyde releasing chemicals ( 105 , 106 ). The Madrid Statement is a document released in 2015 by over 200 scientists worldwide calling to limit the use of fluorinated chemicals in everyday products ( 19 ), and it has previously been shown that fluorinated chemicals are pervasive in personal care products ( 107 ). A database compiled by the Green Science Policy Institute, a group of scientists based in California dedicated to promoting responsible use of chemicals, reveals that there are several perfluorinated and polyfluorinated chemicals that are used in a variety of cosmetics ( 108 ). While there are too many to examine individually, one chemical, polytetrafluoroethylene (PTFE), popularly known as Teflon®, is of particular concern. It is used in a variety of products, including face makeup, eye makeup, men’s shaving cream, and sunscreen, as a bulking agent and slip modifier ( 17 , 108 ). PTFE is traditionally produced by using perfluorooctanoic acid (PFOA) as a surfactant during emulsion polymerization of PTFE, leading to concerns that PTFE is contaminated with PFOA ( 109 ). The adverse health effects of PFOA have been numerously studied. In relation to menopause, a study of 25,957 women aged 18–65 found that the odds of having experienced menopause were significantly higher in the highest quintile of exposure to estrogenic PFOA and perfluorooctane sulfonate (PFOS) compared to the lower quintile of exposure in women aged 42–51 (PFOS odds = 1.4, confidence interval (CI) = 1.1–1.8; PFOA odds =1.4, CI = 1.1–1.8) and in women aged 52 and above (PFOS odds = 2.1, CI=1.6–2.8; PFOA odds = 1.7, CI = 1.3–2.3) ( 110 ). In the NHANES cohort, women in the second and third tertile serum levels of estrogenic polyfluoroalkyl chemicals had significantly higher incidence of menopause than women in the first tertile (hazard ratio = 1.23, 95% CI = 1.04–1.44 for tertile 2; hazard ratio = 1.16, 95% CI = 0.91–1.48 for tertile 3) ( 111 ). For women in the third tertile of serum levels of perfluorooctanoate, perfluorononanoate, and perfluorohexane sulfonate compared to women in the first tertile, the adjusted hazard ratios were 1.36 (95% CI: 1.05, 1.75), 1.47 (95% CI: 1.14, 1.90), and 1.70 (95% CI: 1.36, 2.12), respectively ( 111 ). For women in the second tertile of serum levels of perfluorooctanoate, perfluorononanoate, and perfluorohexane sulfonate compared to women in the first tertile, the adjusted hazard ratios were 1.22 (95% CI: 0.92, 1.62), 1.43 (95% CI: 1.07, 1.91), and 1.42 (95% CI: 1.08, 1.87), respectively ( 111 ). However, a study evaluating serum levels of PFOA from environmental exposure and age at menopause in women aged 40 years and above in the Mid-Ohio Valley community demonstrated no statistically significant associations ( 112 ). Due to the known health hazards of PFOA, PTFE is now produced by many manufacturers with chemicals other than PFOA ( 113 ); however, cosmetic companies do not specify how they manufacture the PTFE that they use.