In pregnancy, the prevalence of vaginal candidiasis is increased compared to those not pregnant [1]. Treatment is antifungal compounds administered locally or systemically depending on the severity of symptoms [2]. In Denmark, antifungal vaginal tablets and antifungal vaginal cream with either the active ingredients miconazole or clotrimazole (both imidiazoles) are recommended as first treatment for vaginal candidiasis. Oral fluconazole (triazole) is only recommended for severe cases when the woman is pregnant [3–9]. Imidazole and triazole compounds are anti-mycotic through inhibition of a specific cytochrome P450 enzyme (CYP51) involved in fungal cell wall synthesis, [10, 11] but they are also known inhibitors of a range of other CYP enzymes including those involved in androgen biosynthesis [11, 12]. Accordingly, anti-androgenic properties have been demonstrated for several conazole fungicides [12, 13].

In pregnancy, hormonal changes increase the risk of vaginal candidiasis, [1, 14, 15] and as the fetus is particularly vulnerable to endocrine disruption, use of anti-fungal medication might affect development of the genitalia [16–18]. A recent Danish study among 1,405,663 pregnant women found an association between use of oral fluconazole during pregnancy and miscarriage [19]. Case reports have linked high-dose, long-term treatment with oral fluconazole during pregnancy to a pattern of skeletal and craniofacial birth defects seen in the offspring. This has raised concern regarding the safety of oral fluconazole use during pregnancy [20–23]. Other studies have examined the association between maternal use of antifungal medicine and malformations in the offspring, but no consistent associations have been found [14, 24, 25]. Subtle signs of anti-androgen action have not been studied in children whose mothers used anti-fungal medication during pregnancy.

Anogenital distance (AGD: distance from anus to genitals) is routinely used in animal toxicology studies and is a sensitive test of exposure of the male fetus to anti-androgenic agents [26–28]. In rodents, AGD has been shown to reflect the amount of androgen to which a male fetus is exposed in early development: in utero exposure to lower levels of androgens results in a shorter AGD, which is one marker of de-masculinisation [26, 28]. A vulnerable fetal masculinisation programming window (MPW) has been identified in rodent models in which androgens must act to masculinise the components of the reproductive tract and to allow the later complete development. This MPW has been identified in human to occur between gestation weeks 8–14 [29].

To our knowledge, no human studies have examined the association between maternal exposure to antifungal medicine and AGD in the male offspring. We therefore prospectively investigated the association between systemic or local use of antifungal medicine during pregnancy and the subsequent AGD and penile width in the male offspring at 3 months of age in the Odense Child Cohort (OCC) study.

Newly pregnant women at gestational age 8–16 weeks residing in Odense Municipality, Denmark, between 2010 and 2012 were recruited: at a voluntary information meeting about ultrasound examinations; at their first antenatal midwife visit; or at their ultrasound examination at Odense University Hospital. All of the pregnant women in the study completed two questionnaires, one during the first and one during the third trimester [30]. Serum samples were collected twice and urine samples once during pregnancy and stored in freezers at the Odense Patient data Explorative Network (OPEN) [30]. The study was carried out in accordance with the Helsinki Declaration II and was approved by the Regional Scientific Ethical Committees for Southern Denmark (S-20090130).

Of the eligible population of 6707 pregnant women, 4017 women were informed about the study and 2421 live born singletons with birth register data participated in the study and are currently being followed up. Participants were better educated (high school +1 year or more) and more often of Danish origin than non-participants [30].

In the questionnaire administered during the first trimester, participants were asked whether they had taken any topical or systemic medications (both prescription and over the counter) and if so to provide the name, dose and gestational week/s of use of each named medication. In the second questionnaire (administered during the third trimester) the women were asked to state whether they had experienced specific diseases or symptoms (including vaginal candidiasis) during the last 3 months of pregnancy. Furthermore, the women were asked to state whether they had ever used any type of medication during their entire pregnancy. They were asked to note names of medication, doses, total number of days of use and in which gestational week(s) it was used.

The answers from both questionnaires were categorised into: overall use of any type of antifungal medicine during pregnancy; use of fluconazole oral tablets; use of antifungal vaginal tablets (including antifungal vaginal cream if used in combination with vaginal tablets); use of antifungal vaginal cream only; and unspecific use of antifungal medicine (women who had not specifically noted which type of antifungal medicine they had used). In addition, we categorised use of antifungal oral and vaginal tablets into exposure in and outside the MPW (gestation weeks 8–14).

Three months after the expected date of birth, regardless of actual gestational age at birth, the children were invited for a clinical examination, which included measurements of length, weight and AGD. The two measurements of AGD, and the measurement of penile width were made using a Vernier caliper, which is the recommended instrument for AGD measurement [31]. The shorter AGD measurement was measured from the centre of anus to the posterior base of scrotum (AGDas) and the longer from the centre of anus to the cephalad insertion of the penis (AGDap). Penile width was measured at the base of the penis. All of these measurements were made three times and the arithmetic mean was calculated. Expert-trained technicians performed the examinations [30]. 13 boys had AGD measured by two examiners. The coefficient of variation (CV) was 3% for all the triplicate AGD measurements. Inter-examiner CV was respectively 4%, 3% and 4% for AGDas, AGDap and penile width indicating a high degree of consistency.

A total of 2421 live born singletons with birth registry data were included in the original OCC dataset. After the initial exclusions, as shown in Fig. 1, the final dataset consisted of 812 mother-son pairs with measurements at 3 months of whom there were 795 AGDap measurements, 810 AGDas measurements and 802 measurements of penile width.

After adjustment for weight-adjusted z-score and post-conceptional age, boys prenatally exposed to oral fluconazole had a significantly shorter AGDas (−6.4 mm; 95% CI: -11.9; −0.9) at their 3-month examination compared to unexposed boys (Table 2, Figs. 2 and 3). In addition, maternal use of antifungal vaginal tablets containing miconazole or clotrimazole was marginally associated with a shorter AGDas (−1.9 mm; 95% CI: -4.3; 0.5). Boys exposed to vaginal tablets containing either miconazole or clotrimazole in the MPW had shorter AGDas than those exposed outside the MPW (Table 2). When dividing use of vaginal tablets into only use of vaginal tablets and vaginal tablet use in combination with cream, boys prenatally exposed to only vaginal tablets had shorter AGDas than those prenatally exposed to a combination (Table 2).

An association was found between unspecific antifungal medicine use and penile width (−0.5 mm; 95%CI: -0.9; −0.1) but no association was seen for penile width and other exposure groups (see Additional file 1). No association was found between prenatal exposure to any antifungal medicine and AGDap.

Four pregnant women taking fluconazole orally gave birth to boys with a significantly shorter AGDas (−6.4 mm) than unexposed boys. In addition, maternal use of antifungal vaginal tablets containing miconazole or clotrimazole was associated with a shorter AGDas, although not significantly, whereas use of the same compounds in the form of vaginal creams did not affect AGDas. AGDap and penile width were not affected apart from a shorter penile width among boys to mothers with unspecific use, which was probably due to chance. Interestingly, use of antifungal medicine in the window of genital development between 8 and 14 weeks of gestation was associated with a larger reduction in AGDas than exposure outside of this window. This is in accordance with the sensitivity of this window to the effects of prenatal anti-androgen exposure in rodent studies [27]. We acknowledge that our rather preliminary findings are based on four cases, but they are biologically plausible and of public health importance as up to 20% of pregnant women experience vaginal candidiasis. In addition, a 6.4 mm shorter AGD corresponds to a 20% decrease and may be of clinical importance as it has been associated with abnormal adult male reproduction [33, 34].

This is to our knowledge the first human study to examine the association between maternal exposure to antifungal medicine and AGD in the offspring. Our findings are biologically plausible, as conazole antifungals have known anti-androgenic properties, [12] and exposure to chemicals with anti-androgenic action (e.g. phthalates) have been associated with shorter AGD in several human studies [32, 35–38]. Rodent studies have shown that an impairment of androgen action within the MPW can result abnormal development and function of reproductive organs [27]. Hence, exposure to anti-androgenic chemicals, including conazole antifungals, during the MPW have resulted in hypospadias, cryptorchidism and shortened AGD in rodent studies [26, 27].

Two large registry-based Danish studies found no association between oral intake of fluconazole during the first trimester and 15 different birth defects in the offspring [24, 25], however, they did not measure AGD. Similar findings have been reported in other smaller studies [14, 15, 24], which also studied hypospadias, but none of the studies measured AGD, which may be a more sensitive marker of prenatal exposure to anti-androgens [32].

The association between maternal use of antifungal vaginal tablets and AGD was weaker than among boys exposed to oral treatment. This may be due to less potent anti-androgenic action of miconazole and clotrimazole and/or lower exposure from vaginal tablets. The same reasoning may be used for the lack of effect of vaginal cream exposure on AGDas, as exposure levels from creams, is expected to be even lower. Miconazole in vaginal tablets have been demonstrated to be absorbed systemically [39]. Women using vaginal tablets without combination with antifungal vaginal cream typically used tablets containing higher doses of miconazole or clotrimazole than women who used both tablets and cream. Thus, the fetus may be exposed to a higher dose after use of vaginal tablets alone compared to the combined treatment of vaginal tables and cream which was seen by a shorter ADGas in this group.

Participation rate was 43% [30] and only 11% of the women reported use of antifungal medicine, which is lower than the estimated 20% of all pregnant women suffering vaginal candidiasis [14, 15]. This could be due to the fact that our study population was higher educated and therefore more aware of the importance of avoiding medication use during pregnancy. Information about use of antifungal medicine was self-reported which might have led to misclassification. This is, however, not likely to be associated with AGD as the women were unaware of this measure when answering the questionnaires leading to an underestimation of the association.

AGD measurements has acceptable intra- and inter-examiner reliability, but with large inter-individual variation in measurements especially in AGDap measured from the anus to the top of the cephalad insertion of the penis. AGDap is therefore more dependent on the size of the child, introducing a higher inaccuracy of the measurement. Also, penile width is small and thus the same absolute measurement error is of relatively larger importance. Confounding by indication may explain our findings, as the overgrowth of Candida albicans organism may be associated with a reduction in AGD, rather than the use of antifungal medication. We adjusted for relevant confounders but we cannot exclude the possibility of residual confounding from for example co-exposure to other environmental chemicals, lifestyle or health behavior.

In male rodents, the shortened AGD persists into adulthood confirming that interference of androgenic activity permanently alters the reproductive tissues [29]. In cross-sectional studies among adult men, AGD has been shown to have a significant positive association with sperm count [34]. In fact, AGD is now considered part of the spectrum of male intrauterine testicular disruption called the testicular dysgenesis syndrome (TDS) [18]. Therefore, the observed reduction in AGD may have potential long-term consequences for male reproductive health.

Based on our findings that four pregnant women taking fluconazole orally gave birth to boys with a significantly shorter AGD we hypothesize that maternal use of antifungal medication may cause anti-androgenic effects. Further studies on maternal antifungal use during pregnancy are needed. In the interim, pregnant women should be advised to use antifungal medicine with caution.