Introduction

In 2021, the Polish Society of Endocrinology (PTE, Polskie Towarzystwo Endokrynologiczne) published an updated version of the guidelines on the management of thyroid diseases in pregnancy [1]. Because the topic is extensive, issues related to fertility disorders versus thyroid disease were not included; the Society made the decision to address them in a separate document. A recommendation grading system was based on the system presented in the 2021 PTE guidelines [1, 2].

Grading system classification:

1 — a strong recommendation, associated with the phrase “recommended”

2 — a weak recommendation, associated with the phrase “suggested” or “advised”.

Strength of evidence classification:

Infertility is defined as the inability to conceive within 12 months of having regular unprotected sexual intercourse, and it affects 8–12.5% of couples of reproductive age [3]. It is estimated that 35% of infertility cases are attributed to the female factor, 30% to the male factor, and 20% to both factors simultaneously, while in 15% of cases, infertility is idiopathic [4]. Thyroid diseases are quite common in women of reproductive age: autoimmune thyroid disorder (AITD) affects about 10%, hypothyroidism 2–3%, and hyperthyroidism 1–2% of this population [5]. Similar epidemiological data were reported in two recently published studies:

Other published data indicate that subclinical hypothyroidism may affect up to 15% of women of reproductive age in some populations [8–10].

Thyroid hormones (THs) have an impact on the female reproductive system through several mechanisms:

Additionally, THs influence male gonadal function and thus male reproduction. The presence of THs receptors has been confirmed in Sertoli and Leydig cells. THs inhibit proliferation and stimulate differentiation of Sertoli cells, thus determining the number of Sertoli cells at puberty, final testicular volume, and the number of spermatozoa in the semen. In addition, triiodothyronine (T3) increases androgen receptor activity, decreases oestrogen receptor activity, and inhibits aromatase activity by blocking the conversion of testosterone to 17b-oestradiol [13]. T3 exerts a direct effect on Leydig cells (by regulating steroidogenesis), and an indirect effect by affecting Sertoli cells. Thyroxine (T4) in vitro also exerts a direct effect on sperm motility [14]. THs affect the bioavailability of sex hormones by stimulating SHBG synthesis by hepatocytes, increasing SHBG blood concentrations, and inhibiting the production of androgen-binding protein (ABP) by Sertoli cells (animal models) [15, 16].

Hypothyroidism and fertility

Hypothyroidism may cause ovulatory disorders associated with insufficient direct stimulation of oocyte maturation by THs inside the ovaries, hyperprolactinaemia, impaired pulsatile GnRH secretion, and decreased SHBG concentrations leading to lower total oestradiol and testosterone concentrations, increased free fraction of sex hormones, and decreased androstenedione and oestrone metabolism [17].

In the contemporary literature, there are no data concerning the effect of overt hypothyroidism on female fertility, probably because this disorder is quite rare. Data from infertility clinics indicate that the frequency of overt hypothyroidism among subfertile women is 0.2–0.4%, which is the same as in the general population [7, 18].

It has been reported that menstrual disturbances are more frequent in women with overt hypothyroidism compared to healthy women: 23–68% vs. 8–12%, mainly in the form of oligomenorrhoea and hypermenorrhoea [4, 19].

In a cross-sectional retrospective study, Lincoln et al. assessed the frequency of hypothyroidism in a group of 704 subfertile women without signs and symptoms of hypothyroidism and found elevated TSH concentrations in 16 of them (2.3%) [20]. In this subgroup, 68% of the subjects showed ovulatory disturbances, and 64% of them became pregnant after starting L-thyroxine treatment. In a Finnish retrospective study comprising 335 women with infertility, screening for hypothyroidism was performed, and elevated TSH concentrations in the range 5.7–32 mIU/L were revealed in 12 of the subjects (4%) [21]. Oligomenorrhoea or amenorrhoea were present in 67% of hypothyroid women vs. 34% of all infertile women studied (p < 0.05).

Spontaneous conception in women with overt hypothyroidism is possible, as shown by Abalovich et al. [22], but there is a high risk of miscarriage. The authors retrospectively evaluated the course of 150 pregnancies in 114 women with hypothyroidism and found that 51 of them (34%) had become pregnant while in uncontrolled hypothyroidism, of whom 16 had been diagnosed with overt hypothyroidism. Due to inadequate substitution treatment, 60% of the women with overt hypothyroidism miscarried and 20% delivered preterm. These complications occurred, respectively, in 71.4% and 7.2% of the analysed women with subclinical hypothyroidism.

In recent years, there have been several publications on the impact of subclinical hypothyroidism on female fertility (Tab. 1). Many of these reports were retrospective, included a selected group of women with infertility, and differed in the range of mean age of the subjects (from 28.5 to 35 years), the criteria for diagnosis of subclinical hypothyroidism, and the aTPO status. In some studies, including 2 studies that did not determine whether hypothyroidism was overt or subclinical [20, 21], hypothyroidism was as common among subfertile women as in the general population (2.3–4% of women), and in a study by Poppe et al., its frequency was even lower (0.9%) [23]. In a retrospective Danish study comprising 9528 women previously untreated for thyroid disease, it was found that subclinical hypothyroidism reduced fertility. A negative correlation was observed between TSH concentrations and elevated aTPO concentrations and the number of children born, as well as older age at first child, compared to euthyroid women, and a higher risk of not having children [24]. However, a major limitation of the study is that the consequences of subclinical hypothyroidism detected in postmenopausal women were related to the reproductive period. In the only prospective study on fertile women, rather than a selected group of women with fertility problems, during a 6-month observation, the pregnancy rates in the group with TSH ≥ 2.5 mIU/L and TSH < 2.5 mIU/L were the same: 68% and 64%, respectively [25]. The study, which was a part of the Effects of Aspirin in Gestation and Reproduction (EAGeR) project, is important because of its prospective and multicentre nature and the large size of the study group, comprising 1193 women. The results suggest that a lower TSH cut-off value of 2.5 mIU/L in women trying to become pregnant is not justified.

However, other studies have indicated a higher prevalence of subclinical hypothyroidism in women with infertility than in the general population: 6.7–13.9%, and its more frequent co-occurrence with specific causes of infertility: in ovulation disturbances, subclinical hypothyroidism was observed in 15–46% of the subjects; in idiopathic infertility, 20%; tubal factor infertility, 18%; and in premature ovarian insufficiency, 40% [26, 27]. Poppe et al. showed that the association of subclinical hypothyroidism with a specific cause of infertility concerned only the women with coexisting elevated aTPO concentrations [23]. The same authors found that the mean TSH concentration in the group of subfertile women was slightly but significantly higher than in the control group (1.3 mIU/L vs. 1.1 mIU/L; p = 0.006), especially in women with ovulation disturbances (1.5 mIU/L vs. 1.1 mIU/L (p < 0.05). Similarly, in the study by Orouji Jokar et al., euthyroid women with idiopathic infertility had higher TSH concentrations than the control group [28]. Two studies on subfertile women assessed the relationship between the presence of subclinical hypothyroidism and diminished ovarian reserve [29, 30], but this was confirmed only in a group of women with genetically determined low ovarian reserve, mainly with Turner syndrome [29]. A second study noted the association between low free triiodothyronine (FT3) concentrations and low antral follicle count [30].

Case-control and cross-sectional studies do not clearly establish a cause-and-effect relationship between subclinical hypothyroidism and fertility problems. Only two prospective studies evaluated the effect of L-thyroxine treatment on reproductive function in women with impaired fertility and coexisting subclinical hypothyroidism. Yoshioka et al. observed 69 subfertile women with subclinical hypothyroidism and TSH concentrations > 3.0 mIU/L treated with L-thyroxine for a mean period of 3.3 years (0.6–6 years) [31]. The therapy aimed at achieving TSH concentrations < 3.0 mIU/L before pregnancy and < 2.5 mIU/L during pregnancy. In 58/69 (84%) of the treated women, pregnancy was achieved, of which 21 naturally and 37 by assisted reproduction techniques; 17 (29%) of the women who became pregnant subsequently had a miscarriage. The presence of aTPO or aTg was found in 42% of subjects, with equal frequency in the group of women who became pregnant and those who did not [31]. Verma et al. administered L-thyroxine to 94 subfertile women with hypothyroidism, among whom, based on unclear criteria, 59 (63%) were diagnosed with subclinical (TSH 4–6 mIU/L) and 35 (37%) with the overt hypothyroidism (TSH > 6.0 mIU/L) [32]. Subjects with tubal factor infertility, endometriosis, and pelvic inflammatory diseases were excluded from the study. The dose of L-thyroxine, depending on TSH concentrations, was 25–150 µg/day. 76.6% of the treated patients became pregnant between 6 weeks and 1 year after the start of treatment. The authors of the abovementioned work concluded that L-thyroxine treatment might increase the chances of pregnancy in subfertile women with subclinical hypothyroidism. However, the lack of a control group not treated with THs makes it impossible to formulate this conclusion unequivocally.

To date, the benefits of screening for hypothyroidism in males treated for fertility disorders have not been clearly confirmed. A study comprising 172 patients with male infertility found no relationship between THs and semen parameters [33]. An earlier study by Poppe et al. showed no difference in the frequency of thyroid dysfunction and elevated anti-thyroid antibodies in groups of men with normal and abnormal semen parameters [34]. However, a study published in 2021 observed a higher prevalence of elevated aTPO concentrations in patients with abnormal semen parameters compared to a control group of men from couples with a non-male infertility factor (men with confirmed normozoospermia) [35]. In patients with abnormal semen parameters (except in cases of non-obstructive azoospermia), TSH concentrations were higher than in the control group [35]. In a Pakistani study, the authors reported a higher prevalence of subclinical hypothyroidism in patients with abnormal semen parameters [36]. In the aforementioned studies, one must note the relatively small size of the study groups, which makes inference difficult [35].

Subclinical hypothyroidism and fertility of couples trying to conceive using ARTs

Assisted reproductive techniques (ARTs) in the initial phase involve controlled ovulation stimulation with clomiphene citrate, gonadotropins or GnRH analogues, followed by hCG administration to mature oocytes. At that time, there is a rapid increase in oestradiol concentrations comparable to advanced pregnancy, which increases thyroxine-binding globulin (TBG) concentrations and can reduce free THs concentrations and, consequently, further increase TSH concentrations. The risk of developing or exacerbating subclinical hypothyroidism is higher in women with positive aTPO and in ovarian hyperstimulation syndrome (OHSS) [37, 38].

Most currently published studies evaluating ART results according to TSH concentrations show no difference in pregnancy and live birth rates between women with TSH < 2.5 mIU/L and TSH in the range of 2.5–5.0 mIU/L. Although in a retrospective study comprising 164 women undergoing ART, Fumarola et al. found a lower pregnancy rate in women with TSH > 2.5 mIU/L compared to women with TSH ≤ 2.5 mIU/L: 8.9% vs. 22.3%, respectively, p = 0.045; the results of subsequent works do not support this finding [39].

Reh et al. retrospectively analysed 1055 women who underwent in vitro fertilisation (IVF), and assessed outcomes according to TSH cut-off values: 2.5 mIU/L and 4.5 mIU/L [40]. Comparing the study groups with TSH ≥ 2.5 mIU/L vs. TSH < 2.5 mIU/L, the authors found no statistically significant differences in the percentages of clinical pregnancy rates (52% vs. 47%), live births (39% vs. 34%), or pregnancy losses (12.5% vs. 13%). Similarly, for subjects with TSH ≥ 4.5 mIU/L vs.< 4.5 mIU/L, there were no significant differences in the rates of clinical pregnancies (54% vs. 48%), live births (43% vs. 34%), and pregnancy losses (9% vs. 13%). Among subjects with TSH > 4.5 mIU/L, the mean TSH concentration was 5.1 mIU/L (TSH concentration range: 4.4–6.7 mIU/L). The above observations were confirmed by Chinese authors in 627 women undergoing an IVF procedure: in the groups with TSH > 4.5 mIU/L and ≤ 4.5 mIU/L, similar rates of clinical pregnancies, live births, and miscarriages were observed [41].

Unuane et al. retrospectively evaluated the percentage of live births among 2406 women, including 333 with elevated aTPO concentrations, undergoing IVF/intracytoplasmic sperm injection (ICSI) [42]. The authors did not observe differences between women with TSH < 2.5 mIU/L and TSH < 5 mIU/L, also after adjusting for aTPO concentrations [42]. Green et al. tried to ascertain the optimal TSH concentration in the range between 0.5 and 2.5 mIU/L, for the best outcomes of the IVF procedure [43]. They tested TSH on day 8 after embryo transfer and analysed TSH concentrations in 0.5 mIU/L increments; however, they found no differences in implantation or miscarriage rates.

Repelaer Van Driel-Delprat et al. conducted a retrospective study among 990 women undergoing an intrauterine insemination (IUI) procedure, to determine whether TSH concentrations in the lower and upper quartile affect the obtained results [44]. They observed no differences in the clinical pregnancy success rates between TSH in the highest quartile (2.35–4.5 mIU/L) versus the 3 lower quartiles (0.3–1.21 mIU/L; 1.22–1.75 mIU/L; 1.76–2.34 mIU/L), respectively: 38.6% vs. 38.6% vs. 33.8% vs. 32.5% (p = 0.376). They reported no differences in miscarriages and live birth rates. Similar results were obtained by Tuncay et al., who evaluated the results of IUI in 302 women according to TSH concentrations: they found no statistically significant differences between the group with TSH 0.38–2.49 mIU/L and the group with TSH 2.5–4.99 mIU/L in the clinical pregnancy rates, miscarriages, and live births, as well as in the condition of the newborns: birth weight and need for neonatal intensive care unit stay [45].

The results of L-thyroxine treatment of women with subclinical hypothyroidism before IVF/ICSI were summarised in a meta-analysis of 3 prospective randomised studies by Velkeniers et al. in 2013. In two of the analysed studies, the criterion for the diagnosis of subclinical hypothyroidism was a TSH concentration > 4.0 mIU/L and > 4.5 mIU/L, and in one, TSH was within the reference range of 0.27–4.2 mIU/L, and they differed in the aTPO status [46]. The treatment strategy in the analysed studies also varied: a fixed dose of L-thyroxine 1 µg/kg body weight (b.w.)/day; 50–100 µg/day and 50 µg/day before ART, subsequent dose modification to achieve TSH < 2.5 mIU/L. A meta-analysis showed that L-thyroxine treatment had no effect on the clinical pregnancy rate (relative risk [RR] of 1.75; 95% CI: 0.90–3.38; p = 0.098 but resulted in a higher number of live births (RR of 2.76; 95% CI: 1.20–6.44; p = 0.018) and fewer miscarriages (RR of 0.45; 95% CI: 0.24–0.82; p = 0.010).

In a recently published prospective cohort study, Cai et al. evaluated the results of L-thyroxine treatment in 270 women with TSH > 4.2 mIU/L undergoing an IVF procedure [47]. The treatment aimed to achieve TSH within the laboratory normal range of 0.2–4.2 mIU/L. There were no differences in the clinical pregnancy rates, miscarriages, and live births between the treated groups with achieved TSH 0.2–2.5 mIU/L and TSH 2.5–4.2 mIU/L [respectively: 47.4% vs. 38.7% (p = 0.436); 7.4% vs. 16.7% (p = 0.379); 43.9% vs. 32.3% (p = 0.288)]. The authors concluded that achieving TSH < 2.5 mIU/L before ARTs does not result in a higher proportion of pregnancy success rates. However, it should be noted that, although there was no statistical relationship, achieving lower TSH values during treatment reduced the risk of miscarriage and increased the number of live births.

According to a systematic review published under the auspices of the Cochrane Library in 2019, which assessed the effect exerted by L-thyroxine treatment in euthyroid women with fertility problems undergoing ART, with coexisting subclinical hypothyroidism or AITD, on live birth rates, clinical pregnancy success, and miscarriage rates, the low quality of the available evidence did not allow for firm conclusions on the beneficial effects of such treatment [48].

The effect of subclinical hypothyroidism in men on IVF/ICSI using their sperm is even less well understood. In 2021, the results of a retrospective study comprising a total of 2511 couples treated for infertility, among which 229 men were diagnosed with subclinical hypothyroidism, were published. There were no differences in the semen quality or sperm DNA fragmentation index between males with subclinical hypothyroidism and euthyroid individuals. However, among couples in which the man suffered from subclinical hypothyroidism, significantly lower clinical pregnancy and embryo implantation rates were observed. When stratified by age, this effect was found only in the case of males aged 35 years or older [49]. There are no available studies evaluating the effect of L-thyroxine treatment of males from infertile couples on the outcome of ARTs.

Anti-thyroid antibodies and fertility

Quite a different area of study is the effect of elevated aTPO and aTg concentrations on fertility in euthyroid women. A possible direct effect of anti-thyroid antibodies on the ovary has been considered: a study in a small group of 17 subfertile women found aTPO and aTg in the follicular fluid, where their concentrations correlated positively with aTPO and aTg concentrations in the blood [50]. The endometrium during the luteal phase and the placenta are also potential sites of negative effects of aTPO since the expression of thyroid peroxidase at the gene and protein level was found in these tissues [51]. Cross-reactivity between aTPO and hCG in accessing the hCG receptor within the zona pellucida was also observed [37, 52]. Other factors potentially negatively affecting the fertility of women with AITD are their older age, higher TSH concentrations compared to healthy women, and the possible co-occurrence of other autoantibodies, e.g. antibodies to ovarian antigens (autoimmune polyglandular syndromes) [53].

Most of the studies have evaluated the effect of aTPO or aTPO/aTg on female fertility and pregnancy complications (Tab. 2), whereas the pathogenic role of the isolated presence of aTg remains unclear and requires further investigation. In addition to the previously mentioned presence of aTg in the follicular fluid in infertile women with AITD, it has been shown that these antibodies can interfere with the thyroid response to hCG stimulation [54], similarly to aTPO [54, 55].

The frequency of anti-thyroid antibodies among women with infertility reaches 18% and appears to be higher than in the general population of women of reproductive age (approximately 10%) [18, 23]. Nineteen per cent of euthyroid women with positive anti-thyroid antibodies in the periconception period develop subclinical hypothyroidism during pregnancy [56]. The presence of antibodies increases the risk of spontaneous pregnancy loss by about 3 to 4 times, and recurrent pregnancy loss and preterm birth by about 2 times [57, 58]. Approximately 50% of women with positive anti-thyroid antibodies in early pregnancy develop postpartum thyroiditis [59, 60].

In a meta-analysis published in 2011, including 4 case-control studies comprising 334 women with positive aTPO/aTg and 1679 women in the control group, the authors found an association of aTPO-positivity with idiopathic infertility (OR of 1.5; 95% CI: 1.1–2.0) [57].

Elevated anti-thyroid antibody concentrations are more common in specific causes of infertility: polycystic ovary syndrome (PCOS) and idiopathic infertility, reaching approximately 25% [61]. In a 2018 meta-analysis of 13 cross-sectional and case-control studies published between 2004 and 2017 and evaluating a total of 1210 women with PCOS and 987 healthy women, Romitti et al. found a significant association between PCOS and AITD (OR of 3.27; 95% CI: 2.32–4.63) [62]. After considering ethnic differences, it was found to be strongest among women of Asian origin (OR of 4.56; 95% CI: 2.47–8.43). A likely reason for the co-occurrence of PCOS and AITD is polymorphism in the fibrillin gene, which regulates transforming growth factor b (TGFb) activity, which in turn affects regulatory T cells (Treg). Reduced TGFb and Treg activity promotes the development of autoimmune diseases. Another predisposing factor is the high oestrogen/progesterone ratio found in women with PCOS, increasing the risk of autoimmune diseases, and vitamin D deficiency [63]. A recently published study carried out by Polish authors did not confirm the association between PCOS and its different phenotypes and positive aTPO, but the results may have been influenced by the frequent occurrence of elevated aTPO concentrations in the control group (23.9% vs. 21.9% women with PCOS, p = 0.07) [64].

The relationship between AITD and endometriosis remains unclear: it has been demonstrated in earlier works, but the results of contemporary studies are inconclusive [23, 65]. Endometriosis may also be associated with immunological disorders, such as the presence of antibodies against endometrial antigens, a reduced number of NKs and the presence of complement deposition in the eutopic endometrium.

Contemporary studies indicate a potential association between high aTPO concentrations and low ovarian reserve (LOvR) assessed by anti-Müllerian hormone (AMH) concentrations or antral follicle count. Chinese authors noted an association between the presence of aTPO and idiopathic LOvR in a group of women with infertility, from which they excluded cases with iatrogenic and genetically determined LOvR (positive aTPO was found in 28.6% of women with low vs. 15.7% with normal vs. 9.5% with high ovarian reserve, p = 0.020) [66]. The authors also found a significantly higher prevalence of aTg in subfertile women with LOvR and endometriosis compared to subfertile women with normal and high ovarian reserve (respectively: 45.5% vs. 9.6%, vs. 14.3%, p = 0.013) [66]. Korevaar et al. observed an increased presence of positive aTPO and lower FT3 concentration in the cases of LOvR and idiopathic infertility (p < 0.01) but did not confirm such a relationship with respect to aTg [30]. Adamska et al. found that aTPO concentrations negatively correlated with AMH concentrations in women with PCOS (r = –0.4; p = 0.02) [64]. In contrast, two studies by Belgian and Japanese authors in groups of 5000 and 153 subfertile women, respectively, found no association between elevated aTPO concentrations and LOvR [29, 67]. The European Society of Human Reproduction and Embryology recommends testing aTPO concentrations in women with premature ovarian insufficiency, defined as loss of ovarian function before the age of 40 years [68].

A recently published US study did not confirm that elevated aTPO/aTg concentrations reduce fertility in healthy women with 1–2 spontaneous pregnancy losses [25]. In the previously mentioned prospective cohort study within EAGeR, the fertility of 154 women with positive and 900 women with negative aTPO/aTg was compared: during the 6-month observation, 74.0% of women with elevated aTPO/aTg concentrations and 72.2% of women in the control group became pregnant (p = 0.64) [25].

The only prospective randomised trial to date — Thyroid Antibodies and Levothyroxine Trial (TABLET) — conducted in euthyroid women with positive aTPO did not show that L-thyroxine treatment affected fertility [69]. This multicentre study included 952 women with elevated aTPO concentrations and a history of pregnancy loss or infertility. During the 12-month observation, conception was achieved in 56.6% of those treated with L-thyroxine vs. 58.3% receiving placebo (RR of 0.97; 95% CI: 0.88–1.07). No differences in the percentage of live births were observed: 37.4% in the L-thyroxine-treated group vs. 37.9% in the placebo group (RR of 0.97; 95% CI: 0.83–1.14; p = 0.74), nor the frequency of pregnancy complications. The authors found no differences in live birth rates between the treatment and placebo groups according to age (< 35 years vs. ≥ 35 years), baseline TSH concentrations (TSH ≤ 2.5 mIU/L vs. > 2.5 mIU/L), aTPO concentrations (very high vs. high), body weight (BMI: ≥ 25 kg/m2 vs. < 25 kg/m2), history of infertility treatment, miscarriage, and ethnicity. Some limitations of this important study are the inclusion of women who became pregnant both in a natural manner and through assisted reproductive techniques, the use of a fixed dose of L-thyroxine of 50 µg per day during the pre-conception period and during pregnancy, and the exclusion of those women who developed thyroid dysfunction during the observation (approximately 10% in each arm), and thus those who could potentially benefit most from treatment.

No studies have been published to date on the effect of selenium use on fertility in women with anti-thyroid antibodies, although several papers have demonstrated its lowering effect on aTPO and aTg [70].

Data on the association of AITD with male infertility is also scarce. In the Tehran Thyroid Study, no association was found between positive aTPO antibodies and fertility problems in men; compared to fertile men, aTPO concentrations in this group were similar [71]. On the other hand, American authors found an increased risk of autoimmune diseases, mainly rheumatoid arthritis, multiple sclerosis, and psoriasis, as well as Hashimoto’s disease and Graves’ disease, in men with infertility. They attribute this association to lower testosterone concentrations in males with infertility and possible cross-reactions between antibodies and gonadotropins [72].

Anti-thyroid antibodies and fertility of couples trying to conceive using ARTs

Studies evaluating the effect exerted by AITD on the possibility of achieving conception through ARTs are mostly retrospective and characterised by a high heterogeneity related to different causes of infertility among the subjects, heterogeneous thyroid functional status (euthyroidism, variously defined subclinical hypothyroidism), and different in vitro fertilisation procedures (IVF, ICSI). Previous studies have indicated that the presence of anti-thyroid antibodies is an independent risk factor for ART failure and is associated with poorer embryo quality, lower pregnancy rates, higher risk of miscarriage, and lower number of live births [73–75], but this is not fully confirmed in contemporary publications. Between 2016 and 2020, 5 meta-analyses on this issue were published.

In a meta-analysis published in 2016, Busnelli et al. evaluated the results of 12 papers (6 prospective and 6 retrospective) published between 1999 and 2015, which examined the effect exerted by aTPO/aTg on IVF/ICSI outcomes [76]. No differences were observed between aTPO/aTg-positive and aTPO/aTg-negative women in conception rates (OR of 1.11; 95% CI: 0.97–1.27; p = 0.13) (3 studies,1082 women), implantation rates (OR of 0.98; 95% CI: 0.73–1.32; p = 0.91) (2 studies, 918 women), or clinical pregnancy rates achieved (OR of 0.90; 95% CI: 0.77–1.06; p = 0.22) (12 studies, 4876 women). However, a higher miscarriage rate (OR of 1.44; 95% CI: 1.06–1.95; p = 0.02) (12 studies, 4876 women) and a lower live birth rate (OR of 0.75; 95% CI: 0.54–0.99; p = 0.04) (9 studies, 4396 women) were observed. The authors concluded that the presence of anti-thyroid antibodies had no effect on fertility in women undergoing ART but increased the risk of pregnancy loss and reduced the chances of delivering a live baby.

In 2018, Poppe et al. presented a meta-analysis on the impact of AITD on ICSI infertility treatment outcomes [61]. The authors analysed 4 studies published between 2013 and 2015 (2 prospective and 2 retrospective) including a total of 290 women with AITD (one study comprising women with positive aTPO, two studies comprising women with positive aTPO and/or aTg, and one study comprising women with positive aTPO and aTg) and 1565 without AITD. In 3 studies, euthyroidism was defined by a TSH cut-off value of 2.5 or 3 mIU/L, while one study did not report TSH concentration results. There were no statistically significant differences between women with AITD and controls in conception rates (OR of 1.02; 95% CI: 0.89–1.16), implantations (OR of 0.98; 95% CI: 0.73–1.32), clinical pregnancies (OR of 0.91; 95% CI: 0.70–1.18), pregnancy losses (OR of 0.95; 95% CI: 0.48–1.87), and live births (OR of 1.12; 95% CI: 0.62–2.03). The authors believe that the ICSI method, used in male and idiopathic infertility, may be particularly beneficial in women with AITD, although there are no studies comparing the results of IVF and ICSI in this group of patients. An experience of Belgian authors and an analysis of the literature [50, 61, 76] resulted in the recommendation by the European Thyroid Association to choose ICSI as a more effective method than IVF in infertile women with AITD [77].

In a 2019 meta-analysis, Grigoriadis et al. evaluated 14 studies published between 2006 and 2018 (6 prospective and 8 retrospective) comparing the effects of ART (IVF/ICSI) in a total of 1279 women with positive aTPO/aTg with the results of 4680 women with negative anti-thyroid antibodies as a control group [78]. The authors conclude that most of the papers (10/14) showed no negative effect of aTPO/aTg on ART outcomes: conception, implantation, clinical pregnancy, and live birth rates, but they noted a trend towards higher miscarriage rates in women with positive aTPO/aTg antibodies. They also noted that coexisting subclinical or overt hypothyroidism may be an important negative factor affecting ART outcomes.

In 2020, Unuane et al. performed a meta-analysis of 15 studies published between 1999 and 2016 (7 prospective and 8 retrospective) grouping a total of 1202 women with AITD and 7073 controls [79]. Eight of the studies assessed aTPO and aTg, while the others assessed only aTPO; 13 papers comprised euthyroid women and in the other two TSH results were not reported. Most of the papers did not find a negative effect of AITD on the clinical pregnancy success rates (13/15), miscarriages (12/15), or live births (12/15).

A meta-analysis by Venables et al. published in 2020 was dedicated to the outcomes of IVF/ICSI in euthyroid women with coexisting AITD [80]. It included 14 studies (7 retrospective and 7 prospective). The authors found no difference between euthyroid women with positive and negative aTPO/aTg in clinical pregnancy rates (OR of 0.88; 95% CI: 0.69–1.12; p = 0.29; 10 studies), miscarriage rates (OR of 1.18; 95% CI: 0.52–2.64; = 0.69; 7 studies), biochemical pregnancy loss (OR of 1.14; 95% CI: 0.48–2.72; p = 0.769; 4 studies), live birth rate per ART cycle (OR of 0.84; 95% CI: 0.67–1.06; p = 0.145; 5 studies), and live birth rate per clinical pregnancy (OR of 0.67; 95% CI: 0.28–1.60; p = 0.369; 4 studies). There was no difference between groups in the number of embryos transferred, oocytes obtained during the procedures, patient age, and TSH concentrations.

Two contemporary studies have observed poorer embryo quality obtained during ARTs in women with positive aTPO/aTg, but one study found no difference in the proportion of the clinical pregnancy rate [81], while the other did not report such data [82]. Poorer embryo quality potentially influences the increased risk of pregnancy loss reported in some studies.

The Belgian authors also evaluated the effect of aTPO/aTg on IUI outcomes and similarly found no differences between women with positive and negative aTPO in terms of pregnancy success rates achieved, miscarriages, and live births (respectively: OR of 0.98; 95% CI: 0.62–1.55; OR of 0.74; 95% CI: 0.23–2.39, and OR of 1.04; 95% CI: 0.63–1.69) [83]. In contrast, in a randomised prospective study, Seungdamrong et al. analysed the relationship between positive aTPO and TSH concentrations (< 2.5 mIU/L and ≥ 2.5 mIU/L) and ovulation stimulation outcomes with subsequent IUI [84]. The study comprised 750 women with PCOS receiving clomiphene or letrozole for ovulation stimulation and 900 women with idiopathic infertility receiving aromatase inhibitors, clomiphene, or gonadotropins. Women with positive aTPO had similar clinical pregnancy rates (OR of 0.86; 95% CI: 0.57–1.30; p = 0.15) but higher miscarriage rates (OR of 2.17; 95% CI: 1.12–4.22; p = 0.02) and lower odds of delivering a live baby (OR of 0.58; 95% CI: 0.35–0.96; p = 0.03), compared with women without AITD. In contrast, TSH concentrations did not affect fertility, pregnancy, or live birth rates.

In the case of women with AITD undergoing ARTs, particular attention should be paid to the risk of hypothyroidism following controlled ovarian stimulation. The reduced thyroid reserve associated with AITD may result in an inadequate increase in THs production in response to rapidly increasing demand [55, 85]. Hence the need for close monitoring of thyroid function, which should consist of assessing TSH concentrations at the time of the second positive hCG result confirming pregnancy, i.e. approximately 6 weeks after the start of the controlled ovarian stimulation procedure and approximately 3 weeks after ovulation induction, and then serially every 4 weeks until mid-pregnancy and at least once around the 30th week of pregnancy [37, 77]. In view of the potential adverse effects of hypothyroidism on pregnancy, in its recently published recommendations the European Thyroid Association (ETA) suggests considering treatment with L-thyroxine 25–50 µg/day before controlled ovarian stimulation in selected cases of euthyroid women with positive aTPO antibodies but with TSH concentrations above 2.5 mIU/L to achieve TSH concentrations < 2.5 mIU/L [67]. The abovementioned management (according to ETA recommendations: weak recommendation, low-quality evidence) should be considered in women over 35 years old with the ovarian factor of infertility or recurrent miscarriages. The question of whether similar management should be implemented in the case of positive aTg or ultrasound features of autoimmune thyroid disease requires further investigation.

The effect of L-thyroxine treatment on pregnancy rates, miscarriage rates, and live births in subfertile women with AITD undergoing ART was evaluated in 2 prospective randomised clinical trials. In 2005, Negro et al. published the results of a study of 72 infertile women with positive aTPO, who were randomised into a placebo group, or a group treated with L-thyroxine 1 µg/kg/day 4 weeks before controlled ovarian stimulation [86]. L-thyroxine treatment was then continued throughout pregnancy. The control group consisted of 412 infertile women with negative aTPO. The IVF results in terms of the percentage of clinical pregnancies achieved did not differ between the 3 groups: 56% in the L-thyroxine-treated group, 49% in the placebo group, and 55% in the control group. A higher risk of miscarriage was found in women with positive aTPO compared with the control group (RR of 2.01; 95% CI: 1.13–3.56; p = 0.028) and in the placebo group compared with the control group (RR of 1.89; 95% CI: 1.2–3.2; p = 0.034). The authors concluded that L-thyroxine treatment in aTPO-positive women undergoing ART did not improve delivery rates [86]. In 2017, Wang et al. published a study evaluating the results of L-thyroxine treatment of infertile women with positive aTPO undergoing IVF/ICSI procedures — Pregnancy Outcomes Study in euthyroid women with Thyroid Autoimmunity after Levothyroxine (POSTAL) [87]. The study comprised 600 women with infertility randomised into a placebo group (n = 300) and a L-thyroxine-treated group (n = 300), in which L-thyroxine was included at an initial dose of 25–50 µg/day 2–4 weeks before controlled ovarian stimulation, depending on TSH and body weight, and then modified so that TSH in the first, second, and third trimester remained within a range of the following values: respectively, 0.1–2.5 mIU/L, 0.2–3.0 mIU/L, and 0.3–3.0 mIU/L. No statistically significant differences were observed between the treatment and control groups in the percentage of clinical pregnancy rates (35.7% vs. 37.7%, p = 0.61), pregnancy losses (10.3% vs. 10.6%, p = 0.94), and live births (31.7% vs. 32.3%, p = 0.86) [87].

Hyperthyroidism and fertility

In hyperthyroidism, elevated concentrations of SHBG, total oestradiol, testosterone, and androstenedione are observed, as well as the increased conversion of testosterone to oestradiol and androstenedione to oestrone. LH concentrations are elevated at baseline and after GnRH stimulation [4]. However, most women with hyperthyroidism have preserved ovulation, as demonstrated by endometrial biopsy studies. Menstrual disturbances accompanying thyrotoxicosis have been reported frequently in the past: 65% vs. 17% among healthy women; contemporary studies indicate a lower proportion of such disorders: 22% vs. 8% [4,76]. The most observed menstrual alterations are hypomenorrhoea and polymenorrhoea. Among women with hyperthyroidism, fertility problems occur with a frequency of 5.8–50% [88, 89], while the prevalence of hyperthyroidism in subfertile women is similar to that in the general population: 2.1% vs. 1–2.0%, respectively.

The recommendations for the management of hyperthyroidism in women with reduced fertility do not differ from those for all women with thyrotoxicosis planning pregnancy and those for pregnant women. These are outlined in the PTE recommendations for the management of thyroid disease in pregnancy published in 2021 [1].

Radioactive iodine 131I treatment and fertility

Differentiated thyroid cancer is one of the most common cancers in individuals of reproductive age, especially women [90]. One of the treatment procedures for differentiated thyroid cancer is radioactive iodine (RAI) therapy, recommended mainly for patients with an intermediate and high risk of cancer recurrence, as well as for patients with metastatic disease [91, 92]. The results of several studies have suggested a negative effect of past RAI treatment for differentiated thyroid cancer on ovarian reserve, especially in the first year after treatment [93–95]. Not all observations agree on this point: according to studies by Giusti et al. [96] and Mittica et al. [97], the main factor negatively affecting AMH concentrations was the age of the patients but not the history of RAI treatment. Retrospective studies also suggested a younger age of menopause in women treated with RAI and suppressive doses of L-thyroxine compared to patients treated with suppressive doses of L-thyroxine for nodular goitre (49.5 years old vs. 51 years old) [98]. According to EANM guidelines, conception attempts should be ceased for 6–12 months after RAI treatment, among others to minimise the risk of miscarriage [99].

In recent years, several meta-analyses have been published on fertility after RAI treatment. The meta-analysis by Anagnostis et al. [100] focussed on the decrease in ovarian reserve as measured by AMH concentrations in women with differentiated thyroid cancer treated with RAI. In this analysis involving 4 prospective studies, there was a significant decrease in AMH concentrations compared to baseline values at 3, 6, and 12 months after RAI administration. No differences in FSH concentrations were observed. In a systematic review, Piek et al. [101] addressed various aspects related to female fertility that could potentially be affected by RAI therapy. The authors found that menstrual alterations occur on average in 12% of women of reproductive age in the first year after RAI (in some analysed studies, the frequency of menstrual disorders reached 31% compared to 14.5% in an age-matched control group [102]). During the first year after RAI administration, secondary amenorrhoea occurs in 8–16% of treated women. In contrast, a meta-analysis comprising more than 18,000 women treated with RAI and more than 15,000 women in the control group found no difference in pregnancy rates (OR of 0.98; 95% CI: 0.72–1.33).

A meta-analysis by Zhang et al. [103] was devoted to the issues of pregnancy and neonatal complications after RAI treatment. The study analysed a total of 7 studies (over 125,000 participants and almost 14,000 pregnancies). The authors found no significant effect of postoperative RAI therapy on the rates of spontaneous miscarriage (OR = 1.05, p = 0.701), pregnancy loss (OR = 1.07, p = 0.098), preterm birth (OR = 1.02, p = 0.756), stillbirth (OR = 1.58, p = 0.364), or birth defects (OR = 1.00, p = 0.986). The cumulative RAI dose (> 3.7 GBq vs. < 3.7 GBq) exerted no effect on the risk of miscarriage or on birth defects. A lower risk of miscarriage was found in women who became pregnant at least one year after RAI administration compared with previous pregnancies (OR = 0.60, p = 0.000).

A paper on male fertility after RAI treatment was also published in 2021 [104]. In a multicentre study comprising 51 men at least 2 years after RAI treatment (mean follow-up period 5.8 years) who received a cumulative RAI dose of at least 3.7 GBq (median 7.4 GBq), no significant long-term effect of RAI treatment on semen quality was found. The proportion of men with poor semen parameters (semen volume, semen concentration, progressive motility, and proportion of motile sperm below the 10th percentile of the WHO norm) did not differ from the general population. The authors concluded that there was no indication for routine semen cryopreservation before planned RAI treatment. A similar position was previously taken by the authors of the EANM guidelines indicating that there is no need for such measures in cases of cumulative 131I activities below 14 GBq [99].

Thyroid diseases and fertility — Guidelines and recommendations

Recommendation 1

In all women diagnosed with fertility problems, evaluation of thyroid function is recommended (determination of TSH, aTPO, and aTg concentrations; in the case of abnormal TSH concentrations – also free thyroid hormones).

Strong recommendation; moderate-quality evidence (1, ●●●○)

Recommendation 2

There are currently no conclusive data to recommend thyroid function assessment in all males diagnosed with fertility problems.

Weak recommendation; low-quality evidence (2, ●●○○)

Recommendation 3

Thyroid function assessment (including anti-thyroid antibody concentrations) may be considered if other indications coexist in a male with fertility problems (autoimmune disease, gynecomastia, erectile dysfunction) and in the case of abnormal semen parameters.

Weak recommendation; low-quality evidence (2, ●●○○)

Recommendation 4

Overt hypothyroidism should be treated with L-thyroxine according to generally accepted principles, both in females and males diagnosed and treated for fertility problems.

Strong recommendation; moderate-quality evidence (1, ●●●○)

Recommendation 5

Subclinical hypothyroidism should be treated with L-thyroxine in women undergoing fertility treatment — especially before planned ART involving controlled ovarian stimulation (irrespective of anti-thyroid antibody concentrations), so that TSH concentrations < 2.5 mIU/L are maintained.

Strong recommendation; moderate-quality evidence (1, ●●●○)

Recommendation 6

Treatment of subclinical hypothyroidism in males with fertility problems should follow the principles for the general population.

Weak recommendation, no evidence (2, ●○○○)

Recommendation 7

Treatment with low doses of L-thyroxine of 25–50 µg/day may be considered prior to planned controlled ovarian stimulation as prophylactic management in women with TSH concentrations between 2.5 mIU/L and the upper limit of normal range (regardless the aTPO/aTg status) in circumstances such as age > 35 years, ovarian factor of infertility, and recurrent miscarriages. L-thyroxine treatment is aimed at achieving a TSH concentration < 2.5 mIU/L at least 4 weeks before planned controlled ovarian stimulation, and at maintaining it in this range throughout pregnancy.

Weak recommendation; low-quality evidence (2, ●●○○)

Recommendation 8

The procedure of controlled ovarian stimulation should be performed at the earliest 4 weeks after optimal thyroid status is reached.

Strong recommendation; moderate-quality evidence (1, ●●●○)

Recommendation 9

Routine treatment with L-thyroxine before assisted reproduction procedures is not recommended for all euthyroid women with elevated aTPO/aTg concentrations.

Strong recommendation; moderate-quality evidence (1; ●●●○)

Recommendation 10

Euthyroid women with elevated aTPO/aTg concentrations, who are undergoing ART, whether treated with L-thyroxine or not, require close monitoring: assessment of TSH concentrations on the day of the second positive hCG result confirming pregnancy, followed by TSH and fT4 testing every 4 weeks until mid-pregnancy and at least once around the 30th week of pregnancy.

Strong recommendation; moderate-quality evidence (1, ●●●○)

Recommendation 11

In women with elevated aTPO/aTg concentrations and euthyroidism, who are not treated with L-thyroxine, this treatment should be initiated if the first TSH test after controlled ovarian stimulation is above the upper range of the reference value [> 3.18 mIU/L, when electrochemiluminescence (ECL) is used]. L-thyroxine treatment aims to achieve a TSH concentration < 2.5 mIU/L.

Strong recommendation; low-quality evidence (1, ●●○○)

Recommendation 12

Women with elevated aTPO/aTg concentrations trying to become pregnant through ART should be offered the ICSI method as potentially more effective than IVF.

Weak recommendation; low-quality evidence (2, ●●○○)

Recommendation 13

There is no indication for L-thyroxine treatment of euthyroid subfertile males with elevated aTPO/aTg antibodies, including pre-ART.

Strong recommendation; low-quality evidence (1, ●●○○)

Recommendation 14

Women with hyperthyroidism and fertility problems should be treated according to the general recommendations for the management of hyperthyroidism in women planning a pregnancy or pregnant. When choosing a treatment method, the procreative plans and age of subfertile couples should be considered.

Strong recommendation; moderate-quality evidence (1, ●●●○)

Recommendation 15

Planned RAI treatment for differentiated thyroid cancer is not an indication for routine gamete cryopreservation.

Weak recommendation; low-quality evidence (2, ●●○○)

Recommendation 16

The decision to implement surgical treatment of subfertile patients with thyroid cancer should not be delayed because of procreative plans.

Strong recommendation; low-quality evidence (1, ●●○○)

Recommendation 17

In the case of a subfertile woman, especially after the age of 35, the potential impact of therapy on fertility should be discussed when deciding on RAI treatment.

Strong recommendation; low-quality evidence (1, ●●○○)

Recommendation 18

There should be an interval between RAI administration and attempting to become pregnant:

Strong recommendation; low-quality evidence (1, ●●○○)