In this study, we analyzed women with type 1 diabetes with and without PCOS and focused on the possible impact of TAI on parameters of ovarian reserve, i.e. hormonal status and ovarian parameters. Comparison of the four groups revealed differences resulting from PCOS characteristics: both PCOS groups had elevated AMH levels compared to the no PCOS + no TPOAb group. Positive titers of TPOAb did not influence AMH concentrations and ovarian parameters across the studied groups. Interestingly, the only significant predictor of AMH levels was insulin dosage.
The incidence of PCOS in women with T1DM is higher than in the general population. In a recently published study of young Indian women with a median duration of T1DM of 12 years, 52% had PCOM, 26% had hyperandrogenism, and 27% met criteria for PCOS.20A good diagnostic marker for PCOM is AMH. AMH is produced by follicles developing from primary to small antral follicles and has a dual negative function in the ovary: it inhibits recruitment of primordial follicles, decreases the sensitivity of antral follicles to FSH, and separates the FSH-independent and FSH-dependent phases of folliculogenesis.twenty oneThere is evidence, as shown in a recently published meta-analysis, that AMH levels are generally reduced in women with type 1 diabetes compared to healthy controls.12Meanwhile, the debate regarding PCOS and elevated AMH levels in women with PCOS and T1DM continues.14Its pathogenesis has not yet been fully elucidated, although a hypothesis regarding the effect of exogenous insulin acting as a cogonadotropin on the ovaries has been suggested.Our previous study confirmed the association of hyperandrogenism and PCOM with the diagnosis of premenstrual T1DM, the timing of initiation of insulin therapy, and the duration of T1DM.FourImportantly, AMH levels of 3.74 ng/ml were predictive of PCOS in patients with T1DM, with a sensitivity of 90.2% and specificity of 70.3%.14.
Published data suggest a possible link between thyroid function and the ovaries. First, the presence of thyroid hormone receptors in granulosa cells has been confirmed.twenty twoThe expression of these receptors increases with the stage of follicular maturation from primordial and primary follicles to mature oocytes, suggesting that triiodothyronine influences oocyte maturation. Furthermore, the presence of type 2 and type 3 deiodinase enzymes in granulosa cells allows the conversion of thyroxine to triiodothyronine.twenty twoFinally, the presence of thyroid peroxidase in human cumulus granulosa cells has recently been detected.9These results indicate that human ovarian follicles may be dependent on thyroid hormone metabolism, but the presence of TPOAb in TAI may inhibit thyroid peroxidase function and impede human oocyte maturation.
In this study, we focused on the effect of TAI and its relationship with ovarian reserve. Although the relationship between TAI and ovarian reserve has been described in several studies, the data were not consistent and the studies were conducted in populations without T1DM. In the literature, the presence of TAI in women with euthyroidism or subclinical hypothyroidism, regardless of levothyroxine treatment, was a significant predictor of pregnancy loss compared with women without thyroid autoantibodies.twenty threeIn a large cross-sectional study by Polyzos et al., TAI and hypothyroidism were not associated with reduced ovarian reserve, as expressed by lower age-adjusted AMH levels, in women from the general population.16However, decreased FT3 levels and positive TPOAb titers were associated with reduced ovarian reserve and reduced antral follicle numbers in infertile women.twenty fourSome authors have suggested that PCOS may be an autoimmune disease with positive titers of organ-specific and non-organ-specific autoantibodies.twenty fiveClinical reports have shown that women with TAI and PCOS have lower pregnancy rates, lower AMH concentrations, and a higher risk of clomiphene citrate resistance than women with PCOS without TAI.26,27Furthermore, in the PCOS group, AMH levels were negatively correlated with TPOAb titers, independent of PCOS phenotype.28In contrast, a study evaluating AMH serum concentrations in TPOAb-positive and -negative PCOS patients found no differences between the groups.29In our study of women with type 1 diabetes, we found no difference in the prevalence of TAI between PCOS and non-PCOS, probably due to the strong coexistence of type 1 diabetes with other autoimmune diseases, mainly TAI and celiac disease.30.
The factors that may affect ovarian reserve in women with T1DM with or without PCOS are still under debate. Our results could not confirm the effect of the presence of TPOAb on AMH values in women with T1DM. Given the similar prevalence of TAI in both PCOS and non-PCOS groups, the presence of TAI is probably not an important factor determining ovarian reserve in this group of women. Other published reports on T1DM patient groups have suggested that insulin administration method or long-acting insulin dose may affect PCOS and AMH concentrations.14,31A meta-analysis of PCOS in type 1 diabetes highlighted the possibility that systemic hyperinsulinemia may be involved in the development of ovarian dysfunction, but individual predisposing factors and other unknown factors also need to be taken into account.1However, the Epidemiology of Diabetes Interventions and Complications (EDIC) study of women with T1DM, with a mean age of 35 years, found that one-third of study participants had elevated AMH levels and the presence of PCOM. Factors such as lower insulin dosage, younger age, non-smoking, and higher T levels were associated with AMH levels.32This is consistent with our study, where we found that TDI was negatively correlated with AMH concentrations in the entire study group of women who were younger than those in EDIC, with a median age of 26 years, but with a similar duration of T1DM. After observing the EDIC group of women with T1DM for 17 years, we found that AMH concentrations declined similarly to women without T1DM and were not associated with time-weighted insulin dose.13It has been suggested that insulin may not act as a co-gonadotropin during later reproductive years, but larger studies are needed to clarify these mechanisms.
Analyzing other variables related to ovarian function, we found that FSH concentrations were decreased in PCOS compared to noPCOS, with or without TPOAb, and estradiol concentrations were decreased in the PCOS + TPOAb group compared to noPCOS + TPOAb. Furthermore, we found that the LH/FSH index correlated with AMH and O-FN. A meta-analysis of studies in women with T1DM focusing on ovarian reserve also found decreased FSH concentrations, but did not correlate with AMH. However, the authors showed that estradiol levels were negatively correlated with daily insulin dose and HbA1c.12It is known that gonadotropins affect AMH production, and high levels of AMH in PCOS are associated with overproduction of androgens through the influence of LH.33A similar explanation may be possible for type 1 diabetes, where insulin enhances the effect of LH on theca cells, promoting the accumulation and growth of antral follicles.twenty oneThe suppression of FSH levels during the follicular phase of the cycle observed in our study supports a role for AMH in regulating the initiation of follicular growth and in setting the threshold for FSH sensitivity before ovulation.33Additionally, there is evidence that AMH reduces aromatase activity in granulosa cells in PCOS, decreasing estradiol production.33On the other hand, there is also an association between high estrogen levels and autoimmunity in PCOS, and Arduc et al. found that TPOAb-positive PCOS women had higher serum estradiol levels than TPOAb-negative women.34However, this result was not replicated in our study.
The present study has several limitations. First, the number of study participants is relatively small. Also, the evaluation of only TPOAbs mainly related to TAI does not address broader autoimmunity. We did not specify AMH levels according to age. There is no information on the duration of TAI, age at TAI occurrence, or time of association with menarche, which could help in the interpretation of the data. There was also no comparison of the study group with women without T1DM. However, our study was conducted in women within the euthyroid range to exclude the influence of hypothyroidism and in women in the follicular phase, which was necessary for the hormone data analysis.
