Serum hormone concentrations in 71 women, with diagnoses unexplained infertility and male infertility, were studied during one arbitrary natural menstrual cycle to investigate the possibility of predicting live birth. There were no significant differences in hormone levels or pregnancy outcome between women with unexplained infertility and women with male infertility.
One reason for increased infertility problems in Sweden is that women tend to delay their pregnancies, which results in a reduced fertility rate . This is also seen in the present study where it is shown that women at the age of 32 or less have the highest chance of a future live birth. Age of the women had the best predictive value of live birth, which is also known from previous studies [28, 29]. The increasing age has also been associated with a shortening length of the menstrual cycle, and the women with cycle length < 26 days has been shown to have significantly lower pregnancy rates after infertility treatment compared with the women with cycle length >34 days . However, this could not been shown in our study, where the menstrual cycle length was not related to the pregnancy outcome.
The presence of one progesterone value ≥ 32 nmol/L was indicative of ovulation . In the present study, an ovulatory cycle as assessed by assay of serum P was a better predictor of a future live birth than levels of any of the other hormones. It has previously been suggested that women with unexplained infertility have diminished ovarian reserve , but this was not noticed in our study.
The use of AMH for determination of ovarian function instead of FSH and/or FSH:LH ratio has increased, and therefore we studied the correlations between these parameters. Serum concentrations of AMH correlated with the FSH:LH ratio, and also with inhibin B levels, which leads to the conclusion that AMH could replace these markers in assessment of ovulatory function. The advantage of AMH is its capacity to maintain relative stable levels during the menstrual cycle  although circadian variations have been observed . The disadvantage of AMH is that there are still no international standard assays for AMH measurement, which makes comparison between different laboratories complicated .
It has previously been shown that FSH:LH ratio can be used as predictor of pregnancy outcome in infertile women , but this was not confirmed in the present study. It has also been demonstrated that AMH is a better marker than age, FSH on cycle day 3 or inhibin B for prediction of IVF pregnancy success , which could neither be shown in the present study. However, our data showed that AMH combined with age and an ovulatory menstrual cycle was predictive for future live birth.
There are only limited data on inhibin B and AMH, and we are not aware of any studies in which groups of infertile women have been compared. However, there was considerable individual variation in serum levels of inhibin B and PRL, which shows that these hormones are not useful in diagnosis and prediction of live birth in women with unexplained and male infertility.
We found no differences in TSH levels between women in the studied groups. Previously, Cramer et al.  demonstrated relatively high TSH levels and relatively low PRL levels in women with male infertility. Conversely, Arojoki et al.  found the lowest TSH levels in women with male infertility, and the highest levels of TSH in women with unexplained infertility, and ovulatory dysfunction. However, menstrual disorders in women with hypothyroidism are more rare than previously reported . In our study TSH as a single variable was not predictive for future live birth, but the combination with age of 32 or less, ovulatory cycle and TSH ≤ 2.5 mIU/L resulted in a predictive value of 88%.
The strength of the present study is the number of hormones measured. Eight different hormones were measured at various occasions during the menstrual cycle and the patients needed to come to the clinic several times during one natural cycle, which may have been challenging for women undergoing in vitro fertilisation. Additionally, our study group was well-defined in two selected groups; unexplained infertile and male infertility, while many previous studies have included patients regardless of cause of infertility. To our knowledge, the long-term probability of live birth in women with unexplained infertility and women with male infertility has not previously been studied.
A weakness of our study is the long recruitment period, which resulted in a prolonged follow-up time. Possible factors (e.g. stress) influencing the serum hormone levels were not taken into account, which may be considered a drawback.