Delay in childbearing age and declining fertility have become increasingly prominent issues. One of the more challenging issues in reproductive health is the irreversible loss in female fertility,which is a key element influencing SA. The risk of SA after pregnancy increases as women age because their ovaries generate fewer collected follicles, lower-quality eggs, and a reduced capacity to produce high-quality oocytes. Age of the female has a substantial link with the likelihood of abortion and is an independent risk factor impacting fertility and pregnancy outcome. Women between the ages of 20 and 29 have the lowest abortion risk, accounting for 12%, while women over 45 have the highest risk,accounting for 65%. The age-related rise in trisomy frequency is thought to be the cause of this association[21]. It is discovered that abnormalities in the embryonic chromosome account for at least half of early SA cases, with older women more likely to have these abnormalities[22][23].The most frequent reason for abortions is trisomy in the embryo. While the risk of other trisomies typically rises rapidly around the age of 35, the risk of trisomy 16 increases linearly from 20 to 40 [28].Around the age of 35, women's natural fertility starts to wane, and infertility typically sets in many years before menstruation totally stops. The loss of natural fertility brought on by postponing childbearing cannot be made up for by ART, despite the fact that it can assist certain couples with infertility issues. It is well accepted that ovarian reserve function may be a good indicator of future fertility and that oocyte quality appears to decrease with age. According to studies conducted on IVF oocytes, the rate of oocyte aneuploidy rises with age. In women under 35, the rate is modest (10%), but it rises to 30% at age 40, 40% at age 43, and 100% in women over 45 [24].To reflect the function of the ovarian reserve, one can use an ovarian reserve test, an ultrasonic measurement, and basic hormone levels. AMH inhibits the early stages of follicular development and is generated by granulosa cells found in preantral and tiny antral follicles. Research has demonstrated that AMH is more sensitive and responsive in the early stages of ovarian function decline, and that it can accurately reflect the number of follicles in non-gonadotropin-dependent stages. It also exhibits good stability throughout the menstrual cycle and across weeks, and it is largely unaffected by objective factors like patients and testers[25].The AMH index of women varies with age; the serum AMH level peaks in adolescence, between the ages of 23 and 25. After that, the level declines with age, declining by roughly 5.6% annually until the menopausal women's AMH level falls or becomes virtually undetectable, which may indicate a shifting correlation between the serum AMH level and ovarian reserve function. In addition to pathogenic variables including genetics, immunology, iatrogenic factors such pelvic surgery, radiation, and chemotherapy, as well as unhealthy lifestyle choices like smoking, ovarian reserve function may decline physically with age.According to Zhang Chunxiao et al. 's[26] literature review, there is a strong positive correlation between the amount of ovarian reactivity and the number of recovered oocytes and the change in serum AMH level. Serum AMH levels with a critical value of 3.21ng/ml had a 94% sensitivity for determining a patient's poor ovarian reactivity, an 88% specificity for diagnosis, and an area under the ROC curve of approximately 0.947.According to findings from other studies, ovarian reserve function can be more accurately determined by AMH level than by FSH index when 8.1 pmol/L is used as the crucial threshold[27]. According to a study, the median amount of serum AMH is 22 pmol/L, while normal values for French women of reproductive age in the first trimester are approximately 2.42 ng/mL[28, 29]. This discrepancy may be caused by race, which should be considered when interpreting AMH value. The age-related drop of Chinese women is even more pronounced than that of European women, with a decrease of 30% and 45% for women aged 28 and 80, respectively, despite the fact that the peak level of AMH in Chinese women at 25 is higher than in European women[30]. Furthermore, compared to white women, African-American women appear to have lower serum AMH levels, but their age dependence diminishes[31, 32]. AMH was revealed to have a role in the process of threatening abortion as evidenced by the dropping blood AMH level in these patients and the subsequent decrease in its expression level with the occurrence of pregnancy failure. The increase of AMH expression may promote the expression of endothelial tissue factor, enhance the recruitment of neutrophils, and further promote the formation of threatened abortion in early pregnancy[33]. Whether an AMH level has prognostic significance for SA is still up for debate. Certain writers have demonstrated a connection between pregnancy loss and low AMH levels. For example, in 2017, Pils et al.[6] confirmed that 62% of women with very early abortion showed low AMH level (≤1 ng/mL). A meta-analysis [34] found that low AMH (< 1.1 ng/ml, RR 3.66, 95% CI 2.1-6.4, P<0.001) significantly increased the risk of miscarriage in the first trimester of pregnancy. Serum AMH could be a useful indicator for estimating the likelihood of an early abortion. According to Bruno et al. [18], patients with lower serum AMH levels had a significantly higher abortion rate. The Box-Tidwell test also confirmed that the abortion rate was present in all age groups and that the impact of AMH on the rate (OR 0.92, 95% CI 0.87-0.98) was unaffected by age (OR 1.08, 95% CI 1.03-1.12). It was discovered in Keane[27] that patients with low AMH levels had the highest abortion rates, and that age dependence rose. The incidence of abortions increased with age. Some authors, on the other hand, contend that there is no clear connection between AMH and miscarriage. A meta-analysis, for instance, found that while AMH level can predict ovarian responsiveness in controlled ovarian hyperstimulation, it cannot increase the incidence of pregnancy or live births. This was determined by reevaluating 20 randomized controlled studies[35]. According to Zarek et al. [36] clinical abortion following a spontaneous conception is unrelated to serum AMH levels. Although AMH concentration is thought to have a function in SA, this investigation found no correlation between clinical pregnancy loss and low (< 1.0 ng/mL) or high (> 3.5 ng/mL) levels of AMH. Thus, more research is required to determine the predictive usefulness of blood AMH level in SA.
Through the creation of a diagnostic experiment AMH could not be used as an effective biological index to predict whether abortion occurred, but it could be used as a screening index for healthy individuals, and a normal level of AMH had a low risk of abortion, according to a meta-analysis of this study's data on the sensitivity and specificity of AMH in predicting spontaneous abortion, which were 0.14 (0.13-0.15) and 0.81 (0.80-0.82). This is in line with the findings of Semih's research [16], which examined the pregnancy outcome of IUI cycles and discovered that there was no variation in the abortion rate across all AMH levels or between women under 35 and those over 35. After IUI, ovulation induction and pregnancy cannot be predicted by female age or AMH. AMH should not be utilized as a measure of oocyte quality, according to Dai et al. [22], who also believe that the hormone has a limited impact in predicting the development potential of embryos in vitro but no role in predicting the development potential of embryos in vivo. In a similar vein, women with AMH < 5.41 pmol/L and women with AMH ≥5.41 pmol/l showed no discernible difference in the rate of biochemical pregnancy or abortion (41.3% vs. 41.6%, P=0.809) in the trial conducted by Alison Richards et al. [23].
The primary cause of study heterogeneity is the threshold effect, which is brought about by various articles' use of distinct diagnostic cutoff values. The updating of test technology and diagnostic technique iterations can lead to changes in the diagnostic cutoff value of the same reagent, conversely, various reagents can detect the same diagnostic test. As a result, it is unavoidable that different publications would have varied diagnostic criteria. As a result, before doing any data analysis, we determine whether a threshold effect exists between the studies. There is currently no agreed-upon threshold for low AMH levels in clinical settings, despite the fact that AMH is frequently utilized to indicate ovarian reserve and forecast ovarian stimulation response. Research indicates that a threshold AMH value of 1.211 ng/mL would be a fair middle ground for differentiating between women with normal ovarian function and those with dysfunctional ovarian reserves[37]. There was a threshold effect among the various studies in this meta-analysis since the included articles established different low AMH level cut-offs (0.7-2 ng/mL) and used different AMH detection reagents. As a result, we use the SROC approach to thoroughly assess the evaluation test's prediction accuracy. This study approach is more accurate since it considers the possibility that the diagnostic efficacy may be impacted by the various AMH cut-off values chosen by various studies. The threshold effect study revealed a Spearman correlation coefficient of 0.800, p=0.000, indicating that varying thresholds were the primary cause of the variability of spontaneous abortion in AMH. Certain research employ the best threshold rather than predetermining it, which could lead to an accuracy inaccuracy in the predictions. According to some research, the SA rate is highest (IRR, 2.1 95% CI 1.1, 3.9) when the outcome of the abortion is at the extreme value of AMH level, that is, AMH is below the 10th percentile (0.4 ng/mL). Additionally, Lyttle et al. [21] showed that among women under 35, those with AMH < 0.4 ng/ml had a 2.3-fold increased incidence of abortion compared to women with AMH ≥1 ng/mL. However, we are unable to conduct a meta-analysis and assess the connection between abnormally low AMH and SA because of the paucity of original research on additional stratification of low AMH level.
Heterogeneity can result from the threshold effect as well as from the selection of various populations and variations in the caliber of the investigation. Age-stratified subgroup analysis yields essentially identical results, with no discernible good predictive value. In order to fit the SROC curve, the young group was included in the study. The Q index was 0.5386, the AUC was less than 0.6, and the AUC was 0.5514. As a result, it is thought that AMH has no prognostic value for spontaneous abortions in young women. A low serum AMH level does not guarantee a poor pregnancy result for young women. It is not sufficient to demonstrate that the quality of follicles influences the result of pregnancy, rather, the low AMH level may represent a decline in follicles and have some bearing on the pregnancy rate. The AMH level is typically at a normal level, indicating that some older women still have great fertility possibilities in terms of ovarian function reserve, and that their pregnancy outcome and fertility potential are good. The physiological characteristics of older women determine the deficiency of ovarian function itself. However, the AMH has no valid prognostic value in this group, and the computed AUC and Q index for older women in this research are 0.5080 and 0.5060, respectively. According to Dai et al. [22], AMH should not be utilized to determine the quality of oocytes in older women. Regardless of age, AMH level cannot be utilized as a reliable indicator to predict SA in this trial. The outcomes of earlier research on the predictive significance of AMH for pregnancy following in vitro fertilization and embryo transfer support it. Although ovarian reserve and reaction do not influence the expected ability of a particular embryo, it is thought that a low AMH level is preferable to low egg quality. Despite this, there is some clinical utility for using low AMH levels as a routine screening tool for women undergoing reproductive treatment[38, 39]. Normal ovulation and high-quality oocytes may be more important in SA than the quantity of ovarian reserves.
One of the main challenges to fertility nowadays is the deterioration of ovarian reserve function. To achieve conception in patients with reproductive needs, ART such as artificial insemination, in vitro fertilization-embryo transfer, and its offshoots are required. With all forms of assisted reproductive technologies, the quality of follicles and their ability to form high-quality embryos vary, and after women take ovulation-promoting drugs, their ovarian responsiveness varies due to differences in ovarian reserve function. These variations also affect the final pregnancy rate and pregnancy outcome. Prior research has indicated that the causes of SA during an assisted reproductive technology pregnancy are numerous and varied, with a rate of approximately 10~20% both domestically and internationally. The factors could include the age of the expectant mothers, the amount of medication used to induce ovulation, ovarian function, embryonic chromosomes, and more[40, 41]. As a result, by integrating naturally pregnant women and doing prospective experimental research, it can more accurately reflect the association between AMH and SA. However, gathering data on naturally pregnant patients in a clinic and maintaining follow-up might be challenging. Furthermore, there was an increase in the percentage of pregnant women with lower AMH levels, and the majority of the literature that was retrieved concentrated on examining the predictive role of AMH in SA incidence in women on antidepressant therapy. The total DOR value of the natural conception study was 1.78 (95%: 1.12, 2.84), and the chance of SA at low AMH level was 1.78 times higher than that at normal or high AMH level, according to the heterogeneity subgroup analysis of conception mode. In patients with natural conception, low AMH may be associated with the development of SA; further experimental research is required to confirm this. According to the SROC curve fitted by ART's conception, AUC is less than 0.6, Q index is 0.5004, and AUC is 0.5005. It is believed that there is no prognostic value for spontaneous abortions caused by ART in AMH. The IUI research exhibits a good prediction value, as seen by its AUC under the SROC curve of 0.8214 and Q index of 0.7553. On the other hand, the area AUC under the SROC curve fitted by embryo transfer is 0.5452 and the Q index is 0.5339, indicating no predictive value. It is believed that the serum AMH level has no bearing on the quality of oocytes, has little bearing on the potential for embryonic development in vivo, and has no bearing on the clinical pregnancy outcome of in vitro fertilization. This is based on the correlation between ART and spontaneous abortion. On the other hand, AMH measurement reflects the ovarian reserve function of patients prior to ovulation induction cycle, but it cannot predict the occurrence of SA. Consequently, it is not advised to regularly check for AMH in order to forecast SA.
In conclusion, a low level of AMH is not regarded as a predictor of SA for screening purposes in order to prevent overtreatment, nor can it be used as a basis for predicting the risk of SA. In diagnosing ovarian status, however, AMH is more useful than estrogen level. It also somewhat reflects women's fertility and may be a useful guide for helping women conceive, creating customized ovulation promotion plans in clinics, minimizing ovarian overstimulation, and lessening the financial burden on patients. Larger-scale epidemiological and clinical investigations are required to further validate the aforementioned conclusions because of the limitations in the amount and quality of included studies.