The women from 30 to 44 years old are at high-risk of ovarian aging, and they are the main population receiving ART . The clinical treatment of these patients generally adopts an "empirical" strategy, in which the choice of COH protocol is based on the patient's age, BMI, ovarian reserve, and clinical experience . Before the implementation of ART, clinicians can predict the patient's ovarian response according to age-related ovarian reserve and provide patients with pre-treatment consultation.
In the test of assessing ovarian reserve function, AMH has extremely high specificity and sensitivity , and is negatively correlated with age . It can reliably predict the ovarian response in ART and the number of retrieved oocytes . Due to the stability of AMH within and between cycles, as well as the convenience of AMH test, AMH is more and more recognized as the preferred clinical biomarker for the ovarian response to controlled ovarian stimulation [26, 27]. Although AMH testing lacks international standards and there are differences between different testing methods , the AMH tests involved in this study all used the same method, which did not affect the reliability of the results.
This retrospective cohort study aims to explore the relationship between AMH and the number of high-quality embryos in the study population. The results showed that younger female, lower bFSH, and higher AFC were associated with higher AMH, which was consistent with the conclusion that AMH was negatively correlated with age and bFSH, and positively correlated with AFC . With the decrease of AMH, the number of retrieved oocytes, the number of MII oocytes, and the number of high-quality embryos decreased significantly, indicating that in this study population, AMH was not only related to the number of retrieved oocytes , but also related to the quality of oocytes and embryo development . In both the minimally-adjusted model and the fully-adjusted model, AMH was positively correlated with the number of high-quality embryos (β = 0.4, 95%CI: 0.3-0.5, P<0.001, and β = 0.3, 95%CI: 0.2-0.4, P <0.001). We also treated AMH as a categorical variable according to the clinical cut-off points to divide the patients into low, medium and high AMH groups. The results showed that the high AMH group (≥8.5 ng/mL) had significantly more high-quality embryos than the low AMH group (<0.7 ng/mL) (P<0.05). Sensitivity analysis showed that, compared with the low AMH group, the number of high-quality embryos in high AMH group increased by 2.2 units for each increase in AMH, and the increase was significant (P=0.003). These results indicate that AMH is closely related to the prognosis of reproduction.
The study from Tian Haiqing et al.  showed that AMH was positively correlated with the number of available embryos in POSEIDON groups 3 and 4 patients. This study emphasized the number of embryos due to the restriction of poor ovarian reserve (POR) population; the ART outcome was the number of embryos, so embryo quality was not studied. The study population included in this study was 30-44 year-old patients. Our results showed that AMH was positively correlated with the number of high-quality embryos, indicating that AMH was related to both the number and quality of embryos. In addition, the above study did not conduct non-linear relationship analysis, threshold effect analysis, or stratification analysis. In our study, we performed generalized additive model (GAM) analysis, and the results showed that there was a nonlinear relationship between AMH and the number of high-quality embryos. The two-piecewise regression model identified the inflection point as AMH=7.1 ng/mL. On the left side of the inflection point (AMH<7.1 ng/mL), AMH was positively correlated with the number of high-quality embryos (β = 0.5, 95%CI: 0.3-0.7, P<0.001); on the right side of the inflection point (AMH≥7.1 ng/mL), the correlation was not significant (β = 0.1, 95%CI: -0.1-0.2, P=0.518). This result demonstrated that the relationship between AMH and high-quality embryo number was not simply linear, but had obvious segmentation effects. The results of stratified analysis and interaction test showed that, after adjusting for confounding factors (female age, COH protocol, AFC), the positive correlation between AMH and the number of high-quality embryos was stable in each subgroup.
AMH is closely related to female age. Several studies [32–34] measured the low AMH values for certain ages: 3.0 ng/mL for 25 years old, 2.5 ng/mL for 30 years old, 1.5 ng/mL for 35 years old, 1 ng/mL for 40 years old, and 0.5 ng/mL for 45 years old. Since AMH is positively correlated with the number of high-quality embryos, for patients with low AMH values, the clinicians should be optimistic about the assessment of their existing ovarian function and the possible number of high-quality embryos in IVF/ICSI. Embryo quality is an important guideline for successful implantation and live birth during the embryo transfer cycle ; it is also a new factor influencing the adverse outcome of newborn . Poor embryo quality is associated with an increased risk of miscarriage, ectopic pregnancy, and low birth weight . On the contrary, high-quality embryos will increase the implantation rate, clinical pregnancy rate, and live birth rate . Therefore, embryo quality is an important surrogate indicator of ART outcome .
This study has certain clinical values. First, after adjusting the female age, COH protocol, and AFC, we observed a nonlinear relationship between AMH and the number of high-quality embryos in the first ART cycle of 30-44 year-old infertile women. Secondly, through the threshold effect analysis, we accurately identified the inflection point of the relationship between AMH and high-quality embryo number. Based on these results, the clinicians can obtain information about the potential reproductive prognosis of the patients, which can help estimate the number of IVF/ICSI cycles needed to get a better outcome.
The advantages of this study are as follows: First, the generalized linear model was used to evaluate the linear relationship between AMH and the number of high-quality embryos, and the generalized additive model (GAM) was used to analyze the non-linear relationship; the threshold effect analysis accurately identified the inflection point and provided reliable information for clinical prediction about the potential reproductive prognosis of patients. Secondly, due to the nature of retrospective study, the confounding factors cannot be avoided; however, our study used strict statistical adjustments to minimize the influence of confounding factors on the relationship between exposure factors and outcome variables. Lastly, we used stratified analysis to assess whether the relationship between AMH and high-quality embryo number is consistent between different subgroups, which improved the reliability of the conclusion. The results showed that there was a non-linear correlation between AMH and the number of high-quality embryos, and AMH=7.1 ng/mL was the inflection point. Moreover, the relationship between the two was stable in different subgroups.
The limitation of this study lies in the nature of retrospective study. The results of this study were obtained from the 30-44 year-old patients who used GnRH-A, Mild /PPOS and Ultra-Short GnRH agonist as the first ovarian stimulation protocol; hence, the results cannot be extrapolated to other populations. In addition, this study is a single-center study; thus, a multi-center prospective study is needed to verify the results.