Anaysis of the association between chromosomal abnormalities in early missed abortion embryos and maternal age and AMH levels based on CNV-Seq

Abstract

Objective

Investigating the association between maternal age, anti-Müllerian hormone (AMH) levels, and chromosomal abnormalities in early missed abortion (EMA) embryos, providing evidence for clinical risk stratification and intervention strategies.

Methods

A retrospective cohort study was conducted, including 990 EMA diagnosed patients between June 2020 and December 2024. All patients underwent chorionic villus copy number variant sequencing (CNV-seq) after curettage. The study groups were defined as follows based on two criteria: (1) maternal age at miscarriage: <35, 35–39, and ≥ 40 years; and (2) serum AMH level within one year of enrollment: <1.1, 1.1–4.5, and ≥ 4.5ng/mL. Chromosomal abnormality detection rates and types were compared across groups.

Results

Overall, chromosomal abnormalities were detected in 58.59% (580/990) of embryos, predominantly Numerical Chromosomal Abnormalities (88.62%), with trisomies of autosomes being the most common (67.24%), especially trisomy 22 (19.48%) and trisomy 16 (14.14%). Detection rates increased significantly with maternal age (< 35: 54.52%, 386/708; 35–39: 66.99%, 140/209; ≥40: 73.97%, 54/73; p < 0.05). The proportion of autosomal trisomies rose with age, while 45,X, polyploidy, and Copy number variants (CNVs) decreased. Similarly, lower AMH levels were associated with higher chromosomal abnormality rates (≥ 4.5 ng/mL: 50.16%, 158/315; 1.1–4.5 ng/mL: 60.92%, 232/381; <1.1 ng/mL: 69.23%, 90/130; p < 0.05). The rate of autosomal trisomy and double trisomy/polysomy increased with declining AMH (p < 0.05). After stratifying by age to control for its confounding effect, a significant inverse association between AMH level and abnormality rate was observed only in the < 35 years subgroup (134/47.69%; 216/57.45%; 36/70.59%; p < 0.05), but not in the ≥ 35 years subgroup (24/70.59%; 116/68.64%; 54/68.35%; p > 0.05). Logistic regression indicated maternal age (OR = 1.039, 95% CI: 1.019–1.059) and AMH (OR = 0.931, 95% CI: 0.902–0.960) were independent predictors of chromosomal abnormalities.

Conclusion

Increased maternal age and decreased AMH levels are closely associated with higher risk of chromosomal abnormalities in EMA embryos. As a sensitive indicator of ovarian reserve, AMH reflects oocyte quality and chromosomal stability, particularly in younger women. Combined assessment of maternal age and AMH may improve risk evaluation and inform preventive and intervention strategies for EMA.

Introduction

Missed abortion (MA) is a distinct type of spontaneous miscarriage, defined in China as the death of an embryo or fetus before 28 weeks of gestation without expulsion from the uterine cavity [1]. The incidence in China is approximately 15% [2], causing significant reproductive and psychological impact. Early missed abortion (EMA), occurring before 12 weeks of gestation, accounts for approximately 80% of spontaneous miscarriages [3, 4]. The etiology is complex and includes embryonic chromosomal abnormalities, parental chromosomal defects, immune disorders, endocrine dysregulation, reproductive structural anomalies, thrombophilic conditions, infections, and environmental or lifestyle factors [5]. Chromosomal abnormalities are implicated in over 50–60% of miscarriages [6, 7].

With delayed childbearing trends, advanced maternal age has become increasingly common. Age negatively impacts reproductive function, particularly ovarian reserve. Serum anti-Müllerian hormone (AMH) is a sensitive biomarker for assessing ovarian reserve, identifying diminished ovarian reserve (DOR), and detecting premature ovarian insufficiency (POI) [8, 9]. This study aims to clarify the relationship between maternal age, AMH levels, and embryonic chromosomal abnormalities in EMA, providing clinical evidence for risk stratification and intervention strategies.

Materials and methods

Study design and participants

A retrospective cohort study was conducted at Jiangxi Maternal and Child Health Hospital, including patients diagnosed with EMA between June 2020 and December 2024.

Inclusion criteria

• Confirmed pregnancy by serum hCG level or ultrasound examination;

• Not more than 12 weeks gestational;

• Crown-rump length (CRL) ≥ 7 mm without cardiac activity;

• Mean gestational sac diameter ≥ 25 mm without embryo;

• Absence of embryo with cardiac activity ≥ 2 weeks after a scan that showed a gestational sac without a yolk sac;

• Absence of an embryo with cardiac activity ≥ 11 days after a scan showing a gestational sac with a yolk sac;

• No history of infection or toxic exposure;

• Underwent uterine curettage followed by low-depth, high-throughput whole-genome copy number variation sequencing (CNV-seq);

• Provided informed consent for genetic testing and data use.

Exclusion criteria

Biochemical pregnancies, ectopic pregnancies, multiple pregnancies, molar pregnancies, pregnancies with preimplantation genetic screening (PGS), or cases with incomplete clinical data.

Ethics approval was obtained from the Institutional Review Board of Jiangxi Maternal and Child Health Hospital (Approval No.: EC-KY-202011).

Data collection and grouping

Chromosomal abnormality classification

Based on CNV-seq results, chromosomal abnormalities were classified as:

1. Numerical Chromosomal Abnormalities: including autosomal trisomy/monosomy, sex chromosome trisomy/polyploidy(e.g., XXX, XXY, XYY et.)/monosomy(45,X), double or multiple trisomy, triploidy, or other polyploidy.

2. Copy number variants: also known as Microdeletions/Microduplications, pathogenic or likely pathogenic copy number variations(CNVs).

3. Complex Abnormalities: a combination of aneuploidy and pathogenic CNVs.

Clinical parameters

Data collected retrospectively included: maternal age, parity, gestational age, ultrasound findings, AMH levels, and method of conception.

Grouping

1. Age Grouping: Participants were categorized into three groups based on maternal age at the time of miscarriage: <35 years, 35–39 years, and ≥ 40 years.

2. AMH Grouping: Patients were stratified based on their serum AMH levels measured within one year prior to enrollment. For patients with multiple test results, the average value was used. The grouping thresholds were defined with reference to relevant clinical guidelines and literature [10–12] as follows: <1.1 ng/mL, 1.1–4.5 ng/mL, and ≥ 4.5 ng/mL.

Sample collection and CNV-seq

Chorionic villus tissue (> 10 mg) was collected post-curettage under sterile conditions, washed with normal saline, and stored at 2–8 °C. Genomic DNA was extracted using the QIAamp DNA Blood Mini Kit (Qiagen, Germany). Subsequently, low-depth, high-throughput whole-genome CNV-seq was performed on the BGISEQ-500 platform (BGI, China) with a resolution of 5 Mb. Sequencing data were aligned to the reference genome GRCh37/hg19 and analyzed according to the American College of Medical Genetics and Genomics (ACMG) guidelines.

Statistical analysis

Statistical analyses were performed using SPSS software (version 23.0; IBM, USA). Categorical data are expressed as number (percentage). Group comparisons were conducted using Pearson’s chi-square test or Fisher’s exact test, as appropriate. Multivariate logistic regression analysis was used to calculate adjusted odds ratios (aOR) with 95% confidence intervals (CI). A two-sided p-value of less than 0.05 was considered statistically significant.

Results

Embryonic chromosomal abnormalities

A total of 990 EMA cases were included (mean age: 31.92 ± 4.87 years; mean parity: 0.35 ± 0.62; mean gestational age: 8.18 ± 1.44 weeks) (Table 1).

Table 1.

The clinical characteristics of missed abortion cases with available cytogenic results (n = 990)

Maternal age(years)	32(28.0–35.0)
<35(n%)	708(71.52)
35–39(n%)	207(20.91)
≥ 40(n%)	75(7.58)
gravidity(n)	2.0(1.0–3.0)
Parity(n)	0.0(0.0–1.0)
AMH	2.98 (1.66–5.12)
<1.1ng/ml(n%)	130(13.13)
1.1–4.1.5ng/ml(n%)	545(55.05)
≥ 4.5ng/ml(n%)	315(31.82)

Categorical variables are presented as n (%),continuous variables with normal distribution are presented as Mean ± SD, and continuous variables without normal distribution are presented as Median (IQR)

Overall, chromosomal abnormalities were detected in 58.59% (580/990) of cases, predominantly Numerical Chromosomal Abnormalities (88.62%, 395/580), followed by CNVs (8.62%, 50/580) and complex abnormalities (2.76%, 16/580).

Autosomal trisomies accounted for 67.24% (390/580) of the abnormalities, with trisomy 22 (19.48%) and trisomy 16 (14.14%) being the most common. Sex chromosome abnormalities were mainly 45,X (7.59%), and polyploidy/triploidy accounted for 6.90% (see Table 2; Figs. 1 and 2).

Table 2.

Distribution of chromosomal abnormalities in villus samples from 990 cases of missed abortion

classification	N	%
Normal chromosomal	410
chromosomal abnormalities	580	58.59
Numerical Chromosomal Abnormalities	514	88.62
Autosomal Chromosomal Aneuploidy	395	68.10
Autosomal trisomy	390	67.24
47, XN,+1	0	0.00
47, XN,+2	10	1.72
47, XN,+3	13	2.24
47, XN,+4	12	2.07
47, XN,+5	6	1.03
47, XN,+6	2	0.34
47, XN,+7	10	1.72
47, XN,+8	12	2.07
47, XN,+9	12	2.07
47, XN,+10	5	0.86
47, XN,+11	6	1.03
47, XN,+12	3	0.52
47, XN,+13	20	3.45
47, XN,+14	10	1.72
47, XN,+15	29	5.00
47, XN,+16	82	14.14
47, XN,+17	2	0.34
47, XN,+18	8	1.38
47, XN,+19	1	0.17
47, XN,+20	4	0.69
47, XN,+21	30	5.17
47, XN,+22	113	19.48
Autosomal monosomy	5	0.86
Sex Chromosome Aneuploidy	45	7.76
45,X	44	7.59
Sex chromosome trisomy/polytomy	1	0.17
double/multiple trisomies	27	4.66
triploidy/polyploidy	40	6.90
Other Numerical Chromosomal Abnormalities	7	1.21
Microdeletion and Microdupliction	50	8.62
Complex abnormality	16	2.76
Total	990

Categorical variables are shown as n (%)

Fig. 1.

Distribution of chromosomal anomalies in chorionic villi from 990 cases of missed abortion

Fig. 2.

Distribution of different types of autosomal trisomy

Additionally, 55 cases of mosaicism were detected. As these were not confirmed by karyotyping or fluorescence in situ hybridization (FISH), they were included in the aneuploidy and CNVs groups. Only pathogenic CNVs were included in this study; variants of uncertain significance (VUS) and benign CNVs were excluded.

The concentric rings depict the spectrum of cytogenetic findings:

• Outer ring: Displays the proportion of normal chromosomes to abnormal chromosomes.

• Middle ring: Shows the distribution of different types of abnormal chromosomes, specifically including Numerical Chromosomal Abnormalities, microdeletions/microduplications, and complex abnormalities.

• Inner ring: Displays the distribution of different types of Numerical Chromosomal Abnormalities, specifically covering autosomal trisomy, autosomal monosomy, sex chromosome trisomy, 45,X, double trisomy/polysomy, triploidy/polyploidy, and other aneuploid abnormalities.

^✱The ratios of each type in the inner ring represent the proportion of that specific type within the total population of chromosomal aneuploidy cases.

Age-stratified analysis

Patients were stratified into three age groups: <35 years, 35–39 years, and ≥ 40 years. The chromosomal abnormality detection rates were 54.52% (386/708), 66.99% (140/209), and 73.97% (54/73), respectively, showing a significant increase with advancing maternal age (χ² = 18.025, p < 0.05).

Further analysis of the abnormality types revealed that the detection rate of autosomal trisomy increased significantly with maternal age (χ² = 34.003, p < 0.05). In contrast, the rates of 45,X, triploidy, and CNVs were highest in the < 35 years group and decreased with maternal age (45,X: χ² = 16.833, p < 0.05; triploidy: χ² = 14.411, p < 0.05; CNVs: χ² = 9.155, p < 0.05). The detection rate of double trisomy was highest in the ≥ 40 years group, though the difference was not statistically significant (χ² = 4.002, p > 0.05). No significant differences were observed among age groups for other types of chromosomal abnormalities (p > 0.05) (Table 3; Figs. 3 and 4).

Table 3.

Relationships between chromosomal abnormality detection rates with maternal ages and AMH levels(ng/mL)

	<35 (n/%)	35–39 (n/%)	≥ 40 (n/%)	X2	P	AMH ≥ 4.50 (n/%)	AMH1.10-4.50.50 (n/%)	AMH<1.10 (n/%)	X2	p
Pathogenic chromosomal abnormalities	386(54.52)	140(66.99)	54(73.97)	18.025	< 0.05	158(50.16)	332(60.92)	90(69.23)	16.512	< 0.05
Numerical Chromosomal Abnormalities	330(85.49)	130(92.86)	54(100.00)	13.172	< 0.05	136(86.08)	294(88.55)	84(93.33)	2.998	0.226
Autosomal Chromosomal Aneuploidy	332(86.01)	115(82.14)	48(88.89)	34.815	< 0.05	95(60.13)	235(70.78)	65(72.22)	6.428	< 0.05
Autosomal trisomy	229(59.33)	113(80.71)	48(88.89)	34.003	< 0.05	93(58.86)	234(70.48)	63(70.00)	6.932	< 0.05
Autosomal monosomy	3(0.78)	2(1.43)	0(0.00)	0.844	0.763	2(1.27)	1(0.30)	2(2.22)	3.813	0.085
Sex Chromosome Aneuploidy	41(10.62)	3(2.14)	1(1.85)	14.089	< 0.05	18(11.39)	22(6.63)	5(5.56)	4.120	0.128
45X	41(10.62)	2(1.43)	1(1.85)	16.833	< 0.05	17(10.76)	22(6.63)	5(5.56)	3.235	0.198
Sex chromosome trisomy	0(0.00)	1(0.71)	0(0.00)	3.373	0.334	1(0.63)	0(0.00)	0(0.00)	2.650	0.428
double/multiple trisomies	14(3.63)	8(5.71)	5(9.26)	4.002	0.118	5(3.16)	13(3.92)	9(10.00)	6.993	< 0.05
triploidy/polyploidy	37(9.59)	3(2.14)	0(0.00)	14.411	< 0.05	14(8.86)	23(6.93)	3(3.33)	2.729	0.261
Other Numerical Chromosomal Abnormalities	6(1.55)	1(0.71)	0(0.00)	0.502	0.840	4(2.53)	1(0.30)	2(2.22)	5.869	< 0.05
Microdeletion and Microdupliction	41(10.62)	9(6.43)	0(0.00)	9.155	< 0.05	15(9.49)	31(9.34)	4(4.44)	2.362	0.328
Complex abnormality	15(3.89)	1(0.71)	0(0.00)	4.684	0.066	7(4.43)	7(2.11)	2(2.22)	2.362	0.307
Total	708	209	73			315	545	130

Categorical variables are presented as n (%), and group comparisons were performed using the χ² test or Fisher’s exact test

Fig. 3.

Distribution of chromosomal abnormality rates by maternal age(ng/mL)

Fig. 4.

Distribution of chromosomal abnormality rates by AMH level group

AMH-Stratified analysis

Based on AMH levels, patients were classified into three groups: ≥4.5 ng/mL, 1.1–4.5 ng/mL, and < 1.1 ng/mL. Chromosomal abnormalities were detected in 50.16% (158/315), 60.92% (232/381), and 69.23% (90/130) of cases in these groups, respectively, with the detection rate increasing significantly as AMH levels declined (χ² = 16.512, p < 0.05) (Table 3; Figs. 3 and 4).

After stratifying by age to control for its confounding effect, a significant inverse association between AMH level and abnormality rate persisted in the < 35 years subgroup (47.69% vs. 57.45% vs. 70.59%; χ² = 11.90, p < 0.05), but not in the ≥ 35 years subgroup (70.59% vs. 68.64% vs. 68.35%; χ² = 0.060, p > 0.05) (Table 4; Fig. 5).

Table 4.

Abnormal chromosome proportion by AMH group after age stratification

		AMH ≥ 1.45	AMH1.10-1.45.45	AMH<1.10	X2	P
<35 years	Normal chromosomal	147(52.31)	160(42.55)	15(29.41)
	pathogenic chromosomal abnormalities	134(47.69)	216(57.45)	36(70.59)	11.9	< 0.05
	Total	281	376	51
≥ 35 years	Normal chromosomal	10(29.41)	53(31.35)	25(31.65)
	pathogenic chromosomal abnormalities	24(70.59)	116(68.64)	54(68.35)	0.06	0.97
	Total	34	169	79

Categorical variables are presented as n (%), and group comparisons were performed using the χ² test or Fisher’s exact test;

Fig. 5.

Stacked Bar chart of chromosome proportions by AMH group and age stratum

Analysis of abnormality types by AMH group indicated that the rates of autosomal trisomy and double trisomy/polysomy increased with declining AMH (58.86% vs. 70.48% vs. 70.00%;χ² = 6.932, p < 0.05༛3.16% vs. 3.92% vs. 10.00%༛χ² = 6.993, p < 0.05). Conversely, the detection rates of 45,X, CNVs, and triploidy were highest in the high AMH group and decreased with lower AMH, but these trends did not reach statistical significance(p > 0.05) (Table 3; Figs. 3 and 4).

Logistic regression

Multivariate logistic regression identified maternal age (OR = 1.035, 95% CI: 1.005–1.066) and AMH level (OR = 0.909, 95% CI: 0.868–0.951) as independent predictors of chromosomal abnormalities. Parity, gestational age, and conception method were not significantly associated. (Table 5).

Table 5.

Results of logistic regression analysis

variable	B	Std.Error	Wald	df	Sig.	Exp(B)	95% CI for Exp(B)
							Lower Bound	Upper Bound
Age	0.034	0.015	5.329	1	0.021	1.035	1.005	1.066
AMH	−0.096	0.023	16.776	1	<0.001	0.909	0.868	0.951
constant	−0.360	0.514	0.493	1	0.483	0.697

Multivariable logistic regression analysis was performed. The significance of variables was assessed using the Wald test, and the results are presented as the regression coefficient (B), standard error (Std. Error), p-value, odds ratio (Exp(B)), and its 95% confidence interval (95% CI)

Discussion

Embryonic chromosomal status is a key determinant of embryo viability. Chromosomal abnormalities are a major cause of missed abortion, accounting for 50–70% of early pregnancy losses [13, 14]. Variations in these rates may result from differences in gestational age, detection techniques, and other variables. Chromosomal analysis of missed abortion specimens holds significant etiologic value.

Traditional G-banding karyotype analysis has notable limitations: it can only detect certain types of chromosomal abnormalities and is constrained by prolonged cell culture requirements, high failure rates, susceptibility to maternal tissue contamination, and low resolution, which restrict its clinical utility.

Recent advances in molecular genetics have addressed these limitations. Low-depth, high-throughput whole-genome copy number variation sequencing (CNV-seq) can detect structural chromosomal abnormalities and microdeletions/duplications smaller than 100 kb, with advantages of short turnaround time, lower cost, and operational simplicity. It has been widely applied in the genetic testing of miscarriage tissues [15].

This study retrospectively collected cases of missed abortion embryos analyzed by CNV-seq to investigate correlations between clinical parameters and embryonic chromosomal abnormalities.

Distribution of chromosomal abnormalities in EMA embryos

In this study, CNV-seq was performed on 990 chorionic villus samples from EMA. The overall detection rate of chromosomal abnormalities was 58.59%, which is consistent with previous reports [13, 14].

Autosomal Chromosomal Aneuploidy was the predominant type of abnormality (68.10%), often resulting from errors during meiosis I in oocytes [16, 17]. Among these, autosomal trisomy was the most common (67.24%), likely due to gene dosage imbalances that affect cell proliferation, differentiation, and organogenesis, ultimately leading to embryonic arrest. For example, the CEBPA gene on chromosome 16 is involved in cell cycle regulation, and the COMT gene on chromosome 22 is closely associated with neurodevelopment. Imbalances in these genes are likely to contribute significantly to the observed phenotypes [18].

Sex chromosome aneuploidy accounted for 7.76% of abnormalities, mainly 45,X (Turner syndrome), which is associated with nondisjunction or lagging of the X chromosome during meiosis. Current evidence indicates distinct behavioral patterns of the X chromosome during oocyte meiosis, wherein differential synapsis and recombination dynamics compared to autosomes may elevate the risk of nondisjunction [19].

Triploidy/polyploidy (6.90%) was usually caused by dispermic fertilization or failure in the second meiotic division. These embryos often fail to survive early development due to severe genomic imbalance [20].

CNVs were detected in 8.62% of cases, suggesting that submicroscopic structural abnormalities can impair developmental gene expression, leading to embryonic arrest [21].

Notably, no trisomy 1 cases were detected, while only one case of trisomy 19 was identified. Given that chromosome 1 carries the highest number of genes, and chromosome 19 exhibits the greatest gene density in the human genome, complete trisomies of these chromosomes are likely to induce critical gene dosage effects, leading to preimplantation lethality and consequent underdetection. This observation underscores the differential impact of gene load (Chr1) versus density (Chr19) on embryonic viability [21].

Maternal age and chromosomal abnormalities

By the 12th week of gestation, oocytes in fetal ovaries have arrested at the diplotene stage of prophase I, a state that can last for decades until ovulation. Upon resumption of meiosis, errors can lead to aneuploidy [22, 23].

A single-center study [24] based on 7,118 abortus specimens from pregnant women aged 23 to 44 showed that the overall detection rate of chromosomal abnormalities gradually increased with advancing maternal age, with an average increase of 0.704% per year from ages 23 to 37, and 2.095% per year from ages 38 to 44. Starting at age 38, the detection rate rose sharply, increasing from 79.01% to 94% by age 44, exhibiting a pronounced maternal age-dependent pattern.

In this study, the detection rate in EMA embryos was 66.99% for women aged 35–39 years and reached 73.97% for those aged ≥ 40 years. Trisomies of chromosomes 15, 16, 21, and 22 were particularly common.

These chromosomes exhibit a pronounced “error-prone propensity” during meiosis, with cytological features including shorter inter-centromeric distance, higher crossover suppression rates, and premature degradation of cohesin proteins [25, 26]. For instance, chromosome 16 has an inter-centromeric distance of only about 1.2 μm, which may impair proper microtubule attachment and lead to mis-segregation [27]. Chromosome 21 shows a crossover suppression rate of 43%, reducing genetic exchange between homologous chromosomes and compromising segregation fidelity [27]. These structural and behavioral characteristics make trisomy of these chromosomes more likely in oocytes of advanced maternal age.

Age-related differences in abnormality types were also observed: younger women showed a higher incidence of X monosomy, polyploidy, and CNVs, whereas older women more frequently had double/multiple trisomies and specific autosomal trisomies [24]. These differences suggest distinct underlying molecular and cellular mechanisms [28].

Aneuploidy can arise through multiple mechanisms, including meiotic I nondisjunction (MI NDJ), precocious separation of sister chromatids (PSSC), and reverse segregation (RS) [29]. The type of meiotic error is closely associated with chromosome size and structure. For instance, aneuploidy involving large metacentric and submetacentric chromosomes, which is more frequent in younger women, often results from MI NDJ [26, 30]. In contrast, aneuploidy of telocentric chromosomes, which is more prevalent in advanced maternal age, is frequently attributable to PSSC and recombination suppression [25, 31].

Monosomy X was also predominantly distributed in the younger group (incidence: 10.62% vs. 1.43% vs. 1.85%). In younger women, monosomy X primarily results from loss of the paternal X chromosome (accounting for 68%), which may be associated with the accumulation of dynamic mutations during spermatogenesis. In contrast, monosomy X in older women is mainly attributed to maternal nondisjunction (accounting for 72%), reflecting an increased error rate in oocytes during meiosis I [32].

Interestingly, CNVs and polyploidy rates decreased with age, possibly due to enhanced selection against severe chromosomal imbalances in embryos from older women, representing a potential age-related protective effect [33, 34].

Ovarian reserve and chromosomal abnormalities

Ovarian reserve, reflecting both the quantity and quality of oocytes, has garnered increasing attention for its association with embryonic chromosomal abnormalities.

Advanced maternal age is a well-established independent risk factor for embryonic chromosomal anomalies and miscarriage, and it often coincides with diminished ovarian reserve. However, whether ovarian reserve itself is independently associated with embryonic chromosomal abnormalities, apart from age, remains unclear.

This study found that lower AMH levels were independently associated with an increased rate of overall chromosomal abnormalities and autosomal trisomies, while the incidence of CNVs decreased. Logistic regression confirmed that both age and AMH were independent predictors, suggesting that ovarian reserve may serve as an important biological indicator of embryonic genetic quality beyond age alone.

From the perspective of karyotype categories, the rates of autosomal trisomy and double trisomy significantly increased with declining AMH, while the incidences of 45,X, CNVs, and triploidy decreased. The distribution pattern highly resembles the effect of advanced age on the spectrum of chromosomal aneuploidy, further supporting the role of AMH as a sensitive indicator of reproductive aging.

Previous studies have shown that AMH correlates with ovarian responsiveness and can identify DOR and POI [7]. Both age and AMH are independent risk factors for miscarriage [33], and low AMH levels are associated with reduced clinical pregnancy rates, lower live birth rates, and increased miscarriage rates in younger women [35, 36].

Our study further confirms at the embryonic chromosomal level that lower AMH is associated with a higher aneuploidy risk. This finding supports the hypothesis that diminished ovarian reserve may impair the accuracy of chromosome segregation through mechanisms related to oocyte aging, thereby compromising embryonic developmental potential [37, 38]. Potential underlying mechanisms may include: (1) reduced chiasma formation, increasing the risk of chromosome non-disjunction and recombination failure [39]; (2) decreased expression of the cohesin complex subunit REC8, impairing centromeric cohesion [40]; (3) dysfunctional phosphorylation of the spindle assembly checkpoint (SAC) key factor BUBR1, allowing mis-attached chromosomes to escape surveillance [41, 42]; (4) aberrant post-translational modifications, such as dysregulated histone H3K9me3 methylation and defective tubulin detyrosination [39]; and (5) insufficient mitochondrial ATP production, affecting spindle microtubule dynamics. Together, these mechanisms contribute to chromosome mis-segregation or anaphase lag [37, 39].

Notably, age-stratified analysis revealed that the impact of decreased AMH on chromosomal abnormalities was more pronounced in women aged < 35 years, while no significant difference was observed in women of advanced age (≥ 35 years). This suggests that in younger women, before overt physiological ovarian aging manifests, AMH may better represent oocyte quality and the follicular microenvironment; whereas in older women, age-related degenerative mechanisms—such as mitochondrial dysfunction and the accumulation of epigenetic errors [37, 39]—may dominate, potentially masking the independent effect of AMH.

It is important to note that the relationship between low AMH and chromosomal abnormalities may not be causal; both are likely regulated by the underlying process of oocyte aging. As indicated by previous studies, the decline in ovarian reserve is essentially a consequence of accumulated molecular damage—such as oxidative stress [43] and telomere shortening [44, 45]—resulting from long-term meiotic arrest at the dictyate stage. AMH can thus be regarded as a phenotypic marker of this process.

Therefore, the clinical assessment of embryonic chromosomal abnormality risk should incorporate both age and AMH levels to better inform reproductive decision-making for women of advanced age and those with diminished ovarian reserve.

This study has several limitations. First, as a single-center retrospective analysis, it is constrained by a limited sample size and an uneven distribution of participants, particularly within the high AMH and younger age groups, which may affect statistical power. Second, ovarian reserve was evaluated solely based on AMH levels without incorporating other established indicators such as antral follicle count (AFC), basal follicle-stimulating hormone (FSH), or inhibin B. Furthermore, potential confounding factors related to male partners—including paternal age and sperm genetic integrity—were not considered. Future multi-center, large-sample prospective studies are warranted, integrating multi-omics approaches to further elucidate the mechanistic links between diminished ovarian reserve and meiotic regulation at the oocyte and microenvironment levels.

Conclusion

Age-related oocyte aging increases the risk of chromosomal abnormalities by interfering with meiotic processes. As a key indicator of ovarian reserve, reduced AMH levels are significantly associated with an elevated risk of embryonic chromosomal abnormalities, demonstrating particularly high predictive value in younger patients. These findings suggest that AMH may serve not only as a marker of ovarian function but also as a potential early biomarker for embryonic chromosomal abnormalities. A deeper understanding of the relationship and underlying mechanisms linking diminished ovarian reserve and embryo chromosomal anomalies is of great clinical and scientific importance, as it may facilitate the identification of high-risk populations and improve reproductive genetic counseling.

Abbreviations

AFC

Antral follicle count

AMH

Anti-Müllerian hormone

CNVs

Copy number variant

CNV-seq

Copy number variation sequencing

CRL

Crown-rump length

DOR

Diminished ovarian reserve

EMA

Early missed abortion

FISH

Fluorescence in situ hybridization

FSH

Follicle-Stimulating Hormone

MA

Missed abortion

MI NDJ

Meiotic I nondisjunction

PGS

Preimplantation Genetic Screening

PGT

Preimplantation genetic testing

PSSC

Precocious separation of sister chromatids

POI

Premature Ovarian Insufficiency

VUS

Variants of uncertain significance

Authors’ contributions

Conception/design: S.H.Huang; Provision of study material or patients: Y.Y.Zhou, B. T. Zeng; Collection and/or assembly of data: D.P. Liu, T.T.Huang, H.Z.Yuan; Data analysis and interpretation: S.H.Huang, Y.Y.Zhou, B. T. Zeng; Manuscript writing: S.H. Huang; Manuscript revision: G. Q. Bai; Final approval of manuscript: All authors.

Funding

This work was supported by grants from the Science and Technology Support Plan of the Health Commission of Jiangxi Province (No.20210887) and the Jiangxi Provincial Administration of Traditional Chinese Medicine Science and Technology Project (No. 2024B0048). The funder, S.H. Huang, had a role in the study.

This study was supported by the Jiangxi Provincial Key Laboratory of Birth Defect for Prevention and Control (Grant No. 20242BCC32086).

This study was supported by Nanchang Key Laboratory of Genetics and Rare Diseases Genetic Testing.

Data availability

All data that support the findings of this study were available from the corresponding author via E-mail due to appropriate request.

Declarations

Ethics approval and consent to participate

Due to the retrospective nature of the study, informed consent was waived, but this study was granted by the Medical Ethics Committee of Jiangxi Maternal and Child Health Hospital and the ethics approval number was EC-KY-202011.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.