Ovarian endometriosis negatively impacts pregnancy outcomes in young infertile women undergoing IVF/ICSI treatment

This retrospective cohort study involved 3256 women who initiated their first IVF or ICSI cycles at the reproductive medicine center of Peking University First Hospital between January 2016 and December 2021. The inclusion criteria specified: (i) women aged 42 years or younger and (ii) no prior IVF cycles. The research enrolled a total of 300 women who had been diagnosed with OE, identified by the existence of the cysts in ovaries at the beginning of the stimulated cycles. The diagnosis of endometriosis was validated through imaging techniques, such as ultrasound and magnetic resonance imaging (MRI) ( n  = 148), or surgical procedures accompanied by histological confirmation ( n  = 152), as previously outlined [ 3 ]. The cyst size was measured between the opposite inner cyst walls in three orthogonal planes according to terms and definitions described by the International Ovarian Tumor Analysis group. All cyst were described regarding the largest diameter observed via ultrasound; in cases where bilateral cysts were found, the analysis utilized the aggregate of the largest diameters when applicable [ 18 , 19 ]. For comparative analysis, the control cohort comprised 2956 infertile women who did not have OE. The following exclusions were implemented in this study: the use of donor oocytes; adenomyosis; uterine malformation; the presence of hydrosalpinx; fibroids measuring 40 mm or larger; intrauterine adhesions; other autoimmune disorders; and hormonal treatment received within 3 months prior to enrollment in this research, such as the combined oral contraceptives, the progesterone only pill, Mirena IUS, ovarian suppression with gonadotrophin releasing hormone (GnRH) therapy [ 3 ]. To investigate the influence of age on IVF/ICSI outcomes, participants were divided into two distinct age groups: a younger group of women under 35 years and an older group of women aged 35 years and above [ 20 ]. Given that the assignment to these groups was not random, potential confounding variables and selection biases were addressed through the application of propensity score matching (PSM) [ 21 ]. The variables utilized for matching included age, body mass index (BMI), and the duration of infertility. The protocols employed for controlled ovarian stimulation during the study cycles included the GnRH agonist protocol, the GnRH antagonist protocol, and alternative stimulation methods, such as the mild ovarian stimulation protocol and the progestin-primed ovarian stimulation protocol. The ovarian stimulation protocol was customized based on each patient’s individual profile. When over three follicles reached 18 mm or more, human chorionic gonadotropin (hCG) was intramuscularly administered, followed by oocyte retrieval using a single lumen needle after 36 h. Fertilization was conducted using standard methods, with embryo culture lasting 3–6 days. Embryos were assessed morphologically using ASEBIR criteria and Gardner’s classification [ 22 ]. The number of embryos transferred, whether one or two, was determined by factors, such as the maternal age, cycle rank, and embryo quality, regardless of the presence of endometriosis. Fresh transfers occurred on days 2 or 3 post-retrieval, while frozen-thawed embryo transfers followed three primary endometrial preparation protocols: natural cycle and hormone replacement therapy, with or without GnRH downregulation. Luteal support was maintained via vaginal or intramuscular progesterone and oral dydrogesterone for 2 week post-transfer, with detailed methodologies outlined in prior research [ 3 ]. The cumulative live birth rate (cLBR) was defined as the percentage of live births obtained following the transfer of all embryos available from a stimulated cycle, whether those embryos were fresh or frozen-thawed [ 23 ]. The implementation of PSM involves several key steps. First, propensity scores are calculated for each participant based on baseline characteristics, which represents the probability of each participant being assigned to a particular group (OE or control), given the influence of these covariates. Second, a matching caliper of 0.02 is applied, along with a 1:1 nearest neighbor matching approach, to ensure comparable relevant variables between the two groups. Finally, post-matching analyses are conducted to evaluate pregnancy outcomes and assess the balance of baseline characteristics achieved through matching [ 24 ]. For continuous variables, the Mann–Whitney U test was utilized, while categorical data were evaluated using the Chi-square test or Fisher’s exact test. To identify independent variables associated with the cLBR per oocyte pickup cycle, logistic regression analysis was performed. The coefficients and standard deviations derived from the model allowed for the calculation of odds ratios (OR) and 95% confidence intervals (CI). The MatchIt package in R (version 3.6.2) was used to perform PSM. A nearest neighbor matching method at a 1:1 ratio was implemented, using a caliper width of 0.02 to ensure comparability between groups and mitigate bias in the analysis [ 24 , 25 ].

A total of 3256 women were enrolled in this study, including 300 women with OE (age < 35 years: n  = 188, age ≥ 35 years: n  = 112) and 2956 women without OE (age < 35 years: n  = 1875, age ≥ 35 years: n  = 1081). After matching age, BMI and duration of infertility, there were 185 women with OE and 185 women without OE among the infertile women younger than 35 years. While among the infertile women 35 and older, there were 109 women with OE and 109 women without OE. The characteristics and ovarian stimulation details are shown in Table  1 . Table 1 Baseline characteristics and ovarian stimulation parameters of women with and without OE by age category Variable Age < 35 years Age ≥ 35 years OE ( n  = 185) Control ( n  = 185) P value OE ( n  = 109) Control ( n  = 109) P value Baseline characteristics  Age (years) 32 (30–33) 32 (30–33) 0.493 37 (35–38) 37 (35–38) 0.955  BMI (kg/m 2 ) 21.48 (19.80–23.10) 21.30 (19.57–23.44) 0.558 22.31 (20.20–24.89) 22.70 (20.54–24.80) 0.511  Duration of infertility (years) 2 (1–4) 2 (2–4) 0.345 3 (2–5) 3 (1–5) 0.472  Type of infertility 0.806 0.343   Primary n (%) 141 (76.2) 143 (77.3) 57 (52.3) 50 (45.9)   Secondary n (%) 44 (23.8) 42 (22.7) 52 (47.7) 59 (54.1)  AMH (ng/ml) 2.68 (1.41–3.57) 3.50 (2.37–5.19)  < 0.001 1.70 (0.86–3.07) 2.44 (1.42–3.50)  < 0.001  AFC ( n ) 7 (4–12) 13 (8–19)  < 0.001 6 (3–10) 10 (7–16)  < 0.001  Basal FSH (IU/l) 8.32 (6.82–11.13) 7.95 (6.53–11.13) 0.827 8.87 (6.67–11.13) 8.15 (6.54–10.46) 0.169  Basal LH (IU/l) 4.23 (3.00–5.14) 4.79 (3.16–5.83) 0.048 4.33 (2.91–5.33) 3.83 (2.96–5.07) 0.489  Basal E2 (pg/ml) 48.00 (33.50–90.00) 48.00 (34.00–78.50) 0.172 54.00 (36.50–217.00) 47.00 (30.00–81.00) 0.108  Basal P (ng/ml) 0.76 (0.47–1.11) 0.68 (0.40–1.11) 0.513 0.67 (0.43–1.11) 0.58 (0.31–0.97) 0.022 Ovarian stimulation parameters  Stimulation Protocol 0.470 0.080   Agonist n (%) 69 (37.3) 65 (35.1) 36 (33.0) 36 (33.0)   Antagonist n (%) 101 (54.6) 110 (59.5) 63 (57.8) 52 (47.7)   Others n (%) 15 (8.1) 10 (5.4) 10 (9.2) 21 (19.3)  Total dose of Gn administered (IU) 3000 (2325–3675) 2400 (1725–3281)  < 0.001 3000 (2325–3900) 3000 (2325–3863) 0.972  Days of Gn administered 10 (9–12) 10 (9–11) 0.884 10 (8–11) 10 (9–11) 0.050  E2 on day of hCG (pg/ml) 2635 (1496–3888) 2940 (2095–4493) 0.001 1615 (888–2940) 2940 (1614–3640)  < 0.001  P on day of hCG (ng/ml) 1.18 (0.87–1.35) 1.14 (0.77–1.49) 0.902 1.00 (0.58–1.21) 1.12 (0.73–1.46) 0.074  Endometrial thickness on day of hCG (mm) 11 (10–12) 10 (10–12) 0.492 10 (9–11) 10 (9–12) 0.578  Number of follicles ≥ 14 mm on day of hCG ( n ) 6 (3–11) 10 (6–15) 0.001 5 (2–8) 7 (3–13)   0.999 1.8 (2/109) 0.0 (0/109) 0.498 Data are shown as n (%), percentage (%) or median (interquartile range) unless otherwise stated OE ovarian endometriosis, BMI body mass index, AMH anti–Müllerian hormone, AFC antral follicular count, FSH follicle–stimulating hormone, LH luteinizing hormone, E2 estradiol, P progesterone, Gn gonadotropin, hCG human chorionic gonadotropin Baseline characteristics and ovarian stimulation parameters of women with and without OE by age category Data are shown as n (%), percentage (%) or median (interquartile range) unless otherwise stated OE ovarian endometriosis, BMI body mass index, AMH anti–Müllerian hormone, AFC antral follicular count, FSH follicle–stimulating hormone, LH luteinizing hormone, E2 estradiol, P progesterone, Gn gonadotropin, hCG human chorionic gonadotropin In both age categories—those younger than 35 and those 35 and older—no significant differences in age, BMI, or infertility duration existed between the groups ( P  > 0.05), supporting the validity of the matching. In the group of age < 35 years, women with OE presented significantly lower AMH levels and AFC compared to those without OE (2.68 ng/ml vs. 3.50 ng/ml, P  < 0.001; 7 vs. 13, P  < 0.001). Similarly, in the group of age ≥ 35 years, women with OE had significantly lower AMH levels and AFC compared to those without OE (1.70 ng/ml vs. 2.44 ng/ml, P  < 0.001; 6 vs. 10, P  < 0.001). Above results indicate a diminished ovarian reserve in OE. On hCG administration day, estradiol (E2) levels and the number of follicles ≥ 14 mm were also lower in women with OE, regardless of whether the age was below 35 years (2635 pg/ml vs. 2940 pg/ml, P  = 0.001; 6 vs. 10, P  = 0.001) or 35 years and above (1615 pg/ml vs. 2940 pg/ml, P  < 0.001; 5 vs. 7, P  < 0.001). Table 2 summarizes the embryological and clinical outcomes. Ovarian sensitivity was calculated as the ovarian sensitivity index (OSI), defined as number of oocytes retrieved divided by the total dose of Gn administered * 1000 [ 26 ]. In the group of age < 35 years, women with OE had fewer retrieved oocytes and a lower OSI (8 vs. 11, P  < 0.001; 2.37 vs. 4.44, P  < 0.001). Similarly, in the group of age ≥ 35 years, women with OE presented significantly lower retrieved oocytes and OSI compared to those without OE (5 vs. 9, P  < 0.001; 1.82 vs. 2.86, P  < 0.001). Above results suggested a reduced response to stimulation in OE. Metrics such as oocyte maturity rates (age < 35 years: 78.7% vs. 83.7%, P  < 0.001; age ≥ 35 years: 77.6% vs. 84.4%, P  = 0.001), fertilization rates (age < 35 years: 77.4% vs. 81.9%, P  = 0.001; age ≥ 35 years: 68.4% vs. 77.2%, P  < 0.001), quantities of top-quality embryos (age < 35 years: 2 vs. 2, P  < 0.001; age ≥ 35 years: 1 vs. 2, P  < 0.001) were also significantly lower in women with OE. Table 2 Embryological data and IVF/ICSI outcomes of women with and without OE by age category Variable Age < 35 years Age ≥ 35 years OE ( n  = 185) Control ( n  = 185) P value OE ( n  = 109) Control ( n  = 109) P value Embryological data  Number of oocytes retrieved ( n ) 8 (4–13) 11 (7–15)  < 0.001 5 (2–8) 9 (6–14)  < 0.001  OSI 2.37 (1.28–4.57) 4.44 (2.15–7.45)  < 0.001 1.82 (0.99–2.84) 2.86 (1.54–4.50)  < 0.001  Oocyte maturity rate % ( n ) 78.7 (1290/1640) 83.7 (1798/2147)  < 0.001 77.6 (489/630) 84.4 (885/1049) 0.001  Mode of fertilization 0.665 0.294   IVF 72.6 (134/183) 57.9 (131/184) 69.2 (74/107) 62.4 (68/109)   ICSI 27.4 (49/183) 42.1 (53/184) 30.8 (33/107) 37.6 (41/109)  Fertilization rate % ( n ) 77.4 (1118/1444) 81.9 (1622/1981) 0.001 68.4 (432/630) 77.2 (810/1049)  < 0.001  Blastocyst rate % ( n ) 54.9 (330/601) 55.6 (533/958) 0.778 40.1 (81/202) 51.7 (230/445) 0.006  Number of embryos ( n ) 5 (2–9) 8 (4–12)  < 0.001 3 (1–6) 6 (4–11)  < 0.001  Number of transplantable embryos ( n ) 5 (2–9) 7 (4–11)  < 0.001 3 (1–6) 6 (3–10)  < 0.001  Number of top–quality embryos ( n ) 2 (0–4) 2 (1–5)  < 0.001 1 (0–2) 2 (0–5)  < 0.001  Top–quality embryo rate (%) 41.0 (424/1034) 43.1 (607/1408) 0.298 42.9 (166/387) 47.9 (309/645) 0.118  Number of ET cycle 257 269 0.128 123 161 0.342   Fresh ET n (%) 64 (24.9) 83 (30.9) 44 (36.0) 49 (33.1)   Frozen ET n (%) 193 (75.1) 186 (69.1) 79 (64.0) 112 (66.9) Clinical outcomes  cCPR % ( n ) 62.8 (113/180) 75.4 (135/179) 0.010 57.5 (61/106) 65.4 (71/108) 0.218  cLBR % ( n ) 56.1 (101/180) 67.6 (121/179) 0.025 45.4 (49/106) 57.4 (62/108) 0.102 Data are shown as n (%), percentage (%) or median (interquartile range) unless otherwise stated OE ovarian endometriosis, OSI Ovarian sensitivity index, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection, ET embryo transfer, cCPR cumulative clinical pregnancy rate, cLBR cumulative live birth rate Embryological data and IVF/ICSI outcomes of women with and without OE by age category Data are shown as n (%), percentage (%) or median (interquartile range) unless otherwise stated OE ovarian endometriosis, OSI Ovarian sensitivity index, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection, ET embryo transfer, cCPR cumulative clinical pregnancy rate, cLBR cumulative live birth rate Younger women with OE exhibited lower cumulative clinical pregnancy rates (cCPR) and cLBR than those in women without OE (62.8% vs. 75.4%, P  = 0.010; 56.1% vs. 67.6%, P  = 0.025), while in the older group, trends did not reach significance (57.5% vs. 65.4%, P  = 0.218; 45.4% vs. 57.4%, P  = 0.102). Univariate and multivariate logistic regression analyses are shown in Table  3 . The univariate logistic regression analysis was used to calculate the unadjusted effect of each candidate risk factor on cLBR. The other confounding factors were adjusted by multivariate logistic regression analysis. The confounding factors included age, cyst size ≥ 40 mm, history of surgery for endometriosis, AMH, AFC, total dose of Gn administered, E2 on day of hCG, number of follicles ≥ 14 mm on day of hCG, number of oocytes retrieved, and number of top-quality embryos. The results revealed that after adjusting above confounding factors, age, AMH and the number of top-quality embryos significantly correlated with cLBR among younger women. The cyst size was correlated with cLBR but not statistically significant. In older women, only age and the number of top-quality embryos were associated with cLBR after adjusting above confounding factors. Table 3 Logistic regression of cumulative live birth in OE group Variable Univariate analysis Multivariate analysis OR (95%CI) P value OR (95%CI) P value Age < 35 years  Age 0.76 (0.65–0.89)  < 0.001 0.79 (0.67–0.93) 0.006  Cyst size ≥ 40 mm 1.72 (0.78–3.81) 0.179 2.15 (0.89–5.24) 0.094  History of surgery for endometriosis 0.55 (0.30–0.99) 0.047  AMH 1.48 (1.21–1.81)  < 0.001 1.35 (1.08–1.70) 0.009  AFC 1.06 (1.00–1.12) 0.041  Total dose of Gn administered 1.00 (1.00–1.00) 0.679  E2 on day of hCG 1.00 (1.00–1.01)  < 0.001  Number of follicles ≥ 14 mm on day of hCG 1.13 (1.06–1.20)  < 0.001  Number of oocytes retrieved 1.11 (1.05–1.18)  < 0.001  Number of top–quality embryos 1.45 (1.22–1.72)  < 0.001 1.31 (1.09–1.57) 0.004 Age ≥ 35 years  Age 0.72 (0.56–0.91) 0.007 0.73 (0.56–0.94) 0.015  Cyst size ≥ 40 mm 0.78 (0.27–2.24) 0.649  History of surgery for endometriosis 0.92 (0.43–1.97) 0.822  AMH 1.14 (0.90–1.45) 0.282  AFC 1.01 (0.94–1.09) 0.750  Total dose of Gn administered 1.00 (1.00–1.00) 0.409  E2 on day of hCG 1.00 (1.00–1.01) 0.003  Number of follicles ≥ 14 mm on day of hCG 1.26 (1.13–1.41)  < 0.001  Number of oocytes retrieved 1.18 (1.07–1.32) 0.002  Number of top–quality embryos 1.53 (1.15–2.04) 0.004 1.50 (1.13–2.01) 0.006 OE ovarian endometriosis, AMH anti-Müllerian hormone, AFC antral follicular count, Gn gonadotropin, E2 estradiol, hCG human chorionic gonadotropin Logistic regression of cumulative live birth in OE group OE ovarian endometriosis, AMH anti-Müllerian hormone, AFC antral follicular count, Gn gonadotropin, E2 estradiol, hCG human chorionic gonadotropin Next, the correlation of AMH with cLBR in younger women with OE was explored using ROC analysis, yielding an area under the curve of 0.679 and an optimal AMH threshold of 1.835 ng/ml with a sensitivity of 50.6% and specificity of 82.2% (Fig.  1 ). The AMH threshold can be expected to be used for identifying the population that can achieve live birth and making appropriate clinical decisions to improve the likelihood of live birth for those below the threshold. Fig. 1 ROC analysis showed the correlation of AMH with cLBR in younger women with OE ROC analysis showed the correlation of AMH with cLBR in younger women with OE Furthermore, among younger women with OE, Spearman's correlation revealed that AMH did not correlate with age ( R  = − 0.129, P  = 0.085, Fig.  2 A), but it negatively correlated with cyst size ( R  = − 0.240, P  = 0.001, Fig.  3 A). In older women, AMH negatively correlated with age ( R  = − 0.253, P  = 0.009, Fig.  2 B), but it did not correlate with cyst size ( R  = − 0.073, P  = 0.459, Fig.  3 B). These findings indicate that cyst size may affect AMH, and then affect the cLBR among women with OE younger than 35 years; while among women with OE aged ≥ 35 years, age may mainly affect the cLBR more than the cyst. Fig. 2 Spearman’s correlation revealed the relationship between age and AMH levels in A women aged < 35 years and B women aged ≥ 35 years Fig. 3 Spearman’s correlation revealed the relationship between cyst size and AMH levels in A women aged < 35 years and B women aged ≥ 35 years Spearman’s correlation revealed the relationship between age and AMH levels in A women aged < 35 years and B women aged ≥ 35 years Spearman’s correlation revealed the relationship between cyst size and AMH levels in A women aged < 35 years and B women aged ≥ 35 years

Endometriosis is defined by the existence of endometrial tissue located outside the uterine cavity, affecting an estimated 5–10% of women during their reproductive years [ 1 ]. This condition is driven by dysregulated hormonal signaling, particularly involving estrogen and progesterone, which promotes the abnormal growth and persistence of ectopic endometrial tissue. Endometriotic tissue generate a localized immune and inflammatory response that contribute to the hallmark symptoms of chronic pelvic pain and infertility [ 1 – 3 ]. Research indicates that approximately 50% of women diagnosed with endometriosis experience infertility [ 4 ]. Despite the high prevalence of infertility in endometriosis patients, the exact pathogenesis of endometriosis in relation to infertility remains unclear. The condition leads to a series of changes, including alterations in pelvic anatomy and microenvironment, changes in hormone levels, and impaired embryo implantation. These changes can impact follicular development as well as endometrial receptivity. As a result, they can adversely affect the growth and maturation of oocytes, fertilization, and embryo implantation, along with other associated processes. Ultimately, such changes can significantly influence the outcomes of pregnancy. The treatment approaches for infertility related to endometriosis typically involve surgical intervention or assisted reproductive technologies (ART). Research indicates that surgical treatment can enhance spontaneous conception rates within 12–18 month post-operation. At the same time, the research pointed out that benefits may depend on factors, such as disease stage and patient age. Among patients with endometriosis who underwent surgical treatment, the lower the severity of the disease stage, the higher the conception rate. The younger (under 35 years) the patient, the higher the conception rate [ 5 ]. In some cases, surgical procedures may lead to reduced ovarian reserve [ 6 , 7 ]. Moreover, age-related decrease in ovarian reserve also needs to be taken into account before the decision to perform surgery [ 8 , 9 ]. Therefore, for women with compromised ovarian function or those over the age of 35, ART is a preferred choice [ 9 , 10 ]. The influence of ovarian endometriosis (OE) on pregnancy outcomes in ART has been the subject of numerous studies, yielding inconsistent findings regarding its effects on live birth rate [ 11 – 13 ]. Recent systematic reviews indicate that the overall pregnancy outcomes following in vitro fertilization (IVF) or intracytoplasmic sperm injection (ICSI) do not exhibit significant differences [ 14 , 15 ]. A retrospective cohort study included 195 women with moderate-to-severe endometriosis and 390 women with other causes of infertility. A significant difference in cumulative live birth rate was not found [ 11 ]. Conversely, a retrospective study included 436 patients with a history of endometriosis confirmed by surgical procedures and 1299 infertile patients with tubal factors, and found a decreased cumulative live birth rate in the endometriosis group [ 13 ]. Possible explanations for these conflicting results are differences in study designs, definitions of endometriosis severity and the control group. Moreover, maternal age remains a critical determinant of pregnancy success rates. A recent prospective randomized controlled trial found that women aged 35 years and older with American Society for Reproductive Medicine (ASRM) stage III–IV endometriosis had a significantly lower live birth rate compared to those with infertility due to tubal factors. Furthermore, differences in live birth rates between the two groups were further extended through stepwise multivariable logistic regression analysis, which was adjusted for age, baseline anti-Müllerian hormone (AMH) levels, antral follicle count (AFC), the number of mature oocytes, and embryos [ 16 ]. In contrast, another retrospective study noted that women within the 36–40 age range diagnosed with endometriosis experienced higher rates of clinical pregnancies and live births [ 17 ]. These conflicting results underscore the complexity of OE's impact on reproductive outcomes. Not only endometriosis itself may affect the live birth rate, but age may also affect the live birth rate as a confounding factor. Specifically, the interaction between OE and age has not been systematically evaluated. The further investigation to clarify the influence of age on the outcomes of IVF and ICSI in women with or without OE is needed. Understanding these potential associations could lead to more tailored and effective treatment strategies for women with OE experiencing infertility, ultimately enhancing their chances of achieving a successful pregnancy. Consequently, this study aimed to explore how maternal age interacts with OE to influence cumulative live birth rates in women undergoing IVF/ICSI treatment.

Currently, the findings regarding the effects of endometriosis on pregnancy outcomes remain inconclusive. Notably, many previous studies have overlooked age as a significant confounding variable that may influence these outcomes. Sharma et al. [ 16 ] found that before adjusting for age and other confounders, LBR was comparable in young women presented with endometriosis and tubal infertility, but lower in endometriosis women after adjusting for age and other confounders. Thus, this research aimed to assess the influence of age on pregnancy results in women with and without OE. Prior investigations into endometriosis-related infertility by Feichtinger et al. included women under 40 years of age [ 27 ], while Invernici et al. [ 28 ] and Scarafia et al. [ 29 ] focused on those aged 43 and younger. In addition, researchers such as Leonardi et al.[ 30 ], Yang et al.[ 30 ], and Benaglia et al. [ 31 ] included women aged 42 and below. Studies have shown that when women are 35 years and above, the number of follicles decreases, the incidence of oocyte aneuploidy increases, and the rate of natural pregnancy decreases [ 32 ]. Franasiak et al. [ 33 ] found that the rate of embryo aneuploidy increases significantly when the age is 35 years and above. Therefore, this study encompassed infertile women aged 42 or younger, choosing 35 years as the threshold to categorize them into two age groups: a younger cohort of those under 35 years and an older cohort of those aged 35 and above. Previous multicenter retrospective study [ 34 ] found that in infertile women, the cLBR decreased progressively with increasing age and BMI. ElMokhallalati et al. [ 35 ] explored the relationship between the duration of infertility and LBR, concluding that a longer duration of infertility corresponds to a lower LBR in IVF/ICSI cycles. Therefore, in this study, we matched participants by age, BMI, and duration of infertility prior to performing a comparative analysis between the OE group and the control group for both younger and older women. Findings indicated that irrespective of age, women with OE consistently exhibited diminished ovarian reserves and impaired responses to stimulation. Among the younger group, both the cCPR and cLBR in women with OE were significantly lower than those in women without OE. In the older cohort, while cCPR and cLBR also trended lower in women with OE, the differences did not reach statistical significance. Current evidence indicates that endometriosis impairs ovarian function [ 36 ]. In this study, women with OE presented lower AMH levels, AFC and OSI compared to those without OE, not only in the group of age < 35 years, but also in the group of age ≥ 35 years. This study showed lower ovarian reserve and response to stimulation in women with OE, which aligns with previous findings [ 37 , 38 ]. Conversely, Marcellin et al. [ 18 ] found that the AMH levels significantly increased with the volume of cysts in cases of ovarian endometriotic cysts. They suggested that larger cysts may stimulate the ovaries to produce and release more AMH into the circulation. However, a recent study by Karadağ et al. [ 39 ] demonstrated a negative correlation between serum AMH levels and cyst size among infertile women diagnosed with endometriosis. The potential reasons for these inconsistent findings may be the heterogeneity of population. Our results confirm that AMH was inversely related to cyst size. AMH was negatively correlated with cyst size among women with OE younger than 35 years, while among women with OE aged ≥ 35 years, AMH was not correlated with cyst size, but was negatively correlated with age, suggesting age has a greater impact on AMH when the age is over 35 years. Therefore, when OE women has compromised ovarian function or those are 35 years and above, the surgical considerations prior to ART should carefully evaluate the potential benefits against the risks of surgery. ART is often the preferred choice, and developing tailored clinical strategies to optimize pregnancy outcomes requires further exploration. Although women with OE in the younger cohort received higher doses of Gn, the total number of oocytes retrieved was significantly lower. Nicolaus et al. [ 40 ] similarly observed reduced oocyte retrieval in women suffering from endometriosis. They believed that endometriosis decreased the response to stimulation, thereby reducing the retrieved oocytes. Meanwhile, changes in the anatomy caused by ovarian endometriotic cysts may make oocyte retrieval more difficult. The gonadotrophins administered for ovarian stimulation could reach the ovarian parenchyma in a less efficient way because of the presence of anatomical distortions [ 41 ]. Ovarian fibrosis could also alter the local vascular supply [ 28 ]. Yan et al. [ 42 ] investigated maternal age’s impact on IVF outcomes and found that older maternal age was linked to higher Gn doses and fewer retrieved oocytes. Consistent with the above research results, this study identified that while Gn doses were elevated in the younger group, the number of oocytes retrieved was still significantly lower in women with OE compared to those without, suggesting that OE impairs ovarian response to stimulation. The influence of endometriosis on oocyte quality remains controversial. Women with OE exhibit lower ovarian reserve and yield fewer retrieved oocytes compared to those without this condition, which partially supports the hypothesis that the endometriotic cyst per se can have an adverse effect on the ovary. Furthermore, a study by Kitajima et al. [ 43 ] revealed that the inflammatory environment within the ovaries of women with endometriosis is likely associated with the cysts, which in turn cause reduced quality of primordial follicles, premature follicle recruitment, and atresia. In addition, excessive presence of reactive oxygen species (ROS) has been observed in the follicular fluid of infertile women with OE. It has been shown that excessive ROS can lead to senescence of granulosa cells, significantly hindering both oocyte retrieval numbers and the maturation process of the oocytes in affected women [ 44 ]. In this study, it was also found that the oocyte maturation rate was reduced in the OE group. Collectively, these results indicate that endometriosis may adversely affect the oocyte microenvironment, ultimately compromising oocyte quality. Recent studies have found that antioxidant treatments can improve oocyte quality in vivo [ 45 ], which has clinical significance for future exploration of antioxidant therapy to improve live birth rates. Moreover, we observed significantly fewer top-quality embryos in the OE cohort, with logistic regression indicating that the number of top-quality embryos serves as a predictor for cLBR. Huang et al. [ 46 ] reached the same conclusion. Previous studies found that the excessive ROS in the intraovarian inflammatory environment affected by cysts can significantly lower the number of oocytes retrieved, and the quality of oocytes and embryos in women with endometriosis [ 43 , 44 ]. While a multivariate analysis found no direct link between endometriosis and embryo quality, it suggested that reduced embryo numbers, rather than quality impairment, primarily account for lower cLBR in affected women [ 47 ]. In addition, previous studies concur that adequate quantities of high-quality embryos result in clinical pregnancy rates comparable to those without endometriosis. Ultimately, enhancing the number of viable embryos may improve reproductive outcomes for women facing this condition. The impact of OE on the outcome of IVF/ICSI remains controversial. Polat et al. [ 48 ] and González-Comadran et al. [ 49 ] found that endometriosis had no impact on LBR. Similarly, Feichtinger et al. [ 27 ] found comparable cCPR and cLBR. In contrast, Zhou et al. [ 13 ] reported a markedly decreased cLBR in women with endometriosis compared to those with tubal factors. However, age stratification was not adequately applied. Numerous studies indicate that advancing age adversely affects ovarian reserve and is an independent variable influencing both clinical pregnancy and live birth rates [ 32 , 33 ]. Previous study has shown [ 32 ] that when the age is ≥ 35 years, the ovarian reserve decreases significantly, the incidence of oocyte aneuploidy increases significantly, and the decline of natural pregnancy rate becomes faster. Franasiak et al. [ 33 ] reported that the aneuploidy rate of embryos was increased when the age was 35 years and above. Therefore, it is common to choose 35 years as threshold for research. Some previous retrospective cohort studies matched the age factor, and found that age can be a confounding factor to affect the overall effect of endometriosis and non-endometriosis on pregnancy outcomes in the primary analysis. Sharma et al. [ 16 ] conducted an evaluation of severe endometriosis and tubal factor infertility concerning pregnancy outcomes while also stratifying participants by age. Their results revealed that for women aged 35 years and older, the live birth rate among those with severe endometriosis was significantly lower than in the control group with age adjusted. Another study found that patients with endometriosis exhibited higher CPRs and LBRs among women aged 36 to 40 years [ 17 ]. This study observed that pregnancy outcomes for women with and without OE in the older age group were similar. However, in the younger age group, women with OE exhibited significantly lower cCPRs and cLBRs compared to those without the condition. Therefore, it is speculated that OE per se could negatively affect pregnancy outcomes, and when the age is over 35 years, the influence of age as a confounder on pregnancy outcome could be greater than the influence of OE per se. A key strength of this study is the application of PSM, which reduces the impact of confounding variables on the outcomes. By approximating random assignment similar to randomized controlled trials, PSM enhances comparability among subjects and improves upon traditional regression methods in observational studies. This methodological approach strengthens the integrity and validity of the findings, allowing for clearer insights into causal relationships and contributing to a more comprehensive understanding of the research question [ 21 ]. In this study, we only analyzed outcomes after the first IVF/ICSI cycle. Most women that fail to get pregnant on the first attempt will probably go through further attempts, that may result in different outcomes. Furthermore, age was considered to affect the overall effect of endometriosis and non-endometriosis on pregnancy outcomes in this study. Retrospective observational design is the limitation of this study. While PSM was used to analyze the effects of endometriosis on IVF/ICSI outcomes, the process led to a reduction in sample size, which may lead to unforeseen effects due to the loss of unmatched cases. The extent to which the findings can be generalized may be limited due to the reduced sample size and its retrospective design. To further clarify the impact of endometriosis itself, it is essential to establish a large randomized clinical cohort that compares pregnancy outcomes of women with and without the condition, thereby providing a clearer understanding of its influence on reproductive success.

Women with OE had lower ovarian reserve than those without OE. Women with OE younger than 35 years had a lower cCPRs and cLBRs than those without OE, and AMH was an independent factor influencing the cLBR. Furthermore, AMH was negatively correlated with cyst size, suggesting that OE negatively impacts pregnancy outcomes in younger women.