Analysis of Risk Factors Affecting Cumulative Live Birth in Couples With Unexplained Infertility Undergoing IVF/ICSI

ABSTRACT

Purpose

To identify the specific risk factors associated with unexplained infertility (UI) among infertile couples.

Methods

We conducted a retrospective cohort study analyzing UI risk factors, focusing on 5465 infertile couples who completed their first full in vitro fertilization/intracytoplasmic sperm injection (IVF/ICSI) cycle out of a total cohort of 62201 infertile couples.

Results

Among 5465 UI couples, 3400 achieved a live birth (cumulative live birth rate = 62.2%). Multivariate logistic regression indicated the key negative predictors included: female age ≥ 35 years (odds ratio [OR]: 0.63; 95% confidence interval [CI]: 0.51–0.78; p < 0.001), male age (OR: 0.98; 95% CI: 0.96–0.99; p = 0.015), Female body mass index ≥ 27.5 kg/m² (OR: 0.78; 95% CI: 0.62–0.97; p = 0.028), low anti‐Müllerian hormone (AMH) (AMH < 1.2 ng/mL) (OR: 0.64; 95% CI: 0.52–0.77; p < 0.001), and a mid‐range endometrial thickness (EMT) (EMT 0.7–1.0 cm) (OR: 0.77; 95% CI: 0.66–0.89; p < 0.001). A higher number of retrieved oocytes positively correlated with cumulative live birth (CLB) (OR: 1.02; 95% CI: 1.01–1.04; p = 0.039).

Conclusion

The prevalence of UI was 8.8% in our population, with 62.2% achieving CLB after one complete IVF/ICSI cycle. Female age and ovarian reserve indicators emerged as the main factors influencing oocyte yield, embryo acquisition, and CLB outcomes in UI patients.

Our study demonstrated a striking prevalence of unexplained infertility (UI) at 8.8% within the study cohort. Notably, patients undergoing a complete IVF/ICSI cycle achieved a remarkable cumulative live birth rate (CLBR) of 62.2%, highlighting the clinical efficacy of assisted reproductive technologies in this population. Determinants of Reproductive Success: Multivariate analysis established female age and ovarian reserve markers as the dominant influence factors on treatment outcomes in UI patients. These factors played a pivotal role across critical reproductive milestones: oocyte retrieval efficiency, embryo development potential, long‐term cumulative live birth success.

1. Introduction

Approximately 15% to 30% of infertile couples are diagnosed with unexplained infertility (UI), which is defined as the failure to achieve a pregnancy after 12 months of regular unprotected intercourse, despite normal findings in standard assessments of ovulation, tubal patency, and semen analysis [1, 2, 3]. As a diagnosis of exclusion, UI presents a clinical challenge, with treatment strategies remaining largely empirical in the absence of a clearly identifiable cause [4, 5, 6]. In vitro fertilization/intracytoplasmic sperm injection (IVF/ICSI) has been widely used to achieve pregnancy in couples with UI [6, 7]. However, reported outcomes for UI patients vary considerably when compared to those with other infertility indications, such as tubal, endometriosis, or male factors, with studies suggesting either superior, equivalent, or inferior success rates [8, 9, 10, 11, 12]. These discrepancies may arise from heterogeneous diagnostic criteria and inconsistent patient selection across studies. Moreover, they highlight the potential influence of undefined factors unique to UI, reflecting its complex etiology.

Advanced male age has been identified as a negative predictor of live birth in IVF/ICSI cycles for idiopathic infertility [13]. While such findings help clarify the role of aging, other UI‐specific factors affecting pregnancy outcomes remain underexplored. Key questions, such as distinguishing UI from age‐related subfertility and evaluating the impact of uterine factors like fibroids or polyps, have been highlighted as research priorities [14]. Furthermore, although cumulative live birth rate (CLBR) is increasingly recognized as a comprehensive measure of treatment success—especially in the context of embryo cryopreservation [15]—robust CLBR data from well‐characterized UI cohorts are still lacking. Nonetheless, there has been limited focus on maternal and neonatal outcomes specifically in UI cases following assisted reproduction.

To address these gaps, we conducted a study focusing exclusively on UI couples undergoing their first complete IVF/ICSI cycle. Using strict inclusion criteria, we selected a well‐defined cohort of 5465 UI couples from a total of 62201 treatment cycles. The study had two primary aims: first, to identify risk factors influencing key treatment stages, including oocyte retrieval, embryo development, and CLBR; and second, to evaluate maternal and neonatal outcomes specifically in this UI population. Our findings aim to provide a clearer clinical profile of UI and improve counseling for affected couples.

2. Methods

2.1. Study Design

This was a retrospective cohort study using data collected at a single center (Hospital for Reproductive Medicine affiliated to Shandong University) from 01/07/2013 to 30/06/2020. To ensure privacy protection, data involving individual information had been desensitized before being included in the research analysis. The study was approved by the Ethics Committee of Reproductive Medicine of Shandong University. The human ethic approval number is [2020] IRB No. (70). The date on which ethical approval was granted was 14/10/2020. Written informed consent was obtained from all couples.

2.2. Patient Characteristics

2.2.1. Inclusion Criteria

Couples were included if they met all of the following criteria:

1. Female partner age between 20 and 40 years.

2. Regular, unprotected sexual intercourse for > 12 months without achieving pregnancy.

3. Evidence of regular ovulation (at least 9 ovulatory cycles per year, confirmed by ultrasonography if needed).

4. At least one fallopian tube confirmed patent by hysterosalpingography or laparoscopy.

5. Male partner with normal semen analysis according to the World Health Organization 5th edition reference standards (no abnormalities in sperm parameters).

2.2.2. Exclusion Criteria

Couples were excluded if any of the following were present:

1. Use of donor eggs or sperm in the cycle.

2. Chromosomal karyotype abnormalities in either or both partners.

3. Use of preimplantation genetic testing in the cycle.

4. Uterine anatomical abnormalities (e.g., uterine septum, unicornuate or bicornuate uterus).

5. Coexisting conditions such as endometriosis, adenomyosis, or intrauterine adhesions.

6. Missing pregnancy outcome records, or cases where no live birth occurred but embryos were still remaining for transfer.

Applying the above criteria, we identified 5465 UI couples from 62201 infertile couples who completed their first complete IVF/ICSI cycle (defined as all fresh and frozen embryo transfer attempts derived from one episode of ovarian stimulation) [2]. Data collection and analysis were performed retrospectively. Figure 1 presents the flow diagram of study structure distribution. Each couple was followed for 2 years unless they achieved a live birth or had no embryos remaining.

FIGURE 1.

Flowchart showing the trial profile and search strategy for couples with UI, recruited from patients with unexplained infertility in Shandong Province, China, 2013–2020.

2.3. Laboratory Procedure

Ovarian stimulation protocols, IVF/ICSI laboratory procedures and the patterns of luteal phase support were individualized for each patient, as previously described [16]. Notably, when paternal sperm failed to meet the IVF criteria on the day of oocyte retrieval or standard IVF encountered unexpected fertilization failure, ICSI was performed.

We evaluated a range of potential risk factors for poor outcomes in these UI couples, focusing on female factors, male factors, and treatment‐related variables. The factors analyzed were:

1. Basic patient characteristics: female age, female body mass index (BMI, categorized as < 18.5, 18.5–22.9, 23.0–27.4, and ≥ 27.5 kg/m² [17]), male age, and sperm concentration.

2. Baseline clinical and medical factors: female ovarian reserve markers (anti‐müllerian hormone [AMH, categorized as < 1.2, 1.2–3.5, and > 3.5 ng/mL] [18, 19, 20, 21, 22], antral follicle count [AFC], basal follicle‐stimulating hormone [FSH], luteinizing hormone [LH], estradiol), thyroid‐stimulating hormone [TSH], infertility diagnosis and type (primary or secondary UI), history of live birth, history of unexplained recurrent spontaneous abortion (URSA), history of uterine polyp removal, history of diagnostic laparoscopy, tubal status (unilateral or bilateral tubal patency), and history of ≥ 2 failed Intrauterine Insemination (IUI) treatments.

3. Treatment cycle variables: controlled ovarian hyperstimulation (COH) protocol type, number of oocytes retrieved, endometrial thickness (EMT) on the day of trigger, fertilization method (IVF or ICSI), normal fertilization rate (NFR), total fertilization failure (TFF), and number of available embryos.

Our approach was not strictly hypothesis‐driven; however, we limited the analysis to the factors listed above to avoid an overly exploratory scope.

2.4. Data Analysis

Our primary outcome was the CLBR per couple. For the purpose of analysis, a “cumulative live birth (CLB)” was defined as having at least one live‐born baby resulting from the first complete cycle (including the fresh embryo transfer and any subsequent frozen–thawed transfers). Baseline patient characteristics were summarized using mean ± standard deviation for continuous variables and frequency (percentage) for categorical variables.

We carried out stepwise analyses to identify risk factors associated with three key outcomes: (a) failure to retrieve any oocytes, (b) failure to obtain at least one usable (transferable) embryo, and (c) failure to achieve a live birth in the first complete cycle. First, we focused on UI patients who underwent ovarian stimulation to determine factors associated with oocyte retrieval (whether at least one oocyte was retrieved in the first cycle). Next, among couples who did retrieve oocytes, we analyzed which factors influenced the availability of at least one transferable embryo. Finally, among couples who proceeded to embryo transfer, we examined factors affecting CLB outcomes.

Univariate logistic regression was used to assess the association of each baseline factor with oocyte retrieval, embryo availability, and CLB outcomes. Variables with a P‐value < 0.05 in univariate analysis, as well as other clinically relevant factors, were then included in multivariate logistic regression models. Backward stepwise regression was subsequently applied to identify the final independent risk factors.

3. Results

3.1. Patient Characteristics

We analyzed a total of 5465 couples with clearly defined UI, representing a UI prevalence of 8.8% among 62201 IVF/ICSI cycles in our center. Among these UI couples, 3400 achieved a live birth following their first complete cycle, corresponding to a CLBR of 62.2% (3400/5465). The baseline characteristics of the cohort (including female age, male age, duration of infertility, hormonal levels, etc., as well as treatment details like COH protocols and fertilization methods) are summarized in Table S1.

3.2. Univariate Analysis of Risk Factors Affecting Oocyte Retrieval

In the first ovarian stimulation cycle of the 5465 UI women, 47 women (0.86%) did not have any oocytes retrieved. Univariate analysis of risk factors for failing to retrieve oocytes showed that increasing female age significantly reduced the likelihood of obtaining oocytes. Women aged ≥ 35 years had a markedly lower chance of successful oocyte retrieval compared to those < 35 years (odds ratio [OR]: 0.51; 95% confidence interval [CI]: 0.28–0.91; p = 0.023). A longer duration of infertility was also associated with a decreased chance of retrieving oocytes (OR: 0.91; 95% CI: 0.83–0.99; p = 0.038). Because only 47 women had zero oocytes retrieved, we did not perform a multivariable regression for this outcome.

Among the 47 women who failed to retrieve any oocytes, 35 (74.5%) had a low AMH level (< 1.2 ng/mL). Low AMH was strongly associated with failure to obtain oocytes (OR: 0.10; 95% CI: 0.04–0.23; p < 0.001). A higher AFC increased the likelihood of oocyte retrieval (OR: 1.28; 95% CI: 1.18–1.38; p < 0.001), whereas a higher basal FSH level was associated with decreased oocyte acquisition (OR: 0.73; 95% CI: 0.68–0.79; p < 0.001) (Table 1).

TABLE 1.

Univariate analysis of risk factors affecting the acquisition of oocytes after ovarian stimulation, recruited from patients with unexplained infertility in Shandong Province, China, 2013–2020.

Variables	None oocytes	≥ 1 oocytes	Univariate analysis of risk factors
	(n = 47)	(n = 5418)	OR, 95% CI	p
Female age (years)
< 35	28 (59.5)	4032 (74.4)	1
≥ 35	19 (40.4)	1386 (25.6)	0.51 (0.28–0.91)	0.023
BMI (kg/m2)
< 18.5	0 (0)	311 (5.7)	/	/
18.5–22.9	20 (42.6)	2562 (47.3)	1
23.0–27.4	17 (36.2)	1881 (34.7)	0.86 (0.45–1.65)	0.658
≥ 27.5	10 (21.3)	664 (12.3)	0.52 (0.24–1.11)	0.092
Duration of infertility (years)	4 (2.5–6.5)	3.5 (2–5)	0.91 (0.83–0.99)	0.043
AFC (n)	7.68 ± 5.04	12.70 ± 5.53	1.28 (1.18–1.38)	0.001
AMH (ng/mL)	0.54 (0.22–1.22)	2.57 (1.36–4.69)	2.69 (1.91–3.79)	0.001
< 1.2	35 (74.5)	1149 (21.2)	0.10 (0.04–0.23)	0.001
1.2–3.5	7 (14.9)	2311 (42.7)	1
3.5	5 (10.6)	1958 (36.1)	1.19 (0.38–3.74)	0.771
FSH (IU/L)	9.93 ± 3.60	7.05 ± 2.20	0.73 (0.68–0.79)	0.001
TSH (μIU/mL)	2.07 (1.32–2.62)	2.17 (1.56–2.94)	1.21 (0.92–1.59)	0.183
Infertility diagnosis, n (%)
Primary	28 (59.6)	3036 (56.0)	1
Secondary	19 (40.4)	2382 (44.0)	1.16 (0.64–2.08)	0.627
URSA, n (%)
Yes	0 (0)	74 (1.4)	/
No	47 (100.0)	5344 (98.6)	/	/
Previous live birth, n (%)
Yes	12 (25.5)	1284 (23.7)	1
No	35 (74.5)	4134 (76.3)	1.10 (0.57–2.13)	0.769
Tub peritoneal factor
Unilateral tubal patency	35 (74.5)	647 (11.9)	1
Bilateral tubal patency	12 (25.5)	4771 (88.1)	1.29 (0.58–2.89)	0.536
Laparoscopic examination, n (%)
Yes	2 (4.3)	121 (2.2)	1
No	45 (95.7)	5297 (97.8)	1.95 (0.47–8.11)	0.361
≥ 2 previous failed IUI, n (%)
Yes	1 (2.1)	296 (5.5)	1
No	46 (97.7)	5122 (94.5)	0.38 (0.05–2.74)	0.334
COH protocol
Long GnRH agonist	3 (6.4)	2787 (51.4)	1
Ultra‐long GnRH agonist	3 (6.4)	155 (2.9)	0.06 (0.01–0.28)	0.001
Short GnRH agonist	19 (40.4)	1465 (27.0)	0.08 (0.03–0.28)	0.001
GnRH antagonist	7 (14.9)	895 (16.5)	0.14 (0.04–0.53)	0.004
Other agonists	15 (31.9)	116 (21.0)	0.008 (0.002–0.029)	0.001

Note: Data are expressed as the mean ± standard deviation, median (interquartile range), percentage or observations number (percentage).

Abbreviations: AFC, antral follicle count; AMH, anti‐Müllerian hormone; BMI, body mass index; COH, controlled ovarian hyperstimulation; FSH, follicle stimulating hormone; GnRH, gonadotropin‐releasing hormone; IUI, intrauterine insemination; TSH, thyroid stimulating hormone; URSA, unexplained recurrent spontaneous abortion.

3.3. Analysis of Risk Factors Affecting Embryo Availability

Of the 5465 UI couples, 5418 (99.1%) had at least one oocyte retrieved in their first cycle, while 47 had none. Among those who retrieved oocytes, 559 couples (10.3%) did not obtain any transferable embryos and therefore had to cancel the embryo transfer, whereas the remaining 4859 couples (89.7%) obtained at least one available embryo for transfer.

Univariate analysis identified several factors significantly associated with the likelihood of obtaining at least one available embryo (p < 0.05 for each). Multivariable logistic regression was performed incorporating both significant univariate factors and clinically relevant covariates (male age, history of failed IUI, female BMI, infertility diagnosis, URSA, prior live birth, tuboperitoneal factors, and fertilization method). Through backward stepwise regression, the following independent factors were identified: female age, male age, duration of infertility, TSH, FSH, AMH, AFC, NFR, and COH protocol.

The odds of obtaining a usable embryo decreased with: advanced female age ≥ 35 (OR: 0.70; 95% CI: 0.52–0.94; p = 0.016), longer infertility duration (OR: 0.95; 95% CI: 0.91–0.98; p = 0.002), higher basal FSH level (OR: 0.91; 95% CI: 0.87–0.94; p < 0.001), higher TSH (OR: 0.92; 95% CI: 0.85–0.99; p = 0.027), lower AMH level (< 1.2 ng/mL vs. higher AMH: OR: 0.72; 95% CI: 0.56–0.93; p = 0.011), and lower AFC (OR: 1.07; 95% CI: 1.05–1.10, p < 0.001). Interestingly, advanced male age showed a positive association with embryo availability (OR: 1.04; 95% CI: 1.01–1.07; p = 0.003). The type of COH protocol also mattered: compared with the long gonadotropin releasing hormone (GnRH) agonist protocol, using a short GnRH agonist protocol was associated with a significantly lower chance of obtaining an embryo (OR: 0.62; 95% CI: 0.48–0.80; p < 0.001), as was the GnRH antagonist protocol (OR: 0.58; 95% CI: 0.44–0.76; p < 0.001). In other words, short agonist and antagonist protocols reduced the probability of having a transferable embryo by ~38% and ~42%, respectively, compared to the ultra‐long protocol. Additionally, a reduced NFR (indicating poorer fertilization success) was associated with a lower probability of obtaining usable embryos. Detailed results of this analysis are presented in Table 2.

TABLE 2.

Analysis of risk factors affecting access to available embryos (n = 5418), recruited from patients with unexplained infertility in Shandong Province, China, 2013–2020.

Variables	None embryos	≥ 1 embryos	Univariate analysis		Multivariate analysis
	(n = 559)	(n = 4859)	OR, 95% CI	p	aOR, 95% CI	p
Female age (years)
< 35	373 (66.7)	3695 (75.3)	1		1
≥ 35	186 (43.3)	1200 (24.7)	0.66 (0.55–0.79)	0.001	0.70 (0.52–0.94)	0.016
Male age (years)	32.81 ± 4.83	32.52 ± 4.67	0.99 (0.97–1.00)	0.112	1.04 (1.01–1.07)	0.003
BMI (kg/m2)
< 18.5	258 (46.2)	2304 (47.4)	1.28 (0.84–1.97)	0.257
18.5–22.9	25 (4.5)	286 (5.9)	1
23.0–27.4	198 (35.4)	1683 (34.6)	0.95 (0.78–1.16)	0.621
≥ 27.5	78 (13.9)	586 (12.1)	0.84 (0.64–1.10)	0.208
Duration of infertility (years)	4 (2–6)	3.5 (2–5)	0.94 (0.91–0.97)	0.001	0.95 (0.91–0.98)	0.002
AFC (n)	10.09 ± 5.27	13.00 ± 5.48	1.12 (1.10–1.15)	0.001	1.07 (1.05–1.10)	0.001
AMH (ng/mL)
< 1.2	123 (22.0)	1223 (25.2)	0.39 (0.31–0.47)	0.001	0.72 (0.56–0.93)	0.011
1.2–3.5	243 (43.5)	1073 (22.1)	1		1
3.5	108 (19.3)	1299 (26.7)	1.31 (1.04–1.65)	0.020	0.94 (0.72–1.22)	0.632
FSH (IU/L)	8.12 ± 3.20	6.95 ± 2.11	0.83 (0.81–0.86)	0.001	0.91 (0.87–0.94)	0.001
TSH (μIU/mL)	2.31 (1.60–3.08)	2.16 (1.56–2.93)	0.96 (0.92–0.99)	0.041	0.92 (0.85–0.99)	0.027
Infertility diagnosis, n (%)
Primary	326 (58.3)	2710 (55.8)	1
Secondary	233 (41.7)	2149 (44.2)	1.11 (0.93–1.33)	0.251
URSA, n (%)
Yes	9 (1.6)	65 (1.3)	1
No	550 (98.4)	4794 (98.7)	1.23 (0.60–2.44)	0.600
Previous live birth, n (%)
Yes	142 (25.4)	1142 (23.5)	1
No	417 (74.6)	3717 (76.5)	1.11 (0.91–1.36)	0.317
Tub peritoneal factor
Unilateral tubal patency	74 (13.2)	573 (11.8)	1
Bilateral tubal patency	485 (86.8)	4286 (88.2)	1.14 (0.88–1.48)	0.319
≥ 2 previous failed IUI, n (%)
Yes	22 (3.9)	274 (5.6)	1		1
No	537 (96.1)	4585 (94.4)	0.69 (0.44–1.07)	0.095	0.68 (0.42–1.09)	0.112
COH protocol
Long GnRH agonist	174 (31.1)	2613 (53.8)	1		1
Extreme‐long GnRH agonist	15 (2.7)	140 (2.9)	0.62 (0.36–1.08)	0.093	0.75 (0.41–1.35)	0.334
Short GnRH agonist	226 (40.5)	1239 (25.5)	0.37 (0.30–0.45)	0.001	0.62 (0.48–0.80)	0.001
GnRH antagonist	94 (16.8)	801 (16.5)	0.57 (0.44–0.74)	0.001	0.58 (0.44–0.76)	0.001
Other agonist	50 (8.9)	66 (1.3)	0.09 (0.06–0.13)	0.001	0.21 (0.13–0.34)	0.001
ART Method
IVF	483 (86.4)	4273 (87.9)	1
ICSI	17 (3.0)	117 (2.4)	0.78 (0.46–1.31)	0.341
IVF/ICSI	59 (10.6)	469 (9.7)	0.89 (0.68–1.20)	0.464
NFR, n (%)
≤ 46	349 (62.4)	1009 (20.7)	0.11 (0.08–0.15)	0.001	0.11 (0.08–0.15)	0.001
47–64	84 (15.0)	1305 (26.9)	0.60 (0.42–0.85)	0.004	0.62 (0.43–0.89)	0.009
65–80	54 (9.7)	1405 (28.9)	1		1
≥ 80	72 (11.8)	1140 (23.5)	0.61 (0.42–0.87)	0.007	0.76 (0.52–1.10)	0.148

Note: Data are expressed as the mean ± standard deviation, median (interquartile range), percentage or observations number (percentage).

Abbreviations: AFC, antral follicle count; AMH, anti‐Müllerian hormone; ART, assisted reproductive technology; BMI, body mass index; COH, controlled ovarian hyperstimulation; FSH, follicle‐stimulating hormone; GnRH, gonadotropin‐releasing hormone; ICSI, intracytoplasmic sperm injection; IUI, intrauterine insemination; IVF, in vitro fertilization; NFR, normal fertilization rate; TSH, thyroid‐stimulating hormone; URSA, unexplained recurrent spontaneous abortion.

3.4. Analysis of Risk Factors Affecting CLB Outcomes

Out of the 4859 UI couples who obtained at least one usable embryo and underwent embryo transfer (fresh or frozen) in their first cycle, 3400 couples achieved a live birth, yielding a CLBR of 62.2% (3400/5465) for UI couples in the first complete cycle. The distribution of clinical characteristics between the live birth group and the non‐live birth group is shown in Table 3.

TABLE 3.

Analysis of risk factors affecting CLB outcomes (n = 4859), recruited from patients with unexplained infertility in Shandong Province, China, 2013–2020.

Variables	CLB (n = 3400)	None CLB (n = 1459)	Univariate analysis		Multivariate analysis
			OR, 95% CI	p	aOR, 95% CI	p
Female age (years)
< 35	2745 (80.7)	914 (62.6)	1		1
≥ 35	655 (19.3)	545 (37.4)	0.40 (0.35–0.46)	0.001	0.63 (0.51–0.78)	0.001
Male age (years)	31.99 ± 4.42	33.78 ± 4.99	0.92 (0.91–0.93)	0.001	0.98 (0.96–0.99)	0.015
BMI (kg/m2)
< 18.5	226 (6.6)	60 (4.1)	1.45 (1.07–1.95)	0.015	1.41 (1.01–2.00)	0.050
18.5–22.9	1664 (48.9)	640 (43.8)	1		1
23.0–27.4	1138 (33.6)	545 (37.4)	0.80 (0.70–0.92)	0.002	0.94 (0.80–1.10)	0.437
≥ 27.5	372 (10.9)	214 (14.7)	0.67 (0.55–0.81)	0.001	0.78 (0.62–0.97)	0.028
Semen concentration (million/mL)	55.30 (36.30–81.50)	54.50 (36.33–80.90)	1.00 (0.99–1.00)	0.672
Duration of infertility (years)	3.5 (2–5)	3.5 (2–5.5)	0.95 (0.93–0.97)	0.001
AFC (n)	13.76 ± 5.47	11.23 ± 5.09	1.10 (1.09–1.11)	0.001	1.01 (1.00–1.03)	0.134
AMH (ng/mL)
< 1.2	442 (13.0)	480 (32.9)	0.38 (0.33–0.45)	0.001	0.64 (0.52–0.77)	0.001
1.2–3.5	1493 (43.9)	618 (42.4)	1		1
3.5	1465 (43.1)	361 (24.7)	1.68 (1.45–1.95)	0.001	1.06 (0.88–1.27)	0.547
FSH (IU/L)	6.75 ± 1.86	7.38 ± 2.42	0.87 (0.84–0.89)	0.001
TSH (μIU/mL)	2.17 (1.57–2.92)	2.13 (1.51–2.95)	1.00 (0.96–1.04)	0.929
Infertility diagnosis, n (%)
Primary	2003 (58.9)	707 (48.5)	1
Secondary	1397 (41.1)	725 (51.5)	0.66 (0.58–0.74)	0.001
Endometrial polypectomy, n (%)
Yes	687 (20.2)	274 (18.8)	1
No	2713 (79.8)	1185 (81.2)	0.91 (0.78–1.07)	0.253
URSA, n (%)
Yes	40 (1.2)	25 (1.7)	1
No	3360 (98.8)	1434 (98.3)	1.46 (0.89–2.42)	0.138
Previous live birth, n (%)
Yes	678 (19.9)	464 (31.8)	1
No	2722 (80.1)	995 (68.2)	1.87 (1.63–2.15)	0.001
Tub peritoneal factor
Unilateral tubal patency	374 (11.0)	199 (13.6)	1
Bilateral tubal patency	3026 (89.0)	1260 (86.4)	1.28 (1.06–1.54)	0.009
≥ 2 previous failed IUI, n (%)
Yes	197 (5.8)	77 (5.3)	1
No	3203 (94.2)	1382 (94.7)	0.91 (0.69–1.19)	0.474
COH protocol
Long GnRH agonist	2017 (59.3)	596 (40.8)	1
Ultra‐long GnRH agonist	106 (3.1)	34 (2.3)	0.92 (0.62–1.37)	0.685
Short GnRH agonist	677 (19.9)	562 (38.6)	0.36 (0.31–0.41)	0.001
GnRH antagonist	573 (16.9)	228 (15.6)	0.74 (0.62–0.89)	0.001
Other agonists	27 (0.8)	39 (2.7)	0.21 (0.12–0.34)	0.001
EMT (cm)
> 1.0	2093 (61.6)	773 (53.0)	1		1
0.7–1.0	1272 (37.4)	666 (45.6)	0.71 (0.62–0.80)	0.001	0.77 (0.66–0.89)	0.001
< 0.7	36 (1.0)	20 (1.4)	0.65 (0.37–1.13)	0.124	1.20 (0.62–2.32)	0.581
Available oocytes, n	11 (8–15)	7 (4–11)	1.17 (1.15–1.18)	0.001	1.02 (1.01–1.04)	0.039
ART Method, n (%)
IVF	2960 (87.6)	1293 (88.6)	1
ICSI	88 (2.6)	29 (2.0)	1.32 (0.86–2.01)	0.204
IVF/ICSI	332 (9.8)	137 (9.4)	1.05 (0.85–1.30)	0.639
Available embryos, n
1	181 (5.3)	449 (30.8)	1		1
2–3	1120 (32.9)	826 (56.6)	3.36 (2.77–4.09)	0.001	3.00 (2.45–3.67)	0.001
4–5	1099 (32.3)	151 (10.3)	18.06 (14.17–23.00)	0.001	13.90 (10.76–17.96)	0.001
≥ 6	1000 (29.5)	33 (2.3)	75.17 (51.03–110.73)	0.001	50.68 (33.71–76.21)	0.001

Note: Data are expressed as the mean ± standard deviation, median (interquartile range), percentage or observations number (percentage).

Abbreviations: AFC, antral follicle count; AMH, anti‐Müllerian hormone; ART, assisted reproductive technology; BMI, body mass index; CLB, cumulative live birth; COH, controlled ovarian hyperstimulation; EMT, endometrial thickness; FSH, follicle stimulating hormone; GnRH, gonadotropin‐releasing hormone; ICSI, intracytoplasmic sperm injection; IUI, intrauterine insemination; IVF, in vitro fertilization; TSH, thyroid stimulating hormone; URSA, unexplained recurrent spontaneous abortion.

Variables significantly associated with CLB (p < 0.05) in univariate analysis were included in the multivariable logistic regression model. Additionally, clinically relevant variables (TSH, history of uterine polyp removal, URSA, and ≥ 2 failed IUIs) were incorporated based on clinical judgment, despite their non‐significance in univariate analysis. Backward stepwise regression identified seven independent factors: female age, male age, BMI, AMH, endometrial thickness on the trigger day, number of oocytes retrieved, and number of usable embryos.

Female age ≥ 35 (OR: 0.63; 95% CI: 0.51–0.78, p < 0.001) was confirmed to be one of the key risk factors. Increasing male age was also associated with a lower likelihood of live birth (OR: 0.98; 95% CI: 0.96–0.99; p = 0.015). Among female factors, obesity (BMI ≥ 27.5 kg/m²) significantly reduced the probability of a live birth (OR: 0.78; 95% CI: 0.62–0.97; p = 0.028). Likewise, a poor ovarian reserve indicated by low AMH (< 1.2 ng/mL) was associated with lower CLB (OR: 0.64; 95% CI: 0.52–0.77; p < 0.001). A thin endometrium also adversely affected outcomes: an EMT in the range 0.7–1.0 cm (vs. > 1.0 cm) was associated with reduced odds of live birth (OR: 0.77; 95% CI: 0.66–0.89; p < 0.001).

On the other hand, the number of oocytes retrieved was a protective factor. Each additional oocyte retrieved slightly increased the chance of achieving a live birth (OR: 1.02; 95% CI: 1.01–1.04; p = 0.039). Similarly, the CLBR was strongly positively correlated with the number of transferable embryos obtained. Compared to UI women with only one available embryo, those with multiple available embryos had dramatically higher odds of having a live birth—in fact, having several embryos increased the likelihood of CLB by up to 50‐fold.

3.5. Maternal and Neonatal Outcomes

Overall, 4122 live newborns were delivered by the 3400 couples who achieved live births. These included 2679 singleton babies and 721 multiples (twins or higher‐order multiples). Excluding 11 newborns with missing sex information, 46.53% of the babies were female and 53.47% were male. Among patients with UI, the risks of preterm birth (PTB), gestational hypertension (GH), and preeclampsia were higher in multiple pregnancies compared to singleton pregnancies (p < 0.05). Similarly, the risks of low birth weight and very low birth weight were also elevated in neonates from multiple pregnancies (p < 0.001), while the risk of macrosomia showed the opposite trend (p < 0.001) (Table 4).

TABLE 4.

Maternal and neonatal outcomes of UI couples after IVF/ICSI, recruited from patients with unexplained infertility in Shandong Province, China, 2013–2020.

Variables	Total outcomes (n = 3400)	Singleton pregnancies (n = 2679)	Multiple pregnancies (n = 721)	p
Complications of pregnancy
Premature labor, n/N (%)	465/3400 (13.7)	179/2679 (6.7)	286/721 (39.7)	< 0.001
Gestational Diabetes, n/N (%)	230/3400 (6.8)	210/2679 (7.8)	50/721 (6.9)	0.418
Gestational hypertension and preeclampsia, n/N (%)	193/3400 (5.7)	126/2679 (4.7)	67/721 (9.3)	< 0.001
Placenta previa, n/N (%)	29/3400 (0.8)	25/2679 (0.9)	4/721 (0.6)	0.327
Neonatal complications
Fetal macrosomia, n/N (%)	249/4108 (6.1)	245/2663 (9.2)	4/1445 (2.8)	< 0.001
Low birth weight infant, n/N (%)	634/4108 (15.4)	96/2663 (3.6)	538/1445 (37.2)	< 0.001
Very low birth weight infant, n/N (%)	51/4108 (1.2)	13/2663 (0.5)	38/1445 (2.6)	< 0.001

Note: p‐value: Comparison of maternal and neonatal complications between singleton and multiple pregnancies.

4. Discussion

The diagnosis of UI imposes significant physical, mental, and economic burdens on affected couples. Identifying risk factors contributing to UI outcomes can greatly benefit patients and their families [23]. Some known risk factors for adverse fertility and pregnancy outcomes—such as Müllerian duct anomalies, chromosomal abnormalities, and endometriosis—are less commonly encountered in UI cases [24]. In this study, we focused on couples with UI undergoing their first IVF/ICSI cycle to determine which factors influence oocyte retrieval, embryo development, and CLB success. We also reported for the first time the incidence of maternal and neonatal complications in UI patients treated with IVF/ICSI.

The prevalence of UI in this study was 8.8%, which is lower than the 15%–30% reported in many previous studies. This difference may be partly explained by our strict inclusion criteria for defining UI. Furthermore, we excluded UI couples who did not complete all embryo transfer cycles and had no live births during follow‐up, which may also reduce the final number of diagnosed UI cases. The CLBR of 62.2% in our cohort (a 2‐year follow‐up) was higher than previously reported rates of 45.9% to 50.6% in unselected infertility patients after one complete IVF/ICSI cycle [2, 9, 10]. A possible explanation for our higher success rate is the younger age of our female participants (20–40 years) and the exclusion of couples with known risk factors for poor outcomes (e.g., Müllerian anomalies, chromosomal abnormalities, endometriosis).

To the best of our knowledge, this study represents the first large‐scale analysis specifically focused on maternal and neonatal outcomes in women with UI. The incidence of PTB (6.7%) and gestational diabetes mellitus (GDM, 7.8%) in singleton pregnancies among our UI cohort falls within the range reported for naturally conceived pregnancies: PTB, 6%–8% [25] and GDM, 5%–10% [26]. Notably, our UI patients demonstrated lower complication rates compared to general IVF/ICSI populations with known infertility etiologies. For instance, the PTB rate in UI singleton pregnancies (6.7%) was lower than the reported range of 7.13%–16.6% in broader IVF/ICSI cohorts [27, 28]. Similarly, the multifetal PTB rate in UI patients (39.7%) was lower than the 52.73% reported in other Assisted Reproductive Technology (ART) populations [29]. When compared to women with polycystic ovary syndrome (PCOS)—a group with distinct endocrine and metabolic abnormalities—UI patients exhibited a markedly more favorable pregnancy risk profile. Specifically, UI patients had lower rates of PTB (6.7% vs. 10.4%–10.5%), GH (4.7% vs. 12.0%–12.2%), preeclampsia (0.9% vs. 3.6%), and GDM (7.8% vs. 24.9%–25.8%) [30, 31]. These findings suggest that UI patients may represent a clinically distinct and relatively healthier subgroup within the infertility population. Their infertility may not stem from overt pathological conditions but rather from factors such as advanced maternal age, suboptimal timing of conception, or subtle physiological variations that evade detection. In contrast, patients undergoing IVF/ICSI for defined etiologies such as PCOS, endometriosis, or tubal factors often carry intrinsic disease burdens that independently elevate the risk of obstetric complications.

It is well known that age is one of the most critical risk factors in female reproductive life. Younger women typically have better ovarian function and respond more favorably to fertility treatments, whereas ovarian oocyte quantity and quality decline with advancing age. Increasing male age is also generally associated with negative effects on fertility and pregnancy outcomes. In our study, the probabilities of obtaining available embryos and achieving a CLB were significantly reduced when women were older than 35 years. In fact, female age ≥ 35 was a pivotal risk factor adversely affecting oocyte retrieval, embryo availability, and ultimately CLB in UI patients undergoing ART. We also found that increasing male age was associated with a lower CLBR, consistent with the findings of Horta et al. [13]. Different from other work, our results suggest that advanced male age might have a mild positive effect on the chance of obtaining available embryos. This inconsistency may be due to heterogeneous inclusion criteria (e.g., different thresholds for semen parameters or definitions of “advanced” male age) and the acknowledgement from the guideline development group that evidence on male age effects is mixed [32]. Notably, our study included only men with normal semen parameters, and we lacked data on male BMI, smoking or alcohol use, and sperm DNA fragmentation—factors that could also influence outcomes. In summary, increasing parental age, especially female age, is linked to poorer IVF/ICSI outcomes. Our findings underscore the importance of assessing fertility at an earlier age and intervening with assisted reproduction in a timely manner if needed.

Ovarian reserve (the quantity and quality of a woman's remaining follicles) is a major determinant of fertility potential. Clinical markers of ovarian reserve include AMH, FSH, and AFC. As expected, we observed a strong positive correlation between ovarian reserve indicators and the chances of retrieving oocytes, obtaining available embryos, and achieving a live birth. Among these markers, AMH is often regarded as a particularly effective predictor of ovarian response and CLB after IVF treatment. In our cohort, a low AMH level was an independent predictor of poorer CLB outcomes in UI couples, reaffirming the importance of ovarian reserve in treatment prognosis.

Female obesity is a well‐known risk factor that adversely affects reproduction, including reduced natural fertility and worse pregnancy outcomes [33]. However, there is still some controversy about the specific impact of obesity on folliculogenesis, embryo development, and endometrial receptivity. In our study, we found that female obesity (BMI ≥ 27.5) was an independent risk factor for a lower CLB in UI patients who underwent embryo transfer. Notably, high BMI was not significantly associated with failure to retrieve oocytes or to obtain embryos, suggesting that obesity's main adverse effect might be on the endometrial receptivity (the uterine environment) rather than on oocyte or embryo quality. Consistent with our findings, another large retrospective study analyzing 22043 autologous frozen–thawed embryo transfer cycles demonstrated a significantly lower live birth rate in obese women (adjusted OR~0.70; 95% CI: 0.62–0.80) after accounting for the potential effect of obesity on oocyte and embryo quality [34]. The negative impact of obesity on endometrial receptivity has been demonstrated in several studies. Using endometrial receptivity array analysis, Bellver et al. reported that obese women (BMI ≥ 30 kg/m²) showed a significantly higher rate of a displaced window of implantation compared to women of normal weight, with a delayed receptivity profile being strongly associated with metabolic disturbances such as dysregulated glucose, insulin, and lipid levels [35]. At a molecular level, obesity has been shown to alter endometrial gene expression, including upregulation of inflammatory mediators (e.g., SDF1, CXCR4), activation of oxidative stress pathways (e.g., NRF2), and dysregulation of lipid metabolism and insulin signaling. These changes contribute to an unfavorable uterine environment that can compromise embryo implantation [36]. Furthermore, in a large retrospective analysis of 2656 first‐time oocyte donation cycles, obese recipients had significantly lower ongoing pregnancy rates compared to normal‐weight recipients, even when receiving high‐quality embryos, indicating that the detrimental effect of obesity on endometrial receptivity is independent of oocyte quality [37]. These results highlight the high priority of adequate weight management counseling for women with UI before they undergo fertility treatment.

Thyroid dysfunction has the potential to affect fertility, and abnormal thyroid function has been implicated as a contributing factor in UI. TSH is the most sensitive and specific indicator for diagnosing and managing thyroid dysfunction. Previous studies have found that TSH levels tend to be higher in infertile women than in fertile women, and that elevated TSH is associated with diminished ovarian reserve. This link may stem from interactions between the hypothalamic–pituitary–ovarian axis and the hypothalamic–pituitary–thyroid axis [38]. In this study, elevated TSH was not an independent influencing factor for the CLBR in UI patients. Cramer et al. [39] also reported no significant association between TSH levels and clinical pregnancy rates following IVF treatment in their prospective study of 509 infertile couples. However, our analysis further identified elevated TSH as an independent risk factor for failure to obtain transferable embryos (following oocyte retrieval) in UI patients. It is noteworthy that Cramer et al. similarly observed significantly higher TSH levels in women with total fertilization failure and noted an association between elevated TSH and fertilization rates below 50%. These findings together suggest that elevated TSH may primarily affect oocyte quality and fertilization efficiency, thereby compromising the availability of viable embryos. It is well recognized that TSH influences oocyte competence primarily through indirect mechanisms. TSH stimulates the release of thyroid hormones, which play a crucial role in regulating the local ovarian environment, including follicular development, oocyte quality, and early embryonic growth [40, 41]. Elevated TSH levels may disrupt the hypothalamic–pituitary–ovarian axis, leading to aberrant FSH and LH secretion and consequently impairing optimal follicular development [42]. Such disruption can result in oocytes with meiotic abnormalities, compromised cytoplasmic maturation, and mitochondrial dysfunction, ultimately affecting fertilization potential [43]. In summary, a high TSH level may interfere with oocyte quality or fertilization processes, thereby reducing the likelihood of obtaining usable embryos.

It is self‐evident that the number of oocytes retrieved and the number of embryos available for transfer are critical factors influencing CLB after ART. Consistent with previous studies, we found that a greater quantity of oocytes and embryos was strongly associated with higher CLB in UI patients. The number of oocytes retrieved reflects the effectiveness of ovarian stimulation; more oocytes typically lead to more embryos and thus more chances for a live birth. In our data, 11.1% of UI couples failed to obtain any oocytes or any usable embryos in their first IVF/ICSI cycle, a rate that is somewhat higher than the 4.7%–10.5% reported in general infertility populations without age restriction [44, 45]. Successful fertilization is a prerequisite for embryo development. In our cohort, the median NFR was 64%, and the incidence of TFF was 4.1%. Consistent with reports focusing on non‐male factor infertility [46, 47], we observed that using ICSI (as opposed to conventional IVF) did not improve the likelihood of obtaining usable embryos or achieving a live birth in UI couples. However, a low NFR was significantly associated with a decreased chance of obtaining usable embryos, highlighting that fertilization efficiency remains important. With advances in reproductive genetics, various genetic factors related to fertilization failure and early embryonic arrest have been identified [48]. Unfortunately, the genetic etiology of UI is still largely unknown. Given that UI is a diagnosis of exclusion, a future direction could be to classify UI subgroups based on underlying reproductive potential or specific defects. For example, screening panels for genetic variants associated with sporadic or familial UI, recurrent TFF, or early embryonic arrest in large cohorts of UI patients may help uncover potential genetic causes of UI.

A subset of UI cases might be explained by subtle impairments in endometrial receptivity. Endometrial receptivity refers to the state of the uterine lining that allows a blastocyst to implant and establish a pregnancy. EMT on the day of ovulation trigger or transfer is one of the most important clinical indicators of endometrial receptivity. In our study, to ensure consistency and minimize time‐related confounding factors, we exclusively utilized trigger‐day EMT measurements for all analyses. This standardized time point reflects a well‐defined physiological stage, providing patients with a stable and comparable assessment metric. Standardizing measurements to this single time point enhances data consistency and strengthens the validity of our risk factor analysis. Researches have shown that a thinner endometrium is associated with lower IVF/ICSI success rates, while women with thicker endometrium generally have better outcomes [49, 50]. Consistent with that, our study found that a thinner EMT was an independent risk factor for a poorer CLB outcome. Endometrial thickness may influence CLBR through several interrelated biological pathways. A thin endometrium often reflects impaired estrogen‐mediated signaling, which can lead to inadequate expression of key receptivity markers such as integrin αvβ3, leukemia inhibitory factor, and HOXA10 [51, 52]. The dysregulation of these molecules compromises endometrial receptivity and impairs embryo adhesion and implantation [52, 53]. In addition, endometrial thickness serves as an indirect indicator of subendometrial vascular perfusion. Inadequate thickness is frequently associated with increased uterine artery resistance and reduced angiogenesis, which may result in local hypoxia and nutrient deficiency—conditions that are unfavorable for embryonic development [54]. Furthermore, a thin endometrium is often characterized by poor development of the functional layer and a lack of synchrony between glandular and stromal maturation, leading to displacement—either advancement, delay, or narrowing—of the window of implantation [55]. Such asynchrony between endometrial and embryonic development can directly lead to implantation failure, even when high‐quality embryos are transferred [56]. In our analysis, EMT proved to be a crucial indicator for assessing endometrial receptivity in predicting reproductive outcomes.

We also considered other factors related to the uterus and prior treatments, such as a history of endometrial polypectomy or a history of failed IUI, but neither showed a significant impact on CLB in our cohort. However, it is noteworthy that the current cohort included patients with a history of endometrial polypectomy. This suggests that any potential mechanical obstruction or inflammatory microenvironment associated with polyps had been surgically addressed prior to enrollment. As a result, the uterine cavity environment in these women was largely restored, rendering their reproductive prognosis comparable to that of women without a history of polyps. Our findings thus indirectly affirm the efficacy of polypectomy in reclassifying such patients from the explained infertility category to the UI group, with no residual adverse impact on CLBRs. Furthermore, our results indicate that a history of unsuccessful IUI does not predict inferior IVF outcomes. IUI success is highly dependent on tubal patency, sperm quality, and adequate sperm–oocyte interaction–mechanical and functional barriers that are effectively bypassed with IVF/ICSI. Therefore, couples who previously responded poorly to IUI may achieve IVF success rates similar to those of UI patients without prior IUI attempts. This underscores that prior IUI failure should not be considered a negative prognostic factor, nor should it discourage patients from proceeding to IVF. For individuals with UI, IVF remains a highly effective treatment option irrespective of IUI history.

This study has several limitations. First, as a retrospective study, it is subject to selection bias, which is an inherent disadvantage. Second, our data were derived from a single medical center, so the generalizability to other settings is uncertain. Future studies should include multi‐center collaborations with diverse patient populations to validate and extend our findings. Third, although all male participants had normal routine semen parameters, data on male BMI, lifestyle habits (e.g., smoking, alcohol use), and advanced sperm function parameters such as DNA fragmentation were not available. These factors have been previously associated with embryonic development and live birth rates [57, 58, 59, 60]. Thus, the results should be interpreted with this limitation in mind, and future studies would benefit from incorporating a broader assessment of male factors. Additionally, because only a small number of UI patients in our cohort failed to retrieve any oocytes, we were unable to perform an in‐depth analysis of risk factors for complete failure of oocyte acquisition. Larger sample sizes or pooled data may be required to explore predictors of oocyte retrieval failure in UI patients.

5. Conclusion

In conclusion, we retrospectively analyzed clinical data from thousands of couples with UI undergoing their first complete IVF/ICSI cycle. We evaluated the risk factors affecting each critical step—oocyte retrieval, acquisition of transferable embryos, and CLB outcomes—as well as maternal and neonatal complication rates.

The prevalence of UI in our cohort was confirmed to be 8.8% (5465 out of 62201 infertility treatment cycles). Among these UI couples, 3400 ultimately achieved live births, resulting in a CLBR of 62.2% (3400/5465). Our findings suggest that classic risk factors, including sociodemographic characteristics like female age and medical factors like ovarian reserve, remain the primary determinants of oocyte yield, embryo development, and live birth outcomes in UI patients. However, the impact of male factors on embryo development and live birth, and the mechanisms underlying those effects, warrant further investigation.

Funding

This work was supported by Natural Science Foundation of Shandong Province, ZR2022MH009. National Natural Science Foundation of China, 82101712. Projects of Medical and Health Technology Development Program in Shandong Province, 202005010520, 202005010523. Health Commission's Science and Technology Development Program in Jinan, 2022‐2‐194, 2022‐TCM‐43.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Basic characteristics of couples with unexplained infertility undergoing the first IVF/ICSI treatment, recruited from patients with unexplained infertility in Shandong Province, China, 2013–2020.

Liu R., Wu T., Ding G., et al., “Analysis of Risk Factors Affecting Cumulative Live Birth in Couples With Unexplained Infertility Undergoing IVF/ICSI,” Reproductive Medicine and Biology 25, no. 1 (2026): e70025, 10.1002/rmb2.70025.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.