Primary versus secondary recurrent pregnancy losses: Clinical findings and live birth rate after comprehensive work-up and personalized management.

Chiara Tersigni: Conceptualization, methodology, software, validation, writing—original draft preparation, project administration, project administration. Marianna Onori: methodology, software, investigation, formal analysis, writing—original draft preparation. Giuliana Beneduce: investigation. Fabio Sannino: methodology, investigation, formal analysis, data curation, writing—original draft preparation. Rita Franco; Conceptualization, validation, investigation, data curation. Andrea Busnelli: software, formal analysis. Chiara Granieri: investigation, resources. Domenico Milardi: investigation. Alfredo Pontecorvi: resources, supervision. Antonio Lanzone: resources, visualization, supervision. Giovanni Scambia: resources, supervision, funding acquisition. Nicoletta Di Simone: resources, writing—review and editing, funding acquisition. All authors have read and agreed to the published version of the manuscript.

The study was approved by the Fondazione Policlinico Universitario A. Gemelli IRCCS, Italy, IRB on October 28, 2016 (approval number 36401/16 ID:1355).

Italian Ministry of Public Health supported this research project (RC 2023).

A cohort of 551 couples were considered for study entry. A total of 93 women were excluded because they experienced: a single late pregnancy loss ( n  = 59), intrauterine fetal demise ( n  = 24) or preterm premature rupture of membranes ( n  = 10). Of the remaining 458 couples, only 295, 89 recruited in Milan and 206 in Rome, completed the entire diagnostic work‐up and were included in the final analysis (Figure  1 ). RPL couples work‐up was performed within an average time of 6 months from first clinical evaluation (or booking). For all couples enrolled in this study, the next pregnancy (or index pregnancy), where achieved, was obtained within 12 months from the end of the work‐up. Recruitment and drop out of the study population. Flow chart summarizing inclusion, exclusion, and drop out of participants of this study. IUFD, Intrauterine fetal demise; pPROM, Preterm premature rupture of membranes; RPL, Recurrent pregnancy losses. Clinical characteristics and obstetrical history of RPL women are reported in Table  1 . A higher frequency of smokers and use or assisted reproductive techniques (ART) was detected among women with pRPL compared with sRPL ( p  < 0.01 and p  < 0.001, respectively). As expected, a significantly higher prevalence of cesarean sections (40%) was observed among sRPL women. No differences were observed between the two cohorts in terms of age, BMI, or distribution of number of previous pregnancy losses. Clinical characteristics and obstetrical history of recurrent pregnancy loss couples enrolled in this study. Note : Continuous variables were defined using median and interquartile range (IQR) while categorical variables were defined as percentages. Abbreviations: ART, assisted reproductive technology; IUI, intrauterine insemination; IVF, in vitro fertilization; pRPL, primary recurrent pregnancy loss; RPL, recurrent pregnancy loss; sRPL: secondary recurrent pregnancy loss. Refers to RPL women's age during work‐up. Prevalence of female and male factors found in the study RPL cohort is summarized in Figure  2 . Female factors associated with pRPL were detected in 81 out of 157 (52%) investigated couples. Combined female/male factors were found in 61 cases of pRPL (61%). Isolated male factors were detected in only eight cases of pRPL (5%). Unexplained RPL was diagnosed in the absence of any recognizable causal or possible explanatory factors and accounted for 6 cases (4%) of our population of pRPL (Figure  2A ). Out of 138 couples with sRPL, in 69% ( n  = 95) of cases only female factors could be found while in remaining 31% ( n  = 43) only isolated male factors were found (Figure  2B ). Prevalence of causes/explanatory factors in the study cohort of RPL couples. (A, B) Pie chart summarizing female, combined or male causes/explanatory factors, or unexplained cases, found in the two cohorts of primary (A, n  = 157) and secondary (B, n  = 138) recurrent pregnancy losses (RPL). Results of female factors found in the two sub‐groups of pRPL compared with the sRPL cohort, are summarized in Table  2 and extensively reported Table  S1 . Main factors identified in RPL women were hormonal ad metabolic factors, autoimmune disorders, acquired uterine/endometrial factors and genital infections. A lower incidence of the following were observed: mullerian anomalies, inherited thrombophilias and karyotype abnormalities. Higher prevalence of hormonal and metabolic factors was found among women with sRPL ( p  < 0.01). In particular, a higher prevalence of hyperprolactinemia ( p  < 0.01), poor ovarian reserve ( p  < 0.0001) and luteal insufficiency ( p  < 0.01) was observed in the sRPL cohort. On the other hand, higher frequency of diabetes mellitus, both type I and II, was found in the pRPL cohort ( p  < 0.05). No significant differences were found in terms of prevalence of autoimmune diseases between the two cohorts. Higher frequency of cervical and endometrial polyps ( p  < 0.001), as well as chronic endometritis ( p  < 0.05) were observed at hysteroscopy among the women with history of pRPL compared with the sRPL group. Otherwise, a higher prevalence of uterine myomas were shown in the sRPL compared with the pRPL cohort ( p  < 0.001). In addition, a higher frequency of vaginal and/or cervical infections were observed in the cohort of sRPL ( p  < 0.0001). Among Mullerian uterine anomalies, only a higher prevalence of uterum septum was detected in the cohort of pRPL ( p  < 0.05). No significant differences in terms of congenital thrombophilias was found between the two cohorts except of a higher prevalence of Protein S deficiency in the sRPL cohort. No significant differences in terms of karyotype abnormalities were observed between the two cohorts. Prevalence of female factors in primary recurrent pregnancy loss (pRPL) vs secondary pregnancy loss (sRPL) cohorts. Note : % refers to the whole cohort of women. Abbreviations: ART, assisted reproductive techniques; FIVET, fertilization in vitro and embryo transfer; ICSI, intracytoplasmic sperm injection; IUI, intrauterine insemination; pPROM, preterm premature rupture of membranes; pRPL, primary recurrent pregnancy loss; RPL, recurrent pregnancy loss; sRPL, secondary recurrent pregnancy loss. Main factors identified in RPL males were sperm infections, abnormal semen analysis and karyotype abnormalities (Table  3 ; Table  S2 ). No significant differences were found in the male population between the two cohorts of pRPL and sRPL in terms of sperm infections or karyotype abnormalities. An isolated higher prevalence of combined asthenozoospermia and teratozoospermia was observed in the pRPL ( p  < 0.05). Prevalence of male factors in primary recurrent pregnancy loss (pRPL) vs secondary recurrent pregnancy loss (sRPL) cohorts. Note : % refers to the whole cohort of men. RPL: recurrent pregnancy loss; pRPL: primary recurrent pregnancy loss; sRPL: secondary recurrent pregnancy loss. Abbreviations: pRPL, primary recurrent pregnancy loss; RPL, recurrent pregnancy loss; sRPL, secondary recurrent pregnancy loss. All RPL women underwent vaginal administration of progesterone 200 mg twice a day from the positive pregnancy test until 12 weeks of gestation. Although administration of progesterone is recommended during the first only in RPL women with vaginal bleeding and threaten miscarriage, 1 in these population, progesterone was administered based on some reports showing that empiric progesterone administration may be of some potential benefit in RPL women. 29 , 30 Vitamin D deficiency was supplemented by oral daily administration of 2.000 UI of vitamin D3 and normalization of serum levels assessed 4 months after starting supplementation. Women with antiphospholipid syndrome were administered with combination of enoxaparin and aspirin according to EULAR recommendations. 31 Women ≥40 years old and/or women with isolated seropositivity for specific autoantibodies (ANA and anti‐DNA), even not associated with clinical signs or symptoms autoimmune diseases, were orally administered with 150 mg of aspirin per day in subsequent pregnancies from early first trimester for the increased risk of preeclampsia. 32 Women with specific inherited thrombophilias (prothrombin mutation, V Leiden mutation, deficient antithrombin III, Protein C or S) were administered with a prophylactic dosage of enoxaparin during pregnancy and puerperium because of the mothers' increased risk of thrombosis. 33 Women diagnosed with celiac disease were pre‐conceptionally put on a gluten‐free diet. Both female and male genital tract infections were treated with specific antibiotics and couples were suggested to looking for a subsequent pregnancy after the infection's eradication. In case of HPV infections couples underwent anti‐HPV vaccination with 9‐valent vaccination in order to facilitate the HPV clearance and to reduce the risk of HPV‐related diseases. 33 , 34 , 35 Abnormal semen analysis were usually related to infectious or anatomic factors (i.e., varicoceles) and significantly recovered after removing the primary causes. Percutaneous endovascular embolization of the testicular varicocele was performed according to EAU Guidelines 37 and the semen analysis was repeated after 4–6 months. Clinical hypothyroidism or subclinical hypothyroidism in presence of anti‐thyroid autoantibodies were treated by oral administration of levothyroxine during pre‐conception period and during pregnancy until the delivery, with median daily dosage of 50 mcg targeting a TSH <2.5 mU/L. 38 Hyperprolactinemia was treated by oral administration of bromocriptine at a median daily dosage of 5 mg. Insulin resistance and hyperinsulinemia was treated by oral administration of metformin (1500 mg/day) pre‐conceptionally and continued during pregnancy until the delivery. DMI and II were treated pre‐conceptionally with insulin and/or metformin according to diabetologist’ indications to obtain target fasting (<92 mg/dL) and post‐prandial (<130 mg/dL at 60 min) glucose levels. All women with diagnosis of uterine septum or T‐shaped uterus underwent metroplasty. Submucosal myomas, polyps or uterine cavity synechiae were removed by operative hysteroscopy. Women with intramural/sub‐serous myomas larger than 5 cm and/or associated to uterine bleeding were removed by myomectomy. Chronic endometritis were managed by oral administration of oral doxycycline (200 mg daily) for 7–10 days 39 and the recovering of the endometritis after treatment assessed by a subsequent hysteroscopy with CD138+ cells count. Among the 295 couples with RPL followed at our centers, overall pregnancy rate after the comprehensive work‐up and therapeutic management of couples in the next pregnancy (or index pregnancy) was 71%. Higher pregnancy rate was observed in pRPL couples (84.7%) compared with sRPL couples (55.8%). Overall pregnancy loss rate of the index pregnancy was 5%, with a significantly higher pregnancy loss rate in the pRPL cohort (8.9%) vs the sRPL cohort 0% ( p  < 0.01). Overall live birth rate in the next pregnancy was 66.4%, with a higher live birth rate in the pRPL cohort (76%) compared with the sRPL cohort (55.8%). Overall use of ART in the index pregnancy was 11%. Higher frequency of ART use was found among pRPL compared with sRPL cohort ( p  < 0.001), although a higher prevalence of oocyte donation was observed in the sRPL group compared with the pRPL ( p  < 0.01). In terms of index pregnancy outcome, more operative vaginal deliveries occurred in the cohort of pRPL compared with sRPL women. Time to pregnancy was significantly longer in pRPL cohort compared with the sRPL one ( p  < 0.0001). A higher frequency of intrauterine fetal growth restriction (IUGR) was observed in the pRPL cohort compared with the sRPL cohort ( p  < 0.05), while higher frequency of placental abruption was found in the sRPL women compared with the pRPL ones (Table  4 ). Obstetric outcomes in primary recurrent pregnancy loss (pRPL) vs secondary pregnancy loss (sRPL) couples. Abbreviations: ART, assisted reproductive technology; IUI, intrauterine insemination; IVF, in vitro fertilization; pRPL, primary recurrent pregnancy loss; RPL, recurrent pregnancy loss; pPROM, preterm premature rupture of membranes; sRPL, secondary recurrent pregnancy loss. Both univariate and multivariate analysis identified AMH >1 ng/mL as the main factor increasing the pregnancy rate in pRPL women (OR 4.59 95% CI 1.39–14.35, p  < 0.05; OR 4.14 95% CI 1.18–13.77; p  40 years, BMI > 30 kg/m 2 , smoking habit). None of the analyzed factors (AMH, age, BMI, smoking) significantly correlated with the pregnancy rate in the sRPL women (Tables  S2 and S4 ). Based on univariate analysis, AMH >1 ng/mL and women age <40 years were identified as factors increasing the chance of live birth rate in pRPL women (OR 3.96 95% CI 1.34–12.52, p  < 0.05 and OR 2.76 95% CI 1.36–5.64, p  < 0.01). Age lower <40 years was also associated with an increased live birth rate in sRPL women (OR 3.23 95% CI 1.55–6.94, p  < 0.01). After adjusting the OR for confounding factors (AMH, BMI and smoking), both pRPL and sRPL women aged 40 years (OR = 2.60, 95% CI, 1.25–5.47, p  < 0.05 and OR 3.83, 95% CI 1.78–8.56, p  < 0.01) (Tables  5 and 6 ). Logistic regression analysis of factors influencing live birth rate in primary recurrent pregnancy loss women. Note : Statistical significance for p  < 0.05. Abbreviations: AMH, anti‐Müllerian hormone; BMI, body mass index; CI, confidence interval; ns, not significant; OR, odds ratio. For women underwent to oocyte donation, the age refers to oocyte's donor. Logistic regression analysis of factors influencing live birth rate in secondary pregnancy loss women. Note : Statistical significance for p  < 0.05. Abbreviations: AMH, anti‐Müllerian hormone; BMI, body mass index; CI, confidence interval; ns, not significant; OR, odds ratio. For women underwent to oocyte donation, the age refers to oocyte's donor.

RPL is a complex and heterogeneous syndrome with a multifactorial etiology. Current recommendations include the investigation of female uterine structural abnormalities, thyroid function and antiphospholipid antibodies. 1 Investigation of female inherited thrombophilias, autoimmunity, sperm DNA fragmentation and couples' karyotype is an additional option for clinicians, 1 however, this work‐up allows to identify a cause for RPL in only 50%–60% of cases. Aiming at personalizing therapeutic strategies in the management of RPL, in this study, we analyzed clinical findings after a comprehensive work‐up investigating several factors potentially interfering with the reproductive outcome of RPL couples, taking into account previous obstetrical history (primary vs secondary RPL). Indeed, history of a secondary obstetric failure is generally more suggestive of an acquired cause compared with a more “constitutive” condition that might be responsible for primary RPL cases. In our cohort of RPL couples, causes or possible explanatory factors could be detected in 93% of pRPL couples and in 100% of sRPL couples. In our population of pRPL, we found out a more frequent “endometrial factor” associated with pregnancy losses, probably negatively interfering with embryo implantation and placental development. In particular, we observed a higher prevalence of acquired endometrial factors (chronic endometritis, endometrial and cervical polyps) potentially leading to the establishment of a hostile endometrial environment. This observation is consistent with the finding of a longer time to pregnancy in the pRPL cohort. On the other hand, a more frequent “ovarian factor” was observed in the sRPL cohort, together with a shorter time to pregnancy, pointing to maternal aging and poor ovarian reserve, as the main risk factor for sRPL. Noteworthy, the latter factor is not currently part of the recommended work‐up for RPL women. In the complex, our observations are coherent with the larger use of homologous ART among the group of pRPL (25% of live births), probably leading to pregnancy because of overcoming, with unknown mechanisms, endometrial receptive defects. Parallelly, a higher prevalence of oocyte donation (4% of live births; 100% of ART procedures) was found in the sRPL group, due to the higher prevalence of poor ovarian reserve in this group of patients. In terms of next pregnancy outcome, in our series, a personalized therapeutic approach, focused to remove modifiable risk and potential interfering factors, lead to a live birth rate in 66.4% of couples, in particular, 76% in pRPL and 56% in sRPL couples. Based on multivariate analysis, in our population, the most important positive predictive factor for live birth rate in both primary and secondary RPL women was age <40 years. Noteworthy, spontaneous pregnancies with live births was observed in 89.2% of RPL couples, in particular in 75.6% of pRPL and in 96.1% of sRPL couples. Consistently to previous reports, 37 we observed a higher frequency of intrauterine growth restriction in the pRPL cohort compared with the sRPL cohort, supporting the hypothesis of a common physiopathology among miscarriage, intrauterine growth restriction and preeclampsia, which represent different clinical presentation of the same phenomenon of poor placentation, in relation to the extent and severity of the defect. 38 Major strengths of this study are (a) the large number of diagnostic factors investigated in a selected population of women with RPL, (b) the detailed stratification and comparison between pRPL and sRPL women in terms of etiology and next pregnancy outcomes, that is currently an under‐investigated topic in the available international literature. Major limitation of this study are (a) its retrospective nature; (b) the administration of some experimental treatments even in the absence of strong recommendations (progesterone, aspirin, heparin, doxycycline), based on retrospective evidences; (c) the lack of a control population. This last issue could not be solved due to the high costs and ethical questions potentially raising in performing such enlarged diagnostic work‐up in a healthy population. Finally, in our RPL population, we have included women over 40 years because of the high average age of Italian women at first pregnancy (35 years), although in many Northern European Countries, the maternal age >40 years is a clinical exclusion criterion to access the RPL‐work‐up in a public health regimen. This choice is a consequence of cultural and demographic characteristics of the study population, and certainly, could provide different results in terms of prevalence of some factors associated with RPL (i.e., poor ovarian reserve, placental insufficiency, endometrial atrophy, etc.) compared with younger populations.

A comprehensive work‐up allows to identify potential causes or possible explanatory factors of RPL in more than 95% of RPL couples. Different etiology and next pregnancy outcomes are expected in pRPL and sRPL women. In our population, a wide screening for RPL‐associated factors and personalized management led to successful pregnancy in 76% of pRPL and 56% of sRPL couples. The most important positive predictive factor for a live birth is an age lower than 40 years. Caution has to be taken in evaluating the results of this retrospective study because of the experimental nature of some treatments administered to our RPL cohort and to draw conclusions about the specific efficacy of each treatment. More research is needed to find out unknown pathogenic mechanisms of RPL and to define the efficacy of customized therapeutic strategies.

Recurrent pregnancy loss (RPL) is defined as two or more consecutive pregnancy losses before 24 weeks of gestation and affects up to 1%–2% of reproductive age couples. 1 RPL is a complex obstetrical syndrome with a heterogeneous etiopathogenesis generating grief to couple and frustration to clinicians. Recognized risk factors for RPL are: advanced maternal age, congenital Müllerian anomalies, parental genetic factors, 1 diminished ovarian reserve, 2 acquired or congenital thrombophilias, 3 , 4 , 5 obesity, 6 hormonal and metabolic disorders, 7 , 8 autoimmune diseases, uterine endometrial factors, 9 maternal vitamin D deficiency, 10 both female and male genital tract infections, sperm abnormalities and/or sperm DNA fragmentation. 11 Meta‐analyzes investigating the role of lifestyle‐related factors such as smoking, alcohol, caffeine and stress have not shown clear increase of risk for RPL. 12 Current recommendations suggest to investigate only female uterine structural abnormalities, thyroid function and antiphospholipid antibodies. 1 Screening of female inherited thrombophilias and autoimmunity, sperm DNA fragmentation and couples' karyotype can also be assessed, 1 although this work‐up allows to identify causal/risk factors in only 50%–60% of RPL couples. On the other hand, there are no different recommendations in clinical management nor in the treatment of women with RPL and no previous live births (primary RPL) and women with RPL after at least one live birth (secondary RPL). Furthermore, it is not clear to clinicians dealing with RPL whether it is expected to have different causes between primary or secondary RPL or different pregnancy outcomes or live birth rate in a subsequent pregnancy (index pregnancy). The aims of this study were to collect clinical data from a cohort of women with history of RPL referred from January 2018 to December 2022 to the Fondazione Policlinico A. Gemelli IRCCS (Rome, Italy) and to the IRCCS Humanitas Research Hospital (Milan, Italy), undergoing a comprehensive diagnostic work‐up, and to investigate: (a) the possible different etiology of pRPL and sRPL; (b) analyze the next pregnancy outcomes and live birth rate in pRPL vs sRPL.

The authors declare no conflicts of interest.

All couples referred from January 2018 to December 2022 to the Fondazione Policlinico A. Gemelli (Rome, Italy) and to the Humanitas Research Hospital (Milan, Italy), for at least two previous pregnancy losses, were considered for study entry. All individuals enrolled in this study gave their informed consent to use, anonymously, their data for research purposes, and the protocol was approved by the ethics committee of the Fondazione Policlinico Universitario A. Gemelli IRCCS, Italy (IRB approval number 36401/16 ID:1355). Couples with a previous history of stillbirth (intrauterine fetal demise ‐ occurring after the 24th week of gestation) or preterm premature rupture of membranes occurring before 24 weeks of gestation, were excluded from this study. Couples who had experienced at least two consecutive pregnancy losses at a gestational age <24 weeks of gestation with the same partner were diagnosed with RPL. Only couples who completed the entire RPL diagnostic work‐up could be included in this study. Couples with RPL who had never given birth to a live infant were defined as affected by primary RPL (pRPL). Couples who had given birth to one or more live infants before pregnancy losses were defined as affected by secondary RPL (sRPL). The work‐up for women with RPL was conducted according to a shared internal protocol of the Fondazione Policlinico A. Gemelli (Rome, Italy) and the Humanitas Research Hospital (Milan, Italy). The protocol was conceived and drawn up on the basis of the RCOG 2011, 13 ASRM 2012 14 and ESHRE 2017 1 guidelines on RPL. In particular, RPL women underwent hormonal tests between the 3th and the 5th day of menstrual cycle. Serum levels of thyroid‐stimulating hormone (TSH), tri‐iodothyronine (FT3), thyroxine (FT4), follicle‐stimulating hormone (FSH), luteinizing hormone (LH), estradiol, prolactin and anti‐Mullerian hormone (AMH) were measured by chemiluminescent immunoassays. RPL women were also tested with a glycemic and insulinemic load curve with 75 g and glucose and insulin levels were assessed by UV hexokinase assay. Serum levels of vitamin 25‐OH‐D were assessed by immunometric assays. Screening for inherited thrombophilias (V Leiden mutation, prothrombin mutation G20210A, Protein C, Protein S. homocysteine, factor VIII, antithrombin III), anti‐phospholipid syndrome (lupus anticoagulant, anti‐beta 2 glycoproteins I IgG and IgM, anti‐cardiolipin IgG and IgM), and autoimmunity (anti‐nuclear antibodies, extractable nuclear antigen antibodies, anti‐liver kidney microsomal antibodies, anti‐neutrophil cytoplasmic antibodies, complement fractions c3 andc4, anti‐dsDNA antibodies, anti‐smooth cells antibodies, anti‐Saccharomyces cerevisiae, anti‐transglutaminase IgA and IgG antibodies, anti‐endomysium IgA, anti‐gliadin IgA, anti‐thyroglobulin, anti‐thyroperossidase, anti‐TSH receptor antibodies) was also performed. Women were evaluated for vaginal dysbiosis by microbial cultures of vaginal swabs. To detect cervical infections for Chlamydia, mycoplasmas or Neisseria gonorrhoeae cervical swab were analyzed by Next Generation Sequencing. To prevent reciprocal contamination between cervical and vaginal niches a sterile speculum and two different sterile swabs were used. Transvaginal gynecological ultrasound with 3D reconstruction of the uterine cavity was performed in all the female study cohort. Briefly, all women were examined in the lithotomy position using a US machine Samsung HS70A, equipped with a 5–9 MHz transvaginal transducer (Samsung V5‐9). The US examination was performed in the luteal phase and began with a conventional 2D transvaginal US followed by acquisition of 3D data using the 3D volume mode. The sweep angle was set to 120° to include the entire uterus and a 3D dataset was then acquired using the high‐quality, slow‐sweep mode. The uterus shape was studied based on coronal view information and with uterine wall thickness used as reference point 15 and the measurement performed according to the recommendations by the Thessaloniki ESHRE/ESGE consensus. 16 The signs of adenomyosis were evaluated and reported according to MUSA revised Criteria. 17 Diagnostic hysteroscopy and endometrial biopsy for CD138+ plasma cells count was also performed in the secretory phase in all the female study cohort. Chronic endometritis was diagnosed in case of ≥5 CD138+ plasma cells/HPF (high‐power fields). 18 Atrophic endometrium was diagnosed in case of hysteroscopic findings of smooth and thin endometrium with direct visualization of the deep endometrial vessels and confirmed by the histopathological examination of the endometrial biopsy. 19 Cervical stenosis was reported in case of a cervix with an obliterated cervical ostium and/or cervical canal at hysteroscopy. 20 Male screening included: spermiogram, sperm culture for aerobic and anaerobic germs and Next Generation Sequencing for Chlamydia, mycoplasmas or Neisseria gonorrhoeae. All 295 couples underwent karyotype analysis. Clinical hypothyroidism was diagnosed for TSH >2.5 μUI/ml and low levels of FT3 (<2.4 pg/mL) and/or FT4 (25 μg/L, confirmed by minutes 0′, 15′ 30′ curve. Poor ovarian reserve was defined for AMH <1 ng/mL. 23 Luteal insufficiency was defined in case of progesterone levels <5 ng/mL in the all three measurements performed in 18th, 21st, 24th days of menstrual cycle (identified by women at home with LH‐test as LH + 4, LH + 7 and LH + 10, respectively). 24 All women were subjected to a 75 g oral glucose tolerance test (OGTT), analyzing glycaemia and insulinemia at minutes 0′, 30′, 60′, 90′, 120 after glucose load. Diabetes was diagnosed based on a HbA1C value ≥6.5% or a fasting plasma glucose value ≥126 mg/dL (7.0 mmol/L) or a 2‐h plasma glucose value after a 75‐g OGTT ≥200 mg/dL (11.1 mmol/L) or occasional plasma glucose value of ≥ 200 mg/dL (≥11.1 mmol/L). 25 Insulin after glucose oral load was expressed as the area under the curve (I‐AUC), calculated according to the trapezoidal rule. Insulin resistance was defined as HOMA‐IR >2.7, 26 while hyperinsulinemia was defined as a value I‐AUC was >16 921 mIU/mL after 180 min. 27 Abnormal semen analysis was defined according to World Health Organization (WHO) guidelines 28 and the semen samples were collected after 3–5 days of sexual abstinence. Data about baseline characteristics and medical history, were collected. Once the distribution characteristics were evaluated, continuous variables were defined using median and interquartile range (IQR). Categorical variables were defined as percentages. We extrapolated the data from the entire study population to estimate the prevalence of seven categories of female RPL‐associated factors (i.e., hormonal and metabolic factors, autoimmune factors, acquired uterine factors, cervical and vaginal infections, congenital Müllerian abnormalities, inherited thrombophilia, karyotype anomalies) and 3 categories of male RPL‐associated factors (sperm infection, sperm abnormalities according to the WHO classification and karyotype anomalies). For each rate, 95% confidence interval (CI) of proportions was estimated through the binomial exact calculation. Comparisons between pRPL and sRPL were performed using Chi‐squared tests, Fisher's exact tests, and Mann–Whitney U (Wilcoxon) statistic, as appropriate. The data were analyzed using IBM SPSS Statistics (Version 27). The comparative analysis must be considered as exploratory since a formal sample size calculation for this analysis was not performed. Statistical significance was set to an alpha level of 0.05. All statistical tests were two‐tailed. Odds Ratio (OR) calculation was performed using Chi‐square test. Confidence intervals were calculated with Baptista‐Pike for OR. Univariate and multivariate logistic regressions were performed using Prism software version 10. For all analyses, p  < 0.05 was considered significant.

Table S1. Prevalence of female factors in pRPL versus sRPL cohorts. Table S2. Prevalence of male factors in pRPL versus sRPL cohorts. Table S3. Logistic regression analysis of factors influencing pregnancy rate in pRPL women. Table S4. Logistic regression analysis of factors influencing pregnancy rate in sRPL women.