Chance of live birth in women with recurrent pregnancy loss referred to a French assisted reproductive technology center.

Observational OA: gold
⚙ AI-generated summary by qwen3.7-flash, 2026-11-18 ⓘ

In women with recurrent pregnancy loss, 75.3% achieved live birth mostly via spontaneous conception, while diminished ovarian reserve and higher miscarriage counts predicted lower success rates.

One-sentence paraphrase of the abstract; not a substitute for reading it. No clinical advice. How this works

⚙ AI-generated deep summary by qwen3.7-flash, 2026-11-18 · read from full text ⓘ

This retrospective cohort study evaluated live birth outcomes in 85 women with recurrent pregnancy loss referred to a French assisted reproductive technology center. The analysis identified that lower anti-Müllerian hormone levels and active paternal smoking were significantly associated with reduced chances of achieving a live birth during the follow-up period. The paper does not explicitly discuss endometriosis or adenomyosis; it was included in the corpus via a keyword match in the upstream search index.

Read from the paper's body, not the abstract. Not a substitute for reading the paper. No clinical advice. How this works

Abstract

BackgroundRecurrent pregnancy loss (RPL) is a multifactorial condition, with over 50% of cases remaining unexplained. The absence of therapeutic consensus raises questions regarding referral for assisted reproductive technology (ART).ObjectiveTo evaluate live birth rates and predictors of success among women with RPL referred to a tertiary ART center.MethodsWe conducted a retrospective cohort study. Women aged 18-42 years with ≥ 3 pregnancy losses before 14 weeks' gestation, referred to the ART center of Amiens-Picardie University Hospital (2014-2022) were included. The primary outcome was achieving a live birth during follow-up. Independent predictors were identified using stepwise logistic regression.ResultsAmong the 85 women with RPL included (median age 29.6 [8.6] years), 49/85 (57.6%) had no underlying infertility issues, and 36/85 (42.3%) had associated infertility issues. Most women were nulliparous (50/85, 58.8%), and had a median of 3 [1] previous miscarriages. Overall, 64/85 (75.3%) achieved at least one live birth: 33/64 (52%) spontaneously and 31/64 (48%) via ART, with median times to conception of 13.6 [10.5] and 14.2 [10.6] months, respectively. ART use was more frequent among infertile women (≥1 IVF/ICSI cycle: 26/36, 72.2% vs. 17/49, 34.7%; p = 0.001), leading to a higher proportion of medically-assisted conceptions in this group (20/24, 83.3% vs. 11/40, 27.5%; p < 0.001). Diminished ovarian reserve and a higher number of previous miscarriages were independently associated with lower live birth rates (aOR 0.26 [0.08-0.89] and 0.76 [0.53-0.96], respectively).ConclusionApproximately three-quarters of women with RPL achieved a live birth, most after spontaneous conception. ART should be reserved for couples with confirmed infertility or when another underlying cause clearly justifies intervention.
Full text 40,200 characters · extracted from pmc-nxml · 5 sections · click to expand

Results

Among the 5,186 women who attended the Department of Reproductive Biology and Medicine at Amiens-Picardie University Hospital for an initial consultation between January 1, 2014, and December 31, 2022, 96 (1.9%) had a history of three or more pregnancy losses before 14 WG. Follow-up data were available for 85 patients (1.6%), who were included in the study (Fig.  1 ). Of these, 49 (57.6%) consulted for a personal history of RPL without infertility, whereas 36 (42.3%) presented with couple infertility associated with RPL in their obstetric history. The demographic, medical, and obstetric characteristics of the patients and their partners are summarized in Table  1 . The median age of women at consultation was 29.6 [8.6] years. Twenty-seven women (31.8%) were active smokers, and 3 (3.1%) indicated regular alcohol consumption. Obesity (body mass index ≥ 30 kg/m²) was recorded for 21 women (24.7%). Thyroid disease was reported for 12 women (14.1%), unilateral tubal surgery for previous ectopic pregnancy for five women (5.9%), and unilateral oophorectomy for ovarian endometriosis for three women (3.5%). Most of the women were nulliparous (50/85, 58.8%). The remaining 35 patients (41.2%) had previously achieved at least one live birth, all following spontaneous conception; half of these births (17/35) involved a different partner. Regarding the number of miscarriages at the time of first consultation, 62 women (72.9%) had experienced three, 9 (10.6%) four, and 14 (16.5%) at least five. Fetal loss beyond 14 WG occurred in 16 women (18.8%), including two instances beyond 22 WG. Fig. 1 Flow chart of patient selection and inclusion. Legend : RPL: recurrent pregnancy loss, WG : week of gestation Flow chart of patient selection and inclusion. Legend : RPL: recurrent pregnancy loss, WG : week of gestation Table 1 Characteristics of patients, partners, and RPL work-up according to live birth outcome Total ( N  = 85) Live birth ( N  = 64) No live birth ( N  = 21) p OR [CI 95%] Couple infertility. n / N (%) 36/85 (42.3) 24/64 (37.5) 12/21 (57.1) 0.11 0.45 [0.17–1.22] Women characteristics  Age ( years ), median [IQR] 29.6 [8.6] n  = 85) 28.7 [8.7] 32.7 [10.1] 0.19  Age ( years ), n/N (%)   - 18–29 39/85 (45.9) 33/64 (51.6) 6/21 (28.6) 0.07 2.66 [0.92–7.73]   - 30–34 19/85 (22.3) 12/64 (18.7) 7/21 (33.4) 0.16 0.46 [0.15–1.39]   - 35–39 18/85 (21.2) 14/64 (21.9) 4/21 (19.0) 1 1.19 [0.34–4.11]   - 40–42 9/85 (10.6) 5/64 (7.8) 4/21 (19.0) 0.21 0.36 [0.09–1.49]  Active smoker, n/N (%) 27/85 (31.8) 20/64 (31.2) 7/21 (33.3) 0.86 0.91 [0.32–2.60]  Excessive alcohol consumption, n/N (%) 3/85 (3.5) 0/64 (0) 3/21 (14.3) NA  BMI ≥ 30 kg/m 2 , n/N (%) 21/85 (24.7) 16/64 (25.0) 5/21 (23.8) 0.91 1.07 [0.34–3.38]  Thyroid disease n/N (%) 12/85 (14.1) 6/64 (9.4) 6/21 (28.6) 0.028 0.26 [0.07–0.92]  Diabetes, n/N (%) 2/85 (2.4) 1/64 (1.6) 1/21 (4.8) NA  Autoimmune disease, n/N (%) 1/85 (1.2) 1/64 (1.6) 0/21 (0) NA  History of ovarian or tubal surgery, n/N (%) 7/85 (8.2) 5/64 (7.8) 2/21 (9.5) 1.0 0.97 [0.14–11.04]  Nulliparity, n/N(%) 50/85 (58.8) 38/64 (59.4) 12/21 (57.1) 0.86 1.1 [0.40–2.97]  ≥ 1 fetal death (> 14 WG), n/N(%) 16/85 (18.8) 10/64 (15.6) 6/21 (28.6) 0.32 0.32 [0.14–1.48]  Total number of miscarriages, median [IQR] 3 [1] n  = 85) 3 [1] 3 [1.25] 0.27  Total number of miscarriages, n/N (%)   - 3 62/85 (72.9) 48/64 (75.0) 14/21 (66.7) 0.64 1.50 [0.51–4.37]   - 4 9/85 (10.6) 8/64 (12.5) 1/21 (4.8) NA   - 5 6/85 (7.1) 4/64 (6.2) 2/21 (9.5) NA   - ≥ 6 8/85 (9.4) 4/64 (6.3) 4/21 (19.0) NA Partner characteristics  Age ( years ), median [IQR] 33.0 [7.0] n  = 85) 32.0 [7.0] 28.5 [5.2] 0.27  Active smoker, n/N (%) 29/85 (34.1) 26/64 (40.6) 3/21 (14.3) 0.031 4.11 [1.11–15.37]  Excessive alcohol consumption, n/N (%) 2/85 (2.4) 2/64 (3.1) 0/21 (0) NA  BMI ≥ 30 kg/m 2 , n/N (%) 10/85 (11.8) 6/64 (9.4) 4/21 (19.0) 0.25 0.44 [0.11–1.74]  ≥ 1 previous live birth, n/N(%) 14/85 (16.5) 10/64 (15.6) 4/21 (19.0) 0.74 0.79 [0.19–3.90] RPL work-up  AMH (ng/mL), median [IQR] 3.25 [2.7] ( n  = 85) 3.80 [2.60] 1.53 [2.41] < 0.001  FSH (IU/L), median [IQR] 7.0 [3.0] ( n  = 85) 7.0 [3.0] 7.0 [2.0] 0.061  LH (IU/L), median [IQR] 5.0 [2.0] ( n  = 85) 5.0 [2.2] 5[2.0] 0.86  Estradiol (pg/mL), median [IQR] 38.0 [16.0] ( n  = 85) 38.0 [17.4] 33.0 [6.0] 0.34  AFC, median [IQR] 13 [12] ( n  = 85) 14 [13] 7 [12] 0.68  DOR, n/N (%) 17/85 (20.0) 9/64 (14.1) 8/21 (38.1) 0.017 0.27 [0.09–0.82]  Positive ANA (titer ≥ 1:80), n/N (%) 12/84 (14.3) 11/63 (17.5) 1/21 (4.8) NA 4.23 [0.51–34.93]  ≥ 1 positive aPL, n/N (%) 6/84 (7.1) 5/63 (7.9) 1/21 (4.8) NA 1.72 [0.19–15.66]  ≥ 1 positive anti-thyroid Ab, n/N (%) 6/82 (7.3) 3/62 (48.4) 3/20 (15.0) 0.31 0.29 [0.05–1.56]  Inherited thrombophilia, n/N (%) 11/84 (13.1) 8/63 (12.7) 3/21 (14.3) 1.0 0.87 [0.21–3.65]  Abnormal parental karyotype, n/N (%) 8/82 (9.8) 7/62 (11.3) 1/20 (5.0) 0.67 2.42 [0.28–20.95]  Endometriosis, n/N (%) 14/85 (16.5) 12/64 (18.8) 2/21 (9.5) 0.50 2.19 [0.45–10.71]  PCOS, n/N (%) 11/85 (12.9) 7/64 (10.9) 4/21 (19.0) 0.46 0.53 [0.14–2.04]  Uterine malformation, n/N (%) 2/85 (23.5) 2/64 (3.1) 0/21 (0) NA  Uterine fibroid or myoma, n/N (%) 13/85 (15.3) 7/64 (10.9) 6/21 (28.6) 0.51 0.31 [0.09–1.05]  Tubal abnormality, n/N (%) 13/85 (15.3) 6/64 (9.4) 7/21 (33.3) 0.021 0.21 [0.06–0.71]  Chronic endometritis, n/N (%) 5/83 (6.0) 4/62 (6.5) 1/21 (4.8) 1.0 0.73 [0.01–7.94]  Abnormal semen analysis, n/N (%) 15/85 (17.6) 10/64 (15.6) 5/21 (23.8) 0.39 0.59 [0.18–1.99]  DFI > 30%, n/N (%) 15/85 (17.6) 10/64 (15.6) 5/21 (23.8) 0.39 0.59 [0.18–1.99] Legend: Ab antibody, ANA antinuclear antibodies, aPL antiphospholipid antibodies, AMH anti-Müllerian hormone, AFC antral follicle count, BMI body mass index, DFI DNA Fragmentation Index, DOR diminished ovarian reserve, NA  not applicable, PCOS polycystic ovary syndrome Bold p -values indicate statistical significance ( p < 0.05) Characteristics of patients, partners, and RPL work-up according to live birth outcome Legend: Ab antibody, ANA antinuclear antibodies, aPL antiphospholipid antibodies, AMH anti-Müllerian hormone, AFC antral follicle count, BMI body mass index, DFI DNA Fragmentation Index, DOR diminished ovarian reserve, NA  not applicable, PCOS polycystic ovary syndrome Bold p -values indicate statistical significance ( p < 0.05) The results of RPL assessments are summarized in Table  1 . Seventeen patients (20.0%) presented with DOR. These patients, as a group, were significantly older than those without DOR (35.1 [7.1] vs. 28.4 [8.3] years, p  < 0.001), however, two patients (2.4%) met the criteria for premature ovarian insufficiency. The median number of miscarriages did not differ significantly between the DOR/non-DOR groups (3[1] in both groups). Among patients with a history of thyroid disease, only one had an abnormal thyroid-stimulating hormone (TSH) level. Six patients out of 84 (7.1%) tested positive for at least one aPL, however, none met all of the criteria for obstetric APS. Most patients ( n  = 84) were screened for hereditary thrombophilia, with at least one abnormality detected in 11 patients (13.1%). None of these patients had a prior history of venous or arterial thromboembolic events. Fifty-three couples (62.3%) were given recommendations related to lifestyle habits, and psychological support was offered to 31 couples (36.5%) (Table  2 ). Table 2 Interventions and follow-up outcomes according to live birth outcome Total ( N  = 85) Live birth ( N  = 64) No live birth ( N  = 21) p OR [CI 95%] Lifestyle modification*, n / N (%) 53/85 (62.3) 41/64 (64.1) 12/21 (57.1) 0.86 0.9 [0.28–2.92] Discontinuation of toxic agents exposure*, n/N (%) 44/85 (51.8) 33/64 (51.6) 11/21 (52.4) 0.95 0.97 [0.36–2.60] Weight reduction*, n/N (%) 27/85 (31.8) 19/64 (29.7) 8/21 (38.1) 0.47 0.69 [0.24–1.92] Psychological counseling*, n/N (%) 31/85 (36.5) 25/64 (39.1) 6/21 (28.6) 0.64 1.31 [0.43–4.01] Antioxidant supplementation, n/N (%) 22/85 (22.9) 19/64 (29.7) 3/21 (14.3) 0.37 2.16 [0.55–8.43] Vitamin D supplementation, n/N (%) 46/85 (54.1) 37/64 (57.8) 9/21 (42.9) 0.52 1.43 [0.48–4.23] Progesterone, n/N (%) 7/85 (8.2) 5/64 (7.8) 2/21 (9.5) 0.63 0.64 [0.11–3.63] Antiplatelet therapy, n/N (%) 31/85 (36.5) 22/64 (34.4) 9/21 (42.9) 0.23 0.51 [0.17–1.53] Prophylactic anticoagulation, n/N (%) 23/85 (27.1) 17/64 (26.6) 6/21 (28.6) 0.58 0.72 [0.23–2.27] Antibiotic therapy**, n/N (%) 4/85 (4.7) 3/64 (4.7) 1/21 (4.8) 1.0 1.19 [0.02–15.95] Hysteroscopic intervention, n/N (%) 6/85 (7.1) 3/64 (4.7) 3/21 (14.3) 0.11 0.25 [0.03–2.08] Surgical intervention, n/N (%) 7/85 (8.3) 4/64 (6.3) 3/21 (14.3) 0.40 0.43 [0.11–1.69] ART procedures, n/N (%)  - ≥ 1 COS cycle with TI 8/85 (9.4) 8/64 (12.5) 0/21 (0) 0.19 NA  - ≥ 1 IUI 13/85 (15.3) 9/64 (14.1) 4/21 (19.0) 0.73 0.70 [0.19–2.54]  - ≥ 1 IVF/ICSI cycle 43/85 (50.6) 27/64 (42.2) 16/21 (76.2) < 0.001 0.05 0.01–0.41]  - ≥ 1 embryo transfer 41/85 (48.2) 27/64 (42.2) 14/21 (66.7) 0.011 0.18 [0.03–0.73] Duration of follow-up (months), median [IQR] 18.5 [21.2] 16.7 [16.1] 37.5 [37.1] 0.004 ≥ 1 further miscarriage, n/N (%) 21/85 (24.7) 15/64 (23.4) 6/21 (28.6) 0.56 0.69 [0.20–2.63] Recurrent implantation failure, n/N (%) 12/41 (29.3) 4/27 (16.7) 8/14 (57.1) 0.06 0.21 [0.03–1.14] ≥ 1 live birth, n/N(%) 64/85 (75.3) Spontaneous conception, n/N (%) 33/64 (51.6) Medically-assisted conception, n/N (%) 31/64 (48.4)  − by COS with TI 1/64 (1.6)  − by IUI 4/64 (6.3)  − by IVF/ICSI 26/64 (40.6) Legend: ART assisted reproductive technology, COS controlled ovarian stimulation, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection, IUI intrauterine insemination, NA not applicable, TI timed intercourse Bold p values indicate statistical significance ( p < 0.05) *Indications provided by the physician to the couple ** for chronic endometritis Interventions and follow-up outcomes according to live birth outcome Legend: ART assisted reproductive technology, COS controlled ovarian stimulation, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection, IUI intrauterine insemination, NA not applicable, TI timed intercourse Bold p values indicate statistical significance ( p < 0.05) *Indications provided by the physician to the couple ** for chronic endometritis The main pharmacological treatments prescribed are summarized in Table  2 . Occasional targeted medical management was provided, such as treatment for diabetes (1/85, 1.2%) and for thyroid dysfunction (1/85, 1.2%). Specific gynecological procedures were also undertaken: six patients (7.1%) underwent hysteroscopy (for fibroid resection or synechiae removal), and seven patients (8.3%) underwent surgical interventions such as endometrioma drainage, tubal surgery, or ovarian drilling. Regarding ART procedures, 8 women (9.4%) received at least one cycle of controlled ovarian stimulation with timed intercourse, 13 (15.3%) received at least one IUI, and 43 (50.6%) underwent at least one cycle of I IVF/ICSI. Of these, 41 women (48.2%) proceeded to at least one embryo transfer, with a median of 2 [3] transfers. Gamete donation was used by three couples (3.5%). Among the eight couples (9.8%) with an abnormal parental karyotype, five underwent preimplantation genetic testing (PGT); of the remaining three couples, one was referred for oocyte donation, one for sperm donation, and one achieved a spontaneous ongoing pregnancy. No other patients underwent PGT during follow-up. During follow-up, 21 patients (24.7%) experienced at least one new miscarriage. Of the 41 women (29.3%) who underwent embryo transfer, RIF was recorded in 12. At least one live birth was achieved for 64 patients (75.3%) (Table  2 ). Among these, 33 women (51.6%) conceived spontaneously ( Table  2 ): 5 (7.8%) between assessment and treatment, 18 (28.1%) after medical and surgical treatment, and 10 (15.6%) during ART procedures. The median time to conception for spontaneous pregnancies was 13.6 [10.5] months. The other 31 pregnancies (48.4%) were achieved through ART, mainly via IVF/ICSI (26/64, 40.6%) (Table  2 ). The median time to conception for assisted pregnancies was 14.2 [10.6] months. Nine patients with a history of RPL without infertility achieved a viable pregnancy through IVF. Four of these patients had an abnormal constitutional karyotype, and benefited from PGT; two had ovarian endometriosis; one had PCOS associated with severe obesity; one had a bicornuate bicervical uterus, and benefited from targeted embryo transfer; and one benefited from oocyte donation abroad. Patients with infertility were slightly older than those without (30.3 [7.4] vs. 28.4 [8.9] years, p  = 0.55; Table S1). Although, as expected, the total number of miscarriages tended to be higher in women without infertility, the difference was not statistically significant. Infertile patients had lower median AMH levels (2.8 [2.5] vs. 3.7 [2.8] ng/mL, p  = 0.08) and slightly reduced AFC (12 [13] vs. 14.5 [12.5], p  = 0.10), although rates of DOR were comparable between groups. Other criteria investigated as part of RPL work-up were similar. Thus, management approaches, including recommendations for lifestyle modifications, psychological support, and medical treatments, did not differ significantly between groups with and without infertility. However, ART was substantially more frequently offered to infertile patients (≥ 1 IVF/ICSI cycle: 26/36, 72.2% vs. 17/49, 34.7%, p  = 0.001), resulting in a greater proportion of medically-assisted conceptions in this group (20/24, 83.3% vs. 11/40, 27.5%, p  < 0.001). Infertility within the couple was more frequent among those who did not achieve a live birth during follow-up (24/64, 37.5% vs. 12/21, 57.1%; p  = 0.11). Although the median age of women was not statistically different between groups (28.7 [8.7] vs. 32.7 [10.1] years, p  = 0.19), younger women (18–29 years) tended to achieve live birth more often, representing more than half of the women achieving a live birth (33/64 (51.6%)) and less than a third of the group not achieving a live birth (6/21 (28.6%), p  = 0.07) (Table  1 ). Thyroid disorders were less frequent among women who achieved live birth (6/64 (9.4%)) than among women who did not achieve live birth (6/21 (28.6%), p  = 0.028). The median number of previous miscarriages was similar between groups achieving or not achieving a live birth (3 [1] vs. 3 [1.25], p  = 0.27) (Table  1 ). Active smoking for which lifestyle advice was offered was significantly more frequent among partners of women who achieved live birth (26/64 (40.6%)) than among partners of women not achieving a live birth (3/21 (14.3%), p  = 0.031) (Table  1 ). Women who achieved a live birth had higher median AMH levels (3.80 [2.60] ng/mL vs. 1.53 [2.41] ng/mL for those not achieving a live birth, p  < 0.001), and DOR was less frequent in the group achieving a live birth (9/64 (14.1%) vs. 8/21 (38.1%), p  = 0.017). Tubal abnormalities were more common among women who did not achieve a live birth (6/64 (9.4%) vs. 7/21 (33.3%), p  = 0.021). No other biological or ancillary test results were significantly associated with live birth (Table  1 ). Lifestyle modification, psychological support, supplementation, medical therapies, and surgical interventions were similar between groups (Table  2 ). In contrast, women who did not achieve a live birth significantly more frequently underwent at least one IVF/ICSI cycle (16/21 (76.2%) vs. 27/64 (42.2%) women who did achieve a live birth, p  < 0.001) (Table  2 ). As would be expected in this context, the median duration of follow-up was significantly longer among those who did not achieve a live birth (16.7 [16.1] vs. 37.5 [37.1] months, p  = 0.004). Stepwise logistic regression using AIC was applied to assess the influence of patient age, DOR, number of previous miscarriages, thyroid disease, and tubal abnormality. The final model retained DOR, number of previous miscarriages, and tubal abnormality (Table  3 ). DOR was associated with a reduced chance of live birth (aOR 0.26 [0.08–0.89], p  = 0.03), and each additional miscarriage decreased the likelihood of live birth by 24% (aOR 0.76 [0.53–0.96], p  = 0.05). Table 3 Factors associated with achieving live birth according to stepwise multivariate logistic regression analysis Variables aOR [IC 95%] p Total number of miscarriage 0.76 [0.53–0.96] 0.05 DOR 0.26 [0.08–0.89] 0.03 Tubal abnormality 0.32 [0.08–1.33] 0.11 Variables entered into the initial model were maternal age, DOR, number of previous miscarriages, thyroid disease, and tubal abnormality. Following stepwise bidirectional selection based on Akaike’s Information Criterion (AIC), the final model retained DOR, number of previous miscarriages, and tubal abnormality Bold p -values indicate statistical significance ( p < 0.05) Legend: DOR diminished ovarian reserve Factors associated with achieving live birth according to stepwise multivariate logistic regression analysis Variables entered into the initial model were maternal age, DOR, number of previous miscarriages, thyroid disease, and tubal abnormality. Following stepwise bidirectional selection based on Akaike’s Information Criterion (AIC), the final model retained DOR, number of previous miscarriages, and tubal abnormality Bold p -values indicate statistical significance ( p < 0.05) Legend: DOR diminished ovarian reserve

Patients

This retrospective cohort study was performed at the Department of Reproductive Biology and Medicine at Amiens-Picardie University Hospital (France). Eligible participants were women aged 18 to 42 years who first consulted between January 1, 2014, and December 31, 2022, allowing for a minimum follow-up of two years. To be eligible for inclusion, women had to have a personal history of at least three spontaneous pregnancy losses before 14 WG prior to consultation [ 7 ], and follow-up data had to be available after the initial consultation. All pregnancy losses had to have resulted from spontaneously conceived pregnancies, and biochemical pregnancies were counted as losses. Patients with recurrent implantation failure (RIF) or repeated biochemical pregnancies following in vitro fertilization (IVF) embryo transfers were excluded. Medical records were retrieved from the computerized database of the Reproductive Biology and Medicine Department (Médifirst ® software, version 1.6.2.8). Data collected included sociodemographic and clinical characteristics (lifestyle factors, medical and gynecological history, and obstetric background), results of the RPL work-up (hormonal, endocrine, and autoimmune investigations, ovarian reserve assessment, thrombophilia testing, parental karyotyping, semen analysis), and pelvic imaging (ultrasound, hysteroscopy, endometrial biopsy). In routine clinical practice at our center, ovarian reserve assessment included serum anti-Müllerian hormone (AMH) measurement, basal hormonal testing (follicle-stimulating hormone (FSH), luteinizing hormone (LH), and estradiol), and antral follicle count (AFC) evaluation by transvaginal ultrasound. Basal hormonal testing and AFC were usually performed during the early follicular phase (cycle days 2–5), whereas AMH was measured irrespective of cycle day. Data collected covered information on medical or surgical management and assisted reproductive procedures, including controlled ovarian stimulation with timed intercourse, intrauterine insemination (IUI), and in vitro fertilization (IVF) with or without intracytoplasmic sperm injection (ICSI). The main outcome monitored was the occurrence of at least one live birth during follow-up. The occurrence of spontaneous miscarriages during follow-up was also evaluated. Infertility was defined as the absence of conception after ≥ 12 months of regular unprotected intercourse [ 12 ]. Diminished ovarian reserve (DOR) was defined by at least one of the following criteria: FSH > 10 IU/L, AMH < 1.2 ng/mL, or total AFC on ultrasound < 5–7 follicles in women with regular menstrual cycles [ 13 ]. RIF was defined as the absence of implantation after at least three transfers of good-quality embryos [ 14 , 15 ]. Categorical variables were expressed as frequencies (n/N, %) and continuous variables as mean (standard deviation [SD]) or median (interquartile range [IQR]), as appropriate. Normality was assessed using the Shapiro–Wilk test. For univariate analyses, continuous variables were compared using the Student’s t-test or Wilcoxon test, and categorical variables were compared using the χ² test or Fisher’s exact test. Multivariate logistic regression analysis was used to identify factors independently associated with the occurrence of at least one live birth. Variables for which data was missing in less than 20% of records (none exceeding 5% individually) were considered. Adjustment variables were informed by the literature and included maternal age and the number of previous miscarriages [ 6 , 8 , 9 ]. These were supplemented by clinically relevant variables that emerged as statistically significant in univariate analyses ( p < 0.05). An initial full model including all eligible predictors was fitted. Stepwise bidirectional selection based on Akaike’s Information Criterion (AIC) was then performed to derive the most parsimonious model. Results are presented as odds ratios (OR) with 95% confidence intervals (95% CI). Model discrimination was assessed using the area under the receiver operating characteristic curve (AUC). All tests were two-sided, with the threshold for statistical significance set at p  < 0.05. All analyses were performed using R software (version 4.4.2). This study was conducted in accordance with the Declaration of Helsinki and French legislation governing retrospective non-interventional research (Law No. 2012 − 300 of March 5, 2012, as amended on April 12, 2018). According to these regulations, ethics committee or Institutional Review Board approval was not required for this study, and therefore ethics approval was waived. The study complied with the MR-004 reference methodology of the French National Commission for Data Protection and Liberties (CNIL), which serves as the competent authority for data protection and research compliance in France. The study was registered under reference number PI2024_843_0027. All patients were informed during routine clinical care that their anonymized data could be used for research purposes, and none objected. In accordance with French regulations, informed consent to participate was therefore waived.

Discussion

In this study, we evaluated records for 85 women with a history of RPL attending a French assisted reproduction center. We focused on reproductive outcomes and how clinical factors and treatment strategies associate with achieving a live birth. Most women (57.6%) were not infertile. Although 24.7% experienced at least one subsequent miscarriage, 75.3% ultimately delivered a live birth. In the non-infertile group, 82% of women achieved a live birth, predominantly (72.5%) through spontaneous conception. Among infertile women, 66.7% achieved a live birth, most frequently (83.3%) following some form of ART. Our findings are consistent with previous cohort studies reporting a generally favorable long-term reproductive prognosis in women with RPL after referral to specialized clinics. A large Danish cohort of 987 women with ≥ 3 pregnancy losses before 22 WG reported live birth rates of 66.7% within five years and 71.1% within fifteen years after the initial consultation [ 8 ]. Prognosis was mainly influenced by maternal age and the number of previous miscarriages [ 8 ]. Similarly, Kling et al. studied 719 women with ≥ 3 pregnancy losses before 22 WG and observed a cumulative ongoing pregnancy rate of 86% and a live birth rate of 65% after five years [ 16 ]. Most pregnancies (69%) occurred spontaneously after a median of 33.7 months, and favorable outcomes were associated with maternal age < 35 years, a shorter time to conception (< 36 months), tubal patency, and the absence of male factor infertility, emphasizing that subfertility rather than recurrent miscarriage itself may be a major determinant of long-term reproductive outcome [ 16 ]. More recent studies have confirmed these observations. Linehan et al. reported reproductive outcomes in 748 women attending a dedicated recurrent miscarriage clinic [ 17 ]. Overall, 77% achieved a subsequent pregnancy, of which 63% resulted in a live birth. Live birth rates were markedly lower in women aged ≥ 40 years and in those with associated infertility. Likewise, in a prospective cohort of 251 women with unexplained RPL, Kaandorp et al . found that 86% conceived within 24 months and that 65% ultimately achieved a live birth. Time to conception was comparable to that observed in healthy fertile women, suggesting that many women with unexplained RPL retain preserved reproductive potential. Interestingly, the number of previous miscarriages was the only independent predictor of time to live birth. More recently, Koki et al. further expanded these findings in a prospective UK cohort of 1,201 women (mean number of previous pregnancy losses: 3.4 ± 1.8) attending recurrent miscarriage clinics [ 18 ]. During three years of follow-up, 53.4% conceived within the first year and a further 22.8% during the second year, whereas only 10.7% failed to conceive [ 18 ]. Overall, approximately 60% of documented pregnancies resulted in a viable pregnancy beyond 24 WG. Increasing maternal age, BMI, PCOS, and the number of previous miscarriages were independently associated with poorer reproductive outcomes, whereas a history of previous live birth was associated with a more favorable prognosis [ 18 ]. Likewise, the prospective Tommy’s Net cohort ( n = 639) observed that 72.8% of women who conceived subsequently achieved a viable pregnancy, whereas maternal smoking, obesity, repeated miscarriage, and subfertility were associated with failure to conceive or achieve a live birth [ 19 ]. Taken together, these studies indicate that the favorable reproductive prognosis observed after RPL is influenced not only by miscarriage history itself but also by underlying fertility potential, maternal characteristics, and modifiable lifestyle factors. Our findings regarding ART are also consistent with the available literature. Qiu et al . assessed the impact of RPL history on the outcomes of infertility treatment in women without chromosomal or uterine abnormalities, studying 29,825 frozen embryo transfer cycles. In their study population, 899 women (3.0%) had a history of RPL [ 20 ]. They found no significant differences in biochemical pregnancy, miscarriage, or live birth rates between women with and without a history of RPL [ 20 ]. In our study, among the 41 women who underwent at least one embryo transfer, 70% achieved at least one ongoing pregnancy leading to a live birth during follow-up. This proportion mirrors that of the overall French population. Thus, a study of couples undergoing IVF found that eight years later, 71% of couples had fulfilled their parental project: 48% through IVF or another medical treatment, 12% through spontaneous conception, and 11% through adoption [ 21 ]. Likewise, a Swedish longitudinal study of 439 women receiving ART reported that, whereas approximately 19% of women remained childless, 80.9% had at least one child within five years: 71.6% thanks to ART, 28% through spontaneous conception, and 6.8% following adoption [ 22 ]. The relatively high live birth rate observed in our cohort is reassuring, and several mechanisms may explain this favorable prognosis. First, many women with RPL retain normal reproductive potential despite their history of miscarriage. In a substantial proportion of cases, pregnancy losses may result from sporadic embryonic chromosomal abnormalities rather than persistent maternal factors affecting all subsequent pregnancies. Furthermore, our cohort was relatively young, most women did not exhibit DOR, and the median number of previous miscarriages was only three, suggesting that severe forms of RPL were uncommon. These factors are all associated with more favorable reproductive outcomes [ 10 , 16 , 23 , 24 ]. Finally, multidisciplinary management in a specialized RPL center may also have contributed to the favorable reproductive outcomes observed in our cohort by providing structured, evidence-based care throughout the diagnostic and therapeutic pathway [ 6 , 7 , 25 ]. A relevant component of the distress experienced by couples with RPL can be linked to confusion and fragmentation of the care pathway. According to the ESHRE, two consecutive miscarriages should not be trivialized and their occurrence warrants specialized evaluation [ 6 ]. However, other professional societies continue to define RPL as three or more pregnancy losses [ 7 , 25 ]. Specialized consultation provides an opportunity to perform a comprehensive evaluation, identify potentially treatable conditions, optimize lifestyle factors, and offer individualized counseling and psychological support [ 6 , 25 ]. In our cohort, lifestyle modification was the most frequent initial intervention: 62.3% of couples received a recommendation to adapt at least one lifestyle factor, with most cases concerning smoking cessation (59%). The detrimental impact of tobacco use on spermatogenesis and male fertility is well recognized [ 26 ]; tobacco use is also associated with embryonic aneuploidy [ 27 ]. Although adherence was difficult to assess, about half of the couples reportedly implemented the recommendations. Interestingly, the proportion of male smokers (initially) was higher among those who ultimately achieved an ongoing pregnancy leading to a live birth. One possible explanation for this positive result is that male partners in this group successfully quit smoking following cessation counseling. The recommended diagnostic evaluation for RPL includes numerous investigations with varying levels of confirmed utility [ 6 , 7 , 25 ]. According to the 2022 ESHRE guidelines, first-line evaluation should include aPL screening, thyroid function assessment, and uterine cavity exploration [ 6 ], with the aim of identifying treatable causes. Hysteroscopic or surgical interventions may be considered when uterine anomalies are revealed, such as septate uterus, cavity-distorting fibroids, hydrosalpinx, tubal pathology, intrauterine adhesions, or endometriosis. However, to date, only septum resection is clearly linked to improved live birth rates [ 6 , 25 ]. Other surgical procedures – such as cervical cerclage, ovarian drilling, or neosalpingostomy – may be considered in specific contexts to optimize ART outcomes [ 28 ]. Anticoagulant therapy is often recommended in cases of RPL, regardless of aPL status. In our cohort, no patient was diagnosed with obstetric aPL syndrome. Nevertheless, 36.5% of women received low-dose aspirin, and 27.1% were prophylactically treated with low-molecular-weight heparin. Neither treatment had a measurable impact on live births. This result add to the debate around the role of thrombophilia, particularly inherited forms, in RPL. Indeed, the efficacy of antiplatelet or anticoagulant therapy has not been demonstrated in either unexplained or thrombophilia-associated RPL [ 29 – 31 ]. A recent meta-analysis of studies involving a total of 1,155 women with ≥ 3 RPL before 24 WG found an overall thrombophilia prevalence of 9.2% (1% acquired thrombophilia and 8.1% hereditary thrombophilia, mainly factor V Leiden and prothrombin gene mutations), with only 0.5% positivity for each type of aPL [ 29 ]. These findings indicate that thrombophilia prevalence in RPL is similar to that of the general population, and suggest that the value of systematic screening is limited. APS diagnostic criteria may therefore need to be revised. Similarly, anticoagulant treatments are frequently prescribed in cases of implantation failure despite a lack of evidence that they improve live birth rates [ 14 ]. The use of ART in women with RPL but no other cause of infertility remains debated, largely due to costs, emotional burden, and delayed access [ 32 , 33 ]. In our cohort, all women without infertility who achieved a pregnancy leading to a live birth following IVF had a clear underlying abnormality justifying recourse to the procedure. However, most cases of RPL remain unexplained, with embryonic aneuploidy being the predominant cause of both sporadic and recurrent miscarriage [ 10 ]. Although aneuploidy can be detected by cytogenetic analysis of the products of conception [ 11 ], the clinical utility of this information remains uncertain and routine implementation of testing is challenging in practice [ 6 , 7 , 25 ]. Several factors — such as parental chromosomal abnormalities, DOR, and high sperm DNA fragmentation — are associated with an increased risk of embryonic aneuploidy [ 24 , 34 ]. For couples having difficulty conceiving, ovarian reserve assessment is of particular interest, representing a point of convergence between RPL and infertility. Jaswa et al [ 24 ]. demonstrated that women with DOR had consistently lower euploid blastocyst rates compared to women with normal ovarian reserves. This difference was independent of age, suggesting DOR was linked to both oocyte quantity and quality. To avoid confusion due to the various definitions of DOR that exist in the literature [ 35 ], we have specified the precise AMH, FSH, and AFC values used here. We found that DOR was more frequent among infertile women and was associated with failure to achieve an ongoing pregnancy leading to a live birth. Similarly, Murugappan et al . reported that an AMH level < 1 ng/mL in women with ≥ 2 miscarriages before 20 WG was associated with a reduced probability of achieving a live birth (OR 0.38 [0.16–0.87]) [ 36 ]. Women with DOR have poorer ovarian responses to stimulation, lower overall pregnancy rates, and consequently, less favorable ART outcomes compared with age-matched women with normal reserves [ 37 ]. Nevertheless, the ESHRE does not currently recommend systematic screening for DOR in women experiencing RPL without infertility, and emphasizes its prognostic rather than its therapeutic value [ 6 ]. Preimplantation genetic testing for aneuploidy (PGT-A) is a subject of controversy in RPL. This technique allows chromosomal assessment of embryos obtained through IVF, allowing the transfer of euploid embryos [ 6 , 25 , 38 ]. In France, PGT is tightly regulated for ethical reasons and restricted to specific indications, primarily the prevention of heritable genetic diseases [ 39 ]. Thus, PGT-A can only be legally authorized in a context of RPL when some form of postnatal constitutional chromosomal abnormality has been identified in one of the partners during genetic counseling [ 7 ]. In other countries, such as Spain, PGT-A may be offered to women with unexplained RPL and normal karyotypes. A recent meta-analysis suggested that PGT-A could help to reduce miscarriage rates and improve rates of live births per transfer [ 38 ]. However, individual studies indicate that embryonic aneuploidy rates are not systematically higher in women with unexplained RPL, and that euploid embryo transfers yield comparable pregnancy and live birth rates to those achieved in women without RPL. Nevertheless, these findings must be interpreted with care, and well-designed randomized trials comparing ART with PGT-A versus expectant management are warranted. In these studies, patients should be stratified by maternal age, cytogenetic findings, and ovarian reserve status [ 38 ]. Current ESHRE guidelines [ 6 ] do not recommend PGT-A for unexplained RPL, and the 2023 RCOG update [ 25 ] similarly concludes that there is insufficient evidence to support its routine use, given the associated costs and risks. Our study has several limitations. The relatively small sample size limited its statistical power and robustness. The retrospective design restricted data accuracy, particularly regarding gestational age at miscarriage. We included women who had had at least three miscarriages, although in some cases losses involved a previous partner, meaning that a few couples had had fewer than three miscarriages at the time of recruitment. In the main analysis, infertile and non-infertile women were analyzed together, despite being clinically distinct. Like in most studies of the population who experience RPL, we were unable to provide a comparison with a control group. Such comparisons remain challenging due to the heterogeneity and the multiple underlying or unexplained factors associated with RPL. In conclusion, approximately three-quarters of women with a history of RPL achieved an ongoing pregnancy leading to a live birth, most after spontaneous conception. The likelihood of achieving such an outcome was higher when infertility was not an issue, and appeared closely linked to the ovarian reserve in terms of both quantity and quality. A history of RPL warrants comprehensive evaluation, including targeted diagnostic testing, lifestyle optimization, psychological support, and clear counseling and reassurance for the couple. ART should be reserved for couples with confirmed infertility or another clearly identified anomaly that justifies its use.

Introduction

Successful reproduction requires a complex interplay between a competent embryo and an endometrium that is both receptive and selective [ 1 – 3 ]. Pregnancy loss, or miscarriage, is defined as the spontaneous termination of a pregnancy prior to fetal viability. It affects approximately 15% of clinically recognized pregnancies, with the majority occurring during the first trimester [ 4 – 6 ]. Early gestational failure is most commonly due to impaired implantation, which may be influenced by genetic, anatomical, hormonal, hematologic, or immunological factors [ 1 , 4 , 5 ]. In 2014, the French National College of Gynecologists and Obstetricians (CNGOF) defined recurrent pregnancy loss (RPL) as three or more pregnancy losses before 14 weeks of gestation (WG) [ 7 ]. Since 2022, the European Society of Human Reproduction and Embryology (ESHRE) recommends that RPL be diagnosed after two or more miscarriages before 22 WG [ 6 ]. RPL affects 2–3% of couples attempting to conceive [ 4 – 6 ]. The principal risk factors are advanced maternal age and the number of prior miscarriages [ 6 , 8 , 9 ]. Although embryonic chromosomal abnormalities represent a major etiology, their prevalence is lower in RPL than in sporadic pregnancy losses, and a substantial proportion of incidences cannot be attributed to genetic causes alone [ 10 , 11 ]. Additional contributing factors include lifestyle, antiphospholipid syndrome (APS), hormonal or metabolic disorders, uterine abnormalities, and sperm-related factors [ 4 – 6 ]. Nevertheless, approximately half of RPL cases remain unexplained. Furthermore, even when an etiology is identified, few targeted interventions—such as optimization of diabetes control, levothyroxine therapy for hypothyroidism, or low-dose aspirin combined with prophylactic low-molecular-weight heparin (LMWH) for obstetric APS—have demonstrably improved live birth rates [ 4 – 6 ]. RPL constitutes a veritable global public health concern [ 5 , 6 ] associated with profound psychological consequences affecting emotional well-being and stability of couples. It also entails substantial economic costs related to diagnostic and therapeutic interventions [ 5 ]. Couples with RPL are frequently referred for assisted reproductive technologies (ART), despite limited evidence of efficacy and an absence of formal recommendations. This referral is therefore not always helpful. In this study, we aimed to determine the live birth rate and potentially associated clinical, demographic, etiological, and therapeutic factors among women with a history of RPL referred to a Reproductive Medicine and Biology department at a French university hospital. Patients were followed-up for a minimum of 2 years.

Supplementary Material

Supplementary Material 1. Supplementary Material 1.

Text is read by the "Ask this paper" AI Q&A widget below. Extraction quality varies by source — PMC NXML preserves structure cleanly, OA-HTML may include some navigation residue, and OA-PDF can have broken hyphenation. The publisher copy (via DOI) is the canonical version.

My notes (saved in your browser only)

⚙ Ask this paper AI returns verbatim quotes from the full text · source: pmc-nxml ⓘ

Answers must be backed by verbatim quotes from this paper's full text. Hallucinated quotes are dropped automatically; if no verbatim passage answers the question, we say so. How this works

Citation neighborhood (no data yet)

We don't have any in-corpus citations linked to this paper yet. This is a recent paper (2026) — citers typically take a year or two to land, and the OpenAlex reference graph may still be filling in.

Source provenance

europepmc
last seen: 2026-11-15T09:34:25.365398+00:00