The impact of maternal chronic viral hepatitis infections on pregnancy and neonatal outcomes in in vitro fertilization: a systematic review and meta-analysis.

The systematic review identified 385 records through database searching, with 14 studies ultimately enrolled after screening (Fig. 1 ) [ 21 ]. Full-text exclusion rationales were detailed in Table S3. Of these, 12 studies investigated maternal HBV infection impacts [ 13 , 22 – 32 ] and 3 examined HCV infection outcomes [ 14 , 31 ]. Table 1 summarizes study characteristics, including twelve conducted in China [ 13 , 22 – 26 , 28 – 33 ], one in Egypt [ 14 ], and one in Slovenia [ 27 ], spanning enrollment periods from 2004 to 2023. The analysis incorporated five case-control studies [ 14 , 22 , 23 , 25 , 31 ] and nine retrospective cohorts [ 13 , 24 , 26 – 30 , 33 , 34 ], comprising 6,195 HBV-infected cases versus 56,662 HBV-uninfected controls, and 283 HCV-infected cases versus 1,748 HCV-uninfected controls. Quality assessment using the revealed two studies with fair quality [ 14 , 22 ] and the other demonstrating high quality based on NOS scores (Additional file 1: Table S4). Fig. 1 PRISMA 2020 flow diagram for new systematic reviews, including searches of databases PRISMA 2020 flow diagram for new systematic reviews, including searches of databases Table 1 Characteristics of enrolled studies Study Area Year study conducted Virus Design Case Control Inclusion criteria Exclusion criteria Adjusted covariables Hanafi et al. 2011 [ 14 ] Egypt Jan. 2004 to Jan. 2008 HCV Case-control 80 40 Women undergoing ICSI. Not mentioned. None Feng et al. 2014 (Ye et al., 2014) [ 22 ] China Feb. 2007 to Dec. 2008 HBV Case-control 38 84 Women receiving ART. Not mentioned. None Lin et al. 2014 (Shi et al., 2014) [ 23 ] China Dec. 2008 to June 2012 HBV Case-control 77 154 First ART cycle of all infertile women. (1) Seropositivity for HCV, HIV, or syphilis; (2) Acute hepatitis diagnosis; (3) Prior antiviral therapy before IVF; (4) Cancelled cycles due to unavailable embryos or ovarian hyperstimulation syndrome. None Lin et al. 2015 (Yang et al., 2015) [ 24 ] China Jan. 2008 to Dec. 2013 HCV Cohort 90 1,256 Women undergoing IVF. (1) Abnormal liver function; (2) Chronic hepatitis diagnosis; (3) Ongoing antiviral treatment. None Shengli et al. 2015 (Lin et al., 2015) [ 33 ] China 2008 to 2012 HBV Cohort 90 199 (1) Women undergoing IVF; (2) Age ≤ 40 years; (3) BMI < 30 kg/m²; (4) Singleton live birth. PGD received. None Linlin et al. 2019 (Wang et al., 2019) [ 25 ] China Jan. 2010 to Apr. 2018 HBV Case-control 894 7,656 First IVF cycle of all infertile women. (1) Missing HBV serostatus; (2) Coinfection with HCV/HIV/syphilis; (3) Age > 38 years; (4) Antagonist protocol treatment; (5) Cycles missing embryo information; (6) Ovarian hyperstimulation syndrome, (7) Missing outcome data. None Yiquan et al. 2021 (Xiong et al., 2021) [ 26 ] China Jan.2014 to Mar. 2019 HBV Cohort 795 6,216 Pregnancies achieved by ART treatment. (1) Gestation > 28 weeks at first antenatal visit; (2) Missing HBsAg serostatus data at first visit, (3) Syphilis or HIV infection; (4) Chronic hypertension, diabetes mellitus, psychosis, hematopathy, epilepsy, hyperthyroidism, hypothyroidism. Female age, multiple gestations, education level, maternal pre-pregnancy BMI and parity. Liu et al. 2022 [ 13 ] China Oct. 2014 to Aug. 2019 HBV Cohort 546 1,024 First ART cycle of all infertile women. (1) Seropositivity for HCV, HIV, or syphilis; (2) Acute hepatitis diagnosis; (3) Significant liver dysfunction; (4) Prior antiviral therapy; (5) Chromosomal abnormalities in either partner; (6) Female history of endometriosis or ovarian surgery; (7) Male partner with asthenospermia or azoospermia; (8) HBV-seropositive male partner. None Petric et al. 2022 [ 27 ] Slovenia Jan. 2011 to Dec. 2019 HBV Cohort 117 10,216 Women receiving ART. (1) Spontaneous cycles; (2) Vitrified/warmed oocytes; (3) Cryopreserved ejaculated semen; (4) TESE-derived spermatozoa, PGT. None Wei et al. 2022 (Yi et al., 2022) [ 28 ] China Oct. 2010 to Mar. 2019 HBV Cohort 224 74 Women undergoing IVF and having delivery. (1) Amniocentesis during pregnancy; (2) Syphilis/HIV/HCV/TORCH coinfection; (3) Family history of congenital fetal malformations; (4) Neoplastic diseases in either partner; (5) Incomplete neonatal Hepatitis B vaccine/HBIG vaccination post-birth for HBV-exposed neonates. None Ling-Ling et al. 2023 (Ruan et al., 2023) [ 29 ] China Jan. 2018 to Apr. 2021 HBV Cohort 793 3,165 First FET cycle of all infertile women. (1) Abnormal karyotyping, (2) HCV, HIV, syphilis infection; (3) History of recurrent abortion or implantation failure; (4) Missing laboratory data; (5) Incomplete live birth records; (6) Missing HBV serostatus. None Ning-Zhao et al. 2023 (Ma et al., 2023) [ 30 ] China Jan. 2014 to Jan. 2019 HBV Cohort 910 19,561 (1) First ART cycle of all infertile women; (2) Age < 38 years; (3) Normal liver function; (4) Controlled ovarian stimulation via GnRH-a long protocol or EFLL protocol. (1) PGS or PGD; (2) HCV/HIV/syphilis coinfection; (3) No embryo transfer or use of donor sperm/oocytes; (4) Incomplete laboratory data or lost to follow-up. Female age, BMI, FSH, AMH, duration of infertility and the E2 and endometrial thickness Fang et al. 2025 (Liu et al., 2025) [ 31 ] China Jan. 2011 to Dec. 2019 HBV Case-control 1,343 5,372 Women undergoing IVF with documented infection history. (1) PGT; (2) Incomplete laboratory data. BMI, duration of infertility, infertility type, cause of infertility, number of total AFC, PCOS, endometrial thickness, fertilization method, day of embryo transfer, quality of the sperm, and number of oocytes retrieved. HCV 113 452 Kaiwen et al. 2025 (Xue et al., 2025) [ 34 ] China June 2012 to Dec. 2023 HBV Cohort 368 2,941 Women receiving ART and undergoing TORCH screening and testing. (1) Chromosomal abnormalities; (2) Varicocele; (3) Long-term medication use; (4) HCV/HIV seropositivity; (5) Surgery or congenital defects (urinary/genital organ-related); (6) Acute hepatitis diagnosis; (7) Antiviral therapy received. None HBV hepatitis B virus, HCV hepatitis C virus, HIV human immunodeficiency virus, PGD pre-implantation genetic diagnosis, PGT pre-implantation genetic testing, PGS preimplantation genetic screening, HBsAg hepatitis B surface antigen, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection, ART assisted reproductive technology, TSES testicular sperm extraction, TORCH Toxoplasma - Others- Rubella.Virus- Cytomegalo.Virus- Herpes.Virus, HBIG hepatitis B immune globulin, BMI body mass index, EFLL early-follicular phase long-acting GnRH agonist long protocol Characteristics of enrolled studies (1) Seropositivity for HCV, HIV, or syphilis; (2) Acute hepatitis diagnosis; (3) Prior antiviral therapy before IVF; (4) Cancelled cycles due to unavailable embryos or ovarian hyperstimulation syndrome. (1) Abnormal liver function; (2) Chronic hepatitis diagnosis; (3) Ongoing antiviral treatment. (1) Women undergoing IVF; (2) Age ≤ 40 years; (3) BMI  38 years; (4) Antagonist protocol treatment; (5) Cycles missing embryo information; (6) Ovarian hyperstimulation syndrome, (7) Missing outcome data. (1) Gestation > 28 weeks at first antenatal visit; (2) Missing HBsAg serostatus data at first visit, (3) Syphilis or HIV infection; (4) Chronic hypertension, diabetes mellitus, psychosis, hematopathy, epilepsy, hyperthyroidism, hypothyroidism. (1) Seropositivity for HCV, HIV, or syphilis; (2) Acute hepatitis diagnosis; (3) Significant liver dysfunction; (4) Prior antiviral therapy; (5) Chromosomal abnormalities in either partner; (6) Female history of endometriosis or ovarian surgery; (7) Male partner with asthenospermia or azoospermia; (8) HBV-seropositive male partner. (1) Spontaneous cycles; (2) Vitrified/warmed oocytes; (3) Cryopreserved ejaculated semen; (4) TESE-derived spermatozoa, PGT. (1) Amniocentesis during pregnancy; (2) Syphilis/HIV/HCV/TORCH coinfection; (3) Family history of congenital fetal malformations; (4) Neoplastic diseases in either partner; (5) Incomplete neonatal Hepatitis B vaccine/HBIG vaccination post-birth for HBV-exposed neonates. (1) Abnormal karyotyping, (2) HCV, HIV, syphilis infection; (3) History of recurrent abortion or implantation failure; (4) Missing laboratory data; (5) Incomplete live birth records; (6) Missing HBV serostatus. (1) First ART cycle of all infertile women; (2) Age < 38 years; (3) Normal liver function; (4) Controlled ovarian stimulation via GnRH-a long protocol or EFLL protocol. (1) PGS or PGD; (2) HCV/HIV/syphilis coinfection; (3) No embryo transfer or use of donor sperm/oocytes; (4) Incomplete laboratory data or lost to follow-up. (1) PGT; (2) Incomplete laboratory data. (1) Chromosomal abnormalities; (2) Varicocele; (3) Long-term medication use; (4) HCV/HIV seropositivity; (5) Surgery or congenital defects (urinary/genital organ-related); (6) Acute hepatitis diagnosis; (7) Antiviral therapy received. HBV hepatitis B virus, HCV hepatitis C virus, HIV human immunodeficiency virus, PGD pre-implantation genetic diagnosis, PGT pre-implantation genetic testing, PGS preimplantation genetic screening, HBsAg hepatitis B surface antigen, IVF in vitro fertilization, ICSI intracytoplasmic sperm injection, ART assisted reproductive technology, TSES testicular sperm extraction, TORCH Toxoplasma - Others- Rubella.Virus- Cytomegalo.Virus- Herpes.Virus, HBIG hepatitis B immune globulin, BMI body mass index, EFLL early-follicular phase long-acting GnRH agonist long protocol Meta-analysis of 12 studies demonstrated no significant association between maternal HBV infections and pregnancy outcomes: live birth (OR: 1.017, 95%CI: 0.936–1.105, 6 studies, I²=0), clinical pregnancy (OR: 0.974, 95%CI: 0.863-1.100, 8 studies, I²=35.8%), implantation rates (OR: 0.867, 95%CI: 0.720–1.045, 4 studies, I²=89.2%), miscarriage (OR: 1.025, 95%CI: 0.822–1.277, 5 studies, I²=50.6%), early abortion (OR: 1.470, 95%CI: 0.743–2.908, 4 studies, I²=80.6%), ectopic pregnancy (OR: 1.097, 95%CI: 0.675–1.784, 2 studies, I²=0), LBW (OR: 1.118, 95%CI: 0.969–1.291, 3 studies, I²=0), and multiple pregnancy (OR: 0.985, 95%CI: 0.885–1.096, 4 studies, I²=0) (Fig.  2 ). There was an increased risk of preterm delivery among HBV-infected mothers (OR: 1.159, 95%CI: 1.022–1.315, 5 studies, I²=0), while GDM (OR: 0.975, 95%CI: 0.683–1.392, 2 studies, I²=68.4%), and caesarean section (OR: 1.041, 95%CI: 0.834–1.299, 2 studies, I²=52.2%) showed no significant associations (Fig.  3 ). None of the research reported ongoing pregnancy rates. Fig. 2 Forest plots for pregnancy outcomes in women with or without HBV infections. a , live birth; b , clinical pregnancy; c , miscarriage; d , implantation rates; e , early abortion; f , ectopic pregnancy; g , multiple pregnancy Forest plots for pregnancy outcomes in women with or without HBV infections. a , live birth; b , clinical pregnancy; c , miscarriage; d , implantation rates; e , early abortion; f , ectopic pregnancy; g , multiple pregnancy Fig. 3 Forest plots for other outcomes in women with or without HBV infections. a , GDM; b , preterm delivery; c , LBW; d , caesarean section. GDM, gestational diabetes mellitus; LBW, low birth weight Forest plots for other outcomes in women with or without HBV infections. a , GDM; b , preterm delivery; c , LBW; d , caesarean section. GDM, gestational diabetes mellitus; LBW, low birth weight Subgroup analysis stratified by study design demonstrated robustness in cohort studies, though case-control studies unexpectedly showed reduced implantation rates (OR: 0.741, 95%CI: 0.674–0.814, 2 studies, I²=0) (Figs.  2 and 3 ). Sensitivity analyses confirmed the robustness of the primary outcomes. The results remained consistent across all iterations, including: (1) leave-one-out cross-validation, where no single study’s exclusion significantly altered the effect estimates (Additional file 1: Fig. S1); (2) the removal of the sole low- or moderate-quality study (NOS score < 7) (Additional file 1: Fig. S2); and (3) the exclusion of studies with outcome definitions inconsistent with this study (Additional file 1: Fig. S3). Sensitivity analysis for preterm delivery using the leave-one-out method showed that the effect estimates remained stable. However, the statistical significance was lost after excluding Yiquan et al. ( p  = 0.110) or Ning-Zhao et al. ( p  = 0.083), suggesting that the statistical power was partly driven by these large-scale studies (Additional file 1: Fig. Sxx). Publication bias assessment revealed no significant asymmetry (Additional file 1: Table S5). Meta-analysis synthesizing 3 studies investigating the impact of maternal HCV infections demonstrated no significant associations between HCV seropositivity and live birth (OR: 0.618, 95%CI: 0.177–2.161, 2 studies, I²=77.1%), clinical pregnancy (OR: 0.585, 95%CI: 0.276–1.241, 3 studies, I²=68.0%), or miscarriage (OR: 1.228, 95%CI: 0.747–2.018, 2 studies, I²=0) (Fig.  4 ). None of the research reported ongoing pregnancy rates. Sensitivity analysis confirmed the robustness of primary outcomes, demonstrating stable effect estimates across all iterations of leave-one-out meta-analytic models, with no single study exclusion altering statistical significance thresholds (Additional file 1: Fig. S4). Publication bias analyses revealed no significant asymmetry through Begg’s test and Egger’s test (Additional file 1: Table S4). After the removal of the one low- or moderate-quality study, the results remained consistent (Additional file 1: Fig. S3). Fig. 4 Forest plots for primary outcomes in women with or without HCV infections. a , live birth; b , clinical pregnancy; c , miscarriage Forest plots for primary outcomes in women with or without HCV infections. a , live birth; b , clinical pregnancy; c , miscarriage

This systematic review protocol was prospectively registered with PROSPERO (CRD420251079149) and conducted in accordance with PRISMA guidelines [ 17 ]. HBV infection was defined as seropositivity for hepatitis B surface antigen (HBsAg) and/or detectable HBV-DNA, while HCV infection was confirmed through anti-HCV antibody seropositivity with recombinant immunoblot assay confirmation or detectable HCV-RNA. A comprehensive literature search was conducted across PubMed, Embase, Cochrane Library, and Web of Science Core Collection databases, with search strategies and results fully documented in Additional file 1: Table S1. The inclusion criteria were restricted to English-language publications up to June 23, 2025, without geographic limitations. Eligible study designs included cohort and case-control studies, while conference reports were excluded. Reference lists of selected articles were systematically screened to identify additional relevant studies that might have been missed in the initial search. Studies were excluded if they did not involve a standard IVF population. This included studies on oocyte donation recipients, in vitro maturation (IVM), or cycles utilizing blastocyst biopsy for preimplantation genetic testing (PGT). We also excluded patients with a diagnosis of congenital or acquired uterine abnormalities, or with uncontrolled hepatitis B or C during the IVF cycle and subsequent pregnancy. Primary outcomes were defined as ongoing pregnancy beyond 12 weeks, live birth (delivery of ≥ 1 viable infant after 24 gestational weeks), and clinical pregnancy (ultrasound-confirmed gestational sac with cardiac activity at 7 weeks). Secondary outcomes included implantation rate (number of intrauterine gestational sacs per total number of transferred embryos), miscarriage (pregnancy loss before 24 weeks), early abortion (< 12 weeks), multiple pregnancy (≥ 2 gestational sacs), gestational diabetes mellitus (GDM), ectopic pregnancy, preterm delivery (< 37 weeks), low birth weight (LBW, < 2500 g), and caesarean section rates. Two independent investigators (Di Mao and Wenhui Nan) performed study selection, with discrepancies resolved through consensus discussions or third-party arbitration (Kai-Lun Hu). Data extraction encompassed first author’s name, publication year, year study conducted, country of origin, sample size, study population, study design, adjusted covariables, outcome measures, and regression analysis results. Two investigators (Di Mao and Wenhui Nan) independently extracted data, which were subsequently validated by Kai-Lun Hu. Pregnancy and neonatal outcomes were quantified separately for case and control groups. A summary list of outcome definitions in included studies was presented in Additional file 1: Table S2. Adjusted odds ratios (OR) from multivariable analyses were preferentially utilized in meta-analyses. The selection between fixed- and random-effects models was based on study homogeneity; fixed-effects models were applied to subgroups with similar populations and designs, while random-effects models were used for more heterogeneous groups (e.g., those containing both case-control and cohort studies).Subgroup analyses were restricted to cohort or case-control studies to assess robustness. For outcomes reported in ≥ 3 studies, we assessed potential publication bias through Begg’s rank correlation test and Egger’s linear regression test [ 18 , 19 ], with statistical significance threshold set at p < 0.05. Sensitivity analyses were conducted employing three methods: (1) leave-one-out cross-validation, (2) exclusion of low- or moderate-quality studies (NOS score < 7), and (3) exclusion of studies whose outcome definitions were inconsistent with those of the present study. Methodological quality was evaluated using the Newcastle-Ottawa Scale (NOS) by two independent reviewers (Di Mao and Wenhui Nan) [ 20 ]. All quality assessments underwent secondary review by Kai-Lun Hu to resolve discrepancies and ensure consistency in scoring criteria application.

Our meta-analysis incorporated 12 studies examining maternal HBV infections and 3 studies addressing HCV infections. Neither HBV nor HCV infection demonstrated statistically significant associations with live birth, or clinical pregnancy. Notably, maternal HBV infection was associated with an increased risk of preterm delivery, while no significant correlations emerged for other adverse pregnancy outcomes or neonatal complications. These associations persisted in sensitivity analyses. A distinctive strength of this study lies in its rigorous differentiation between maternal and paternal infections, a critical distinction often obscured in prior analyses on parental infection [ 15 , 16 ]. This stratification is biologically imperative: while paternal infection primarily impacts reproductive outcomes via compromised sperm DNA integrity and fertilization potential [ 35 , 36 ], maternal infection might exerts a far more complex influence, encompassing oocyte quality, the systemic inflammatory milieu, and the delicate immune tolerance at the maternal-fetal interface. By specifically isolating the maternal phenotype, our meta-analysis addresses this substantial methodological gap, providing precise evidence that maternal HBV infection does not compromise live birth rates following IVF. While our results align with the majority of included studies regarding live birth outcomes, they diverge from the findings of Liu et al., who reported significantly lower clinical pregnancy rates in HBV-positive women [ 13 ]. This discrepancy, however, warrants cautious interpretation due to methodological limitations in the primary data. Notably, the analysis by Liu et al. did not adjust for critical ovarian reserve markers, such as AFC and AMH. Given that these variables exhibited marked intergroup disparities in our own cohort and are established independent predictors of IVF success, their omission may have introduced residual confounding, potentially biasing their observed associations.In naturally conceived pregnancies, maternal HBV infection has been postulated to be associated with increased risks of adverse obstetric outcomes, including GDM [ 37 , 38 ], intrahepatic cholestasis of pregnancy (ICP) [ 39 ], antepartum hemorrhage [ 40 ], LBW [ 41 ], preterm delivery [ 37 ], and elevated maternal and fetal morbidity in women with cirrhosis [ 42 ]. However, evidence regarding these associations remains inconsistent. For instance, while some studies suggest elevated risks, others have reported conflicting findings: Jiang et al. observed a similar incidence of overall preterm delivery but an increased risk of early preterm delivery [ 43 ]; Cong et al. found no significant difference in preterm delivery rates [ 37 ]; and Chen et al. even reported a lower incidence of preterm delivery alongside similar rates of antepartum/intrapartum hemorrhage [ 38 ]. Similarly, Weng et al. reported a decreased incidence of antepartum hemorrhage [ 41 ]. In contrast to these heterogeneous findings in the general obstetric population, our meta-analysis in the IVF population identified no significant associations between HBV infection and GDM or other metabolic complications, with the notable exception of preterm delivery. This discrepancy may be attributed to the inherent selection bias within the IVF population. Unlike the general population, infertile women undergo rigorous pre-conception screening. Candidates exhibiting significant liver dysfunction or active viral replication are typically advised to defer IVF cycle until their clinical status stabilizes. Consequently, the enrolled IVF cohort likely represents a subset of HBV carriers with well-preserved hepatic function, which may attenuate the risks of liver-related metabolic complications such as GDM. This distinct clinical profile highlights the necessity of differentiating IVF cohorts from the general obstetric population when interpreting the risks associated with viral hepatitis. Regarding maternal HCV infection, our analysis revealed no significant correlations with either live birth, or clinical pregnancy rates. However, these conclusions should be interpreted cautiously given the limited number of included studies (only two and three studies, respectively). Notably, Hanafi’s investigation lacked transparency in case group selection criteria and failed to adjust for confounding variables, compromising the reliability of their findings [ 14 ]. Our meta-analysis demonstrates that maternal HBV or HCV infections do not significantly compromise live birth, clinical pregnancy, or most pregnancy and neonatal outcomes in IVF populations. Furthermore, evidence from a long-term follow-up study indicated that the HBsAg presence in oocytes and embryos may not induce vertical HBV transmission in carriers’ offspring in IVF populations [ 44 ]. Consequently, supplementary interventions for HBV/HCV-positive women appear clinically unjustified. This conclusion is particularly relevant given the substantial psychological burdens associated with ART treatment. A meta-analysis indicates that 50% of infertile couples experience some degree of mild, moderate, or severe depression [ 45 ], while an international online survey revealed that 55% of respondents strongly agreed that infertility causes emotional strain [ 46 ]. In addition, seropositive women who are seeking for IVF treatment may endure heightened distress due to concerns related to their infections. Our findings help alleviate this psychological burden by demonstrating the limited clinical impact of maternal infections. Collectively, these insights advocate for minimizing non-routine clinical measures, which may reduce financial expenditures and psychological stress while optimizing perinatal outcomes and healthcare resource utilization. The association between HBV infections and GDM remains controversial. While no increased GDM risk in ART-treated women with HBV was found in the population-based study [ 26 ], multiple investigations have identified HBsAg positivity as an independent GDM risk factor [ 47 – 49 ]. Notably, this relationship appears attenuated in IVF populations, potentially due to stringent liver function requirements for infertility treatment candidates. However, limited sample sizes and few dedicated studies preclude definitive conclusions. Future research should examine how liver function parameters modulate GDM risk in HBV-positive IVF patients. Although our study suggests no increased risk of GDM in HBV-positive IVF women, the potential modulation of pregnancy outcomes by liver function parameters remains an area for future investigation. Current studies often lack granular data on alanine aminotransferase (ALT), aspartate transaminase (AST) and bile acids. Future research should prioritize large-scale prospective cohorts that include detailed liver function panels to definitively exclude any subtle impacts of hepatic metabolic alterations on GDM and other pregnancy complications in the ART setting. Our meta-analysis identified a significant association between maternal HBV infection and preterm delivery, indicating a 15.9% higher odds among infected women. In terms of absolute risk, given a baseline preterm birth rate of approximately 12% in IVF/ICSI populations [ 50 ], this corresponds to a modest excess risk of about 1–2 additional cases per 100 pregnancies. Clinically, while this magnitude of effect does not necessitate aggressive interventions or contraindicate ART, it establishes HBV infection as a relevant risk marker that warrants enhanced obstetric surveillance. It is noteworthy that while this association was directionally consistent across studies (I 2 = 0), our leave-one-out sensitivity analysis revealed that statistical significance relied on the inclusion of the largest cohorts (Yiquan et al. and Ning-Zhao et al.). This suggests that the effect size is modest and requires substantial statistical power to detect, though the potential clinical risk should not be overlooked. This relationship may be mediated by placental inflammatory responses, as HBV DNA has been detected in placental tissue and trophoblast cells of HBV-infected women [ 51 , 52 ]. The HBV target the CD 133-negative cells to activate the Smad signalling and induce epithelial-mesenchymal transition (EMT), which eventually lead to Lower proliferation and mobility, and higher apoptosis in trophoblast [ 53 ]. HbxAg also may promote HBV replication in trophoblasts via downregulation of Smc5/6, activates the EGFR promoter and inhibits trophoblast apoptosis via the PI3K/p-AKT downstream signalling pathway, thereby increasing the risk of HBV intrauterine infection [ 54 ]. It was hypothesized that TLRs in trophoblasts are likely involved in the prevention of HBV intrauterine transmission via the recruitment of the effector protein MyD88 as well, which then triggers the activation of NF-kappa B and subsequently generates an inflammatory response against HBV [ 55 ]. With the abnormal development of trophoblast in early stage, the risks of preterm delivery increase [ 56 ]. Furthermore, although infrequent, maternal liver dysfunction, including ICP and hepatitis flares may necessitate medically indicated preterm delivery, which may also account for this association [ 57 ]. In addition, potential confounding factors, including antiviral medication use and socioeconomic disparities, may also contribute to preterm labor risk. Subsequent studies should adjust for these covariates to clarify causation, while basic research ought to elucidate the molecular mechanisms linking HBV infection to preterm birth pathophysiology. From a clinical perspective, these findings underscore the need for increased vigilance regarding the risk of preterm birth in HBV-infected pregnant women. Implementing a multidisciplinary care approach involving obstetrics, neonatology, and pediatrics is essential to optimize outcomes for potentially preterm infants. Our analysis identified only one study investigating the correlation between HBV-DNA viral load and pregnancy outcomes, which demonstrated no significant differences in live birth, clinical pregnancy, or miscarriage rates across varying HBV-DNA levels [ 13 ]. However, these findings should be interpreted with caution due to the limited sample size. Current evidence suggests HBV infections may adversely affect maternal-fetal immunotolerance through multiple pathways, including increased regulatory B-cell frequency, reduced CD3⁺CD4⁺ helper T-cell counts, and diminished peripheral natural killer (NK) cell activity [ 58 ]. The observed disruption of immune homeostasis at the maternal-fetal interface could potentially contribute to adverse pregnancy outcomes, though this relationship remains unclear in clinical studies. Notably, the most of current studies of defining cases solely based on HBsAg positivity without considering HBV-DNA quantification introduces substantial heterogeneity in case populations, potentially obscuring dose-dependent effects of viral replication. Future investigations should prioritize large-scale studies incorporating quantitative HBV-DNA stratification and standardized reporting of antiviral therapy status to clarify these associations. To overcome the limitations of current retrospective designs, future investigations must adopt rigorous multivariable adjustments. Specifically, subsequent studies should prioritize the collection and analysis of key potential confounders to clarify causation. These include comprehensive liver function parameters, including ALT, AST, and bile acids, and quantitative HBV-DNA viral loads, as distinguishing between inactive carriers and those with active viral replication is crucial for assessing fetal risk. Furthermore, the impact of antiviral therapy status must be stratified to decouple the physiological effects of the virus from the potential influence of pharmaceutical interventions. Finally, given the metabolic and reproductive complexities often present in IVF populations, rigorous adjustment for maternal metabolic profiles, such as pre-pregnancy BMI, PCOS status, and insulin resistance, as well as obstetric history including parity and history of adverse pregnancy outcomes, is imperative to exclude residual confounding and ensure precise risk estimation. Evidence regarding maternal HCV infection’s impact on pregnancy outcomes remains scarce and heterogeneous. The non-significant findings in this meta-analysis warrant cautious interpretation due to limited power. Rigorously designed prospective studies with standardized outcome definitions are needed to establish clinical associations. Additionally, current evidence remains insufficient to assess the potential impact of maternal HCV infections on neonatal outcomes in women undergoing IVF procedures, as no dedicated studies have specifically investigated this association to date. This critical knowledge gap underscores the necessity for prospective cohort studies with longitudinal follow-up to elucidate potential virological effects on perinatal health outcomes in this population. This meta-analysis was prospectively registered on PROSPERO and implemented a comprehensive search strategy with strictly defined case, control, and outcome criteria. Study screening and data extraction were conducted independently by two investigators. The analysis incorporated a substantial pooled population (HBV: 6,195 cases vs. 56,662 controls; HCV: 283 cases vs. 1,748 controls), with statistical conservatism ensured through random-effects models and comprehensive sensitivity analyses. Pre-specified subgroup analyses stratified by study design (cohort vs. case-control) and quality assessment (NOS scoring) confirmed the robustness of primary findings. Publication bias evaluation via Begg’s rank correlation and Egger’s regression tests revealed no significant asymmetry across outcomes. However, several limitations warrant consideration: First, all included studies utilized retrospective observational designs, limiting causal inference. Furthermore, substantial heterogeneity was observed in secondary outcomes, particularly implantation and early abortion rates (I 2 > 80). This variability is likely attributable to inherent clinical and methodological diversity across the included studies. Regarding implantation rates, the heterogeneity appears driven by the evolution of ART protocols from cleavage-stage to blastocyst transfers over the decade-long study period. Similarly, divergence in early abortion rates reflects disparities in cycle types (fresh vs. frozen) as well as “small-study effects,” where earlier unadjusted studies (Feng et al.) reported higher risks than recent large-scale, multivariate-adjusted cohorts. Second, HCV-related conclusions are constrained by the scarcity of evidence ( n = 3 studies), particularly regarding neonatal outcomes. Consequently, the non-significant findings should be interpreted with caution, as the limited sample size may result in insufficient statistical power to detect subtle adverse effects; thus, non-significance does not definitively equate to no effect. Third, the absence of individual participant data precluded adjustment for critical confounding variables and procedural heterogeneity. This limitation is also particularly pertinent regarding the inability to distinguish between spontaneous and medically indicated preterm delivery. Furthermore, variations in ART protocols constitute a significant source of residual confounding. Fertilization methods (IVF vs. ICSI) and embryo transfer strategies (fresh vs. frozen) are known to independently influence obstetric outcomes; for instance, frozen transfers are often associated with higher birth weights compared to fresh cycles. In our meta-analysis, the majority of included studies combined these modalities without providing stratified data, which prevented specific subgroup analyses. Consequently, we could not fully dissect whether the observed associations were modulated by the ART procedure itself, underscoring the need for future studies to explicitly stratify outcomes by cycle type. Fourth, the geographic concentration of the included studies, predominantly from China (12/14), restricts generalizability to diverse populations. While this distribution aligns with global epidemiology given that China accounts for approximately one-third of the global HBV burden [ 59 ], it limits applicability to populations with different HBV genotypes and healthcare systems [ 60 ]. Notably, China has implemented rigorous public health strategies and standardized protocols for preventing mother-to-child transmission [ 2 , 61 , 62 ]. Consequently, the largely favorable outcomes observed in this meta-analysis may reflect this optimized clinical management, which may not fully extrapolate to regions with less stringent viral control or limited access to antiviral prophylaxis during ART. Therefore, future multi-center studies involving diverse ethnic populations are essential to validate these findings globally.

Neither maternal HBV nor HCV infection is associated with live birth, clinical pregnancy, and most of pregnancy and neonatal outcomes in IVF populations. Current evidence does not warrant routine clinical interventions specifically targeting HBV/HCV in infertile women undergoing IVF treatment.

Chronic hepatitis B (HBV) and hepatitis C (HCV) infections represent a substantial global health burden with prevalence at around 3.2% globally [ 1 ]. Notably, HBV prevalence among pregnant women reached 4.8% (95% CI: 3.8–5.8) [ 2 ]. Similarly, a global analysis of 98 studies ( n = 236,964) reported a prevalence of HCV infections at 1.8% (95% CI: 1.4–2.3) [ 3 ]. The Global Burden of Disease (GBD) 2019 study quantified mortality attributed to these infections, reporting 365,240 HBV-related deaths and 399,490 HCV-related deaths globally in 2019 [ 4 ]. Notably, apart from risks of progressive hepatic dysfunction, cirrhosis, and liver cancer, pregnant women with chronic hepatitis may have elevated risks of obstetric complications, including intrahepatic cholestasis of pregnancy, antepartum hemorrhage, gestational diabetes mellitus, fetal growth restriction, spontaneous abortion, and preterm delivery [ 5 , 6 ]. These findings underscore the significant healthcare burden imposed by viral hepatitis. Concurrently, infertility rates demonstrate concerning trends. Between 1990 and 2021, the global age-standardized prevalence rate increased annually by 0.49% (95% CI: 0.34–0.63) for males and 0.68% (95% CI: 0.51–0.86) for females [ 7 ]. As assisted reproductive technology (ART) advances, growing numbers of infertile patients with chronic HBV/HCV infections pursue pregnancy through in vitro fertilization (IVF). This convergence necessitates urgent investigation into how parental hepatitis infections affect ART pregnancy and neonatal outcomes. Accumulating evidence has demonstrated the presence of HBV and HCV in ovarian tissue and follicular fluid [ 8 – 12 ], establishing a biological plausibility for viral impacts on reproductive physiology. Specifically, HBV-infected patients exhibit impaired ovarian reserve, as evidenced by reduced antral follicle counts (AFC), fewer retrieved oocytes during controlled ovarian stimulation, and suppressed anti-Müllerian hormone (AMH) levels [ 13 ], and HCV-infected individuals experience higher rates of failed ART cycles [ 14 ]. A meta-analysis found no significant impact of parent HBV infection (one of the partners) on live birth rates (OR: 1.12, 95% CI: 0.88–1.41) or clinical pregnancy rates (OR: 1.25, 95% CI: 0.82–1.90) [ 15 ], though it did not stratify results by parental sex. In contrast, a 2024 analysis of male HBV infection demonstrated significantly reduced fertilization rates (OR: 0.86, 95% CI: 0.76–0.99) [ 16 ]. Crucially, no systematic review has specifically examined maternal HBV/HCV infections’ effects on pregnancy and neonatal outcomes in IVF populations. Therefore, this systematic review aims to synthesize evidence on maternal HBV and HCV infections’ effects on pregnancy and neonatal outcomes in patients undergoing IVF, addressing this knowledge gap to inform clinical practice and future research.

Supplementary Material 1. Supplementary Material 1.