Association between G6PD gene variants and adverse pregnancy outcomes.

The following data were collected for all neonates: G6PD gene variants, G6PD enzyme activity values, presence of jaundice, intrauterine distress, gestational age at birth, preterm birth status, and birth weight. For mothers, data collection included age, parity, history of adverse pregnancy (abortion, preterm birth), internal medical conditions (cytopenia, gestational diabetes mellitus, thyroid dysfunction, hypertensive disorders, and cholestasis), gynecological conditions (uterine adhesions, uterine fibroids, endometriosis, uterine malformations, placental abnormalities, cervical insufficiency, and chorioamnionitis), as well as G6PD gene variants. In accordance with the “Neonatal Disease Screening Management Measures” [ 23 ] issued by the National Ministry of Health in 2009, heel blood was collected from the lateral plantar surface of neonates 72 h after birth and applied onto dried filter paper cards (Whatman™ 903, General Type, Shanghai Jie’er Biotechnology Co., Ltd.) to produce dried blood spot specimens for G6PD enzyme activity measurement and genetic variant testing. For mothers, 2 mL of peripheral blood was obtained for G6PD variant testing. Preliminary screening of G6PD enzyme activity was conducted on neonatal dried blood spots using the G6PD fluorescence analysis kit from PerkinElmer Finland. Experimental procedures and result interpretations were performed strictly in accordance with the manufacturer’s instructions. A cutoff value of 2.5 U/mL was applied: samples with values > 2.5 U/mL were considered negative, while those with values ≤ 2.5 U/mL were considered positive. Dried blood spots that screened positive underwent G6PD genetic testing. According to the diagnostic criteria of our hospital, G6PD enzyme activity values were further classified into three grades: Grade 2 (positive) for values ≤ 2.0 U/mL, Grade 1 (weakly positive) for values > 2.0 U/mL and ≤ 2.5 U/mL, and Grade 0 (negative) for values > 2.5 U/mL. Genetic testing of initially screening-positive samples was conducted using the fluorescent PCR with a melting-curve assay and the G6PD gene variant detection kit (PerkinElmer). Each sample was analyzed through two PCR systems, each incorporating four-color fluorescent probes. The assay was designed to detect 12 G6PD gene variants commonly observed in the Chinese population (c.95 A > G, c.383T > C, c.392G > T, c.487G > A, c.517T > C, c.592 C > T, c.871G > A, c.1004 C > A, c.1024 C > T, c.1360 C > T, c.1376G > T, c.1388G > A) by assessing differences in the melting temperatures of target–probe hybridization products. Genotyping results were interpreted according to the manufacturer’s instructions. Statistical analyses were conducted using SPSS 24.0 software. Measurement data are presented as mean ± standard deviation ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\stackrel{̄}{x}\pm\:s$$\end{document} ) or median M (P 25 , P 75 ) and were analyzed using the t -test or rank-sum test. Enumeration data are expressed as percentages, and intergroup comparisons were performed using the chi-squared test or Fisher’s exact probability method. p  < 0.05 was considered statistically significant. This is a retrospective cohort study aimed at investigating the association between G6PD genotype and adverse pregnancy outcomes. Inclusion criteria: Male neonates born between August 2020 and August 2021 at Fujian Provincial Maternity and Children’s Hospital (hereinafter referred to as “our hospital”) who underwent neonatal screening, along with their mothers were consecutively selected as research participants, and were divided into two groups: (1) Study group ( G6PD variant group): 225 male neonates diagnosed with G6PD gene variants and their mothers; (2) Control group: 121 healthy male neonates without G6PD gene variants and their mothers. All participants are of Chinese ethnicity. Exclusion criteria: (1) neonates with chromosomal abnormalities, (2) fetuses with structural abnormalities, (3) multiple pregnancies, (4) mothers with cardiovascular or autoimmune diseases, (5) those with incomplete clinical data records. The study was conducted in accordance with the principles of the Declaration of Helsinki. Peripheral blood samples were collected from all participants after informed consent was obtained from the neonates’ guardians through the neonatal disease screening consent form. For neonates diagnosed with G6PD deficiency, a genetic testing consent form was additionally signed by their guardians. All clinical data were reviewed, and approval was granted by the Ethics Committee of Fujian Provincial Maternity and Children’s Hospital (Ethics Approval No.: 2021KR031).

A comparative analysis of the clinical data of 225 cases in the G6PD study group and 121 cases in the control group was conducted, as presented in Table  1 . The gestational age at birth of neonates in the G6PD study group was significantly lower than that of the control group ( p  < 0.05), whereas no significant difference was observed in birth weight between the groups. The median G6PD enzyme activity value in the G6PD study group was 1.20 U/mL, consistent with previously reported values, and was significantly lower than that of healthy neonates ( p  < 0.01). The rates of preterm birth and jaundice in the G6PD study group were 11.81 times and 3.07 times higher, respectively, than those in the control group, with statistically significant differences ( p  < 0.01). No significant difference was identified in the incidence of intrauterine distress between the two cohorts. Analysis of maternal factors demonstrated no significant differences in age, parity, history of abortion, internal medical conditions, or gynecological conditions between the groups. While, the history of preterm birth demonstrated a statistically significant difference between the two groups ( P  < 0.05). These findings indicate that G6PD gene variants may be associated with neonatal jaundice and preterm birth, as well as with a history of preterm birth in the mothers. Table 1 Comparative analysis of clinical data between the study group and the control group Object Clinical phenotype Study group ( n  = 225) Control group ( n  = 121) Statistics P Newborn Phenotype Gestational age (Week, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{\mathrm x}\pm\mathrm s$$\end{document} ) 38.97 ± 1.53 39.29 ± 1.00 2.07 a 0.039 Birth weight (g, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{\mathrm x}\pm\mathrm s$$\end{document} ) 3274.17 ± 463.26 3346.20 ± 363.40 1.46 a 0.15 G6PD enzyme activity value (U/mL , M (P25 , P75)) 1.20 (0.81 , 3.33) 57.2 (50.10 , 64.30) 15.20 b < 0.01 Preterm birth (% , (number of positive cases, number of negative cases)) 8.44% (19, 206) 0.83% (1, 120) 8.39 c 0.004 Jaundice (% , (number of positive cases, number of negative cases)) 50.7% (114, 111) 16.53% (20, 101) 38.64 c < 0.01 Intrauterine distress (% , (number of positive cases, number of negative cases)) 10.70% (24, 201) 9.92% (12, 109) 1.62 c 0.45 Maternal body Phenotype Maternal age (year, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{\mathrm x}\pm\mathrm s$$\end{document} ) 30.31 ± 4.10 30.19 ± 3.64 0.68 b 0.79 Parity (%, (number of parity=1, number of parity=2, number of parity≥3)) 51.56%, 40.89%, 7.56% (116, 92, 17) 55.37%, 38.02%, 6.61% (67, 46, 8) 0.48 c 0.79 History of preterm birth (% , (number of positive cases, number of negative cases)) 9.80% (22, 203) 0.83% (1, 120) 10.06 c < 0.01 History of miscarriage (% , (number of positive cases, number of negative cases)) 16.44% (37, 188) 16.53% (20, 101) 0.00 c 0.98 Maternal internal medical diseases d (%, (number of positive cases, number of negative cases)) 48% (108, 117) 37.20% (45, 76) 3.73 c 0.054 Maternal gynecological diseases e (%, (number of positive cases, number of negative cases)) 19.56% (44, 181) 21.49% (26, 95) 0.18 c 0.67 a: The birth weights of both groups of newborns were normally distributed, using t-test; b: After performing the Shapiro-Wilk test, it was found that the distribution of G6PD enzyme activity values in both groups was skewed ( P  < 0.01). Therefore, the Mann-Whitney U test was used for analysis; c: Using the chi-square test; d: Maternal internal medical diseases include cytopenia, gestational diabetes, abnormal thyroid function, hypertension, and cholestasis; e: Maternal gynecological diseases include uterine adhesions, uterine fibroids, endometriosis, uterine malformations, placental abnormalities, cervical insufficiency, and chorioamnionitis Comparative analysis of clinical data between the study group and the control group Newborn Phenotype G6PD enzyme activity value (U/mL , M (P25 , P75)) 1.20 (0.81 , 3.33) 57.2 (50.10 , 64.30) 8.44% (19, 206) 0.83% (1, 120) 50.7% (114, 111) 16.53% (20, 101) 10.70% (24, 201) 9.92% (12, 109) Maternal body Phenotype 51.56%, 40.89%, 7.56% (116, 92, 17) 55.37%, 38.02%, 6.61% (67, 46, 8) 9.80% (22, 203) 0.83% (1, 120) 16.44% (37, 188) 16.53% (20, 101) 48% (108, 117) 37.20% (45, 76) 19.56% (44, 181) 21.49% (26, 95) a: The birth weights of both groups of newborns were normally distributed, using t-test; b: After performing the Shapiro-Wilk test, it was found that the distribution of G6PD enzyme activity values in both groups was skewed ( P  < 0.01). Therefore, the Mann-Whitney U test was used for analysis; c: Using the chi-square test; d: Maternal internal medical diseases include cytopenia, gestational diabetes, abnormal thyroid function, hypertension, and cholestasis; e: Maternal gynecological diseases include uterine adhesions, uterine fibroids, endometriosis, uterine malformations, placental abnormalities, cervical insufficiency, and chorioamnionitis A risk analysis was performed for factors potentially leading to abortion (Table  2 ) and preterm birth (Table  3 ). Carrying a G6PD variant was not identified as a risk factor for abortion (OR = 0.99, p  = 0.98). However, gestational diabetes mellitus significantly increased the risk of abortion among carriers of G6PD variants (OR = 2.634, p  = 0.02) (Table  2 ). A strong association was observed between G6PD variants and preterm birth (OR = 13.01, p  = 0.013) (Table  3 ). In addition, cholestasis (OR = 10.114, p  = 0.014) and cytopenia (OR = 3.186, p  = 0.046) in study group demonstrated significant positive correlations with preterm birth (Table  3 ). Table 2 Analysis of risk factors that May lead to miscarriage Category G6PD mutation carrier neonates Miscarriage risk analysis Healthy newborn Miscarriage risk analysis OR Statistics P OR Statistics P Carrying a G6PD gene mutation 0.99 a 0.99 a 0.98 a / a / a / a Maternal gynecological diseases Uterine adhesions 0 b / b / b / c / c / c Uterine fibroids 2.073 1.06 0.30 0.554 0.268 0.605 Endometriosis 0.723 0.073 0.79 0 b / b / b Uterine malformations 2.892 0.65 0.42 0 b / b / b Placental abnormalities 0.751 0.11 0.74 1.751 0.365 0.546 Cervical insufficiency / c / c / c / d / d / d Chorioamnionitis 1.737 0.16 0.69 0 b / b / b Maternal internal medical diseases Cytopenia 0.757 0.26 0.61 1.395 0.144 0.704 Gestational diabetes 2.634 5.42 0.02 0.649 0.388 0.533 Abnormal thyroid function 1.204 0.11 0.74 0 b / b / b Hypertensive disease 0.86 0.4 0.84 0 b / b / b Cholestasis 1.803 0.39 0.53 0 b / b / b Data in this table were derived from logistic regression analyses a: G6PD deficiency carrier neonates were analyzed together with the cohort of control neonates; b: Due to the small number of samples, the exposure group has a value of 0, making the calculation meaningless; c: Due to the small number of samples, abnormal values​appear in the OR calculation; d: Due to the small number of samples, the non-exposed group has a value of 0 and cannot be calculated Analysis of risk factors that May lead to miscarriage Data in this table were derived from logistic regression analyses a: G6PD deficiency carrier neonates were analyzed together with the cohort of control neonates; b: Due to the small number of samples, the exposure group has a value of 0, making the calculation meaningless; c: Due to the small number of samples, abnormal values​appear in the OR calculation; d: Due to the small number of samples, the non-exposed group has a value of 0 and cannot be calculated Table 3 Analysis of risk factors that May lead to preterm birth Category Neonates carrying G6PD deficiency Preterm birth risk analysis Healthy newborn Preterm birth risk analysis OR Statistics P OR Statistics P Carrying a G6PD gene mutation 13.01 a 6.22 a 0.013 a / a / a / a Maternal gynecological diseases Uterine adhesions 2.525 0.202 0.653 1.201 0 1 Uterine fibroids 1.134 0.01 0.921 0 b / b / b Endometriosis 0 b / b / b 0 b / b / b Uterine malformations 0 b / b / b 150672308.6 c 0 1 Placental abnormalities 2.803 2.262 0.133 0 b / b / b Cervical insufficiency 9.421 2.356 0.125 / d / d / d Chorioamnionitis 0 b 0 0.999 1.567 0 1 Maternal internal medical diseases Cytopenia 3.186 3.997 0.046 0 b 0 0.999 Gestational diabetes 1.809 1.164 0.281 / d / d / d Abnormal thyroid function 0.656 0.286 0.593 143906683.3 c 0 0.996 Hypertensive disease 1.497 0.201 0.654 0 b 0 0.999 Cholestasis 10.114 5.987 0.014 1.511 0 1 Data in this table were derived from logistic regression analyses a: Analyzing the cohort of newborns carrying G6PD deficiency together with the cohort of healthy newborns; b: Due to the small sample size, the exposed group has values of 0, rendering calculations meaningless; c: Anomalous values were obtained in the calculation of odds ratios due to the small sample size; d: Due to the small sample size, the non-exposed group has values of 0, making calculations impossible Analysis of risk factors that May lead to preterm birth Data in this table were derived from logistic regression analyses a: Analyzing the cohort of newborns carrying G6PD deficiency together with the cohort of healthy newborns; b: Due to the small sample size, the exposed group has values of 0, rendering calculations meaningless; c: Anomalous values were obtained in the calculation of odds ratios due to the small sample size; d: Due to the small sample size, the non-exposed group has values of 0, making calculations impossible Among the 225 neonatal G6PD carriers, 10 variant sites were identified (Fig.  1 ). The variants carried by the neonates were entirely consistent with those carried by their mothers. The most frequent variants were c.1376G > T (47.11%) and c.1388G > A (26.67%), collectively accounting for 73.78% of the total variants. These were followed by c.95 A > G (6.22%), c.1024 C > T (6.22%), c.392G > T (4.89%), c.871G > A (4.00%), c.1360 C > T (2.22%), c.487G > A (1.78%), c.1004 C > A (0.44%), and c.517T > C (0.44%). For neonatal G6PD deficiency carriers, the proportion with abnormal G6PD enzyme activity was 74.7% (168/225). Fig. 1 Distribution of gene variants in 225 neonates with G6PD deficiency Distribution of gene variants in 225 neonates with G6PD deficiency The interrelationship between G6PD gene variant sites and G6PD enzyme activity values in neonates was assessed through data analysis. As depicted in Fig. 2 , enzyme activity values varied substantially across different variant sites, and all were lower than those of healthy neonates, consistent with diagnostic indicators for G6PD deficiency [ 1 ]. The highest median enzyme activity value was observed for c.487G > A (4.30 U/mL), followed by c.392G > T (2.48 U/mL). The median values for the most frequent variants c.1376G > T, c.1388G > A, and c.95 A > G were relatively low, at 0.80 U/mL, 1.25 U/mL, and 1.07 U/mL, respectively. With the exception of c.1004 C > A and c.517T > C, which were represented by a single case each, the mean enzyme activity values for the other eight variants, arranged from lowest to highest, were: c.1360 C > T (0.28 ± 0.12 U/mL), c.1376G > T (1.74 ± 2.01 U/mL), c.1024 C > T (1.96 ± 0.31 U/mL), c.871G > A (2.11 ± 1.43 U/mL), c.95 A > G (3.04 ± 3.65 U/mL), c.1388G > A (3.32 ± 3.39 U/mL), c.487G > A (4.77 ± 3.17 U/mL), and c.392G > T (5.63 ± 5.61 U/mL). Fig. 2 G6PD enzyme activity values across different variant sites G6PD enzyme activity values across different variant sites The number of neonates corresponding to each G6PD enzyme activity grade was compiled for every variant site with detailed data presented in Table  4 . Variants associated with Grade 0 enzyme activity were primarily c.1376G > T (40.35%) and c.1388G > A (36.84%). Grade 1 included c.1024 C > T (40%), c.392G > T (33.33%), c.1376G > T (20%), and c.1004 C > A (6.67%). Grade 2 was mainly associated with c.1376G > T (52.29%), c.1388G > A (25.49%), and c.95 A > G (6.54%). Fisher’s exact test indicated a significant association between variant sites and enzyme activity grades ( p  < 0.001). Table 4 Distribution of gene mutations in 225 children with G6PD deficiency Gene mutation type Level 0 (Negative) Level 1 (weak positive) Level 2 (Positive) Total Proportion of level 2 (%) Total positive rate (%) χ 2 P 1376G > T 23 3 80 106 75 78 60.72 A 21 0 39 60 65 65 95 A > G 4 0 10 14 71 71 1024 C > T 0 6 8 14 57 100 392G > T 4 5 2 11 19 64 871G > A 2 0 7 9 78 78 1360 C > T 0 0 5 5 100 100 487G > A 3 0 1 4 25 25 1004 C > A 0 1 0 1 0 100 517T > C 0 0 1 1 100 100 Total 57 15 153 225 68 75 Distribution of gene mutations in 225 children with G6PD deficiency A grading analysis of G6PD enzyme activity was conducted for the total sample of 346 neonates (225 from the G6PD study group and 121 from the control group). Among these, 178 cases were classified as Grade 0, 15 cases as Grade 1, and 153 cases as Grade 2. As presented in Table  5 , the distribution of enzyme activity grades demonstrated statistical significance ( p  < 0.01). Further pairwise comparisons indicated significant differences among all grade pairs (Grade 0 vs. Grade 1, Grade 0 vs. Grade 2, and Grade 1 vs. Grade 2) ( p  < 0.01). Table 5 Analysis between G6PD enzyme activity grading and neonatal clinical phenotypes Clinical phenotype Enzyme activity classification Statistics P Level 0 Level 1 Level 2 G6PD enzyme activity value (U/mL, M (P25, P75)) 50.40 (8.03, 61.28) 2.22 (2.14, 2.40) 0.93 (0.75, 1.21) H = 266.58 a < 0.01 b Birth gestational age (Week, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{\mathrm x}\pm\mathrm s$$\end{document} ) 39.16 ± 1.20 39.16 ± 0.81 39.02 ± 1.52 0.39 c 0.68 Birth weight (g, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{\mathrm x}\pm\mathrm s$$\end{document} ) 3284.67 ± 431.83 3196.82 ± 330.66 3254.501 ± 640.37 0.23 c 0.80 Preterm birth (%, (number of positive cases, number of negative cases)) 3.93% (7, 171) 6.67% (1, 14) 7.84% (12, 141) 2.60 d 0.23 Jaundice (%, (number of positive cases, number of negative cases)) 26.11% (47, 133) 53.85% (7, 6) 52.29% (80, 73) 25.18 d < 0.01 e Intrauterine distress (%, (number of positive cases, number of negative cases)) 10.00% (18, 162) 15.38% (2, 11) 10.46% (16, 137) 0.34 d 0.84 a: According to the Shapiro-Wilk test, G6PD enzyme activity levels exhibit skewed distribution ( P  < 0.01). Kruskal-Wallis test was conducted to calculate H value and corresponding P value b: Further multiple pairwise comparisons were performed, and statistical differences were found between every two groups ( P  < 0.01) c: Birth gestational age and birth weight of newborns follow a normal distribution. Fisher test was used for multiple groups to calculate F value and corresponding P value d: For categorical data, chi-square test was conducted to calculate χ 2 values and corresponding P values e: Further multiple chi-square tests were performed, revealing statistical differences only between levels 0 and 2 ( P  < 0.01) Analysis between G6PD enzyme activity grading and neonatal clinical phenotypes G6PD enzyme activity value (U/mL, M (P25, P75)) 50.40 (8.03, 61.28) 2.22 (2.14, 2.40) 0.93 (0.75, 1.21) Birth gestational age (Week, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{\mathrm x}\pm\mathrm s$$\end{document} ) Birth weight (g, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overline{\mathrm x}\pm\mathrm s$$\end{document} ) Preterm birth (%, (number of positive cases, number of negative cases)) 3.93% (7, 171) 6.67% (1, 14) 7.84% (12, 141) 26.11% (47, 133) 53.85% (7, 6) 52.29% (80, 73) Intrauterine distress (%, (number of positive cases, number of negative cases)) 10.00% (18, 162) 15.38% (2, 11) 10.46% (16, 137) a: According to the Shapiro-Wilk test, G6PD enzyme activity levels exhibit skewed distribution ( P  < 0.01). Kruskal-Wallis test was conducted to calculate H value and corresponding P value b: Further multiple pairwise comparisons were performed, and statistical differences were found between every two groups ( P  < 0.01) c: Birth gestational age and birth weight of newborns follow a normal distribution. Fisher test was used for multiple groups to calculate F value and corresponding P value d: For categorical data, chi-square test was conducted to calculate χ 2 values and corresponding P values e: Further multiple chi-square tests were performed, revealing statistical differences only between levels 0 and 2 ( P  < 0.01) Statistical analyses were performed for neonatal clinical indicators, including gestational age at birth, birth weight, preterm birth, neonatal jaundice, and intrauterine distress, using the grading of G6PD enzyme activity values as the primary variable. The χ² test indicated that only the incidence of jaundice showed significant variation across the enzyme activity grades ( p   0.05) (Table  5 ). Since all enzyme activity grades in the control group were classified as Grade 0, a comparative analysis was subsequently performed between neonates in the G6PD study group with Grade 0 enzyme activity and those in the control group (Supplementary Table S1). Significant differences were identified in G6PD enzyme activity values, gestational age at birth, birth weight, preterm birth, and neonatal jaundice ( p  < 0.05). No significant differences were observed for intrauterine distress ( p  = 0.96). These findings suggested that neonates in the G6PD study group with Grade 0 enzyme activity had distinct characteristics in several clinical phenotypes compared with the control group. Statistical analysis was conducted on neonatal outcomes and maternal abortion, internal medical conditions, and gynecological conditions according to G6PD variant sites. Detailed results are presented in Supplementary Table S2. Among the 225 carriers, preterm infants accounted for 9.8% (22/225). Within this subgroup, the most frequent variants were c.1376G > T (10 cases, 45.5%) and c.1388G > A (7 cases, 31.8%). No preterm births were observed in carriers of c.1360 C > T, c.487G > A, c.1004 C > A, or c.517T > C variants. The overall incidence of neonatal jaundice among carriers of G6PD variants was 50.7%. Variant-specific incidences exceeding 50% were observed for c.1376G > T (57 cases, 50.0%), c.1388G > A (31 cases, 27.2%), c.1024 C > T (8 cases, 57.1%), c.392G > T (6 cases, 54.5%), and c.487G > A (3 cases, 75%). The variant site with the highest incidence of intrauterine distress was c.95 A > G (21.4%). Among the 225 mothers carrying G6PD variants, 16.4% (37/225) reported a history of abortion, of whom 73.0% (27/37) experienced one abortion and 27.0% (10/37) experienced two or more. The highest proportion of reported history of abortion was observed among mothers carrying the c.1376G > T variant (25/106). Among mothers carrying the c.1388G > A variant, 5 of 60 had a history of abortion, whereas 3 of 14 mothers carrying the c.95 A > G variant reported abortion. The χ² test demonstrated significant differences in abortion rates among mothers of neonates with different variant sites ( p  < 0.05, Supplementary Table S2). Comparative analysis was conducted to examine the association between variant sites and maternal internal medical conditions (Supplementary Table S2). Significant differences were observed across variant sites ( p  < 0.05); however, for each variant, at least one-third of the mothers presented with internal medical conditions. Among the five most frequent variant sites, the highest incidences were observed in mothers with the c.95 A > G variant (85.71%) and the c.392G > T variant (72.73%). Pairwise comparisons further indicated statistically significant differences in the prevalence of internal medical conditions between variant sites: c.95 A > G vs. c.1376G > T (χ² = 7.01, p  = 0.008), c.95 A > G vs. c.1388G > A (χ² = 8.81, p  = 0.003), and c.95 A > G vs. c.1024 C > T (χ² = 7.34, p  = 0.007).

Although most patients with G6PD deficiency do not exhibit any symptoms, ingestion of specific foods or drugs may trigger hemolytic reactions, which can be life-threatening in severe cases. This study demonstrated that, compared with the control group without G6PD variants, mothers in the study group exhibited an increased risk of abortion when gestational diabetes mellitus was present and an increased risk of preterm birth when cytopenia or cholestasis was present. These findings highlight the importance of early diagnosis and timely prevention strategies for G6PD deficiency. In China, neonatal screening for G6PD deficiency is primarily performed by measuring G6PD enzyme activity. Abnormally low enzyme activity values serve as key clinical indicators for the diagnosis of G6PD deficiency and contribute to the etiological assessment of neonatal jaundice, thereby reducing the occurrence of hemolytic events and related complications. Among the 225 neonatal samples carrying G6PD variants in this study, 57 exhibited enzyme activity values within the normal range and were confirmed only through genetic testing. The findings further indicated that preterm birth, abortion, and neonatal jaundice were associated with G6PD variants themselves, rather than solely with enzyme activity. Therefore, for G6PD screening, both enzyme activity measurement and genetic variant testing such as carrier screening and neonatal genetic screening should be strengthened to minimize the risk of missed diagnoses. Based on G6PD gene variants identified in the Fujian region of southern China, this study systematically investigated the association between G6PD genotypes and pregnancy-related clinical phenotypes, characterized the variant spectrum of G6PD deficiency in this region, and examined the molecular mechanisms by which G6PD deficiency synergizes with maternal complications to influence abortion and preterm birth. The relatively small clinical sample size in this study limited the ability to fully elucidate the intrinsic relationship between the complete G6PD variant spectrum and clinical phenotypes such as maternal abortion and preterm birth. Furthermore, enzyme activity was measured only in neonates; due to challenges in follow-up, the enzyme activity of mothers could not be assessed. Future studies with larger clinical sample sizes will enable more comprehensive exploration of the molecular mechanisms linking G6PD deficiency to adverse pregnancy outcomes, as well as the associations between genotype and clinical phenotypes. Such research will provide a stronger scientific basis for improving the management and prevention of this condition.

G6PD deficiency is a common hereditary hematological disorder in southern China and represents one of the major genetic diseases in the Fujian region. The carrier rate of G6PD deficiency in China demonstrates a geographic trend of higher prevalence in the south and lower prevalence in the north, with notable variation across different ethnic groups. Literature data on G6PD deficiency were primarily collected from southern regions, including Hainan Province, Guangxi Province, Guangzhou City in Guangdong Province, Chengdu City in Sichuan Province, Yunnan Province, Ningbo City in Zhejiang Province, and Fujian Province (Supplementary Table S3). Epidemiological studies have reported that the incidence of G6PD deficiency in neonates in the Fujian region ranges from 0.45% to 1.62% [ 6 ]. In terms of ethnicity, the carrier rate of G6PD deficiency among ethnic minorities in southern regions (Zhuang, Li, Dai, Jingpo, and She) is significantly higher than that of the Han population. Within Fujian, the carrier rate among the She ethnicity is 1.62%, which is significantly higher than that observed in the local Han population. Consequently, G6PD testing has become a routine component of neonatal screening in the Fujian region. Such screening enables timely dietary and medical interventions to reduce the occurrence of acute and chronic hemolysis. Even among carriers of G6PD variants, an adequate quality of life can be maintained. Currently, only a limited number of clinical studies have proposed an association between G6PD deficiency and abortion or preterm birth [ 1 , 2 ]. However, the rates of abortion and preterm birth among pregnant women and neonates carrying G6PD variants, as well as the principal risk factors contributing to these outcomes, have not been systematically investigated. Research in this area could support the development of proactive prenatal and perinatal management strategies for pregnant women and fetuses carrying G6PD variants, thereby contributing to the reduction of adverse pregnancy outcomes. This study focused on G6PD variants prevalent in the Chinese population, all of which are missense variants. These missense variants cause amino acid substitutions that impair protein stability or reduce catalytic activity, leading to decreased G6PD enzyme activity [ 2 ]. The three most frequent variants identified in this study were c.1376G > T, c.1388G > A, and c.95 A > G, which are highly consistent with the G6PD deficiency variant spectrum reported across various regions of Fujian Province [ 6 , 16 , 24 ]. These three variants accounted for 80% of the total variants observed in this study, consistent with the G6PD deficiency variant spectrum in China, where they collectively represent 70–80% of all variants [ 25 , 26 ]. The comparison of these results indicates that, although the sample size in this study was relatively limited, the findings are reliable and representative of the G6PD deficiency variant spectrum in the Fujian region. Consistent with previous studies, G6PD variants and the resulting abnormal enzyme activity were closely associated with the occurrence of neonatal jaundice. The incidence of preterm birth and jaundice in neonates carrying G6PD variants with normal enzyme activity (Grade 0) was 18.4 times ( p  < 0.01) and 2.8 times ( p  < 0.01) higher, respectively, than that in neonates without G6PD variants (control group). Within the G6PD study group, the preterm birth rate in the Grade 0 subgroup was higher than that in the abnormal enzyme activity subgroup (15.25% vs. 7.83%, p  = 0.099). These findings indicate that the presence of a G6PD variant alone increases the risk of neonatal preterm birth and jaundice. Further analysis demonstrated that G6PD variants were a risk factor for preterm birth (OR = 13.01, p  = 0.013). In addition, maternal cholestasis (OR = 10.114, p  = 0.014) and cytopenia (OR = 3.186, p  = 0.046) were identified as risk factors that further increased the probability of preterm birth when occurring in conjunction with G6PD variants. From the maternal perspective, no association was identified between neonatal G6PD variants or enzyme activity and abortion. However, G6PD variant combined with maternal gestational diabetes mellitus was associated with an increased risk of abortion (OR = 2.634, p  = 0.02). Higher proportions of preterm birth were observed in pregnant women carrying the following variants: c.392G > T (2/11, 18.18%), c.1388G > A (7/60, 11.67%), c.871G > A (1/9, 11.11%), and c.1376G > T (10/106, 9.43%). Higher proportions of reported history of abortion were observed among women carrying the c.1376G > T (25/106, 23.58%), c.95 A > G (3/14, 21.43%), and c.392G > T (2/11, 18.18%) variants. Although the above data reveals a potential association between G6PD variants and adverse pregnancy outcomes. However, it must be made clear that the occurrence of adverse pregnancy outcomes is often the result of multiple confounding factors. We did not systematically collect and compare all known high-risk factors for adverse pregnancy outcomes between the two groups, and therefore did not rule out the contributions of these risk factors towards the miscarriages or preterm births. G6PD variants may further amplify the risk of adverse pregnancy outcomes under the synergistic influence of other factors. Therefore, the conclusions of this study need to be interpreted with caution. A literature review on the mechanisms linking G6PD deficiency to abortion and preterm birth indicates that research in this area remains limited and the mechanisms are not fully understood. G6PD functions as the sole rate-limiting enzyme in the PPP. By catalyzing the first reaction of the PPP, it generates ribose-5-phosphate required for DNA synthesis and nicotinamide adenine dinucleotide phosphate (NADP+/NADPH), the primary hydrogen donor in biosynthetic reactions [ 27 ]. G6PD exists in monomeric, dimeric, and tetrameric forms, with only the dimer exhibiting catalytic activity. Interconversion among these forms is essential for enzyme function [ 28 , 29 ]. Huang et al. [ 22 ] reported that GPR120 promotes decidualization by upregulating glucose uptake and activation of the PPP in human endometrial stromal cells, thereby exerting a protective effect on pregnancy. The PPP represents a major glucose metabolic pathway that supplies ribose-5-phosphate and NADPH for the biosynthesis of nucleic acids, fatty acids, amino acids, and other biomolecules, thus supporting cellular proliferation, differentiation, and activation [ 30 ]. When G6PD , the key enzyme of the PPP, is upregulated by GPR120, the decidualization-promoting effect of GPR120 is enhanced, whereas inhibition of the PPP results in impaired endometrial decidualization and implantation of fertilized eggs. This mechanism provides a plausible explanation for the higher incidence of abortion among females carrying G6PD variants and offers a basis for analyzing the pathways by which G6PD variants contribute to abortion. On this basis, the molecular mechanisms underlying the G6PD –adverse pregnancy association were examined, as illustrated in Fig. 3 . Fig. 3 Proposed molecular mechanisms linking G6PD deficiency to adverse pregnancy Proposed molecular mechanisms linking G6PD deficiency to adverse pregnancy Along with its role in the PPP in supporting decidualization, G6PD contributes to the activity of superoxide dismutase, glutathione peroxidase (GSHPX), and catalase in processing hydrogen peroxide (H₂O₂) generated during aerobic metabolism. When G6PD expression is reduced, NADPH levels decrease correspondingly, leading to oxidative stress and accumulation of H₂O₂. Under the conditions of the Haber–Weiss and Fenton reactions, this accumulation results in the production of large quantities of hydroxyl radicals ( - OH), which damage DNA, proteins, and fatty acids essential for maintaining cellular integrity, thereby impairing embryonic development [ 31 ]. Taken together, the increased susceptibility to adverse pregnancy in females with G6PD deficiency may be explained by dual mechanisms: impaired decidualization and induction of oxidative stress, leading to H₂O₂ accumulation and subsequent adverse effects on embryonic development, ultimately resulting in adverse pregnancy. Data analysis indicated that the preterm birth rate in the G6PD study group was 11.2 times higher than that in the control group, with a statistically significant difference between the two groups (χ² = 10.06, p = 0.02). Furthermore, regardless of whether the G6PD variant altered enzyme activity, an elevated incidence of preterm birth was observed. These findings indicate a specific intrinsic association between G6PD variants and preterm birth. One clinical study reported that neonatal G6PD deficiency combined with preterm birth constituted a risk factor for severe hyperbilirubinemia, increasing its incidence by 10-fold (OR = 10.2, 95% CI 1.35–76.9). However, the causes of preterm birth and severe hyperbilirubinemia remain unclear [ 32 ]. Maternal cholestasis and cytopenia were closely associated with preterm birth, indicating that maternal factors may play a primary role in its occurrence. Due to the lack of more detailed clinical data, the underlying relationship between G6PD mutant genotypes and the preterm birth phenotype could not be fully elucidated in this study and will require further investigation. Exposure to specific oxidants, such as primaquine and chloroquine (used for malaria prophylaxis), or ingestion of fava beans may induce acute hemolytic episodes in patients with G6PD deficiency [ 33 ]. Strengthened clinical screening and management are therefore particularly important. In clinical practice, initial management should involve genetic screening of pregnant women for G6PD variants, monitoring for potential complications during pregnancy, and caution regarding the use of known hemolytic agents. Supplementation with antioxidants such as vitamin E, vitamin C, beta-carotene, lipoic acid, riboflavin, selenium, and zinc has been recommended [ 34 ]. Antioxidants in the body are derived from endogenous sources, including catalase and GSHPX, as well as exogenous dietary sources, including vitamin E, vitamin C, beta-carotene, lipoic acid, riboflavin, selenium, and zinc. In G6PD deficiency, adequate dietary intake of antioxidants may mitigate the effects of increased oxidant stress, potentially reducing the risk of abortion [ 35 , 36 ]. Whether such strategies are effective in preventing preterm birth remains to be determined. However, evidence indicates that pregnant women carrying G6PD variants should be closely monitored for the risk of preterm birth.

The glucose-6-phosphate dehydrogenase ( G6PD ) gene (OMIM ID: 305900) encodes glucose-6-phosphate dehydrogenase, an oxidoreductase. It is located on the X chromosome (Xq28), spanning approximately18 kb and comprising 13 exons. The gene encodes a protein of 515 amino acids and catalyzes the oxidation of D-glucose-6-phosphate (G6P) to 6-phosphoglucono-β-lactone, while reducing nicotinamide adenine dinucleotide phosphate (NADP) to reduced nicotinamide adenine dinucleotide phosphate (NADPH) [ 1 , 2 ]. Pathogenic or likely pathogenic G6PD variants result in G6PD deficiency, an X-linked incompletely dominant genetic disorder and the most common erythrocyte enzyme-related disease worldwide [ 3 ]. The global incidence of G6PD deficiency is estimated at 4.9% [ 4 ]. In China, the condition is highly prevalent, with incidence rates decreasing from south to north and showing considerable variation among different ethnic groups [ 5 – 7 ]. Currently, approximately 217 G6PD deficiency variants have been reported globally, with at least 35 variants identified in the Chinese population [ 8 – 10 ]. The 14 common G6PD variants in China are c.95 A > G, c.392G > T, c.487G > A, c.493 A > G, c.592G > T, c.871G > A, c.1004 C > T, c.1024 C > T, c.1360 C > T, c.1376G > T, c.1387 C > T, c.1388G > A, c.1381G > A, and c.1311 C > T ( NM_001042351.2 ) [ 11 , 12 ]. G6PD deficiency affects approximately 500 million people worldwide. Most affected patients remain asymptomatic but may develop symptoms when triggered by specific factors, such as ingestion of fava beans, certain drugs, infections, or metabolic stressors [ 13 , 14 ]. Clinical presentations include: (1) neonatal jaundice; (2) acute hemolytic anemia triggered by fava beans, drugs, or infections; and (3) chronic nonspherocytic hemolytic anemia [ 15 ]. Neonatal jaundice associated with G6PD deficiency may progress to kernicterus in severe cases, representing a life-threatening condition. Optimal management relies on proactive prevention through avoidance of known triggers. Cost-effective, simple, and reliable strategies such as neonatal screening allow early identification of affected neonates, thereby facilitating timely diagnosis and intervention [ 16 , 17 ]. G6PD is an enzyme active in the initial stage of the pentose phosphate pathway (PPP). The PPP is recognized as a critical glucose oxidation pathway responsible for generating reduced NADPH and ribose-5-phosphate. NADPH contributes to the reduction of oxygen-derived free radicals that are continuously produced during physiological processes, enabling the PPP to play an essential role in energy metabolism during embryonic development [ 18 – 20 ]. In 1985, Toncheva et al. [ 21 ] reported that the rate of spontaneous abortion in early pregnancy among females heterozygous for G6PD deficiency (21.7%) was significantly higher compared with unaffected females (9.3%), indicating reduced viability of fertilized eggs in heterozygous carriers of G6PD variants. In 2017, Huang et al. demonstrated that upregulation of G6PD, a key enzyme in the PPP, by GPR120 enhanced its decidualization-promoting effect, whereas inhibition of the PPP impaired endometrial decidualization and fertilized egg implantation [ 22 ]. Although early clinical studies have examined the effects of G6PD deficiency on abortion and embryonic development, no systematic investigation has assessed the association between specific genetic variants, their corresponding G6PD enzyme activities, and pregnancy comorbidities or adverse gestational outcomes such as abortion and preterm birth. This study conducted a statistical analysis of the incidence of abortion and preterm birth among pregnant patients with comorbidities who carried pathogenic G6PD variants. In addition, perinatal clinical data were compared between neonates carrying G6PD variants and healthy neonates without such variants to assess the relationship between G6PD deficiency, neonatal jaundice, maternal clinical phenotypes during pregnancy, and adverse outcomes including abortion and preterm birth.

Supplementary Material 1. Supplementary Material 1. Supplementary Material 2. Supplementary Material 2. Supplementary Material 3. Supplementary Material 3.