Risk of neurodevelopmental disorders in the offspring of mothers with a history of endometriosis in Taiwanese women

Pei-Yin Yang: Writing – original draft, Conceptualization. Ching-Ming Wang: Writing – review & editing, Writing – original draft, Conceptualization. Pei-Lun Liao: Software, Project administration, Investigation, Formal analysis, Data curation, Conceptualization. Jing-Yang Huang: Supervision, Methodology. Keng-Wei Liang: Software, Project administration. Yu-Hsuan Lin: Data curation. Shun-Fa Yang: Visualization, Project administration. Po-Hui Wang: Writing – review & editing, Validation, Supervision, Project administration, Methodology, Conceptualization.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Figure 1 In total, 166,500 mother–offspring pairs (including 148,000 offspring of mothers without endometriosis and 18,500 offspring of mothers with endometriosis) were included in this analysis. Table 1 presents the basic characteristics of the two groups of mother–offspring pairs matched for maternal age at delivery, infant sex, and delivery date (within 180 days) and compares the offspring characteristics, maternal characteristics, and paternal age. Table 1 Baseline characteristics of the mothers and offspring of the case cohort (mother–offspring pairs with mothers who had endometriosis before becoming pregnant) and the matched comparison cohort a . Table 1 Offspring of mothers without endometriosis Offspring of mothers with endometriosis P value ASD b Total number 148,000 18,500 Characteristics of offspring Child's birth year 0.8809 0.0561  2009 2128 (1.44 %) 257 (1.39 %)  2010 9041 (6.11 %) 1162 (6.28 %)  2011 14,089 (9.52 %) 1746 (9.44 %)  2012 18,782 (12.69 %) 2299 (12.43 %)  2013 21,448 (14.49 %) 2724 (14.72 %)  2014 24,414 (16.5 %) 3070 (16.59 %)  2015 27,983 (18.91 %) 3468 (18.75 %)  2016 30,115 (20.35 %) 3774 (20.4 %) Child's sex 1.0000 0.0000  Female 71,080 (48.03 %) 8885 (48.03 %)  Male 76,920 (51.97 %) 9615 (51.97 %) Birth weeks <0.0001 0.1965  41 3282 (2.22 %) 343 (1.85 %)  Mean ± SD 38.28 (±1.57) 38.00 (±1.80) Birth weight, gram <0.0001 0.1088  <2500 9820 (6.64 %) 1688 (9.12 %)  2500–3500 115,968 (78.36 %) 14,568 (78.75 %)  ≥3500 22,212 (15.01 %) 2244 (12.13 %)  Mean ± SD 3081.05 (±434.69) 3023.97 (±460.38) Apgar score at 1 min < 7 <0.0001 0.0719  ASA <7 3075 (2.08 %) 598 (3.23 %)  7 ≤ ASA 144,925 (97.92 %) 17,902 (96.77 %) Apgar score at 5 min < 7 <0.0001 0.0434  ASB <7 509 (0.34 %) 120 (0.65 %)  7 ≤ ASB 147,491 (99.66 %) 18,380 (99.35 %) Characteristics of mother s Mother's age at delivery 0.0207 0.0408  <30 22,759 (15.38 %) 2804 (15.16 %)  30–39 116,279 (78.57 %) 14,481 (78.28 %)  ≥40 8962 (6.06 %) 1215 (6.57 %)  Mean ± SD 33.48 (±3.95) 33.57 (±4.00) Urbanization 0.0304 0.0000  Urban 98,466 (66.53 %) 12,418 (67.12 %)  Sub-urban 42,643 (28.81 %) 5296 (28.63 %)  Rural 6891 (4.66 %) 786 (4.25 %) Insurance unit <0.0001 0.0344  Public sector insurance 11,216 (7.58 %) 1560 (8.43 %)  Labor insurance 114,988 (77.69 %) 14,680 (79.35 %)  Agriculture 5800 (3.92 %) 672 (3.63 %)  Low-income household insurance 159,96 (10.81 %) 1588 (8.58 %) Mode of delivery <0.0001 0.3187  Vaginal delivery 93,119 (62.92 %) 8744 (47.26 %)  Cesarean section 54,881 (37.08 %) 9756 (52.74 %) Mother's comorbidities  Asthma 1740 (1.18 %) 186 (1.01 %) 0.0412 0.0164  Allergic rhinitis 10,458 (7.07 %) 1345 (7.27 %) 0.3079 0.0079  Atopic dermatitis 23,247 (15.71 %) 3216 (17.38 %) <0.0001 0.0451  Hypertension 2962 (2 %) 359 (1.94 %) 0.5770 0.0044  Diabetes mellitus 4427 (2.99 %) 535 (2.89 %) 0.4538 0.0059  Hyperlipidemia 903 (0.61 %) 119 (0.64 %) 0.5867 0.0042  Urinary tract infection 16,744 (11.31 %) 2529 (13.67 %) <0.0001 0.0713  Gestational diabetes mellitus 10,562 (7.14 %) 1411 (7.63 %) 0.0149 0.0188  Eclampsia or preeclampsia 5904 (3.99 %) 833 (4.50 %) 0.0008 0.0255  Anxiety 4598 (3.11 %) 642 (3.47 %) 0.0076 0.0204  Sleep disorder 7471 (5.05 %) 1061 (5.74 %) <0.0001 0.0304  Rheumatoid arthritis 132 (0.09 %) 23 (0.12 %) 0.1396 0.0108  Systemic lupus erythematosus 284 (0.19 %) 59 (0.32 %) 0.0003 0.0252  Sjogren's syndrome 443 (0.30 %) 92 (0.50 %) <0.0001 0.0314 Characteristics of fathers Father's age at delivery <0.0001 0.0345  <30 13,775 (9.31 %) 1636 (8.84 %)  30–39 10,3128 (69.68 %) 13,219 (71.45 %)  ≥40 24,307 (16.42 %) 3080 (16.65 %)  Mean ± SD 34.26 (±4.79) 35.39 (±4.67) SD, standard deviation. a The selected 18,500 mother–offspring pairs with mothers who had endometriosis (the case cohort) were matched with 148,000 mother–offspring pairs with mothers who did not have endometriosis (the comparison cohort) for maternal age at delivery, infant sex, and delivery date (within 180 days) at a 1:8 ratio. b ASD, absolute standardized differences. An ASD <0.1 indicates that the characteristics were similar between the two groups. Baseline characteristics of the mothers and offspring of the case cohort (mother–offspring pairs with mothers who had endometriosis before becoming pregnant) and the matched comparison cohort a . SD, standard deviation. The selected 18,500 mother–offspring pairs with mothers who had endometriosis (the case cohort) were matched with 148,000 mother–offspring pairs with mothers who did not have endometriosis (the comparison cohort) for maternal age at delivery, infant sex, and delivery date (within 180 days) at a 1:8 ratio. ASD, absolute standardized differences. An ASD <0.1 indicates that the characteristics were similar between the two groups. The analysis of offspring characteristics revealed a higher incidence of preterm birth (gestational age <37 weeks, 10.64 % vs. 7.44 %, P  < 0.0001, ASD = 0.1965; Table 1 ) and low birth weight (<2500 g, 9.12 % vs. 6.64 %, P  < 0.0001, ASD = 0.1088), as well as lower Apgar scores (<7) at 1 min (3.23 % vs. 2.08 %, P  < 0.0001, ASD = 0.0719) and 5 min (0.65 % vs. 0.34 %, P  < 0.0001, ASD = 0.0434) in the case group compared with the comparison group. With regard to maternal characteristics, mothers with a history of endometriosis had a higher mean maternal age at delivery (33.57 vs. 33.48 years, P  = 0.0207, ASD = 0.0408), a higher urbanization percentage (67.12 % vs. 66.53 %, P  = 0.0304, ASD = 0.0000), different insurance unit ( insured entity attributes ; P  < 0.0001, ASD = 0.0344), and a higher cesarean section rate (52.74 % vs. 37.08 %, P  < 0.0001, ASD = 0.3187) than mothers without endometriosis. In terms of maternal comorbidities, mothers with endometriosis had a lower incidence of asthma (1.01 % vs. 1.18 %, P  = 0.0412, ASD = 0.0164) and higher incidence rates of atopic dermatitis (17.38 % vs. 15.71 %, P  < 0.0001, ASD = 0.0451), urinary tract infection (13.67 % vs. 11.31 %, P  < 0.0001, ASD = 0.0713), gestational diabetes (7.63 % vs. 7.14 %, P  = 0.0149, ASD = 0.0188), eclampsia or preeclampsia (4.50 % vs. 3.99 %, P  = 0.0008, ASD = 0.0255), systemic lupus erythematosus (0.32 % vs. 0.19 %, P  = 0.0003, ASD = 0.0252), Sjogren's syndrome (0.50 % vs. 0.30 %, P  < 0.0001, ASD = 0.0314), anxiety (3.47 % vs. 3.11 %, P  = 0.0076, ASD = 0.0204), and sleep disorder (5.74 % vs. 5.05 %, P  < 0.0001, ASD = 0.0304) than did mothers without endometriosis. In terms of paternal age, the husbands of mothers with a history of endometriosis had a higher mean age at delivery than the husbands of mothers without endometriosis (35.39 vs. 34.26, P  0.1 to detect statistical significance, significant between-group differences were observed in offspring characteristics, specifically in the incidence of preterm birth (ASD = 0.1965) and low birth weight (ASD = 0.1088), as well as in maternal characteristics, specifically in the rate of cesarean section (ASD = 0.3187, Table 1 ). No difference was detected in the incidence of maternal anxiety (ASD = 0.0204) or sleep disorders (ASD = 0.0304) between the case and comparison groups. The primary outcomes, the relationships between maternal endometriosis and neurodevelopmental disorders in offspring, including developmental delay, cerebral palsy, intellectual disabilities, and other neurodevelopmental and psychiatric disorders, were compared between the case and comparison groups ( Table 2 ). The relationships between maternal endometriosis and developmental delay in offspring were significant without adjustment ( P  < 0.0001), after adjustment in model 1 ( P  < 0.0001) and after adjustment in model 2 ( P  < 0.0001). The relationships between maternal endometriosis and cerebral palsy were significant without adjustment ( P  = 0.0269), after adjustment in model 1 ( P  = 0.0256) but not significant after adjustment in model 2 ( P  = 0.9372). However, there were no significant differences between maternal endometriosis and intellectual disabilities in offspring without adjustment ( P  = 0.3753), after adjustment in model 1 ( P  = 0.3816) and after adjustment in model 2 ( P  = 0.1131). There were also no significant differences between maternal endometriosis and other neurodevelopmental and psychiatric disorders in offspring without adjustment ( P  = 0.0530), after adjustment in model 1 ( P  = 0.1290) and after adjustment in model 2 ( P  = 0.2675). Table 2 Outcomes of neurodevelopmental disorders in the offspring of mothers with endometriosis and the comparison cohort a . Table 2 Offspring of mothers without endometriosis Offspring of mothers with endometriosis P Developmental delay  N 148,000 18,500  Observed person-months 9,893,129 1,223,517  Event of case 12,906 1853  Incidence rate b (95 % CI) 1.30 (1.28–1.33) 1.51 (1.45–1.59)  Crude HR (95 % CI) reference 1.16 (1.11–1.22) <0.0001  Adjusted HR (95 % CI) model 1 c reference 1.16 (1.11–1.22) <0.0001  Adjusted HR (95 % CI) a model 2 d reference 1.11 (1.06–1.17) <0.0001 Cerebral palsy  N 148,000 18,500  Observed person-months 10,348,897 1,290,100  Event of case 326 56  Incidence rate b (95 % CI) 0.03 (0.03–0.04) 0.04 (0.03–0.06)  Crude HR (95 % CI) reference 1.38 (1.04–1.83) 0.0269  Adjusted HR (95 % CI) model 1 c reference 1.39 (1.04–1.85) 0.0256  Adjusted HR (95 % CI) model 2 d reference 1.01 (0.76–1.36) 0.9372 Intellectual disabilities  N 148,000 18,500  Observed person-months 1,0347,440 1,291,073  Event of case 619 69 Incidence rateb (95 % CI) 0.06 (0.06–0.06) 0.05 (0.04–0.07)  Crude HR (95 % CI) reference 0.89 (0.70–1.15) 0.3753  Adjusted HR (95 % CI) model 1 c reference 0.89 (0.69–1.16) 0.3816  Adjusted HR (95 % CI) model 2 d reference 0.81 (0.62–1.05) 0.1131 Other neurodevelopmental and psychiatric disorders  N 148,000 18,500  Observed person-months 10,338,206 1,288,140  Event of case 1025 151  Incidence rate b (95 % CI) 0.10 (0.09–0.11) 0.12 (0.10–0.14)  Crude HR (95 % CI) reference 1.00 (1.00–1.40) 0.0530  Adjusted HR (95 % CI) model 1 c reference 1.15 (0.96–1.37) 0.1290  Adjusted HR (95 % CI) model 2 d reference 1.11 (0.93–1.32) 0.2675 CI, confidence interval. a The selected 18,500 mother–offspring pairs with mothers who had endometriosis (the case cohort) were matched with 148,000 mother–offspring pairs with mothers who did not have endometriosis (the comparison cohort) for maternal age at delivery, infant sex, and delivery date (within 180 days) at a 1:8 ratio. b Incidence rate, per 1000 person-months. c Model 1: Adjusted hazard ratio (aHR), adjusted for child's birth year, child's sex, mother’s age at delivery, urbanization, insurance unit, father’s age, and mother’s comorbidity. d Model 2: aHR, adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father’s age, mother’s comorbidity, mode of delivery, birth weeks, birth weight, and Apgar score. Outcomes of neurodevelopmental disorders in the offspring of mothers with endometriosis and the comparison cohort a . CI, confidence interval. The selected 18,500 mother–offspring pairs with mothers who had endometriosis (the case cohort) were matched with 148,000 mother–offspring pairs with mothers who did not have endometriosis (the comparison cohort) for maternal age at delivery, infant sex, and delivery date (within 180 days) at a 1:8 ratio. Incidence rate, per 1000 person-months. Model 1: Adjusted hazard ratio (aHR), adjusted for child's birth year, child's sex, mother’s age at delivery, urbanization, insurance unit, father’s age, and mother’s comorbidity. Model 2: aHR, adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father’s age, mother’s comorbidity, mode of delivery, birth weeks, birth weight, and Apgar score. Considering secondary outcomes and incidence rate, the results revealed that the incidence rate and crude HR and adjusted HRs in model 1 and model 2 of developmental delay in the offspring of mothers with a history of endometriosis were higher than those in the offspring of mothers without endometriosis (incidence 1.51 vs. 1.30 per 1000 person-months, crude HR = 1.16, 95 % CI = 1.11–1.22; adjusted HR = 1.16, 95 % CI = 1.11–1.22 in model 1; and adjusted HR = 1.11, 95 % CI = 1.06–1.17 in model 2; Table 2 ). In view of the sensitivity analysis, no matter crude HR (1.16, 95 % CI = 1.11–1.22), and aHRs in model 1 (1.16, 95 % CI = 1.11–1.22) or model 2 (1.11, 95 % CI = 1.06–1.17), the offspring of mothers with a history of endometriosis exhibited a significantly higher risk of developmental delay than offspring of mothers without endometriosis. The incidence rate and crude HR and adjusted HR in model 1 of cerebral palsy in the offspring of mothers with a history of endometriosis were higher than those in the offspring of mothers without endometriosis (incidence 0.04 vs. 0.03 per 1000 person-months, crude HR = 1.38, 95 % CI = 1.04–1.83; adjusted HR = 1.39, 95 % CI = 1.04–1.85 in model 1), but not higher in adjusted HR in model 2 (adjusted HR = 1.01, 95 % CI = 0.76–1.36). With regard to intellectual disabilities, the incidence rate and HRs in the offspring of mothers with a history of endometriosis were not higher than those in the offspring of mothers without endometriosis (incidence 0.05 vs. 0.06 per 1000 person-months, crude HR = 0.89, 95 % CI = 0.70–1.15; adjusted HR = 0.89, 95 % CI = 0.69–1.16 in model 1; and adjusted HR = 0.81, 95 % CI = 0.62–1.05 in model 2; Table 2 ). Regarding other neurodevelopmental and psychiatric disorders, although their incidence (0.12 vs. 0.10 per 1000 person-months) and crude HR (1.00, 95 % CI = 1.00–1.40) tended to be higher in the offspring of mothers with a history of endometriosis than in the offspring of mothers without endometriosis, no significantly different aHRs were observed in model 1 (adjusted HR = 1.15, 95 % CI = 0.96–1.37) or in model 2 (adjusted HR = 1.11, 95 % CI = 0.93–1.32) between the two cohorts. The Kaplan–Meier plot curve and log rank test revealed that the offspring of mothers with a history of endometriosis had significantly higher risks of developmental delay ( P  < 0.0001) and cerebral palsy ( P  = 0.0265, Fig. 1 ) than those of mothers without endometriosis, but such increased risks were not detected for intellectual disabilities and other neurodevelopmental and psychiatric disorders ( P  = 0.375 and 0.0527, respectively). Fig. 1 Kaplan–Meier curves for the cumulative incidence rates of developmental delay (1A, P  < 0.0001), cerebral palsy (1B, P  = 0.0265), intellectual disabilities (1C, P  = 0.375), and other neurodevelopmental and psychiatric disorders (1D, P  = 0.0527) in the offspring of mothers with a history of endometriosis and the comparison cohort. The log-rank test was used to compare the differences of neurodevelopment disorders to determine statistical significance between two cohorts . Fig. 1 Kaplan–Meier curves for the cumulative incidence rates of developmental delay (1A, P  < 0.0001), cerebral palsy (1B, P  = 0.0265), intellectual disabilities (1C, P  = 0.375), and other neurodevelopmental and psychiatric disorders (1D, P  = 0.0527) in the offspring of mothers with a history of endometriosis and the comparison cohort. The log-rank test was used to compare the differences of neurodevelopment disorders to determine statistical significance between two cohorts . In assessing the time-varying effect of maternal endometriosis on the neurodevelopmental disorders of offspring after they were born, it revealed crude HRs of 1.38, 1.16, and 1.09 for developmental delay from the index date (the date when the offspring was born) to 12 months, from 12 to 48 months, and from 48 months onwards, respectively ( Table 3 ). It also revealed that the significantly highest risk estimate for developmental delay was for the period from 12 to 48 months after the index date in model 2, with an aHR of 1.12 (95 % CI = 1.05–1.19). The crude HR was 1.16 (95 % CI = 1.09–1.23) and the aHR in model 1 was 1.15 (95 % CI = 1.09–1.23) during this period ( Table 3 ). The significantly highest risk estimate for cerebral palsy in offspring was for the period from the index date to 36 months after the index date, with a crude HR of 1.41 (95 % CI = 1.03–1.92). The aHR was 1.45 (95 % CI = 1.06–1.98) in model 1 and 1.04 (95 % CI = 0.76–1.43) in model 2 during this period ( Table 4 ). However, maternal endometriosis showed no time-varying effects on the risks of intellectual disabilities ( Supplementary Table 1 ) and other neurodevelopmental and psychiatric disorders ( Supplementary Table 2 ) in offspring. Table 3 Time-varying effect of maternal endometriosis on the incidence of developmental delay in offspring. Table 3 Follow-up time interval Comparisons Offspring of mothers with endometriosis HR (95 % CI) aHR (95 % CI) Model 1 b aHR (95 % CI) Model 2 c Person-months Event Incidence a (95 % CI) Person-months Event Incidence a (95 % CI) From index date to 12 months 1,766,943 1084 0.61 (0.58–0.65) 220,426 186 0.84 (0.73–0.97) 1.38 (1.18–1.61) 1.39 (1.19–1.63) 1.16 (0.99–1.36) 12–48 months 4,942,667 8305 1.68 (1.64–1.72) 612,149 1195 1.95 (1.84–2.07) 1.16 (1.09–1.23) 1.15 (1.09–1.23) 1.12 (1.05–1.19) After 48 months 3,183,519 3517 1.10 (1.07–1.14) 390,942 472 1.21 (1.10–1.32) 1.09 (0.99–1.20) 1.11 (1.00–1.22) 1.09 (0.99–1.20) CI, confidence interval. a Incidence rate, per 1000 person-months. b Model 1: Adjusted hazard ratio (aHR), adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father's age, and mother's comorbidity. c Model 2: aHR, adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father’s age, mother’s comorbidity, mode of delivery, birth weeks, birth weight, and Apgar score. Table 4 Time-varying effect of maternal endometriosis on the incidence of cerebral palsy in offspring. Table 4 Comparisons Offspring with maternal Endometriosis HR (95 % CI) aHR (95 % CI) Model 1 b aHR (95 % CI) Model 2 c Follow-up time interval Person-months Event Incidence a (95 % CI) Person-months Event Incidence a (95 % CI) From index date to 36 months 5,307,858 268 0.05 (0.04–0.06) 662,818 47 0.07 (0.05–0.09) 1.41 (1.03–1.92) 1.45 (1.06–1.98) 1.04 (0.76–1.43) After 36 months 5,041,039 58 0.01 (0.01–0.01) 627,282 9 0.01 (0.01–0.03) 1.25 (0.62–2.52) 1.11 (0.53–2.32) 0.85 (0.40–1.81) CI, confidence interval. a Incidence rate, per 1000 person-months. b Model 1: Adjusted hazard ratio (aHR), adjusted for child's birth year, child's sex, mother’s age at delivery, urbanization, insurance unit, father’s age, and mother’s comorbidity. c Model 2: aHR, adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father’s age, mother’s comorbidity, mode of delivery, birth weeks, birth weight, and Apgar score. Time-varying effect of maternal endometriosis on the incidence of developmental delay in offspring. CI, confidence interval. Incidence rate, per 1000 person-months. Model 1: Adjusted hazard ratio (aHR), adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father's age, and mother's comorbidity. Model 2: aHR, adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father’s age, mother’s comorbidity, mode of delivery, birth weeks, birth weight, and Apgar score. Time-varying effect of maternal endometriosis on the incidence of cerebral palsy in offspring. CI, confidence interval. Incidence rate, per 1000 person-months. Model 1: Adjusted hazard ratio (aHR), adjusted for child's birth year, child's sex, mother’s age at delivery, urbanization, insurance unit, father’s age, and mother’s comorbidity. Model 2: aHR, adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father’s age, mother’s comorbidity, mode of delivery, birth weeks, birth weight, and Apgar score.

We conducted a nationwide population-based cohort analysis involving infants born in Taiwan from January 1, 2009 to December 31, 2016 using data from the Taiwan Maternal and Child Health Database (TMCHD). This database consists of four linked nationwide databases, namely the National Health Insurance Research Database (NHIRD), the Birth Certificate Application Database (BCAD), the Birth Registration Database (BRD), and the National Register of Death (NRD). The NHIRD was established in Taiwan in 1995, and more than 99.8 % of Taiwan's residents are enrolled in it. This database gives detailed clinical information of outpatients and inpatients, including dates of visits, disease diagnoses, prescriptions, and medical orders. The TMCHD has data on more than 99.78 % of births in Taiwan since 2004, including specific information about newborns and their family (i.e., parental identity, child identity, and birth registration data). The BRD and NRD provide data on the birth and survival status of infants. Through these four databases, data on infants/children and their parents can be obtained using unique identification numbers. The study has been carried out in accordance with the recommendations of the World Medical Association Declaration of Helsinki and the International Committee of Medical Journal Editors, and includes the representative human populations as per those recommendations. This work was performed in compliance with relevant laws and was approved by the institutional review board at Chung Shan Medical University Hospital (approval number CS2-22054) with a waiver for informed consent. The privacy rights of human subjects were observed. The current study is a population–based retrospective cohort research in which data from children born between 2009 and 2016 in Taiwan were analyzed. The total number of eligible mother–newborn pairs was 1,459,093. Of these, we excluded pairs with mothers of foreign nationalities (n = 88,837) because the study analyzes the relationship between the endometriosis history in Taiwanese mothers and neurodevelopmental disorders in the offspring, abortion or stillbirths (n = 14,639) because they could not be followed long enough for the occurrence of neurodevelopmental disorders and this research focused on the long term outcome or condition of live offspring, infants with missing identification (n = 67,274) because their clinical data could not be followed completely, and multiple pregnancy (n = 42,218) because the exact relationship between maternal endometriosis and neurodevelopmental disorders could not be defined accurately, especially if the disorders of multiple pregnancies were not presented consistently ( Supplementary Fig. 1 ). Of the total 1,246,125 eligible mother–newborn pairs, we identified 18,500 mothers with a history of clinically diagnosed endometriosis and 1,227,625 mothers without endometriosis. The mothers with and without endometriosis were matched for maternal age at delivery, infant sex, and delivery date (within a 180-day window) at a 1:8 ratio (18,500 vs. 148,000) in order to enhance statistical stability, reduce sampling error impact, and maximize research benefits when control group data is abundant yet cost-effective and therefore to expand the matching ratio typically providing a more comprehensive representation of population characteristics. ( Iwagami and Shinozaki, 2022 ; Stuart, 2010 ). The identification of mothers with a history of endometriosis was enrolled based on the International Classification of Disease (ICD), Ninth/Tenth Revision, Modification code. Mothers with endometriosis were enrolled when they had 2 outpatient visits or any hospitalization ( Huang et al., 2021 , 2023 ) based on the diagnosis codes ICD-9-CM: 617 and ICD-10-CM: N80. The operation history of endometriosis-related procedures (coding according to national health insurance drug benefit items and payment standards) was further used for the validation of the diagnosis of endometriosis, including laparoscopic fulguration or excision of pelvic endometriosis (code 80014, 80029, 81032, 80031, 81033), myomectomy (code 80402, 80415, 80420, 80425, 97013), salpingo-oophorectomy or oophorectomy and partial or complete adnexectomy (code 80807, 80802, 80811, 80812), and endometrial ablation or transcervical endometrial resection (code 81023). The operation records recruited the classification of endometriosis based on the American Society for Reproductive Classification of Endometriosis, with operation methods from the mildest laparoscopic electrocautery to the more severe myomectomy (adenomyomctomy). Adenomysis cases were included for analysis. Transcervical endometrial resection may find adenomyoma incidentally. Salpingo-oophorectomy or oophorectomy and partial or complete adnexectomy may find pelvic or ovarian endometriosis, ovarian endometrioma or hydrosalpinx caused by endometriosis. The characteristics of the offspring, mothers, and fathers were extracted and assessed. The baseline characteristics of the offspring included the year of birth, infant sex, gestational age, birth weight, and the Apgar score. Maternal characteristics included age at delivery, urbanization, insurance unit, mode of delivery, and comorbidities including asthma, allergic rhinitis, atopic dermatitis, hypertension, diabetes mellitus, hyperlipidemia, urinary tract infection, gestational diabetes mellitus, preeclampsia or eclampsia, anxiety, sleep disorder, rheumatoid arthritis, systemic lupus erythematosus, and Sjogren's syndrome. The study applied database consists of four linked nationwide databases, NHIRD, BCAD, BRD, and NRD. Because the objective of this study was mainly to investigate the impact of maternal endometriosis on the neurodevelopmental disorders in offspring and the data were de-identified dataset, only the patient's age could be included in the analysis, other characteristics of father could not be obtained. All offspring were followed until the onset of disease, death, or December 31, 2019. The primary outcome was to analyze the relationship between maternal endometriosis and neurodevelopmental disorders in offspring in Taiwanese women. The neurodevelopmental disorders included developmental delay (ICD-9-CM:315, ICD-10-CM: F80-F89), cerebral palsy (ICD-9-CM:343, ICD-10-CM:G80), intellectual disabilities which was necessarily measured using IQ assessments in the offspring (ICD-9-CM:317–319, ICD-10-CM:F70-F79), and other neurodevelopmental and psychiatric disorders (ICD-9-CM:312, 313) including impulse disorder (ICD-10-CM:F63), conduct disorder (ICD-10-CM:F91), emotional disorders with onset specific to childhood (ICD-10-CM:F93), and disorders of social functioning with onset specific to childhood and adolescence (ICD-10-CM:F94). The secondary outcomes were further to explore the associations between maternal endometriosis and specific neurodevelopmental disorders in offspring using hazard ratios (HRs). In addition to the crude hazard ratio (HR), two adjusted models were used to evaluate the associations between maternal endometriosis and neurodevelopmental disorders in offspring. The baseline characteristics of the offspring of mothers with and without maternal endometriosis were compared. Absolute standardized differences (ASDs) were used to evaluate the difference between the case and comparison cohorts. An ASD <0.1 was considered to indicate similar characteristics between the two groups. The Kaplan–Meier method and log-rank test were used to compare the cumulative probability of offspring neurodevelopmental disorders between the case and comparison groups. A Cox proportional hazards model was used to estimate the associations of maternal endometriosis diagnosed before pregnancy and neurodevelopmental disorders in the offspring, which were reported as crude HRs and adjusted HRs (aHR) by adjusting associated covariates and 95 % confidence intervals (95 % CIs). Adjusted hazard ratio (aHR) was adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father's age, and mother's comorbidity in model 1. Moreover, adjusted hazard ratio (aHR) was adjusted for child's birth year, child's sex, mother's age at delivery, urbanization, insurance unit, father's age, mother's comorbidity, mode of delivery, birth weeks, birth weight, and Apgar score in model 2. Furthermore, the time-varying effect was used to analyze the effect of maternal endometriosis after the offspring were born on their neurodevelopment and was assessed at different age . The time-varying effect is exclusively analyzed using a Cox proportional hazards model for crude hazard, aHR in model 1 and aHR in model 2 in this study. A two-tail P value of less than 0.05 was regarded as a statically significant difference. SAS 9.4 (SAS Institute Inc., Cary, NC) was used for the statistical analyses.

The study's aim was to analyze the relationships between maternal endometriosis and neurodevelopmental disorders in offspring in Taiwanese women . The novel findings demonstrated that the offspring of mothers with a history of endometriosis had significantly higher risks of developmental delay and cerebral palsy in the long term follow-up than the offspring of mothers without endometriosis. However, our results did not demonstrate statistically significant associations between a history of maternal endometriosis and the incidence of intellectual disabilities which was necessarily measured using IQ assessments in the offspring. We speculated it may be because IQ assessments are more accurate for older children. Most main IQ tests were applied after 2 years olds such as Stanford-Binet Intelligence Scales or the Weschler Preschool and Primary Scale of Intelligence-III (WPPSI). For preschool-aged children, observing developmental milestones may provide a more reliable measure of cognitive abilities. Thus, intellectual disabilities could not be easily detected in children before 2 years old in our study. Despite the lack of an association between endometriosis and intellectual disabilities in our study, it cannot be definitively concluded that there was no association between maternal endometriosis and the incidence of intellectual disabilities in offspring. Moreover, this study also revealed no association between maternal endometriosis and other neurodevelopmental and psychiatric disorders in offspring. Our results suggest that developmental delay in offspring is associated with maternal endometriosis. MIA triggered by both acute and chronic systemic inflammation is hypothesized to be one of the mechanisms involved in neurodevelopmental disorders in offspring ( Jiang et al., 2018 ; Spann et al., 2018 ; Sun et al., 2022 ). An experimental study conducted in 2020 on adolescent rat offspring found that a single lipopolysaccharide injection in a MIA model increased cytokine levels and oxidative stress markers and caused deficits in pre- and postsynaptic proteins in the cerebral cortex. These changes disrupted synaptic structure, function, and plasticity, potentially leading to behavioral abnormalities linked to autism and other neurodevelopmental disorders ( Cieślik et al., 2020 ). Thus, developmental delay in offspring could be associated with MIA induced by maternal endometriosis. Several studies have found that certain cytokines, such as IL-1β, Il-6, and TNF-α, are expressed in various locations in women with endometriosis ( Boka et al., 1994 ; Copray et al., 2001 ; DeLeo et al., 1996 ; Heijmans-Antonissen et al., 2006 ; Klein et al., 1997 ; Schäfers et al., 2003 ). Endometriosis shares many similarities with autoimmune diseases such as rheumatoid arthritis, Crohn's disease, and psoriasis. These similarities include elevated levels of cytokines as described previously, decreased cell apoptosis, and T- and B-cell abnormalities ( Shigesi et al., 2019 ; Yoshii et al., 2021 ). The use of immunomodulators and inflammatory modulators have been proven to be an effective approach to the medical management of these autoimmune diseases, and similar therapies may prove useful in treating endometriosis ( Donnez and Cacciottola, 2022 ; Kapoor et al., 2021 ). Several large-scale studies have also revealed that advanced maternal age, maternal thyroid problems, preterm birth, assisted reproductive technology infertility treatments, multiple pregnancy, severe unintentional injury, and some infections or inflammation, such as measles, rubella, and severe bacterial infections, increase the risk of cerebral palsy in children ( Ahlin et al., 2013 ; Himmelmann et al., 2011 ; Nelson and Ellenberg, 1986 ; Ogoke, 2022 ). However, our study found that the offspring of mothers with a history of endometriosis had a higher crude HR and aHR (after model 1 and 2 adjustment) of cerebral palsy than offspring of mothers without endometriosis. The brain insult/injury causing cerebral palsy is non-progressive and can occur in the prenatal, perinatal, or postnatal periods. The etiology is often multifactorial. Some studies have indicated that perinatal pathology is not of much importance in the etiology of cerebral palsy. However, prenatal factors seem to be responsible for nearly 75 % of cases, whereas infant and neonatal-period risk factors account for 10 %–18 % of all cerebral palsy cases ( Ogoke, 2022 ; Paul et al., 2022 ). Several pathways have been found to be related to cerebral palsy, including prenatal inflammation, prematurity, twining, and placental pathologies ( Korzeniewsky et al., 2018 ; Ogoke, 2022 ). Of these, prenatal inflammation may be relevant to our study. Several studies involving neuroimaging and inflammatory markers, such as IL-6, IL-β, and TNF-α, have indicated that pronounced inflammation plays a vital role in the mechanism of cerebral palsy ( Yoon et al., 1997 , 2000 ). A cytokine-mediated cerebral white matter injury in preterm infants is also an implicated mechanism of cerebral palsy ( Ogoke, 2022 ). It is reasonable that our findings revealed that the offspring of mothers with a history of endometriosis had an elevated risk of cerebral palsy, because certain cytokines, such as IL-1β, IL-6, and TNF-α, are expressed in various locations in women with endometriosis ( Boka et al., 1994 ; Copray et al., 2001 ; DeLeo et al., 1996 ; Heijmans-Antonissen et al., 2006 ; Klein et al., 1997 ; Schäfers et al., 2003 ). Accumulating evidence indicates that maternal exposure to infectious and inflammatory insults is associated with the development of major neuropsychiatric disorders, such as bipolar disorder, schizophrenia, and autism spectrum disorder, in offspring ( Brown and Meyer, 2018 ). However, few studies have explored the relationships between maternal health and the incidence of other neuropsychiatric conditions in offspring, such as impulse disorders, conduct disorders, emotional disorders with childhood onset, and social functioning disorders that begin in childhood or adolescence. Our novel findings revealed no significant association between maternal endometriosis and these conditions in offspring. However, further research is needed to explore this area. Our study demonstrated the novel finding that in the nationwide population of Taiwan, the offspring of mothers with a history of endometrioses had higher incidence rates of developmental delay and cerebral palsy with a time-varying effect of maternal endometriosis. Furthermore, the significantly highest risk estimates for developmental delay and cerebral palsy were for the period from 12 to 48 months after the index date and the period from the index date to 36 months after the index date. We hypothesized that fetuses exposed to elevated maternal inflammatory conditions, such as maternal endometriosis, have elevated risks of developmental delay and cerebral palsy in childhood. Other potential mechanisms in developmental delay and cerebral palsy are similar to the susceptible genetic disorders of endometriosis that mothers and offspring may share in these disease. Endometriosis can run in families, so genetic factors (heritability) play a role in how it develops in some women but not in others. It has been reported that endometriosis exhibits complex genetic interactions and is inherited in a polygenic/multifactorial manner ( Rahmioglu et al., 2023 ). However, timely neurodevelopmental assessments are crucial for the early diagnosis and intervention of disorders to improve outcomes. Therefore, further mechanistic studies are needed to clarify the relationship of maternal endometriosis with early life changes in offspring that affect their risk of developing developmental delay and cerebral palsy in childhood. In summary, we infer some mechanisms by which maternal endometriosis leads to developmental impairment and cerebral palsy in the offspring. Inflammation is the core of endometriosis, and inflammatory cytokines and immune response in endometriosis primarily occurs in endometriotic stroma cells, endometriotic lesions and peritoneal fluid within peritoneal cavity ( Bhanoori et al., 2005 ; D'Hooghe et al., 2004 ; Kim et al., 2015 ; Thanatsis et al., 2021 ). Peritoneal macrophages can produce excessive cytokines (TNF-α, IL-1β, and IL-6) ( Brown and White, 2018 ; Doe, 2019 ). IL-6 has been found to can transfer through the placenta (Zaretsky et et al, 2004). Moreover, placenta extracellular vesicles in blood can interact with maternal endothelial cells and potentially transfer their contents correlated with inflammatory cytokines including TNF-α, IL-6, IFN-γ those are associated with endometriosis inflammation through placenta ( Black and Horowitz, 2018 ; Rodrigo and Glastras, 2018 ). The vascular/lymphatic spread theory, one of the pathophysiology of endometriosis may also explain the systemic dissemination form placenta to fetus via placenta ( Smith and Johnson, 2020 ; Brown and White, 2018 ). This has been further demonstrated by several studies that elevated cytokine levels were found in the umbilical cord, amniotic fluid, and fetal blood and then they were linked to cerebral palsy and impaired neurodevelopment ( Browns et al., 2014 ; Buka et al., 2001 ; Canetta et al., 2014 ). The most commonly implicated cytokines include TNF-α, IFN-γ, IL-1, IL-6, and IL-8 ( Andersen, 2022 ; Browns et al., 2014 ; Buka et al., 2001 ; Canetta et al., 2014 ; Garay et al., 2013 ; Jiang et al., 2018 ; Meyer, 2014 ). Moreover, in some related rodent models, MIA can not only induce pro-inflammatory cytokines in maternal serum but also increase the levels of IL-1, IL-6, IL-10, and TNF-α within the fetal brain compared with pregnant ones without inducing MIA ( Garay et al., 2013 ; Meyer, 2014 ). It was hypothesized that these cytokines could induce the occurrence of developmental delay and cerebral palsy in the offspring, whose mother had endometriosis history through placenta in this study. Because developmental milestones may provide a more reliable measure of cognitive abilities in preschool-aged children and cerebral palsy could be assessed after the offspring were born, the impacts of maternal endometriosis on the developmental delay and cerebral palsy in their offspring were assessed at different age after they were born. The follow-up duration was stratified into three-age based intervals: 0–12 months, 12–48 months, and beyond 48 months for developmental delay evaluation. For cerebral palsy assessment, two follow-up intervals were established: 0–36 months and beyond 36 months. Within each time interval, we applied separate Cox proportional hazard model to estimate HR and 95 % CI for the association between maternal endometriosis and developmental delay or cerebral palsy. This approach allows us to examine potential time-varying effects of maternal endometriosis exposure on the outcomes. The significantly highest risk estimates for developmental delay and cerebral palsy were demonstrated to be 12 to 48 months and index date to 36 months after the offspring were born, respectively. This study also has some limitations. First, the database used in this study lacks information on education, biochemical blood tests, immunological function, genetic background, and endometriosis treatment before pregnancy. Second, due to the limited study period, we could not evaluate the condition of the children after 10 years of age. Third, our study found no association between maternal endometriosis and the incidence of intellectual disabilities in offspring measured using IQ assessments. This may be because IQ assessments are more accurate for older children. Thus, intellectual disabilities could not be easily detected in preschool-aged children in our study. Despite the lack of an association between endometriosis and intellectual disabilities in our study, it cannot be definitively concluded that there is no association between maternal endometriosis and the incidence of intellectual disabilities in offspring. Fourth, a significant limitation of the study is the absence of biological data. A future study should be conducted to verify the biological factors caused by maternal endometriosis leading to neurodevelopmental disorders in the offspring. Finally, all of the data were from individuals of Chinese ethnicity, which restricts the external power of the study; thus, the results cannot be generalized to other ethnic groups. However, this study has some strengths. First, this was a nationwide cohort study with a large sample size and thus possesses high statistical power. Second, we emphasized on the ASD but not P value in considering the baseline characteristics of the mothers and offspring of the case cohort and the matched comparison cohort. When the sample size is extremely large, even minor differences can attain statistical significance ( P   < 0.05). In such situations, the statistical significance may not be of clinical relevance because these small differences may lack substantive impact. Therefore, a big data research typically includes the ASD ( Birkeland et al., 2017 ; Huang et al., 2023 ; Sullivan and Feinn, 2012 ), a statistical metric used to assess the differences between two cohorts that is not influenced by sample size, unlike the P value. Third , the database used in this study links birth registration, birth certificate, death registration, and national health insurance, providing detailed information on delivery, childbirth condition, medication, and parents’ basic data. Fourth, the data were obtained for newborns delivered from 2008 to 2019, which encompasses more than 10 years, a reliably long duration for tracing the time-to-event for neurodevelopmental disorders.

This study is the first to demonstrate that in Taiwanese women, the offspring of mothers with a history of endometriosis have higher risks of developmental delay and cerebral palsy in the long term follow-up than the offspring of mothers without endometriosis. It may be hypothesized that endometriosis associated chronic inflammation and immune response in mothers may elevate the levels of TNF-α, IL-6, IFN-γ. Then, placenta components mainly extracellular vesicles may interact with maternal endothelial cells and potentially transfer their contents correlated with inflammatory cytokines including TNF-α, IL-6, IFN-γ through placenta, as well as lead to the elevation of these cytokine levels in the umbilical cord, amniotic fluid, and fetal blood and brain, and thus induce the occurrence of developmental delay and cerebral palsy in their offspring. The significantly highest risk estimates for developmental delay and cerebral palsy were for the period from 12 to 48 months after the index date and the period from the index date to 36 months after the index date, respectively. The offspring, whose mothers have endometriosis history, have elevated risks of developmental delay and cerebral palsy in childhood and should be carefully followed, monitored and treated early, especially when they are in these high risk periods. These may prevent and reduce the occurrence of developmental delay and cerebral palsy. However, a randomized controlled trial should be conducted to verify this observational research. Moreover, this research lacks the biological data, and biological analyses are necessary to strengthen our findings by the demonstration of biological factors caused by maternal endometriosis that pass through the placenta into fetal blood and brain in the future.

Generative artificial intelligence was not used at any point during the writing of this manuscript.

Neurodevelopmental disorders are multifaceted conditions characterized by impairments in cognition, communication, behavior, and/or motor skills resulting from abnormal brain development ( Mullin et al., 2013 ). Timely neurodevelopmental assessments are crucial for diagnosing disorders such as intellectual disability, for which early intervention may improve outcomes. Disruptive, impulse-control, and conduct disorders, commonly diagnosed in childhood, stem from a combination of neural development and environmental factors during key developmental periods ( Dohrmann and Schneider, 2023 ). As we know, many studies have recognized that inflammation is an important contributor to central nervous system (CNS) injury in the developing brain ( Andersen, 2022 ; Deverman and Patternson, 2009 ; Jiang et al., 2018 ; Spann et al., 2018 ). The brain is especially vulnerable in utero and insults that occur during this critical period may lead to long term damage ( Andersen, 2022 ; Ellul et al., 2023 ; Jiang et al., 2018 ). In published studies, several neurodevelopmental disorders, including cerebral palsy, epilepsy, cognitive impairment, and autism spectrum disorder, have been linked to early life immune activation and inflammation ( Ellul et al., 2023 ; Knuesel et al., 2014 ; Kuban et al., 2015 ; Nelson et al., 1998 ). Maternal immune activation (MIA) is the activation of maternal innate and adaptive immune systems due to infection, stress, autoimmunity, asthma, allergies, or inflammation. MIA is hypothesized to play an important role in fetal neurodevelopment ( Ellul et al., 2023 ; Knuesel et al., 2014 ). Cells within the developing CNS use cytokines for autocrine and paracrine signaling, and as many of these cytokines also function as immune modulators, normal cytokine-mediated developmental processes are susceptible to disruption by cytokine imbalances ( Goeden et al., 2016 ; Nelson et al., 1998 ; Spann et al., 2018 ). Pro-inflammatory cytokine induction in response to maternal or early life infection has been shown to affect neurodevelopment ( O'Shea et al., 2014 ; Spann et al., 2018 ). In contrast, anti-inflammatory or regulatory cytokines have been reported to alleviate these adverse effects ( Andersen, 2022 ; Jiang et al., 2018 ; Muller and Schwarz, 2008 ). In some related rodent models, MIA can not only induce pro-inflammatory cytokines in maternal serum but also increase the levels of interleukin (IL)-1, IL-6, IL-10, and tumor necrosis factor-alpha (TNF-α) within the fetal brain compared with pregnant ones without inducing MIA ( Garay et al., 2013 ; Meyer, 2014 ). Observational studies have also reported that the expression levels of IL-6 and C-reactive protein in maternal third trimester are negatively associated with neonatal functional connectivity ( Browns et al., 2014 ; Buka et al., 2001 ). Moreover, epidemiological studies have linked elevated cytokine levels in the umbilical cord, amniotic fluid, and fetal blood to cerebral palsy and impaired neurodevelopment ( Browns et al., 2014 ; Buka et al., 2001 ; Canetta et al., 2014 ). The most commonly implicated cytokines include TNF-α, interferon gamma (IFN-γ), IL-1, IL-6, and IL-8 ( Andersen, 2022 ; Browns et al., 2014 ; Buka et al., 2001 ; Canetta et al., 2014 ; Garay et al., 2013 ; Jiang et al., 2018 ; Meyer, 2014 ). It is well known that endometriosis is a chronic inflammatory condition in which various inflammatory mediators are secreted, including prostaglandins, vascular endothelial growth factors, TNF-α, nerve growth factor, and ILs ( Donnez and Cacciottola, 2022 ; Kapoor et al., 2021 ; Machairiotis et al., 2021 ; Malvezzi et al., 2019 ). The exact pathophysiology of endometriosis is still unclear, but it may be caused by multiple factors, including genetic, immune, hormonal, and environmental factors ( Mariadas et al., 2025 ). There are several theories that attempt to explain the mechanisms of endometriosis development ( Brown and White, 2018 ; Smith and Johnson, 2020 ). The most widely known retrograde menstruation/endometriotic disease theory indicates the spread of endometrial cells to ectopic sites but cannot explain cases in nonmenstruating individuals. The vascular/lymphatic spread theory supports systemic dissemination, whereas coelomic/methothelial cell metaplasia theory represents transformation of peritoneal cells into endometrium-like tissues. Mülleran remnants/embryonic theory and genetic/epigenetic theory further influence disease susceptibility and progression. Repeated lesion bleeding promotes chronic inflammation, adhesion formation, and pain ( Brown and White, 2018 ; Doe, 2019 ). Inflammation is at the core of endometriosis, and immune dysfunction can exacerbate the severity of the condition. Peritoneal macrophages produce excessive cytokines (TNF-α, IL-1β, and IL-6), while reduced nature killer cells activity results in ectopic cells to escape immune clearance. NF-κB overactivation and generation of reactive oxygen species further damage tissues and exacerbate chronic pain and inflammation ( Brown and White, 2018 ; Doe, 2019 ) . The JAK/STAT pathway is involved in cytokine signaling and immune responses in endometriosis and the immune responses primarily occurs in endometriotic stromal cells, endometriotic lesions, and peritoneal fluid within peritoneal cavity ( Bhanoori et al., 2005 ; D'Hooghe et al., 2004 ; Kim et al., 2015 ; Thanatsis et al., 2021 ). The dysregulated pathways—PI3K/Akt, Wnt/β-catenin, and JAK/STAT—trigger proliferation, immune evasion, and angiogenesis ( Brown and White, 2018 ; Doe, 2019 ; Mariadas et al., 2025 ). IL-6 has been found to can transfer through the placenta ( Zaretsky et al., 2004 ). Moreover, placenta extracellular vesicles in blood can interact with maternal endothelial cells and potentially transfer their contents correlated with inflammatory cytokines including TNF-α, IL-6, IFN-γ those are associated with endometriosis inflammation through placenta ( Black and Horowitz, 2018 ; Rodrigo and Glastras, 2018 ), thus supposed to induce the neurodevelopmental disorders in their offspring. Some studies have found that MIA and certain cytokines, such as IL-1β, IL-6, and TNF-α, are also expressed in some locations of the nervous system, including the dorsal root ganglion, spinal cord, or injured nerves, and hence have been associated with various pathological processes including impaired neurodevelopment ( Boka et al., 1994 ; Copray et al., 2001 ; Deleo et al., 1996 ; Heijmans-Antonissen et al., 2006 ; Klein et al., 1997 ; Schäfers et al., 2003 ). Because of co-occurrence of inflammatory mediators/cytokines and immune response in endometriosis and neurodevelopmental disorders, connection between maternal endometriosis and neurodevelopmental disorders in offspring could be proposed. In this study, the primary objective was to analyze the relationship between maternal endometriosis and neurodevelopmental disorders in offspring in Taiwanese women.

I have nothing to declare.

The data used in this nationwide population-based cohort study were obtained from the Taiwan Maternal and Child Health Database, which comprises four linked nationwide databases, namely the National Health Insurance Research Database, the, the Birth Registration Database, and the National Register of Deaths. The research data are unavailable to access because they include sensitive or confidential information.