Introduction
Maternal prenatal psychological distress, which includes depression and anxiety, is known to be a risk factor for autism spectrum disorder (ASD) among children through fetal programing.Reference Beversdorf, Stevens, Margolis and Van de Water1,Reference Fine, Zhang and Stevens2,Reference Robinson, Lahti-Pulkkinen, Heinonen, Reynolds and Räikkönen3,Reference Scheinost, Sinha and Cross4 However, there is no consistent knowledge regarding at which term during pregnancy psychological distress affects the risk of ASD among children.Reference Beversdorf, Stevens and Jones5 Despite these issues, no independent studies on this topic have been conducted in Japan.
Over the recent years, Japan has been conducting the Japan Environment and Children’s Study (JECS), which is a nationwide birth cohort study involving 100,000 pairs of parents and their children, to investigate children’s development and environment.Reference Kawamoto, Nitta and Murata6.Reference Michikawa, Nitta and Nakayama7 In this study, we used this dataset to examine the association between maternal prenatal psychological distress and ASD among 3-year-old children.
Materials and methods
Design and participants
The JECS protocol has been described previously.Reference Kawamoto, Nitta and Murata6,Reference Michikawa, Nitta and Nakayama7 Recruitment to the JECS occurred between January 2011 and March 2014, and it included pregnant women nationwide. The JECS is currently underway and plans to continue until the children are 13 years of age. A dataset containing the results of this test for all 3-year-old children was provided in 2021. In this study, we used the jecs-ta-20190930 dataset, which was revised in June 2021. Among the 104,062 records in this dataset, the records of 78,745 women were analyzed. Because this study investigated single pregnancies, records of twin or triplet pregnancies were excluded from the analysis.
Maternal psychological distress
The JECS protocol was designed to administer the six-item Kessler Psychological Distress Scale (K6) twice during the pregnancy: the first (M-T1) and the second (M-T2) half of pregnancy.Reference Michikawa, Nitta and Nakayama7 The K6 has been used widely to assess psychological distress during the perinatal and postnatal periods.Reference Kessler, Andrews and Colpe8,Reference Kessler, Barker and Colpe9 It is a self-administered questionnaire comprising six questions that evaluate depressive moods and anxiety according to the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV), over the preceding four weeks on a scale of 0 to 4. The total score was the sum of the six items, and it ranged from 0 to 24. We used the Japanese version of the K6 with a cutoff of ≥5 to identify cases of psychological distress, as used in previous studies involving populations and affected communities in Japan.Reference Furukawa, Kawakami and Saitoh10,Reference Kuroda, Goto and Koyama11,Reference Sakurai, Nishi, Kondo, Yanagida and Kawakami12
We classified the participants into four groups based on their K6 scores of ≥5 at M-T1 and M-T2: (1) K6 scores of ≤4 at M-T1 and M-T2, (2) K6 scores of ≤4 at M-T1 and ≥5 at M-T2, (3) K6 scores of ≥5 at M-T1 and ≤4 at M-T2, and (4) K6 scores of ≥5 at M-T1 and M-T2.
Outcome: ASD among 3-year-old children
Based on data obtained from the C-3y questionnaire (when the child was three years of age), which was self-reported by the participants, we estimated the incidence of ASD among 3-year-old children. In the questionnaire, caregivers were asked, “Has your child been diagnosed with ASD by physicians?”, and children whose parents answered “Yes” were defined as diagnosed with ASD. ASD was diagnosed among children aged between 2 and 3 years. The diagnostic categories for neurodevelopmental disorders, which caregivers were asked, were based on the International Statistical Classification of Diseases and Related Health Problems Tenth Revision (ICD-10) (codes: F84.0, childhood autism; F84.1, atypical autism; F84.5, Asperger’s syndrome; F84.8, other pervasive developmental disorders; F84.9, pervasive developmental disorder unspecified; F84.2, Rett syndrome; and F84.3, childhood disintegrative disorder).
Statistical analysis and covariables
We analyzed the data to determine the association between K6 scores of ≥5 and ASD among 3-year-old children. Crude and multivariate logistic regression analyses were used to obtain odds ratios (ORs) and 95% confidence intervals (CI). The multivariate logistic regression analyses were adjusted for maternal age at delivery, paternal age at conception, maternal body mass index (kg/m2) before pregnancy, parity, marital status, treatment for infertility, unexpected pregnancies, maternal and paternal academic history, maternal job during pregnancy, maternal and paternal smoking during pregnancy, maternal alcohol consumption during pregnancy, household income (×103 yen/year) during pregnancy, maternal neuropsychiatric disorders, maternal autism spectrum quotient Japanese version (AQ-J) 10 ≥7,Reference Kurita, Koyama and Osada13 psychoactive drug use, folic supplement use, multivitamin supplement use, pregnancy complications, obstetric labor complications, intrauterine infections, gender of the child, birth weight of the child, chromosome abnormalities of the child, gestational week, and breastfeeding. These covariates were also analyzed in previous studies.Reference Class, Abel and Khashan14,Reference Li, Vestergaard and Obel15,Reference Li, Francis, Hinkle, Ajjarapu and Zhang16,Reference Roberts, Lyall, Rich-Edwards, Ascherio and Weisskopf17 For each confounder, “no answer” was analyzed as a single item. The AQ-J10, which is a self-reported questionnaire, was designed to measure autistic traits distributed among the general population, and 9 of the 10 items referred to social communication difficulties. The cutoff of the AQ-J10 score was ≥7,Reference Kurita, Koyama and Osada13 which we defined as higher autistic traits among mothers in this study.Reference Hosozawa, Cable and Ikeda18 We conducted maternal AQ-J10 during the second or third trimester. Maternal neuropsychiatric disorders included depression, anxiety disorders, obsessive-compulsive disorder, panic disorder, schizophrenia, epilepsy, migraines, autonomic dysreflexia, attention deficit hyperactivity disorder, learning disabilities, pervasive developmental disorders, Asperger’s syndrome, and ASD.
All statistical analyses were performed using SPSS version 27 (IBM Corp., Armonk, NY, USA).
Results
Of the 104,062 records in the provided dataset, records from 78,745 children were analyzed (Fig. 1). The characteristics of the participants are listed in Table 1. At M-T1, the maternal prenatal K6 was estimated at a median of 15.1 (interquartile range 12.3–18.9) weeks of gestation. At M-T2, it was estimated at a median of 27.4 (interquartile range 25.3–30.1) weeks of gestation.Reference Iwai-Shimada, Nakayama and Isobe19
The flow chart for selected research participants.
The characteristics of participants
Abbreviations: Autism spectrum disorder (ASD); body mass index (BMI; kg/m2); the Kessler 6-item psychological distress scale (K6).
M-T1: median 15.1 (interquartile range 12.3–18.9) pregnant weeks; M-T2: median 27.4 (interquartile range 25.3–30.1) pregnant weeks.
Participants were divided into four groups: (1) 46,463 mothers (59.0%) had K6 scores of ≤4 at M-T1 and M-T2, (2) 7697 mothers (9.8%) had K6 scores of ≤4 at M-T1 and ≥5 at M-T2, (3) 10,629 mothers (13.5%) had K6 scores of ≥5 at M-T1 and ≤4 at M-T2, and (4) 13,956 mothers (17.7%) had K6 scores of ≥5 at both M-T1 and M-T2. A total of 355 children (0.45%) were diagnosed with ASD, which was based on the survey questions presented to the participants.
Multivariate logistic regression analyses demonstrated that a maternal K6 score of ≥5 at both M-T1 and M-T2 was significantly associated with ASD among 3-year-old children (adjusted odds ratio [AOR], 1.440; 95% CI, 1.104–1.877) compared to a maternal K6 score of ≤4 at both M-T1 and M-T2 (Table 2). There was no significant difference in the group with a maternal K6 score of ≥5 only at either M-T1 or M-T2 (Table 2).
Maternal K6 and ASD among 3-year-old children (n = 78,745)
Abbreviations: The Kessler 6-item psychological distress scale (K6), autism spectrum disorder (ASD), crude odds ratio (COR), confidence interval (CI), adjusted odds ratio (AOR).
M-T1: median 15.1 (interquartile range 12.3–18.9) pregnant weeks; M-T2: median 27.4 (interquartile range 25.3–30.1) pregnant weeks.
Adjusted for maternal age at delivery, paternal age at conception, maternal body mass index (kg/m2) before pregnancy, parity, marital status, treatment for infertility, unexpected pregnancies, maternal academic history, paternal academic history, maternal job during pregnancy, maternal smoking during pregnancy, paternal smoking during pregnancy, maternal alcohol consumption during pregnancy, household income during pregnancy, maternal neuropsychiatric disorders, maternal Autism Spectrum Quotient Japanese version 10 ≥7, psychoactive drugs use, folic supplements use, multivitamin supplements use, diabetes or gestational diabetes, pregnancy complications, obstetric labor complications, intrauterine infections, gender of children, birth weight of children, chromosome abnormalities of children, gestation week, and breast feeding.
Discussion
The group with continuous maternal psychological distress from the first to the second half of pregnancy showed a risk of ASD among 3-year-old children compared to other groups. Contrarily, in the group without persistent maternal psychological distress during pregnancy, there was no significant association with ASD among 3-year-old children.
Previous studies reported inconsistent findings on the impact of maternal stress during pregnancy on the risk of ASD among children.Reference Beversdorf, Stevens, Margolis and Van de Water1 A population-based cohort study, which followed prenatal exposure to ice storms in Quebec, Canada, suggested that first trimester prenatal objective stress increased the risk of ASD.Reference Walder, Laplante, Sousa-Pires, Veru, Brunet and King20 A population-based cohort study that followed prenatal exposure to hurricanes and tropical storms in Louisiana, United States suggested a significantly increased risk of ASD at a gestational age of 5–6 months during storm or hurricane exposure.Reference Kinney, Miller, Crowley, Huang and Gerber21 A retrospective survey in the United States suggested that a higher prevalence of prenatal stressors was found in ASD at 21–32 weeks of gestation, with a peak at 25–28 weeks.Reference Beversdorf, Manning and Hillier22 A population-based cohort study in Sweden suggested that third-trimester prenatal stress increased the risk of ASD.Reference Class, Abel and Khashan14 A cohort study in China suggested that the second-trimester might be the sensitive period for exposure to prenatal stress, thereby increasing the risk of autistic-like behaviors.Reference Chen, Strol and Wy23
In our nationwide birth cohort study, the risk of ASD was only observed in the group with maternal psychological distress at both approximately 15 and 27 weeks of pregnancy. This indicated that if psychological distress did not continue during pregnancy, the risk of the onset of ASD among the children could be reduced.
This study has certain limitations. First, the K6 was a self-administered questionnaire. Hence, it did not mean that the psychological distress was medically diagnosed. Second, gene polymorphisms were not studied. Third, in this study, the diagnosis of ASD by the physicians was based on participant self-reports. Diagnostic categories of the questionnaire were based on ICD-10. Therefore, misclassification may have occurred as the data were self-reported. However, our study targeted 3-year-old children in Japan, and the prevalence of ASD diagnosis was 0.45%. In a previous Japanese study that evaluated the cumulative incidence of ASD using the ICD-10 among 5-year-old children, the prevalence of ASD was 0.27%.Reference Honda, Shimizu, Imai and Nitto24 Previous population-based cohort studies conducted in Denmark,Reference Li, Francis, Hinkle, Ajjarapu and Zhang16 Sweden,Reference Li, Vestergaard and Obel15 and the United StatesReference Roberts, Lyall, Rich-Edwards, Ascherio and Weisskopf17 reported that the prevalence of ASD was approximately 0.16–0.83%. These prevalences were regarded as not significantly deviating from our results.
The strength of this study lies in the fact that it is the first to analyze the effect of maternal psychological distress on ASD among children in Japan using a large sample from a nationwide birth cohort study.
Conclusion
Continuous maternal psychological distress from the first to the second half of pregnancy was associated with ASD among 3-year-old children. Contrarily, in the group without persistent maternal psychological distress during pregnancy, there was no significant association. This indicates that if maternal mental health is assessed during the early stages of a pregnancy and psychological distress is detected, it may be possible to prevent negative effects on the children through appropriate interventions to improve the distressing situations experienced by mothers.
Supplementary materials
For supplementary material for this article, please visit https://doi.org/10.1017/S2040174422000411
Acknowledgments
The authors are grateful to all the study participants. The members of the JECS group as of 2021 are as follows: Michihiro Kamijima (principal investigator, Nagoya City University, Nagoya, Japan), Shin Yamazaki (National Institute for Environmental Studies, Tsukuba, Japan), Yukihiro Ohya (National Center for Child Health and Development, Tokyo, Japan), Reiko Kishi (Hokkaido University, Sapporo, Japan), Nobuo Yaegashi (Tohoku University, Sendai, Japan), Koichi Hashimoto (Fukushima Medical University, Fukushima, Japan), Chisato Mori (Chiba University, Chiba, Japan), Shuichi Ito (Yokohama City University, Yokohama, Japan), Zentaro Yamagata (University of Yamanashi, Chuo, Japan), Hidekuni Inadera (University of Toyama, Toyama, Japan), Takeo Nakayama (Kyoto University, Kyoto, Japan), Hiroyasu Iso (Osaka University, Suita, Japan), Masayuki Shima (Hyogo Medical University, Nishinomiya, Japan), Hiroshige Nakamura (Tottori University, Yonago, Japan), Narufumi Suganuma (Kochi University, Nankoku, Japan), Koichi Kusuhara (University of Occupational and Environmental Health, Kitakyushu, Japan), and Takahiko Katoh (Kumamoto University, Kumamoto, Japan).
Financial support
This study was funded by the Ministry of the Environment, Japan. The findings and conclusions of this article are solely the opinions of the authors, and do not represent the official views of the above government.
Conflict of interest
None.
Ethical standards
This study was conducted according to the guidelines laid down in the Declaration of Helsinki. It was reviewed and approved by the Ministry of the Environment’s Institutional Review Board on Epidemiological Studies (no. 100910001) and by the ethics committees of all participating institutions. Written informed consent was obtained from all participants.
