Introduction

Maternal prenatal and postnatal psychological distress, including depression and anxiety, are well-known risk factors for autism spectrum disorder (ASD). These variables impact children by means of fetal programming, identified in the theory of the developmental origins of health and disease1,2,3,4,5,6,7,8. However, there is no consistent evidence regarding which period of pregnancy is most susceptible to the impact of maternal psychological distress1,9,10,11,12. Previous studies have reported that postnatal maternal psychological distress is associated with ASD among children6,7,8. However, there are no reports in Japan that examine this issue together with prenatal and postnatal psychological distress in mothers. Furthermore, there are sex-specific differences in vulnerability to the effects of maternal psychological distress on children’s neurodevelopment1,13,14; and, to the best of our knowledge, no studies have examined sex-specific differences in the association between maternal psychological distress and ASD in children, either in Japan or internationally.

Therefore, this study aimed to examine the association between a combination of maternal prenatal and postnatal psychological distress and ASD in 6-year-old children, accounting for sex-specific differences, using data from the JECS15,16.

Methods

Design and participants

The JECS protocol has been previously described15,16. Recruitment to the JECS occurred between January 2011 and March 2014 and included pregnant women nationwide. The JECS is currently underway and plans to continue until children are 18 years old. In this study, we used the dataset of “fixed data up to age 6,” which was provided in March 2025. As this study investigated single pregnancies, records of twin or triplet pregnancies were excluded from the analysis.

Maternal psychological distress

The JECS administered the six-item version of the Kessler Psychological Distress Scale (K6) during the first (M-T1) and second (M-T2) halves of pregnancy and at 1 year postpartum (C-1y)13. The K6 has been widely used to assess psychological distress during the perinatal and postnatal periods17,18. According to the Diagnostic and Statistical Manual of Mental Disorders (4th Edition), it is a self-administered questionnaire comprising six questions that evaluate depressive mood and anxiety over the preceding 4 weeks, using a 0–4 scale. The total score is the sum of the six items ranging from 0 to 24. We used the Japanese version of the K6 with a cut-off of ≥ 5 to identify cases of psychological distress, as used in previous studies involving populations and affected communities in Japan12,13,19,20,21.

We classified the participants into eight trajectory groups based on K6 scores ≥ 5, and K6 scores ≤ 4 at M-T1, M-T2, and C-1y, as in our previous study (Table 1)13.

Table 1 The characteristics of participants.
Full size table

Outcome: autism spectrum disorder among 6-year-old children

Based on previous studies of JECS for ASD22,23,24,25, ASD up to age 6 was identified using self-reported data from the C-3y, C-4y, C-5y, and C-6y questionnaires (completed when children were 3, 4, 5, and 6 years old, respectively). Participants were asked the following question via the questionnaire: “Has your child ever been diagnosed by a doctor with autism spectrum disorder (e.g., autism, pervasive developmental disorder, Asperger’s syndrome) since the age of 3 (including ongoing visits and treatment)?” The response options were “yes” or “no”12,22,23,24,25. In Japan, as a national policy, pre-school health checks including screening for neurodevelopmental disorders are conducted for children aged 5–6 years. Children identified as suspected of having autism spectrum disorder (ASD) through pediatric examinations and parent interviews are referred to specialized medical institutions for further evaluation and diagnosis. While diagnoses based on parental reports may introduce variability in diagnostic criteria, it is assumed that most diagnoses in this study population were made with reference to the DSM-5, which was the standard diagnostic framework in Japan during the study period26.

Statistical analyses and covariables

We analyzed the data to determine the association between K6 scores of ≥ 5, and children diagnosed with ASD up to age 6. The control group consisted of children who had not been diagnosed with ASD up to age 6. Crude and multivariate logistic regression analyses were performed to obtain the 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 histories, 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 during pregnancy27, psychoactive drug use during pregnancy, folic supplement use during pregnancy, multivitamin supplement use during pregnancy, pregnancy complications, obstetric labor complications, birth weight of the children, chromosome abnormalities of the children, gestational age at delivery, and breastfeeding at 6 months postpartum. These covariates were also analyzed in both previous studies and in our study12,28,29,30,31. For each covariable, “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 nine of the 10 items referred to social communication difficulties. The cut-off of the AQ-J10 score was ≥ 727, which we defined as higher autistic traits among mothers in this study32. We conducted a maternal AQ-J10 assessment during the second half of pregnancy. Maternal neuropsychiatric disorders included depression, anxiety, obsessive–compulsive disorder, panic disorder, schizophrenia, epilepsy, migraine, autonomic dysreflexia, attention-deficit hyperactivity disorder (ADHD), learning disabilities, pervasive developmental disorders, Asperger’s syndrome, and ASD. No multicollinearity was found in this analysis (variance inflation factor < 2).

Second, multiple logistic regression analyses were adjusted for variables selected through a stepwise method, with the significance level for entry into the model set at 0.20 and for staying in the model at 0.15.

All statistical analyses were performed using SPSS version 25 (IBM Corp., Armonk, NY, USA).

Results

Of the 104,043 records in the provided dataset, records from 63,413 mothers and their children, 32,417 boys and 30,996 girls, were analyzed (Fig. 1). Among them, 1219 (3.76%) boys and 322 (1.04%) girls were diagnosed with ASD up to 6 years of age (Table 1); and overall, 2.43% of the children were diagnosed with ASD up to 6 years of age. Participants’ characteristics are presented in Table 1. At M-T1, the maternal prenatal K6 was estimated at medians of 15.0 (interquartile range 12.1–18.6) weeks of gestation for boys and 15.0 (interquartile range 12.3–18.6) weeks of gestation for girls. At M-T2, the maternal prenatal K6 was estimated at medians of 27.3 (interquartile range 25.1–30.0) weeks of gestation for boys and 27.3 (interquartile range 25.1–30.0) weeks of gestation for girls28.

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

Participant selection process flow chart.

The participants were divided into eight trajectory groups (Table 1). The results of the crude analyses are shown in Table 2.

Table 2 Maternal psychological distress (K6) and ASD among 6-year-old children; Crude analyses.
Full size table

Boys

Multivariate logistic regression analyses without the stepwise method demonstrated that the groups (1) maternal K6 score of ≥ 5 at M-T1 and K6 score ≤ 4 at both M-T2 and C-1y (adjusted OR [AOR] 1.23; 95% CI 1.01–1.50); (2) maternal K6 score of ≥ 5 at C-1y and K6 score ≤ 4 at both M-T1 and M-T2 (AOR 1.51; 95% CI 1.20–1.90); (3) maternal score of ≥ 5 at both M-T1 and M-T2 and K6 score ≤ 4 at C-1y (AOR 1.32; 95% CI 1.07–1.64); (4) maternal K6 score ≤ 4 at M-T2 and score of ≥ 5 at both M-T1 and C-1y (AOR 1.68; 95% CI 1.29–2.20); (5) maternal K6 score ≤ 4 at M-T1 and score of ≥ 5 at both M-T2 and C-1y (AOR 1.85; 95% CI 1.37–2.49); and (6) maternal K6 score of ≥ 5 at M-T1 and M-T2 and C-1y (AOR 1.72; 95% CI 1.43–2.07), were associated with ASD among 6-year-old boys compared with the group with a K6 score ≤ 4 for both M-T1 and M-T2 and C-1y as reference (Table 3).

Table 3 Maternal psychological distress (K6) and ASD among 6-year-old children; Multivariate logistic regression analyses.
Full size table

Multivariate logistic regression analyses with the stepwise method demonstrated that the groups (1) maternal K6 score of ≥ 5 at M-T1 and K6 score ≤ 4 at both M-T2 and C-1y (AOR 1.24; 95% CI 1.01–1.51); (2) maternal K6 score of ≥ 5 at C-1y and K6 score ≤ 4 at both M-T1 and M-T2 (AOR1.54; 95% CI 1.22–1.93); (3) maternal score of ≥ 5 at both M-T1 and M-T2 and K6 score ≤ 4 at C-1y (AOR 1.31; 95% CI 1.06–1.62); (4) maternal K6 score ≤ 4 at M-T2 and score of ≥ 5 at both M-T1 and C-1y (AOR 1.70; 95% CI 1.30–2.22); (5) maternal K6 score ≤ 4 at M-T1 and score of ≥ 5 at both M-T2 and C-1y (AOR 1.86; 95% CI 1.38–2.51); and (6) maternal K6 score of ≥ 5 at M-T1 and M-T2 and C-1y (AOR 1.74; 95% CI 1.45–2.08), were associated with ASD among 6-year-old boys compared with the group with a K6 score ≤ 4 for both M-T1 and M-T2 and C-1y as reference (Table 3).

Girls

Multivariate logistic regression analyses without the stepwise method demonstrated that the groups (1) maternal K6 score ≤ 4 at M-T2 and score of ≥ 5 at both M-T1 and C-1y (AOR 1.87; 95% CI 1.15–3.06); (2) maternal K6 score ≤ 4 at M-T1 and score of ≥ 5 at both M-T2 and C-1y (AOR 1.81; 95% CI 1.03–3.17); and (3) maternal K6 score of ≥ 5 at M-T1 and M-T2 and C-1y (AOR 1.91; 95% CI 1.37–2.66), were associated with ASD among 6-year-old girls compared with the group with a K6 score ≤ 4 for both M-T1 and M-T2 and C-1y as reference (Table 3).

Multivariate logistic regression analyses with the stepwise method demonstrated that the groups (1) maternal K6 score ≤ 4 at M-T2 and score of ≥ 5 at both M-T1 and C-1y (AOR 1.93; 95% CI 1.18–3.14); (2) maternal K6 score ≤ 4 at M-T1 and score of ≥ 5 at both M-T2 and C-1y (AOR 1.82; 95% CI 1.04–3.17); and (3) maternal K6 score of ≥ 5 at M-T1 and M-T2 and C-1y (AOR 2.06; 95% CI 1.49–2.84), were associated with ASD among 6-year-old girls compared with the group with a K6 score ≤ 4 for both M-T1 and M-T2 and C-1y as reference (Table 3).

Discussion

This study demonstrated an association between maternal psychological distress and ASD among 6-year-old children, considering maternal psychological distress during the first and second halves of pregnancy and 1 year postpartum. To the best of our knowledge, this is the first study to examine this issue separately in boys and girls.

For boys, notably, (1) maternal psychological distress in the first half of pregnancy alone (median of 15 weeks of gestation) was associated with ASD, and (2) maternal psychological distress at 1 year postpartum alone was associated with ASD.

For girls, maternal psychological distress coexisting during pregnancy and at 1 year postpartum was associated with ASD. However, there was no significant association between maternal psychological distress during pregnancy (even if mothers had psychological distress at both the first and second halves of pregnancy) and ASD if there was no maternal psychological distress at 1 year postpartum. Furthermore, maternal psychological distress at 1 year postpartum alone was not significantly associated with ASD.

First, we discussed the sex-specific differences. We conducted sex-stratified analyses based on multiple lines of evidence. ASD is significantly more common in boys (approximately four times more than in girls), suggesting the possibility of differences in biological susceptibility and developmental pathways between the sexes33,34,35. Further, in this study, 3.76% of boys and 1.04% of girls were diagnosed with ASD up to 6 years of age. Furthermore, research on other neurodevelopmental outcomes suggests that the effects of perinatal environmental exposures, including maternal psychological distress and stress, on children’s neurodevelopment may differ by sex1,36,37,38,39. These findings indicate that the impact of maternal psychological distress on ASD risk varies depending on sex-specific susceptibility. Our study found that in boys, even if the mother had no psychological distress at 1 year postpartum, ASD was associated with maternal psychological distress (1) in the first half of pregnancy alone or (2) in both the first and second halves of pregnancy. However, in girls, even when mothers had psychological distress in both the first and second halves of pregnancy, there was no significant association with ASD if there was no maternal psychological distress at 1 year postpartum. This suggests that boys may be more susceptible to the effects of maternal psychological distress during pregnancy than girls. The mechanism underlying this sex-specific difference remains unclear, and further studies are required. To the best of our knowledge, this topic has not been investigated previously. For reference, although unrelated to ASD, there are reports indicating that boys are more sensitive than girls to maternal psychological distress during pregnancy, in relation to various neurodevelopmental disorders in children. A longitudinal study in Australia, which followed prenatal exposure to the Queensland Flood, reported that higher levels of objective flood exposure during pregnancy were associated with irritable temperament in boys, but not in girls40. A prospective case–control study in France reported that prenatal maternal depression was associated with generalized anxiety, activity/impulsivity, and sleep problems, particularly in boys41. A community-based birth cohort study in the Netherlands reported that prenatal maternal anxiety was associated with hyperactivity/inattention in boys42. A prospective longitudinal study in China reported that severe stress in the second trimester had a significantly increased risk of developing ADHD symptoms in boys, but not in girls43. A population-based cohort study in Denmark suggested that maternal exposure to severe stress during pregnancy increased the risk of ADHD in boys, but not in girls44. A longitudinal study in the United States reported that prenatal exposure to Sandy Storms distinctively elevated the risk of attention-deficit/disruptive behavioral disorders, including ADHD, conduct disorder, and oppositional-defiant disorder, among boys59. Our previous JECS study found that persistent maternal psychological distress from the first half of pregnancy to 1 year postpartum had a disadvantageous association with verbal cognitive development in boys, but not in girls13.

Second, we discussed the specific period of prenatal maternal psychological distress associated with ASD in children. Unfortunately, we found no previous studies on sex-specific differences for this issue; therefore, the following discussion treats the findings of the study as overall children. Previous studies have reported inconsistent findings on which period of pregnancy is more susceptible to psychological distress (including maternal stress, depression, and anxiety), regarding the risk of the child developing ASD1,9,10,11. A population-based cohort study in Canada, which followed prenatal exposure to ice storms in Quebec, suggested that first trimester (definition is up to 13 weeks of gestation) prenatal stress increased the risk of ASD45. A cohort study in China suggested that the sensitive period for exposure to prenatal stress might be the second trimester (definition is from 14 to 27 weeks of gestation) for the risk of autistic-like behaviors58. A population-based cohort study in the United States, which followed prenatal exposure to hurricanes and tropical storms in Louisiana, suggested that a gestational age of 5–6 months during storm or hurricane exposure increased the risk of ASD46. A retrospective study in United States suggested that the peak prevalence for prenatal stressors in ASD was between 25 and 28 weeks of gestation60. A population-based cohort study in Sweden suggested that the third trimester (defined as 28 or later weeks of gestation) increased the risk of ASD28. Our study demonstrated that maternal psychological distress in the first half of pregnancy (first to second trimester) (median, 15.0; interquartile range 12.1–18.6 weeks of gestation) alone was associated with ASD in boys. However, maternal psychological distress in the second half of pregnancy (second to third trimester) (median, 27.3; interquartile range 25.1–30.0 weeks of gestation) alone was not significantly associated with ASD in boys. Our findings suggest that in boys, maternal distress in the first half of pregnancy (first to second trimester) is more sensitive to the risk of ASD than in the second half of pregnancy (second to third trimester).

Third, we discussed the association between postnatal maternal psychological distress and ASD in children. Our study demonstrated that maternal psychological distress, including depression and anxiety, at 1 year postpartum alone was associated with ASD in boys. Unfortunately, we found no previous studies on sex-specific differences for this issue; therefore, the following discussion considers the findings of the study to refer to children in general. These results were consistent with those of previous studies. A nationwide study in Sweden, using the data from the Swedish Medical Birth register, demonstrated that mothers of children with ASD had high rates of depression47. A case–control study in Turkey suggested that maternal postpartum depression increased the risk of ASD among their children50. A nationwide study in Taiwan, using the data from the Taiwan National Health Insurance Research Database, demonstrated that maternal postpartum depression was associated with the risk of ASD among children6. Hence, the following hypotheses are made regarding this association. The genetic vulnerability of mothers who experience postpartum depression may be associated with ASD risk in their children48,49. In addition, maternal role dysfunctions associated with postpartum depression, such as bonding failure in children and mutual mother–child attachment failure, may be associated with an increased likelihood of subsequent ASD50,51,52.

Clinical implications: Our findings suggest that a one-size-fits-all approach to preventing maternal prenatal/postnatal psychological distress may be insufficient. Rather, patterns of sex-specific vulnerability indicate that targeted interventions that consider the timing of exposure during the perinatal period and the child’s sex may be more effective. In this study, the risk of ASD was observed in the group with maternal psychological distress in the first half of pregnancy (first to second trimester) for boys. This indicates the importance of preventing maternal psychological distress, including depression and anxiety before conception, to reduce the risk of children developing ASD. Moreover, maternal psychological distress at 1 year postpartum alone is associated with ASD in boys. Therefore, even if the mother does not experience psychological distress during pregnancy, proper measures may be important to prevent maternal postpartum psychological distress and reduce the risk of ASD in boys. For girls, maternal psychological distress coexisting during pregnancy and at 1 year postpartum was associated with ASD. However, there was no significant association between maternal psychological distress during pregnancy and ASD if there was no maternal psychological distress at 1 year postpartum. This suggests that even if a mother experiences psychological distress during pregnancy, the risk of her daughter developing ASD may be reduced by preventing postpartum onset of maternal psychological distress.

This study has some limitations. First, the K6 is a self-administered questionnaire. Hence, this implies that psychological distress was not medically diagnosed. Second, gene polymorphisms were not studied. Third, the diagnosis of ASD by physicians was based on the participants’ self-reports. It is assumed that most diagnoses in this study population were made with reference to the DSM-5. However, in our study in Japan, the prevalence of ASD diagnosis among 6-year-old children was 2.43% (boys, 3.76%; girls, 1.04%). The global prevalence estimate of overall ASD is approximately 2%53,54,55,56. A recent epidemiological study in Japan reported that the prevalence of ASD at 5 years of age was estimated to be 3.22%57; and these prevalence rates are not significantly different from our results.

The strength of this study is that it is the first analysis in Japan of the effects of a combination of prenatal and postnatal maternal psychological distress on children with ASD, using a large sample from a nationwide birth cohort study. In addition, we identified differences between boys and girls in the risk of development of ASD, according to the timing of maternal psychological distress.

Conclusion

Maternal psychological distress during the first half of pregnancy is associated with ASD in boys. This finding suggests that the risk of ASD because of maternal distress during pregnancy is high in the first half of pregnancy in boys. This indicates the importance of preventing maternal psychological distress, including depression and anxiety, even preconception, to reduce the risk of the development of ASD among children.

Moreover, maternal psychological distress at 1 year postpartum is associated with ASD in boys. For girls, maternal psychological distress coexisting during pregnancy and at 1 year postpartum was associated with ASD in girls. However, there was no significant association between maternal psychological distress during pregnancy and ASD if there was no maternal psychological distress at 1 year postpartum. This indicates the need to prevent the onset of postpartum maternal psychological distress to reduce the risk of children developing ASD.