Introduction
Autism spectrum disorder (ASD) and attention-deficit/hyperactivity disorder (ADHD) are neurodevelopmental disorders that exhibit high comorbidity rates with psychiatric disorders (Hossain et al., Reference Hossain, Khan, Sultana, Ma, McKyer, Ahmed and Purohit2020; Larson, Russ, Kahn, & Halfon, Reference Larson, Russ, Kahn and Halfon2011; Simonoff et al., Reference Simonoff, Pickles, Charman, Chandler, Loucas and Baird2008). Such comorbidities have also been observed in mild cases that do not meet the full diagnostic criteria of these disorders (Kim et al., Reference Kim, Cho, Kim, Kim, Shin and Kim2009; Lundström et al., Reference Lundström, Chang, Kerekes, Gumpert, Råstam, Gillberg, Lichtenstein and Anckarsäter2011; Rai et al., Reference Rai, Culpin, Heuvelman, Magnusson, Carpenter, Jones, Emond, Zammit, Golding and Pearson2018), indicating that children with subthreshold symptoms often exhibit equivalent clinical support needs (Balázs & Keresztény, Reference Balázs and Keresztény2014; Kamio et al., Reference Kamio, Moriwaki, Takei, Inada, Inokuchi, Takahashi and Nakahachi2013).
As mental-health–related problems during school age have long-term effects on educational, occupational, and social functioning in adulthood (Copeland, Wolke, Shanahan, & Costello, Reference Copeland, Wolke, Shanahan and Costello2015), early intervention is a critical public health priority. However, the specific temporal pathways linking neurodevelopmental disorder traits to secondary mental health problems and the mechanisms for this association remain poorly understood. Moreover, minimal research has been conducted on children with subthreshold symptoms, and their support systems remain underdeveloped.
This study addresses this gap by using large-scale longitudinal data from Japanese elementary and middle-school students to examine longitudinal associations between ASD and ADHD traits, social adjustment, and mental health outcomes from childhood through adolescence. By treating neurodevelopmental characteristics as continuous spectra and examining their effects in a nonclinical population, this study aims to inform early intervention strategies and preventive support practices beyond traditional diagnostic frameworks.
Comorbidities of ASD/ADHD traits and mental health problems
A British study of children aged 10–14 years found that 70% of children with ASD had at least one additional diagnosis per the Diagnostic and Statistical Manual of Mental Disorders (DSM), whereas 40% had two or more comorbid conditions, with anxiety disorders and oppositional defiant disorder being particularly prevalent (Simonoff et al., Reference Simonoff, Pickles, Charman, Chandler, Loucas and Baird2008). Similarly, a large-scale epidemiological study in the United States examining children aged 6–17 years revealed that only 33.1% of children who had been diagnosed with ADHD presented without comorbidities or complications, with most exhibiting co-occurring conditions such as conduct disorder, anxiety disorders, and depression (Larson et al., Reference Larson, Russ, Kahn and Halfon2011).
Comparable patterns have been observed for subthreshold ASD and ADHD traits, which are associated with elevated risks of anxiety, depressive symptoms, and externalizing problems (Kim et al., Reference Kim, Cho, Kim, Kim, Shin and Kim2009; Lewinsohn, Shankman, Gau, & Klein, Reference Lewinsohn, Shankman, Gau and Klein2004; Lundström et al., Reference Lundström, Chang, Kerekes, Gumpert, Råstam, Gillberg, Lichtenstein and Anckarsäter2011; Moriwaki & Kamio, Reference Moriwaki and Kamio2013; Noda et al., Reference Noda, Okada, Tani, Ohnishi, Naoto, Nakajima and Tsujii2013; Rai et al., Reference Rai, Culpin, Heuvelman, Magnusson, Carpenter, Jones, Emond, Zammit, Golding and Pearson2018). These findings highlight the importance of examining mental health among children who present with elevated traits but do not meet diagnostic thresholds.
Processes linking ASD/ADHD traits and mental health
Although elevated ASD and ADHD traits are associated with increased mental health difficulties, the pathways underlying these associations remain unclear. One possible explanation for this link is that neurodevelopmental traits and psychiatric symptoms share biological underpinnings. For example, family and twin studies suggest genetic overlap between ADHD tendencies and emotional problems (Chen et al., Reference Chen, Ji, Wang, Lichtenstein, Larsson and Chang2016; Cole et al., Reference Cole, Ball, Martin, Scourfield and McGuffin2009). However, genetic explanations alone may not fully account for the heterogeneity and high co-occurrence of symptoms, and evidence for shared biological mechanisms remains inconclusive.
Another possibility is that neurodevelopmental traits impede social adaptation, resulting in secondary mental health problems. For example, children with elevated ASD traits often experience challenges in peer interactions (Kasari, Locke, Gulsrud, & Rotheram-Fuller, Reference Kasari, Locke, Gulsrud and Rotheram-Fuller2011; Whitehouse, Durkin, Jaquet, & Ziatas, Reference Whitehouse, Durkin, Jaquet and Ziatas2009). Similarly, children with elevated ADHD traits are prone to academic underachievement and interpersonal difficulties owing to concentration problems and impulsivity (Herman, Lambert, Ialongo, & Ostrander, Reference Herman, Lambert, Ialongo and Ostrander2007; Humphreys et al., Reference Humphreys, Katz, Lee, Hammen, Brennan and Najman2013; Powell et al., Reference Powell, Riglin, Hammerton, Eyre, Martin, Anney, Thapar and Rice2020). Such adaptive difficulties may lead to decreased self-esteem, social isolation, and chronic stress, potentially resulting in secondary emotional problems. Nonetheless, causal pathways cannot be presumed on the basis of these associations alone, and alternative explanations, including moderation by contextual factors, remain plausible.
Social domains relevant to adjustment
Considering the practical importance of identifying school-based targets for support, this study focuses on three domains of social adjustment that have been consistently linked with ASD/ADHD traits and mental health outcomes.
Peer relationships
Children with ASD tend to have fewer friendships and lower friendship quality and often occupy peripheral positions in classroom social networks (Kasari et al., Reference Kasari, Locke, Gulsrud and Rotheram-Fuller2011; Whitehouse et al., Reference Whitehouse, Durkin, Jaquet and Ziatas2009). Furthermore, population-based studies indicate that elevated ASD traits are linked with difficulties in maintaining friendships and increased risks of bullying victimization (Hsiao, Tseng, Huang, & Gau, Reference Hsiao, Tseng, Huang and Gau2013; Nakajima et al., Reference Nakajima, Ito, Myogan, Takayanagi, Murayama, Hamada, Katori and Tsujii2021; Tanaka et al., Reference Tanaka, Ito, Takayanagi, Hrada, Noda, Otake, Nkajima, Mochizuki and Tsujii2014; Tanaka et al., Reference Tanaka, Ito, Murayama, Noda, Nakajima, Hamada, Katagiri, Takayanagi and Tsujii2015). Similarly, children with ADHD face heightened risks of peer rejection and bullying. Impulsivity and aggression contribute to exclusion, interpersonal conflict, and social isolation, with a substantial proportion of children with ADHD experiencing victimization regularly (Bagwell, Molina, Pelham, & Hoza, Reference Bagwell, Molina, Pelham and Hoza2001; Becker, Mehari, Langberg, & Evans, Reference Becker, Mehari, Langberg and Evans2017).
Relationships with teachers and parents
Children with ADHD are often reprimanded for impulsivity and noncompliance, which can strain parent–teacher relationships. These patterns are linked to increased parenting stress and harsher disciplinary practices, which may impair emotional regulation and increase internalization and externalization problems (Matsuoka et al., Reference Matsuoka, Okada, Tani, Onishi, Nakajima and Tsujii2011; Rogers & Tannock, Reference Rogers and Tannock2018). Conversely, children with ASD often show reduced attention to social cues and fewer attachment behaviors, making connections with supportive adults more difficult (Grzadzinski, Luyster, Spencer, & Lord, Reference Grzadzinski, Luyster, Spencer and Lord2014; Jones, Carr, & Klin, Reference Jones, Carr and Klin2008). Other research has highlighted that adult and peer support act as protective factors against depression, bullying, and school refusal (Jackson & Warren, Reference Jackson and Warren2000; Murayama et al., Reference Murayama, Ito, Otake, Katagiri, Hamada, Nakajima, Uemiya, Nomura, Takayanagi, Myogan and Tsujii2016; Williams, Connolly, Pepler, & Craig, Reference Williams, Connolly, Pepler and Craig2005).
Academic problems
Children with ADHD often struggle with core academic skills such as reading, writing, and mathematics (Barkley, Anastopoulos, Guevremont, & Fletcher, Reference Barkley, Anastopoulos, Guevremont and Fletcher1991), with higher trait severity linked with poorer performance (DuPaul & Stoner, Reference DuPaul and Stoner2003; Kawabata, Tseng, & Gau, Reference Kawabata, Tseng and Gau2012). Classroom studies show greater off-task behavior and shorter attention spans among children with ADHD compared to their peers (Kofler, Rapport, & Alderson, Reference Kofler, Rapport and Alderson2008; Rapport et al., Reference Rapport, Kofler, Alderson, Timko and Dupaul2009). Moreover, large-scale research has confirmed higher rates of absenteeism, dropout, and poor outcomes among children with ADHD (Fleming et al., Reference Fleming, Fitton, Steiner, McLay, Clark, King, Mackay and Pell2017). These findings highlight the broad academic impact of ADHD traits. Meanwhile, children with ASD may face difficulties in comprehension, limited interest in tasks, or avoidance of new content due to anxiety (Uno, Reference Uno2020). Although some studies have reported discrepancies between intellectual ability and actual achievement in individuals with ASD without intellectual disability (Estes et al., Reference Estes, Rivera, Bryan, Cali and Dawson2011), evidence on the link between ASD traits and academic problems remains limited.
Study aim
While many previous studies have examined the associations between neurodevelopmental disorders and psychosocial problems, few have explored the specific social domains that may help explain how ASD and ADHD traits are associated with mental health outcomes. In particular, few studies have simultaneously modeled ASD and ADHD traits while longitudinally examining multiple domains of social adjustment (peer relationships, academic performance, and relationships with teachers and parents). To address this gap, this study analyzes six-wave longitudinal data from children in grades 4–9 to examine whether ASD and ADHD traits show longitudinal associations with internalizing (depression and anxiety) and externalizing problems (anger and aggression), and how these associations relate to difficulties in peer, academic, and adult–child relationship domains. The study does not test causal mediation but instead investigates temporal associations among these constructs.
Hypotheses
Social difficulties have been widely discussed as potential pathways linking neurodevelopmental traits and mental health. Longitudinal studies indicate that depressive symptoms in children with elevated ASD traits are associated with negative peer experiences, including bullying (Rai et al., Reference Rai, Culpin, Heuvelman, Magnusson, Carpenter, Jones, Emond, Zammit, Golding and Pearson2018). Therefore, the following hypothesis was proposed:
Hypothesis 1: ASD traits are longitudinally associated with internalizing problems (depression and anxiety), and these associations are related to difficulties in peer relationships.
Academic challenges are particularly salient for ADHD. Children with elevated ADHD traits often experience persistent learning difficulties, which are associated with later emotional problems (Powell et al., Reference Powell, Riglin, Hammerton, Eyre, Martin, Anney, Thapar and Rice2020). Therefore, the following hypothesis was proposed:
Hypothesis 2: ADHD traits are longitudinally associated with internalizing problems (depression and anxiety), and these associations are related to academic difficulties.
ADHD traits have also been linked with externalizing problems in the context of interpersonal and environmental interactions. Reviews suggest that difficulties in emotion regulation contribute to aggression and oppositional behavior (Kraft, Ebner, Leo, & Lindenberg, Reference Kraft, Ebner, Leo and Lindenberg2023). Moreover, stressful or conflictual relationships with adults may exacerbate behavioral problems (Mak et al., Reference Mak, Yin, Li, Cheung and Oon2020). These findings highlight the importance of relational contexts. Thus, the following hypothesis was proposed:
Hypothesis 3: ADHD traits are longitudinally associated with externalizing problems (anger and aggression), and these associations are related to difficulties in relationships with parents and teachers.
Methods
Participants and procedures
This study was conducted as part of an ongoing longitudinal cohort project. It aimed to elucidate the mechanisms underlying development and adaptation in children and adolescents through a comprehensive population-based survey encompassing all public elementary and junior high schools in a medium-sized city in Chubu, Japan.
The surveyed municipality is located within commuting distance of a major metropolitan area while maintaining robust industrial and agricultural sectors. Consequently, it encompasses households with diverse socioeconomic backgrounds, including families with members employed in urban centers and those engaged in locally based occupations. With a population of approximately 90,000, the municipality is nearly equivalent to the national average for cities and wards (approximately 100,000), excluding government-designated cities and Tokyo’s 23 special wards.
Data were analyzed from longitudinal surveys conducted from FY2012 to FY2023. Information was collected from all enrolled students, their parents, and their homeroom teachers. To ensure the validity of the longitudinal analysis, students who participated in only one wave or none were excluded. The final sample included 9,619 students from an initial 11,549 enrolled students (participation rate: 83.3%). Details of cohort composition and grade progression are provided in Supplementary Table S1.
The data in this study comprise both complete longitudinal data from participants who continued across multiple waves and incomplete longitudinal data from those who participated in only some waves. Given the nature of school-based surveys, not all students could participate in every wave because of absences, school transfers, family circumstances, and other factors, leading to incomplete longitudinal data.
Comparisons of sex, student grade, and mean values of key variables showed no significant differences between the included and excluded students. All standardized mean differences were d < .20, suggesting that selection bias was limited. The average missing rate per wave was 11.6%, primarily due to absences and school transfers. This pattern is typical of school-based surveys, and the missingness was judged unlikely to violate the missing-at-random (MAR) assumption. To address the mixed presence of complete and incomplete longitudinal data, missing values were handled using full information maximum likelihood (FIML), using all available information for estimation.
Before data collection, informed consent was obtained from the students and parents via explanatory documents stating that participation was voluntary and that no disadvantages would result from nonparticipation. This study was reviewed and approved by our institution’s Ethics Committee.
Survey content
Developmental characteristics
ASD and ADHD tendencies were measured using parent-report measures: the shortened Japanese versions of the Autism Spectrum Screening Questionnaire (ASSQ; Ito, Matsumoto, et al., Reference Ito, Matsumoto, Takayanagi, Harada, Otake, Mochizuki, Nakajima, Noda, Tanaka and Tsujii2014a; Ito, Tanaka, et al., Reference Ito, Tanaka, Murayama, Nakajima, Takayanagi, Noda, Mochizuki, Matsumoto and Tsujii2014b) and the ADHD Rating Scale-IV (DuPaul, Power, Anastopoulos, & Reid, Reference DuPaul, Power, Anastopoulos and Reid1998; Ichikawa & Tanaka, Reference Ichikawa and Tanaka2008; Tanaka et al., Reference Tanaka, Ito, Murayama, Noda, Nakajima, Hamada, Katagiri, Takayanagi and Tsujii2015). The ASSQ primarily captures Asperger-like social and behavioral characteristics, which represent a narrower subset of the broader ASD phenotype described in DSM-5 and ICD-11. Both scales are suitable for assessing dimensional variation in neurodevelopmental traits within nonclinical school populations.
Social maladjustment
Social maladjustment was measured using the Social Maladjustment Scale for Elementary and Junior High School Students (Ito, Matsumoto, et al., Reference Ito, Matsumoto, Takayanagi, Harada, Otake, Mochizuki, Nakajima, Noda, Tanaka and Tsujii2014a; Ito, Tanaka, et al., Reference Ito, Tanaka, Murayama, Nakajima, Takayanagi, Noda, Mochizuki, Matsumoto and Tsujii2014b), a self-report questionnaire that captures children’s subjective sense of maladjustment. The scale evaluates difficulties in three domains: academic problems (e.g. ‘Has difficulty keeping up with classwork’), problems with teachers (e.g. ‘Often conflicts with teachers’), and family problems (e.g. ‘Frequent arguments with parents’).
Peer problems were assessed using the five-item ‘peer relationships’ subscale from the Japanese version of the Strengths and Difficulties Questionnaire (SDQ; Harada et al., Reference Harada, Ito, Mochizuki, Tanaka, Otake, Takayanagi and Tsujii2014). The ratings obtained from students, teachers, and parents were z-standardized and then combined into a composite variable (α = .898). The multi-informant SDQ approach reduces rater bias and captures peer difficulties across home and school contexts. The SDQ Peer Problems subscale is internationally validated and complements aspects of social functioning not fully represented in the Social Maladjustment Scale.
Mental health
Depression was measured using the shortened Japanese Birleson Depression Self-Rating Scale for Children (DSRS-C; Murata, Shimizu, Mori, & Oshima, Reference Murata, Shimizu, Mori and Oshima1996; Namikawa et al., Reference Namikawa, Tani, Wakita, Kumagai, Nakane, Noguchi and Tsujii2011). Aggression was assessed using the shortened Hostility-Aggression Questionnaire for Children (HAQ-C; Sakai et al., Reference Sakai, Yamasaki, Soga, Oashi, Shimai and Otake2000). Teacher and parent SDQ ratings of emotional symptoms and conduct problems were integrated with DSRS-C and HAQ-C to form composite variables for internalizing (α = .861) and externalizing problems (α = .895).
Statistical analysis
We used structural equation modeling (SEM) to estimate cross-lagged panel models across six waves. A conceptual diagram of the longitudinal associations examined in this study is presented in Supplementary Figure S1.
Analyses were conducted using JASP (Version 0.18.3; JASP Team, 2024). Parameters were estimated with robust maximum likelihood (MLR), which provides standard errors and test statistics that are robust to non-normality. Missing data were handled using FIML, incorporating all available data under the MAR assumption.
The cross-lagged panel model included autoregressive paths for each construct across adjacent waves; cross-lagged paths among ASD and ADHD traits; social maladjustment variables (peer, academic, teacher, and family problems); internalizing and externalizing problems; and within-wave correlations among all variables. All primary variables were z-standardized to facilitate comparison of standardized coefficients. Because the model was saturated (df = 0), fit indices were not informative and therefore were not reported. To reduce potential confounding, sex and school grade were included as time-invariant covariates in all models.
Results
Descriptive statistics and correlation coefficients
Summary statistics, distributional characteristics, reliability coefficients, and correlations for all major variables are provided in Supplementary Tables S2–S4 and Supplementary Figure S2. Overall, the variables showed right-skewed distributions with most participants scoring in the lower range.
Tables 1–3 present descriptive statistics and correlation coefficients for each fourth- and ninth-grade variable. ASD traits showed a slight decrease from fourth to ninth grade, whereas ADHD traits demonstrated a more pronounced decline. In contrast, academic problems and depressive symptoms increased across development. ASD and ADHD traits were moderately correlated at both time points, and ASD traits showed strong associations with peer problems. Peer problems were also strongly associated with internalizing symptoms.
Descriptive statistics for each variable
Table 1. Long description
The table presents Mean (M) and Standard Deviation (SD) for various psychological and behavioral traits across six grade levels (4th, 5th, 6th, 7th, 8th, and 9th).
* A S D traits (Parent informant): Means decrease slightly from 13.14 in 4th grade to 12.90 in 9th grade.
* A D H D traits (Parent informant): Means decrease from 2.93 in 4th grade to 1.69 in 9th grade.
* Social maladaptation categories:
- Peer problems (Child): Decreases from 2.68 to 2.02.
- Peer problems (Teacher): Fluctuates between 6.05 and 6.22.
- Peer problems (Parent): Remains stable around 1.60.
- Teacher problems (Child): Decreases from 0.98 to 0.77.
- Family problems (Child): Decreases from 2.03 to 1.59.
- Academic problems (Child): Increases from 3.40 in 4th grade to 4.45 in 9th grade.
* Internalizing problems:
- Depression (Child): Increases from 4.22 in 4th grade to 5.19 in 9th grade.
- Emotional symptoms (Teacher): Remains stable between 5.60 and 5.84.
- Emotional symptoms (Parent): Decreases from 1.66 to 1.41.
* Externalizing problems:
- Aggression (Child): Decreases from 16.00 in 4th grade to 14.70 in 9th grade.
- Conduct problems (Teacher): Decreases from 6.05 to 5.80.
- Conduct problems (Parent): Decreases from 1.76 to 1.43.
ADHD, attention-deficit/hyperactivity disorder; ASD, autism spectrum disorder; SDQ, Strengths and Difficulties Questionnaire.
Correlation coefficients among variables at fourth grade
Table 2. Long description
The table presents correlation coefficients for eight variables. The variables are: 1. A S D traits, 2. A D H D traits, 3. Peer problems, 4. Teacher problems, 5. Family problems, 6. Academic problems, 7. Internalizing problems, and 8. Externalizing problems. All reported correlations are significant at the p < .001 level, indicated by three asterisks.
* Row 2 (A D H D traits): Correlates with variable 1 at .559.
* Row 3 (Peer problems): Correlates with variable 1 at .550 and variable 2 at .348.
* Row 4 (Teacher problems): Correlates with variable 1 at .159, variable 2 at .229, and variable 3 at .243.
* Row 5 (Family problems): Correlates with variable 1 at .166, variable 2 at .195, variable 3 at .240, and variable 4 at .430.
* Row 6 (Academic problems): Correlates with variable 1 at .128, variable 2 at .222, variable 3 at .227, variable 4 at .364, and variable 5 at .389.
* Row 7 (Internalizing problems): Correlates with variable 1 at .399, variable 2 at .286, variable 3 at .541, variable 4 at .201, variable 5 at .271, and variable 6 at .269.
* Row 8 (Externalizing problems): Correlates with variable 1 at .329, variable 2 at .452, variable 3 at .388, variable 4 at .395, variable 5 at .341, variable 6 at .317, and variable 7 at .315.
Note: *p < .05 **p < .01 ***p < .001.
Correlation coefficients among variables at ninth grade
Table 3. Long description
A lower triangular correlation matrix with 8 numbered columns and 8 corresponding rows. All reported correlations are significant at the p < .001 level, indicated by three asterisks.
* Row 1: A S D traits. Column 1 is a dash.
* Row 2: A D H D traits. Correlation with Column 1 (A S D traits) is .518.
* Row 3: Peer problems. Correlation with Column 1 is .573; Column 2 is .310.
* Row 4: Teacher problems. Correlation with Column 1 is .094; Column 2 is .146; Column 3 is .173.
* Row 5: Family problems. Correlation with Column 1 is .155; Column 2 is .166; Column 3 is .223; Column 4 is .387.
* Row 6: Academic problems. Correlation with Column 1 is .076; Column 2 is .167; Column 3 is .110; Column 4 is .258; Column 5 is .220.
* Row 7: Internalizing problems. Correlation with Column 1 is .375; Column 2 is .270; Column 3 is .597; Column 4 is .199; Column 5 is .307; Column 6 is .211.
* Row 8: Externalizing problems. Correlation with Column 1 is .270; Column 2 is .380; Column 3 is .370; Column 4 is .304; Column 5 is .318; Column 6 is .227; Column 7 is .362.
Note: *p < .05, **p < .01, ***p < .001.
Note: *p < .05 **p < .01 ***p < .001.
Cross-lagged effects
Cross-lagged structural equation models were used to examine longitudinal associations among variables. Regression coefficients were interpreted using Orth et al.’s (Reference Orth, Meier, Bühler, Dapp, Krauss, Messerli and Robins2024) guidelines (~.03 = small, .07 = medium, .12 = large). All effects reported below reflect temporal associations rather than causal influence.
The major cross-lagged effects are shown in Figure 1, and the full set of cross-lagged coefficients is summarized in Table 4.
Cross-lagged path models for ASD traits and ADHD traits.
Figure 1. Long description
The figure consists of two panels.
Left panel: A vertical path model with three boxes. At the top is A S D Traits, in the middle is Peer Problems, and at the bottom is Internalizing Problems. Bidirectional vertical arrows connect A S D Traits and Peer Problems with beta values of .096 to .164 upward and .146 to .190 downward. Bidirectional vertical arrows connect Peer Problems and Internalizing Problems with beta values of .043 to .107 upward and .046 to .121 downward. A long curved arrow points from A S D Traits directly to Internalizing Problems with a beta of .068 to .097.
Right panel: A vertical path model with six boxes. From top to bottom, they are A D H D Traits, Externalizing Problems, Teacher Problems, Academic Problems, Family Problems, and Internalizing Problems.
- Between A D H D Traits and Externalizing Problems: beta .032 to .082 upward and .064 to .118 downward.
- Between Externalizing Problems and Teacher Problems: beta .011 to .047 upward and .103 to .262 downward.
- Between Teacher Problems and Academic Problems: beta .082 to .128 upward and .019 to .047 downward.
- Between Academic Problems and Family Problems: beta .090 to .161 upward and .105 to .163 downward.
- Between Family Problems and Internalizing Problems: beta .032 to .092 upward and .025 to .092 downward.
- Long-range arrows: A D H D Traits to Academic Problems beta .015 to .069. Externalizing Problems to Internalizing Problems beta .042 to .126. Family Problems to Externalizing Problems beta .029 to .093.
Cross-lagged effects of each variable
Table 4. Long description
A complex statistical table detailing the cross-lagged effects of eight variables: Internalizing problems, Externalizing problems, Peer problems, Teacher problems, Family problems, Academic problems, A S D traits, and A D H D traits. The data is organized into six time-lagged columns: 4th_5th, 5th_6th, 6th_7th, 7th_8th, and 8th_9th. Each column provides the beta coefficient and standard error S E for the relationship between independent and dependent variables.
Key findings include:
* Internalizing problems show high stability with beta values ranging from .490 to .545 across all time points.
* A S D traits consistently predict Internalizing problems with beta values between .068 and .097, all significant at p < .001.
* Externalizing problems show high stability with beta values ranging from .545 to .611.
* A D H D traits are a strong predictor of Externalizing problems, with beta values between .064 and .118, all significant at p < .001.
* Peer problems are significantly predicted by A S D traits across all intervals, with beta values ranging from .146 to .190.
* Academic problems are consistently predicted by Family problems, with beta values increasing from .090 in 4th_5th to .161 in 8th_9th.
* A S D and A D H D traits themselves show high longitudinal stability, with beta coefficients consistently above .600.
Significance levels are indicated by asterisks: * p < .05, ** p < .01, and *** p < .001.
Note: *p < .05 **p < .01 ***p < .001.
Internalizing problems
ASD traits showed small-to-medium longitudinal associations with later internalizing problems (β = .068–.097). Associations were relatively stronger from fourth to fifth grade (β = .093) and seventh to eighth grade (β = .097). Peer problems also showed small-to-large associations with later internalizing problems (β = .046–.121), with the largest association observed from fifth to sixth grade (β = .121). Academic problems during elementary school showed small associations with later internalizing problems (β = .055–.090). These findings indicate that ASD traits, peer difficulties, and academic challenges each show temporal associations with internalizing symptoms across development.
Externalizing problems
ADHD traits showed small-to-medium longitudinal associations with later externalizing problems (β = .064–.118). Associations were strongest from fifth to sixth grade (β = .118) and fourth to fifth grade (β = .107). Family problems also showed small-to-medium associations with later externalizing problems (β = .029–.093). These patterns suggest that ADHD traits and family-related difficulties are consistently associated with externalizing behaviors over time.
Peer problems
ASD traits showed large longitudinal associations with later peer problems (β = .146–.190). Associations were consistently strong across all grade transitions. Internalizing problems also showed small-to-medium associations with later peer problems (β = .043–.107). These results indicate stable bidirectional associations between ASD traits, emotional difficulties, and peer relationship challenges.
Teacher problems
Externalizing problems showed medium-to-large associations with later teacher relationship problems (β = .103–.262), with the strongest association from seventh to eighth grade (β = .262). Academic problems also showed medium-to-large associations with later teacher problems (β = .082–.128). These findings suggest that behavioral and academic difficulties are consistently associated with strained teacher relationships.
Family problems
Academic problems showed medium-to-large associations with later family problems (β = .105–.163). Externalizing problems also showed small-to-medium associations with later family problems (β = .042–.126). These results indicate that academic and behavioral difficulties are associated with increased family-related challenges over time.
Academic problems
ADHD traits showed small-to-medium associations with later academic problems (β = .015–.069). Family problems also showed medium-to-large associations with later academic problems (β = .090–.161). These findings highlight consistent associations between ADHD traits, family context, and academic functioning.
ASD traits
Peer problems showed medium-to-large associations with later ASD traits (β = .096–.164), with the strongest associations from sixth to seventh grade (β = .164) and eighth to ninth grade (β = .145). These results suggest reciprocal associations between ASD traits and peer difficulties across development.
ADHD traits
Externalizing problems showed small-to-medium associations with later ADHD traits (β = .032–.082). Associations were strongest from seventh to eighth grade (β = .082). These findings indicate reciprocal associations between ADHD traits and behavioral difficulties.
Discussion
This study examined longitudinal associations among ASD traits, ADHD traits, social adjustment, and mental health outcomes in a large cohort of Japanese children and adolescents. The findings highlight distinct patterns of association for ASD and ADHD traits across multiple social domains and underscore the importance of considering dynamic, bidirectional processes rather than unidirectional causal pathways.
Association between ASD traits and internalizing problems
The results showed that ASD traits were associated with peer relationship difficulties over time. Moreover, peer relationship difficulties were subsequently associated with increases in internalizing symptoms, supporting Hypothesis 1. However, these associations do not establish mediation or causal direction. This pattern aligns with prior research reporting that peer problems are related to emotional difficulties among youth with elevated ASD traits (Rai et al., Reference Rai, Culpin, Heuvelman, Magnusson, Carpenter, Jones, Emond, Zammit, Golding and Pearson2018).
These associations were particularly pronounced during the upper elementary years, a developmental period characterized by increasing importance of peer group belonging and social competence. Difficulties in social communication and peer interactions associated with ASD traits may place children at heightened risk for emotional distress, especially as peer relationships become more central to daily life. These findings emphasize the importance of school-based support systems that strengthen peer relationships, such as structured social skills programs and peer-mediated interventions, which may help reduce emotional distress during this sensitive developmental stage.
Association between ADHD traits and internalization
Although ADHD traits were significantly associated with internalizing problems at several grade intervals, the magnitude of these associations was generally small, suggesting limited direct links. Instead, ADHD traits were more consistently associated with academic difficulties, which themselves were associated with later internalizing symptoms. This pattern is consistent with prior work showing that inattention and impulsivity are linked with academic challenges (Barkley et al., Reference Barkley, Anastopoulos, Guevremont and Fletcher1991; DuPaul & Stoner, Reference DuPaul and Stoner2003; Kawabata et al., Reference Kawabata, Tseng and Gau2012).
The results provide partial support for Hypothesis 2: academic difficulties may represent one pathway through which ADHD traits relate to internalizing symptoms. However, these associations were not consistent across all developmental periods and do not indicate formal mediation. These findings suggest that academic support – such as individualized instruction, organizational skills training, and self-efficacy enhancement – may help reduce emotional distress among children with elevated ADHD traits.
Association between ADHD traits and externalizing problems
ADHD traits showed strong and consistent longitudinal associations with externalizing problems, particularly during elementary school. This pattern is consistent with the core behavioral characteristics of ADHD, including impulsivity and difficulties with behavioral regulation, which are often linked with oppositional and aggressive behaviors.
Difficulties in relationships with teachers and parents were also associated with externalizing problems, although these associations were weaker than those involving ADHD traits, providing only partial support for Hypothesis 3. These findings highlight the importance of early behavioral interventions targeting self-regulation, combined with collaborative support between families and schools, to help reduce the escalation of behavioral difficulties.
Bidirectional associations between social maladjustment and mental health
A key contribution of this study is the identification of bidirectional associations between social adjustment and mental health. Peer difficulties and internalizing problems showed mutual associations, as did teacher–student and parent–child relationship problems with externalizing behaviors. These findings suggest that social difficulties and mental health symptoms may reinforce one another over time, forming dynamic cycles rather than simple one-way pathways.
For example, peer isolation may be associated with increases in depressive symptoms, while emotional distress may, in turn, be associated with further withdrawal or victimization. Similarly, strained relationships with teachers or parents may be associated with increases in oppositional behavior, which may contribute to additional relational conflict. These patterns underscore the importance of school-based mental health services, including counseling and early identification systems, to interrupt these cycles and provide continuous support throughout adolescence.
Interaction between developmental traits and social environment
The findings also indicate that ASD and ADHD traits are not static characteristics; rather, their subsequent levels may be associated with ongoing interactions with social environments. Peer difficulties were associated with later increases in ASD traits, and externalizing problems were associated with later increases in ADHD traits, suggesting reciprocal associations between traits and social experiences. These results align with the International Classification of Functioning, Disability and Health (ICF) framework, which conceptualizes functioning as an interaction between individual characteristics and environmental contexts.
This perspective is also consistent with Bronfenbrenner’s (Reference Bronfenbrenner1979) ecological systems theory, in which peer relationships, teacher–student interactions, and academic contexts represent key microsystems that shape developmental outcomes. Understanding ASD and ADHD traits within these dynamic systems highlights the importance of interventions that address both individual needs and environmental supports, including classroom accommodations, teacher training, and family–school collaboration.
Limitations
Several limitations should be acknowledged. First, clinical diagnostic information, including ASD, ADHD, and internalizing/externalizing disorders, was unavailable, and the study relied on dimensional trait measures. Given the size of the sample, it is likely that some children with clinical diagnoses were included, which may have influenced the distribution or variability of the measured constructs. However, because the analyses focused on dimensional traits rather than diagnostic groups, the inclusion of clinically diagnosed cases does not undermine the validity of the dimensional approach. Second, although six waves of data were analyzed, the follow-up intervals varied and the design did not constitute a fully closed cohort, which limits causal interpretation despite the prospective ordering of variables.
Third, the high correlations observed among some constructs, particularly between ASD traits and internalizing problems, may reflect item overlap, shared method variance, or rater biases. Future studies should incorporate multiple measurement approaches, including observational methods and performance-based assessments, to reduce shared-method dependencies. Fourth, the distribution of ASD traits deviated from normality, showing the positively skewed pattern typically observed in general-population samples. This non-normality may have influenced parameter estimates, including regression coefficients and variance estimation.
Fifth, formal tests of longitudinal measurement invariance across waves could not be conducted because identical item-level data were not available across all waves. This raises the possibility that changes in psychometric properties may have affected longitudinal estimates. Sixth, some variables were assessed through self-report, which may be influenced by subjective perceptions and response styles. In addition, the composite variables used in the analyses required integrating information across informants; although the construction procedures were described in detail, the weighting scheme may still introduce measurement heterogeneity.
Seventh, although this study examined prospective associations, it did not formally test mediation or indirect effects. The social adaptation indicators included in the model were treated as correlates rather than mediators, and the study did not evaluate whether these variables moderated the associations between developmental traits and mental health outcomes. Thus, the possibility that social contexts may buffer or amplify these associations remains an important direction for future research.
Eighth, assumptions related to missing data were based on the MAR framework, but MAR could not be empirically verified. Similarly, although standardized mean differences for attrition analyses were generally below 0.20, indicating limited selection bias, these thresholds should be interpreted cautiously. Finally, the sample was drawn from mainstream classrooms in a single region of Japan, which may limit the generalizability of the findings to other cultural or educational contexts.
Conclusions
This study examined longitudinal associations among ASD traits, ADHD traits, social adjustment, and mental health outcomes using large-scale cohort data. The findings showed that social environmental maladjustment was not a simple mediating mechanism but was a part of a broader set of temporal associations linking developmental traits and mental health. In particular, bidirectional associations between social adjustment and mental health suggest that these processes unfold dynamically over time rather than through a single, unidirectional pathway.
Furthermore, the results indicate that social contexts, such as peer relationships, academic environments, and interactions with teachers and parents, play a meaningful role in shaping developmental trajectories. These contexts were associated with both later mental health outcomes and subsequent levels of ASD and ADHD traits, underscoring the importance of understanding developmental characteristics within their environmental settings.
Overall, these findings highlight the need for comprehensive, context-sensitive support strategies that address both individual characteristics and the social environments in which children develop. Interventions that strengthen peer relationships, enhance academic support, and improve family–school collaboration may help promote healthier developmental pathways. By emphasizing dynamic interactions between traits and environments, this study provides a conceptual foundation for developing integrated support systems that respond to both social and individual needs.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S003329172610539X.
Data availability statement
The data that support the findings of this study are available from the corresponding author, EH, upon reasonable request.
Acknowledgements
We extend our deepest gratitude to the research area’s board of education and the faculty and staff of the participating public elementary and middle schools for their invaluable cooperation. We are equally grateful to the students and parents whose continued participation made this longitudinal study possible. Finally, we sincerely thank the research collaborators who supported data collection, organization, and analysis, along with the experts who provided thoughtful peer review and guidance in the preparation of this manuscript.
Funding statement
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Competing interests
The authors declare none.
Ethical standards
The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. This study was reviewed and approved by the Ethics Committee of Chukyo University (approved March 2007, reference number 18–166).
Informed consent
Informed consent was obtained from the parents of all participants before their inclusion in the study.
Open access
