Perinatal and neonatal factors and mental disorders in children and adolescents: looking for the contributions of the early environment to common and dissociable aspects of psychopathology.

High rates of co-occurrence of mental disorders have been hypothesized to represent a result of common susceptibility to overall psychopathology. The purpose of this study is to test the hypothesis that commonalities among psychiatric disorders might be partially driven by sharable perinatal and neonatal environmental factors for mental disorders. Participants were 6-14 years of age children and their parents. Primary caregivers provided data on perinatal and neonatal information assessed retrospectively (n = 2231). Psychiatric disorders diagnoses were assessed using the Development and Well Being Behavior Assessment (DAWBA). We used bifactor models to disentangle common from dissociable aspects of psychopathology. These models allow modeling psychiatric disorders as the result of a common domain of psychopathology (p-factor) and three dissociable domains (fear, distress, and externalizing symptoms). Associations were tested using linear and tobit regression models. The p-factor was associated with male sex, low socioeconomic status, gestational smoking, gestational drinking, low levels of maternal education and presence of mental disorder in the mother. Associations with specific factors also emerged suggesting some risk factors might also have some role for fear, distress and externalizing factors. Our study supports the hypothesis that overall susceptibility to psychopathology might be partially driven by sharable perinatal and neonatal factors.

The synergistic effect of genetic and environmental factors in the development of attention-deficit/hyperactivity disorder symptoms in children and adolescents.

Gene-environment interactions (GxE) have been increasingly explored in psychiatry but with low replication rates. Attention-deficit/hyperactivity disorder (ADHD) is a suitable candidate for studying GxE due to its high heritability and well-defined environmental risk factors. Here, we explored GxE using polygenic risk score (PRS) to represent the genetic liability to ADHD (ADHD-PRS) and environmental risk score (ERS) to represent the combined effects of environmental risk factors. We analyzed longitudinal data of 2,046 individuals (6-14 years of age at baseline and 14-23 at the last follow-up) from the Brazilian High-Risk Cohort Study for Psychiatric Disorders. Psychiatric evaluation included the Child Behavior Checklist and the Strength and Difficulties Questionnaire. Statistical analyses were performed using mixed-effects models. We observed statistically significant interactions between ADHD-PRS and ERS, suggesting that environmental and genetic factors act synergistically in the development of ADHD symptoms. These effects were not present for depression or anxiety symptoms. No evidence of GxE correlation was detected. Mechanistically, our findings suggest that environmental stressors modulate the genetic risk for ADHD. Future studies should investigate whether the reduction of environmental risks can prevent the development of symptoms of ADHD, especially in children with a family history of the disorder.

The association between duration of breastfeeding and the trajectory of brain development from childhood to young adulthood: an 8-year longitudinal study.

Breastfeeding has been associated with several short- and long-term health benefits, including positive cognitive and behavioral outcomes. However, the impact of breastfeeding on structural brain development over time remains unclear. We aimed to assess the association between breastfeeding duration in childhood and the developmental trajectory of overall cortical thickness, cortical area, and total intracranial volume during the transition from childhood to early adulthood. Participants included 670 children and adolescents with 1326 MRI scans acquired over 8 years from the Brazilian High-Risk Cohort for Mental Conditions (BHRCS). Breastfeeding was assessed using a questionnaire answered by the parents. Brain measures were estimated using MRI T1-weighted images at three time points, with 3-year intervals. Data were evaluated using generalized additive models adjusted for multiple confounders. We found that a longer breastfeeding duration was directly associated with higher global cortical thickness in the left (edf = 1.0, F = 6.07, p = 0.01) and right (edf = 1.0, F = 4.70, p = 0.03) hemispheres. For the total intracranial volume, we found an interaction between duration of breastfeeding and developmental stage (edf = 1.0, F = 6.81, p = 0.009). No association was found between breastfeeding duration and brain area. Our study suggests that the duration of breastfeeding impacts overall cortical thickness and the development of total brain volume, but not area. This study adds to the evidence on the potential impact of breastfeeding on brain development and provides relevant insights into the mechanisms by which breastfeeding might confer cognitive and mental health benefits.

Childhood poverty and mental health disorders in early adulthood: evidence from a Brazilian cohort study.

BACKGROUND: We examined the association between childhood poverty and mental health disorders (MHD) in childhood and early adulthood. We also investigated whether the association between poverty in childhood and MHD is mediated by exposure to stressful life events (SLE). METHODS: We used data from a prospective community cohort of young people assessed at baseline (M = 9.7 years, SD = 1.9), first (M = 13.5 years, SD = 1.9), and second (M = 18.2 years, SD = 2.0) follow-ups (N = 1,590) in Brazil. Poverty was assessed using a standardized classification. Exposure to 20 different SLE was measured using the Life History instrument. Psychiatric diagnoses were evaluated using the Development and Well-Being Assessment. Latent growth models investigated the association between poverty at baseline and the growth of any MHD, externalizing, and internalizing disorders. Mediation models evaluated whether the association between childhood poverty and MHD in early adulthood was mediated by exposure to SLE. RESULTS: Poverty affected 11.4% of the sample at baseline and was associated with an increased propensity for presenting externalizing disorders in adolescence or early adulthood (standardized estimate = 0.27, p = 0.016). This association was not significant for any disorder or internalizing disorders. Childhood poverty increased the likelihood of externalizing disorders in early adulthood through higher exposure to SLE (OR = 1.07, 95 CI% 1.01-1.14). Results were only replicated among females in stratified analyses. CONCLUSIONS: Childhood poverty had detrimental consequences on externalizing MHD in adolescence, especially among females. Poverty and SLE are preventable risk factors that need to be tackled to reduce the burden of externalizing disorders in young people.

Long-term stability of the cortical volumetric profile and the functional human connectome throughout childhood and adolescence.

There is compelling evidence showing that between-subject variability in several functional and structural brain features is sufficient for unique identification in adults. However, individuation of brain functional connectomes depends on the stabilization of neurodevelopmental processes during childhood and adolescence. Here, we aimed to (1) evaluate the intra-subject functional connectome stability over time for the whole brain and for large scale functional networks and (2) determine the long-term identification accuracy or 'fingerprinting' for the cortical volumetric profile and the functional connectome. For these purposes, we analysed a longitudinal cohort of 239 children and adolescents scanned in two sessions with an interval of approximately 3 years (age range 6-15 years at baseline and 9-18 years at follow-up). Corroborating previous results using short between-scan intervals in children and adolescents, we observed a moderate identification accuracy (38%) for the whole functional profile. In contrast, identification accuracy using cortical volumetric profile was 95%. Among the large-scale networks, the default-mode (26.8%), the frontoparietal (23.4%) and the dorsal-attention (27.6%) networks were the most discriminative. Our results provide further evidence for a protracted development of specific individual structural and functional connectivity profiles.

Neuroimaging Association Scores: reliability and validity of aggregate measures of brain structural features linked to mental disorders in youth.

In genetics, aggregation of many loci with small effect sizes into a single score improved prediction. Nevertheless, studies applying easily replicable weighted scores to neuroimaging data are lacking. Our aim was to assess the reliability and validity of the Neuroimaging Association Score (NAS), which combines information from structural brain features previously linked to mental disorders. Participants were 726 youth (aged 6-14) from two cities in Brazil who underwent MRI and psychopathology assessment at baseline and 387 at 3-year follow-up. Results were replicated in two samples: IMAGEN (n = 1627) and the Healthy Brain Network (n = 843). NAS were derived by summing the product of each standardized brain feature by the effect size of the association of that brain feature with seven psychiatric disorders documented by previous meta-analyses. NAS were calculated for surface area, cortical thickness and subcortical volumes using T1-weighted scans. NAS reliability, temporal stability and psychopathology and cognition prediction were analyzed. NAS for surface area showed high internal consistency and 3-year stability and predicted general psychopathology and cognition with higher replicability than specific symptomatic domains for all samples. They also predicted general psychopathology with higher replicability than single structures alone, accounting for 1-3% of the variance, but without directionality. The NAS for cortical thickness and subcortical volumes showed lower internal consistency and less replicable associations with behavioural phenotypes. These findings indicate the NAS based on surface area might be replicable markers of general psychopathology, but these links are unlikely to be causal or clinically useful yet.

Relationships between childhood maltreatment, impairment in executive functions and disruptive behavior disorders in a community sample of children.

Evidence points to an independent relationship among childhood maltreatment, impairments in executive functions (EF) and disruptive behavior disorders (DBD). However, it is still not fully understood how these three factors are interrelated. This study evaluated the association between childhood maltreatment and DBD testing the role of EF performance as a mediator or moderator. We studied a probabilistic school-based sample of 2016 children from 6 to 12 years. Mental disorders were assessed using the Development and Well-Being Assessment with parents and children. Children answered questions about exposure to child maltreatment and were evaluated with a set of cognitive tasks addressing inhibitory control, working memory, cognitive flexibility and planning. Childhood maltreatment was strongly associated with DBD (OR = 7.7, CI 95% 4.5-12.9). No association was found between childhood maltreatment and EF performance. Children with DBD showed worse performance in cognitive flexibility, which was not identified as a mediator or moderator of the association between childhood maltreatment and DBD. Results indicate that the association between maltreatment and disruptive behavior occurs regardless of performance in executive function in a community sample. Future studies are essential to confirm these findings and elucidate the cognitive mechanisms involved in this association.

Heterotypic trajectories of dimensional psychopathology across the lifespan: the case of youth-onset attention deficit/hyperactivity disorder.

BACKGROUND: Recent studies have demonstrated the existence of a distinct late-onset attention deficit/hyperactivity disorder (ADHD) trajectory. Our objective is to test if there are distinct ADHD trajectories regarding age of onset from childhood to adolescence and to compare clinical manifestations, cognitive functions and genetic risk for ADHD among distinct longitudinal groups. METHOD: Nine hundred and twenty four children and adolescents from the community participated in the study. We compared clinical, cognitive features and genetic risk among four groups of participants: (a) childhood-limited, (b) youth-onset, (c) childhood-onset with youth persistence, and (d) community comparisons without ADHD. Symptomatic and diagnostic assessments were performed using the Development and Well-Being Behavior Assessment, the Strengths and Difficulties Questionnaire, and the Child Behavior Checklist. Cognitive functions were measured using a battery of standardized tests. Genetic risk for ADHD was calculating using summary statistics from the Psychiatric Genomics Consortium. RESULTS: Half of the adolescents (52%) with ADHD had their symptom onset in adolescence. The impairment level of this group in adolescence is similar to the persistent group. Despite not having ADHD, the youth-onset group already presented in childhood more symptoms from other domains of psychopathology, higher shared variance in psychiatric symptomatology (p-factor), school impairment, and executive dysfunctions than community comparisons. Furthermore, the youth-onset group presented lower levels of genetic risk for ADHD compared to other cases. CONCLUSIONS: A significant proportion of adolescents with ADHD were youth-onset cases and presented similar impairment levels as those cases with early-onset ADHD. The presence of cognitive impairments and higher levels of clinical symptoms in the youth-onset group already at childhood speaks in favor of a heterotypic trajectory of psychopathology suggesting that youth-onset ADHD might be an artificial consequence of categorizing dimensional psychopathology into discrete diagnostic groups.

Associations between children's family environment, spontaneous brain oscillations, and emotional and behavioral problems.

The family environment in childhood has a strong effect on mental health outcomes throughout life. This effect is thought to depend at least in part on modifications of neurodevelopment trajectories. In this exploratory study, we sought to investigate whether a feasible resting-state fMRI metric of local spontaneous oscillatory neural activity, the fractional amplitude of low-frequency fluctuations (fALFF), is associated with the levels of children's family coherence and conflict. Moreover, we sought to further explore whether spontaneous activity in the brain areas influenced by family environment would also be associated with a mental health outcome, namely the incidence of behavioral and emotional problems. Resting-state fMRI data from 655 children and adolescents (6-15 years old) were examined. The quality of the family environment was found to be positively correlated with fALFF in the left temporal pole and negatively correlated with fALFF in the right orbitofrontal cortex. Remarkably, increased fALFF in the temporal pole was associated with a lower incidence of behavioral and emotional problems, whereas increased fALFF in the orbitofrontal cortex was correlated with a higher incidence.

Cross-Sectional and Longitudinal Associations of Temperament and Mental Disorders in Youth.

Here we evaluate the cross-sectional and longitudinal associations between temperament and mental disorders in adolescents. Temperament was assessed in a cohort of 1540 youths by the revised self-report Early Adolescence Temperament Questionnaire (EATQ-R) at baseline and confirmatory factor analyses were used to test the best empirical model. Mental disorders were assessed by parental interview using the Development and Well-Being Behavior Assessment at baseline and at 3-year follow-up. Participants were grouped into Typically Developing Comparisons, Phobias, Distress, Attention-Deficit/Hyperactivity Disorder (ADHD) and Disruptive Behavior Disorders (DBD). Logistic regression models tested the effects of temperament on incidence and remission of mental disorders. The bifactor model of EATQ-R presented the best fit. Distress, ADHD and DBD have lower levels of effortful control in baseline. Adjusted longitudinal analysis showed that effortful control predicted lower incidence of Phobias (OR 0.74; p = 0.018), distress (OR 0.74; p = 0.014) and DBD (OR 0.68; p = 0.037). Temperament factors did not predicted remission rates.

Debate: Dimensions of mania in youth: possibly bipolar, probably risk indicators, certainly impairing.

We comment on the utility of latent manic dimensions to elucidate the prevalence of youth Bipolar Disorder in epidemiological samples, irrespective of sample's cultural background or clinician's predetermined prototypical definitions of a bipolar case.

The association between psychotic experiences and traumatic life events: the role of the intention to harm.

BACKGROUND: Previous work showed traumatic life events (TLE) with intention to harm, like bullying and abuse, to be more strongly associated with psychotic experiences (PE) than other types of trauma, like accidents. However, this association is subject to reporting bias and can be confounded by demographic characteristics and by differences in dose of exposure across different trauma categories. We studied the association between TLE with and without intention to harm and PE, taking into account potential confounders and biases. METHODS: A total of 2245 children and adolescents aged 6-14 years were interviewed by psychologists. The interview included the presence of 20 PE (both self-report and psychologist evaluation). In addition, parents provided information on child exposure to trauma, mental health and PE. RESULTS: Results showed no significant association between TLE without intention to harm only and PE for the three methods of assessment of PE (self-report, parent report and psychologist rating). On the other hand, there was a positive association between PE and TLE in groups exposed to traumatic experiences with intention to harm (with intention to harm only and with and without intention to harm). Results remained significant after controlling for demographic and clinical confounders, but this positive association was no longer significant after adjusting for the number of TLE. CONCLUSIONS: TLE with intention to harm display a stronger association with PE than TLE without intention to harm, and this difference is likely reducible to a greater level of traumatic exposure associated with TLE with intention to harm.

Fine motor ability and psychiatric disorders in youth.

Impaired fine motor ability has been linked to several domains of psychopathology. However, studies validating the specificity of this association among several categorical psychiatric disorders are still needed. The aim of this study was to assess differences in fine motor ability performance among four non-overlapping groups of psychiatric disorders and a group of typical developing comparisons (TDC). Our sample consisted of 2035 subjects aged 6-14-year-old. Diagnoses of psychiatric disorders were performed with the Development and Well-Being Assessment (DAWBA). Five non-overlapping groups without comorbidity were formed: phobic disorders (n = 101), distress disorders (n = 82), attention deficit hyperactivity disorder (ADHD) (n = 133), oppositional defiant disorder/conduct disorder (ODD/CD) (n = 52) and one group of typical developing comparisons (TDC) (n = 1667). Fine motor ability was evaluated by three tasks: sequential finger-thumb opposition, Oseretsky, and pronation/supination tests. Each task was assessed by total time to perform the movement and levels of accuracy, fluency, symmetry, precision, and coordination. We found that, when compared to TDC, the ADHD group performed more poorly in total fine motor ability (mean difference = - 0.28; p = 0.014), time to complete the task (mean difference = 0.36; p < 0.001), accuracy (mean difference = - 0.30; p = 0.005), fluency (mean difference = - 0.26; p = 0.03), symmetry (mean difference - 0.25; p = 0.04), precision (mean difference = - 0.26; p = 0.026), and coordination (mean difference = - 0.25; p = 0.042). The ADHD group also took more time to complete the task than the ODD/CD group (mean difference = 0.45; p = 0.037). No other significant between-group differences were found. Additional analyses showed those differences were statistically significant for all ADHD dimensions. Our results suggest that children with ADHD present impairments in all aspects of fine motor abilities compared to TDC, a difference not found in other psychiatric disorders.

The Social Aptitudes Scale: looking at both "ends" of the social functioning dimension.

PURPOSE: Dimensional approaches are likely to advance understanding of human behaviors and emotions. Nevertheless, it is unclear whether instruments in psychiatry capture variability at the full spectrum of these dimensions. We aimed to investigate this issue for two scales assessing distinct aspects of social functioning: the Social Aptitudes Scale (SAS), a "bidirectional" scale constructed to investigate both "ends" of social functioning; and the social Child Behavior Checklist (CBCL-social), a "unidirectional" scale constructed to assess social problems. METHODS: We investigated 2512 children and adolescents aged 6-14. Item response theory was used to investigate on which range of the trait each scale captures information. We performed quantile regressions to investigate if correlations between SAS and CBCL-social vary within different levels of social aptitudes dimension and multiple logistic regressions to investigate associations with negative and positive clinical outcomes. RESULTS: SAS was able to provide information on the full range of social aptitudes, whereas CBCL-social provided information on subjects with high levels of social problems. Quantile regressions showed SAS and CBCL-social have higher correlations for subjects with low social aptitudes and non-significant correlations for subjects with high social aptitudes. Multiple logistic regressions showed that SAS was able to provide independent clinical predictions even after adjusting for CBCL-social scores. CONCLUSIONS: Our results provide further validity to SAS and exemplify the potential of "bidirectional" scales to dimensional assessment, allowing a better understanding of variations that occur in the population and providing information for children with typical and atypical development.

Default mode network maturation and psychopathology in children and adolescents.

BACKGROUND: The human default mode (DMN) is involved in a wide array of mental disorders. Current knowledge suggests that mental health disorders may reflect deviant trajectories of brain maturation. METHOD: We studied 654 children using functional magnetic resonance imaging (fMRI) scans under a resting-state protocol. A machine-learning method was used to obtain age predictions of children based on the average coefficient of fractional amplitude of low frequency fluctuations (fALFFs) of the DMN, a measure of spontaneous local activity. The chronological ages of the children and fALFF measures from regions of this network, the response and predictor variables were considered respectively in a Gaussian Process Regression. Subsequently, we computed a network maturation status index for each subject (actual age minus predicted). We then evaluated the association between this maturation index and psychopathology scores on the Child Behavior Checklist (CBCL). RESULTS: Our hypothesis was that the maturation status of the DMN would be negatively associated with psychopathology. Consistent with previous studies, fALFF significantly predicted the age of participants (p < .001). Furthermore, as expected, we found an association between the DMN maturation status (precocious vs. delayed) and general psychopathology scores (p = .011). CONCLUSIONS: Our findings suggest that child psychopathology seems to be associated with delayed maturation of the DMN. This delay in the neurodevelopmental trajectory may offer interesting insights into the pathophysiology of mental health disorders.

Attention-deficit/hyperactivity disorder dimensionality: the reliable 'g' and the elusive 's' dimensions.

The best structural model for attention-deficit/hyperactivity disorder (ADHD) symptoms remains a matter of debate. The objective of this study is to test the fit and factor reliability of competing models of the dimensional structure of ADHD symptoms in a sample of randomly selected and high-risk children and pre-adolescents from Brazil. Our sample comprised 2512 children aged 6-12 years from 57 schools in Brazil. The ADHD symptoms were assessed using parent report on the development and well-being assessment (DAWBA). Fit indexes from confirmatory factor analysis were used to test unidimensional, correlated, and bifactor models of ADHD, the latter including "g" ADHD and "s" symptom domain factors. Reliability of all models was measured with omega coefficients. A bifactor model with one general factor and three specific factors (inattention, hyperactivity, impulsivity) exhibited the best fit to the data, according to fit indices, as well as the most consistent factor loadings. However, based on omega reliability statistics, the specific inattention, hyperactivity, and impulsivity dimensions provided very little reliable information after accounting for the reliable general ADHD factor. Our study presents some psychometric evidence that ADHD specific ("s") factors might be unreliable after taking common ("g" factor) variance into account. These results are in accordance with the lack of longitudinal stability among subtypes, the absence of dimension-specific molecular genetic findings and non-specific effects of treatment strategies. Therefore, researchers and clinicians might most effectively rely on the "g" ADHD to characterize ADHD dimensional phenotype, based on currently available symptom items.

Decreased centrality of subcortical regions during the transition to adolescence: a functional connectivity study.

Investigations of brain maturation processes are a key step to understand the cognitive and emotional changes of adolescence. Although structural imaging findings have delineated clear brain developmental trajectories for typically developing individuals, less is known about the functional changes of this sensitive development period. Developmental changes, such as abstract thought, complex reasoning, and emotional and inhibitory control, have been associated with more prominent cortical control. The aim of this study is to assess brain networks connectivity changes in a large sample of 7- to 15-year-old subjects, testing the hypothesis that cortical regions will present an increasing relevance in commanding the global network. Functional magnetic resonance imaging (fMRI) data were collected in a sample of 447 typically developing children from a Brazilian community sample who were submitted to a resting state acquisition protocol. The fMRI data were used to build a functional weighted graph from which eigenvector centrality (EVC) was extracted. For each brain region (a node of the graph), the age-dependent effect on EVC was statistically tested and the developmental trajectories were estimated using polynomial functions. Our findings show that angular gyrus become more central during this maturation period, while the caudate; cerebellar tonsils, pyramis, thalamus; fusiform, parahippocampal and inferior semilunar lobe become less central. In conclusion, we report a novel finding of an increasing centrality of the angular gyrus during the transition to adolescence, with a decreasing centrality of many subcortical and cerebellar regions.

Temporal stability of network centrality in control and default mode networks: Specific associations with externalizing psychopathology in children and adolescents.

Abnormal connectivity patterns have frequently been reported as involved in pathological mental states. However, most studies focus on "static," stationary patterns of connectivity, which may miss crucial biological information. Recent methodological advances have allowed the investigation of dynamic functional connectivity patterns that describe non-stationary properties of brain networks. Here, we introduce a novel graphical measure of dynamic connectivity, called time-varying eigenvector centrality (tv-EVC). In a sample 655 children and adolescents (7-15 years old) from the Brazilian "High Risk Cohort Study for Psychiatric Disorders" who were imaged using resting-state fMRI, we used this measure to investigate age effects in the temporal in control and default-mode networks (CN/DMN). Using support vector regression, we propose a network maturation index based on the temporal stability of tv-EVC. Moreover, we investigated whether the network maturation is associated with the overall presence of behavioral and emotional problems with the Child Behavior Checklist. As hypothesized, we found that the tv-EVC at each node of CN/DMN become more stable with increasing age (P < 0.001 for all nodes). In addition, the maturity index for this particular network is indeed associated with general psychopathology in children assessed by the total score of Child Behavior Checklist (P = 0.027). Moreover, immaturity of the network was mainly correlated with externalizing behavior dimensions. Taken together, these results suggest that changes in functional network dynamics during neurodevelopment may provide unique insights regarding pathophysiology.

An Evaluation of Item Harmonization Strategies Between Assessment Tools of Psychopathology in Children and Adolescents.

Data aggregation in mental health is complicated by using different questionnaires, and little is known about the impact of item harmonization strategies on measurement precision. Therefore, we aimed to assess the impact of various item harmonization strategies for a target and proxy questionnaire using correlated and bifactor models. Data were obtained from the Brazilian High-Risk Study for Mental Conditions (BHRCS) and the Healthy Brain Network (HBN; N = 6,140, ages 5-22 years, 39.6% females). We tested six item-wise harmonization strategies and compared them based on several indices. The one-by-one (1:1) expert-based semantic item harmonization presented the best strategy as it was the only that resulted in scalar-invariant models for both samples and factor models. The between-questionnaires factor correlation, reliability, and factor score difference in using a proxy instead of a target measure improved little when all other harmonization strategies were compared with a completely at-random strategy. However, for bifactor models, between-questionnaire specific factor correlation increased from 0.05-0.19 (random item harmonization) to 0.43-0.60 (expert-based 1:1 semantic harmonization) in BHRCS and HBN samples, respectively. Therefore, item harmonization strategies are relevant for specific factors from bifactor models and had little impact on p-factors and first-order correlated factors when the child behavior checklist (CBCL) and strengths and difficulties questionnaire (SDQ) were harmonized.

The interplay between ADHD and school shift on educational outcomes in children and adolescents: A cross-sectional and longitudinal analysis.

Many countries implement a double-shift schooling system, offering morning or afternoon shifts, driven by diverse factors. Young people with ADHD may face educational problems attending morning shifts compared to afternoon shifts. To investigate this, we used data from a Brazilian school-based cohort (n = 2.240, 6-14 years old, 45.6% female; 50.2% in the morning shift; 11.2% with ADHD). ADHD was determined by child psychiatrists using semi-structured interview. Educational outcomes were measured cross-sectionally and three years later (80% retention) and included reading and writing ability, performance in school subjects, and any negative school events (repetition, suspension, or dropout). Generalized regression models tested the interaction between ADHD and school shift and were adjusted for age, sex, race/ethnicity, intelligence, parental education, socioeconomic status, and site. Attrition was adjusted with inverse probability weights. We used two dimensional measures of attentional problems as sensitivity analysis. ADHD and morning shift were independently associated with lower reading and writing ability and with higher odds for negative school events cross sectionally. ADHD independently predicted lower performance in school subjects and higher negative school events at follow-up. Interaction was found only at the cross-sectional level in a way that those studying in the afternoon present better educational outcomes compared with those studying in the morning only if they have lower ADHD symptom. Thus, ADHD was not associated with poorer educational outcomes among those studying in the morning. However, participants studying in the afternoon with lower levels of attentional problems presented better educational, despite these associations fade away over time.