TYPE Original Research PUBLISHED 12 April 2023 DOI 10.3389/fpsyt.2023.1136994 The impact of attention-deficit/ OPEN ACCESS hyperactivity disorder and specific learning disorders on academic EDITED BY José Carlos Núñez, University of Oviedo, Spain REVIEWED BY Debora Areces, performance in Spanish children University of Oviedo, Spain Belen Rosello Miranda, University of Valencia, Spain from a low-middle- and a Celestino Rodríguez, Universidad de Oviedo Mieres, Spain high-income population *CORRESPONDENCE Mireia Pagerols mireia.pagerols@sjd.es Gemma Español-Martín 1,2,3,4, Mireia Pagerols 5,6*, Raquel Prat 5,7, SPECIALTY SECTION Cristina Rivas 5, Josep Antoni Ramos-Quiroga 1,2,3,4, This article was submitted to Miquel Casas 3,5 and Rosa Bosch 4,5 ADHD, a section of the journal 1 Servei de Psiquiatria, Vall d’Hebron Hospital Universitari, Barcelona, Spain, 2 Grup de Psiquiatria, Salut Frontiers in Psychiatry Mental i Addiccions, Vall d’Hebron Institut de Recerca (VHIR), Vall d’Hebron Hospital Universitari, Barcelona, Spain, 3 Departament de Psiquiatria i Medicina Legal, Universitat Autònoma de Barcelona, RECEIVED 03 January 2023 Bellaterra, Spain, 4 Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Instituto de ACCEPTED 20 March 2023 Salud Carlos III, Madrid, Spain, 5 Programa MIND Escoles, Hospital Sant Joan de Déu, Institut de Recerca PUBLISHED 12 April 2023 Sant Joan de Déu, Esplugues de Llobregat, Spain, 6 Unitat de Farmacologia, Departament de Fonaments CITATION Clínics, Facultat de Medicina i Ciències de la Salut, Universitat de Barcelona (UB), Barcelona, Spain, Español-Martín G, Pagerols M, Prat R, Rivas C, 7 Centre for Health and Social Care Research (CEES), University of Vic-Central University of Catalonia Ramos-Quiroga JA, Casas M and (UVic-UCC), Vic, Spain Bosch R (2023) The impact of attention-deficit/ hyperactivity disorder and specific learning disorders on academic performance in Spanish children from a low-middle- and a high- income population. Introduction: Past research has demonstrated that attention-deficit/hyperactivity Front. Psychiatry 14:1136994. disorder (ADHD), specific learning disorders (SLD), and socioeconomic status (SES) doi: 10.3389/fpsyt.2023.1136994 affect a host of educational outcomes. However, there are no studies examining COPYRIGHT whether SES moderates the association between these neurodevelopmental © 2023 Español-Martín, Pagerols, Prat, Rivas, Ramos-Quiroga, Casas and Bosch. This is an disorders (ND) and the academic achievement of children and adolescents. open-access article distributed under the terms The present investigation examined the impact of ADHD and SLD on academic of the Creative Commons Attribution License performance in 1,287 Spanish students aged 5–17 from a low-middle (LM)- and (CC BY). The use, distribution or reproduction in other forums is permitted, provided the a high-income population, when adjusted for comorbidity and demographic original author(s) and the copyright owner(s) factors that may influence educational functioning. are credited and that the original publication in this journal is cited, in accordance with Methods: Parents completed a questionnaire regarding demographic data accepted academic practice. No use, along with the Strengths and Difficulties Questionnaire. Additionally, teachers distribution or reproduction is permitted which provided information on learning difficulties trough the Protocol for Detection does not comply with these terms. and Management of Dyslexia. Teacher’s Version. Academic performance across multiple domains (i.e., first language, foreign language, mathematics) was obtained from school records. ND were determined using standardized diagnostic methods based on the Diagnostic and Statistical Manual of Mental Disorders criteria. To examine the effects of ADHD and SLD on academic achievement and the potential moderating role of SES, a series of ordinal logistic regressions were conducted. Results: Emotional/behavioral problems, learning difficulties, and ND were more frequent among individuals from the LM-income population. After controlling for gender, age, parental divorce/separation, grade retention, frequency of screen use, and daily meals, both ADHD and SLD were associated with worse educational outcomes. Lower SES also increased the risk for academic impairment, although the interactions with ADHD or SLD were not significant. Frontiers in Psychiatry 01 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 Conclusion: These findings indicate that ADHD and SLD exert a pervasive impact on academic performance across different socioeconomic backgrounds. Therefore, early detection and effective intervention strategies aimed at students with these ND are crucial to improve their educational functioning and mitigate the negative consequences related to academic problems. KEYWORDS Strengths and Difficulties Questionnaire (SDQ), attention-deficit/hyperactivity disorder (ADHD), specific learning disorders (SLD), socioeconomic status (SES), academic performance Introduction with SLD have also been found to earn poorer GPA, require more educational support, and have lower high school and postsecondary Neurodevelopmental disorders (ND) emerge at an early age and completion rates (22, 23). In addition, poor readers are more affect normal development, producing delays in the expected social, vulnerable to mental health problems and negative outcomes than emotional, language, and cognitive performance. Attention-deficit/ their peers, including low self-concept, anxiety, depression, attempted hyperactivity disorder (ADHD) and specific learning disorders (SLD) suicide, unemployment, and incarceration (22–26). Consequently, are among the most common ND in school-age children worldwide, children exhibiting reading, math, or spelling difficulties in with prevalence estimates ranging from 2 to 7% and 5 to 15%, combination with attentional deficits could be much more impaired respectively (1–3). ADHD is characterized by high levels of in learning than those with SLD or ADHD alone (27). In this vein, inattention, hyperactivity, impulsivity, and disruptive behavior. some authors found that subjects who meet criteria for both ADHD Depending on the predominance of the symptoms, it can be classified and SLD usually display more severe neurocognitive deficits in into three clinical presentations: predominantly inattentive, executive functions and working memory, negative academic predominantly hyperactive–impulsive, and combined if six or more experience, and higher risk of school retention than those with either symptoms of each type are present for at least 6 months. Children with isolated condition (18, 28, 29). SLD, on the other hand, exhibit significant and persistent difficulties On the other hand, demographic characteristics such as gender, in learning academic skills, such as reading, written expression, and/ SES, stressful events (e.g., adoption, parental divorce/separation, grade or mathematics, despite adequate instruction and intelligence. In retention), and lifestyle behaviors (e.g., diet, screen time, sleep particular, dyslexia, which is one of the most common SLD, involve duration) are also important variables that affect academic deficits in word reading accuracy, reading fluency, and reading performance in children (30–35). In particular, school failure and comprehension. Dysgraphia can manifest as impairment in spelling poor academic achievement are strongly related to socioeconomic accuracy, grammar/punctuation, and clarity or organization of written factors operating at individual and community levels (36–38). In this expression. Finally, children with impairment in mathematics may sense, multiple studies have reported the beneficial effects of living in show problems in basic number processing, arithmetic facts, a high-SES neighborhood on school readiness, verbal ability, reading calculation skills, and math reasoning (1). skills, math achievement, and GPA, even after accounting for Past research has demonstrated a strong negative relation between individual and family characteristics (36–38). Neighborhood affluence ADHD symptoms and academic outcomes in clinical and community has also been positively associated with youth’s chances of completing samples, with attention problems being the primary predictor of high school, attending college, and years of schooling completed (36). poorer achievement from early childhood into adulthood (4–6). These By contrast, children and adolescents from low SES communities are associations remain even after controlling for intelligence, psychiatric at high risk for negative educational outcomes, including reading comorbidity, and socioeconomic status (SES) (7–9). Specifically, impairment, school drop-out, and academic underachievement (36, children with ADHD perform worse on standardized academic tests, 38, 39). Of note, Fluss et al. (39) examined the prevalence of reading have lower grade point average (GPA), and increased rates of disabilities in 1,062 children distributed across three educational absenteeism, grade retention, special education need, and school zones (i.e., low, medium, and high-SES) based on multiple social and drop-out (4, 10, 11), which place them at risk for educational and demographic indicators, such as the parents’ professional background, occupational difficulties during adulthood. Indeed, a growing body of the rate of unemployment in the district area, the percentage of literature has documented a wide range of long-term consequences disadvantaged families, and the proportion of non-native speakers associated with ADHD such as low self-esteem, high-risk behaviors, living in the proximity of the school. The authors revealed that reading disrupted relationships, delinquency, substance use, unemployment, impairment was highly influenced by neighborhood SES, with lifelong disadvantage, and the development of other mental disorders estimates ranging from 3.3% in high SES to 24.2% in low SES areas (e.g., antisocial disorders, depression, anxiety, learning disabilities) (39). Besides, a strong association exists between living in (12–16). For instance, studies suggest that between 25 and 50% of disadvantaged communities and the occurrence of other psychiatric children with ADHD have a comorbid learning disorder, with rates disturbances, including behavioral/emotional problems, ADHD, and ranging from 18 to 45% for reading disabilities, 9 to 63% for writing depression (36, 40–42). difficulties, and 11 to 30% for math-related deficits (17–21). This Thus, literature consistently shows that ADHD, SLD, and SES comorbidity augments the risk for academic failure, since subjects affect a host of educational outcomes in children and adolescents (4, Frontiers in Psychiatry 02 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 22, 36). However, there are still controversial issues that warrant The schools were equally distributed between two different further investigation. First, most studies featured relatively small territories of the Barcelona Metropolitan Area: Rubí, a LM-income samples, focused on academic attainment at the end of compulsory area with 74,536 inhabitants, and Sant Cugat del Vallès, a high-income schooling, and relied on screening measures or self-reports for area with 87,830 inhabitants (43). Socioeconomic level was established assessing the presence of ND instead of standardized diagnostic based on a territorial socioeconomic indicator (TSI) that measures the methods. Second, previous research on the relation between ADHD, socioeconomic characteristics of the Catalan population according to SLD, and academic impairment has rarely examined these ND the employment status (i.e., percentage of employed residents, simultaneously. Likewise, additional confounding factors, such as percentage of employed residents in blue-collar occupations), gender, age, stressful events and lifestyle behaviors, have not been educational level (i.e., percentage of illiterate people or with primary accounted for in the majority of investigations. Last, there have been education, percentage of residents with a junior school degree or no studies examining whether SES moderates the association between below), immigration (i.e., percentage of foreign people from ADHD, SLD, and the academic achievement of youth on specific LM-income countries), and household income of the individuals who domains (e.g., first language, foreign language, mathematics). To live in a small geographic unit (44). A reference value for Catalonia is address this gap in the literature and overcome some of the 100 and a value for each territorial unit is established in comparison methodological limitations of earlier investigations, the current with the average Catalan value. In 2015, the mean TSI for the research aimed to evaluate for the first time the impact of ADHD and LM-income and high-income area was 97.8 and 124.7, respectively, SLD on academic performance in 1,287 Spanish children distributed which corresponded to the fourth and tenth decile. Data for both across a low-middle (LM)-income and a high-income population, territorial units were obtained from the Statistical Institute of when adjusted for comorbidity and demographic variables that may Catalonia (45). influence educational functioning (i.e., gender, age, nationality, The LM- and the high-income sample included 836 and 694 adoption, parental divorce/separation, grade retention, dietary habits, participants, respectively. Overall, 51.8% (n = 793) were boys and screen use, sleep duration). The specific objectives were as follows: (a) 48.2% were girls, with a mean age of 9.83 years (SD = 2.92; to compare the prevalence of psychopathology, learning difficulties, range = 5–17 years). Demographic characteristics for the total sample and ND across a LM-income and a high-income population, (b) to and by SES are summarized in Table 1. examine whether SES, ADHD, and SLD were significantly related to academic performance, and (c) to determine whether the association between these ND and educational outcomes was modified by Procedure SES. We hypothesized that mental health problems, including psychiatric symptoms, learning difficulties, and ND would be more The Department of Health and the Department of Education common among individuals from the LM-income group. We also (Generalitat de Catalunya, Spain) authorized the project, and ethical hypothesized that lower SES would be associated with poor approval was granted by the Ethics Committee of the Vall d’Hebron performance across multiple domains. Further, we expected that Hospital Universitari (PR(AG)72-2012), in Barcelona. Six schools ADHD and SLD would increase the risk for academic impairment. from the Barcelona Metropolitan Area were contacted in two Finally, we hypothesized that SES would moderate the relation academic years (i.e., 2015–2016, 2016–2017) and invited to participate between ADHD, SLD, and educational outcomes, such that the after explaining the study to the school staff. All of them accepted, negative effect of ND on academic performance would be mitigated which included four public primary and two secondary schools for children living in high SES areas. In the context of the extensive equally distributed across a LM-income and a high-income area. All early and long-term consequences associated with such ND, it is students enrolled in the selected schools, with ages comprised between important to determine potential factors that may buffer or worsen 5 and 17 years, were considered for study inclusion, resulting in 2,135 the academic difficulties faced by students with ADHD and SLD in eligible subjects. A two-stage procedure was applied: in the first phase order to design and implement targeted interventions that improve of the project (i.e., from September to April), families were informed the academic achievement of those particularly vulnerable. Thus, and we obtained informed consent for 1,699 children (participation results from this study will serve a practical use for clinicians, rate = 79.6%), 693 of whom were at least 11 and also gave permission. education researchers, and policymakers in their efforts to prevent Parents of the participating students received a questionnaire school failure and promote students’ mental health by identifying regarding demographic data and school-related factors along with the those individuals who are at higher risk of underperforming. Strengths and Difficulties Questionnaire (SDQ) (46), which was completed at home. Additionally, the main classroom teachers provided information on reading and writing difficulties through the Materials and methods Protocol for Detection and Management of Dyslexia. Teacher’s Version (PRODISCAT) (47), and youth over 11 years completed the Participants self-reported SDQ. Cases with missing values in the SDQ were removed. Thus, the final sample of this first phase comprised 1,530 The current cross-sectional population-based sample is part of a children and adolescents. Of them, 767 were identified as potential larger research called INSchool, which started in 2011 with the aim of cases in accordance with the following criteria: (a) a score in the identifying children and adolescents’ mental health problems in a clinical range on any of the parent-reported SDQ problem scales (i.e., school setting. For the current investigation, we used data from six Emotional symptoms, Conduct problems, Hyperactivity/inattention, different schools, which participated in this ongoing research during Peer problems); (b) five or more high-risk indicators on the the academic years 2015–2016 and 2016–2017. PRODISCAT; or (c) a previous diagnosis of ND from a medical Frontiers in Psychiatry 03 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 TABLE 1 Demographic characteristics of the study samples. LM-income High-income Total sample sample sample Statistic Effect size Value of p (n=1,530) (n=836) (n=694) Gender (n, %) χ2 = 4.53 Cramer’s V = 0.054 0.033  Boys 793 (51.8) 454 (54.3) 339 (48.8)  Girls 737 (48.2) 382 (45.7) 355 (51.2) Age (M, SD) 9.83 (2.92) 9.74 (2.89) 9.95 (2.95) NS NS NS Educational stage (n, %) NS NS NS  Primary 918 (60.0) 517 (61.8) 401 (57.8)  Secondary 612 (40.0) 319 (38.2) 293 (42.2) Nationality (n, %) χ2 = 14.8 Cramer’s V = 0.098 < 0.001  Spanish 1,462 (95.8) 814 (97.6) 648 (93.6)  Foreign origin 64 (4.19) 20 (2.40) 44 (6.36) Adoption (n, %) χ2 = 13.7 Cramer’s V = 0.097 < 0.001  No 1,434 (98.8) 786 (99.7) 648 (97.6)  Yes 18 (1.24) 2 (0.25) 16 (2.41) Parental divorce/separation (n, %) χ2 = 7.41 Cramer’s V = 0.071 0.006  No 1,140 (78.0) 602 (75.3) 538 (81.3)  Yes 321 (21.0) 197 (24.7) 124 (18.7) Frequency of screen use (M, SD) 2.11 (0.72) 2.12 (0.73) 2.10 (0.71) NS NS NS Sleeping hours (M, SD) 9.55 (0.98) 9.47 (0.99) 9.67 (0.95) t = −3.66 Cohen’s d = 0.208 < 0.001 Three meals a day (n, %) χ2 = 14.1 Cramer’s V = 0.096 < 0.001  No 134 (8.78) 94 (11.3) 40 (5.79)  Yes 1,392 (91.2) 741 (88.7) 651 (94.2) Differences in sample sizes across variables are due to missing data. For χ2 tests with degrees of freedom equal to 1, the effect size is regarded as small for values > 0.10, moderate for values between 0.30 and 0.50, and strong for values > 0.50. With regard to Cohen’s d, a value of 0.20 is considered small, 0.50 medium, and 0.80 large. LM, low-middle; M, mean; NS, non-significant; SD, standard deviation. professional as reported by parents. Subjects who screened positive included students with a negative screening score and those who were subsequently invited to participate in the second phase of the underwent the diagnostic assessment (n = 1,287). Finally, the project (i.e., from January to June). Parents provided consent for 558 information on academic performance was obtained from school of them (participation rate = 72.8%) and children were interviewed for records at the end of the academic year. Figure 1 describes the study diagnostic confirmation based on the Diagnostic and Statistical design and data collection process. Manual of Mental Disorders (DSM) criteria. Interviews were carried out on separate days within the school facilities so all students could complete the clinical assessment, which avoided sample attrition. Measures Participants along with their parents met a psychiatrist of the research team, who determined the presence of ADHD using the Kiddie Demographic factors Schedule for Affective Disorders and Schizophrenia Present and Demographic data included child’s gender, age, country of birth, Lifetime Version (K-SADS/PL) (48). Besides, students were granted adoption, parental divorce/separation, and whether he or she had an additional appointment with an experienced neuropsychologist repeated any grade. Parents were asked if their child had ever received who conducted a comprehensive neuropsychological battery for the a diagnosis of ADHD, SLD, or another developmental disorder from evaluation of SLD, in addition to a full exploration of their cognitive a medical professional. They also provided information about their abilities using the Wechsler Intelligence Scale for Children, fourth child’s eating habits by indicating how often they had breakfast, lunch (WISC-IV) (49) or fifth edition (WISC-5) (50). The following and dinner (1 = never, 2 = occasionally, and 3 = always), and reported standardized tests were administered to evaluate the reading and how frequently the student played video games, used the cell phone, writing performance of youth: Battery for the Evaluation of Reading and social networks. Each of these three categories had four possible Processes, Revised (PROLEC-R) (51), Battery for the Evaluation of responses (1 = never, 2 = sometimes, 3 = often, and 4 = always) and were Reading Processes in Junior and Senior High-School Students, summed to compute the mean frequency of screen use. Finally, sleep Revised (PROLEC-SE-R) (52), Test for the Analysis of Reading and duration was derived from the answer to the following questions: “At Writing (TALE) (53), and Battery for the Evaluation of Writing what time does your child usually go to bed?” and “At what time does Processes (PROESC) (54). In total, the second-phase population your child usually wake up?” Frontiers in Psychiatry 04 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 children aged 5–17 by Español-Martín et al. (55), who also provided Spanish normative data according to the child’s gender, age and type of informant. The frequency of subjects with a score in the clinical range on each SDQ subscale was estimated based on the bands developed by those authors (55). Ordinal alpha coefficients yielded adequate reliability estimates across all subscales, with values ranging from 0.76 to 0.89, both for the parent and adolescent version. Protocol for detection and management of dyslexia: Teacher’s version The PRODISCAT (47) is a screening instrument aimed at teachers and professors of primary and secondary education to detect children who present reading and/or writing difficulties. It is available in five different versions, depending on the educational stage, with good internal consistency (Cronbach’s alpha = 0.75–0.95) (56). These include 23–44 items, some of which are high-risk indicators that require intervention (e.g., “He/she has difficulties in lexical access when speaking,” “He/she makes many spelling mistakes compared to the class group,” “His/her reading speed is slow compared to the class group”). Each item has two response options (0 = no, 1 = yes). In the present study, subjects with five or more high- risk indicators on the PRODISCAT were regarded as potential cases of dyslexia. Kiddie schedule for affective disorders and schizophrenia present and lifetime version The K-SADS/PL (48) is a semi-structured interview that assesses current and past psychopathology in children and adolescents according to the DSM. It contains an introductory interview with questions about basic demographic characteristics, complaints, and prior psychiatric problems, followed by an 82-symptom screen interview and five diagnostic supplements: (a) Affective disorders, (b) Psychotic disorders, (c) Anxiety disorders, (d) ADHD and behavioral disorders; and (e) Substance abuse, tic, eating, and elimination disorders. Items are scored using a 0- to 3-point scale, where 0 means no information is available, 1 suggests the symptom is not present, 2 FIGURE 1 Flowchart of the study design and data collection process. *Cases indicates subthreshold levels of symptomatology, and 3 represents were defined according to: a score in the clinical range on any of the threshold criteria. The diagnostic supplement for a given area is only parent-reported SDQ problem scales; five or more high-risk completed if the child receives at least one threshold rating on any of indicators on the PRODISCAT; or a previous diagnosis of neurodevelopmental disorder from a medical professional. the symptoms surveyed in that section of the screen interview. The Spanish version of the K-SADS/PL, which has shown good psychometric properties (57, 58), was administered to parents and children/adolescents separately. Strengths and difficulties questionnaire The SDQ (46) is a screening instrument that can Battery for the evaluation of reading processes, be administered to parents of children aged 4–17 or as self-reports revised in subjects over 11 years old. The questionnaire covers common The PROLEC-R (51) is one of the most extensively used areas of social, emotional, and behavioral functioning through 25 instruments to assess reading performance in Spanish children aged items distributed across five subscales: Emotional symptoms, 6–12 years. The battery explores the perceptual, lexical, syntactic, and Conduct problems, Hyperactivity/inattention, Peer problems, and semantic processes involved in reading comprehension through nine Prosocial behavior. Each dimension consists of five items with three tasks, namely: Name or sound of letters, Equal-different, Words response options (0 = not true, 1 = somewhat true, and 2 = certainly reading, Pseudo-words reading, Grammatical structures, Punctuation, true), thus yielding a score that ranges between 0 and 10. Sentences comprehension, Text comprehension, and Listening Furthermore, a Total difficulties score may be derived by summing comprehension. Adequate values of internal consistency, indexed by the first four subscales (range = 0–40). The parent and adolescent ordinal alpha, have been reported for all subtests (range = 0.76–0.95) versions of the SDQ used in the present study have been validated (51). Reading accuracy and speed are measured by the number of as feasible instruments to identify mental health problems in correct answers and the time spent to complete each task. Frontiers in Psychiatry 05 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 Battery for the evaluation of reading processes in test. The skewness and kurtosis values were all within ±2, suggestive junior and senior high-school students, revised of normal distribution, and Levene’s test determined the homogeneity The PROLEC-SE-R (52) evaluates the reading ability and the of variances (p > 0.05) (59, 60). Therefore, the Student’s t-test was underlying lexical, syntactic, and semantic processes of adolescents applied for comparison of means, and Cohen’s d was reported as the from 12 to 18 years. The battery includes 13 tests (i.e., Lexical selection, effect size measure. Semantic categorization, Word reading, Pseudo-word reading, Grammatical structures I, Grammatical judgment, Grammatical Data analytic plan structures II, Punctuation, Expository comprehension, Narrative In order to determine whether there was evidence of clustering comprehension, Pure reading comprehension, Mnemonic reading within the data, we first ran an intercept-only model for each comprehension, Listening comprehension), with ordinal alpha educational outcome (i.e., grades on first language, foreign language, coefficients ranging from 0.73 to 0.99 (52). mathematics) with school as a random effect. Results from the unconditional models justified the use of a single-level approach, since Test for the analysis of reading and writing the variance components for school were not significant, indicating The TALE (53) allows to determine the general level and specific that grades within schools were not different than grades between characteristics of reading and writing in children from first to fourth schools. A series of backward regressions were then performed to grade of primary school (6–10 years). It contains a battery of subtests, remove non-significant demographic predictors, followed by multiple which evaluate the reading of letters, syllables, words and texts, as well ordinal logistic regressions that retained the significant demographic as reading comprehension. The writing assessment consist of copying predictors identified in the preliminary analysis and added the syllables, words and sentences, a dictation task, and writing a primary variables of interest as described below. Analyses were composition. In addition, it includes a part to examine aspects related conducted separately for each educational outcome, with A as the to graphology, such as letters size and form, separation between lowest category and D as the highest. words, etc. The number of mistakes and the time required to complete each task are compared to those from the general population. Preliminary analyses Gender, age, nationality, adoption, parental divorce/separation, Battery for the evaluation of writing processes grade retention, frequency of screen use, sleep duration, and having The PROESC (54) aims to evaluate the main writing processes in three meals a day were entered into a series of backward regressions. children from third grade of primary to adolescents in secondary Significant predictors in each model (p < 0.05) were retained. education (8–15 years old). The battery is composed of six subtests (i.e., Dictation of syllables, Dictation of words, Dictation of pseudo- Primary analyses words, Dictation of sentences, Writing a story, Writing an essay) and To examine the impact of ADHD and SLD on academic has demonstrated good internal consistency for the total score achievement, and the hypothesized moderating role of SES, a series of (Cronbach’s alpha = 0.82) (54). The student’s writing performance is ordinal logistic regressions were performed. As noted above, the determined by comparing the number of correct answers in each significant demographic predictors from the preliminary analyses subtest to the scale for the corresponding school year. were entered in Step 1. ADHD and SLD (Step 2), and SES (Step 3) were then added to test main effects, while controlling for comorbidity. Academic performance Last, we included interaction terms (i.e., ADHD × SES, SLD × SES) Academic performance on four core subjects (i.e., Catalan, into the final model (Step 4) and conducted stratified analyses in case Spanish, English as foreign language, and mathematics) was obtained of significant interactions. The Nagelkerke R2 was used to measure the from school records at the end of the academic year. However, schools global predictive capacity of each model. Odds ratio (OR) and their varied in the way they scored those subjects and, thus, grades were corresponding 95% confidence interval (CI) were reported. A converted into a 4-point scale from D to A (D = unsatisfactory two-sided p-value of 0.05 was set as significance level in all tests. achievement, fail, 0–4.9; C = satisfactory achievement, pass, average, 5–6.9; B = good achievement, above average, 7–8.9; and A = excellent achievement, 9–10). Given the high correlation between Catalan and Results Spanish grades (r = 0.75, p < 0.001), we calculated the academic performance on first language as the mean of the two scores. Sample characteristics The LM- and the high-income sample included 836 and 694 Statistical analyses subjects, respectively. Overall, 51.8% (n = 793) of the participants were boys and 48.2% were girls, with a mean age of 9.83 years (SD = 2.92; All analyses were performed with SPSS 22.0. Descriptive statistics range = 5–17 years). The proportion of boys was higher in the LM- were calculated to illustrate the demographic characteristics, clinically than in the high-SES sample (54.3% vs. 48.8%, p = 0.033), while foreign relevant cases based on the SDQ, academic performance, and origin was more prevalent among children from the high-income prevalence of ADHD and SLD in the total sample and by group (6.36% vs. 2.40%, p < 0.001). The proportion of participants SES. Individuals from the LM-income and the high-income reporting parental divorce/separation, adoption, and not having three population were compared with the χ2 test for categorical variables meals a day was larger in the LM-income population (Table 1). Finally, and Cramer’s V was used to estimate the strength of the significant no significant differences were detected in the frequency of screen use associations. For continuous variables, we examined the normality (Table 1), although subjects from the high-SES sample slept more and homoscedasticity of data using skewness, kurtosis, and Levene’s hours than those from the LM-income group (9.67 vs. 9.47, p < 0.001). Frontiers in Psychiatry 06 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 Prevalence of psychiatric symptoms and between the LM- and the high-income sample (25.6% vs. 17.3%, learning difficulties p < 0.001). Specifically, 123 students were identified as having ADHD, which represented an overall prevalence rate of 9.56%. Of these, 63 Table 2 presents the proportion of children who fell within the (4.90%) met criteria for the combined presentation, 4.20% (n = 54) clinical range on each SDQ subscale in the total sample and by had the predominantly inattentive presentation, and 0.47% (n = 6) SES. According to both parents’ and adolescents’ perception, were diagnosed with the predominantly hyperactive–impulsive LM-income students exhibited a higher prevalence of clinically presentation. The prevalence of SLD was 16.4% (n = 211), with reading relevant symptoms in all the SDQ subscales, with the exception of the and writing difficulties being present in 15.4% (n = 198) and 6.29% Hyperactivity/inattention scale (Table 2). Besides, we identified a total (n = 81) of the total sample, respectively. Notably, children from the of 349 (28.0%) participants who struggled in reading or writing based LM-income population appeared to have a significantly higher on the PRODISCAT. Of note, learning difficulties were significantly prevalence of SLD (20.7% vs. 11.0%, p < 0.001) and reading difficulties more frequent among individuals with lower SES (34.7% vs. 18.5%, (19.7% vs. 10.0%, p < 0.001), while the rate of writing difficulties p < 0.001). (LM-income: 7.40%; high-income: 4.90%, p = 0.067) and ADHD (LM-income: 8.66%; high-income: 10.7%, p = 0.220) was similar across samples. Last, over 18% (n = 52) of the students who received Academic performance a diagnosis of ND suffered from both ADHD and SLD. Thus, the overall prevalence was 4.04% and did not differ by SES (LM-income: As shown in Table 3 there were significant differences on academic 3.77%; high-income: 4.38%, p = 0.582). performance according to SES, although repeater students were equally distributed across the LM- and the high-income population (5.28% vs. 4.01%, p = 0.249). In particular, students who achieved a Demographic predictors of academic grade of A or B were over-represented within the high-income sample performance for all educational outcomes (i.e., first language, foreign language, mathematics). By contrast, children from the LM-income population As a preliminary analysis, we removed non-significant received lower grades, except on foreign language where the number demographic predictors of academic performance following the of students with a grade of D was greater in the high-income group. backward elimination process for each educational outcome (i.e., grades on first language, foreign language, mathematics). Results indicated that grades on first language and foreign language were Prevalence rates of ADHD and SLD negatively associated with male gender (p < 0.001), parental divorce/ separation (p = 0.005 and p < 0.001, respectively), grade retention According to parent reports, the prevalence of students who (p < 0.001), frequency of screen use (p = 0.008 and p = 0.038, already had a diagnosis of ND prior to the start of the study was 9.25% respectively), and not having three meals a day (p < 0.001 and p = 0.012, (n = 119) and did not differ by SES (LM-income: 10.3%; high-income: respectively); age, nationality, and adoption, on the other hand, were 7.88%, p = 0.131). Based on the diagnostic assessment conducted in not retained (p > 0.05) for the primary analyses described below. the second phase, however, a total of 282 (21.9%) subjects met criteria Significant predictors of poor academic performance on mathematics for at least ADHD or SLD and significant differences were found included age (p < 0.001), parental divorce/separation (p < 0.001), grade TABLE 2 Clinically relevant cases according to the parent- and adolescent-reported SDQ. LM-income High-income Total sample Statistic Effect size Value of p sample sample SDQ parent (n, %)  Emotional symptoms 244 (15.9) 163 (19.5) 81 (11.7) χ2 = 17.3 Cramer’s V = 0.106 < 0.001  Conduct problems 269 (17.6) 174 (20.8) 95 (13.7) χ2 = 13.3 Cramer’s V = 0.093 < 0.001  Hyperactivity/inattention 212 (13.9) 125 (15.0) 87 (12.5) NS NS NS  Peer problems 234 (15.3) 143 (17.1) 91 (13.1) χ = 4.67 2 Cramer’s V = 0.055 < 0.031  Total difficulties 176 (11.5) 120 (14.4) 56 (8.07) χ2 = 14.7 Cramer’s V = 0.098 < 0.001 SDQ adolescents (n, %)  Emotional symptoms 80 (14.3) 53 (18.5) 27 (9.93) χ2 = 8.41 Cramer’s V = 0.123 0.004  Conduct problems 89 (15.9) 64 (22.4) 25 (9.19) χ = 18.1 2 Cramer’s V = 0.180 < 0.001  Hyperactivity/inattention 71 (12.7) 38 (13.3) 33 (12.1) NS NS NS  Peer problems 93 (16.7) 57 (19.9) 36 (13.2) χ = 4.50 2 Cramer’s V = 0.090 0.034  Total difficulties 63 (11.3) 41 (14.3) 22 (8.09) χ = 5.43 2 Cramer’s V = 0.099 0.020 For χ2 tests with degrees of freedom equal to 1, the effect size is regarded as small for values > 0.10, moderate for values between 0.30 and 0.50, and strong for values > 0.50. LM, low-middle; NS, non-significant; SDQ, Strengths and Difficulties Questionnaire. Frontiers in Psychiatry 07 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 TABLE 3 Academic performance on first language, foreign language, and mathematics. LM-income High-income Total sample sample sample Statistic Effect size Value of p (n=1,530) (n=836) (n=694) Grade retention (n, %) NS NS NS  No 1,419 (95.3) 772 (94.7) 647 (96.0)  Yes 70 (4.70) 43 (5.28) 27 (4.01) First language (n, %) χ2 = 27.6 Cramer’s V = 0.135 < 0.001  A 216 (14.3) 102 (12.3) 114 (16.6)  B 516 (34.1) 249 (30.0) 267 (39.0)  C 732 (48.3) 444 (53.6) 288 (42.0)  D 50 (3.30) 34 (4.10) 16 (2.34) Foreign language (n, %) χ2 = 32.0 Cramer’s V = 0.146 < 0.001  A 227 (15.0) 107 (13.0) 120 (17.5)  B 502 (33.2) 243 (29.5) 259 (37.8)  C 691 (45.8) 432 (52.4) 259 (37.8)  D 90 (5.96) 43 (5.21) 47 (6.86) Mathematics (n, %) χ2 = 40.1 Cramer’s V = 0.163 < 0.001  A 227 (15.0) 87 (10.5) 140 (20.4)  B 464 (30.7) 238 (28.8) 226 (33.0)  C 708 (46.9) 432 (52.4) 276 (40.3)  D 111 (7.35) 68 (8.24) 43 (6.28) Differences in sample sizes across variables are due to missing data. For χ tests with degrees of freedom equal to 2, the effect size is regarded as small for values > 0.07, moderate for values 2 between 0.21 and 0.35, and strong for values > 0.35. LM, low-middle; NS, non-significant. retention (p < 0.001), and not having three meals a day (p = 0.038), Foreign language while gender, nationality, adoption, frequency of screen use, and sleep After controlling for gender, parental divorce/separation, grade duration were not retained (p > 0.05) (Tables 4-6; Step 1). retention, frequency of screen use, and daily meals (Table 5; Step 1), ADHD and SLD increased the risk for academic impairment (Table 5; Step 2), while SES failed to predict ADHD, SLD, and SES as predictors of underachievement (Table 5; Step 3). However, with the interaction academic performance terms included in the model (Step 4), subjects from the LM-income population were found to achieve lower grades on foreign language First language than those with higher SES (OR = 1.36, 95% CI = 1.05–1.77, After controlling for gender, parental divorce/separation, grade p = 0.020). Furthermore, the effects of ADHD and SLD were still retention, frequency of screen use, and daily meals (Table 4; Step 1), significant in Step 4 (ADHD: OR = 2.58, 95% CI = 1.40–4.73, ADHD and SLD were significant risk factors for academic p = 0.002; SLD: OR = 5.56, 95% CI = 3.54–8.75, p < 0.001) and did underachievement (Table 4; Step 2). Specifically, children who not differ by SES, since no significant interactions were observed received a diagnosis of ADHD had ~4 times the odds of achieving (Table 5). lower grades on first language (OR = 3.94, 95% CI = 2.48–6.26, p < 0.001), while those with SLD were over 6 times more likely to Mathematics perform poorly (OR = 6.39, 95% CI = 4.33–9.45, p < 0.001). SES was After controlling for age, parental divorce/separation, grade also found to predict academic achievement (Table 4; Step 3), retention, and daily meals (Table 6; Step 1), ADHD and SLD were suggesting that LM-income students performed significantly worse negatively related to academic achievement on mathematics than those from the high-income population (OR = 1.41, 95% (Table 6; Step 2). Step 3 also revealed a significant main effect of CI = 1.12–1.78, p = 0.003). These findings remained in Step 4 (ADHD: SES, since LM-income students tended to perform worse than OR = 3.83, 95% CI = 2.02–7.29, p < 0.001; SLD: OR = 6.89, 95% those from the high-income population (OR = 1.80, 95% CI = 1.46– CI = 4.26–11.1, p < 0.001; SES: OR = 1.38, 95% CI = 1.08–1.78, 2.22, p < 0.001). These associations remained mainly the same p = 0.011), where we added interaction terms into the final model to when interaction terms were introduced in Step 4 (ADHD: examine the potential moderating role of SES. However, there were no OR = 2.91, 95% CI = 1.72–4.91, p < 0.001; SLD: OR = 3.02, 95% significant interactions between ADHD or SLD and SES (Table 4), CI = 2.09–4.36, p < 0.001; SES: OR = 1.85, 95% CI = 1.47–2.31, indicating that these ND exert a pervasive impact on educational p < 0.001), although the interaction between ADHD or SLD and outcomes across different socioeconomic backgrounds. SES was not significant (Table 6). Thus, higher SES did not Frontiers in Psychiatry 08 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 TABLE 4 Predictors of poor academic performance on first language. Nagelkerke R2 OR (95% CI) Value of p Step 1 0.146  Gender (ref. Girls) 1.75 (1.37–2.25) < 0.001  Parental divorce/separation (ref. No) 1.57 (1.14–2.14) 0.005  Grade retention (ref. No) 13.6 (6.52–28.3) < 0.001  Frequency of screen use 1.27 (1.07–1.52) 0.008  Three meals a day (ref. Yes) 2.46 (1.54–3.94) < 0.001 Step 2 0.277  Gender (ref. Girls) 1.77 (1.40–2.25) < 0.001  Parental divorce/separation (ref. No) 1.47 (1.09–1.98) 0.011  Grade retention (ref. No) 10.3 (4.98–21.1) < 0.001  Frequency of screen use 1.28 (1.08–1.52) 0.004  Three meals a day (ref. Yes) 2.34 (1.49–3.66) < 0.001  ADHD (ref. No) 3.94 (2.48–6.26) < 0.001  SLD (ref. No) 6.39 (4.33–9.45) < 0.001 Step 3 0.283  Gender (ref. Girls) 1.74 (1.37–2.20) < 0.001  Parental divorce/separation (ref. No) 1.46 (1.09–1.96) 0.012  Grade retention (ref. No) 10.4 (5.07–21.3) < 0.001  Frequency of screen use 1.28 (1.08–1.52) 0.004  Three meals a day (ref. Yes) 2.18 (1.39–3.41) 0.001  ADHD (ref. No) 4.14 (2.60–6.59) < 0.001  SLD (ref. No) 6.07 (4.10–8.96) < 0.001  SES (ref. High-income sample) 1.41 (1.12–1.78) 0.003 Step 4 0.283  Gender (ref. Girls) 1.73 (1.36–2.19) < 0.001  Parental divorce/separation (ref. No) 1.44 (1.07–1.94) 0.016  Grade retention (ref. No) 10.4 (5.00–21.4) < 0.001  Frequency of screen use 1.28 (1.08–1.52) 0.004  Three meals a day (ref. Yes) 2.20 (1.40–3.45) 0.001  ADHD (ref. No) 3.83 (2.02–7.29) < 0.001  SLD (ref. No) 6.89 (4.26–11.1) < 0.001  SES (ref. High-income sample) 1.38 (1.08–1.78) 0.011  ADHD × SES NS NS  SLD × SES NS NS All steps are incremental. First, the significant demographic predictors were entered in Step 1. ADHD and SLD (Step 2), and SES (Step 3) were then added to test main effects, while controlling for comorbidity. Last, we included interaction terms (i.e., ADHD × SES, SLD × SES) into the final model (Step 4) to examine the potential moderating role of SES. ADHD, attention-deficit/ hyperactivity disorder; CI, confidence interval; NS, non-significant; OR, odds ratio; SES, socioeconomic status; SLD, specific learning disorders. attenuate the risk for academic underachievement conferred by and whether the association was modified by SES, when controlling these ND. for comorbidity and demographic factors that may influence educational functioning. In accordance with previous studies (36, 39–42, 61), we observed Discussion a higher prevalence of emotional/behavioral problems, learning difficulties, and ND among children from the LM-income population, The current research examined whether ADHD and SLD were which supported the hypothesis that low SES would be associated with significantly related to academic achievement across specific subjects negative health outcomes. Interestingly, a prior investigation (i.e., first language, foreign language, mathematics) in a sample of conducted in a representative sample of Catalan children aged 4–14 1,287 Spanish children from a LM- and a high-income population, also showed that those from disadvantaged families were at risk of Frontiers in Psychiatry 09 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 TABLE 5 Predictors of poor academic performance on foreign language. Nagelkerke R2 OR (95% CI) Value of p Step 1 0.156  Gender (ref. Girls) 1.72 (1.37–2.16) < 0.001  Parental divorce/separation (ref. No) 1.70 (1.29–2.26) < 0.001  Grade retention (ref. No) 17.2 (9.37–31.8) < 0.001  Frequency of screen use 1.18 (1.01–1.39) 0.038  Three meals a day (ref. Yes) 1.72 (1.13–2.63) 0.012 Step 2 0.293  Gender (ref. Girls) 1.73 (1.37–2.19) < 0.001  Parental divorce/separation (ref. No) 1.61 (1.21–2.16) 0.001  Grade retention (ref. No) 15.5 (7.94–30.3) < 0.001  Frequency of screen use 1.20 (1.01–1.41) 0.033  Three meals a day (ref. Yes) 1.58 (1.02–2.46) 0.042  ADHD (ref. No) 3.50 (2.27–5.40) < 0.001  SLD (ref. No) 6.68 (4.60–9.70) < 0.001 Step 3 0.296  Gender (ref. Girls) 1.70 (1.33–2.17) < 0.001  Parental divorce/separation (ref. No) 1.60 (1.18–2.17) 0.002  Grade retention (ref. No) 15.5 (7.74–31.1) < 0.001  Frequency of screen use 1.19 (1.00–1.42) 0.045  Three meals a day (ref. Yes) NS NS  ADHD (ref. No) 3.60 (2.29–5.67) < 0.001  SLD (ref. No) 6.43 (4.35–9.51) < 0.001  SES (ref. High-income sample) NS NS Step 4 0.299  Gender (ref. Girls) 1.71 (1.34–2.18) < 0.001  Parental divorce/separation (ref. No) 1.60 (1.18–2.17) 0.002  Grade retention (ref. No) 14.9 (7.41–30.0) < 0.001  Frequency of screen use 1.19 (1.01–1.42) 0.044  Three meals a day (ref. Yes) NS NS  ADHD (ref. No) 2.58 (1.40–4.73) 0.002  SLD (ref. No) 5.56 (3.54–8.75) < 0.001  SES (ref. High-income sample) 1.36 (1.05–1.77) 0.020  ADHD × SES NS NS  SLD × SES NS NS All steps are incremental. First, the significant demographic predictors were entered in Step 1. ADHD and SLD (Step 2), and SES (Step 3) were then added to test main effects, while controlling for comorbidity. Last, we included interaction terms (i.e., ADHD × SES, SLD × SES) into the final model (Step 4) to examine the potential moderating role of SES. ADHD, attention-deficit/ hyperactivity disorder; CI, confidence interval; NS, non-significant; OR, odds ratio; SES, socioeconomic status; SLD, specific learning disorders. worse mental health than their counterparts, as evidenced by higher been hypothesized that children from more disadvantaged scores on most of the SDQ subscales (41). On the other hand, these communities might have less access to learning materials, cultural findings replicate the increased likelihood of learning disabilities resources and stimulating environments, which would limit their found in low SES communities (23, 40, 62, 63), since the rate of SLD reading skills and literacy experiences, particularly if parents lack the and reading difficulties was approximately two-fold higher within the tools to supplement their education (36, 37, 39, 63). Besides, the LM-income students (20.7% and 19.7%, respectively). Likewise, Fluss current research adds to a wealth of data documenting the positive et al. (39) reported that reading impairment was highly influenced by relation between SES and academic achievement (36–38). As neighborhood SES, with estimates ranging from 3.3% in high SES to we expected, students within the high-income sample had better 24.2% in low SES areas, and poor children have shown to be 1.5 times grades across all educational outcomes, and lower SES increased the more likely to have a learning disability (39, 64). In this sense, it has likelihood of poor academic performance. Frontiers in Psychiatry 10 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 TABLE 6 Predictors of poor academic performance on mathematics. Nagelkerke R2 OR (95% CI) Value of p Step 1 0.130  Age 1.12 (1.08–1.16) < 0.001  Parental divorce/separation (ref. No) 1.88 (1.43–2.47) < 0.001  Grade retention (ref. No) 7.35 (4.18–12.9) < 0.001  Three meals a day (ref. Yes) 1.51 (1.02–2.23) 0.038 Step 2 0.211  Age 1.11 (1.07–1.15) < 0.001  Parental divorce/separation (ref. No) 1.80 (1.40–2.32) < 0.001  Grade retention (ref. No) 5.96 (3.48–10.2) < 0.001  Three meals a day (ref. Yes) NS NS  ADHD (ref. No) 2.71 (1.88–3.92) < 0.001  SLD (ref. No) 3.47 (2.58–4.68) < 0.001 Step 3 0.231  Age 1.11 (1.07–1.16) < 0.001  Parental divorce/separation (ref. No) 1.78 (1.37–2.31) < 0.001  Grade retention (ref. No) 5.99 (3.45–10.4) < 0.001  Three meals a day (ref. Yes) NS NS  ADHD (ref. No) 2.97 (2.03–4.33) < 0.001  SLD (ref. No) 3.20 (2.35–4.35) < 0.001  SES (ref. High-income sample) 1.80 (1.46–2.22) < 0.001 Step 4 0.231  Age 1.11 (1.07–1.16) < 0.001  Parental divorce/separation (ref. No) 1.79 (1.37–2.32) < 0.001  Grade retention (ref. No) 5.96 (3.42–10.4) < 0.001  Three meals a day (ref. Yes) NS NS  ADHD (ref. No) 2.91 (1.72–4.91) < 0.001  SLD (ref. No) 3.02 (2.09–4.36) < 0.001  SES (ref. High-income sample) 1.85 (1.47–2.31) < 0.001  ADHD × SES NS NS  SLD × SES NS NS All steps are incremental. First, the significant demographic predictors were entered in Step 1. ADHD and SLD (Step 2), and SES (Step 3) were then added to test main effects, while controlling for comorbidity. Last, we included interaction terms (i.e., ADHD × SES, SLD × SES) into the final model (Step 4) to examine the potential moderating role of SES. ADHD, attention-deficit/ hyperactivity disorder; CI, confidence interval; NS, non-significant; OR, odds ratio; SES, socioeconomic status; SLD, specific learning disorders. Our findings also supported the hypothesis that ADHD would academic underachievement, especially for language subjects, increase the risk for academic impairment and are consistent with suggesting that learning disorders may compromise the academic previous literature, showing that youth with ADHD experience a performance of students by impeding or slowing their acquisition of variety of academic difficulties, including poor grades, low academic knowledge through reading (22). Thus, children with diagnosed achievement, special education need, grade retention and failure to learning disabilities have been found to earn lower GPA, require more complete high school, even after adjusting for intelligence, SES, and educational support, and have higher dropout rates than their peers learning disabilities (10, 11, 65). Interestingly, though, the present (22, 23, 66). study confirms for the first time the association between ADHD Overall, these results confirmed that ADHD, SLD and, to a lesser diagnosis and academic performance in Spain and supports a prior extent, low SES are significant predictors of poor performance across analysis conducted by Pagerols et al. (33), which found that students multiple academic domains, even after adjustment for comorbidity with increased levels of attention problems were more likely to and demographic variables that may influence educational functioning perform poorly, regardless of other risk factors such as comorbid (i.e., gender, age, parental divorce/separation, grade retention, psychopathology, sociodemographic characteristics, stressful events, frequency of screen use, dietary habits). However, studies that examine and lifestyle behaviors (33). Moreover, the current investigation the role of SES on academic achievement among children who suffer showed that SLD contributed independently to heighten the risk of from these ND are surprisingly sparse. Indeed, the current research Frontiers in Psychiatry 11 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 was the first to investigate whether high SES might attenuate the causality. Students who screened negative did not undergo the association between ADHD, SLD and adverse educational outcomes. clinical assessment and, therefore, false negatives might have Given the lack of significant interactions with SES and the strongest occurred. In addition, we cannot discount the possibility of a main effect observed for ND in all the academic domains assessed, our selection bias with regard to subjects included and excluded from findings indicate that SES does not provide the hypothesized buffer the analysis. Last, the SES measure used in our study may have against the negative impact of ADHD and SLD on academic misrepresented the individual SES of participants and resulted in performance. Similarly, a recent study, which evaluated whether the different patterns of associations (75, 76). Yet, area-level indicators association between a clinical diagnosis of depression and lower of SES provide useful information on contextual factors that are educational attainment was modified by SES, did not find significant relevant to health beyond individual or family-level characteristics moderating effects (67). (75, 76). For instance, there is evidence that neighborhood Considering that ADHD and SLD, unlike SES, are more easily conditions may affect a variety of outcomes, including educational modifiable risk factors, these results have large public health attainment and employment, through the availability and quality of implications as the ADHD- and SLD-related academic problems local services (e.g., child care centers, preschools, public schools, can lead to both short- and long-term consequences, including afterschool programs, health care facilities), physical and social lower productivity, unemployment, economic hardship, stress, environment stressors (e.g., air pollution, exposure to heavy metals, anxiety, depression, and low self-esteem (5, 68–70). Thus, early crime and violence levels, noise), and neighborhood-based networks identification and effective intervention strategies aimed at (36, 77). children with these ND might improve their educational functioning and mitigate later negative outcomes. In this sense, there is evidence for a positive short-term effect of ADHD Conclusions and future directions medications on some aspects of school performance (e.g., improvements in classroom behaviors, seatwork productivity, Overall, the present investigation adds to the evidence for GPA and achievement testing), although long-term studies are mental health disparities across socioeconomic backgrounds, and scarce and effects are generally smaller or more diverse (4, 71). expands previous research on the relation between ND and This suggests that additional supports both in the family and academic performance by analyzing children from LM-income and academic setting may be necessary to effectively remediate the high-income populations in order to determine whether SES academic difficulties faced by students with ADHD and moderates the association. Our results confirmed the higher SLD. Particularly, teachers may play an important role in the prevalence of psychiatric symptoms, learning difficulties, and ND educational development of affected children as evidenced by observed in disadvantaged communities. We also found that studies showing that school-based interventions enhance the ADHD, SLD and, to a lesser extent, low SES are significant academic achievement of children and adolescents with ADHD predictors of poor performance across multiple academic domains, mainly by optimizing their classroom behavior, self-regulation, even after adjustment for comorbidity and demographic variables organizational skills, and homework performance (72, 73). that may influence educational functioning (i.e., gender, age, The current research should be interpreted in the light of some parental divorce/separation, grade retention, frequency of screen strengths and limitations. One of the main strengths is that, for the use, dietary habits). However, SES did not modify the association, first time to our knowledge, we investigated the impact of ADHD suggesting that students with a clinical diagnosis of ADHD or SLD and SLD on children’s academic performance across specific are more likely to achieve lower grades on first language, foreign domains (i.e., first language, foreign language, mathematics) in language, and mathematics regardless of their socioeconomic Spain, and whether SES moderates the association between ND and background. Given the extensive consequences associated with educational outcomes after adjusting for comorbidity and academic problems, these findings highlight the need for early demographic variables that may influence educational functioning. detection and intervention strategies aimed at students who suffer Additional advantages include the large size and age range of the from ND in order to improve their functioning at school. In this sample, composed of 1,287 students aged 5–17 years, and the sense, integrating mental health services within the school setting comprehensive case identification through the administration of or in routine pediatric revisions could be helpful, especially for standardized screening instruments and DSM-based clinical children from low SES communities who might have limited access interviews by trained psychiatrics and neuropsychologists. to health care. Moreover, the identification of factors that may Moreover, our study involved the use of real-life measures of attenuate the risk for academic underachievement conferred by academic performance in a school-based sample, which allowed the ADHD and SLD is imperative to design and implement targeted detection of undiagnosed children and provides a more valid interventions for particularly vulnerable individuals. Therefore, reflection of the general population than a clinical sample. Finally, future research should explore other potential moderators, such as we controlled for multiple demographic risk factors and limited parental involvement, personal attributes and social networks, shared method variance by using different informants for clinical educational support, or treatment type and length. Alternatively, diagnoses and academic achievement (74). Nevertheless, alternative studies examining whether the relation between ND and academic explanations for the observed relations cannot be excluded since performance is modified by SES should include participants from other possible confounding variables, such as children’s intelligence more financially disadvantaged backgrounds, and combine quotient, executive functioning, treatment status or other individual measures of SES with data at the community level to psychiatric comorbidities, were not considered. Similarly, the cross- simultaneously estimate individual and contextual effects. Further, sectional design of the study prevents from drawing conclusions on a broader selection of covariates (e.g., intelligence, executive Frontiers in Psychiatry 12 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 functioning, psychiatric comorbidities, treatment status) is required Acknowledgments to fully validate the results of the current investigation. We are grateful to families, teachers, and staff of the public primary schools (i.e., Joan Maragall, Pins del Vallès, Ramon Llull, and Data availability statement Rivo Rubeo) and secondary schools (i.e., Angeleta Ferrer i Sensat and Duc de Montblanc), who kindly participated in this study, as well as The raw data supporting the conclusions of this article will to Departament d’Educació (Generalitat de Catalunya). be made available by the authors, without undue reservation. Conflict of interest Ethics statement GE-M has received travel grants from Angelini Pharma, The studies involving human participants were reviewed and Laboratorios Rubió, Lundbeck, and Takeda for participating in approved by the Ethics Committee of the Vall d’Hebron Hospital psychiatric meetings. JR-Q has served on the speakers’ bureau and Universitari, the Department of Health, and the Department of acted as consultant for Bial, Bristol Myers Squibb, Janssen-Cilag, Education (Generalitat de Catalunya, Spain). Written informed Laboratorios Raffo, Laboratorios Rubió, Medice, Novartis, Shionogi, consent to participate in this study was provided by the participants’ Shire, Sincrolab, Takeda, Tecnofarma, and Uriach. He has received legal guardian/next of kin. travel awards from Bial, Janssen-Cilag, Laboratorios Rubió, Medice, Shionogi, Shire, and Takeda for participating in psychiatric meetings. The Department of Psychiatry chaired by JR-Q has received Author contributions unrestricted educational and research support from Janssen-Cilag, Laboratorios Rubió, Oryzon, Psious, Roche, and Shire in the past GE-M, MP, and RB: study conception and design. GE-M, RP, and 3 years. MC has received travel grants and research support from Eli CR: data collection. GE-M, MP, and RP: data curation. GE-M and MP: Lilly and Co., Janssen-Cilag, and Shire. He has been on the advisory data analysis and interpretation of results. JR-Q, MC, and RB: study board and served as a consultant for Eli Lilly and Co., Janssen-Cilag, supervision. MC: funding acquisition. GE-M and MP: writing–draft and Shire. manuscript preparation. JR-Q, MC, and RB: writing–review and The remaining authors declare that the research was conducted in editing. All authors contributed to the article and approved the the absence of any commercial or financial relationships that could submitted version. be construed as a potential conflict of interest. Funding Publisher’s note This work was funded by Fundació “la Caixa,” Diputació de All claims expressed in this article are solely those of the authors Barcelona, and Departament de Salut (Generalitat de Catalunya). The and do not necessarily represent those of their affiliated organizations, funders were not involved in the study design, collection, analysis, or those of the publisher, the editors and the reviewers. Any product interpretation of data, the writing of this article or the decision to that may be evaluated in this article, or claim that may be made by its submit it for publication. manufacturer, is not guaranteed or endorsed by the publisher. References 1. American Psychiatric Association. Diagnostic and statistical manual of mental trajectories in middle childhood. J Clin Child Adolesc Psychol. (2020) 49:912–29. doi: disorders. 5th ed. Washington, DC: American Psychiatric Publishing. (2013). 10.1080/15374416.2019.1644645 2. Bosch R, Pagerols M, Rivas C, Sixto L, Bricollé L, Español-Martín G, et al. 8. Garner AA, O’connor BC, Narad ME, Tamm L, Simon J, Epstein JN. The Neurodevelopmental disorders among Spanish school-age children: prevalence and relationship between ADHD symptom dimensions, clinical correlates, and sociodemographic correlates. Psychol Med. (2021) 52:3062–72. doi: 10.1017/ functional impairment. J Dev Behav Pediatrics. (2013) 34:469–77. doi: 10.1097/ S0033291720005115 DBP.0b013e3182a39890 3. Sayal K, Prasad V, Daley D, Ford T, Coghill D. ADHD in children and young 9. Polderman TJ, Boomsma DI, Bartels M, Verhulst FC, Huizink AC. A systematic people: prevalence, care pathways, and service provision. Lancet Psychiatry. (2018) review of prospective studies on attention problems and academic achievement. Acta 5:175–86. doi: 10.1016/S2215-0366(17)30167-0 Psychiatr Scand. (2010) 122:271–84. doi: 10.1111/j.1600-0447.2010.01568.x 4. Keilow M, Holm A, Fallesen P. Medical treatment of attention deficit/hyperactivity 10. DuPaul GJ, Morgan PL, Farkas G, Hillemeier MM, Maczuga S. Eight-year latent disorder (ADHD) and children’s academic performance. PLoS One. (2018) 13:e0207905. class trajectories of academic and social functioning in children with attention-deficit/ doi: 10.1371/journal.pone.0207905 hyperactivity disorder. J Abnorm Child Psychol. (2018) 46:979–92. doi: 10.1007/ s10802-017-0344-z 5. Miranda A, Berenguer C, Colomer C, Roselló R. Influence of the symptoms of attention deficit hyperactivity disorder (ADHD) and comorbid disorders on 11. Fried R, Petty C, Faraone SV, Hyder LL, Day H, Biederman J. Is ADHD a risk factor functioning in adulthood. Psicothema. (2014) 26:471–6. doi: 10.7334/ for high school dropout? A controlled study. J Atten Disord. (2016) 20:383–9. doi: psicothema2014.121 10.1177/1087054712473180 6. Sayal K, Washbrook E, Propper C. Childhood behavior problems and academic 12. Barkley RA. Recent longitudinal studies of childhood attention-deficit/ outcomes in adolescence: longitudinal population-based study. J Am Acad Child Adolesc hyperactivity disorder: important themes and questions for further research. J Abnorm Psychiatry. (2015) 54:360–368.e2. doi: 10.1016/j.jaac.2015.02.007 Psychol. (2016) 125:248–55. doi: 10.1037/abn0000125 7. Evans SC, Cooley JL, Blossom JB, Pederson CA, Tampke EC, Fite PJ. Examining 13. Erskine HE, Norman RE, Ferrari AJ, Chan GC, Copeland WE, Whiteford HA, ODD/ADHD symptom dimensions as predictors of social, emotional, and academic et al. Long-term outcomes of attention-deficit/hyperactivity disorder and conduct Frontiers in Psychiatry 13 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 disorder: a systematic review and meta-analysis. J Am Acad Child Adolesc Psychiatry. 37. Tienken CH, Colella A, Angelillo C, Fox M, McCahill KR, Wolfe A. Predicting (2016) 55:841–50. doi: 10.1016/j.jaac.2016.06.016 middle level state standardized test results using family and community demographic data. RMLE Online. (2017) 40:1–13. doi: 10.1080/19404476.2016.1252304 14. Fleming M, Fitton CA, Steiner MFC, McLay JS, Clark D, King A, et al. Educational and health outcomes of children treated for attention-deficit/hyperactivity disorder. 38. Wang C, Fan X, Pugalee DK. Impacts of school racial composition on the JAMA Pediatr. (2017) 171:e170691. doi: 10.1001/jamapediatrics.2017.0691 mathematics and reading achievement gap in post unitary Carlotte-Mecklenburg schools. Educ Urban Soc. (2020) 52:1112–1132. doi: 10.1177/0013124519894970 15. May F, Ford T, Janssens A, Newlove-Delgado T, Emma Russell A, Salim J, et al. Attainment, attendance, and school difficulties in UK primary schoolchildren with 39. Fluss J, Ziegler JC, Warszawski J, Ducot B, Richard G, Billard C. Poor reading in probable ADHD. Br J Educ Psychol. (2021) 91:442–62. doi: 10.1111/bjep.12375 French elementary school: the interplay of cognitive, behavioral, and socioeconomic factors. J Dev Behav Pediatr. (2009) 30:206–16. doi: 10.1097/DBP.0b013e3181a7ed6c 16. Uchida M, Spencer TJ, Faraone SV, Biederman J. Adult outcome of ADHD: an overview of results from the MGH longitudinal family studies of pediatrically and 40. Billard C, Bricout L, Ducot B, Richard G, Ziegler J, Fluss J. Evolution of competence psychiatrically referred youth with and without ADHD of both sexes. J Atten Disord. in reading, spelling and comprehension levels in low socioeconomic environments and (2018) 22:523–34. doi: 10.1177/1087054715604360 impact of cognitive and behavioral factors on outcome in two years. Rev Épidemiol Sante Publique. (2010) 58:101–10. doi: 10.1016/j.respe.2009.11.002 17. Brimo K, Dinkler L, Gillberg C, Lichtenstein P, Lundström S, Åsberg JJ. The co- occurrence of neurodevelopmental problems in dyslexia. Dyslexia. (2021) 27:277–93. 41. Rajmil L, López-Aguilà S, Mompart Penina A, Medina Bustos A, Rodríguez Sanz doi: 10.1002/dys.1681 M, Brugulat GP. Socio-economic inequalities in children’s mental health in Catalonia. An Pediatr (Barc). (2010) 73:233–40. doi: 10.1016/j.anpedi.2010.02.022 18. Crisci G, Caviola S, Cardillo R, Mammarella IC. Executive functions in neurodevelopmental disorders: comorbidity overlaps between attention deficit and 42. Russell AE, Fort T, Williams R, Russell G. The association between socioeconomic hyperactivity disorder and specific learning disorders. Front Hum Neurosci. (2021) disadvantage and attention deficit/hyperactivity disorder (ADHD): a systematic review. 15:594234. doi: 10.3389/fnhum.2021.594234 Child Psychiatry Hum Dev. (2016) 47:440–58. doi: 10.1007/s10578-015-0578-3 19. Farias AC, Cordeiro ML, Felden EP, Bara TS, Benko CR, Coutinho D, et al. 43. Instituto Nacional de Estadística. Estadística del padrón continuo a 1 de enero de Attention-memory training yields behavioral and academic improvements in children 2015. Datos por municipio (n.d.). Available at: https://www.ine.es/jaxi/Tabla.htm?path=/ diagnosed with attention-deficit hyperactivity disorder comorbid with a learning t20/e245/p05/a2015/l0/&file=00008001.px&L=0 disorder. Neuropsychiatr Dis Treat. (2017) 13:1761–9. doi: 10.2147/NDT.S136663 44. Institut d’Estadística de Catalunya. Índex socioeconòmic territorial 2015-2018. 20. Molitor SJ, Langberg JM, Evans SW. The written expression abilities of adolescents Anàlisi de resultats (2021). Available at: https://www.idescat.cat/serveis/biblioteca/docs/ with attention-deficit/hyperactivity disorder. Res Dev Disabil. (2016) 51-52:49–59. doi: cat/ist20152018.pdf 10.1016/j.ridd.2016.01.005 45. Institut d’Estadística de Catalunya. Índex socioeconòmic territorial (IST). Índex 21. Sahoo MK, Biswas H, Padhy SK. Psychological co-morbidity in children with Catalunya=100 (2021). Available at: https://www.idescat.cat/pub/?id=ist&n=14034&by specific learning disorders. J Family Med Prim Care. (2015) 4:21–5. doi: =mun&t=201500%3AP 10.4103/2249-4863.152243 46. Goodman R. The strengths and difficulties questionnaire: a research note. J Child 22. Holopainen L, Hakkarainen A. Longitudinal effects of reading and/or Psychol Psychiatry. (1997) 38:581–6. doi: 10.1111/j.1469-7610.1997.tb01545.x mathematical difficulties: the role of special education in graduation from upper 47. Col·legi de Logopedes de Catalunya. PRODISCAT. Protocol de detecció i actuació secondary education. J Learn Disabil. (2019) 52:456–67. doi: 10.1177/0022219419865485 en la dislèxia. Àmbit educatiu. Barcelona: Departament d’Ensenyament de la Generalitat 23. Weis R, Erickson CP, Till CH. When average is not good enough: students with de Catalunya (2011). learning disabilities at selective, private colleges. J Learn Disabil. (2017) 50:684–700. doi: 48. Kaufman J, Birmaher B, Brent D, Rao U, Flynn C, Moreci P, et al. Schedule for 10.1177/0022219416646706 affective disorders and schizophrenia for school-age children-present and lifetime 24. Francis DA, Caruana N, Hudson JL, McArthur GM. The association between poor version (K-SADS-PL): initial reliability and validity data. J Am Acad Child Adolesc reading and internalising problems: a systematic review and meta-analysis. Clin Psychol Psychiatry. (1997) 36:980–8. doi: 10.1097/00004583-199707000-00021 Rev. (2019) 67:45–60. doi: 10.1016/j.cpr.2018.09.002 49. Wechsler D. Wechsler intelligence scale for children-fourth edition (WISC-IV). San 25. Lindén-Boström M, Persson C. Disparities in mental health among adolescents Antonio, TX: The Psychological Corporation (2003). with and without impairments. Scand J Public Health. (2015) 43:728–35. doi: 50. Wechsler D. Wechsler intelligence scale for children-fifth edition (WISC-V). San 10.1177/1403494815589219 Antonio, TX: NCS Pearson (2014). 26. McArthur G, Castles A. Helping children with reading difficulties: some things 51. Cuetos F, Rodríguez B, Ruano E, Arribas D. PROLEC-R: Batería de evaluación de we have learned so far. NPJ Sci Learn. (2017) 2:7. doi: 10.1038/s41539-017-0008-3 los procesos lectores. Revisada. Madrid: TEA Ediciones (2007). 27. Turker S, Seither-Preisler A, Reiterer SM, Schneider P. Cognitive and behavioural 52. Cuetos F, Arribas D, Ramos JL. PROLEC-SE-R. Batería de evaluación de los weaknesses in children with reading disorder and AD(H)D. Sci Rep. (2019) 9:15185. doi: procesos lectores en secundaria y bachillerato – revisada. Madrid: TEA Ediciones (2016). 10.1038/s41598-019-51372-w 53. Toro J, Cervera M. Test de análisis de la lecto-escritura. Madrid: Aprendizaje-Visor (1991). 28. Sahu A, Patil V, Sagar R, Bhargava R. Psychiatric comorbidities in children with specific learning disorder-mixed type: a cross-sectional study. J Neurosci Rural Pract. 54. Cuetos F, Ramos JL, Ruano E. PROESC. Evaluación de los procesos de escritura. (2019) 10:617–22. doi: 10.1055/s-0039-1697879 Madrid: TEA Ediciones. (2002). 29. Scott MN, Hunter SJ, Joseph RM, O’Shea TM, Hooper SR, Allred EN, et al. 55. Español-Martín G, Pagerols M, Prat R, Rivas C, Sixto L, Valero S, et al. Strengths Neurocognitive correlates of attention-deficit hyperactivity disorder symptoms in and difficulties questionnaire: psychometric properties and normative data for Spanish children born at extremely low gestational age. J Dev Behav Pediatr. (2017) 38:249–59. 5- to 17-year-olds. Assessment. (2021) 28:1445–58. doi: 10.1177/1073191120918929 doi: 10.1097/DBP.0000000000000436 56. Prat R, Pagerols M, Español-Martín G, Rivas C, Sala M, Casas M, et al. 30. Burrows T, Goldman S, Pursey K, Lim R. Is there an association between dietary PRODISCAT-R: new approach for detecting reading disorders among children and intake and academic achievement: a systematic review. J Hum Nutr Diet. (2017) adolescents. Dyslexia. (2023) Under review 30:117–40. doi: 10.1111/jhn.12407 57. de la Peña FR, Villavicencio LR, Palacio JD, Félix FJ, Larraguibel M, Viola L, et al. 31. Faught EL, Ekwaru JP, Gleddie D, Storey KE, Asbridge M, Veugelers PJ. The Validity and reliability of the kiddie schedule for affective disorders and schizophrenia combined impact of diet, physical activity, sleep and screen time on academic present and lifetime version DSM-5 (K-SADS-PL-5) Spanish version. BMC Psychiatry. achievement: a prospective study of elementary school students in Nova Scotia, Canada. (2018) 18:193. doi: 10.1186/s1288-018-1773-0 Int J Behav Nutr Phys Act. (2017) 14:29. doi: 10.1186/s12966-017-0476-0 58. Ulloa RE, Ortiz S, Higuera F, Nogales I, Fresán A, Apiquian R, et al. Interrater reliability of the Spanish version of the schedule for affective disorders and schizophrenia 32. Hughes JN, Cao Q, West SG, Allee Smith P, Cerda C. Effect of retention in for school-age children-present and lifetime version (K-SADS-PL). Actas Esp Psiquiatr. elementary grades on dropping out of school early. J Sch Psychol. (2017) 65:11–27. doi: (2006) 18:36–40. doi: 10.1186/s12888-018-1773-0 10.1016/j.jsp.2017.06.003 59. George D, Mallery M. SPSS for windows step by step: A simple guide and reference, 33. Pagerols M, Prat R, Rivas C, Español-Martín G, Puigbó J, Pagespetit È, et al. The 17.0 update. 10th ed. Boston, MA: Pearson (2010). impact of psychopathology on academic performance in school-age children and adolescents. Sci Rep. (2022) 12:4291. doi: 10.1038/s41598-022-08242-9 60. Hair JF, Hult GTM, Ringle CM, Sarstedt M. A primer on partial least squares structural equation modeling (PLS-SEM). 3rd ed. London: Sage (2022). 34. Pedersen ML, Holen S, Lydersen S, Martinsen K, Neumer SP, Adolfsen F, et al. School functioning and internalizing problems in young schoolchildren. BCM Psychol. 61. Liu Y, Liu Z, Liang R, Luo Y. The association between community-level (2019) 7:88. doi: 10.1186/s40359-019-0365-1 socioeconomic status and depressive symptoms among middle-aged and older adults in China. BMC Psychiatry. (2022) 22:297. doi: 10.1186/s12888-022-03937-9 35. Yamada M, Sekine M, Tatsuse T, Asaka Y. Association between lifestyle, parental smoke, socioeconomic status, and academic performance in Japanese elementary school 62. van Zwieten A, Teixeira-Pinto A, Lah S, Nassar N, Craig JC, Wong G. children: the super diet education project. Environ Health Prev Med. (2019) 24:22. doi: Socioeconomic status during childhood and academic achievement in secondary school. 10.1186/s12199-019-0776-x Acad Pediatr. (2021) 21:838–48. doi: 10.1016/j.acap.2020.10.013 36. Devenish B, Hooley M, Mellor D. The pathways between socioeconomic status and 63. Taanila A, Yliherva A, Kaakinen M, Moilanen I, Ebeling H. An epidemiological adolescent outcomes: a systematic review. Am J Community Psychol. (2017) 59:219–38. study on finish school-aged children with learning difficulties and behavioral problems. doi: 10.1002/ajcp.12115 Int J Circumpolar Health. (2011) 70:59–71. doi: 10.3402/ijch.v70i1.17799 Frontiers in Psychiatry 14 frontiersin.org Español-Martín et al. 10.3389/fpsyt.2023.1136994 64. Duncan GJ, Brooks-Gunn J. Poverty, welfare reform, and children’s achievement 71. Baweja R, Mattison RE, Waxmonsky JG. Impact of attention-deficit hyperactivity In: B Biddle, editor. Social class, poverty, and education. New York, NY: Routledge Almer disorder on school performance: what are the effects of medication? Pediatr Drugs. (2001). 49–76. (2015) 17:459–77. doi: 10.1007/s40272-015-0144-2 65. Colomer C, Berenguer C, Roselló B, Baixauli I, Miranda A. The impact of 72. DuPaul GJ, Evans SW, Owens JS, Cleminshaw CL, Kipperman K, Fu Q, et al. School- inattention, hyperactivity/impulsivity symptoms, and executive functions on learning based intervention for adolescents with attention-deficit/hyperactivity disorder: effects on behaviors of children with ADHD. Front Psychol. (2017) 8:540. doi: 10.3389/ academic functioning. J Sch Psychol. (2021) 87:48–63. doi: 10.1016/j.jsp.2021.07.001 fpsyg.2017.00540 73. DuPaul GJ, Gormley MJ, Daffner-Deming M. School-based interventions for 66. Bergey BW, Deacon SH, Parrila RK. Metacognitive reading and study strategies elementary school students with attention-deficit/hyperactivity disorder. Child Adolesc and academic achievement of university students with and without a history of reading Psychiatr Clin N Am. (2022) 31:149–66. doi: 10.1016/j.chc.2021.08.003 difficulties. J Learn Disabil. (2017) 50:81–94. doi: 10.1177/0022219415597020 74. Ringoot AP, Tiemeier H, Jaddoe VW, So P, Hofman A, Verhulst FC, et al. Parental 67. Wickersham A, Fort T, Stewart R, Downs J. Estimating the impact of child and depression and child well-being: young children’s self-reports helped addressing biases early adolescent depression on subsequent attainment; secondary analysis of an existing in parent reports. J Clin Epidemiol. (2015) 68:928–38. doi: 10.1016/j.jclinepi.2015.03.009 data linkage. Epidemiol Psychiatr Sci. (2021) 30:e76. doi: 10.1017/S2045796021000603 75. Moss JL, Johnson NJ, Yu M, Altekruse SF, Cronin KA. Comparisons of individual- 68. Condo JS, Chan ESM, Kofler MJ. Examining the effects of ADHD symptoms and and area-level socioeconomic status proxies for individual-level measures: evidence parental involvement on children’s academic achievement. Res Dev Disabil. (2022) from the mortality disparities in American communities study. Popul Health Metrics. 122:104156. doi: 10.1016/j.ridd.2021.104156 (2021) 19:1. doi: 10.1186/s12963-020-00244-x 69. Döpfner M, Mandler J, Breuer D, Schürmann S, Dose C, Walter D, et al. 76. Pardo-Crespo MR, Narla NP, Williams AR, Beebe TJ, Sloan J, Yawn BP, et al. Children with attention-deficit/hyperactivity disorder grown up: an 18-year follow- Comparison of individual-level versus area-level socioeconomic measures in assessing up after multimodal treatment. J Atten Disord. (2021) 25:1801–17. doi: health outcomes of children in Olmsted County. Minnesota J Epidemiol Community 10.1177/1087054720948133 Health. (2013) 67:305–10. doi: 10.1136/jech-2012-201742 70. Pelham WE, Page TF, Altszuler AR, Gnagy EM, Molina BSG, Pelham WE. The 77. Ellen IG, Mijanovich T, Dillman KN. Neighborhood effects on health: exploring long-term financial outcome of children diagnosed with ADHD. J Consult Clin Psychol. the links and assessing the evidence. J Urban Aff. (2001) 23:391–408. doi: (2020) 88:160–71. doi: 10.1037/ccp0000461 10.1111/0735-2166.00096 Frontiers in Psychiatry 15 frontiersin.org