Examining the comorbidity between learning (dis)abilities and attention deficit hyperactivity disorder

Examining the comorbidity between learning (dis)abilities and attention deficit hyperactivity disorder

First Author: Callie Little -- University of New England
Keywords: Comorbidity, Learning disability or difficulty, Literacy, Mathematics
Abstract / Summary: 

Research has demonstrated that learning (dis)abilities and Attention-Deficit Hyperactivity Disorder co-occur at higher rates than expected by chance. Children with an identified learning problem are more likely to exhibit higher rates of another learning disability or behavior problems like ADHD than their typically developing peers. However, our understanding of how these skills and behaviors co-influence each other both longitudinally and across the continuum of ability is limited, as well as our knowledge of which etiological influences are general across these domains versus those that are domain-specific. These research questions can be appropriately examined by implementing a range of methods across multiple samples and measures, which this symposium addresses through a combination of advanced statistical approaches implemented across several labs. Results from latent growth curve modeling, dual change score modeling, bi-factor models, and genetically-sensitive quantile regression methods will be presented and discussed to examine the comorbidity between learning dis(abilities) and ADHD.

Symposium Papers: 

The relationship between ADHD and reading comprehension throughout the achievement spectrum.

First Author/Chair:Jeffrey Shero -- Florida State University
Additional authors/chairs: 
Sara Hart; Jessica Logan

Purpose - The negative impact of Attention Deficit Hyperactivity Disorder (ADHD) on reading achievement is well established. This study aimed to deepen the understanding of this relationship, by examining how this impact varies at different levels of achievement.
Methods - Participants were 437 twin pairs (mean age = 11 years), drawn from the Florida Twin Project on Reading. Using quantile regression, the impact of ADHD related inattention and hyperactivity behaviors, as measured by the SWAN inattention and hyperactivity subscales, on reading comprehension, measured by the Florida Comprehensive Assessment Test, was assessed throughout the distribution of achievement.
Results - Showed that the correlation between ADHD behaviors and reading comprehension was highest at lower quantiles of achievement and gradually decreased when moving higher in the achievement distribution. For both inattention and hyperactivity subscales the correlation between ADHD behaviors and reading comprehension was over double the magnitude at the lowest quantiles (10th-20th)of achievement (approximately 0.6 and 0.4 SD’s respectively) compared to the highest quantiles (80th-90th) (approximately 0.2 and 0.15 SD’s respectively). This study then extended its examination into a behavioral genetics framework. A new model was developed to test if the genetic correlation between ADHD behaviors and reading comprehension followed a similar non-linear pattern as the phenotypic relationship. Initial evidence suggested that the genetic correlation is indeed higher at the lower levels of reading achievement.
Conclusion – This study suggests that the association of ADHD behaviors to reading comprehension is far larger for low-achieving readers compared to their higher achieving counterparts.

How specific are learning disabilities?

First Author/Chair:Lauren McGrath -- University of Denver
Additional authors/chairs: 
Erik Willcutt; Janice Keenan; Richard Olson; Bruce Pennington; Robin Peterson

Purpose: Academic skills are strongly correlated across the curriculum leading to high rates of comorbidity between “specific” learning disabilities (SLDs) in reading, writing, and mathematics. The purpose of this study was to quantify the evidence for generality vs. specificity of difficulties across academic domains. In other words, how specific are SLDs?

Method: Participants included 686 children aged 8 to 16 who were over-selected for learning disabilities (28%). Children completed 18 measures of basic and complex skills in reading, mathematics, and writing.

Results: We used bifactor structural equation models to identify variance shared across academic domains (academic g) as well as variance specific to reading, mathematics, and writing. In the bifactor model, all measures loaded strongly on academic g. Basic reading and mathematics maintained variance distinct from academic g, consistent with the notion of SLDs in these domains. Evidence for reading comprehension-specific variance was mixed, and writing did not maintain specific variance apart from academic g.

Conclusions: The strong effect of academic g across academic domains highlights the pervasiveness of children who struggle across the curriculum and are difficult to classify in current diagnostic systems because their academic difficulties are general and not specific. In addition to these generalized effects, there was also evidence for specificity of learning disabilities in basic reading and math, but no evidence for specific writing difficulties. This is not to say that children do not experience writing difficulties, but that these difficulties rarely occur in isolation from other academic challenges

Exploring reading and numeracy development using latent growth curve modeling

First Author/Chair:Sally Larsen -- University of New England
Additional authors/chairs: 
Callie Little; William Coventry

Purpose. This study investigated trajectories of reading and numeracy skills in Australian students in Grades 3, 5, 7 and 9, and examined whether attention behaviors predicted growth in either of these skills. Method. Two longitudinal cohorts of students (aged 8-16) from the NAPLAN Twin Study were selected for latent growth curve analyses (N=378). Reading and numeracy were measured by scores on the National Assessment Program: Literacy and Numeracy (NAPLAN). Attention behaviors at Grade 3 assessed by the Strengths and Weaknesses of ADHD-Symptom and Normal-Behavior Scale (SWAN) were used to predict intercept and slope factors. Results demonstrated significant individual differences in mean reading and numeracy achievement at Grade 3, and significant individual variation in the growth trajectories of both skills between Grades 3 and 9. Negative correlations between intercept and growth factors in reading (-.70) and numeracy (-.41) suggested compensatory growth patterns in both skills. Attention (.43-.44) and SES (.27-.30) predicted achievement in both skills at Grade 3, and negatively predicted growth in reading over time (-.28), but not numeracy. Conclusions. Compensatory growth trajectories indicate that students who achieved poorly in Grade 3 developed more quickly in their reading and numeracy skills than those who started with higher scores. The negative relation between attention and growth in reading skills indicates an additional compensatory growth pattern whereby students with poorer attention in Grade 3 have a steeper growth trajectory in reading than those with better attention.

Exploring the co-development of attention deficit hyperactivity disorder and language ability

First Author/Chair:Callie Little -- University of New England
Additional authors/chairs: 
Elsje van Bergen; Eveline de Zeeuw

Purpose. The development of language ability and Attention Deficit Hyperactivity Disorder (ADHD) has been found to be highly related, suggesting that these skills develop concurrently. The present study explores the co-development of language ability and ADHD across middle childhood (ages 7-12) using a dual change score model (DCSM). The DCSM allows for the examination of the extent to which change in one developmental domain influences subsequent changes in another related domain. Method. Teacher ratings on ADHD assessed with the Conner’s Rating Scales - Revised and teacher ratings of language ability were measured at 3 waves (ages 7, 10, and 12) in 13,803 twin pairs from the Netherlands Twin Register. One member of each twin pair was selected for these analyses. Results. Estimates of latent growth on both constructs were small and positive, with a greater magnitude for language ability (.24) over ADHD (.16). Proportional change for both measures was negative and small with a slightly smaller magnitude indicated for ADHD (.14) versus language (.18). Additionally, cross-lagged estimates indicated a negative and moderate influence of initial language ability on subsequent change in ADHD behaviors (-.33) and a small influence from initial ADHD on subsequent change in language ability (-.14). Discussion. While language ability is a leading indicator of change in ADHD, ADHD also appears to have some influence on change in language ability. This suggests a bi-directional influence between the two, potentially from shared genetic and environmental influences between them.

Discussant - Examining the comorbidity between learning (dis)abilities and attention deficit hyperactivity disorder

First Author/Chair:DISCUSSANT Erik Willcutt