Developmental brain trajectories in children with ADHD and controls: a longitudinal neuroimaging study

As individuals enter adolescence, the presentation of some ADHD symptoms and neuropsychological functioning appears to change [9‐12]. Inattentive symptoms remain relatively constant from ages 9–12 years, whereas hyperactive/impulsive symptoms decline although they do not normalize [13, 14]. The transition to adolescence is an important developmental shift with major environmental and biological changes potentially exerting influence on functional status. This includes the transition to high school which has been associated with an interruption in the decline in ADHD symptomotology [9]. It also corresponds with the major physiological and emotional effects of puberty.

Numerous cross-sectional studies have examined brain differences between individuals with and without ADHD. Functional imaging studies have highlighted several abnormalities in individuals with ADHD, particularly in the prefrontal cortex and striatum (fronto-striatal circuits) and the parietal cortex [15‐19]. Structural imaging studies have reported ADHD-related anomalies in the prefrontal cortex, cerebellum, striatum and basal ganglia, corpus callosum, and the parietal cortex [20‐27]. There is debate as to whether these differences represent specific brain abnormalities characteristic of ADHD, or a delay of normal development (i.e., a maturation lag) [28]. However, there have been marked inconsistencies in previous studies, attributable to the use of small, homogenous clinical samples with considerable between-study variation in subtypes, gender and age. Generalizability to the larger/wider population of children with ADHD is therefore limited.

While ADHD symptoms vary considerably with age and neuroimaging abnormalities have been described in children with ADHD, it is unclear whether the changes in brain structure and function parallel symptom changes. It is also unknown whether deviations from typical trajectories of brain development are associated with differential outcomes, such as the persistence or remission of ADHD symptoms and academic, cognitive, social and mental health outcomes. A recent meta-analysis [29], identified the need for longitudinal designs and larger samples to advance the field. For our understanding of the neural underpinnings of ADHD to progress and generate knowledge that will inform treatments, there is a need to establish working models of neurodevelopment. This can best be achieved by linking serial measures of brain development, using multiple state-of-the-art neuroimaging methods, with detailed phenotypic and functional outcomes indices.

NICAP will collect single site, multimodal neuroimaging on a high-resolution 3 Tesla scanner to link neurobiological structure and function to academic, cognitive, social, and mental health outcomes. We will assess children with and without ADHD as they progress through puberty at mean ages 10, 11.5 and 13 years. This design will enable us to map trajectories in brain growth and how they differ between typically developing children and those with ADHD. We will be able to ascertain whether such changes are reflected in ADHD symptomatology and functional abilities.

The primary aims are to 1) Describe how brain structure (whole-brain volume, grey-matter volume, white-matter volume, cortical thickness, diffusion indices) and function (resting state connectivity) change across late childhood to early adolescence (brain growth trajectories) for children with and without ADHD; and 2) Examine whether differences in trajectories of brain structure and function reflect differential outcomes for children with ADHD and non-ADHD controls. Outcomes to be assessed include the persistence of ADHD, ADHD symptom severity and functional outcomes (academic, cognitive, social, and mental health).

Secondary aims will explore the impact of sex, ADHD presentation, ADHD severity, comorbidities and medication use on brain growth trajectories.

Our sample size is primarily based on the feasibility of recruitment from the pre-existing CAP cohort. Based on recruitment of equal numbers of controls to ADHD cases, with a participation rate of 65 %, accounting for exclusions due to MRI incompatibility (e.g., dental braces), and an estimated attrition of 5 % at each timepoint, we expect ~100, 95 and 90 participants per group at timepoints 1, 2 and 3 respectively. It is difficult to estimate with precision the power this will provide for detecting group differences in trajectories of development. However, a study of the sample size required in longitudinal MRI studies of brain volume in adults [31] suggests 80 % power detection of a 5 % difference between two groups for change in even small subcortical structures (e.g., the caudate) from a sample size of 90–104 per group.

At each data collection time-point participating families will attend a 3.5 h assessment session at The Royal Children’s Hospital, Melbourne, Australia. Assessment sessions involve a structured diagnostic interview, parent questionnaire, child cognitive assessment and MRI scanning. Saliva samples will also be collected for future research questions around genetics and pubertal hormones. With parent consent, questionnaires will be sent to the child’s classroom teacher.

Measures are summarized in Table 1. Children will be assessed in their usual classroom condition, therefore, if the child is currently using medication, they are not asked to cease medication for the assessment, but details of medication history and dosage are recorded. Research staff conducting assessments will be blinded to the child’s diagnostic status.

NIMH Diagnostic Interview Schedule for Children-IV [32]: At baseline (10 years) and 36 months (13 years), parents complete the well-validated and widely used DISC-IV diagnostic interview (60–90 mins) to determine the participants’ ADHD status and comorbid mental health problems including anxiety, mood and externalizing disorders.

Questionnaires assess several domains covering a range of predictors and outcome variables. Key measures are described below and summarized in Table 1. Parent and teachers complete questionnaires pertaining to the child’s ADHD symptom severity and their social and emotional functioning. Parents also complete a series of questionnaires about the child’s functioning including emotional, physical, social and school quality of life, the child’s peer victimization, a screening measure for autism spectrum disorder symptoms, the child’s general health, the use of allied health services and medication history. Questionnaires concerning the home environment include a measure of family quality of life, stressful life events, and parents’ mental health. Numerous scales are drawn from the Longitudinal Study of Australian Children (LSAC; [33]) with items assessing parenting and the parent couple relationship. Retrospective questions regarding potentially relevant pre- and post-natal factors are also assessed, such as maternal alcohol use and smoking during pregnancy, gestational diabetes, preclampsia, stress/anxiety/depression/stressful life events during pregnancy, birth weight, gestational age, intensive care following birth and maternal postnatal depression. Teachers are asked questions around the child’s academic competence, the student-teacher relationship, as well as details on the teacher characteristics and education services. Those not described below have been described in detail elsewhere [8].

The Pubertal Development Scale (PDS) [34] is a parent-reported measure assessing development on five indices of pubertal growth. Parents are asked about whether the child’s secondary sexual characteristics have 1) not yet started, 2) barely started, 3) definitely started, 4) seems complete. Parents of males are asked about changes to voice and growth of facial hair and parents of females are asked about breast development and about the onset and age of menstruation.

The Tanner Sexual Maturation Scale (SMS) is a parent-reported measure used to assess the child’s pubertal stage. It comprises drawings of five progressive stages of pubertal development of secondary sexual characteristics, from stage 1 (pre-adolescent) through to stage 5 (adult appearance). For males, five drawings combining pubic hair and genital development are presented. For females, breast development and pubic hair development are presented in different drawings. Tanner staging has historically been considered the gold standard for puberty measurement [35].

NICAP includes direct assessments of working memory, inhibition, sustained attention, cognitive flexibility, spatial attention and visuo-motor integration. The first four tasks are computer-based, enabling measurement of properties such as reaction time.

The Stop Signal Task assesses response inhibition [36]. Subjects perform a choice reaction task and on a random selection of the trials, an auditory stop signal instructs subjects to withhold their response.

The Sustained Attention to Response Task (SART) is a measure of sustained attention [37]. The fixed version of SART is a repeating sequence of digits (1–9). Using a button press, participants respond to every digit (go-trial) except ‘3’ (no-go trial).

The Spatial N-Back is a widely used measure of working memory that requires flexible updating capabilities. This includes a spatial 1-back and 2-back version. The 1-back requires maintaining and updating one location at a time, whereas the more difficult 2-back requires maintaining and updating two locations.

The Set Shifting task assesses cognitive flexibility. Two target pictures are presented that vary along two dimensions (e.g., shape and color). Participants are cued with a letter to respond to the target pictures, according to one dimension.

The Landmark Task is a paper-based task measuring spatial attention [38]. Participants are presented with 20 examples of a bisected line, half are bisected exactly in the middle, while the remainder are bisected slightly offset to the left or right. Participants indicate which side of the line is shorter. Leftward or rightward spatial biases can be ascertained.

The Grooved Pegboard test (Lafayette Instruments, Lafayette, IN) is a timed motor test to assess complex visual-motor coordination for both the dominant and non-dominant hand. Participants place grooved pegs into a pegboard unit in an ordered pattern of 25 holes, requiring the participant to match the groove of the peg with the groove of the board.

Quality control procedures are important at a number of steps. Regarding MRI motion artifacts, several steps are taken in order to minimize movement during the scanning and to assess the data quality afterwards. The mock scanner session prior to the scan is vital to assist participants be aware of movements and become comfortable in the scanner environment. During all scans except for the rs-fMRI participants watch a movie of their choice distracting them from the scanning environment. During scanning, movement is monitored and participants are reminded to keep still when necessary. Poor images due to motion are repeated if time permits. At the scanner level, scanner stability is monitored weekly with the standard functional Brain Imaging Research Network (FBIRN) QA protocol. In addition, T1 and T2 are performed to examine signal-to-noise and image uniformity.

Research staff and students who conduct the assessments are trained to a high level of competence in the scanning and assessment procedures, and are observed for the first two assessments. Written standard operation procedures were developed for standardized assessments, cognitive testing, mock scan and saliva collection. Fortnightly supervision meetings take place with a registered clinical psychologist (ES) in order to maintain consistency across cognitive and diagnostic assessments.