Braincharting autism and ADHD reveals age- and sex- specific effects

396 

Abstract Submission 

Saashi Bedford¹, Meng-Chuan Lai², Michael Lombardo³, Bhismadev Chakrabarti⁴, Amber Ruigrok⁵, John Suckling⁶, Evdokia Anagnostou⁷, Jason Lerch⁸, Margot Taylor⁹, Rob Nicolson¹⁰, Georgiades Stelios¹¹, Jennifer Crosbie¹², Russell Schachar⁹, Elizabeth Kelley¹³, Jessica Jones¹³, Paul Arnold¹⁴, Eric Courchesne¹⁵, Karen Pierce¹⁶, Lisa Eyler¹⁷, Kathleen Campbell¹⁶, Cynthia Carter Barnes¹⁶, Jakob Seidlitz¹⁸, Aaron Alexander-Bloch¹⁸, Edward Bullmore⁶, Simon Baron-Cohen⁶, Richard Bethlehem¹⁹

¹University of Cambridge, Cambridge, Select State/Province, ²Centre for Addiction and Mental Health, Toronto, Ontario, ³Laboratory for Autism and Neurodevelopmental Disorders, Istituto Italiano di Tecnologia, Rovereto, Italy, ⁴University of Reading, Reading, Berkshire, ⁵University of Manchester, Manchester, Lancashire, ⁶University of Cambridge, Cambridge, United Kingdom, ⁷University of Toronto, Toronto, Ontario, ⁸Wellcome Centre for Integrative Neuroimaging, University of Oxford, Oxford, Oxford, ⁹Hospital for Sick Children, Toronto, Ontario, ¹⁰University of Western Ontario, London, Ontario, ¹¹McMaster University, Hamilton, Ontario, ¹²The Hospital for Sick Children, Toronto, Ontario, ¹³Queen’s University, Kingston, Kingston, Ontario, ¹⁴University of Calgary, Calgary, Alberta, ¹⁵University of California San Diego, San Diego, CA, ¹⁶University of California San Diego, San Diego, California, ¹⁷University of California San Diego, La Jolla, CA, ¹⁸University of Pennsylvania, Philadelphia, PA, ¹⁹Autism Research Centre, Department of Psychiatry, University of Cambridge, Cambridge, United Kingdom

Saashi Bedford  
University of Cambridge
Cambridge, Select State/Province

Meng-Chuan Lai  
Centre for Addiction and Mental Health
Toronto, Ontario

Michael Lombardo  
Laboratory for Autism and Neurodevelopmental Disorders, Istituto Italiano di Tecnologia
Rovereto, Italy

Bhismadev Chakrabarti  
University of Reading
Reading, Berkshire

Amber Ruigrok  
University of Manchester
Manchester, Lancashire

John Suckling  
University of Cambridge
Cambridge, United Kingdom

Evdokia Anagnostou  
University of Toronto
Toronto, Ontario

Jason Lerch, PhD  
Wellcome Centre for Integrative Neuroimaging, University of Oxford
Oxford, Oxford

Margot Taylor  
Hospital for Sick Children
Toronto, Ontario

Rob Nicolson  
University of Western Ontario
London, Ontario

Georgiades Stelios  
McMaster University
Hamilton, Ontario

Jennifer Crosbie  
The Hospital for Sick Children
Toronto, Ontario

Russell Schachar  
Hospital for Sick Children
Toronto, Ontario

Elizabeth Kelley  
Queen’s University, Kingston
Kingston, Ontario

Jessica Jones  
Queen’s University, Kingston
Kingston, Ontario

Paul Arnold  
University of Calgary
Calgary, Alberta

Eric Courchesne  
University of California San Diego
San Diego, CA

Karen Pierce  
University of California San Diego
San Diego, California

Lisa Eyler, PhD  
University of California San Diego
La Jolla, CA

Kathleen Campbell  
University of California San Diego
San Diego, California

Cynthia Carter Barnes  
University of California San Diego
San Diego, California

Jakob Seidlitz  
University of Pennsylvania
Philadelphia, PA

Aaron Alexander-Bloch  
University of Pennsylvania
Philadelphia, PA

Edward Bullmore  
University of Cambridge
Cambridge, United Kingdom

Simon Baron-Cohen  
University of Cambridge
Cambridge, United Kingdom

Richard Bethlehem  
Autism Research Centre, Department of Psychiatry, University of Cambridge
Cambridge, United Kingdom

Autism and attention-deficit/hyperactivity disorder (ADHD) are heterogeneous neurodevelopmental conditions with complex underlying neurobiology, and neuroanatomical alterations have been reported in both [1–3]. Both conditions show significant sex and age modulations on neuroanatomy [3,4], which are not yet fully understood. Normative modelling is an emerging technique that provides a unified framework for studying age- and sex-specific divergence in brain development in a common space [5]. We aimed to characterise regional cortical and global neuroanatomy in autism and ADHD, as well as sex and age differences, benchmarked against models of typical brain development based on a sample of over 75,000 individuals.

We combined T1-weighted MRIs from 49 sites across 7 datasets, for a total dataset of 4255 participants after quality control (1869 controls, 987 ADHD, 1399 autism; ages 2-64 years). Images were processed using FreeSurfer [6] 6.0.1, and regional cortical estimates extracted based on the Desikan-Killiany [7] atlas, for cortical thickness (CT), volume (CV) and surface area (SA). We also examined total grey and white matter volume, subcortical grey matter volume, ventricular volume, mean CT and total SA.

We used normative reference models [8] previously generated by our group to map neuroanatomical developmental trajectories across the lifespan for global and regional neuroanatomical measures, accounting for age, sex and site/scanner. Out-of-sample normative centile scores for our study sample were generated based on these models, quantifying divergence from normative brain development in our sample. We examined diagnostic group differences in centile scores, and the interaction between diagnosis and sex, for all global volumes and regional cortical measures using multiple linear regressions.

To examine age differences, we conducted a sliding-window age analysis to examine diagnostic group differences across development, in age intervals of 5 years, starting at 2 years and sliding by 1 year. We selected a random subset of 70 participants per group, bootstrapped 1000 times, and examined diagnostic group differences at each age interval, averaged across the 1000 bootstraps. This allowed us to determine change in group differences across development, and the age window with the maximum diagnostic group difference.

All results were significant at 5% false discovery rate. For global measures, individuals with ADHD had significantly lower total cortical and subcortical grey and white matter volumes, and total cortical SA centile scores (d = -0.13- -0.18), but greater mean CT centiles (d = 0.09) relative to controls. Autistic individuals had significantly larger ventricular volume centiles relative to controls (d = 0.15). Individuals with ADHD showed regional CT increases (d = 0.09-0.10) but lower CV and SA centiles (d = -0.07- -0.18) across much of the cortex. Autistic individuals showed greater regional CT and CV localised to the superior temporal cortex (STG; d = 0.13-0.15). There was a significant sex-by-diagnosis interaction, and distinct diagnostic differences by sex for autism, but not for ADHD, where autistic males had greater CV and CT in the STG, but autistic females had lower SA in the fusiform gyrus, relative to controls.

For autism, diagnostic group differences for the STG were strongest in the youngest age windows, across all cortical measures. For SA, and to a slightly less extent CV, this was also the case in most frontal and parietal regions. For CT, most other regions had the strongest effects in later ages, around 15-20 years. Group differences in ADHD were typically strongest at the latest age windows, starting at 14-15 years of age. Some regions showed the strongest differences in the youngest windows, including the STG for SA, and some frontal and parietal areas.

These results indicate distinct cortical differences in autism and ADHD that are differentially impacted by age and sex.

Neurodevelopmental/ Early Life (eg. ADHD, autism) ¹

Cortical Anatomy and Brain Mapping ²

Attention Deficit Disorder

Autism

Computational Neuroscience

Cortex

Development

DISORDERS

MRI

Open Data

Psychiatric

STRUCTURAL MRI