An

Empirical Evaluation

of Executive Function

in Autism Spectrum Disorder

A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

Eleni Andrea Demetriou BScPsych (Hons), MPsychol (Clin), M.B.M.Sc.

Brain and Mind Centre Faculty of Medicine and Health The University of Sydney

2019

Dedication

This dissertation

is dedicated to

these extraordinary people

I have been truly blessed to have in my life

my

grandmother

Eleni Yiangoulli

and

my parents

Andreas and Aggeliki Demetriou whose selfless dedication at great personal sacrifice

to the

education of their children

inspired my lifelong quest for knowledge

and to my

remarkable nephew

Andreas Demetriou

who is showing us all

there are no boundaries to knowledge

ii Acknowledgements

It has been a privilege to undertake my research in the autism programme led by my research supervisor, Professor Adam Guastella. I would like to express my deepest appreciation and gratitude to Adam for his supervisory support, commitment and direction, and for sharing his incredible knowledge and insight while guiding me through my candidature. I would not have been able to complete my dissertation without his professional guidance and his understanding of personal challenges I faced during my studies.

I would also like to express my gratitude to my auxiliary supervisor Professor Sharon Naismith.

Your guidance, support and genuine interest in my research and me have been invaluable.

To all my wonderful research participants, I am very grateful for your interest and commitment to my study. I hope through this research to contribute in some small way to a better understanding of the challenges you face, as well as your strengths.

Arriving to this point would have been much harder without the good humour, understanding and support of fellow students:

Anne Masi and Cristina Cacciotti-Saija you have been great role models I hope to emulate.

Karen Pepper and Shin Ho Park, thank you for all your assistance and always been willing to answer even my most banal questions!

Marilena de Mayo, my deepest gratitude for your wonderful friendship, constant encouragement and for always helping me see the big picture whenever I was overwhelmed by the minutiae of research!

iii My appreciation to Lisa Whittle and Daniel Quintana, who provided me with expert knowledge and support through my first tentative steps in my candidature, and to the wonderful research interns, Magdalena Durrant and Karen Gould, whom I am privileged to now call my friends.

To Amit Lampit, thank you for your incredible patience, guidance and for making meta- analysis statistics seem almost fun!

I have been very fortunate to share part of this journey with Christine Song, thank you for your amazing support, for believing in me and for your constant encouragement and understanding.

It is not often that you meet a person who guides you throughout your professional career.

Michael McFadden, you have been an amazing mentor since my early steps out of university into a professional career; your incredible faith in me and continuous friendship, have been invaluable.

My deepest appreciation to my wonderful team leaders at the Department of Human Services,

Mirna Saad, Linda Gomez and Charlette Yaako-Khanania. If not for your wonderful understanding, encouragement and incredible support of my quests for learning this journey would have been very arduous and may have never been completed.

To all my friends and colleagues who have encouraged me throughout this quest, you are too many to mention individually but to each and every one of you my deepest gratitude.

I would like to acknowledge and thank Sally Pope for providing proof reading services and

Marianna Koulias for assistance with formatting.

Finally, I would never have had the privilege of now submitting my doctoral dissertation if not for the selfless support of my family.

iv From a very young age, my parents instilled in me a love of learning and taught me to always reach for the stars and strive for excellence but to always be grateful for the journey.

Mum, I am so sorry you are not here to witness this milestone that you always encouraged me to reach, and that your unconditional love and selfless dedication to me made possible.

Dad, I hope in some small way I can reflect your consummate commitment for learning, sharing knowledge and inspiring people to achieve their dreams.

To my brother and sister in law, Chris and Viki Demetriou, your continuous support, encouragement and understanding of my passion for learning has made this journey so much easier, and to my brother George Demetriou, your boundless generosity in setting aside your dreams so I can reach mine will never be forgotten.

My aunty Aggeliki Koulias, you have been an incredible source of support, thank you for your love and dedication.

Most importantly, to our beautiful boy, my wonderful nephew Andreas, you have brought incredible joy to our lives and you have taught me more than any education could. You were my inspiration for starting this journey and I will continue to strive for knowledge with you.

v Declaration of Originality

This thesis is submitted to the University of Sydney in fulfilment of the requirement for the

Degree of Doctor of Philosophy.

The work presented in this thesis is to my best knowledge and belief, original except as acknowledged in the text. I hereby declare that I have not submitted this material, either in full or in part for a degree at this or any other institution.

Eleni Andrea Demetriou

December 2018

vi Co-Author Declaration

We the undersigned acknowledge the following statement:

This thesis principally represents the work of Eleni Andrea Demetriou.

Ms Demetriou was responsible for:

 Research co-ordination and recruitment of research participants

 conceptual and study design

 data extraction

 data collection for cognitive and executive function test battery for

o Autism Spectrum Disorder, Social Anxiety Disorder and neurotypical control

participants

 significant contribution for data collection for the social cognition test battery

 data analysis

 data interpretation

 manuscript preparation of Chapters 2 to 6 and,

 critical revision of manuscripts

Professor Adam Guastella provided considerable supervisory support and input on the conceptual design, data interpretation, manuscript synthesis and critical review for intellectual content of each of the empirical chapters 2 to 6 and for the Introduction and Discussion chapters of this thesis.

vii Professor Sharon Naismith provided supervisory support and input on the conceptual design of Chapter 2.

Each co-author contributed to the critical evaluation of the relevant chapter(s) in this thesis by reviewing the written work prior to journal submission or preparation for journal submission.

Additional assistance was provided as follows:

Chapter 2: Dr Amit Lampit provided support and considerable input for data analysis, data interpretation and production of statistical figures. Dr Daniel Quintana provided input for conceptual design. Mr Jonathon Pye undertook data validation.

Chapter 6: Mr Nicholas Ho provided input on data analysis and completed data analysis and production of figures. Mr Shin Ho Park undertook data analysis and production of figures.

Ms Emma Thomas assisted with diagnostic assessments for the Autism Spectrum Disorder cohort.

Ms Karen Pepper contributed to significant data collection for the social cognition test battery and assisted with recruitment of research participants.

Professor Adam Guastella and Dr Alice Norton supervised clinical interns for diagnostic assessments for the Social Anxiety Disorder cohort.

Mr Django White provided administration support for the YMH test battery.

Aspects of the neuropsychological test battery assessed in this study were originally designed and developed by the Youth Mental Health (YMH) team, Brain and Mind Centre, University of Sydney.

viii Professor Adam Guastella

Professor Sharon Naismith

Professor Ian Hickie

Professor Daniel Hermens

Professor Liz Pellicano

Mr Nicholas Ho

Dr Amit Lampit

Dr Alice Norton

Dr Daniel Quintana

ix Dr Christine Song

Mr Shin Ho Park

Ms Karen Pepper

Mr Jonathon Pye

Ms Emma Thomas

Mr Django White

x Publications and Research Papers Associated with this Thesis

Autism Spectrum Disorders: A meta-analysis of executive function.

Demetriou, E. A., Lampit, A., Quintana, D. S., Naismith, S. L., Song, Y. J. C., Pye, J. E.,

Hickie, I.B., Guastella, A. J. (2018). Molecular Psychiatry, 23(5), 1198–1204. http://doi.org/10.1038/mp.2017.75

Autism, Early Psychosis, and Social Anxiety Disorder: A transdiagnostic examination of executive function cognitive circuitry and contribution to disability.

Demetriou, E.A., Song, Y. J. C., Park, S.H., Pepper, K.L., Naismith, S.L., Hermens, D.F.,

Hickie, I.B., Thomas, E.E., Norton, A.E., White, D., Guastella, A.J. Translational Psychiatry volume 8, Article number: 233 (2018) https://doi.org/10.1038/s41398-018-0193-8.

Autism Spectrum Disorder: An examination of sex differences in neuropsychological and self-report measures of cognitive and executive function.

Demetriou, E.A., Pepper, K.L., Park, S.H., Pellicano, L., Naismith, S.L., Song, Y. J. C.,

Hickie, I.B., Thomas, E.E., Guastella, A.J.

(under review, Journal of Autism and Developmental Disorders).

A transdiagnostic examination of anxiety and stress on executive function outcomes in disorders with social impairment.

Demetriou, E.A., Park, S.H., Pepper, K.L., Naismith, S.L., Song, Y. J. C., Hickie, I.B.,

Guastella, A.J.

(to be submitted for review).

xi Machine learning for differential diagnosis between clinical disorders with social impairment.

Demetriou, E.A.1, Park, S.H1, Ho, N., Pepper, K.L., Song, Y. J. C., Naismith, S.L., Thomas,

E.E., Hickie, I.B., Guastella, A.J.

1Joint first authors

(under review, Schizophrenia Bulletin).

xii Additional Research Papers Co-Authored during Candidature

Autism, Early Psychosis, and Social Anxiety Disorder: Understanding the role of social cognition and its relationship to disability in young adults with disorders characterised by social impairment.

Pepper, K.L., Demetriou, E.A., Song, Y.J. C., Park, S.H., Hickie, I.B., Thomas, E.E,

Cacciotti-Saija, C., Langdon, R., Piguet, O., Kumfor, F., and Guastella, A.J.

Translational Psychiatry volume 8, Article number: 233 (2018) https://doi.org/10.1038/s41398-018-0282-8

Disability functioning and quality of life among treatment seeking young autistic adults and its relation to depression, anxiety and stress.

Park, S.H., Song, Y.J. C., Demetriou, E.A., Pepper, K.L., Norton, A.E., Thomas, E.E.,

Hickie, I.B., Hermens, D.F., Glozier, N., Guastella, A.J.

Autism https://doi.org/10.1177/1362361318823925

Validation of the self-report World Health Organization Disability Assessment Schedule

II (WHODAS-II) among individuals with Autism Spectrum Disorder.

Park, S.H., Demetriou, E.A., Pepper, K.L., Song, Y.J.C., Thomas, E.E., Glozier, N., Hickie,

I.B., Guastella, A.J.

Autism Research 000: 1-11, (2019). https://doi.org/10.1177/1362361318823925

Distress, quality of life and disability in treatment-seeking young adults with Social

Anxiety Disorder.

Park, S.H., Song, Y.J.C., Demetriou, E.A., Pepper, K.L, Hickie, I.B., Glozier, N., Hermens,

D.F., Scot, E.M., Guastella, A.J.

(under review in ‘Early Intervention in Psychiatry’).

xiii Validation of the 21-item Depression Anxiety and Stress Scales (DASS-21) in individuals with Autism Spectrum Disorder.

Park, S.H., Song, Y.J.C., Demetriou, E.A., Pepper, K.L., Thomas, E.E., Hickie, I.B., Guastella,

A.J.

(under review in ‘Autism’).

Validation of the Social Functioning Scale: Comparison and evaluation in First-Episode

Psychosis, Autism Spectrum Disorder and Social Anxiety Disorder.

Chan, E., Song, Y.J.C., Demetriou, E.A., Kong, D., Glozier, N., Hickie, I.B., Guastella, A.J.

Psychiatry Research (2019) 276: 45-55, https://doi.org/10.1016/j.psychres.2019.03.037

.Neuropsychological predictors of observed participation in group cognitive-behaviour therapy and social skills treatment among young adults with Autism Spectrum

Disorder: Norton, A.E, Demetriou, E.A., Pepper, K.L., Park, S.H., Song, Y.J.C., Hickie,

I.B., Thomas, E.E., & Guastella, A.J.

(manuscript to be submitted for peer review).

xiv Conference Abstracts and Presentations during Candidature

Society for Mental Health Research Annual Conference, November 2018, Noosa,

Australia

Symposium: Mental health in vulnerable populations

Oral presentation: Autism, Early Psychosis, and Social Anxiety Disorder: a transdiagnostic examination of executive function cognitive circuitry and contribution to disability

Brain and Mind Centre, University of Sydney, Annual Symposium, November

2018, Sydney Australia

Poster presentation: Executive function, anxiety and stress as transdiagnostic markers in psychopathology

NSW Health Western Sydney Local Health District, Family and Carer Mental

Health Programme, February 2018, Sydney Australia

Oral presentation: Adults with neurodevelopmental disorders, research challenges and life outcomes

Brain and Mind Centre, University of Sydney, Annual Symposium, November

2017, Sydney, Australia

Poster presentation: Autism, early psychosis and social anxiety disorder, the relationship between executive function and disability

xv Allied Health Professionals Conference, Commonwealth Department of Human

Services, June 2017, Sydney Australia

Oral Presentation: Autism: Aetiology, diagnosis, novel interventions and quality of life

Midwife Association of Australia, Sleep Safe Conference, August 2016, Sydney

Australia

Oral presentation: Oxytocin and Autism

Australasian Society for Bipolar and Depressive Disorders Annual Conference,

November 2015, Sydney Australia

Oral presentation: Developing novel interventions from translational models to improve cognitive rigidity, anxiety and social responsiveness in hard to treat illness

Biological Psychiatry Australia Annual Meeting, October 2015, Sydney Australia

Poster presentation: Executive dysfunction in Autism Spectrum Disorders: A meta- analysis

12th World Congress of Biological Psychiatry, June 2015, Athens Greece

Oral presentation: Executive dysfunction in Autism Spectrum Disorders: A meta- analysis

xvi Research Funding obtained during Candidature

National Health and Medical Research Council of Australia (NHMRC)

Postgraduate Research Scholarship GNT1056587

xvii TABLE OF CONTENTS

DEDICATION ...... II ACKNOWLEDGEMENTS ...... III DECLARATION OF ORIGINALITY ...... VI CO-AUTHOR DECLARATION...... VII PUBLICATIONS AND RESEARCH PAPERS ASSOCIATED WITH THIS THESIS ...... XI ADDITIONAL RESEARCH PAPERS CO-AUTHORED DURING CANDIDATURE ...... XIII CONFERENCE ABSTRACTS AND PRESENTATIONS DURING CANDIDATURE ...... XV RESEARCH FUNDING OBTAINED DURING CANDIDATURE ...... XVII TABLE OF CONTENTS ...... XVIII LIST OF FIGURES ...... XX LIST OF TABLES ...... XXI LIST OF COMMON ABBREVIATIONS ...... XXII THESIS ABSTRACT ...... 1 CHAPTER 1: INTRODUCTION...... 2

1.1 PREFACE ...... 3 1.2 EXECUTIVE FUNCTION ...... 5 1.3 MEASUREMENT OF EXECUTIVE FUNCTION ...... 11 1.4 RESEARCH OF EXECUTIVE FUNCTION IN AUTISM SPECTRUM DISORDER ...... 20 1.5 MODERATORS OF EXECUTIVE FUNCTION IN AUTISM SPECTRUM DISORDER ...... 28 1.6 EFFICACY OF EXECUTIVE FUNCTION AS A COGNITIVE INTERMEDIATE PHENOTYPE ...... 31 1.7 PROGRAMME OF RESEARCH ...... 35

CHAPTER 2: AUTISM SPECTRUM DISORDERS: A META-ANALYSIS OF EXECUTIVE FUNCTION ...... 41

2.1 INTRODUCTION ...... 44 2.2 METHODS ...... 46 2.3 RESULTS...... 49 2.4 DISCUSSION...... 51

CHAPTER 3: SEX DIFFERENCES IN EXECUTIVE FUNCTION IN AUTISM SPECTRUM DISORDER ...... 59

3.1 INTRODUCTION ...... 62 3.2 METHOD...... 66 3.3 RESULTS...... 68 3.4 DISCUSSION...... 69

CHAPTER 4: A TRANSDIAGNOSTIC EXAMINATION OF ANXIETY AND STRESS ON EXECUTIVE FUNCTION OUTCOMES IN DISORDERS WITH SOCIAL IMPAIRMENT ...... 79

4.1 INTRODUCTION ...... 83

xviii 4.2 METHOD...... 87 4.3 RESULTS...... 89 4.4 DISCUSSION...... 91

CHAPTER 5: AUTISM, EARLY PSYCHOSIS, AND SOCIAL ANXIETY DISORDER: A TRANSDIAGNOSTIC EXAMINATION OF EXECUTIVE FUNCTION COGNITIVE CIRCUITRY AND CONTRIBUTION TO DISABILITY ...... 100

5.1 INTRODUCTION ...... 103 5.2 METHOD...... 105 5.3 RESULTS...... 107 5.4 DISCUSSION...... 110

CHAPTER 6: MACHINE LEARNING FOR DIFFERENTIAL DIAGNOSIS BETWEEN CLINICAL DISORDERS WITH SOCIAL IMPAIRMENT ...... 122

6.1 INTRODUCTION ...... 125 6.2 METHOD...... 128 6.3 RESULTS...... 132 6.4 DISCUSSION...... 134

CHAPTER 7: GENERAL DISCUSSION ...... 145

7.1 SUMMARY OF STUDIES IN PROGRAMME OF RESEARCH ...... 146 7.2 CONCLUSIONS FROM EMPIRICAL STUDIES IN THE PROGRAMME OF RESEARCH ...... 148 7.3 SYNTHESIS OF FINDINGS OF THE PROGRAMME OF RESEARCH...... 150 7.4 LIMITATIONS OF RESEARCH ...... 158 7.5 FUTURE RESEARCH DIRECTIONS ...... 160 7.6 FINAL COMMENT ...... 161

REFERENCES ...... 162 APPENDIX 1: RESEARCH PUBLICATIONS ...... 265 APPENDIX 2: SUPPLEMENTARY DATA, CHAPTER 2 ...... 283 APPENDIX 3: SUPPLEMENTARY DATA, CHAPTER 3 ...... 355 APPENDIX 4: SUPPLEMENTARY DATA, CHAPTER 4 ...... 360 APPENDIX 5: SUPPLEMENTARY DATA, CHAPTER 5 ...... 363

xix List of Figures

Figure 2.1. Developmental changes in executive function and associated impairment

in ASD ………………………………………………………………………….55

Figure 2.2. Subgroup analysis of executive function domains …………………………….56

Figure 2.3. Subgroup analysis of executive function domains by age ……………………..57

Figure 3.1. Summary of cognitive domains and assessment measures …………….72

Figure 6.1. Flow chart of training and learning algorithms ………………………………140

Figure 6.2. Heat map of comparisons across cohorts ……………………………………..141

Figure 6.3. Venn Diagram of distinguishing features between cohorts …………………..142

Figure 7.1. Research framework for executive function in ASD ………………………...158

xx List of Tables

Table 3.1. Demographic information …………………………………………………….…73

Table 3.2. Neurocognitive and executive function domains ……………………...... 74

Table 3.3. Self-report measures of executive function ………………………………….….76

Table 4.1a. Regression Analysis on executive function factor ……………………………....94

Table 4.1b. Regression Analysis on Verbal/Auditory short term memory factor ……….…..95

Table 4.1c. Regression Analysis on Visuospatial short term memory factor ………….…….96

Table 4.2a. Regression Analysis on BRIEF BRI index ……………………………………...97

Table 4.2b. Regression Analysis on BRIEF MI index ……………………………………….98

Table 5.1. Demographic characteristics and participants responses to the Depression

Anxiety Stress Scale (DASS) …………………………………………………..113

Table 5.2. Group characteristics of executive function neuropsychological and self-report

measures and disability self-report ratings ……………………………………..115

Table 5.3. Effects of executive function neuropsychological predictors on disability ……118

Table 5.4. Additive effect of executive function self-report predictors on disability ……...119

Table 6.1. Demographic descriptive statistics by diagnosis ……………………………….143

Table 6.2. Classification performance on repeated cross-validation test sets ……………..144

Table 6.3. Venn diagram tables for differentiating SAD, ASD and EP …………………...145

xxi List of Common Abbreviations

ABAS Adaptive Behavioural Assessment System

ACT Attention Control Theory

ACTr Autism Clinic for Translational Research

ADHD Attention Deficit Hyperactivity Disorder

ADI Autism Diagnostic Interview

ADOS Autism Diagnostic Observation Schedule

ANOVA Analysis of Variance

ASD Autism Spectrum Disorder

BASC Behavioural Assessment System for Children

BRI Behavioural Regulation Index

BRIEF Behavioural Rating Inventory of Executive Function

CANTAB Cambridge Neuropsychological Test Automated Battery

CMA Comprehensive Meta-Analysis

COWAT Controlled Oral Word Association Test

DASS Depression, Anxiety and Stress Scale

DCCS Dimensional Change Card Sort

D-KEFS Delis-Kaplan Executive Function System dlPFC Dorsolateral prefrontal cortex

DSM Diagnostic and Statistical Manual of Mental Disorders

EF Executive function

EP Early Psychosis

FIST Flexible Item Selection Task

HFA High Functioning Autism

IED Intra/Extra Dimensional

ICD International Classification for Diseases

IQ Intelligence quotient

xxii LM 1 Logical Memory I

LM II Logical Memory II

MANOVA Multivariate Analysis of Variance

MI Metacognition Index

MR Multiple Regression

MOOSE Meta-analysis of Observational Studies in Epidemiology

OL Overlap percentage statistic

PAL Paired Associate Learning

PeQoL Paediatric Quality of Life Index

PFA Principal Factor Analysis

PFC Prefrontal cortex

PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses

RAVLT Ray Auditory Verbal Learning Test

RDoC Research Domain Criteria

RVP Rapid Visual Processing

SAD Social Anxiety Disorder

STAI State and Trait Anxiety Inventory

SPSS Statistical Package for the Social Sciences

SSP Spatial Span

STM Short Term Memory

SWM Spatial Working Memory

TMT Trails Making Test

SRS Social Responsiveness Scale

ToH Tower of Hanoi

ToL Tower of London

TYP Neurotypical control

WAIS Wechsler Adult Intelligence Scale

xxiii WCST Wisconsin Card Sorting Test

WHODAS World Health Organisation Disability Assessment Scale

WM Working memory

WMS Wechsler Memory Scale

WTAR Wechsler Test of Adult Reading

xxiv Thesis Abstract

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition characterised by social impairment and restricted/repetitive behaviours. It is associated with significant disability and poor life outcomes with increasing interest in the factors that contribute to this disability. A significant body of research has focused on investigating the role of executive function (EF) in

ASD. This thesis presents a series of studies that aim to advance the knowledge of EF in ASD.

The studies considered factors that may moderate EF and investigated the role of EF in diagnosis and predicting disability. The research cohort comprised of youth and adults with

ASD. Comparisons were made with clinical groups diagnosed with Social Anxiety Disorder and Early Psychosis. The first empirical study presented a meta-analysis of the extant literature on EF in ASD across the lifespan. Empirical studies 2 and 3, explored the role of moderators on EF including sex differences and affective states. Empirical study 4, examined the role of

EF in differential diagnosis and in predicting disability. The final study utilised a machine learning paradigm and examined whether EF discriminated the ASD cohort from the comparison groups.

The research results point to broad executive dysfunction in ASD not influenced by moderator variables or sex differences. Affective states moderated EF across all comparison groups, suggesting a transdiagnostic influence. EF differentiated the ASD cohort from comparison groups and was a unique predictor of disability for the ASD group only. The studies presented in this thesis highlight the importance of a multifaceted evaluation of EF in ASD.

This will allow evaluation of unique and shared factors influencing disability outcomes, acknowledge the contribution of mental health factors to EF and facilitate targeted intervention and remediation programmes. Importantly, the cross diagnostic relevance of these factors could facilitate resource allocation and social inclusion.

Page | 1

Chapter 1: Introduction

Page | 2

1.1 Preface

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder that has been traditionally defined by difficulties in at least two domains: social communication and social interaction, and restricted, repetitive patterns of behaviour, interests or activities1. Social communication and interaction includes difficulties in reciprocal social interaction2, non-verbal social communication3, 4 and ability to develop, maintain and understand relationships5. Symptoms associated with the restricted and repetitive behaviours manifest across motor, verbal, non- verbal and sensory modalities6 and include motor stereotypies, echolalia, insistence on sameness, ritualised behaviours, narrow interests and hyper- or hypo-reactivity to sensory stimuli1.

Autism was first introduced as a diagnostic category in the 3rd edition of the Diagnostic and

Statistical Manual of Mental Disorders (DSM-III)7. Prior to this edition, the DSM diagnostic criteria8, 9 focused on differentiating a subgroup of children with schizophrenia characterised by extreme social withdrawal10. In the DSM-III7, the discrete diagnostic category of Infantile

Autism focused on symptoms associated with language deficits and lack of responsiveness.

Onset of the condition was defined as prior to 36 months of age. In the revised DSM-III-R11, diagnostic criteria were broadened to encompass symptom clusters associated with social deficits and repetitive/stereotypic behaviours. Age of onset expanded to include individuals with onset over 36 months. Broadening of diagnostic criteria continued in successive editions of the DSM1, 12-14 and led to a higher percentage of high functioning individuals having a diagnosis of Autism Spectrum Disorder15. In the DSM-51, previous autism classification categories (e.g. Autistic Disorder, Asperger’s Syndrome) were subsumed under the broader diagnosis of ASD with three incremental levels of symptom severity. These changes led to the diagnosis of ASD in individuals with a broad range of intellectual and cognitive functioning, language ability and symptom severity16.

Page | 3

A number of theoretical cognitive models5, 17, 18 have been developed to explain the occurrence of these symptom clusters in autism. One of the most widely recognised cognitive explanations comes from the observation that people with autism tend to show executive function (EF) difficulties across the lifespan19, 20. Although there is considerable conceptual divergence on the definition of executive function21-23, broadly defined, EF refers to the capacity to engage in

‘independent, purposive, self-serving behaviour’21 (p.35) and is guided by discrete EF sub- component processes24-26. Impaired EF component processes in ASD include set-shifting, planning, response inhibition and working memory27. Evidence, however, of executive impairment across the autism spectrum is mixed28, 29. This may be partially due to the changes and broadening of the diagnostic criteria for ASD over time1, 30 and theoretical divergence in the conceptualisation of EF and its assessment26.

Early research of EF in ASD focused primarily on executive impairment across distinct domains and also investigated its relationship to stereotypic and repetitive behaviours31.

Emerging evidence suggests, however, that EF has a broader influence on the ASD phenotype and may embrace social cognition32, 33, mental health34 and disability35, 36, and contributes to life outcomes37. Research support of a relationship between EF processes and motor atypicalities38, 39, a key characteristic in ASD, lends further support to the relevance of EF in this cohort. Conversely, EF performance in ASD may be moderated by a number of factors including general intellectual functioning40 and affective states41, 42. Given the likely multifaceted influence of EF in ASD, a detailed exploration of this relationship could lead to a significant contribution to the study of autism. Research on EF, however, in both normative literature and ASD, is confounded by conceptual and methodological limitations that translate to considerable variability between studies43. This chapter will discuss the development of the

EF concept, theoretical models, conceptual and methodological limitations, and application of

EF to the study of autism.

Page | 4

1.2 Executive function

Definition:

The study of EF has a long history, the term, however, was first proposed in the mid-20th century to explain functions associated with the frontal cortex44. The frontal lobes were of interest due to observations that frontal lobe damage was associated with impairment of discrete functions (e.g. planning, organisation and self-regulation) with the intellectual level of functioning remaining mostly intact. These observations were documented in detail in an early case study of Phineas Gage described by Harlow in 184845. Following traumatic injury to the frontal lobes, Phineas Gage exhibited disinhibition, poor planning and personality changes, while no significant differences were noted in his general cognition. Harlow referred to this pattern of changes as the ‘frontal lobe syndrome’45. This and subsequent case studies46 led to the conclusions that the frontal lobes have a primary role in organising higher order functions47.

Much of the subsequent research of EF focused primarily on this brain region and the functions inferred to be associated with it48, 49.

Despite a focus on the frontal cortex, there is no consensus on the definition of EF, theorised models and underpinning mechanisms. Commonly reported definitions of EF26 refer to an overarching regulation of higher order cognitive processes that is goal directed22, 50, future oriented51 and that guides self-regulation26, 50. Many researchers fractionate EF to individual component domains (EFs) 24 25. While there is some agreement on the definition of the broad

EF construct26, individual EFs reported in the literature range from two52 to more than thirty26.

In the fractionated model, the most commonly reported or core EFs25, 53 are response inhibition

(ability to inhibit a dominant response), working memory (updating information in short term memory) and set shifting (ability to shift mind set to new concepts)24, 26. Some researchers propose different levels of complexity with core EFs contributing to higher order EF constructs including reasoning, planning and problem solving25.

Page | 5

The literature diverges further when characterising EF and may conflate underpinning mechanisms and behavioural outcomes26. Executive function(s) have thus been operationalised along cognitive, behavioural or self-regulatory processes and individual EF components are variably classified in the literature as domains, cognitive processes, behavioural skills or tasks26. Definitions of EF usually reflect different theoretical and methodological approaches, while in some studies, they are researcher specific and atheoretical26. In this thesis, the consensus definition of EF as derived from a recent meta- analysis26 will be adopted. Executive function is defined as an overarching goal directed supervisory process that regulates distinct EF sub-component processes. The sub-component

EF processes will be primarily referred to as constructs or domains to minimise undue inferences about underpinning mechanisms or behavioural outcomes.

Theoretical models of executive function

Models of EF draw on different theoretical paradigms and levels of analyses and include cognitive, clinical and neurobiological frameworks54. This has in part contributed to the observed disparity in theorised models and underpinning mechanisms26. The following overview, while not exhaustive, aims to draw on some of the main distinctions between EF models and their pertinence to ASD research. Theoretical models of EF are broadly divided into unitary versus multifactorial models with the latter proposing multiple EF domains.

Single factor models of EF focus on unitary executive control and are mostly theorised from a cognitive framework. Prominent models include Baddeley’s central executive hypothesis55,

Posner’s theory on cognitive attention56 and the supervisory attentional system (SAS)57.

Baddeley’s58 model applies the concept of unitary executive control to the processes of working memory. It proposes that a central executive mechanism supervises information flow to working memory component processes55 and is responsible for behaviour regulation58, 59. The unitary model proposed by Posner60 focuses on the executive control role of attention.

Page | 6

Regulation of cognitive and non-cognitive functions was theorised to be maintained by attentional processes. Attention was divided into orientating, alerting and cognitive components, the latter responsible for executive cognitive control. The supervisory attentional system (SAS)57 also focused on attention as the main feature of executive control. A distinction was made between routine or habituated actions versus non-routine actions. Non-routine actions require the individual to disengage from habituated behaviour and make a novel response. The model proposes that the attentional system exerts supervisory control in novel situations where routine behaviours must be inhibited.

The unifactor models described above do not make specific inference on the localisation of the executive control process although they are generally assumed to be in the frontal cortex.

Unifactor models predict that executive dysfunction results in broad impairment in behaviour regulation. The SAS model more specifically predicts perseverative behaviours, distractibility and apathy due to disrupted inhibitory control61.

Alexander Luria62 proposed one of the early models of EF that focuses on neurobiological rather than cognitive level of analysis. Luria’s model does not readily align with either the unitary or multifactorial models of EF. Although the model prescribes an overarching regulatory role to the frontal lobes, it also advocates integration of multiple brain systems and functions to achieve this. Luria’s model thus suggests a framework that bridges between the unitary models described above and multifactorial frameworks of EF. Luria, did not explicitly use the term EF, however, his conceptualisation of higher order cognitive processes regulated by the frontal lobes aligns with later definitions of executive function63. Luria described behaviour as the outcome of the functional organisation and integration of three brain functional units labelled as the ‘first’, ‘second’ and ‘third’ functional units64. The first and second functional units were concerned with alertness and sensory information processing, guided by the parietal, temporal and occipital lobes and some brain stem structures. The third

Page | 7 functional unit was formed by the frontal lobes and was responsible for complex information processing64. Damage to the third unit was associated with the frontal lobe syndrome63 characterised by disinhibition, inability to follow a sequence of instructions and repetitive motor movements.

Multifactorial models of EF can be distinguished on several levels. A number of models differentiate EF into discrete domains24, 25, while others emphasise integration of distinct EF domains with other cognitive or psychological processes to achieve goal directed outcomes22

51. Individual EF domains may be inferred from clinical observations following frontal lobe damage65, theorised from impaired performance on cognitive assessments of frontal lobe function or derived empirically from experimental paradigms24.

A major contribution to the multifactorial framework of EF was made by Miyake’s unity in diversity model52, 66. Utilising factor analytic statistical procedures, the three commonly studied core EF domains (response inhibition, working memory and set-shifting26) were empirically verified67. The statistical analysis confirmed that the EF domains hypothesised above corresponded with three empirically derived factors of set-shifting, inhibition and working memory67. The study also showed that these core EF domains were not completely independent but shared a common factor67. The three-factor model was revised to two factors retaining set-shifting and working memory as core EF domains, while the shared factor was ascribed to response inhibition66. Further research to elucidate mechanisms of the common factor indicated that individual variability observed in EF performance may be attributed to genetic influences on the shared factor68. Furthermore, these differences could not be explained by differences in intellectual level of functioning. This theoretical framework of EF may be particularly relevant in the study of EF in ASD cohorts without intellectual disability. The condition is characterised by genetic variability69, heterogeneity28 and impairment in EF performance70, 71 despite intact intellectual functioning. Observations of broad executive

Page | 8 dysfunction in ASD72 may be attributed to the shared common factor, with genetic variability explaining individual differences in performance68, 73.

An alternative framework25 proposes six independent domains of executive function. These include the core domains of set-shifting, response inhibition and working memory, and three higher order complex executive functions: planning, decision-making and problem-solving.

Much of the research associated with this model derives from neuropsychological assessments associated with frontal lobe damage and cognitive remediation programmes24, 25. The model fractionates EF to six independent domains; in ASD, it could explain distinct profiles of executive dysfunction.

The Delis-Kaplan model65 evolved from clinical observations of functions sensitive to frontal lobe damage and proposes nine EF processes. Its key attribute is that it differentiates selective

EF domains based on processing mode, discriminating between visuospatial and verbal modalities. A test battery (Delis-Kaplan Executive Function System-D-KEFS) has been developed specific to the individual EF processes proposed by the model. The Delis-Kaplan model incorporates the three core EF domains, and in addition proposes problem-solving, fluency, categorical processing, deductive reasoning and verbal abstraction as EF components.

Given that the model proposes discrete EF domains, it could explain a fractionated profile of

EF deficits in ASD cohorts.

The unitary and multifactorial models described above characterise EF(s) as cognitive process(es) that can be measured based on performance in EF assessments. In contrast to the above, the following two models proposed by Stuss22, 74 and Barkley51, incorporate multiple processes (including emotional, behavioural and self-regulation) that are not always performance-based or characterised as EF processes.

Page | 9

The model proposed by Stuss identifies task-setting and monitoring as EF processes regulated by the frontal lobes. These EFs interact with other frontal lobe non-EF processes to accomplish behavioural control towards accomplishment of novel tasks22. This distinction was made in order to dissociate frontal lobe function equating with EF processes only. The other frontal lobe functions were described as energisation, behavioural/emotional self-regulation and metacognition. Energisation referred to processing speed when completing cognitive tasks.

Behavioural/emotional self-regulation was defined as a non-EF frontal lobe function. It may require, however, activation of the EF components (task-setting and monitoring) to manifest.

Metacognition was described as a higher order supervisory process of the other executive and non-executive functions. It has an integrative role important in accomplishing novel complex tasks. It is of note that the non-executive metacognition function has been characterised as an

EF process by other researchers26.

The model proposed by Barkley23 is qualitatively different from the models discussed above as it focuses on observed behaviours and not performance-based cognitive measures of executive function. The model was empirically derived from rating behaviours considered to tap executive regulation51. It comprises five EF factors that are surmised to be influenced by external (cultural/societal factors) as well as intra-individual processes. Barkley proposed that each EF represents complex processes of self-regulation aimed at achieving future goals.

These were self-management in time, self-organisation and problem solving, self-restraint, self-motivation and self-regulation of emotion51.

In summary, models of EF described above increase in complexity ranging from unitary attentional control to composite integration of EF and non-EF processes. Earlier models56, 57,

59 focus primarily on single process of cognitive executive control to guide executive function.

Multifactorial models extend executive control to multiple but distinct cognitive processes25,

65. The unity in diversity model24 in part encompasses mechanisms of unitary models by

Page | 10 proposing a shared EF factor. Recent models advocate metacognitive, cultural and societal influences on executive function and incorporate emotional and behavioural components22, 23.

Discriminating between these models is outside the scope of this thesis; however, they will form a basis for later discussion on how findings on EF in ASD may be explained.

1.3 Measurement of executive function

The measurement of EF reflects the different theoretical approaches discussed above. A number of assessment tools were designed to assess frontal lobe function 65 75 while others focus on behaviour correlates of executive function 76 51. The measurement of EF has been subject to considerable debate, particularly in relation to broader issues of validity and reliability, and concerns on the task purity of associated measurement tools.

Validity and reliability

Validity reflects whether a specific EF assessment tool measures the theorised EF construct50,

77, whereas reliability refers to the consistency of the measurement50. A number of EF assessments have been developed based on functions observed to be impaired following injury to the frontal lobes75. Observed functional deficits were operationalised and measured by standardised psychometric tests and experimental tasks65, 78. This contributed to the frontal lobe hypothesis of EF49, 62, 79, where in some research, functions impaired following frontal lobe damage were defined as executive. This approach has been critiqued on the premise that it introduces heuristic flaws in the study of executive function. It is argued that a circular logic applies to this conceptualisation of EF23 and its measurement. It assumes that tests sensitive to frontal lobe damage measure EF processes and are therefore valid assessments of executive function51. Further, it amalgamates cognitive and neurobiological levels of analysis to validate a cause-effect relationship23, 80, when each only represents a different measurement level of the theorised construct. Associated with this is the critique that there is a lack of clear distinction

Page | 11 between functions classed as executive from other non EF processes51. This likely provided scope for broad divergence on the definition of EFs, evident in the literature by identification of more than thirty EF domains26.The reliability of measuring EF has also been questioned based in part on observations of generally low correlations between repeat administrations of an EF test (low test-retest reliability)50, 51. Explanations include suggestions that as EF domains reflect novel processes, results are compromised in the second administration of a test81. In addition, participants are likely to engage in different problem-solving strategies at different test administrations.

Task-purity

A further criticism of measuring EF reflects on the task-purity of the assessment measures18,

24. Task-purity assumes that an assessment tool measures a unique EF component. Research evidence suggest that multiple EF and non-EF processes may be captured by psychometric tests of proposed discrete EF domains. For example, assessment tests considered classic measures of set-shifting have been reported to also measure other core EFs including inhibition and working memory82. This criticism also applies to behavioural rating scales where it is acknowledged that they assess a composite set of executive behaviours and do not represent pure measures of the specific domain51. A number of neuroimaging studies, while reporting null findings on neuropsychological measures of EF, observed significant differences in brain activation patterns between comparison groups83, 84. This reflects the task-purity limitations discussed above and suggests that performance measures of EF do not assess unique EF components processes.

Developmental trajectory of executive function

Executive functions are characterised by developmental changes, with most significant differences occurring between early childhood and middle adolescence85. Core EFs24 emerge in late infancy and

Page | 12 early childhood with response inhibition and working memory preceding the development of set shifting. It has been suggested that this may be due to reliance on the former EF processes for appropriate expression of set shifting86. Response inhibition is observed from preschool years and is reasonably well formed by age eight86. Improvements in response inhibition tasks continue into late adolescence85, 86. Similarly, working memory emerges in early childhood86 with performance levels plateauing by mid86 to late adolescence87. Evidence for the developmental trajectory of set-shifting are mixed. While there is consensus of set-shifting emerging in early-mid childhood86, 87, maturation levels vary from late childhood87 to early adulthood86. Differences may be in part due to the specific measuring tool employed in the study86, 87.

The divergent developmental trajectories of EF domains49 and atypicalities in EF development reported in ASD88 may contribute to methodological limitations in measuring executive functions in this group. Much of the EF research is cross-sectional and thus may not adequately capture developmental changes or differences between comparison groups. In ASD, this is of particular concern as a number of studies include mixed age samples89-91, thus limiting conclusions on EF comparisons and associated processes.

Assessment instruments of executive function

Measurement of EF traditionally focused on cognitive level of assessment and was primarily based on neuropsychological assessments sensitive to frontal lobe damage65, 75. These assessments were deemed objective measures with some evidence of validity and reliability92,

93 despite some criticism78, 82 as noted earlier. Measurement of EF traditionally focused on cognitive level of assessment and was primarily based on neuropsychological assessments sensitive to frontal lobe damage65, 75. These assessments were deemed objective measures with some evidence of validity and reliability92, 93 despite criticism78, 82 as noted earlier.

Neuropsychological tests measure EF in structured experimental settings. They are performance-based measures of the participant’s effectiveness and efficiency to achieve the

Page | 13 maximal test outcome. Test effectiveness or accuracy, measures the participant’s correct responses while efficiency assesses the quality of the responses (e.g. total errors made or time taken prior to identifying a correct strategy). In addition to concerns regarding construct validity discussed above, concerns about their ecological validity have also been raised.

Ecological validity refers to the extent that assessment outcomes align with conditions in the natural environment94 and is usually evaluated by the standards of representativeness and generalisability95. Representativeness reflects how the assessment items of an EF test correspond in format to situations outside the laboratory96. Generalisability indicates whether test performance reflects impairment in the real world95. There are emerging trends to develop new neuropsychological tests that meet the ecological standards described above. Examples include the Multiple Errands and the Six Elements tasks97 where test content reflects every day activities (e.g. complete a shopping task). Virtual environment assessment platforms have been proposed as the appropriate future direction for enhancing test ecological validity98. A number of studies have demonstrated the efficacy of this approach in Autism Spectrum Disorder99.

Experimental tasks have also been utilised and they purport to measure unique component processes and to represent a purer measure of executive function24. Experimental tasks are primarily research specific although some paradigms are commonly used. For example the

Go/No-Go100 and Eriksen flanker tasks101 are common measures of response inhibition. The n-back task102 is widely utilised in assessing working memory. Variability between studies may reflect experiment specific methodological differences. These may include stimulus presentation mode, e.g. computerised versus traditional or stimulus type, e.g. verbal versus pictorial, or the temporal parameters of stimulus may vary, e.g. duration of stimulus presentation or time lapse between stimuli. These limit generalisability of study outcomes and conclusion that may be drawn on validity and reliability.

Page | 14

The development of behavioural rating scales of EF51, 103 aimed to provide more ecologically valid assessments51, 95 104 that focus on executive regulation of everyday behaviours. A number of studies support their ecological validity36, 105-107, although the underpinning processes are unclear. It has been argued that behavioural ratings do not tap the same underlying processes as measured by cognitive assessments of executive function108. It is asserted that behavioural scales reflect the individual’s motivation to evaluate and achieve goals and expectations regarding goal outcomes108. By comparison, psychometric assessments and experimental tasks are performance-based measures with optimal outcomes that are predetermined. Different cognitive mechanisms may therefore be responsible for determining optimal performance outcomes compared to self-evaluations of goal-related processes.

It is also of note that behavioural self-ratings may be subject to response biases and may be influenced by a number of factors including affective states109 and insight110. These can result in biased self-reporting including on behavioural ratings of executive function. Fear of negative evaluation and biased self-attention is a characteristic of individuals with high trait levels of social anxiety. A significant association between these factors and biased self- reporting on subjective quality of life was demonstrated in the study by Dryman and associates109. A second study showed that individuals with Social Anxiety Disorder (SAD) report high levels of executive dysfunction despite intact achievement on performance-based measures35. A number of cognitive theories including theory of mind5 indicate that some clinical cohorts, including those with ASD, experience difficulties in understanding their own mental state 32, 33. These factors may result in different evaluative processes that may not reflect on the EF construct and may negate comparisons between groups. Support for this premise is provided by evidence that atypical connectivity in the default mode network, which largely influences self-referential processes, was associated with poorer insight on self-report measures of psychopathology110.

Page | 15

Definition and measurement of six executive function domains

Research on EF diverges on several levels including theoretical models, conceptual definitions and operationalisation of the EF construct. To facilitate a uniform terminology, six EF component domains will be defined below and these definitions will be applied in the empirical studies of this dissertation (Chapters 2 to 6). Assessment tools that have been primarily associated with each EF domain will also be discussed with some discussion of research linking cognitive measures of EF with underlying neural circuitry. Focus will be on discrete domains of EF as these are the most extensively studied in both normative and ASD literature26, 111. A summary of the key EF domains and associated developmental changes in ASD is presented in Appendix 2, Table S2.1.

Set Shifting/Concept Formation

Set-shifting or concept formation is defined as the capacity to shift between mental processes to form new concepts and identify the conceptual relationships shared by stimuli24, 65. Other commonly used terminology for set-shifting includes concept formation and cognitive or mental flexibility112. Theorised mechanisms for set-shifting have included switching between mental processes. It is argued, however, that set-switching113 represents a distinctly different

EF component that needs to be differentiated from set-shifting (discussed below). Set-shifting is usually assessed by psychometric tests and experimental tasks that require the participant to deduce a new concept. Some commonly used assessments of set-shifting include the

Wisconsin Card Sorting Test (WCST)82, the Intra/Extra Dimensional (IED) shift test from the

Cambridge Neuropsychological Test Automated Battery (CANTAB)87, the Dimensional

Change Card Sort Test (DCCS)114 and the Flexible Item Selection Task (FIST)115. The DCCS and FIST were developed with the aim to obtain a purer measure of set-shifting by reducing working memory demands and to simplify administration. A recent meta-analysis of the

WCST in individuals with ASD reported medium to large effect sizes for the commonly

Page | 16 reported outcome measures (stages completed and perseveration errors) suggesting that it is a reliable measure of set-shifting in this group116. Neuroimaging studies using the WCST indicate that performance is associated with activity of the lateral prefrontal cortex, anterior cingulate cortex and inferior parietal lobule82. Neuroimaging research in children with ASD using a modified version of the WCST117 showed activation of the right insula at the set-shifting stage. In this study, increases in activation in the control group during the extra-dimensional shift compared to decreased activation in the ASD group differentiated between the two groups and suggest frontal region involvement in this task.

Mental Flexibility/Set Switching

Set-switching is defined here as the capacity to switch between mental processes (multiple tasks, operations or mental sets) in response to changing demands113, 118. Different experimental designs may be utilised in the study of set-switching; one traditional assessment is the Trails Making Test (Trails B)119. This requires the participant to switch between two mental sets, in this case numbers and letters. A number of studies demonstrated the involvement of the prefrontal cortex (PFC) and frontoparietal areas of the brain during a set- switching task120 82 121.

Fluency

Fluency is defined as the capacity to generate verbal and non-verbal stimuli including ideas122, designs65 and words123. Verbal fluency is a frequently studied measure of executive functioning48 and is distinguished into phonemic (generativity for unrelated words) and semantic fluency (generativity for semantically-related words or categories)124. Psychometric assessments of fluency include the Controlled Oral Word Association Test (COWAT)125 and the Design Fluency test from the D-KEFS battery65. There is some debate as to whether phonemic and semantic fluency represent EF50, 123 or language processes126. However, as a number of studies suggest that verbal fluency is reliant on core EF processes73, 127 and this is

Page | 17 supported by neuroimaging findings48, it is classified here as an EF domain. Neuroimaging research suggests extended brain network connectivity between dorsal and ventral brain networks in fluency tasks with activation in the inferior frontal gyrus (IFG) and left dorsolateral prefrontal cortex48. A differentiation between dorsal and ventral brain networks was observed between phonemic and semantic fluency tasks124 respectively, supporting that these represent different EF processes.

Planning

Planning is defined as the capacity to execute a sequence of actions so that a desired goal is achieved50. The Tower of London (ToL)128, the Tower of Hanoi (ToH)129 and the CANTAB

One Touch Stockings of Cambridge87 are commonly used psychometric tests of planning.

Extended brain network involvement is reported during completion of planning tasks including activation of the dorsolateral prefrontal cortex (dlPFC), the anterior and posterior cingulate areas and the parietal cortex130. Superior performance is associated with a spatially extended activation of the left dlPFC compared to average performance. A meta-analysis of neuroimaging studies using the ToL task reinforced the above finding and also identified bilateral contributions of mid‐dlPFC as well as additional frontal (e.g. frontal eye fields, supplementary motor area) and parietal regions (e.g. inferior parietal cortex)131.

Response Inhibition

Response inhibition primarily refers to the ability to inhibit a previously learned or prepotent response53. Two additional EF component processes, resistance to distractor interference and resistance to proactive interference complement prepotent inhibition70. The former refers to the ability to process a target stimulus while ignoring irrelevant information presented at the same time. The latter refers to the ability to efficiently process distractors from recently activated memory stimuli. Resistance to proactive interference is classified in some research as a working memory process and it is not included here as a response inhibition component

Page | 18 process. Psychometric measures of response inhibition include the Stroop test132 and the

Colour-Word Interference test65 from the D-KEFS battery. Both are based on colour-word conflict. Experimental measures include the Go/no-Go task 100 and the Hayling test133 for measuring prepotent response inhibition and the Eriksen flanker task101 for measuring resistance to distractor interference. A study utilising the Go/no-Go task identified a large right-lateralised region of activity in the right prefrontal cortex82, while another study contradicts this and suggests no specific inhibitory processes within the frontal lobes.

Instead, involvement of an extended functional network134 that reflects both inhibitory and non-inhibitory processes was reported. The researchers suggested that inadequate assessment of other cognitive non-inhibitory mechanisms led to a disproportionate focus of frontal lobe involvement in response inhibition. This may reflect the task purity limitation discussed earlier.

Working Memory

The concept of working memory is sometimes used interchangeably with that of short-term memory (STM), although different processes relate to each. Working memory refers to the capacity to store and dynamically manipulate information in temporary short term memory50.

A range of psychometric tests and experimental tasks have been developed for the study of working memory. Psychometric tests include subtests from larger cognitive test batteries, for example the Letter Sequencing task and the Digits Backwards task from the Wechsler Memory

Scale (WMS)135 and the Spatial Working Memory (SWM) from the CANTAB battery. The n- back task102 is the most frequently used experimental task and it can be presented by verbal or pictorial stimuli. The n-back task requires participants to remember and recall a stimulus presented n-back trials to the target trial. Activation of frontal regions during a verbal n-back task was demonstrated in a number of neuroimaging studies117 (bilateral superior and middle

Page | 19 frontal gyri, bilateral frontal polar regions, precuneous gyrus) while a pictorial n-back task activated inferior frontal areas.

The above overview of EF domains and their measurement reflects the variability inherent in the study of executive function. Differences in conceptual definitions of EF domains and the wide range of assessment tools for each EF domain may impact on generalisability and comparisons between studies. There is overlap in the definition of some EF domains (e.g. set- shifting, cognitive flexibility, set-switching) and while neuroimaging studies show that many

EF processes are associated with the frontal lobes there is no unique involvement. The following section will review the study of EF in ASD and relate some of the above themes to the research conducted in individuals with Autism Spectrum Disorder.

1.4 Research of executive function in Autism Spectrum Disorder

Evidence of impairment in EF processes in autism have been documented since the early description of autism by Kanner136. These were mostly based on case studies and described perseverative behaviours and impairment in abstract aptitude18. Following the introduction of the autism classification in the DSM-III7, there has been a proliferation of empirical studies examining executive functions, partly in response to observed perseverative behaviours137.

Much of the research adopted a multifactorial approach assessing distinct EF domains112. Few studies, however, assessed EF across multiple domains and less examined adult cohorts. This limits conclusions, whether executive dysfunction in ASD characterised by broad atypicalities or whether specific domains have a differential impact on performance. The following review focuses on research completed in individuals with a diagnosis of ASD without intellectual disability.

An early study by Hoffman and associates138 compared a group of autistic children to neurotypical controls. Deficits in the EF domains of set-shifting and set-switching were

Page | 20 identified while there were no group differences on non-EF visuospatial tests. The study concluded that findings correlate with frontal lobe impairment in autism and was one of the early studies to link autism to aberrant neural circuitry of the frontal lobes. These findings were replicated in a study on set-shifting that included children with borderline levels of intellectual functioning139. In a series of studies, Rumsey and associates140-142 reported deficits in set-shifting, set-switching and planning in adults with autism. Importantly, the studies showed that the differences were not moderated by level of intellectual functioning or learning disability. Findings of impaired set-shifting and planning were replicated in a mixed younger age cohort of 7-18 years61. In this study, executive dysfunction was attributed to the mechanisms underpinning the SAS model. That is, impairment in executive attentional control resulted in poor responding in novel situations and poor plan formation.

A series of studies completed by Ozonoff, Pennington and colleagues143 144 92 91 identified deficits in ASD on discrete EF domains. In a study assessing set-shifting and planning, in a mixed age group (range 8-20 years)143, the ASD cohort was significantly impaired on measures of set-shifting and planning with the impaired performance persistent at a three year follow-up143, 144. The findings on set-shifting and planning were replicated in a mixed age group91 with ASD utilising traditional and computer-based assessments. Contrary to the consistent significant findings of impaired set-shifting and planning, an early study examining response inhibition in children145 found no differences between groups while a study on verbal fluency in children146 identified impairment in generating abstract but not concrete lists of words (phonemic versus semantic fluency)146. The extant literature on EF in ASD was formalised in theoretical reviews18, 111 and contributed to the executive dysfunction hypothesis111.

Page | 21

Executive Dysfunction Hypothesis

One of the early references to executive dysfunction as a theoretical model for ASD was made by Pennington and Ozonoff18. In their review of neurodevelopmental disorders, EF was proposed as a marker that can differentiate between neurodevelopmental conditions, such as

Attention Deficit Hyperactivity Disorder (ADHD) and ASD, on discrete EF cognitive processes. The ASD cohort was significantly impaired compared to neurotypical controls on set-shifting and working memory but were intact on response inhibition. By comparison, the

ADHD participants were impaired on response inhibition only. Thus, it was proposed that individuals with ASD may be differentiated from a neurotypical comparison and ADHD cohorts based on EF measures.

Empirical findings on EF deficits in ASD were formalised in the executive dysfunction hypothesis111, 112 proposed by Hill in an effort to review and integrate the cumulative literature of EF in autism. Hill referred to EF as an ‘umbrella term’ and identified six EF domains. The selected domains were based on functions shown to be sensitive to frontal lobe damage. They were planning, working memory, impulse control, inhibition, set-shifting and, initiation and action monitoring. The review focused on four EF domains: planning, mental flexibility, inhibition and self-monitoring. These were assumed to represent core EF processes. Mental flexibility, set-shifting and cognitive flexibility were used interchangeably as representing the same EF domain processes.

In her review, Hill identified a trend but not unequivocal support for EF impairment in autism.

It was suggested that EF may reflect a broader cognitive phenotype in autism. Impairment in

EF was observed in ASD individuals and (to a lesser degree) to relatives of probands with autism. While highlighting the relevance of EF in the study of autism, Hill also identified a number of limitations that may impact on the generalisability of the findings to this cohort.

Page | 22

Executive function deficits were not exclusive to ASD but were observed in other neurodevelopmental conditions, for example, ADHD and schizophrenia. This limited the utility of EF as a unique diagnostic marker for autism. Second, there was lack of consensus between ASD studies on the affected EF domains and third, EF deficits were not homogeneous within the autism cohort. Further limitations referred to the small number of longitudinal studies and the use of mixed age groups in many of the cross-sectional studies.

The following overview represents a selection of studies of EF in ASD with focus on the three core EF domains of set-shifting, response inhibition and working memory67. The studies reviewed were drawn from the larger meta-analysis reported in this thesis72 and represent approximately thirty percent of the studies for each of the age groups of children (under 13), adolescents (13-18), and adult (over 18 years) participants. The overall conclusion from this overview concur with the limitations raised by Hill and indicate that despite extensive research in EF in ASD, there are still areas requiring further investigation.

Set-Shifting

A range of outcome measures were utilised in the study of set-shifting. The most commonly reported are number of successful shifts to new categories (stages completed) and perseverative errors116. Other measures included total number of errors, non-perseverative errors as well as experimenter derived composite measures. In children with ASD, a number of studies reported impaired performance on stages completed115, 147, 148, perseverative errors147-150, total errors151 and a composite measure assessing shifting flexibility152, 153. By comparison, the following studies reported no differences on comparable measures including stages completed154-157, perseverative errors155, total errors157 and a composite measure switching cost157. In adolescents impaired performance was reported for perseverative errors158-160 and categories completed160. Studies by Bolte161 and Happe162, however, reported no significant differences between ASD and neurotypical comparison groups on the computerised version of the WCST

Page | 23 and the IED test respectively on perseverative errors161 and trials to reach criterion162. Studies involving adults with ASD also report mixed findings. Rumsey and Hamburger163, 164 reported impaired performance across stages completed and perseverative errors while Dichter and

Belger165 and Maes et al found no differences166.

Although all studies utilised tests purporting to assess set-shifting, methodological differences may have contributed to some of the variability observed between studies. Neuropsychological tests of set-shifting differ on degree of explicit instruction provided to participants regarding the task. The WCST75 and IED87 test both require the participant to identify the shift stage and derive the new rule. The DCCT114 provides explicit instructions on the new rule, whereas during administration of the FIST115, participants are provided with a cue to shift but no explicit instruction on what the new rule is. Differences between studies utilising these tests may reflect different mechanisms underpinning these test instructions. The latter two tests may better reflect set-switching rather than set-shifting processes. Related to test administration are differences observed based on feedback provided during test completion. A study by

Broadbent and Stokes167 identified impaired performance in individuals with ASD under selective feedback conditions. No differences between groups were reported when feedback to participants was provided only when making correct responses. However, when feedback was given for both correct and incorrect responses, individuals with ASD performed significantly worse. The authors concluded that negative feedback enhanced perseverative responses leading to worse performance.

Differences in sample characteristics and choice of outcome dependent measures may also result in variability in EF performance. In early studies of EF, participants with autism141, 142 were diagnosed based on narrower criteria that may also have contributed to a more circumscribed level of cognitive functioning. The broadening of the DSM diagnostic criteria1,

168 allowed greater heterogeneity in cognitive function and may be contributing to some of the

Page | 24 null findings156, 157, 161. Some studies that reported no impairment in EF utilised matching criteria based across all three IQ indices (verbal, performance and full scale IQ)165. By comparison, two studies that matched participants on non-verbal IQ only reported significant differences between groups147, 153. Variability in EF performance outcomes has also been observed in studies that have utilised the same EF test, outcome measures (i.e. IED test/stages completed)157, 169 and a neurotypical comparison group. The studies, however, differed on the composition of the ASD cohort. The study by Ozonoff et al169 found differences in individuals diagnosed with Asperger’s syndrome, whereas in both studies, no differences were observed for individuals diagnosed with high functioning autism.

A meta-analysis on cognitive flexibility synthesised extant research and showed that ASD cohorts experience impairment across age groups71. The meta-analysis considered together a number of studies with mixed age cohorts, and was less conclusive on group differences on the developmental trajectory of cognitive flexibility. A broader concept of cognitive flexibility was adopted in the study that included assessments of set-shifting, set-switching and some measures of inhibitory control processes. It reflects on the difficulties inherent in the study of

EF including lack of task purity in EF measures and differences between studies on sample and task characteristics.

Response Inhibition

Studies of response inhibition utilised mostly experimental tasks based on the Go/No-Go100 and Eriksen flanker task101 paradigms. The Stroop132 was the most commonly used psychometric test. Many researchers, however, utilised study specific variations of the above tasks introducing considerable variability between studies. Outcome measures of response inhibition depend on the task used. The Go/No-Go task outcome measures include successful

‘hits’ on Go trials, ‘omission errors’ or mistakes on Go trials and ‘false alarms’ or ‘commission errors’ mistakes on No-Go trials. Most of the studies reviewed below reported mixed results.

Page | 25

Christ et al170 reported no differences between groups utilising the Stroop and Eriksen flanker tests. A dissociation, however, was reported for the Go/No-Go task. Children with ASD were not impaired on error rates on No-Go trials but were impaired on Go trials. In adolescents, studies by Adams and Jarrold171 and Christ et al172 reported no differences for prepotent response inhibition (stop signal tasks) but significant differences for distractor interference

(Eriksen flanker task). In the study by Johnson et al173 differences between ASD and comparison control group depended on the stimulus presentation sequence order, fixed (not significant) or random (significant). Similarly, in adults with ASD, Langen et al174 found significant differences on No-Go error rates and no differences on Go trial error rates. Johnston et al175 reported differences for the Hayling task - these were attributed to the verbal presentation of stimuli - but none for the Stroop test. Significantly impaired performance across all measures of response inhibition was reported in children176 and adolescents177 while conversely, the following studies reported no significant differences for children157, 178-181, adolescents182-184 and adults99, 185, 186.

A recent meta-analysis70 synthesised the literature on response inhibition focusing on the sub- component processes of prepotent response inhibition and interference control. The studies’ participants were mostly children and youth. The overall findings indicated that prepotent response inhibition was significantly impaired in the ASD cohort, impairment however, was attenuated with increase in age. Deficits in interference control, however, persisted across the lifespan.

Working Memory

A similar mixed profile of findings is reported for the working memory (WM) domain. In children, Williams et al187 reported impaired performance in clinical test assessing spatial WM but not in the corresponding verbal WM test. Conversely, in the study Williams et al conducted in adults using the n-back task, the reverse findings were reported188. These may reflect

Page | 26 differences in the type of task used but also different stages of the developmental trajectory of working memory. A study by Cox et al99 utilising a study specific WM ecological task99, found differences dependent on the outcome measure used. Significant deficits were reported in children with ASD utilising a range of verbal189, 190, 191, 192 and spatial WM tasks157. One research reported impaired spatial WM in adolescents193 utilising the CANTAB SWM test. In adults with ASD, impairment was noted on the WAIS Letter-Number sequencing task194, on spatial span and letter sequencing tasks195 and on the WM index from the WAIS test battery.

In a number of studies, no deficits were reported in children when utilising verbal196 or pictorial181, 197 n-back stimuli. Null findings were also reported for the digit span198 from the

WAIS battery and the SWM test197. Urbain et al199 found no differences between an ASD cohort and neurotypical controls on behavioural measures of working memory using a pictorial n-back task. Significant differences were observed, however, in brain synchronisation across frontoparietal areas using a magnetoencephalography paradigm. Adolescents with ASD did not differ from controls on a verbal and spatial WM (assessed with a number counting task and the Corsi spatial tapping backwards task)182 or a mental rotation task84. In adults with ASD, no significant differences were observed on a pictorial n-back face and letter task respectively90,

200 or on the verbal WM index from the WMS test battery200. The above overview reflects the considerable variability in the choice of WM assessment tasks. The observation that significant differences and null findings are reported for the same tasks and for similar age ranges suggests that factors other than task characteristics may be contributing to the disparate findings.

A systematic review201 of working memory in ASD showed mixed results consistent with the research summarised above. A recent meta-analysis202 of working memory primarily in children and young adolescents identified impairment in WM with spatial WM more impaired that verbal WM. There was no effect of age, but this may be because a primarily younger aged cohort was included.

Page | 27

In summary, findings in the extant literature of EF performance in ASD are mixed across development. A number of systematic reviews and meta-analyses attempted to synthesise the

ASD literature on EF to identify common themes. The meta-analyses discussed above evaluated individual domains of EF. While this contributed to current understanding of impairment in ASD on specific EF domains, they do not directly address whether individual

EF domains are differentially impaired in autism.

The conclusions drawn from the above overview is that methodological differences between studies may be contributing to the mixed findings observed on EF in autism. Studies differ in the following ways: selection of the ASD sample156, 157, 160 and comparison groups157, 158, choice of EF domains to be studied, measurement tools115, 148, 153 and outcome dependent measures152,

156, 157. A number of studies also use mixed age samples203-205, potentially masking developmental differences between groups. Other variables that may influence EF research outcomes include presence of other co-morbid conditions (e.g. Attention Deficit Hyperactivity

Disorder - ADHD), other developmental syndromes including various learning disorders, and mental ill health. There has been no systematic evaluation of these or other potential moderators of EF in the ASD literature. Moderators may include but not be limited to level of intellectual functioning206, sample and task characteristics, and differences due to biological sex161, 207. Affective states, and in particular anxiety208 and stress209 have also been shown to influence EF performance in clinical and non-clinical cohorts. The following section provides an overview of potential moderators that may influence EF in Autism Spectrum Disorder.

1.5 Moderators of executive function in Autism Spectrum Disorder

General intellectual functioning

Early studies of EF were based partly on the observation that higher cognitive processes such as planning and concept formation may be impaired despite intact intellectual functioning.

Despite this general observation, there is some empirical support for the concept that

Page | 28 intellectual ability may moderate performance on neuropsychological assessments of executive function210. This is pertinent in the study of ASD where differences between specific indices of intelligence (verbal versus perceptual versus full scale intelligence quotient – VIQ versus

PIQ versus FSIQ) have been reported for the clinical subgroups of Autistic Disorder and

Asperger’s Syndrome206.

Sample characteristics

The sample characteristics may be prescribed in a range of ways by different researchers.

These include ASD sub-type, choice of comparison groups, age, and criteria matching the ASD cohort to the comparison group. The diagnostic criteria for ASD have broadened significantly since the first inclusion of autism in the DSM-III7. In DSM-51 discrete diagnostic categories

(Autistic Disorder, Asperger’s Syndrome) have been merged into a single spectrum. However, much of the earlier research did not always make clear diagnostic distinctions. Many study cohorts were comprised of mixed diagnostic classifications151, 211-213 while some studies included the informal classification of High Functioning Autism (HFA)143. The HFA group comprised of autistic individuals with no intellectual disability, (IQ greater than 70). However, the minimal level of intellectual functioning was arbitrarily selected between studies ranging from borderline157, low average156 to average147. This would have contributed to greater variability in intellectual and executive functioning and potentially led to heterogeneous study outcomes.

Although most studies relied on standardised diagnostic assessments of autism (Autism

Diagnostic Observation Schedule – ADOS, Autism Diagnostic Interview – ADI) and DSM based diagnostic criteria, some studies relied primarily on screening tools147, 214 or earlier classification criteria not directly related drawn from the DSM139. This would also contribute to heterogeneity in intellectual and EF performance and variability in research outcomes.

Selection of comparison control groups also varies between studies and is a further factor for

Page | 29 consideration. Although most studies utilised neurotypical comparison groups other comparisons included to non-affected siblings170, 215 or clinical groups only144, 216, 217.

Task characteristics

Given the broad range of EF domains and measurement tools discussed above, a number of potential moderators need to be considered. These include assessment type (psychometric tests, experimental tasks and behavioural rating scales) and administration mode (traditional versus computerised format). There are some evidence that individuals with ASD perform better on computerised versus traditional administration of EF tests92, 104, although this is not unequivocal149. It does suggest, however, that equivalence between the two formats cannot be assumed. Further, the stimulus and response presentation formats (verbal stimuli versus pictorial stimuli) may be important moderators. This applies particularly, given research support of superior performance individuals with ASD in visuoperceptual tasks requiring attention to detail218.

Sex differences

Autism Spectrum Disorder is a neurodevelopmental condition with a strong male bias, currently at about three males to every one female219. A number of theories have been proposed to explain this difference. These are based on genetic and/or neurobiological differences between males and females as described for example in the imprinted-X liability model220, the male brain theory221 and the female protective effect theory222, 223. There is growing interest in identifying the characteristics that might differentiate autistic male and female individuals including EF performance. Given the strong neurobiological component described above, it may be that neurocognitive processes such as EF may be influencing diagnostic outcomes and observed prevalence rates. Comparisons, however, of male and female subjects with ASD on neuropsychological assessments and self/informant appraisals of EF has been limited. Some research findings224-226 suggest that EF performance in ASD is dependent on sex while others

Page | 30 report no differences227, 228. A potential confounding factor that may explain the mixed findings is that not all studies use a normative comparison group. Thus, sex differences observed in the general population are not always methodologically controlled. Biological sex and inclusion of a neurotypical comparison group are two further factors that need to be considered in autism research.

Affective states

The influence of affective states, especially anxiety, depression and stress as moderators of EF is of particular interest given shared influence on prefrontal cortex circuitry. This has been well documented at least for anxiety208 and stress229. However, empirical research examining the influence of anxiety on EF has shown mixed results. Some studies show that trait anxiety impairs performance on EF tasks230-232, whereas other studies suggest that moderating factors may attenuate performance. Similarly, some studies examining the influence of state anxiety have shown a positive relationship with cognitive performance233, whereas others demonstrated impairment230. Overall, research to date suggests a moderating effect of anxiety on cognitive function in non-clinical samples of high anxious individuals234 with mixed findings reported in clinical populations. In ASD, anxious arousal negatively correlated with test performance on neuropsychological assessments of concept formation34. An empirically derived factor of anxiety correlated with impaired performance on neuropsychological measures of inhibition, mental flexibility and shifting235. Thus, normative literature and research in ASD indicate the relevance of investigating the role of affective states in moderating executive function.

1.6 Efficacy of executive function as a cognitive intermediate phenotype

Endophenotypes or intermediate phenotypes236 are characteristics that represent a vulnerability in a particular population. These characteristics represent a link between genes, brain processes and the observed behavioural phenotype. Traditionally, intermediate phenotypes are

Page | 31 characterised by genetic and biological processes236. A broader definition of endophenotypes has been proposed, encompassing processes - including neurocognitive function237 - that may direct an intervention238. Executive function can be considered an example of a cognitive endophenotype215. It has measurable outcomes, is linked to neurobiological239 and genetic processes236 and may be modified following interventions. For example, functional imaging studies have demonstrated that neuropsychological assessments of EF are linked with specific brain activation areas236 and EF outcomes may be modified by remedial intervention programmes240.

Additional criteria must be met for a particular function to be studied as an endophenotype215.

These include that the endophenotype should be heritable, thus it must cluster within families of affected individuals and it should be observed in non-affected family members at a higher rate than the general population. Literature findings suggest that EF meets the endophenotype criteria for study in autism. First, genetic studies suggest that part of the variance in EF is due to genetic influences68. Further, there is considerable literature support for executive dysfunction in ASD70, 71. Lastly there is empirical support of EF difficulties in relatives of probands with ASD at a higher rate than the general population241. The examination of EF as a potential endophenotype in ASD is important for a number of reasons. Given that EF impairment is observed in neurodevelopmental conditions with genetic overlap with ASD (e.g. schizophrenia), it provides opportunities to explore its utility as a transdiagnostic cognitive endophenotype242 and potential marker for diagnosis. Further, given the observed heterogeneity within ASD243 it may help characterise clusters within the autism spectrum and thus better inform diagnostic and intervention models of autism.

Executive function and diagnostic clusters of Autism Spectrum Disorder

The relationship between EF and diagnostic criteria of ASD has been subject to empirical evaluation with particular focus on the diagnostic cluster of restricted and repetitive

Page | 32 behavioural patterns, and more recently in relation to social cognition and theory of mind.

Early reviews of the literature244, 245 and empirical studies reported a correlation between neuropsychological156, 246, 247 and behavioural measures248 of executive impairment and severity of everyday repetitive behaviours. This relationship was reported for specific EF domains, (i.e. cognitive flexibility, response inhibition and working memory) and conversely, was not influenced by other EF domains such as planning and fluency249. Another study distinguished that EF deficits were specific to repetitive but not restricted behaviour patterns250.

Contradicting these findings, a recent study reported that theory of mind rather than EF measures positively correlated with both repetitive/restrictive behaviours and social communication in youth with Autism Spectrum Disorder32.

Theory of mind5 is one of the prominent cognitive explanations for impaired social cognition in ASD. The theory proposes that impaired ability to attribute mental states to self and others contributes to a range of deficits including those observed in the social communication cluster251. A number of studies, however, suggest that EF moderates theory of mind processes

33, 252 and may thus form a putative link to the social communication phenotype. For example, reduced WM has been linked to impaired social communication skills in ASD, due to impaired ability to maintain social cues during social interactions253. This relationship although not extensively investigated lends support to the broader influence of EF in Autism Spectrum

Disorder.

Executive function and disability in Autism Spectrum Disorder

Social impairment and disability has been linked to EF performance in ASD254, 255 suggesting that EF may be a moderator of disability in cohorts with ASD. A correlational study by Panerai et al256, showed a linear relationship between EF and adaptive functioning in children with

ASD. Specifically, neuropsychological measures of set-shifting, verbal fluency, planning and response inhibition positively correlated with adaptive functioning as assessed by the Vineland

Page | 33

Adaptive Behavioural Scales257. Notably, however, there were no significant differences between the ASD and the neurotypical control group. Interestingly in this study, the behavioural rating measure of EF did not correlate with adaptive functioning despite expectations that an ecological measure of EF will better assess disability.

In a study of children and youth with ASD258 the relationship between carer ratings of EF based on the BRIEF and social functioning - assessed by the Social Responsiveness Scale (SRS)259 - were evaluated. The findings showed that a distinct profile associated with metacognition domains of the BRIEF was unique in predicting social impairment in the ASD cohort. De

Vries et al260 in a study of children aged 8-12 years, reported that executive dysfunction assessed by the BRIEF significantly predicted quality of life, assessed by the Paediatric Quality of Life Index, (PeQoL)261. In a study of youth and young adults with ASD262, neuropsychological measures of response inhibition and set-switching did not predict adaptive functioning, based on parental ratings on the Behavioural Assessment System for Children

(BASC)263. In adults, studies utilising behavioural measures of EF identified mostly a linear relationship between better executive functioning and quality of life. Bishop-Fitzpatrick et al37, using EF informant ratings on the BRIEF, reported that better executive function was associated with better life outcomes. Wallace et al41 reported that difficulties in set-shifting, verbal fluency, planning and response inhibition, predicted adaptive functioning, assessed by the Adaptive Behavioural Assessment System (ABAS-II)264.

The overview of the literature suggests that there have been a limited number of studies examining EF as a predictor of adaptive functioning and disability. Most studies are of children with few investigating adult populations and the majority utilise behavioural ratings of executive function. In these studies, assessment of EF is mostly reliant on informant ratings.

The overview also highlights that a broad range of disability assessments were utilised by different studies. A number of studies use assessments of adaptive functioning and quality of

Page | 34 life (e.g. Vineland Scales, ABAS and PeQoL), with others using specific measures of social impairment (SRS). These methodological differences were highlighted in a recent meta- analysis of life outcomes in adults with ASD as likely contributing to conflicting outcomes265.

The DSM-51 recommends the World Health Organisation Disability Assessment Scale

(WHODAS-II)266 as the best current measure of daily adaptive functioning. To date, however, limited studies have examined the relationship between EF measures and the WHODAS in autism. Studies that have assessed overall disability in individuals with ASD using the

WHODAS indicate that it is an appropriate disability measure for this group267, 268.

In summary, the previous overview highlights the potential value of the study of EF in ASD in identifying specific or broad areas of impairment, in contributing to understanding poor functional outcomes in ASD and informing the study of EF across clinical cohorts with shared symptomatology. These commonalities and influences on EF have contributed to the programme of research described below.

1.7 Programme of research

This programme of research has as its primary focus the study of EF in a cohort of youth and adults with Autism Spectrum Disorder. Within this programme of research, the extant literature on EF in ASD was synthesised in a meta-analysis and four empirical studies assessed the efficacy of EF as an indicator in the study of autism. The focus of the study was on youth and adults with ASD as this cohort has received less attention in the literature. It represents, however, a developmental stage where EF processes may have crystallised and optimal EF performance levels may have been reached. It thus provides an opportunity to identify potential markers that may be modified and contribute to better life outcomes.

Page | 35

Aim of thesis

The overall aim of this thesis is to examine whether EF makes a unique contribution in characterising Autism Spectrum Disorder. The specific objectives of the programme of research were to: synthesise the extant literature on EF in ASD; evaluate potential moderator influences of sex, anxiety and stress on EF; and investigate the efficacy of EF in discriminating

ASD from other clinical cohorts and neurotypical control groups and in predicting disability.

Research methodology

Autism Spectrum Disorder sample

A sequential sample of individuals with ASD presenting for social skills training was recruited into the study. For inclusion in the study, participants required a recent formal diagnosis of

ASD based on standardised diagnostic instruments (ADOS and ADI/ADI-R). In addition, participants were required to be at least sixteen years of age, stabilised on medication for at least eight weeks prior to commencement of study and to have adequate reading and comprehension skills. Potential participants were excluded from the study if they had, intellectual disability (IQ<70), current clinical levels of substance or alcohol use, experienced active psychotic symptoms, indicated current suicide risk or reported a history of head injury with loss of consciousness greater than 30 minutes. All participants completed standardised self-report screening instruments to assess levels of depression, social anxiety, general anxiety and stress.

Comparison samples

The ASD cohort was compared to two clinical cohorts and a neurotypical comparison group.

The inclusion /exclusion criteria described below applied to each of the three comparison groups.

Page | 36

Comparison group participants were required to be at least sixteen years of age, stabilised on medication for at least eight weeks prior to commencement of study and to have adequate reading and comprehension skills. Potential participants were excluded from the study if they had, intellectual disability (IQ<70), current clinical levels of substance or alcohol use, experienced active psychotic symptoms, indicated current suicide risk or reported a history of head injury with loss of consciousness greater than 30 minutes.

In addition, the following criteria applied for each of the three comparison groups:

Social Anxiety Disorder (SAD) – for study inclusion, participants were required to have a formal diagnosis of SAD based on the Anxiety Disorders Interview Schedule (ADIS-4 and

ADIS-5).

Early Psychosis (EP) – for study inclusion, participants were required to have a formal diagnosis of symptoms associated with EP based on diagnostic assessments on the Structured

Clinical Interview for DSM-IV (SCID-IV), the Scale for the Assessment of Positive Symptoms

(SAPS) and the Scale for the Assessment of Negative Symptoms (SANS).

Neurotypical control group – for study inclusion participants were required to have no past or current formalcurrent formal diagnosis of a psychiatric condition and must have scored below clinical cut-off scores on screening tools for ASD and SAD.

A sequential sample of individuals with ASD presenting for social skills training was recruited into the study. For inclusion in the study participants required a formal diagnosis of ASD based on standardised diagnostic instruments (ADOS and ADI). Exclusion criteria were intellectual disability (IQ<70) and current clinical levels of substance or alcohol use.

Page | 37

Assessment battery

A research test protocol developed for use in the Autism Clinic for Translational Research

(ACTr) and headspace clinics was submitted and approved by the University of Sydney Ethics

Committee (Protocol number 2013/352). The research protocol consisted of diagnostic assessments, an assessment of pre-morbid general intellectual functioning, neuropsychological tests and behavioural ratings of executive and cognitive function, and self-appraisals of symptom severity and disability. A detailed summary of assessment measures is presented in

Appendix 5, Table S5.1.

The EF test battery comprised neuropsychological assessments of the domains of set-shifting, set-switching, phonemic fluency and semantic fluency. Neuropsychological measures of attention and psychomotor speed were also included as they contribute to EF models and underpinning mechanisms. The behavioural rating scale assessed nine domains of EF: inhibit, shift, emotional control, self-monitor, initiate, working memory, plan/organise, task monitor and organisation of materials.

The inclusion of EF domains was informed in part by the findings of the meta-analysis

(empirical study 1), literature findings on EF and study-specific considerations. Set-shifting is consistently identified as an area of impairment in ASD and was included in the study. Set- switching represents a different EF process that is not always distinguished between studies.

Its inclusion provides an opportunity to examine whether it differentially influences EF performance in individuals with Autism Spectrum Disorder. Distinguishing between phonemic and semantic fluency is also of interest in ASD given that the former has been linked to perseverative behaviours. Neuropsychological measures of psychomotor speed and sustained attention complemented the EF test battery given that attention processes and processing speed feature in EF models. Performance measures of response inhibition were not included, based on findings of the unity in diversity model where response inhibition was

Page | 38 identified as a common shared factor. Response inhibition and working memory were assessed by behavioural ratings.

Empirical studies

Chapter 2: Autism Spectrum Disorder: a meta-analysis of executive function

The review of literature findings on EF in ASD indicated considerable variability between studies, partly due to conceptual and methodological differences. The primary aim of this study was to synthesise the extant literature on EF in ASD in a meta-analysis and evaluate the differential contribution of distinct domains and moderator variables.

Chapter 3: Sex differences in executive function in Autism Spectrum Disorder

Significant sex differences are observed in ASD in prevalence rate and age of initial diagnosis.

Differences in EF performance may be one factor that contributes to this variability. The primary aim of this study was to investigate sex differences in EF, while controlling for differences observed on the normative population by inclusion of a neurotypical comparison group.

Chapter 4: Affective states as moderators of executive function in Autism Spectrum Disorder

Affective states, in particular anxiety and stress, are reported to moderate executive function outcomes. The primary aim of this study was to examine whether EF may be moderated by depression, anxiety and stress and whether affective states have a differential impact on EF performance in the ASD cohort compared to other comparison groups.

Chapter 5: Autism, Early Psychosis, and Social Anxiety Disorder: a transdiagnostic examination of executive function cognitive circuitry and contribution to disability

Poor life outcomes and disability are commonly reported in ASD; however, the influence of

EF in predicting and mediating these has received less attention. This is particularly true in the youth and adult population. The aims of this study were to ascertain the role of performance

Page | 39 and behavioural measures of EF in differentiating between ASD, clinical and neurotypical comparison groups, and in predicting disability.

Chapter 6: Machine learning for differential diagnosis between clinical disorders with social impairment

The aim of this study was to examine whether EF makes a unique contribution in discriminating between ASD and other clinical and control cohorts when considered together with an extensive battery of measures across multiple cognitive domains.

Chapter 7: General discussion

The findings of the studies presented in this thesis are synthesised in Chapter 7. Conclusions from individual studies and their contribution in informing the study of EF are reviewed and methodological limitations are addressed. Finally, the study’s contribution in advancing the knowledge of EF is discussed within a research framework for the study of EF in Autism

Spectrum Disorder.

Page | 40

Chapter 2: Autism Spectrum Disorders: A meta-analysis of executive function

Study objectives:

The study objectives were:

(a) Synthesise the EF literature since the first introduction of autism as a diagnostic

category in the DSM with a focus on studies comparing ASD to a neurotypical control

group;

(b) Quantify differences between the comparison groups on overall EF and core EF

domains;

(c) Assess the influence of moderating variables on sample and task characteristics and;

(d) Quantify the clinical sensitivity and specificity of individual EF measures

Page | 41

Autism Spectrum Disorders: A meta-analysis of executive function

Demetriou, E. A., Lampit, A., Quintana, D. S., Naismith, S. L., Song, Y. J. C., Pye, J. E.,

Hickie, I.B., Guastella, A. J. (2018). Molecular Psychiatry, 23(5), 1198–1204.

http://doi.org/10.1038/mp.2017.75

See Appendix 1 for published version

Page | 42

Abstract

Evidence of executive dysfunction in Autism Spectrum Disorders (ASD) across development remains mixed and establishing its role is critical for guiding diagnosis and intervention. The primary objectives of this meta-analysis is to analyse executive function (EF) performance in

ASD, the fractionation across EF sub-domains, the clinical utility of EF measures and the influence of multiple moderators (e.g. age, gender, diagnosis, measure characteristics). The

Embase, Medline and PsychINFO databases were searched to identify peer reviewed studies published since inclusion of Autism in DSM-III (1980) up to end June 2016 which compared

EF in ASD to neurotypical controls. The random-effects model was used and moderators were tested using subgroup analysis. The primary outcome measure was Hedges’ g effect size for

EF and moderator factors. Clinical sensitivity was determined by overlap percentage statistic

(OL%). Results were reported according to the PRISMA guidelines. Two-hundred and thirty- five studies comprising 14181 participants were included (N, ASD =6816, Control =7265). A moderate overall effect size for reduced EF (Hedges’ g=0.48, CI 95% 0.43 - 0.53) was found with similar effect sizes across each domain. The majority of moderator comparisons were not significant. Only a small number of EF measures achieved clinical sensitivity. This study confirms a broad executive dysfunction in ASD that is relatively stable across development.

The fractionation of executive dysfunction into individual sub-domains was not supported, nor was diagnostic sensitivity. Development of feasible EF measures focusing on clinical sensitivity for diagnosis and treatment studies should be a priority.

Page | 43

2.1 Introduction

Autism Spectrum Disorder (ASD) is a neurodevelopmental condition defined by deficits in social communication and interaction and restricted and repetitive patterns of behaviour269.

Although genetic and neurobiological factors contribute to the ASD phenotype, neurocognitive functions also play an important role in the core behaviours of ASD. Executive function (EF) has long been of interest given its proposed role in contributing to specific impairments in ASD in the areas of theory of mind270 and social cognition, social impairment271, restricted and repetitive behaviour patterns272, as well as broader impacts on quality of life273. Executive function encompasses a broad range of purposeful higher-order neuropsychological domains, including goal directed behaviour, abstract reasoning, decision making and social regulation50.

It is generally accepted that EF difficulties have an important role in ASD, described as poor regional co-ordination and integration of pre frontal executive processes which integrate with other emotion and social circuits274. In ASD, brain abnormalities have been observed in cortical volume and thickness in both frontal and other cortical brain regions275. Aberrant functional network connectivity influencing EF have also been reported between prefrontal and other cortical and sub-cortical areas276 which may be influenced by different EF subdomains. A summary of key EF domains, associated brain areas and related ASD phenotype is presented in Appendix 2, Table S2.1. Figure 2.1 illustrates developmental changes in EF and observed impairment in ASD.

Despite extensive research, however, including a number of meta-analyses277, 278 and reviews201, 279 the role of EF in ASD remains unclear. Individual research studies place different emphasis on the EF constructs of interest and few studies evaluate EF across most of the accepted domains of interest. The identification of a cognitive profile of executive dysfunction could provide a better understanding of the neural circuitry underpinning ASD, may assist with clinical utility, diagnosis and treatment. Previous published ASD meta-analyses and systematic

reviews of EF focus on one or two specific subdomains and thus an overall framework of the executive dysfunction profile in ASD has not been established.

Other factors may contribute to these mixed findings. Studies inconsistently control for potential moderators of the relationship between EF and ASD and the observed high inter- individual cognitive variability within the spectrum280. Moderators considered in ASD studies include variables that impact on sample selection or task characteristics and may influence the observed relationship between executive dysfunction and ASD. Selection of the ASD and comparison samples varies between studies depending on the ASD classification(s) of interest and choice of a clinical or typical comparison group (or a combination). Matching criteria between ASD and comparison groups also vary and may be based on a range of variables including cognitive measures (different IQ indices), age (chronological/mental age) and gender. Sample characteristics including age and gender may moderate EF performance given that EF domains may follow a differential developmental trajectory in the typically developing brain281, 282 and those with ASD283. However, many studies do not examine developmental trajectories in ASD and utilise mixed age cohorts205, 284 making outcomes on EF performance difficult to interpret. Finally, task characteristics may vary between studies on a range of variables including assessment type (psychometric tests vs. experimental tasks), features of presented stimulus (verbal vs. visuospatial), presentation format (computerised vs. traditional) and the response type required from the participant (verbal or motor response). Yet these potential moderators have not been systematically studied285.

The observed inter-individual cognitive variability within the spectrum and differences in EF performance that may be differentially modulated by distinct EF domains (and associated brain areas) and/or different mediating factors may translate to the need for a more individualised approach for diagnostic measures and clinical interventions. Thus, research is needed to

Page | 45 explore the clinical utility of group-based EF measures (based on standardised psychometric tests and experimental tasks) in discriminating between ASD and comparison typical populations286.

The objectives of the study were: a) to examine evidence for executive dysfunction in ASD including the individual contribution of EF subdomains; b) to assess the influence of moderating variables based on sample or task characteristics and c) to review the clinical sensitivity of individual EF measures. We hypothesised that overall EF will be impaired in

ASD, individual EF subdomains will make a differential contribution to executive dysfunction and this will be correlated with improved clinical sensitivity in associated behavioural and informant EF measures. An exploratory approach was taken for reviewing moderator impact and no specific hypotheses were made.

2.2 Methods

The PRISMA287 and MOOSE288 guidelines (Appendix 2, Table S2.5 and Figure S2.1) were followed in conducting this study.

2.2.1 Study Selection

Included in the meta-analysis were studies published in peer reviewed journals in the English language with an a priori aim to assess EF in ASD. The selected publication date was between

1980 (first inclusion of Autism diagnosis in the DSM-III289) and end June 2016. The majority of selected studies utilised a cross-sectional design. For data extracted from clinical trials or longitudinal study designs, only the baseline data was included in the meta-analysis. Eligible studies included participants with a diagnosis of ASD (based on DSM or ICD classifications) and/or a diagnosis of ASD based on structured and validated diagnostic instruments (Autism

Diagnostic Observations Schedule – ADOS, and/or the Autism Diagnostic Interview - ADI).

Given that the search period ranged from 1980 - 2016, the diagnostic criteria for the selected

Page | 46 studies varied depending on the edition of DSM and ICD publications. Studies with participants less than six years of age were excluded from the meta-analysis to account for the qualitative differences in the types of assessment instruments used in younger aged groups290. Eligible studies evaluated one or more of six key EF domains, (Concept Formation/Set-shifting, Mental

Flexibility/Set Switching, Fluency, Planning, Response Inhibition and Working Memory,

(refer Appendix 2, supplementary Table S2.1). These EF domains were selected as they have been widely investigated in the ASD literature.

2.2.2 Search strategy and study variables

The literature search was conducted on the computerised databases of Medline, EMBASE and

PsycINFO using search criteria based on EF domains and measures of interest. The first author

(EAD) screened search results for initial eligibility based on title and abstract. Full text versions of the potentially eligible studies were then assessed and included if satisfying the selection criteria. Coding of individual outcomes into EF domains was done by the first author based on accepted neuropsychological categorisation50, 291 and verified by a second independent reviewer (JEP). Reported outcomes were extracted as means and standard deviations (SDs), F- test value or t-test value for each group at a single time point. In order to avoid selective data extraction, when studies included more than one measure of EF (either within the same domain or for more than one domain), all relevant outcomes were extracted. This was based on the assumption that within assessment measures are at least moderately correlated, and to avoid selective data reporting.

2.2.3 Moderator analysis:

Age: A stratified approach (based on the mean age reported in the study plus or minus one SD) was utilised to categorise each study in one of the following age categories: “Children≤12”,

“Youth >12<18”, “Adults≥18”, “Mixed age<18 years” and “Mixed age≥18”.

Page | 47

Gender: A comparison between studies which included female or male participants only.

Diagnostic group: Participants were grouped based on their study classification (Autism

Diagnosis, Asperger, or ASD Combined (including a combination of two or more of the above classifications).

Control type: A comparison between studies utilising neurotypical controls vs. sibling controls.

Diagnostic Tool: Studies were classified based on the assessment tool(s) utilised for the diagnosis. These may have included one or more of the following, DSM, ICD (International

Classification of Diseases), ADOS, ADI.

Sample Matching Criteria: A comparison between studies which used one or more matching criteria for sample selection.

IQ differences: A comparison based on whether a significant IQ difference was observed between the study groups.

Assessment tool format: A comparison between computer vs. traditional administration of assessments.

Stimulus Processing Mode: A comparison based on the presentation features of test stimuli, verbal vs. non-verbal.

Response Mode: A comparison based on the response mode required from the participants, verbal vs. Motor.

Study Appraisal and Risk of Bias in Individual Studies

Quality review was based on the Quality Assessment Tool292 and was completed by two independent assessors (see acknowledgements in published study), not involved in any other aspects of the study. To assess risk of publication bias, funnel plots for overall outcomes as well as for each cognitive domain were inspected for asymmetry and formally assessed using

Egger’s regression test.

Page | 48

2.2.4 Data analysis

Data analysis was performed on Comprehensive Meta-analysis (CMA) version 3 (Biostat, NJ) using the random-effects model. The unit of analysis was standardised mean difference (SMD, calculated as Hedges’ g) on each measure between ASD and healthy controls. When more than one control group was reported in the study, the control groups were combined following established statistical procedures293. A positive effect size indicated that the control group performed better on the EF measure compared to the ASD group.

The data analysis was planned a priori and was completed in three stages. The initial analysis combined all EF outcomes to assess the overall EF effect size in ASD. The second analysis examined subgroup comparison of the individual EF domains. In the final step, subgroup analyses were conducted to examine between study variability and moderator impact.

Hedges’ g effect sizes ≤0.30, >0.30 and <0.60 and ≥0.60 are described as small, moderate or large following the same convention applied to Cohen’s d effect sizes. Heterogeneity across studies was assessed using the I2 statistic with 95% confidence intervals. I2 values of 25%, 50% and 75% define small, moderate and large heterogeneity. Between-subgroup heterogeneity was tested using the Cochrane's Q statistic.

Clinical sensitivity was determined by the overlap percentage statistic (OL%) based on

Cohen’s294 idealised distributions. This can be converted to a percentage representing the degree that the performance of the ASD group overlaps with the control group, OL< 15% represents clinical marker criteria.

2.3 Results

The literature search resulted in 235 studies (see Appendix 2, Table S2.2) which satisfied the selection criteria with a total of 14081 participants (ASD: N = 6816, Control: N =7265).

Page | 49

2.3.1 Overall effect of executive function:

The overall effect of EF was large and statistically significant, (k=235, g=0.60, 95% CI [0.53-

0.67], p<0.001). True heterogeneity across studies was large (I2=75.5%). The forest plot revealed that studies including results based on self or carer-reported ratings had higher effect sizes with the majority of results ranging from large to very large (0.64

A statistical comparison of “Assessment Tool Type” revealed significant differences between the different tool types (psychometric test vs experimental task vs questionnaire) with the questionnaire format having the largest effect size (g=1.84, 95% CI [1.48 to 2.20], p<0.001).

Comparably, true heterogeneity was highest for the questionnaire format (I2=89.6) compared to experimental tasks (I2=56.7) and psychometric tests (I2=50.4). It is of note that most studies including self/informant reports used the Behavioural Rating Inventory of Executive Function

– BRIEF76. A sensitivity analysis revealed that by excluding questionnaire outcomes, the revised effect size was moderate (g=0.49, 95% CI [0.44 to 0.55], p<0.001). Homogeneity was similarly reduced to I2=54.2%. Given the above results, the questionnaire data was excluded from the remainder of the meta-analysis and results are reported based on dependent measures assessed by psychometric tests and/or experimental tasks only.

The forest plot revealed two conspicuous outliers with both g>5295, 296. Following removal of the two outliers there was a marginal reduction in effect size (k=221, g=0.48, CI 95% [0.43.to

0.53], p<0.001) and heterogeneity was comparably reduced (I2=46.2%). The funnel plot suggested evidence of small study effect (Egger’s intercept = 1.21, p<0.001). A trim and fill

Page | 50 analysis however did not result in imputation of any studies and the overall effect size remained the same.

2.3.2 EF domain specific effects:

Small to moderate effect sizes were observed for each of the EF domains of interest (see Figure

2.2 and Appendix 2, Table S2.3). The subgroup analysis between EF domains was not significant (p>0.05).

Moderator analysis: Figure 2.3 summarises the EF subdomain analysis by age. A detailed summary of subgroup analysis for moderator effects is presented in Appendix 2, Table S2.3.

All within subgroup analyses on moderator effects were significant with effect sizes ranging from small-moderate to large, however the majority of the between subgroup analyses assessing moderator impact were not significant. Significant between group effects were observed for the subgroup age comparison for the Working Memory domain. This was driven by lack of significant difference between ASD and controls for the Youth age grouping. The subgroup comparison between computer vs. traditional assessment format was also significant with presentation of tasks by computer having an attenuating effect on EF.

2.3.3 Clinical specificity and sensitivity

Only a very limited number of measures achieved the criterion of clinical sensitivity as defined by Cohen’s ‘idealised population distribution’. The majority of the measures reaching clinical sensitivity were based on the BRIEF76 questionnaire (see Appendix 2, Table S2.4).

2.4 Discussion

The meta-analysis extracted all EF data since ASD was introduced as a psychiatric diagnosis and showed consistent evidence of an overall moderate effect size of executive dysfunction in

ASD. Individuals with a diagnosis of ASD performed on average significantly worse on EF in

Page | 51 comparison to neurotypical controls. However, contrary to our prediction that individual EF subdomains would be differentially impaired, no significant differences in effect sizes were observed between these. Moderate effect sizes were observed for all of the established individual EF subdomains of interest. These findings suggest that there is relative equivalence of EF impairments in ASD across the constructs that were examined. This was further supported in this study by the largely homogeneous impact of most moderators on EF outcomes.

These findings are consistent with the largely linear trajectory observed in the development of

EF in ASD (see Appendix2, Table S2.1) and recent trends in ASD research focusing on aberrant brain connectivity in predicting cognitive deficits and symptom severity in ASD297,

298. A global impairment due to either under or over connectivity between brain networks broadly contributing to EF, as opposed to discrete anatomical deficits, could account for the lack of differences between subdomains of EF. The age comparison was only significant for the working memory domain. The age effect for working memory may relate to the developmental trajectory reported for some EF in neurotypical populations where performance may decline around puberty due to synapse re-organisation299. Thus the lack of significant difference in working memory observed in adolescents may reflect the underlying neural changes observed in this developmental period that may contribute to a decrease in EF performance in typically developing individuals. Differences therefore in this subdomain between the two groups may be least pronounced in this age range.

The generally smaller effect sizes on EF observed for the adult ASD group support other research that, either due to developmental maturity and/or increased use of compensatory strategies, adults with ASD perform better in EF than younger age groups while residual executive dysfunction still persists. In addition, a smaller effect size between ASD and controls

Page | 52 was observed when only one of the ADOS or ADI was used for diagnosis. Given the variability across ASD and recommendations for multifactorial assessment, use of a single diagnostic tool may result in a less severe cohort meeting much broader ASD criteria.

It was hypothesised that the different diagnostic classifications of ASD may introduce variability between studies of EF, because of potential heterogeneity in cognitive function reflecting different classifications. However, our results failed to find differences in effect sizes between different diagnostic groups. This lends support to the recent focus on individual variability within the spectrum rather than between classification groups guiding EF outcomes28.

Similarly, an evaluation of the potential differences between different matching criteria did not reach significance although the largest effect size was observed for matching based on chronological age. Differences in EF are likely to be more pronounced between experimental and comparison groups within the same age range when no other moderators such as IQ or mental age are taken into account.

Our findings on the clinical utility of EF measures show that the majority of EF measures did not achieve clinical utility in differentiating between ASD and typical controls, with mostly informant based measures based on the BRIEF76 achieving absolute clinical marker criteria.

This lends further support to the proposition that measures with ecological validity (i.e. based on more representative environmental situations) may be more appropriate especially in clinical practice. Informant measures such as the BRIEF, may offer greater clinical utility, however further investigations are needed to consider whether outcomes represents higher validity or might be influenced by demand or reporter characteristics. Based on the results of this study, however, the superior ecological validity of informant measures286 suggests their use for circuitry based models (RDoC)300 and clinical staging models301 (matching

Page | 53 developmental stage of impairment with clinical intervention and risk factors302) and for diagnostic and intervention frameworks. In addition, laboratory based EF neuropsychological tests should be chosen based on feasibility and ease of use, given the relative equivalence of performance across domains. Taken together, these findings suggest that the focus of diagnostic and intervention measures needs to shift to a more ecologically and clinically valid framework while taking into account the likely individual differences within the spectrum.

A number of limitations may have influenced the findings of this study. The self- or informant- report questionnaires were excluded from the majority of analyses given the significant differences in effect sizes compared against psychometric tests and experimental tasks. In addition, only accuracy-based measures were included in the analysis with all reaction time variables excluded given the direction and specification of the reaction time variable to EF can be unclear. Furthermore, we did not explore the impact that intra-individual variability within the spectrum may have on the observed findings. Finally, although we attempted to consider a comprehensive number of moderators, there remain a number of factors that may influence EF in ASD. These may relate to task characteristics (e.g. task complexity, open ended vs. structured task format285) or participant characteristic including symptom severity, emotional states (e.g. depression/anxiety) or co-morbidities (e.g. Attention Deficit Hyperactivity

Disorder). Anxiety in particular has been noted to have a strong association with poor EF performance in ASD populations303.

Page | 54

2.4.1 Conclusion and further directions

In this meta-analysis we conducted an evaluation of the role of EF in ASD including an assessment of a large range of potential moderators. Our findings across a large number of research participants with ASD suggest there is an overall effect of executive dysfunction and this applies evenly across individual domains, where moderate effect sizes were observed.

Predictions of a differential profile of executive dysfunction based on subdomain performance were not supported and our review of moderators mostly returned null results. Taken together, these findings suggest that ASD populations are impaired in EF but this reflects an overall and not fractionated impairment in EF performance and may be best accounted by the observed aberrant long range and local over and under connectivity between brain networks in ASD.

Further work is needed to identify feasible and sensitive EF general markers for use in diagnosis and clinical trials. Given the stability of EF performance in ASD across neurodevelopment, early intervention may provide the best opportunity to alter trajectories over the lifetime to improve outcomes for people with ASD.

Page | 55

Figure 2.1: Developmental changes in executive function and associated impairment in Autism Spectrum Disorder

Page | 56

Figure 2.2: Subgroup analysis of executive function domains

Page | 57

Figure 2.3: Subgroup analysis executive function domains by age

Page | 58

Chapter 3: Sex differences in executive function in Autism Spectrum Disorder

Study objectives:

Empirical study 1 reviewed a number of moderator variables most of which did not influence

EF performance. The objectives of empirical study 2 are to:

(a) investigate sex differences in performance and behavioural measures of EF in

individuals with autism spectrum disorder

(b) investigate whether normative sex differences influence EF outcomes in Autism

Spectrum Disorder

Page | 59

Autism Spectrum Disorder: An examination of sex differences in neuropsychological

and self-report measures of cognitive and executive function

Demetriou, E.A., Pepper, K.L., Park, S.H., Pellicano, L., Naismith, S.L., Song, Y. J. C.,

Hickie, I.B., Thomas, E.E., Guastella, A.J.

(under review in the ‘Journal of Autism and Developmental Disorders’).

Page | 60

Abstract

Autism spectrum disorders (ASD) have higher prevalence amongst males than females. Sex differences in ASD may in part be understood, by an atypical sex profile of cognitive and executive function (EF). In this study, we compared females and males with ASD against neurotypical individuals (TYP) on neuropsychological and self-report measures to examine if any sex differences in cognitive and EF might be unique to ASD. Our study showed a significant overall female advantage for measures of psychomotor speed, cognitive flexibility, verbal learning and memory and semantic fluency. No sex differences were observed on the self-report measure of EF. Our results suggest that while females show different cognitive performance to males, these sex differences were not specific to the ASD cohort.

Page | 61

3.1 Introduction

Autism spectrum disorder is a neurodevelopmental condition with a strong male bias, currently at about three males to every one female219. This sex bias may be due to a range of factors.

These include genetic and/or neurobiological differences as for example described in the

‘imprinted-X liability model’220, ‘the male brain theory’221 and the ‘female protective effect theory’ 222, 223. These aetiological factors may thus contribute to the observed differences in the symptom profile of ASD females where diagnosis is generally made later in life222, 304 and is characterised by more severe symptomatology with significant co-morbidities222, 305.

Alternatively, difference may be due to diagnostic or clinician biases in the characterisation of

ASD306, sex-related sociocultural and environmental moderators307 and gender roles308 or, methodological differences in study design309. Sex differences have also been reported on informant ratings of ASD symptomatology310, adaptive behaviour and EF311, however the factors contributing to these differences are still undetermined.

To better understand the aetiology of sex differences, there has been growing interest in identifying the characteristics that might differentiate autistic males and females including neurocognitive performance. However, despite extensive research on group differences in

ASD, comparison of sex differences on neuropsychological assessments and self/informant appraisals of cognitive function has been limited. Examination of neurocognitive performance is particularly pertinent given the reported relationship between neuropsychological measures and ASD aetiology such as theory of mind252. Further, sex differences in the subjective self- assessments of ASD may inform on underlying processes guiding this and provide an additional framework to understand the factors that lead to a higher male prevalence in autism.

Broadly, the study of cognition is subsumed under discrete cognitive functions and a distinction is generally made between executive and non-executive function domains. Executive function

(EF) refers to the capacity to respond adaptively and engage in goal directed, purposeful

Page | 62

behaviours21. Core executive functions include mental set-shifting, working memory and response inhibition67, which contribute to higher order EF domains including fluency/generativity, concept formation and planning25. Together with EF, the current edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM)1 defines five non EF domains - complex attention, learning and memory, visuo-perceptual functions, language and social cognition - as the key neurocognitive domains important in normative behaviour and psychopathology. In ASD there is a well established pattern for overall impairment in EF compared to neurotypical populations27 however research on sex differences in ASD is limited.

In terms of cognition, most sex-related research in ASD has focused on neuropsychological assessments of children, with fewer studies assessing cognitive sex differences in adults, where overall the research findings are equivocal. In a cohort of ASD participants under the age of

18, Frazier and colleagues showed that females had significantly lower verbal/non-verbal and overall IQ in comparison to males223. Although these findings are somewhat disputed as they represent a very specific genetic and geographical cohort and thus may be subject to ascertainment bias312 comparable outcomes for verbal IQ were reported in a second study of children with ASD recruited from the general community313. Another study of autistic children aged 7-14 years, showed that females perform poorly on cognitive flexibility measures in comparison to males225. In autistic adult males and females226 it was shown that females performed better on verbal fluency tests of executive function whereas males showed superior verbal abilities as assessed by verbal intelligence quotient (IQ) tasks. In contrast, Lai and colleagues did not find any sex differences across any IQ indices or on a response inhibition task227. One possible reason for the mixed findings in the ASD literature is that some studies have failed to include a gender matched neurotypical comparison group. For example, cognitive sex differences in three-year-old children with ASD314 dissipated once they were compared to a comparison group of neurotypical children of similar age and intellectual ability.

Page | 63

In a study of adults aged 19 and over which included comparisons to neurotypical adults228, no significant main or interaction effects for gender and diagnosis were observed for the EF domains of planning, mental flexibility and working memory. Two other studies however, reported a significant sex with diagnosis interaction effect for EF measures of response inhibition207 and cognitive flexibility161 . Overall, there is emerging evidence that females might show lower IQ and poorer verbal abilities, while there are mixed findings regarding executive function skills.

In contrast to the mixed ASD findings, cognitive sex differences in the normative population traditionally report a female advantage for verbal domains and a male advantage for visuospatial tasks309, 315. This finding has been confirmed in recent meta-analyses316, although the causal explanations of such differences and the research itself is still debated. For example, contrary to the traditional view of biological heritability of sex differences, it has been proposed that ‘inherited’ socio-environmental factors have a pivotal role in maintaining sex differences with biological sex contributing to individual ‘intra-generational’ variability317. Potential biases however, may also influence the study of normative sex differences and include gender socialisation roles308 or methodological differences in study design such as type of assessment measures and task characteristics308, 318, 319. Given the observation of sex differences in the normative population, reported sex differences in ASD may be magnified in the absence of a comparison neurotypical group.

Consistent with the above observation, a recent meta-analysis of thirteen studies which included comparisons to neurotypical individuals320 did not identify a distinct cognitive sex profile in autism. Specifically, no sex differences were identified for IQ (effect size synthesized on 13 studies) or for the key ASD domains of social communication (synthesized on five studies) and restricted and repetitive patterns and interests (synthesized on four studies). In

Page | 64

their systematic review, the authors320 reported trends for sex differences for cognitive flexibility and response inhibition (female advantage) and a visuospatial task (male advantage).

In addition to neuropsychological assessments, subjective ratings of cognitive ability have been of increasing interest in the autism field particularly as they may represent a more ecologically- valid assessment of cognitive difficulties41, 321. Again, the study of sex differences has been limited. A study which utilised informant ratings311 found that parents rated females to have significantly higher levels of executive difficulties despite performing at comparable levels of intellectual functioning as males. A second study310 reported a trend for girls to be rated by parents as more impaired on planning and organising ability but did not identify any other sex differences on informant ratings of EF. These sex differences may be explained by the more severe symptomatology observed in females with ASD304 but may also reflect biases on the expected behaviours of females compared to males.

In summary, extant research on sex differences in autism is limited with only few studies focusing on potential underlying factors such as neurocognitive functioning that may contribute to the behavioural phenotype. Methodological differences in the autism literature of cognitive sex differences limit generalisation of findings. The goal of this study was to assess for sex differences in autism across a range of cognitive and EF domains and evaluate whether they contribute to an atypical cognitive profile in autism. We aimed to address limitations of previous research by including a neurotypical comparison group, selecting a homogeneous age cohort over the age of sixteen to control for potential sex-related developmental differences and by utilising measures across both verbal and visuospatial domains of cognitive and EF.

Our measures consisted of both neuropsychological assessments and self-evaluations of cognitive and EF, the latter based on the Behavioural Rating Inventory of Executive Function

– BRIEF103. Given the mixed findings in the extant literature, we based our hypotheses on the overall findings of the meta-analysis and systematic review by Hull and colleagues.

Page | 65

First, our primary hypothesis was that we should see a significant interaction effect for sex by diagnosis for our measures of psychomotor speed and cognitive flexibility. Second, aligned with trends in the typical population, we expect an overall female advantage on neuropsychological measures of verbal domains and a male advantage for visuospatial tasks.

Third, consistent with normative findings for the EF self-report measure103, we expect a significant gender effect for the BRIEF clinical scales of ‘emotional control’, i.e. the capacity to regulate and modulate emotional responses and ‘initiate’, i.e. the capacity to commence a task and engage in appropriate problem solving strategies103. A male advantage was reported for the former and a female advantage for the latter scale.

3.2 Method

3.2.1 Participants

The University of Sydney Ethics Committee approved the research protocol (No: 2013/352) for this project. Informed consent was obtained directly from each participant prior to inclusion in the study. A cohort of youth and adults (N= 128; Age: M = 24.8, SD = 6.6, 41.4% female) were recruited for the study. The Autism Spectrum Disorder group (ASD: N = 69, Age: M =

24.4, SD = 7.5, 36.2% female) was a convenience sample presenting for treatment and/or social skills development at the Autism Clinic for Translational Research (ACTr) and headspace

Brain and Mind Centre clinics. Neurotypical comparison participants (TYP; N = 59, Age: M

= 25.3, SD = 5.2, 48.1% female) were recruited through advertising at a number of different university websites. All participants were screened and excluded from the study if they had intellectual disability (IQ<70) or reported current substance dependence. Neurotypical participants were excluded if they reported a mental health diagnosis (past or current) or were currently experiencing significant levels of depression or anxiety as measured by standardised instruments. Participants met criteria for the primary presenting disorder of ASD based on

Page | 66

clinical diagnoses made by qualified clinicians at the ACTr, on standardised clinical interviews and diagnostic assessment instruments.

The assessment battery comprised of a combination of diagnostic assessments, neuropsychological tests and self-report measures of cognitive and EF. Trained postgraduate research students conducted all assessments in person over two sessions at the ACTr clinic.

Participants were individually tested in dedicated test rooms to ensure minimal distraction and appropriate environmental conditions (lighting/sound). All participants completed the Social

Responsiveness Scale (SRS)259. The ASD diagnosis was confirmed following completion of the on the Autism Diagnostic Observation Schedule – 2nd edition (ADOS-2)322.

The neuropsychological test battery comprised of assessments of pre-morbid IQ based on the

Wechsler Test of Adult Reading (WTAR)323 and assessments of non-EF and EF. Non-EF measures consisted of tests of psychomotor speed, visuospatial and verbal learning and short- term memory (STM). EF measures assessed the domains of mental flexibility, sustained attention, attentional shifting and phonetic and semantic fluency (see Figure 3.1 and Appendix

3, Table S3.1 for details of specific measures). The self-report ratings of EF were based on the

BRIEF103, refer Appendix 3, Table S3.1.

3.2.2 Data analysis

Data analysis was performed using the IBM SPSS Statistics version 24. Univariate (ANOVA) examined for differences across the demographics of Age, intellectual functioning (IQ) and

Years of Education (Education). The Pearson Chi Square Test Statistic of Independence, (χ2), assessed sex representation across the diagnostic groups. Multivariate Analysis of Variance

(MANOVA) examined main effect differences for diagnosis, sex and the interaction of diagnosis with sex, across all measures of cognitive and executive function. The raw scores

Page | 67

attained in each measure were utilised in all statistical analyses. As this was an exploratory study, significant findings are reported for p<0.05.

3.3 Results

3.3.1 Demographics

Results are summarised in Table 3.1. There were no significant differences in the sex representation between diagnostic groups, χ2=1.65, p=0.199. A significant difference for

Education was observed, F(1, 126)=22.89, p<0.001, with the ASD group (M=12.54, SD=2.05) having significantly lower level of education compared to the TYP group (M=14.31, SD=2.13).

No significant effects were observed for Age, F(1, 126)=0.558, p=0.457 or IQ, F(1, 126)=0.120, p=0.730. Further, there were no statistical differences between males and females on Age, IQ and Education.

3.3.2 Sex differences across ASD and TYP cohorts

Results are summarised in Table 3.2. A significant overall main effect for diagnosis for the

2 neuropsychological measures (Hotelling’s Trace: F(12, 112)=3.45, p<0.001, ηp =.270) revealed broad difficulties in cognitive and executive functioning for the ASD group compared to the

TYP group. A summary of these findings and a detailed discussion has been reported elsewhere35. A significant overall main effect for sex was also observed (Hotelling’s Trace:

2 F(12, 112)=2.52, p=0.006, ηp =.212), however, the diagnosis by sex interaction effect was not

2 significant, (Hotelling’s Trace: F(12, 112)=0.62, p=0.826, ηp =.062). The interaction effect sizes

2 for almost all of the individual neurocognitive measures were very small (ηp <0.04). Post-hoc power analysis (conducted with G*Power software) indicated that a sample size of N>4000 would be required for the above interaction effects to reach significance. This suggests that our non-significant sex by interaction finding is robust for our cohort.

Page | 68

The between subjects main effects for sex, revealed that the following neuropsychological measures reached significance: psychomotor speed (TMT-A): F(1, 123)=4.05, p=0.046,

2 2 ηp =.032; set-switching (TMT-B): F(1, 123)=6.01, p=0.016, ηp =.047; verbal learning-structured

2 (LM I): F(1, 123)=6.30, p=0.013, ηp =.049; verbal learning-unstructured (RAVLTA): F(1,

2 123)=6.22, p=0.014, ηp =.048 and semantic fluency (COWATSemantic): F(1, 123)=4.62, p=0.034,

2 ηp =.036. For each measure, females performed significantly better than males.

A significant overall effect for diagnosis was also observed for the two clinical indices of the

2 BRIEF (Hotelling’s Trace: F(2, 81)=18.36, p<0.001, ηp =.312), with each of the MI, F(1,

2 2 82)=24.88, p<0.001, ηp =.233 and BRI, F(1, 82)=36.13, p<0.001, ηp =.306, indices reaching

2 significance. The sex main effect (Hotelling’s Trace: F(2, 81)=1.09, p=0.340, ηp =.026) and the

2 sex by diagnosis interaction, (Hotelling’s Trace: F(2, 81)=1.17, p=0.316, ηp =.028) was not significant for the BRIEF MI and BRI indices respectively, or the clinical scales, (Sex:

2 Hotelling’s Trace: F(9, 73)=1.83, p=0.078, ηp =.184; Sex*Diagnosis: Hotelling’s Trace: F(9,

2 73)=1.29, p=0.255, ηp =0.138), results are summarised in Table 3.3.

3.4 Discussion

Our study demonstrated broad sex differences in cognitive function; however, importantly these were not different for individuals in the ASD group. Contrary to our predictions, we did not find a significant sex by interaction effect for our measures of psychomotor speed and cognitive flexibility. Instead, an overall female advantage was observed on measures of psychomotor speed, cognitive flexibility, two measures of verbal learning and memory and for semantic fluency. The findings on cognitive flexibility (as measured by reasoning speed) and verbal tasks are consistent with trends in the general normative population315. It is of note however, that there was no persistent female advantage across all verbal tasks supporting the proposal that sex differences may be influenced by broader ‘intergenerational or other sociocultural factors317 with individual biological sex differences further augmenting

Page | 69

variability in cognitive performance. The influence of intra-generational sex variability317 is supported by recent research identifying neuroanatomical male and female brain phenotypes independent of biological sex324, 325 which influence sensitivity and specificity of ASD diagnosis and may thus influence other behavioural phenotypes.

The female advantage in verbal tasks and cognitive flexibility observed in our study may in part be explained, by sex differences in neural circuitry. A functional neuroimaging study326 of working memory tasks, identified different areas of activation by males and females. Males recruited more parietal/visuospatial areas of the brain, whereas females recruited prefrontal/hippocampal areas associated with phonological tasks suggesting that females may perform better on tests of verbal short-term memory. Comparably, Ingalhalikar and colleagues327 reported sex differences in cortical and cerebellar connectivity which they related to enhanced perception in males and superior analytical/intuitive processes in females, the latter potentially accounting for our observed sex differences in cognitive flexibility. A recent neuroimaging study identified a significant sex by diagnosis interaction for frontal tracks328 which may also in part explain our findings for EF domains of fluency and cognitive flexibility.

The lack of expected normative trends for male advantage for visuospatial tasks may be in part due to differences in the visuospatial tasks used here compared to other studies329. Other moderating factors leading to null results may relate to converging trends in gender socialisation330, or atypical brain cognitive processes where cognitive processes in some males and females may be more aligned with the opposite sex than their biological sex325.

Further, our hypothesis for sex differences on the self-report measure was also not supported.

An examination of differences within the ASD cohort revealed similar self-reports of executive dysfunction that was significantly higher than the neurotypical comparison group but did not differ by sex. The reported significant (albeit small) normative sex difference for the ‘emotional control’ and ‘initiate’ clinical scales103 was not extended to the ASD individuals in this study.

Page | 70

These results suggest that males and females with ASD report similar difficulties in executive functioning in real world situations. Our findings also suggest that the reported sex differences in EF based on parental ratings311 may reflect informant cognitive biases in the evaluation of males and females with ASD.

Commensurate with our overall findings, recent research suggests sex differences in ASD reflect normative trends thus highlighting the importance of including a neurotypical comparison group. Such trends have been reported for specific cognitive tasks, e.g. visuospatial rotational task331, broader social traits such as social motivation and friendship332 and, brain neurobiology324.

3.4.1 Limitations

There are a number of limitations to our study. Our ASD group did not include individuals with intellectual difficulties and overall they showed a relatively high IQ, thus our results cannot extend to the broader ASD population with intellectual difficulties. Our cognitive test battery was not exhaustive, other cognitive domains that could be assessed include working memory, and higher-order EF such as planning and “hot” executive’ functions. Further, we did not control for use of medication or for mood states that may have a confounding effect on test performance. We note this is a cross-sectional study and a longitudinal analysis across development would better determine whether sex differences in cognitive profiles emerge at different stages of development and may explain the observed variability in research findings.

Emerging evidence suggest that there are indeed gender specific trajectories at least in the area of autistic social traits333, further research addressing neurocognitive factors will provide additional insight on cognitive sex differences.

Page | 71

3.4.2 Conclusion and future directions

Our study suggests sex differences in cognitive/EF in adults with ASD as evidenced by an overall female advantage on measures of psychomotor speed, cognitive flexibility, verbal learning and memory and semantic fluency. These differences were not, however, unique to

ASD and followed a similar sex difference pattern for typically developing individuals.

Contrary to other research advocating a distinct male brain profile221 our findings, do not support the view of atypical sex differentiation in ASD on neurocognitive or self-report measures. Future studies of sex differences in ASD should incorporate a multidimensional assessment protocol and a neurotypical comparison group, to better evaluate potential sex differences in cognitive performance and their relation to other core ASD domains. Enhancing the research design to incorporate a more comprehensive assessment protocol across both neuropsychological and subjective measures and controlling for potential neurobiological and sociocultural moderators can contribute to a better understanding of sex differences and inform diagnostic and intervention strategies.

Page | 72

Figure 3.1: Cognitive and executive function domains

Page | 73

Table 3.1: Demographics

ASD (N=69) TYP (N=59)

Mean SD Mean SD

Age 24.4 7.5 25.3 5.2 F(1, 126)= 0.56, p=0.457

Predicted IQ 107.7 8.4 107.2 8.3 F(1, 126)= 0.12, p=0.730

Years of 12.5 2.1 14.3 2.1 F(1, 126)= 22.89, p<0.001

Education

Page | 74

Table 3.2: Neurocognitive and executive function domains ASD TYP Gender Between Subject Factors Mean SD Mean SD Sex Main Effect Sex by Diagnosis Interaction

Processing speed and attention

2 2 TMT-A female 31.1 14.5 20.8 4.8 F(1, 123)= 4.05, p=0.046, η =0.032 F(1, 123)= 0.160, p=0.690, ηp =.001

male 37.3 20.2 23.3 6.9

2 2 RVP-A female .875 .071 .924 .040 F(1, 123)= 0.442, p=0.508, η =0.004 F(1, 123)= 0.488, p=0.486, ηp =.004

male .888 .053 0931 .041

Verbal STM

2 2 LM I female 42.0 12.8 43.8 8.4 F(1, 123)= 6.30, p=0.013, η =0.049 F(1, 123)= 0.588, p=0.445, ηp =.005

male 35.1 12.9 41.1 10.5

2 2 LM II female 24.4 11.5 28.1 7.2 F(1, 123)= 3.53, p=0.063, η =0.028 F(1, 123)= 0.206, p=0.651, ηp =.002

male 20.4 9.9 26.9 7.9

2 2 RAVLTA female 54.6 11.3 59.5 6.7 F(1, 123)= 6.2, p=0.014, η =0.048 F(1, 123)= 0.509, p=0.477, ηp =.004

male 48.7 12.4 57.3 7.3

2 2 RAVLTB female 11.5 3.9 12.9 2.9 F(1, 123)= 2.46, p=0.119, η =0.020 F(1, 123)= 0.553, p=0.458, ηp =.004

male 10.0 4.1 12.6 2.7

Page | 75

Visuospatial STM

2 2 SSP female 6.2 1.4 7.6 1.1 F(1, 123)= 1.80, p=0.182, η =0.014 F(1, 123)= 0.044, p=0.834, ηp <.001

male 6.6 1.3 7.8 1.2

2 2 PALTOTAL female 7.9 9.9 5.6 8.3 F(1, 123)= 3.88, p=0.051, η =0.031 F(1, 123)= 4.103, p=0.045, ηp =.032

ERRORS

male 23.1 33.7 3.7 4.6

Executive Function

2 2 TMT-B female 71.8 42.7 45.8 10.2 F(1, 123)= 6.0, p=0.016, η =0.047 F(1, 123)= 0.0, p=0.998, ηp <.001

male 85.8 39.2 54.7 17.8

2 2 IEDerrors female 23.3 20.5 15.6 14.2 F(1, 123)= 0.422, p=0.517, η =0.003 F(1, 123)= 0.415, p=0.521, ηp =.003

male 28.6 30.4 13.7 11.8

2 2 COWATphonemic female 34.1 13.6 41.8 7.8 F(1, 123)= 0.111, p=0.740, η =0.001 F(1, 123)= 1.043, p=0.309, ηp =.008

male 31.5 9.6 43.7 11.4

2 2 COWATsemantic female 21.0 5.8 23.1 4.4 F(1, 123)= 4.62, p=0.034, η =0.036 F(1, 123)= 0.331, p=0.566, ηp =.003

male 18.5 4.9 21.7 5.3

Page | 76

Table 3.3: Self-report measures of executive function

BRIEF ASD TYP Gender

Mean SD Mean SD Main Effect Interaction

2 2 BRI female 65.2 11.7 39.9 10.0 F(1, 82)= 0.428, p=0.515, η =0.005 F(1, 82)= 1.93, p=0.168, ηp =.023

male 59.8 13.9 40.8 8.9

2 2 MI female 82.0 19.6 54.3 11.8 F(1, 82)= 0.226, p=0.635, η =0.003 F(1, 82)= 0.274, p=0.602, ηp =.003

male 82.0 16.1 56.2 15.5

Clinical scales

2 2 Inhibit female 15.8 4.2 11.0 2.4 F(1, 81)= 0.116, p=0.735, η =0.001 F(1, 81)= 0.789, p=0.377, ηp =.010

male 14.9 3.5 11.2 2.4

2 2 Shift female 14.5 2.9 8.0 2.5 F(1, 81)= 0.901 p=0.345, η =0.011 F(1, 81)= 3.45, p=0.065, ηp =.041

male 12.7 2.8 8.5 1.9

2 2 Emotional female 22.5 5.3 13.7 4.1 F(1, 81)= 0.796, p=0.375, η =0.010 F(1, 81)= 1.52, p=0.222, ηp =.018 control

male 20.0 5.9 13.8 4.1

2 2 Self monitor female 12.4 2.4 7.3 1.9 F(1, 81)= 0.209, p=0.649, η =0.003 F(1, 81)= 0.952, p=0.332, ηp =.012

male 12.1 3.5 8.1 1.8

Page | 77

2 2 Initiate female 17.9 4.8 10.6 2.5 F(1, 81)= 0.595, p=0.443, η =0.007 F(1, 81)= 1.54, p=0.219, ηp =.019

male 17.5 4.0 12.2 3.9

2 2 Working female 17.5 4.7 10.7 3.4 F(1, 81)= 0.429, p=0.514, η =0.005 F(1, 81)= 1.09, p=0.298, ηp =.013 memory

male 16.1 3.8 10.7 2.5

2 2 Plan/Organise female 20.4 5.2 12.7 2.9 F(1,81)= 1.43, p=0.235, η =0.017 F(1, 81)= 1.08, p=0.302, ηp =.013

male 20.6 4.7 14.6 4.7

2 2 Task/Monitor female 11.4 3.3 8.7 2.2 F(1, 81)= 0.320, p=0.573, η =0.004 F(1, 81)= 0.101, p=0.752, ηp =.001

male 11.9 2.2 8.7

2 2 Organisation/ female 14.9 4.5 11.6 2.0 F(1, 81)= 0.154, p=0.696, η =0.002 F(1, 81)= 0.728, p=0.396, ηp =.009 materials

male 15.9 3.9 11.2 3.0

Page | 78

Chapter 4: A transdiagnostic examination of anxiety and stress on executive function

outcomes in disorders with social impairment

Study objectives:

Empirical study 3 continues investigation of moderator influences on cognitive performance.

The objectives of empirical study 3 are to:

(a) investigate the role of anxiety, stress and depression on EF and non-EF performance

and behavioural measures in a cohort characterised by social impairment

(b) evaluate the transdiagnostic efficacy of EF performance and behavioural measures

Page | 79

A transdiagnostic examination of anxiety and stress on executive function outcomes in

disorders with social impairment

Demetriou, E.A., Park, S.H., Pepper, K.L., Naismith, S.L., Song, Y. J. C., Thomas, E.E.,

Hickie, I.B., Guastella, A.J.

(manuscript in preparation)

Page | 80

Abstract

The objective of the study was to examine the moderating influence of anxiety and stress on executive function (EF) across two clinical conditions characterised by social impairment,

Autism Spectrum Disorder (ASD) and Social Anxiety Disorder (SAD). In Study 1, neuropsychological measures collected from a large case-control cohort (N=210) were subject to factor analytic modelling to identify EF and non-EF latent factor structure(s). The larger sample was utilised in the factor analysis to attain better power for the study. The EF and non-

EF factors identified in Study 1 were subject to further statistical analyses in Study 2. Study

2, comprised of a smaller cohort of participants (N=138) who also completed behavioural self- appraisals of EF and affective states. Multiple regression analyses investigated the predictive value of depression, anxiety and stress on EF and non-EF cognitive domains and on composite indices of EF behavioural ratings.

Study participants met standardised diagnostic criteria for ASD (Study1: N=69; Study 2:

N=52), or SAD (Study1: N=82, Study 2: N=63) and were included in the study if they had no intellectual disability (IQ>70), neurological condition or substance dependence. Neurotypical control participants (TYP: Study1: N=59, Study 2: N=31) were excluded if they reported a mental health diagnosis (past or current) or were currently experiencing significant symptom severity for depression or social anxiety, as assessed by standardised screening instruments.

Our results showed that trait anxiety was associated with better performance on neuropsychological measures of EF while state-based stress was associated with lower EF performance. A dissociation was observed between trait anxiety and state stress in predicting self-appraisal ratings of EF on the behavioural indices. Depression, anxiety and stress did not predict performance on non-EF cognitive domains.

Page | 81

The findings demonstrate that trait anxiety and state-based stress moderate EF processes across disorders with social impairment. The transdiagnostic efficacy of these moderating factors on

EF can facilitate remediation strategies and may also contribute to individuals with ASD gaining better access to mental health services.

Page | 82

4.1 Introduction

Executive function (EF) broadly refers to goal directed cognitive processes and behaviours21 and is generally assumed to be distinguished across discrete domains. Core EF domains include set-shifting, response inhibition and working memory26, 67. Impairment of EF has been reported across psychopathology73, 334, 335 and is considered a distinguishing characteristic of

Autism Spectrum Disorder (ASD)27, 336. Although EF has been studied across many psychiatric conditions72, 334, 337, 338 moderators of EF have received less attention. Anxiety339, 340 and stress340, 341 are two factors thought to moderate EF partly due to their attenuating influence on top-down executive processes of the prefrontal cortex342.

Examining the relationship between EF, anxiety and stress is important given research support that EF impairment may be a predisposing factor34, 343 and/or consequence of anxiety and stress209, 334. The pertinence of this relationship is augmented further by current trends advocating development of cross diagnostic research domain criteria (RDoC)344 for the study of psychiatric344, 345 and neurodevelopmental conditions346. Anxiety347 and EF348 have both been identified as having cross-diagnostic relevance. Their potential relationship may be of particular interest to investigate in conditions with common behavioural outcomes but varying degrees of executive dysfunction. Both ASD and SAD provide populations for examining this relationship. Both diagnoses are characterised by social impairment but EF deficits are consistently noted only in ASD72 while EF processes in SAD appear mostly intact349.

Theoretical models of anxiety propose a multifactorial construct with at least a distinction between state and trait based factors including state and trait anxiety350. Similarly a distinction is made between state/acute and trait/chronic based stress209. A number of cognitive models of anxiety351 and stress352 propose that they exert their influence on the executive control processes of the prefrontal cortex209, 351. Attentional Control Theory (ACT)351 asserts that trait208 and state339 anxiety will impair top-down executive control processes. Specifically,

Page | 83

highly anxious individuals will shift attention focus from top-down executive control to bottom-up processing of a stimulus thus impairing performance on EF tasks. The ACT model further predicts that anxiety will not influence non-EF processes such as phonological and visuospatial short term memory (STM)232. A further premise proposed by ACT is that motivation, may moderate EF performance in highly anxious individuals353, 354. Acute stress209 is also thought to moderate top-down executive control impairing EF processes342 which are further attenuated by chronic influences of stress341.

Empirical studies examining the relationship between anxiety, stress and EF, have shown mixed results230, 231, 355. One possible explanation for these mixed results may be due to diverging effects of state/trait and acute/chronic measures of anxiety and stress, as well as the specific type of EF under investigation. A study by Pachego et al230 showed that trait but not state anxiety was associated with impairment in the executive control network.. A meta- analysis investigating acute stress on EF domains209 showed that acute stress impaired the domains of set-shifting and working memory but not response inhibition. Differences may also be due to the type of stimulus presented in the assessment tasks. Many studies examine the relationship of anxiety and EF in experimental conditions where affective rather than neutral stimuli are presented. Given that different top-down cognitive control processes are believed to be involved in the processing of emotional stimuli356, this research study focuses only on the influence of anxiety and stress on neutral measures of EF.

Empirical findings on the relationship between EF, anxiety and stress are mixed in clinical and non-clinical samples230-232. A comparison of individuals with high versus low trait anxiety, indicated that the high trait cohort had reduced performance on a working memory task232 while state anxiety did not exert a significant influence. A more complex pattern of results has been reported by other studies230, 231. In a non-clinical sample, high trait anxiety was associated with impairment in executive control of attention whereas high state anxiety only influenced non-

Page | 84

executive attentional processes. A study by Visu-Petra et al231 showed that high trait anxious participants performed better on working memory EF tasks. A negative relationship however was reported between state and trait anxiety on the EF tasks of response inhibition and set- shifting. The positive association between anxiety and working memory was attributed to higher motivation to perform well by subjects showing greater sensitive to social evaluation.

In a study of neurotypical children aged 9-10 years and classified as high state versus low state anxious357, no relationship was shown between state anxiety and performance measures on working memory tasks. An association however was observed between state anxiety and processing efficiency. More effort was required to complete the working memory tasks in the highly anxious children357. In summary, studies in non-clinical samples suggest a possible dissociation between anxiety and EF, dependent on the anxiety measure, EF domain and motivational factors.

Most studies that have recruited participants diagnosed with SAD, focus on the influence of anxiety on executive cognitive control in the context of threat related stimuli. Few studies have investigated this relationship using neutral stimuli. In a cohort diagnosed with SAD, a negative relationship was reported between symptom severity of social anxiety and performance on a set-shifting task358. Levels of social anxiety did not influence other EF (fluency, set-switching) and non-EF measures (STM). In contrast, Heeren et al355 found a positive association between severity of SAD symptoms and non-EF attentional processes. No relationship was observed between EF attentional processes with state/trait anxiety or severity of SAD symptom.

Studies that have recruited participants diagnosed with ASD also report mixed findings. A study of adults with ASD reported a negative association between state anxiety and performance on a concept formation task34 but not with state measures of depression and stress.

In a study of adolescents, anxiety correlated with impaired performance on EF measures of inhibition, set-switching and set-shifting but not with working memory235. A small number of

Page | 85

studies in ASD cohorts examined the relationship between self-appraisal ratings of EF and anxiety. Wallace et al41 reported a positive association between poor set-shifting self- appraisals and anxiety but not with any of the other behavioural measures of EF studied. In summary, there is a body evidence that supports anxiety influences EF in clinical populations however, the specific impact of state and trait anxiety is inconclusive. Moreover, there has been limited investigation of the role of closely related affective states, depression and stress, in these relationships.

The aim of this study was to examine the influence of anxiety and stress on EF in two cohorts that report high levels of anxiety but with differing levels of EF performance. A moderately sized clinical population of participants was recruited with known EF impairments (ASD) and those with known high level of anxiety but likely intact EF (SAD). In addition, we recruited a sample of participants with no likely impairments in EF and no evidence of heightened anxiety symptoms (TYP).

The main hypothesis is that the influence of anxiety would be greatest for those with impaired

EF function (ASD) and for those who report more extreme levels of anxiety (SAD). In contrast, no relationship of anxiety on EF is expected for the cohort that show low levels of anxiety and no EF impairment (TYP). Further, it is expected that these moderating influence of anxiety would be specific to EF measures and not on other non-EF measures of cognitive function or depression. Finally, it is proposed that high levels of state stress will impair EF performance with no specific prediction made on the influence of stress on non-EF cognitive measures as none are made by cognitive models of stress.

Page | 86

4.2 Method

Methodology

In Study 1, EF and non-EF cognitive measures collected from a cohort of 210 participants

(Study 1: N= 210; Age: M = 23.4, SD = 6.7) were subject to a factor analysis. The objective of Study 1 was to extract the cognitive and EF factor(s) structure of the neuropsychological assessments completed by the participants. In Study 2, multiple regression analyses were conducted on the extracted factors utilising a smaller cohort of ASD, SAD and TYP participants (Study 2: N= 146; Age: M = 23.4, SD = 6.7). These participants had also completed a self-report measure of EF, a trait anxiety measure and state measures of anxiety, depression and stress. The objective of Study 2 was to examine the predictive influence of affective states, in particular trait and state anxiety, on the extracted cognitive and executive function factors.

Participants

The University of Sydney Ethics Committee approved the research protocol (2013/352) for this project. Informed consent was obtained directly from each participant prior to inclusion in the study. A cohort of youth and adults, with a diagnosis of ASD or SAD, were sequentially recruited from referrals for treatment or social skills development at the Autism (Autism Clinic for Translational Research - ACTr) or headspace clinics at the Brain and Mind Centre,

University of Sydney. A comparison neurotypical control group (TYP) was recruited via the

University of Sydney website. Clinicians at the Autism Clinic who completed training on standardised diagnostic assessment instruments assessed referrals and made the diagnoses of

ASD and SAD. Diagnosis of ASD was based on the Autism Diagnostic Observation Schedule

(ADOS)322 and diagnosis of SAD was based on the Anxiety Diagnostic Interview Schedule

(ADIS)359. Participants met criteria for the primary presenting disorder of either ASD; (Study

1: N = 69, Study 2: N=52) or SAD: (Study1: N = 82, Study 2: N=63). Neurotypical control

Page | 87

participants (TYP: Study 1: N = 59, Study 2: N=31) were recruited through advertising at a number of different university websites. All participants were screened and excluded from the study if they had intellectual disability (IQ<70) or current substance dependence. Furthermore, control participants were excluded if they reported a mental health diagnosis (past or current), were currently experiencing significant levels of depression, anxiety or stress, (assessed with the DASS-21)360 or if they scored above the screening cut-offs for Social Anxiety (Social

Interaction Anxiety Scale - SIAS)361 and/or Autism Spectrum (Autism Quotient - AQ)362.

Outcome Measures:

A detailed description of all measures is summarised in Appendix 5, Table S5.1. The assessment battery comprised of a combination of neuropsychological tests and self-report ratings of cognitive and EF and symptom severity measures. Intelligence quotient (IQ) and psychomotor speed were estimated based on the Wechsler Test of Adult Reading (WTAR)323 and the Trail Making Test (TMT-A)119 respectively. Verbal learning and STM for unstructured information was assessed with the Rey Auditory Verbal Learning Test (RAVLT) whereas verbal memory for structured information were assessed with the Logical Memory Stories I

(LM I) and II (LM II) from the Wechsler Memory Scale, 3rd edition. Visuospatial short term memory and learning were assessed with the Spatial Span test (SSP) and the Paired Associates

Learning test (PAL) from the Cambridge Neuropsychological Test Automated Battery

(CANTAB)87. The EF tests assessed the sub-domains of ‘set-shifting’ (CANTAB, Intra/Extra

Dimensional Shift Test-IED)87, ‘set-switching’ (Trail Making Test, TMT-B)119, ‘verbal fluency’ (Controlled Oral Word Association Test - COWAT) and ‘sustained attention’

(CANTAB, Rapid Visual Processing test - RVP)87. Self-report ratings of EF were assessed by the BRIEF103 and specifically the Behavioural Regulation Index (BRI) and the Metacognition

Index (MI). Trait anxiety was assessed by the trait scale of the State-Trait Anxiety Inventory

Page | 88

(STAI)363. State measures of depression, anxiety and stress were assessed by the respective scales of the DASS-21360 self-report measure.

Data Analysis:

Data analysis was performed using the IBM SPSS Statistics version 24. In Study 1, we investigated the underlying factor structure of the neuropsychological test battery. A Principal

Axis Factor Analysis (PFA) was conducted on all cognitive measures with oblique promax rotation to ascertain the latent factors of the cognitive assessment battery. This was followed by confirmatory factor analysis of the designated EF measures only. Hierarchical multiple regression analyses (MR) were undertaken to examine the influence of affective measures on the EF and non-EF cognitive factors and behavioural EF indices. Demographic moderators were entered in Step 1, overall ratings of trait/state anxiety and state depression and stress, were entered in Step 2.

4.3 Results

Demographics:

For the overall sample, no significant gender differences (43% female) were observed between

2 the diagnostic and control groups, χ 2, N=210) = 2.13, p>0.05. Significant effects were obtained for

Education (F2, 207=24.23, p<0.001) and IQ (F2, 207=6.30, p=0.002). Pairwise comparisons with

Bonferroni corrections (α=0.01) revealed that the ASD and SAD groups had significantly fewer years of education compared to the control group (p<0.001), whereas the SAD group had significantly higher IQ compared to the control group (p<0.001). The between group differences for the neuropsychological and self-report measures have been reported in

Demetriou et al35 and are also reported in Chapter 5 of this thesis. The ASD and SAD groups reported comparable levels of anxiety, depression and stress which were significantly higher than the neurotypical group. DASSDepression overall ANOVA, (F2, 124=31.2, p<0.001, post-hoc analysis with Bonferroni adjustment, ASD vs SAD, p>0.05, ASD vs TYP, p<0.001, SAD vs

Page | 89

TYP, p<0.001. DASSAnxiety overall ANOVA, (F2, 124=19.8, p<0.001, post-hoc analysis with

Bonferroni adjustment, ASD vs SAD, p>0.05, ASD vs TYP, p<0.001, SAD vs TYP, p<0.001.

DASSStress overall ANOVA, (F2, 124=34.2, p<0.001, post-hoc analysis with Bonferroni adjustment, ASD vs SAD, p>0.05, ASD vs TYP, p<0.001, SAD vs TYP, p<0.001.

Study 1: Factor Analysis

The results of the Factor Analyses are presented in Appendix 4, Tables S4.1 and S4.2. Principal axis factoring (PAF) of the neuropsychological test measures identified three principal factors

(with eigen values exceeding 1) which accounted for around 62% of the test variance. The pattern matrix revealed that the EF tests (except IED), SSP and TMT-A, loaded on Factor 1.

The verbal/auditory learning/STM measures loaded on Factor 2, this was designated as the verbal/auditory factor. Factor 3 comprised primarily of the visual memory/new learning (PAL) measures and was designated as the visuospatial factor. A follow-up factor analysis of the measures theorised to contribute to EF processes, (IED, TMT-B, COWAT and RVP) identified only one factor, this was designated as the EF factor.

Study 2:

Regression analyses

Regression analyses were completed on the combined sample of ASD, SAD and TYP participants for the designated EF, verbal/auditory and visuospatial factors. Detailed statistics for the neuropsychological EF and non-EF factors and the behavioural indices of the BRIEF are presented in Tables 4.1a-4.2b.

Executive function neuropsychological measures

The overall model for Step 1 was significant and accounted for 37.5% of the variance,

R2=0.375, F (4, 133)=19.97, p<0.001. Education (β=.233, p=0.003) and IQ (β=.377, p<0.001) were the significant predictors. Step 2 accounted for an additional 9.7% of the variance

ΔR2=0.097, F (4, 129)=5.92, p<0.001. Education, (β=.214, p=0.005), IQ (β=.275, p<0.001),

Page | 90

DASSStress (β=-.339, p=0.004) and STAI (β=.385, p=0.003 were the significant predictors. The interaction effect between diagnosis with each of the affective measures was not significant.

Non-executive function test measures

The overall model for Step 1, was significant for each of the non-EF factors, (Factor 2, R2=.26,

F(2, 135)=23.77, p<0.001 and Factor 3, R2=.075, F(2, 135)=5.51, p=0.005; whereas adding the affective measures in Step 2, did not contribute to a significant increment in variance for any of the non-EF factors. IQ was a significant predictor for Factor 2, (β=.454, p<0.001). The interaction effect between diagnosis with each of the affective measures was not significant (p

> 0.05).

Executive function self-report measures

Separate regression analyses were completed for the BRI and MI indices. For the BRI both

Step 1, R2=0.397, F (4, 103)=16.94, p<0.001 and Step 2 ΔR2=0.279, F (4, 99)=21.28, p<0.001 were significant. ASD (β=.674, p<0.001) and SAD (β=.424, p<0.001) diagnoses, were the only significant predictors in Step 1, whereas the DASSStress (β=.536, p<0.001) was the only significant predictor in Step 2.

For the MI index, Step 1, R2=0.386, F (4, 103)=16.16, p<0.001 and Step 2 ΔR2=0.174, F (4,

99)=9.82, p<0.001 were significant. Education (β=-.300, p=0.001), ASD (β=.546, p<0.001) and SAD (β=.529, p<0.001) diagnoses, were the significant predictors in Step 1, whereas in

Step 2, Education (β=-.223, p=0.006) and STAI (β=.420, p=0.003) were the only significant predictors. The interaction effect between diagnosis with each of the affective measures was not significant (p> 0.05).

4.4 Discussion

Our results showed that, regardless of group assignment, trait anxiety and state stress were associated with performance on the composite EF factor. Interestingly, however, trait anxiety and state stress showed opposing influences on these measures of EF. While trait anxiety was

Page | 91

associated with better EF performance, state stress was associated with poorer EF performance.

Higher trait anxiety and state stress also predicted worse self-appraisal evaluation of EF for the

MI and BRI indices respectively but not the verbal and visuospatial STM factors.

The findings partially supported our main hypotheses. While trait anxiety influenced EF, in contrast to our expectation, higher trait anxiety was associated with better EF performance.

This finding may be explained, in terms of ACT’s tenet of an interaction between task characteristics and motivation. That is, when presented with demanding tasks, high anxious individuals are likely to engage in compensatory strategies and recruit more attentional neural resources to improve performance354. Contrary to our predictions state stress rather than state anxiety, was a significant predictor of EF performance. Specifically, state-based stress impaired EF performance across both clinical groups. A recent meta-analysis noted that acute stress was associated with impaired performance in the EF domains of working memory, cognitive flexibility and cognitive inhibition209. The attenuating effect of state stress on EF in this study is also consistent with findings by Edwards and colleagues364 where high levels of state stress impaired cognitive performance on a working memory task. The lack of association between anxiety/stress and non-EF factors is also consistent with predictions made by ACT where only the executive control component of working memory is expected to be impaired and not their corollary visuospatial and phonological STM components232.

The dissociation observed between state stress, (associated with the BRI index) and trait anxiety (associated with the MI index) may be explained by the EF processes specific to each index. A closer examination of the clinical scales of the two indices indicate that the BRI is composed mainly of domains relating to emotional control and self-monitoring103. By comparison, the MI relates to the individual’s ability to cognitively manage attention and problem solve via planning and organisation103. Higher levels of stress/anxious arousal may be expected to correlate with poorer emotional control as reflected in the BRI index.

Page | 92

Conversely, as predicted by the ACT model, high trait anxiety likely attenuated executive control thus influencing higher order EF domains such as problem solving25 contributing to impairment in the MI index.

An important finding in this study is that diagnostic classification did not moderate the association between anxiety and EF performance. This is despite both the SAD and TYP groups showing adequate performance on the neuropsychological measures of EF. This lends support to the transdiagnostic influence of anxiety (and stress) on EF. Our findings complement recent research on the transdiagnostic correlates of anxiety with EF processes.

Shanmugan and colleagues365 reported that an empirically derived anxiety factor (based on a structured clinical interview and thus likely reflecting a stable anxiety dimension), was associated with better memory performance and hyperactivation of the executive neural network365. A second study that utilised a machine learning paradigm, identified six psychopathology subtypes366 drawn from a composite clinical cohort. The researchers found that the psychopathology subtype associated with anxious arousal/state anxiety factor, was associated with greatest level of neurocognitive impairment, particularly for working memory366.

Limitations

There are a number of limitations associated with our study. Firstly, we utilised a cross sectional design and thus causal relationships between anxiety and EF cannot be established.

Second, our sample size was moderate, extending this methodology to a larger sample particularly in relation to the self-report measure of the BRIEF and, inclusion of other clinical cohorts with shared phenotype will further strengthen the transdiagnostic relevance of our results. Third, we utilised only self-report measures of anxiety, depression and stress and we did not induce anxiety or stress experimentally which may lead to different relationships.

Finally, our study did not include performance measures of working memory, a key process in

Page | 93

the ACT theory that may have contributed to more complex relationships between EF and affective states.

Conclusions

The findings indicate that anxiety and stress influence EF processes including capacity to form new concepts, generate ideas and flexibly switch between mental processes. The influence of anxiety and stress on EF appears to be similar across diagnostic groups. On neuropsychological measures of EF, a bi-directional relationship was observed. Trait anxiety improves EF performance, and this may relate to the role motivation on processing efficiency and effectiveness. In contrast, state stress impairs performance on neuropsychological assessments of EF and this may be related to the role of stress in modulating prefrontal activation and re- allocating resources from EF processes to limbic system processes367. On self-report measures of EF, direct relationships were observed for high trait anxiety and impaired MI index and high state stress and impaired BRI index. Findings of bi-directional influence of trait anxiety and state stress on EF highlight the importance of a multifactorial research design. Affective moderators should be studied in conjunction with EF measures so that interaction effects may be adequately assessed. Future research is now required to determine whether state and trait affective moderators influence EF in other psychiatric conditions and whether manipulation of these factors alters EF.

Page | 94

Table 4.1a: Regression Analysis on executive function factor

B SE B β

Constant -5.891 .963

IQ .044 .009 .377**

Education .089 .030 .233*

ASD -.438 .195 -.240

SAD .020 .190 .011

Step 2

Constant -5.164 .959

IQ .032 .008 .275**

Education .082 .029 .214*

ASD -.347 .219 -.190

SAD -.067 .224 -.037

DASSDepression .000 .008 -.002

DASSAnxiety -.013 .009 -.150

DASSStress -.027 .009 -.339*

STAI .026 .008 .385*

* p < 0.01, **p<0.001

Page | 95

Table 4.1b: Regression Analysis on Verbal/Auditory STM factor

B SE B β

Constant -6.363 1.130

IQ .053 .010 .422**

Education .052 .035 .125

ASD -.262 .228 -.132

SAD -.022 .222 -.011

Step 2

Constant -6.559 1.193

IQ .047 .010 .372**

Education .059 .036 .141

ASD -.425 .273 -.214

SAD -.312 .279 -.161

DASSDepression -.007 .010 -.089

DASSAnxiety -.003 .011 -.029

DASSStress -.011 .011 -.123

STAI .025 .010 .349

* p < 0.01, **p<0.001

Page | 96

Table 4.1c: Regression Analysis on Visuospatial STM factor

B SE B β

Constant 2.078 1.221

IQ -.013 .011 -.100

Education -.065 .038 -.156

ASD .537 .247 .269

SAD -.094 .240 -.048

Step 2

Constant 2.134 1.284

IQ -.008 .011 -.060

Education -.072 .039 -.172

ASD .811 .294 .406*

SAD .286 .300 .147

DASSDepression -.013 .011 -.177

DASSAnxiety .005 .012 .055

DASSStress .011 .012 .128

STAI -.015 .011 -.206

* p < 0.01, **p<0.001

Page | 97

Table 4.2a: Regression Analysis on BRIEF BRI index

B SE B β

Constant 68.683 17.591

IQ -.113 .159 -.061

Education -1.040 .498 -.183

ASD 18.950 3.116 .674**

SAD 12.282 3.154 .424**

Step 2

Constant 33.467 14.167

IQ .065 .127 .035

Education -.617 .386 -.108

ASD 3.815 2.960 .136

SAD -1.978 3.082 -.068

DASSDepression .023 .121 -.022

DASSAnxiety .213 .122 .165

DASSStress .633 .124 .356**

STAI .107 .120 .104

* p < 0.01, **p<0.001

Page | 98

Table 4.2b: Regression Analysis on BRIEF MI index

B SE B β

Constant 69.082 24.182

IQ .198 .219 .078

Education -2.329 .684 -.300*

ASD 20.904 4.284 .546**

SAD 20.849 4.336 .529**

Step 2

Constant 42.641 22.475

IQ .143 .202 .056

Education -1.727 .612 -.223

ASD 3.499 4.696 .091

SAD 1.501 4.889 .038

DASSDepression .044 .192 .030

DASSAnxiety .127 .193 .072

DASSStress .222 .196 .138

STAI .586 .190 .420*

* p < 0.01, **p<0.001

Page | 99

Chapter 5: Autism, Early Psychosis, and Social Anxiety Disorder: A transdiagnostic

examination of executive function cognitive circuitry and contribution to disability

Study objectives:

The objectives of empirical study 4 are to investigate:

(a) if EF behavioural and performance measures differentiate individuals with ASD from

two clinical cohorts with EP and SAD and a neurotypical comparison group

(b) the efficacy of EF behavioural and performance measures in predicting disability

Page | 100

Autism, Early Psychosis, and Social Anxiety Disorder: A transdiagnostic examination of

executive function cognitive circuitry and contribution to disability

Demetriou, E.A., Song, Y. J. C., Park, S.H., Pepper, K.L., Naismith, S.L., Hermens, D.F.,

Hickie, I.B., Thomas, E.E., Norton, A.E., White, D., Guastella, A.J. Translational Psychiatry,

8, Article number 200 (2018)

https://doi.org/10.1038/s41398-018-0193-8.

See Appendix 1 for published version

Page | 101

Abstract

The disability burden in clinical cohorts with social impairment is significant leading to poor functional outcomes. Some of this impairment has been linked to executive dysfunction. In this study, a transdiagnostic approach was taken to identify executive function (EF) processes in young adults that may underpin social impairment and evaluate their contribution to disability.

Comparisons were made between three prominent disorders that are characterised by social impairment, Autism Spectrum Disorder (ASD), Early Psychosis (EP) and Social Anxiety

Disorder (SAD) as well as a neurotypically developing group (TYP). We examined whether overall disability could be predicted by neuropsychological and self-report assessments of EF.

Our study showed that ASD participants demonstrated impaired performance on most domains of EF compared to the TYP group (mental flexibility, sustained attention and fluency) while the EP group showed impairment on sustained attention and attentional shifting. The SAD participants showed EF impairment on self-report ratings even though their objective performance was intact. Self-reports of EF explained a significant percentage (17%) of disability in addition to the variance explained by other predictors, and this was particularly important for ASD. This is the first study to compare EF measures across clinical groups of social impairment and suggests unique cognitive-circuitry that underpins disability within groups. Impairments in EF were broad in ASD and predicted disability, EP impairments were specific to attentional processes and SAD impairments likely relate to negative self-monitoring.

Self-report, as opposed to performance-based EF, provided best capacity to predict disability.

These findings contribute to transdiagnostic circuitry models and intervention strategies.

Page | 102

5.1 Introduction

The disability burden in clinical cohorts with social impairment is significant and typically associated with poor functional outcomes265, 368, 369. Social impairment and disability have been linked to poor executive function (EF) performance255, 358 in these groups and EF may be a useful cognitive marker to predicting disability. Assessment of EF is traditionally based on neuropsychological (objective) measures of the level of performance across cognitive domains21. More recently, standardized scales of self/informant (subjective) based ratings of

EF have been introduced with empirical support that these may be more ecologically valid assessments of EF104. Understanding how EF and underlying cognitive circuitry may contribute to disability in clinical groups with social impairment and more specifically, ascertaining the contributions of objective and subjective measures of EF may be pivotal for diagnosis and functional outcomes. Such research is particularly important in early adulthood370 when brain development and transition into higher education, work, and adult social relationships coincide with establishing lifelong roles..

Autism Spectrum Disorder (ASD), Psychotic Disorders and Social Anxiety Disorder (SAD) are the three most common and recognised psychiatric conditions characterised by social impairment. Traditionally, the influence of EF has been examined within disorder clusters.

Numerous researchers have proposed common aetiologies and maintaining factors underpinning disability in these cohorts, raising potential for common circuitry processes across psychiatric disorders that may predict disability348, 371. In the neurodevelopmental cluster

(i.e. ASD, Psychosis), EF is believed to result from differential brain development372 which may contribute to disability373. Empirical support for a link between EF and social impairment has been reported for ASD255, 258 and psychotic disorders374 including the most common psychotic presentation in young adults, Early Psychosis (EP). These EF impairments may contribute to poor functional outcomes374, 375. Neural circuitry studies show involvement of the

Page | 103

prefrontal cortex (PFC) and in particular the dorsolateral and ventrolateral prefrontal cortex

(DLPFC, VLPFC) in EF performance for both schizophrenia and ultra-high risk psychotic populations376,338.

Within ASD, a recent meta-analysis suggested more global executive dysfunction27 with little evidence of selective dysfunction in specific EF domains (although few studies examined adult samples). Similarly, the underlying brain circuitry in ASD suggests widespread functional and anatomical differences377 although neuroimaging studies have also identified specific deficits in the PFC378 and fronto-striatal circuitry379. To date however, the relationship of disability to

EF has been explored by few studies41 and mostly in children or young adolescents36, 258. There has also been no transdiagnostic examination of the influence of EF to disability in neurodevelopmental cohorts.

Where the primary presenting condition is SAD, the relationship between neuropsychological difficulties and disability is more tenuous349, 358. Mixed results noted potential deficits in mental flexibility, verbal fluency349 and in the areas of sustained attention and concept formation358.

There is no empirical support, however for impaired neural circuitry specific to EF in SAD.

For SAD, underlying aetiology and maintaining factors are believed to focus primarily on fear circuitry380, maladaptive cognitions, negative evaluation381 and avoidance109. Research suggests that these may be modulated by impaired top-down connectivity between the ventromedial prefrontal cortex (VMPFC) the amygdala382 and aberrant processing in the default mode network (DMN), leading to disturbed self-evaluative and self-referential processing382. Significant impairment in social functioning has been reported in SAD109 however, the relationship between EF and disability in SAD has not been specifically studied.

Page | 104

5.1.1 Study aims

In this study, we aim to address the relationship between EF and disability in young adults with disorders characterised by social impairment. As far as we are aware, no study has yet compared EF performance across these disorders. The first aim of this study was to determine

EF outcomes in treatment seeking young adults with presenting primary diagnosis of ASD,

SAD, and EP, and a neurotypical control group. It was predicted that young adults diagnosed with ASD and EP would show broad EF impairments in comparison to those with SAD and neurotypical controls. The second aim of this study was to evaluate the predictive value of EF to disability across and within these cohorts. It was predicted that EF would predict disability for participants with ASD and EP but not SAD. Finally, we were interested in the degree that performance measures or self-report measures predicted disability.

5.2 Method

5.2.1 Participants

The University of Sydney Ethics Committee approved the research protocol (No: 2013/352) for this project and informed consent was obtained directly from each participant prior to inclusion in the study. A cohort of young adults (N= 253; Age: M = 23.16, SD = 5.80) presenting for treatment and/or social skills development at the Autism Clinic for Translational

Research (ACTr) and Headspace Brain and Mind Centre clinics were recruited into the study.

Participants met criteria for the primary presenting disorder of either Autism Spectrum

Disorder (ASD; N = 60), First Episode Psychosis (EP; N =58), or Social Anxiety Disorder

(SAD; N = 76). Clinical diagnoses were made by qualified clinicians at the ACTr, based on standardised clinical interviews and diagnostic assessment instruments. Neurotypical control participants (TYP; N = 59) were recruited through advertising at a number of different university websites. All participants were screened and excluded from the study if they had

Page | 105

intellectual disability (IQ<70) or current substance dependence. The TYP participants were excluded if they reported a mental health diagnosis (past or current), were currently experiencing significant levels of depression, anxiety or stress, (DASS-21), or if they scored at clinical cut-offs for SAD, (SIAS) and/or ASD (AQ).

5.2.2 Measures

A detailed description of all measures is provided in Supplementary Table S5.1. The assessment battery comprised of a combination of diagnostic assessments, neuropsychological tests and self-report measures of EF, mood and disability. Diagnostic assessments consisted of standardised clinical interviews for the diagnosis of ASD (ADOS), EP (SCID-I) and SAD

(ADIS), psychotic symptoms were assessed with the SAPS and SANS scales. Neuropsychological tests comprised of assessments of pre-morbid IQ (WTAR) and performance measures of EF assessing the domains of ‘Set Shifting’ (CANTAB-IED), ‘Mental Flexibility’ (TMT-B) and

‘Verbal Fluency’ (COWAT). The cognitive domains of ‘Sustained Attention’ (CANTAB-

RVP) and ‘Psychomotor Speed’ (TMT-A) were also assessed. Self-report ratings of EF and disability were based on the BRIEF and WHODAS respectively. Symptom severity was assessed by the DASS-21, AQ and the SIAS.

5.2.3 Data analysis

Data analysis was performed using the IBM SPSS Statistics version 24. Univariate (ANOVA) and Multivariate Analysis of Variance (MANOVA) examined differences between the groups on EF and disability. Two-sided tests with Bonferroni confidence interval adjustments were made for all statistical comparisons, alpha was set at p=0.01 to control for multiple comparisons. Multiple Regression (MR) analyses examined the predictive value of the EF measures on disability with each EF measure’s raw score entered individually in the model.

The EP group did not complete the BRIEF and MR analysis for this group were based on

Page | 106

performance only measures of EF. For each of the MR analyses, IQ, Education, Depression and EF performance measures were entered in Step 1 and where available, the BRIEF was entered in Step 2.

5.3 Results

5.3.1 Statistics

Sample size was based on the number of participants tested at the conclusion of the study. Our sample size per group exceeded the suggested minimum (n=30)383 for violations against the assumptions of normality and equality of variance in MANOVA. The MANOVA was carried out with and without the two covariates (IQ, Education) and the statistical outcomes were comparable. Results are reported for the analysis including covariates. All assumptions for multiple regression were met with the exception of multivariate outliers as assessed by

Mahalanobis Distance. An examination of the largest outliers together with the Statistic for

Cook’s Distance indicated that they do not exert an influence on the model383 and no further action was taken.

5.3.2 Demographics

Participant demographic information is presented in Table 5.1. No significant gender

2 differences were observed between the diagnostic and control groups, χ (3, N=253) = 2.73, p>0.05.

The overall significant effect for Age (F3, 249=4.33, p=0.005) did not hold for pairwise comparisons. Significant overall effects were observed for Education, (F3, 249=13.85, p=<0.001), IQ, (F3, 249=18.48, p<0.001), and Depression, (F3, 224=34.09, p<0.001). Pairwise comparisons showed that: the TYP group had significantly more years of Education compared to each of the clinical groups; the EP group had significantly lower IQ compared to each of the other groups. The clinical groups reported higher levels of depression compared to the TYP group with no differences between the clinical groups.

Page | 107

5.3.3 Group comparison of EF performance/self-report measures and disability self- report measures:

For the distribution of participant responses for the neuropsychological and self-report measures of EF, refer Figure 5.1. Participants’ responses on all measures are summarised in

Table 5.2.

For the corrected model of neuropsychological measures, a significant overall MANOVA effect was observed for the overall diagnosis, (Hotelling’s T: F=4.23, p<0.001). Follow-up analyses showed significant overall effects for the domains of Psychomotor Speed, (F3,

247=11.37, p<0.001), Mental Flexibility, (F3, 247=7.46, p<0.001), Sustained Attention, (F3,

247=9.37, p<0.001), Shifting, (F3, 247=4.34, p=0.005), Phonemic Fluency, (F3, 247=7.09, p<0.001) and Semantic Fluency, (F3, 247=3.90, p=0.009).

Post-hoc pairwise comparisons indicated that the ASD group showed the greatest level of EF impairments. The ASD group performed significantly worse compared to at least one or more of the other groups on the domains of: Psychomotor Speed, (ASD < EP, ASD < SAD and ASD

< TYP); Mental Flexibility, (ASD < TYP and ASD < SAD); Sustained Attention, (ASD <

TYP); Phonemic Fluency, (ASD < TYP) and Semantic Fluency, (ASD < TYP). The EP group was significantly impaired on Sustained Attention (FEP

(EP

The ANOVA analysis comparing differences between groups on the BRIEF (overall score) was significant, (F2, 108=20.05, p<0.001). Pairwise comparisons revealed that both ASD and

SAD groups reported similar and significantly higher levels of EF impairment compared to the

TYP group. The overall MANOVA analysis for Diagnosis was significant for the BRIEF clinical scales, (Hotelling’s T: F=5.63, p<0.001). Pairwise comparisons showed that the ASD group reported significantly impaired EF on all clinical scales compared to the TYP group and

Page | 108

on the Self-Monitor subscale compared to the SAD group. The SAD group also reported significant impairment on most clinical scales compared to the TYP group. The non-significant results were for the scales of Inhibit, Self-Monitor and Organization of Materials.

A significant overall ANOVA effect was observed for Diagnosis for the WHODAS (overall score), (F3, 218=42.23, p<0.001). Pairwise comparisons indicated that each of the clinical groups reported significantly higher levels of overall disability compared to the TYP group with no differences between the clinical groups. The overall MANOVA analysis of the WHODAS domains was also significant (Hotelling’s T: F=10.71, p<0.001). Each of the clinical groups reported significant impairment compared to the TYP group for each of the six WHODAS domains, in addition the SAD group demonstrated significantly higher levels of impairment on the ‘Getting along’ domain compared to the EP group.

5.3.4 Effects of EF neuropsychological predictors on disability:

The first multiple regression examined the relationship of the predictors of Diagnosis, EF performance measures, IQ, Education and Depression on disability (Table 5.3). The model for

EF performance measures and disability was significant across the study cohort F (10, 209=

37.11, p<0.001) and explained 64.0% of the total variance, with significant predictors

Education (β=-0.169, p<0.001) and Depression (β=0.726, p<0.001). As Diagnosis was not a significant predictor no follow up interaction models were examined.

5.3.5 Additive effect of EF self-report predictor on disability

A second regression analysis was completed for the participants that completed the self-report measure of EF (excludes EP group). The procedure outlined above was followed for Step 1, and the self-report measure (BRIEF) was entered in Step 2 (Table 5.4). Model 1 accounted

64.0% of the total variance (F10, 97=17.22, p<0.001) with the only significant predictor

Page | 109

Depression (β=0.726, p<0.001). When the BRIEF was entered in Step 2 (method Enter), Model

2 accounted for an additional 17.0% of the variance, (F1,96=86.14, p<0.001) with the following significant predictors: Diagnosis, (β=0.583, p<0.001), Education, (β=0.161, p=0.009), IQ,

(β=0.227, p<0.001), IEDTotal errors, (β=-0.146, p=0.007), Depression, (β=0.325, p<0.001) and

BRIEF (β=0.967, p<0.001). A follow-up analysis was performed to examine the interaction of diagnosis with the five significant predictors (Diagnosis x Education, Diagnosis x IQ,

Diagnosis x IEDTotal errors, Diagnosis x Depression and Diagnosis x BRIEF. Depression was a significant predictor of disability for the ASD (B=0.621, p<0.001) and the SAD (B=0.711, p<0.001) groups. The BRIEF significantly predicted disability for the ASD group only

(B=0.248, p=0.001). None of the other interaction comparisons were significant.

5.4 Discussion

This study is the first to compare underlying cognitive markers for three disorders characterised by social impairment, EP, SAD, and ASD, and specifically evaluate EF and its role in predicting disability. Our study showed that, despite no apparent impairments in intellectual function, ASD participants showed significant and broad impairments across the domains of

Mental Flexibility, Sustained Attention and Fluency, as well as Psychomotor Speed.

Participants with EP only showed impairments on Sustained Attention and Set Shifting.

Interestingly, despite adequate EF performance on the objective measures, SAD participants self-reported EF impairment that was of a similar degree to those diagnosed with ASD. Our findings suggest that different cognitive-circuitry is associated with social impairment in these cohorts and in particular, broad EF connectivity377 in ASD, attentional switching338 in EP and self-referential processes382 in SAD. Further research is now required linking the neurobiological underpinnings of these to this disability.

The second aim was to establish which EF measures predicted disability. Results showed that the self-report measure of EF and depression predicted disability overall. The relationship

Page | 110

between depression and disability was significant across both the ASD and SAD groups however, the self-report measure of EF explained additional proportion of the disability variance for the ASD group only. This study therefore provides evidence that neuropsychological deficits in EF are particularly pronounced in ASD and may be particularly important in terms of predicting disability.

For the ASD participants, the majority of EF performance measures differentiated these participants from the TYP group. While we acknowledge there was variability even within the

ASD group itself, this study supports assertions of broad executive dysfunction in young adults with ASD that are likely neuro-developmentally driven. We have previously argued that broad connectivity models of brain development, rather than region specific brain processes, underpin these impairments27. Interestingly, the EP group performed significantly worse than the TYP group on the RVP and IED, tests which purport to assess attentional processes87. This finding is in line with involvement of the DLPFC384. These domains have previously been shown to predict conversion of clinical high risk individuals to psychosis385 and may reflect a higher level of impairment. The lack of any further EF impairment in this group supports empirical findings that broad EF in early stages of psychosis remains intact386. There was no evidence of

EF impairment in SAD on neuropsychological measures, which was expected given that the underlying aetiology for SAD is focused on fear circuits380 and maladaptive cognitions109.

While we did not have data available for EP participants, both SAD and ASD participants reported similar and significant impairment on the EF self-report measure. This was despite the SAD group showing objective performance on all EF measures that was similar to neurotypical controls. A number of factors may contribute to this finding. Firstly, it may extend the adverse influence of maladaptive cognitions387 and self-focused attention388 beyond social performance to other areas of self-evaluation including EF. That is, individuals with SAD negatively evaluate their own adaptive functioning in relation to EF. This might suggest that

Page | 111

self-report measures of EF are significantly influenced by anxiety driven biases and exploration of the role of anxiety across all of these cohorts on EF self-report measures is required.

Secondly, it may reflect the assertion that self-report ratings of EF assess a different cognitive construct that is not constrained by performance outcomes of objective cognitive measures but instead is goal oriented and reflects the individual’s beliefs relevant to these goals108. This may be particularly pertinent in individuals with SAD given their intact performance on objective measures of EF and may augment reporting of executive dysfunction in ASD individuals.

Partial support for this proposal was found by the observed low correlation between BRIEF domains and neuropsychological measures of EF389 (supplementary Table S5.2), suggesting they might tap into different cognitive processes. Alternatively, rating measures of EF may be more ecologically valid and thus better capture functional outcomes associated with EF 104.

All of our young adult clinical groups presented to our specialist assessment clinics and reported significant disability compared to the neurotypical control group as measured by the

WHODAS. Results showed that in addition to ‘Depression’, the self-report measure of EF was the strongest predictor of disability in this cohort but, within groups, this relationship was significant only for the ASD group. In contrast, neuropsychological measures of EF were largely unrelated to disability in all groups. The lack of relationship between EF and disability in the other clinical groups indicates that other factors may be more important including affective states349 and co-morbid conditions390. Further research is needed to elucidate the specific contribution of these factors to disability, which ultimately is linked to participation in society, educational and vocational outcomes265, 391.

Our study has several limitations. Our performance measures of EF although broad did not encompass all commonly accepted subdomains of EF. We do note, however, our previous work that has highlighted the relative equivalence of EF domains in ASD and across development27.

Further, the BRIEF was only completed for part of our cohort (ASD/SAD/TYP). Although our

Page | 112

results on the relationship between the BRIEF and disability are quite robust, they rely on a self-report measure of disability. Further examination of these against objective measures of disability and on larger samples would further inform the significance of these findings.

5.4.1 Conclusions

This is the first study to examine both objective and subjective markers of EF across three clinical groups of young adults with social impairment and to evaluate their relationship to disability. The study lends support to the importance of executive dysfunction in the ASD population both in differentiating between clinical groups and in self-reported EF predicting disability. It also indicates EF deficits are not likely primary contributors to disability for SAD and EP, despite the SAD group reporting EF impairment on self-report measures and the EP group performing worse on tests of sustained attention and attentional shifting. The EF outcomes across our cohort support impaired self-referential processing that may relate to

DMN processes in SAD, broad executive dysfunction that may be primarily driven by aberrant connectivity in ASD and impaired attentional processes driven by the DLPFC in EP. Overall, these findings have treatment implications for young adults with social impairment and suggest that for ASD therapeutic support may need to include cognitive training of EF, whereas in SAD and EP focus on maladaptive cognitions and attentional processes respectively may be more appropriate.

Page | 113

Table 5.1. Demographic characteristics and participants responses to the Depression, Anxiety and Stress Scale (DASS)

ASDa FEPb SADc TYPd Statistical analysis

Gender N % N % N % N % Pearson Chi-Square

2 Male 38 63.3 37 63.8 41 53.9 31 52.5 χ (3, N=253)=2.73, p>0.05

Demographic Mean SD Mean SD Mean SD Mean SD ANOVA Post-hoc Multiple Comparisons with Bonferroni characteristics adjustment (p=0.01)

N 60 58 76 59 a vs b a vs c a vs d b vs c b vs d c vs d

Age (years) 24.11 7.27 21.79 4.05 22.11 5.64 24.88 5.30 F(3,249)=4.33* ns ns ns ns ns ns

Education (years) 12.43 1.96 12.24 2.20 12.61 1.84 14.37 2.19 F(3,249)=13.85** ns ns ** ns ** **

IQ 107.56 8.66 100.71 9.37 111.28 6.56 106.99 8.30 F(3,249)=18.48** ** ns ns ** ** ns

ASDa FEPb SADc TYPd Statistical analysis

DASS-21 Mean SD Mean SD Mean SD Mean SD ANOVA Post-hoc Multiple Comparisons with Bonferroni

adjustment (p=0.01)

N 55 49 65 59 a vs b a vs c a vs d b vs c b vs d c vs d

Depression Anxiety 19.38 13.44 17.38 12.24 22.77 12.03 3.73 4.83 F(3,224)=34.09** ns ns ** ns ** **

Stress Scale –

Depression Score

Note: *p≤0.01, **p<0.001

Table 5.2. Group characteristics of executive function neuropsychological and self-report measures and disability self-report ratings

ASDa FEPb SADc TYPd Statistical analysis

Neuropsychological Mean SD Mean SD Mean SD Mean SD MANOVA Post-hoc Multiple Comparisons with measures Bonferroni adjustment (p=0.01)

N=60 N=58 N=76 N=59 a vs b a vs c a vs d b vs c b vs d c vs d

TMT-A 35.61 19.37 28.41 9.22 26.67 9.58 21.94 5.96 F(3,247)=11.37** ** ** ** ns ns ns

TMT-B 80.38 40.74 72.95 38.64 60.47 19.90 50.45 15.15 F(3,247)= 7.46** ns * ** ns ns ns

RVP-A' 0.88 0.06 0.87 0.05 0.91 0.05 0.93 0.04 F(3,247)= 9.37** * ns ** ns ** ns

IEDtotal errors 27.04 27.88 35.61 32.14 19.08 18.07 14.62 12.89 F(3,247)= 4.34* ns ns ns ns * ns

COWATphonemic 32.72 11.58 36.09 10.14 37.28 9.27 42.29 9.91 F(3,247)= 7.09** ns ns ** ns ns ns

COWATsemantic 19.33 5.42 19.29 4.17 20.49 4.86 22.47 4.71 F(3,247)= 3.90** ns ns * ns ns ns

ASDa FEPb SADc TYPd Statistical analysis

Self-report Mean SD Mean SD Mean SD Mean SD ANOVAa / Post-hoc Multiple Comparisons with measures MANOVAb Bonferroni adjustment (p=0.01)

BRIEF (N) 48 n/a 35 30 a vs b a vs c a vs d b vs c b vs d c vs d

a BRIEFGEC 144.79 28.98 135.5 23.62 94.33 19.80 F(2,108)=20.05** n/a ns ** n/a n/a **

b BRIEFInhibit 15.35 3.81 14.40 3.29 10.97 2.33 F(2,108)= 7.32** n/a ns ** n/a n/a ns

b BRIEFShift 13.60 3.03 11.60 3.06 11.53 3.57 F(2,108)=19.50** n/a ns ** n/a n/a **

b BRIEFEmotional Control 20.96 5.78 19.49 5.61 13.40 3.63 F(2,108)=11.48** n/a ns ** n/a n/a **

b BRIEFSelf Monitor 12.25 3.14 9.40 2.40 7.63 1.88 F(2,108)=17.72** n/a ** ** n/a n/a ns

b BRIEFInitiate 17.60 4.28 17.69 3.09 11.00 2.59 F(2,108)=22.37** n/a ns ** n/a n/a **

b BRIEFWorking Memory 16.79 4.36 16.06 4.04 10.53 2.91 F(2,108)=12.95** n/a ns ** n/a n/a **

b BRIEFPla Organize 20.79 4.82 20.43 4.46 13.27 3.31 F(2,108)=14.71** n/a ns ** n/a n/a **

b BRIEFTask Monitor 11.90 2.65 12.11 2.41 8.40 1.96 F(2,108)=11.72** n/a ns ** n/a n/a **

BRIEFOrganization 15.54 4.27 14.94 3.96 11.30 2.69 F(2,108)= 5.79** n/a ns * n/a n/a ns

b Materials

ASDa FEPb SADc TYPd Statistical analysis

Self-report Mean SD Mean SD Mean SD Mean SD ANOVAa / Post-hoc Multiple Comparisons with measures MANOVAb Bonferroni adjustment (p=0.01)

WHODAS (N) 54 50 62 58 a vs b a vs c a vs d b vs c b vs d c vs d

a WHODAStotal 35.59 18.67 37.26 15.84 38.01 16.18 7.51 6.91 F(3,218)=42.23** ns ns ** ns ** **

b WHODASDomain1 39.74 22.74 40.18 18.17 40.21 18.89 10.52 11.87 F(3,218)=26.77** ns ns ** ns ** **

b WHODASDomain2 21.88 21.29 24.88 21.89 18.25 21.18 2.37 5.59 F(3,218)=10.73** ns ns ** ns ** **

b WHODASDomain3 18.64 21.78 15.87 16.89 13.87 16.63 1.38 3.95 F(3,218)= 8.39** ns ns ** ns ** **

b WHODASDomain4 49.81 24.08 40.54 27.85 58.57 22.84 10.63 14.29 F(3,218)=43.34** ns ns ** ** ** **

b WHODASDomain5 40.19 31.71 40.20 28.10 42.10 26.44 10.00 14.51 F(3,218)=14.56** ns ns ** ns ** **

b WHODASDomain6 35.87 25.51 44.85 18..18 42.35 21.46 6.25 9.57 F(3,218)=37.55** ns ns ** ns ** **

*p≤0.01, **p≤0.001

Note: TMT-A=Trail Making Test A; TMT-B=Trail Making Test B; RVP=Rapid Visual Processing; IED=Intra/Extra Dimensional Shift; COWAT=Controlled

Oral Word Association Test; BRIEF=Behavioral Rating Inventory of Executive Function; WHODAS= WHO Disability Assessment Scale

Table 5.3. Effects of executive function neuropsychological predictors on disability

B SE B β

Constant 49.673 20.081

Diagnosis 0.662 0.882 0.036

Education -1.521 0.414 -0.169**

IQ 0.044 0.112 0.020

TMT-A 0.176 0.088 0.114

TMT-B -0.035 0.035 -0.057

RVP-A -25.174 19.108 -0.071

IEDerrors 0.010 0.037 0.012

Fluencyphonetic -0.068 0.100 -0.037

Fluencysemantic -0.093 0.209 -0.023

DASSdepression 1.078 0.068 0.726**

** p < 0.001

Page | 119

Table 5.4. Additive effect of executive function self-report predictors on disability

B SE B β

Step 1

Constant 49.673 29.476

Diagnosis 0.662 1.295 0.036

Education -1.521 0.607 -0.169

IQ 0.044 0.164 0.020

TMT-A 0.176 0.129 0.114

TMT-B -0.035 0.052 -0.057

RVP-A -25.174 28.048 -0.071

IEDerrors 0.010 0.055 0.012

Fluencyphonemic -0.068 0.147 -0.037

Fluencysemantic -0.093 0.306 -0.023

DASSdepression 1.078 0.099 0.726**

Step 2

Constant -119.902 28.223

Diagnosis 10.578 1.426 0.583**

Education 1.456 0.547 0.161*

IQ 0.500 0.130 0.227**

TMT-A -0.152 0.101 -0.099

TMT-B 0.073 0.039 0.120

RVP-A -16.501 20.490 -0.046

IEDerrors -0.117 0.042 -0.146*

Fluencyphonetic -0.286 0.110 -0.155

Fluencysemantic -0.188 0.224 -0.047

DASSdepression 0.482 0.097 0.325**

BRIEF 0.585 0.063 0.967**

** p < 0.001

Chapter 6: Machine learning for differential diagnosis between clinical disorders with

social impairment

Study objectives:

The objectives of empirical study 5 were to examine the following:

(a) the sensitivity and specificity of the cognitive domains of EF, social cognition,

visuomotor learning and memory, in differentiating between cohorts with social

impairment and a neurotypical control group

(b) whether discriminating profiles of cognitive measures can be identified that

discriminate between the neurodevelopmental cohort of individuals with ASD and EP

from a clinical group (SAD) also characterised with social impairment.

(c) whether a neuropsychological profile can distinguish each of the ASD, SAD and EP

cohorts from each other.

Page | 122

Machine learning for differential diagnosis between clinical disorders with social

impairment

Demetriou, E.A.1, Park, S.H1, Ho, N., Pepper, K.L., Song, Y. J. C., H Naismith, S.L.,

Thomas, E.E., Hickie, I.B., Guastella, A.J.

1Joint first authors

(under review in Schizophrenia Bulletin).

Page | 123

ABSTRACT

Differential diagnosis in adult cohorts with social difficulty is confounded by comorbid mental health conditions, common aetiologies and shared phenotypes. Identifying shared and discriminating profiles can facilitate intervention and remediation strategies. The objective of the study was to identify salient features of a composite test battery of cognitive and mood measures using a machine learning paradigm in clinical cohorts with social interaction difficulties compared to neurotypical controls. We recruited clinical participants who met standardised diagnostic criteria for Autism Spectrum Disorder (ASD: N=62), Early Psychosis

(EP: N=48) or Social Anxiety Disorder (SAD: N=83) and compared them with a neurotypical comparison group (TYP: N=43). Using five machine-learning algorithms and repeated cross- validation, we trained and tested classification models using measures of cognitive and executive function, lower and higher order social cognition and mood severity. Performance metrics were the AUC and Brier Scores. Sixteen features successfully differentiated between the groups. The control versus social impairment cohorts were differentiated by social cognition visuospatial memory and mood measures. Importantly, a distinct profile cluster drawn from social cognition, visual learning, executive function and mood, distinguished the neurodevelopmental cohort (EP and ASD) from the SAD group. The mean AUC range was between 0.892 and 0.915 for social impairment versus control cohorts and, 0.731 to 0.777 for

SAD vs neurodevelopmental cohorts.

This is the first study that compares an extensive battery of neuropsychological and self-report measures using a machine learning protocol in clinical and neurodevelopmental cohorts characterised by social impairment. Findings are relevant for diagnostic, intervention and remediation strategies for these groups.

Page | 124

6.1 Introduction

Machine learning paradigms have facilitated the evaluation of complex datasets392, 393 and provide a dynamic framework to enhance comparisons between groups that may share neurodevelopmental, clinical or cognitive profiles394. Machine learning draws from statistical estimation and pattern recognition theories and aims to identify useful patterns in large amounts of data through automated computational algorithms. In medicine and psychology, the resultant algorithms have led to insights in clinical classification within395 and between clinical cohorts396, transdiagnostic subtyping of mental health symptoms366 and comparative lifetime health outcomes397. Such research may contribute to improved profiling of other cohorts such as Autism Spectrum Disorder (ASD) and Schizophrenia (SCH) given shared genetic liability398 and theorised common aetiologies associated with social cognition399 and EF72, 400 processes.

The clinical sub-groups of ASD and SCH have drawn much debate about similarities and differences that might exist between the two diagnoses401, 402. There is considerable empirical support of shared genetic, neurocognitive and behavioural pathways between ASD and SCH403-

405. In both, co-morbidities appear higher than expected population outcomes406 and impairments in cognitive function appear similarly in domains of social cognition and executive function (EF)35, 407. For ASD, diagnosis may be made as early as 18 months of age but the developmental course of psychosis is vastly different, with a slow progression beginning with social withdrawal and early psychosis (EP) that typically begins in later adolescence and early adulthood385. In these cases, a third of people who develop EP will go on to develop SCH408. There has been limited research exploring cognitive markers that may assist differential diagnosis. Such comparisons are particularly useful in early adulthood prior to the chronic manifestation of SCH symptoms to permit early differentiation of these disorders.

Page | 125

The Diagnostic and Statistical Manual of Mental Disorders (DSM-5)1 defines six cognitive domains as key domains for the assessment of neurocognitive disorders. These are complex attention, executive function (EF), learning and memory, language, perceptual–motor function and social cognition. While there is evidence of difficulties across some of these cognitive domains for ASD and EP/SCH, few studies have directly compared the two cohorts.

In studies of complex attention, impairment in sustained attention has been reported in ASD409 and on a composite battery of attention measures in EP410. A recent comparison between ASD and EP35 showed the latter was significantly more impaired on attentional processes and this may relate to atypical processing in the prefrontal cortex384, 409. Empirical findings in other cognitive domains are mixed and, in part dependent on the modality studied (verbal versus visuospatial). Studies with participants diagnosed with ASD have reported impaired performance in verbal learning411, visuospatial short term memory (STM)409, whilst others, noted superior visual19 and comparable verbal STM19, 411. In psychotic populations, verbal

STM and learning have also been noted to be impaired400, 412, 413 but there have been mixed results for visual learning412, 413. A study examining language domain measures in ASD with a neurotypical comparison group414 found no differences between them. For the perceptual- motor domain difficulties have been reported for ASD40 and EP413, based primarily on performance on the Rey Complex Figure Test (RCFT).

There is, however, a greater amount of research examining social cognition and EF domains and their contribution to symptoms and disability. Lower and higher order social cognition5 performance has been shown to be reduced in participants diagnosed with either ASD or EP.

These include performance on tests of emotion recognition415, 416 and theory of mind tasks144,

417, 413. In relation to EF, reduced performance has been reported for ASD40, 70, 72 and EP418, 419.

A recent meta-analysis across six EF domains72 points to broad executive problems in ASD, likely characterised by aberrant neural network connectivity420. For EP, impairment in

Page | 126

attentional shifting has been reported35 with mixed findings across other domains including working memory and abstract thinking412, 421.

In summary, despite considerable evidence of shared genetic liability and common phenotype profiles for ASD and EP, there has been limited direct research comparing the two groups.

Such comparisons may also present empirical evidence that contribute to the debate of the neurodevelopmental basis of SCH. Such analysis should incorporate a comprehensive range of assessment measures across cognitive domains and neurodevelopmental and comparison cohorts of interest. The emergent trend for utilising machine learning algorithms for complex databases is an appropriate statistical approach for such analysis.

The broad goal of this study was to use machine learning on a large dataset of multiple cognitive domain measures and mood self-appraisals. The aims were to identify differentiating profiles between neurodevelopmental (ASD, EP), clinical (Social Anxiety Disorder-SAD) and neurotypical (TYP) comparison groups. The SAD group presents an important comparison cohort given that EP422 and ASD423 are both associated with substantially elevated levels of social anxiety. Furthermore, there is a period of prodromal features that are difficult to distinguish between social anxiety and early psychosis424. Identifying discriminating profiles between the two conditions would facilitate early intervention. Our assessment battery included multiple measures across the domains of complex attention, executive function, learning and memory, perceptual-motor function and social cognition. Self-report measures of depression, anxiety and stress were also included in the study given research evidence demonstrating high levels of co-morbid depression in SAD425 and depression and anxiety comorbidities in ASD254 and EP426. To our knowledge, this is the first study to compare ASD and EP with clinical and non-clinical comparison groups across broad cognitive domains and affective states. The first aim was to identify a profile that may distinguish between the combined social impairment cohort and control group. The second aim was to identify

Page | 127

variables that differentiated the neurodevelopmental cohort from the SAD group. The third aim was to determine whether each of the ASD, SAD and EP groups could be distinguished on a subset of measures from the other clinical groups and from each other. We predicted that the neurotypical control group would be distinguished from the social impairment cohort on self- appraisal measures of depression and anxiety, given the reported high comorbidity rates in the clinical groups. We did not expect any of the cognitive measures to contribute to a discriminating profile between the clinical groups, given the mixed cognitive profile of the social impairment cohort. Second, we predicted that the neurodevelopmental cohort would be distinguished from the clinical comparison group on measures of attention, psychomotor speed, social cognition, executive function and visuo-motor performance. This is based on literature findings that these domains are generally intact in SAD349 but impaired in ASD409, 411 and EP413.

Third, we predicted that the ASD and EP groups would be distinguished from each other on measures of complex attention given empirical support for impaired neural circuitry underpinning attention networks in EP338. No specific predictions were made for the comparisons of ASD versus SAD/EP and EP versus SAD/ASD.

6.2 Method

Ethics

Ethics approval was given by the University of Sydney Ethics Committee (Protocol number

2013/352). Informed consent was obtained from each of the participants by postgraduate research students and trained clinicians.

Participants

Our dataset consists of clinical participants who have met standardized diagnostic criteria for

ASD (N=62), EP (N=48) or SAD (N = 83). Participants were referrals from the Autism Clinic for Translational Research, Anxiety Clinic and headspace clinics, at the Brain and Mind Centre,

Page | 128

University of Sydney. Neurotypical control study volunteers (TYP=43), were recruited separately through advertising at university websites.

Participants for the social impairment cohort were sequentially recruited from a cohort of young adults who presented for treatment and/or social skills development at our Brain and

Mind Centre clinics at the University of Sydney. They were assessed using standardized diagnostic instruments and clinical case files, and assigned to a clinical group if they met criteria for that condition. The Autism Diagnostic Observation Schedule – 2nd edition (ADOS-

2)427 was used to assess participants for the ASD group, the Anxiety Diagnostic Interview

Schedule (ADIS-IV/V)359, 428 was used to assess those for the SAD group, and the Structured

Clinical Interview for DSM-IV Axis I Disorders (SCID-I)429 was used to assess those for the

EP group. All participants were screened for IQ using either the two subtest version of the

Wechsler Abbreviated Scale of Intelligence (WASI)430 or the Wechsler Test of Adult Reading

(WTAR)431, and were excluded if IQ was below 70. Prospective TYP participants were excluded if they reported past or current mental health diagnosis, or of they scored above cut- off scores on the Depression, Anxiety and Stress Scale-(DASS-21)432, the Social Interaction

Anxiety Scale (SIAS)433 or the short-form of the Autism Quotient (AQ-10)434.

Assessment battery

In this study, we utilized both neuropsychological (objective) and self-report (subjective) measures of social cognition, cognitive and executive function as well as self-report measures of affective states (depression, anxiety and stress). A detailed summary is presented in

Appendix 5, Table S5.1

Data selection

Participants and variables with more than 50% missing values were removed from the dataset and the remaining missing data values were imputed with multivariate imputation with chained equations (MICE)435 with 10 iterations using predictive mean matching for missing values.

Page | 129

Plausibility and consistency of the imputed values was visually inspected through density plots of the observed and imputed data, and the first imputed dataset was selected for downstream analysis.

Machine learning

We applied machine learning to build models that can classify between our groups of interest.

In particular, we selected five algorithms that can also perform variable selection in order to ascertain a variable’s contribution to the model. The five algorithms are Area Under the Curve

Random Forests (AUCRF)436, Boruta437, Lasso regression438, Elastic net regression439 and

Bayesian Additive Regression Trees (BART)440.

The AUCRF and Boruta algorithms are both based on the Random Forest (RF)441 algorithm.

RF uses bootstraps of samples to build a forest of decision trees with variables as nodes of the tree. Furthermore, RF has an internal variable importance ranking system that describes the decrease in node impurity. A higher-ranking variable is one that splits the samples into more pure groups.

AUCRF recursively builds RF models whilst eliminating the lowly ranked variables. The optimal set of variables are those used in the RF model with the best performance. For AUCRF, the metric for performance is the Area under the Receiver-Operator Curve (AUC), which describes the model’s true positive rate (sensitivity) and false positive rate (1-specificity) across different thresholds for binary classification.

Boruta uses RF to compare a variable’s original importance score to its importance score from a permutation of that variable. Permutations break the relationship between the predictor and the response variables and, hence, are expected to decrease the predictive value of a variable.

Variables with higher importance scores than in its permuted form are considered important.

RF models were built with the optimal sets of variables as identified by AUCRF and Boruta, and tested to obtain performance metrics. For each RF model in AUCRF and Boruta and for

Page | 130

each final RF model, we generated 1000 decision trees and, at each split of each decision tree, the best variable to use for splitting was identified from a random set of √푝 variables where 푝 is the total number of variables.

Lasso regression uses L1-regularisation that penalises coefficients with large absolute values in order to reduce overfitting. Lasso regression shrinks the coefficient of unimportant variables to zero and hence, effectively, performs variable selection. In contrast, Elastic net regression employs a linear combination of L1-regularisation and L2-regularisation, which penalises coefficients with large squared values. We used internal 5-fold cross validation to identify the optimal value for 휆 which controls the strength of the penalisation in Lasso and Elastic net, and

훼 which controls the balance between L1- and L2-regularisation in Elastic net.

In contrast to RF where trees are built from random bootstraps and independently, BART employs a sum-of-trees approach. The Bayesian foundations of BART allows for the specification of regularisation priors that ensures that each tree is weak and the use of Bayesian back-fitting442 to fit trees iteratively. Variable selection with BART involves comparing the variable’s inclusion proportions, which reflects the frequency of which the variable is chosen to be the split node, against a null distribution created from multiple permutations of the variable.

Performance and cross-validation

To measure the performance of these machine learning models, we used the AUC metric and the Brier Score443, which is the mean of squared probability errors. The AUC is a suitable statistic for inferring model accuracy and summarizes the likelihood of the model predicting a higher probability for true positive cases than true negative cases444. The AUC metric however, does not evaluate the error boundaries of the probability estimates. It has been criticised as not been clinically relevant445 in part because changes in AUC values do not directly correlate with

Page | 131

improvements in prediction446. To improve the prediction accuracy of the model it is recommended that the AUC statistic is used in combination with a second metric such as the

Brier score447. The Brier score summarises the magnitude of the error of the probability estimates and in combination with the AUC statistic presents a more accurate measure of each model’s performance.

In order to better estimate the true performance of these models on unseen data, 10-fold cross- validation was applied (Figure 6.1) repeated 10 times. In each fold, 90% of the samples were used as the training set and the remaining 10% were used for testing. In each training set, the response variable was balanced through random under-sampling of the majority class whilst the test set remained unmodified in order to represent the distribution of participants in the clinic.

Diagnostic classifications

We applied the above methodology to discriminate between different diagnostic groups.

Firstly, we classified between neurotypical controls and the combined social impairment cohort and identified important variables in the models constructed. Secondly, we repeated the methodology to classify each clinical cohort against the other two to identify discriminative features unique to that condition.

6.3 Results

Sample description

Demographic characteristics are summarized in Table 6.1. In total, 236 participants met all inclusion/exclusion criteria and had all required demographic information. The mean age of participants was 22.72 years (SD 5.83), and 96 (40.7%) were female. No significant differences were observed between the cohorts with regard to age, gender and years of education (p>0.05).

Page | 132

Model performance

Classification performance for classifying between neurotypical controls and social impairment cohorts was good with mean AUCs greater than 0.89 (Table 6.2). All five algorithms performed similarly well with BART providing the highest mean AUC (0.915) and

Boruta providing the lowest mean Brier Score (0.138). The optimal score for AUC is 1, which is indicative of a perfect test whereas a score of 0.5 reflects that the test’s performance is equivalent to random chance. Brier scores range from zero, which indicates a perfect prediction with zero probability errors, to one, which indicates the worst prediction performance.

For classification between clinical and neurodevelopmental groups, the mean AUCs were lower than that between neurotypical controls and the combined social impairment cohort, which reflects the challenge in developing a classification tool between disorders. Mean AUCs for discrimination between the ASD group and SAD and EP groups ranged from 0.643 to 0.748 with Elastic-net providing the highest mean AUC and lowest mean Brier Scores.

Variable selection

Figure 6.2, illustrates the frequency that each variable was selected across the five algorithms with repeated cross-validation. Self-report measures of depression, anxiety and stress (DASS), social cognition measures of EQ-social skills, and the cognitive measure of visuospatial STM, best discriminated between clinical and control groups.

Figure 6.3 and Table 6.3 presents the top 14 variables identified from the variables input into the three models for differentiating diagnosis between the disorders of social impairments by all five algorithms. Variables that were selected at least twenty times by all five algorithms were included. Discriminating variables were identified across cognitive domains and affective states.

Page | 133

6.4 Discussion

In this study, we used machine learning algorithms on a composite assessment battery to identify cognitive profiles that discriminate between clinical, neurodevelopmental and neurotypical comparison groups. Our results showed that a reduced set of assessment measures differentiated between the comparison groups with good discriminative ability (AUC>=0.7 and

Brier score = 0.19-0.24). Our first hypothesis was confirmed, depression, anxiety and stress discriminated the combined social impairment cohort from the comparison control group. Two measures drawn from the social cognition and learning/memory domains (social skill and visuospatial short-term memory) complemented this profile. Our second hypothesis received partial support. Three of the predicted five cognitive domains (visual learning, social cognition and EF), featured in the optimized profile discriminating between the neurodevelopmental and the SAD group. Depression was the other distinguishing feature. Finally, contrary to our third hypothesis, psychomotor speed rather than complex attention, distinguished between the ASD and EP groups. Taken together our research outcomes support and extend literature findings on distinguishing features of clinical (SAD) and neurodevelopmental (ASD/EP) groups. The results are particularly compelling given the high discriminative performance of the optimized profiles that emerged from an extensive battery across multiple cognitive domains and affective states.

The first finding of interest is that cognitive domains in addition to EF and social cognition featured in the optimized profiles. The learning/memory domain measure of visuospatial memory contributed to the combined social impairment cohort versus control discriminating profile. This is surprising given that cognitive function in SAD349 is generally intact and thus not expected to differentiate this cohort from a neurotypical control group. The finding suggests that our combined social impairment cohort shares atypicalities in maintaining visual information in short term memory. There is evidence of reduced visual working memory

Page | 134

capacity in ASD188, 448 and EP449 and our findings may in part reflect this. The shared profile with SAD however, points to processes that are more complex. A number of cognitive models predict that anxiety attenuates cognitive control and impairs working memory processes450 including visual working memory232. Our combined social impairment cohort is characterised by high levels of anxiety, and our findings may reflect the influence of anxiety on executive control.

Measures from learning, attention and psychomotor speed domains featured in the optimized profiles that discriminated between clinical cohorts. Visual associative learning contributed to discriminating the neurodevelopmental from the SAD cohort. A closer examination of this profile indicated that although all groups were comparable on overall visual learning performance, the neurodevelopmental cohort made more errors. This may reflect impaired processes specific to ASD and EP including impaired visual working memory448, 449 and slow processing speed451, 452. Attentional processes were the most salient features that discriminated the EP group from the combined ASD/SAD cohort and EP from the SAD groups. Attentional neural circuitry in EP is clearly impaired in the course of illness384 and indicates that it may have a unique role in early detection and differentiation. Psychomotor processing speed was the only distinguishing feature discriminating between ASD and EP groups. Research supports that processing speed is impaired in both groups413, 451 however, different patterns of reaction time changes may apply. There is some evidence that processing speed in EP/SCH deteriorates in later age453 whilst in ASD, processing speed has matured by adolescence 454 and is significantly impaired compared to neurotypical controls451. The discriminating profile identified here may reflect different trajectory changes. The absence of measures from other cognitive domains in the EP/ASD comparison support that these two groups have a shared phenotype across most cognitive domains.

Page | 135

The second finding of interest is that phonemic fluency was the only EF measure that contributed to a profile discriminating between SAD and the neurodevelopmental cohort.

Phonemic fluency performance is thought to be positively associated with intact frontal lobe function455 and results may indicate frontal lobe alterations in ASD377 and EP338. Given that impairment in EF is noted for both ASD40, 71 and EP385 cohorts, greater prominence of EF measures would be expected. The limited role of our other EF measures in differentiating between the clinical groups suggests EF may have greater relevance as a transdiagnostic dimension of neurodevelopment348.

Social cognition was a distinguishing feature for a number of optimized profiles. These measures featured in all profiles that included participants diagnosed with ASD, except for the

ASD/EP direct comparison. Self-appraisals for social skill (a sub-scale of the EQ questionnaire that measures difficulty in social situations), differentiated the clinical cohort from the control group. Co-morbidity with SAD has been reported for each of the ASD423 and EP456 groups and our finding of a shared profile feature likely reflects this. The neurodevelopmental cohort was distinguished from the SAD group on measures of basic emotion recognition (RMET task), identifying emotions in the absence of salient cues (movie stills task) and, in experiencing an appropriate emotion in response to another (self-appraisal of emotional reactivity/empathy).

Finally, the ASD versus SAD profile distinguished between the two groups on the overall level of empathy (EQ questionnaire). These findings highlight the salience of social cognition in the neurodevelopmental cohort and particularly for the ASD group. Considered together with the limited prominence of EF features despite known EF deficits, it suggests that social cognition is a more important domain for discriminating the ASD group from other cohorts.

The prominence of mental health features (depression, anxiety and stress) in the profile discriminating between the combined social impairment cohort and the control group, reflects the high levels of co-morbid depression425, 426 and anxiety456 reported for SAD, EP and ASD.

Page | 136

The inclusion of depression and stress self-appraisals in discriminating between the three clinical cohorts warrants further discussion and suggests that nuanced differences differentiate between the groups. The SAD group reported the highest levels of depression in our clinical cohort and EP the lowest levels of stress. Depression was the only affective state that discriminated the SAD versus neurodevelopmental cohort. This may reflect the high levels of co-morbid depression characterizing SAD425. The lower levels of stress differentiating the EP group from ASD/SAD may reflect differences in symptom severity levels on presentation to our services. Acute positive psychotic symptoms in the EP cohort were controlled prior to inclusion to our services. The SAD and ASD participants however, would be experiencing a more acute profile of their respective symptoms. This may translate to the lower levels of distress reported by the EP group. Alternatively, lower stress in EP may reflect different levels of insight. There is research support that individuals with EP, (particularly those with more impaired cognitive function) have lower levels of insight457 and may therefore report lower levels of stress.

Limitations:

There are a number of limitations in our study. First, the relatively small sample size of our cohort reduces the parameters for trainability and cross validations of our data. Second, the cognitive/affective measures included in our final analysis were limited a-priori to measures available across all comparison groups. The discriminating profiles identified here, may have been modified further, by inclusion of the additional measures. Third, our findings can only be attributed to ASD individuals without intellectual disability, as we did not include any participants with an IQ below 70. Fourth, our findings include a number of features based on self-appraisals and there is some question whether self-report appraisals by individuals with

ASD are comparable to other cohorts35, 407. Fifth, a number of participants in the diagnostic groups

Page | 137

were being treated with medication, however, we were not able to control for medication use in this study.

Conclusions

The optimized profiles identified in our study highlight the importance of evaluating multiple cognitive domains when determining discriminating profiles between clinical groups. Further, they demonstrate that our combined social impairment cohort (ASD, EP and SAD), is characterized by both shared and discriminating features. This has implications for diagnostic, intervention and remediation strategies. The discriminating profiles can thus facilitate differential diagnosis particularly when clinical cohorts are characterised by comorbid mental health conditions and shared phenotypes. Conversely, the shared profile features, provide a framework for identifying transdiagnostic dimensions for intervention and remediation programmes. The unique discriminating features (empathy and attention) that respectively characterised our ASD and EP cohorts potentially identify key target areas for early intervention programmes. To our knowledge, this is the first study that utilized measures across multiple cognitive domains and affective states. Our findings provide a framework for further research on shared and differentiating profiles of neurodevelopmental cohorts and cohorts characterized by social impairment.

Page | 138

Figure 6.1: Flow chart of training and learning algorithms

Page | 139

Table 6.1. Demographic descriptive statistics by diagnosis

All Control SAD ASD EP Significance (n = 236) (n = 43) (n = 83) (n = 62) (n = 48) Age in years 22.72 23.21 22.34 22.63 23.08 H(3) = 2.567, p = 0.463 (5.83) (5.84) (6.15) (5.55) (5.76) Gender female 96 21 28 21 26 χ2 (3) = 7.654, p = 0.054 (%) (40.7%) (48.8%) (33.7%) (33.9%) (54.2%) Education in 13.02 12.86 12.85 13.37 13.07 H(3) = 1.714, p = 0.634 years (2.19) (2.00) (2.30) (2.45) (1.81)

H: Kruskal-Wallis Test for non-normal data

Page | 140

Table 6.2. Classification performance on repeated cross-validation test sets Test set AUC’s mean (SD)

Control vs SAD vs Neuro ASD vs EP vs EP vs ASD EP vs SAD SAD vs ASD Clinical - SAD and EP ASD and developmenta SAD l (ASD and EP) AUCRF 0.894 (0.073) 0.739 (0.115) 0.643 (0.117) 0.773 (0.122) 0.740 (0.139) 0.823 (0.122) 0.719 (0.132) Boruta 0.895 (0.069) 0.744 (0.107) 0.648 (0.122) 0.780 (0.125) 0.736 (0.133) 0.835 (0.115) 0.729 (0.116) Lasso 0.892 (0.080) 0.731 (0.114) 0.729 (0.128) 0.720 (0.146) 0.723 (0.144) 0.789 (0.137) 0.809 (0.105) Elastic-net 0.898 (0.073) 0.751 (0.099) 0.748 (0.121) 0.726 (0.149) 0.722 (0.150) 0.789 (0.131) 0.802 (0.106) BART 0.915 (0.060) 0.777 (0.099) 0.734 (0.106) 0.782 (0.119) 0.754 (0.151) 0.832 (0.116) 0.787 (0.101)

Test set Brier Scores mean (SD) AUCRF 0.141 (0.032) 0.212 (0.049) 0.242 (0.035) 0.196 (0.034) 0.207 (0.048) 0.176 (0.048) 0.220 (0.051) Boruta 0.138 (0.032) 0.209 (0.043) 0.240 (0.039) 0.194 (0.038) 0.211 (0.050) 0.169 (0.042) 0.212 (0.043) Lasso 0.157 (0.049) 0.224 (0.064) 0.226 (0.063) 0.232 (0.057) 0.206 (0.083) 0.162 (0.056) 0.193 (0.048) Elastic-net 0.156 (0.048) 0.210 (0.052) 0.213 (0.056) 0.236 (0.057) 0.217 (0.089) 0.159 (0.056) 0.195 (0.052) BART 0.150 (0.023) 0.198 (0.034) 0.214 (0.026) 0.205 (0.025) 0.090 (0.029) 0.139 (0.029) 0.160 (0.030)

Page | 141

Figure 6.2: Heat map of comparisons across cohorts

Page | 142

Figure 6.3: Venn diagram of distinguishing features

Page | 143

Table 6.3. Variables discriminating between ASD, EP, SAD and neurotypical controls

Cohort N Variables TYP∩ASD/EP/SAD 5 DASS Depression, DASS Anxiety, DASS Stress, EQ Social Skill, SSP, RVP-A SAD∩ASD/EP 6 EQ Emotional Reactivity, Movie Stills - No Face, RMET, DASS Depression, PAL Total Errors, COWAT Phonemic EP∩ASD/SAD 3 IED Total Errors, IED EDS Errors, DASS Stress ASD∩EP/SAD 2 EQ Cognitive Empathy, Picture Sort - Social Script ASD ∩ SAD 1 EQ Total ASD ∩ EP 1 TMT-A EP ∩ SAD 1 RVP-A

Page | 144

Chapter 7: General discussion

Page | 145

7.1 Summary of studies in programme of research

The aim of the programme of research was to examine the efficacy of EF in the study of ASD, by synthesising the extant literature, evaluating potential moderator influences, and examining the contribution of EF in differentiating ASD from other cohorts and in predicting disability.

The study focused on youth and adults over the age of sixteen, as this age group has received less attention in the ASD literature. Developmental trajectories of EF domains are also expected to have mostly matured, thus providing a framework to assess lifelong impacts.

Chapter 1

Chapter 1 provides an overview of EF in normative and clinical populations including an evaluation of conceptual, methodological and measurement issues in the study of EF. Unitary and multifactorial models of EF were discussed in relation to underpinning mechanisms and relevance to the study of autism. A conceptual framework was proposed to unify terminology for some of the commonly studied EF domains. The executive dysfunction hypothesis as an explanatory model for ASD was reviewed in relation to findings for children, youth and adults with autism. Evidence was presented addressing moderator influences on EF including intellectual functioning, sex differences, affective states, sample selection and task characteristics.. Finally, the role of EF as an intermediate cognitive phenotype for ASD was reviewed.

Chapter 2

Chapter 2 (empirical study 1) reported the findings of the first meta-analysis to examine in a single study, six key EF domains, and to statistically evaluate a large range of potential moderator variables of executive function. The meta-analysis synthesised the literature on EF in ASD since the first introduction of autism as a diagnostic category in the DSM. The included studies compared ASD individuals with a neurotypical comparison group. Data was synthesised for neuropsychological, experimental and behavioural rating measures of

Page | 146

executive function. Further, studies were classified within three age groupings (children, youth and adults) to examine developmental differences in EF performance.

Chapter 3

Chapter 3 (empirical study 2) examined the potential moderating impact of sex differences on

EF performance in autism. This study addressed gaps in the extant literature by including a neurotypical comparison group in order to control for sex differences typically observed in the normative population. Sex differences were evaluated across the cognitive domains of EF, learning and memory, and for psychomotor processing speed.

Chapter4

Chapter 4 (empirical study 3) assessed the influence of affective states in moderating executive function. This study compared individuals with ASD, SAD and a neurotypical comparison group. Drawing on predictions of attentional control theory, the study evaluated the moderating effect of trait and state anxiety, depression and stress, on neuropsychological measures and behavioural ratings of executive function.

Chapter 5

Chapter 5 (empirical study 4), explored the diagnostic utility of EF in discriminating between

ASD and other clinical cohorts, and assessed its efficacy in predicting disability. Differences between ASD and three comparison groups (SAD, EP and neurotypical controls) were assessed on neuropsychological and behavioural measures of executive function.

Chapter 6

Chapter 6 (empirical study 5) applied machine learning algorithms on a large dataset of measures drawn across five cognitive domains (complex attention, learning and memory, visuo-motor, EF and social cognition). The aim of the study was to determine whether an

Page | 147

optimised set of cognitive and affective measures can discriminate between ASD, clinical and neurodevelopmental cohorts.

7.2 Conclusions from empirical studies in the programme of research

Empirical study 1

The meta-analysis (empirical study 1) identified broad impairment of EF in ASD that persists across the lifespan. Unlike previous meta-analyses where focus was on specific EF domains, this study synthesised data on six domains of EF to evaluate whether discrete domains differentially impact EF in autism. The lack of significant differences between the six EF domains points to broad executive dysfunction in ASD cohorts. Importantly, the meta-analysis showed that executive dysfunction in ASD continued into adulthood. Meta-regression analyses of potential moderator variables (intellectual functioning, age, sample and task characteristics) were not significant, lending further support to the premise that EF is a stable characteristic in autism. Furthermore, large effect sizes were observed for behavioural ratings of EF compared to small/medium effects for neuropsychological tests and experimental tasks. Behavioural ratings also demonstrated higher sensitivity and specificity in discriminating between ASD and the neurotypical comparison group. This finding highlights the importance of including both types of measures when researching EF in ASD, as they likely make different contributions to the study of EF. The final conclusion drawn from the meta-analysis is that a significant proportion of differences between the ASD and neurotypical comparison group could not be attributed to moderators included in the study. Two factors not considered that may further influence EF were sex and affective states. Their potential moderating effect was assessed in empirical studies 2 and 3.

Page | 148

Empirical study 2

This study concluded that sex differences for neuropsychological assessments of EF were not dependent on diagnosis and noted no sex differences for EF self-appraisal ratings. A significant female advantage was noted for the EF sub-domains of set-shifting and semantic fluency, and for learning/memory and psychomotor speed. The findings highlight the importance of including appropriate comparison groups in research in autism. These results also suggest that irrespective of possible ascertainment bias in the diagnosis of females with ASD, once diagnosed, both sexes have comparable difficulties in executive function.

Empirical study 3

Empirical study 3 examined the role of state and trait anxiety, depression and stress in moderating EF in autism. The study identified that affective states influence EF but have no moderating effects on non-EF verbal and visuospatial domains. Furthermore, the moderating influence was independent of diagnosis. The study pointed to divergent influence of trait anxiety and state stress on performance measures of EF, whereas for behavioural ratings both measures had an attenuating effect. Considered together, these findings lend support to the potential relevance of EF as a transdiagnostic marker given that diagnosis was not a significant predictor in this study.

Empirical study 4

This study reinforced the findings of empirical study 1 of a broad impairment in EF that is not domain specific. Individuals with ASD were impaired across most neuropsychological measures of EF compared to the clinical and control comparison groups. The second finding of this study is that self-appraisals show sensitivity to everyday difficulties in executive function even when objective performance is intact. Third, this study demonstrated the potential utility of behavioural ratings as a predictor of disability, but only for the ASD group, suggesting that EF self-appraisals may be a unique feature in adults with autism.

Page | 149

Empirical study 5

This study makes a unique contribution to the study of EF as it evaluates the relative influence of EF within a broad range of measures drawn from multiple cognitive domains. The study identified that EF contributes to a cognitive profile that discriminates a neurodevelopmental cohort (ASD, early psychosis) from individuals with Social Anxiety Disorder. This profile also includes social cognition and visual learning measures. Furthermore, the study indicates that social cognition and not EF is the prominent distinguishing feature in individuals diagnosed with Autism Spectrum Disorder.

7.3 Synthesis of findings of the programme of research

The series of studies undertaken in this programme of research sought to advance the knowledge of EF by investigating multiple neuropsychological and behavioural measures across multiple domains in a cohort of youth and adults with Autism Spectrum Disorder.

Comparisons with neurodevelopmental, clinical and neurotypical comparison groups allowed inferences on the efficacy of EF as a cognitive marker for differential diagnosis and as a predictor of disability outcomes. The overarching conclusion points to broad executive dysfunction in ASD across multiple domains. This dysfunction is evident in performance- based and self-appraisals of executive function, and persists into adulthood. The findings reported here concur with the executive dysfunction hypothesis111 and extend its conclusions to behavioural ratings of EF, an area that is receiving increasing focus in the study of autism.

Broad executive dysfunction in Autism Spectrum Disorder

The programme of research makes a number of novel contributions to the study of EF in autism.

The meta-analysis (empirical study 172), provides strong support for broad executive dysfunction with no significant differential contribution by individual EF domains. This novel conclusion is made possible by investigating six EF domains in a single study and extends the

Page | 150

findings of earlier meta-analytic studies that focused on individual domains70, 71. Support for broad executive dysfunction is provided by the empirical findings demonstrating no moderating influence of general intellectual functioning, sample or task characteristics, 72 or differences between male and female participants with autism458. Empirical study 472 also identified comparable impairment across multiple EF domains and complements the results of the meta-analysis. The cumulative findings of empirical studies 1, 2 and 4 suggest that the mechanisms underpinning executive dysfunction in ASD impact across multiple cognitive processes of EF contributing to the broad impairment reported in this thesis. It is important to highlight that as a group, individuals with ASD were significantly impaired compared to the neurotypical comparison group. However, there was considerable variability in their performance on neuropsychological measures of executive function. This was also evident in the broader literature by the large heterogeneity (I2 statistic) reported in the meta-analysis72 and the large standard deviations reported for the ASD group in empirical studies 2 and 435, 458.

Interestingly, only small variability was observed for the ASD group on self-report appraisals of executive function35, 458. This suggests that in individuals with ASD, their everyday experience of executive difficulties is comparable and is not commensurate with ability shown on EF performance measures. The overall findings reported here concur with models of EF that advocate a unitary or a shared factor directing executive function24, 56. Most of these models, however, with the exception of the unity in diversity model66, 68, do not adequately explain the observed variability in EF performance measures observed in the ASD cohort.

Measurement of executive function in Autism Spectrum Disorder

The second significant contribution of this research pertains to insights on the measurement of executive function. Neuropsychological measures of EF identified broad impairment in ASD but were characterised by considerable variability between and within studies compared to behavioural ratings of executive function. The large effect sizes and high sensitivity and

Page | 151

specificity of behavioural ratings compared to performance measures reported in empirical study 1 indicate that behavioural ratings better reflect differences in the experience of EF difficulties of the ASD group. Complementing the meta-analysis findings, empirical studies 2 and 4 demonstrated that individuals with ASD report a high degree of executive dysfunction in their day-to-day experiences. Furthermore, compared to performance measures of EF, self- appraisals were characterised by a small degree of variability within the ASD cohort, indicating comparable levels of difficulty irrespective of EF performance. Interestingly, self-appraisals of EF predicted disability for the ASD group only (empirical study 4), despite the SAD cohort reporting comparable levels of executive dysfunction on behavioural measures. Considered together, these findings indicate that behavioural ratings have a complementary role to performance measures in the broader research on executive function. Second, the findings indicate that behavioural ratings likely reflect different cognitive processes given the dissociation between performance measures and self-appraisals in the SAD group where EF performance was intact. In relation to ASD, these findings suggest that self-appraisals of EF likely make a unique contribution to the study of disability in this group.

The cognitive processes that may guide performance on neuropsychological measures of EF are summarised in the unitary and multifactor models discussed in Chapter 1. A number of these models were developed based on neuropsychological measures observed to be sensitive to frontal lobe functions and thus do not readily reflect on self- (or informant appraisals) of executive function. It has been proposed that self-report ratings of EF assess a different cognitive construct that is not constrained by performance outcomes of objective cognitive measures. Instead they are goal oriented and reflect the individual’s beliefs relevant to these goals108. Partial support for the argument that different cognitive processes may be involved is found in the low correlation observed between clinical scales of the BRIEF and neuropsychological measures of EF reported in empirical study 435. In SAD, these beliefs may

Page | 152

be influenced by maladaptive cognitions387 and self-focused attention388 and result in negative self-appraisals on behavioural measures of executive function. In ASD, self-appraisals of executive dysfunction are positively correlated with self-reports of disability. This suggests that unlike SAD, where self-appraisals may reflect negative self-beliefs only, executive difficulties in ASD likely reflect real impairment in day-to-day functioning.

Transdiagnostic efficacy of executive function

The third significant contribution of this research is that it advances our understanding of the efficacy of EF as a cognitive marker in ASD and across clinical cohorts. Research findings of empirical studies 1, 2 and 4 indicate that neuropsychological measures of EF effectively differentiate the ASD group from the neurotypical and other comparison groups. These findings indicate that EF is an appropriate cognitive marker for identifying EF difficulties in autism. Behavioural ratings of EF also contributed to differentiating the ASD from the neurotypical comparison group even though executive dysfunction was comparable to that reported by the SAD group. This suggests that self-appraisals may tap different cognitive processes in clinical cohorts and may be an appropriate transdiagnostic marker of executive function. The findings of empirical study 3 indicated that the moderating influences of anxiety and stress on EF were independent of diagnostic group classification, lending further support to EF also having transdiagnostic efficacy with some EF features or moderating influences not unique to Autism Spectrum Disorder.

Empirical study 5 investigated further whether EF makes a unique contribution in ASD by evaluating a composite assessment battery with measures drawn from five cognitive domains.

The findings indicated that EF performance measures contributed to discriminating the neurodevelopmental cohort (individuals with ASD and EP) from the SAD and neurotypical comparison groups. Importantly, despite the ASD group having significantly greater executive

Page | 153

dysfunction (empirical studies 2 and 4), social cognition and not EF was the prominent cognitive domain that discriminated ASD from the other groups. The findings here indicate that although ASD is characterised by significantly higher levels of broad executive dysfunction, aspects of EF are shared with other cohorts. The assessment of EF in ASD may thus have a specific role in identifying impairment at a group level, contributing to targeted interventions but also as a transdiagnostic marker that can facilitate specific interventions across clinical groups.

Cognitive models of executive function

The findings summarised above will be discussed in relation to theoretical models of EF and the efficacy of the EF construct in advancing the study of autism. These findings corroborate earlier research on performance-based measures of EF as summarised in the executive dysfunction hypothesis111 and support other research reporting real world difficulties in EF in autism104, 107. The executive dysfunction hypothesis18, 111 presents a framework for EF difficulties in ASD; it does not, however, specify underpinning mechanisms. A number of models of EF discussed in Chapter 1 account for the findings reported here of impaired performance on neuropsychological measures.

The unitary models of EF57, 59, 459 and the unity in diversity model53 predict broad EF deficits.

Unitary models of EF do not fractionate EF processes to distinct components and can thus account for the observed broad impairment in executive function. A limitation, however, of unitary models is that they do not make predictions for individual differences in EF performance. The mechanisms proposed in the revised unity in diversity model67, 68 account for broad executive dysfunction and heterogeneity in EF performance. The empirical findings of a broad impairment in EF (empirical studies 1 and 4) are consistent with the above model and likely reflect the common shared factor. The variability in performance by the ASD cohort

Page | 154

noted in individual domains of EF (empirical study 2 and 4) is also explained by the model’s prediction of genetic influences on the common shared EF factor contributing to individual differences in EF performance68. In addition, the model states that genetic influences on EF are independent of level of intellectual functioning. This concurs with the findings of the meta- analysis (empirical study 1) of no moderating effect of intellectual functioning on executive function.

Cognitive models of EF do not distinguish between neuropsychological and self-appraisal measures and do not make specific predictions on the cognitive mechanisms that may guide self-appraisals of executive function. The composite model proposed by Stuss provides a framework that may reflect on self-ratings of executive function. The model22 incorporates the metacognition factor, which has a supervisory function in integrating the energisation, motivational/emotional and executive factors towards achievement of novel tasks. Self- appraisal ratings may reflect the motivational and emotional components. In SAD, these could translate to low self-ratings given the possible role of maladaptive cognitions and negative self- appraisals evident in the condition. It is assumed that the energisation (psychomotor processing speed) and executive factors will not be impaired in SAD based on literature findings349 and evidence presented here, of intact neuropsychological function. In individuals with ASD, however, all four factors discussed above are likely to be impaired and this may translate to poor self-appraisals but also to the poor disability outcome for ASD reported in this study.

A research framework for the study of executive function

The introductory chapter outlined a broad spectrum of theoretical models and associated conceptual and methodological limitations in the study of executive function. The preceding discussion outlined key findings pertaining to the study of EF in ASD and presented a discussion of likely explanatory models. It is evident from the extant literature and the

Page | 155

cumulative research findings reported here that the study of EF would benefit from a more unified research framework. Such a framework would integrate the factors identified to be important in the research of EF and contribute to advancing knowledge on the role of EF in the

ASD cohort. A research framework for the study of EF in ASD is proposed below. It integrates literature and research findings reported in this thesis and may assist in better understanding of

EF processes in youth and adults on the autism spectrum.

It is proposed that at a first level of analysis, the research framework should include measures from both neuropsychological and behavioural assessments of EF, self-report measures of depression, anxiety and stress, and objective and self-report appraisals of disability or functional outcomes. These reflect the primary findings of this thesis reported in empirical studies 1-5. This level of analysis also draws on literature findings of executive dysfunction in

ASD70, 71, high levels of co-morbid depression and anxiety460 and disability461. It also highlights that performance on neuropsychological assessments likely reflect different cognitive processes compared to self/informant behavioural ratings108, 462.

At the second level of analysis, genetic mapping should be incorporated where possible to investigate the contribution of genetic variability in EF performance. This concurs with conclusions from the review of EF models presented in Chapter 1. Specifically, that a cognitive model as proposed by Miyake and associates68 on genetic variability influencing EF, would reflect best the marked heterogeneity in EF performance in ASD243. Figure 7.1 summarises the proposed research framework. Although all of the above factors have been identified by different research studies, their cumulative contribution in advancing knowledge on EF in ASD has not been integrated. The arrows summarise the specific findings of this thesis. The phenotype would represent behavioural outcomes based on the degree of influence of each individual component of the model. Phenotype clustering can then form the basis for targeted interventions.

Page | 156

Figure 7.1: Research framework for executive function in Autism Spectrum Disorder

Page | 157

Conclusions The series of studies reported here provide a cumulative body of evidence that brings further insights on the study of EF in Autism Spectrum Disorder. First, impairment in EF in youth and adults with ASD is significant and is not moderated by level of intellectual functioning, sex differences or sample and task characteristics, the moderating influences of anxiety and stress are not specific to ASD diagnosis. Second, the study of EF in ASD is better informed by incorporating assessments of neuropsychological measures and behavioural ratings of executive function. Third, behavioural ratings of EF are likely better predictors of disability.

Fourth, EF distinguishes a neurodevelopmental cohort (ASD and EP) from other comparison groups. It makes a smaller contribution, however, compared to other cognitive domains, particularly social cognition, in discriminating the ASD group alone from other cohorts.

Executive function may thus have greater efficacy in guiding remediation and targeted intervention strategies in ASD and as a transdiagnostic marker.

7.4 Limitations of research

A number of methodological limitations apply to the reported studies.

7.4.1 Meta-analysis

The self/informant-report questionnaires were excluded from the majority of subsequent analyses given the significant differences in effect sizes compared against psychometric tests and experimental tasks. Second, only accuracy measures were included in the analysis.

Reaction time variables were excluded given that the interpretation of reaction time variables of EF can vary between studies and experimental tasks. Finally, although a comprehensive number of moderators was considered, some factors remain that were not directly addressed in the meta-analysis but that may influence EF in ASD. These may relate to task characteristics

(e.g. task complexity, open-ended versus structured task format) or participant characteristic

Page | 158

including symptom severity, emotional states (e.g. depression/anxiety) or co-morbidities (e.g.

Attention Deficit Hyperactivity Disorder).

7.4.2 Population studied

The ASD sample in this thesis excluded individuals with intellectual disability and thus the relevance of EF findings to this population cannot be determined. A significant percentage of individuals with ASD present with intellectual disability. Prevalence rates vary depending on the study and year conducted with 30-40 percent incidence rate typically reported463, 464. In

Australia, over 30 percent of individuals with ASD also have intellectual disability465.

Addressing cognitive and EF factors in this cohort can contribute improved lifelong outcomes.

The clinical cohorts of ASD, SAD and EP self-selected to present for research at the Autism and headspace clinics and may not be representative of the wider community population with these diagnoses. A number of barriers may impede access to mental health services for people with ASD. These include difficulties in differentiating mental health symptoms in ASD from the typical diagnostic features of the condition466, 467 468 reluctance by autistic individuals to seek mental health services468 and/or lack of autism-specific training469 or support from appropriate professionals468. Given the above factors, the ASD sample presenting for this research may not be representative of the wider ASD cohort.

Autism diagnostic clusters, symptom severity, comorbidities and medication use were not specifically addressed in this research. The restricted/repetitive diagnostic cluster has been associated with impaired executive function. A closer examination of this relationship including symptom severity with the EF measures reported in this thesis may contribute to intervention and remediation strategies. Co-morbid mental health conditions were not formally diagnosed in the ASD group and may have better informed these findings. It is of note however that empirical study 3 demonstrated that the influence of anxiety and depression on EF was not

Page | 159

dependent on diagnosis. Psychotropic medication use in children470 and adults471 with ASD is likely overprescribed possibly to address behavioural challenges. Future studies should consider whether medication use influences EF outcomes in Autism Spectrum Disorder.

Furthermore, intra-individual variability within the ASD spectrum was not explored in detail but could contribute to further insights on the reported heterogeneity for Autism Spectrum

Disorder.

7.4.3 Study methodology

This is a cross-sectional study and thus it cannot inform on developmental changes that may have led to the EF profile reported in the study sample of youth and adults. The neuropsychological test battery did not address all of the typically accepted EF domains; the study design, however, was guided in part by the findings of the meta-analysis that noted no differential contribution of distinct EF domains. The EP group did not complete the BRIEF self-report measure, thus findings on self-appraisals of EF cannot be extended to this group.

The study findings on subjective measures would be further supported, if informant

(carer/parent) measures were utilised. Finally, the findings on the relevance of EF in ASD would be strengthened with an examination of the relationship between EF and Theory of

Mind.

7.5 Future research directions

Within the current classification framework (i.e. DSM-5, ICD-10), the use of appropriate comparison groups is of paramount importance. Future studies should ensure that comparison groups include a neurotypical control group so that expected normative differences are experimentally controlled. Further, studies should aim to incorporate appropriate EF behavioural measures. Self and informant questionnaires are both particularly relevant and the utility of subjective measures in predicting disability in ASD has been demonstrated in this

Page | 160

thesis. Extension of EF research to ASD cohorts with intellectual disability is also important as it may contribute to targeted interventions and markers of disability. Research in ASD would particularly benefit from further exploration of the EF phenotype in a transdiagnostic framework. This approach may best account for the heterogeneity observed within the spectrum and guide intervention strategies. Intervention strategies need to be cognisant of the relevance of the EF phenotype in designing remediation and therapeutic programmes. Finally, use of advanced research platforms such as machine learning approaches would enhance understanding of the relevance of the EF phenotype, its diagnostic specificity and sensitivity, and its importance as a transdiagnostic marker in relation to other cognitive domains.

7.6 Final comment

Executive function is an important factor in the study of ASD and has utility in predicting life outcomes. The study of EF to date, however, has maintained a more narrow focus on neuropsychological assessments of EF with relatively recent interest in behavioural self- referential evaluations of EF. The role of affective states in EF in ASD has received very limited attention, yet the empirical findings reported in this thesis show that depression, anxiety and stress are significant factors in ASD. Perhaps the most significant contribution of the empirical findings reported here is to draw attention to the importance of a multifaceted evaluation of EF in ASD. The framework for the study of EF in ASD proposed in response to the present findings may guide intervention strategies as specific components of the framework may be targeted for intervention. Importantly, the transdiagnostic influence of some of the factors indicate that interventions may be applied across different disorders facilitating resource allocation and contributing to social inclusion.

Page | 161

References

1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders. 5th edn. American Psychiatric Association: Washington, DC, 2013.

2. Wagner RE, Zhang Y, Gray T, Abbacchi A, Cormier D, Todorov A et al. Autism-

related variation in reciprocal social behavior: a longitudinal study. Child

Development 2018; 0(0).

3. Papagiannopoulou EA, Chitty KM, Hermens DF, Hickie IB, Lagopoulos J. A

systematic review and meta-analysis of eye-tracking studies in children with autism

spectrum disorders. Social Neuroscience 2014; 9(6): 610-632.

4. Caruana N, Stieglitz Ham H, Brock J, Woolgar A, Kloth N, Palermo R et al. Joint

attention difficulties in autistic adults: An interactive eye-tracking study. Autism 2017;

22(4): 502-512.

5. Baron-Cohen S, Leslie AM, Frith U. Does the autistic child have a “theory of mind” ?

Cognition 1985; 21(1): 37-46.

6. South M, Ozonoff S, McMahon WM. Repetitive behavior profiles in Asperger

syndrome and high-functioning autism. Journal of Autism and Developmental

Disorders 2005; 35(2): 145-158.

7. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders

Page | 162

3rd edn. American Psychiatric Association: Washington, DC, 1980.

8. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders

1st edn. American Psychiatric Association: Washington, DC, 1952.

9. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders

2nd edn. American Psychiatric Association: WAshington, DC, 1968.

10. Bleuler E. Dementia Praecox or the group of schizophrenias. International

Universities: New York, 1950 [1911].

11. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders

3rd Revised edn. American Psychiatric Association: Washington, DC, 1987.

12. Volkmar FR, Cicchetti DV, Bregman J, Cohen DJ. Three diagnostic systems for

autism: DSM-III, DSM-III-R, and ICD-10. Journal of Autism and Developmental

Disorders 1992; 22(4): 483-492.

13. Volkmar FR, McPartland JC. From Kanner to DSM-5: Autism as an evolving

diagnostic concept. Annual Review of Clinical Psychology 2014; 10(1): 193-212.

Page | 163

14. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders

4th edn. American Psychiatric Association: Washington, DC, 1994.

15. Masi A, DeMayo MM, Glozier N, Guastella AJ. An overview of autism spectrum

disorder, heterogeneity and treatment options. Neuroscience Bulletin 2017; 33(2):

183-193.

16. Klopper F, Testa R, Pantelis C, Skafidas E. A cluster analysis exploration of autism

spectrum disorder subgroups in children without intellectual disability. Research in

Autism Spectrum Disorders 2017; 36: 66-78.

17. Happé F, Frith U. The weak coherence account: detail-focused cognitive style in

autism spectrum disorders. Journal of Autism and Developmental Disorders 2006;

36(1): 5-25.

18. Pennington BF, Ozonoff S. Executive Functions and Developmental

Psychopathology. Journal of Child Psychology and Psychiatry 1996; 37(1): 51-87.

19. Lever AG, Geurts HM. Age-related differences in cognition across the adult lifespan

in autism spectrum disorder. Autism Research 2016; 9(6): 666-676.

20. Olde Dubbelink LME, Geurts HM. Planning skills in autism spectrum disorder across

the lifespan: a meta-analysis and meta-regression. Journal of Autism and

Developmental Disorders 2017; 47(4): 1148-1165.

Page | 164

21. Lezak M, Howieson D, Bigler E, Tranel D. Neuropsychological Assessment. 5th

edition edn. Oxford University Press: New York, 2012.

22. Stuss DT. Functions of the frontal lobes: relation to executive functions. Journal of

the International Neuropsychological Society 2011; 17(5): 759-765.

23. Barkley R, A. Executive functions: What they are, why they work and why they

evolved. The Guildford Press: 72 Spring Street, New York, NY 10012, 2012.

24. Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD. The

unity and diversity of executive functions and their contributions to complex “frontal

lobe” tasks: a latent variable analysis. Cognitive Psychology 2000; 41(1): 49-100.

25. Diamond A. Executive functions. Annual Review of Psychology 2013; 64: 135-168.

26. Baggetta P, Alexander Patricia A. Conceptualization and operationalization of

executive function. Mind, Brain, and Education 2016; 10(1): 10-33.

27. Demetriou EA, Lampit A, Quintana DS, Naismith SL, Song YJC, Pye JE et al.

Autism spectrum disorders: a meta-analysis of executive function. Mol Psychiatry

2017.

Page | 165

28. Geurts HM, Sinzig J, Booth R, Happe F. Neuropsychological heterogeneity in

executive functioning in autism spectrum disorders. Int J Dev Disabil 2014; 60(3):

155-162.

29. Geurts HM, Corbett B, Solomon M. The paradox of cognitive flexibility in autism.

Trends in Cognitive Sciences 2009; 13(2): 74-82.

30. Volkmar F, Bregman J, Cohen D, Cicchetti D. DSM-III and DSM-III-R diagnoses of

autism. American Journal of Psychiatry 1988; 145(11): 1404-1408.

31. South M, Ozonoff S, McMahon WM. The relationship between executive

functioning, central coherence, and repetitive behaviors in the high-functioning

autism spectrum. Autism 2007; 11(5): 437-451.

32. Jones CRG, Simonoff E, Baird G, Pickles A, Marsden AJS, Tregay J et al. The

association between theory of mind, executive function, and the symptoms of autism

spectrum disorder. Autism Research 2018; 11(1): 95-109.

33. Kouklari E-C, Tsermentseli S, Auyeung B. Executive function predicts theory of

mind but not social verbal communication in school-aged children with autism

spectrum disorder. Research in Developmental Disabilities 2018; 76: 12-24.

34. Zimmerman D, Ownsworth T, O’Donovan A, Roberts J, Gullo MJ. Associations

between executive functions and mental health outcomes for adults with autism

spectrum disorder. Psychiatry Research 2017; 253(Supplement C): 360-363.

Page | 166

35. Demetriou EA, Song CY, Park SH, Pepper KL, Naismith SL, Hermens DF et al.

Autism, Early Psychosis, and Social Anxiety Disorder: a transdiagnostic examination

of executive function cognitive circuitry and contribution to disability. Translational

Psychiatry 2018; 8(1): 200.

36. Pugliese CE, Anthony LG, Strang JF, Dudley K, Wallace GL, Naiman DQ et al.

Longitudinal examination of adaptive behavior in autism spectrum disorders:

influence of executive function. Journal of Autism and Developmental Disorders

2016; 46(2): 467-477.

37. Bishop-Fitzpatrick L, Hong J, Smith LE, Makuch RA, Greenberg JS, Mailick MR.

Characterizing objective quality of life and normative outcomes in adults with autism

spectrum disorder: an exploratory latent class analysis. Journal of Autism and

Developmental Disorders 2016; 46(8): 2707-2719.

38. Asonitou K, Koutsouki D, Kourtessis T, Charitou S. Motor and cognitive

performance differences between children with and without developmental

coordination disorder (DCD). Research in Developmental Disabilities 2012; 33(4):

996-1005.

39. Asonitou K, Koutsouki D. Cognitive process-based subtypes of developmental

coordination disorder (DCD). Human Movement Science 2016; 47: 121-134.

Page | 167

40. Lai CLE, Lau Z, Lui SSY, Lok E, Tam V, Chan Q et al. Meta-analysis of

neuropsychological measures of executive functioning in children and adolescents

with high-functioning autism spectrum disorder. Autism Research 2017; 10(5): 911-

939.

41. Wallace GL, Kenworthy L, Pugliese CE, Popal HS, White EI, Brodsky E et al. Real-

world executive functions in adults with autism spectrum disorder: profiles of

impairment and associations with adaptive functioning and co-morbid anxiety and

depression. Journal of Autism and Developmental Disorders 2016; 46(3): 1071-1083.

42. Gordon-Lipkin E, Marvin AR, Law JK, Lipkin PH. Anxiety and mood disorder in

children with autism spectrum disorder and ADHD. Pediatrics 2018.

43. Banich M, T. . Executive function: the search for an integrated account. Current

Directions in Psychological Science 2009; 18(2): 89-94.

44. Pribram MI, Snyder CRR. Attention and cognitive control. In: Solso R (ed).

Information processing and cognition: The Loyola symposium. Lawrence Erlbaum:

Hillsdale, NJ, 1975, pp 55-85.

45. Harlow J. Passage of an iron bar through the head. Boston Medical and Surgical

Journal 1848; 39: 389-393.

46. Stuss DT, Benson DF. The frontal lobes. Raven: New York, 1986.

Page | 168

47. Szczepanski Sara M, Knight Robert T. Insights into human behavior from lesions to

the prefrontal cortex. Neuron 2014; 83(5): 1002-1018.

48. Alvarez JA, Emory E. Executive function and the frontal lobes: A Meta-analytic

review. Neuropsychology Review 2006; 16(1): 17-42.

49. Otero TM, Barker LA. The frontal lobes and executive functioning. In: Goldstein S,

Naglieri JA (eds). Handbook of Executive Functioning. Springer New York: New

York, NY, 2014, pp 29-44.

50. Lezak MD, Howieson DB, Bigler ED, Tranel D. Neuropsychological Assessment,

2012.

51. Barkley RA. The assessment of executive functioning using the Barkley Deficits in

Executive Functioning scales. In: Goldstein S, Naglieri JA (eds). Handbook of

Executive Functioning. Springer New York: New York, NY, 2014, pp 245-263.

52. Miyake A, Friedman NP. The nature and organization of individual differences in

executive functions: four general conclusions. Current directions in psychological

science 2012; 21(1): 8-14.

53. Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD. The

unity and diversity of executive functions and their contributions to complex "frontal

lobe" tasks: a latent variable analysis. Cognitive Psychology 2000; 41(1): 49-100.

Page | 169

54. Hunter S, Sparrow E. Models of executive functioning. In: Hunter S, Sparrow E (eds).

Executive Function and Dysfunction. Cambridge University Press: Cambridge, 2012.

55. Baddeley A. Exploring the central executive. The Quarterly Journal of Experimental

Psychology Section A 1996; 49(1): 5-28.

56. Posner MI. Orienting of attention. Quarterly Journal of Experimental Psychology

1980; 32(1): 3-25.

57. Shallice T. Fractionation of the supervisory system. In: Stuss D, Knight R (eds).

Principles of frontal lobe function. Oxford University Press: New York, 2002, pp

261-277.

58. Baddeley A. Working memory. Clarendon Press: Oxford, 1986.

59. Baddeley A. Working memory: Theories, models, and controversies. Annual Review

of Psychology 2012; 63(1): 1-29.

60. Petersen SE, Posner MI. The attention system of the human brain: 20 years after.

Annual Review of Neuroscience 2012; 35(1): 73-89.

61. Hughes C, Russell J, Robbins TW. Evidence for executive dysfunction in autism.

Neuropsychologia 1994; 32(4): 477-492.

62. Luria AR. Higher cortical functions in man. Basic Books: Oxford England, 1966.

Page | 170

63. Luria AR. Frontal lobe syndromes. In: Vinken P, Bruyn G (eds). Handbook of

Clinical Neurology, vol. 2. North Holland: Amsterdam, 1969, pp 725-757.

64. Luria AR. The Working Brain. Penguin Books: Middlesex, England, 1973.

65. Delis DC, Kaplan E, Kramer JH. D-KEFS Executive function system: Examiners

Manual. Psychological Corporation: San Antonio, TX, 2001.

66. Friedman NP, Miyake A. Unity and diversity of executive functions: Individual

differences as a window on cognitive structure. Cortex 2017; 86: 186-204.

67. Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A. The unity and

diversity of executive functions and their contributions to complex “frontal lobe”

tasks: a latent variable analysis. Cognitive Psychology 2000; 41: 49-100.

68. Friedman NP, Miyake A, Young SE, DeFries JC, Corley RP, Hewitt JK. Individual

differences in executive functions are almost entirely genetic in origin. Journal of

Experimental Psychology: General 2008; 137(2): 201-225.

69. Tick B, Bolton P, Happé F, Rutter M, Rijsdijk F. Heritability of autism spectrum

disorders: a meta‐analysis of twin studies. Journal of Child Psychology and

Psychiatry 2016; 57(5): 585-595.

Page | 171

70. Geurts HM, Bergh SFWM, Ruzzano L. Prepotent Response Inhibition and

Interference Control in Autism Spectrum Disorders: Two Meta‐Analyses. Autism

Research 2014; 7(4): 407-420.

71. Leung RC, Zakzanis KK. Brief report: cognitive flexibility in autism spectrum

disorders: a quantitative review. Journal of Autism and Developmental Disorders

2014; 44(10): 2628-2645.

72. Demetriou EA, Lampit A, Quintana DS, Naismith SL, Song YJC, Pye JE et al.

Autism spectrum disorders: a meta-analysis of executive function. Molecular

Psychiatry 2018; 23(5): 1198-1204.

73. Snyder HR, Miyake A, Hankin BL. Advancing understanding of executive function

impairments and psychopathology: bridging the gap between clinical and cognitive

approaches. Frontiers in Psychology 2015; 6: 328.

74. Henri-Bhargava A, Stuss DT, Freedman M. Clinical assessment of prefrontal lobe

functions. CONTINUUM: Lifelong Learning in Neurology 2018; 24(3): 704-726.

75. Heaton RK, Chelune GJ, Talley JL, Kay GG, Curtis G. Wisconsin Card Sorting Test

(WCST) manual: revised and expanded. Psychological Assessment Resources Inc.:

Odessa, 1993.

76. Gioia G, Isquith PK, Guy SC, Kenworthy L. Behavior rating inventory of executive

function. Child Neuropsychol 2000; 6(3): 235-238.

Page | 172

77. DeVon HA, Block ME, Moyle-Wright P, Ernst DM, Hayden SJ, Lazzara DJ et al. A

psychometric toolbox for testing validity and reliability. Journal of Nursing

Scholarship 2007; 39(2): 155-164.

78. Keifer E, Tranel D. A neuropsychological investigation of the Delis-Kaplan Executive

Function System. Journal of Clinical and Experimental Neuropsychology 2013;

35(10): 1048-1059.

79. Goldstein S, Naglieri JA, Princiotta D, Otero TM. Introduction: A History of

Executive Functioning as a Theoretical and Clinical Construct. In: Goldstein S,

Naglieri JA (eds). Handbook of Executive Functioning. Springer New York: New

York, NY, 2014, pp 3-12.

80. Denckla MB. A theory and model of executive function: A neuropsychological

perspective. Attention, memory, and executive function. Paul H Brookes Publishing:

Baltimore, MD, US, 1996, pp 263-278.

81. Chan RCK, Shum D, Toulopoulou T, Chen EYH. Assessment of executive functions:

Review of instruments and identification of critical issues. Archives of Clinical

Neuropsychology 2008; 23(2): 201-216.

82. Buchsbaum B, R., Greer S, Chang WL, Berman Karen F. Meta‐analysis of

neuroimaging studies of the Wisconsin Card‐Sorting task and component processes.

Human Brain Mapping 2005; 25(1): 35-45.

Page | 173

83. Larson MJ, South M, Clayson PE, Clawson A. Cognitive control and conflict

adaptation in youth with high-functioning autism. Journal of child psychology and

psychiatry, and allied disciplines 2012; 53(4): 440-448.

84. Silk TJ, Rinehart N, Bradshaw JL, Tonge B, Egan G, O'Boyle M et al. Visuospatial

processing and the function of prefrontal-parietal networks in autism spectrum

disorders: a functional MRI study. The American Journal of Psychiatry 2006; 163(8):

1440-1443.

85. Blakemore S-J, Choudhury S. Development of the adolescent brain: implications for

executive function and social cognition. Journal of Child Psychology and Psychiatry

2006; 47(3-4): 296-312.

86. Best JR, Miller PH. A Developmental Perspective on Executive Function. Child

Development 2010; 81(6): 1641-1660.

87. De Luca CR, Wood SJ, Anderson V, Buchanan J-A, Proffitt TM, Mahony K et al.

Normative data from the Cantab. I: development of executive function over the

lifespan. Journal of Clinical and Experimental Neuropsychology 2003; 25(2): 242-

254.

88. O'Hearn K, Asato M, Ordaz S, Luna B. Neurodevelopment and executive function in

autism. Development and Psychopathology 2008; 20(4): 1103-1132.

Page | 174

89. Sachse M, Schlitt S, Hainz D, Ciaramidaro A, Schirman S, Walter H et al. Executive

and visuo-motor function in adolescents and adults with autism spectrum disorder.

Journal of Autism and Developmental Disorders 2013; 43(5): 1222-1235.

90. Koshino H, Kana RK, Keller TA, Cherkassky VL, Minshew NJ, Just MA. fMRI

investigation of working memory for faces in autism: Visual coding and

underconnectivity with frontal areas. Cerebral Cortex 2008; 18(2): 289-300.

91. Ozonoff S, Cook I, Coon H, Dawson G, Joseph RM, Klin A et al. Performance on

Cambridge Neuropsychological Test Automated Battery subtests sensitive to frontal

lobe function in people with autistic disorder: Evidence from the Collaborative

Programs of Excellence in Autism Network. Journal of Autism and Developmental

Disorders 2004; 34(2): 139-150.

92. Ozonoff S. Reliability and validity of the Wisconsin Card Sorting Test in studies of

autism. Neuropsychology 1995; 9(4): 491-500.

93. Homack S, Lee D, Riccio CA. Test Review: Delis-Kaplan Executive Function

System. Journal of Clinical and Experimental Neuropsychology 2005; 27(5): 599-

609.

94. Schmuckler MA. What Is Ecological Validity? A Dimensional Analysis. Infancy

2001; 2(4): 419-436.

Page | 175

95. Burgess PW, Alderman N, Forbes C, Costello A, M-A.Coates L, Dawson DR et al.

The case for the development and use of “ecologically valid” measures of executive

function in experimental and clinical neuropsychology. Journal of the International

Neuropsychological Society 2006; 12(2): 194-209.

96. Wallisch A, Little LM, Dean E, Dunn W. Executive function measures for children: a

scoping review of ecological validity. OTJR: Occupation, Participation and Health

2017; 38(1): 6-14.

97. Shallice T, Burgess Paul W. Deficits in strategy application following frontal lobe

damage in man. Brain 1991; (114): 727-741.

98. Parsons TD, Carlew AR. Bimodal virtual reality Stroop for assessing distractor

inhibition in autism spectrum disorders. Journal of Autism and Developmental

Disorders 2016; 46(4): 1255-1267.

99. Cox SM, Cox DJ, Kofler MJ, Moncrief MA, Johnson RJ, Lambert AE et al. Driving

simulator performance in novice drivers with autism spectrum disorder: the role of

executive functions and basic motor skills. Journal of Autism and Developmental

Disorders 2016; 46(4): 1379-1391.

100. Gomez P, Ratcliff R, Perea M. A model of the go/no-go task. Journal of experimental

psychology General 2007; 136(3): 389-413.

Page | 176

101. Eriksen CW. The flankers task and response competition: A useful tool for

investigating a variety of cognitive problems. Visual Cognition 1995; 2(2-3): 101-

118.

102. Kane MJ, Conway ARA, Miura TK, Colflesh GJH. Working memory, attention

control, and the n-back task: A question of construct validity. Journal of Experimental

Psychology: Learning, Memory, and Cognition 2007; 33(3): 615-622.

103. Roth RM, Isquith PK, Gioia G. BRIEF-A: Behavior Rating Inventory of Executive

Function-Adult Version. PAR: Lutz, FL 33549, 2005.

104. Kenworthy L, Yerys BE, Anthony LG, Wallace GL. Understanding executive control

in Autism Spectrum Disorders in the lab and in the real world. Neuropsychology

Review 2008; 18(4): 320-338.

105. Gioia GA, Isquith PK, Kenworthy L, Barton RM. Profiles of Everyday Executive

Function in Acquired and Developmental Disorders. Child Neuropsychology 2002;

8(2): 121-137.

106. van den Bergh SFWM, Scheeren AM, Begeer S, Koot HM, Geurts HM. Age Related

Differences of Executive Functioning Problems in Everyday Life of Children and

Adolescents in the Autism Spectrum. Journal of Autism and Developmental

Disorders 2014; 44(8): 1959-1971.

Page | 177

107. Lawson RA, Papadakis AA, Higginson CI, Barnett JE, Wills MC, Strang JF et al.

Everyday executive function impairments predict comorbid psychopathology in

autism spectrum and attention deficit hyperactivity disorders. Neuropsychology 2015;

29(3): 445-453.

108. Toplak ME, West RF, Stanovich KE. Practitioner Review: Do performance-based

measures and ratings of executive function assess the same construct? Journal of

Child Psychology and Psychiatry 2013; 54(2): 131-143.

109. Dryman MT, Gardner S, Weeks JW, Heimberg RG. Social anxiety disorder and

quality of life: How fears of negative and positive evaluation relate to specific

domains of life satisfaction. Journal of Anxiety Disorders 2016; 38: 1-8.

110. Hogeveen J, Krug MK, Elliott MV, Solomon M. Insula-Retrosplenial Cortex

Overconnectivity Increases Internalizing via Reduced Insight in Autism. Biological

Psychiatry 2018; 84(4): 287-294.

111. Hill EL. Evaluating the theory of executive dysfunction in autism. Developmental

Review 2004; 24(2): 189-233.

112. Hill EL. Executive dysfunction in autism. Trends in Cognitive Sciences 2004; 8(1):

26-32.

Page | 178

113. Ravizza SM, Carter CS. Shifting set about task switching: behavioral and neural

evidence for distinct forms of cognitive flexibility. Neuropsychologia 2008; 46(12):

2924-2935.

114. Zelazo PD. The Dimensional Change Card Sort (DCCS): a method of assessing

executive function in children. Nature Protocols 2006; 1: 297.

115. Yerys BE, Wolff BC, Moody E, Pennington BF, Hepburn SL. Brief report: Impaired

Flexible item Selection task (FIST) in school-age children with autism spectrum

disorders. Journal of Autism and Developmental Disorders 2012; 42(9): 2013-2020.

116. Landry O, Al-Taie S. A meta-analysis of the Wisconsin Card Sort Task in Autism.

Journal of Autism and Developmental Disorders 2016; 46(4): 1220-1235.

117. Taylor MJ, Donner EJ, Pang EW. fMRI and MEG in the study of typical and atypical

cognitive development. Neurophysiologie Clinique/Clinical Neurophysiology 2012;

42(1): 19-25.

118. Arbuthnott K, Frank J. Trail Making Test, Part B as a measure of executive control:

Validation using a set-switching paradigm. Journal of Clinical and Experimental

Neuropsychology 2000; 22(4): 518-528.

119. Reitan RM, Wolfson D. Trail Making Test: Manual for Administration and Scoring.

Neuropsychological Press: Tucson, AZ, 1985.

Page | 179

120. Chmielewski WX, Beste C. Action control processes in autism spectrum disorder –

Insights from a neurobiological and neuroanatomical perspective. Progress in

Neurobiology 2015; 124: 49-83.

121. Simmonds DJ, Pekar JJ, Mostofsky SH. Meta-analysis of Go/No-go tasks

demonstrating that fMRI activation associated with response inhibition is task-

dependent. Neuropsychologia 2008; 46(1): 224-232.

122. Bishop DV, Norbury CF. Executive functions in children with communication

impairments, in relation to autistic symptomatology I: Generativity. Autism 2005;

9(1): 7-27.

123. Aita S, Boettcher A, Slagel B, Holcombe J, Espenan M, King M et al. The relation

between verbal fluency and executive functioning: An exploratory factor analysis

(EFA) approach. Archives of Clinical Neuropsychology 2016; 31(6): 656-656.

124. Costafreda Sergi G, Fu Cynthia HY, Lee L, Everitt B, Brammer Michael J, David

Anthony S. A systematic review and quantitative appraisal of fMRI studies of verbal

fluency: Role of the left inferior frontal gyrus. Human Brain Mapping 2006; 27(10):

799-810.

125. Ruff RM, Light RH, Parker SB, Levin HS. Benton controlled oral word association

test: Reliability and updated norms. Archives of Clinical Neuropsychology 1996;

11(4): 329-338.

Page | 180

126. Whiteside DM, Kealey T, Semla M, Luu H, Rice L, Basso MR et al. Verbal fluency:

language or executive function measure? Applied Neuropsychology: Adult 2016;

23(1): 29-34.

127. Piatt AL, Fields JA, Paolo AM, Tröster AI. Action (verb naming) fluency as an

executive function measure: convergent and divergent evidence of validity.

Neuropsychologia 1999; 37(13): 1499-1503.

128. Unterrainer JM, Rahm B, Kaller CP, Leonhart R, Quiske K, Hoppe-Seyler K et al.

Planning abilities and the Tower of London: is this task measuring a discrete

cognitive function? J Clin Exp Neuropsychol 2004; 26(6): 846-856.

129. Welsh MC. Rule-guided behavior and self-monitoring on the Tower of Hanoi disk-

transfer task. Cognitive Development 1991; 6(1): 59-76.

130. Cazalis F, Valabrègue R, Pélégrini‐Issac M, Asloun S, Robbins TW, Granon S.

Individual differences in prefrontal cortical activation on the Tower of London

planning task: implication for effortful processing. European Journal of Neuroscience

2003; 17(10): 2219-2225.

131. Nitschke K, Köstering L, Finkel L, Weiller C, Kaller Christoph P. A Meta‐analysis

on the neural basis of planning: Activation likelihood estimation of functional brain

imaging results in the Tower of London task. Human Brain Mapping 2016; 38(1):

396-413.

Page | 181

132. Stroop JR. Studies of interference in serial verbal reactions. Journal of Experimental

Psychology 1935; 18(6): 643-662.

133. Burgess PW, Shallice T. Response suppression, initiation and strategy use following

frontal lobe lesions. Neuropsychologia 1996; 34(4): 263-272.

134. Erika-Florence M, Leech R, Hampshire A. A functional network perspective on

response inhibition and attentional control. Nature Communications 2014; 5: 4073.

135. Wechsler D. Wechsler Memory Scale -Third Edition. The Psychological Corporation:

San Antonio, Texas, 1997.

136. Kanner L. Autistic disturbances of affective contact. Nervous Child 1943; 2: 217-250.

137. Boucher J. Alternation and sequencing behaviour, and response to novelty in autistic

children. Journal of Child Psychology and Psychiatry 1977; 18(1): 67-72.

138. Hoffmann WL, Prior MR. Neuropsychological dimensions of autism in children: A

test of the hemispheric dysfunction hypothesis. Journal of Clinical Neuropsychology

1982; 4(1): 27-41.

139. Prior M, Hoffmann W. Brief report: Neuropsychological testing of autistic children

through an exploration with frontal lobe tests. Journal of Autism and Developmental

Disorders 1990; 20(4): 581-590.

Page | 182

140. Rumsey JM. Conceptual problem -solving in highly verbal non-retarded autistic men.

Journal of Autism and Developmental Disorders 1985; 15: 23-26.

141. Rumsey JM, Hamburger SD. Neuropsychological findings in high-functioning autistic

men with infantile autism, residual state. Journal of Clinical and Experimental

Neuropsychology 1988; 10: 201-221.

142. Rumsey JM, Hamburger SD. Neuropsychological divergence of high-level autism and

severe dyslexia. Journal of Autism & Developmental Disorders 1990; 20(2): 155-168.

143. Ozonoff S, Pennington BF, Rogers SJ. Executivefunction deficits in high‐

functioning autistic individuals: relationship to theory of mind. Journal of Child

Psychology and Psychiatry 1991; 32(7): 1081-1105.

144. Ozonoff S, McEvoy RE. A longitudinal study of executive function and theory of

mind development in autism. Development and Psychopathology 1994; 6: 415-431.

145. Ozonoff S, Strayer D. Inhibitory function in nonretarded children with autism.

Journal of Autism and Developmental Disorders 1997; 27(1): 59-77.

146. Boucher J. Word fluency in high-functioning autistic children. Journal of Autism and

Developmental Disorders 1988; 18(4): 637-645.

147. Lam YG. Re-examining the cognitive phenotype in autism: A study with young

Chinese children. Research in Developmental Disabilities 2013; 34: 4591-4598.

Page | 183

148. Pascualvaca DM, Fantie BD, Papageorgiou M, Mirsky AF. Attentional capacities in

children with autism: Is there a general deficit in shifting focus? Journal of Autism

and Developmental Disorders 1998; 28(6): 467-478.

149. Williams D, Jarrold C. Assessing planning and set-shifting abilities in autism: Are

experimenter-administered and computerised versions of tasks equivalent? Autism

Research 2013; 6(6): 461-467.

150. Williams D, Boucher J, Lind S, Jarrold C. Time-based and event-based prospective

memory in autism spectrum disorder: The roles of executive function and theory of

mind, and time-estimation. Journal of Autism and Developmental Disorders 2013;

43(7): 1555-1567.

151. Yerys BE, Wallace GL, Sokoloff JL, Shook DA, James JD, Kenworthy L. Attention

deficit/hyperactivity disorder symptoms moderate cognition and behavior in children

with autism spectrum disorders. Autism Research 2009; 2(6): 322-333.

152. Van Eylen L, Boets B, Steyaert J, Evers K, Wagemans J, Noens I. Cognitive

flexibility in autism spectrum disorder: Explaining the inconsistencies? Research in

Autism Spectrum Disorders 2011; 5(4): 1390-1401.

153. Durrleman S, Franck J. Exploring links between language and cognition in autism

spectrum disorders: Complement senctences, false belief, and executive functioning.

Journal of Communication Disorders 2015; 54: 15-31.

Page | 184

154. Nyden A, Gillberg C, Hjelmquist E, Heiman M. Executive function/attention deficits

in boys with Asperger syndrome, attention disorder and reading/writing disorder.

Autism 1999; 3(3): 213-228.

155. Schneider SG, Asarnow RF. A comparison of cognitive/neuropsychological

impairments of nonretarded autistic and schizophrenic children. Journal of Abnormal

Child Psychology 1987; 15(1): 29-46.

156. Yerys BE, Wallace GL, Harrison B, Celano MJ, Giedd JN, Kenworthy LE. Set-

shifting in children with autism spectrum disorders: Reversal shifting deficits on the

intradimensional/extradimensional Shift Test correlate with repetitive behaviors.

Autism 2009; 13(5): 523-538.

157. Goldberg MC, Mostofsky SH, Cutting LE, Mahone EM, Astor BC, Denckla MB et al.

Subtle executive impairment in children with autism and children with ADHD.

Journal of Autism and Developmental Disorders 2005; 35(3): 279-293.

158. Brunsdon VE, Colvert E, Ames C, Garnett T, Gillan N, Hallett V et al. Exploring the

cognitive features in children with autism spectrum disorder, their co-twins, and

typically developing children within a population-based sample. Journal of Child

Psychology and Psychiatry 2015; 56(8): 893-902.

159. Floris DL, Chura LR, Holt RJ, Suckling J, Bullmore ET, Baron Cohen S et al.

Psychological correlates of handedness and corpus callosum asymmetry in autism: the

Page | 185

left hemisphre dysfunction theory revisited. Journal of Autism & Developmental

Disorders 2013; 43(8): 1758-1772.

160. Sawa T, Kodaira M, Oiji A, Sasayama D, Iwadare Y, Ushijima H et al. Dysfunction

of orbitofrontal and dorsolateral prefrontal cortices in children and adolescents with

high-functioning pervasive developmental disorders. Annals of General Psychiatry

2013; 12(1).

161. Bölte S, Duketis E, Poustka F, Holtmann M. Sex differences in cognitive domains and

their clinical correlates in higher-functioning autism spectrum disorders. Autism 2011;

15(4): 497-511.

162. Happe F, Booth R, Charlton R, Hughes C. Executive function deficits in autism

spectrum disorders and attention-deficit/hyperactivity disorder: Examining profiles

across domains and ages. Brain and Cognition 2006; 61(1): 25-39.

163. Rumsey JM. Conceptual problem-solving in highly verbal, nonretarded autistic men.

Journal of Autism and Developmental Disorders 1985; 15(1): 23-36.

164. Rumsey JM, Hamburger SD. Neuropsychological divergence of high autism and

severe dyslexia. Journal of Autism and Developmental Disorders 1990; 20: 155-168.

165. Dichter GS, Belger A. Social stimuli interfere with cognitive control in autism.

NeuroImage 2007; 35(3): 1219-1230.

Page | 186

166. Maes JHR, Eling PATM, Wezenberg E, Vissers CTWM, Kan CC. Attentional set

shifting in autism spectrum disorder: Differentiating between the role of

perseveration, learned irrelevance, and novelty processing. Journal of Clinical and

Experimental Neuropsychology 2011; 33(2): 210-217.

167. Broadbent J, Stokes MA. Removal of negative feedback enhances WCST

performance for individuals with ASD. Research in Autism Spectrum Disorders 2013;

7(6): 785-792.

168. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders

4th Text revision edn. American Psychiatric Association: Washington, DC, 2000.

169. Ozonoff S, South M, Miller JN. DSM-IV-Defined Asperger Syndrome: Cognitive,

Behavioral and Early History Differentiation from High-Functioning Autism. Autism

2000; 4(1): 29-46.

170. Christ SE, Holt DD, White DA, Green L. Inhibitory control in children with autism

spectrum disorder. Journal of Autism and Developmental Disorders 2007; 37(6):

1155-1165.

171. Adams NC, Jarrold C. Inhibition in autism: Children with autism have difficulty

inhibiting irrelevant distractors but not prepotent responses. Journal of Autism and

Developmental Disorders 2012; 42(6): 1052-1063.

Page | 187

172. Christ SE, Kester LE, Bodner KE, Miles JH. Evidence for selective inhibitory

impairment in individuals with autism spectrum disorder. Neuropsychology 2011;

25(6): 690-701.

173. Johnson KA, Robertson IH, Kelly SP, Silk TJ, Barry E, Daibhis A et al. Dissociation

in performance of children with ADHD and high-functioning autism on a task of

sustained attention. Neuropsychologia 2007; 45(10): 2234-2245.

174. Langen M, Leemans A, Johnston P, Ecker C, Daly E, Murphy CM et al. Fronto-

striatal circuitry and inhibitory control in autism: Findings from diffusion tensor

imaging tractography. Cortex: A Journal Devoted to the Study of the Nervous System

and Behavior 2012; 48(2): 183-193.

175. Johnston K, Madden AK, Bramham J, Russell AJ. Response inhibition in adults with

autism spectrum disorder compared to attention deficit/hyperactivity disorder. Journal

of Autism and Developmental Disorders 2011; 41(7): 903-912.

176. Chan RC, Hu Z-y, Cui J-f, Wang Y, McAlonan GM. Social attribution in children

with high functioning autism and Asperger syndrome: An exploratory study in the

Chinese setting. Research in Autism Spectrum Disorders 2011; 5(4): 1538-1548.

177. Ames CS, Jarrold C. The problem with using eye-gaze to infer desire: a deficit of cue

inference in children with Autism Spectrum Disorder? Journal of Autism and

Developmental Disorders 2007; 37: 1761-1775.

Page | 188

178. Ambrosino S, Bos D, van Raalten T, Kobussen N, van Belle J, Oranje B et al.

Functional connectivity during cognitive control in children with autism spectrum

disorder: An independent component analysis. Journal of Neural Transmission 2014;

121(9): 1145-1155.

179. Brandimonte MA, Filippello P, Coluccia E, Altgassen M, Kliegel M. To do or not to

do? Prospective memory versus response inhibition in autism spectrum disorder and

attention-deficit/hyperactivity disorder. Memory 2011; 19(1): 56-66.

180. Goldberg MC, Spinelli S, Joel S, Pekar JJ, Denckla MB, Mostofsky SH. Children

with high functioning autism show increased prefrontal and temporal cortex activity

during error monitoring. Developmental Cognitive Neuroscience 2011; 1(1): 47-56.

181. Kuijper SJM, Hartman CA, Hendriks P. Who is he? Children with ASD and ADHD

take the listener into account in their production of ambiguous pronouns. PLoS ONE

2015; 10(7).

182. Morsanyi K, Holyoak KJ. Analogical reasoning ability in autistic and typically

developing children. Developmental Science 2010; 13(4): 578-587.

183. Russell J, Jarrold C, Hood B. Two intact executive capacities in children with autism:

Implications for the core executive dysfunctions in the disorder. Journal of Autism

and Developmental Disorders 1999; 29(2): 103-112.

Page | 189

184. Vara AS, Pang EW, Doyle-Thomas KAR, Vidal J, Taylor MJ, Anagnostou E. Is

inhibitory control a 'no-go' in adolescents with autism spectrum disorder? Molecular

Autism 2014; 5(1).

185. Altgassen M, Koch A. Impact of inhibitory load on remembering delayed intentions

in autism. International Journal of Developmental Disabilities 2014; 60(3): 198-204.

186. Shafritz KM, Bregman JD, Ikuta T, Szeszko PR. Neural systems mediating decision-

making and response inhibition for social and nonsocial stimuli in autism. Progress in

Neuro-Psychopharmacology and Biological Psychiatry 2015; 60: 112-120.

187. Williams DL, Goldstein G, Minshew NJ. The profile of memory function in children

with autism. Neuropsychology 2006; 20(1): 21-29.

188. Williams DL, Goldstein G, Carpenter PA, Minshew NJ. Verbal and spatial working

memory in autism. Journal of Autism and Developmental Disorders 2005; 35(6): 747.

189. Englund JA, Decker SL, Allen RA, Roberts AM. Common cognitive deficits in

children with attention-deficit/hyperactivity disorder and autism: Working memory

and visual-motor integration. Journal of Psychoeducational Assessment 2014; 32(2):

95-106.

190. Andersen PN, Skogli EW, Hovik KT, Geurts H, Egeland J, Øie M. Working memory

arrest in children with high-functioning autism compared to children with attention-

Page | 190

deficit/hyperactivity disorder: Results from a 2-year longitudinal study. Autism 2015;

19(4): 443-450.

191. Russell J, Jarrold C, Henry L. Working memory in children with autism and with

moderate learning difficulties. Journal of Child Psychology and Psychiatry and Allied

Disciplines 1996; 37(6): 673-686.

192. Vogan VM, Morgan BR, Lee W, Powell TL, Smith ML, Taylor MJ. The neural

correlates of visuo-spatial working memory in children with autism spectrum

disorder: Effects of cognitive load. Journal of Neurodevelopmental Disorders 2014;

6(19).

193. Steele SD, Minshew NJ, Luna B, Sweeney JA. Spatial working memory deficits in

autism. Journal of Autism and Developmental Disorders 2007; 37(4): 605-612.

194. Bucaille A, Grandgeorge M, Degrez C, Mallégol C, Cam P, Botbol M et al. Cognitive

profile in adults with Asperger syndrome using WAIS-IV: Comparison to typical

adults. Research in Autism Spectrum Disorders 2016; 21: 1-9.

195. Crane L, Goddard L, Pring L. Autobiographical memory in adults with autism

spectrum disorder: the role of deperssed mood, rumination, working memory and

theory of mind. Autism 2013; 17(2): 205-219.

Page | 191

196. Cui J, Gao D, Chen Y, Zou X, Wang Y. Working memory in early-school-age

children with asperger's syndrome. Journal of Autism and Developmental Disorders

2010; 40(8): 958-967.

197. Matsuura N, Ishitobi M, Arai S, Kawamura K, Asano M, Inohara K et al.

Distinguishing between autism spectrum disorder and attention deficit hyperactivity

disorder by using behavioral checklists, cognitive assessments, and

neuropsychological test battery. Asian Journal of Psychiatry 2014; 12(1): 50-57.

198. Geurts HM, De Wit S. Goal-directed action control in children with autism spectrum

disorders. Autism 2014; 18(4): 409-418.

199. Urbain C, Vogan VM, Ye AX, Pang EW, Doesburg SM, Taylor MJ.

Desynchronization of fronto-temporal networks during working memory processing

in autism. Human Brain Mapping 2016; 37(1): 153-164.

200. Williams DL, Goldstein G, Minshew NJ. Impaired memory for faces and social

scenes in autism: Clinical implications of memory dysfunction. Archives of Clinical

Neuropsychology 2005; 20(1): 1-15.

201. Kercood S, Grskovic JA, Banda D, Begeske J. Working memory and autism: A

review of literature. Res Autism Spectr Disord 2014; 8(10): 1316-1332.

Page | 192

202. Wang Y, Zhang Y-b, Liu L-l, Cui J-f, Wang J, Shum DHK et al. A meta-analysis of

working memory impairments in autism spectrum disorders. Neuropsychology Review

2017; 27(1): 46-61.

203. Kleinhans NM, Muller RA, Cohen DN, Courchesne E. Atypical functional

lateralization of language in autism spectrum disorders. Brain Research 2008; 1221:

115-125.

204. Miller HL, Ragozzino ME, Cook EH, Sweeney JA, Mosconi MW. Cognitive set

shifting deficits and their relationship to repetitive behaviors in autism spectrum

disorder. Journal of Autism and Developmental Disorders 2015; 45(3): 805-815.

205. Griebling J, Minshew NJ, Bodner K, Libove R, Bansal R, Konasale P et al.

Dorsolateral prefrontal cortex magnetic resonance imaging measurements and

cognitive performance in autism. J Child Neurol 2010; 25(7): 856-863.

206. Chiang H-M, Tsai LY, Cheung YK, Brown A, Li H. A meta-analysis of differences in

IQ profiles between individuals with Asperger’s disorder and high-functioning

autism. Journal of Autism and Developmental Disorders 2014; 44(7): 1577-1596.

207. Lemon JM, Gargaro B, Enticott PG, Rinehart NJ. Brief report: Executive functioning

in autism spectrum disorders: A gender comparison of response inhibition. Journal of

Autism and Developmental Disorders 2011; 41(3): 352-356.

Page | 193

208. Eysenck MW, Derakshan N. New perspectives in attentional control theory.

Personality and Individual Differences 2011; 50(7): 955-960.

209. Shields GS, Sazma MA, Yonelinas AP. The effects of acute stress on core executive

functions: A meta-analysis and comparison with cortisol. Neuroscience &

Biobehavioral Reviews 2016; 68: 651-668.

210. Liss M, Fein D, Allen D, Dunn M, Feinstein C, Morris R et al. Executive functioning

in high-functioning children with autism. The Journal of Child Psychology and

Psychiatry and Allied Disciplines 2001; 42(2): 261-270.

211. Nakahachi T, Iwase M, Takahashi H, Honaga E, Sekiyama R, Ukai S et al.

Discrepancy of performance among working memory-related tasks in autism

spectrum disorders was caused by task characteristics, apart from working memory,

which could interfere with task execution. Psychiatry and Clinical Neurosciences

2006; 60(3): 312-318.

212. Verte S, Geurts HM, Roeyers H, Oosterlaan J, Sergeant JA. Executive functioning in

children with an autism spectrum disorder: can we differentiate within the spectrum?

Journal of Autism and Developmental Disorders 2006; 36(3): 351-372.

213. Mackinlay R, Charman T, Karmiloff-Smith A. High functioning children with autism

spectrum disorder: A novel test of multitasking. Brain and Cognition 2006; 61(1): 14-

24.

Page | 194

214. Montgomery JM, Stoesz BM, McCrimmon AW. Emotional intelligence, theory of

mind, and executive functions as predictors of social outcomes in young adults with

Asperger syndrome. Focus on Autism and Other Developmental Disabilities 2013;

28(1): 4-13.

215. Nyden A, Hagberg B, Gousse V, Rastam M. A cognitive endophenotype of autism in

families with multiple incidence. Research in Autism Spectrum Disorders 2011; 5(1):

191-200.

216. Szatmari P, Tuff L, J. Finlayson MA, Bartolucci G. Asperger's Syndrome and Autism:

Neurocognitive Aspects. Journal of the American Academy of Child & Adolescent

Psychiatry 1990; 29(1): 130-136.

217. Bennetto L, Pennington BF, Rogers SJ. Intact and Impaired Memory Functions in

Autism. Child Development 1996; 67(4): 1816-1835.

218. Joseph RM, Keehn B, Connolly C, Wolfe JM, Horowitz TS. Why is visual search

superior in autism spectrum disorder? Developmental Science 2009; 12(6): 1083-

1096.

219. Loomes R, Hull L, Mandy WPL. What Is the male-to-female ratio in autism spectrum

disorder? a systematic review and meta-analysis. Journal of the American Academy of

Child & Adolescent Psychiatry 2017; 56(6): 466-474.

Page | 195

220. Skuse DH. Imprinting, the X-Chromosome, and the Male Brain: Explaining sex

differences in the liability to autism. Pediatric Research 2000; 47: 9.

221. Baron-Cohen S, Knickmeyer RC, Belmonte MK. Sex dfferences in the brain:

Implications for explaining autism. Science 2005; 310(5749): 819-823.

222. Ferri SL, Abel T, Brodkin ES. Sex differences in autism spectrum disorder: a review.

Current Psychiatry Reports 2018; 20(2): 9.

223. Frazier TW, Georgiades S, Bishop SL, Hardan AY. Behavioral and cognitive

characteristics of females and males With Autism in the Simons Simplex Collection.

Journal of the American Academy of Child & Adolescent Psychiatry 2014; 53(3):

329-340.e323.

224. Lai M-C, Lombardo MV, Ruigrok ANV, Chakrabarti B, Wheelwright SJ, Auyeung B

et al. Cognition in males and females with Autism: similarities and differences. PLOS

ONE 2012; 7(10): e47198.

225. Memari AH, Ziaee V, Shayestehfar M, Ghanouni P, Mansournia MA, Moshayedi P.

Cognitive flexibility impairments in children with autism spectrum disorders: Links to

age, gender and child outcomes. Research in Developmental Disabilities 2013;

34(10): 3218-3225.

226. Lehnhardt F-G, Falter CM, Gawronski A, Pfeiffer K, Tepest R, Franklin J et al. Sex-

related cognitive profile in autism spectrum disorders diagnosed late in life:

Page | 196

implications for the female autistic phenotype. Journal of Autism and Developmental

Disorders 2016; 46(1): 139-154.

227. Lai M-C, Lombardo MV, Ruigrok AN, Chakrabarti B, Auyeung B, Szatmari P et al.

Quantifying and exploring camouflaging in men and women with autism. Autism

2017; 21(6): 690-702.

228. Kiep M, Spek AA. Executive functioning in men and women with an autism spectrum

disorder. Autism Research 2017; 10(5): 940-948.

229. Liston C, McEwen BS, Casey BJ. Psychosocial stress reversibly disrupts prefrontal

processing and attentional control. Proceedings of the National Academy of Sciences

of the United States of America 2009; 106(3): 912-917.

230. Pacheco-Unguetti AP, Acosta A, Callejas A, Lupiáñez J. Attention and anxiety:

different attentional functioning under state and trait anxiety. Psychological Science

2010; 21(2): 298-304.

231. Visu-Petra L, Miclea M, Visu-Petra G. Individual differences in anxiety and executive

functioning: A multidimensional view. International Journal of Psychology 2013;

48(4): 649-659.

232. Eysenck M, Payne S, Derakshan N. Trait anxiety, visuospatial processing, and

working memory. Cognition and Emotion 2005; 19(8): 1214-1228.

Page | 197

233. Ursache A, Raver CC. Trait and state anxiety: relations to executive functioning in an

at-risk sample. Cognition & Emotion 2014; 28(5): 845-855.

234. Salthouse TA. How general are the effects of trait anxiety and depressive symptoms

on cognitive functioning? Emotion 2012; 12(5): 1075.

235. Hollocks MJ, Jones CRG, Pickles A, Baird G, Happé F, Charman T et al. The

association between social cognition and executive functioning and symptoms of

anxiety and depression in adolescents with autism spectrum disorders. Autism

Research 2014; 7(2): 216-228.

236. Rommelse NNJ, Geurts HM, Franke B, Buitelaar JK, Hartman CA. A review on

cognitive and brain endophenotypes that may be common in autism spectrum disorder

and attention-deficit/hyperactivity disorder and facilitate the search for pleiotropic

genes. Neuroscience & Biobehavioral Reviews 2011; 35(6): 1363-1396.

237. Gottesman II, Gould TD. The endophenotype concept in psychiatry: Etymology and

strategic intentions. American Journal of Psychiatry 2003; 160(4): 636-645.

238. Beversdorf DQ, Missouri Autism Summit C. Phenotyping, etiological factors, and

biomarkers: toward precision medicine in autism spectrum disorders. Journal of

Developmental and Behavioral Pediatrics 2016; 37(8): 659-673.

239. Jurado MB, Rosselli M. The elusive nature of executive functions: a review of our

current understanding. Neuropsychology Review 2007; 17(3): 213-233.

Page | 198

240. de Vries M, Verdam MGE, Prins PJM, Schmand BA, Geurts HM. Exploring possible

predictors and moderators of an executive function training for children with an

autism spectrum disorder. Autism 2017; 22(4): 440-449.

241. Lien VE, Bart B, Nele C, Hilde P, Jean S, Johan W et al. Executive functioning and

local‐global visual processing: candidate endophenotypes for autism spectrum

disorder? Journal of Child Psychology and Psychiatry 2017; 58(3): 258-269.

242. Walsh P, Elsabbagh M, Bolton P, Singh I. In search of biomarkers for autism:

scientific, social and ethical challenges. Nature Reviews Neuroscience 2011; 12: 603.

243. Geurts H, Sinzig J, Booth R, HappÉ F. Neuropsychological heterogeneity in

executive functioning in autism spectrum disorders. International Journal of

Developmental Disabilities 2014; 60(3): 155-162.

244. Turner M. Towards an executive dysfunction account of repetitive behaviour in

autism. Autism as an executive disorder. Oxford University Press: New York, NY,

US, 1997, pp 57-100.

245. Turner M. Annotation: repetitive behaviour in autism: a review of psychological

research. The Journal of Child Psychology and Psychiatry and Allied Disciplines

1999; 40(6): 839-849.

Page | 199

246. Ozonoff S, Rogers SJ, Pennington BF. Asperger's syndrome: Evidence of an

empirical distinction from high-functioning autism. Journal of Child Psychology and

Psychiatry and Allied Disciplines 1991; 32(7): 1107-1122.

247. Hughes C, Russell J. Autistic children's difficulty with mental disengagement from an

object: Its implications for theories of autism. Developmental Psychology 1993;

29(3): 498-510.

248. Kenworthy L, Black DO, Harrison B, della Rosa A, Wallace GL. Are executive

control functions related to autism symptoms in high-functioning children? Child

Neuropsychology 2009; 15(5): 425-440.

249. Lopez BR, Lincoln AJ, Ozonoff S, Lai Z. Examining the relationship between

executive functions and restricted, repetitive symptoms of Autistic Disorder. Journal

of Autism and Developmental Disorders 2005; 35(4): 445-460.

250. Boyd BA, McBee M, Holtzclaw T, Baranek GT, Bodfish JW. Relationships among

repetitive behaviors, sensory features, and executive functions in high functioning

autism. Research in Autism Spectrum Disorders 2009; 3(4): 959-966.

251. Mazza M, Mariano M, Peretti S, Masedu F, Pino MC, Valenti M. The Role of Theory

of Mind on social information processing in children with autism spectrum disorders:

a mediation analysis. Journal of Autism and Developmental Disorders 2017; 47(5):

1369-1379.

Page | 200

252. Pellicano E. Links between theory of mind and executive function in young children

with autism: clues to developmental primacy. Developmental Psychology 2007;

43(4): 974-990.

253. McEvoy RE, Rogers SJ, Pennington BF. Executive function and social

communication deficits in young autistic children. Journal of Child Psychology and

Psychiatry 1993; 34(4): 563-578.

254. Zimmerman D, Ownsworth T, O’Donovan A, Roberts J, Gullo MJ. High-functioning

autism spectrum disorder in adulthood: A systematic review of factors related to

psychosocial outcomes. Journal of Intellectual & Developmental Disability 2018;

43(1): 2-19.

255. Lieb RW, Bohnert AM. Relations Between Executive Functions, Social Impairment,

and Friendship Quality on Adjustment Among High Functioning Youth with Autism

Spectrum Disorder. Journal of Autism and Developmental Disorders 2017; 47(9):

2861-2872.

256. Panerai S, Tasca D, Ferri R, D G, #x, Arrigo V et al. Executive Functions and

Adaptive Behaviour in Autism Spectrum Disorders with and without Intellectual

Disability. Psychiatry Journal 2014; 2014: 11.

257. Burger-Caplan R, Saulnier CA, Sparrow SS. Vineland Adaptive Behavior Scales. In:

Kreutzer J, DeLuca J, Caplan B (eds). Encyclopedia of Clinical Neuropsychology.

Springer International Publishing: Cham, 2017, pp 1-5.

Page | 201

258. Leung RC, Vogan VM, Powell TL, Anagnostou E, Taylor MJ. The role of executive

functions in social impairment in autism spectrum disorder. Child Neuropsychology

2016; 22(3): 336-344.

259. Constantino J, Gruber C. Social Responsiveness Scale. Western Psychological Press:

Los Angeles, 2005.

260. de Vries M, Geurts H. Influence of autism traits and executive functioning on quality

of life in children with an autism spectrum disorder. Journal of Autism and

Developmental Disorders 2015; 45(9): 2734-2743.

261. Bastiaansen D, Koot HM, Bongers IL, Varni JW, Verhulst FC. Measuring quality of

life in children referred for psychiatric problems: Psychometric properties of the

PedsQLTM 4.0 generic core scales. Quality of Life Research 2004; 13(2): 489-495.

262. Montgomery JM, Stoesz BM, McCrimmon AW. Emotional intelligence, theory of

mind, and executive functions as predictors of social outcomes in young adults with

Asperger syndrome. Focus on Autism and Other Developmental Disabilities 2012;

28(1): 4-13.

263. Reynolds C, Kamphaus R. BASC-2: Behavior assessment system for children second

edition; Manual supplement for the self-report of personality-interview. AGS

Publishing: Circle Pines, MN, 2005.

Page | 202

264. Oakland T, Harrison P. Adaptive Behaviour Assessment System II: Clinical use and

interpretation. Elsevier - Academic Press: San Diego CA, 2008.

265. Howlin P, Magiati I. Autism spectrum disorder: outcomes in adulthood. Current

Opinion in Psychiatry 2017; 30(2): 69-76.

266. Ustun T, Kostanjsek N, Chatterji S, FRehm J. Measuring health and Disability:

Manual for WHO Disability Assessment Schedule, WHODAS 2.0. World Health

Organisation: Malta, 2010.

267. Schmidt L, Kirchner J, Strunz S, Broźus J, Ritter K, Roepke S et al. Psychosocial

functioning and life satisfaction in adults with autism spectrum disorder without

intellectual impairment. Journal of Clinical Psychology 2015; 71(12): 1259-1268.

268. Sizoo B, van den Brink W, Koeter M, Gorissen van Eenige M, van Wijngaarden-

Cremers P, van der Gaag RJ. Treatment seeking adults with autism or ADHD and co-

morbid Substance Use Disorder: Prevalence, risk factors and functional disability.

Drug and Alcohol Dependence 2010; 107(1): 44-50.

269. American, Psychiatric, Association. Diagnostic and Statistical Manual of Mental

Disorders. 5th edn. American Psychiatric Association: Washington DC, 2013.

270. Pellicano E. Links between theory of mind and executive function in young children

with autism: clues to developmental primacy. Dev Psychol 2007; 43(4): 974-990.

Page | 203

271. Leung RC, Vogan VM, Powell TL, Anagnostou E, Taylor MJ. The role of executive

functions in social impairment in Autism Spectrum Disorder. Child Neuropsychol

2016; 22(3): 336-344.

272. Mostert-Kerckhoffs MA, Staal WG, Houben RH, Jonge MV. Stop and change:

Inhibition and flexibility skills are related to repetitive behavior in children and young

adults with autism spectrum disorders. J Autism Dev Disord 2015; 45(10): 3148-

3158.

273. de Vries M, Geurts H. Influence of Autism Traits and Executive Functioning on

Quality of Life in Children with an Autism Spectrum Disorder. J Autism Dev Disord

2015; 45(9): 2734-2743.

274. Maximo JO, Cadena EJ, Kana RK. The implications of brain connectivity in the

neuropsychology of autism. Neuropsychol Rev 2014; 24(1): 16-31.

275. Libero LE, DeRamus TP, Lahti AC, Deshpande G, Kana RK. Multimodal

neuroimaging based classification of autism spectrum disorder using anatomical,

neurochemical, and white mattter correlates. Cortex 2015; 66: 46-59.

276. Nomi JS, Uddin LQ. Developmental changes in large-scale network connectivity in

autism. Neuroimage: Clinical 2015; 7: 732-741.

Page | 204

277. Geurts HM, van den Bergh SF, Ruzzano L. Prepotent response inhibition and

interference control in autism spectrum disorders: Two Meta-Analyses. Autism Res

2014; 7(4): 407-420.

278. Willcutt EG, Doyle AE, Nigg JT, Faraone SV, Pennington BF. Validity of the

executive funciton theory of attention-deficit hyperactivity disorder: a meta-analytic

review. Biol Psychiatry 2006; 57(11): 1336-1346.

279. Barendse EM, Hendriks MP, Jansen JF, Backes WH, Hofman PA, Thoonen G et al.

Working memory deficits in high-functioning adolescents with autism spectrum

disorders: Neuropsychological and neuroimaging correlates. J Neurodev Disord 2013;

5(1): 1-11.

280. Czermainski FR, Bosa CA, Mina CS, Meimes MA, Miranda MC, Carim D et al.

Performance of children/adolescents with autism spectrum disorders in executive

function: Study of case series. Psychol Neurosci 2015; 8(3): 305-320.

281. Anderson P. Assessment and development of executive function (EF) during

childhood. Child Neuropsychol 2002; 8(2): 71-82.

282. Jurado MB, Rosselli M. The elusive nature of executive functions: a review of our

current understanding. Neuropsychol Rev 2007; 17: 213-233.

283. O'Hearn K, Asato M, Ordaz S, Luna B. Neurodevelopment and executive function in

autism. Dev Psychopathol 2008; 20(4): 1103-1132.

Page | 205

284. Gonzalez-Gadea ML, Chennu S, Bekinschtein TA, Rattazzi A, Beraudi A, Tripicchio

P et al. Predictive coding in autism spectrum disorder and attention deficit

hyperactivity disorder. J Neurophysiol 2015; 114(5): 2625-2636.

285. van Eylen L, Boets B, Steyaert J, Wagemans J, Noens I. Executive functioning in

autism spectrum disorders: influence of task and sample characteristics and relation to

symptom severity. Eur Child Adolesc Psychiatry 2015; 24(11): 1399-1417.

286. Leung RC, Zakzanis KK. Brief report: Cognitive flexibility in autism spectrum

disorders: A quantitative review. J Autism Dev Disord 2014; 44(10): 2628-2645.

287. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting system for

systematic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009.

288. Stroup DF, Berlin DA, Morton SC, Ilkin I, Williamson GD, Rennie D et al. Meta-

analysis of observational studies in epidemiology: a proposal for reporting. JAMA

2000; 283(15): 2008-2012.

289. American Psychiatric Association. Diagnostic and Statistical Manual of Mental

Disorders. 3rd edn: Washington, DC, 1980.

290. Espy KA. Using developmental, cognitive and neuroscience approaches to understand

executive control in young children. Dev Neuropsychol 2004; 26(1): 379-384.

Page | 206

291. Miyaki A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD. The unity

and diversity of executive functions and their contributions to complex "frontal lobe"

tasks: a latent variable analysis. Cogn Psychol 2000; 41(1): 49-100.

292. Armijo-Olivo S, Stiles CR, Hagen NA, Biondo PD, Cummings GG. Assessment of

study quality for systematic reviews: a comparison of the Cochrane Collaboration

Risk of Bias Tool and the Effective Public Health Practice Project Quality

Assessment Tool: methodological research. J Eval Clin Pract 2012; 18(1): 12-18.

293. Higgins JPT, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistencies in

meta-analyses. BMJ 2003; 327: 557-560.

294. Cohen J. Statistical Power Analysis for the Behavioral Sciences. Hillsdale, N.J.

Erlbaum Associates1988.

295. Scott FJ, Baron-Cohen S. Imagining real and unreal things: Evidence of a dissociation

in autism. J Cogn Neurosci 1996; 8(4): 371-382.

296. Kushki A, Drumm E, Pla Mobarak M, Tanel N, Dupuis A, Chau T et al. Investigating

the Autonomic Nervous System Response to Anxiety in Children with Autism

Spectrum Disorders. PLoS ONE 2013; 8(4): e59730.

doi:59710.51371/journal.pone.0059730.

Page | 207

297. Vissers EM, Cohen MX, Geurts HM. Brain connectivity and high functioning autism:

A promising path of research that needs refined models, methodological convergence,

and stronger behavioural links. Neurosci Biobehav Rev 2012; 36: 604-625.

298. Aoki Y, Abe O, Nippashi Y, Yamasue H. Comparison of white matter integrity

between auism spectrum disorder subjects and typically developing individuals: a

meta-analysis of diffusion tensor imaging tractography studies. Mol Autism 2013;

4(25).

299. Blakemore S-J, Choudhury S. Development of the adolescent brain: implications for

executive function and social cognition. J Child Psychol Psychiatry 2006; 47(3): 296-

312.

300. Foss-Feig JH, McPartland JC, Anticevic A, Wolf J. Re-conceptualizing ASD Within a

dimensional framework: Positive, negative, and cognitive feature clusters. J Autism

Dev Disord 2016; 46(1): 342-351.

301. McGorry PD. Staging in neuropsychiatry: a heuristic model for understanding

prevention and treatment. Neurotox Res 2010; 18: 244-255.

302. McGorry P, Keshavan M, Goldstone S, Amminger P, Allott K, Berk M et al.

Biomarkers and clinical staging in psychiatry. World Psychiatry 2014; 13: 211-223.

303. Hollocks MJ, Jones CR, Pickles A, Baird G, Happe F, Charman T et al. The

association between social cognition and executive functioning and symptoms of

Page | 208

anxiety and depression in adolescents with autism spectrum disorders. Autism Res

2014; 7(2): 216-228.

304. Jamison R, Bishop S, Huerta M, Halladay AK. The clinician perspective on sex

differences in autism spectrum disorders. Autism 2017; 21(6): 772-784.

305. Dworzynski K, Ronald A, Bolton P, Happé F. How different are girls and boys above

and below the diagnostic threshold for autism spectrum disorders? Journal of the

American Academy of Child & Adolescent Psychiatry 2012; 51(8): 788-797.

306. Van Wijngaarden-Cremers PJM, van Eeten E, Groen WB, Van Deurzen PA,

Oosterling IJ, Van der Gaag RJ. Gender and age differences in the core triad of

impairments in autism spectrum disorders: a systematic review and meta-analysis.

Journal of Autism and Developmental Disorders 2014; 44(3): 627-635.

307. Kreiser NL, White SW. ASD in females: are we overstating the gender difference in

diagnosis? Clinical Child and Family Psychology Review 2014; 17(1): 67-84.

308. Reilly D, Neumann DL, Andrews G. Sex and sex-role differences in specific

cognitive abilities. Intelligence 2016; 54: 147-158.

309. Hyde JS. Sex and cognition: gender and cognitive functions. Current Opinion in

Neurobiology 2016; 38: 53-56.

Page | 209

310. Torske T, Nærland T, Øie MG, Stenberg N, Andreassen OA. Metacognitive aspects

of executive function are highly associated with social functioning on parent-rated

measures in children with autism spectrum disorder. Frontiers in Behavioral

Neuroscience 2018; 11(258).

311. White EI, Wallace GL, Bascom J, Armour AC, Register-Brown K, Popal HS et al.

Sex differences in parent-reported executive functioning and adaptive behavior in

children and young adults with autism spectrum disorder. Autism Research 2017;

10(10): 1653-1662.

312. Howe YJ, Yatchmink Y, Viscidi EW, Morrow EM. Ascertainment and gender in

autism spectrum disorders. Journal of the American Academy of Child and Adolescent

Psychiatry 2014; 53(6): 698-700.

313. van Ommeren TB, Koot H, M., Scheeren A, M., Begeer S. Sex differences in the

reciprocal behaviour of children with autism. Autism 2016; 21(6): 795-803.

314. Zwaigenbaum L, Bryson SE, Szatmari P, Brian J, Smith IM, Roberts W et al. Sex

differences in children with autism spectrum disorder identified within a high-risk

infant cohort. Journal of Autism and Developmental Disorders 2012; 42(12): 2585-

2596.

315. Gur RC, Richard J, Calkins ME, Chiavacci R, Hansen JA, Bilker WB et al. Age

group and sex differences in performance on a computerized neurocognitive battery in

children age 8−21. Neuropsychology 2012; 26(2): 251-265.

Page | 210

316. Miller DI, Halpern DF. The new science of cognitive sex differences. Trends in

Cognitive Sciences 2014; 18(1): 37-45.

317. Fine C, Dupré J, Joel D. Sex-linked behavior: Evolution, stability, and variability.

Trends in Cognitive Sciences 2017; 21(9): 666-673.

318. Reed JL, Gallagher NM, Sullivan M, Callicott JH, Green AE. Sex differences in

verbal working memory performance emerge at very high loads of common

neuroimaging tasks. Brain and Cognition 2017; 113: 56-64.

319. Lejbak L, Crossley M, Vrbancic M. A male advantage for spatial and object but not

verbal working memory using the n-back task. Brain and Cognition 2011; 76(1): 191-

196.

320. Hull L, William M, Petrides K. Behavioural and cognitive sex/gender differences in

autism spectrum condition and typically developing males and females. Autism 2016;

21(6): 706-727.

321. Albein-Urios N, Youssef GJ, Kirkovski M, Enticott PG. Autism spectrum traits linked

with reduced performance on self-report behavioural measures of cognitive flexibility.

Journal of Autism and Developmental Disorders 2018.

322. Lord C, Rutter M, DiLavore P, Risi S, Gotham K. Autism diagnostic observation

schedule-2nd edition (ADOS-2). Western: Los Angeles, CA., 2012.

Page | 211

323. Wechsler D. Wechsler Test of Adult Reading. Psychological Corporation: San

Antonio, Texas, 2001.

324. Ecker C, Andrews DS, Gudbrandsen CM, et al. Association between the probability

of autism spectrum disorder and normative sex-related phenotypic diversity in brain

structure. JAMA Psychiatry 2017; 74(4): 329-338.

325. Svedholm-Häkkinen AM, Ojala SJ, Lindeman M. Male brain type women and female

brain type men: Gender atypical cognitive profiles and their correlates. Personality

and Individual Differences 2018; 122: 7-12.

326. Hill AC, Laird AR, Robinson JL. Gender differences in working memory networks: A

BrainMap meta-analysis. Biol Psychol 2014; 102: 18-29.

327. Ingalhalikar M, Smith A, Parker D, Satterthwaite TD, Elliott MA, Ruparel K et al.

Sex differences in the structural connectome of the human brain. Proceedings of the

National Academy of Sciences 2014; 111(2): 823-828.

328. Zeestraten EA, Gudbrandsen MC, Daly E, de Schotten MT, Catani M, Dell'Acqua F

et al. Sex differences in frontal lobe connectivity in adults with autism spectrum

conditions. Translational Psychiatry 2017; 7: e1090.

Page | 212

329. Pletzer B, Scheuringer A, Scherndl T. Global-local processing relates to spatial and

verbal processing: implications for sex differences in cognition. Scientific Reports

2017; 7(1): 10575.

330. Halari R, Hines M, Kumari V, Mehrotra R, Wheeler M, Ng V et al. Sex differences

and individual differences in cognitive performance and their relationship to

endogenous gonadal hormones and gonadotropins. Behavioral Neuroscience 2005;

119(1): 104-117.

331. Rohde MS, Georgescu AL, Vogeley K, Fimmers R, Falter-Wagner CM. Absence of

sex differences in mental rotation performance in autism spectrum disorder. Autism

2017; 0(0): 1362361317714991.

332. Sedgewick F, Hill V, Yates R, Pickering L, Pellicano E. Gender differences in the

social motivation and friendship experiences of autistic and non-autistic adolescents.

Journal of Autism and Developmental Disorders 2016; 46(4): 1297-1306.

333. Mandy W, Pellicano L, St Pourcain B, Skuse D, Heron J. The development of autistic

social traits across childhood and adolescence in males and females. Journal of Child

Psychology and Psychiatry 2018; 0(0).

334. Snyder HR, Kaiser RH, Warren SL, Heller W. Obsessive-compulsive disorder is

associated with broad impairments in executive function: A meta-analysis. Clinical

psychological science : a journal of the Association for Psychological Science 2015;

3(2): 301-330.

Page | 213

335. Rock PL, Roiser JP, Riedel WJ, Blackwell AD. Cognitive impairment in depression: a

systematic review and meta-analysis. Psychological Medicine 2013; 44(10): 2029-

2040.

336. Prouteau A, Atzeni T, Tastet H, Bergua V, Destaillats J-M, Verdoux H.

Neurocognitive insight and executive functioning in schizophrenia. Cognitive

Neuropsychiatry 2015; 20(1): 64-71.

337. Mosiołek A, Gierus J, Koweszko T, Szulc A. Cognitive impairment in schizophrenia

across age groups: a case–control study. BMC Psychiatry 2016; 16(1): 37.

338. Wood SJ, Pantelis C, Velakoulis D, Yücel M, Fornito A, McGorry PD. Progressive

changes in the development toward schizophrenia: studies in subjects at increased

symptomatic risk. Schizophrenia Bulletin 2008; 34(2): 322-329.

339. Derakshan N, Smyth S, Eysenck MW. Effects of state anxiety on performance using a

task-switching paradigm: An investigation of attentional control theory. Psychonomic

Bulletin & Review 2009; 16(6): 1112-1117.

340. Edwards MS, Edwards EJ, Lyvers M. Cognitive trait anxiety, stress and effort interact

to predict inhibitory control. Cognition and Emotion 2017; 31(4): 671-686.

Page | 214

341. Butler K, Klaus K, Edwards L, Pennington K. Elevated cortisol awakening response

associated with early life stress and impaired executive function in healthy adult

males. Hormones and Behavior 2017; 95: 13-21.

342. Arnsten AFT, Raskind MA, Taylor FB, Connor DF. The effects of stress exposure on

prefrontal cortex: Translating basic research into successful treatments for post-

traumatic stress disorder. Neurobiology of Stress 2015; 1: 89-99.

343. Scult MA, Knodt AR, Radtke SR, Brigidi BD, Hariri AR. Prefrontal executive control

rescues risk for anxiety associated with high threat and low reward brain function.

Cerebral Cortex 2017: 1-7.

344. Insel TR. The NIMH Research Domain Criteria (RDoC) Project: Precision medicine

for psychiatry. American Journal of Psychiatry 2014; 171(4): 395-397.

345. Insel TR. Rethinking schizophrenia. Nature 2010; 468: 187-193.

346. Casey BJ, Oliveri ME, Insel T. A neurodevelopmental perspective on the Research

Domain Criteria (RDoC) framework. Biological Psychiatry 2014; 76(5): 350-353.

347. Cuthbert BN, Insel TR. Toward the future of psychiatric diagnosis: the seven pillars

of RDoC. BMC Medicine 2013; 11(1): 126.

348. McGrath LM, Braaten EB, Doty ND, Willoughby BL, Wilson HK, O'Donnell EH et

al. Extending the ‘cross-disorder’ relevance of executive functions to dimensional

Page | 215

neuropsychiatric traits in youth. Journal of Child Psychology and Psychiatry 2016;

57(4): 462-471.

349. O'Toole MS, Pedersen AD. A systematic review of neuropsychological performance

in social anxiety disorder. Nordic Journal of Psychiatry 2011; 65(3): 147-161.

350. Spielberger C. Assessment of state and trait anxiety: Conceptual and methodological

issues. Southern Psychologist 1985; 2(4): 6-16.

351. Eysenck MW, Derakshan N, Santos R, Calvo MG. Anxiety and cognitive

performance: Attentional control theory. Emotion 2007; 7(2): 336-353.

352. Gagnon SA, Wagner AD. Acute stress and episodic memory retrieval:

neurobiological mechanisms and behavioral consequences. Annals of the New York

Academy of Sciences 2016; 1369(1): 55-75.

353. Edwards EJ, Edwards MS, Lyvers M. Individual differences in trait anxiety and goal-

commitment predict updating efficiency on the reading span task. Motivation and

Emotion 2016; 40(6): 936-945.

354. Berggren N, Derakshan N. Attentional control deficits in trait anxiety: Why you see

them and why you don’t. Biol Psychol 2013; 92(3): 440-446.

Page | 216

355. Heeren A, Maurage P, Philippot P. Revisiting attentional processing of non-emotional

cues in social anxiety: A specific impairment for the orienting network of attention.

Psychiatry Research; 228(1): 136-142.

356. Rolls ET. The functions of the orbitofrontal cortex. Brain and Cognition 2004; 55(1):

11-29.

357. Hadwin JA, Brogan J, Stevenson J. State anxiety and working memory in children: A

test of processing efficiency theory. Educational Psychology 2005; 25(4): 379-393.

358. Fujii Y, Kitagawa N, Shimizu Y, Mitsui N, Toyomaki A, Hashimoto N et al. Severity

of generalized social anxiety disorder correlates with low executive functioning.

Neuroscience Letters 2013; 543: 42-46.

359. Brown TA, DiNardo PA, Barlow DH. Anxiety Disorders Interview Schedule for

DSM-IV. Oxford University Press: 198 Madison Avenue, New York, New York,

10016, 1994.

360. Lovibond PF, Lovibond SH. The structure of negative emotional states: Comparison

of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and

Anxiety Inventories. Behaviour Research and Therapy 1995; 33(3): 335-343.

361. Mattick RP, Clarke JC. Development and validation of measures of social phobia

scrutiny fear and social interaction anxiety. Behaviour Research and Therapy 1998;

36(4): 455-470.

Page | 217

362. Baron-Cohen S, Wheelwright S, Skinner R, Martin J, Clubley E. The Autism-

Spectrum Quotient (AQ): Evidence from Asperger Syndrome/High-Functioning

Autism, Malesand Females, Scientists and Mathematicians. Journal of Autism and

Developmental Disorders 2001; 31(1): 5-17.

363. Spielberger C. State-Trait Anxiety Inventory: Bibliography. 2nd edition edn.

Consulting Psychologists Press: Palo Alto, CA, 1989.

364. Edwards EJ, Edwards MS, Lyvers M. Interrelationships between trait anxiety,

situational stress and mental effort predict phonological processing efficiency, but not

effectiveness. Emotion 2016; 16(5): 634-646.

365. Shanmugan S, Wolf DH, Calkins ME, Moore TM, Ruparel K, Hopson RD et al.

Common and dissociable mechanisms of executive system dysfunction across

psychiatric disorders in youth. The American journal of psychiatry 2016; 173(5): 517-

526.

366. Grisanzio KA, Goldstein-Piekarski AN, Wang M, Rashed Ahmed AP, Samara Z,

Williams LM. Transdiagnostic symptom clusters and associations with brain,

behavior, and daily function in mood, anxiety, and trauma disorders. JAMA

Psychiatry 2018; 75(2): 201-209.

Page | 218

367. McKlveen JM, Myers B, Herman JP. The Medial Prefrontal Cortex: Coordinator of

autonomic, neuroendocrine and behavioural responses to stress. Journal of

Neuroendocrinology 2015; 27(6): 446-456.

368. Murray B. Stein, Yin M. Kean. Disability and quality of life in social phobia:

epidemiologic findings. American Journal of Psychiatry 2000; 157(10): 1606-1613.

369. Wiersma D, Wanderling J, Dragomirecka E, Ganev K, Harrison G, An Der Heiden W

et al. Social disability in schizophrenia: its development and prediction over 15 years

in incidence cohorts in six European centres. Psychological Medicine 2000; 30(5):

1155-1167.

370. Nelson EE, Jarcho JM, Guyer AE. Social re-orientation and brain development: An

expanded and updated view. Developmental Cognitive Neuroscience 2016;

17(Supplement C): 118-127.

371. Carpenter WT. The RDoC controversy: alternate paradigm or dominant paradigm.

American Journal of Psychiatry 2016; 173(6): 562-563.

372. Hickie IB, Hermens DF, Naismith SL, Guasatella AJ, Glozier N, Scott J et al.

Evaluating differential developmental trajectories in to adolescent-onset mood and

psychotic disorders. BMC Psychiatry 2013; 13: 303.

Page | 219

373. Lee RSC, Hermens DF, Redoblado-Hodge MA, Naismith SL, Porter MA, Kaur M et

al. Neuropsychological and socio-occupational functioning in young psychiatric

outpatients: a longitudinal investigation. PLOS ONE 2013; 8(3): e58176.

374. Woon PS, Chia MY, Chan WY, Sim K. Neurocognitive, clinical and functional

correlates of subjective quality of life in Asian outpatients with schizophrenia.

Progress in Neuro-Psychopharmacology and Biological Psychiatry 2010; 34(3): 463-

468.

375. Fu S, Czajkowski N, Rund BR, Torgalsbøen A-K. The relationship between level of

cognitive impairments and functional outcome trajectories in first-episode

schizophrenia. Schizophrenia Research 2017.

376. Eisenberg DP, Berman KF. Executive function, neural circuitry, and genetic

mechanisms in schizophrenia. Neuropsychopharmacology 2009; 35(1): 258-277.

377. Libero LE, DeRamus TP, Lahti AC, Deshpande G, Kana RK. Multimodal

neuroimaging based classification of autism spectrum disorder using anatomical,

neurochemical, and white matter correlates. Cortex 2015; 66: 46-59.

378. Yerys BE, Antezana L, Weinblatt R, Jankowski KF, Strang J, Vaidya CJ et al. Neural

correlates of set-shifting in children with autism. Autism Research 2015; 8(4): 386-

397.

Page | 220

379. D'Cruz AM, Mosconi MW, Ragozzino ME, Cook EH, Sweeney JA. Alterations in the

functional neural circuitry supporting flexible choice behavior in autism spectrum

disorders. Translational Psychiatry 2016; 6(10): e916.

380. Davidson RJ. Anxiety and affective style: role of prefrontal cortex and amygdala.

Biological Psychiatry 2002; 51(1): 68-80.

381. Whitfield-Gabrieli S, Ghosh S, Nieto-Castanon A, Saygin Z, Doehrmann O, Chai X et

al. Brain connectomics predict response to treatment in social anxiety disorder.

Molecular Psychiatry 2016; 21: 680-685.

382. Cremers HR, Roelofs K. Social anxiety disorder: a critical overview of

neurocognitive research. WIREs Cognitive Science 2016; 7: 218-232.

383. Allen P, Bennett KP, Heritage B. SPSS Statistics version 22: a practical guide. 3rd

edition edn. Cengage Learning Australia: South melbourne, Victoria, 2014.

384. Silton RL, Heller W, Towers DN, Engels AS, Spielberg JM, Edgar JC et al. The time

course of activity in dorsolateral prefrontal cortex and anterior cingulate cortex during

top-down attentional control. NeuroImage 2010; 50(3): 1292-1302.

385. Seidman LJ, Shapiro DI, Stone WS, Woodberry KA, Ronzio A, Cornblatt BA et al.

Association of neurocognition with transition to psychosis: Baseline functioning in

the second phase of the north american prodrome longitudinal study. JAMA

Psychiatry 2016; 73(12): 1239-1248.

Page | 221

386. Ayesa-Arriola R, Rodríguez-Sánchez JM, Suero ES, Reeves LE, Tabarés-Seisdedos

R, Crespo-Facorro B. Diagnosis and neurocognitive profiles in first-episode non-

affective psychosis patients. European Archives of Psychiatry and Clinical

Neuroscience 2016; 266(7): 619-628.

387. Goldin PR, Ziv M, Jazaieri M, Hahn K, Heimberg R, Gross JJ. Impact of cognitive

behavioral therapy for social anxiety disorder on the neural dynamics of cognitive re-

appraisal of negative self-beliefs. JAMA Psychiatry 2013; 70(10): 1048-1056.

388. Boehme S, Miltner WHR, Straube T. Neural correlates of self-focused attention in

social anxiety. Social Cognitive and Affective Neuroscience 2015; 10(6): 856-862.

389. Toplak ME, Bucciarelli SM, Jain U, Tannock R. Executive functions: Performance-

based measures and the Behavior Rating Inventory of Executive Function (BRIEF) in

adolescents with attention deficit/hyperactivity disorder (ADHD). Child

Neuropsychology 2008; 15(1): 53-72.

390. Moss Lowengrub K, Stryjer R, Birger M, Iancu I. Social anxiety disorder comorbid

with schizophrenia: the importance of screening for this underrecognised and

underrated condition. Israeli Journal of Psychiatry and related Science 2015; 52(1):

40-46.

391. van Heijst B, Geurts H. Quality of life in autism across the lifespan: A meta-analysis.

Autism 2015; 19(2): 158-167.

Page | 222

392. Chekroud AM. Bigger data, harder questions—opportunities throughout mental health

care. JAMA Psychiatry 2017; 74(12): 1183-1184.

393. Huys QM. Advancing clinical improvements for patients using the theory-driven and

data-driven branches of computational psychiatry. JAMA Psychiatry 2018; 75(3):

225-226.

394. Kanchanatawan B, Sriswasdi S, Thika S, Stoyanov D, Sirivichayakul S, Carvalho

André F et al. Towards a new classification of stable phase schizophrenia into major

and simple neuro-cognitive psychosis: Results of unsupervised machine learning

analysis. Journal of Evaluation in Clinical Practice 2018; 0(0).

395. Bone D, Bishop Somer L, Black Matthew P, Goodwin Matthew S, Lord C, Narayanan

Shrikanth S. Use of machine learning to improve autism screening and diagnostic

instruments: effectiveness, efficiency, and multi-instrument fusion. Journal of Child

Psychology and Psychiatry 2016; 57(8): 927-937.

396. Duda M, Haber N, Daniels J, Wall DP. Crowdsourced validation of a machine-

learning classification system for autism and ADHD. Translational Psychiatry 2017;

7: e1133.

397. Bishop-Fitzpatrick L, Movaghar A, Greenberg Jan S, Page D, DaWalt Leann S,

Brilliant Murray H et al. Using machine learning to identify patterns of lifetime health

problems in decedents with autism spectrum disorder. Autism Research 2018; 0(0).

Page | 223

398. Anney RJL, Ripke S, Anttila V, Grove J, Holmans P, Huang H et al. Meta-analysis of

GWAS of over 16,000 individuals with autism spectrum disorder highlights a novel

locus at 10q24.32 and a significant overlap with schizophrenia. Molecular Autism

2017; 8(1): 21.

399. Tin LNW, Lui SSY, Ho KKY, Hung KSY, Wang Y, Yeung HKH et al. High-

functioning autism patients share similar but more severe impairments in verbal

theory of mind than schizophrenia patients. Psychological Medicine 2017; 48(8):

1264-1273.

400. Addington J, Brooks BL, Addington D. Cognitive functioning in first episode

psychosis: initial presentation. Schizophrenia Research 2003; 62(1): 59-64.

401. Goldstein G, Minshew NJ, Allen DN, Seaton BE. High-functioning autism and

schizophrenia A comparison of an early and late onset neurodevelopmental disorder.

Archives of Clinical Neuropsychology 2002; 17(5): 461-475.

402. Chisholm K, Lin A, Abu-Akel A, Wood SJ. The association betweeen autism and

schizophrenia spectrum disorders: a review of eight alternate models of co-

occurrence. Neuroscience and Biobehavioural Reviews 2015; 55: 173-183.

403. Park MTM, Raznahan A, Shaw P, Gogtay N, Lerch JP, Chakravarty MM.

Neuroanatomical phenotypes in mental illness: identifying convergent and divergent

Page | 224

cortical phenotypes across autism, ADHD and schizophrenia. Journal of psychiatry &

neuroscience : JPN 2018; 43(3): 201-212.

404. DiLalla L F, McCrary M, Diaz E. A review of endophenotypes in schizophrenia and

autism: The next phase for understanding genetic etiologies. American Journal of

Medical Genetics Part C: Seminars in Medical Genetics 2017; 175(3): 354-361.

405. O'Connell KS, McGregor NW, Lochner C, Emsley R, Warnich L. The genetic

architecture of schizophrenia, bipolar disorder, obsessive-compulsive disorder and

autism spectrum disorder. Molecular and Cellular Neuroscience 2018; 88: 300-307.

406. Lugo Marín J, Alviani Rodríguez-Franco M, Mahtani Chugani V, Magán Maganto M,

Díez Villoria E, Canal Bedia R. Prevalence of schizophrenia spectrum disorders in

average-iq adults with autism spectrum disorders: a meta-analysis. Journal of Autism

and Developmental Disorders 2018; 48(1): 239-250.

407. Pepper KL, Demetriou EA, Park SH, Song YC, Hickie IB, Cacciotti-Saija C et al.

Autism, early psychosis, and social anxiety disorder: understanding the role of social

cognition and its relationship to disability in young adults with disorders characterized

by social impairments. Translational Psychiatry 2018; 8(1): 233.

408. Fusar-Poli P, Bonoldi I, Yung AR, et al. Predicting psychosis: Meta-analysis of

transition outcomes in individuals at high clinical risk. Archives of General Psychiatry

2012; 69(3): 220-229.

Page | 225

409. Chien YL, Gau SSF, Shang CY, Chiu YN, Tsai WC, Wu YY. Visual memory and

sustained attention impairment in youths with autism spectrum disorders.

Psychological Medicine 2015; 45(11): 2263-2273.

410. Pina-Camacho L, Garcia-Prieto J, Parellada M, Castro-Fornieles J, Gonzalez-Pinto

AM, Bombin I et al. Predictors of schizophrenia spectrum disorders in early-onset

first episodes of psychosis: a support vector machine model. European Child &

Adolescent Psychiatry 2015; 24(4): 427-440.

411. Bowler DM, Limoges E, Mottron L. Different verbal learning strategies in autism

spectrum disorder: Evidence from the Rey Auditory Verbal Learning Test. Journal of

Autism and Developmental Disorders 2009; 39(6): 910-915.

412. Carrión RE, Walder DJ, Auther AM, McLaughlin D, Zyla HO, Adelsheim S et al.

From the psychosis prodrome to the first-episode of psychosis: No evidence of a

cognitive decline. Journal of Psychiatric Research 2018; 96: 231-238.

413. Ayesa-Arriola R, Setién-Suero E, Neergaard KD, Ferro A, Fatjó-Vilas M, Ríos-Lago

M et al. Evidence for trait related Theory of Mind impairment in First Episode

Psychosis patients and its relationship with processing speed: A 3 Year follow-up

study. Frontiers in Psychology 2016; 7(592).

414. Williams DL, Minshew NJ, Goldstein G. Further understanding of complex

information processing in verbal adolescents and adults with autism spectrum

disorders. Autism 2015; 19(7): 859-867.

Page | 226

415. Uljarevic M, Hamilton A. Recognition of emotions in autism: A formal meta-analysis.

Journal of Autism and Developmental Disorders 2013; 43(7): 1517-1526.

416. Healey KM, Bartholomeusz CF, Penn DL. Deficits in social cognition in first episode

psychosis: A review of the literature. Clinical Psychology Review 2016; 50: 108-137.

417. Murray K, Johnston K, Cunnane H, Kerr C, Spain D, Gillan N et al. A new test of

advanced theory of mind: The “Strange Stories Film Task” captures social processing

differences in adults with autism spectrum disorders. Autism Research 2017; 10(6):

1120-1132.

418. Seidman LJ, Shapiro DI, Stone WS, et al. Association of neurocognition with

transition to psychosis: Baseline functioning in the second phase of the north

american prodrome longitudinal study. JAMA Psychiatry 2016; 73(12): 1239-1248.

419. Burton CZ, Harvey PD, Patterson TL, Twamley EW. Neurocognitive insight and

objective cognitive functioning in schizophrenia. Schizophrenia Research 2016;

171(1): 131-136.

420. Deshpande G, Libero L, Sreenivasan K, Deshpande H, Kana R. Identification of

neural connectivity signatures of autism using machine learning. Frontiers in Human

Neuroscience 2013; 7(670).

Page | 227

421. Kanchanatawan B, Thika S, Anderson G, Galecki P, Maes M. Affective symptoms in

schizophrenia are strongly associated with neurocognitive deficits indicating disorders

in executive functions, visual memory, attention and social cognition. Progress in

Neuro-Psychopharmacology and Biological Psychiatry 2018; 80: 168-176.

422. Michail M, Birchwood M. Social anxiety disorder in first-episode psychosis:

incidence, phenomenology and relationship with paranoia. British Journal of

Psychiatry 2018; 195(3): 234-241.

423. Maddox BB, White SW. Comorbid social anxiety disorder in adults with autism

spectrum disorder. Journal of Autism and Developmental Disorders 2015; 45(12):

3949-3960.

424. Birchwood M. Pathways to emotional dysfunction in first-episode psychosis. British

Journal of Psychiatry 2018; 182(5): 373-375.

425. Adams GC, Balbuena L, Meng X, Asmundson GJG. When social anxiety and

depression go together: A population study of comorbidity and associated

consequences. Journal of Affective Disorders 2016; 206: 48-54.

426. Dai J, Du X, Yin G, Zhang Y, Xia H, Li X et al. Prevalence, demographic and clinical

features of comorbid depressive symptoms in drug naïve patients with schizophrenia

presenting with first episode psychosis. Schizophrenia Research 2018; 193: 182-187.

Page | 228

427. Lord C, Rutter M, DiLavore PC, Risi S, Gotham K, Bishop SL. Autism Diagnostic

Observation Schedule (ADOS-2). 2nd edn. Western Psychological Services: Torrance,

CA, 2012.

428. Brown TA, Barlow DH. Anxiety and Related Disorders Interview Schedule for DSM-

5 (ADIS-5). Oxford University Press: Oxford, UK, 2014.

429. First MB, Spitzer RL, Gibbon M, Williams JBW. Structured Clinical Interview for

DSM-IV Axis I Disorders — Patient Edition (SCID-I/P, Version 2.0), Biometrics

Research. New York State Psychiatric Institute: New York, 1995.

430. Wechsler D. Wechsler Abbreviated Scale of Intelligence (WASI) Manual.

Psychological Corportation: San Antonio, Texas, 1999.

431. Wechsler D. Wechsler Test of Adult Reading: WTAR. The Psychological Corporation:

San Antonio, Texas, 2001.

432. Lovibond PF, Lovibond SH. The structure of negative emotional states: comparison

of the Depression Anxiety Stress Scales (DASS) with the Beck Depression and

Anxiety Inventories. Behav Res Ther 1995; 33(3): 335-343.

433. Heimberg RG, Mueller GP, Holt CS, Hope DA, Liebowitz MR. Assessment of

anxiety in social interaction and being observed by others: The Social Interaction

Anxiety Scale and the Social Phobia Scale. Behav 1992; 23(1): 53-73.

Page | 229

434. Allison C, Auyeung B, Baron-Cohen S. Toward brief "Red Flags" for autism

screening: The Short Autism Spectrum Quotient and the Short Quantitative Checklist

for Autism in toddlers in 1,000 cases and 3,000 controls [corrected]. J Am Acad Child

Adolesc Psychiatry 2012; 51(2): 202-212 e207.

435. van Buuren S, Groothuis-Oudshoorn K. mice: Multivariate Imputation by Chained

Equations in R. Journal of Statistical Software 2011; 45(3): 1-67.

436. Calle ML, Urrea V, Boulesteix AL, Malats N. AUC-RF: A New Strategy for

Genomic Profiling with Random Forest. Human Heredity 2011; 72(2): 121-132.

437. Kursa MB, Rudnicki WR. Feature Selection with the Boruta Package. 2010 2010;

36(11): 13.

438. Tibshirani R. Regression shrinkage and selection via the Lasso. Journal of the Royal

Statistical Society Series B (Methodological) 1996; 58(1): 267-288.

439. Zou H, Hastie T. Regularization and variable selection via the elastic net. Journal of

the Royal Statistical Society: Series B (Statistical Methodology) 2005; 67(2): 301-

320.

440. Chipman HA, George EI, McCulloch RE. BART: Bayesian additive regression trees.

Ann Appl Stat 2010; 4(1): 266-298.

441. Breiman L. Random Forests. Machine Learning 2001; 45(1): 5-32.

Page | 230

442. Hastie T, Tibshirani R. Bayesian Backfitting. Statistical Science 2000; 15(3): 196-

213.

443. Brier GW. Verification of forecasts expressed in terms of probability. Monthly

Weather Review 1950; 78(1): 1-3.

444. Vanderlooy S, Hüllermeier E. A critical analysis of variants of the AUC. Machine

Learning 2008; 72(3): 247-262.

445. Pepe MS, Janes HE. Gauging the Performance of SNPs, Biomarkers, and Clinical

Factors for Predicting Risk of Breast Cancer. JNCI: Journal of the National Cancer

Institute 2008; 100(14): 978-979.

446. Kruppa J, Ziegler A, König IR. Risk estimation and risk prediction using machine-

learning methods. Human genetics 2012; 131(10): 1639-1654.

447. Wu Y-C, Lee W-C. Alternative performance measures for prediction models. PloS

one 2014; 9(3): e91249-e91249.

448. Cardillo R, Menazza C, IC M. Visuoconstructive abilities and visuospatial memory in

autism spectrum disorder without intellectual disability: Is the role of local bias

specific to the cognitive domain tested? Neuropsychology 2018; 32(7): 822-834.

Page | 231

449. Wannan CMJ, Bartholomeusz CF, Cropley VL, Van Rheenen TE, Panayiotou A,

Brewer WJ et al. Deterioration of visuospatial associative memory following a first

psychotic episode: a long-term follow-up study. Psychological Medicine 2017; 48(1):

132-141.

450. Bishop S. Trait anxiety and impoverished prefronatl control of attention. Nature

Neuroscience 2009; 12(1): 92-98.

451. Haigh SM, Walsh JA, Mazefsky CA, Minshew NJ, Eack SM. Processing speed is

impaired in adults with autism spectrum disorder, and relates to social communication

abilities. Journal of Autism and Developmental Disorders 2018; 48(8): 2653-2662.

452. Liu C-C, Hua M-S, Hwang T-J, Chiu C-Y, Liu C-M, Hsieh MH et al. Neurocognitive

functioning of subjects with putative pre-psychotic states and early psychosis.

Schizophrenia Research 2015; 164(1): 40-46.

453. Mathias SR, Knowles EEM, Barrett J, Leach O, Buccheri S, Beetham T et al. The

Processing-Speed Impairment in Psychosis Is More Than Just Accelerated Aging.

Schizophrenia Bulletin 2017; 43(4): 814-823.

454. Luna B, Doll SK, Hegedus SJ, Minshew NJ, Sweeney JA. Maturation of Executive

Function in Autism. Biological Psychiatry 2007; 61(4): 474-481.

Page | 232

455. Tupak SV, Badewien M, Dresler T, Hahn T, Ernst LH, Herrmann MJ et al.

Differential prefrontal and frontotemporal oxygenation patterns during phonemic and

semantic verbal fluency. Neuropsychologia 2012; 50(7): 1565-1569.

456. Michail M, Birchwood M. Social anxiety disorder in first-episode psychosis:

incidence, phenomenology and relationship with paranoia. British Journal of

Psychiatry 2009; 195(3): 234-241.

457. Saeedi H, Addington J, Addington D. The association of insight with psychotic

symptoms, depression, and cognition in early psychosis: A 3-year follow-up.

Schizophrenia Research 2007; 89(1): 123-128.

458. Demetriou EA, Pepper KL, Park SH, Pellicano E, Song CY, Naismith SL et al.

Autism Spectrum Disorder: an examination of sex differences in neuropsychologcal

and self-report measures of cognitive and executive function. under review.

459. Petersen SE, Posner MI. The attention system of the human brain: 20 years after.

Annual rerview of Neuroscience 2012; 35: 73-89.

460. Park SH, Song YJC, Demetriou EA, Pepper KL, Norton A, Thomas EE et al.

Disability, functioning, and quality of life among treatment-seeking young autistic

adults and its relation to depression, anxiety, and stress. Autism 2019:

1362361318823925.

Page | 233

461. Park SH, Demetriou EA, Pepper KL, Song YJC, Thomas EE, Hickie IB et al.

Validation of the 36-item and 12-item self-report World Health Organization

Disability Assessment Schedule II (WHODAS-II) in individuals with autism

spectrum disorder. Autism Research 2019; 0(0).

462. Demetriou EA, Song CY, Park SH, Pepper KL, Naismith SL, Hermens DF et al.

Autism, Early Psychosis, and Social Anxiety Disorder: a transdiagnostic examination

of executive function cognitive circuitry and contribution to disability. Translational

psychiatry 2018; 8(1): 200-200.

463. Christensen DL, Braun KVN, Baio J, Bilder D, Charles J, Constantino JN et al.

Prevalence and Characteristics of Autism Spectrum Disorder Among Children Aged 8

Years - Autism and Developmental Disabilities Monitoring Network, 11 Sites, United

States, 2012. Morbidity and mortality weekly report Surveillance summaries

(Washington, DC : 2002) 2018; 65(13): 1-23.

464. La Malfa G, Lassi S, Bertelli M, Salvini R, Placidi GF. Autism and intellectual

disability: a study of prevalence on a sample of the Italian population. Journal of

Intellectual Disability Research 2004; 48(3): 262-267.

465. Australian, Bureau, of, Statistics. Disability ageing and carers: summary of findings,

Australia 2015. ABS cat. no 4430.0 Canberra ABS2016.

466. Kerns CM, Kendall PC. The Presentation and Classification of Anxiety in Autism

Spectrum Disorder. Clinical Psychology: Science and Practice 2012; 19(4): 323-347.

Page | 234

467. Kerns CM, Kendall PC, Berry L, Souders MC, Franklin ME, Schultz RT et al.

Traditional and Atypical Presentations of Anxiety in Youth with Autism Spectrum

Disorder. Journal of Autism and Developmental Disorders 2014; 44(11): 2851-2861.

468. Crane L, Adams F, Harper G, Welch J, Pellicano E. ‘Something needs to change’:

Mental health experiences of young autistic adults in England. Autism 2018:

1362361318757048.

469. Foley K, Trollor J. Management of mental ill health in people with autism spectrum

disorder. Australian Family Physician 2015; 44: 784-790.

470. Mandell DS, Morales KH, Marcus SC, Stahmer AC, Doshi J, Polsky DE.

Psychotropic Medication Use Among Medicaid-Enrolled Children With Autism

Spectrum Disorders. Pediatrics 2008; 121(3): e441.

471. Cvejic RC, Arnold SRC, Foley K-R, Trollor JN. Neuropsychiatric profile and

psychotropic medication use in adults with autism spectrum disorder: results from the

Australian Longitudinal Study of Adults with Autism. BJPsych Open 2018; 4(6): 461-

466.

472. Willcutt EG, Doyle AE, Nigg JT, Faraone SV, Pennington BF. Validity of the

executive funciton theory of attention-deficit hyperactivity disorder: a meta-analytic

review. Biological Psychiatry 2006; 57(11): 1336-1346.

Page | 235

473. Geurts HM, van den Bergh SF, Ruzzano L. Prepotent response inhibition and

interference control in autism spectrum disorders: Two Meta-Analyses. Autism

Research 2014; 7(4): 407-420.

474. Kercood S, Grskovic JA, Banda D, Begeske J. Working memory and autism: A

review of literature. Research in Autism Spectrum Disorders 2014; 8(10): 1316-1332.

475. Thomas BH, Ciliska D, Dobbins M, Muicucci S. A process for systematically

reviewing the literature: providing the research evidence for public health nursing

interventions

Worldviews on Evidence-Based Nursing 2004; 1: 176-184.

476. Higgins JPT, Green S. Cochrane Handbook of Systematic Reviews of interventions.

The Cochrane Collaboration2010.

477. Cohen J. Statistical Power Analysis for the Behavioural Science. Routledge2013,

567pp.

478. Maximo JO, Cadena EJ, Kana RK. The implications of brain connectivity in the

neuropsychology of autism. Neuropsychology Review 2014; 24(1): 16-31.

479. Bakhtiari R, Zürcher NR, Rogier O, Russo B, Hippolyte L, Granziera C et al.

Differences in white matter reflect atypical developmental trajectory in autism: A

Tract-based Spatial Statistics study. NeuroImage: Clinical 2012; 1(1): 48-56.

Page | 236

480. Ikuta T, Shafritz KM, Bregman J, Peters BD, Gruner P, Malhotra AK et al. Abnormal

cingulum bundle development in autism: A probabilistic tractography study.

Psychiatry Research - Neuroimaging 2014; 221(1): 63-68.

481. Vissers EM, Cohen MX, Geurts HM. Brain connectivity and high functioning autism:

A promising path of research that needs refined models, methodological convergence,

and stronger behavioural links. Neuroscience and Biobehavioral Reviews 2012; 36:

604-625.

482. Anderson P. Assessment and development of executive function (EF) during

childhood. Child Neuropsychology 2002; 8(2): 71-82.

483. Taylor SJ, Barker LA, Heavey L, McHale S. The typical development trajectory of

social and executive functions in late adolescence and early childhood.

Developmental Psychology 2013; 49(7): 1253-1265.

484. Huizinga M, Dolan CV, van der Molen MW. Age related change in executive

function: Developmental trends and a laten variable analysis. Neuropsychologia 2006;

44: 2017-2036.

485. van den Bergh SF, Scheeren AM, Begeer S, Koot HM, Geurts HM. Age related

differences of executive functioning problems in everyday life of children and

adolescents in the autism spectrum. Journal of Autism and Developmental Disorders

2014; 44(8): 1959-1971.

Page | 237

486. Berger HJC, Van Spaendonck KPM, Horstink MWIM, Buytenhuijs EL, Lammers

PWJM, Cools AR. Cognitive shifting as a predictor of progress in social

understanding in high-functioning adolescents with autism: A prospective study.

Journal of Autism and Developmental Disorders 1993; 23(2): 341-360.

487. Pennington BF, Ozonoff S. Executive functions and developmental psychopathology.

Child Psychology & Psychiatry & Allied Disciplines 1996; 37(1): 51-87.

488. Hunter SJ, Edidin JP, Hinkle CD. The developmental neuropsychology of executive

function. In: Hunter SJ, Sparrow EP (eds). Executive Function and Dysfunction.

Cambridge University Press2012, pp 17-36.

489. Schmitz N, Rubia K, Daly E, Smith A, Williams S, Murphy DG. Neural correlates of

executive function in autistic spectrum disorders. Biological Psychiatry 2006; 59(1):

7-16.

490. Mostert-Kerckhoffs MA, Staal WG, Houben RH, Jonge MV. Stop and change:

Inhibition and flexibility skills are related to repetitive behavior in children and young

adults with autism spectrum disorders. Journal of Autism and Developmental

Disorders 2015; 45(10): 3148-3158.

491. Romine C, Reynolds C. A model of the development of frontal lobe functioning:

findings from a meta-analysis. Applied Neuropsychology 2005; 12(190-201).

Page | 238

492. Blakemore S-J, Choudhury S. Development of the adolescent brain: implications for

executive function and social cognition. Journal of Child Psychology and Psychiatry

2006; 47(3): 296-312.

493. Kawakubo Y, Kuwabara H, Watanabe K, Minowa M, Someya T, Minowa I et al.

Impaired prefrontal hemodynamic maturation in Autism and unaffected siblings.

PLoS ONE 4(9): e6881 doi:101371/journalpone0006881 2009.

494. Pellicano E. The development of core cognitive skills in Autism: A 3-year prospective

study. Child Development 2010; 81(5): 1400-1416.

495. Czermainski FR, dos Santos Riesgo R, Guimaraes LSP, de Salles JF, Bosa CA.

Executive functions in children and adolescents with autism spectrum disorders.

Paideia 2014; 24(57): 85-94.

496. Bramham J, Ambery F, Young S, Morris R, Russell A, Xenitidis K et al. Executive

functioning differences between adults with attention deficit hyperactivity disorder

and autistic spectrum disorder in initiation, planning and strategy formation. Autism

2009; 13(3): 245-264.

497. Rubia K, Smith AB, Taylor E, Brammer M. Linear age-correlated functional

development in right inferior fronto-striato-cerebellar netweorks during response

inhibition and anterior cingulate during error-related processes. Human Brain

Mapping 2007; 28: 1163-1177.

Page | 239

498. Solomon M, Yoon JH, Ragland J, Niendam TA, Lesh TA, Fairbrother W et al. The

development of the neural substrates of cognitive control in adolescents with autism

spectrum disorders. Biological Psychiatry 2014; 76(5): 412-421.

499. Kretschmer A, Lampmann S-A, Altgassen M. Relations between moral reasoning,

theory of mind and executive functions in children with autism spectrum disorders.

International Journal of Developmental Disabilities 2014; 60(3): 174-183.

500. Tamnes CK, Walhovd KB, Grydeland H, Holland D, Ostby Y, Dale AM et al.

Longitudinal working memory development is related to structural maturation of

frontal and parietal cortices. Journal of Cognitive Neuroscience 2013; 25(1611-1623):

1611-1623.

501. Rosenthal M, Wallace GL, Lawson R, Wills MC, Dixon E, Yerys BE et al.

Impairments in real-world executive function increase from childhood to adolescence

in autism spectrum disorders. Neuropsychology 2013; 27(1): 13-18.

502. Lever AG, Werkle-Bergner M, Brandmaier AM, Richard Ridderinkhof K, Geurts

HM. Atypical working memory decline across the adult lifespan in autism spectrum

disorder? Journal of Abnormal Psychology 2015; 124(4): 1014-1026.

503. Geurts HM, Vissers ME. Elderly with autism: Executive functions and memory.

Journal of Autism and Developmental Disorders 2012; 42(5): 665-675.

Page | 240

504. Gilotty L, Kenworthy L, Sirian L, Black DO, Wagner AE. Adaptive skills and

executive function in autism spectrum disorders. Child Neuropsychology 2002; 8(4):

241-248.

505. Adamo N, Huo L, Adelsberg S, Petkova E, Castellanos FX, Di Martino A. Response

time intra-subject variability: commonalities between children with autism spectrum

disorders and children with ADHD. European Child & Adolescent Psychiatry 2014;

23: 69-79.

506. Adams NC, Jarrold C. Inhibition and validity of the Stroop task for children with

autism. Journal of Autism and Developmental Disorders 2009; 39(8): 1112-1121.

507. Alderson-Day B. Verbal problem-solving in autism spectrum disorders: A problem of

plan construction? Autism Research 2011; 4(6): 401-411.

508. Alderson-Day B, McGonigle-Chalmers M. Is it a bird? Is it a plane? Category use in

problem-solving in children with autism spectrum disorders. Journal of Autism and

Developmental Disorders 2011; 41(5): 555-565.

509. Alderson-Day B. Verbal problem-solving difficulties in autism spectrum disorders

and atypical language development. Autism Research 2014; 7(6): 720-730.

510. Altgassen M, Koban N, Kliegel M. Do adults with Autism Spectrum Disorders

compensate in naturalistic prospective memory tasks? Journal of Autism and

Developmental Disorders 2012; 42(10): 2141-2151.

Page | 241

511. Ambery FZ, Russell AJ, Perry K, Morris R, Murphy DGM. Neuropsychological

functioning in adults with Asperger syndrome. Autism 2006; 10(6): 551-564.

512. Andersen PN, Hovik KT, Skogli EW, Egeland J, Øie M. Symptoms of ADHD in

children with high-functioning autism are related to impaired verbal working memory

and verbal delayed recall. PLoS ONE 2013; 8(5).

513. Andersen PN, Skogli EW, Hovik KT, Egeland J, Oie M. Associations among

symptoms of autism, symptoms of depression and executive functions in children

with high-functioning autism: A 2 year follow-up study. Journal of Autism and

Developmental Disorders 2015; 45(8): 2497-2507.

514. Baez S, Rattazzi A, Gonzalez-Gadea ML, Torralva T, Vigliecca NS, Decety J et al.

Integrating intention and context: Assessing social cognition in adults with Asperger

syndrome Frontiers in Human Neuroscience 2012; 6(302): 1-30.

515. Barneveld PS, De Sonneville L, Van Rijn S, Van Engeland H, Swaab H. Impaired

response inhibition in autism spectrum disorders, a marker of vulnerability to

schizophrenia spectrum disorders? Journal of the International Neuropsychological

Society 2013; 19(6): 646-655.

516. Barron-Linnankoski S, Reinvall O, Lahervuori A, Voutilainen A, Lahti-Nuuttila P,

Korkman M. Neurocognitive performance of children with higher functioning autism

spectrum disorders on the NEPSY-II. Child Neuropsychology 2015; 21(1): 55-77.

Page | 242

517. Beacher FDCC, Radulescu E, Minati L, Baron-Cohen S, Lombardo MV, Lai MC et

al. Sex differences and Autism: Brain function during verbal fluency and mental

rotation. PLoS ONE 2012; 7(6).

518. Begeer S, Wierda M, Scheeren AM, Teunisse J-P, Koot HM, Geurts HM. Verbal

fluency in children with autism spectrum disorders: Clustering and switching

strategies. Autism 2014; 18(8): 1014-1018.

519. Biscaldi M, Bednorz N, Weissbrodt K, Saville CWN, Feige B, Bender S et al.

Cognitive endophenotypes of attention deficit/hyperactivity disorder and intra-subject

variability in patients with autism spectrum disorder. Biol Psychol 2016; 118: 25-34.

520. Bodner KE, Beversdorf DQ, Saklayen SS, Christ SE. Noradrenergic moderation of

working memory impairments in adults with autism spectrum disorder. Journal of the

International Neuropsychological Society 2012; 18(3): 556-564.

521. Brady DI, Schwean VL, Saklofske DH, McCrimmon AW, Montgomery JM, Thorne

KJ. Conceptual and perceptual set-shifting executive abilities in young adults with

Asperger's syndrome. Research in Autism Spectrum Disorders 2013; 7(12): 1631-

1637.

522. Brenner LA, Shih VH, Colich NL, Sugar CA, Bearden CE, Dapretto M. Time

reproduction performance is associated with age and working memory in high-

functioning youth with autism spectrum disorder. Autism Research 2015; 8(1): 29-37.

Page | 243

523. Chan AS, Cheung M-c, Han YM, Sze SL, Leung WW, Man HS et al. Executive

function deficits and neural discordance in children with autism spectrum disorders.

Clinical Neurophysiology 2009; 120(6): 1107-1115.

524. Chan AS, Han YM, Leung WW-m, Leung C, Wong VC, Cheung M-c. Abnormalities

in the anterior cingulate cortex associated with attentional and inhibitory control

deficits: A neurophysiological study on children with autism spectrum disorders.

Research in Autism Spectrum Disorders 2011; 5(1): 254-266.

525. Chan AS, Sze SL, Han YM. An intranasal herbal medicine improves executive

functions and activates the underlying neural network in children with autism.

Research in Autism Spectrum Disorders 2014; 8(6): 681-691.

526. Chantiluke K, Barrett N, Giampietro V, Brammer M, Simmons A, Rubia K. Disorder-

dissociated effects of fluoxetine on brain function of working memory in attention

deficit hyperactivity disorder and autism spectrum disorder. Psycological medicine

2015; 45(6): 1195-1205.

527. Chantiluke K, Barrett N, Giampietro V, Santosh P, Brammer M, Simmons A et al.

Inverse fluoxetine effects on inhibitory brain activation in non-comorbid boys with

ADHD and with ASD. Psychopharmacology 2015; 232(12): 2071-2082.

Page | 244

528. Chen S, Chien Y, Wu C, Shang C, Wu Y, Gau S. Deficits in executive functions

among youths with autism spectrum disorders: An age-stratified analysis.

Psychological Medicine 2016; 46(8): 1625-1638.

529. Corbett BA, Constantine LJ, Hendren R, Rocke D, Ozonoff S. Examining executive

functioning in children with autism spectrum disorder, attention deficit hyperactivity

disorder and typical development. Psychiatry Research 2009; 166(2-3): 210-222.

530. Costecu CA, Vanderborght B, David DO. Reversal learning task in children with

Autism Spectrum Disorder: A robot based approach. Journal of Autism and

Developmental Disorders 2015; 45(11): 3715-3725.

531. Davids RC, Groen Y, Berg IJ, Tucha OM, Balkom ID. Executive functions in older

adults with autism spectrum disorder: Objective performance and subjective

complaints. Journal of Autism and Developmental Disorders Jun 2016; (Pagination):

No Pagination Specified.

532. De Vries M, Geurts HM. Beyond individual differences: Are working memory and

inhibition informative specifiers within ASD? Journal of Neural Transmission 2014;

121(9): 1183-1198.

533. Dichter GS, Felder JN, Bofish JW. Autism is characterised by dorsal anterior

cingulate hyperactivation during social target detection. Social Cognirive and

Affective Neuroscience 2009; 4(3): 215-226.

Page | 245

534. Dichter GS, Lam KSL, Turner-Brown LM, Bodfish JW. Generativity abilities predict

communication deficits but not repetitive behaviors in autism spectrum disorders.

Journal of Autism & Developmental Disorders 2009; 39(9): 1298-1304.

535. Dichter GS, Radonovich KJ, Turner-Brown LM, Lam KS, Holtzclaw TN, Bodfish

JW. Performance of children with autism spectrum disorders on the dimension-

change card sort task. Journal of Autism and Developmental Disorders 2010; 40(4):

448-456.

536. Edgin JO, Pennington BF. Spatial cognition in autism spectrum disorders: superior,

impaired, or just intact? Journal of autism and developmental disorders 2005; 35(6):

729-745.

537. Endedijk H, Denessen E, Hendriks AW. Relationships between executive functioning

and homework difficulties in students with and without autism spectrum disorder: An

analysis of student- and parent-reports. Learning and Individual Differences 2011;

21(6): 765-770.

538. Faja S, Murias M, Beauchaine TP, Dawson G. Reward-based decision making and

electrodermal responding by young children with Autism Spectrum Disorders dring a

gambling task. Autism Research 2013; 6(6): 494-505.

539. Fitch A, Fein DA, Eigsti I-M. Detail and gestalt focus in individuals with optimal

outcomes from autism spectrum disorders. Journal of Autism & Developmental

Disorders 2015; 45(6): 1887-1896.

Page | 246

540. Garcia-Villamisar D, Sala SD. Dual-task perfromance in adults with autism.

Cognitive Neuropsychiatry 2002; 7(1): 63-74.

541. Geurts HM, Verte S, Oosterlaan J, Roeyers H, Sergeant JA. How specific are

executive functioning deficits in attention deficit hyperactivity disorder and autism?

Journal of Child Psychology and Psychiatry 2004; 45(4): 836-854.

542. Gioia GA, Isquith PK, Kenworthy L, Barton RM. Profiles of everyday executive

funcitons in acquired and developmental disorders. Child Neuropsychology 2002;

8(2): 121-137.

543. Goddard L, Dritschel B, Robinson S, Howlin P. Development of autobiographical

memory in children with autism spectrum disorders: Deficits, gains, and predictors of

performance. Development and Psychopathology 2014; 26(1): 215-228.

544. Goldstein G, Johnson CR, Minshew NJ. Attentional processes in autism. Journal of

Autism and Developmental Disorders 2001; 31(4): 433-440.

545. Gonzalez-Gadea ML, Baez S, Torralva T, Castellanos FX, Rattazzi A, Bein V et al.

Cognitive variability in adults with ADHD and AS: Disentangling the roles of

executive functions and social cognition. Research in Developmental Disabilities

2013; 34(2): 817-830.

Page | 247

546. Gonzalez-Gadea ML, Tripicchio P, Rattazzi A, Baez S, Marino J, Roca M et al. Inter-

individual cognitive variability in children with Asperger's syndrome. Frontiers in

Human Neuroscience Vol 8 Jul 2014, ArtID 575 2014; 8.

547. Gonzalez-Gadea ML, Chennu S, Bekinschtein TA, Rattazzi A, Beraudi A, Tripicchio

P et al. Predictive coding in autism spectrum disorder and attention deficit

hyperactivity disorder. Journal of Neurophysiology 2015; 114(5): 2625-2636.

548. Griebling J, Minshew NJ, Bodner K, Libove R, Bansal R, Konasale P et al.

Dorsolateral prefrontal cortex magnetic resonance imaging measurements and

cognitive performance in autism. Journal of Child Neurology 2010; 25(7): 856-863.

549. Haebig E, Kaushanskaya M, Ellis Weismer S. Lexical processing in school-age

children with autism spectrum disorder and children with specific language

impairment: the role of semantics. Journal of Autism and Developmental Disorders

2015; 45(12): 4109-4123.

550. Hanson LK, Atance CM. Brief report: Episodic foresight in autism spectrum disorder.

Journal of Autism and Developmental Disorders 2014; 44(3): 674-684.

551. Hill EL, Bird CM. Executive processes in Asperger syndrome: Patterns of

performance in a multiple case series. Neuropsychologia 2006; 44(14): 2822-2835.

Page | 248

552. Holdnack J, Goldstein G, Drozdick L. Social perception and WAIS-IV performance

in adolescents and adults diagnosed with asperger's syndrome and Autism.

Assessment 2011; 18(2): 192-200.

553. Hooper SR, Poon KK, Marcus L, Fine C. Neuropsychological characteristics of

school-age children with High-Functioning Autism: Performance on the NEPSY.

Child Neuropsychology 2006; 12(4-5): 299-305.

554. Ikeda Y, Okuzumi H, Kokubun M. Inhibitory control in children with intellectual

disabilities with and without autism spectrum disorders in animal size tests.

International Journal of Developmental Disabilities 2014; 60(2): 80-88.

555. Inokuchi E, Kamio Y. Qualitative analyses of verbal fluency in adolescents and young

adults with high-functioning autism spectrum disorder. Research in Autism Spectrum

Disorders 2013; 7(11): 1403-1410.

556. Irvine CA, Eigsti I-M, Fein DA. Uh, um, and autism: Filler disfluencies as pragmatic

markers in adolescents with optimal outcomes from autism spectrum disorder.

Journal of Autism and Developmental Disorders 2016; 46(3): 1061-1070.

557. Ishii-Takahashi A, Takizawa R, Nishimura Y, Kawakubo Y, Kuwabara H,

Matsubayashi J et al. Prefrontal activation during inhibitory control measured by

near-infrared spectroscopy for differentiating between autism spectrum disorders and

attention deficit hyperactivity disorder in adults. NeuroImage: Clinical 2014; 4: 53-

63.

Page | 249

558. Iwanami A, Okajima Y, Ota H, Tani M, Yamada T, Hashimoro R et al. Task

dependent prefrontal dysfunction in persons with Asperger's disorder investigated

with multi-channel near-infrared spectroscopy. Research in Autism Spectrum

Disorders 2011; 5(3): 1187-1193.

559. Joseph RM, McGrath LM, Tager-Flusberg H. Executive dysfunction and its relation

to language ability in verbal school-age children with autism. Developmental

Neuropsychology 2005; 27(3): 361-378.

560. Kado Y, Sanada S, Yanagihara M, Ogino T, Ohno S, Watanabe K et al. Executive

function in children with pervasive developmental disorder and attention-

deficit/hyperactivity disorder assessed by the Keio version of the Wisconsin Card

Sorting Test. Brain and Development 2012; 34(5): 354-359.

561. Kaland N, Smith L, Mortensen EL. Brief report: Cognitive flexibility and focused

attention in children and adolescents with Asperger syndrome or high-functioning

autism as measured on the computerized version of the Wisconsin card sorting test.

Journal of Autism and Developmental Disorders 2008; 38(6): 1161-1165.

562. Kalbfleisch ML, Loughan AR. Impact of IQ discrepancy on executive function in

high-functioning autism: Insight into twice exceptionality. Journal of Autism and

Developmental Disorders 2012; 42(3): 390-400.

Page | 250

563. Kaufmann L, Zotter S, Pixner S, Starke M, Haberlandt E, Steinmayr-Gensluckner M

et al. Brief report: CANTAB performance and brain structure in pediatric patients

with asperger syndrome. Journal of Autism and Developmental Disorders 2013;

43(6): 1483-1490.

564. Keary CJ, Minshew NJ, Bansal R, Goradia D, Fedorov S, Keshavan MS et al. Corpus

callosum volume and neurocognition in autism. Journal of Autism and Developmental

Disorders 2009; 39(6): 834-841.

565. Kimhi Y, Shoam-Kugelmas D, Agam Ben-Artzi G, Ben-Moshe I, Bauminger-Zviely

N. Theory of mind and executive function in preschoolers with typical development

versus intellectually able preschoolers with autism spectrum disorder. Journal of

Autism and Developmental Disorders 2014; 44(9): 2341-2354.

566. Kloosterman PH, Kelley EA, Parker JD, Craig WM. Executive functioning as a

predictor of peer victimization in adolescents with and without an Autism Spectrum

Disorder. Research in Autism Spectrum Disorders 2014; 8(3): 244-254.

567. Koolen S, Vissers CT, Egger JI, Verhoeven L. How stimulus and task complexity

affect monitoring in high-functioning adults with autism spectrum disorder. Journal

of Autism and Developmental Disorders 2014; 44(10): 2499-2513.

568. Kushki A, Drumm E, Pla Mobarak M, Tanel N, Dupuis A, Chau T et al. Investigating

the Autonomic Nervous System Response to Anxiety in Children with Autism

Spectrum Disorders. PLoS ONE 2013; 8(4).

Page | 251

569. Kushki A, Brian J, Dupuis A, Anagnostou E. Functional autonomic nervous system

profile in children with autism spectrum disorder. Molecular Autism 2014; 5(1).

570. Landa RJ, Goldberg MC. Language, social, and executive functions in high

functioning autism: A continuum of performance. Journal of Autism and

Developmental Disorders 2005; 35(5): 557-573.

571. Lee PS, Yerys BE, Della Rosa A, Foss-Feig J, Barnes KA, James JD et al. Functional

connectivity of the inferior frontal cortex changes with age in children with autism

spectrum disorders: A fcMRI study of response inhibition. Cerebral Cortex 2009;

19(8): 1787-1794.

572. Li J, Zhu L, Liu J, Li X. Social and non-social deficits in children with high-

functioning autism and their cooperative behaviors. Research in Autism Spectrum

Disorders 2014; 8(12): 1657-1671.

573. Low J, Goddard E, Melser J. Generativity and imagination in autism spectrum

disorder: Evidence from individual differences in children's impossible entity

drawings. British Journal of Developmental Psychology 2009; 27(2): 425-444.

574. Mackie MA, Fan J. Reduced efficiency and capacity of cognitive control in autism

spectrum disorder. Autism Research 2016; 9(3): 403-414.

Page | 252

575. Maister L, Simons JS, Plaisted-Grant K. Executive functions are employed to process

episodic and relational memories in children with autism spectrum disorders.

Neuropsychology 2013; 27(6): 615-627.

576. Mashal N, Kasirer A. Thinking maps enhance metaphoric competence in children

with autism and learning disabilities. Research in Developmental Disabilities 2011;

32(6): 2045-2054.

577. McCrimmon AW, Schwean V, Saklofske DH, Montgomery JM, Brady DI. Executive

functions in Asperger's syndrome: An empirical investigation of verbal and nonverbal

skills. Research in Autism Spectrum Disorders 2012; 6(1): 224-233.

578. McLean RL, Johnson Harrison A, Zimak E, Joseph RM, Morrow EM. Executive

function in probands with autism with average IQ and their unaffected first-degree

relatives. Journal of the American Academy of Child & Adolescent Psychiatry 2014;

53(9): 1001-1009.

579. McCrory E, Henry LA, Happe F. Eye-witness memory and suggestibility in children

with Asperger syndrome. Journal of Child Psychology and Psychiatry 2007; 48(5):

482-489.

580. Minshew NJ, Goldstein G, Siegel DJ. Neuropsychologic functioning in autism:

Profile of a complex information processing disorder. Journal of the International

Neuropsychological Society 1997; 3(4): 303-316.

Page | 253

581. Minshew NJ, Meyer J, Goldstein G. Abstract reasoning in autism: A dissociation

between concept formation and concept identification. Neuropsycology 2002; 16(3):

327-333.

582. Narzisi A, Muratori F, Calderoni S, Fabbro F, Urgesi C. Neuropsychological profile

in high functioning autism spectrum disorders. Journal of Autism and Developmental

Disorders 2013; 43(8): 1895-1909.

583. Oerlemans AM, Droste K, van Steijn DJ, de Sonneville LM, Buitelaar JK, Rommelse

NN. Co-segregation of social cognition, executive function and local processing style

in children with ASD, their siblings and normal controls. Journal of Autism and

Developmental Disorders 2013; 43(12): 2764-2778.

584. Oerlemans AM, Hartman CA, Franke B, Buitelaar JK, Rommelse NNJ. Does the

cognitive architecture of simplex and multiplex ASD families differ? Journal of

Autism and Developmental Disorders 2016; 46(2): 489-501.

585. O'Shea AG, Fein DA, Cillessen AH, Klin A, Schultz RT. Source memory in children

with autism spectrum disorders. Developmental Neuropsychology 2005; 27(3): 337-

360.

586. Ozonoff S, Jensen J. Specific executive function profiles in three neurodevelopmental

disorders. Journal of Autism and Developmental Disorders 1999; 29(2): 171-177.

Page | 254

587. Park S, Park J-E, Cho S-C, Kim B-N, Shin M-S, Kim J-W et al. No association of the

norepinephrine transporter gene (SLC6A2) and cognitive and behavioural phenotypes

of patients with autism spectrum disorder. European Archives of Psychiatry and

Clinical Neuroscience 2014; 264(6): 507-515.

588. Planche P, Lemonnier E. Children with high-functioning autism and Asperger's

syndrome: Can we differentiate their cognitive profiles? Research in Autism Spectrum

Disorders 2012; 6(2): 939-948.

589. Pooragha F, Kafi SM, Sotodeh SO. Comparing response inhibition and flexibility for

two components of executive functioning in children with autism spectrum disorder

and normal children. Iranian Journal of Pediatrics 2013; 23(3): 309-314.

590. Radulescu E, Minati L, Ganeshan B, Harrison NA, Gray MA, Beacher FD et al.

Abnormalities in fronto-striatal connectivity within language networks relate to

differences in grey-matter heterogeneity in Asperger syndrome. NeuroImage: Clinical

2013; 2(1): 716-726.

591. Rajendran G, Law AS, Logie RH, Van Der Meulen M, Fraser D, Corley M.

Investigating multitasking in high-functioning adolescents with autism spectrum

disorders using the virtual errands task. Journal of Autism and Developmental

Disorders 2011; 41(11): 1445-1454.

592. Richard AE, Lajiness-O'Neill R. Visual attention shifting in autism spectrum

disorders. J Clin Exp Neuropsychol 2015; 37(7): 671-687.

Page | 255

593. Rinehart NJ, Bradshaw JL, Moss SA, Brereton AV, Tonge BJ. Brief report: Inhibition

of return in young people with autism and Asperger's disorder. Autism 2008; 12(3):

249-260.

594. Robinson S, Goddard L, Dritschel B, Wisley M, Howlin P. Executive functions in

children with autism spectrum disorders. Brain and Cognition 2009; 71(3): 362-368.

595. Rommelse N, Langerak I, van der Meer J, de Bruijn Y, Staal W, Oerlemans A et al.

Intelligence may moderate the cognitive profile of patients with ASD. PLoS ONE Vol

10(10), Oct 2015, ArtID e0138698 2015; 10(10).

596. Rommelse N, Langerak I, van der Meer J, de Bruijn Y, Staal W, Oerlemans A et al.

Intelligence May Moderate the Cognitive Profile of Patients with ASD. PLoS ONE

[Electronic Resource] 2015; 10(10): e0138698.

597. Samyn V, Roeyers H, Bijttebier P, Rosseel Y, Wiersema JR. Assessing effortful

control in typical and atypical development: Are questionnaires and

neuropsychological measures interchangeable? A latent-variable analysis. Research in

Developmental Disabilities 2015; 36: 587-599.

598. Sanderson C, Allen ML. The specificity of inhibitory impairments in autism and their

relation to ADHD-type symptoms. Journal of Autism and Developmental Disorders

2013; 43(5): 1065-1079.

Page | 256

599. Schneider K, Regenbogen C, Pauly KD, Gossen A, Schneider DA, Mevissen L et al.

Evidence for gender-specific endophenotypes in high-functioning autism spectrum

disorder during empathy. Autism Research 2013; 6(6): 506-521.

600. Scott FJ, Baron-Cohen S. Imagining real and unreal things: Evidence of a dissociation

in autism. Journal of Cognitive Neuroscience 1996; 8(4): 371-382.

601. Semrud-Clikeman M, Walkowiak J, Wilkinson A, Butcher B. Executive functioning

in children with asperger syndrome, ADHD-combined type, ADHD-predominately

inattentive type, and controls. Journal of Autism and Developmental Disorders 2010;

40(8): 1017-1027.

602. Semrud-Clikeman M, Fine JG, Bledsoe J. Comparison among children with children

with autism spectrum disorder, nonverbal learning disorder and typically developing

children on measures of executive functioning. Journal of Autism and Developmental

Disorders 2014; 44(2): 331-342.

603. Semrud-Clikeman M, Fine JG, Bledsoe J. Social functioning using direct and indirect

measures with children with high functioning autism, nonverbal learning disability,

and typically developing children. Child Neuropsychology 2015; 22(3): 318-335.

604. Shu BC, Lung FW, Tien AY, Chen BC. Executive function deficits in non-retarded

autistic children. Autism 2001; 5(2): 165-174.

Page | 257

605. Sinzig J, Morsch D, Bruning N, Schmidt MH, Lehmkuhl G. Inhibition, flexibility,

working memory and planning in autism spectrum disorders with and without

comorbid ADHD-symptoms. Child and Adolescent Psychiatry and Mental Health

2008; 2(4).

606. Sinzig J, Vinzelberg I, Evers D, Lehmkuhl G. Executive function and attention

profiles in preschool and elementary school children with autism spectrum disorders

or ADHD. International Journal of Developmental Disabilities 2014; 60(3): 144-154.

607. Solomon M, Ozonoff SJ, Cummings N, Carter CS. Cognitive control in autism

spectrum disorders. International Journal of Developmental Neuroscience 2008;

26(2): 239-247.

608. Solomon M, Ozonoff SJ, Ursu S, Ravizza S, Cummings N, Ly S et al. The neural

substrates of cognitive control deficits in autism spectrum disorders.

Neuropsychologia 2009; 47(12): 2515-2526.

609. South M, Larson MJ, Krauskopf E, Clawson A. Error processing in high-functioning

autism spectrum disorders. Biol Psychol 2010; 85(2): 242-251.

610. Spek A, Schatorje T, Scholte E, van Berckelaer-Onnes I. Verbal fluency in adults

with high functioning autism or Asperger syndrome. Neuropsychologia 2009; 47(3):

652-656.

Page | 258

611. Sumiyoshi C, Kawakubo Y, Suga M, Sumiyoshi T, Kasai K. Impaired ability to

organize information in individuals with autism spectrum disorders and their siblings.

Neuroscience Research 2011; 69(3): 252-257.

612. Taddei S, Contena B. Brief report: Cognitive performance in autism and asperger's

syndrome: What are the differences? Journal of Autism and Developmental Disorders

2013; 43(12): 2977-2983.

613. Torralva T, Gleichgerrcht E, Roca M, Ibanez A, Marenco V, Rattazzi A et al.

Impaired theory of mind but intact decision-making in Asperger syndrome:

Implications for the relationship between these cognitive domains. Psychiatry

Research 2013; 205: 282-284.

614. Troyb E, Rosenthal M, Eigsti I-M, Kelley E, Tyson K, Orinstein A et al. Executive

functioning in individuals with a history of ASDs who have achieved optimal

outcomes. Child Neuropsychology 2014; 20(4): 378-397.

615. Tsuchiya E, Oki J, Yahara N, Fujieda K. Computerized version of the Wisconsin Card

Sorting Test in children with high-functioning autistic disorder or attention-

deficit/hyperactivity disorder. Brain & Development 2005; 27: 233-236.

616. Tye C, Asherson P, Ashwood K, Azadi B, Bolton P, McLoughlin G. Attention and

inhibition in children with ASD, ADHD and co-morbid ASD + ADHD: An event-

related potential study. Psychological medicine 2014; 44(5): 1101-1116.

Page | 259

617. Vaidya CJ, Foss-Feig J, Shook DA, Kaplan L, Kenworthy L, Gaillard WD.

Controlling attention to gaze and arrows in childhood: an fMRI study of typical

development and autism spectrum disorders. Developmental Science 2011; 14(4):

911-924.

618. Van Eylen L, Boets B, Steyaert J, Wagemans J, Noens I. Executive functioning in

autism spectrum disorders: influence of task and sample characteristics and relation to

symptom severity. European Child & Adolescent Psychiatry 2015; 24(11): 1399-

1417.

619. Vanegas SB, Davidson D. Investigating distinct and related contributions of Weak

Central Coherence, Executive Dysfunction, and Systemizing theories to the cognitive

profiles of children with Autism Spectrum Disordersand typically developing

children. Research in Autism Spectrum Disorders 2015; 11: 77-92.

620. Vanmarcke S, Van Der Hallen R, Evers K, Noens I, Steyaert J, Wagemans J. Ultra-

rapid categorization of meaningful real-life scenes in adults with and without ASD.

Journal of Autism and Developmental Disorders 2016; 46(2): 450-466.

621. Velazquez J, Barcelo F, Hung Y, Leshchenko Y, Nenadovic V, Belkas J et al.

Decreased brain coordinated activity in autism spectrum disorders during executive

tasks: Reduced long-range synchronization in the fronto-parietal networks.

International Journal of Psychophysiology 2009; 73(3): 341-349.

Page | 260

622. Verte S, Geurts HM, Roeyers H, Oosterlaan J, Sergeant JA. Executive functioning in

children with autism and Tourette syndrome. Development and Psychopathology

2005; 17(2): 415-445.

623. Voelbel GT, Bates ME, Buckman JF, Pandina G, Hendren RL. Caudate nucleus

volume and cognitive performance: are they related in childhood psychopathology?

Biological Psychiatry 2006; 60(9): 942-950.

624. Wallace GL, Silvers JA, Martin A, Kenworthy LE. Brief report: Further evidence for

inner speech deficits in autism spectrum disorders. Journal of Autism and

Developmental Disorders 2009; 39(12): 1735-1739.

625. Weismuller B, Thienel R, Youlden A-M, Fulham R, Koch M, Schall U.

Psychophysiological correlates of developmental changes in healthy and autistic boys.

Journal of Autism and Developmental Disorders 2015; 45(7): 2168-2175.

626. White SJ, Burgess PW, Hill EL. Impairments on “open-ended” executive function

tests in autism. Autism Research 2009; 2(3): 138-147.

627. Williams MA, Moss SA, Bradshaw JL, Rinehart NJ. Random number generation in

autism. Journal of Autism and Developmental Disorders 2002; 32(1): 43-47.

628. Williams DM, Bowler DM, Jarrold C. Inner speech is used to mediate short-term

memory, but not planning, among intellectually high-functioning adults with autism

spectrum disorder. Development and Psychopathology 2012; 24(1): 225-239.

Page | 261

629. Williams DL, Mazefsky CA, Walker JD, Minshew NJ, Goldstein G. Associations

between conceptual reasoning, problem solving, and adaptive ability in high-

functioning autism. Journal of Autism and Developmental Disorders 2014; 44(11):

2908-2920.

630. Wilson C, Happe F, Wheelwright SJ, Ecker C, Lombardo MV, Johnston P et al. The

neuropsychology of male adults with high-functioning Autism or Asperger syndrome.

Autism Research 2014; 7(5): 568-581.

631. Winsler A, Abar B, Feder MA, Schunn CD, Rubio DA. Private speech and executive

functioning among high-functioning children with autistic spectrum disorders.

Journal of Autism and Developmental Disorders 2007; 37(9): 1617-1635.

632. Easson A, Woodbury-Smith M. The role of prenatal immune activation in the

pathogenesis of autism and schizophrenia: A literature review. Research in Autism

Spectrum Disorders 2014; 8(3): 312-316.

633. Xiao T, Xiao Z, Ke X, Hong S, Yang H, Su Y et al. Response inhibition impairment

in high functioning autism and attention deficit hyperactivity disorder: evidence from

near-infrared spectroscopy data. PLoS ONE 2012; 7(10).

634. Yang J, Zhou S, Yao S, Su L, McWhinnie C. The relationship between theory of mind

and executive function in a sample of children from mainland China. Child psychiatry

and human development 2009; 40(2): 169-182.

Page | 262

635. Yeung MK, Han YM, Sze SL, Chan AS. Abnormal frontal theta oscillations underlie

the cognitive flexibility deficits in children with high-functioning autism spectrum

disorders. Neuropsychology 2016; 30(3): 281-295.

636. Yi L, Fan Y, Joseph L, Huang D, Wang X, Li J et al. Event-based prospective

memory in children with autism spectrum disorder: The role of executive function.

Research in Autism Spectrum Disorders 2014; 8(6): 654-660.

637. Yoran-Hegesh R, Kertzman S, Vishne T, Weizman A, Kotler M. Neuropsychological

mechanisms of Digit Symbol Substitution Test impairment in Asperger Disorder.

Psychiatry Research 2009; 166(1): 35-45.

638. Zhang L, Tang J, Dong Y, Ji Y, Tao R, Liang Z et al. Similarities and differences in

decision-making impairments between autism spectrum disorder and schizophrenia.

Frontiers in Behavioral Neuroscience 2015; 9(September).

639. Zinke K, Fries E, Altgassen M, Kirschbaum C, Dettenborn L, Kliegel M. Visuospatial

short-term memory explains deficits in tower task planning in high-functioning

children with autism spectrum disorder. Child Neuropsychology 2010; 16(3): 229-

241.

640. First MB, Gibbon M. The Structured Clinical Interview for DSM-IV Axis I Disorders

(SCID-I) and the Structured Clinical Interview for DSM-IV Axis II Disorders (SCID-

Page | 263

II). Comprehensive handbook of psychological assessment, Vol. 2: Personality

assessment. John Wiley & Sons Inc: Hoboken, NJ, US, 2004, pp 134-143.

641. Andresen NC, Flaum M, Arndt S, Alliger R, Swayze VW. Positive and negative

symptoms: Assessment and validity. Negative versus positive schizophrenia, 1992.

Page | 264

Appendix 1: Research Publications

Page | 265

Page | 266

Page | 267

Page | 268

Page | 269

Page | 270

Page | 271

Page | 272

Page | 273

Page | 274

Page | 275

Page | 276

Page | 277

Page | 278

Page | 279

Page | 280

Page | 281

Page | 282

Appendix 2: Supplementary Data, Chapter 2

Page | 283

Supplementary Materials eAppendix S2.1: Study protocol eTable S2.1: Key executive function domains and their relationship to

developmental stage and ASD phenotype eTable S2.2: Studies included in the meta-analysis eTable S2.3: Hedges’ g values of EF and moderator outcomes eTable S2.4: Clinical sensitivity / Cohen’s d overlap statistic eTable S2.5: MOOSE checklist eFigure S2.1: PRISMA flow chart

Page | 284

Autism Spectrum Disorders - A protocol for a meta-analysis of executive function

Demetriou, EA., Lampit,A., Quintana, D.S., Naismith, S.L., Song, Y.J.C., Pye, J.E., Hickie,

I.B., Guastella, A.J.

Page | 285

Abstract

Background: Executive function (EF) in Autism Spectrum Disorders (ASD) has been subject to considerable research due to its proposed role in explaining at least partially the ASD phenotype and in particular the stereotypic and repetitive patterns of behaviour and its potential aetiological contribution to the observed deficits associated with theory of mind. Although generally accepted that there are EF deficits in ASD generalisations have been difficult due to the variability between studies including the specific EF domains selected for investigation, methodological differences on sample selection including differences in the comparison groups used, heterogeneity between measures of EF and within task characteristics and different demand expectations placed on the participants.

Methods/Design: This study will be a meta-analysis of EF in ASD and will be designed following the PRISMA guidelines. The study will synthesise research on EF in ASD. A systematic literature review will be carried out in Medline, Embase and PsychINFO using a selection of relevant terms and will cover the period of January 1980 to June 2016. Eligible studies will be primary empirical studies published in peer reviewed journals in the English language and will compare ASD clinical group(s) with a neurotypical control group. One reviewer will perform eligibility assessment and data extraction which will be verified by a second reviewer on a sample of the data. The Quality Assessment Tool for Quantitative Studies

(Effective Public Health Practice Project, EPHPP) will be used to evaluate the methodological quality of the included studies. Data analysis will be performed on Comprehensive Meta- analysis (CMA) version 3 (Biostat, NJ) using the random-effects model.

Discussion: The meta-analysis will yield an overall effect size for EF as well as individual effect sizes for unitary EF constructs. Subgroup analyses will evaluate the mediating effects of moderators on EF and the clinical sensitivity of measures. Results will be discussed in the context of their contribution to the study of EF.

Page | 286

Introduction

Background

Autism Spectrum Disorders (ASD) are a group of neurodevelopmental conditions whose core symptoms include deficits in social communication and social interaction and restricted and repetitive patterns of behaviour269. Although genetic and neurobiological factors have provided a key focus in explaining the ASD phenotype, cognitive factors have also been postulated to contribute to the expression of the core behaviours in ASD and the role of executive functions

(EF) has been one of the key areas of study.

Within the ASD literature EF has been assessed in one or more of the EF constructs of fluency, cognitive flexibility, planning, response inhibition and working memory111 472-474although there is inconsistency between studies on the terminology used. Research findings have been equivocal and a number of factors may account for this including methodological differences on sample selection, type of outcome measures, age classification and presentation format of

EF measures.

Rationale:

The present study was undertaken to fill a gap in the existing literature by reviewing in a single study individual EF domains of interest as well as providing an evaluation of EF as a unitary construct. To our knowledge this is the first study within the framework of the PRISMA guidelines (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) which synthesises in a quantitative review the EF constructs of fluency, planning and working memory and reviews cognitive flexibility separately under the specific domains of concept formation/set shifting and mental flexibility/set switching. In addition, the planned moderator analysis extends reviews of moderators assessed to date. The study will also examine the clinical sensitivity of EF measures. Finally, the study adds to the existing literature by reviewing studies published up to June 2016.

Page | 287

Objectives

The aims of this meta-analysis are to synthesise the evidence of executive dysfunction in populations with ASD by examining evidence for overall impairment of EF as well as within individual domains of EF and further, to explore the influence of moderating variables that may mediate performance on EF tasks. A secondary aim is to review EF outcome measures which may be useful clinical markers in differentiating between ASD and typical populations.

Protocol and Registration

The design and methods used to inform this meta-analysis protocol comply with the Preferred

Reporting Items for Systematic Reviews and Meta-Analyses Protocols (PRISMA-P). The protocol has not been registered prior to this publication.

Methods

Eligibility criteria

To be considered for inclusion studies must be peer review articles published in the English language with no restriction on geographical location. Longitudinal and cross sectional study designs will be considered for inclusion. The experimental and comparison group samples will consist of individuals over the age of six with either a diagnosis of ASD or identified to have neurotypical development. Studies must have included outcome measures of one of more of the following executive function classifications of concept formation, cognitive flexibility, mental flexibility, fluency, set shifting, set switching, response inhibition, planning and working memory. There will be no restriction on the type of outcome measures. These may include one or more measures derived from psychometric tests, experimental tasks and/or self/informant questionnaires.

Information sources:

Information sources will be primarily based on Embase, Medline and PsychINFO databases as well as manual search of articles reported in review papers. The search period will be from

Page | 288

1980 to June 2016. The start date of 1980 is considered appropriate as this is the first time a diagnosis of Autism was included in the Diagnostic and Statistical Manual of Mental Disorders

(3rd edition)289

Search strategy:

The search strategy will include terms and keywords based on initial scoping of a sample of journals and authors’ expertise. Keywords will include diagnostic criteria of ASD (e.g.

“Autism” “Autism Spectrum Disorders”, “Asperger’s”), domains of EF (e.g. “cognitive flexibility”, “fluency”, “working memory,) and assessment measures (e.g. “Wisconsin Card

Sorting Test”, “Go-no Go task” “Stroop test”).

Study Records:

Data Management

Search output will be managed through EndNote, a bibliography specific software.

Selection process

One reviewer (EAD) will independently scan primary titles to select articles for further scrutiny, deleting any duplicate titles. Abstracts of potential eligible studies will then be read to determine eligibility for coding. When the title and abstract cannot be rejected, the full text of the article is obtained and reviewed for inclusion using a coding form. A structured data abstraction coding form will be designed to ensure consistency of appraisal for each study.

Inclusion or exclusion is then determined.

Data collection process

Two independent reviewers will complete the data extraction. If any inconsistencies are noted, the data extraction coding form will be reviewed by both reviewers to ensure consistency is met. If there is still disagreement a third reviewer will consider the data coding and make the final decision.

Page | 289

Data items

Group level summary data (e.g. sample size, means, standard deviations, F-values) will be extracted for all measures (i.e. psychometric tests, experimental tasks and self/informant questionnaires) reporting outcomes for executive function as well as for moderator variables.

Risk of bias in individual studies

Quality assessment

Quality review of the selected studies will be completed based on the Quality Assessment Tool for Quantitative Studies – Effective Public Health Practice Project475. Studies are reviewed on eight components (“selection bias”, “study design”, “confounders”, “blinding”, “data collection methods”, “withdrawals and drop outs”, “intervention strategy” and “analysis”) plus a global rating on study quality. Components specifically pertaining to intervention studies

(e.g. intervention integrity) will not be rated. The quality review will be completed by two independent assessors not involved in any other aspects of the study. Study quality will be analysed in a meta-regression.

Data

Data will be presented in table format and will include details of study (authors, title and year of publication), group level sample characteristics (age, gender and sample size) and group level outcome measures of EF. The search process and outcomes of inclusion/exclusion criteria will be summarised on a PRISMA flow chart.

Synthesis:

Data analysis will be performed on Comprehensive Meta-analysis (CMA) version 3 (Biostat,

NJ) using the random-effects model. The unit of analysis will be the standardised mean difference (SMD, calculated as Hedges’ g) on each measure between the ASD and neurotypical comparison group. When data on more than one comparison group is reported in the study, the comparison groups will be combined following established statistical procedures476. The same

Page | 290

procedure will be applied to combine more than one experimental group when all are compared to a single control group. A positive effect size will indicate that the control group performed better on the EF measure of interest compared to the ASD group. Analyses will be performed for an overall effect size of EF (combining all EF outcomes) as well as separate effect sizes for each individual EF domain of interest.

The data analysis is planned a priori and will be completed in three stages. The initial analysis will combine all EF outcomes to determine if there is an overall difference in EF between ASD and comparison groups. The second stage will examine differences in effect sizes between

ASD and control groups for each individual EF domain of interest. In the final stage, subgroup analyses will be conducted to identify any mediating effects by the predefined moderators.

Hedges’ g effect sizes ≤0.30, >0.30 and <0.50 and ≥0.50 will be deemed as small, moderate or large following the same convention applied to Cohen’s d effect sizes477. Heterogeneity across studies will be assessed using the I2 statistic with 95% confidence intervals. I2 values of 25%,

50% and 75% define small, moderate and large heterogeneity.

Meta-bias(es)

In order to examine risk of publication bias, funnel plots for overall outcomes as well as for each cognitive domain were inspected for asymmetry and formally assessed using Egger’s regression test with probability of significance set at p=0.1.

Discussion:

Study outcomes will be discussed in the context of the role of overall EF in ASD and the contribution of unitary EF constructs in differentiating within the spectrum and at different age classifications. The potential influence of moderating variables and the clinical utility of EF measures in discriminating between ASD and neurotypical populations will also be reviewed.

Page | 291

Update: there are no previous versions of this protocol

Registration: this protocol has not been registered

Support: The authors have not received any financial support for the preparation of this protocol and study

Contributions

Protocol Development: Eleni Andrea Demetriou

Amit Lampit

Sharon Naismith

Daniel Quintana

Adam Guastella

Literature Search: Eleni Andrea Demetriou

Identification relevant titles, abstracts and studies: Eleni Andrea Demetriou

Data extraction: Eleni Andrea Demetriou

Benedikt Langenbach

Review of studies: Jonathon Edward Pye

Quality appraisal: Magdalena Durrant

Karen Gould

Data analysis and/or interpretation: Eleni Andrea Demetriou

Amit Lampit

Adam Guastella

Sharon Naismith

Yun Ju Christine Song

Draft review: Eleni Andrea Demetriou

Amit Lampit

Page | 292

Daniel Quintana

Sharon Naismith

Yun Ju Christine Song

Ian Hickie

Adam Guastella

Page | 293 eTable S2.1: Key executive function domains and their relationship to developmental stage and ASD phenotype

EF domain Developmental stage and Developmental changes observed in ASD Associated impairment in the ASD

associated brain area(s) in phenotype

typical development

Global  Functional connectivity  Functional connectivity  Aberrant connectivity is noted to changes . Age related increase in . Age related decrease in white matter contribute to the core behavioural

481 white matter volume and volume and associated long range deficits in ASD

associated long range connectivity with over-connectivity

connectivity478 observed in children and under-

connectivity observed in adults478

. Children (age<11) display hyper-

connectivity within brain networks

and hypo-connectivity between

brain networks276

. Adolescents (age 11-18 have

comparable within network

connectivity but hypo-connectivity

Page | 294

for between brain networks

compared to neurotypical controls276

. Adults with ASD show no within or

between brain network

differences276

. Decreased fractional anisotropy

(FA) observed in adolescents with

ASD in tracts including the inferior

fronto-occipital fasciculus and the

inferior and superior longitudinal

fasciculi but no differences in FA

were observed between the adult

comparison groups479.

. Dysregulation in cingulum bundle

white matter development in late

adolescence and early adulthood in

ASD with lower FA associated with

Page | 295

overall executive dysfunction as

measured by the BRIEF480

Concept  Emerges in early childhood  Parent reported ratings of concept  Impaired ability in concept formation formation/set and matures in adolescence, 482 formation (based on the Shift subscale is associated with shifting of the BRIEF) showed no changes in restricted/repetitive behaviours &

The capacity  Functional peak observed in concept formation between younger and impairment in social communication to shift mid adolescence (17 years) older children (6-8 & 9-11) and and social interaction249 between followed by decline (18-19 adolescents (12-14 & 15-18)485 mental years)483  Deficits in concept formation/set processes to  Age related improvements in concept shifting are associated with restricted form new  Adult levels of set shifting formation in adolescents (12-16) and repetitive behaviours31, 156 concepts and observed in 8-10 year olds87 compared to children (8-11) with identify the but also in adolescence484 ASD162  Set shifting performance correlates conceptual

with development of social relationships  Recruitment of prefrontal  Children (7-14) displayed greater understanding486 shared by cortex (PFC) areas including

65 activation in mid-dorsal Anterior stimuli dorsolateral prefrontal cortex

Cingulate Cortex (ACC), superior

Page | 296

(DLPFC), parietal lobe and frontal gyrus (SFG), left medial frontal

cerebellum in childhood with gyrus (MFG), frontal pole and right

increased activation in left inferior frontal gyrus (IFG) compared to

posterior parietal cortex and controls378.

right middle frontal gyrus in

adulthood

Mental  Increased activation of left inferior and right  Impaired performance on task switching flexibility/set  Emerges in early childhood mesial parietal cortex during ‘switch’ task in predicted motor and sensory stereotypic switching and matures in adolescence482, 489 adults with ASD compared to controls behaviours490 The capacity 488 to switch between mental processes

(multiple tasks, operations or

Page | 297

mental sets) in response to changing demands53, 487

Fluency  Emerges in early childhood  Age related improvements in semantic  Fluency deficits are associated with

and matures in early and design fluency in adolescents (12- impairment in social communication122

The capacity adolescence,482, 488 16) compared to children (8-11) with to generate ASD162 novel ideas  Greatest period of

(ideational development in early to mid-  Attenuated recruitment of anterior PFC fluency) and childhood (5-8) with continued during a verbal fluency task in adults responses improvement into early with ASD compared to controls with no

(phonemic and adulthood491 comparable differences observed in semantic children, reflecting dysregulated fluency122. developmental trajectory of prefrontal

May be activity in ASD493

Page | 298

assessed by  Significantly higher activation verbal and of left inferior frontal cortex in non-verbal children compared to adults492 tasks.

Planning  Emerges and significantly  Parent reported behavioural problems  Impaired planning performance relates to

The capacity develops in early childhood, on planning (based on the BRIEF) were inability to adapt in social interactions to to execute a some research suggests brief greatest for older children (9-11) and the demands of the situation and plan in sequence of regression of skills in adolescents (12-14) compared to response to social demands495 actions so that adolescence, matures in early younger children with ASD (6-8)485 a desired goal adulthood482, 488  Impaired planning performance relates to is achieve50.  Age related improvements in planning restricted and repetitive behaviours496

 Significant improvement in in adolescents (12-16) compared to

late adolescence (15-19) with children (8-11) with ASD162

optimal performance in early

adulthood (20-29)87

Page | 299

 Significant longitudinal improvement in

 Greatest period of planning ability from younger age (5-6)

development in early to mid- to older age (8-9)494

childhood (5-8) with continued

improvement into early

adulthood491

 Activation of frontal parietal

and premotor networks in

adulthood483 with significant

role of DLPC

Page | 300

Response  Emerges in early childhood,  Impaired performance in response

Inhibition matures in late childhood to inhibition across development, optimal  Impaired performance on response

The capacity early adolescence488 ability matures in mid adolescence454 inhibition task(s) predicted motor and to inhibit a sensory stereotypic behaviours249, 490 previously  Greatest period of  Parent reported behavioural problems learned development in early to mid- on inhibition (based on the BRIEF)  Inhibitory control is inversely related to response53. childhood (5-8) with continued were greatest for younger children with moral competence in children with

improvement into early ASD (6-8) compared to older children ASD499

adolescence491 (9-11) and adolescents (12-14)485

 Adult levels of response  Age related improvements in response

inhibition achieved in late inhibition in adolescents (12-16)

childhood (age 11)484 compared to children (8-11) with

ASD162

 in children noted recruitment

of ACC, OFC, inferior and  Reduced fronto-cerebellar connectivity

middle gyri with higher in adolescents compared to children

Page | 301

volume of activation in in with ASD consistent with typical

DLPFC compared with development498

comparable recruitment of  Recruitment of reactive control

brain areas in adults but with processes (increased functional

higher activation of OFC492 connectivity between ventrolateral

prefrontal cortex -VLPFC and anterior

 Recruitment of frontal and cingulate cortex – ACC) contrasting

parietal areas and networks with recruitment of proactive control

incorporating thalamus and processes (DLPFC and parietal cortex)

cerebellum in adulthood497 in typically developing adolescents498

Page | 302

Working  Emerges in early childhood  WM improvement from childhood to  Impaired WM associated with low

Memory and matures in early early adolescence162, 454 with protracted performance on communication and

(WM) adolescence488, 492 development from adolescence to early socialisation domains as assessed by

The capacity adulthood454, 501 behavioural informant questionnaires504 to store and  Peak improvement in late manipulate adolescence (15-19)  Working memory performance  WM performance stable across information in maintained in early adulthood negatively correlated with restricted and adulthood and old age502 decline in WM temporary (20-29)87 repetitive behaviours249 in old age503 short term storage for  WM continued to develop into  Parent reported ratings on WM (based complex early adulthood484 on the BRIEF) showed no age related cognitive deficits across children (6-8 & 9-11) and 5 manipulations  Increased activation of left and adolescents (12-14 &15-18)485 0 . right PFC and left and right

posterior parietal cortex (PPC)  Lack of longitudinal improvement in from adolescence to WM in children/youth with ASD190 adulthood492

Page | 303

 Developmental relationship

between improved WM

performance and decreased

bilateral cortical thickness in

frontal and parietal regions500

Page | 304 eTable S2.2: Summary of studies included in meta-analysis

Study name & ASD Sample Control Hedges' g, p-value= Age Gender Diagnostic Domain(s)

year of size Sample (95% CI) Years (% Group

publication size (Mean) Male)

Adamo et al, 46 36 0.17 0.4554 10 no data ASD Response Inhibition

2014505 (-0.27-0.60) Combined

Adams & 24 24 0.15 0.6068 13.5 83 Autism Response Inhibition

Jarrold, 2009506 (-0.41-0.71) Diagnosis

Adams & 15 15 0.18 0.6221 14.5 Response inhibition

Jarrold, 2012171 (-0.53-0.88)

Alderson-Day, 21 21 0.76 0.0152 13.8 95 ASD Concept Formation

2011507 (0.15-1.38) Combined

Alderson-Day 14 14 0.82 0.0327 13.3 86 ASD Concept Formation

& McGonigle- (0.07-1.57) Combined

Chalmers,

2011508

Page | 305

Alderson-Day, 30 15 0.59 0.0612 13.0 93 ASD Concept Formation

2014509 (-0.03-1.21) Combined Fluency

Altgassen et al, 25 25 0.55 0.0518 21.8 80 ASD Mental Flexibility

2012510 (0.00-1.11) Combined Planning

Response Inhibition

Altgassen & 22 22 0.32 0.2903 25.8 91 ASD Response Inhibition

Koch, 2014185 (-0.27-0.90) Combined

Ambery et al, 27 20 0.42 0.1516 37.6 82 Asperger Concept Formation

2006511 (-0.15-1.00) Syndrome Fluency

Response Inhibition

Ames and 15 15 1.28 0.0011 14.2 no data Autism Response Inhibition

Jarrold 2007177 (0.51-2.05) diagnosis

Anbrosino et al, 19 19 0.43 0.1786 11.5 100 Autism Response Inhibition

2014178 (-0.20-1.06) Diagnosis

Andersen et al, 38 50 1.20 0.0000 12.0 82 ASD Working Memory

2013512 (0.74-1.65) Combined

Page | 306

Andersen et al 34 45 1.10 0.0000 11.6 83 Autism Response Inhibition

2015513 (0.62-1.57) Diagnosis

Baez et al, 15 15 0.38 0.2977 35.5 73 Asperger Concept Formation

2012514 (-0.33-1.09) Syndrome Fluency

Mental Flexibility

Response Inhibition

Working Memory

Barneveld et al, 29 40 0.57 0.0197 14.7 72 Autism Concept Formation

2013515 (0.09-1.06) Diagnosis Response Inhibition

Barron- 30 60 0.22 0.3263 9.1 93 Asperger Fluency

Linnankoski et (-0.22-0.67) Syndrome Response Inhibition

al, 2015516

Beacher et al, 29 32 0.36 0.1623 32.8 52 Asperger Fluency

2012517 (-0.14-0.86) Syndrome

Begeer et al, 26 26 -0.05 0.8575 13.8 88 Austism Fluency

2014518 (-0.58-0.49) Diagnosis

Page | 307

Biscaldi et al 28 33 0.95 0.0006 10.7 82 ASD Response Inhibition

2016519 (0.40-1.49) Combined Working Memory

Bishop & 14 18 0.75 0.0383 8.3 100 Autism Fluency

Norbury, (0.04-1.46) Diagnosis

2005122

Bodner et al 14 13 1.18 0.0037 18.9 71 Autism Response Inhibition

2012520 (0.38-1.98) Diagnosis Working Memory

Bolte et al, 21 35 0.39 0.1603 14.3 0 ASD Concept Formation

2011a161 (-0.15-0.93) Combined Planning

Mental Flexibility

Bolte et al, 35 23 0.48 0.0735 14.0 100 ASD Concept Formation

2011b161 (-0.05-1.01) Combined Planning

Mental Flexibility

Boucher, 7 7 0.22 0.6676 14.2 100 Autism Fluency

1988146 (-0.77-1.20) Diagnosis

Boyd et al, 60 64 2.63 0.00 10.2 93 ASD BRIEF Global EF

2009250 (2.15-3.12) Combined

Page | 308

Brady et al, 34 34 0.42 0.0821 18.9 77 Asperger Concept Formation

2013521 (-0.05-0.90) Syndrome

Bramham et al, 45 31 0.42 0.0742 32.8 85 ASD Fluency

2009496 (-0.04-0.88) Combined Planning

Brandimonte et 10 10 0.34 0.4352 8.3 70 Autism Response Inhibition

al, 2011179 (-0.51-1.18) Diagnosis

Brenner et al, 27 25 0.44 0.1104 12.7 85 Autism Working Memory

2015522 (-0.10-0.98) Diagnosis

Broadbent & 25 26 -0.41 0.1450 26.9 31 Asperger Concept Formation

Stokes, 2013a167 (-.95-0.14) Syndrome

Broadbent & 25 24 0.44 0.1190 26.9 31 Asperger Concept Formation

Stokes, 2013b167 (-0.11-1.00) Syndrome

Brunsdon et al, 149 155 0.37 0.0014 13.5 83 Autism Concept Formation

2015158 (0.14-0.60) Fluency

Response Inhibition

Bucaille et al, 16 16 1.14 0.0022 26.6 69 Asperger Working Memory

2016194 (0.41-1.88) Syndrome

Page | 309

Chan et al, 13 29 0.84 0.0150 10.8 85 Autism BRIEF Global EF

2009523 (0.16-1.52) Diagnosis Response Inhibition

Chan et al, 20 20 0.72 0.0247 10.8 95 ASD Response Inhibition

2011a524 (0.09-1.35) Combined

Chan et al, 16 19 0.70 0.0411 8.0 100 ASD Concept Formation

2011b176 (0.03-1.37) Combined Fluency

Response Inhibition

Chan et al, 15 15 0.33 0.3644 11.9 87 ASD Mental Flexibility

2014525 (-0.38-1.04) Combined Planning

Chantiluke et al, 17 22 -0.02 0.9390 15.2 100 Autism Working Memory

2015a526 (-0.65-0.60) Diagnosis

Chantiluke et al, 19 25 0.33 0.2767 14.7 100 Autism Response Inhibition

2015b527 (-0.26-0.92) Diagnosis

Chen et al 53 63 0.51 0.0072 10.0 91 Autism Concept Formation

2016a528 (0.14-0.88) Diagnosis Planning

Working Memory

Page | 310

Chen et al 58 51 0.33 0.0922 14.7 98 Autism Concept Formation

2016b528 (-0.05-0.70) Diagnosis Planning

Working Memory

Christ et al, 48 15 1.00 0.0011 8.2 89 Autism Response Inhibition

2007170 (0.40-1.59) Diagnosis

Christ et al, 28 49 0.46 0.0543 13.1 96 Autism Response Inhibition

2011172 (-0.01-0.92) Diagnosis

Corbett et al, 18 18 1.00 0.0045 9.4 67 ASD Concept Formation

2009529 (0.31-1.69) Combined Mental Flexibility

Planning

Response Inhibition

Working Memory

Costecu et al, 41 40 0.05 0.8148 8.4 no data Autism Mental Flexibility

2015530 (-0.38-0.49) Diagnosis

Cox et al 201699 13 26 0.09 0.8007 18.3 100 ASD Working Memory

(-0.58-0.75) Combined

Page | 311

Crane et al, 28 28 0.66 0.0152 41.6 50 ASD Working Memory

2013195 (0.13-1.19) Combined

Cui et al, 12 29 0.04 0.9091 7.5 92 Asperger Working Memory

2010196 (-0.62-0.70) Syndrome

Czermainski et 11 19 0.88 0.0232 11.7 82 ASD Fluency

al, 2014495 (0.12-1.64) Combined Mental Flexibility

Response Inhibition

Working Memory

Davids et al 36 36 0.47 0.0495 58.6 83 Autism BRIEF

2016531 (0.00-0.94) Diagnosis Fluency

Planning

de Vries & 77 45 0.13 0.49 10.7 87 Autism Working Memory

Geurts, 2014532 (-0.24-0.50) Diagnosis

de Vries & 120 76 3.30 0.00 10.2 no data Autism BRIEF Global EF

Geurts, 2015260 (2.87-3.74) Diagnosis

Dichter & 14 15 0.79 0.0383 22.9 94 ASD Concept Formation

Belger, 2007165 (0.04-1.53) Combined

Page | 312

Dichter et al 15 19 0.82 0.0204 23.3 93 Autism Concept Formation

2009a533 (0.13-1.50) Diagnosis

Dichter et al 39 39 0.20 0.3717 9.7 97 Autism Fluency

2009b534 (-0.24-0.65) Diagnosis

Dichter et al, 32 34 0.49 0.0482 10.0 97 Autism Concept Formation

2010535 (0.00-0.97) Diagnosis

Durrleman & 17 17 -0.21 0.5372 9.2 no data Autism Concept Formation

Franck, 2015153 (-0.87-0.45) Diagnosis

Edgin & 24 34 -0.25 0.3435 11.5 Autism Concept Formation

Pennington (-0.77-0.27) 47 Diagnosis Working Memory

2005536

Endedijk et al, 68 84 0.64 0.0001 13.9 87 ASD BRIEF Global EF

2011537 (0.32-0.97) Combined

Englund 2014189 33 79 1.28 0.0000 10.8 73 ASD Working Memory

(0.84-1.72) Combined

Faja et al, 21 21 1.55 0.0000 6.8 71 Autism BRIEF Global EF

2013538 (0.87-2.23) diagnosis

Page | 313

Fitch et al 20 17 0.57 0.0828 12.9 no data Autism Planning

2015539 (-0.07-1.22) diagnosis Response Inhibition

Floris et al 40 40 0.71 0.0127 14.5 100 Autism Concept Formation

2013159 (0.15-1.27) diagnosis

Garcia- 16 16 1.54 0.0001 23.5 50 Autism Concept Formation

Villamisar & (0.76-2.33) diagnosis Fluency

Sala, 2002540 Working Memory

Geurts et al, 41 41 0.69 0.0024 9.4 100 Autism Concept Formation

2004541 (0.24-1.13) Diagnosis Fluency

Planning

Response Inhibition

Working Memory

Geurts & 23 23 0.32 0.2825 63.6 78 ASD Concept Formation

Vissers 2012503 (-0.26-0.89) Combined Fluency

Mental Flexibility

Planning

Page | 314

Geurts & De 24 24 0.41 0.1526 10.5 83 Autism Working Memory

Wit 2014198 (-0.15— Diagnosis

0.97)

Gioia et al, 54 208 1.46 0.0000 10.8 85.4 Autism BRIEF Global EF

2002542 (1.13-1.78) Diagnosis

Goddard et al, 63 63 0.39 0.0289 12.6 81 ASD Concept Formation

2014543 (0.04-0.74) Combined Fluency

Planning

Response Inhibition

Goldberg et al, 17 32 0.33 0.2729 10.3 76 Autism Concept Formation

2005157 (-0.26-0.91) Diagnosis Planning

Response Inhibition

Working Memory

Goldberg et al, 11 15 1.69 0.0002 10.4 73 Autism Response Inhibition

2011180 (0.81-2.58) Diagnosis

Page | 315

Goldstein et al 103 103 0.39 0.0052 18.2 86 Autism Concept Formation

2001544 (0.12-0.67) Diagnosis Mental Flexibility

Response Inhibition

Gonzalez- 23 21 0.19 0.5322 33.0 63 Asperger Concept Formation

Gadea et al, (-0.40-0.77) Syndrome Mental Flexibility

2013545 Working Memory

Gonzalez- 19 19 0.30 0.3533 11.9 95 ASD Mental Flexibility

Gadea et l, (-0.33-0.92) Combined Response Inhibition

2014546 Working Memory

Gonzalez- 24 19 0.46 0.1326 10.4 96 Autism Fluency

Gadea et al, (-1.14-1.06) Diagnosis Mental Flexibility

2015547 Working Memory

Griebling et al, 24 38 0.70 0.0070 17.9 92 Autism Concept Formation

2010548 (0.19-1.20) Diagnosis Planning

Haebig et al 27 30 0.32 0.2306 9.5 85 Autism Concept Formation

2015549 (-0.20-0.83) Diagnosis Working Memory

Page | 316

Hanson & 25 25 0.25 0.3796 6.0 88 ASD Concept Formation

Atance 2014550 (-0.30-0.80) Combined Fluency

Planning

Response Inhibition

Working Memory

Happe et al, 13 16 0.59 0.1353 9.2 100 ASD Concept Formation

2006(a)162 (-0.18-1.37) Combined Fluency

Planning

Response Inhibition

Happe 14 10 0.03 0.9351 13.2 100 ASD Concept Formation

2006(b)162 (-0.71-0.77) Combined Fluency

Planning

Response Inhibition

Hill & Bird 22 22 0.46 0.1272 31.1 72 Asperger Concept Formation Fluency 2006551 (-0.13-1.05) Syndrome Mental Flexibility Planning Response /Inhibition

Page | 317

Holdnack et al, 43 43 0.74 0.0009 22.0 85 ASD Working Memory

2011552 (0.30-1.17) Combined

Hooper et al, 23 23 1.21 0.0009 9.6 83 Autism Planning

2006553 (0.59-1.83) Diagnosis

Ikeda et al, 9 21 -0.61 0.1245 15.4 67 Autism Response Inhibition

2014554 (-1.39-0.17) Diagnosis

Ikuta et al, 21 21 5.60 0.0000 18.1 86 ASD BRIEF Global EF

2014480 (4.25-6.94) Combined

Inokuchi & 30 18 0.46 0.1253 19.2 83 ASD Fluency

Kamio, 2013555 (-0.13-1.04) Combined

Irvine et al 24 16 0.25 0.4324 12.8 88 Autism BRIEF

2016556 (-0.37-0.87) Diagnosis Fluency

Ishii-Takahashi 21 21 0.23 0.4561 30.8 38 ASD Fluency

et al, 2014557 (-0.37-0.83) Combined Response Inhibition

Iwanami et al, 20 18 0.86 0.0100 27.2 70 Asperger Fluency

2011558 (0.21-1.51) Syndrome

Page | 318

Johnson et al 21 18 0.65 0.0100 10.5 95 Autism Response Inhibition

2007173 (0.01-1.29) Diagnosis

Johnston et al, 24 14 0.49 0.1432 27.8 79 Autism Response Inhibition

2011175 (-0.17-1.14) Diagnosis

Joseph et al, 37 31 0.46 0.0621 7.9 86 ASD Planning

2005(a)559 (-0.02-0.93) Combined Response Inhibition

Working Memory

Kado et al, 52 52 0.52 0.0088 9.9 75 ASD Concept Formation

2012560 (0.13-0.91) Combined

Kaland et al, 13 13 0.50 0.1931 16.4 100 ASD Concept Formation

2008561 (-0.26-1.26) Combined

Kalbfleisch & 19 21 1.90 .0000 12.5 84 Autism BRIEF

Loughan (1.16-2.65) Diagnosis

2011562

Kaufmann et al, 10 10 0.07 0.8687 14.7 80 Asperger Concept Formation

2013563 (-0.78-0.92) Syndrome Planning

Working Memory

Page | 319

Kawakubo et al, 14 14 0.20 0.5933 12.8 86 ASD Fluency

2009a493 (-0.52-0.92) Combined

Kawakubo et al, 13 13 0.39 0.3087 26.8 69 ASD Fluency

2009b493 (-0.36-1.14) Combined

Keary et al, 32 34 0.86 0.0008 19.8 94 Autism Concept Formation

2009564 (0.36-1.36) Diagnosis Planning

Kimhi et al, 29 30 0.53 0.0410 5.0 86 ASD Planning

2014565 (0.02-1.05) Combined

Kleinhans et al, 14 14 1.68 0.0001 23.8 100 ASD Fluency

2008203 (0.836-2.53) Combined

Kloosterman et 30 40 1.95 0.0000 14.9 100 ASD BRIEF Global EF

al, 2014566 (1.38-2.52) Combined

Koolen et al, 15 15 -0.06 0.8591 37.5 93 ASD Fluency

2014567 (-0.76-0.64) Combined Mental Flexibility

Response Inhibition

Working Memory

Page | 320

Koshino et al, 11 11 -0.48 0.2527 24.5 100 Autism Working Memory

200890 (-1.31-0.35) Diagnosis

Kretschmer et 21 21 0.41 0.1775 10.2 86 Autism Response Inhibition

al, 2014499 (-0.19-1.01) Diagnosis Working Memory

Kuijper et al, 43 38 0.40 0.0709 9.3 87 ASD Working Memory

2015181 (-0.03-0.84) Combined

Kushki et al, 12 17 5.95 0.0000 11.3 83 ASD Response Inhibition

2013568 (4.26-7.64) Combined

Kushki et al, 40 34 0.67 0.0048 12.0 83 Autism Response Inhibition

2014569 (0.20-1.13) Diagnosis Working Memory

Lai et al, 32 32 0.47 0.0655 27.0 100 ASD Fluency

2012a224 (-0.03-0.96) Combined Response Inhibition

Lai et al, 32 32 0.51 0.0441 28.1 0 ASD Fluency

2012b224 (0.01-1.00) Combined Response Inhibition

Lam 2013147 16 16 1.11 0.0032 8.9 75 Autism Concept Formation

(0.37-1.85) Diagnosis

Page | 321

Landa & 19 19 0.38 0.2530 11.0 no data Autism Concept Formation

Goldberg, (-0.27-1.04) Diagnosis Planning

2005570 Working Memory

Langen et al, 21 22 1.09 0.0007 25.5 100 Autism Response Inhibition

2012174 (0.46-1.73) Diagnosis

Larson et al 28 36 -0.07 0.7914 13.0 93 Autism Response Inhibition

201283 Diagnosis

(-0.55-0.42)

Lee et al, 11 10 -0.15 0.7321 10.2 75 Autism Response Inhibition

2009571 (-0.98-0.69) Diagnosis

Lemon et al, 13 14 0.66 0.0851 11.0 0 ASD Response Inhibition

2011207 (-0.09-1.41) Combined

Leung et al, 70 71 2.28 0.0000 11.1 87 Autism BRIEF Global EF

2016258 1.86-2.71) Diagnosis

Li et al, 2014572 37 31 0.63 0.0104 9.6 79 ASD Concept Formation

(0.15-1.12) Combined

Page | 322

Lopez et al, 17 17 0.62 0.0729 29.1 83 Autism Concept Formation

2005249 (-0.06-1.29) Diagnosis Fluency

Response Inhibition

Low et al, 27 27 0.54 0.0465 8.3 85 ASD Fluency

2009573 (0.01-1.08) Combined

Mackie & Fan 15 15 0.67 0.0692 26.0 100 Autism Working Memory

2016574 (-0.05-1.40) Diagnosis

Mackinlay et al, 14 16 0.91 0.0159 11.9 100 ASD Mental Flexibility

2006213 (0.17-1.65) Combined Planning

Maes et al 17 19 -0.05 0.8829 38.4 76 ASD Concept Formation

2011166 (-0.69-0.59) Combined

Maister et al, 14 14 0.26 0.4873 12.2 100 Autism Concept Formation

2013575 (-0.47-0.98) Diagnosis Fluency

Response Inhibition

Mashal & 20 20 1.91 0.0000 13.0 90 Autism Fluency

Kasirer 2011576 (1.17-2.65) Diagnosis

Page | 323

Matsuura et al, 11 19 0.29 0.4386 12.0 100 ASD Working Memory

2014197 (-0.44-1.02) Combined

McCrimmon et 33 33 0.30 0.2217 18.8 79 Asperger Concept Formation

al, 2012577 (-0.18-0.79) Syndrome Fluency

Mental Flexibility

Planning

Response Inhibition

McLean et al, 65 76 1.34 0.0000 9.4 no data ASD BRIEF Global EF

2014a578 (0.97-1.70) Combined

McLean et al, 35 57 0.69 0.0016 9.4 no data ASD Concept Formation

2014c578 (0.26-1.12) Combined Response inhibition

McCrory et al, 23 27 0.42 0.1434 13.0 92 Asperger Fluency

2007579 (-0.14-0.98) Syndrome Response Inhibition

Miller et al, 60 55 0.21 0.2725 15.1 83 Autism Concept Formation

2015204 (-0.17-0.60) Diagnosis

Page | 324

Minshew et al, 33 33 0.32 0.1885 20.9 88 Autism Concept Formation

1997580 (-0.16-0.80) Diagnosis Mental Flexibility

Minshew et al 90 107 0.64 0.0000 21.4 no data Autism Concept Formation

2002581 (0.35-0.93) Diagnosis Mental Flexibility

Montgomery et 25 0.42 0.0392 18.2 80 Asperger Mental Flexibility

al, 2013214 (0.02-0.82) Syndrome Response Inhibition

Morsanyi & 23 49 -0.07 0.7710 13.5 83 Autism Working Memory

Holyoak,2010182 (-0.56-0.42) Diagnosis

Nakahachi et al, 16 28 0.12 0.6890 28 75 ASD Working Memory

2006211 (-0.48-0.73) Combined

Narzisi et al, 22 44 1.16 0.0000 9.8 100 ASD Concept Formation

2013582 (0.62-1.71) Combined Fluency

Planning

Response Inhibition

Nyden et al, 37 13 0.44 0.1733 13.3 76 ASD Mental Flexibility

2011215 (-0.19-1.06) Combined Planning

Page | 325

Nyden et al 10 10 0.41 0.3439 10.1 100 Asperger Concept Formation

1999154 (-0.44-1.27) Syndrome

Oerlemans et al, 140 127 0.12 0.3395 12.4 82 ASD Response Inhibition

2013583 (-0.12-0.36) Combined Working Memory

Oerlemans et al 54 77 0.27 0.1254 12.3 85 ASD Concept Formation

2016a584 (-0.08-0.62) Combined Response Inhibition

Working Memory

Oerlemans et al 91 46 0.27 0.1339 11.6 71.4 ASD Concept Formation

2016b584 (-0.08-0.63) Combined Response Inhibition

Working Memory

O’Shea 2005585 21 21 0.66 0.0349 10.9 81 ASD Planning

(0.05-1.27) Combined

Ozonoff et al, 23 20 0.87 0.0066 12.1 91 ASD Concept Formation

1991246 (0.24-1.50) Combined Planning

Ozonoff 10 11 -0.20 0.6265 11.9 90 ASD Concept Formation

1995a92 (-1.03-0.62) Combined

Page | 326

Ozonoff 24 24 1.19 0.0001 12.0 96 ASD Concept Formation

1995b92 (0.59-1.80) Combined

Ozonoff & 40 29 0.63 0.0112 12.6 no data Autism Concept Formation

Jensen, 1999586 (0.14-1.11) Diagnosis Planning

Response Inhibition

Ozonoff et al, 79 70 0.47 0.00 15.7 91 Autism Concept Formation

200491 Diagnosis Planning

Park et al, 75 77 0.34 0.0361 7.9 86 ASD Concept Formation

2014587 (0.02-0.66) Combined Response Inhibition

Parsons et al 8 10 1.28 0.0106 22.3 no data Autism Response Inhibition

201698 (0.30-2.25) Diagnosis

Pascualvaca et 23 23 0.40 0.1890 8.7 65 Autism Concept Formation

al, 1998148 (-0.20-1.00) Diagnosis

Planche & 30 15 0.39 0.2149 8.0 83 ASD Planning

Lemmonier, (-0.23-1.01) Combined Working Memory

2012588

Page | 327

Pooragha et al, 15 15 1.10 0.0052 9.3 93 ASD Concept Formation

2013589 (0.33-1.86) Combined Response Inhibition

Prior & 12 12 0.72 0.0767 13.8 75 Autism Concept Formation

Hoffman, (-0.08-1.53) Diagnosis

1990139

Radulescu et al, 22 26 0.37 0.1939 32.5 50 Asperger Fluency

2013590 (-0.19-0.94) Syndrome

Rajendran et al, 18 18 1.49 0.00 13.9 89 Autism Concept Formation

2011591 (0.76-2.21) Diagnosis Planning

Raymakers et al 39 29 -0.52 0.0351 11.3 85 Aspergers Response Inhibition

2006 (-1.00-(-

0.04)

Richard & 20 20 0.76 0.0181 10.5 50 Autism Mental Flexibility

Lajiness- (0.13-1.39) Diagnosis

O’Neill 2015592

Rinehart et al, 12 12 0.43 0.2925 10.6 92 Autism Response Inhibition

2008(a)593 (-0.37-1.22) Diagnosis

Page | 328

Rinehart et al, 12 12 0.32 0.4159 13.4 83 Asperger Response Inhibition

2008(b)593 (-0.46-1.10) Syndrome

Robinson et al, 54 54 0.46 0.0188 12.5 78 ASD Concept Formation

2009594 (0.08-0.84) Combined Fluency

Planning

Response Inhibition

Rommelse et al 27 22 0.20 0.4858 12.1 57 ASD Mental Flexibility

2015a595 (-0.36-0.75) Combined Response Inhibition

Working Memory

Rommelse et al 56 53 0.15 0.4398 12.2 65 ASD Mental Flexibility

2015b595 (-0.23-0.52) Combined Response Inhibition

Working Memory

Rommelse et al 40 70 0.15 0.4533 11.4 85 ASD Mental Flexibility

2015c596 (-0.24-0.54) Combined Response Inhibition

Working Memory

Page | 329

Rumsey 1985163 9 10 1.12 0.0204 27.0 100 Autism Concept Formation

(0.17-2.07) Diagnosis

Rumsey & 10 25 2.21 0.0000 29.0 100 Autism Concept Formation

Hamburger (1.31-3.10) Diagnosis

1990142

Russell et al 22 22 0.93 0.0035 12.5 no data Autism Working Memory

1996191 (0.31-1.56) Diagnosis

Russell 1999183 19 19 0.68 0.0378 13.8 no data Autism Response Inhibition

(0.04-1.32) Diagnosis

Sachse et al, 30 28 0.54 0.0410 19.2 90 ASD Concept Formation

201389 (0.02-1.06) Combined Planning

Response Inhibition

Working Memory

Samyn et al, 31 148 0.27 0.1781 12.8 100 Autism Response Inhibition

2015597 (-0.12-0.65) Diagnosis

Page | 330

Sanderson & 31 28 -0.07 0.7806 13.7 84 Autism Response Inhibition

Allen 2013598 (-0.59-0.44) Diagnosis

Sawa et al, 19 19 0.89 0.0074 13.2 90 ASD Concept Formation

2013160 (0.24-1.55) Combined

Schmitz et al, 10 12 0.21 0.6191 38.0 100 ASD Response Inhibition

2006489 (-0.61-1.02) Combined Mental Flexibility

Schneider & 15 28 0.23 0.4695 10.7 93 Autism Concept Formation

Asarnow (-0.39-0.85) Diagnosis

1987155

Schneider et al 19 26 0.05 0.8679 31 Fluency

2013a599 (-0.53-0.63) Mental Flexibility

Working Memory

Schneider et al 14 12 0.17 0.6587 34 Fluency

2013b599 (-0.58-0.92) Mental Flexibility

Working Memory

Page | 331

Schneider et al 7 12 0.01 0.9792 30 Fluency

2013c (-0.93-0.96) Mental Flexibility

Working Memory

Scott & Baron- 15 15 67.05 0.0000 13.0 No data Autism Fluency

Cohen, 1996600 (45.00- Diagnosis

89.09)

Semrud- 15 32 (1.17) 0.0002 10.6 53 Autism BRIEF Global EF

Clikeman et al, 1.28 Diagnosis Fluency

2010601 (0.61-1.94) Planning

Response Inhibition

Semrud- 37 40 0.62 0.0080 12.8 74 Autism Concept Formation

Clikeman et al, (0.16-1.07) Diagnosis Mental Flexibility

2014602

Semrud- 36 36 1.61 0.0000 12.1 no data Autism BRIEF Global EF

Clikerman et al, (1.02-2.26) Diagnosis

2015603

Page | 332

Shafritz et al, 15 15 0.34 0.3445 18.1 80 ASD Response Inhibition

2015186 (-0.36-1.04) Combined

Shu et al, 26 52 1.35 0.0000 no data no data Autism Concept Formation

2001604 (0.82-1.88) diagnosis

Silk et al, 7 9 0.13 0.7916 14.7 100 ASD Working Memory

200684 (-0.81-1.06) Combined

Sinzig et al, 20 20 0.14 0.6584 14.3 80 ASD Concept Formation

2008605 (-0.47-0.75) Combined Planning

Response Inhibition

Working Memory

Sinzig et al, 26 29 0.45 0.0969 6.7 100 ASD Response Inhibition

2014606 (-0.08-0.98) Combined

Solomon et al, 31 32 0.26 0.3037 12.3 91 ASD Response Inhibition

2008607 (-0.23-0.75) Combined

Solomon et al, 20 23 0.80 0.0102 9.1 77 ASD Response Inhibition

2009608 (0.19-1.42) Combined

Page | 333

South et al, 24 21 0.53 0.0785 14.0 92 Autism Response Inhibition

2010609 (-0.06-1.11) Diagnosis

Spek et al, 62 30 0.54 0.0171 39.7 92 ASD Fluency

2009610 (0.10-0.97) Combined

Steele et al, 29 29 0.73 0.0068 14.8 100 Autism Working Memory

2007193 (0.20-1.25) Diagnosis

Sumiyoshi et al, 22 15 0.89 0.0097 26.5 86 ASD Concept Formation

2011611 (0.22-1.56) Combined

Taddei & 38 15 3.06 0.0000 13.0 79 ASD Planning

Contena, (2.22-3.91) Combined Response Inhibition

2013612

Torralva et al, 25 25 0.45 0.1082 33.9 72 Asperger Concept Formation

2013613 (-0.10-1.01) Syndrome Fluency

Mental Flexibility

Working Memory

Page | 334

Troyb et al, 38 32 0.93 0.0003 13.9 91 Autism BRIEF Global EF

2014614 (0.43-1.42) Diagnosis Mental Flexibility

Planning

Response Inhibition

Working Memory

Tsuchiya et al, 17 25 1.13 0.0007 12.5 94 Autism Concept Formation

2005615 (0.48-1.79) Diagnosis

Tye et al, 19 26 0.19 0.5255 11.7 100 Autism Response Inhibition

2014616 (-0.39-0.77) Diagnosis

Urbain et al 17 20 0.27 0.4051 11.2 65 Autism Working Memory

2016199 (-0.37-0.91) Diagnosis

Vaidya et al 18 18 0.61 0.0816 10.8 73 Autism Response Inhibition

2011617 (-0.08-1.29) Diagnosis van Eylen et al, 40 40 0.39 0.0802 11.3 90 Autism Concept Formation

2011152 (-0.05-0.83) Diagnosis

Page | 335

van Eylen et al, 50 50 0.65 0.0018 12.2 60 Autism BRIEF

2015618 (0.24-1.05) Diagnosis Concept Formation

Fluency

Mental Flexibility

Planning

Response Inhibition

Working Memory

Vanegas & 24 25 0.46 0.1290 9.7 no data ASD BRIEF

Davidson (-0.13-1.05) Combined Concept Formation

2015619

Vanmarcke et al 24 24 0.99 0.1290 20.6 79 Autism BRIEF

2016620 (0.40-1.59) Diagnosis

Vara et al, 15 15 0.46 0.2033 15.5 80 Autism Response Inhibition

2014184 (-0.25-1.17) Diagnosis

Velazquez et al, 15 16 0.82 0.0255 10.8 94 Asperger Concept Formation

2009621 (0.10-1.53) Syndrome Response Inhibition

Page | 336

Verte et al, 60 24 0.46 0.0555 9.1 93 Autism Concept Formation

2005622 (-0.01-0.94) Diagnosis Fluency

Mental Flexibility

Planning

Response Inhibition

Working Memory

Verte et al, 87 47 0.64 0.0005 8.6 91 ASD Concept Formation

2006a212 (0.28-1.00) Combined Fluency

Planning

Response Inhibition

Working Memory

Verte et al, 66 82 0.77 0.0000 8.7 92 ASD Response Inhibition

2006b212 (0.43-1.10) Combined Working Memory

Voelbel et al, 38 13 0.58 0.0731 10.2 100 ASD Concept Formation

2006623 (-0.05-1.21) Combined Fluency

Response Inhibition

Page | 337

Vogan et al, 19 17 0.74 0.0279 11.1 84 Autism Working Memory

2015192 (0.08-1.41) Diagnosis

Wallace et al, 28 25 0.62 0.0258 15.7 96 ASD Planning

2009624 (0.07-1.16) Combined

Weismuller et 15 12 1.03 0.0105 9.4 100 Autism Fluency

al, 2015625 (0.24-1.82) Diagnosis Planning

White et al, 45 27 0.51 0.0381 9.6 91 ASD Planning

2009626 (0.03-0.99) Combined Response Inhibition

Williams & 21 21 0.55 0.0759 10.5 no data ASD Concept Formation

Jarrold 2013149 (-0.06-1.16) Combined Planning

Williams et al, 14 14 1.03 0.0086 23.2 86 Autism Response Inhibition

2002627 (0.26-1.80) Diagnosis

Williams et al, 29 34 0.19 0.4393 28.7 90 Autism Working Memory

2005c200 (-0.30-0.68) Diagnosis

Page | 338

Williams et al, 38 38 0.17 0.4468 11.7 no data Autism Working Memory

2006187 (-0.27-0.62) Diagnosis

Williams et al, 17 17 0.07 0.8299 42.1 no data ASD Planning

2012628 (-0.58-0.73) Combined

Williams et al 21 21 0.50 0.1042 10.6 no data ASD Concept Formation

2013150 (-0.10-1.11) Combined

Williams et al, 17 17 0.49 0.1545 31.1 82 ASD Working Memory

2014629 (-0.18-1.16) Combined

Wilson et al, 87 88 0.49 0.0017 26.0 100 ASD Fluency

2014630 (0.18-0.80) Combined Response Inhibition

Winsler et al, 33 28 1.16 0.0000 11.0 97 ASD BRIEF Global EF

2007631 (0.60-1.72) Combined Concept Formation

Woodbury- 23 23 0.52 0.0815 29.7 87 Autism Planning

Smith et al (-0.06-1.10) Diagnosis Response Inhibition

2005632

Xiao et al, 19 16 0.61 0.0753 10.1 100 Autism Response Inhibition

2012633 (-0.06-1.10) Diagnosis

Page | 339

Yang et al, 20 30 0.52 0.0748 8.1 90 ASD Concept Formation

2009634 (-0.05-1.08) Combined Response Inhibition

Working Memory

Yerys et al, 28 21 0.99 0.0013 9.7 71 ASD BRIEF

2009a151 (0.38-1.59) Combined Response Inhibition

Working Memory

Yerys et al, 42 84 0.20 0.2851 10.2 79 Autism Concept Formation

2009(b)156 (-0.17-0.57) Diagnosis

Yerys et al, 22 22 0.51 0.0884 8.5 82 ASD Concept Formation

2012115 (-0.08-1.10) Combined

Yerys et al, 20 19 0.86 0.0090 11.3 80 Autism Mental Flexibility

2015378 (0.22-1.51) Diagnosis

Yeung et al, 25 25 0.77 0.0075 10.1 76 Autism Concept Formation

2016635 (0.21-1.34) Diagnosis Fluency

Mental Flexibility

Page | 340

Yi et al, 2014636 25 28 0.52 0.0631 7.7 78 Autism Concept Formation

(-0.03-1.06) Diagnosis Response Inhibition

Working Memory

Yoran-Hegesh 23 43 0.22 0.3917 15.1 100 Asperger Response Inhibition

et al, 2009637 (-0.28-0.72) Syndrome

Zhang et al, 37 80 -0.12 0.5563 18.9 84 Autism Concept Formation

2015638 (-0.50-0.27) Diagnosis Mental Flexibility

Working Memory

Zinke et al, 15 17 0.95 0.0093 9.0 87 Autism Planning

2010639 (0.23-1.67) Diagnosis

Page | 341 eTable S2.3: Hedges’ g values for EF and moderator outcomes Measure Hedges’ g (p-value) I2% Egger’s Number of Studies (CI 95%) Regression Qbetween, df, p-value test (p-value) Overall EF 0.60 p<0.001 75.5 1.53 (incl self/informant questionnaire (0.53-0.67) (p<0.1) data) k=235 Qbetween=948.7, df=234, p<0.001

Overall EF 0.48 p<0.001 46.2 1.21 (excl self/informant questionnaire (0.43-0.53) (p<0.1) data) and outliers k=221 Qbetween=408.7, df=220, p<0.001

EF Domain Hedges’ g (p-value) I2% Egger’s (No of Studies) (CI 95%) (CI 95%) Regression

Qbetween=2.1092, df=5, p- test value=0.83 (p-value) Fluency 0.45 p<0.001 45.0% 1.75 (k=56) (0.35-0.55) (p=0.002)

Concept Formation/Set Shifting 0.48 p<0.001 55.70% 1.11 (k=92) (0.39-0.56) (p<0.03)

Mental Flexibility/Set Switching 0.48 p<0.001 47.30% 1.45 (k=38) (0.36-0.60) (p<0.06)

Planning 0.55 p<0.001 57.1% 0.94 (k=51) (0.44-0.66) (p=0.32)

Response Inhibition 0.46 p<0.001 52.94% 1.15 (k=103) (0.38-0.53) (p<0.1)

Working Memory 0.47 p<0.001 59.1% 0.97 (k=70) (0.37-0.57) (p=0.15)

Page 342 of 393

EF Domain Moderator Hedges’ g p-value I2 grouping (CI 95%) (No of Studies) (Number of Studies) Q , df, p-value= between

Children<1 0.51 p<0.001 47.19% 2 (0.43-0.59) (k=86) Overall EF 0.43 p<0.001 47.89% Youth (0.30-0.57) (k=171) >12<18 (k=33) 0.46 p<0.001 29.1% Qbetween=1.2095, df=2, p=0.546 (0.36-0.56) Adults>18 (k=52)

Children<1 0.61 p<0.001 29.09% 2 (0.45-0.77) (k=17) Fluency 0.49 p=0.06 78.83%

Youth (-0.02- (k=48) >12<18 0.99)

(k=6) p<0.001 0.00% Q =4.3417, df=2, p=0.114 between 0.40

Adults>18 (0.29-0.51) (k=25)

Children<12 0.48 p<0.001 53.19% (k=36) (0.35-0.61)

Concept Formation Youth 0.48 p<0.001 53.78%

>12<18 (0.27-0.69) (k=67) (k=14)

0.57 p<0.001 73.13% Q =0.3289, df=2, p=0.8484 between Adults>18 (0.28-0.86)

(k=17)

Children<1 0.60 p<0.001 63.38% 2 (0.35-0.85) (k=13) Mental Flexibility 0.47 p=0.16 67.59%

Youth (-0.19- (k=27) >12<18 1.14)

(k=2) p<0.001 0% Q =0.6417, df=2, p=0.7255 between 0.48

Adults>18 (0.31-0.65) (k=12)

Page | 343

Children<1 0.58 p<0.001 37.55% 2 (0.42-0.74) (k=19) Planning 0.41 p=0.006 56.69%

Youth (0.12-0.70) (k=36) >12<18

(k=8) 0.27 p=0.02 34.69% Q =4.9891, df=2, p=0.0825 between (0.05-0.49)

Adults>18 (k=9)

Children<1 0.48 p<0.001 47.32% 2 (0.37-0.60) Response Inhibition (k=45) 46.62%

0.38 p<0.001 (k=83) Youth (0.20-0.55)

>12<18 Q =1.1530, df=2, p=0.5618 between (k=17) 0.49 p<0.001 28.77%

(0.34-0.64) Adults>18 (k=21)

Children<1 0.62 p<0.001 54.57% 2 (0.48-0.75) Working Memory (k=30)

0.20 p=0.21 41.2% (k=52) Youth (-0.11-

>12<18 0.51) Q =6.8478, df=2, p=0.0325 between (k=6) p<0.001 48.36%

0.40 Adults>18 (0.17-0.63) (k=16)

Page | 344

Moderator Moderator Hedges’ g p-value I2 grouping (CI 95%) (CI 95%) (No of Studies) (Number of Studies) Q , df, p-value between Female 0.44 p=0.001 0% Gender (k=4) (0.13-0.75)

k=52 Male 0.54 p<0.001 38.68%

(k=34) (0.41-0.68) Q =0.3665, df=1, p=0.5449 between

ASD combined 0.49 p<0.001 35.76% group (0.42-0.56) (k=94) ASD Diagnostic group Asperger’s 0.42 p<0.001 20.4% k=221 syndrome (0.29-0.55) (k=24) Qbetween=1.0977, df=2, p=0.5776 Autism Diagnosis 0.49 p<0.001 56.45% (k=103) (0.41-0.58)

Siblings 0.42 p<0.001 0% (k=8) (0.27-0.57) Control type k=221 Typical 0.49 p<0.001 47.21%

(k=213) (0.43-0.54) Q =0.6373, df=1, p=.4247 between

Experimental Task 0.46 p<0.001 56.71% (k=88) (0.37-0.55)

Assessment type Psychometric Test 50.44%

(k=155) 0.50 p<0.001 k=263 (0.44-0.57)

Questionnaire Q =53.1375, df=2, p<0.001 between (k=20) 1.84 p<0.001 89.64% (1.48-2.20)

Traditional 0.54 p<0.001 51.0% presentation (0.47-0.61) Assessment format (k=134) 36.2% k=237 Computer 0.42 p<0.001 presentation (0.35-0.49) Qbetween=6.0909, df=1, p<0.014 (k=103)

Page | 345

ADOS/ADI 0.41 p<0.001 25.9% (k=39) (0.32-0.50)

Clinician 0.35 p=0.02 62.20 (k=10) (0.05-0.64)

Diagnostic tool DSM/ICD 0.56 p<0.001 65.8% k=218 (k=80) (0.45-0.67)

Qbetween=4.9546, df=4, p=0.2919 DSM/ICD plus 0.46 4.0% ADOS/ADI (0.40-0.52) p<0.001 (k=84)

Screening test 0.44 p=0.008 6.13 k=5 (0.11-0.76)

Age plus IQ 0.47 p<0.001 25.07% (k=62) (0.38-0.55)

Chronological Age 0.59 p<0.001 46.06% (k=25) (0.43-0.75) Sample matching criteria

IQ only 0.50 p=0.003 44.78% k=104 (k=7) (0.17-0.83)

Q =5.4635, df=3, p=0.1408 between Mental Age 0.29 (k=10) (0.08 – p=0.006 17.86% 0.49)

Non Verbal 0.48 p<0.001 47.51% (k=172) (0.43-0.54) Stimulus Processing Mode

Verbal 0.50 p<0.001 64.34% k=283 (k=111) (0.41-0.58)

Q =0.0843, df=1, p=0.7715 between

Motor 0.49 p<0.001 48.28% (k=186) (0.43-0.54) Response mode Verbal 0.49 p<0.001 63.81% k=292 (k=106) (0.41-0.58) Qbetween=0.0013, df=1, p=0.9717

Page | 346

Asia East 0.49 p=0.002 44.25% (k=28) (0.35-0.64)

Asia Middle East 0.91 p=0.011 80.51% (k=4) (0.21-1.62)

Australia 0.47 Geographical location (k=10) (0.17-0.77) p=0.002 40.75%

k=221 Europe 0.45

(k=96) (0.38-0.53) p<0.001 48.0% Q =3.5404, df=5, p=0.6173 between North America (k=77) 0.53 p<0.001 44.97% (0.45-0.62)

South America (k=6) 0.42 p=0.001 0% (0.16-0.69)

Page | 347

eTable S2.4: Clinical sensitivity / Cohen’s d overlap statistic

Dependent Measure Cohen’s d 95% CI (x1-x2) Overlap

(Number of Studies) Statistic

(OL%)

BRIEF – Self Report Emotional Control 5.83 (4.44 – 7.22) 0.18

(k=1)

STROOP CAS Interference 3.95 (2.99 – 4.91) 2.48

(k=1)

BRIEF – Self Report Shift 3.60 (1.38 – 5.81) 3.74

(k=2)

BRIEF – Self Report Inhibit 3.58 (0.19 – 6.97) 3.81

(k=2)

Stanford - Binet Problem Situations 3.15 (2.11 – 4.19) 6.10

(k=1)

BRIEF – Self Report Global Executive 3.12 (2.18 – 4.05) 6.34

(k=1)

BRIEF Caregiver Shift 2.43 (2.13 – 2.72) 12.67

(k=6)

BRIEF Caregiver GE 2.40 (1.51 – 3.28) 13.06

(k=4)

CAS Planning 2.27 (1.53 – 3) 14.74

(k=1)

N-Back task errors 2.26 (1.62 – 2.91) 14.82

(k=1)

BRIEF – Self Report Behavioural Regulation Index 2.15 (0.71-3.6) 16.44

(k=4)

Page | 348

Stanford-Binet Verbal Absurdities 1.99 (0.36 – 3.62) 18.97

(k=2)

BRIEF Caregiver Behavioural Regulation Index 1.85 (1.34 – 2.36) 21.54

(k=6)

BRIEF Caregiver Metacognition 1.77 (1.25 – 2.29) 23.15

(k=6)

WCST Conceptual Level Responses 1.73 (0.69 – 2.78) 23.96

(k=3)

BRIEF Caregiver Self Monitor 1.66 (1.09 – 2.22) 25.62

(k=5)

BRIEF Caregiver Initiate 1.59 (1.35 – 1.84) 27.09

(k=4)

BRIEF Caregiver Working Memory 1.56 (1.35 – 1.77) 27.76

(k=6)

BRIEF Caregiver Emotional Control 1.53 (1.31 – 1.76) 28.46

(k=5)

BRIEF Caregiver Plan/Organise 1.42 (0.89 – 1.95) 31.41

(k=7)

Page | 349

eTable S2.5: MOOSE checklist

Criteria Brief description of how the criteria were handled in the

meta-analysis

Reporting of background should include

 Problem definition The role of executive dysfunction in Autism Spectrum

Disorders (ASD) throughout development is unclear. Empirical

evidence for the differential contribution of discrete executive

function (EF) domains and role of mediating factors is mixed.

Establishing the role of EF in ASD to guide diagnostic

measures and clinical treatments is of critical importance.

.

 Hypothesis statement Overall EF will be impaired in ASD, individual EF

subdomains will make a differential contribution to executive

dysfunction and this will be correlated with improved clinical

sensitivity in associated behavioural and informant EF

measures.

 Description of study outcomes Executive function

 Type of exposure or intervention Autism Spectrum Disorder

used

 Type of study designs used Cross sectional and longitudinal observational studies

 Study population Autism Spectrum Disorder with comparative typical control

group.

Reporting of search strategy should include

 Qualifications of searchers The qualifications of investigator EAD are indicated in the

author list.

Page | 350

 Search strategy, including time Embase from 1980 – June 2016

period included in the synthesis Medline from 1980 – June 2016

and keywords PsychINFO from 1980 – June 2016

 Databases and registries searched Embase, Medline, PsychINFO

 Search software used, name and No specific search software was employed. EndNote was used

version, including special features to combine search results and remove duplicates.

 Use of hand searching Reviews were searched for additional references.

 List of citations located and those Details of the literature search process are outlined in the flow

excluded, including justifications chart.

 Method of addressing articles Articles not published in the English language were excluded.

published in languages other than

English

 Method of handling abstracts and Only peer reviewed articles were included in the meta-analysis.

unpublished studies

 Description of any contact with Due to the large scope of the study, individual authors were not

authors contacted.

Reporting of methods should include

 Description of relevance or Inclusion and exclusion criteria were described in the methods

appropriateness of studies section.

assembled for assessing the

hypothesis to be tested

 Rationale for the selection and Data was extracted from each of the studies based on the

coding of data identified population and comparison group characteristics,

outcome measures and moderator variables.

 Assessment of confounding Based on quality measurement instrument ratings.

 Assessment of study quality, Based on quality measurement instrument ratings.

including blinding of quality

Page | 351

assessors; stratification or

regression on possible predictors

of study results

 Assessment of heterogeneity Study heterogeneity was assessed using Cochrane’s Q test of

heterogeneity and I2 statistic.

 Description of statistical methods Description of methods of meta-analyses, sensitivity analyses,

in sufficient detail to be replicated clinical sensitivity measures and assessment of publication bias

are detailed in the methods.

 Provision of appropriate tables A comprehensive reporting of data analysis is presented in

and graphics supplementary materials.

Reporting of results should include

 Graph summarizing individual Due to the large number of studies, summary statistics are

study estimates and overall included in eTable 2

estimate

 Table giving descriptive eTable 2

information for each study

included

 Results of sensitivity testing Reported in results section and supplementary materials

 Indication of statistical 95% confidence intervals were presented with all summary

uncertainty of findings estimates, I2 values and results of sensitivity analyses

Reporting of discussion should include

 Quantitative assessment of bias Sensitivity analyses indicate heterogeneity in strengths of the

association due to most common biases in observational

studies.

 Justification for exclusion Excluded studies that reported EF measures based on affective

stimuli.

Page | 352

 Assessment of quality of included Quality analysis is available from the first author, potential

studies reasons for the observed heterogeneity are discussed in the

paper.

Reporting of conclusions should include

 Consideration of alternative Discussion was made of potential moderators (e.g. symptom

explanations for observed results severity, mood states, comorbidities) as alternative explanation

for observed results.

 Generalization of the conclusions Conclusions apply to the broad spectrum of autism

populations.

 Guidelines for future research The lack of fractionated EF differences guide future research

directions.

 Disclosure of funding source No separate funding was necessary for the undertaking of this

meta-analysis.

Page | 353

eFigure S2.1: PRISMA flow chart

3508 Records identified through database search

1980 Records after duplicates removed

1980 Records screened based on title and 1506 Records excluded abstract

255 Full-text articles excluded

240 Did not meet inclusion criteria 474 Full-text articles assessed for eligibility 13 Insufficient data

2 Emotional stimuli

219 Studies included in the meta-analysis 235 study samples analysed

Page | 354

Appendix 3: Supplementary data, Chapter 3

Page | 355

Table S3.1: Summary of diagnostic, neuropsychological and self-report assessments

Measure Domain Characteristics

Clinical and screening measures

ADOS-2 Autism Spectrum Semi-structured, standardized assessment of:

Autism Diagnostic Observation Disorder  social interaction and communication,

Schedule – 2nd edition  restricted and repetitive behaviours

322

SRS-2 Autism Spectrum Self-report measure of autistic traits and capacity to identify, understand social cues and

Social Responsiveness Scale Disorder engage in social interaction

259

Page | 356

Neuropsychological measures

WTAR Premorbid IQ Reading of fifty words and assessment of correct pronunciation may be subject to

Wechsler Test of Adult Reading323 regional language variations in pronunciation. Outcome measure predicted IQ from total

number of words read correctly

TMT-A Psychomotor speed Outcome measure is completion time in seconds, a higher score indicates worse

Trail Making Test-A119 performance

TMT-B Mental flexibility Outcome measure is completion time in seconds, a higher score indicates worse

Trail Making Test-B119 performance

LM – WMS-III

Logical Memory Test

Wechsler Memory Scale 3rd Verbal narrative memory Outcome measure is number of correct phrases at Immediate and Delayed recall edition135

RAVLT Verbal learning and Outcome measure is number of correct words at Immediate and Delayed recall

Rey Auditory Verbal Learning memory

Test21

Page | 357

SSP Visuospatial memory Outcome measure is the number of visual sequences of objects retained in short term

Spatial Span Test87 memory

PAL Visuospatial learning and Capacity to learn where a sequence of patterns is located, the outcome measure is total

Paired Associate Learning87 memory number of errors

IED Set shifting and flexibility Outcome measures are:

Intra-Extra Dimensional Shift of attention  stages completed, the total number of stages completed successfully (range 1-9)

Test87  total errors (adjusted), the expected maximum number of errors irrespective of

whether the participant completed all 9 stages (range 0-225).

RVP Sustained attention Outcome measure is ‘RVP-A Prime, derived from Signal Detection Theory (SDT) and is

Rapid Visual Processing Test87 a measure of sensitivity to the target (range 0.00-1.00). A score of ‘1’ indicates that the

participant always detected the target.

COWAT Phonemic and Semantic  Phonemic fluency, the letters “F”, “A”, and “S” were used

Controlled Oral Word Association fluency  Semantic fluency, the “animal” category was used

Test21 Participants were required to name as many words as possible in each letter and animal

categories within 60 seconds.

Outcome measures are total sum of words for the Phonemic and Semantic categories. A

higher score indicates better performance.

Page | 358

Self-report-measures

BRIEF Executive function A 75 item self-report questionnaire that assesses executive function. It consists of nine

Behavioural Rating Inventory of clinical scales that in combination provide an overall score of EF and two index scores:

Executive Function103  GEC - Global Executive Composite

 BRI - Behavioural Regulation Index (derived from the clinical scales Inhibit,

Shift, Emotional Control and Self Monitor)

 MCI - Metacognition Index (derived from the clinical scales Initiate, Working

Memory, Plan/Organize, Task Monitor, Organization of Materials)

Page | 359

Appendix 4: Supplementary Data, Chapter 4

Page | 360

Table 4.1: Factor structure of neuropsychological tests

Test measure Factor 1 Factor 2 Factor 3 Factor 4

COWATFluency .766

TMT-B -.647

COWATSemantic .633

RVP A’ .584

TMT-A -.568

SSP 0.509

LM II .952

LM I .903

RAVLTLearning .493

RAVLTDelayed Recall .434

PALTotal Errors 1.006

PAL6 Shape Adj .876

IED Total errors .319

Total variance: 62.28%

Page | 361

Table 4.2: Executive function/attention factor

Communalities Initial

RVP-A .271

TMT-B .190

IEDTotal Errors .408

COWATFluency .382

COWATSemantic .333

Page | 362

Appendix 5: Supplementary Data, Chapter 5

Page | 363

Table S5.1: Clinical, neuropsychological and disability measures.

Measure Domain Characteristics

Clinical Measures

ADOS-2 Autism Spectrum Semi-structured, standardized assessment of:

Autism Diagnostic Observation Disorder  social interaction and communication,

Schedule – 2nd edition322  restricted and repetitive behaviours

ADIS Social Anxiety Disorder Semi-structured interview based on DSM-5 criteria for current assessment of mental

Anxiety Diagnostic Interview health conditions and for differential diagnosis among them:

Schedule359  anxiety

 mood

 obsessive-compulsive disorder

 trauma

 related disorders (e.g., somatic symptom, substance use)

SCID-I First Episode Psychosis Semi-structured interview for the major DSM-IV Axis I, II diagnoses.

The Structured Clinical Interview for DSM-IV Axis I Disorders640

Page | 364

SAPS

Scale for the Assessment of

Positive Symptoms641

SANS

Scale for the Assessment of

Negative Symptoms641

Neuropsychological measures

WTAR Premorbid IQ

Wechsler Test of Adult Reading323

TMT-A Psychomotor speed Outcome measure is completion time in seconds, a higher score indicates worse

Trail Making Test-A119 performance

TMT-B Mental flexibility Outcome measure is completion time in seconds, a higher score indicates worse

Trail Making Test-B119 performance

Page | 365

COWAT Phonemic and Semantic  Phonemic fluency, the letters “F”, “A”, and “S” were used

Controlled Oral Word Association fluency  Semantic fluency, the “animal” category was used

Test21 Participants were required to name as many words as possible in each letter and animal

categories within 60 seconds.

Outcome measures are total sum of words for the Phonemic and Semantic categories. A

higher score indicates better performance.

IED Set shifting and flexibility Outcome measures are:

Intra-Extra Dimensional Shift of attention  stages completed, the total number of stages completed successfully (range 1-9)

Test87  total errors (adjusted), the expected maximum number of errors irrespective of

whether the participant completed all 9 stages (range 0-225).

RVP Sustained attention Outcome measure is ‘RVP-A’, derived from Signal Detection Theory (SDT) and is a

Rapid Visual Processing Test87 measure of sensitivity to the target (range 0.00-1.00). A score of ‘1’ indicates that the

participant always detected the target.

Self-report-measures

BRIEF Executive function A 75 item self-report questionnaire that assesses executive function. It consists of nine

Behavioural Rating Inventory of clinical scales that in combination provide an overall score of EF and two index scores:

Executive Function103  GEC - Global Executive Composite

Page | 366

 BRI - Behavioural Regulation Index (derived from the clinical scales Inhibit,

Shift, Emotional Control and Self Monitor)

 MCI - Metacognition Index (derived from the clinical scales Initiate, Working

Memory, Plan/Organize, Task Monitor, Organization of Materials)

WHODAS-2 Disability A 36-item self-report questionnaire that assesses the disability burden of mental and physical health problems overall and across six domains: World Health Organisation  Cognition Disability Assessment Schedule-  Mobility 2266  Self-Care  Getting Along  Life Activities  Participation

The overall score range from 0-79. Higher scores indicate greater functional disability.

DASS-21 A 21 item self-report questionnaire that assesses Depression, Anxiety and Stress over the Depression Anxiety Stress Scale360 last week. Higher scores indicate greater impairment.

Page | 367

Table S5.2: Bivariate correlation between EF performance and self-report measures

Measure 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

1. TMT-A - 2. TMT-B .600** - 3. RVP-A' -.465** -.467** -

4. IEDErrors .290** .337** -.350** -

5. FluencyPhonetic -.353** -.315** .339** -.250** -

6. FluencySemantic -.316** -.330** .287** -.224** .572** -

7. BRIEFInhibit .350** .213* -.200* .204* -.120 -.121 -

8. BRIEFShift .336** .361** -.288** .230* -.245** -.215* .720** -

9. BRIEFEmotional Control .336** .347** -.299** .395** -.267** -.134* .609** .776** -

10. BRIEFSelf Monitor .441** .366** -.239* .385** -.238* -.168 .741** .720** .721** -

11. BRIEFInitiate .276** .189* -.167 .081 -.195* -.164 .736** .702** .540** .518** -

12. BRIEFWorking Memory .263** .291** -.229 .155 -.120 -.151 .810** .798** .603** .584** .796** -

13. BRIEFPlan Organize .328** .258** -.164 .199* -.153 -.122 .792** .733** .618** .650** .809** .809** -

14. BRIEFTask Monitor .297** .171 -.185* .105 -.097 -.096 .739** .643** .506** .515** .775** .807** .806** -

15. BRIEFOrganization .210* .123 -.036 .091 -.119 -.076 .667** .557** .546** .613** .623** .661** .775** .659** -

Materials

Page | 368

Page | 369