Personalised Mental Health Care for Young People

Personalised Mental Health Care for

Young People: Using Past Outcomes to Build

Future Solutions

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

Doctor of Philosophy

Frank Iorfino

Brain and Mind Centre

Faculty of Medicine and Health

The University of Sydney

2018

Acknowledgements

Firstly, thank you to all the incredible young people who have taken part in this clinical research program over the years. Your generosity has been admirable and I hope that this work will play some role in bringing about a healthier future for all young people.

I would like to thank my supervisor Professor Daniel Hermens, for the incredibly valuable advice and support you have given me over the past few years. Your guidance has extended way beyond this thesis to other areas of academic life that will certainly stay with me forever. Thank you for always reminding me that the map is the treasure—it always provided a sense of composure that inspired discovery and creativity. To Professor Ian Hickie, you have been an inspirational mentor whose passion and genuine interest in making a difference for young people with mental health disorders has been a constant motivation. I have thoroughly enjoyed working with you both, and look forward to continuing our work in the future.

Much of the work in this thesis could not have been completed without contributions from my co-authors who each had a major role in getting the work done. I’d like to give special mention to Dr Shane Cross who provided such valuable clinical insights. Your clinical and health service experience has really been key to ensuring the focus of this work was always directed towards having a real-world clinical impact. To Tracey Davenport, I would like to thank you for always challenging me to be creative and providing me with so many opportunities to learn and develop new skills. Special thanks also to Laura-

Ospina Pinillos. I feel incredibly lucky to have shared this PhD experience with you, and while your clinical expertise has always been so helpful, your friendship and support over the course of this thesis is something I will always remember. To everyone who helped gather such incredible data for this project, specifically, Natalia, Alissa and Josine, I could not have done any of this without your incredible efforts, so thank you very much.

Thank to my other friends at the BMC who have made my time here so memorable, particularly Courtney, Kate, Julie, Mac, Amelia, Django, Sarah and

Haley. These friendships have been so important to me and made coming to work every day so enjoyable. Specific thanks to Courtney, I am very grateful for the friendship we have developed over coffee breaks and endless debates about sport, which were always the perfect way to clear the head and break up a day of writing.

Thank you to all my friends and family who have been so very patient and supportive during this time. Specifically, my parents, sister Tania and Michael, who despite having no clue what it is I do, constantly provided encouragement and support. Hopefully this document will provide some insight into what it is

I have been doing all these years. To my wonderful group of friends; Matt,

Steph, Sam, Dom, Luke, Britt, Alex and Jenny, thank you for the major role you have each played in helping me maintain a healthy balance over the past few years through a plentiful supply of food, wine and good times. Finally, thank you to Charlotte. You have been by my side every single day with your unwavering belief, love and encouragement. Your endless support has kept me going and is something I will always cherish. The work you do every day with young people is truly inspirational and I would be happy if this thesis could have even a quarter of the real-world impact that your work has in a single day.

Now, I’m off to read Dune on a beach somewhere.

Table of Contents

Acknowledgements ...... 2 Declaration of originality ...... 5 Authorship attribution statement...... 6 Publications presented for examination ...... 7 Additional relevant peer-reviewed publications ...... 8 Conference presentations during candidature ...... 9 List of common abbreviations ...... 11 Abstract ...... 12 Thesis overview ...... 15 Chapter 1: Introduction ...... 16 1.1. Mental health disorders and the vulnerability of young people ...... 16 1.2. The broad impact of mental health disorders ...... 18 1.3. Early intervention and youth mental health services ...... 21 1.4. Addressing the needs of young people—challenges in practice ...... 23 1.5. Aims of the current thesis ...... 26 Chapter 2: Trajectories of social and occupational function ...... 28 Chapter 3: Suicide attempts and long-term outcomes ...... 39 Chapter 4: Responding to suicidality using technology ...... 47 Chapter 5: Neurobiological correlates and multidimensional outcomes ...... 62 Chapter 6: Discussion ...... 101 6.1. Summary of findings...... 101 6.1.1. Delineating long-term outcomes ...... 101 6.1.2. Future solutions for personalised mental health care ...... 102 6.2. Early intervention—not early enough ...... 103 6.3. Individual care needs over time ...... 107 6.4. Integrated and outcomes-focused care ...... 111 6.5. Technology-enhanced mental health care ...... 115 6.6. Limitations of the current work ...... 117 6.6.1. Clinical cohort ...... 117 6.6.2. Data collection methodology ...... 118 6.6.3. Limited treatment and evaluation data ...... 119 6.7. Future directions ...... 120 6.8. Concluding remarks ...... 123 References ...... 124

Declaration of originality

To the best of my knowledge, this thesis contains no copy or paraphrase of work published by another person, accept where duly acknowledged in the text. This thesis contains no material that has been presented for a degree at the

University of Sydney, or any other University.

Frank Iorfino Date: 28th April, 2018

Authorship attribution statement

This thesis principally represents the work of Frank Iorfino. The data forming the basis of the empirical studies was collected under protocol conceptualised and designed by Prof. Ian B. Hickie, Prof. Daniel F. Hermens,

Prof. Sharon L. Naismith, A/Prof Elizabeth M. Scott, Prof Adam Guastella and Ms Tracey Davenport. Mr Iorfino contributed significantly to data collection throughout the candidature, and was primarily responsible for all data analyses, interpretation, and preparation of manuscripts presented in this thesis. Assistance with data collection and interpretation across the studies presented here was provided by Ms Laura Ospina-Pinillos, Dr Shane

Cross, Ms Natalia Zmicerevska, Ms Alissa Nichles and Ms Josine Groot.

Assistance with data analysis for chapter two was provided by Ms Caro-Anne

Badcock. Each author provided comments on manuscript drafts and approved the final versions of papers on which they are listed as co-authors.

As supervisor for the candidature upon which this thesis is based, I can confirm that the authorship attribution statements above are correct.

Prof. Daniel Hermens Date: 28th April, 2018

Publications presented for examination

The following papers form the basis of this thesis:

. Iorfino, F., Hermens, D. F., Cross, S. P., Zmicerevska, N., Nichles, A.,

Badcock, C. A., Groot J., Scott, E., & Hickie, I. B. (2018). Delineating the

trajectories of social and occupational functioning of young people

attending early intervention mental health services in Australia: a

longitudinal study. BMJ Open, 8(3), e020678.

Presented in Chapter 2 of this thesis.

. Iorfino, F., Hermens, D. F., Cross, S. P., Zmicerevska, N., Nichles, A.,

Groot J., Guastella, A.G., Scott, E., & Hickie, I.H. (2018). Prior suicide

attempts predict worse clinical and functional outcomes in young people

attending a mental health service. Journal of Affective Disorders. 238: 563-

569.

Presented in Chapter 3 of this thesis.

. Iorfino, F., Davenport, T. A., Ospina-Pinillos, L., Hermens, D. F., Cross,

S., Burns, J., & Hickie, I. B. (2017). Using new and emerging technologies

to identify and respond to suicidality among help-seeking young people:

a cross-sectional study. Journal of Medical Internet Research, 19(7).

Presented in Chapter 4 of this thesis.

. Iorfino, F., Hickie, I.B., Lee, R.S., Lagopoulos, J & Hermens, D.F. (2016).

The underlying neurobiology of key functional domains in young people

with affective disorders: A systematic review. BMC Psychiatry, 16(1): 1.

Presented in Chapter 5 of this thesis.

Additional relevant peer-reviewed publications

A substantial contribution was also made to the following papers, as part of research studies, which were complementary to this thesis. The outcomes of these studies helped to inform the theoretical framework and interpretations of the data reported in this thesis.

. Scott, E., Iorfino, F., Cross, S. P., Hermens, D. F., L., White, Guastella,

A.G., Carpenter, J.S., Scott, J., McGorry, P., & Hickie, I. B. Predictive

validity of the early phases of a clinical staging framework in young

people presenting to early intervention services for common mental

disorders. Acta Psychiatrica Scandinavica. [in submission].

. Ospina-Pinillos, L., Davenport, T. A., Iorfino, F., Cross, S., Burns, J., &

Hickie, I. B. Testing new and innovative technologies to assess clinical

stage in early intervention youth mental health services. Journal of

Medical Internet Research. In press.

. Tickell, A.M., Scott, E.M., Davenport, T.A., Iorfino, F., Ospina-Pinillos,

L., Harel, K., Parker, L., Hickie, I. B. & Hermens, D. F. Neurocognitive

clusters: A pilot study of young people with affective disorders in an

inpatient facility. Journal of Affective Disorders [under review; revisions

submitted]

. Tickell, A.M., Scott, E.M., Davenport, T.A., Iorfino, F., Ospina-Pinillos,

L., White, D., Harel, K., Parker, L., Hickie, I. B. & Hermens, D. F.

Developing Neurocognitive Standard Clinical Care: A Study of Young

Adult Inpatients. PloS One [Under review]

. Iorfino, F., Alvares, G. A., Guastella, A. J., & Quintana, D. S. (2015). Cold

face test-induced increases in heart rate variability are abolished by

engagement in a social cognition task. Journal of Psychophysiology.

30(1): 38-46.

Conference presentations during candidature

. Iorfino, F., Cross, S.P.M & Hickie, I.B. The use of emerging technologies

to generate new clinical findings that guide improvements to service

provision. 7th Annual E-Mental Health Conference. 2nd – 3rd February, 2018,

Vancouver, Canada. Joint keynote oral presentation.

. Iorfino, F., Hermens, D.F., Cross, S.P., Zmicerevska, N., Nichles, A.,

Scott, E., & Hickie, I.B. Trajectories of social and occupational functioning

among young people with emerging mental illness. 39th Annual Society

for Mental Health Research Conference. 6th – 8th December, 2017, Canberra,

Australia. Poster presentation.

. Davenport, T., Iorfino, F., Cross, S.P., Ospina-Pinillos, L., Scott, E., &

Hickie, I.B. Project Synergy: Striving for technology-enabled mental

health service reform. 39th Annual Society for Mental Health Research

Conference. 6th – 8th December, 2017, Canberra, Australia. Oral

presentation.

. Iorfino, F., Cross, S.P., Ospina-Pinillos, L., Davenport, T.D., Hermens,

D.F., English, A., Naismith, S., Guastella, A., Carpenter, J., Scott, E.M.,

Buus, N., Glozier, N., Hickie, I.B. Using new and emerging technologies

to design and develop an online clinical shared decision making tool for

personalised mental health care. 4th International conference on Youth

Mental Health. 24th – 26th September, 2017, Dublin, Ireland. Table top

presentation.

. Iorfino, F., Davenport, T.D., Ospina-Pinillos, L., Cross, S.P., & Hickie, I.B.

Using new and emerging technologies to identify and respond to

suicidality among young people seeking help within primary care and

community settings. The Annual National Suicide Prevention Conference.

26th – 29th July, 2017, Brisbane, Australia. Oral presentation.

. Iorfino, F., Hickie, I.B., Davenport, T.D., & Hermens, D.F. The

implementation of a personalised assessment system to characterise

clinical and functional profiles in young people with affective disorders .

18th Annual Conference of the International Society for Bipolar Disorders Held

Jointly with the 8th Biennial Conference of the International Society for

Affective Disorders. 13th – 16th July, 2016, Amsterdam, The Netherlands.

Poster presentation.

. Iorfino, F., Hickie, I.B., Davenport, T.D., & Hermens, D.F. The

development of a personalised assessment and feedback system for

young people with emerging mood disorders. 37th Annual Society for

Mental Health Research Conference. 2nd – 4th December, 2015, Brisbane,

Australia. Poster presentation.

. Iorfino, F., Hickie, I.B., & Hermens, D.F. The design and development of

a personalised assessment and feedback system for young people with

emerging mood disorders. Australasian Society for Bipolar & Depressive

Disorders Conference. 6th – 8th December, 2015, Sydney, Australia. Invited

oral presentation.

. Iorfino, F., Hickie, I.B., & Hermens, D.F. Neurobiological aspects of key

functional outcomes associated with affective disorders in young people:

A systematic review. Inaugural Inter-university Neuroscience & Mental

Health Conference. 29th – 30th September, 2014, Sydney, Australia. Poster

presentation.

List of common abbreviations

DSM - Diagnostic and Statistical Manual of Mental Disorders

ICD - International Classification of Diseases

NEET - Not in Education, Employment or Training

RDOC - Research Domain Criteria

SOFAS - Social and Occupational Functional Assessment Scale

WHO - World Health Organisation

Abstract

Mental health disorders present one of the most serious public health challenges we face in the 21st century. Their prevalence, early age of onset and chronicity contribute to the substantial burden and secondary risks associated with these disorders. They tend to emerge during adolescence and young adulthood, a period characterised by major physical and social change, and so the effects of these disorders can often have long-term consequences on the adult lives of young people. Previously young people have had poor access to health services that address their needs, however the development of youth-focused health services has been a critical step forward to addressing this gap, providing better access to mental health care for adolescents and young adults. The nature of mental health disorders among young people, however, means that there are still major challenges in providing quality mental health care that addresses the broad range of health, social and functional needs of young people.

The overall aim of this thesis was to examine the long-term outcomes of young people attending early intervention youth mental health services to inform the development and delivery of personalised mental health care that address the needs of young people. The first two studies utilise real-world longitudinal data from a trandiagnostic cohort of young people after their initial presentation to an early intervention mental health service. The first of these studies investigated the longitudinal course of social and occupational functioning and identified that young people with emerging mental health disorders have significant functional impairment at presentation for care, and for the majority, it persists over the course of clinical care. The second study examined the relationship between prior suicidal attempts and long-term clinical and functional outcomes. Prior suicide attempts were associated with a higher risk of further suicidal behaviours, the emergence of bipolar disorder, and poorer health and functional outcomes. This study illustrates that these young people

12 require active system-level strategies to reduce suicidal thoughts and behaviours as well as more intensive management to reduce longer-term secondary risks.

The third study of this thesis assesses the utility of a technology-enabled assessment that was integrated with a health service to enable an appropriate and timely response to young people reporting higher levels of suicidality. This approach meant that clinicians could utilise other clinical information such as the presence of comorbid issues, including psychosis-like symptoms, and a history of suicide plans and attempts to make informed clinical decisions. This study also has implications for the use of personalised assessment and feedback that responds to the broader needs of young people presenting for care. Finally, the fourth study of this thesis is a systematic review, which examines the relationship between measures of brain function (i.e. neuropsychology, neuroimaging) and multidimensional outcomes that may be used to enhance further assessment in clinical practice that identifies potential treatment targets.

The overall findings of this thesis are fourfold. The first is that there is still work to be done to ensure young people are receiving mental health care earlier in the course of these illnesses and prior to the emergence of significant problems.

Further research is still needed in this area to determine which strategies are most effective for improving help-seeking behaviours and promoting early intervention. The second is that health service strategies should be in place to identify and respond to individual health and social needs young people present with, particularly those that are associated with poorer long-term outcomes, such as substantial functional impairment and suicidal behaviours.

The third is that early identification followed by current standard care provisions do not appear to inevitably result in improved outcomes. Instead, it is clear that the development and evaluation of specific, integrated care

13 packages may be needed to reduce the morbidity and mortality due to early- onset major mental health disorders. Finally, there is a role for new technologies in mental health reform and the delivery of personalised mental health care, and future studies should focus on implementation strategies that facilitate the evaluation of these technologies in real world settings.

The youth mental health landscape has undergone dramatic transitions over the past decade, in terms of awareness, access and resource. These advancements are a crucial step in the right direction for addressing mental health disorders and their associated burden in young people. The new insights generated by this thesis have been crucial to inform the next step in this revolution—to ensure that young people who access care, receive quality, personalised mental health that addresses their multitude of needs early in the course of life so that they can lead fulfilled and contributing lives in adulthood.

Thesis overview

Chapter one aims to establish why the mental health of young people is one of the most important public health issues today and provide a general overview of the current challenges that face the delivery of youth mental health care.

Specifically, this introduction aims to highlight the results from population- based studies that emphasise the broader health, social and economic risks of young people with mental health disorders. This is followed by an overview of the development of early intervention youth mental health services—how they have increased access to mental health services and the challenges that remain for improving clinical and functional outcomes. This introduction sets the context for chapters two and three, which are two empirical papers that generate new knowledge with regard to the long-term health, social and economic outcomes of help-seeking young people who attend an early intervention mental health service. This outcome data was used to delineate which young people are particularly at risk for worse clinical and functional outcomes and guide the development of new approaches that aim to improve the delivery of youth mental health care. Chapter four is a third empirical paper, which assesses the role of a technology-enabled clinical protocol to identify and respond to suicidal thoughts and behaviours at service entry. This study has been critical to the development of real-world service provisions that address this specific health care need with further implications for broader use. Chapter five is a systematic review which collates the findings from over 130 studies to determine the neurobiological correlates of these broader health, social and economic outcomes to inform future clinical research and the integration of further assessment into the delivery of youth mental health care. Finally, chapter six aims to collate these findings in an overall discussion about the current long-term outcomes of young people attending early intervention services and the direct implications for future service delivery and mental health reform.

Chapter 1: Introduction

1.1. Mental health disorders and the vulnerability of young people

Mental health disorders present one of the most serious public health challenges we face in the 21st century. They affect fundamental aspects of human life—our ability to work, function and develop quality social relationships, and too often lead to the early loss of life. Approximately one out of every three individuals are affected by mental health disorders over their lifetime and studies continuously show that they are a leading cause of years lost to disability

(Whiteford et al., 2013, Vos et al., 2015, Steel et al., 2014). Globally, these disorders are also associated with high mortality with over 800,000 deaths a year being attributed to suicide, and up to 20 times more people attempting to take their own life (World Health Organization, 2016). Young people in particular are at risk with over 75% of adult mental health disorders emerging before the age of 25 (Jones, 2013, Kessler et al., 2005), and over 45% of the total burden of disease for those aged 10 to 24 being attributed to mental health disorders (Gore et al., 2011). These are the chronic illnesses of young people and if not adequately addressed their impact can have effects that endure a lifetime

(Insel and Fenton, 2005, Patel et al., 2007).

Much of this burden can be attributed to the prevalence and early age-of-onset associated with these disorders. Lifetime prevalence estimates indicate that up to one-third of young people meet diagnostic criteria for a mental health disorder (Merikangas et al., 2009), while the 12-month prevalence among

Australian young people aged 16 to 24 was the largest across all age groups at about 26% (Slade et al., 2009). The emergence of mental health disorders tends to be associated with development from childhood to adolescence with most anxiety and neurodevelopmental disorders occurring earlier in life, while depression, bipolar and psychosis tend to emerge later in adolescence and young adulthood (Merikangas et al., 2009, Kim-Cohen J et al., 2003, Costello et

16 al., 2005). Studies consistently report such high prevalence rates for mental health disorders prior to the age of 25, irrespective of whether or not they adhere to strict diagnostic rules about symptom thresholds and impairment

(Copeland et al., 2011, Roberts et al., 2015, Angst et al., 2005). The prevalence of mental health disorders during youth poses a risk for future health and wellbeing outcomes precisely due to the timing at which they tend to emerge.

Adolescence and young adulthood is a critical period of biological and social development—a time of change and transitions. Usually beginning with the onset of puberty, it is a time of major physical and neurobiological change, characterised by the development of higher order cognitive and emotional functions that if suboptimal or disrupted can have a significant impact on behaviour (Paus et al., 2008). In support of this, recent evidence suggests that distinct developmental trajectories of heightened amygdala reactivity emerge during adolescence as a function of positive family history of depression or stressful life events, and prior to the emergence of clinical symptoms or disorder

(Swartz et al., 2015). Importantly, it is the interaction between gene expression and exposure to early life stress that are particularly crucial for influencing the development of key brain circuits responsible for depression and anxiety, and disruptions to the development of these circuits are suggested to underlie disorder (Gross and Hen, 2004, Heim et al., 2008, Uys et al., 2006, Perry et al.,

1995, Sheline et al., 2010, Kendler et al., 2001).

This period is also characterised by significant social development as young people embark on the early phases of their careers via education and employment, and engage in more complex relationships with friends, family, romantic and sexual partners. During this time young people face increasing diversity in potential life trajectories as they move from relatively restricted and homogenous environments, such as school and home life, to environments

17 characterised by greater independence and variability in new educational, employment and social environments (Shanahan, 2000). For some this transition may occur without major disruption, however for others this change in context and roles can lead to maladaptive functioning or manifest and amplify underlying problems (Schulenberg et al., 2004). Thus, the biological and social complexity of this period means that young people are particularly vulnerable to the onset of mental health problems and the long-term impact they can have on outcomes in adulthood (Merikangas et al., 2010b, Paus et al., 2008, Walker et al., 2004).

1.2. The broad impact of mental health disorders

The impact of mental health disorders extends beyond the symptoms that define them. They affect many aspects of human life which typically have their foundations established during adolescence and young adulthood (see chapter

1.1). Thus, it is not surprising that 22 years of age is identified as the age at which the maximum negative impact of a disabling illness occurs (Murray et al., 1996). This next section aims to provide an overview of the broader impact of mental health disorders on the lives of young people and their transition into adulthood. Specifically, this section highlights their impact on social and economic outcomes, the emergence of suicidal thoughts and behaviours, and their comorbidity with alcohol or substance misuse and physical health problems.

Prospective longitudinal studies have demonstrated that the presence of a mental health disorder before the age of 16 affects the economic potential (i.e. reduced income by 28%) and social relationships (i.e. marriage instability) of adults aged 50 years (Goodman et al., 2011). Furthermore, individuals who had a mental health disorder between the ages of 18 and 25 were more likely to have lower occupational participation, income and living standards by age 30 (Gibb

18 et al., 2010). The effect of mental health disorders on these outcomes is greater than the impact of physical health problems among young people (Goodman et al., 2011), and are present even when the mental health disorder is subthreshold or has remitted (Copeland et al., 2015). This illustrates that even though adolescent mental health disorders may dissipate by adulthood, due to the nature of the adolescent and young adulthood period, they may have already affected an individual’s ability to function economically and socially over the course of their life.

These disorders are consistently associated with suicidal thoughts and behaviours, which increases the risk of death by suicide and the overall burden of these disorders (Harris and Barraclough, 1997). Lifetime prevalence estimates indicate that almost 10% of the population across the world engage in suicide ideation and approximately 5% engage in suicidal behaviours (i.e. plans or attempts) with the peak age of onset occurring during adolescence and young adulthood (Nock et al., 2008, Nock et al., 2013). The overall rate of these behaviours is up to three times higher among young people than older adults, yet young people are also less likely to die from a suicide attempt indicating the potential for these behaviours to have a wider, ongoing impact on young people (Crosby et al., 2011, Bunney et al., 2002). Specifically, population-based studies have shown that suicidal thoughts and behaviours among young people are associated with the onset and persistence of mental health problems, physical health comorbidity, and poorer social and occupational outcomes in adulthood (Goldman-Mellor et al., 2014, Fergusson et al., 2005, Brière et al.,

2015, Copeland et al., 2015).

The risk these disorders confer for the persistence and onset of other comorbid mental and physical health problems also contributes to their substantial burden. Many mental health disorders that emerge early in life persist into

19 adulthood, particularly those with longer duration of illness or recurrent episodes in adolescence (Patton et al., 2014, Ormel et al., 2015, Merikangas et al., 2010b, Kessler et al., 2012). It is common for adult disorders to be preceded by the same disorder in childhood or adolescence, however these early disorders are also associated with progression or transitions from one class of disorder to another (Copeland et al., 2009, Costello et al., 2006). Early mental health problems, such as anxiety, behavioural and developmental disorders that are evident in children before the age of 12 predict the full spectrum of later depressive, bipolar and psychotic disorders (Kim-Cohen J et al., 2003). These data emphasise that a trandiagnostic approach to understanding mental health disorders and their associated outcomes may be most appropriate for young people (Insel et al., 2010, Casey et al., 2013, Hickie et al., 2013d).

The comorbidity between different mental health disorders has been well- established in both clinical and community samples (Angold et al., 1999,

Seligman and Ollendick, 1998, Kessler et al., 2003, Merikangas et al., 2007).

Approximately 40% of adolescents with one mental health disorder meet criteria for two or more lifetime disorders (Merikangas et al., 2010b), and studies consistently show that major depression and anxiety disorders are particularly comorbid conditions (Kessler et al., 2015a). The presence of comorbidity is associated with greater illness severity (Fichter et al., 2010,

Ormel et al., 2015), poorer response to treatment (Jakubovski and Bloch, 2014), greater role impairment (Ormel et al., 2008, Kessler et al., 2015a) and higher rates of suicidality (Nock et al., 2010, May et al., 2012, Hoertel et al., 2015). Thus, young people with mental health disorders are not only at risk for persistent mental health disorders in adulthood, but at risk for multiple mental health disorders that often lead to poorer outcomes.

The pervasive impact of mental health disorders extends to other chronic physical illnesses, such as diabetes and cardiovascular disease (Scott and

Happell, 2011, Patten et al., 2008) and alcohol or substance use problems

(Compton et al., 2007, Kessler et al., 2007). All mental health disorders are associated with an increased risk for the onset of a range of chronic physical health conditions with population estimates suggesting that the onset of up to

13% of physical health conditions could be attributed to specific mental health disorders (Scott et al., 2016). The comorbidity between physical and mental health conditions is associated with an increased burden of disease with evidence suggesting that such comorbidity is more burdensome than physical health conditions alone or in combination (Moussavi et al., 2007). Similarly, up to 15% of young people with a prior mental health disorder engage in alcohol or substance abuse with the rates being particularly high for those with anxiety and behavioural disorders (Conway et al., 2016). This contributes to the overall impact of these disorders, since alcohol and substance abuse tends to be associated with greater disability (Dawson et al., 2008) and impaired productivity and interpersonal functioning (Rehm et al., 2009).

Collectively these studies underscore the broader impact of mental health disorders on the capacity for young people to make the transition into adulthood and reach their actual health, social and economic potential. These results emphasise the extent to which the treatment of mental health disorders should not be limited to a discrete set of symptoms, but rather focus more broadly on outcomes across multiple domains to truly address the needs of young people with these emerging disorders.

1.3. Early intervention and youth mental health services

Despite the prevalence and substantial burden associated with mental health disorders among young people, these disorders tend to remain undetected until

21 later in life with a large gap between illness onset and access to sufficient care

(Patel et al., 2007, Kohn et al., 2004). Delayed access to care means that young people are vulnerable to the development of chronic mental health problems and secondary risks (see chapter 1.2), such as social and occupational impairment and comorbid alcohol or substance misuse (Patton et al., 2007,

Jones, 2013, Hickie, 2011). A major cause for poor access can be linked to health service infrastructures that are inadequately designed to address the needs of young people (Patton et al., 2007). Poor continuity between child and adolescent mental health services and adult mental health services is compounded by the insufficient focus of these services on the actual needs of young people

(McDonagh and Viner, 2006, Yung, 2016). These services tend to rely on traditional mental health diagnoses, which are largely based on adult presentations of these disorders. Yet, these diagnostic systems typically do not apply to young people who often present with a range of milder, sub-threshold mental health problems (Hickie et al., 2013b).

These health service limitations have been met by the development of early intervention services in Australia, Canada, Ireland and the United Kingdom, to prevent or delay the onset and progression of these disorders and reduce secondary morbidity (McGorry et al., 2013, Iyer et al., 2015). In Australia, headspace is a primary care-based model for youth mental health care that aims to be highly accessible to young people and provide care that addresses a range of social, physical and mental health problems with locally available specialist service partnerships (McGorry et al., 2007). This service model does away with traditional diagnostic criteria and arbitrary symptom thresholds in favour of recognising the complex and varied needs of young people. These services have already demonstrated successes in reducing the stigma associated with mental health disorders, particularly for engaging young men, and increasing overall access to mental health care among young people (Muir et al., 2009).

Psychological symptoms, such as depression or anxiety, are often reported as the main reason for presentation to these services, however educational, vocational, social and physical health problems are also prevalent issues

(Rickwood et al., 2014, O'Reilly et al., 2015). Initial studies have demonstrated that up to 60% of the population attending these services either have mild mental health problems or attenuated/sub-threshold mental health disorders, and despite not meeting criteria for a Diagnostic and Statistical Manual of

Mental Disorders (DSM) or International Classification of Diseases (ICD) diagnosis, they are often significantly impaired and have substantial comorbidity (Scott et al., 2012, Rickwood et al., 2014). Young people attending these services also tend to report higher rates of disengagement from education, employment or training (up to 24%) than the general population (approximately

11%) (Hamilton et al., 2011, O'Dea et al., 2014). Though, for those engaged in employment or education, functional impairment due to mental ill-health tends to manifest as higher rates of days out of these roles (Scott et al., 2014). This work supports the utility of these services to attract and engage young people who are early in the course of illness yet have a clear need for clinical care.

1.4. Addressing the needs of young people—challenges in practice

A major challenge for early intervention services is the heterogeneous pattern of illness and diverse needs that young people present with. The emergence of mental health disorders during adolescence and young adulthood tend to be characterised by a non-specific set of symptoms, which may or may not develop into distinct syndromes or disorders over time. Current data from recent community studies that assess people longitudinally from childhood or adolescence, emphasise that many clinical phenomena (i.e. anxiety, depressed mood, psychotic-like experiences) are shared, both at the onset and over the trajectory of illness (Merikangas et al., 2012, Merikangas et al., 2010a,

Merikangas et al., 2008, Kelleher et al., 2012, Murray and Jones, 2012).

While diagnostic manuals are used to guide the diagnosis and treatment of these disorders, these have been heavily criticised for their lack of validity and inability to account for the huge heterogeneity of these disorders, particularly among young people due to the emerging and transient nature of symptoms and comorbidities (Scott et al., 2017, Hansell et al., 2012, Scott et al., 2013, Cuthbert,

2014, Casey et al., 2013). This challenge is highlighted by recent data which suggests that the majority of individuals attending these services experience persistent symptoms and functional impairment over short-term follow-up

(Cross et al., 2018, Rickwood et al., 2015b). Consequently, there has been a major shift towards trandiagnostic approaches among these younger populations to better match diagnosis and treatment with the common developmental psychopathology of mental health disorders and specific service settings (Insel et al., 2010, Casey et al., 2013).

Despite major advancements in the development of psychological and pharmacological treatments to treat mental health disorders, the extent to which current service provisions improve long-term outcomes and reduce secondary morbidity remains largely unclear (Jorm et al., 2017). Systematic reviews and meta-analyses have consistently demonstrated the effectiveness of psychological therapies and pharmacotherapies to treat mental health disorders among young people (Weisz and Kazdin, 2010, Compton et al., 2004, Cox et al.,

1993, McDermut et al., 2001), which suggests that the problem may be largely translational. Treatment studies tend to be conducted in more controlled research settings, include restricted sample groups and usually focus on specific syndromes, which means they are generally not suited to address questions about the clinical and functional outcome processes in real-world clinical settings where young people present with greater complexity (Weisz et al.,

2005, Lambert, 2013, Westen et al., 2004). Consequently, studies conducted in these real-world settings tend to report poorer outcomes in symptoms and

24 functioning with many evidence-based treatments failing to outperform usual care (Garland et al., 2006, Weisz et al., 2006).

This challenge has largely driven the development of new personalised approaches (see box 1 below) for identifying and treating common mental health disorders that are not only consistent with developmental epidemiology and neurobiology but also useful when applied in everyday clinical practice

(Insel et al., 2010, Kozak and Cuthbert, 2016, Hickie et al., 2013c, Insel, 2009).

These approaches emphasise that beyond the clinical syndromes, young people are particularly vulnerable to worse health, social and functional outcomes (see chapter 1.2), and that interventions should be targeted to improve these outcomes. The importance placed on diverse outcomes aligns with the substantial burden associated with these disorders and the needs reported by young people and their families, which are reflected in population-based national mental health policies (Saxena et al., 2013). The World Health

Organisations’ (WHO) mental health plan (2017 – 2020) emphasises the need for mental health care to transcend a wholly medical model to address the social determinants of mental health, educational and employment opportunities and psychosocial disability so that people can achieve their highest potential for health and participate fully in society.

This approach reiterates that for truly personalised mental health care in young people, a move away from categorically defined disorders toward a focus on clinically meaningful differentiations that improve outcomes transdiagnostically are urgently needed (Kapur et al., 2012). Understanding whether existing youth mental health services address outcomes across multiple domains is critical to determine their effectiveness and guide the development of new strategies for the delivery of quality mental health care

(National Mental Health Commission, 2014).

Box 1. The nomenclature of ‘personalised’

Current uses of the term ‘personalised’ vary depending on the medical

discipline and context. A primarily biological and genetic perspective would

limit the use of ‘personalised’ to describe the use of unique interventions

which have been customised to an individual, such is the case for

personalised vaccines (Belli et al., 2002). In psychiatry, the term personalised

has been replaced with terms such as ‘stratified’, to better reflect the

subtyping illnesses on the basis of salient treatment-relevant characteristics

(Kapur et al., 2012) or ‘precision’ medicine to place greater emphasis on the

exactness of measurement in psychiatry (Fernandes et al., 2017). A broader,

yet related, concept is ‘person-centred’ medicine which is commonly used to

describe a holistic view of the individual with an emphasis on the role of the

person in treatment (Mezzich et al., 2011). The term ‘personalised mental

health care’ in this thesis aims to encompass various facets from the

definitions described above to more broadly describe the notion that the

assessment and intervention of mental health disorders are tailored to the

individual and their specific needs over time.

1.5. Aims of the current thesis

The epidemiological and clinical data summarised above demonstrates the substantial need for quality mental health care among young people. Young people are particularly vulnerable given the onset, persistence and comorbidity of mental health disorders during such a critical time for biological and social development. Previously there was a huge gap in mental health service provisions for young people, however youth-focused health services, such as headspace, have been a critical step forward to addressing this gap, providing access to mental health care for adolescents and young adults. The nature of mental health disorders among young people, however, means that there are still major challenges in providing quality mental health care that addresses the

26 broad range of health, social and functional needs of young people.

Understanding the extent of mental health problems and secondary risks among a transdiagnostic cohort of young people presenting to these services is critical to guide the future development of solutions that improve long-term outcomes.

1. Investigating the longitudinal course of clinical and functional

outcomes for a trandiagnostic cohort of young people after their

initial presentation to an early intervention mental health service;

and

2. Assessing the utility of technology-enabled assessment and

neurobiological measures as ‘future solutions’ to improve the

delivery of youth mental health care.

27 Chapter 2: Trajectories of social and occupational function

Iorfino, F., Hermens, D. F., Cross, S. P., Zmicerevska, N., Nichles, A.,

Badcock, C. A., Groot J., Scott, E.,& Hickie, I. B. (2018). Delineating the trajectories of social and occupational functioning of young people attending early intervention mental health services in Australia: a longitudinal study.

BMJ Open, 8(3), e020678.

28 Open Access Research Delineating the trajectories of social and occupational functioning of young people attending early intervention mental health services in Australia: a longitudinal study

Frank Iorfino,1 Daniel F Hermens,1,2 Shane, PM Cross,1 Natalia Zmicerevska,1 Alissa Nichles,1 Caro-Anne Badcock,3 Josine Groot,1 Elizabeth M Scott,1 Ian B Hickie1

To cite: Iorfino F, Hermens DF, Abstract Strengths and limitations of this study Cross S,PM, et al. Delineating Objectives Mental disorders typically emerge during the trajectories of social and adolescence and young adulthood and put young people occupational functioning ►► This study used a rich dataset of 554 participants at risk for prolonged socioeconomic difficulties.This of young people attending with between two and nine observations per per- early intervention mental study describes the longitudinal course of social and son (median=4; approximately 2200 data points) health services in Australia: a occupational functioning of young people attending up to 5 years after initial presentation and applied longitudinal study. BMJ Open primary care-based, early intervention services. a novel group-based trajectory modelling procedure 2018;8:e020678. doi:10.1136/ Design A longitudinal study of young people receiving to characterise the pattern of change in functional bmjopen-2017-020678 mental healthcare. impairment over time. Setting Data were collected between January 2005 ►► Prepublication history and ►► This study is one of the first to report on the long-term additional material for this and August 2017 from a designated primary care-based functional outcomes for young people attending primary paper are available online. To mental health service. care-based, early intervention mental health services. Its view these files, please visit Participants 554 young people (54% women) aged naturalistic design provides valuable insight into the extent the journal online (http://dx.doi. 12–32 years. of functional impairment over the course of these common org/10. 1136/bmjopen- 2017- mental disorders and raises specific questions about how Measures A systematic medical file audit collected to improve health service and individual intervention strat- 020678). clinical and functional information at predetermined time egies to monitor, target and improve these outcomes. ► Since this was a naturalistic cohort study, there may be Received 16 November 2017 intervals (ie, 3 months to 5+ years) using a clinical pro ► some factors that account for the trajectories or differenc- Revised 6 February 2018 forma. Group-based trajectory modelling (GBTM) was used es in functional outcome that were not collected in this Accepted 16 February 2018 to identify distinct trajectories of social and occupational functioning over time (median number of observations per study, such as socioeconomic status, the type and intensity of interventions an individual received. Since these factors person=4; median follow-up time=23 months). were not uniformly collected, it is difficult to make specific Results Between first clinical contact and time last conclusions about the effect of specific intervention or ser- seen, 15% of young people had reliably deteriorated, vice models on these trajectories or outcomes. 23% improved and 62% did not demonstrate substantive ►► Since this study focuses on individuals who were contin- change in function. Of the whole cohort, 69% had ually engaged in clinical care and represents 18% of the total research register, it is unclear how representative functional scores less than 70 at time last seen, indicative this sample is of the whole population presenting to these of ongoing and substantive impairment. GBTM identified services. Similarly, there is a lack of information about the six distinct functional trajectories whereby over 60% had differences between those who continually engage in care versus those who may have disengaged. 1 moderate-to-serious functional impairment at entry and Youth Mental Health Team, remained chronically impaired over time; 7% entered with Brain and Mind Centre, The serious impairment and deteriorated further; a quarter University of Sydney, Sydney, were mildly impaired at entry and functionally recovered sophisticated and more intensive individual-level and New South Wales, Australia organisational strategies may be required to achieve 2Mind and Neuroscience and only a small minority (4%) presented with serious significant and sustained functional improvements. Thompson Institute, University impairments and functionally improved over time. Not of the Sunshine Coast, Sunshine being in education, employment or training, previous Coast, QLD, Australia hospitalisation and a younger age at baseline emerged as 3 Statistical Consulting, The significant predictors of these functional trajectories. Introduction University of Sydney, Sydney, Conclusion Young people with emerging mental disorders Mental disorders consistently rank among New South Wales, Australia have significant functional impairment at presentation for the leading causes of death and disability 1–3 Correspondence to care, and for the majority, it persists over the course of worldwide. These disorders typically Mr Frank Iorfino; clinical care. In addition to providing clinical care earlier emerge during adolescence and young frank. iorfino@sydney. edu.au in the course of illness, these data suggest that more adulthood and put these young people at

Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 1 Open Access risk for prolonged socioeconomic difficulties over their significant changes in functional impairment,34 the long- lifetime, even when their mental ill health subsides or is term patterns of functional impairment among young subthreshold.4–7 There are major direct healthcare costs people engaged in primary mental healthcare remains attributed to diagnosis and treatment; however, it is their largely unknown. indirect costs linked to income loss through mortality, Understanding the changes in social and occupational disability and regular absences from education or work functioning over time in real-world clinical cohorts is that impact future income potential and have substantial crucial for guiding the development of mental health global economic consequences.8 9 The significant overlap service provisions that meet the individual needs of young between these disorders, economic inactivity and func- people with emerging mental disorders. This study exam- tional impairment reiterates the need to recognise and ines the longitudinal course of social and occupational address the common health and economic vulnerabilities functioning for a cohort of young people after their initial of these young people.10 presentation to a primary mental healthcare service. We The long-term outcomes for most major mental disor- report on the overall rate of change in social and occupa- ders often include high rates of recurrence, and slow tional functioning, and aim to determine whether there or incomplete functional recovery, even among those are distinct long-term trajectories (via modelling) of func- who may have symptomatically remitted.11–14 Long-term tioning over the course of care. follow-up studies among older adults indicate that functional impairment often persists with most people experiencing some degree of disability during the majority Methods of the long-term follow-up period,15 while it is common Participants for those within a primary care setting to spend up to Study participants were drawn from a larger longitu- 16 one-third of the long-term follow-up period off work. dinal cohort of young people (n=3087; 59% female, These patterns are also evident among young people, mean age=18.52±3.8) presenting to the Brain and Mind since most medical and psychological treatments devel- Centre’s youth mental health clinics in the Sydney suburbs oped to address depression do not consistently improve of Camperdown and Campbelltown. These clinics consist 17–19 functioning in these populations. Of the few studies of an integrated mix of primary-level services branded as that report long-term functional outcomes for young headspace35 as well as more specialised services including people, most adolescents treated for depression experi- psychiatric services. These clinics primarily attract young enced positive functional outcomes up to 3 years later; people with a range of mental health problems, including however, persistent functional impairment was common those with subthreshold and full-threshold mental disor- for those with comorbidity and recurrence of depres- ders, who may have been self-referred, referred via a 20 sion. Similarly, young people with psychosis tend to family member or friend or else via the community experience significant social disability that persists over including external general practitioner, schools or univer- time and may be indicative of the difficulty of achieving sity.29 The young people in this study were recruited to a 21 functional recovery in these groups. For many of these research register for mood, psychotic, developmental and severe mental disorders, the onset of functional deteri- other mental disorders between January 2005 and August oration tends to occur prior to the onset of illness and 2017. All young people received clinician-based case suggests that there is the capacity to address these prob- management and relevant psychological, social and/or 22 23 lems early. medical interventions over the duration of their time in Early intervention services and models of care have care, which may also include referral to/from higher tier been designed to respond to the early phases of these mental health services or hospitalisation for those whose disorders, their associated comorbidities and impair- needs exceed the capacity of the primary care services. ment, to prevent or delay the progression of illness and Individuals were included in the present study if they met 24–26 reduce the burden for those at risk. Although many the following inclusion criteria: (1) between 12 and 32 young people present with subthreshold syndromes, they years of age at the time of initial assessment; (2) were seen frequently report significant functional impairment (ie, by a clinician on at least two separate occasions. Exclu- reduced functioning in social, occupational or other sion criteria for all potential participants were: medical areas of daily life) and a high rate of disengagement from instability or lack of capacity to give informed consent education, employment or training (not in education, (as determined by a psychiatrist), history of neurolog- 24 27–29 employment or training (NEET)). Over time, func- ical disease (eg, tumour, head trauma, epilepsy), medical tional impairment tends to be associated with symptom illness known to impact cognitive and brain function (eg, remission; however, the overall level of impairment and cancer, electroconvulsive therapy in last 3 months) and/ rate of disengagement remains high compared with the or clinically evident intellectual disability and/or insuffi- 30–32 community. This is particularly the case for those with cient English to participate in the research protocol. more severe presentations who, despite receiving more intensive initial interventions, are unlikely to function- Data collection process ally recover in relatively short-term care environments.33 Trained research psychologists and medical officers While the first 12 months of care are characterised by conducted a medical file audit to collect demographic,

2 Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 Open Access

Figure 1 The frequency of the total number of time points recorded for each participant (median=4; light grey bar). clinical and functional information at predetermined time Scale (SOFAS)36 and engagement in part-time or full- intervals using a clinical pro forma (see details next). The time education, employment or training, used to deter- first available clinical assessment at the service was taken mine NEET status). The SOFAS is a clinician-rated as the baseline time point for each participant and the measure that assesses functioning on a 0–100 scale, with date of this assessment was used to determine each of the lower scores suggesting more severe impairment. The follow-up time points: 3 months, 6 months, 12 months, 2 instructions emphasise that the rater should aim to avoid years, 3 years, 4 years and 5 years. If no clinical notes were confounding the rating with clinical symptoms. available within ±1 month of the 3-month and 6-month time points, or ±3 months of the yearly time points, then Statistical analyses this particular entry was left missing. A ‘time last seen’ Statistical analyses were performed using SAS Software. entry was also used to capture final clinical information Overall changes in functioning (ie, ‘improvement’, ‘no that did not align with one of the specified time points change’ and ‘deterioration’) between baseline and time to ensure that every participant had data entered for the last seen were determined using a Reliable Change Index total time they were engaged with the clinical service. score of 10 points, and a clinically significant cut-off of 32 34 37 When data were available for a specified time point, all equal to or above 69 was used. To characterise the clinical notes from the preceding pro forma entry, up to pattern of change in functional impairment over time, we and including the current pro forma entry were used to used Group-Based Trajectory Modelling (GBTM) using a 38 complete the pro forma. procedure called PROC TRAJ. This method estimates multiple trajectory groups within the population and uses Clinical pro forma a maximum-likelihood method to calculate the proba- The clinical pro forma captures key clinical information bility of membership within each trajectory for each partic- about the current episode and specific illness course ipant. We first fit the null model (one group model), and characteristics, and an earlier version has been used in progressively increased the number of groups until we previous studies.24 29 The pro forma collects informa- reached the optimal number of trajectory groups, which tion about: (1) basic demographics (age, gender, receipt was determined using the Bayesian Information Criterion of government benefits); (2) mental health diagnoses (BIC). A higher number (ie, smaller negative number) (based on Diagnostic and Statistical Manual of Mental indicates a better balance between model complexity and Disorders (DSM-V) criteria); (3) clinical course infor- model fit. The shape of each trajectory was examined by mation (hospitalisations, childhood diagnoses); (4) modelling three parameters (linear, quadratic, cubic) and comorbidities (physical health diagnoses, such as auto- then, starting with the higher order polynomials, drop- immune, endocrine, metabolic and so on, and suicidal ping non-significant parameters from the model. If all thoughts and behaviours); and (5) functioning (assessed three parameters were not significant, the linear param- using the Social Occupational Functional Assessment eter was retained. Finally, to explore which baseline factors

Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 3 Open Access

Figure 2 The distribution of the total follow-up time for each participant in months. The bars have been shaded into quartiles (median=23 months). The majority of participants (50%) were followed up between 9 months and 49 months (ie, 4 years) after initial presentation, while 25% were followed up between 1 and 8 months, and the remaining 25% followed up between 50 months (ie, 4 years) and 126 months (ie, 10 years). were associated with each trajectory group, we used step- for entry and exclusion was set at p=0.15 and based on the wise logistic regression, which included baseline demo- likelihood ratio statistic. graphic and clinical characteristics: age, gender, receipt of government benefits, NEET status, mental health diagnosis, medical diagnosis, childhood mental health diag- Results nosis, hospitalised (ever), suicide ideation (ever), suicide Sample characteristics planning (ever) and suicide attempts (ever). An α level The sample consisted of 554 young people, 54% (297/554) were female and the mean age was 19.83 (SD=3.77). At baseline, 20% (113/554) identified as Table 1 Criteria for selecting the number of trajectories NEET, 17% (95/554) were currently receiving government benefits and the majority (78%; 423/542) were in Number of the clinical range of functional impairment (ie, SOFAS groups BIC Null model BIC change score <69). The most common primary diagnosis was 1 −8773.03 0 – depression (43%; 237/548), followed by bipolar disorder 2 −8372.01 1 401.022 (20%; 108/548), and then anxiety (18%; 99/548) with 3 −8243.31 2 128.695 comorbid mental health problems identified in 79% 4 −8215.51 3 27.802 (428/544) of participants. Physical health comorbidities were reported in 26% (142/554) of participants, 23% 5 −8207.80 4 7.710 (127/554) had previously been hospitalised due to a 6 −8166.46 5 41.339 mental health problem and 14% (75/554) had a mental 7 −8164.58 6 1.882 health or behavioural diagnosis in childhood. 8 −8162.05 7 2.528 Changes in functional impairment between baseline and time 9 −8155.80 8 6.251 last seen Bayesian Information Criterion (BIC) change presents the changes The number of follow-up time points recorded for an indi- in the BIC value as the number of trajectory group’s increases. vidual varied between 2 and 9 (median=4) (figure 1) and Large changes in BIC from 1 to 6 groups justified moving towards the number of months between baseline and time last seen the more complex model; however, changes in BIC from 7 to 9 groups were rather small and compromised the balance between was between 1 and 126 (median=23 months) (figure 2). complexity and fit. Six trajectory groups were deemed to be the The occurrence of time last seen was spread with 38% most parsimonious model. (208/554) occurring within the first 12 months after

4 Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 Open Access

Table 2 Model selection for each functional impairment trajectory group Trajectory group Parameter Model 1 Model 2 Model 3 Model 4 1 Intercept 51.61208 51.77906 51.21822 50.92215 Serious impairment— Linear −0.84458*** −0.86418*** −0.50281*** −0.49666*** deterioration Quadratic 0.02424** 0.02483** 0.00607** 0.00599** Cubic −0.00022* −0.00022 – – 2 Intercept 54.98897 54.95892 54.54367 54.75505 Serious impairment— Linear −0.19938 −0.18538 0.02760 −0.03218 chronic Quadratic 0.00966 0.00901 −0.00110 – Cubic −0.00012* −0.00012 – – 3 Intercept 41.08481 42.22558 42.03591 42.21444 Serious impairment— Linear 1.76596*** 1.26818*** 1.26797*** 1.25871*** improvement Quadratic −0.03534* −0.01123*** −0.01116*** −0.01106*** Cubic 0.00028 – – – 4 Intercept 61.20176 61.32354 61.52807 61.44346 Moderate Linear 0.09497 0.04047 0.01924 0.02027 impairment—chronic Quadratic −0.00309 −0.00039 – – Cubic 0.00003 – – – 5 Intercept 67.79146 68.08779 68.12046 68.11021 Mild impairment— Linear 0.46038*** 0.31975*** 0.32482*** 0.32399*** improvement Quadratic −0.01202* −0.00468*** −0.00478*** −0.00477*** Cubic 0.00009 – – – 6 Intercept 77.35888 77.40056 77.94966 77.93924 Slight impairment— Linear 0.19581 0.13170 0.04127 0.04153 stable Quadratic −0.00575 −0.00168 – – Cubic 0.00005 – – – Model fit BIC −8166.462 −8156.357 −8148.227 −8145.595

Parameter estimates are shown. Significant values are in bold. The first model identified that the cubic parameters for trajectories 3, 4, 5 and 6 were not significant and were thus dropped for model 2. Model 2 identified that the quadratic parameters for trajectories 4 and 6 were not significant, and that the cubic parameters for trajectories 1 and 2 were not significant and were dropped for model 3. Model 3 identified that the quadratic parameter for trajectory 2 was not significant and was dropped for model 4. The final model (model 4) had the highest Bayesian Information Criterion (BIC) and contained quadratic parameters for trajectories 1, 3 and 5 and linear parameters for trajectories 2, 4, and 6. *P<0.05. **P<0.01. ***P<0.001. baseline and 62% (346/554) occurring more than 1 year all three parameters in the model (linear, quadratic and after baseline. Overall, between baseline and time last cubic). The final model (model 4) had the highest BIC seen, 15% (79/538) had reliably deteriorated, 23% and contained quadratic parameters for trajectories 1, 3 (122/538) reliably improved and 62% (337/538) did not and 5 and linear parameters for trajectories 2, 4 and 6. reliably change, while 69% (370/538) were below the Figure 3 shows SOFAS score over a 5-year period for clinical cut-off (SOFAS<69) at time last seen. the six trajectories included in the final model (see Identifying functional impairment trajectories online supplementary figure 1 for individual-level GBTM identified that six distinct trajectories provided the trajectories for each group). Three trajectories start best balance between model complexity and model fit for out with serious functional impairment at baseline but the data (table 1). The BIC continued to increase as the differ in the type of change in functioning over time. number of groups increased; however, the BIC change The first was the second largest group of the entire from seven to nine trajectories was small and resulted in sample (29%; 158/554) and included individuals who trajectory groups with very small sample sizes that did not followed a chronic course of serious functional impair- add useful information beyond that provided by the six ment with little to no change in functioning over trajectories. Table 2 shows the model selection process for time (‘serious impairment—chronic’). The second the shape of each of the six trajectories. We started with trajectory was quadratic and included individuals who

Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 5 Open Access

Figure 3 The six distinct trajectories identified for Social Occupational Functional Assessment Scale (SOFAS) score over a 5-year period. The thickness of each line represents the sample size of that particular trajectory, relative to all others. The dotted line represents the clinical impairment cut-off, which is set at a SOFAS score of 69. Slight impairment—stable (n=24, 4%), intercept equal to 78 and linear trend over time; mild impairment—improvement (n=129, 23%), intercept equal to 68 and quadratic trend over time; moderate impairment—chronic (n=185, 33%), intercept equal to 61 and linear trend over time; serious impairment—chronic (n=158, 29%), intercept equal to 55 and linear trend over time; serious impairment—improvement (n=19, 4%), intercept equal to 42 and quadratic trend over time; serious impairment—deterioration (n=39, 7%), intercept equal to 51 and quadratic trend over time. significantly deteriorated in the first 12 months before remaining in the functional recovered population over plateauing between 12 and 60 months (‘serious impair- time (‘mild impairment—improvement’). The final ment—deterioration’), while the third trajectory was trajectory group characterised the small number of indi- also quadratic and included the small minority who viduals who were functioning well with no more than improved significantly over the first 24 months to slight impairment at baseline and whose functioning mild levels of functional impairment before slightly was stable over time (‘slight impairment—stable’). tapering off with mild to no functional impairment (‘serious impairment—improvement’). By contrast, Differentiating between functional impairment trajectories the remaining three trajectories each started out with The aim of these analyses was to identify any demographic moderate-to-mild levels of functional impairment. The and clinical differences at baseline between the trajec- first included the largest number of people across the tory groups. The serious impairment—chronic trajec- entire sample (33%; 185/554) who presented with tory was chosen as the reference group because, of the moderate impairment and followed a chronic course most impaired groups at entry; this group was the largest of moderate impairment over time (‘moderate impair- group and followed a stable/chronic trajectory over ment—chronic’). The second trajectory was quadratic time. Of the demographic and clinical variables at base- and characterised by individuals who were mildly line (table 3), NEET status, age and previous hospitalisa- impaired at baseline, but improved/functionally recov- tions emerged as the factors that differentiated trajectory ered in the first 6–12 months before tapering off and groups and were included in the reduced model. NEET

6 Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 Open Access 4 (18) 2 (10) 9 (41) 5 (24) 1 (4) 7 (29) 1 (4) 0 (0) 6 (25) 5 (21) 12 (50) 78.13 (7.56) 2 (8) 1 (4) 19 (83) 20.29 (4.23) 24 (4) Slight impairment— stable 4 (3) 4 (3) 17 (15) 16 (14) 59 (50) 18 (16) 18 (15) 28 (22) 11 (9) 26 (21) 27 (21) 59 (47) 11 (9) 70 (58) 129 (23) Mild impairment— improvement 61.39 (5.24) 68.06 (5.35) 19.75 (3.88) 20.12 (3.35) 22 (13) 29 (18) 89 (53) 40 (23) 21 (12) 51 (28) 23 (13) 14 (8) 38 (21) 36 (20) 72 (39) 27 (15) 32 (17) 103 (60) 185 (33) Moderate impairment— chronic 5 (28) 5 (29) 10 (59) 7 (37) 4 (22) 5 (26) 4 (21) 1 (5) 3 (16) 1 (5) 10 (53) 4 (21) 9 (47) 10 (56) 19 (4) Serious impairment— improvement oup (n=554) 36 (27) 38 (29) 76 (57) 48 (35) 23 (17) 41 (26) 22 (14) 10 (6) 31 (20) 24 (15) 70 (45) 35 (22) 47 (30) 77 (52) 158 (29) Serious impairment— chronic Serious impairment— deterioration 19.83 (3.77) 20.26 (4.05) 19.68 (3.70) 18.37 (4.76) 75 (14) 8 (24) 70 (13) 9 (24) 108 (20) 4 (11) 99 (18) 6 (16) 95 (17) 16 (41) 113 (20) 20 (51) 297 (54) 18 (49) 554 (100) 39 (7) Total sample Total Baseline characteristics by functional impairment trajectory gr Baseline characteristics by functional impairment trajectory

Suicide attempts, n (%) 91 (16) 7 (20) Suicide planning, n (%) 94 (17) 4 (12) Suicide ideation, n (%) 258 (47) 15 (44) Hospitalised, n (%) 127 (23) 9 (27) Childhood diagnosis, n (%) Medical diagnosis, n (%) 142 (26) 10 (26) Other, n (%) Other, Psychosis, n (%) 34 (6) 5 (13) Bipolar, n (%) Bipolar, Anxiety, n (%) Anxiety, Depression, n (%)Depression, 237 (43) 14 (37) SOFAS score, mean (SD) score, SOFAS 60.45 (9.19) 50.61 (7.25) 54.90 (5.63) 43.83 (7.05) Receiving government benefits, n (%) NEET, n (%) NEET, Female, n (%) Age, mean (SD) N (%) Column percentages are shown, except for the first row which shows the sample size for each trajectory group as a percentage of the total N. NEET, not in education, employment or training; of the total N. NEET, as a percentage row which shows the sample size for each trajectory group shown, except for the first are Column percentages Social and Occupational Functional Assessment Scale. SOFAS, Table 3 Table

Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 7 Open Access status distinguished between most trajectories, whereby that deterioration occurs early and often persists over those on the serious impairment—chronic trajectory longer periods. were less likely to be engaged in education, employment While the overall rate of change is important, this study or training compared with moderate impairment— examined the longer term patterns of change (ie, over chronic (OR 0.47, 95% CI 0.27 to 0.83, p<0.01), mild a 5-year period), which were informed by multiple time impairment—improvement (OR 0.08, 95% CI 0.03 to points. This revealed that across all levels of impairment 0.23, p<0.001) and slight impairment—stable (OR 0.09, there were high rates of chronicity with many individ- 95% CI 0.01 to 0.70, p<0.05). Regarding age, those on uals remaining at similar levels of functioning over the the serious impairment—chronic trajectory were: older course of care. For some who may have been on a path of than those on the serious impairment—improvement deterioration prior to presentation for care, maintaining trajectory (OR 0.83, 95% CI 0.71 to 0.98, p<0.05), and a consistent level of impairment may reflect a positive younger than those on the mild impairment—improve- outcome whereby engagement with care stabilised their ment trajectory (OR 1.08, 95% CI 1.00 to 1.16, p<0.05). situation or prevented further deterioration or wors- For previous hospitalisation, those on the serious impair- ening. For others, however, not being able to return to ment—chronic trajectory were more likely to have been work or education, or improve social functioning could previously hospitalised than those on the mild impair- be detrimental to their future health and socioeconomic ment—improvement trajectory (OR 2.72, 95% CI 1.39 to well-being and may reflect a lack of sufficient integrated 5.33, p<0.01). psychological and vocational interventions to directly address these outcomes.39 40 These results suggest that for those who present with Discussion mild functional impairment, functional improvement Young people with emerging mental disorders have is likely to occur relatively quickly (ie, evident from the significant functional impairment that is dynamic and quadratic trend towards improvement within the first 6 chronic over the course of clinical care. Improvement months); however, for those with more serious impair- occurs throughout the course of care; however, the rate ment, there may be the need for more intensive strate- of clinical impairment and functional deterioration gies delivered over a longer period of time to prevent or remains high for a large number of people. The results address ongoing functional impairment. Previous research also indicate that while individual trajectories may be has shown that only a small number of young people highly variable, there are distinct patterns of social and attending these primary mental health services received occupational functioning that are differentiated by the specific vocational support in the previous year,30 despite level of functioning at entry and rate of change over evidence to suggest that adjunctive interventions targeting the course of clinical care. Over 60% of the sample vocational activity can have a positive impact on functional had moderate-to-serious functional impairment at outcomes.41 42 Even among those with severe, comorbid entry and remained chronically impaired over time, disorders, early intervention combined with focused a further 7% entered with serious impairment and social recovery has demonstrated clinical utility over early deteriorated further, while approximately a quarter intervention alone for improving functional outcomes.43 of the sample were mildly impaired at entry and were Together, this reiterates the need for early intervention able to improve and functionally recover. Only a small and ongoing care that does more to directly address func- minority (4%), the youngest of the trajectory groups, tional impairment over longer periods, particularly for presented with serious impairments and were able to those who present with substantial functional impairment. functionally improve over time. This may reflect the For health services and clinicians, determining when to benefits of early intervention; however, this requires adopt these intervention strategies and for whom is crit- further investigation. These distinct trajectories high- ical. The general trajectories observed in this study are light the need for improving mental health service characterised by substantial individual variation from one and individual intervention strategies to monitor and time point to the next (see online supplementary figure directly target these problems over the course of care 1). This individual variability highlights the challenge to facilitate clinical, social and occupational recovery.10 health professionals often face when planning effective The overall rate of reliable change in this study was long-term interventions in a cohort with emerging mental comparable to studies conducted in similar cohorts health disorders. Being NEET, previous hospitalisation that were followed for relatively short-term occasions and a younger age at entry was associated with the serious of service. The rate of reliable improvement in this impairment trajectories compared with the moderate, study (23%) is consistent with a similar cohort of young mild and slight impairment trajectories; however, the people followed for approximately 6 months (25%)34 long-term predictive utility of these characteristics is still and slightly lower than an Australian national study of limited. Thus, there is a need to improve health service young people attending headspace followed for approxi- approaches to help clinicians identify and track individual mately 3 months (31%).32 Interestingly, the rate of reli- functional outcomes and trajectories over the course of able deterioration in this study was consistent with the care, so that the appropriate interventions can be strate- national study at approximately 15%, which suggests gically implemented.

8 Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 Open Access

One solution may be the development and integra- and occupational impairment does not result in lifetime tion of new and emerging technologies that use routine socioeconomic burden. outcome measurement and feedback within health services, to deliver more personalised interventions that Acknowledgements We would like to thank all the young people who have 44 45 participated in this study, and all the staff in the Youth Mental Health Team at the respond to an individual’s needs. Regular feedback to Brain and Mind Centre, past and present, who have contributed to this work. clinicians and individuals can provide important insights Contributors FI, DFH, SC and IBH designed the study, interpreted the results and about functional impairment over time as well as the drafted the manuscript. FI and C-AB conducted the statistical analyses. FI, NZ, AN, effectiveness of particular interventions for addressing JG and EMS were involved in study coordination and data collection. All authors key clinical and functional outcomes.46 These approaches contributed to and have approved the final manuscript. could also make use of assessments that aim to identify Funding This study was supported by the National Health & Medical Research underlying characteristics, such as cognition, which have Council (NHMRC) Centre of Research Excellence grant (No. 1061043). IBH was demonstrated some utility in predicting changes in func- supported by the NHMRC Research fellowship (No. 1046899). FI was supported by an Australian Postgraduate Award (APA). tioning over time.47–49 Competing interests IH has been a Commissioner in Australia’s National Mental This study has some limitations. The sample used for Health Commission since 2012. He is the Co-Director, Health and Policy at the this study focuses on individuals who were continually Brain and Mind Centre (BMC) University of Sydney. The BMC operates an early engaged in clinical care, which means that the overall intervention youth services at Camperdown under contract to headspace. IH has rate of improvement or deterioration among those previously led community-based and pharmaceutical industry-supported (Wyeth, Eli Lily, Servier, Pfizer, AstraZeneca) projects focused on the identification and who disengaged is unknown. Furthermore, the overall better management of anxiety and depression. He is a Board Member of Psychosis rate of improvement and deterioration in functioning Australia Trust and a member of Veterans Mental Health Clinical Reference group. at time last seen is imperfect given that many young He was a member of the Medical Advisory Panel for Medibank Private until October people may be still engaged in care and so time last 2017. He is the Chief Scientific Advisor to, and an equity shareholder in, Innowell. InnoWell has been formed by the University of Sydney and PwC to administer the seen may not align with a complete period of care. This $A30 million Australian Government Funded Project Synergy. Project Synergy is a is where the GBTM is beneficial over the overall rate 3-year programme for the transformation of mental health services through the use of change, since it accounts for the overall trends to of innovative technologies. EMS is the Medical Director, Young Adult Mental Health provide a clearer picture of change over time. While Unit, St Vincent’s Hospital Darlinghurst, Discipline Leader of Adult Mental Health, School of Medicine, University of Notre Dame, Research Affiliate, The University of we know that this sample represents approximately Sydney and Consultant Psychiatrist. She has received honoraria for educational 18% of the research register (554/3087), it is unclear seminars related to the clinical management of depressive disorders supported by what proportion of the whole population attending Servier and Eli-Lilly pharmaceuticals. She has participated in a national advisory these services this sample represents. Moreover, given board for the antidepressant compound Pristiq, manufactured by Pfizer. She was that the study was conducted within the context of the National Coordinator of an antidepressant trial sponsored by Servier. normal clinical service, the clinical and functional Patient consent Not required. information available for particular individuals was Ethics approval University of Sydney Human Research Ethics Committee. diverse and while the option for ‘not enough infor- Provenance and peer review Not commissioned; externally peer reviewed. mation available’ was provided to raters, it is unclear Data sharing statement No additional data available. how the type of information available impacted on the Open Access This is an Open Access article distributed in accordance with the completion of the clinical pro forma. Finally, there Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which may be other factors that account for these trajecto- permits others to distribute, remix, adapt, build upon this work non-commercially, ries or differences in functional outcome that were not and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/ collected, such as, but not limited to, socioeconomic licenses/by- nc/4. 0/ status, the type and intensity of interventions an indi- © Article author(s) (or their employer(s) unless otherwise stated in the text of the vidual received or pre-existing undiagnosed learning article) 2018. All rights reserved. No commercial use is permitted unless otherwise or developmental disorders. It is important for future expressly granted. work to determine the effectiveness of specific interventions on functional impairment trajectories and improving these outcomes to determine the reliability and validity of the medical file audit process used in References 1. Walker ER, McGee RE, Druss BG. Mortality in mental disorders and this study. global disease burden implications: a systematic review and meta- This study provides valuable insights into the long- analysis. JAMA Psychiatry 2015;72:334–41. term functional trajectories of young people engaged 2. Whiteford HA, Degenhardt L, Rehm J, et al. Global burden of disease attributable to mental and substance use disorders: in primary mental healthcare. The significant chronicity findings from the Global Burden of Disease Study 2010. The Lancet observed in this clinical cohort reiterates that ongoing 2013;382:1575–86. 3. Vos T, Barber RM, Bell B, et al. Global, regional, and national functional impairment is prevalent among young people incidence, prevalence, and years lived with disability for 301 acute with emerging mental health disorders and should be a and chronic diseases and injuries in 188 countries, 1990-2013: a primary focus of intervention, in addition to symptom- systematic analysis for the Global Burden of Disease Study 2013. Lancet 2015;386:743–800. atic improvement. The substantial variability in individ- 4. Goodman A, Joyce R, Smith JP. The long shadow cast by childhood uals’ trajectories over time highlights the need for better physical and mental problems on adult life. Proceedings of the National Academy of Sciences 2011;108:6032–7. health service and individual intervention strategies that 5. Merikangas KR, He JP, Burstein M, et al. Lifetime prevalence of monitor and target these outcomes so that early social mental disorders in U.S. adolescents: results from the National

Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 9 Open Access

Comorbidity Survey Replication--Adolescent Supplement (NCS-A). J 28. O'Dea B, Glozier N, Purcell R, et al. A cross-sectional exploration Am Acad Child Adolesc Psychiatry 2010;49:980–9. of the clinical characteristics of disengaged (NEET) young people in 6. Gibb SJ, Fergusson DM, Horwood LJ. Burden of psychiatric disorder primary mental healthcare. BMJ Open 2014;4:e006378. in young adulthood and life outcomes at age 30. Br J Psychiatry 29. Scott EM, Hermens DF, Glozier N, et al. Targeted primary care- 2010;197:122–7. based mental health services for young Australians. Med J Aust 7. Copeland WE, Wolke D, Shanahan L, et al. Adult functional outcomes 2012;196:136–40. of common childhood psychiatric problems: a prospective, 30. O'Dea B, Lee RS, McGorry PD, et al. A prospective cohort study longitudinal study. JAMA Psychiatry 2015;72:892–9. of depression course, functional disability, and NEET status in 8. Trautmann S, Rehm J, Wittchen Hans‐Ulrich. The economic costs of help-seeking young adults. Soc Psychiatry Psychiatr Epidemiol mental disorders. EMBO Rep 2016;17:1245–9. 2016;51:1395–404. 9. Bloom D, Cafiero E, Jané-Llopis E, et al. The global economic burden 31. Judd LL, Akiskal HS, Zeller PJ, et al. Psychosocial disability during of noncommunicable diseases: Program on the Global Demography the long-term course of unipolar major depressive disorder. Arch Gen of Aging, 2012. Psychiatry 2000;57:375–80. 10. Scott J, Fowler D, McGorry P, et al. Adolescents and young adults 32. Rickwood DJ, Mazzer KR, Telford NR, et al. Changes in who are not in employment, education, or training: British Medical psychological distress and psychosocial functioning in young people Journal Publishing Group, 2013. visiting headspace centres for mental health problems. Med J Aust 11. Coryell W, Scheftner W, Keller M, et al. The enduring psychosocial 2015;202:537–42. consequences of mania and depression. Am J Psychiatry 33. Cross SP, Hermens DF, Hickie IB. Treatment patterns and short-term 1993;150:720-7. outcomes in an early intervention youth mental health service. Early 12. Furukawa TA, Takeuchi H, Hiroe T, et al. Symptomatic recovery Interv Psychiatry 2016;10:88–97. and social functioning in major depression. Acta Psychiatr Scand 34. Cross SP, Scott J, Hermens DF, et al. Clinical outcomes for youth 2001;103:257–61. with subthreshold severe mental disorders accessing an early 13. Kennedy N, Abbott R, Paykel ES. Remission and recurrence intervention service. Psychiatric Services In Press. of depression in the maintenance era: long-term outcome in a 35. McGorry PD, Tanti C, Stokes R, et al. headspace: Australia's National Cambridge cohort. Psychol Med 2003;33:827–38. Youth Mental Health Foundation--where young minds come first. 14. Riihimäki KA, Vuorilehto MS, Melartin TK, et al. Five-year outcome Med J Aust 2007;187:S68. of major depressive disorder in primary health care. Psychol Med 36. Goldman HH, Skodol AE, Lave TR. Revising axis V for DSM- 2014;44:1369–79. IV: a review of measures of social functioning. Am J Psychiatry 15. Judd LL, Schettler PJ, Solomon DA, et al. Psychosocial disability and 1992;149:1148–56. work role function compared across the long-term course of bipolar 37. Falkenström F. Does psychotherapy for young adults in routine I, bipolar II and unipolar major depressive disorders. J Affect Disord practice show similar results as therapy in randomized clinical trials? 2008;108:49–58. Psychother Res 2010;20:181–92. 16. Riihimäki K, Vuorilehto M, Isometsä E. A 5-year prospective study 38. Nagin DS, Odgers CL. Group-based trajectory modeling in clinical of predictors for functional and work disability among primary care research. Annu Rev Clin Psychol 2010;6:109–38. patients with depressive disorders. Eur Psychiatry 2015;30:51–7. 39. Power E, Clarke M, Kelleher I, et al. The association between 17. Vitiello B, Rohde P, Silva S, et al. Functioning and quality of life in the economic inactivity and mental health among young people: a Treatment for Adolescents with Depression Study (TADS). J Am Acad longitudinal study of young adults who are not in employment, Child Adolesc Psychiatry 2006;45:1419–26. education or training. Ir J Psychol Med 2015;32:155–60. 18. Wagner KD, Ambrosini P, Rynn M, et al. Efficacy of sertraline in 40. Rodwell L, Romaniuk H, Nilsen W, et al. Adolescent mental health the treatment of children and adolescents with major depressive and behavioural predictors of being NEET: a prospective study of disorder: two randomized controlled trials. JAMA 2003;290:1033–41. young adults not in employment, education, or training. Psychol Med 19. Brent DA, Holder D, Kolko D, et al. A clinical psychotherapy trial for 2017:1–11. adolescent depression comparing cognitive, family, and supportive 41. Burns T, Catty J, Becker T, et al. EQOLISE Group. The effectiveness therapy. Arch Gen Psychiatry 1997;54:877–85. of supported employment for people with severe mental illness: a 20. Peters AT, Jacobs RH, Feldhaus C, et al. Trajectories of functioning randomised controlled trial. Lancet 2007;370:1146–52. into emerging adulthood following treatment for adolescent 42. Drake RE, McHugo GJ, Bebout RR, et al. A randomized clinical trial depression. J Adolesc Health 2016;58:253–9. of supported employment for inner-city patients with severe mental 21. Hodgekins J, Birchwood M, Christopher R, et al. Investigating disorders. Arch Gen Psychiatry 1999;56:627–33. trajectories of social recovery in individuals with first-episode 43. Fowler D, Hodgekins J, French P, et al. Social recovery therapy psychosis: a latent class growth analysis. Br J Psychiatry in combination with early intervention services for enhancement 2015;207:536–43. of social recovery in patients with first-episode psychosis 22. Santesteban-Echarri O, Paino M, Rice S, et al. Predictors of (SUPEREDEN3): a single-blind, randomised controlled trial. Lancet functional recovery in first-episode psychosis: A systematic Psychiatry 2018;5:41–50. review and meta-analysis of longitudinal studies. Clin Psychol Rev 44. Boswell JF, Kraus DR, Miller SD, et al. Implementing routine outcome 2017;58:59–75. monitoring in clinical practice: benefits, challenges, and solutions. 23. Fowler D, Hodgekins J, Painter M, et al. Cognitive behaviour therapy Psychother Res 2015;25:6–19. for improving social recovery in psychosis: a report from the ISREP 45. Cross SP, Hickie I. Transdiagnostic stepped care in mental health. MRC Trial Platform Study (Improving Social Recovery in Early Public Health Res Pract 2017;27. Psychosis). Psychol Med 2009;39:1627–36. 46. Carlier IV, Meuldijk D, Van Vliet IM, et al. Routine outcome monitoring 24. Hickie IB, Scott EM, Hermens DF, et al. Applying clinical staging and feedback on physical or mental health status: evidence and to young people who present for mental health care. Early Interv theory. J Eval Clin Pract 2012;18:104–10. Psychiatry 2013;7:31–43. 47. Lee RSC, Hermens DF, Scott J, et al. A transdiagnostic study of 25. McGorry PD, Hickie IB, Yung AR, et al. Clinical staging of psychiatric education, employment, and training outcomes in young people with disorders: a heuristic framework for choosing earlier, safer and more mental illness. Psychol Med 2017;47:2061–70. effective interventions. Aust N Z J Psychiatry 2006;40:616–22. 48. Iorfino F, Hickie IB, Lee RS, et al. The underlying neurobiology of 26. McGorry P, Bates T, Birchwood M. Designing youth mental health key functional domains in young people with mood and anxiety services for the 21st century: examples from Australia, Ireland and disorders: a systematic review. BMC Psychiatry 2016;16:1. the UK. Br J Psychiatry Suppl 2013;54:s30–s35. 49. Lee RS, Hermens DF, Redoblado-Hodge MA, et al. 27. Scott J, Scott EM, Hermens DF, et al. Functional impairment in Neuropsychological and socio-occupational functioning in young adolescents and young adults with emerging mood disorders. Br J psychiatric outpatients: a longitudinal investigation. PLoS One Psychiatry 2014;205:362–8. 2013;8:e58176.

10 Iorfino F,et al . BMJ Open 2018;8:e020678. doi:10.1136/bmjopen-2017-020678 Chapter 3: Suicide attempts and long-term outcomes

Iorfino, F., Hermens, D. F., Cross, S. P., Zmicerevska, N., Nichles, A., Groot

J., Guastella, A.G., Scott, E., & Hickie, I.H. (2018). Prior suicide attempts predict worse clinical and functional outcomes in young people attending a mental health service. Journal of Affective Disorders. 238: 563-569.

39 Journal of Affective Disorders 238 (2018) 563–569

Contents lists available at ScienceDirect

Journal of Aﬀective Disorders

journal homepage: www.elsevier.com/locate/jad

Research paper Prior suicide attempts predict worse clinical and functional outcomes in T young people attending a mental health service ⁎ Frank Iorﬁnoa, , Daniel F. Hermensa,b, Shane P.M. Crossa, Natalia Zmicerevskaa, Alissa Nichlesa, Josine Groota, Adam J. Guastellaa, Elizabeth M. Scotta, Ian B. Hickiea a Brain and Mind Centre, University of Sydney, 94 Mallet Street, Camperdown, Sydney, NSW, Australia b Sunshine Coast Mind and Neuroscience Thompson Institute, University of the Sunshine Coast, Birtinya, Queensland, Australia

ARTICLE INFO ABSTRACT

Keywords: Background: Mental disorders and suicidal thoughts and behaviours are common in help-seeking youth. Few Depression studies report the longitudinal associations between these phenomena and clinical and functional outcomes. This Bipolar study examined whether prior suicide attempts predict poorer outcomes in mental health service attendees. Disability Methods: Clinical and functional data from 1143 individuals (aged 12–30) attending a primary care-based Self-harm mental health service in Australia were collected over 3–60 months (median = 21 months). Odds ratios (OR) Alcohol with 95% conﬁdence intervals for the eﬀect of a prior suicide attempt on follow-up outcomes were estimated (adjusted for confounders). Results: Prior suicide attempts were common (n = 164; 14%) and prospectively associated with suicidal thoughts (OR = 1.71), suicide attempts (OR = 2.59), self-harm (OR = 1.71), an increased likelihood of being diagnosed with bipolar disorder (OR = 2.99), and the onset of an alcohol/substance use disorder (OR = 2.87). Over the course of care, no suicide attempts were reported in 1052 (92%) individuals, but 25 (2%) had recurrent attempts, and 66 (6%) had new onset of an attempt. New onset was associated with being female and previous suicidal ideation or self-harm; recurrent attempts were associated with being older and comorbid alcohol/ substance use disorder. Limitations: The cohort includes only individuals who remained in clinical contact, and the consistency of their documentation varied (across clinicians and over time). Conclusions: Young people with prior suicide attempts are vulnerable to ongoing suicidal behaviours, and poorer clinical and functional outcomes. More intensive management strategies may be needed to directly address these behaviours and the long-term risks they confer. These behaviours also emerge over the course of care among those with no previous history, which has important implications for active service-level strategies that target these behaviours for all of those who present to such services.

1. Introduction Previous studies have demonstrated that suicidal behaviours are associated with a broad range of concurrent mental and physical health Suicidal behaviour is a major contributor to death and disability problems and functional impairment (Bridge et al., 2006; Lewinsohn among young people. Lifetime estimates suggest that approximately 3 et al., 1996). More recently, a number of population-based longitudinal to 5% of people will attempt suicide and that the initial onset and peak studies have investigated the long-term health and functional outcomes of these behaviours typically occurs during adolescence and young of young people who have attempted suicide. These studies have shown adulthood (Kessler et al., 1999; Nock et al., 2013; Patton et al., 2009). that suicidal behaviours are associated with poorer relationship out- Although a history of suicide attempt is the most important predictor of comes, associated with aggression and partner violence (Kerr and subsequent attempts and suicide (Joiner et al., 2005; O'connor et al., Capaldi, 2011), the onset and persistence of mental health problems, 2013), these behaviours are also associated with a substantial economic physical health comorbidity, and poorer social and occupational out- burden among young people through loss of productivity in education comes (Brière et al., 2015; Copeland et al., 2015; Fergusson et al., 2005; and employment that often impact outcomes in later life (Shepard et al., Goldman-Mellor et al., 2014). The extent to which suicidal behaviours 2016). directly contribute to these outcomes in adulthood is contentious given

⁎ Corresponding author. E-mail address: frank.iorfi[email protected] (F. Iorfino). https://doi.org/10.1016/j.jad.2018.06.032 Received 4 March 2018; Received in revised form 7 May 2018; Accepted 12 June 2018 Available online 14 June 2018 0165-0327/ Crown Copyright © 2018 Published by Elsevier B.V. All rights reserved. fi F. Ior no et al. Journal of Affective Disorders 238 (2018) 563–569 that when other risk factors associated with suicidal behaviours and from higher tier mental health services or hospitalisation for those poorer health and functional outcomes are accounted for these re- whose needs exceed the capacity of the primary care services. lationships tend to diminish (Copeland et al., 2015; Goldman-Mellor 1143 individuals from this larger cohort were included in the pre- et al., 2014). Despite this, these studies suggest suicidal behaviours may sent study based on the following inclusion criteria: (i) between 12 and be a marker of long-term vulnerability for a range of poorer health and 30 years of age at the time of initial assessment; and (ii) were followed functional outcomes, suggesting the need for more intensive, or a up at least 3 months after baseline. Exclusion criteria for the research broader a range of, interventions strategies (King, 2017). register were: medical instability or lack of capacity to give informed Youth mental health services have been developed in Australia, consent (as determined by a psychiatrist), a history of neurological Canada, Ireland and the United Kingdom, to prevent or delay the onset disease (e.g. tumour, head trauma, epilepsy), medical illness known to and progression of these disorders and reduce secondary morbidity impact cognitive and brain function (e.g. cancer, ECT in last 3 months) (Iyer et al., 2015; McGorry et al., 2013). These services target young or clinically evident intellectual disability and/or insu fficient English to people who are at risk of a range of poorer health, economic and social participate in the research protocol. The study was approved by the outcomes including, disengagement from education or work, con- University of Sydney Human Research Ethics Committee. current alcohol or other substance misuse and/or those who experience suicidal thoughts or behaviours (STB) (McGorry et al., 2013; McGorry 2.2. Data collection process et al., 2014; McGorry et al., 2007). In Australia, they aim to identify and respond to emerging mental health problems, and relevant comorbid- Trained research psychologists and medical officers conducted a ities, not only to prevent or delay the progression of these mental dis- medical file audit to collect demographic, clinical and functional in- orders but also to reduce the longer-term impacts on functional and formation at predetermined time intervals using a clinical proforma other health outcomes (Hickie et al., 2013b). Despite many of these (see details below). The first available clinical assessment at the service young people being earlier in the course of illness, STBs are common was taken as the baseline time point for each participant and the date of even among those without an established mental health disorder this assessment was used to determine each of the follow up time (Iorfino et al., 2017; Scott et al., 2012b) and typically two to three times points: 3 months, 6 months, 12 months, 2 years, 3 years, 4 years, and higher than the general population risk (Husky et al., 2012). Similarly, 5 years. If clinical notes were not available within +/− 1 month of the the rate of suicidal attempts and suicide among individuals engaged in 3 and 6 month time points, or +/− 3 months of the yearly time points clinical care for a mental health disorder is much higher than the then this particular entry was left missing. A ‘time last seen’ entry was general population (Bostwick and Pankratz, 2000; Chen and Dilsaver, also used to capture final clinical information that did not always align 1996). Despite the higher prevalence of these behaviours among young with one of the specified time points to ensure that every participant people attending care, few studies have examined the longer-term im- had data entered for the total time they were engaged with the clinical plications of STB within a clinical sample of young people, beyond the service. When data was available for a specified time point, all clinical risk for future suicidal behaviours. Delineating the predictive sig- notes from the preceding proforma, up to and including the current nificance of these behaviours on multiple outcome domains is im- proforma entry were used to inform and then complete the current portant for prevention and intervention strategies that not only address proforma. The clinician researcher team responsible for collecting the suicidality but these other outcomes among young people. data consulted regularly with the each other and the lead clinician (IH) This study tests some of the longitudinal associations between sui- about ambiguous cases and how to deal with these consistently. These cide attempts and longer-term outcomes using a cohort of young people meetings involved case examples and scenarios that were resolved to- (aged 12–30) attending primary care-based, early intervention mental gether to determine the guiding principles for data collection. health services in Sydney, Australia. It tests the hypothesis that a suicide attempt history predicts a general vulnerability to poorer outcomes 2.3. Clinical proforma across multiple domains, despite the provision of ongoing clinical care. Specifically, we aim to: (i) determine whether a suicide attempt history The clinical proforma captures key clinical information about the at presentation is associated with poorer mental health and functional current episode and specific illness course characteristics, and an earlier outcomes, increased suicidality and greater comorbidity; and (ii) ex- version has been used in previous studies (Hickie et al., 2013b; Scott amine the course of suicide attempts after entry into care. et al., 2012a). The current proforma has been used previously (Iorfino et al., 2018) and the interrater reliability for this proforma 2. Methods currently ranges from moderate to substantial agreement (k = 0.48–0.78) depending on the variable of interest. The proforma 2.1. Participants collects information about: (i) demographics (age, sex, and receipt of government benefits); (ii) mental health diagnoses, which were based Study participants were drawn from a larger longitudinal cohort of on DSM-V criteria; (iii) clinical course information that included hos- young people (n = 3445; 59% female, mean age = 18.55 ± 3.8) re- pitalisations due to a mental health problem, previous use of medica- cruited to a research register for mood, psychotic, developmental and tions, any known childhood disorders (e.g. anxiety, attention deficit/ other mental health disorders. Young people presenting to the Brain hyperactivity disorder, autism spectrum disorders) and the presence of and Mind Centre's youth mental health clinics in the Sydney suburbs of any at-risk mental health states, such as psychosis-like experiences (e.g. Camperdown and Campbelltown were recruited to this research reg- hallucinations, delusions) and mania-like experiences (e.g. elevated ister between January 2005 and January 2018. These clinics consist of mood, racing thoughts). These at-risk mental health states were in- an integrated mix of primary-level services branded as headspace cluded to record the clinical phenomenon associated with mania and (McGorry et al., 2007; Scott et al., 2012a) as well as more specialised psychosis commonly experienced by young people yet may not ne- psychiatric clinics. These clinics primarily attract young people with a cessarily meet DSM criteria for manic or psychotic disorders; (iv) range of mental health problems, including those with sub-threshold physical health comorbidities that were more persistent conditions and full threshold mental disorders, who may have been self-referred, identified by the treating clinician, and usually meant the young person referred via a family member or friend, or else via the community in- had a specific diagnosis (e.g. autoimmune, endocrine and metabolic cluding external general practitioner, schools or university (Scott et al., disorders); (v) self-harm, suicidal thoughts and behaviours, which were 2012a). All young people received clinician-based case management primarily based on the individuals self-reported behaviours to the and relevant psychological, social and/or medical interventions over clinician. A suicide attempt was recorded when a young person had the duration of their time in care, which may also include referral to/ actually taken steps to take their own life, that is, an act with the

564 fi F. Ior no et al. Journal of Affective Disorders 238 (2018) 563–569 intention of taking their own life. The distinction between self-harm Table 1 and suicide attempts was based on the intent and was typically very Sample characteristics at baseline (n = 1143). clear in the clinical notes. If they harmed themselves via cutting, hitting Characteristic themselves, burning themselves or scratching with the intention to self- harm only and not to take their life, then this was included as self-harm Age, mean (sd) 18.76 (3.82) and not a suicide attempt. The reasons for self-harm were usually de- Female, n (%) 651 (57%) ffi “ NEET, n (%) 197 (17%) scribed as releasing emotion, distraction from di cult emotions, or to Receiving government benefits, n (%) 164 (14%) feel something”; and (vi) functioning, which was assessed at each time SOFAS score, mean (sd) 60.79 (9.64) point using the Social Occupational Functional Assessment Scale Primary diagnosis, n (%) (SOFAS) (Goldman et al., 1992) and current engagement in part-time or Depression 530 (46%) fulltime education, employment or training, used to determine Not in Bipolar disorder 143 (13%) Psychosis 68 (6%) Education, Employment or Training (NEET) status. Anxiety 228 (20%) Neurodevelopmental 57 (5%) 2.4. Statistical analyses Comorbid mental health disorder, n (%) 860 (75%) Alcohol/substance use disorder, n (%) 97 (9%) Mania-like experiences, n (%) 215 (19%) All statistical analyses were conducted in SPSS Version 24. Logistic Psychosis-like experiences, n (%) 283 (25%) regression models were used to examine the association between a Childhood diagnosis, n (%) 126 (20%) suicide attempt history at presentation and outcomes over the course of Physical health diagnosis, n (%) 238 (21%) care. These models assessed the difference in odds for each dichot- Hospitalised, n (%) 203 (18%) omous outcome (e.g. NEET, suicide ideation or bipolar onset) by each Prescribed medication, n (%) 631 (55%) ‘ ’ Self-harm, n (%) 438 (38%) level of the predictor variable; suicide attempt history at baseline ( yes Suicide ideation, n (%) 519 (45%) vs. ‘no’). These analyses aggregated outcomes across all follow up ob- Suicide attempt, n (%) 164 (14%) servations for an individual to determine the occurrence of a particular outcome at any time during follow up. Adjusted models included; sui- Note. The percentages are rounded to nearest whole numbers. NEET = Not in cide attempt history at baseline, age, sex, primary diagnosis at baseline education, employment or training. (i.e. depression, anxiety, bipolar disorder, psychosis, neurodevelopmental disorder, other), mental health comorbidity at baseline, the poorer mental health outcomes, comorbid alcohol or substance use outcome of interest at baseline, childhood diagnosis history and total disorder, and an increased likelihood of suicidal thoughts and beha- follow up time, to test whether a suicide attempt history at presentation viours (Fig. 1). Prior suicide attempts were prospectively associated had a long-term effect on outcomes over and above primary diagnosis/ with an increased likelihood for the onset of bipolar disorder – pre-existing mental health disorders and the outcome of interest at (OR = 2.99, 95% CI 1.78 5.00, p < 0.001), the onset of alcohol or – baseline. The odds ratios (ORs) and 95% confidence intervals (CIs) are substance use disorder (OR = 2.87, 95% CI 1.54 5.37, p < 0.001), and – reported. Based on the combination of suicide attempt history at psychosis-like experiences (OR = 1.78, 95% CI 1.12 2.82, p < 0.01). A baseline (‘yes’ vs. ‘no’) and suicide attempt during care (‘yes’ vs. ‘no’), suicide attempt history also predicted the persistence of suicidality over individuals were also categorised into one of four suicide attempt the course of care with both an increased likelihood for suicide ideation – course groups; never, new onset, recurrent and remitted. These sec- (OR = 1.72, 95% CI 1.14 2.60, p < 0.01) and recurrent suicide at- – ondary analyses were conducted using chi-squared tests (categorical tempts (OR = 2.59, 95% CI 1.52 4.41, p < 0.001) among these in- variables) and Analysis of Variance (ANOVA; continuous variables) to dividuals. Similarly, self-harm behaviour was higher among these in- examine demographic and clinical differences between these suicide dividuals compared to those with no suicide attempt history – attempt groups at baseline. The significance level was set at 0.05 in all (OR = 1.72, 95% CI 1.14 2.60, p < 0.01). Table 2 shows the associa- analyses. tion between prior suicide attempts and each of the follow up outcomes before and after controlling for the outcome of interest at baseline and 3. Results highlights that the associations described above remain even after controlling for the outcome of interest at baseline. In contrast the re- 3.1. Sample characteristics lationship between prior suicide attempts and a number of other outcomes, namely NEET status, physical health diagnoses and health care As shown in Table 1, the sample mean age was 18.76 years utilisation variables, dissipate when controlling for the outcome of in- (s.d. = 3.82) and 57% (n = 651) of the sample was female. The ma- terest at baseline. jority had a primary diagnosis of depression (46%; n = 530), followed by anxiety disorders (20%; n = 228) with a large majority having a 3.1.2. The course of suicide attempts comorbid mental health problem (75%; 860). In terms of suicidal Of the 1143 participants, 913 (80%) never attempted suicide, 25 thoughts and behaviours, 45% (n = 519) had a history of suicidal (2%) had recurrent attempts during care, 139 (12%) remitted from ideation, while 14% (n = 164) had a suicide attempt history at base- attempts during care, and 66 (6%) had new onset of an attempt line. Previous health care utilisation was evident in a number of in- (Table 3). We compared those who never attempted suicide to those dividuals with just over half the sample having been prescribed medi- who had a new onset during care to determine what characterised the cation in the past (55%; n = 631), and 18% (n = 203) having a emergence of this behaviour following presentation to a mental health previous hospitalisation for a mental health problem. service. At baseline, those with a new onset during care were more The median follow up time was 21 months (range, 3–127) and the likely to be female (73% vs. 56%, χ2(1) = 7.22, p = 0.007), have a median number of observations per participant was 4 (range, 2–9). history of suicide ideation (50% vs. 37%, χ2(1) = 4.57, p = 0.03) and a There were no differences in total follow up time (F(1,1141) = 0.04, history of self-harm (46% vs. 31%, χ2(1) = 5.92, p = 0.02). Finally, we p = 0.84) or number of observations (F(1,1141) = 1.53, p = 0.22) be- sought to determine what characteristics at baseline differentiated between those who did and did not have a suicide attempt history at tween those who remitted from suicidal behaviour versus those who presentation. continued to engage in suicidal behaviour. This analysis identified that those who continued to engage in suicidal behaviour were more likely 3.1.1. Associations between suicide attempt history and follow-up outcomes to be older (20.88 vs 19.22, F = 4.90, p = 0.03) and have an alcohol or Young people who had a suicide attempt history at baseline had substance use disorder (28% vs. 10%, χ2(1) = 5.66, p = 0.03).

565 ﬁ F. Ior no et al. Journal of Affective Disorders 238 (2018) 563–569

Fig. 1. NEET = Not in education employment or training. Each of the odds ratios and 95% CIs from adjusted model 2 (suicide attempt history at baseline, age, sex, primary diagnosis at baseline, comorbidity at baseline, outcome of interest at baseline, childhood diagnosis history and total follow up time) are presented for each follow up outcome, in descending order (increased odds to reduced odds). The values between 0 and 1 on the x-axis are logarithmic.

4. Discussion psychosis-like experiences) and comorbid alcohol and substance use disorder among those with a prior history of suicide attempts was in- This study highlights that young people who present to care with a dependent of these speciﬁc outcomes at baseline. This study also suicide attempt history are vulnerable to not only ongoing STB but also identiﬁed that new onset suicidal behaviour was more common than worse clinical and functional outcomes. Importantly, these poorer recurrent suicidal behaviour, and was associated with being female and outcomes were observed despite the remission of suicidal behaviour having a past history of suicide ideation and self-harm, while recurrent among over 80% (139/164) of these individuals over the course of care. suicidal behaviour was associated with being older and having an al- Of particular interest here is that the increased likelihood for ongoing cohol and substance use disorder. STB, self-harm, more severe illness (i.e. bipolar or unstable mood type, These results are consistent with previous population-based studies

Table 2 Associations between a suicide attempt history at baseline and follow up outcomes.

Outcomes No. (%) OR (95% CI)

No SA (n = 979) SA (n = 164) Adjusted model 1 Adjusted model 2

Suicidal thoughts and behaviours Suicide attempt 66 (7%) 25 (15%) 2.50 (1.47–4.25)*** 2.59 (1.52–4.41)*** Suicide ideation 376 (38%) 104 (63%) 2.97 (2.03–4.34)*** 1.72 (1.14–2.60)** Self-harm 234 (24%) 69 (42%) 2.68 (1.83–3.93)*** 1.71 (1.13–2.58)** Clinical features/diagnoses Bipolar disorder onset 73 (8%) 29 (18%) 2.98 (1.77–5.00)*** 2.99 (1.78–5.00)*** Psychosis-like experiences 239 (24%) 60 (37%) 2.16 (1.45–3.21)*** 1.78 (1.12–2.82)** Mania-like experiences 196 (20%) 55 (34%) 1.73 (1.12–2.66)* 1.58 (1.00–2.50) Depression onset 117 (12%) 11 (7%) 1.04 (0.49–2.21) 1.06 (0.50–2.27) Psychosis onset 34 (4%) 6 (4%) 0.89 (0.29–2.68) 0.90 (0.30–2.73) Anxiety onset 130 (13%) 15 (9%) 0.55 (0.31–1.00) 0.56 (0.31–1.00) Health care utilisation Hospitalised 172 (18%) 56 (34%) 2.66 (1.78–3.97)*** 1.56 (0.98–2.46) Prescribed medication 735 (75%) 142 (87%) 1.85 (1.10–3.11)* 1.14 (0.64–2.02) Alcohol and substance use Alcohol/substance use disorder onset 46 (5%) 18 (11%) 2.56 (1.38–4.76)** 2.87 (1.54–5.37)*** Functioning NEET 304 (31%) 77 (47%) 1.89 (1.31–2.74)*** 1.44 (0.94–2.21) Physical health Physical health diagnosis 266 (27%) 65 (40%) 1.54 (1.05–2.25)* 1.27 (0.74–2.16)

Note. The percentages are rounded to nearest whole numbers. SA = Suicide attempt history at baseline; NEET = Not in education, employment or training; ‘onset’ variables refer to diagnoses that were identiﬁed during the follow up period yet absent at baseline. OR = Odds ratios; * p < 0.05, ** p < 0.01, ***p < 0.001; Adjusted model 1: suicide attempt history at baseline, age, sex, primary diagnosis at baseline (depression, anxiety, bipolar disorder, psychosis, neurodevelopmental disorders, other), comorbidity at baseline, childhood diagnosis history and total follow up time; Adjusted model 2: same as model 1 with the outcome of interest at baseline included.

566 ﬁ F. Ior no et al. Journal of Affective Disorders 238 (2018) 563–569

Table 3 impairment – notably in terms of receipt of external income support Cross tab of suicide attempt history at baseline and suicide attempts during and NEET status (O'Dea et al., 2014). This work has demonstrated that care. for the majority of these young people, functional impairment tends to Suicide attempt during persist over the course of clinical care with only 23% demonstrating follow up notable improvement (Iorfino et al., 2018). This study identifies that suicidal behaviours may be another and somewhat unexpected pre- No Yes Total dictor of these undesirable longer-term outcomes, alongside other fac- Suicide attempt history at No 913 (80%) 66 (6%) 979 (86%) tors such as delayed access to care, more severe depressive or psychotic baseline symptoms, and reduced neuropsychological performance (Cotter et al., Yes 139 (12%) 25 (2%) 164 (14%) 2017; Iorfino et al., 2016; Lee et al., 2017; Lee et al., 2013; Robillard Total 1052 (92%) 89 (8%) 1143 (100%) et al., 2016; Scott et al., 2014). The course of suicidal behaviours is heterogeneous with some Note. Percentages out of the whole sample are presented in brackets and are fi rounded to nearest whole numbers. people experiencing signi cant ongoing risks that are associated with greater impairment and poorer social adjustment during adulthood fi which have identified that beyond the risk for repeated suicide at- (Goldston et al., 2016). The present study identi ed that the overall rate tempts, suicide attempts during childhood, adolescence and young of suicide attempts over the course of care was 8% (89/1143), con- adulthood were associated with an increased vulnerability for poorer sistent with other clinical studies (Chang et al., 2015; Robinson et al., long-term outcomes (Copeland et al., 2015; Goldman-Mellor et al., 2009). While these studies primarily focussed on young people with 2014). This long-term vulnerability is consistent with data which sug- psychosis, the present study included a much broader clinical cohort fi gests that individuals who have attempted suicide have a reduced life (dominated by mood disorder presentations) and identi ed that a expectancy of up to 18 years and that the majority of the excess mor- previous suicide attempt was an important indicator of suicide attempts fl tality was not due to suicide (Jokinen et al., 2017). The present study during care. This re ects the persistent vulnerability to suicidal beha- extends this prior work by focusing on a clinical cohort of young people viour and suicide that tends to endure over time (Suominen et al., who are engaged in clinical care over time. Despite receiving ongoing 2004). In addition to the increased likelihood for future attempts clinical care this youth cohort remains not only at high risk of ongoing among those who have a clinical history of STB, a larger overall pro- suicidality, but are also vulnerable to poorer long-term outcomes and portion of suicide attempts during care was comprised of those with no fl these trajectories are evident early within the clinical course. While prior history of such behaviour. This re ects not only the common these results do not demonstrate any causal relationship between prior emergence of these behaviours during adolescence and young adult- fi suicide attempts and poorer follow-up outcomes they do highlight the hood but, more speci cally, the increased vulnerability to STB among potential value of a suicide attempt history as a marker of individuals those presenting to early intervention services. As this is a high-risk who may benefit from more intensive, ongoing, and multidisciplinary cohort, and not simply a few high-risk individuals within that cohort, it care regardless of baseline diagnosis. argues for active system-level suicide prevention strategies that target The longitudinal association between a clinical history of at- STB across the whole group. The large number of individuals who re- tempting suicide and the onset of bipolar disorder (OR = 2.99), with mitted over the course of care suggests that perhaps engagement with ff related increases in psychosis-like experiences (OR = 1.78), as distinct clinical services was somewhat protective or that e ective intervention from a depressive, psychotic or anxiety disorder is intriguing. strategies were able to target these behaviours. Further examination of Importantly, this ambulatory-care based cohort is characterised by a these suicidality subgroups is needed to better understand the distinct degree of depressive symptoms and mood instability, with associated characteristics of these groups at entry and over time. functional impairment, at presentation to care. This suggests that in The results of this study should be considered in light of some addition to any other clinical features (e.g. hypomanic or manic-like limitations. This selected cohort focuses on those who continue to re- fl features, family history) that may help to identify those at specific risk ceive clinical care, presumably re ecting non-clinical factors that sup- of developing bipolar disorder (Hickie et al., 2013a), a clinical history port ongoing engagement as well as a bias towards the more severe, of attempting suicide may be relevant. Current predictive models persistent or recurrent cases. The results are based on clinical in- fi (Conus et al., 2014; Duffy et al., 2016; Vieta et al., 2018) designed to formation collected via a medical le audit, and so the consistency of assist earlier identification and the design of early intervention studies clinical and functional information varied and may be biased by the may need to include this association. documentation available. This limits the reliability of the information The longitudinal association between a history of attempting suicide collected since it is unclear whether the absence of clinical information and the onset of alcohol or other substance use disorder (OR = 2.56) about a particular outcome or parameter means the information was may be expected in such a cohort, and could reflect a shared vulner- unavailable/missing, not assessed by the clinician or simply not clini- ability to greater risk-taking, impulsive or poorly-regulated behavioural cally relevant. This increases the potential for false negatives for par- control rather than a causative mechanism. However, it does indicate ticular outcomes, since it is possible that some clinical or functional the clear possibility of developing specific secondary prevention stra- information may have been present but not recorded by a clinician or tegies to thwart this undesirable pattern of comorbidity in this selected missed in the clinical notes. Thus there may be other factors that ac- sub-group. This is particularly important given that young people who count for the relationship between suicide attempts and poorer out- were older and had an alcohol or substance use disorder were more comes that weren't collected, such as, but not limited to, socio-eco- likely to engage in recurrent suicide attempts compared to those whose nomic status, method of attempt and the severity of current mental suicidal behaviour remitted. Thus, the co-occurrence of suicidality and health symptoms. Further, it is unclear what proportion of the total alcohol or substance use problems may be an important indicator for population were excluded based on having medical instability, and so future suicidal behaviours and poorer outcomes, which reiterates the this may have resulted in more conservative estimates for outcomes need for tailored interventions that address the multitude of risks as- associated with these presentations. The current study was also not fi ff sociated with ongoing suicidal behaviours (Clarke et al., 2014; suited to make speci c conclusions about the e ectiveness of particular Wahlbeck et al., 2011). intervention strategies since little information detailed about the type In much of our current work with this cohort presenting to desig- and intensity of interventions received were not systematically col- nated early intervention services, we have highlighted the extent to lected. Future research should aim to understand whether multi- which they have already demonstrated significant functional disciplinary intervention strategies that aim to address suicidal behaviours and these broader functional outcomes have a positive impact.

567 ﬁ F. Ior no et al. Journal of Affective Disorders 238 (2018) 563–569

While the final adjusted model includes the outcome of interest at Contributors baseline and provides some insight into the directionality of the relationship between suicide attempts and these follow up outcomes, the FI, DFH and IBH designed the study and drafted the manuscript. FI, directionality of the relationship between suicide attempts and out- DFH, SC, AG and IBH interpreted the results. FI conducted the statistical comes prior to entry into care remains unclear for those variables analyses. FI, NZ, AN, JG and ES were involved in study coordination whereby both a suicide attempt and an outcome occurred before and data collection. All authors contributed to and have approved the baseline. Specifically, for some of the variables of interest, namely final manuscript. NEET, physical health diagnoses, and health care utilisation variables, it is unclear whether a suicide attempt preceded or proceeded these Acknowledgements outcomes, since these variables were all recorded at baseline. So, although the adjusted models demonstrate that suicide attempts do not We would like to thank all the young people who have participated predict some of these outcomes over the course of care when the in this study, and all the staff in the Youth Mental Health Team at the baseline outcome is controlled for, it is still unclear which occurred Brain and Mind Centre, past and present, who have contributed to this first; prior to entry into care. Future work should aim to untangle some of work. these issues to determine the nature of the relationship between suicide attempts and poorer outcomes. Role of funding This is one of few studies to examine the longer-term outcomes and course of suicide attempts among a broad clinical cohort (who largely This study was supported by the National Health & Medical share depressive symptoms and related functional impairment) of Research Council (NHMRC) Centre of Research Excellence grant (No. young people presenting to an early intervention mental health service. 1061043). Professor Ian Hickie was supported by the NHMRC Research Young people with a suicide attempt history are particularly vulnerable fellowship (No. 1046899). Frank Iorfino was supported by an to poorer clinical and functional outcomes with a clear need for in- Australian Postgraduate Award (APA). The funding source provided tensive ongoing care. The new emergence and overall course of these grant funds for the project but was not involved in data analysis, in- behaviours during care reiterates the need for active system-level terpretation, or manuscript preparation. strategies that target these behaviours and related risks at service entry and over time. These results are important for guiding the development References of health service provisions that are able to identify these vulnerable individuals so that these behaviours and the risk they confer can be Bostwick, J.M., Pankratz, V.S., 2000. Affective disorders and suicide risk: a reexamina- better addressed early in the clinical course. tion. Am. J. Psychiatry 157, 1925–1932. Bridge, J.A., Goldstein, T.R., Brent, D.A., 2006. Adolescent suicide and suicidal behavior. J. Child Psychol. Psychiatry 47, 372–394. Brière, F.N., Rohde, P., Seeley, J.R., Klein, D., Lewinsohn, P.M., 2015. Adolescent suicide Conflict of interest attempts and adult adjustment. Depress. Anxiety 32, 270–276. Chang, W.C., Chen, E.S., Hui, C.L., Chan, S.K., Lee, E.H.M., Chen, E.Y., 2015. Prevalence and risk factors for suicidal behavior in young people presenting with first-episode Professor Ian Hickie has been a Commissioner in Australia's psychosis in Hong Kong: a 3-year follow-up study. Soc. Psychiatry Psychiatr. National Mental Health Commission since 2012. He is the Co-Director, Epidemiol. 50, 219–226. Health and Policy at the Brain and Mind Centre (BMC) University of Chen, Y.-W., Dilsaver, S.C., 1996. Lifetime rates of suicide attempts among subjects with bipolar and unipolar disorders relative to subjects with other Axis I disorders. Biol. Sydney. The BMC operates an early-intervention youth services at Psychiatry 39, 896–899. Camperdown under contract to headspace. Professor Hickie has pre- Clarke, M.C., Coughlan, H., Harley, M., Connor, D., Power, E., Lynch, F., Fitzpatrick, C., Cannon, M., 2014. The impact of adolescent cannabis use, mood disorder and lack of viously led community-based and pharmaceutical industry-supported – fi education on attempted suicide in young adulthood. World Psychiatry 13, 322 323. (Wyeth, Eli Lily, Servier, P zer, AstraZeneca) projects focused on the Conus, P., Macneil, C., McGorry, P.D., 2014. Public health significance of bipolar dis- identification and better management of anxiety and depression. He is a order: implications for early intervention and prevention. Bipolar Disord. 16, – Board Member of Psychosis Australia Trust and a member of Veterans 548 556. Copeland, W.E., Wolke, D., Shanahan, L., Costello, E.J., 2015. Adult functional outcomes Mental Health Clinical Reference group. He was a member of the of common childhood psychiatric problems: a prospective, longitudinal study. JAMA Medical Advisory Panel for Medibank Private until Oct 2017. He is the Psychiatry 72, 892–899. Chief Scientific Advisor to, and an equity shareholder in, Innowell. Cotter, J., Lin, A., Drake, R.J., Thompson, A., Nelson, B., McGorry, P., Wood, S.J., Yung, A.R., 2017. Long-term employment among people at ultra-high risk for psychosis. InnoWell has been formed by the University of Sydney and PwC to Schizophr. Res. 184, 26–31. administer the $30 M Australian Government Funded Project Synergy. Duffy, A., Jones, S., Goodday, S., Bentall, R., 2016. Candidate risks indicators for bipolar Project Synergy is a three year program for the transformation of disorder: early intervention opportunities in high-risk youth. Int. J. Neuropsychopharmacol. 19. mental health services through the use of innovative technologies. A/ Fergusson, D.M., Horwood, L.J., Ridder, E.M., Beautrais, A.L., 2005. Suicidal behaviour in Prof Elizabeth Scott is the Medical Director, Young Adult Mental Health adolescence and subsequent mental health outcomes in young adulthood. Psychol. Unit, St Vincent's Hospital Darlinghurst, Discipline Leader of Adult Med. 35, 983–993. Mental Health, School of Medicine, University of Notre Dame, Research Goldman-Mellor, S.J., Caspi, A., Harrington, H., Hogan, S., Nada-Raja, S., Poulton, R., Moffitt, T.E., 2014. Suicide attempt in young people: a signal for long-term health Affiliate, The University of Sydney and Consultant Psychiatrist. She has care and social needs. JAMA Psychiatry 71, 119–127. received honoraria for educational seminars related to the clinical Goldman, H.H., Skodol, A.E., Lave, T.R., 1992. Revising axis V for DSM-IV: a review of – management of depressive disorders supported by Servier and Eli-Lilly measures of social functioning. Am. J. Psychiatry 149, 1148 1156. Goldston, D.B., Erkanli, A., Daniel, S.S., Heilbron, N., Weller, B.E., Doyle, O., 2016. pharmaceuticals. She has participated in a national advisory board for Developmental trajectories of suicidal thoughts and behaviors from adolescence the antidepressant compound Pristiq, manufactured by Pfizer. She was through adulthood. J. Am. Acad. Child Adolesc. Psychiatry 55, 400–407 e401. the National Coordinator of an antidepressant trial sponsored by Hickie, I.B., Hermens, D.F., Naismith, S.L., Guastella, A.J., Glozier, N., Scott, J., Scott, E.M., 2013a. Evaluating differential developmental trajectories to adolescent-onset Servier. All remaining authors declare no support from any organisation mood and psychotic disorders. BMC Psychiatry 13, 303. for the submitted work besides the acknowledged financial support; no Hickie, I.B., Scott, E.M., Hermens, D.F., Naismith, S.L., Guastella, A.J., Kaur, M., Sidis, A., financial relationships with any organisations that might have an in- Whitwell, B., Glozier, N., Davenport, T., Pantelis, C., Wood, S.J., McGorry, P.D., 2013b. Applying clinical staging to young people who present for mental health care. terest in the submitted work in the previous three years; no other re- Early Intervention Psychiatry 7, 31–43. lationships or activities that could appear to have influenced the sub- Husky, M.M., Olfson, M., He, J.-P., Nock, M.K., Swanson, S.A., Merikangas, K.R., 2012. mitted work. Twelve-month suicidal symptoms and use of services among adolescents: results from the National Comorbidity Survey. Psychiatr. Serv. 63, 989–996. Iorfino, F., Davenport, T.A., Ospina-Pinillos, L., Hermens, D.F., Cross, S., Burns, J., Hickie,

568 ﬁ F. Ior no et al. Journal of Affective Disorders 238 (2018) 563–569

I.B., 2017. Using new and emerging technologies to identify and respond to suicid- Nock, M.K., Green, J.G., Hwang, I., McLaughlin, K.A., Sampson, N.A., Zaslavsky, A.M., ality among help-seeking young people: a cross-sectional study. J. Med. Internet Kessler, R.C., 2013. Prevalence, correlates, and treatment of lifetime suicidal beha- Res. 19. vior among adolescents: results from the National Comorbidity Survey Replication Iorfino, F., Hermens, D.F., Shane, P., Zmicerevska, N., Nichles, A., Badcock, C.-A., Groot, Adolescent Supplement. JAMA Psychiatry 70, 300–310. J., Scott, E.M., Hickie, I.B., 2018. Delineating the trajectories of social and occupa- O'connor, R.C., Smyth, R., Ferguson, E., Ryan, C., Williams, J.M.G., 2013. Psychological tional functioning of young people attending early intervention mental health ser- processes and repeat suicidal behavior: a four-year prospective study. J. Consult. Clin. vices in Australia: a longitudinal study. BMJ Open 8, e020678. Psychol. 81, 1137. Iorfino, F., Hickie, I.B., Lee, R.S., Lagopoulos, J., Hermens, D.F., 2016. The underlying O'Dea, B., Glozier, N., Purcell, R., McGorry, P.D., Scott, J., Feilds, K.-L., Hermens, D.F., neurobiology of key functional domains in young people with mood and anxiety Buchanan, J., Scott, E.M., Yung, A.R., 2014. A cross-sectional exploration of the disorders: a systematic review. BMC Psychiatry 16, 1. clinical characteristics of disengaged (NEET) young people in primary mental Iyer, S., Boksa, P., Lal, S., Shah, J., Marandola, G., Jordan, G., Doyle, M., Joober, R., healthcare. BMJ Open 4, e006378. Malla, A., 2015. Transforming youth mental health: a Canadian perspective. Irish J. Patton, G.C., Coffey, C., Sawyer, S.M., Viner, R.M., Haller, D.M., Bose, K., Vos, T., Psychol. Med. 32, 51–60. Ferguson, J., Mathers, C.D., 2009. Global patterns of mortality in young people: a Joiner Jr, T.E., Conwell, Y., Fitzpatrick, K.K., Witte, T.K., Schmidt, N.B., Berlim, M.T., systematic analysis of population health data. Lancet 374, 881–892. Fleck, M., Rudd, M.D., 2005. Four studies on how past and current suicidality relate Robillard, R., Hermens, D.F., Lee, R.S., Jones, A., White, D., Naismith, S.L., Southan, J., even when" everything but the kitchen sink" is covaried. J. Abnorm. Psychol. 114 (2), Whitwell, B., Scott, E.M., Hickie, I.B., 2016. Sleep–wake profiles predict longitudinal 291–303. changes in manic symptoms and memory in young people with mood disorders. J. Jokinen, J., Talbäck, M., Feychting, M., Ahlbom, A., Ljung, R., 2017. Life expectancy after Sleep Res. 25, 549–555. the first suicide attempt. Acta Psychiatrica Scand. Robinson, J., Cotton, S., Conus, P., Graf Schimmelmann, B., McGorry, P., Lambert, M., Kerr, D.C., Capaldi, D.M., 2011. Young men's intimate partner violence and relationship 2009. Prevalence and predictors of suicide attempt in an incidence cohort of 661 functioning: long-term outcomes associated with suicide attempt and aggression in young people with first-episode psychosis. Aust. New Zealand J. Psychiatry 43, adolescence. Psychol. Med. 41, 759–769. 149–157. Kessler, R.C., Borges, G., Walters, E.E., 1999. Prevalence of and risk factors for lifetime Scott, E.M., Hermens, D.F., Glozier, N., Naismith, S.L., Guastella, A.J., Hickie, I.B., 2012a. suicide attempts in the National Comorbidity Survey. Arch. Gen. Psychiatry 56, Targeted primary care-based mental health services for young Australians. Med. J. 617–626. Aust. 196, 136–140. King, R.A., 2017. Adolescent Suicidal Thoughts/Behaviors Are a Marker of Long-Term Scott, E.M., Hermens, D.F., Naismith, S.L., White, D., Whitwell, B., Guastella, A.J., Vulnerability to Poor Adult Outcomes. Elsevier. Glozier, N., Hickie, I.B., 2012b. Thoughts of death or suicidal ideation are common in Lee, R., Hermens, D., Scott, J., O'Dea, B., Glozier, N., Scott, E., Hickie, I., 2017. A young people aged 12 to 30 years presenting for mental health care. BMC Psychiatry transdiagnostic study of education, employment, and training outcomes in young 12, 1. people with mental illness. Psychol. Med. 1–10. Scott, J., Scott, E.M., Hermens, D.F., Naismith, S.L., Guastella, A.J., White, D., Whitwell, Lee, R., Hermens, D.F., Redoblado-Hodge, M.A., Naismith, S.L., Porter, M.A., Kaur, M., B., Lagopoulos, J., Hickie, I.B., 2014. Functional impairment in adolescents and White, D., Scott, E.M., Hickie, I.B., 2013. Neuropsychological and socio-occupational young adults with emerging mood disorders. Br. J. Psychiatry 205, 362–368. functioning in young psychiatric outpatients: a longitudinal investigation. PLoS One Shepard, D.S., Gurewich, D., Lwin, A.K., Reed, G.A., Silverman, M.M., 2016. Suicide and 8, e58176. suicidal attempts in the United States: costs and policy implications. Suicide Life Lewinsohn, P.M., Rohde, P., Seeley, J.R., 1996. Adolescent suicidal ideation and attempts: Threat. Behav. 46, 352–362. prevalence, risk factors, and clinical implications. Clin. Psychol. 3, 25–46. Suominen, K., Isometsä, E., Suokas, J., Haukka, J., Achte, K., Lönnqvist, J., 2004. McGorry, P., Bates, T., Birchwood, M., 2013. Designing youth mental health services for Completed suicide after a suicide attempt: a 37-year follow-up study. Am. J. the 21st century: examples from Australia, Ireland and the UK. Br. J. Psychiatry 202, Psychiatry 161, 562–563. s30–s35. Vieta, E., Salagre, E., Grande, I., Carvalho, A.F., Fernandes, B.S., Berk, M., Birmaher, B., McGorry, P.D., Goldstone, S.D., Parker, A.G., Rickwood, D.J., Hickie, I.B., 2014. Cultures Tohen, M., Suppes, T., 2018. Early intervention in bipolar disorder. Am. J. Psychiatry for mental health care of young people: an Australian blueprint for reform. Lancet appi. ajp. 2017.17090972. Psychiatry 1, 559–568. Wahlbeck, K., Westman, J., Nordentoft, M., Gissler, M., Laursen, T.M., 2011. Outcomes of McGorry, P.D., Tanti, C., Stokes, R., Hickie, I.B., Carnell, K., Littlefield, L.K., Moran, J., Nordic mental health systems: life expectancy of patients with mental disorders. Br. J. 2007. headspace: Australia's National Youth Mental Health Foundation–where young Psychiatry 199, 453–458. minds come first. Med. J. Aust. 187, S68–S70.

569 Chapter 4: Responding to suicidality using technology

Iorfino, F., Davenport, T. A., Ospina-Pinillos, L., Hermens, D. F., Cross, S.,

Burns, J., & Hickie, I. B. (2017). Using new and emerging technologies to identify and respond to suicidality among help-seeking young people: a cross-sectional study. Journal of Medical Internet Research, 19(7).

47 JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

Original Paper Using New and Emerging Technologies to Identify and Respond to Suicidality Among Help-Seeking Young People: A Cross-Sectional Study

Frank Iorfino, BSc (Psych), MBMSc; Tracey A Davenport, BA (Hons), eMBA; Laura Ospina-Pinillos, MD; Daniel F Hermens, PhD; Shane Cross, BPsy (Hons), MPsych (Clinical), PhD; Jane Burns, PhD; Ian B Hickie, AM, MD, FRANZCP, FASSA Brain and Mind Centre, The University of Sydney, Sydney, Australia

Corresponding Author: Frank Iorfino, BSc (Psych), MBMSc Brain and Mind Centre The University of Sydney 94 Mallett St, Camperdown Sydney, 2050 Australia Phone: 61 02 9351 0827 Email: [email protected]

This is a corrected version. See correction statement: http://www.jmir.org/2017/10/e310/

Abstract

Background: Suicidal thoughts are common among young people presenting to face-to-face and online mental health services. The early detection and rapid response to these suicidal thoughts and other suicidal behaviors is a priority for suicide prevention and early intervention efforts internationally. Establishing how best to use new and emerging technologies to facilitate person-centered systematic assessment and early intervention for suicidality is crucial to these efforts. Objective: The aim of this study was to examine the use of a suicidality escalation protocol to respond to suicidality among help-seeking young people. Methods: A total of 232 young people in the age range of 16-25 years were recruited from either a primary mental health care service or online in the community. Each young person used the Synergy Online System and completed an initial clinical assessment online before their face-to-face or online clinical appointment. A suicidality escalation protocol was used to identify and respond to current and previous suicidal thoughts and behaviors. Results: A total of 153 young people (66%, 153/232) reported some degree of suicidality and were provided with a real-time alert online. Further levels of escalation (email or phone contact and clinical review) were initiated for the 35 young people (15%, 35/232) reporting high suicidality. Higher levels of psychological distress (P<.001) and a current alcohol or substance use problem (P=.02) predicted any level of suicidality compared with no suicidality. Furthermore, predictors of high suicidality compared with low suicidality were higher levels of psychological distress (P=.01), psychosis-like symptoms in the last 12 months (P=.01), a previous mental health problem (P=.01), and a history of suicide planning or attempts (P=.001). Conclusions: This study demonstrates the use of new and emerging technologies to facilitate the systematic assessment and detection of help-seeking young people presenting with suicidality. This protocol empowered the young person by suggesting pathways to care that were based on their current needs. The protocol also enabled an appropriate and timely response from service providers for young people reporting high suicidality that was associated with additional comorbid issues, including psychosis-like symptoms, and a history of suicide plans and attempts.

(J Med Internet Res 2017;19(7):e247) doi:10.2196/jmir.7897

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.1 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

KEYWORDS suicidal ideation; mental health; primary health care; telemedicine; health services

exposure to acutely suicidal patients in a clinician’s daily work Introduction and a lack of systematic or organizational processes that directly Suicidal thoughts are common among young people presenting address suicidal thoughts and behaviors [17]. to traditional face-to-face mental health services and engaging New and emerging technologies (eg, mobile and Internet-based with online mental health services [1,2]. Young people apps and e-tools) may be able to improve the systematic presenting to such services are also more likely to engage in assessment and response to suicidality at a service and individual suicidal behaviors (such as planning or attempts), which are level so that those at risk can receive the appropriate care sooner among the strongest predictors of completed suicide [3-6]. [18,19]. Evidence indicates that online assessments are preferred Suicidal thoughts and behaviors are also associated with and accurate for identifying suicidal thoughts and behaviors complex comorbid mental health problems [7], alcohol, or other and other sensitive information [20,21], and online screening substance use problems [8], as well as social and economic has demonstrated utility for facilitating access to treatment, difficulties that contribute to greater disability [9]. Together, especially when integrated with professional services [22-24]. this highlights the need for suicide prevention and early The integration of these technologies with traditional services intervention strategies that facilitate the early detection and is crucial, and understanding how best to utilize the benefits of rapid response to suicidality for those help-seeking young people new and emerging technologies in terms of accessibility is an [10,11]. important goal for the ongoing development of effective early This is a particularly pertinent issue given that almost half of intervention strategies that target suicidal thoughts and those who have died by suicide had contact with a primary care behaviors. provider within one month of the suicide [12], and one-quarter The aim of this study was to examine the use of a suicidality of those with depression who die by suicide are likely to have escalation protocol embedded within the Synergy Online System been in active engagement with mental health services at the (Textbox 1) that identifies and responds to suicidal thoughts time of death [13-16]. This emphasizes the challenge mental and behaviors experienced by young people (aged 16-25 years) health professionals and services face when trying to identify seeking help through primary mental health care and community and respond to those at high risk of engaging in harmful suicidal settings and to identify specific predictors of suicidality. behaviors. This may be influenced by the relatively limited

Textbox 1. Synergy Online System. The Synergy Online System is a personalized Internet-based resource designed to help people manage their physical, mental, and social wellbeing using a mixture of evidence-based apps, e-tools, and online and face-to-face services. One of the cornerstone principles of the Synergy Online System is a focus on the entire spectrum of health and well-being, from those who simply want to achieve goals to improve their daily habits, to those experiencing serious mental health problems. A key feature of the Synergy Online System is that it’s configurable (ie, can rearrange or turn on or off different components within the system as well as tailor content), which allows it to easily adapt and thus meet the needs of end users. The System aims to transform the provision of mental health services by delivering readily accessible, affordable, and equitable mental health care through an increased focus on prevention and early intervention and improving the management of mental disorders across settings.

participants were recruited for a trial of the MHeC of the Methods Synergy Online System embedded with primary mental health Participants care services (headspace). Participants in this study included young people aged 16-25 Community sample: Participants were recruited from three years who had access to the Internet and were either seeking urban, regional, and rural communities in New South Wales help through primary mental health care services (headspace) that have a number of geographical, social, and economic or online in the community for the first time. Participants were vulnerabilities (ie, Central Coast, Western Sydney, and the Far recruited into one of three groups as follows: West). Participants were recruited through targeted advertising in each of these communities (including posters and postcards Primary care sample 1: Participants were recruited from a group in local businesses, paid Facebook advertisements, and of young people presenting for the first time to headspace advertisements on organizational social media channels) from Camperdown or headspace Campbelltown (both in Sydney, March 2016 to June 2016. Young people were invited to Australia) from July 2015 to July 2016. These participants were participate in the study if they were currently living in one of recruited for the initial “proof of concept” trial of the Mental these communities and had regular access to a mobile phone health eClinic (MHeC) of the Synergy Online System. and the Internet. Primary care sample 2: Participants were recruited from a group Ethics of young people presenting for the first time to any headspace service in the Central and Eastern Sydney Primary Health The University of Sydney Human Research Ethics Committees Network (ie, Ashfield, Bondi Junction, Camperdown, Hurstville, approved these studies and all participants gave written or online and Miranda) from September 2016 to February 2017. These informed consent when they first accessed the Synergy Online System and before completing the initial clinical assessment. http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.2 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

Measures to 20 corresponds to “low current suicidal ideation”, and a score All participants were invited to complete an initial clinical of 21 to 50 corresponds to “high current suicidal ideation”. The assessment (accessed via the MHeC of the Synergy Online scale has strong internal reliability (Cronbach alpha=.91). System). Participants from primary care sample 1 were provided Lifetime suicidal thoughts and behaviors (ideation, planning, with a URL to the alpha version of the MHeC and asked to and attempts) were assessed by three questions from the Youth complete the initial clinical assessment online before either a Risk Behaviors Survey [29,30]; (1) “Have you ever seriously video visit or face-to-face appointment with a clinician. thought about killing yourself?” (2) (1) “Have you ever seriously Participants from primary care sample 2 were provided with a thought about killing yourself?” (2) “Have you ever made a URL to the beta version of the MHeC (with the video visit plan about how you would kill yourself?” and (3) “How many “turned off”) and asked to complete the initial assessment before times have you actually tried to kill yourself?”. their scheduled face-to-face appointment with a clinician. Functioning Participants from the community sample either navigated An item from the Brief Disability Questionnaire (BDQ) was themselves to the MHeC or were automatically directed (via an used to assess participant’s inability to carry out daily tasks over e-tool embedded within the Synergy Online System) to the beta the previous month [31]. Specifically, participants were asked version of the MHeC (with the video visit “turned on”) if they “Over the past month, how many days in total were you unable were expressing psychological distress. For all participants using to carry out your usual daily activities fully?” This enabled a the MHeC, a “need help now” button was always displayed to calculation of “days out of role in the past month.” provide the details of relevant emergency and helpline services for those who sought immediate help. Alcohol and Substance Use The initial clinical assessment assesses a range of mental health Two questions about alcohol and substance use were used to outcomes, as well as comorbid and associated risk factors. Being assess the presence of a current comorbid alcohol or substance administered online and using smart skips, the full assessment use problem. Specifically, participants were asked “Have you takes approximately 45 min to complete (median, 47.5 min) recently thought that you should cut down on alcohol or other and includes 14 modules (in the following order): demographics; addictive drugs?” (derived from the CAGE questionnaire; [32]) current education and employment participation; mental health and “Have you recently had a friend, relative or doctor suggest concerns; self-harm and suicidal behaviors; tobacco, alcohol, that you should cut down on alcohol or other addictive drugs?” and other substance use; physical activity; sleep-wake behaviors; (derived from the Alcohol Use Disorders Identification Test; lifetime disorders; physical and mental health history; cognition; [33]). Participants who answered “no” to one or both of these eating behaviors and body image; social connectedness; and questions were categorized as “no problem”, and participants puberty. Participants completed all modules. For the purposes who answered “yes” to both questions were categorized as of this study, the following measures were specifically selected “likely problem” [34]. and included for analysis. Statistical Analyses Demographics All statistical analyses were performed using Statistical Package Participants’ age, gender, highest level of education, and current for the Social Sciences (SPSS 22.0 for Windows). Group education, employment, and training status (used to determine differences between the three sample groups (primary care not in education, employment or training [NEET] status). sample 1, primary care sample 2, and community sample) were assessed using the Kruskal-Wallis H test for continuous variables Mental Health and the chi-square test for categorical variables. The sample Current psychological distress was assessed using the Kessler-10 was then split by suicidality group (see Textbox 2; “no (K10) questionnaire [25] that is a well-validated measure of suicidality”, “low suicidality”, and “high suicidality”) to assess general psychological symptoms and distress widely used in group differences using the Kruskal-Wallis H test for continuous adult and adolescent populations in both clinical and community variables and the chi-square test for categorical variables. To settings. Hypomania-like symptoms over the last 12 months examine the independent predictors of suicidality, two separate were assessed using a screener derived from the Altman logistic regressions were conducted. The first model compares self-rating scale [26]. Psychosis-like symptoms over the last 12 the “no suicidality” group with the “any suicidality” group (low months were assessed using a screener derived from Community and high suicidality groups combined). The second model Assessment of Psychotic Experiences-Positive Symptoms scale compares the “low suicidality” and “high suicidality” groups. [27]. Participants were also asked “Have you ever experienced For both models, variables were entered using a forward a major mental health or behavioral problem that has affected forced-entry method with demographic variables (age, gender, your everyday life?” and this was used as a proxy for a previous education, and NEET status) entered in the first block, current mental health problem. mental health variables (K10, hypomania-like symptoms, psychosis-like symptoms, and alcohol or substance use) entered Suicidality in the second block, mental health history variables (previous The Suicide Ideation Attributes Scale (SIDAS) is a 5-item scale mental health problem, suicide plans or attempts history) entered assessing suicidal ideation over the past month [28]. The scale in the third block, and functioning (days out of role) entered in assesses frequency, controllability, closeness to attempt, distress, the final block. To control for sample groups in the analyses, and interference with daily activities on a 10-point Likert scale. “sample” was also entered in the final block. Only models with A score of 0 corresponds to “no current ideation”, a score of 1 http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.3 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al nonsignificant Hosmer-Lemeshow goodness of fit tests were included.

Textbox 2. Suicidality escalation protocol. The suicidality escalation protocol involves multiple levels of action, dependent on the participants’ responses to the initial assessment (Figure 1). Every young person completes the initial clinical assessment, and at the end of the suicidality module the digitally smart algorithms assess current and past suicidality. The algorithm assigns them to one of the three groups: “no suicidality” (SIDAS score of 0 and no lifetime suicidal behaviors), “low suicidality” (SIDAS score of 1-20 and/or lifetime suicidal behaviors), and “high suicidality” (SIDAS score of 21 to 50). For the “no suicidality” group, no action is taken. For those in the “low suicidality” or “high suicidality” groups, an automatic real-time alert is immediately presented on the young person’s screen. The alert displays information regarding both crisis and non-crisis services so the young person can access immediate support if needed. For those in the “high suicidality” group, two additional actions are initiated. First, a notification is sent to the clinical research team who initiate email contact with the participant within 24 h. This email aims to provide further information that encourages the young person to seek help and requests they inform the clinical research team how they are going by replying to the email or calling. Second, for those currently in contact with a service, the young person’s data is forwarded to the clinical service for review, and a decision is made regarding further follow-up or escalation. Further follow-up or escalation involves one or more of the following: contact over the phone, rescheduling the young person’s appointment or an online “video visit” within the subsequent 72 h. Importantly, this suicidality escalation protocol is designed to respond in real-time to the suicidality expressed by the young person and is not used to determine suicide risk. Formal suicide risk is determined by a health professional or a multidisciplinary team after making contact with the young person and reviewing the young person’s data.

Figure 1. Suicidality escalation protocol.

Results Suicidality Escalation in Primary Care—A Proof of Concept Sample Characteristics The first use of the suicidality escalation protocol in a primary The demographic and behavioral characteristics for each sample mental health care setting occurred at headspace Camperdown are presented in Tables 1 and 2. A total of 232 participants were and headspace Campbelltown and was rolled out entirely by included in the analyses (95 from primary care sample 1, 105 the clinical research team. Of the entire primary care sample 1, from primary care sample 2, and 32 from the community 33% (31/95) were identified as “no suicidality” and so no action sample). The mean age of the entire sample was 20.44 years was initiated, 51% (49/95) were identified as “low suicidality” (standard deviation [SD]=2.59; median=21 years), 69% and were presented with a real-time alert only, and 16% (15/95) (160/232) were female, and 37% (87/232) were classified as were identified as “high suicidality,” which initiated the NEET. Across all three samples, the mean K10 score was in real-time alert and an additional two escalation actions. All 15 the severe range of psychological distress (x̅=28.99, SD=8.86; individuals were contacted via email by the clinical research median=30); 17% (40/232) reported no psychological distress, team and had their data reviewed. Of these 15, 7 had their entry 13% (31/232) reported mild psychological distress, 19% into clinical care escalated (ie, their initial clinical assessment (43/232) reported moderate psychological distress, and 51% appointment was brought forward). Clinicians reported that the (118/232) reported severe psychological distress. The mean decision to escalate an individual was influenced by the SIDAS score was 8.25 (SD=11.52; median=2); 39% (90/232) following: (1) concerns over specific suicidal ideation attributes reported no current suicidal ideation, 46% (107/232) reported such as little of control over suicidal thoughts (5/7 participants) low current suicidal ideation, and 15% (35/232) reported high and closeness to making an attempt (5/7 participants), (2) current suicidal ideation. The only statistically significant concerns over the presence of hypomania or psychosis-like differences identified between the three sample groups were for symptoms (1/7 participants), (3) recent plans to make an attempt “days out of role” over the past month (χ2 =16.2, P<.001). that were identified upon follow-up (1/7 participants), (4) few 2 protective factors identified upon follow-up (1/7 participants), (5) few protective factors identified at follow-up (1/7 http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.4 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al participants), and (6) recent self-harm (1/7 participants). All 7 brought forward due to most, or all of the above, factors being participants were escalated due to one or more of these factors absent or because their clinical appointment was scheduled being present, and the clinician decided that their initial within a few days (range=0-5 days), which was deemed to be appointment for care was too long to wait. The remaining 8 sufficient by the reviewing clinician. participants did not have the initial assessment appointment Table 1. Demographic characteristics by sample group (N=232). Characteristics Primary care 1 Primary care 2 Community P value (n=95) (n=105) (n=32)

Age, mean (SDa) 20.39 (2.56) 20.41 (2.53) 20.66 (2.90) .88 Gender, n (%) .76 Female 68 (72) 71 (68) 21 (66) Male 27 (28) 34 (32) 11 (34)

Educationb, n (%) .70 Secondary 44 (46) 49 (48) 16 (55) Tertiary 51 (54) 54 (52) 13 (45)

NEETc status, n (%) .12 Non-NEET 62 (65) 59 (56) 24 (75) NEET 33 (35) 46 (44) 8 (25) aSD: standard deviation. b“no formal education” and “primary education” groups were left out due to insufficient cell counts (n=5 cases missing). cNEET: not in education, employment or training. clinical research team and had their data forwarded for review Suicidality Escalation Scaled Up for Use in Primary to the clinical service responsible so that specific service Care protocols could be initiated. The suicidality escalation protocol was scaled up and rolled out Of the entire community sample, 37.5% (12/32) young people across all headspace services in the Central and Eastern Sydney were identified as “no suicidality” and so no action was initiated, Primary Health Network. Of the entire primary care sample 2, 37.5% (12/32) were identified as “low suicidality” and were 34% (36/105) were identified as “no suicidality” and so no presented with a real-time alert, and 25% (8/32) were identified action was initiated, 55% (57/105) were identified as “low as “high suicidality”, which initiated the real-time alert and an suicidality” and were presented with a real-time alert only, and additional two escalation actions. All 8 individuals were 11% (12/105) were identified as “high suicidality,” which contacted via email by the clinical research team, and had their initiated the real-time alert and an additional two escalation data reviewed. actions. All 12 individuals were contacted via email by the

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.5 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

Table 2. Behavioral characteristics by sample group (N=232). Characteristics Primary care 1 Primary care 2 Community P value (n=95) (n=105) (n=32)

K10a score, mean (SDb) 29.28 (8.16) 29.75 (8.28) 25.59 (11.76) .11 K10 category, n (%) .08 No 13 (14) 15 (14) 12 (38) Mild 15 (16) 13 (13) 3 (9) Moderate 18 (19) 20 (19) 5 (16) Severe 49 (51) 57 (54) 12 (37)

SIDASc score, mean (SD) 7.93 (11.50) 7.52 (9.71) 11.59 (16.05) .87 SIDAS category, n (%) .34 No ideation 37 (39) 40 (38) 13 (41) Low ideation 43 (45) 53 (51) 11 (34) High ideation 15 (16) 12 (11) 8 (25) Hypomania-like symptoms, last 12 months, n (%) .06 No 68 (72) 88 (84) 22 (69) Yes 27 (28) 17 (16) 10 (31) Psychosis-like symptoms, last 12 months, n (%) .51 No 65 (68) 71 (68) 25 (78) Yes 30 (32) 34 (32) 7 (22) Days out of role in past month, mean (SD) 7.53 (7.22) 8.04 (8.47) 2.34 (2.66) <.001 Alcohol and substance use, current, n (%) .22 No problem 60 (73) 87 (83) 25 (78) Likely problem 26 (27) 18 (17) 7 (22) Previous mental health problem, ever, n (%) .24 No 27 (28) 31 (29) 14 (44) Yes 68 (72) 74 (71) 18 (56) Suicide plans or attempts, ever, n (%) .18 No 67 (71) 66 (63) 17 (53) Yes 28 (29) 39 (37) 15 (47) aK10: Kessler-10. bSD: standard deviation. cSIDAS: Suicide Ideation Attributes Scale. however, significant differences were identified between the Predictors of Suicidality three suicidality groups for psychological distress (χ2 =48.5, The overall sample was split according to “no suicidality”, “low 2 P<.001), hypomania-like symptoms in the last 12 months suicidality”, and “high suicidality” to examine the demographic 2 and behavioral differences between these groups (Tables 3 and (χ 2=12.9, P=.002), psychosis-like symptoms in the last 12 2 2 4). No differences were identified between the sample groups months (χ 2=29.2, P<.001), alcohol or substance use (χ 2=8.3, (P=.33) or for the demographic variables; age (P=.08), gender P=.02), and previous mental health problem (χ2 =15.8, P<.001). (P=.74), or NEET status (P=.29); however, significant 2 Significant differences between the “low suicidality” and “high differences were identified for highest level of education suicidality” groups were also identified for history of suicide (χ2 =8.6, P=.01). In terms of behavioral characteristics, no 2 plans or attempts (χ2 =22.3, P<.001). differences were identified for days out of role (P=.09); 1

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.6 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

Table 3. Demographic characteristics by suicidality group (N=232). Characteristics Suicidality P value No Low High (n=79) (n=118) (n=35) Sample, n (%) .33 Primary care 1 31 (39) 49 (42) 15 (43) Primary care 2 36 (46) 57 (48) 12 (34) Community 12 (15) 12 (10) 8 (23)

Age, mean (SD)a 20.32 (2.66) 20.75 (2.52) 19.66 (2.53) .08 Gender, n (%) .74 Female 57 (72) 79 (67) 24 (69) Male 22 (28) 39 (33) 11 (31)

Educationb, n (%) .01 Secondary 32 (41) 53 (46) 24 (71) Tertiary 46 (59) 62 (54) 10 (29)

NEETc status, n (%) .29 Non-NEET 48 (61) 71 (60) 26 (74) NEET 31 (39) 47 (40) 9 (26) aSD: standard deviation. b“no formal education” and “primary education” groups were left out due to insufficient cell counts (n=5 cases missing). CNEET: not in education, employment or training.

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.7 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

Table 4. Behavioral characteristics by suicidality group (N=232). Characteristics Suicidality P value No Low High (n=79) (n=118) (n=35) Sample, n (%) .33 Primary care 1 31 (39) 49 (42) 15 (43) Primary care 2 36 (46) 57 (48) 12 (34) Community 12 (15) 12 (10) 8 (23)

K10a score, mean (SDb) 24.30 (8.04) 29.92 (8.22) 36.43 (6.41) <.001 K10 category, n (%) <.001 No 26 (33) 14 (12) 0 (0) Mild 14 (18) 16 (14) 1 (3) Moderate 17 (21) 22 (18) 4 (11) Severe 22 (28) 66 (56) 30 (86) Hypomania-like symptoms, last 12 months, n (%) .002 No 67 (85) 92 (78) 19 (54) Yes 12 (15) 26 (22) 16 (46) Psychosis-like symptoms, last 12 months, n (%) <.001 No 67 (85) 82 (70) 12 (34) Yes 12 (15) 36 (30) 23 (66) Days out of role in past month, mean (SD) 6.22 (7.63) 7.18 (7.69) 8.46 (7.35) .09 Alcohol and substance use, current, n (%) .02 No problem 70 (89) 87 (74) 24 (69) Likely problem 9 (11) 31 (26) 11 (31) Previous mental health problem, ever, n (%) <.001 No 34 (43) 36 (30) 2 (6) Yes 45 (57) 82 (70) 33 (94)

Suicide plans or attempts, ever, n (%) <.001c No 79 (100) 67 (57) 4 (11) Yes 0 (0) 51 (43) 31 (89) aK10: Kessler-10. bSD: standard deviation. cThis P value refers to the 2x2 comparison between the low and high groups. By definition the “no suicidality” group has 0 “yes” responses. Further analyses using logistic regression were conducted to 2 2 compared with “no suicidality” (χ 12=57.7, P<.001, R =0.22). (1) identify predictors of “no suicidality” compared with “any Model 2 identified that higher psychological distress, any suicidality” (low and high suicidality groups combined) (Model psychosis-like symptoms in the last 12 months, a previous 1, Table 5), and (2) to identify predictors of “low suicidality” mental health problem, and a history of suicide plans or attempts compared with “high suicidality” (Model 2, Table 5). Model 1 were all predictors of “high suicidality” compared with “low identified that higher psychological distress and a current alcohol suicidality” (χ2 =67.0, P<.001, R2=0.36). or substance use problem were predictors of “any suicidality” 13

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.8 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

Table 5. Logistic regression models showing predictors of suicidality (N=232).

No suicidality versus any suicidalitya Low suicidality versus high suicidalityb

Beta (SEc) OR (95% CI) P value Beta (SE) OR (95% CI) P value Age .05 (0.07) 1.05 (0.91-1.22) .47 −.17 (0.15) 0.84 (0.63-1.13) .26 Gender Female 1.00 1.00 Male .65 (0.38) 1.92 (0.92-4.02) .08 .66 (0.64) 1.94 (0.55-6.81) .30

Educationd Secondary 1.00 1.00 Tertiary −.22 (0.39) 0.81 (0.37-1.75) .58 −.91 (0.69) 0.40 (0.10-1.57) .19

NEETe status NEET 1.00 1.00 Non-NEET .30 (0.35) 1.35 (0.68-2.67) .39 .92 (0.68) 2.50 (0.66-9.51) .18

K10f score .11 (0.02) 1.12 (1.07-1.17) <.001 .11 (0.04) 1.12 (1.03-1.21) .01 Hypomania-like symptoms, last 12 months No 1.00 1.00 Yes .14 (0.45) 1.16 (0.48-2.76) .75 .41 (0.60) 1.50 (0.47-4.84) .50 Psychosis-like symptoms, last 12 months No 1.00 1.00 Yes .80 (0.41) 2.22 (1.00-4.95) .05 1.54 (0.58) 4.68 (1.51-14.53) .01 Alcohol and substance use, current No problem 1.00 1.00 Likely problem 1.04 (0.46) 2.84 (1.15-7.05) .02 −.12 (0.63) 0.89 (0.26-3.04) .85 Previous mental health problem No 1.00 1.00 Yes .42 (0.35) 1.52 (0.77-3.03) .23 2.43 (0.99) 11.34 (1.64-78.30) .01 Suicide plans or attempts, ever No 1.00

Yes N/Ag N/A N/A 2.34 (0.70) 10.41 (2.65-40.83) .001 Days out of role, past month −.02 (0.02) 0.98 (0.94-1.03) .42 .01 (0.04) 1.00 (0.92-1.09) .93 Sample Community 1.00 1.00 Primary care 1 −.22 (.57) 0.80 (0.26-2.43) .69 .25 (0.86) 1.28 (0.24-6.83) .77 Primary care 2 −.27 (.57) 0.76 (0.25-2.30) .63 −.51 (0.92) 0.60 (0.10-3.67) .58 a 2 2 Model 1 : R =0.22 (Cox and Snell), 0.31 (Nagelkerke). Model χ 12=57.7, P<.001. b 2 2 Model 2 : R =0.36 (Cox and Snell), 0.55 (Nagelkerke). Model χ 13=67.0, P<.001. cSE: standard error. d“no formal education” and “primary education” groups were left out due to insufficient cell counts (n=5 cases missing) . eNEET: not in education, employment or training. fK10: Kessler-10. gN/A: Not applicable, this comparison is invalid since the “no suicidality” group, by definition, has no history of suicide plans or attempts and therefore was left out of the model.

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.9 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

as reported in this paper is critical in reaching the high numbers Discussion of at-risk youth in the community who are not presenting to Principal Findings traditional face-to-face services. Importantly, with the rapid increase in new and emerging technologies for mental health We identified that two-thirds of help-seeking young people care, there is a significant need for effective suicide escalation reported some degree of suicidality, and the protocol provided protocols that can appropriately and efficiently manage risk. A these young people with a real-time alert online. Further levels real-time mapping system to (local) mental health services might of escalation (email or phone contact and clinical review) were be useful for those in the community who seek help online to initiated for the 15% (35/232) of young people who reported ensure the system effectively facilitates help-seeking behavior, high suicidality. Higher levels of psychological distress and a which is a crucial unresolved issue for online assessment and current alcohol or substance use problem predicted any level feedback systems [40-42]. Similarly, further follow-up through of suicidality (compared with no suicidality). In addition to partnerships with specific local or national suicide prevention higher levels of psychological distress, psychosis-like symptoms organizations may be needed to increase help-seeking behavior in the last 12 months, a previous mental health problem, and a for those identified as at-risk or in need of care in the history of suicide plans or attempts were specific predictors of community. high suicidality (compared with low suicidality). These results support the use of new and emerging technologies to facilitate Psychological distress differentiated between each level of the systematic assessment and detection of young people suicidality identified, which is consistent with the established experiencing suicidal thoughts with additional comorbidities relationship between distress and suicidality [43,44]. The only and enable an appropriate and timely response from service other predictor that differentiated between no suicidality and providers. any suicidality was a current alcohol or substance use problem. This reflects the common relationship between alcohol and The use of the suicidality escalation protocol of the Synergy substance use and suicidal thoughts and behaviors, particularly Online System as an adjunct to traditional primary mental health among young people with mental health problems [45]. Young care services assisted clinical decision-making about suicide people reporting high suicidality were also more likely to report risk and the need for care among those young people reporting psychosis-like symptoms in the last 12 months, a previous higher levels of suicidality. Of the young people in primary mental health problem, and a history of suicide plans or attempts. care, 13.5% (27/200) had their case escalated to clinical review Together, this confirms the significant comorbidity that by a clinician or clinical team before their entry into care. help-seeking young people initially present with and reiterates Importantly, none of these young people were referred to crisis the need for services to be equipped to respond to the differing services but instead had their entry into care facilitated due to individual needs a young person has when they first present to a clinically perceived higher need for immediate care. This care. escalation process ensured that individuals presenting to primary care services with increased suicidality were not delayed by a The ongoing development of the Synergy Online System would service waitlist, which commonly arises from a mismatch benefit from employing methodologies that utilize longitudinal between service demand and capacity [35]. Instead, the Synergy outcomes to improve the existing algorithms accuracy for Online System was able to deploy many immediate actions to identifying individual cases of suicidality that should be ensure the suicidality risk is addressed in a timely and efficient escalated and followed up immediately by a clinician and manner. The use of this System has already had major service. Machine learning methodologies are increasingly used implications on actual health service practices for the youth in psychiatric research as they facilitate individual-level mental health services that have adopted Synergy; specifically, prediction of unseen observations, which makes them suitable improving patient and workforce management through for the development of clinically useful digital tools [46]. Recent systematic assessment, automatic escalation of an individual’s evidence has demonstrated the use of these algorithms to utilize data, and assisting clinical team review and decision-making clinical and demographic variables to predict suicide attempters processes. among a group of mood disorder patients with accuracy comparable with most breast cancer prediction algorithms Importantly, the results here also highlight the benefits of [47,48], whereas another study demonstrated the utility of such offering online services to young people by allowing mental algorithms to differentiate between suicidal and nonsuicidal health care and the service to be brought to the young person patients [49]. Employing these approaches could improve the when they need it, wherever they live, rather than relying on personalization of care beyond simple cut-off scores and include young people to present initially to a face-to-face service which key risk factors specific to a particular individual. Similar has many barriers to overcome [36]. Notably, there were approaches have been employed by Facebook who have comparable levels of suicidal ideation in the community sample developed an online tool that uses machine learning to identify compared with those presenting to primary mental health care. users at risk of suicide by assessing their posts and comments These young people may never have presented to a face-to-face and provides the user with a number of options for how to get service either because of common barriers to help-seeking or help [50]. These semiautomated approaches require rigorous because a service was not available locally [37,38]. The use of evaluation and validation using qualitative person-centered the online service meant that a service could “come to them” approaches such as user acceptance testing, in addition to more when they needed it and in a manner that is preferable to some traditional quantitative methods to determine whether they are young people [39]. The use of new and emerging technologies appropriate and effective. This is important for the development http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.10 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al of clinically useful and scalable suicide prevention and early the relatively small sample size of the community sample, intervention efforts that are integrated with existing services compared with the two primary care sample groups, means that and practices. the sample characteristics were somewhat biased toward the primary care groups and limits the generalizability of these Limitations results to young people in the community who seek help online. For the future development of the protocol, some limitations Finally, the use of the K10 as a measure of general psychological need to be addressed. First, the initiation of the suicidality distress may be limited primarily to depression and anxiety escalation protocol is dependent on when the young person symptoms and less useful for other mental health problems completes the online assessment. So young people at-risk who common in adolescence. don’t complete the online assessment immediately cannot be identified and spend a longer period under distress and not in Conclusions care. Second, the outcome for those who had their entry into This study contributes to the research and knowledge about the care escalated is unclear, so it is difficult to determine the impact use of new and emerging technologies to identify and respond of the suicidality escalation protocol on their clinical outcome. to increased suicidality among help-seeking young people. This was beyond the scope of this particular study, but it is an Young people with increased suicidality were more likely to important focus for future research to establish the long term present with a number of comorbid issues including impact of this protocol on engagement with services and clinical psychosis-like symptoms and a history of plans or attempts, trajectory. Another key focus for this work would be to which emphasizes the need for these young people to receive determine whether the protocol missed individuals who would appropriate and timely care. become high risk or later engage in suicidal behaviors. Third,

Acknowledgments Frank Iorfino is supported by an Australian Postgraduate Award (APA). Data collected for primary care sample 1 was supported by the Young and Well CRC (2011-16), and data collected for primary care sample 2 and the community sample was supported by a Commonwealth Government of Australia investment in Project Synergy — the design, build, and evaluation of an innovative system of care that embeds new and emerging technology into Australia’s youth mental health services (2014-16). This project was carried out through a partnership between the Young and Well CRC and The University of Sydney’s Brain and Mind Centre. When Young and Well CRC closed on June 30, 2016, Project Synergy was novated to The University of Sydney. Thanks to the entire Youth Mental Health and Technology Team at the Brain and Mind Centre.

Conflicts of Interest Professor Ian Hickie has been a commissioner in Australia’s National Mental Health Commission since 2012. He is the co-director, Health and Policy at The University of Sydney’s Brain and Mind Centre. The Brain and Mind Centre operates an early-intervention youth services at Camperdown under contract to headspace. Professor Hickie has previously led community-based and pharmaceutical industry-supported (Wyeth, Eli Lily, Servier, Pfizer, AstraZeneca) projects focused on the identification and better management of anxiety and depression. He is a member of the Medical Advisory Panel for Medibank Private, a board member of Psychosis Australia Trust, and a member of Veterans Mental Health Clinical Reference group. He is the chief scientific advisor to, and an equity shareholder in, Innowell. Innowell has been formed by The University of Sydney and PricewaterhouseCoopers (PwC) to deliver the $30m Australian Government-funded “Project Synergy.” Project Synergy is a 3- year program for the transformation of mental health services through the use of new and innovative technologies. Professor Jane Burns is the CEO of, and an equity shareholder in, Innowell.

References 1. Hemelrijk E, van BW, Donker T, van SA, Kerkhof A. Internet-based screening for suicidal ideation in common mental disorders. Crisis 2012;33(4):215-221. [doi: 10.1027/0227-5910/a000142] [Medline: 22713975] 2. Scott EM, Hermens DF, Naismith SL, White D, Whitwell B, Guastella AJ, et al. Thoughts of death or suicidal ideation are common in young people aged 12 to 30 years presenting for mental health care. BMC Psychiatry 2012 Dec 26;12:234 [FREE Full text] [doi: 10.1186/1471-244X-12-234] [Medline: 23268688] 3. Shaffer D, Pfeffer CR. Practice parameter for the assessment and treatment of children and adolescents with suicidal behavior. Journal of the American Academy of Child & Adolescent Psychiatry 2001;40(7):24. [Medline: 11434483] 4. Gould MS, Greenberg T, Velting DM, Shaffer D. Youth suicide risk and preventive interventions: a review of the past 10 years. J Am Acad Child Adolesc Psychiatry 2003 Apr;42(4):386-405. [doi: 10.1097/01.CHI.0000046821.95464.CF] [Medline: 12649626] 5. Ranieri WF, Steer RA, Lavrence TI, Rissmiller DJ, Piper GE, Beck AT. Relationships of depression, hopelessness, and dysfunctional attitudes to suicide ideation in psychiatric patients. Psychol Rep 1987 Dec;61(3):967-975. [doi: 10.2466/pr0.1987.61.3.967] [Medline: 3438422]

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.11 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

6. Schulberg HC, Bruce ML, Lee PW, Williams JW, Dietrich AJ. Preventing suicide in primary care patients: the primary care physician's role. Gen Hosp Psychiatry 2004;26(5):337-345. [doi: 10.1016/j.genhosppsych.2004.06.007] [Medline: 15474633] 7. Gould MS, King R, Greenwald S, Fisher P, Schwab-Stone M, Kramer R, et al. Psychopathology associated with suicidal ideation and attempts among children and adolescents. J Am Acad Child Adolesc Psychiatry 1998 Sep;37(9):915-923. [doi: 10.1097/00004583-199809000-00011] [Medline: 9735611] 8. Fombonne E. Suicidal behaviours in vulnerable adolescents. Time trends and their correlates. Br J Psychiatry 1998 Aug;173:154-159. [Medline: 9850228] 9. van Spijker Bregje A J, van SA, Kerkhof Ad J F M, Hoeymans N, Smit F. Disability weights for suicidal thoughts and non-fatal suicide attempts. J Affect Disord 2011 Nov;134(1-3):341-347 [FREE Full text] [doi: 10.1016/j.jad.2011.05.020] [Medline: 21652085] 10. Mann JJ, Apter A, Bertolote J, Beautrais A, Currier D, Haas A, et al. Suicide prevention strategies: a systematic review. JAMA 2005 Oct 26;294(16):2064-2074. [doi: 10.1001/jama.294.16.2064] [Medline: 16249421] 11. Rihmer Z. Strategies of suicide prevention: focus on health care. J Affect Disord 1996 Jul 08;39(2):83-91. [Medline: 8827416] 12. Luoma JB, Martin CE, Pearson JL. Contact with mental health and primary care providers before suicide: a review of the evidence. Am J Psychiatry 2002 Jun;159(6):909-916 [FREE Full text] [doi: 10.1176/appi.ajp.159.6.909] [Medline: 12042175] 13. Barraclough B, Bunch J, Nelson B, Sainsbury P. A hundred cases of suicide: clinical aspects. Br J Psychiatry 1974 Oct;125:355-373. [Medline: 4425774] 14. Hagnell O, Lanke J, Rorsman B. Suicide rates in the Lundby study: mental illness as a risk factor for suicide. Neuropsychobiology 1981;7(5):248-253. [Medline: 7312144] 15. Pirkis J, Burgess P. Suicide and recency of health care contacts. A systematic review. Br J Psychiatry 1998 Dec;173:462-474. [Medline: 9926074] 16. Appleby L. University of Manchester. 2012. National Confidential Inquiry into Suicide and Homicide by People with Mental Illness Annual Reportngland, Wales, Scotland, and Northern Ireland URL: http://www.webcitation.org/6qoEw0eUz [accessed 2017-05-29] [WebCite Cache ID 6qoEw0eUz] 17. Erlangsen A, Nordentoft M. Psychosocial therapy for people at risk of suicide. Lancet Psychiatry 2016 Jun;3(6):494-495. [doi: 10.1016/S2215-0366(16)30021-9] [Medline: 27017085] 18. Hetrick SE, Robinson J, Spittal MJ, Carter G. Effective psychological and psychosocial approaches to reduce repetition of self-harm: a systematic review, meta-analysis and meta-regression. BMJ Open 2016 Sep 22;6(9):e011024 [FREE Full text] [doi: 10.1136/bmjopen-2016-011024] [Medline: 27660314] 19. Cosgrave EM, Robinson J, Godfrey KA, Yuen HP, Killackey EJ, Baker KD, et al. Outcome of suicidal ideation and behavior in a young, help-seeking population over a 2-year period. Crisis 2007;28(1):4-10. [doi: 10.1027/0227-5910.28.S1.4] [Medline: 17555027] 20. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001 Sep;16(9):606-613 [FREE Full text] [Medline: 11556941] 21. Kobak KA, Greist JH, Jefferson JW, Katzelnick DJ. Computer-administered clinical rating scales. A review. Psychopharmacology (Berl) 1996 Oct;127(4):291-301. [Medline: 8923563] 22. Williams A, Larocca R, Chang T, Trinh N, Fava M, Kvedar J, et al. Web-based depression screening and psychiatric consultation for college students: a feasibility and acceptability study. Int J Telemed Appl 2014;2014:580786 [FREE Full text] [doi: 10.1155/2014/580786] [Medline: 24799895] 23. Kim E, Coumar A, Lober WB, Kim Y. Addressing mental health epidemic among university students via web-based, self-screening, and referral system: a preliminary study. IEEE Trans Inf Technol Biomed 2011 Mar;15(2):301-307. [doi: 10.1109/TITB.2011.2107561] [Medline: 21257386] 24. Lawrence ST, Willig JH, Crane HM, Ye J, Aban I, Lober W, et al. Routine, self-administered, touch-screen, computer-based suicidal ideation assessment linked to automated response team notification in an HIV primary care setting. Clin Infect Dis 2010 Apr 15;50(8):1165-1173 [FREE Full text] [doi: 10.1086/651420] [Medline: 20210646] 25. Kessler RC, Andrews G, Colpe LJ, Hiripi E, Mroczek DK, Normand S L T, et al. Short screening scales to monitor population prevalences and trends in non-specific psychological distress. Psychol Med 2002 Aug;32(6):959-976. [Medline: 12214795] 26. Altman EG, Hedeker D, Peterson JL, Davis JM. The Altman Self-Rating Mania Scale. Biol Psychiatry 1997 Nov 15;42(10):948-955. [doi: 10.1016/S0006-3223(96)00548-3] [Medline: 9359982] 27. Stefanis NC, Hanssen M, Smirnis NK, Avramopoulos DA, Evdokimidis IK, Stefanis CN, et al. Evidence that three dimensions of psychosis have a distribution in the general population. Psychol Med 2002 Feb;32(2):347-358. [Medline: 11866327] 28. van Spijker Bregje A J, Batterham PJ, Calear AL, Farrer L, Christensen H, Reynolds J, Kerkhof AJ. The suicidal ideation attributes scale (SIDAS): Community-based validation study of a new scale for the measurement of suicidal ideation. Suicide Life Threat Behav 2014 Aug;44(4):408-419. [doi: 10.1111/sltb.12084] [Medline: 24612048] 29. Brener ND, Collins JL, Kann L, Warren CW, Williams BI. Reliability of the Youth Risk Behavior Survey Questionnaire. Am J Epidemiol 1995 Mar 15;141(6):575-580. [Medline: 7900725]

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.12 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

30. Kann L, Kinchen S, Shanklin SL, Flint KH, Kawkins J, Harris WA, Centers for Disease Control Prevention (CDC). Youth risk behavior surveillance--United States, 2013. MMWR Surveill Summ 2014 Jun 13;63 Suppl 4:1-168 [FREE Full text] [Medline: 24918634] 31. Von KM, Ustun TB, Ormel J, Kaplan I, Simon GE. Self-report disability in an international primary care study of psychological illness. J Clin Epidemiol 1996 Mar;49(3):297-303. [Medline: 8676177] 32. Mayfield D, McLeod G, Hall P. The CAGE questionnaire: validation of a new alcoholism screening instrument. Am J Psychiatry 1974 Oct;131(10):1121-1123. [doi: 10.1176/ajp.131.10.1121] [Medline: 4416585] 33. Saunders JB, Aasland OG, Babor TF, de la Fuente J R, Grant M. Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on Early Detection of Persons with Harmful Alcohol Consumption--II. Addiction 1993 Jun;88(6):791-804. [Medline: 8329970] 34. Hickie IB, Davenport TA, Hadzi-Pavlovic D, Koschera A, Naismith SL, Scott EM, et al. Development of a simple screening tool for common mental disorders in general practice. Med J Aust 2001 Jul 16;175 Suppl:S10-S17. [Medline: 11556430] 35. Kenis P. Waiting lists in Dutch health care: an analysis from an organization theoretical perspective. J Health Organ Manag 2006;20(4):294-308. [doi: 10.1108/14777260610680104] [Medline: 16921814] 36. Gulliver A, Griffiths KM, Christensen H. Perceived barriers and facilitators to mental health help-seeking in young people: a systematic review. BMC Psychiatry 2010;10:113 [FREE Full text] [doi: 10.1186/1471-244X-10-113] [Medline: 21192795] 37. Caldwell TM, Jorm AF, Dear Keith B G. Suicide and mental health in rural, remote and metropolitan areas in Australia. Med J Aust 2004 Oct 4;181(7 Suppl):S10-S14. [Medline: 15462636] 38. Saxena S, Thornicroft G, Knapp M, Whiteford H. Resources for mental health: scarcity, inequity, and inefficiency. Lancet 2007 Sep 8;370(9590):878-889. [doi: 10.1016/S0140-6736(07)61239-2] [Medline: 17804062] 39. Ellis LA, Collin P, Hurley PJ, Davenport TA, Burns JM, Hickie IB. Young men's attitudes and behaviour in relation to mental health and technology: implications for the development of online mental health services. BMC Psychiatry 2013;13:119 [FREE Full text] [doi: 10.1186/1471-244X-13-119] [Medline: 23601273] 40. Batterham PJ, Calear AL, Sunderland M, Carragher N, Brewer JL. Online screening and feedback to increase help-seeking for mental health problems: population-based randomised controlled trial. BJPsych Open 2016 Jan;2(1):67-73 [FREE Full text] [doi: 10.1192/bjpo.bp.115.001552] [Medline: 27703756] 41. Gould MS, Marrocco FA, Hoagwood K, Kleinman M, Amakawa L, Altschuler E. Service use by at-risk youths after school-based suicide screening. J Am Acad Child Adolesc Psychiatry 2009 Dec;48(12):1193-1201 [FREE Full text] [doi: 10.1097/CHI.0b013e3181bef6d5] [Medline: 19858758] 42. Wintersteen MB. Standardized screening for suicidal adolescents in primary care. Pediatrics 2010 May;125(5):938-944. [doi: 10.1542/peds.2009-2458] [Medline: 20385627] 43. Chamberlain P, Goldney R, Delfabbro P, Gill T, Dal GL. Suicidal ideation. The clinical utility of the K10. Crisis 2009;30(1):39-42. [doi: 10.1027/0227-5910.30.1.39] [Medline: 19261567] 44. O'Connor SS, Beebe TJ, Lineberry TW, Jobes DA, Conrad AK. The association between the Kessler 10 and suicidality: a cross-sectional analysis. Compr Psychiatry 2012 Jan;53(1):48-53. [doi: 10.1016/j.comppsych.2011.02.006] [Medline: 21489424] 45. Vijayakumar L, Kumar MS, Vijayakumar V. Substance use and suicide. Current Opinion in Psychiatry 2011;24(3):197-202. [doi: 10.1097/YCO.0b013e3283459242] [Medline: 21430536] 46. Kessler R, van Loo HM, Wardenaar KJ, Bossarte RM, Brenner LA, Cai T, et al. Testing a machine-learning algorithm to predict the persistence and severity of major depressive disorder from baseline self-reports. Molecular Psychiatry 2016;21(10):1366-1371. [Medline: 26728563] 47. Anothaisintawee T, Teerawattananon Y, Wiratkapun C, Kasamesup V, Thakkinstian A. Risk prediction models of breast cancer: a systematic review of model performances. Breast Cancer Res Treat 2011 Nov 11;133(1):1-10. [doi: 10.1007/s10549-011-1853-z] [Medline: 22076477] 48. Passos IC, Mwangi B, Cao B, Hamilton JE, Wu M, Zhang XY, et al. Identifying a clinical signature of suicidality among patients with mood disorders: A pilot study using a machine learning approach. Journal of Affective Disorders 2016 Mar;193:109-116. [doi: 10.1016/j.jad.2015.12.066] [Medline: 26773901] 49. Pestian JP, Sorter M, Connolly B, Bretonnel Cohen K, McCullumsmith C, Gee JT, et al. A machine learning approach to identifying the thought markers of suicidal subjects: a prospective multicenter trial. Suicide and life-threatening behavior 2016;47(1):112-121. [Medline: 27813129] 50. Facebook uses Artifical Intelligence to identify suicide risk.: News.com.au; 2017. URL: http://www.webcitation.org/ 6rQjDeGIV [accessed 2017-06-23] [WebCite Cache ID 6rQjKsTDp]

Abbreviations BDQ: Brief Disability Questionnaire K10: Kessler-10 Questionnaire MHeC: Mental health eClinic NEET: Not in Educational, Employment or Training

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.13 (page number not for citation purposes) JOURNAL OF MEDICAL INTERNET RESEARCH Iorfino et al

SIDAS: Suicide Ideation Attributes Scale SPSS: Statistical Package for the Social Sciences

Edited by G Eysenbach; submitted 23.04.17; peer-reviewed by P Batterham, AE Aladağ, D Reidenberg, S MacLean; comments to author 17.05.17; revised version received 26.05.17; accepted 26.05.17; published 12.07.17 Please cite as: Iorfino F, Davenport TA, Ospina-Pinillos L, Hermens DF, Cross S, Burns J, Hickie IB Using New and Emerging Technologies to Identify and Respond to Suicidality Among Help-Seeking Young People: A Cross-Sectional Study J Med Internet Res 2017;19(7):e247 URL: http://www.jmir.org/2017/7/e247/ doi:10.2196/jmir.7897 PMID:28701290

©Frank Iorfino, Tracey A Davenport, Laura Ospina-Pinillos, Daniel F Hermens, Shane Cross, Jane Burns, Ian B Hickie. Originally published in the Journal of Medical Internet Research (http://www.jmir.org), 12.07.2017. This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.

http://www.jmir.org/2017/7/e247/ J Med Internet Res 2017 | vol. 19 | iss. 7 | e247 | p.14 (page number not for citation purposes) Chapter 5: Neurobiological correlates and multidimensional outcomes

Iorfino, F., Hickie, I.B., Lee, R.S., Lagopoulos, J & Hermens, D.F. The underlying neurobiology of key functional domains in young people with affective disorders: A systematic review. BMC Psychiatry, 16(1): 1.

62 Iorfino et al. BMC Psychiatry (2016) 16:156 DOI 10.1186/s12888-016-0852-3

RESEARCH ARTICLE Open Access The underlying neurobiology of key functional domains in young people with mood and anxiety disorders: a systematic review Frank Iorfino, Ian B. Hickie, Rico S. C. Lee, Jim Lagopoulos and Daniel F. Hermens*

Abstract Background: Mood and anxiety disorders are leading causes of disability and mortality, due largely to their onset during adolescence and young adulthood and broader impact on functioning. Key factors that are associated with disability and these disorders in young people are social and economic participation (e.g. education, employment), physical health, suicide and self-harm behaviours, and alcohol and substance use. A better understanding of the objective markers (i.e. neurobiological parameters) associated with these factors is important for the development of effective early interventions that reduce the impact of disability and illness persistence. Methods: We systematically reviewed the literature for neurobiological parameters (i.e. neuropsychology, neuroimaging, sleep-wake and circadian biology, neurophysiology and metabolic measures) associated with functional domains in young people (12 to 30 years) with mood and/or anxiety disorders. Results: Of the one hundred and thirty-four studies selected, 7.6 % investigated social and economic participation, 2. 1 % physical health, 15.3 % suicide and self-harm behaviours, 6.9 % alcohol and substance use, whereas the majority (68.1 %) focussed on clinical syndrome. Conclusions: Despite the predominance of studies that solely examine the clinical syndrome of young people the literature also provides evidence of distinct associations among objective measures (indexing various aspects of brain circuitry) and other functional domains. We suggest that a shift in focus towards characterising the mechanisms that underlie and/or mediate multiple functional domains will optimise personalised interventions and improve illness trajectories. Keywords: Depression, Anxiety, Bipolar, Functional outcomes, Biomarkers, Neurobiology, Neuropsychology, Personalised psychiatry

Background Whilst, some specific diagnoses have been successfully Depression and anxiety are associated with the greatest treated with certain interventions (e.g. CBT for social burden of disease of all neurological, psychiatric and sub- anxiety disorder; [4]), there are limitations to the op- stance use disorders [1]. The early onset of psychiatric ill- timal treatment for unipolar, bipolar and comorbid ness plays a key role in such disability with approximately mood disorders. This is particularly evident for those 75 % of these disorders occurring before the age of 25 with emerging illnesses who often experience mixed years [2]. Despite this, our current capacity to provide tai- states and/or subthreshold symptoms [5]. Since these lored early interventions and prevent the progression of mood states typically arise during adolescence and illness or slow the pathway to disability is lacking [3]. young adulthood, a period of critical brain development and functional independence, the impact of ill- * Correspondence: [email protected] ness can lead to greater disability and worse illness Clinical Research Unit, Brain and Mind Centre, University of Sydney, 94 Mallet outcomes [6–8]. Thus, the search for objective Street, Camperdown, NSW 2050, Australia

© 2016 Iorfino et al. Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 2 of 38

markers of early risk states with predicative capacity in common brain circuits involved in cognition and in regards to disability and mortality requires rigorous behaviour are responsible for the development of psy- investigation so that appropriate interventions can be chopathology and general dysfunction [23]. In this trialled and delivered as early as possible to reduce view, quantifying the integrity of such brain systems the impact of disability [3, 9]. (e.g. via neuroimaging, neurophysiology or circadian Traditionally, there has been a focus on the ‘clinical biology) along with their behavioural concomitants syndrome’ defined as identifying distinct clinical cat- (e.g. neuropsychology, social cognition or sleep-wake egories or disorders (based on ICD or DSM diagnos- patterns) may lead to the identification of objective tic criteria) with specific thresholds and the impact of markers of early risk states and also serve as treat- these on functioning. However, even at a subthreshold ment targets. For example, in a longitudinal study by symptom level significant contributors to disability our group, neuropsychological performance at base- and mortality include social and economic disability line was the single best predictor of socio- [10, 11], poor physical health (e.g. diabetes) [12], high occupational functioning at follow up, over and above suicide and self-harm behaviours [10, 13–15], and diagnosis and symptom severity [24]. Such findings risky alcohol and substance use [16, 17]. Given the demonstrate the relevance of objective ‘brain’ markers clinical impact of these factors for young people with (in this case, a cognitive phenotype) to provide im- emerging mood and anxiety disorders, we have identi- portant insights about a crucial functional domain, fied them as key functional domains that, we argue, which cut across diagnostic categories to direct effect- should be the focus of targeted personalised assess- ive treatment strategies at the pathophysiological ment and intervention. Although the term ‘functional driver of poor patient outcomes. domain’ has traditionally, often referred to outcomes Here, we present a systematic review of the neuro- relating to occupational (i.e. employment and educa- biological and neurocognitive correlates, of five function) status, here, we use the term to include other tional domains in young people with mood and key factors that have significant (often concomitant) anxiety disorders. We focus specifically on major impacts on levels of functioning in young people. depression, bipolar disorder (I, II, not otherwise speci- These domains largely align with the framework pro- fied; NOS) and anxiety disorders (excluding post- vided by the ‘International classification of function- traumatic stress disorder), since these most closely ing, disability and health’ [18] for conceptualising relate to the common developmental trajectories of health and health related states. These include: (i) so- emerging mood disorders in young people [25]. In cial and economic participation (i.e. engagement and this review we evaluate the relationship between the stability in employment, education and social relation- functional domains (described above), and evidence ships); (ii) physical health; (iii) suicide and self-harm from neuropsychology, neuroimaging, sleep-wake and behaviours; (iv) alcohol and substance use; and (v) circadian biology, neurophysiology and metabolic clinical syndrome (i.e. diagnostic category, stage of ill- studies. A wide age range was chosen (12–30 years) ness and severity of symptoms [3]). These domains to focus on the adolescent and young adult popula- are priority areas for service models in Australia (e.g. tion; referred to collectively as ‘young people’,tobet- headspace [19, 20]), which recognise the need for ter understand the primary age group that are early interventions that aim to target specific out- vulnerable and present to primary youth mental comes associated with illness persistence and greater health services. The primary objective of this study is disability [21]. Importantly, a focus on these five do- to establish the current status of the literature of mains recognises the need to evaluate multiple (often young people with mood and anxiety disorder with interacting) aspects of an individual to better charac- respect to neurobiological investigations addressing terise their specific phenotype and, as a result, at- any of the proposed five functional domains. Whilst, tempt to predict their potential illness trajectory. we expect that the large majority of identified studies To overcome some of the limitations associated would investigate clinical syndromes and a smaller with current diagnostic approaches that link poorly to number would investigate the remaining functional neurobiological risk factors or patterns of treatment domains, it is expected that unique associations be- response it is important to characterise the neurobiol- tween neurobiological parameters and a functional ogythatmayunderlieormediateobservablefunc- domain, not accounted for by the clinical syndrome, tional impairment(s) [9, 22]. This emphasises the will become clearer. The aim of our approach is to need to focus on the four remaining functional do- ultimately provide a framework for guiding the devel- mains in addition to the traditional focus on the clin- opment of personalised assessment and interventions ical syndrome to optimise personalised interventions. to prevent or delay significant disability in young Models of psychopathology suggest that breakdowns mental health patients. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 3 of 38

Methods Study characteristics and selection Methods of review regarding eligibility criteria, data Using a pro forma, the first author (FI) checked the collection and synthesis were specified in advance in abstract and/or full texts of each paper for the follow- the form of a review protocol. We followed the ing inclusion criteria; i) a mean age between 12 and guidelines for conducting and reporting a systematic 30 years; ii) at least one group of subjects was re- review set out by ‘the PRISMA statement’ [26], and ported as having a primary mood and anxiety (i.e. de- the ‘Cochrane Handbook for Systematic Reviews’ [27]. pression, bipolar, anxiety) disorder (according to DSM-IV or ICD-10 criteria) or syndrome (e.g. ‘at risk’, current depressive symptoms); iii) at least one of the Eligibility criteria following functional domains: (a) social and economic Report characteristics and information sources participation; (b) physical health; (c) suicide and self- We searched PubMed databases for unique records harm behaviours; (d) alcohol and substance use; and/ using the following criteria: (i) published in the last or (e) clinical syndrome, was measured/quantified; 20 years (i.e. between January 1994 and March, 2014, and iv) a statistical (i.e. correlational, regression, etc.) to coincide with the release of DSM-IV since this ver- association between the functional domain and at sion introduced the use of clinical significance ratings least one neurobiological parameter (i.e. neuropsych- related to the impact of illness on areas of function- ology, brain imaging, sleep-wake and circadian biol- ing); (ii) the study was reported in English; and (iii) ogy, neurophysiology and metabolic) was reported. had keyword combinations (see Table 1 for full search Review articles and case studies were excluded from terms). The reference lists of studies identified by our the final synthesis. Studies were labelled ‘yes’ if they PubMed search were not utilised as an additional in- fulfilled all four criteria, ‘no’ if they failed to meet all formation source. four criteria or ‘possibly’ if it was unclear whether all

Table 1 Full list of search terms used according to each topic area Topic area Pub Med Terms Population of interest Mood and anxiety syndrome or profile Anxiety disorder OR anxiety OR depression OR depressive disorders OR depressive disorder [MeSH Terms] OR major depressive disorder OR MDD OR disorder, bipolar [MeSH Terms] OR bipolar disorder OR affective disorder OR mood disorder OR affective syndrome OR manic syndrome OR depressive syndrome OR anxious syndrome Youth Adolescents OR young people OR adolescence OR adolescent [MeSH Terms] OR youth OR young adult Functional domain Social and economic participation Socio-occupational functioning OR functioning OR social functioning OR occupational participation OR economic participation Physical health Physical health OR metabolic rate OR obesity OR blood pressure OR CVD OR fitness OR cardiovascular disease OR BMI or body mass index OR waist measurement OR blood glucose OR smoking rate OR physical activity OR cholesterol levels Suicide and self-harm behaviours Suicide [MeSH Terms] OR suicide ideation OR self-harm OR suicide risk Alcohol and substance use Substance use disorder [MeSH Terms] OR alcohol use OR drug use Clinical syndrome Illness progression OR syndrome progression OR symptom severity Neurobiological parameter Neuropsychology Neuropsychology OR neuropsychological test [MeSH Terms] Imaging Brain imaging [MeSH Terms] OR imaging OR neuroimaging OR fMRI OR DTI OR MRI OR MRS Sleep-wake and circadian biology Actigraphy [MeSH Terms] OR melatonin secretion OR circadian rhythms OR DLMO OR sleep-wake and circadian biology Neurophysiology Neurophysiology OR EEG OR electroencephalography OR ERP OR event-related potentials Metabolic BMI OR waist measurement OR blood pressure OR cholesterol Note. Terms within each cell in column 3 (above) used the ‘OR’ function, whilst the ‘AND’ function was used to combine terms between the cells of column 3 Iorfino et al. BMC Psychiatry (2016) 16:156 Page 4 of 38

criteria were fully met. Any disagreement with these for each study were taken as any evidence of an associ- rules was resolved by consensus with the senior au- ation between a particular neurobiological measure and thor (DH). a functional domain. In order to achieve this, the various scales, tests, and assessments were collapsed into broader categories of key measures (e.g. specific neuro- Identification of studies psychological subtests grouped into a cognitive domain; Figure 1 displays the series of steps undertaken as we see Tables 3, 4, 5, 6 and 7). Given the variability in meth- identified studies for this systematic review. First, of the odology and the large and varied outcomes of identified 3975 studies identified by the searches (see Table 1 for studies it was not appropriate to carry out a meta- search terms), 565 titles and abstracts were examined for analysis [28]. eligibility. At this stage, 188 studies were excluded on the basis of not meeting one or more of the eligibility Results criteria specified. The eligibility stage involved the as- A total of 134 studies were included in this systematic sessment 377 full texts (i.e. the published manuscript) to review (see Fig. 1); 10 of these studies were featured evaluate whether these studies were suitable, which led more than once in the data synthesis to make a total of to a further 243 studies being excluded. The remaining 144 reported results. As summarised in Table 2, the in- 134 studies were included in the final synthesis (see cluded studies were categorized according to functional Table 2 for a summary of these studies). domain in the following proportions: 7.6 % (k = 11) investigated social and economic participation, 2.1 % (k = Synthesis of results 3) physical health, 15.3 % (k = 22) suicide and self-harm For each of the included studies, the reviewer (FI) behaviours, 6.9 % (k = 10) alcohol and substance use, collated data with respect to the study design (i.e. cross- and 68.1 % (k = 98) clinical syndrome. In regards to sectional, longitudinal; see Table 2), sample characteris- neurobiological parameters, 19.4 % (k = 28) focused on tics (i.e. age, sample sizes), aims, key measures (e.g. neuropsychology, 43.1 % (k = 62) on neuroimaging, 16 % neuropsychological, circadian, clinical) and key findings (k = 23) on sleep-wake and circadian biology, 14.6 % (k = (presented in Tables 3, 4, 5, 6 and 7; one table per 21) on neurophysiology and 6.9 % (k = 10) on metabolic neurobiological parameter). To clarify, the key findings measures. The range of the mean ages for patient groups

Fig. 1 This figure communicates the flow of studies through the systematic review process and identifies the number of studies excluded at each phase as well as the reason for exclusion Iorfino et al. BMC Psychiatry (2016) 16:156 Page 5 of 38

Table 2 Overview of the included studies organised according to functional domain and neurobiological parameter investigated Neurobiological parameter Functional domain Neuropsychology Imaging Sleep-wake and circadian biology Neurophysiology Metabolic Social & economic participation Beiderman (2011) Perlman (2012) Abelson (1996) Kaur (2013) Taylor (2008)* Fujii (2013)* Goodyer (1998)* Korhonen (2002) Granger (1994) Lee (2013c) Lee (2013a) Physical health Bond (2011) Jarworska (2011)* Mannie (2013) Suicide & self-harm behaviours Bridge (2012) Ehrlich (2004) Coplan (2000)* Ashton (1994) Apter (1999) Miranda (2012) Ehrlich (2005) Mathew (2003) Graae (1996) De Berardis (2013) Ohmann (2008) Goodman (2011)* McCracken (1997)* Pechtel (2013) Plana (2010) Oldershaw (2009) Pan (2011) Soreni (1999) Pan (2013b)* Pan (2013a) Tyano (2006) Pan (2013b)* Alcohol & substance use Harvey (2007) Chitty (2013) Chitty (2014) Goldstein (2008) Hermens (2013a) Cornelius (2010) Ehlers (2011) De Bellis (2005) Jarvis (2008) Medina (2007) Clinical syndrome Andres (2007) Adler (2007) Adam (2010) Bakker (2011) Pine (2001) Andres (2008) Aghajani (2013) Ankers (2009) Carrasco (2013a) Taylor (2008)* Basso (2001) Bitter (2011) Armitage (1997) Carrasco (2013b) Cataldo (2005) Chang (2008) Coplan (2000)* Croarkin (2014) Fleck (2008) Chu (2013) Doane (2013) Dai (2012) Fujii (2013)* Diler (2013) Ellenbogen (2006) El Badri (2001) Gunther (2004) Forbes (2006) Ellenbogen (2010) Hajcak (2008) Han (2012) Forbes (2010) Goodyer (1998)* Houston (2003) Kıvırcık (2003)* Gabbay (2012) Harkness (2011) Jarworska (2011)* Klimkeit (2011) Gabbay (2013) Landsness (2011) Jarworska (2013) Okasha (2000)* Gao (2013) McCracken (1997)* Kıvırcık (2003)* Pavuluri (2010a) Gilbert (2000) Murray (2012) Okasha (2000)* Schmid (2013) Gilbert (2009) Rao (1996) Stern (2010) Simons (2009) Goodman (2011)* Rao (2008) Vaidyanathan (2014) Torres (2010) Gruner (2012) Robillard (2013a) Wall (2013) Hatton (2012) Robillard (2013b) Henderson (2013) Scott (2014) Ho (2014) Huang (2012) Huyser (2011) Huyser (2013) Ladouceur (2011) Lagopoulos (2012) Lagopoulos (2013a) Lagopoulos (2013b) Lazaro (2008) Iorfino et al. BMC Psychiatry (2016) 16:156 Page 6 of 38

Table 2 Overview of the included studies organised according to functional domain and neurobiological parameter investigated (Continued) Neurobiological parameter Functional domain Neuropsychology Imaging Sleep-wake and circadian biology Neurophysiology Metabolic Clinical syndrome Lazaro (2012) Lisy (2011) MacMaster (2006) MacMillian (2003) McClure (2007) Meng (2013) Pannekoek (2014) Patel (2008) Pavuluri (2010b) Pavuluri (2011) Phan (2013) Rauch (2002) Reynolds (2014) Rosenberg (1997) Rosenberg (2000) Rosso (2005) Schienle (2011) Schneider (2012) Strawn (2012) Wegbreit (2011) Yucel (2008) Zarei (2011) Zuo (2013) Note. * = indicates the study appears more than once, BOLD = longitudinal study, italicized = Study conducted by the Brain and Mind Centre across all the included studies was 11.7 to 31.7 years, investigated physical health. The relationship between glo- since those studies that had comparison groups both in- bal deficits in cognition and social and economic participa- side and outside the inclusion criteria of 12 to 30 were tion in mood disorders is unclear as there were only two still included. studies; one reporting a positive relationship [24], and the other no relationship [29]. The former study utilised a Neuropsychology mixed psychiatric sample that consisted of mood disorder There were 28 studies (a total of 2877 participants; and psychosis patients, which may have influenced the re- 58.5 % female) that utilised neuropsychology and across sults given the well-supported relationship between social these studies 69 % (2037/2877) were patients and 29 % and economic participation and neuropsychology in pa- (804/2877) were healthy controls. Among the patient tients with psychosis [30]. However, the latter study only group 47 % (966/2037) had depression, 15 % (301/2037) investigated MDD (N = 16) and utilised the Global Assess- had bipolar, 12 % (239/2037) had anxiety, and 26 % ment of Functioning scale (GAF) as a measure of partici- (531/2037) were classified as other. pation, which could be problematic as the rating can be made on the basis of symptoms or functioning. Thus, it is Functional domains: social and economic participation, clear that more studies are needed in this population to physical health, suicide and self-harm & alcohol and resolve or clarify such findings. substance use Studies exploring specific neuropsychological capabil- Our systematic search found an association between neuro- ities have provided greater insight into how these relate psychology and three functional domains (i.e. social and to functional domains that can be ambiguous when in- economic participation, suicide and self-harm and alcohol vestigating global cognition. Of the neuropsychological and substance use). No studies that met our criteria studies reviewed, executive function appears to be Iorfino et al. BMC Psychiatry (2016) 16:156 Page 7 of 38

Table 3 Neuropsychological studies evaluating the five functional domains in young people (12–30 yrs) with a mood and/or anxiety disorder Outcome Study Age (mean ± SD) Sample (N) Aims Key measures Key findings measure Social and [40] HC: 13.6 ± 2.1, HC (47M; 34F), Evaluate the clinical NΨ: Executive function BPD-I: ↓ executive economic impact of executive deficits (CPT, CVLT-C, RCF, function ~ ↓ social and participation HC-EFD: 13.9 ± 2.3, HC-EFD (12M; 5F), function deficits in youth SCWT, WCST, WAIS-III- economic participation BPD-I: 13.7 ± 2.1, BPD-I (52M; 24F), with BPD-I disorder. FFD) BPD-I-EFD: 12.8 ± 2.4 BPD-I-EFD (49M; 13F) Functional: GAF, WRAT-III, placement in special class [39]* SAD: 23.9 ± 6.7; SAD (20M; 10F) Assess the NΨ: Executive function SAD: ↓ executive neuropsychological (CPT, TMT-B, WCST), Pro- function ~ ↓ social and HC: 25.6 ± 5.6 HC (20M; 10F) function of SAD without cessing speed (TMT-A), economic participation co-morbidity Verbal learning & mem- (and ↑ SAD severity) ory (AVLT) Functional: GAF [29] MDD: 18.9 ± 2.0, MDD (4M; 12F), Investigate the NΨ: Executive function MDD: No significant ~ NΨ association between (SCWT, TMT-B), Verbal HC: 16.9 ± 1.9 HC (11M; 14F) MDD: ↓ social and cognitive performance learning & memory economic participation and MDD. (WMS-SR, LLT, RCF-3min), General intellect (WAIS-III- S & V), Attention (WAIS-III- DS, BD & DSp) Functional: GAF FED: 22.00 ± 4.9 FED (8M; 12F) Assess the effectiveness NΨ: Executive function FED & FEP: CR ~ ↑ of CR in patients with a (CANTAB-IED; -FAS, TMT- immediate learning and FEP: 23.30 ± 3.9 FEP (20M; 13F) first-episode of either B), Processing speed memory, and ↑ social major depression or (TMT-A, category fluency), and economic psychosis Attention and working participation (mediated memory (LDSF, LDSB, by ↑ delayed learning CANTAB-SSP;-RVP, mental and memory) control), Immediate learning and memory (Logical Memory I, RAVLT- tot, CANTAB-PAL), Delayed learning and memory (LM-Ret, RCF-3min, RAVLT-Ret) Functional: SFS [24] MHP: 21.6 ± 4.5 MDD (34) Identify cognitive markers NΨ : Executive function MHP: ↑ BL general NΨ that predict later socio- (CANTAB-IED, TMT-B), Pro- ~ ↑ social and economic BPD (29) occupational functioning. cessing speed (TMT-A, participation at FUP PSD (30) CANTAB-FAS), Attention and working memory (of the total 93, (CANTAB-RVP), Verbal 52 % were male) learning & memory (LM- Ret, RAVLT-ret), Visual learning & memory (CAN- TAB- SSP;- PAL) Functional: SOFAS Suicide and [31] SA: 15.5 ± 1.4 SA (10M; 30F) Examine decision-making NΨ: Decision making SA: ↓ decision making ~ self-harm processes in suicide (IGT) suicide attempt history attempters and never- PC: 15.6 ± 1.4 PC (10M; 30F)suicidal comparison Functional: CSHF, PSIS subjects [32] SA: 18.31 ± 0.63 SA (1M; 12F) Examine whether NΨ: Executive function SA: ↓ BL cognitive cognitive inflexibility can (WCST; perseverative inflexibility ~ ↑ suicide NSA: 18.31 ± 0.78 NSA (9M; 23F) differentially and errors) ideation at 6-month FUP. prospectively predict Functional: BSS, SBS, suicidal ideation. SHBQ [35] SIB: 15.5 ± 1.3 SIB (99) Investigate the NΨ: Executive function Null findings neuropsychological (SCWT, WCST) NSIB: 15.1 ± 1.4 NSIB (77) differences between Iorfino et al. BMC Psychiatry (2016) 16:156 Page 8 of 38

Table 3 Neuropsychological studies evaluating the five functional domains in young people (12–30 yrs) with a mood and/or anxiety disorder (Continued) psychiatric patients with Functional: Clinical and without SIB. interview [33] HC: 15.8 ± 1.5 HC (11M; 46F) Assess decision making NΨ: Decision making (IGT, DSH: ↓ decision making and problem solving MEPS) ~ current, but not past ability in adolescents with DSH PC: 15.7 + 1.3 PC (2M; 20F)current or past self-harm Functional: Clinical interview DSH: 15.8 + 1.5 DSH (5M; 49F) [34]* SA: 16.20 ± 0.78 SA (4M; 11F) Measure neural activity NΨ: Decision making SA: ↑ decision making ~ during performance on (IGT-mod) suicide attempt history the IGT in adolescents. PC: 15.79 ± 1.58 PC (7M; 7F) Functional: C-CASA, CSHF, SIQ, SIS HC: 15.15 ± 1.46 HC (8M; 5F) Alcohol and [164] CU: 16.2 CU (28M; 42F) Investigate the non-acute NΨ: Intelligence (WASI), CU: ↓ attention, spatial substance use (13.5 – 18.4) relationship between can- Executive function working memory and nabis use and cognitive (CANTAB-IED), (CANTAB- learning. function MS), Attention and working memory CU: was independent predictor of performance (CANTAB-RVP;-SWM; -SSP, on the working memory DS, SDMT), Immediate and strategy measures learning and memory (RAVLT, CANTAB-PAL) Functional: TLFB [165] HC-NB: 22.9 ± 3.1 HC-NB (7M; 14F) Compare the cognition in NΨ: Intelligence (WTAR), MDD-B: ↓ visual learning binge drinkers with Psychomotor speed & memory and overall HC-B: 23.0 ± 2.5; HC-B (13M; 11F) depression to those with (TMT-A) Executive pattern of ↓ NΨ MDD-NB: 21.7 ± 3.2 MDD-NB (16M; 32F) depression alone or function (TMT-B), Verbal functioning. binge drinking alone. learning and memory MDD-B: 21.8 ± 3.4 MDD-B (24M; 19F) (RAVLT), Attention (CANTAB- RVP), working memory (CANTAB-SSP, Visuospatial learning and memory (CANTAB-PAL) Functional: AUDIT Clinical [50] OCD: 13.84 ± 2.78 OCD (18M; 17F) Investigate the influence NΨ: Intelligence (WISC-R: OCD: ↓ verbal and visual syndrome of clinical variables Vo), Visual organisation memory and velocity. HC: 13.81 ± 2.74 HC (18M; 17F) treatment on cognitive (WISC-R:-BD), Attention (Neuropsychological performance in OCD (WISC-R: -DS;-Co), Verbal impairment was not patients learning and memory related to obsessive- (WMS-III- LM1 & 2, compulsive severity) RAVLT), Visual learning and memory (WMS-III: VR 1 & 2, RCFT), Processing speed (TMT-A), Cognitive flexibility (TMT-B, WCST, SCWT), Verbal fluency (COWAT) Clinical: CDI, Y-BOCS [51] OCD: 13.46 ± 2.83 OCD (16M; 13F) Explore the evolution of NΨ: Intelligence (WISC-R: OCD: ↓ memory, speed cognitive dysfunction in Vo), Visual organisation of information processing HC: 13.06 ± 2.84 HC (12M; 10F) children and adolescents (WISC-R:-BD), Attention and cognitive flexibility. with OCD after treatment (WISC-R: -DS;-Co), Verbal (After treatment the learning and memory cognitive profile of the (WMS-III- LM1 & 2, OCD group was RAVLT), Visual learning normalized) and memory (WMS-III: VR 1 & 2, RCFT), Processing speed (TMT-A), Cognitive flexibility (TMT-B, WCST, SCWT), Verbal fluency (COWAT) Clinical: Y-BOCS Iorfino et al. BMC Psychiatry (2016) 16:156 Page 9 of 38

Table 3 Neuropsychological studies evaluating the five functional domains in young people (12–30 yrs) with a mood and/or anxiety disorder (Continued) [53] OCD: 29.70 ± 10.74 OCD (12M; 8F) Examine the impact of NΨ: VCAT, Verbal Fluency OCD: cognitive flexibility depression on executive (COWAT), Processing deficits ~ co-morbid HC: 30.06 ± 10.06 HC (11M; 21F) function deficits in OCD speed (TMT-A), Cognitive depression severity flexibility (TMT-B, WCST) Clinical: MMPI-D [47] HC: 12.5 ± 2.4 HC (11M; 10F) Compare impulsivity at NΨ: Cognitive style DD: ↑ symptom severity the neuropsychological (MFFT), Verbal fluency ~ ↑ reaction time, ↓ in DD: 11.7 ± 2.3 DD (11M; 10) and behavioural level in (VFT), Decision making commission errors. young depressed patients (WDWT), cognitive DD: ↑ conservative and healthy controls. flexibility (SCWT), Impulsivity (CPT) response styles & attention problems, ↓ Clinical: HDRS, CDI, CPRS- reaction times & R:L response initiation [166] HC: 28.2 ± 7.9 HC (20M; 28F) Investigate the effect of NΨ: Intelligence (NAART), EUT: ↑ cognitive flexibility syndrome state or course Cognitive flexibility than MEM. Performed EUT: 30.0 ± 7.2 EUT (11M; 14F) on executive dysfunction (WCST) similarly to FEM FEM: 25.7 ± 9.2 FEM (11M; 10F) Clinical: YMRS, HDRS MEM: 28.2 8.6 MEM (16M; 18F) [39]* SAD: 23.9 ± 6.7; SAD (20M; 10F) Assess the NΨ: Executive function SAD: ↓ executive neuropsychological (CPT, TMT-B, WCST), Pro- function ~ ↑ SAD severity HC: 25.6 ± 5.6 HC (20M; 10F) function of SAD without cessing speed (TMT-A), co-morbidity Verbal learning & memory (AVLT) Clinical: GAF [44] HC: 12.8 ± 2.5 HC (15M; 18F) Examine basic NΨ: Intelligence (WISC-III), DD: ↓ verbal learning and performance Verbal learning and memory compared to HC ANX: 12.4 ± 2.3 ANX (19M; 15F) neuropsychological memory (RAVLT), and ANX. DD: 13.5 ± 2.6 DD (17M; 14F) performance in children Attention (go-no go task) and adolescents with Clinical: CDI anxiety disorder or depressive disorder and in healthy subjects under drug-free condition [48] HC: 17.46 ± 1.59, HC (14M, 16F) MDD Investigate whether NΨ: Inhibitory control MDD: ↑ depression MDD: 17.32 ± 1.59 (12M, 19F) major depression in (CPT, go-no go task), At- symptom severity ~ ↓ adolescence is tention (ANT), Decision cognitive control of characterized by making (IGT), Verbal emotion processing neurocognitive deficits in learning and memory attention, affective (RAVLT), Attention (go-no decision making, and go task) cognitive control of emotion processing Clinical: BDI [167]* OCD: 27 ± 9.8 OCD (15M; 16F) Characterize the cognitive NΨ: cognitive flexibility Null findings for functions of the patients (SCWT, TMT-B), Processing neuropsychological tests. HC: 27.4 ± 9.1 HC (14M; 16F) with OCD by utilizing speed (TMT-A), Design ERPs and fluency test, Verbal flu- neuropsychological tests ency (CWAT) Clinical: HDRS [49] MDD: 15.3 ± 1.6 MDD (5M; 17F) Investigate verbal fluency, NΨ: Verbal fluency DD: ↓ WM & VF. MDD: cognitive speed, motor (COWAT), Processing ↓WM & processing speed DD: 15.6 ± 1.5 DD (6M; 6F) speed, and executive speed (Inspection time HC: 15.8 ± 1.2 HC (9M; 24F) functions in adolescents task), Working memory with unipolar depression. (Serial choice reaction time task), Set shifting (Local-global task) Clinical: [52] OCD: 24.06 ± 5 OCD (21M; 9F) Assess the relationship NΨ: Intelligence (WAIS- Results showed a between cognitive BD; -S), Cognitive defective visuospatial HC: Matched HC (21M; 9F) dysfunction, clinical status flexibility (WCST) recognition, which and severity in OCD. worsens with chronicity, Clinical: YBOCS deteriorated set-shifting Iorfino et al. BMC Psychiatry (2016) 16:156 Page 10 of 38

Table 3 Neuropsychological studies evaluating the five functional domains in young people (12–30 yrs) with a mood and/or anxiety disorder (Continued) abilities, overfocused attention to irrelevant stimuli and delayed selective attention to relevant tasks. Mild cases showed better selective attention than severe cases. Obses- sive cases had a defective visual memory, while compulsive cases had delayed perception of task relevant stimuli. Mixed cases showed disturbed information-processing both early and late. [55] HC: 12.4 ± 3.3, HC (15M; 9F) Examine the treatment NΨ: Attention (TMT-A, BPD: ↑ Working memory impact of lamotrigine on CPT), cognitive flexibility and verbal memory BPD: 13 ± 3.1 BPD (18M; 16F) the neurocognitive profile (TMT-B), Verbal fluency following treatment (to of patients with pediatric (COWT), Working memory levels similar to HC) bipolar disorder (WMS; DS, SS), Verbal memory (CVLT) Clinical: YMRS [168] MDD: 26.93 ± 5.33 MDD (14M; 14F) Assess the association NΨ: cognitive flexibility MDD: Poor BL inhibition between executive (CWIT, TMT-B), Verbal flu- and switching ~ ↑ relapse HC: 26.93 ± 5.18 HC (14M; 14F) function and relapse ency (VFT), Processing at FU speed (TMT-A), Clinical: MADRS [45] CS: 28 ± 7.9 CS (642) Examine whether NΨ: Information CS: Poor BL episodic cognitive deficits predict processing (SCWT, CST, memory ~ ↑ depressive current and/or follow-up LDST), Episodic memory symptoms at FUP (sub)clinical depressive (AVLT) symptoms in the general population Clinical: SCL-90 [169] BD: 22.2 ± 3.9 BD (23M; 22F) Determine whether NΨ: Intelligence (NAART), BPD: ↓ learning/memory, neuropsychological Visual spatial reasoning spatial/nonverbal HC: 22.5 ± 4.8 HC (12M; 13F) impairments are present (K-BIT), Attention/ reasoning, executive in clinically stable processing speed (TMT-A, function, and some patients with bipolar CANTAB-RVP, CVLT), aspects of attention disorder shortly after Learning and memory resolution of their first (CVLT- recall, CANTAB- manic episode SRM;-PRM;-PAL), cognitive flexibility (TMT-B, CANTAB-IED;-SWM), Ver- bal fluency (COWT), Clinical: PANSS, HDRS, BPRS, GAF, YMRS [56] MDD: 16.2 ± 1.1 MDD (7M; 11F) Investigate the NΨ: Verbal learning and MDD: ↓ Depressive neurocognitive outcome memory (CAVLT), symptoms were in adolescents who were Cognitive flexibility (D- associated with ↑ in treated with TMS KEFS, TMT) immediate and delayed verbal memory. Clinical: CDRS-R Note. Sample: ANX Anxiety disorder, BPD Bipolar Disorder, BPD-I Bipolar Disorder I, BPD-I-EFD Bipolar Disorder I with Executive Function Deficits, CS Community Sample, DD Depressive disorder, CU Cannabis user, DSH Deliberate Self-Harm, EUT euthymic, FED First-Episode depression, FEM first episode mania, FEP First- Episode Psychosis, HC Healthy Controls, HC-B Healthy Control Binge drinker, HC-EFD Healthy Control with Executive Function Deficits, HC-NB Healthy Control Non Binge drinker, MDD Major Depression Disorder, MDD-B Major Depression Disorder Binge drinker, MDD-NB Major Depression Disorder Non Binge drinker, MEM multiple episode mania, MHP Mental Health Patients (mixed diagnosis sample), NSA No Suicide Attempt, NSIB No Suicide Ideation Behaviour, OCD Obsessive Compulsive Disorder, PC Psychiatric Control (i.e. psychiatric diagnosis but no suicide attempt), PSD Psychotic Spectrum Disorder, SA Suicide Attempters, SAD Social Anxiety Disorder, SIB Suicide Ideation Behaviour Measures: ANT Attention network test, AUDIT Alcohol Use Disorder Identification Test, AVLT Auditory Verbal Learning Test, BDI Beck Depression Inventory, BPRS Brief psychiatric rating scale, BSS Beck Scale for Suicidal Ideation, CANTAB Cambridge Neuropsychological Test Automated Battery [subsets include: FAS Fluency and semantic test, IED Intra/Extra dimensional Set Shift Errors, MS Motor Screening, PAL Paired associates learning, PRM Pattern recognition memory, RVP Rapid Visual Processing hits score, SRM Spatial recognition memory, SSP Spatial span task, SWM Spatial working memory), C-CASA Columbia Classification Algorithm of Suicide Assessment, CDI Children’s Depression Inventory, COWAT Controlled Oral Word Association Task, CPRS-R:L Conners Parent Rating Scale- Revised: Long Iorfino et al. BMC Psychiatry (2016) 16:156 Page 11 of 38

Version, CPT Continuous Performance Test, CSHF Colombia Suicide History Form, CST Concept shifting test, CVLT California Verbal Learning Test for Children, D-KEFS Delis–Kaplan Executive Function System, GAF Global Assessment of Functioning, HDRS Hamilton depression rating scale, IGT Iowa Gambling Task, K-BIT Kaufman Brief Intelligence Test, LDSB Longest Digit Span Backward, LDSF Longest Digit Span Forward, LDST letter digit substitution test, LM-Ret logical memory- percentage retention, MADRS Montgomery-Asberg Depression Rating Scale, MEPS means-ends problem-solving procedure, MFFT matching familiar figures test, MMPI-D Minnesota Multiphasic Personality Test, depression subscale, NAART North American Adult Reading Test, NΨ neuropsychological, PANSS positive and negative syndrome scale, PSIS pierce suicide intent scale, RAVLT rey auditory verbal learning test; total score; retention; and/or 20min score, RCF Rey-Osterrieth Complex Figure, SBS Suicide Behavior Screening, SCWT stroop colour and word test, SCL-90 symptom checklist, SCWT stroop colour and word test, SDMT symbol digit modality test, SFS social functioning scale, SHBQ self-harm behavior questionnaire, SIQ suicide ideation questionnaire, SIS suicide intent scale, SOFAS social and occupational functioning assessment scale, = time-line followback, TMT-A trail making test – part A, TMT-B trail making test – part B, VFT verbal fluency task, WAIS-III Wechsler Adult Intelligence Scale [subsets include: S similarities, V Vocabulary, DS Digit symbol, BD Block design, FFD freedom from distractibility], WASI Wechsler abbreviated scale of Intelligence, WISC Wechsler Intelligence Scale For Children, WCST Wisconsin Card Sorting Test, WDWT Walk don’t-walk test, WMS Wechsler Memory Scale [subsets include: DS Digit Span, LM logical memory 1 & 2, SR = story recall, LLT List Learning Test, SS Spatial span, VR Visual Reproduction], WRAT-III wide range achievement test – third edition, WTAR wechsler test of adult reading, Y-BOCS Yale–Brown obsessive-compulsive scale, YMRS young mania rating scale Findings: ↑ = Increased, Improved or Higher, ↓ = Decreased, Reduced or Lower, ~ = ‘is associated with’, FUP follow-up *indicates that the study features more than once in the data synthesis particularly associated with social and economic partici- learning and memory are needed however the evidence pation as well as suicide and self-harm behaviours. More to date suggests that there may be an important role for specifically, three of the five included studies have shown this particular neuropsychological domain with regards decision-making and conceptual flexibility impairments to to social and economic participation. be predictive of suicide and self-harm behaviours [31–33]. Taken together, these studies identify a shared pathophysi- Functional domain: clinical syndrome ology among those who have previously attempted suicide, The association between cognitive function and the clin- those who were current suicide attempters and those who ical syndromes of mood and anxiety disorders in adoles- were currently self-harming; that is, they all showed char- cents and young adults has previously been reviewed acteristic deficits in decision-making and cognitive inflex- extensively (see [42, 43]), and some of the findings of ibility. There is also contrary evidence whereby suicide such reviews have been reiterated by the present system- attempters performed better on the decision making task atic review (see below). Interestingly, it has been re- than depressed patients who hadn’t attempted suicide and ported that impaired verbal memory is significantly healthy controls [34], and there were no neuropsycho- associated with depression (not specified as MDD) [44] logical differences between self-harmers and non-self- as well as the development of depressive symptoms, in a harmers [35]. However, these findings may be attributed community sample [45]. These findings implicate a dys- to methodological differences and/or a modest sample function in memory occurring earlier in the course of size, especially since the latter study [35] did not distin- depressive syndrome development while, poor executive guish between current self-harmers and previous self- function may be associated with more persistent MDD harmers. The notion that impaired decision making and [46]. In terms of delineating symptom severity, individ- conceptual flexibility may predispose one to suicidal be- uals diagnosed with unipolar depression with higher haviours is supported by evidence showing neurobiological levels of depressive symptoms also show increased (i.e. de- changes in the areas thought to subserve these functions. layed) choice reaction time [47], lower cognitive control of Namely, structural and functional dysfunction in the orbi- emotional processing [48], and processing speed deficits tofrontal prefrontal cortex have been identified which [49] suggesting that a broader range of cognitive (includ- imply that suicidal behaviour may be associated with defi- ing social) measures are also associated with depression cits in the attribution of importance to stimuli [36–38], al- and may also be sensitive to the severity of illness. though changes in other regions, such as the dorsolateral Similar deficits in cognition have been observed prefrontal cortex, have been implicated as well [38]. among those with anxiety disorders. Greater Social Anx- Similar deficits in executive function, particularly in iety Disorder (SAD) symptom severity was associated conceptual flexibility, was associated with impaired so- with poor executive function, specifically cognitive in- cial and economic participation [39, 40], whereas studies flexibility in SAD patients [39]. Most studies have identi- that investigated verbal learning and memory reported fied cognitive deficits associated with OCD, such as, conflicting results regarding social and economic partici- impaired verbal and visual memory [50], information pation. Two studies [24, 41] identified a positive rela- processing [51, 52], and cognitive flexibility [51–53]. As tionship with this functional domain, whilst another two observed in unipolar depression, increased symptom studies did not find any association [29, 39]. Notably, lo- severity in those with OCD is associated with worse gical memory retention (an index of structured learning selective attention [52], however this has also not been and memory) was a common measure identified as sig- uniformly reported [50]. Comorbidity has been identi- nificant in the positive studies, but it was not utilised in fied as another factor related to the cognitive deficits the other two studies; additional studies of structured in OCD, whereby comorbid depression was associated Iorfino et al. BMC Psychiatry (2016) 16:156 Page 12 of 38

Table 4 Imaging studies evaluating the five functional domains young people (12-30 yrs) with a mood and/or anxiety disorder Outcome Study Age (mean ± SD) Sample (N) Aims Key measures Key findings measure Social and [57] MDD: 15.7 ± 1.5; MDD (8M; 6F) Assess amygdala activation and Imaging: fMRI MDD: ↓ amygdala–seeded economic connectivity during an emotional connectivities ~ ↓ social and participation HC: 15.1 ± 1.6 HC (8M; 6F)regulation task. Functional: CGAS economic participation Physical [58] BPD-O: 23.8 ± 4.5 BPD-O (9M; 11F) Examine the relationship between BMI Imaging: sMRI BPD: ↑ BMI ~ ↓ WMV and health and brain volumes in mania. TLV BPD-N: 22.2 ± 4.4 BPD-N (19M; 18F) Functional: BMI HC: ↑ BMI ~ ↓ TBV and GMV. HC-O: 22.0 ± 3.8 HC-O (12M; 5F) HC-N: 22.3 ± 3.5 HC-N (19M; 19F) Suicide and [170] MHP: 14.6 ± 3.4 PSD (18M; 5F) Compare WMH in psychiatrically Imaging: sMRI MDD: ↑ WMH ~ suicide self-harm hospitalized youth with and without a attempt history, but not BPD (26M; 9F) history of suicide attempt Functional: PRS ideation MDD (33M; 15F) PC (34M; 12F) [171] MDD: 26.7 ± 5.5 MDD (34M; 68F) Compare the prevalence and location Imaging: sMRI MDD: ↑ PVH, not DWMH, ~ of WMH in young MDD inpatients suicide attempt history, but Functional: with and without histories of suicide not ideation Clinical records attempts [59]* HC: 16.2 ± 0.8 HC (4M; 9F) Evaluate the ACC volumes of MDD/ Imaging: sMRI MDDx: ↓ BA24 volumes ~ ↑ borderline personality patients with number of suicide attempts MDDx: 15.8 ± 1.1 MDDx (2M; 11F)and without a suicide attempt history Functional: (and ↑ borderline severity, Clinical interview but not depression) [62] SA: 16.20 ± 0.78 SA (4M; 11F) Evaluate the association between Imaging: fMRI PC: ↑ activity in right ACG PC: 15.87 ± 1.55 neural activity during performance of compared to SA (but SA not HC: 15.21 ± 1.42 PC (7M; 8F)the go no-go task and suicide history. Functional: CSHF different from HC) HC (8M; 6F) [60] SA: 16.21 ± 0.80 SA (4M; 10) Measure neural activity during Imaging: fMRI SA: ↑ dorsal ACG activity processing of emotional faces in when viewing angry faces, PC: 15.87 ± 1.55 PC (7M; 8F)adolescents with a history of Clinical: C-CASA, and ↓ visual, sensory, CSHF, SIQ, SIS HC: 15.27 ± 1.39 HC (8M; 7F) depression and suicide attempt prefrontal, ACG activity to intense happy and neutral faces ~ suicide attempt history. [34]* SA: 16.20 ± 0.78 SA (4M; 11F) Measure neural activity during Imaging: fMRI PC: ↑ hippocampal activity performance on the IGT in compared to HC. (HC and PC: 15.79 ± 1.58 PC (7M; 7F) Functional: C- adolescents. SA did not differ, evidence CASA, CSHF, SIQ, of ↓ activation) HC: 15.15 ± 1.46 HC (8M; 5F) SIS Alcohol and [172] BPD-L: 23.7 ± 3.6 BPD-L (14M; 5F) Assess the effects of alcohol use on Imaging: MRS BPD-H: ↓ GSH substance use GSH in young people with BPD. BPD-H: 23.4 ± 3.1 BPD-H (12M; 2F) Functional: AUDIT HC: 23.6 ± 2.8 HC (13M; 4F) [65] MDD: 21.7 ± 2.0 MDD (5M; 1F) Examine the effect of cannabis use on Imaging: fMRI MDD: ↑ CU ~ ↓amygdala threat-related amygdala reactivity. reactivity Functional: SCID (presence of dependence) [66] AUD: 17 ± 2.1 AUD (8M; 6F) Compare prefrontal-thalamic- Imaging: sMRI AUD: ↓ PFC & PFC WMV. cerebellar measures of adolescents ↓ ↑ HC: 16.9 ± 2.3 HC (16M; 12F)and young adults with adolescent- Fucntional: ACQ, AUD: PFC GM ~ alcohol LHAUI, SCID consumption onset alcohol use disorders AUD(M): ↓ CV PFC volume variables ~ measures of alcohol consumption [64] BPD: 16 ± 2, BPD (5M; 9F) Compare brain morphometry in Imaging: sMRI BPD: CUD ~ ↓ LFG GMV & ↑ bipolar adolescents with co-occurring RC, PCG GMV substance and alcohol disorders Functional: ASI, SCID, SAC Iorfino et al. BMC Psychiatry (2016) 16:156 Page 13 of 38

Table 4 Imaging studies evaluating the five functional domains young people (12-30 yrs) with a mood and/or anxiety disorder (Continued) [63] CU: 18 ± 0.7 CU (12M; 4F) Examine the relationship between Imaging: sMRI CU: ↓ WMV ~ ↑ depressive brain volumes, depression and symptoms HC: 18 ± 0.9 HC (11M; 5F) Functional: BDI, cannabis use. CDDR, HDRS, TLFB, Clinical [173] BPD: 19.9 ± 7.9 BPD (15M; 18F) Explore structural brain changes in Imaging: VBM BPD: ↑ volume in left syndrome first-episode bipolar patients thalamus and fusiform and HC: 21.5 ± 4.3 HC (19M; 14F) Clinical: SCID, cerebellum bilaterally. ↑ AC, KSADS PPS GMD. ↑ middle/superior temporal and posterior cingulate gyri, GMV & GMD. [90] DD: 15.6 ± 1.4, DD (4M; 21F) Investigate WM microstructure in a Imaging: DTI DD: ↓ FA and ↑ RD, MD in sample of clinically depressed corpus callosum. ↑ FA & AD, HC: 14.7 ± 1.6 HC (3M; 18F) Clinical: ADIS, adolescents relative to matched and ↓ RD in uncinate CDI, CBCL, controls. fasciculus. RCADS, YSRS [46] BPD: 15.8 ± 1.8 BPD (6M; 11F) Compare amygdala Imaging: sMRI BPD: ↑ BL amygdala neurodevelopment among BPD, volumes ~ symptomatic ADHD: 16.3 ± 1.7 ADHD (13M; 11F)ADHD, and healthy adolescents Clinical: KSADS, recover compared to those LIFE, YMRS, who did not achieve HC: 16.3 ± 1.8 HC (13M; 10F) HDRS recovery. No increase in amygdala volume over time. [85] BPD: 15.9 ± 1.4 BPD (4M; 4F) Evaluate the effect of lamotrigine Imaging: fMRI BPD: clinical improvement treatment on amygdalar activation ~ ↓ right amygdalar Clinical: CDRS activation [99] BPD: 25 ± 9 BPD (7M; 14F) Investigate the distribution of lactate Imaging: MRS BPD: ↑ Lac/NAA & Lac/Cr in bipolar and healthy brains ratio HC: 25 ± 6 HC (5M, 5F) Clinical: SCID [174] BPD: 15.6 ± 0.9 BPD (2M; 8F) Explore the neural correlates of Imaging: fMRI BPD: After treatment, ↓ left depression at baseline and after 6 occipital cortex activity in HC: 15.6 ± 1.2 HC (2M; 8F)weeks of open as usual treatment Clinical: KSADS, the intense fearful CDRS, SCARED, experiment, but ↑ left insula, YMRS left cerebellum, and right ventrolateral PFC in the intense happy experiment. ↑ improvement in depression ~ ↑ BL activity in ventral ACC to mild happy faces [175] MDD: 14.73 ± 1.49 MDD (3M; 11F) Examine behavioral and neural Imaging: fMRI MDD: ↓ activation in the responses to reward in young people ACC, bilateral caudate, and HC: 14.45 ± 1.79 HC (7M; 10F)with depressive disorders using a Clinical: KSADS, inferior OFC bilaterally CBCL, SCARED, reward decision-making task during reward decision/ CDI, BDI anticipation and reward outcome. [176] MDD: 12.9 ± 2.3 MDD (4M; 9F) Evaluate reward-related brain function Imaging: fMRI MDD: severity, anxiety and as a predictor of treatment response depression symptoms ↓ Clinical: KSADS, in adolescents with MDD following treatment. ↑ MFQ, SCARED, reward related striatal CGI function before treatment ~ ↑ clinical severity, ↓ anxiety symptoms and faster improvement in anxiety symptoms after treatment. ↑ mPFC function before ~ slower improvements in anxiety symptoms. [68] MDD: 16.7 ± 2.7 MDD (9M; 12F) Test whether ACC GABA levels are Imaging: MRS, MDD & HC: ↑ ACC GABA ~ ↓ decreased in adolescents with MDD sMRI anhedonia scores. HC: 16.2 ± 1.6 HC (6M; 15F) Clinical: KSADS, MDD: ↓ ACC GABA, ↓ ACC CDRS-R, BDI WM [177] MDD: 17.1 ± 2.5 MDD (9M; 12F) Assess striatum-based circuitry in rela- Imaging: fMRI MDD: ↑ iFC between all tion to categorical diagnosis of MDD striatal regions bilaterally HC: 16.3 ± 1.4 HC (9M; 12F)and anhedonia severity Clinical: KSADS, and DmPFC, RVC and ACC. CDRS-R, BDI Iorfino et al. BMC Psychiatry (2016) 16:156 Page 14 of 38

Table 4 Imaging studies evaluating the five functional domains young people (12-30 yrs) with a mood and/or anxiety disorder (Continued) MDD severity ~ iFC between the striatum and the precuneus, posterior cingulate cortex and dmPFC. Anhedonia severity ~ Pregenual ACC, subgenual ACC, supplementary motor area, and supramarginal gyrus iFC. [178] BPD: 15.1 ± 1.81 BPD (6M; 12F) Investigate the brain structural Imaging: DTI BPD: ↓ GMV in left changes in BPD children and hippocampus. ↓ FA value in HC: 14.1 ± 1.61 HC (6M; 12F) Clinical: KSADS, adolescents rACC. YMRS, MFQ ↓ hippocampal volume ~ ↑ YMRS score [91] OCD: 12.35 ± 2.93 OCD (7M; 14F) Measure neuroanatomical changes in Imaging: MRI OCD: ↑ thalamic volumes in the thalamus of patients with OCD treatment naïve patients. ↓ HC: 12.47 ± 8.33 HC (7M; 14F) Clinical: KSADS, near the onset of illness, and before thalamic volumes (to YBOCS, HDRS and after treatment. comparable levels with controls) ~ paroxetine monotherapy. ↓ thalamic volumes ~ ↓ OCD symptom severity [179] OCD: 13.1 ± 2.5 OCD (11M; 7F) Examine whether overlapping but Imaging: fMRI OCD: ↓ activity in right symptom dimension-specific neural insula, putamen, thalamus, HC: 13.6 ± 2.4 HC (11M; 7F)activity patterns in adults are apparent Clinical: YBOCS dorsolateral prefrontal cortex in youths and left orbitofrontal cortex, and right thalamus and right insula. ↑ OCD symptom related measures were significantly predictive of ↓ neural activity in the right dorsolateral prefrontal cortex during the contamination experiment. [59]* HC: 16.2 ± 0.8 HC (4M; 9F) Evaluate the ACC volumes of MDD/ Imaging: sMRI MDDx: ↓ BA24 volumes ~ ↑ borderline personality patients with borderline severity, but not MDDx: 15.8 ± 1.1 MDDx (2M; 11F) Clinical: Clinical and without a suicide attempt history depression interview [94] OCD: 14.3 ± 2.1 OCD (13M; 10F) Investigate white matter abnormalities Imaging: DTI ODD: ↑ FA in splenium ~ ↑ in pediatric obsessive-compulsive obsession severity HC: 14.2 ± 2.2 HC (12M; 11F)disorder. Clinical: YBOCS, KSADS [180] MHP: 22.3 ± 3.7 MHP (50M; 83F) Examine the relationship between Imaging: sMRI MHP: ↓ GMV in left anterior anterior insula GMV, clinical symptom insula. Changes (↑ or ↓)in HC: 23.8 ± 2.4 HC (13M; 26F) Clinical: BPRS, severity and neuropsychological right anterior insula GMV ~ ↑ HDRS, SOFAS performance. symptom severity. [95] MDD: 16.8 ± 2.2 MDD (9M; 8F) Investigate WM microstructure Imaging: DTI MDD: ↓ WM integrity in the in MDD using diffusion tensor imaging genu of corpus callosum, HC: 16.4 ± 1.4 HC (6M; 10F) Clinical: KSADS, anterior thalamic radiation, BDI, CDRS-R, anterior cingulum and MASC sagittal stratum ~ ↑ depression severity. [181] MDD: 15.8 ± 1.4 MDD (8M; 11F) Investigate sgACC FC in adolescent Imaging: fMRI MDD: ↑ sgACC- amygdala depression during negative emotional Functional connectivity and HC: 16.1 ± 1.2 HC (8M; 11F) Clinical: BDI processing. ↓ sgACC-fusiform gyrus, sgACC-precuneus, sgACC- insula, and sgACC-middle frontal gyrus functional connectivity. ↓ sgACC-precuneus functional connectivity ~ ↑ depression severity. [93] HC: 16 ± 2.74 HC (6M; 7F) Evaluate whether the observed WM Imaging: DTI MDD at FUP ~ ↓ FA in the disruptions are associated with superior longitudinal fasciculi Iorfino et al. BMC Psychiatry (2016) 16:156 Page 15 of 38

Table 4 Imaging studies evaluating the five functional domains young people (12-30 yrs) with a mood and/or anxiety disorder (Continued) MT: 15.89 ± 2.79 MT (5M; 14F)increased vulnerability to Clinical: KSADS, & the right cingulum- psychopathology during prospective FH-RDC, CDRS-R, hippocampal Projection. follow-up HDRS, CGAS, BDI SUD at FUP ~ ↓ FA in the DUSI right cingulum-hippocampal projection. [182] OCD: 13.95 ± 2.52 OCD (9M; 16F) Investigate the development of the Imaging: fMRI OCD: ↑ ACC activity during ACC and its associations with error responses in bilateral HC: 13.71 ± 2.85 HC (9M; 16F) Clinical: YBOCS, psychopathology. insular cortex during high ADIS, CDI, STAI- conflict tasks C, CBCL [81] OCD: 13.78 ± 2.58 OCD (11M; 18F) Identify differences in regional brain Imaging: VBM OCD: ↑ Orbitofrontal GMV volume between medication-free after treatment ~ ↑ symptom HC: 13.6 ± 2.73 HC (11M; 18F)pediatric OCD patients and controls Clinical: YBOCS, improvement ADIS, CDI, STAI- and examine changes after cognitive C, CBCL behavioural therapy [183] BPD-I: 14.57 ± 1.98 BPD-I (11M; 7F) Examine patterns of activity and Imaging: fMRI BP-I: ↑ activity in amygdala connectivity in youth with BPD. and VMPFC regulation BPD-NOS: 12.59 BPD-NOS (11M; 7F) Clinical: KSADS, regions to happy faces and ± 2.27 MFQ, SCARED, reduced DLPFC activity to CALS HC: 13.67 ± 2.55 HC (7M; 11F) fearful faces compared to HC. BPD-NOS: ↓ PFC activity to neural faces compared to HC. [71] HC: 23.9 ± 2.3 HC (12M; 21F) Evaluate patterns of grey matter Imaging: sMRI ST-2/3: ↓ GMV in frontal changes very early in the course of brain regions ST-1: 20.4 ± 5.2 ST-1 (8M; 15F)affective illness compared to those Clinical: HDRS, SOFAS, BPRS ST-2/3: 23.5 ± 3.5 ST-2/3 (14M; 10F) with discrete disorders and/or illness persistence [96] HC: 23.82 ± 2.52 HC (15M; 24F) Examine the association between Imaging: DTI ST-2/3: ↓ FA within the left microstructural WM changes and anterior corona radiata ST-1B: 21.36 ± 3.51 ST-1B (24M; 49F)different stages of psychiatric illness. Clinical: HDRS, compared to HC. BPRS, SOFAS ST-2/3: 22.45 ± 4.35 ST-2/3 (37M; 32) ST-1B: pattern of ↓ FA within the left anterior corona radiate (less WM involvement than ST-2/3) [92] BPD: 23.03 ± 5.04 BPD (23M; 35F) Examine WM microstructural changes Imaging: DTI BPD: ↓ FA in the genu, body in BPD. and splenium of the corpus HC: 24.05 ± 2.92 HC (12M; 28F) Clinical: HDRS, callosum as well as the YMRS, SOFAS, superior and anterior corona BPRS radiata. ↑ radial diffusivity. [184] OCD: 13.1 ± 2.7 OCD (7M; 5F) Investigate possible regional brain Imaging: fMRI OCD: ↑ activation bilaterally dysfunction in premotor cortico- in the middle frontal gyrus. HC: 13.7 ± 2.8 HC (7M; 5F) Clinical: ChIPS, Y- striatal activity, correlate brain activa- Clinical improvement BOCS, CDI, STAI- tion with severity of obsessive- following pharmacological compulsive symptomatology; And, de- C, LOI-CV treatment ~ ↓ activation in tect possible changes in brain activity left insula and left putamen after pharmacological treatment [98] OCD: 12.5 ± 2.9 OCD (6M; 5F) Measure neurometabolite Imaging: MRS OCD: ↓ total Cho in left concentrations in anterior cingulate- striatum (this ↓ did not HC: 14.5 ± 2.8 HC (5M; 7F)medial frontal cortex and right and left Clinical: Y-BOCS, change over time and striatum of drug naïve children and CDI, STAI-C, LOI- persisted at follow-up CV adolescents with OCD Assessment) [185] BPD: 27 ± 10 BPD (26M; 32F) Assess changes in GMV in BPD. Imaging: sMRI BPD: ↑ GMV in portions of the VLPFC and hippocamps HC 27 ± 10 HC (21M; 27F) Clinical: SCID, complex. ↑ GMV in KSADS amygdala proper and caudate. ↑ number of depressive episodes ~ ↑ GMV in the right cingulate gyrus bilaterally and right thalamus and bilateral lenticulate nuclei, and left cerebellar vermis. ↑ illness Iorfino et al. BMC Psychiatry (2016) 16:156 Page 16 of 38

Table 4 Imaging studies evaluating the five functional domains young people (12-30 yrs) with a mood and/or anxiety disorder (Continued) duration ~ ↓ GMV in left cerebellar vermis. [83] OCD: 12.79 ± 2.64 OCD (10M; 21F) Measure pituitary gland volume in Imaging: MRI OCD: ↓ pituitary gland OCD volume ~ ↑ compulsive HC: 12.89 ± 2.66 HC (10M; 21F) Clinical: YBOCS, symptom severity (more HAMA, HDRS pronounced in males). [74]MDD:8– 17years MDD (10M; 13F) Examine temporal lobe anatomy in Imaging: MRI MDD: ↑ left and right pediatric patients with MDD near the amygdala: hippocampus HC: 8–17 years HC (10M; 13F) Clinical: CDRS-R, onset of illness before treatment volume ratios ~ ↑ severity of HAMA anxiety (but not ↑ depression severity or duration of illness) [87] ANX: 11.8 ± 1.8 ANX (6M; 6F) Examine the relationships between Imaging: fMRI ANX: ↑ left amygdala pretreatment amygdala activity and activation pre-treatment ~ treatment response in a sample of Clinical: KSADS, treatment response to CBT anxious children and adolescents CGI or medication. (no association between pre-treatment symptom severity and pre- treatment amygdala activity) [78] SAD: 21.80 ± 3.68 SAD (14M; 6F) Explore the GMD deficits in drug-naïve Imaging: VBM SAD: ↓ GMD in bilateral adult SAD patients thalami, right amygdala, and HC: 21.58 ± 3.72 HC (13M; 6F) Clinical: HAMA, right precuneus. ↓ right HDRS, LSAS, amygdala GMD ~ ↑ disease SCID duration and ↓ age of onset. [69] DD: 15.4 ± 1.5 DD (3M; 23F) Examine GMV in brain areas putatively Imaging: VBM DD: ↓ bilateral dorsal ACC involved in affective psychopathology. volume. No association with HC: 14.7 ± 1.5 HC (3M; 23F) Clinical: ADIS, clinical severity of CDI, RCADS, YSR, depression or anxiety. CBCL [100] BPD: 15.5 ± 1.5, BPD (5M; 23F) Compare in vivo neurometabolite Imaging: MRS BPD: ↑ NAA in the ACC and concentrations in bipolar adolescents VLPFC. ↑ Cho and Cr in the HC: 14.6 ± 1.8 HC (4M; 6F)with a depressed episode Clinical: KSADS, VLPFC. CDRS-R [186] BPD: 14.3 ± 1.1 BPD (6M; 11F) Investigate the effects of Imaging: fMRI BPD: YMRS improvement ~ ↓ pharmacotherapy on brain function VMPFC activity. HC: 14.1 ± 2.4 HC (7M; 7F) Clinical: YMRS, underlying affect dysregulation and Normalization of activity in KSADS, CDI, cognitive function in pediatric bipolar the inferior frontal gyrus disorder. CDRS-R following pharmacological treatment. [187] BPD: BPD (16M; 8F) Determine the relative effects of Imaging: fMRI BPD: Divalproex treatment risperidone and divalproex on brain ~ ↑ activity in left MPFC HC: 13.9 ± 3.4 HC (7M; 7F) Clinical: KSADS, function in pediatric mania relative and modulation of CDRS, YMRS positive emotions to risperidone. ↑ pre-treatment right amygdala activity with negative and positive condition in the risperidone group, and left amygdala with positive condition in divalproex group predicted poor response on YMRS. [86] gSP: 25.91 ± 5.50 gSP (8M; 13F) Examine the change in amygdala- Imaging: fMRI gSP: SSRI treatment ~ ↓ insula-medial frontal function during amygdala reactivity to fearful HC: 26.95 ± 8.11 HC (10M; 9F) Clinical: SCID, perception of social threat cues before faces (which was ↑ pre- LSAS, HDRS, BDI, and after SSRI treatment treatment) and ↑ ventral STAI MPF activity to angry faces (which was ↓ Pre-treatment treatment). No correlations with symptom improvement. [188] OCD: 28.8 ± 8.2 OCD (4M; 5F) Identify neuroimaging predictors of Imaging: PET OCD: ↓ rCBF in OFC and ↑ medication response in rCBF values in PCC predicted Clinical: AAS, contamination-related obsessive com- better fluvoxamine OCDAS pulsive disorder OCD treatment response. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 17 of 38

Table 4 Imaging studies evaluating the five functional domains young people (12-30 yrs) with a mood and/or anxiety disorder (Continued) [70] HC: 17.19 ± 1.87 HC (7M; 9F) Investigate the role of dysregulation of Imaging: sMRI MDD: ↑ right and left rostral frontal-limbic circuits in the sympto- MFG, and left caudal anterior MDD: 16.89 ± 2.01 MDD (9M; 21F) Clinical: CDRS, mology of this disorder cingulate cortex thickness. ↑ KSADS, BDI age ~ ↓ left MFG thickness. [84] OCD: 12.70 ± 3.11 OCD (13M; 8F) Investigate the regional morphology Imaging: sMRI OCD: ↑ corpus callosum of the CC in OCD. (except the isthmus). ↑ CC HC: 12.74 ± 3.12 HC (13M; 8F) Clinical: YBOCS, area, genu, anterior body, HAMA, HDRS, posterior body, isthmus and KSADS anterior splenium ~ ↑ compulsive symptom severity [79] OCD: 12.89 ± 3.23 OCD (5M; 6F) Evaluate neuroanatomic changes in Imaging: sMRI OCD: No significant change the thalamus of OCD patients near in thalamic volume after CBT Clinical: YBOCS, illness onset before and after cognitive behavioral therapy HDRS, HAMA, KSADS [73] MDD: 15.35 ± .34, MDD (3M; 17F) Examine amygdala and hippocampus Imaging: sMRI MDD: ↓ left and right volumes in pediatric MDD. amygdala volumes. No HC: 14.08 ± .31 HC (8M; 16F) Clinical: HDRS, correlations with symptom FH-RDC, KSADS severity, age of onset or illness duration. [75] GAD: 22.9 ± 4.1, GAD (16F) Investigate the neural substrates Imaging: sMRI GAD: ↑ amygdala and associated with excessive and DMPFC volumes. ↑ symptom HC: 23.7 ± 3.7 HC (15F) Clinical: SCID, persistent worrying in GAD severity ~ ↑ DMPFC and ACC BDI, MCQ volumes [189] BPD: 14.6 ± 2.2 BPD (11M; 12F) Examine the neurofunctional effects of Imaging: fMRI BPD: Ziprasidone treatment ziprasidone in manic adolescents ~ ↑ in right BA 11 and 47 HC: 15.0 ± 1.8 HC (6M; 4F) Clinical: YMRS, activation. No association CGI, KSADS with symptom improvement. ↓ BL right BA 47 activation ~ ↑ improvement of YMRS score. [101] BPD-R: 15.4 ± 1 BPD-R (4M; 3F) Evaluate the in vivo effects of Imaging: MRS BPD-r: ↓ BL Glx in LVLPFC. extended-release divalproex sodium Change in LVLPFC Glu ~ BPD-NR: 14.1 ± 2.2 BPD-NR (6M; 1F) Clinical: KSADS, on the glutamatergic system in ado- change in YMRS score CDRS, CGI, YMRS HC: 14.4 ± 1.6 HC (6M; 9F) lescents with BPD and neurochemical predictors of clinical remission. [88] BPD-RE: 13.5 ± 2.4 BPD-RE (13M; 9F) Determine functional connectivity Imaging: fMRI BPD-RE: ↑ connectivity of among patients with pediatric BPD the amygdala before and BPD-NRE: BPD-NRE (6M; 6F)who are responders to Clinical: KSADS, after treatment compared to 13.3 ± 2.0 pharmacotherapy and those who are YMRS, CDRS-R BPD-NRE. ↑ right amygdala HC: 14.2 ± 3.1 HC (7M; 7F) nonresponders, functional connectivity after treatment ~ ↑ improvement in mania symptoms [72] MDD: 28.8 ± 10.7 MDD (35M; 30F) Evaluate the early effects of Imaging: sMRI MDD: >3 untreated antidepressant therapy, as well as of depressive episodes ~ ↓ HC: 28.4 ± 10.7 HC (37M; 56F) Clinical: YMRS, key clinical variables, on ACC volume subcallosal gyrus volumes GAF, HDRS, SCID compared to HC. [82] OCD: 16.6 ± 1.5 OCD (14M; 12F) Identify structural GM and WM Imaging: sMRI, OCD: ↑ symptom severity microstructure changes in pediatric DTI ~ ↑ GM volume in right OCD insula, posterior orbitofrontal HC: 16.5 ± 1.4 HC (14M; 12F) Clinical: YBOCS cortex, brainstem and cerebellum, [80] OCD: 22.0 ± 5.2 OCD (3M; 5F) Evaluated resting brain metabolism Imaging: PET, OCD: ↑ clinical and treatment response in OCD MRI improvement ~ ↑ changes in patients. bilateral dosal ACC and in HC: 21.5 ± 5.9 HC (8F) Clinical: YBOCS, the right middle occipital HDRS gyrus Note. Sample: ADHD attention deficit hyperactivity disorder, ANX anxiety disorder, AUD alcohol use disorder, BPD bipolar disorder, BPD-I bipolar disorder I, BPD-O bipolar disorder with obesity, BPD-L bipolar disorder with low alcohol use, BPD-H bipolar disorder with high alcohol use, BPD-N bipolar disorder without obesity, BPD-NOS bipolar disorder not otherwise specified, BPD-R bipolar disorder remitters, BPD-NR bipolar disorder non remitters, BPD-RE bipolar disorder responders to pharmacotherapy, BPD-NRE bipolar disorder non responders to pharmacotherapy, DD depressive disorder, CU cannabis user, GAD generalised anxiety disorder, gSP generalised social phobia, HC healthy controls, HC-O healthy controls with obesity, HC-N healthy controls without obesity, MDD major depression disorder, MDDx Iorfino et al. BMC Psychiatry (2016) 16:156 Page 18 of 38

major depression disorder with borderline personality disorder, MHP mental health patients (mixed diagnosis sample), MT childhood maltreatment, OCD obsessive compulsive disorder, PC psychiatric control (i.e. psychiatric diagnosis but no suicide attempt), PSD psychotic spectrum disorder, SA suicide attempters, ST stage of illness; 1B, 2, & 3, SAD social anxiety disorder Measures: AAS anxiety analogue scale, ACQ alcohol consumption questionnaire, ADIS anxiety disorders interview schedule, ASI addictions severity index, AUDIT alcohol use disorder identification test, BDI beck depression inventory, BMI body mass index, PRS brief psychiatric rating scale, CALS child affect liability scale, CBCL child behaviour checklist, CDI children’s depression inventory, CDRS children’s depression rating scale; R revised, CGAS children’s global assessment scale, CGI clinical global impression scale, ChIPS children’s interview for psychiatric syndromes, CDDR customary drinking and drug use record, CGAS child global assessment scale, C-CASA Columbia Classification Algorithm of Suicide Assessment, CSHF Colombia Suicide History Form, DTI diffuse tensor imaging, DUSI drug use screening inventory, FH-RDC family history-research diagnostic criteria, fMRI functional magnetic resonance imaging, GAF global assessment of functioning, HAMA Hamilton anxiety rating scale, HDRS Hamilton depression rating scale, K-SADS kiddie schedule for affective disorders and schizophrenia, LHAUD lifetime history of alcohol use disorder, LIFE modified longitudinal interval follow-up examination, LOI-CV Leyton Obsessive Inventory-Child Version, LSAS Liebowitz social anxiety scale, MASC multidimensional anxiety scale for children, MCQ meta cognition questionnaire, MFQ Mood frequencies questionnaire, MRS magnetic resonance spectroscopy, PET positron emission tomography, PRS Pfeffer rating scale, OCDAS obsessive compulsive disorder analogue scale, RCADS the revised child anxiety and depression scale, SAC substance abuse course-modified life II, SCARED screen for child anxiety related disorders, SCID structured clinical interview for DSM, SIQ suicide ideation questionnaire, SIS suicide intent scale, SOFAS social and occupational functioning assessment scale, sMRI structural magnetic resonance imaging, STAI-C state- trait anxiety inventory – child version, TLFB time-line followback, VBM voxel-based morphometry, Y-BOCS Yale–Brown obsessive-compulsive scale, YMRS young mania rating scale, YSRS the youth self-report scale Findings: ↑ = Increased, Improved or Higher, ↓ = Decreased, Reduced or Lower, ~ = ‘is associated with’, ACC anterior cingulate cortex, AD Axial diffusivity, ACG Anterior Cingulate Gyrus, BA Broadman Area -24, BL baseline, CV cerebellar vermis, DmPFC dorsomedial prefrontal cortex, DWMH deep white matter hyperintensities, FA fractional anisotropy, GABA gamma-aminobutyric acid, GM grey matter, GMV grey matter volumes, GSH glutathione, iFC intrinsic functional connectivity, LFG left fusiform gyrus, MD mean diffusivity, MFG middle frontal gyus, MPFC medial prefrontal cortex, OFC orbitofrontal cortex, RD radial diffusivity, PVH periventricular hyperintensities, PCG precentral Gyrus, PFC prefrontal cortex, TBV total brain volumes, TLV temporal lobe volume, VMPFC ventromedial prefrontal cortex, WMH white matter hyperintensities, WMV White Matter Volumes *indicates that the study features more than once in the data synthesis with executive function deficits in these patients [53]. Functional domains: social and economic participation, It is clear that the literature for anxiety disorders is physical health, suicide and self-harm & alcohol and less consistent with regard to the pattern of cognitive substance use deficits and their relationship to anxiety and the sever- Results indicate that neuroimaging is a particularly use- ity of illness. The findings regarding GAD and cogni- ful modality for investigating suicide and self-harm be- tion are unsurprisingly similar to the cognitive deficits haviours as well as alcohol and substance use. In observed in depression, and provide support for a contrast, the utility of neuroimaging for investigating so- shared underlying neurobiology for these disorders cial and economic participation [57] and physical health [54]. [58] outcomes is yet to be determined due to a lack of Clinical trials utilizing neuropsychological function studies exploring these relationships. Moreover, consist- as an outcome measure have demonstrated that ver- ent with the findings previously discussed (see 'Neuro- bal memory improves following: (i) lamotrigine treat- psychology') linking poor cognitive flexibility and ment in bipolar patients [55]; and (ii) Transcranial decision making to suicidal behaviours, reduced Anterior Magnetic Stimulation (TMS) treatment in depressed Cingulate Cortex (ACC) volume was associated with a patients [56]. Furthermore, the improved verbal mem- higher number of suicide attempts in patients with co- ory performance was significantly associated with im- morbid MDD and borderline personality disorder [59]. provements in clinical symptoms of mania and Furthermore, a study investigating ACC function using depression, in the former study, and with reductions fMRI demonstrated that individuals with a suicide at- in hallmark symptoms of depression, in the latter. tempt history had increased dorsal ACC activity when This adds to the close and complex relationship ob- viewing angry faces, and reduced visual, sensory, preserved between cognition and mood disorders in frontal ACC activity to intense happy and neutral faces young people, and reiterates the need for future re- compared to both healthy and psychiatric controls [60]. search to closely examine the direction of these It is suggested that the ACC is an important area in- relationships. volved in attentional control that regulates both cognitive and emotional processes [61]. Structural and Neuroimaging functional abnormalities in this area may be indicative of There were 62 studies (a total of 3069 participants; attentional control deficits that affect normal cognitive 55.5 % female) that utilised neuroimaging and across and emotional processes that are associated with an in- these studies 62 % (1894/3069) were patients and 38 % creased risk for suicidal behaviours in this population. (1175/3069) were healthy controls. Among the patient However further evidence for this theory is needed since group 28 % (534/1894) had depression, 27 % (520/1894) some fMRI studies could not distinguish between suicide had bipolar, 15 % (288/1894) had anxiety, and 29 % attempters and healthy controls using similar decision (552/1894) were classified as other (i.e. mixed psychiatric making and cognitive flexibility tasks. Psychiatric con- samples, ADHD, alcohol use disorders, substance use trols demonstrated increased activity in the ACC during disorder). the go-no go task [62] and hippocampus during the Iorfino et al. BMC Psychiatry (2016) 16:156 Page 19 of 38

Table 5 Sleep-wake and circadian biology studies evaluating the five functional domains in young people (12-30 yrs) with a mood and/or anxiety disorder Outcome Study Age (mean ± SD) Sample (N) Aims Key measures Key findings measure Social and [103] PD: 30.6 ± 6.1 PD (8M; 12F) Determine whether HPA SWC: 24-hour PD: ↑ cortisol secretion pre- economic activity can predict FUP cortisol sam- treatment ~ ↓ social and eco- participation functional status. ples, ACTH nomic participation (better profiles, CRH than pre-treatment clinical stimulation severity) test Functional: SDS [102]* MDD (M): 12.8 ± 2.6, MDD (22M; 33F) Investigate whether diurnal SWC: MDD: ↑ cortisol/DHEA ratios MDD (F): 13.6 ± 1.9 changes in cortisol and DHEA Cortisol/ at BL ~ ↓ social and levels are associated with the DHEA ratio, economic participation at occurrence of undesirable life FUP. events. Functional: Semi- structured interview [104] MHP: 12.1 MHP (62M; 40F) Investigate whether cortisol SWC: Cortisol MHP: ↑ cortisol secretion (7 – 17.9 years) reactivity is associated with level during the social interaction internalizing problem task ~ ↓ social and economic behaviour Functional: participation CBCL, SASC, CDI Suicide and [105]* MDD: 25.19 ± 2.42 MDD (33M; 23F) Examine baseline SWC: Sleep MDD: ↑ BL GH secretion self-harm neuroendocrine predictors of EEG, GH during first 4 hours of sleep HC: 25.92 ± 2.16 HC (10M; 11F) follow up clinical features secretion, ~ a suicide attempt during blood FUP cortisol Functional: Clinical interview [106] MDD: 25.19 ± 2.42 MDD (33M; 23F) Assess whether any SWC: Sleep MDD: ↑ BL cortisol secretion premorbid cortisol EEG, GH in the late evening hours ~ HC: 25.92 ± 2.16 HC (10M; 11F) abnormalities were associated secretion, suicide attempts during FUP with depressive course of blood illness cortisol Functional: Clinical interview [107]* MDD: 16 ± 0.3 MDD (6M; 14F) Compare sleep EEG profiles SWC: Sleep MDD: ↓ Delta sleep variable of a sample of outpatient EEG, blood ~ ↑ suicidality (and HC: 15.6 ± 0.6 HC (7M; 6F) adolescents samples depression severity). Functional: HDRS Clinical [108] CS: 17.04 ± 0.36 CS (57M; 173F) Examine whether individual SWC: Salivary CS: ↑ cortisol after waking at syndrome differences in the CAR serve cortisol BL ~ ↑ risk of developing as a premorbid risk factor for MDD at FUP MDD Clinical: SCID, LSI [118] HYP: 20.91 ± 3.72 HYP (8M; 23F) Assess circadian activity and SWC: HYP: ↑ variability in duration, sleep in individuals at Actigraphy fragmentation and efficiency behavioral high-risk of hypo- of sleep, ↓ sleep duration and HC: 22.12 ± 2.83 HC (8M; 16F) Clinical: SCID, mania/bipolar disorders later more variable be times. HPS, HIQ, ISS [120] MDD: 12 ± 1.9 MDD (2M; 4F) Explore the effects of SWC: Sleep MDD: ↑ stage 1 sleep, fluoxetine on sleep EEG EEG arousals and REM density ~ fluoxetine treatment Clinical: K- SADS, CDRS, BDI, WSAS Iorfino et al. BMC Psychiatry (2016) 16:156 Page 20 of 38

Table 5 Sleep-wake and circadian biology studies evaluating the five functional domains in young people (12-30 yrs) with a mood and/or anxiety disorder (Continued) [105]* MDD: 25.19 ± 2.42 MDD (33M; 23F) Examine baseline SWC: Sleep MDD: Premorbidly, earlier neuroendocrine predictors of EEG, GH and more steep GH secretion HC: 25.92 ± 2.16 HC (10M; 11F) follow up clinical features secretion, at sleep onset blood cortisol Clinical: Clinical interview [114] MDD: 17.04 ± 0.35 MDD (4M; 7F) Examine the associations SWC: Salivary P-MDD & MDD/ANX: flatter between MDD and anxiety cortisol diurnal cortisol slopes disorders, and HPA- axis ANX: 17.04 ± 0.37 ANX (8M; 21F) Clinical: functioning MASQ, LSI MDD/ANX: 16.85 ± MDD-ANX 0.21 (4M; 8F) P-MDD: 17.13 ± 0.37 P-MDD (11M; 45F) P-ANX: 17.02 ± 0.38 P-ANX (6M; 2F) [109] HR: 16.8 ± 1.7 HR (14M; 15F) Examine the cortisol increase SWC: Salivary HR: ↑ daytime cortisol in their after awakening and basal cortisol natural environment. cortisol levels hypothesis that LR: 16.6 ± 2.1 LR (14M; 15F) Clinical: CDI, high-risk offspring are more CBCL, PANAS reactive to psychosocial stress than low-risk offspring [110] HR: 18.3 ± 2.6 HR (12M; 12F) Determine whether HR SWC: Salivary HR: ↑ afternoon cortisol levels individuals exhibit elevated cortisol in their natural environment cortisol levels relative to LR LR: 18.0 ± 2.3 LR (11M; 11F) Clinical: BDI, individuals during two weeks of daily sampling CDI, PSWQ, CBCL, RLEQ [102]* MDD (M): 12.8 ± 2.6 MDD (22M; 33F) Investigate whether diurnal SWC: MDD: ↑ cortisol/DHEA ratios changes in cortisol and DHEA Cortisol/ at BL ~ persistent major levels are associated with the DHEA ratio, depression at FUP occurrence of undesirable life MDD (F): 13.6 ± 1.9 Clinical: events. Semi- structured interview [113] Mild: 14.73 ± 2.30 Mild (10M; 20F) Moderate Examine cortisol reactivity to SWC: Salivary Moderate/severe depression: Moderate: 15.69 ± (7M; 9F) a psychological stress cortisol ↓ cortisol response regardless 1.58 challenge in depressed of child maltreatment history Clinical: adolescents. Severe: 16.00 ± 2.00 Severe (6M; 19F) CECA, BDI-II, K-SADS [119] MDD: 23.94 ± 2.31 MDD (8M; 9F) Investigate the effect of SWC: Sleep MDD: ↑ overnight dissipation reducing slow waves during EEG of SWA predicted ↓ in sleep on depression depressive symptoms. Clinical: QIDS, symptomology HDRS [107]* MDD: 16 ± 0.3 MDD (6M; 14F) Compare sleep EEG profiles SWC: Sleep MDD: ↓ Delta sleep variable of a sample of outpatient EEG, blood ~ ↑ depression severity. HC: 15.6 ± 0.6 HC (7M; 6F) adolescents samples Clinical: HDRS [117] DD: 15.35 ± 1.85 DD (18M; 28F) Assess sleep disturbances SWC: DD: ↓ sleep efficiency and pain and pubertal Actigraphy total time asleep, ↑ time development in adolescent awake after sleep onset. ↑ HC: 14.83 ± 1.76 HC (17M; 43F)depressive disorders Clinical: K- pain intensity and depressive SADS, PDS, symptoms predicted worse CES-D, BPD sleep quality [111] MDD: 22.4 ± 1.5 MDD (9M; 17F) Examine the relationship SWC: Sleep MDD: recurrent illness ~ ↑ between longitudinal clinical EEG plasma cortisol near sleep course, sleep and cortisol in onset at BL. HC: 21.9 ± 1.7 HC (13M; 20F) adolescent depression Iorfino et al. BMC Psychiatry (2016) 16:156 Page 21 of 38

Table 5 Sleep-wake and circadian biology studies evaluating the five functional domains in young people (12-30 yrs) with a mood and/or anxiety disorder (Continued) Clinical: K- HC: high density REM and ↓ SADS REM latency at BL ~ the development of depression a FUP [112] MDD: 15.6 ± 1.4 MDD (6M; 10F) Examine EEG sleep and HPA SWC: NUFC, MDD: ↓ NUFC excretion changes during MDD sleep EEG during remission episodes and recovery HC: 15.8 ± 1.9 HC (7M; 9F) Clinical: PRS, HDRS, K- SADS [115] UPD: 21.8 ± 4.3 UPD (5M; 13F) Evaluate the potential of SWC: BPD:↓ and later onset of circadian measures as early Actigraphy, melatonin secretion markers of mood disorders DLMO subtypes BPD: 22.8 ± 4.8 BPD (3M; 11F) Clinical: Psychiatric interview (DSM-IV criteria), BDI [116] HC: 24.8 ± 2.5 HC (8M; 12F) Investigate objectively the 24- SWC: BPD: 62 % had delayed sleep h sleep–wake cycle in adoles- Actigraphy (during a depressive phase), cents and young adults with and later sleep offset UPD: 20.1 ± 4.7 UPD and BPD (28M; 47F)mood disorders Clinical: compared to UPD and HC Psychiatric BPD: 23.2 ± 4.3 interview UPD: 30 % had delayed sleep (DSM-IV HC: 10 % had delayed sleep criteria) [121] Stage 1a: 17.6 ± 4.0 Stage 1a (7M; 11F) Determine if disturbed sleep– SWC: Stage 1b & 2: ↑ delayed sleep wake cycle patterns in young Actigraphy schedule, especially on people with emerging mental weeknights Stage 1b: 19.1 ± 4.1 Stage 1b (44M; 38F) Clinical: disorder are associated with Psychiatric Stage 2+: 22.4 ± 4.3 Stage 2+ (27M; 27F)stages of illness Stage 1a & 2+: ↓ sleep interview efficiency HC: 24.4 ± 3.1 HC (11M; 12F) (DSM-IV criteria) Note. Sample: ANX anxiety disorder, BPD bipolar disorder, CS community sample, DD depressive disorder, HC healthy controls, HR high risk participants (offspring of parents with bipolar disorder), HYP hypomanic participants, LR low risk participants (offspring of parents without a mental disorder), MDD-ANX comorbid Major depressive disorder and anxiety disorder, MDD major depression disorder, MHP mental health patients (mixed diagnosis sample), P-MDD past major depressive disorder, P-ANX past anxiety disorder, PD panic disorder Measures: ACTH adrenocorticotropic hormone, BDI beck depression inventory, BPD body pain diagram, CBCL child behaviour checklist, CDI children’s depression inventory, CDRS children’s depression rating scale, CECA childhood experience of case and abuse contextual semi-structured interview and rating system, CES-D Center for Epidemiologic Studies Depression, CRH corticotropin-releasing hormone, DHEA dehydroepiandrosterone, DLMO dim light melatonin onset, DSM-IV diagnostic and statistical manual of mental disorders IV, EEG electroencephalography, GH growth hormone, HDRS Hamilton depression rating scale, HPS hypomanic personality scale, HIQ hypomanic interpretations questionnaire, ISS internal state scale, K-SADS schedule for affective disorders and schizophrenia for school age children, LSI life stress interview, MASQ mood and anxiety symptom questionnaire, NUFC nocturnal urinary free cortisol, PANAS positive and negative affect scale, PDS pubertal developmental scale, PRS Pfeffer rating scale, PSWQ Penn state worry questionnaire, QIDS quick inventory of depressive symptomatology, RLEQ recent life events questionnaire, SASC social anxiety scale for children, SCID structured clinical interview for DSM, SDS Sheehan disability scale, SWC sleep- wake and circadian biology, WSAS work and social adjustment scale Findings: ↑ = Increased, Improved or Higher, ↓ = Decreased, Reduced or Lower, ~ = ‘is associated with’, BL baseline, FUP follow-up, NUFC nocturnal urinary free cortisol, REM rapid eye movement, SWA slow wave activity *indicates that the study features more than once in the data synthesis

Iowa Gambling Task (IGT) [34], compared to suicide reduced total white matter volumes [63], and reduced attempters who were comparable to healthy controls. It left fusiform gyrus grey mater volumes [64]. Whilst, the may be that the ACC is associated with attentional con- only fMRI study investigating cannabis use reported that trol related to the emotive processing that has been lower amygdala reactivity was associated with higher linked to suicide rather than the higher cognitive pro- rates of cannabis use in MDD patients [65]. One study cesses investigated in these latter studies. [66] investigated alcohol use via sMRI and identified that Alcohol and substance use appears to affect multiple lower prefrontal cortex white matter and overall grey brain structures with most studies indicating that alco- matter volumes were associated with greater levels of hol and substance use is associated with a pattern of re- alcohol consumption. Collectively, these findings are ductions in brain volume and impairments in brain mostly consistent with the aforementioned neuropsycho- function. Cannabis use was investigated by two struc- logical studies (see 'Neuropsychology') and previous evi- tural MRI (sMRI) studies where it was associated with dence regarding the neurobiological effects of alcohol Iorfino et al. BMC Psychiatry (2016) 16:156 Page 22 of 38

Table 6 Neurophysiological studies evaluating the five functional domains in young people (12-30 yrs) with a mood and/or anxiety disorder Outcome Study Age (mean ± SD) Sample (N) Aims Key measures Key findings measure Social and [190] MHP: 22.1 ± 4.0 BPD (18) Determine the longitudinal Nα: MMN BPD & PSD: ↑ BL MMN ~ ↑ social and economic relationship between economic participation at FUP participation PSD (13)MMN/P3a and functional Functional: outcomes in patients. SOFAS, WHO- DAS-II Physical [191]* MDD: 17.1 ± 0.6 MDD (8F) Investigate the effect of Nα: EEG MDD: ↓ rPR theta & ↓ smoking health nicotine on resting EEG withdrawal, craving and physical Functional: HONC activity and affect. symptoms ~ acute nicotine administration. Suicide and [122] SA: 29.5 ± 13.3, HC: SA (24M; 16F) Investigate the trait Nα: EEG and SA: ↓ CNV and whole blood 5-HT ~ self-harm 34 ± 13.3 predisposing to DSH by blood samples multiple episodes of self-harm. examining EEG and HC (13M; 14F) Functional: HLS, peripheral monoamine activity. MADRS, SIS [123] SA: 14 (12 – 17yrs) SA (16F) Examine EEG alpha Nα: EEG alpha SA: ↑ posterior alpha asymmetry ~ asymmetry among high- asymmetry suicidal intent (not depression risk adolescents severity) HC: 14 (12 – 17yrs) HC (22F) Functional: HASS, SIS [124] rMDD + CSA: rMDD + CSA (15F) Examine the association Nα: EEG rMDD + CSA: ↑ subgenual ACC 31.60 ± 10.98 between CSA, MDD and activation during reward based maladaptive behaviour. decision making, ↓ reaction time rMDD: 24.81 ± 3.94 rMDD: (16F) Functional: YRBS during incentive-based trials ~ ↑ fre- (adult version) HC: 30.44 ± 10.78 HC (18F) quency of self harm/suicidal behaviours. Alcohol and [192] BPD-L: 21.8 ± 3.9 BPD-L (5M; 11F) Investigate the effects of Nα: MMN BPD-H: ↓ temporal MMN substance use alcohol use on MMN in BP. BPD-H: 22.6 ± 3.4 BPD-H (9M; 17F) Functional: AUDIT HC-L: 22.4 ± 2.6 HC-L (6M; 14F) HC-H: 23.4 ± 3.2 HC-H (6M; 8F) [193] AD: 24 ± 3.77 AD (44M; 47F) Explore the use of a startle Nα: Startle, ERP AD: ↑ facilitation, ↓ inhibition of the paradigm and its N4S component by pre pulse BD: 24.6 ± 5.76 BD (23M; 18F) Functional: association with alcohol stimuli. SSAGA, FHAM AFF: 22.9 ± 3.94 AFF (32M; 65F) use. DD: 23.5 ± 3.17 DD (51M; 61F) Clinical [128] ANX : 12.9 ± 2.6 ANX (7M; 13F) Examine the relationship Nα: Multiple ANX: ↓ in multiple muscle ASR ~ ↓ in syndrome between ASR, symptom muscle ASR anxiety symptoms. reduction and treatment ↑ HC: 12.0 ± 2.5 HC (10M; 15F)success. Clinical: ADIS-C/P, ANX: multiple muscle ASR SCAS predicted CBT treatment response [133] OCD: 13.9 ± 2.4 OCD (18M; 22F) Assess ERN as a biomarker Nα: ERN OCD & SIB: ↑ ERN at Cz for OCD (independent of symptom severity, SIB: 13.9 ± 2.4 SIB (13M; 6F) Clinical: Y-BOCS, current diagnostic status and CBCL, MASC, CDI HC: 13.8 ± 2.3 HC (20M; 20F) treatment effects). [134] ANX: 11.8 ± 2.3 ANX (3M; 10F) Demonstrate ERN Nα: ERN ANX & OCD: ↑ ERN at Cz amplitude is increased in (independent of symptom severity, OCD: 12.7 ± 2.2 OCD (8M; 18F) Clinical: Y-BOCS, young anxiety patients. current diagnostic status and CBCL, MASC, CDI HC: 12.4 ± 2.2 HC (14M; 13F) treatment effects). [194] RES: 14.1 ± 2.8 RES (2M; 6F) Examine the relationship Nα: TMS NoRES: ↑ deficits in pre-treatment between TMS with LICI NoRES: 13.1 ± 1.6 NoRES (5M; 3F) Clinical: CDRS-R, subsequent treatment response QIDS, CGI-severity scale [195] HC: 25.54 ± 3.41 HC (28M; 16F) Investigate the intensity Nα: MDD: ↑ intensity scores for sad faces evaluation of social stimuli ERP (N170, P1, P2) compared with HC, ↑ reaction times in depression for all faces and ↑ P1 & P2 DEP: 25.96 ± 4.58 DEP (9M; 15F) Clinical: SCID, BDI, amplitude for sad faces HDRS, BAI MDD: 26.58 ± 4.16 MDD (10M; 14F) DEP: ↓ scores for happy and neutral faces, ↑ reactions times and ↑ P1 & Iorfino et al. BMC Psychiatry (2016) 16:156 Page 23 of 38

Table 6 Neurophysiological studies evaluating the five functional domains in young people (12-30 yrs) with a mood and/or anxiety disorder (Continued) P2 amplitude for happy faces compared to sad faces. [196] HC: 27.7 ± 7.0 HC (14M; 12F) Assess brain function Nα: Resting EEG BPD: ↑ power in all wave bands. impairments in bipolar Marked increases in right temporal BPD: 30.7 ± 6.1 BPD (10M; 19F) Clinical: BDI patients. theta and left occipital beta. [135] OCD: 13.3 ± 2.8, OCD (13M; 5F) Examine ERN in paediatric Nα: ERN OCD: ↑ ERN pre-treatment and after patients with OCD treatment. No relationship with HC: 11.9 ± 2.6 HC (8M; 10F) Clinical : Y-BOCS symptom severity or changes in symptom severity [197] HC: 17 ± 1.6 HC (43F) Evaluate the effects of Nα: ERP (P300) DD: ↓ P300 amplitude. No effect of depression and a family family history of alcohol or drug HC-FHA: 16.5 ± 1.3 HC-FHA (31F)history of alcohol or Clinical: SSAGA, dependence. MAST, PANAS HC-FHD: 16.1 ± 1.5 HC-FHD (27F) substance dependence on P300. DEP: 17.2 ± 1.4 DD (12F) DEP-FHA: 17.3 ± 1.5 DD-FHA (9F) DEP-FHD: 16.3 ± 1.3 DD-FHD (8F) [191]* MDD: 17.1 ± 0.6 MDD (8F) Investigate the effect of Nα: EEG MDD: Nicotine ↓ theta amplitude in acute nicotine right parietal region. No associations Clinical: BDI, administration on resting with mood. HONC, PANAS EEG activity and affect [129] MDD: 30.4 ± 11.8 MDD (28M; 23F) Assess the utility of baseline Nα: LDAEP MDD: steep N1 sLORETA-LDAEP at LDAEP predicting response BL ~ treatment response. ↑ P2 to antidepressants. Clinical: HDRS, sLORETA-LDAEP slope at week 1 ~ MADRS treatment response. [167]* OCD: 27 ± 9.8 OCD (15M; 16F) Characterize the cognitive Nα: ERP (P300) OCD: ↓ P300 duration. ↓ stroop functions of the patients duration ~ ↑ P300 amplitude in HC: 27.4 ± 9.1 HC (14M; 16F)with OCD by utilizing ERPs Clinical: HDRS occipital, parietal and temporal and neuropsychological anterior regions. tests [52]* OCD: 24.06 ± 5 OCD (21M; 9F) Assess the relationship Nα: ERP (N100, OCD: ↑ P200 amplitude, unrelated to between cognitive P200, N200, P300) neither severity nor chronicity of HC: Matched HC (21M; 9F) dysfunction, clinical status illness. ↓ N200 amplitude (worsens Clinical: YBOCS and severity in OCD. with ↑ severity). ↓ N100 and P200 ~ ↑ chronicity [136] OCD-U: 25 ± 8.0 OCD-U (9M; 10F) Examine the effects of Nα: ERN OCD: ↑ ERN, irrespective of chronic medication on medication use. OCD-M: 30.8 ± 9.5 OCD-M (9M; 10F)error responses in OCD. Clinical: HDRS, HAMA, YBOCS PC-M: 31.7 ± 10.6 PC-M (8M; 11F) HC & PC: ↑ anxiety and depression ~ ↑ ERN amplitude HC: 25.3 ± 7.5 HC (11M; 10F) [127] DEP: 20.9 ± 0.55 DEP (515) Examine whether recurrent Nα: ASR DEP: ↑ ASR was associated with major depression is associated multiple (more than 1) depressive Clinical: SCID with abnormal startle episode. Note. Sample : AFF affective disorder (not specified), AD alcohol dependence, ANX anxiety disorder, BD behavioural disorder, BPD bipolar disorder, BPD-L bipolar disorder with low alcohol use, BPD-H bipolar disorder with high alcohol use, DD depressive disorder, DD-FHA depressive disorder with family history of alcohol dependence, DD-FHD depressive disorder with family history of drug dependence, DrDep drug dependence, HC healthy controls, HC-FHA healthy control with family history of alcohol dependence, HC-FHD healthy control with family history of drug dependence, HC-L health control with low alcohol use, HC-H healthy control with high alcohol use, MDD major depression disorder, MHP mental health patients (mixed diagnosis sample), NoRES treatment non responders, OCD obsessive compulsive disorder, OCD-M obsessive compulsive disorder patient medicated, OCD-U obsessive compulsive disorder patients unmedicated, PC-M psychiatric control patient medicated, PSD psychotic spectrum disorder, RES treatment responders, rMDD remitted major depression disorder, rMDD+CSA remitted major depression disorder with childhood sexual abuse history, SA suicide attempters, SIB suicide ideation behaviour Measures : ADIS-C/P anxiety disorders interview schedule for children, ASR auditory startle reflex, AUDIT Alcohol Use Disorder Identification Test, BAI beck anxiety inventory, BDI beck depression inventory, CBCL child behaviour checklist, CDI children’s depression inventory, CDRS children’s depression rating scale, CGI clinical global impression scale, EEG electroencephalography, ERP event related potential, ERN event related negativity, FHAM family history assessment module, HAMA Hamilton anxiety rating scale, HASS Harkavy Asnis suicide scale, HDRS Hamilton depression rating scale, HLS beck hopelessness scale, HONC hooked on nicotine checklist, LDAEP loudness dependant auditory evoked potential, MASC multidimensional anxiety scale for children, MADRS Montgomery-Asberg depression rating scale, MMN mismatch negativity, MAST Michigan Alco- holism Screening Test, PANAS positive and negative affect scale, QIDS quick inventory of depressive symptomatology, SCID structured clinical interview for DSM, SCAS Spence children’s anxiety scale, SIS suicide intent scale, SOFAS social and occupational functioning assessment scale, SSAGA semi-structured assessment for the genetics of alcoholism, TMS transcranial magnetic stimulation, WHO-DAS-II World Health Organisation Disability Assessment Scale II, Y-BOCS, Yale–Brown obsessive-compulsive scale, YRBS youth risk behaviour survey Findings : ↑ = Increased, Improved or Higher, ↓ = Decreased, Reduced or Lower, ~ = ‘is associated with’, 5-HT serotonin, BL baseline, CBT cognitive behaviour therapy, CNV contingent negative variation, FUP follow-up, N4S late wave frontal ERP component responses, rPR right Parietal Region, RT reaction time * indicates that the study features more than once in the data synthesis Iorfino et al. BMC Psychiatry (2016) 16:156 Page 24 of 38

and substance misuse on the structure and function of treatment and how this may have influenced the course of frontal and temporal brain regions [67]. illness. Compared to healthy controls, treatment naïve patients with OCD had larger thalamic volumes, which normalised following paroxetine treatment. These reductions Functional domain: clinical syndrome were also associated with a decrease in OCD symptoms. From the fifty neuroimaging studies investigating clinical While, CBT treatment for OCD was not associated with syndrome, multiple regions of interest have been studied change in thalamic volumes [79], greater symptom im- using a variety of imaging methods. provement following CBT was associated with a normalised metabolism in the ACC [80], and with increased prefrontal Structural magnetic resonance imaging (sMRI) In de- grey matter volumes [81]. These studies are consistent with pressive disorders the majority of studies examining the association between the lower symptom severity and brain structure have focused on frontal and limbic re- greater prefrontal grey matter volumes [82]. Higher com- gions with mixed findings; although some promising pulsive symptom severity was associated with reduced pitu- patterns emerge when the severity and/or clinical course itary gland volume in OCD patients in males, compared to of specific disorders are considered. Reduced ACC vol- healthy controls [83], and larger corpus callosum area [84]. umes [68, 69], and increased ACC thickness [70], have Collectively these OCD studies seem to indicate that been identified in MDD patients compared to healthy pharmacological treatment for OCD may be particularly controls. Whilst, no significant association with clinical useful for targeting deficits in thalamus structure and func- severity or symptoms was found in these studies, re- tion, whilst CBT may be better for targeting clinical features duced ACC volume was associated with higher border- associated with prefrontal structures. line personality disorder symptom severity but not depression, in patients diagnosed with comorbid MDD Functional magnetic resonance imaging (fMRI) Stud- and borderline personality disorder [59]. Some further ies investigating brain function using fMRI have also lines of inquiry provide greater detail about how the se- predominantly focused on frontal and limbic regions. verity of the clinical syndrome may influence or be influ- Clinical improvement following lamotrigine treatment enced by particular brain structures. Decreased grey for bipolar disorder [85], and SSRI treatment for gener- matter volume in frontal brain regions were evident in alised social phobia [86], were associated with reduc- patients with a discrete or persistent affective illness tions in amygdala activation whilst viewing negative compared to those with attenuated syndromes and valanced emotional pictures. Similarly, higher activa- healthy controls [71], and MDD patients who experi- tion of the amygdala to emotionally fearful faces com- enced more than three untreated depressive episodes pared to happy faces was associated with treatment had reduced subcallosal gyrus volumes, an ACC sub- response to CBT or medication in anxiety patients division [72]. Collectively, these studies reiterate the [87]. Bipolar disorder pharmacotherapy treatment re- relationship between reductions in the ACC and depres- sponders compared to non-responders had greater sion, and suggest that greater reductions in the ACC amygdala functional connectivity within the frontolim- may be associated with more severe illness. bic network, and higher amygdala functional connect- Compared to healthy controls, MDD patients had ivity within this network after treatment was lower amygdala volumes, and no association with clin- associated with greater improvements in mania symp- ical severity or illness duration [73], however MDD pa- toms [88]. These studies consistently demonstrate a re- tients near the onset of their illness had increased lationship between higher amygdala activity and amygdala-hippocampal volume ratios that were associ- patterns of treatment response for both bipolar and ated with higher severity of anxiety, but not depression anxiety disorders. This may be indicative of a shared severity [74]. This is consistent with evidence indicating neurobiological vulnerability for heightened amygdala that larger amygdala volumes in GAD patients are asso- reactivity associated with stress and the emergence of ciated with greater symptom severity [75]. This suggests particular affective disorders [89]. that common forms of depression and anxiety may share similar biological processes and genetic liability [76, 77], Diffusion Tensor Imaging (DTI) Theevidencefrom yet differ in their phenotypic expression. Although dis- these DTI studies collectively indicate that poorer tinct pathophysiology may underlie the development of white matter integrity is associated with affective dis- SAD, since lower amygdala grey matter density was as- orders that may be an early marker of disorder. A sociated with greater disease duration and earlier age of diagnosis of a depressive disorder, compared to onset in this group [78]. healthy controls was associated with lower fractional Changes to areas of the brain following a course of treat- anisotropy, a measure indicating poorer white matter ment can provide valuable insight into the success of integrity, and higher mean and radial diffusivity in the Iorfino et al. BMC Psychiatry (2016) 16:156 Page 25 of 38

Table 7 Metabolic studies evaluating the five functional domains in young people (12-30 yrs) with a mood and/or anxiety disorder Outcome Study Age (mean ± SD) Sample (N) Aims Key measures Key findings measure Social and [140]* MHP: 28.74 ± 10.38 MHP (38M; 28F) Identify changes in the Metabolic: BMI MHP: ↑ BMI ~ ↑ social and participation rates of obesity in economic participation (40 MDD, 26 BPD)never-treated patients Functional: GAF with mood disorder over 4 years of follow- up. Physical [198] FH+: 18.9 ± 1.0 FH+ (32M, 53F) Determine whether Metabolic/Functional: FH+: ↑ peripheral and central health young people with a glucose, lipids and high- BP, arterial stiffness and ↓ HC: 19.1 ± 0.1 HC (27M; 42F) family history of sensitivity CRP. BP, arterial insulin sensitivity depression have altered stiffness and waking cor- metabolic markers. tisol concentration. Suicide and [138] SUC: 15.93 ± 1.48 SUC (15M; 32F) Examine the Metabolic: blood serum SUC: ↑ cholesterol ~ current self-harm relationship between samples suicide behaviour (within the serum cholesterol levels SUC group, ↑ serum PC: 16.22 ± 1.95 PC (58M; 47F) Functional: SPI and suicidal behaviours cholesterol ~ ↓ severity of SUC, but not ~ symptom severity) PD: 25.3 ± 3.3 PD (37M; 35F) Elucidate the Metabolic: BMI, blood PD: ↓ HDL-C and ↑ VLDL-C ~ relationships between serum samples higher suicide ideation alexithymia, suicide Functional: SSI ideation and serum lipid levels. [137] SA: 15.44 ± 1.99 SA (17M; 49F) Explore the associations Metabolic: blood serum SA: ↓ cholesterol levels ~ between cholesterol samples attempted suicide history and suicidal behaviour PC: 15.19 ± 1.68 PC (15M; 39F) Functional: hospital records [199] SA: 16.8 SEM = .74 SA (3M; 6F) Investigate platelet PBR Metabolic: blood serum SA: ↓ platelet PBR density density in suicidal teens samples PC: 16.5 SEM = .5 PC (7M; 3F) Functional: hospital records, SPI, SRS [155] SA: 15.87 ± 1.56 SA(10M; 25F) Evaluate the relationship Metabolic: blood serum SA: ↓ plasma 5-HT level ~ ↑ between plasma samples suicidality. (5-HT did not dis- serotonin levels and criminate between the psychi- PC: 16.29 1.81 PC (19M; 11F)psychometric measures Functional: hospital atric diagnostic categories) records, SPI ER: 16.91 2.47 ER (13M; 38F) in suicidal adolescents HC: 15.26 1.41 HC (45M; 50F) Alcohol and [200] BPD-O: 12.9 ± 3.1 BPD-O (77M; 68F) Investigate obesity in Metabolic: BMI BPD: SUD ~ 2.8 fold increased substance paediatric bipolar prevalence of BPD-OB. use BPD-NO: 13.3 ± 3.0 BPD-NO (108M; 95F)patients and notable Functional: K-SADS-P correlates Clinical [162] MDD: 24.1 ± 3.2 MDD (45M; 44F) Examine the association Metabolic: BMI MDD: ↑ BMI at FUP (in syndrome between MDD in adulthood) HC: 22.2 ± 2.9 HC (43M; 45F)childhood and BMI in Clinical: KSADS, adulthood [140]* MHP: 28.74 ± 10.38 MHP (38M; 28F) Identify changes in the Metabolic: BMI MHP: clinical improvement ~ ↑ rates of obesity in never- BMI (40 MDD, 26 BPD) Clinical: HAMD treated patients with mood disorder over 4 years of follow-up. Note. Sample : BPD bipolar disorder, BPD-O bipolar disorder with obesity, BPD-NO bipolar disorder without obesity, ER emergency room patients admitted for suicide attempt, HC healthy controls, FH+ family history of depression, MDD major depression disorder, MHP mental health patients (mixed diagnosis sample), PC psychiatric control (i.e. psychiatric diagnosis but no suicide attempt), PD panic disorder, SA suicide attempters, SUC suicidal tendencies (either ideation, threat or attempt) Measures : BMI body mass index, BP blood pressure, CRP C-reactive protein, GAF global assessment of functioning, HDRS Hamilton depression rating scale, K- SADS schedule for affective disorders and schizophrenia for school age children, SPI suicide potential interview, SRS suicide risk scale, SSI scale of suicide ideation Findings : ↑ = Increased, Improved or Higher, ↓ = Decreased, Reduced or Lower, ~ = ‘is associated with’, 5-HT serotonin, FUP follow-up, PBR peripheral-type benzodiazepine receptors * indicates that the study features more than once in the data synthesis Iorfino et al. BMC Psychiatry (2016) 16:156 Page 26 of 38

corpus callosum, whilst higher fractional anisotropy monitoring (e.g. actigraphy) or indicators (e.g. cortisol and axial diffusivity and lower radial diffusivity in the secretion) to determine sleep-wake and circadian uncinated fasciculus [90]. Lower fractional anisotropy function. There were 23 studies (a total of 1609 par- in the ACC [91], and genu, body and splenium of the ticipants; 59.3 % female) that utilised sleep-wake and corpus callosum as well as the superior and anterior circadian biology and across these studies 84 % corona radiata [92] is evident in bipolar disorder (1352/1609) were patients and 16 % (257/1609) were compared to healthy controls. For those who experi- healthy controls. Among the patient group 55 % enced maltreatment during childhood, compared to (747/1352) had depression, 3 % (45/1352) had bipolar, healthy controls, MDD at follow up was associated 5 % (69/1352) had anxiety, and 36 % (491/1352) were with lower fractional anisotropy in the superior longi- classified as other. tudinal fasciculi and the right cingulum-hippocampal projection, whilst substance use disorder at follow up Functional domains: social and economic participation, was associated with lower fractional anisotropy in the physical health, suicide and self-harm & alcohol and right cingulum-hippocampal projection [93]. Greater substance use obsession symptom severity in OCD patients was Sleep-wake and circadian biology appears to be useful associated with higher fractional anisotropy in the for characterising two functional domains in young splenium [94]. Lower white matter integrity in the people, namely social and economic participation, and genu of the corpus callosum, anterior thalamic radi- suicide and self-harm behaviours. Three sleep-wake and ation, anterior cingulum and sagittal stratum was as- circadian studies investigated the relationship between sociated with higher depression severity in MDD salivary cortisol secretion and social and economic par- patients [95]. Having a discrete or persistent psychi- ticipation. Two of these studies [102, 103] were longitu- atric illness was associated with lower fractional an- dinal investigations that identified that increased salivary isotropy in the left anterior corona radiata compared cortisol at baseline [102], and before alprazolam treat- to healthy controls, whilst a similar pattern of lower ment for patients with panic disorder [103] predicted fractional anisotropy within this region was associated poorer social and economic participation at follow up. with attenuated syndromes of psychiatric illness [96]. Such results suggest that increased HPA activity, indexed by a greater salivary cortisol response, may be indicative of HPA axis deregulation with prognostic significance Magnetic Resonance Spectroscopy (MRS) In terms of and not simply a cross sectional marker of stress and ac- MRS studies, the major metabolites that were investigated tive illness. Similarly, the one cross-sectional study [104] include N-acetyl aspartate - a measure of neuronal integ- found that increased cortisol secretion during a social rity; choline - involved in cell membrane production, lac- interaction task was associated with poorer social func- tate - marks glycolysis has been initiated in an oxygen tioning. Alone, this study would seem to demonstrate deficient environment; creatine - indicates metabolism of that increased cortisol secretion is a state marker of so- brain energy; glutamate/glutamine -an excitatory neuro- cial stress, however in light of the previous longitudinal transmitter involved in neural activation; and GABA - an studies, it is possible that heightened HPA activity is in- inhibitory neurotransmitter involved in reducing neuronal dicative of a persistent dysregulated stress response to excitability [97]. Lower total choline in the left striatum social specific cues. Notably, all three of these studies was associated with OCD and remained consistent over were published over 15 years ago, suggesting the need the course of illness [98]. Compared to healthy controls, for new evidence to explore the relationships identified bipolar disorder was associated with a higher lactate to N- by the longitudinal studies. acetyl aspartate and lactate to creatine ratios [99], and Three sleep-wake and circadian studies were longitu- higher N-acetyl aspartate in the ACC and higher N-acetyl dinal and identified a relationship with suicide outcomes aspartate, choline and creatine in the ventral lateral pre- in MDD patients. At baseline, higher growth hormone frontal cortex [100]. Moreover, bipolar disorder re- secretion during the first 4 h of sleep [105], and higher sponders had lower left ventral lateral prefrontal cortex cortisol secretion in the late hours of sleep [106] were glutamate/glutamine [101]. For both MDD patients and both associated with the emergence of a suicide attempt healthy controls, higher ACC GABA was associated with at follow-up. The final study [107] identified that re- lower anhedonia scores, and lower ACC GABA was asso- duced delta sleep activity was associated with higher ciated with MDD [68]. levels of suicidality as well as depression severity. Similarly, to the social and economic participation stud- Sleep-wake and circadian biology ies, these biological substrates also point to HPA dysreg- This section entails studies that have utilised either ulation as a predictor of later suicide attempts in MDD sleep physiology (e.g. sleep EEG) and/or sleep-wake patient groups. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 27 of 38

Functional domain: clinical syndrome Neurophysiology Of the eighteen sleep-wake and circadian biology studies, There were 21 studies (a total of 2034 participants; nine studies investigated clinical syndrome utilising corti- 69.6 % female) that utilised neurophysiology and across sol responses. Whilst, the timing of cortisol secretion var- these studies 74 % (1510/2034) were patients and 26 % ied between studies, the findings consistently indicate that (524/2034) were healthy controls. Among the patient increased cortisol response is associated with the develop- group 56 % (851/1510) had depression, 9 % (130/1510) ment [108–110] and persistence [102, 111] of MDD, had bipolar, 21 % (313/1510) had anxiety, and 14 % whilst remission [112] is associated with reductions in cor- (216/2034) were classified as other. tisol measures. The heterogeneity of depression becomes clearer with evidence of moderate-to-severe depression Functional domains: social and economic participation, being associated with significant blunting of the cortisol physical health, suicide and self-harm & alcohol and response compared to those with mild depression, who substance use had increased cortisol secretion during a stress task [113]. Evidence from the included studies indicated that neuro- These results suggest that the neurobiological systems me- physiology may be particularly useful for characterising diating the stress response are functioning very differently, suicide and self-harm behaviours, and alcohol and sub- and may indicate distinct pathophysiological drivers of de- stance use. Three separate neurophysiological studies pression for these two groups. Specifically, genetically me- found an association with specific types of suicide and diated depression associated with increased severity and self-harm behaviours. Specifically, lower contingent chronic stress leading to a desensitization of glucocortic- negative variation was associated with multiple episodes oid receptors versus mild to moderate depression arising of deliberate self-harm [122], increased posterior EEG from predominately environment risk factors with typical alpha asymmetry was associated with suicidal intent and HPA abnormalities [113]. This pattern of findings appears the lethality of suicide attempt, but not depression sever- to differ for comorbid MDD and anxiety, which was asso- ity, in suicide attempters [123], and slower reaction ciated with flatter diurnal cortisol slopes (daytime cortisol times during incentive based decision making tasks was activity) [114], suggesting that the presence of anxiety in- associated with increased frequency for suicide and self- fluences cortisol function in a way that contrasts to de- harm behaviours in remitted MDD patients [124]. To- pression alone. gether these findings link abnormal brain functions im- Studies that investigated the relationship between the plicated in decision making processes to different types sleep-wake cycle and clinical syndrome reported quite of self-harm and suicidal behaviours [125, 126]. consistent findings. When compared to unipolar depression, bipolar disorder is associated with delayed onset and lower levels of melatonin secretion [115], as well as Functional domain: clinical syndrome increased rates of delayed sleep [116]. Similarly, poor With regard to the fifteen neurophysiology studies, a sleep efficiency and lower sleep duration was reported in number of these examined the relationship between both unipolar depression [117] and hypomanic individ- treatment response and neurophysiological markers uals [118] compared to healthy controls. Increased de- using a number of methods. Firstly, increased startle re- pression severity was associated with reduced delta sleep sponse was associated with multiple episodes of depres- [107], while higher overnight dissipation of slow wave sion in depressed individuals [127]. Increased startle sleep predicted a reduction in depressive symptoms response was also associated with the presence of an [119]. Similarly, greater high density REM and lower anxiety disorder compared to healthy controls [128]. In REM latency at baseline was associated with the devel- this study, a reduction in acoustic startle response was opment of depression at follow in healthy controls [111], associated with a reduction in anxiety symptoms follow- whilst another found that greater high density REM was ing CBT, and a higher startle response at baseline pre- associated with fluoxetine treatment in patients with dicted treatment response. Similarly, steep N1 of the MDD [120]. Whilst these sleep-wake cycle deficits are Loudness Dependency of Auditory Evoked Potentials evident in those with a full threshold disorders (both (LDAEP) at baseline and higher P2 LDAEP at week 1 unipolar and bipolar disorder), increased rates of delayed predicted anti-depressant treatment response in patients sleep are also evident in individuals with either a discrete with MDD [129]. These findings are consistent with the disorder or an attenuated syndrome compared to indi- phenomenon described in this paper (see 'Neuroimag- viduals with mild symptoms and healthy controls [121]. ing') that link maladaptive processes associated with in- This indicates that a similar pattern of sleep-wake deficit creased threat processing measured at a neural, are also evident in those with subthreshold disorders psychological and behavioural level [130]. These findings and highlight that such deficits can be identified earlier implicate the amygdala as the predominate brain region in the course of illness. involved in threat processing, however it functions as Iorfino et al. BMC Psychiatry (2016) 16:156 Page 28 of 38

part of a wider brain circuit involving the dorsal medial increase in the prevalence rates for overweight/obesity, prefrontal (anterior cingulate) cortex [131, 132]. consistent with the finding of the former study. There Four studies utilised Event Related Negativity (ERN) to are a number of illness related factors that could explain characterise the clinical syndrome of OCD patient this relationship which include a return of normal appe- groups. All four studies found that increased ERN was tite, medication use, self-modulation of mood by over- associated with OCD compared to healthy controls eating [141] as well as biological factors implicated in [133–136], however none identified a relationship with mood, and metabolic function and weight maintenance, symptoms severity, treatment status or medication use. such as leptin [142] and neurotransmitter abnormalities Conversely the N100 ERP had an inverse relationship [143]. with symptom severity, whilst both the N100 and P200 had an inverse relationship with illness chronicity [52]. Functional domain: clinical syndrome These studies indicate that ERN may be a trait measure See Table 7 for individual results. associated with OCD independent of symptoms severity, diagnosis and/or treatment effects, while the ERP N100 and P200 may be measures sensitive to illness specific Discussion factors (e.g. chronicity). As expected, there is a predominate focus in the literature on clinical syndrome in young patients compared to the other four functional domains (i.e. social and eco- Metabolic nomic participation, physical health, suicide and self- There were 10 studies (a total of 1,385 participants; 43 % harm behaviours and alcohol and substance use). Whilst female) that utilised metabolic measures and across the neurobiology of these clinical syndromes have been these studies 80 % (1133/1385) were patients and 18 % extensively reviewed previously [131, 144], we provide (252/1385) were healthy controls. Among the patient an overview of these findings. Typically biomarkers of group 22 % (254/1133) had depression, 35 % (400/1133) the clinical syndrome alone do not readily provide a had bipolar, 6 % (72/1133) had anxiety, and 36 % (407/ complete understanding of disability and the risk factors 1133) were classified as other. that put young people at greater risk for a worse illness trajectory [3]. This review demonstrates the use of these Functional domains: social and economic participation, biomarkers to investigate multiple functional domains in physical health, suicide and self-harm & alcohol and addition to the clinical syndrome. It is clear that the na- substance use ture of the relationship between the underlying neuro- Three metabolic studies investigated the relationship be- biology and functional domains is an issue that needs to tween suicide and self-harm behaviours and cholesterol. be resolved in this area. However, overall this review ex- Lower total cholesterol levels were associated with a sui- hibits the usefulness of neurobiological parameters to as- cide attempt history [137], and increased severity of sui- sess these additional functional domains and identify cidal behaviour (e.g. ideations, gestures) among those treatment targets, in addition to the traditional focus on who were currently suicidal [138], while lower high clinical syndrome, to optimise interventions and im- density lipid cholesterol was associated with higher sui- prove illness trajectories. cide ideation [139]. Together these findings seem to indicate that lower cholesterol is associated with the Implications for personalised psychiatry spectrum of suicidal behaviours, particularly in males. The search for gold standard screening or diagnostic However, this is an area of contention since higher total tests has ultimately been unsuccessful, however the in- cholesterol was associated with current suicidal behav- creasing emphasis on personalised (or ‘stratified’) psych- iour versus those not currently suicidal. These findings iatry has the potential to make significant advances in highlight the complex association between suicidal be- terms of clinical validity and applicability [145]. Whilst, haviours and cholesterol, and indicate that its usefulness conventional diagnostic methods remain entirely rele- a biological marker needs further clarification. vant, the addition of the neurobiological markers that in- Two studies examined the correlates of the risk of dicate prognosis or potential treatment targets are obesity, indexed by BMI, among young people with essential for advancing personalised psychiatry. Utilising mood disorders. Substance use was associated with an individual characteristics to guide treatment decisions is 2.8 fold increased risk of overweight/obesity among bi- key to personalised medicine and providing person- polar patients and increased BMI was associated with centred care. This review exhibits the utility of the better social and economic participation over a 4-year RDoC approach to investigate individual characteristics longitudinal study in mood disorder patients (unipolar that extend to functional domains that contribute to on- and bipolar) [140]. The latter study also reported an going disability and poorer outcomes. By collating the Iorfino et al. BMC Psychiatry (2016) 16:156 Page 29 of 38

Fig. 2 The potential neurobiological targets associated with mental illness trajectory organised by functional domain. This is a model illustrating the potential neurobiological targets for each functional domain identified by this systematic review. At the centre is ‘mental illness trajectory’ that is a term employed here to encompass all the characteristic features of illness, and its associated short term and long term outcomes. Surrounding this are the five functional domains investigated by this review to illustrate how a young person’s mental illness trajectory is made up of and influenced by each of these functional domains. Within the respective sections of the figure, the potential neurobiological targets for each functional domain identified by this systematic review are highlighted in the boxes. The absence of any neurobiological targets in the physical health section of the figure reflects the lack of investigation into this particular functional domain emphasising the need for more research into this area. The clinical syndrome section is the most populated which reflects the predominant focus in the literature on this particular domain. Despite less focus on the remaining three functional domains, the evidence for potential neurobiological targets for these areas is promising for clinical utility and future research available evidence, this systematic review provides a cognitive tests that characterise memory and executive basis for future investigations to further evaluate the functions of young people with emerging illnesses may clinical utility of specific neurobiological markers and be particularly useful for identifying individuals who are their relationship with these functional domains (Fig. 2). at risk of poor social and economic participation outcomes. Unlike diagnosis, neuropsychological function in- Social and economic participation dependently predicts social and economic participation Our findings suggest that those with more severe im- outcomes in young people with emerging mood pairments in memory and executive functioning are at disorders [146]. The evidence suggests that neuropsych- the greatest risk for diminished participation in educa- ology, specifically memory and executive function, may tion, employment and social settings. The use of mediate these outcomes in young people. Therefore, Iorfino et al. BMC Psychiatry (2016) 16:156 Page 30 of 38

those with identified weaknesses in these areas are likely Suicide and self-harm behaviours to require more intensive intervention targeted at these Studies investigating suicide and self-harm behaviours deficits to improve this functional domain. Such treat- spanned across all five neurobiological parameters, ment practise has demonstrated clinical utility for im- and yielded consistent findings across these parame- proving memory performance in adult groups with ters. Across studies, those with specific executive affective disorders [147]. Moreover, cognitive training function impairments in the domains of decision- was comparable to pharmacological treatment for im- making and conceptual flexibility appear to be more proving the depression state, whilst also associated with likely to engage in suicidal thinking or behaviours. additional benefits such as, no adverse side effects, im- Therefore, identifying these deficits may be particu- proved levels of anxiety, and better academic perform- larly important for recognising at risk patients and ance [148]. Cognitive training may slow or prevent the providing effective interventions. Furthermore, our impact of cognitive impairment on the social and eco- findings provide converging evidence from different nomic participation domain for young people with an measures of brain structure and function that suicide emerging affective disorder. and self-harm behaviours are associated with signifi- Notably, the common indicators of social and eco- cant disruptions in decision-making ability. The re- nomic functioning in the reviewed studies varied duced ACC volumes and activity in the identified greatly with some measures (e.g. SOFAS, academic neuroimaging studies as well as neurophysiological grades) being more useful than others (e.g. GAF) that evidence of lower accuracy at differentiating correct conflate symptoms and functioning. Of the reviewed and incorrect responses on a decision-making task studies, there was a lack of focus on social function- corroborate the neuropsychological findings regarding ing and the relationship with certain neurobiological the relationship between suicide and self-harm behav- parameters. The only study to specifically examine iours and impaired decision making and conceptual this relationship identified that HPA dysregulation, flexibility. indexed by an enhanced cortisol response during a Importantly, these findings have major implications social task was predictive of social functioning [104]. for the assessment and intervention of suicide and Although the aforementioned study was cross sec- self-harm behaviours. Firstly, decision making ability tional, longitudinal evidence supporting the predictive should be incorporated into the assessment of young validity of HPA dysregulation and greater disability people with mood and anxiety disorders to stratify (including social disability), provides further support young people on the basis of risk for suicidal and for the role of HPA dysregulation in social and eco- self-harms behaviours. Whilst, there are certainly nomic participation [102, 103]. However, it is clear other risk factors involved in suicide and self-harm that the role of HPA functioning is complicated given behaviours [150, 151], the findings from this review its implications in multiple functional domains includ- indicate that decision-making may be a mediator of ing suicide and self-harm behaviours and clinical suicide and self-harm outcomes. Secondly, cognitive syndrome. remediation interventions aimed at improving decision-making ability in those young people that are Physical health identified as having significant impairments may Physical health as a functional domain for young people prove to be a successful early intervention to reduce with mood disorders has been notably understudied or prevent elevated risk for suicide and self-harm be- (only three studies met the inclusion criteria) compared haviours [152]. to the other four functional domains. This is not com- The relationship between metabolic studies investi- pletely unexpected considering that the majority of gatingbloodcholesterolandsuicideandself-harmbe- neurobiological modalities (neuropsychology, neuroim- haviours was another major area identified by this aging, sleep-wake and circadian biology and neurophysi- review. Three studies investigating cholesterol consist- ology) are not recognised as being the traditional ently identified an association between suicide and method for investigating this particular domain [149]. Of self-harm behaviours and lower cholesterol levels. note, given that metabolic measures are used to classify Such findings are consistent with evidence that de- physical health outcomes (i.e. BMI is both a physical creased cholesterol levels in the brain may be associ- health outcome and a metabolic measure) it was not ated with reduced synaptic plasticity and impaired deemed appropriate to carry out searches for metabolic neurobiological functioning [153]. Furthermore, lower measures and physical health outcomes. Future studies serotoninergic activity has also been associated with may look to improve the methods and/or key terms used reduced cholesterol and implicated in the affective to investigate the best available measures to assess and disorders [154], which was evident in another meta- track physical health outcomes in this population. bolic study identified by this review [155]. It has been Iorfino et al. BMC Psychiatry (2016) 16:156 Page 31 of 38

suggested that these changes to cholesterol levels may control methods to investigate the neurobiological char- be the result of HPA axis dysfunction, however fur- acteristics that separate discrete diagnostic cases from ther studies are needed to explore these associations healthy controls, however these distinctions often pro- and investigate the treatment implications. vide limited clinically useful information. Those studies that sought to describe how specific neurobiological Alcohol and substance use characteristics are related to illness severity (or per- Converging evidence from neuropsychology, neuroim- haps more importantly, employed a case-case control aging and neurophysiology indicate that alcohol use is method to delineate between cases) were particularly associated with global impairments. Unsurprisingly, all useful for identifying neurobiological or neurocognitive the neuroimaging studies suggest that widespread im- risk factors of poorer illness trajectories. Using a multi pairment across frontal and temporal areas of the parameter approach, this review has been able to brain are associated with alcohol and substance use. collate evidence that covers a number of ‘columns’ in Specifically, reduced brain volume and function are the RDoC matrix (i.e. circuits, physiology, behaviours) particularly prominent in these areas and these results to study the pathophysiological drivers of each func- reflect the findings of neuropsychological and neuro- tional domain, which is crucial for identifying treat- physiology studies that suggest alcohol and substance ment targets [157]. use to be associated with cognitive impairments and This review has provided evidence to indicate that ver- poor attention. Independently each individual study’s bal memory problems may be an early indicator associ- limitation of small sample sizes and lack of replica- ated with the emergence of depression since these tion limit their clinical applicability, however together deficits were evident in young people with depression, the evidence points to similar phenomena. not identified as MDD, and the emergence of depressive An important issue is to determine the how these symptoms in a community sample) [44, 45]. However, findings can be used to understand the risk factors for those young people with more persistent MDD and associated with alcohol and substance use. Many of more severe depression or anxiety symptoms executive the impairments may be the result of alcohol and function deficits seem to be more prominent. Similarly, substance use rather than a mediating factor involved another line of evidence emerging from this review indi- in the risk of engaging in these behaviours. These cates that greater illness severity is associated with problems have major implications in terms of assess- greater reductions in the ACC, a prefrontal brain region ment and intervention. Whilst, the aforementioned associated with executive function. Considering the neuroimaging, neuropsychology and neurophysiology previously discussed results regarding the relationship studies may be useful for tracking changes in the ef- between executive function, and another two functional fects of these behaviours overtime, it is still unclear domains (i.e. social and economic participation and what assessment measures may be particularly useful suicide and self-harm behaviours), this reiterates the in the early identification of individuals at risk for en- role of executive function as a mediating factor that is gaging in these harmful behaviours. More specific longi- associated with poorer illness outcomes across mul- tudinal studies before young people engage in alcohol tiple functional domains that should be a clear treat- and substance use are needed to differentiate between ment target at both a brain circuit and behaviour level. pre-existing and subsequent effects of alcohol and sub- The benefits of targeting neuropsychological function stance use. These longitudinal investigations will help are made clear by treatment studies that have shown model the role that risk factors such as risk taking, im- that improved cognitive function following a form of pulsivity, social occupational factors and decision- treatment, namely, TMS or pharmacological, has making play in the development of poorer alcohol and improve hallmark symptoms in both bipolar and de- substance use outcomes. pression. Again, this is reiterated by evidence from neuroimaging that treatments increasing ACC function Clinical syndrome are associated with clinical improvements in depres- The final functional domain addressed by this review sion and bipolar. has been extensively reviewed elsewhere [131, 144], and The role of comorbid anxiety in depression and bipo- therefore we provide an overview of these findings spe- lar disorder is notable and associated with a substantial cific to young people and in the context of the other increase in morbidity and mortality [158–160]. From a four functional domains. The primary focus of this par- circuitry point of view, the amygdala is one of the pri- ticular domain is to identify features of the clinical syn- mary brain regions in a broader network that is involved drome that may help characterise a young person’s in depression and anxiety [131, 132]. Specifically, evi- illness phenotype and/or stage of illness [25, 156]. The dence from this review indicates that increased amygdala majority of studies identified by this review utilised case- volumes are associated with increased anxiety symptom Iorfino et al. BMC Psychiatry (2016) 16:156 Page 32 of 38

severity in MDD as well as a diagnosis of GAD, reiterat- conclusions. For the field of psychiatry to make new ing theories that these conditions share similar patho- ground regarding the underlying neurobiology of psy- physiology that cuts across diagnostic boundaries [161]. chopathology and its associated outcomes major consoli- Importantly, other neuroimaging work utilising fMRI dation of the common standardised measures for these has demonstrated that heightened amygdala activity is key functional domains and neurobiological parameters associated with both depression and anxiety [132]. This should be implemented. Admittedly differences in scien- is evidenced by the neurophysiology study identified by tific or clinical motivations will affect the widespread this review whereby successful CBT treatment for anx- adoption of common measures, however much like the iety disorder was associated with a reduction in anxiety RDoC initiatives focus on key neurobiological domains symptoms and the startle response, another index of of interest similar efforts should be made to maximise amygdala reactivity [128]. Together these findings impli- the standardised measurement of the key functional do- cate the amygdala as a treatment target that may reduce mains of interest across multiple diagnostic groups in neurobiological substrates of anxiety, commonly impli- clinical and research settings. In doing this review, we cated in the emergence of depression and a problematic have provided an overview of the essential functional do- feature in bipolar disorder. To reiterate the value of mains and current neurobiological evidence associated amygdala reactivity as a treatment target for anxiety in with these domains, the next step will be to outline the affective disorders, studies in the present review demon- standard measures that should be drawn upon to pro- strated that reduced amygdala activation to negative mote better consolidation of findings in the psychiatric emotional stimuli, either using pharmacological treat- and neurobiological study of mood and anxiety ments or CBT, were associated with clinical improve- disorders. ment in both bipolar and anxiety disorders. The lack of emphasis on all of these key functional do- From the sleep-wake and circadian biology studies, mains and their role in disorder onset, persistence and sleep dysfunction and cortisol secretion have consist- impact from a publication, reporting and/or research ently demonstrated a relationship with the clinical syn- priority point of view is problematic for the clinical drome features that allude to differential illness translation of psychiatric research. Our study highlights trajectories. For example, sleep dysfunction, charac- this pertinent issue so that future research may better terised by a number of different sleep parameters, was account for these factors and their relationship to key not only associated with the presence of a discrete de- neurobiological parameters and disorders to improve pressive or bipolar illness, but similar dysfunctions were our understanding of how these functional domains also evident in those with hypomanic or attenuated syn- interact or relate to mental illness trajectories. dromes. This is a critical finding since it presents sleep dysfunction as a primary treatment target to prevent illness progression in these affective illnesses. Similarly, in- Limitations and future directions creased cortisol secretion was consistently implicated in Some limitations of this review should be considered. the emergence of depression [102, 108–110, 112], whilst First, few studies investigated a particular functional do- two metabolic studies suggest that increased BMI is as- main (i.e. physical health) or utilised particular neurobio- sociated with the chronic course of mood disorders logical parameters, namely sleep-wake and circadian [140, 162]. Neurobiologically these findings implicate biology or metabolic, which limits the synthesis of these HPA dysfunction as being a core feature involved in findings and caution is advised when interpreting these mood disorders, and so interventions aimed at improv- results. Secondly, the restricted use of search terms for ing the deficits in these brain circuits may be useful to the functional domains or neurobiological parameters address these clinical outcomes [163]. may have limited the identification of key studies, particularly favouring studies reporting current primary dis- Moving towards greater clinical translation orders rather than lifetime diagnoses. Best efforts were One of the clear problems identified by this systematic made to be inclusive of as many studies as possible to review is the lack of consistently used patient groups carry out a complete overview of the literature, however and assessment measures. For any given functional do- future studies should look to expand on this work by main, multiple self-report or clinician rated scales, neu- adding key search terms that may have been missed to roimaging techniques, and cognitive tests were further the advancement of this growing literature focus- implemented that often assess the same or similar out- ing on the functional domains of mood and anxiety dis- comes, whilst the selection of patient groups varies dra- orders in young people. While we focused on RDoC matically for each study. This fundamentally limits the levels of analysis that correspond to ‘circuits’, ‘physiology’, capacity for strong comparisons to be made between ‘behaviour’,and‘self-report’, future studies may want to studies, or arrive at meaningful and clinically relevant include levels of analysis that include genetics, molecules Iorfino et al. BMC Psychiatry (2016) 16:156 Page 33 of 38

and cells for a deeper understanding of these neurobio- Funding logical factors. Funding for this study was provided by a National Health & Medical Research Council (NHMRC) Centres of Research Excellence (No. 1061043; ‘Optymise’) Furthermore, future researchers may want to consider grant. FI is supported by an Optymise PhD scholarship. RSCL is supported by an including borderline personality disorder into these in- Optymise post-graduate scholarship. IBH is supported by a NHMRC Senior vestigations, given its significant affective component Principal Research Fellowship (No. 1046899). The funders had no further role in study design; in the collection, analysis and interpretation of data; in the writing and relationship with these primary mood and anxiety of the report; and in the decision to submit the paper for publication. disorders. Limitations associated with the systematic review process should be considered when interpreting the Availability of data and materials The methods and results presented and discussed in this article are available present findings, namely the use of one independent re- in Tables 1–7. searcher for study selection and lack of a systematic risk of assessment bias since these may have impacted on the Authors’ contributions reliability and validity of data synthesis. Finally, the wide FI conducted the systematic review, and prepared the initial draft manuscript. DFH supervised FI during this process. FI, IBH and DFH, age range selected for this study is an important caveat conceived the study design. RSCL, IBH, JL & DH provided interpretation of that should be addressed in future studies since there the results. All authors contributed significantly to the writing of this are significant developmental/neurobiological changes manuscript, as well as having read and approved the final manuscript. during this dynamic period of brain development with Competing interests respect to grey (e.g. pruning) and white (myelination; IBH is a Commissioner in Australia’s new National Mental Health Commission connectivity) matter processes. from 2012. He was a director of headspace: the national youth mental health foundation until January 2012. He was previously the chief executive officer (till 2003) and clinical adviser (till 2006) of beyondblue, an Australian National Depression Initiative. He is the Co -Director, Health and Policy at the Brain and Conclusions Mind Centre which operates two early-intervention youth services under contract Mood and anxiety disorders are especially difficult to to headspace. He has led a range of community-based and pharmaceutical characterise and treat in young people (age 12 – 30 industry-supported depression awareness and education and training programs. He has led projects for health professionals and the community supported by years) when confounds of normal development and governmental, community agency and pharmaceutical industry partners (Wyeth, changing environmental influences are prominent. This Eli Lily, Servier, Pfizer, AstraZeneca) for the identification and management of review identified a predominant focus in the literature depression and anxiety. He has received honoraria for presentations of his own work at educational seminars supported by a number of non-government on the clinical syndrome, which in our view does not ad- organisations and the pharmaceutical industry (including Servier, Pfizer, equately address key individual characteristics, such AstraZeneca, and Eli Lilly). He is a member of the Medical Advisory Panel for as suicide and self-harm behaviours or alcohol and sub- Medibank Private and also a Board Member of Psychosis Australia Trust. He leads an investigator-initiated study of the effects of agomelatine on circadian stance use, that are involved in disability and persistent parameters (supported in part by Servier) and has participated in a multicentre illness. Based on the synthesis of results from multiple clinical trial of the effects of agomelatine on sleep architecture in depression and neurobiological modalities, we provide a detailed sum- a Servier-supported study of major depression and sleep disturbance in primary care settings. DFH has received honoraria for educational seminars from Janssen- mary of how the clinical utility of neurobiological mea- Cilag and Eli Lilly. sures may be improved by focussing on personalised assessment of these additional functional outcomes. We Consent for publication Not applicable. suggest that a shift in focus towards characterising the mechanisms that underlie and/or mediate multiple func- Ethics approval and consent to participate tional domains will optimise personalised interventions This study was performed in accordance with the declaration of Helsinki, and it was approved by the Tarbiat Modares Institutional ethical review and improve illness trajectories. board (Reference Number: 1789965214). All participants signed informed consent before the start of the study. Abbreviations ACC: anterior cingulate cortex; BMI: body mass index; CBT: cognitive Received: 22 October 2015 Accepted: 8 May 2016 behavioural therapy; DLMO: dim-light melatonin onset; DSM: diagnostic and statistical manual of mental disorders; EEG: electroencephalography; ERN: error related negativity; GABA: gamma amino butyric acid; GAF: global References assessment functioning; HPA: hypothalamus-pituitary-adrenal; 1. Collins PY, Patel V, Joestl SS, March D, Insel TR, Daar AS, Bordin IA, ICD: international classification of diseases; IGT: Iowa gambling task; Costello EJ, Durkin M, Fairburn C. Grand challenges in global mental LDAEP: loudness dependant auditory evoked potential; LICI: long-interval health. Nature. 2011;475(7354):27–30. cortical inhibition; MDD: major depressive disorder; MMN: mismatch 2. Gore FM, Bloem PJ, Patton GC, Ferguson J, Joseph V, Coffey C, Sawyer SM, negativity; MRI: magnetic resonance imaging; MRS: magnetic resonance Mathers CD. Global burden of disease in young people aged 10–24 years: a spectroscopy; OCD: obsessive compulsive disorder; PRISMA: Preferred systematic analysis. Lancet. 2011;377(9783):2093–102. Reporting Items for Systematic Reviews and Meta-Analyses; RDoC: research 3. Hickie IB, Scott J, Hermens DF, Scott EM, Naismith SL, Guastella AJ, Glozier domain criteria; REM: rapid eye movement; SAD: social anxiety disorder; N, McGorry PD. Clinical classification in mental health at the cross-roads: SOFAS: social occupational functioning assessment scale; SSRI: selective which direction next? BMC Med. 2013;11(1):125. serotonin reuptake inhibitor; TMS: transcranial magnetic stimulation. 4. Heimberg RG. Cognitive-behavioral therapy for social anxiety disorder: current status and future directions. Biol Psychiatry. 2002;51(1):101–8. 5. Scott J, Leboyer M, Hickie I, Berk M, Kapczinski F, Frank E, Kupfer D, McGorry P. Acknowledgments Clinical staging in psychiatry: a cross-cutting model of diagnosis with heuristic None. and practical value. Br J Psychiatry. 2013;202(4):243–5. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 34 of 38

6. Paus T, Keshavan M, Giedd JN. Why do many psychiatric disorders emerge 30. Green MF, Kern RS, Heaton RK. Longitudinal studies of cognition and during adolescence? Nat Rev Neurosci. 2008;9(12):947–57. functional outcome in schizophrenia: implications for MATRICS. Schizophr 7. Carrión RE, McLaughlin D, Goldberg TE, Auther AM, Olsen RH, Olvet DM, Res. 2004;72(1):41–51. Correll CU, Cornblatt BA. Prediction of functional outcome in individuals at 31. Bridge JA, McBee-Strayer SM, Cannon EA, Sheftall AH, Reynolds B, clinical high risk for psychosis. JAMA Psychiatry. 2013;70(11):1133–42. Campo JV, Pajer KA, Barbe RP, Brent DA. Impaired decision making in 8. McGorry PD, Hickie IB, Yung AR, Pantelis C, Jackson HJ. Clinical staging of adolescent suicide attempters. J Am Acad Child Adolesc Psychiatry. psychiatric disorders: a heuristic framework for choosing earlier, safer and 2012;51(4):394–403. more effective interventions. Aust N Z J Psychiatry. 2006;40(8):616–22. 32. Miranda R, Gallagher M, Bauchner B, Vaysman R, Marroquin B. 9. Insel TR, Sahakian BJ, Voon V, Nye J, Brown VJ, Altevogt BM, Bullmore ET, Cognitive inflexibility as a prospective predictor of suicidal ideation Goodwin GM, Howard RJ, Kupfer DJ. Drug research: a plan for mental among young adults with a suicide attempt history. Depress Anxiety. illness. Nature. 2012;483(7389):269. 2012;29(3):180–6. 10. Scott EM, Hermens DF, Glozier N, Naismith SL, Guastella AJ, Hickie IB. Targeted 33. Oldershaw A, Grima E, Jollant F, Richards C, Simic M, Taylor L, Schmidt U. primary care-based mental health services for young Australians. Med J Aust. Decision making and problem solving in adolescents who deliberately self- 2012;196(2):136–40. harm. Psychol Med. 2009;39(1):95–104. 11. Scott J, Scott EM, Hermens DF, Naismith SL, Guastella AJ, White D, Whitwell 34. Pan LA, Segreti A, Almeida J, Jollant F, Lawrence N, Brent D, Phillips M. B, Lagopoulos J, Hickie IB. Functional impairment in adolescents and young Preserved hippocampal function during learning in the context of risk in adults with emerging mood disorders. Br J Psychiatry. 2014;205(5):362–8. adolescent suicide attempt. Psychiatry Res. 2013;211(2):112–8. 12. Stuart MJ, Baune BT. Depression and type 2 diabetes: inflammatory 35. Ohmann S, Schuch B, Konig M, Blaas S, Fliri C, Popow C. Self-injurious mechanisms of a psychoneuroendocrine co-morbidity. Neurosci behavior in adolescent girls. Association with psychopathology and Biobehav Rev. 2012;36(1):658–76. neuropsychological functions. Psychopathology. 2008;41(4):226–35. 13. Moran P, Coffey C, Romaniuk H, Olsson C, Borschmann R, Carlin JB, Patton 36. Mann JJ. Neurobiology of suicidal behaviour. Nat Rev Neurosci. 2003;4(10):819–28. GC. The natural history of self-harm from adolescence to young adulthood: 37. Bechara A, Damasio AR, Damasio H, Anderson SW. Insensitivity to future a population-based cohort study. Lancet. 2012;379(9812):236–43. consequences following damage to human prefrontal cortex. Cognition. 14. Hawton K, Saunders KE, O'Connor RC. Self-harm and suicide in adolescents. 1994;50(1):7–15. Lancet. 2012;379(9834):2373–82. 38. Van Heeringen C, Bijttebier S, Godfrin K. Suicidal brains: a review of 15. De Leo D, Heller TS. Who are the kids who self-harm? An Australian self- functional and structural brain studies in association with suicidal behaviour. report school survey. Med J Aust. 2004;181(3):140–4. Neurosci Biobehav Rev. 2011;35(3):688–98. 16. Merikangas K, Mehta RL, Molnar BE, Walters EE, Swendsen JD, Aguilar- 39. Fujii Y, Kitagawa N, Shimizu Y, Mitsui N, Toyomaki A, Hashimoto N, Kako Y, Gaziola S, Bijl R, Borges G, Caraveo-Anduaga JJ, Dewit DJ. Comorbidity Tanaka T, Asakura S, Koyama T, et al. Severity of generalized social anxiety of substance use disorders with mood and anxiety disorders: results of disorder correlates with low executive functioning. Neurosci Lett. 2013;543: the International Consortium in Psychiatric Epidemiology. Addict Behav. 42–6. 1998;23(6):893–907. 40. Biederman J, Petty CR, Wozniak J, Wilens TE, Fried R, Doyle A, Henin A, 17. Hermens DF, Scott EM, White D, Lynch M, Lagopoulos J, Whitwell BG, Bateman C, Evans M, Faraone SV. Impact of executive function deficits in Naismith SL, Hickie IB. Frequent alcohol, nicotine or cannabis use is youth with bipolar I disorder: a controlled study. Psychiatry Res. 2011;186(1): common in young persons presenting for mental healthcare: a cross- 58–64. sectional study. BMJ Open. 2013;3(2). doi: 10.1136/bmjopen-2012- 41. Lee R, Redoblado-Hodge M, Naismith S, Hermens D, Porter M, Hickie I. 002229 Cognitive remediation improves memory and psychosocial functioning in first- 18. WHO. International Classification of Functioning, Disability, and Health: episode psychiatric out-patients. Psychol Med. 2013;43(06):1161–73. Children & Youth Version: ICF-CY. Geneva: World Health Organization; 2007. 42. Baune BT, Fuhr M, Air T, Hering C. Neuropsychological functioning in 19. McGorry PD, Tanti C, Stokes R, Hickie IB, Carnell K, Littlefield LK, Moran J. adolescents and young adults with major depressive disorder–A review. Headspace: Australia’s National Youth Mental Health Foundation–where Psychiatry Res. 2014;218:261–71. young minds come first. Med J Aust. 2007;187(7 Suppl):S68–70. 43. Castaneda AE, Tuulio-Henriksson A, Marttunen M, Suvisaari J, Lönnqvist J. A 20. Howe D, Coates D, Batchelor S. Headspace Gosford data: The local review on cognitive impairments in depressive and anxiety disorders with a application of a National model. Australas Psychiatry. 2014;22(4):374–7. focus on young adults. J Affect Disord. 2008;106(1):1–27. 21. Sahakian BJ, Malloch G, Kennard C. A UK strategy for mental health and 44. Gunther T, Holtkamp K, Jolles J, Herpertz-Dahlmann B, Konrad K. Verbal wellbeing. Lancet. 2010;375(9729):1854–5. memory and aspects of attentional control in children and adolescents 22. Hyman SE. The diagnosis of mental disorders: the problem of reification. with anxiety disorders or depressive disorders. J Affect Disord. 2004; Annu Rev Clin Psychol. 2010;6:155–79. 82(2):265–9. 23. Buckholtz JW, Meyer-Lindenberg A. Psychopathology and the human 45. Simons CJ, Jacobs N, Derom C, Thiery E, Jolles J, van Os J, Krabbendam L. connectome: toward a transdiagnostic model of risk for mental illness. Cognition as predictor of current and follow-up depressive symptoms in Neuron. 2012;74(6):990–1004. the general population. Acta Psychiatr Scand. 2009;120(1):45–52. 24. Lee R, Hermens DF, Redoblado-Hodge MA, Naismith SL, Porter MA, 46. Bitter SM, Mills NP, Adler CM, Strakowski SM, DelBello MP. Progression of Kaur M, White D, Scott EM, Hickie IB. Neuropsychological and socio- amygdala volumetric abnormalities in adolescents after their first manic occupational functioning in young psychiatric outpatients: a episode. J Am Acad Child Adolesc Psychiatry. 2011;50(10):1017–26. longitudinal investigation. PLoS One. 2013;8(3):e58176. 47. Cataldo MG, Nobile M, Lorusso ML, Battaglia M, Molteni M. Impulsivity in 25. Hickie IB, Hermens DF, Naismith SL, Guastella AJ, Glozier N, Scott J, Scott depressed children and adolescents: a comparison between behavioral and EM. Evaluating differential developmental trajectories to adolescent-onset neuropsychological data. Psychiatry Res. 2005;136(2-3):123–33. mood and psychotic disorders. BMC Psychiatry. 2013;13(1):303. 48. Han G, Klimes-Dougan B, Jepsen S, Ballard K, Nelson M, Houri A, Kumra S, 26. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gøtzsche PC, Ioannidis JP, Cullen K. Selective neurocognitive impairments in adolescents with major Clarke M, Devereaux P, Kleijnen J, Moher D. The PRISMA statement for depressive disorder. J Adolesc. 2012;35(1):11–20. reporting systematic reviews and meta-analyses of studies that evaluate 49. Klimkeit EI, Tonge B, Bradshaw JL, Melvin GA, Gould K. Neuropsychological health care interventions: explanation and elaboration. Ann Intern Med. deficits in adolescent unipolar depression. Arch Clin Neuropsychol. 2011; 2009;151(4):W-65–94. 26(7):662–76. 27. Higgins JP, Green S. Cochrane handbook for systematic reviews of 50. Andres S, Boget T, Lazaro L, Penades R, Morer A, Salamero M, Castro- interventions, vol. 5. Hoboken: Wiley Online Library; 2008. Fornieles J. Neuropsychological performance in children and adolescents 28. Deeks JJ, Higgins J, Altman DG. Analysing Data and Undertaking Meta‐ with obsessive-compulsive disorder and influence of clinical variables. Biol Analyses. In: Cochrane Handbook for Systematic Reviews of Interventions: Psychiatry. 2007;61(8):946–51. Cochrane Book Series. 2008. p. 243–96. 51. Andres S, Lazaro L, Salamero M, Boget T, Penades R, Castro-Fornieles J. 29. Korhonen V, Laukkanen E, Antikainen R, Peiponen S, Lehtonen J, Viinamaki Changes in cognitive dysfunction in children and adolescents with H. Effect of major depression on cognitive performance among treatment- obsessive-compulsive disorder after treatment. J Psychiatr Res. 2008;42(6): seeking adolescents. Nord J Psychiatry. 2002;56(3):187–93. 507–14. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 35 of 38

52. Okasha A, Rafaat M, Mahallawy N, Nahas GE, Dawla A, Sayed M, Kholi SE. 73. Rosso IM, Cintron CM, Steingard RJ, Renshaw PF, Young AD, Yurgelun-Todd Cognitive dysfunction in obsessive‐compulsive disorder. Acta Psychiatr DA. Amygdala and hippocampus volumes in pediatric major depression. Scand. 2000;101(4):281–5. Biol Psychiatry. 2005;57(1):21–6. 53. Basso MR, Bornstein RA, Carona F, Morton R. Depression accounts for 74. MacMillan S, Szeszko PR, Moore GJ, Madden R, Lorch E, Ivey J, Banerjee SP, executive function deficits in obsessive-compulsive disorder. Rosenberg DR. Increased amygdala: hippocampal volume ratios associated with Neuropsychiatry Neuropsychol Behav Neurol. 2001;14(4):241–5. severity of anxiety in pediatric major depression. J Child Adolesc 54. Gorwood P. Generalized anxiety disorder and major depressive disorder Psychopharmacol. 2003;13(1):65–73. comorbidity: an example of genetic pleiotropy? Eur Psychiatry. 2004; 75. Schienle A, Ebner F, Schafer A. Localized gray matter volume abnormalities 19(1):27–33. in generalized anxiety disorder. Eur Arch Psychiatry Clin Neurosci. 2011; 55. Pavuluri MN, Passarotti AM, Mohammed T, Carbray JA, Sweeney JA. 261(4):303–7. Enhanced working and verbal memory after lamotrigine treatment in 76. Kendler KS, Neale MC, Kessler RC, Heath AC, Eaves LJ. Major depression and pediatric bipolar disorder. Bipolar Disord. 2010;12(2):213–20. generalized anxiety disorder: same genes,(partly) different environments? 56. Wall CA, Croarkin PE, McClintock SM, Murphy LL, Bandel LA, Sim LA, Arch Gen Psychiatry. 1992;49(9):716. Sampson SM. Neurocognitive effects of repetitive transcranial magnetic 77. Roy M-A, Neale M, Pedersen N, Mathe A, Kendler K. A twin study of stimulation in adolescents with major depressive disorder. Front Psychiatry. generalized anxiety disorder and major depression. Psychol Med. 1995; 2013;4:165. 25(05):1037–49. 57. Perlman G, Simmons AN, Wu J, Hahn KS, Tapert SF, Max JE, Paulus MP, 78. Meng Y, Lui S, Qiu C, Qiu L, Lama S, Huang X, Feng Y, Zhu C, Gong Q, Brown GG, Frank GK, Campbell-Sills L, et al. Amygdala response and Zhang W. Neuroanatomical deficits in drug-naive adult patients with functional connectivity during emotion regulation: a study of 14 depressed generalized social anxiety disorder: a voxel-based morphometry study. adolescents. J Affect Disord. 2012;139(1):75–84. Psychiatry Res. 2013;214(1):9–15. 58. Bond DJ, Lang DJ, Noronha MM, Kunz M, Torres IJ, Su W, Honer WG, 79. Rosenberg DR, Benazon NR, Gilbert A, Sullivan A, Moore GJ. Thalamic Lam RW, Yatham LN. The association of elevated body mass index with volume in pediatric obsessive-compulsive disorder patients before and after reduced brain volumes in first-episode mania. Biol Psychiatry. 2011;70(4): cognitive behavioral therapy. Biol Psychiatry. 2000;48(4):294–300. 381–7. 80. Zuo C, Ma Y, Sun B, Peng S, Zhang H, Eidelberg D, Guan Y. Metabolic imaging 59. Goodman M, Hazlett EA, Avedon JB, Siever DR, Chu KW, New AS. of bilateral anterior capsulotomy in refractory obsessive compulsive disorder: Anterior cingulate volume reduction in adolescents with borderline an FDG PET study. J Cereb Blood Flow Metab. 2013;33(6):880–7. personality disorder and co-morbid major depression. J Psychiatr Res. 81. Huyser C, van den Heuvel OA, Wolters LH, de Haan E, Boer F, Veltman DJ. 2011;45(6):803–7. Increased orbital frontal gray matter volume after cognitive behavioural 60. Pan LA, Hassel S, Segreti A, Nau S, Brent D, Phillips M. Differential patterns therapy in paediatric obsessive compulsive disorder. World J Biol Psychiatry. of activity and functional connectivity in emotion processing neural circuitry 2013;14(4):319–31. to angry and happy faces in adolescents with and without suicide attempt. 82. Zarei M, Mataix-Cols D, Heyman I, Hough M, Doherty J, Burge L, Winmill L, Psychol Med. 2013;43(10):2129–42. Nijhawan S, Matthews PM, James A. Changes in gray matter volume and white 61. Bush G, Luu P, Posner MI. Cognitive and emotional influences in anterior matter microstructure in adolescents with obsessive-compulsive disorder. Biol cingulate cortex. Trends Cogn Sci. 2000;4(6):215–22. Psychiatry. 2011;70(11):1083–90. 62. Pan LA, Batezati-Alves SC, Almeida JR, Segreti A, Akkal D, Hassel S, 83. MacMaster FP, Russell A, Mirza Y, Keshavan MS, Banerjee SP, Bhandari R, Lakdawala S, Brent DA, Phillips ML. Dissociable patterns of neural activity Boyd C, Lynch M, Rose M, Ivey J, et al. Pituitary volume in pediatric during response inhibition in depressed adolescents with and without obsessive-compulsive disorder. Biol Psychiatry. 2006;59(3):252–7. suicidal behavior. J Am Acad Child Adolesc Psychiatry. 2011;50(6):602–611. 84. Rosenberg DR, Keshavan MS, Dick EL, Bagwell WW, MacMaster FP, e603. Birmaher B. Corpus callosal morphology in treatment-naive pediatric 63. Medina KL, Nagel BJ, Park A, McQueeny T, Tapert SF. Depressive symptoms obsessive compulsive disorder. Prog Neuropsychopharmacol Biol in adolescents: associations with white matter volume and marijuana use. J Psychiatry. 1997;21(8):1269–83. Child Psychol Psychiatry. 2007;48(6):592–600. 85. Chang KD, Wagner C, Garrett A, Howe M, Reiss A. A preliminary functional 64. Jarvis K, DelBello MP, Mills N, Elman I, Strakowski SM, Adler CM. Neuroanatomic magnetic resonance imaging study of prefrontal‐amygdalar activation comparison of bipolar adolescents with and without cannabis use disorders. J changes in adolescents with bipolar depression treated with lamotrigine. Child Adolesc Psychopharmacol. 2008;18(6):557–63. Bipolar Disord. 2008;10(3):426–31. 65. Cornelius JR, Aizenstein HJ, Hariri AR. Amygdala reactivity is inversely related 86. Phan KL, Coccaro EF, Angstadt M, Kreger KJ,MaybergHS,LiberzonI,SteinMB. to level of cannabis use in individuals with comorbid cannabis dependence Corticolimbic brain reactivity to social signals of threat before and after sertraline and major depression. Addict Behav. 2010;35(6):644–6. treatment in generalized social phobia. Biol Psychiatry. 2013;73(4):329–36. 66. De Bellis MD, Narasimhan A, Thatcher DL, Keshavan MS, Soloff P, Clark DB. 87. McClure EB, Adler A, Monk CS, Cameron J, Smith S, Nelson EE, Leibenluft E, Prefrontal cortex, thalamus, and cerebellar volumes in adolescents and Ernst M, Pine DS. fMRI predictors of treatment outcome in pediatric anxiety young adults with adolescent-onset alcohol use disorders and comorbid disorders. Psychopharmacology (Berl). 2007;191(1):97–105. mental disorders. Alcohol Clin Exp Res. 2005;29(9):1590–600. 88. Wegbreit E, Ellis JA, Nandam A, Fitzgerald JM, Passarotti AM, Pavuluri 67. Hermens DF, Lagopoulos J, Tobias-Webb J, De Regt T, Dore G, Juckes L, Latt N, MN, Stevens MC. Amygdala functional connectivity predicts Hickie IB. Pathways to alcohol-induced brain impairment in young people: a pharmacotherapy outcome in pediatric bipolar disorder. Brain review. Cortex. 2013;49:3–17. Connectivity. 2011;1(5):411–22. 68. Gabbay V, Mao X, Klein RG, Ely BA, Babb JS, Panzer AM, Alonso CM, 89. Hariri AR, Drabant EM, Munoz KE, Kolachana BS, Mattay VS, Egan MF, Shungu DC. Anterior cingulate cortex gamma-aminobutyric acid in Weinberger DR. A susceptibility gene for affective disorders and the depressed adolescents: relationship to anhedonia. Arch Gen Psychiatry. response of the human amygdala. Arch Gen Psychiatry. 2005;62(2): 2012;69(2):139–49. 146–52. 69. Pannekoek JN, van der Werff SJ, van den Bulk BG, van Lang ND, Rombouts SA, 90. Aghajani M, Veer IM, van Lang ND, Meens PH, van den Bulk BG, Rombouts van Buchem MA, Vermeiren RR, van der Wee NJ. Reduced anterior cingulate gray SA, Vermeiren RR, van der Wee NJ: Altered white-matter architecture in matter volume in treatment-naive clinically depressed adolescents. NeuroImage treatment-naive adolescents with clinical depression. Psychol Med. 2013:1- Clin. 2014;4:336–42. 12. Epub ahead of print. 70. Reynolds S, Carrey N, Jaworska N, LangevinLM,YangXR,MacmasterFP.Cortical 91. Gilbert AR, Moore GJ, Keshavan MS, Paulson LAD, Narula V, Mac Master FP, thickness in youth with major depressive disorder. BMC Psychiatry. 2014;14:83. Stewart CM, Rosenberg DR. Decrease in thalamic volumes of pediatric 71. Lagopoulos J, Hermens DF, Naismith SL, Scott EM, Hickie IB. Frontal lobe patients with obsessive-compulsive disorder who are taking paroxetine. changes occur early in the course of affective disorders in young people. Arch Gen Psychiatry. 2000;57(5):449–56. BMC Psychiatry. 2012;12(1):4. 92. Lagopoulos J, Hermens DF, Hatton SN, Tobias-Webb J, Griffiths K, Naismith 72. Yucel K, McKinnon MC, Chahal R, Taylor VH, Macdonald K, Joffe R, SL, Scott EM, Hickie IB. Microstructural white matter changes in the corpus MacQueen GM. Anterior cingulate volumes in never-treated patients with callosum of young people with bipolar disorder: a diffusion tensor imaging major depressive disorder. Neuropsychopharmacology. 2008;33(13):3157–63. study. PLoS One. 2013;8(3):e59108. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 36 of 38

93. Huang H, Gundapuneedi T, Rao U. White matter disruptions in adolescents 113. Harkness KL, Stewart JG, Wynne-Edwards KE. Cortisol reactivity to social exposed to childhood maltreatment and vulnerability to psychopathology. stress in adolescents: role of depression severity and child maltreatment. Neuropsychopharmacology. 2012;37(12):2693–701. Psychoneuroendocrinology. 2011;36(2):173–81. 94. Gruner P, Vo A, Ikuta T, Mahon K, Peters BD, Malhotra AK, Ulug AM, Szeszko 114. Doane LD, Mineka S, Zinbarg RE, Craske M, Griffith JW, Adam EK. Are flatter PR. White matter abnormalities in pediatric obsessive-compulsive disorder. diurnal cortisol rhythms associated with major depression and anxiety Neuropsychopharmacology. 2012;37(12):2730–9. disorders in late adolescence? The role of life stress and daily negative 95. Henderson SE, Johnson AR, Vallejo AI, Katz L, Wong E, Gabbay V. A emotion. Dev Psychopathol. 2013;25(3):629–42. preliminary study of white matter in adolescent depression: 115. Robillard R, Naismith S, Rogers N, Scott E, Ip T, Hermens D, Hickie I. Sleep- relationships with illness severity, anhedonia, and irritability. Front wake cycle and melatonin rhythms in adolescents and young adults with Psychiatry. 2013;4:152. mood disorders: Comparison of unipolar and bipolar phenotypes. Eur 96. Lagopoulos J, Hermens D, Hatton S, Battisti R, Tobias-Webb J, White D, Psychiatry. 2013;28:412–6. Naismith S, Scott E, Ryder W, Bennett M. Microstructural white matter 116. Robillard R, Naismith SL, Rogers NL, Ip TK, Hermens DF, Scott EM, Hickie IB. changes are correlated with the stage of psychiatric illness. Translat Delayed sleep phase in young people with unipolar or bipolar affective Psychiatry. 2013;3(4):e248. disorders. J Affect Disord. 2013;145(2):260–3. 97. Bustillo JR. Use of proton magnetic resonance spectroscopy in the 117. Murray CB, Murphy LK, Palermo TM, Clarke GM. Pain and sleep-wake treatment of psychiatric disorders: a critical update. Dialogues Clin Neurosci. disturbances in adolescents with depressive disorders. J Clin Child Adolesc 2013;15(3):329–37. Psychol. 2012;41(4):482–90. 98. Lazaro L, Bargallo N, Andres S, Falcon C, Morer A, Junque C, Castro-Fornieles J. 118. Ankers D, Jones SH. Objective assessment of circadian activity and sleep Proton magnetic resonance spectroscopy in pediatric obsessive-compulsive patterns in individuals at behavioural risk of hypomania. J Clin Psychol. disorder: longitudinal study before and after treatment. Psychiatry Res. 2012; 2009;65(10):1071–86. 201(1):17–24. 119. Landsness EC, Goldstein MR, Peterson MJ, Tononi G, Benca RM. 99. Chu WJ, Delbello MP, Jarvis KB, Norris MM, Kim MJ, Weber W, Lee JH, Antidepressant effects of selective slow wave sleep deprivation in major Strakowski SM, Adler CM. Magnetic resonance spectroscopy imaging depression: a high-density EEG investigation. J Psychiatr Res. 2011;45(8): of lactate in patients with bipolar disorder. Psychiatry Res. 2013; 1019–26. 213(3):230–4. 120. Armitage R, Emslie G, Rintelmann J. The effect of fluoxetine on sleep EEG in 100. Patel NC, Cecil KM, Strakowski SM, Adler CM, DelBello MP. Neurochemical childhood depression: a preliminary report. Neuropsychopharmacology. alterations in adolescent bipolar depression: a proton magnetic resonance 1997;17(4):241–5. spectroscopy pilot study of the prefrontal cortex. J Child Adolesc 121. Scott EM, Robillard R, Hermens DF, Naismith SL, Rogers NL, Ip TK, White Psychopharmacol. 2008;18(6):623–7. D, Guastella A, Whitwell B, Smith KL. Dysregulated sleep-wake cycles in 101. Strawn JR, Patel NC, Chu WJ, Lee JH, Adler CM, Kim MJ, Bryan HS, Alfieri DC, young people are associated with emerging stages of major mental Welge JA, Blom TJ, et al. Glutamatergic effects of divalproex in adolescents disorders. Early Interv Psychiatry. 2014;10:63–70. with mania: a proton magnetic resonance spectroscopy study. J Am Acad 122. Ashton CH, Marshall EF, Hassanyeh F, Marsh VR, Wright-Honari S. Biological Child Adolesc Psychiatry. 2012;51(6):642–51. correlates of deliberate self-harm behaviour: a study of 102. Goodyer IM, Herbert J, Altham PM. Adrenal steroid secretion and major electroencephalographic, biochemical and psychological variables in depression in 8- to 16-year-olds, III. Influence of cortisol/DHEA ratio at parasuicide. Acta Psychiatr Scand. 1994;90(5):316–23. presentation on subsequent rates of disappointing life events and 123. Graae F, Tenke C, Bruder G, Rotheram M-J, Piacentini J, Castro-Blanco D, persistent major depression. Psychol Med. 1998;28(2):265–73. Leite P, Towey J. Abnormality of EEG alpha asymmetry in female 103. Abelson JL, Curtis GC. Hypothalamic-pituitary-adrenal axis activity in panic adolescent suicide attempters. Biol Psychiatry. 1996;40(8):706–13. disorder: prediction of long-term outcome by pretreatment cortisol levels. 124. Pechtel P, Pizzagalli DA. Disrupted reinforcement learning and Am J Psychiatry. 1996;153(1):69–73. maladaptive behavior in women with a history of childhood sexual 104. Granger DA, Weisz JR, Kauneckis D. Neuroendocrine reactivity, internalizing abuse: a high-density event-related potential study. JAMA Psychiatry. behavior problems, and control-related cognitions in clinic-referred children 2013;70(5):499–507. and adolescents. J Abnorm Psychol. 1994;103(2):267–76. 125. Davis CE, Hauf JD, Wu DQ, Everhart DE. Brain function with complex decision 105. Coplan JD, Wolk SI, Goetz RR, Ryan ND, Dahl RE, Mann JJ, Weissman MM. making using electroencephalography. Int J Psychophysiol. 2011;79(2):175–83. Nocturnal growth hormone secretion studies in adolescents with or 126. Macar F, Vidal F. The CNV peak: an index of decision making and temporal without major depression re-examined: integration of adult clinical follow- memory. Psychophysiology. 2003;40(6):950–4. up data. Biol Psychiatry. 2000;47(7):594–604. 127. Vaidyanathan U, Welo EJ, Malone SM, Burwell SJ, Iacono WG. The effects of 106. Mathew SJ, Coplan JD, Goetz RR, Feder A, Greenwald S, Dahl RE, Ryan ND, recurrent episodes of depression on startle responses. Psychophysiology. Mann JJ, Weissman MM. Differentiating depressed adolescent 24 h cortisol 2014;51(1):103–9. secretion in light of their adult clinical outcome. 128. Bakker M, Tijssen M, Koelman J, Boer F. Normalization of the auditory startle Neuropsychopharmacology. 2003;28(7):1336–43. reflex after symptom reduction in children with anxiety disorders. J Psychiatr 107. McCracken JT, Poland RE, Lutchmansingh P, Edwards C. Sleep Res. 2011;45(6):796–802. electroencephalographic abnormalities in adolescent depressives: effects of 129. Jaworska N, Blondeau C, Tessier P, Norris S, Fusee W, Blier P, Knott V. scopolamine. Biol Psychiatry. 1997;42(7):577–84. Response prediction to antidepressants using scalp and source-localized 108. Adam EK, Doane LD, Zinbarg RE, Mineka S, Craske MG, Griffith JW. loudness dependence of auditory evoked potential (LDAEP) slopes. Prog Prospective prediction of major depressive disorder from cortisol Neuro-Psychopharmacol Biol Psychiatry. 2013;44:100–7. awakening responses in adolescence. Psychoneuroendocrinology. 130. Robinson OJ, Vytal K, Cornwell BR, Grillon C. The impact of anxiety upon 2010;35(6):921–31. cognition: perspectives from human threat of shock studies. Front Hum 109. Ellenbogen MA, Hodgins S, Walker CD, Couture S, Adam S. Daytime cortisol Neurosci. 2013;7:203. and stress reactivity in the offspring of parents with bipolar disorder. 131. Price JL, Drevets WC. Neural circuits underlying the pathophysiology of Psychoneuroendocrinology. 2006;31(10):1164–80. mood disorders. Trends Cogn Sci. 2012;16(1):61–71. 110. Ellenbogen MA, Santo JB, Linnen AM, Walker CD, Hodgins S. High cortisol 132. Etkin A. Functional neuroanatomy of anxiety: a neural circuit perspective. In: levels in the offspring of parents with bipolar disorder during two weeks of Behavioral neurobiology of anxiety and its treatment. Berlin: Springer; 2010. p. daily sampling. Bipolar Disord. 2010;12(1):77–86. 251–77. 111. Rao U, Dahl RE, Ryan ND, Birmaher B, Williamson DE, Giles DE, Rao R, 133. Carrasco M, Harbin SM, Nienhuis JK, Fitzgerald KD, Gehring WJ, Hanna Kaufman J, Nelson B. The relationship between longitudinal clinical GL. Increased error-related brain activity in youth with obsessive- course and sleep and cortisol changes in adolescent depression. Biol compulsive disorder and unaffected siblings. Depress Anxiety. 2013; Psychiatry. 1996;40(6):474–84. 30(1):39–46. 112. Rao U, Poland RE. Electroencephalographic sleep and hypothalamic- 134. Carrasco M, Hong C, Nienhuis JK, Harbin SM, Fitzgerald KD, Gehring WJ, Hanna pituitary-adrenal changes from episode to recovery in depressed GL. Increased error-related brain activity in youth with obsessive-compulsive adolescents. J Child Adolesc Psychopharmacol. 2008;18(6):607–13. disorder and other anxiety disorders. Neurosci Lett. 2013;541:214–8. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 37 of 38

135. Hajcak G, Franklin ME, Foa EB, Simons RF. Increased error-related brain 158. Keller M. Follow-up and family study of anxious depression. Am J Psychiatry. activity in pediatric obsessive-compulsive disorder before and after 1991;148(1):1512–7. treatment. Am J Psychiatry. 2008;165(1):116–23. 159. Joffe RT, Bagby RM, Levitt A. Anxious and nonanxious depression. Am J 136. Stern ER, Liu Y, Gehring WJ, Lister JJ, Yin G, Zhang J, Fitzgerald KD, Himle JA, Psychiatr. 1993;150:1257. Abelson JL, Taylor SF. Chronic medication does not affect hyperactive error 160. Fava M, Rush AJ, Alpert JE, Carmin CN, Balasubramani G, Wisniewski SR, responses in obsessive-compulsive disorder. Psychophysiology. 2010;47(5): Trivedi MH, Biggs MM, Shores-Wilson K. What clinical and symptom 913–20. features and comorbid disorders characterize outpatients with anxious 137. Plana T, Gracia R, Mendez I, Pintor L, Lazaro L, Castro-Fornieles J. major depressive disorder: a replication and extension. Can J Psychiatry. Total serum cholesterol levels and suicide attempts in child and 2006;51(13):823–35. adolescent psychiatric inpatients. Eur Child Adolesc Psychiatry. 2010; 161. Watson D. Rethinking the mood and anxiety disorders: a quantitative 19(7):615–9. hierarchical model for DSM-V. J Abnorm Psychol. 2005;114(4):522. 138. Apter A, Laufer N, Bar-Sever M, Har-Even D, Ofek H, Weizman A. Serum 162. Pine DS, Goldstein RB, Wolk S, Weissman MM. The association between cholesterol, suicidal tendencies, impulsivity, aggression, and depression in childhood depression and adulthood body mass index. Pediatrics. 2001; adolescent psychiatric inpatients. Biol Psychiatry. 1999;46(4):532–41. 107(5):1049–56. 139. De Berardis D, Campanella D, Serroni N, Moschetta FS, Di Emidio F, Conti C, 163. Holsboer F. Stress, hypercortisolism and corticosteroid receptors in Carano A, Acciavatti T, Di Iorio G, Martinotti G. Alexithymia, suicide risk and depression: implicatons for therapy. J Affect Disord. 2001;62(1):77–91. serum lipid levels among adult outpatients with panic disorder. Compr 164. Harvey MA, Sellman JD, Porter RJ, Frampton CM. The relationship between Psychiatry. 2013;54(5):517–22. non-acute adolescent cannabis use and cognition. Drug Alcohol Rev. 2007; 140. Taylor V, Macdonald K, McKinnon MC, Joffe RT, MacQueen GM. Increased 26(3):309–19. rates of obesity in first-presentation adults with mood disorders over the 165. Hermens DF, Lee RS, De Regt T, Lagopoulos J, Naismith SL, Scott EM, Hickie course of four-year follow-up. J Affect Disord. 2008;109(1-2):127–31. IB. Neuropsychological functioning is compromised in binge drinking 141. Kilbourne AM, Rofey DL, McCarthy JF, Post EP, Welsh D, Blow FC. Nutrition young adults with depression. Psychiatry Res. 2013;210:256–62. and exercise behavior among patients with bipolar disorder1. Bipolar 166. Fleck DE, Shear PK, Madore M, Strakowski SM. Wisconsin Card Sorting Test Disord. 2007;9(5):443–52. performance in bipolar disorder: effects of mood state and early course. 142. Lu X-Y, Kim CS, Frazer A, Zhang W. Leptin: a potential novel antidepressant. Bipolar Disord. 2008;10(4):539–45. Proc Natl Acad Sci. 2006;103(5):1593–8. 167. Kıvırcık BB, Yener GG, Alptekin K, Aydın H. Event-related potentials and 143. Wildes JE, Marcus MD, Fagiolini A. Obesity in patients with bipolar disorder: neuropsychological tests in obsessive–compulsive disorder. Prog Neuro- a biopsychosocial-behavioral model. J Clin Psychiatry. 2006;67(6):904–15. Psychopharmacol Biol Psychiatry. 2003;27(4):601–6. 144. Chamberlain SR, Sahakian BJ. The neuropsychology of mood disorders. Curr 168. Schmid M, Hammar A. A follow-up study of first episode major depressive Psychiatry Rep. 2006;8(6):458–63. disorder. Impairment in inhibition and semantic fluency-potential predictors 145. Kapur S, Phillips A, Insel T. Why has it taken so long for biological psychiatry for relapse? Front Psychol. 2013;4:633. to develop clinical tests and what to do about it&quest. Mol Psychiatry. 169. Torres IJ, DeFreitas VG, DeFreitas CM, Kauer-Sant'Anna M, Bond DJ, Honer WG, 2012;17(12):1174–9. Lam RW, Yatham LN. Neurocognitive functioning in patients with bipolar I 146. Lee R, Hermens DF, Naismith SL, Lagopoulos J, Jones A, Chitty KM, disorder recently recovered from a first manic episode. J Clin Psychiatry. 2010; White D, Robillard R, Scott E, Hickie I. Neuropsychological and 71(9):1234–42. functional outcomes in recent-onset major depression, bipolar disorder 170. Ehrlich S, Noam GG, Lyoo IK, Kwon BJ, Clark MA, Renshaw PF. White matter and schizophrenia-spectrum disorders: a longitudinal cohort study. hyperintensities and their associations with suicidality in psychiatrically Transl Psychiatry. 2015;5:e555. hospitalized children and adolescents. J Am Acad Child Adolesc Psychiatry. 147. Naismith SL, Redoblado-Hodge MA, Lewis SJ, Scott EM, Hickie IB. Cognitive 2004;43(6):770–6. training in affective disorders improves memory: a preliminary study using 171. Ehrlich S, Breeze JL, Hesdorffer DC, Noam GG, Hong X, Alban RL, Davis SE, the NEAR approach. J Affect Disord. 2010;121(3):258–62. Renshaw PF. White matter hyperintensities and their association with 148. Alvarez LM, Cortés Sotres JF, León SO, Estrella J, Sánchez Sosa JJ. Computer suicidality in depressed young adults. J Affect Disord. 2005;86(2):281–7. program in the treatment for major depression and cognitive impairment 172. Chitty KM, Lagopoulos J, Hickie IB, Hermens DF. Risky alcohol use in young in university students. Comput Hum Behav. 2008;24(3):816–26. persons with emerging bipolar disorder is associated with increased 149. De Hert M, Cohen D, Bobes J, Cetkovich-Bakmas M, Leucht S, Ndetei DM, oxidative stress. J Affect Disord. 2013;150(3):1238–41. Newcomer JW, Uwakwe R, Asai I, Möller H-J. Physical illness in patients with 173. Adler CM, DelBello MP, Jarvis K, Levine A, Adams J, Strakowski SM. Voxel- severe mental disorders. II. Barriers to care, monitoring and treatment based study of structural changes in first-episode patients with bipolar guidelines, plus recommendations at the system and individual level. World disorder. Biol Psychiatry. 2007;61(6):776–81. Psychiatry. 2011;10(2):138. 174. Diler RS, Ladouceur CD, Segreti A, Almeida JR, Birmaher B, Axelson DA, 150. Miret M, Ayuso-Mateos JL, Sanchez-Moreno J, Vieta E. Depressive disorders Phillips ML, Pan LA. Neural correlates of treatment response in depressed and suicide: epidemiology, risk factors, and burden. Neurosci Biobehav Rev. bipolar adolescents during emotion processing. Brain Imaging Behav. 2013; 2013;37(10):2372–4. 7(2):227–35. 151. WHO. Mental health: new understanding, new hope. Geneva: World Health 175. Forbes EE, Christopher May J, Siegle GJ, Ladouceur CD, Ryan ND, Carter CS, Organization; 2001. Birmaher B, Axelson DA, Dahl RE. Reward-related decision-making in 152. Cáceda R, Nemeroff CB, Harvey PD. Toward an understanding of decision pediatric major depressive disorder: an fMRI study. J Child Psychol making in severe mental illness. 2014. Psychiatry. 2006;47(10):1031–40. 153. Pfrieger F. Cholesterol homeostasis and function in neurons of the central 176. Forbes EE, Olino TM, Ryan ND, Birmaher B, Axelson D, Moyles DL, Dahl RE. nervous system. Cell Mol Life Sci. 2003;60(6):1158–71. Reward-related brain function as a predictor of treatment response in 154. Papakostas GI, Öngür D, Iosifescu DV, Mischoulon D, Fava M. Cholesterol in adolescents with major depressive disorder. Cogn Affect Behav Neurosci. mood and anxiety disorders: review of the literature and new hypotheses. 2010;10(1):107–18. Eur Neuropsychopharmacol. 2004;14(2):135–42. 177. Gabbay V, Ely BA, Li Q, Bangaru SD, Panzer AM, Alonso CM, Castellanos FX, 155. Tyano S, Zalsman G, Ofek H, Blum I, Apter A, Wolovik L, Sher L, Sommerfeld E, Milham MP. Striatum-based circuitry of adolescent depression and Harell D, Weizman A. Plasma serotonin levels and suicidal behavior in anhedonia. J Am Acad Child Adolesc Psychiatry. 2013;52(6):628–641.e613. adolescents. Eur Neuropsychopharmacol. 2006;16(1):49–57. 178. Gao W, Jiao Q, Qi R, Zhong Y, Lu D, Xiao Q, Lu S, Xu C, Zhang Y, Liu 156. Hickie IB, Scott EM, Hermens DF, Naismith SL, Guastella AJ, Kaur M, Sidis A, X, et al. Combined analyses of gray matter voxel-based morphometry Whitwell B, Glozier N, Davenport T, et al. Applying clinical staging to young and white matter tract-based spatial statistics in pediatric bipolar mania. people who present for mental health care. Early Interv Psychiatry. 2013;7(1): J Affect Disord. 2013;150(1):70–6. 31–43. 179. Gilbert AR, Akkal D, Almeida JR, Mataix-Cols D, Kalas C, Devlin B, Birmaher B, 157. Insel TR, Cuthbert BN, Garvey MA, Heinssen RK, Pine DS, Quinn KJ, Sanislow CA, Phillips ML. Neural correlates of symptom dimensions in pediatric obsessive- Wang PS. Research domain criteria (RDoC): toward a new classification compulsive disorder: a functional magnetic resonance imaging study. J Am framework for research on mental disorders. Am J Psychiatr. 2010;167(7):748–51. Acad Child Adolesc Psychiatry. 2009;48(9):936–44. Iorfino et al. BMC Psychiatry (2016) 16:156 Page 38 of 38

180. Hatton SN, Lagopoulos J, Hermens DF, Naismith SL, Bennett MR, Hickie IB. 199. Soreni N, Apter A, Weizman A, Don-Tufeled O, Leschiner S, Karp L, Gavish M. Correlating anterior insula gray matter volume changes in young people Decreased platelet peripheral-type benzodiazepine receptors in adolescent with clinical and neurocognitive outcomes: an MRI study. BMC Psychiatry. inpatients with repeated suicide attempts. Biol Psychiatry. 1999;46(4):484–8. 2012;12(1):45. 200. Goldstein BI, Birmaher B, Axelson DA, Goldstein TR, Esposito-Smythers C, 181. Ho TC, Yang G, Wu J, Cassey P, Brown SD, Hoang N, Chan M, Connolly CG, Strober MA, Hunt J, Leonard H, Gill MK, Iyengar S, et al. Preliminary findings Henje-Blom E, Duncan LG, et al. Functional connectivity of negative emotional regarding overweight and obesity in pediatric bipolar disorder. J Clin processing in adolescent depression. J Affect Disord. 2014;155:65–74. Psychiatry. 2008;69(12):1953–9. 182. Huyser C, Veltman DJ, Wolters LH, de Haan E, Boer F. Developmental aspects of error and high-conflict-related brain activity in pediatric obsessive-compulsive disorder: a fMRI study with a Flanker task before and after CBT. J Child Psychol Psychiatry. 2011;52(12):1251–60. 183. Ladouceur CD, Farchione T, Diwadkar V, Pruitt P, Radwan J, Axelson DA, Birmaher B, Phillips ML. Differential patterns of abnormal activity and connectivity in the amygdala-prefrontal circuitry in bipolar-I and bipolar- NOS youth. J Am Acad Child Adolesc Psychiatry. 2011;50(12):1289.e1272. 184. Lazaro L, Caldú X, Junqué C, Bargalló N, Andrés S, Morer A, Castro-Fornieles J. Cerebral activation in children and adolescents with obsessive– compulsive disorder before and after treatment: A functional MRI study. J Psychiatr Res. 2008;42(13):1051–9. 185. Lisy ME, Jarvis KB, DelBello MP, Mills NP, Weber WA, Fleck D, Strakowski SM, Adler CM. Progressive neurostructural changes in adolescent and adult patients with bipolar disorder. Bipolar Disord. 2011;13(4):396–405. 186. Pavuluri MN, Passarotti AM, Parnes SA, Fitzgerald JM, Sweeney JA. A pharmacological functional magnetic resonance imaging study probing the interface of cognitive and emotional brain systems in pediatric bipolar disorder. J Child Adolesc Psychopharmacol. 2010; 20(5):395–406. 187. Pavuluri MN, Passarotti AM, Lu LH, Carbray JA, Sweeney JA. Double-blind randomized trial of risperidone versus divalproex in pediatric bipolar disorder: fMRI outcomes. Psychiatry Res. 2011;193(1):28–37. 188. Rauch SL, Shin LM, Dougherty DD, Alpert NM, Fischman AJ, Jenike MA. Predictors of fluvoxamine response in contamination-related obsessive compulsive disorder: a PET symptom provocation study. Neuropsychopharmacology. 2002;27(5):782–91. 189. Schneider MR, Adler CM, Whitsel R, Weber W, Mills NP, Bitter SM, Eliassen J, Strakowski SM, Delbello MP. The effects of ziprasidone on prefrontal and amygdalar activation in manic youth with bipolar disorder. Isr J Psychiatry Relat Sci. 2012;49(2):112–20. 190. Kaur M, Lagopoulos J, Lee RSC, Ward PB, Naismith SL, Hickie IB, Hermens DF. Longitudinal associations between mismatch negativity and disability in early schizophrenia-and affective-spectrum disorders. Prog Neuro- Psychopharmacol Biol Psychiatry. 2013;46:161–9. 191. Jaworska N, McIntosh J, Villeneuve C, Thompson A, Fisher D, Milin R, Knott V. Effects of nicotine on electroencephalography and affect in adolescent females with major depressive disorder: a pilot study. J Addict Med. 2011; 5(2):123–33. 192. Chitty KM, Kaur M, Lagopoulos J, Hickie IB, Hermens DF. Risky alcohol use predicts temporal mismatch negativity impairments in young people with bipolar disorder. Biol Psychol. 2014;99:60–8. 193. Ehlers CL, Phillips E, Criado JR, Gilder DA. N4 component responses to pre- pulse startle stimuli in young adults: relationship to alcohol dependence. Psychiatry Res. 2011;188(2):237–44. 194. Croarkin PE, Nakonezny PA, Husain MM, Port JD, Melton T, Kennard BD, Emslie GJ, Kozel FA, Daskalakis ZJ. Evidence for Pretreatment LICI Deficits Among Depressed Children and Adolescents With Nonresponse to Fluoxetine. Brain Stimul. 2014;7(2):243–51. 195. Dai Q, Feng Z. More excited for negative facial expressions in depression: Evidence from an event-related potential study. Clin Neurophysiol. 2012; 123(11):2172–9. 196. El-Badri SM, Ashton CH, Moore PB, Marsh VR, Ferrier IN. Electrophysiological and cognitive function in young euthymic patients with bipolar affective disorder. Bipolar Disord. 2001;3(2):79–87. 197. Houston RJ, Bauer LO, Hesselbrock VM. Depression and familial risk for substance dependence: a P300 study of young women. Psychiatry Res. 2003;124(1):49–62. 198. Mannie ZN, Williams C, Diesch J, Steptoe A, Leeson P, Cowen PJ. Cardiovascular and metabolic risk profile in young people at familial risk of depression. Br J Psychiatry. 2013;203(1):18–23. Chapter 6: Discussion

The work presented in this thesis had two key aims. The first was to investigate the longitudinal outcomes of young people attending early intervention youth mental health services. The second was to examine the utility of technology- enabled assessment and neurobiological measures as future solutions to inform the delivery of personalised youth mental health care. While, the results from individual studies have been discussed in their respective discussion sections

(chapters two to five), the first section of this chapter summarises the main findings of these studies in relation to the aims of this thesis. The sections that follow collectively discuss the broader implications of, and themes derived from, these studies in the context of the current literature. The final sections of this discussion focus on the overall limitations and future directions of this research, before closing with the concluding remarks.

6.1. Summary of findings

6.1.1. Delineating long-term outcomes

Chapters two and three address the first aim of this thesis and generate new insights into the long-term outcomes of young people attending early intervention services. Chapter two highlights that the majority of those who present early in the course of mental health disorders already have established impairment in major areas of educational, employment or social function. While such findings have been common in the at-risk to psychosis and first-episode psychosis literature, they have not been as extensively delineated in a transdiagnostic population largely comprised of early mood disorders.

Critically, this chapter also highlights that a large number of these young people remain functionally impaired over the course of care. Individual trajectories of social and occupational functioning were highly variable, however the group- based trajectory modelling used in this study identified six trajectories that could effectively describe subgroups of individuals similar in terms of level of

101 functioning at entry and their rate of change over the course of clinical care. Of those individuals who presented with moderate to severe functional impairment, a small minority (4%) were able to functionally recover, while the vast majority (>60%) remained chronically impaired.

Chapter three highlights that individuals who present with a history of suicidal behaviours are particularly vulnerable to poorer outcomes. While epidemiologically-based reports show that such behaviours in the community can have a wider impact on individuals, the results from this chapter show that among a clinical population of help-seeking young people with early disorders, these behaviours are not only a determinant of immediate distress and additional impairment but also a predictor of later onset of more severe (i.e. bipolar or unstable mood type) illness, greater functional impairment and comorbidity. Importantly, at entry to services, these behaviours are not restricted to those with more severe illness types and emerge among those with no previous history over the course of care. These findings have important implications for an enhanced focus on reducing suicidal thoughts and behaviours throughout the course of clinical care for all of those who present to such services – not just those who are rated as ‘higher-risk’.

6.1.2. Future solutions for personalised mental health care

Chapters four and five address the second aim of this thesis and examine the utility of future solutions that attempt to inform the delivery of personalised youth mental health care. The study presented in chapter four achieved two primary goals. The first was to extend our understanding of the nature of suicidal thoughts and behaviours among help-seeking young people with greater insights into the clinical and functional correlates of current suicide ideation at initial presentation. The second was to demonstrate the impact technology-enabled assessment can have on service provisions. The

102 development and integration of online assessment and feedback with existing clinical service infrastructures enabled a more timely response to individuals reporting high levels of suicidality. Young people who otherwise would have been on a waitlist to see a face-to-face clinician were able to complete an online assessment, and when needed for particular individuals, an appropriate action could be enacted by the clinical service to ensure timely access to care. This process was not entirely online or automated, but rather, a semi-automated integration whereby data was automatically collected and then algorithmically derived feedback was forwarded to clinicians and used to drive clinical decisions.

Chapter five presents the results of an extensive systematic review of 134 studies that provides a better understanding of the neurobiological correlates across multiple outcome domains among young people with emerging mental health disorders. The review highlights that there is a major focus on clinical syndromes and symptoms and argues for a broader focus on identifying secondary risks, such as social and economic participation and suicidal thoughts and behaviours. Nonetheless the findings provide an overview of relationship between measures of brain function (i.e. neuropsychology, neuroimaging) and multidimensional outcomes that may be used to enhance further assessment in practice. More specifically, the results from this review reveal the extent to which specific neurobiological or psychological processes, such as neurocognition and circadian dysregulation, are associated with poorer outcomes among these transdiagnostic cohorts, which may help to identify individuals at risk for poorer outcomes.

6.2. Early intervention—not early enough

The past decade has seen a major shift towards early intervention and preemptive psychiatry, which recognises that trajectories of mental health

103 disorders and associated impairment are often not fixed, but instead malleable to change (Insel, 2007, Henry et al., 2010, Harrison et al., 2001, Hickie et al.,

2013a). This has been accompanied by major investment in the development of early intervention services for young people which aim to intervene prior to the onset of major mental health disorders and secondary morbidity (McGorry et al., 2014). A consistent finding across all three empirical papers presented here

(chapters two, three & four) is that impairment and concurrent morbidity are well-established among young people presenting to early intervention services.

Rates of disengagement from education, employment and training were already up to twice the national average (AIHW, 2017) and the rate of prior suicide attempts were at least four times higher than the general population (Johnston et al., 2009). The extent of mental ill-health and secondary risks indicate that there are a large number of young people who are presenting to these services much later in the course of illness. It is precisely these longer periods of untreated illness that increase the complexity of treatment and risk for ongoing problems at greater costs to the health system (Ghio et al., 2014, Murru et al.,

2015, Marshall et al., 2005, Harris et al., 2005).

Prior work suggests that young people who present with greater impairment and more developed mental health syndromes require more intensive treatments and resources, yet are less likely to recover. In contrast, those with milder symptoms and impairment are more likely to recover after a briefer period of care (Cross et al., 2016). These findings are consistent with those identified in chapter two, whereby young people who presented with mild impairment typically recovered relatively quickly (approx. 6 months), while those with more significant impairment were more likely to experience chronic impairment. The only exception to this was a small subgroup, the youngest and only group under 19 years of age, who entered with significant impairment and were able to functionally recover. Critically though, the time to recovery for this smaller subgroup seems to have been prolonged over an extended period of up

104 to and beyond two years, which may reflect the long-term care needs of those with substantial impairments (discussed further in chapter 6.3). Together, these results reiterate the potential for earlier intervention, both for those with milder symptoms and impairment or for those younger in age, to have a major impact on improving future outcomes.

Improving early intervention and prevention involves addressing key barriers to help-seeking and accessing services among these populations. Younger people continue to be less likely to access mental health services then older adults (Zachrisson et al., 2006, Harris et al., 2015). Thus, addressing individual barriers to help-seeking such as stigma, self-reliance, and poor mental health literacy remains a priority for facilitating access into care (Gulliver et al., 2010,

Clement et al., 2015). The type of contact individuals have with health services can act as critical junctions for either shortening or delaying the pathway into care (Cabassa et al., 2018). So, in addition to these individual factors, it is important to address common barriers at a health system level such as long- wait lists, limited opening hours and high costs to improve early intervention practices (Rickwood et al., 2015a, McCann and Lubman, 2012). The development of youth-friendly services have been effective to engage young people earlier by addressing many of these individual and health system barriers through the use of service design principles that make these service more acceptable and appropriate to young people (Muir et al., 2012, Hilferty et al., 2015, Walker, 2010). Though, given that many young people still experience extended periods of untreated illness and present with substantial impairment, it is clear that further developments in this area are needed to break down these barriers and improve access to care.

Increasingly, online technologies are becoming a major source of information and a tool for people to access mental health care (Ybarra and Eaton, 2005,

105

Griffiths and Christensen, 2007, Burns et al., 2013). The widespread availability and ease of access to these technologies provide a major opportunity to break down some of the major individual and health service barriers to engage young people earlier in the course of care. Initial studies in this area demonstrate the effectiveness of online psychoeducation to increase mental health literacy and reduce stigma to promote help-seeking intentions among young people (Taylor-

Rodgers and Batterham, 2014, Costin et al., 2009), yet these approaches do not always result in increased help-seeking behaviours (Kauer et al., 2014, Han et al., 2017). Some evidence suggests that approaches to addressing stigma and literacy may need to be personalised to specific groups less likely to seek help

(i.e. males, younger people), and specific mental health problems (i.e. suicidal thoughts or anxiety) for better outcomes (Taylor-Rodgers and Batterham, 2014,

Calear et al., 2014, Calear et al., 2017). Additionally, these online technologies offer the opportunity to overcome many practical barriers to accessing care traditional health care by offering brief, online support to address mild concerns and facilitate entry into care through integration with the existing health system (Burns and Birrell, 2014, Burns et al., 2010, Boydell et al., 2014).

Together this work highlights that there is still work to be done to ensure young people are receiving mental health care earlier in the course of these illnesses and prior to the emergence of significant problems. Further research is still needed in this area to determine which strategies are most effective for improving help-seeking behaviours and promoting early intervention. The use of personalised approaches, online technologies and the integration with existing health services seem to be important strategies for breaking down these barriers and improving early intervention practices. If we manage to effectively engage young people earlier then the next challenge is to determine how to provide mental health care that addresses the needs of young people and improve long-term outcomes.

106

6.3. Individual care needs over time

Breaking down these individual-level and health system barriers should be met with mental health care that addresses the needs that young people present with. As described in chapter 1.2., mental health disorders among young people have a much broader impact on the lives of young people and the studies presented in this thesis demonstrate the extent of mental health problems and comorbidities that exist at initial presentation to these early intervention services. Approximately one-third of the sample in chapter four reported symptoms associated with at-risk mental health states (i.e. psychosis and mania symptoms), previous suicidal behaviours and/or alcohol or substance misuse.

The average days out of role in the past month was just over one week and approximately one-third identified as NEET, which is consistent with the substantial impairment at initial presentation identified in chapter two. Since young people are more likely to present with multidimensional needs (Burnett-

Zeigler et al., 2012), health service strategies should be in place to identify and respond to a range of health and social issues with the appropriate type and intensity of intervention (Fleury et al., 2016).

The study in chapter four presents a potential solution that aims to improve timely access to care for those with higher levels of suicidality. This study showed that a technology-enabled protocol that integrates with existing service provisions can be used to provide access to assessment outside the typical operating hours and enables an individual to complete it in their own time, prior to their first face-to-face appointment. The use of this type of ‘pre-visit’ screening provides two important advantages in practice. The first is that it provides detailed results to the clinician that inform the initial face-to-face appointment and reduce the time spent on assessment allowing more time to discuss intervention and treatment planning (Fothergill et al., 2013)(discussed

107 further in chapter 6.5). The second is that it facilitates the identification of individuals who may need to access care sooner, bypassing typical wait times.

While the protocol does not predict an individual’s actual risk of engaging in suicidal behaviours, it does present a personalised solution whereby health services can be informed by data reported in real-time to make an important clinical decision about how to manage individuals and ensure the safety of young people presenting for care.

While the study in chapter four escalated individuals based on short-term risks associated with high suicidality, future applications of this technology could extend to the early identification of other markers of need. Screening for a range of health and social problems may be beneficial to identify and target treatment at issues that are not readily discussed or easily recognised, yet are associated with greater complexity and predict poorer long-term outcomes (Tylee et al.,

2007, Burnett-Zeigler et al., 2012). Specifically, this includes factors identified within this thesis such as, major functional impairment (chapter two), suicidal behaviours (chapter three), and other known markers, such as at-risk mental health states (i.e. psychosis and hypomania) and substantial comorbidity. Of the six trajectories of functioning identified in chapter two, three of them were characterised by significant impairment at entry (i.e. SOFAS scores below 55) and the overwhelming majority of these individuals remained chronically impaired over time. One of these groups deteriorated even further over time, while a second group functionally recovered. This second group was the smallest (4% of total sample), and individuals in this group only managed to recover over a longer-term period of care (approx. over a two-year period).

Together these results illustrate that individuals with substantial functional impairment at entry may require substantial, long-term health care to functionally recover over time.

108

Similarly, the results from chapter three highlight the long-term vulnerability of young people who have a history of suicide attempts. The results from this study suggest that these young people, in particular, may require long-term care needs that go beyond reducing suicidality and include the need for secondary support to reduce long-term burden. These long-term risks remained even after controlling for primary diagnoses which suggests that this vulnerability may be transdiagnostic. Similar to other cohorts of more severe mental health disorders or population-based samples, we identified that a previous suicide attempt was associated with repeated attempts among this transdiagnostic group of young people (Robinson et al., 2009, Chang et al.,

2015). A somewhat unexpected finding was the emergence of bipolar and alcohol or substance use disorders among those with a history of suicide attempts. These results are consistent with the findings of chapter four, whereby alcohol and substance misuse was associated with the presence of any suicidality, and supports the ongoing risk of suicidal thoughts and behaviours associated with substance use among those with depression and anxiety

(Batterham and Christensen, 2012). Thus, young people presenting with suicidal thoughts and behaviours not only require direct intervention that target these behaviours, but it does suggest that this group may require much more assertive or intensive clinical care that address these long-term secondary risks.

The use of assessments to identify predictors of poorer outcomes, or subgroups of individuals who have differential care needs may be important in the delivery of personalised youth mental health care. A higher severity of needs tends to be associated poorer outcomes over time, and so strategies that address the extent of these problems may be able to improve long-term outcomes

(Fleury et al., 2013). A recent systematic review demonstrated that of the few studies that assessed multiple health domains (e.g. mental health, alcohol use, sleep) in primary care, screening facilitated the opportunity to provide targeted

109 interventions and led to better health outcomes (Webb et al., 2016, Tylee et al.,

2007). Though, critical to the effectiveness of these strategies is the context and implementation of these approaches. Population-based studies have demonstrated that screening and feedback do not promote health service use, and may actually have negative effects on help seeking behaviours among people with depression and social anxiety (Batterham et al., 2016, Gilbody et al., 2008). This suggests that for screening and feedback to be effective, these approaches may need to be accompanied by structured or enhanced care to ensure that young people are able to receive the appropriate interventions.

Figure 1. A proposed schematic of personalised youth mental health care that utilises the findings of this thesis and approaches discussed in this chapter.

Chapter four presents an approach that includes a comprehensive assessment at first contact with a service, which immediately produces feedback for the

110 individual and their clinicians. This assessment and feedback is embedded within clinical service supports that are able to respond to the domains that warrant further assessment or intervention and make specific recommendations about what to do next (figure 1). This type of screening assessment when integrated with existing mental health services can become a powerful tool that aides clinical decision making by identifying the mental health needs of young people, and providing the means for directly responding to these needs

(discussed further in chapter 6.5). Chapter four is a good example of the integration of these technologies with mental health services, and the further application outlined in this section illustrates it future potential to transform the delivery of mental health services. Critical to this transformation is an approach that matches the individual to the right type and intensity of care until a specific outcome is achieved and the individuals needs have been addressed.

6.4. Integrated and outcomes-focused care

Equally important to the earlier identification of mental health disorders, is the provision of quality mental health care that effectively targets the underlying problems and improves outcomes. A recent report regarding Australia’s mental health system recommended a fundamental shift away from more costly acute care and hospitals, to improving primary and community mental health services with a focus on multidisciplinary care, early detection and recovery focussed care (National Mental Health Commission, 2014). This type of shift was motivated by the recognition that providing the appropriate interventions, earlier in the course of illness and prior to the onset of major morbidity can lower the long-term health costs to individuals, the health system and society

(Vigo et al., 2016).

Unfortunately, the studies presented in chapters two and three suggest that early identification followed by current standard care provisions do not appear

111 to inevitably result in improved outcomes. This particular cohort represent a population of young people, described as the ‘missing middle’, who have needs that are beyond the capabilities of primary care, yet not severe enough for more specialist and expensive health services (McGorry and Hamilton, 2017). For these individuals, the typical six to ten sessions that enhanced primary care services (i.e. headspace) can provide are probably insufficient to meet the chronic and multidimensional needs of these young people (Hetrick et al., 2017).

Instead, it is clear that the development and evaluation of specific, integrated care packages may be needed to reduce the morbidity and mortality due to early-onset major mental health disorders (Fleury et al., 2014, Fleury and

Mercier, 2002). Inherent in these approaches is outcomes-focused care that move beyond symptom and risk management to an emphasis on improving outcomes and reduce secondary risks over extended periods of care.

The substantial gains in the rate of functional recovery among young people with first-episode psychosis illustrate the potential benefits of early intervention services (Fowler et al., 2009a, Rinaldi et al., 2010). This was largely realised due to a shift in the model of care and service response to readily meet the needs of these young people, so that now psychotic disorders are viewed as amenable to positive functional outcomes (Malla et al., 2016). Further research has demonstrated that targeted social recovery as an adjunct to early intervention, particularly for those with more severe disorders, is more beneficial than early intervention alone for improving social and functional outcomes (Fowler et al., 2009b, Fowler et al., 2018). Other targeted interventions, such as individual placement and support, can also be highly effective for improving educational and employment outcomes among young people with severe mental health disorders (Bond et al., 2016, Bond et al., 2015) and have proven effective for a broader range of mental health disorders

(Killackey et al., 2017). These targeted interventions are particularly important in early intervention settings, since early functional recovery predicts long-term

112 functional recovery better than symptomatic recovery (Killackey, 2015,

Alvarez-Jimenez et al., 2012). This is consistent with the findings of chapter two whereby long-term functional recovery was observed for the subgroup of individuals who functionally improved within the first six months of care.

In addition to these targeted interventions, indirect interventions that address other key predictors of poorer outcomes may prove valuable. The systematic review presented in chapter five provides evidence of the indirect relationship between underlying measures of brain circuits and behaviours and multidimensional outcomes that may be potential treatment targets. These treatments would target specific neurobiological or psychological processes associated with poorer outcomes among these transdiagnostic cohorts such as neurocognition, social avoidance, circadian dysregulation and depressive symptoms. More specifically, this review highlighted the utility of assessing for memory and executive function deficits to identify individuals at risk for poorer long term social and occupational functioning. This is consistent with evidence which suggests that these specific cognitive functions may be critical to the development of common mental health disorders (Lee et al., 2012, Lee et al.,

2014).

Similarly, across both imaging and neuropsychological studies reviewed in chapter five, results indicated that suicidal thoughts and behaviours were typically associated with impaired decision making and conceptual flexibility in major mood and anxiety disorders. This relationship may reflect state or trait processes that involve deficits in higher order cognitive control and impulsivity, commonly associated with suicidality (Gvion et al., 2015). These findings are largely consistent with the results of chapter three whereby the emergence of alcohol or substance use disorders and bipolar disorder, characterised by impulsivity and dysregulated behavioural control, was higher

113 among individuals with a history of suicide attempts. Other studies provide further support for the relationship between these higher order cognitive functions and suicidal thoughts and behaviours (Szanto et al., 2015, van

Heeringen et al., 2017), and even suggest that neurocognition may help explain the transition of suicidal thoughts to behaviours (Saffer and Klonsky, 2018).

These results, along with the other findings discussed in chapter five, highlight the potential use of further assessments to identify these deficits and provide indirect interventions that target these processes for improving outcomes. This is largely in line with the focus of other major initiatives, such as RDOC in the

United States, which aim to identify the underlying neurobiological correlates of brain processes typically evident across major mental health disorders (Insel et al., 2010). This is an important focus for future clinical research since many cognitive, behavioural and pharmacological interventions have demonstrated efficacy for addressing one or more of these underlying factors. Specifically, cognitive and behavioural interventions are associated with substantial long- term benefits in terms of symptomatic and functional recovery among young people at risk for developing severe mental health disorders (McGorry et al.,

2017) and for reducing symptoms of depression and anxiety and improving quality of life (Carpenter et al., 2018, Rochat et al., 2017). These cognitive and behaviour interventions have also demonstrated efficacy for reducing the risk of suicidal thoughts and behaviours (Tarrier et al., 2008, Rudd et al., 2015), while the use of adjunctive psychotherapy that targets cognitive reappraisal is effective for reducing suicidal behaviours and self-harm among young people with bipolar disorder (Inder et al., 2016).

Much like the progress made for early intervention and integrated care for psychosis, similar efforts have been made to apply these principles to more common mental health disorders. The data discussed here suggest that the

114

“missing middle” remain and that further efforts are required to determine whether integrated, multidisciplinary care can improve the long-term outcomes of young people. Many of the necessary assessments and interventions have been developed, yet identifying the right subgroups of individuals and effectively implementing this type of outcomes focused care remains a challenge in practice.

6.5. Technology-enhanced mental health care

Many of the solutions discussed in the preceding sections involve the use of online technologies—largely due to the unavoidable promise and potential of new technologies. As we move further into the 21 st century these technologies are becoming an enormous presence in our lives; from the way we socialise, organise daily life, work, access information and services, and they will continue to revolutionise many other areas of life. The mental health sector, like many others, is undergoing immense change as the technology revolution transforms the way people can access and manage their health (Piwek et al.,

2016). This is evident in the dramatic increase in the number of mobile applications, internet based resources and platforms that target mental health problems (Anthes, 2016). While it remains a major priority to develop and assess effective online interventions, an equally important task is to determine how to integrate these technologies with health services in way that improves the delivery of personalised mental health care (Cross and Hickie, 2017).

The data presented in chapters two and three of this thesis were collected retrospectively via a medical file audit, a method which has some inherent limitations (see chapter 6.6). Putting these limitations aside for a moment, and engaging in simple a thought experiment raises some interesting questions about the potential opportunities for the future of technology-enabled youth mental health care. For example, if this data (i.e. SOFAS score) was routinely

115 collected in “real-time” and presented back to an individual or clinician, would this have had an impact on intervention strategies, or more specifically, would data regarding deterioration in functioning reorient care to target functioning earlier and before major impairment? Evidence from previous research regarding this type of feedback suggests that the answer may be yes, since the use of routine outcome monitoring does have an impact on outcomes. Namely, monitoring an individual’s progress during care can reduce deterioration and improve treatment effects by notifying individuals and clinicians to positive and negative changes to a particular treatment (Boswell et al., 2015, Schmidt et al.,

2006). This facilitates the opportunity to alter the treatment plan accordingly and actively engage young people who may have disengaged or not adhering to treatment. This type of regular personalised feedback about an individual’s progress has been associated with preventing treatment failure and improving outcomes across multiple studies involving young people in outpatient settings

(Shimokawa et al., 2010).

This personalised approach lends itself to prioritising and timing the sequence of the specific interventions proposed in chapter 6.4. By leveraging the advantages of technology, mental health services could offer regular standardised assessment and immediate feedback to individuals and clinicians to guide clinical decision making, at scale. These approaches are particularly suitable for young people with many co-occurring conditions since interventions can be effectively coordinated in a way that addresses the most impairing problems first (Bearman and Weisz, 2015). This also facilities the use of integrated or conceptually unified approaches that incorporate multiple sources of information about causal or maintaining factors to develop interventions that target these processes (Barlow, 2014, Ehrenreich-May and

Bilek, 2012). Chapter two illustrates that while statistical modelling may be able to group similar individual trajectories together, the within-individual variability in outcomes over time emphasises the challenge clinician’s face

116 every day when making clinical decisions. Some patients do deteriorate during treatment, and some require long-term care, so being able to identify an individual’s progress during treatment may facilitate decisions about the effectiveness of current interventions and the timing of new ones (Harmon et al., 2007).

6.6. Limitations of the current work

6.6.1. Clinical cohort

The clinical samples used in chapters two and three represent a smaller proportion of the overall clinical research register of individuals who had opted in to participate in research, and this register itself represents a smaller proportion of the overall sample attending these early intervention services. As such, the samples used in these two studies potentially represent a more severe group of the population attending these services, since young people engaged with these services for relatively minor health, social or functional problems may be less likely to participate in research. Specifically, compared to a large national study, these samples tended to be older in age, have less females and present with lower overall functional impairment. Despite these differences, the rate of change in functioning between the samples in this thesis and these other cohorts were comparable (see chapter two). While the first study (chapter two) only represented approximately 18% of the total research register, this increased to 37% for the second study (chapter three) improving the overall generalisability of these results to the broader cohort.

Another limitation related to the nature of this clinical cohort is the absence of systematic follow-up procedures used to collect the outcome data, and so the availability of follow-up data relied on real-world clinical engagements. This naturalistic method presents some obvious limitations in terms of missing and incomplete data, inconsistent content of data records and variable follow-up

117 periods. This limited the type and comprehensiveness of analyses that were conducted, which may have affected the conclusions about health service models and interventions. Similarly the inclusion criteria for these studies was fairly unrestrictive and open, which means that the heterogeneity of the sample was quite significant. Despite the limitation of huge variability in presentation characteristics and real-world data collection, this type of non-selective recruitment is consistent with other recommendations which suggest that cohorts recruited from a common service setting, who demonstrate variance along the relevant dimensions of interests, are more likely to yield clinically useful insights (Casey et al., 2013). Similarly, much of the data generated from this thesis is being used to drive real-world clinical and research transformations that lend themselves to more systematic analyses and evaluation in the future.

6.6.2. Data collection methodology

A clinical file audit methodology was used to collect data for chapters two and three, which means there are a number of factors that may influence the type and quality of the data collected. Since this methodology relies on clinical notes recorded by the treating clinicians, the consistency of information available in these notes varied by the young person, clinician assessing the individual and entering the information and the time point of entry. This limits the reliability of the information collected since it is unclear whether the absence of clinical information about a particular outcome or parameter means the information was unavailable/missing, not assessed by the clinician or simply not clinically relevant. This increases the potential of false negatives for particular outcomes, since it is possible that some clinical or functional information may have been present but not recorded by a clinician or missed in the clinical notes. This limitation is not so problematic for the outcomes endorsed as ‘positive’ since the inclusion of information about these outcomes in the clinical notes typically

118 means that this was identified or reported by the individual during assessment or a focus of intervention.

Despite these limitations, efforts were made to address these concerns and limit their impact on the results of these studies. Specifically, the clinical file audit was conducted by trained clinician researchers, using established and validated criteria for diagnosis (ie. DSM-5 criteria) and functioning (ie. SOFAS). Other clinical and functional outcomes were limited to reporting the discrete presence or absence of specific outcomes (e.g. suicide attempts) when explicitly stated in the clinical notes to avoid variability in interrater judgements of specific constructs or outcomes. All available clinical notes for an individual case were used to complete each time point entry to ensure the data collected was comprehensive and completely informed by the individual’s clinical history. So, for example, all clinical notes between three months and six months were used to complete the six month entry, likewise all notes between twelve months and two years were used to complete the two year entry. The clinician researcher team responsible for collecting the data consulted regularly with the each other and the lead clinician (IH) about ambiguous cases and how to deal with these consistently. These meetings involved case examples and scenarios that were resolved together to determine the guiding principles for data collection.

6.6.3. Limited treatment and evaluation data

An obvious limitation of this thesis is that limited data was available to draw stronger conclusions about the impact of treatments on long-term outcomes, and the effectiveness of technology-enabled assessment to improve outcomes.

The papers presented in chapters two and three both report on the long-term outcomes of young people attending early intervention health services, however there is little information presented about the type and interventions received and how these might have impacted on the outcomes observed. These

119 studies were used to make general conclusions about the impact of current early intervention service provisions to improve these outcomes based on related studies, however specific conclusions about the effectiveness of particular intervention strategies were not possible. Similarly, although chapter four reports on the use of technology-enabled assessment to identify and respond to suicidality, the lack of follow-up data limits the conclusions that can be made about the effectiveness of this protocol to reduce risks and improves outcomes.

Although the systematic review presented in chapter five does not include data, the evidence synthesised was largely drawn from cross-sectional studies with only a few longitudinal studies identified and included. Due to the limited number of longitudinal studies, the conclusions drawn from this synthesis of the literature largely focus on the cross-sectional relationship between neurobiological measures and functional outcomes with limited insights into the prospective predictive value of these measures.

6.7. Future directions

A key focus for future research should be to continue to collect and evaluate this type of real-world outcome data to inform the development of new public mental health strategies, mental health service reform and new interventions for young people. This type of clinical research will be complicated due to the challenge of recruiting individuals and delivering these complex interventions in the real-world. Despite the challenge, this type of translational research is needed to develop strategies that are effective for individuals attending these services in the real-world. The work of this thesis highlights some intriguing avenues for further research that could test whether some of the solutions proposed here improve long-term outcomes. Namely;

i. Determine whether the use of technology to deliver initial assessment

and screening improves the allocation of care to address an individual’s

presenting needs;

120

ii. Evaluate whether integrated treatment strategies can be effectively

implemented in real-world clinical settings and improve long-term

multidimensional outcomes; iii. Assess the effectiveness of outcome monitoring and feedback to improve

long-term multidimensional outcomes and reduce deterioration.

Increasingly, the use of data is fundamental to practice in many sectors—and mental health care is no different. Future work in this area should focus on continuing to improve the quality of data collected and refining the analytical approaches employed. This thesis highlights the value of quality real-world data about clinical and functional outcomes to report on the practice of current clinical models and the outcomes of service attendees. The availability of such data, in real-time, may prove to be invaluable to service stakeholders to evaluate the effectiveness of mental health service delivery and guide decisions about resource allocation and implementation of strategies that address the needs of a particular service population. The previous section (chapter 6.6) highlights some of the limitations of this dataset that could be improved upon in future research to improve the quality and generalisability of the data to make such key decisions.

In addition to improving data quality, new data analytic methodologies may be needed to resolve some of these more complex real-world research questions.

The nature of data collected in these real-world settings means that some of the problems related to missing data and variable follow up periods are problematic when dealing with traditional frequentist approaches. The limitations of these datasets often lead to predictive models that are specific to discrete populations, or vary based on the variables available to model. This leads to problems related to irreproducibility and limits predictive performance

(Kohn et al., 2018). Thus, there may be a need for alternate approaches, such a

121

Bayesian modelling and machine learning, which leverage data science and analytic expertise to build better predictive models that can be translated to the real world. Future research should aim to employ both methodologies and aim to compare the relative capabilities of different approaches.

The future use of the online assessment and feedback presented in chapter four is to identify critical areas of an individual’s health and wellbeing that may need to be addressed and monitored during their care. The ongoing development of this tool would benefit from employing methodologies that aim to improve the performance of the existing algorithms. Currently, the tool uses static rule-based decision algorithms (based on expert opinion or established cut-offs) to inform the display of assessment results, however future research should examine the use of machine learning methodologies to build better predictive models that enable these results to become personalised to individuals and specific health service populations. Machine learning methodologies are increasingly used in mental health research as they facilitate individual-level prediction of unseen observations, which makes them suitable for the development of clinically useful tools. These methodologies have been used to analyse neuroimaging data to distinguish between Alzheimer’s disease and normal aging (Klöppel et al., 2009), predict individuals with psychosis versus controls (Koutsouleris et al., 2009), and depressed treatment responders from non-responders (Fu et al., 2008). Other studies have applied machine learning to clinical information to develop accurate risk-stratification models to predict depression course (Kessler et al., 2016, Wardenaar et al., 2014) and suicidality (Passos et al., 2016, Kessler et al., 2015b). Though, these semiautomated approaches require rigorous evaluation and validation to determine whether they are appropriate and effective for clinical use.

122

6.8. Concluding remarks

The youth mental health landscape has undergone dramatic transitions over the past decade, in terms of awareness, access and resource. These advancements are a crucial step in the right direction for addressing mental health disorders and their associated burden in young people. The next important step in this revolution is to ensure that young people who access care, receive quality, personalised mental health care that addresses their multitude of needs early in the course of life so that they can lead fulfilled and contributing lives in adulthood. With major mental health reform on the agenda for Australia, the insights gained from this thesis are essential to understanding the nature and scope of current challenges regarding the delivery of mental health care to young people.

123

References

Aihw 2017. Australia’s welfare 2017. Australia’s welfare. Canberra: Australian Institute of Health and Welfare. Alvarez-Jimenez, M., Gleeson, J., Henry, L., Harrigan, S., Harris, M., Killackey, E., Bendall, S., Amminger, G., Yung, A. & Herrman, H. 2012. Road to full recovery: longitudinal relationship between symptomatic remission and psychosocial recovery in first-episode psychosis over 7.5 years. Psychological medicine, 42, 595-606. Angold, A., Costello, E. J. & Erkanli, A. 1999. Comorbidity. Journal of child psychology and psychiatry, 40, 57-87. Angst, J., Gamma, A., Neuenschwander, M., Ajdacic-Gross, V., Eich, D., Rössler, W. & Merikangas, K. R. 2005. Prevalence of mental disorders in the Zurich Cohort Study: a twenty year prospective study. Epidemiology and Psychiatric Sciences, 14, 68-76. Anthes, E. 2016. Pocket psychiatry: mobile mental-health apps have exploded onto the market, but few have been thoroughly tested. Nature, 532, 20- 24. Barlow, D. H. 2014. Clinical handbook of psychological disorders: A step-by-step treatment manual, Guilford publications. Batterham, P. J., Calear, A. L., Sunderland, M., Carragher, N. & Brewer, J. L. 2016. Online screening and feedback to increase help-seeking for mental health problems: population-based randomised controlled trial. British Journal of Psychiatry Open, 2, 67-73. Batterham, P. J. & Christensen, H. 2012. Longitudinal risk profiling for suicidal thoughts and behaviours in a community cohort using decision trees. Journal of affective disorders, 142, 306-314. Bearman, S. K. & Weisz, J. R. 2015. Comprehensive treatments for youth comorbidity–evidence‐guided approaches to a complicated problem. Child and adolescent mental health, 20, 131-141. Belli, F., Testori, A., Rivoltini, L., Maio, M., Andreola, G., Sertoli, M. R., Gallino, G., Piris, A., Cattelan, A. & Lazzari, I. 2002. Vaccination of metastatic melanoma patients with autologous tumor-derived heat shock protein gp96-peptide complexes: clinical and immunologic findings. Journal of Clinical Oncology, 20, 4169-4180. Bond, G., Drake, R. & Luciano, A. 2015. Employment and educational outcomes in early intervention programmes for early psychosis: a systematic review. Epidemiology and psychiatric sciences, 24, 446-457. Bond, G., Drake, R. E. & Campbell, K. 2016. Effectiveness of individual placement and support supported employment for young adults. Early intervention in psychiatry, 10, 300-307.

124

Boswell, J. F., Kraus, D. R., Miller, S. D. & Lambert, M. J. 2015. Implementing routine outcome monitoring in clinical practice: Benefits, challenges, and solutions. Psychotherapy research, 25, 6-19. Boydell, K. M., Hodgins, M., Pignatiello, A., Teshima, J., Edwards, H. & Willis, D. 2014. Using technology to deliver mental health services to children and youth: a scoping review. Journal of the Canadian Academy of Child and Adolescent Psychiatry, 23, 87. Brière, F. N., Rohde, P., Seeley, J. R., Klein, D. & Lewinsohn, P. M. 2015. Adolescent suicide attempts and adult adjustment. Depression and anxiety, 32, 270-276. Bunney, W., Kleinman, A., Pellmar, T. & Goldsmith, S. 2002. Reducing suicide: A national imperative, National Academies Press. Burnett-Zeigler, I., Walton, M. A., Ilgen, M., Barry, K. L., Chermack, S. T., Zucker, R. A., Zimmerman, M. A., Booth, B. M. & Blow, F. C. 2012. Prevalence and correlates of mental health problems and treatment among adolescents seen in primary care. Journal of Adolescent Health, 50, 559-564. Burns, J. & Birrell, E. 2014. Enhancing early engagement with mental health services by young people. Psychology research and behavior management, 7, 303. Burns, J., Davenport, T., Christensen, H., Luscombe, G., Mendoza, J., Bresnan, A., Blanchard, M. & Hickie, I. 2013. Game on: Exploring the impact of technologies on young men’s mental health and wellbeing. Findings from the first Young and Well National Survey. Melbourne: Young and Well Cooperative Research Centre, 1-76. Burns, J., Davenport, T., Durkin, L. A., Luscombe, G. M. & Hickie, I. B. 2010. The internet as a setting for mental health service utilisation by young people. Medical Journal of Australia, 192, S22. Cabassa, L. J., Piscitelli, S., Haselden, M., Lee, R. J., Essock, S. M. & Dixon, L. B. 2018. Understanding Pathways to Care of Individuals Entering a Specialized Early Intervention Service for First-Episode Psychosis. Psychiatric Services, appi. ps. 201700018. Calear, A. L., Banfield, M., Batterham, P. J., Morse, A. R., Forbes, O., Carron- Arthur, B. & Fisk, M. 2017. Silence is deadly: a cluster-randomised controlled trial of a mental health help-seeking intervention for young men. BMC public health, 17, 834. Calear, A. L., Batterham, P. J. & Christensen, H. 2014. Predictors of help-seeking for suicidal ideation in the community: risks and opportunities for public suicide prevention campaigns. Psychiatry Research, 219, 525-530. Carpenter, J. K., Andrews, L. A., Witcraft, S. M., Powers, M. B., Smits, J. A. & Hofmann, S. G. 2018. Cognitive behavioral therapy for anxiety and

125

related disorders: A meta‐analysis of randomized placebo‐controlled trials. Depression and anxiety. Casey, B., Craddock, N., Cuthbert, B. N., Hyman, S. E., Lee, F. S. & Ressler, K. J. 2013. DSM-5 and RDoC: progress in psychiatry research? Nature Reviews Neuroscience, 14, 810. Chang, W. C., Chen, E. S., Hui, C. L., Chan, S. K., Lee, E. H. M. & Chen, E. Y. 2015. Prevalence and risk factors for suicidal behavior in young people presenting with first-episode psychosis in Hong Kong: a 3-year follow- up study. Social psychiatry and psychiatric epidemiology, 50, 219-226. Clement, S., Schauman, O., Graham, T., Maggioni, F., Evans-Lacko, S., Bezborodovs, N., Morgan, C., Rüsch, N., Brown, J. & Thornicroft, G. 2015. What is the impact of mental health-related stigma on help- seeking? A systematic review of quantitative and qualitative studies. Psychological medicine, 45, 11-27. Commission, N. M. H. 2014. Contributing Lives, Thriving Communities: Report of the National Review of Mental Health Programmes and Services. Sydney, Australia. Compton, S. N., March, J. S., Brent, D., Albano, A. M., Weersing, V. R. & Curry, J. 2004. Cognitive-behavioral psychotherapy for anxiety and depressive disorders in children and adolescents: an evidence-based medicine review. Journal of the American Academy of Child & Adolescent Psychiatry, 43, 930-959. Compton, W. M., Thomas, Y. F., Stinson, F. S. & Grant, B. F. 2007. Prevalence, correlates, disability, and comorbidity of DSM-IV drug abuse and dependence in the United States: results from the national epidemiologic survey on alcohol and related conditions. Archives of general psychiatry, 64, 566-576. Conway, K. P., Swendsen, J., Husky, M. M., He, J.-P. & Merikangas, K. R. 2016. Association of lifetime mental disorders and subsequent alcohol and illicit drug use: results from the National Comorbidity Survey– Adolescent Supplement. Journal of the American Academy of Child & Adolescent Psychiatry, 55, 280-288. Copeland, W. E., Shanahan, L., Costello, E. J. & Angold, A. 2009. Childhood and adolescent psychiatric disorders as predictors of young adult disorders. Archives of general psychiatry, 66, 764-772. Copeland, W. E., Shanahan, L., Costello, E. J. & Angold, A. 2011. Cumulative prevalence of psychiatric disorders by young adulthood: a prospective cohort analysis from the Great Smoky Mountains Study. Journal of the American Academy of Child & Adolescent Psychiatry, 50, 252-261. Copeland, W. E., Wolke, D., Shanahan, L. & Costello, E. J. 2015. Adult functional outcomes of common childhood psychiatric problems: a prospective, longitudinal study. JAMA psychiatry, 72, 892-899.

126

Costello, E. J., Egger, H. & Angold, A. 2005. 10-year research update review: the epidemiology of child and adolescent psychiatric disorders: I. Methods and public health burden. Journal of the American Academy of Child & Adolescent Psychiatry, 44, 972-986. Costello, E. J., Foley, D. L. & Angold, A. 2006. 10-year research update review: the epidemiology of child and adolescent psychiatric disorders: II. Developmental epidemiology. Journal of the American Academy of Child & Adolescent Psychiatry, 45, 8-25. Costin, D. L., Mackinnon, A. J., Griffiths, K. M., Batterham, P. J., Bennett, A. J., Bennett, K. & Christensen, H. 2009. Health e-cards as a means of encouraging help seeking for depression among young adults: randomized controlled trial. Journal of medical Internet research, 11. Cox, B. J., Swinson, R. P., Morrison, B. & Lee, P. S. 1993. Clomipramine, flouxetine, and behavior therapy in the treatment of obsessive- compulsive disorder: a meta-analysis. Journal of Behavior Therapy and Experimental Psychiatry, 24, 149-153. Crosby, A., Gfroerer, J., Han, B., Ortega, L. & Parks, S. E. 2011. Suicidal thoughts and behaviors among adults aged> ̲18 Years--United States, 2008-2009. Cross, S. P., Hermens, D. F. & Hickie, I. B. 2016. Treatment patterns and short- term outcomes in an early intervention youth mental health service. Early Intervention in Psychiatry, 10, 88-97. Cross, S. P. & Hickie, I. 2017. Transdiagnostic stepped care in mental health. Public health research & practice, 27. Cross, S. P., Scott, J. L., Hermens, D. F. & Hickie, I. B. 2018. Variability in Clinical Outcomes for Youths Treated for Subthreshold Severe Mental Disorders at an Early Intervention Service. Psychiatric services (Washington, D.C.), appips201700046-appips201700046. Cuthbert, B. N. 2014. The RDoC framework: facilitating transition from ICD/DSM to dimensional approaches that integrate neuroscience and psychopathology. World Psychiatry, 13, 28-35. Dawson, D. A., Li, T.-K., Chou, S. P. & Grant, B. F. 2008. Transitions in and out of alcohol use disorders: their associations with conditional changes in quality of life over a 3-year follow-up interval. Alcohol & Alcoholism, 44, 84-92. Ehrenreich-May, J. & Bilek, E. L. 2012. The development of a transdiagnostic, cognitive behavioral group intervention for childhood anxiety disorders and co-occurring depression symptoms. Cognitive and Behavioral Practice, 19, 41-55. Fergusson, D. M., Horwood, L. J., Ridder, E. M. & Beautrais, A. L. 2005. Suicidal behaviour in adolescence and subsequent mental health outcomes in young adulthood. Psychological Medicine, 35, 983-993.

127

Fernandes, B. S., Williams, L. M., Steiner, J., Leboyer, M., Carvalho, A. F. & Berk, M. 2017. The new field of ‘precision psychiatry’. BMC medicine, 15, 80. Fichter, M., Quadflieg, N., Fischer, U. C. & Kohlboeck, G. 2010. Twenty‐five‐ year course and outcome in anxiety and depression in the Upper Bavarian Longitudinal Community Study. Acta Psychiatrica Scandinavica, 122, 75-85. Fleury, M.-J., Bamvita, J.-M., Grenier, G., Schmitz, N., Piat, M. & Tremblay, J. 2016. Adequacy of help received by individuals with severe mental disorders after a major healthcare reform in Quebec: predictors and changes at 5-year follow-up. Administration and Policy in Mental Health and Mental Health Services Research, 43, 799-812. Fleury, M.-J., Grenier, G., Bamvita, J.-M., Piat, M. & Tremblay, J. 2014. Adequacy of help received among individuals with severe mental disorders. Administration and Policy in Mental Health and Mental Health Services Research, 41, 302-316. Fleury, M.-J., Grenier, G., Bamvita, J.-M., Tremblay, J., Schmitz, N. & Caron, J. 2013. Predictors of quality of life in a longitudinal study of users with severe mental disorders. Health and quality of life outcomes, 11, 92. Fleury, M.-J. & Mercier, C. 2002. Integrated local networks as a model for organizing mental health services. Administration and Policy in Mental Health and Mental Health Services Research, 30, 55-73. Fothergill, K. E., Gadomski, A., Solomon, B. S., Olson, A. L., Gaffney, C. A. & Wissow, L. S. 2013. Assessing the impact of a web-based comprehensive somatic and mental health screening tool in pediatric primary care. Academic pediatrics, 13, 340-347. Fowler, D., Hodgekins, J., French, P., Marshall, M., Freemantle, N., Mccrone, P., Everard, L., Lavis, A., Jones, P. B., Amos, T., Singh, S., Sharma, V. & Birchwood, M. 2018. Social recovery therapy in combination with early intervention services for enhancement of social recovery in patients with first-episode psychosis (SUPEREDEN3): a single-blind, randomised controlled trial. Lancet Psychiatry, 5, 41-50. Fowler, D., Hodgekins, J., Howells, L., Millward, M., Ivins, A., Taylor, G., Hackmann, C., Hill, K., Bishop, N. & Macmillan, I. 2009a. Can targeted early intervention improve functional recovery in psychosis? A historical control evaluation of the effectiveness of different models of early intervention service provision in Norfolk 1998–2007. Early Intervention in Psychiatry, 3, 282-288. Fowler, D., Hodgekins, J., Painter, M., Reilly, T., Crane, C., Macmillan, I., Mugford, M., Croudace, T. & Jones, P. 2009b. Cognitive behaviour therapy for improving social recovery in psychosis: a report from the ISREP MRC Trial Platform study (Improving Social Recovery in Early Psychosis). Psychological medicine, 39, 1627-1636.

128

Fu, C. H., Mourao-Miranda, J., Costafreda, S. G., Khanna, A., Marquand, A. F., Williams, S. C. & Brammer, M. J. 2008. Pattern classification of sad facial processing: toward the development of neurobiological markers in depression. Biological psychiatry, 63, 656-662. Garland, A. F., Hurlburt, M. S. & Hawley, K. M. 2006. Examining psychotherapy processes in a services research context. Clinical Psychology: Science and Practice, 13, 30-46. Ghio, L., Gotelli, S., Marcenaro, M., Amore, M. & Natta, W. 2014. Duration of untreated illness and outcomes in unipolar depression: a systematic review and meta-analysis. Journal of affective disorders, 152, 45-51. Gibb, S. J., Fergusson, D. M. & Horwood, L. J. 2010. Burden of psychiatric disorder in young adulthood and life outcomes at age 30. The British Journal of Psychiatry, 197, 122-127. Gilbody, S., Sheldon, T. & House, A. 2008. Screening and case-finding instruments for depression: a meta-analysis. Canadian Medical Association Journal, 178, 997-1003. Goldman-Mellor, S. J., Caspi, A., Harrington, H., Hogan, S., Nada-Raja, S., Poulton, R. & Moffitt, T. E. 2014. Suicide attempt in young people: a signal for long-term health care and social needs. JAMA psychiatry, 71, 119-127. Goodman, A., Joyce, R. & Smith, J. P. 2011. The long shadow cast by childhood physical and mental problems on adult life. Proceedings of the National Academy of Sciences, 108, 6032-6037. Gore, F. M., Bloem, P. J., Patton, G. C., Ferguson, J., Joseph, V., Coffey, C., Sawyer, S. M. & Mathers, C. D. 2011. Global burden of disease in young people aged 10–24 years: a systematic analysis. The Lancet, 377, 2093- 2102. Griffiths, K. M. & Christensen, H. 2007. Internet‐based mental health programs: A powerful tool in the rural medical kit. Australian Journal of Rural Health, 15, 81-87. Gross, C. & Hen, R. 2004. The developmental origins of anxiety. Nature Reviews Neuroscience, 5, 545-552. Gulliver, A., Griffiths, K. M. & Christensen, H. 2010. Perceived barriers and facilitators to mental health help-seeking in young people: a systematic review. BMC psychiatry, 10, 113. Gvion, Y., Levi-Belz, Y., Hadlaczky, G. & Apter, A. 2015. On the role of impulsivity and decision-making in suicidal behavior. World journal of psychiatry, 5, 255. Hamilton, B. A., Naismith, S. L., Scott, E. M., Purcell, S. & Hickie, I. B. 2011. Disability is already pronounced in young people with early stages of affective disorders: Data from an early intervention service. Journal of Affective Disorders, 131, 84-91.

129

Han, J., Batterham, P. J., Calear, A. L., Wu, Y., Xue, J. & Van Spijker, B. A. 2017. Development and pilot evaluation of an online psychoeducational program for suicide prevention among university students: A randomised controlled trial. Internet Interventions. Hansell, N. K., Wright, M. J., Medland, S. E., Davenport, T. A., Wray, N. R., Martin, N. G. & Hickie, I. B. 2012. Genetic co-morbidity between neuroticism, anxiety/depression and somatic distress in a population sample of adolescent and young adult twins. Psychological Medicine, 42, 1249-1260. Harmon, S. C., Lambert, M. J., Smart, D. M., Hawkins, E., Nielsen, S. L., Slade, K. & Lutz, W. 2007. Enhancing outcome for potential treatment failures: Therapist–client feedback and clinical support tools. Psychotherapy research, 17, 379-392. Harris, E. C. & Barraclough, B. 1997. Suicide as an outcome for mental disorders: a meta-analysis. The British Journal of Psychiatry, 170, 205-228. Harris, M. G., Henry, L. P., Harrigan, S. M., Purcell, R., Schwartz, O. S., Farrelly, S. E., Prosser, A. L., Jackson, H. J. & Mcgorry, P. D. 2005. The relationship between duration of untreated psychosis and outcome: an eight-year prospective study. Schizophrenia research, 79, 85-93. Harris, M. G., Hobbs, M. J., Burgess, P. M., Pirkis, J. E., Diminic, S., Siskind, D. J., Andrews, G. & Whiteford, H. A. 2015. Frequency and quality of mental health treatment for affective and anxiety disorders among Australian adults. The Medical Journal of Australia, 202, 185-189. Harrison, G., Hopper, K., Craig, T., Laska, E., Siegel, C., Wanderling, J., Dube, K. C., Ganev, K., Giel, R., An Der Heiden, W., Holmberg, S. K., Janca, A., Lee, P. W., Leon, C. A., Malhotra, S., Marsella, A. J., Nakane, Y., Sartorius, N., Shen, Y., Skoda, C., Thara, R., Tsirkin, S. J., Varma, V. K., Walsh, D. & Wiersma, D. 2001. Recovery from psychotic illness: a 15- and 25-year international follow-up study. Br J Psychiatry, 178, 506-17. Heim, C., Newport, D. J., Mletzko, T., Miller, A. H. & Nemeroff, C. B. 2008. The link between childhood trauma and depression: insights from HPA axis studies in humans. Psychoneuroendocrinology, 33, 693-710. Henry, L. P., Amminger, G. P., Harris, M. G., Yuen, H. P., Harrigan, S. M., Prosser, A. L., Schwartz, O. S., Farrelly, S. E., Herrman, H., Jackson, H. J. & Mcgorry, P. D. 2010. The EPPIC follow-up study of first-episode psychosis: longer-term clinical and functional outcome 7 years after index admission. J Clin Psychiatry, 71, 716-28. Hetrick, S. E., Bailey, A. P., Smith, K. E., Malla, A., Mathias, S., Singh, S. P., O'reilly, A., Verma, S. K., Benoit, L. & Fleming, T. M. 2017. Integrated (one-stop shop) youth health care: best available evidence and future directions. The Medical journal of Australia, 207, S5-S18.

130

Hickie, I. 2011. Youth mental health: we know where we are and we can now say where we need to go next. Early Intervention in Psychiatry, 5, 63-69. Hickie, I., Scott, E. M., Hermens, D. F., Naismith, S. L., Guastella, A. J., Kaur, M., Sidis, A., Whitwell, B., Glozier, N., Davenport, T., Pantelis, C., Wood, S. J. & Mcgorry, P. D. 2013a. Applying clinical staging to young people who present for mental health care. Early intervention in psychiatry, 7, 31- 43. Hickie, I., Scott, J., Hermens, D. F., Scott, E. M., Naismith, S. L., Guastella, A. J., Glozier, N. & Mcgorry, P. D. 2013b. Clinical classification in mental health at the cross-roads: which direction next? BMC medicine, 11, 125. Hickie, I., Scott, J., Hermens, D. F., Scott, E. M., Naismith, S. L., Guastella, A. J., Glozier, N. & Mcgorry, P. D. 2013c. Clinical classification in mental health at the cross-roads: which direction next? Bmc Medicine, 11. Hickie, I. B., Scott, J., Hermens, D. F., Scott, E. M., Naismith, S. L., Guastella, A. J., Glozier, N. & Mcgorry, P. D. 2013d. Clinical classification in mental health at the cross-roads: Which direction next? BMC Medicine, 11. Hilferty, F., Cassells, R., Muir, K., Duncan, A., Christensen, D., Mitrou, F., Gao, G., Mavisakalyan, A., Hafekost, K. & Tarverdimamaghani, Y. 2015. Is headspace making a difference to young people’s lives? Final report of the independent evaluation of the headspace program. Hoertel, N., Franco, S., Wall, M., Oquendo, M., Kerridge, B., Limosin, F. & Blanco, C. 2015. Mental disorders and risk of suicide attempt: a national prospective study. Molecular psychiatry, 20, 718-726. Inder, M. L., Crowe, M. T., Luty, S. E., Carter, J. D., Moor, S., Frampton, C. M. & Joyce, P. R. 2016. Prospective rates of suicide attempts and nonsuicidal self-injury by young people with bipolar disorder participating in a psychotherapy study. Australian & New Zealand Journal of Psychiatry, 50, 167-173. Insel, T. 2007. The arrival of preemptive psychiatry. Early Intervention in Psychiatry, 1, 5-6. Insel, T. 2009. Translating scientific opportunity into public health impact: a strategic plan for research on mental illness. Archives of general psychiatry, 66, 128-133. Insel, T., Cuthbert, B., Garvey, M., Heinssen, R., Pine, D. S., Quinn, K., Sanislow, C. & Wang, P. 2010. Research domain criteria (RDoC): toward a new classification framework for research on mental disorders. Am J Psychiatry., 167, 748-51. Insel, T. & Fenton, W. 2005. Psychiatric epidemiology: it’s not just about counting anymore. Archives of general psychiatry, 62, 590-592. Iyer, S., Boksa, P., Lal, S., Shah, J., Marandola, G., Jordan, G., Doyle, M., Joober, R. & Malla, A. 2015. Transforming youth mental health: a Canadian perspective. Irish Journal of Psychological Medicine, 32, 51-60.

131

Jakubovski, E. & Bloch, M. H. 2014. Prognostic subgroups for citalopram response in the STAR* D trial. The Journal of clinical psychiatry, 75, 738. Johnston, A. K., Pirkis, J. E. & Burgess, P. M. 2009. Suicidal thoughts and behaviours among Australian adults: findings from the 2007 National Survey of Mental Health and Wellbeing. Australian & New Zealand Journal of Psychiatry, 43, 635-643. Jones, P. B. 2013. Adult mental health disorders and their age at onset. The British Journal of Psychiatry, 202, s5-s10. Jorm, A. F., Patten, S. B., Brugha, T. S. & Mojtabai, R. 2017. Has increased provision of treatment reduced the prevalence of common mental disorders? Review of the evidence from four countries. World Psychiatry, 16, 90-99. Kapur, S., Phillips, A. & Insel, T. 2012. Why has it taken so long for biological psychiatry to develop clinical tests and what to do about it&quest. Molecular psychiatry, 17, 1174-1179. Kauer, S. D., Mangan, C. & Sanci, L. 2014. Do online mental health services improve help-seeking for young people? A systematic review. Journal of Medical Internet Research, 16. Kelleher, I., Keeley, H., Corcoran, P., Lynch, F., Fitzpatrick, C., Devlin, N., Molloy, C., Roddy, S., Clarke, M. C., Harley, M., Arseneault, L., Wasserman, C., Carli, V., Sarchiapone, M., Hoven, C., Wasserman, D. & Cannon, M. 2012. Clinicopathological significance of psychotic experiences in non-psychotic young people: evidence from four population-based studies. The British Journal of Psychiatry, 201, 26-32. Kendler, K. S., Thornton, L. M. & Gardner, C. O. 2001. Genetic risk, number of previous depressive episodes, and stressful life events in predicting onset of major depression. American Journal of Psychiatry, 158, 582-586. Kessler, R. C., Angermeyer, M., Anthony, J. C., De Graaf, R., Demyttenaere, K., Gasquet, I., De Girolamo, G., Gluzman, S., Gureje, O. & Haro, J. M. 2007. Lifetime prevalence and age-of-onset distributions of mental disorders in the World Health Organization's World Mental Health Survey Initiative. World psychiatry, 6, 168. Kessler, R. C., Avenevoli, S., Costello, E. J., Georgiades, K., Green, J. G., Gruber, M. J., He, J.-P., Koretz, D., Mclaughlin, K. A. & Petukhova, M. 2012. Prevalence, persistence, and sociodemographic correlates of DSM-IV disorders in the National Comorbidity Survey Replication Adolescent Supplement. Archives of general psychiatry, 69, 372-380. Kessler, R. C., Berglund, P., Demler, O., Jin, R., Koretz, D., Merikangas, K. R., Rush, A. J., Walters, E. E. & Wang, P. S. 2003. The epidemiology of major depressive disorder: results from the National Comorbidity Survey Replication (NCS-R). Jama, 289, 3095-3105.

132

Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R. & Walters, E. E. 2005. Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the National Comorbidity Survey Replication. Archives of general psychiatry, 62, 593-602. Kessler, R. C., Sampson, N. A., Berglund, P., Gruber, M., Al-Hamzawi, A., Andrade, L., Bunting, B., Demyttenaere, K., Florescu, S. & De Girolamo, G. 2015a. Anxious and non-anxious major depressive disorder in the World Health Organization World Mental Health Surveys. Epidemiology and psychiatric sciences, 24, 210-226. Kessler, R. C., Van Loo, H. M., Wardenaar, K. J., Bossarte, R. M., Brenner, L. A., Cai, T., Ebert, D. D., Hwang, I., Li, J., De Jonge, P., Nierenberg, A. A., Petukhova, M. V., Rosellini, A. J., Sampson, N. A., Schoevers, R. A., Wilcox, M. A. & Zaslavsky, A. M. 2016. Testing a machine-learning algorithm to predict the persistence and severity of major depressive disorder from baseline self-reports. Mol Psychiatry. Kessler, R. C., Warner, C. H., Ivany, C., Petukhova, M. V., Rose, S., Bromet, E. J., Brown, M., Cai, T., Colpe, L. J. & Cox, K. L. 2015b. Predicting suicides after psychiatric hospitalization in US Army soldiers: the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS). JAMA psychiatry, 72, 49-57. Killackey, E. 2015. Resignation not accepted: employment, education and training in early intervention, past, present and future. Early intervention in psychiatry, 9, 429-432. Killackey, E., Allott, K., Woodhead, G., Connor, S., Dragon, S. & Ring, J. 2017. Individual placement and support, supported education in young people with mental illness: an exploratory feasibility study. Early intervention in psychiatry, 11, 526-531. Kim-Cohen J, A., C., E., M. T., H., H., J., M. B. & R., P. 2003. Prior juvenile diagnoses in adults with mental disorder: Developmental follow-back of a prospective-longitudinal cohort. Archives of General Psychiatry, 60, 709- 717. Klöppel, S., Chu, C., Tan, G., Draganski, B., Johnson, H., Paulsen, J., Kienzle, W., Tabrizi, S. J., Ashburner, J. & Frackowiak, R. S. 2009. Automatic detection of preclinical neurodegeneration Presymptomatic Huntington disease. Neurology, 72, 426-431. Kohn, D., Glozier, N., Hickie, I. B., Durrant-Whyte, H. & Cripps, S. 2018. Irreproducibility; Nothing is More Predictable. arXiv preprint arXiv:1803.04481. Kohn, R., Saxena, S., Levav, I. & Saraceno, B. 2004. The treatment gap in mental health care. Bulletin of the World health Organization, 82, 858-866. Koutsouleris, N., Meisenzahl, E. M., Davatzikos, C., Bottlender, R., Frodl, T., Scheuerecker, J., Schmitt, G., Zetzsche, T., Decker, P. & Reiser, M. 2009.

133

Use of neuroanatomical pattern classification to identify subjects in at- risk mental states of psychosis and predict disease transition. Archives of general psychiatry, 66, 700-712. Kozak, M. J. & Cuthbert, B. N. 2016. The NIMH Research Domain Criteria Initiative: Background, issues, and pragmatics. Psychophysiology, 53, 286- 97. Lambert, M. J. 2013. Outcome in psychotherapy: the past and important advances. Lee, R. S., Hermens, D. F., Porter, M. A. & Redoblado-Hodge, M. A. 2012. A meta-analysis of cognitive deficits in first-episode major depressive disorder. Journal of affective disorders, 140, 113-124. Lee, R. S., Hermens, D. F., Scott, J., Redoblado-Hodge, M. A., Naismith, S. L., Lagopoulos, J., Griffiths, K. R., Porter, M. A. & Hickie, I. B. 2014. A meta- analysis of neuropsychological functioning in first-episode bipolar disorders. Journal of psychiatric research, 57, 1-11. Malla, A., Iyer, S., Mcgorry, P., Cannon, M., Coughlan, H., Singh, S., Jones, P. & Joober, R. 2016. From early intervention in psychosis to youth mental health reform: a review of the evolution and transformation of mental health services for young people. Social psychiatry and psychiatric epidemiology, 51, 319-326. Marshall, M., Lewis, S., Lockwood, A., Drake, R., Jones, P. & Croudace, T. 2005. Association between duration of untreated psychosis and outcome in cohorts of first-episode patients: a systematic review. Archives of general psychiatry, 62, 975-983. May, A. M., Klonsky, E. D. & Klein, D. N. 2012. Predicting future suicide attempts among depressed suicide ideators: a 10-year longitudinal study. Journal of psychiatric research, 46, 946-952. Mccann, T. V. & Lubman, D. I. 2012. Young people with depression and their experience accessing an enhanced primary care service for youth with emerging mental health problems: a qualitative study. BMC psychiatry, 12, 96. Mcdermut, W., Miller, I. W. & Brown, R. A. 2001. The efficacy of group psychotherapy for depression: A meta‐analysis and review of the empirical research. Clinical Psychology: Science and Practice, 8, 98-116. Mcdonagh, J. E. & Viner, R. M. 2006. Lost in transition? Between paediatric and adult services. British Medical Journal Publishing Group. Mcgorry, P., Bates, T. & Birchwood, M. 2013. Designing youth mental health services for the 21st century: examples from Australia, Ireland and the UK. The British Journal of Psychiatry, 202, s30-s35. Mcgorry, P., Goldstone, S. D., Parker, A. G., Rickwood, D. J. & Hickie, I. B. 2014. Cultures for mental health care of young people: an Australian blueprint for reform. Lancet Psychiatry, 1, 559-568.

134

Mcgorry, P. & Hamilton, M. 2017. Broken promises and missing steps in mental health reform. The Medical Journal of Australia, 206, 487-489. Mcgorry, P., Nelson, B., Markulev, C., Yuen, H. P., Schäfer, M. R., Mossaheb, N., Schlögelhofer, M., Smesny, S., Hickie, I. B. & Berger, G. E. 2017. Effect of ω-3 polyunsaturated fatty acids in young people at ultrahigh risk for psychotic disorders: the NEURAPRO randomized clinical trial. JAMA psychiatry, 74, 19-27. Mcgorry, P., Tanti, C., Stokes, R., Hickie, I. B., Carnell, K., Littlefield, L. K. & Moran, J. 2007. headspace: Australia's National Youth Mental Health Foundation-where young minds come first. Medical Journal of Australia, 187, S68. Merikangas, K., Akiskal, H. S., Angst, J., Greenberg, P. E., Hirschfeld, R. M., Petukhova, M. & Kessler, R. C. 2007. Lifetime and 12-month prevalence of bipolar spectrum disorder in the National Comorbidity Survey replication. Archives of general psychiatry, 64, 543-552. Merikangas, K., Cui, L., Kattan, G., Carlson, G. A., Youngstrom, E. A. & Angst, J. 2012. Mania with and without depression in a community sample of U.S. adolescents. Arch Gen Psychiatry., 69, 943-51. Merikangas, K., He, J.-P., Burstein, M., Swanson, S. A., Avenevoli, S., Cui, L., Benjet, C., Georgiades, K. & Swendsen, J. 2010a. Lifetime prevalence of mental disorders in U.S. adolescents: Results from the national comorbidity survey replication adolescent supplement (NCS-A). J Am Acad Child Adolesc Psychiatry, 49, 980-989. Merikangas, K., He, J.-P., Burstein, M., Swanson, S. A., Avenevoli, S., Cui, L., Benjet, C., Georgiades, K. & Swendsen, J. 2010b. Lifetime prevalence of mental disorders in US adolescents: results from the National Comorbidity Survey Replication–Adolescent Supplement (NCS-A). Journal of the American Academy of Child & Adolescent Psychiatry, 49, 980- 989. Merikangas, K., Herrell, R., Swendsen, J., Rossler, W., Ajdacic-Gross, V. & Angst, J. 2008. Specificity of bipolar spectrum conditions in the comorbidity of mood and substance use disorders: results from the Zurich cohort study. Archives of General Psychiatry, 65, 47-52. Merikangas, K., Nakamura, E. F. & Kessler, R. C. 2009. Epidemiology of mental disorders in children and adolescents. Dialogues in clinical neuroscience, 11, 7. Mezzich, J. E., Snaedal, J., Van Weel, C., Botbol, M. & Salloum, I. 2011. Introduction to person‐centred medicine: from concepts to practice. Journal of evaluation in clinical practice, 17, 330-332. Moussavi, S., Chatterji, S., Verdes, E., Tandon, A., Patel, V. & Ustun, B. 2007. Depression, chronic diseases, and decrements in health: results from the World Health Surveys. The Lancet, 370, 851-858.

135

Muir, K., Mcdermott, S., Gendera, S., Flaxman, S., Patulny, R., Sitek, T., Abello, D., Oprea, I. & Katz, I. 2009. Independent evaluation of headspace: the National Youth Mental Health Foundation: interim evaluation report. Sydney: Social Policy Reseach Centre UNSW. Muir, K., Powell, A. & Mcdermott, S. 2012. ‘They don’t treat you like a virus’: youth‐friendly lessons from the Australian National Youth Mental Health Foundation. Health & social care in the community, 20, 181-189. Murray, C. J., Lopez, A. D. & Organization, W. H. 1996. The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020: summary. Murray, G. K. & Jones, P. B. 2012. Psychotic symptoms in young people without psychotic illness: mechanisms and meaning. Br J Psychiatry., 201, 4-6. Murru, A., Primavera, D., Oliva, M., Meloni, M., Vieta, E. & Carpiniello, B. 2015. The role of comorbidities in duration of untreated illness for bipolar spectrum disorders. Journal of affective disorders, 188, 319-323. Nock, M. K., Borges, G., Bromet, E. J., Alonso, J., Angermeyer, M., Beautrais, A., Bruffaerts, R., Chiu, W. T., De Girolamo, G. & Gluzman, S. 2008. Cross- national prevalence and risk factors for suicidal ideation, plans and attempts. The British Journal of Psychiatry, 192, 98-105. Nock, M. K., Green, J. G., Hwang, I., Mclaughlin, K. A., Sampson, N. A., Zaslavsky, A. M. & Kessler, R. C. 2013. Prevalence, correlates, and treatment of lifetime suicidal behavior among adolescents: results from the National Comorbidity Survey Replication Adolescent Supplement. JAMA psychiatry, 70, 300-310. Nock, M. K., Hwang, I., Sampson, N. A. & Kessler, R. C. 2010. Mental disorders, comorbidity and suicidal behavior: results from the National Comorbidity Survey Replication. Molecular psychiatry, 15, 868-876. O'dea, B., Glozier, N., Purcell, R., Mcgorry, P. D., Scott, J., Feilds, K.-L., Hermens, D. F., Buchanan, J., Scott, E. M., Yung, A. R., Killacky, E., Guastella, A. J. & Hickie, I. B. 2014. A cross-sectional exploration of the clinical characteristics of disengaged (NEET) young people in primary mental healthcare. Bmj Open, 4. O'reilly, A., Illback, R., Peiper, N., O'keeffe, L. & Clayton, R. 2015. Youth engagement with an emerging Irish mental health early intervention programme (Jigsaw): participant characteristics and implications for service delivery. Journal of Mental Health, 24, 283-288. Organization, W. H. 2016. Preventing Suicide: A Global Imperative. Geneva: WHO. 2014. Ormel, J., Petukhova, M., Chatterji, S., Aguilar-Gaxiola, S., Alonso, J., Angermeyer, M. C., Bromet, E. J., Burger, H., Demyttenaere, K. & De Girolamo, G. 2008. Disability and treatment of specific mental and

136

physical disorders across the world. The British Journal of Psychiatry, 192, 368-375. Ormel, J., Raven, D., Van Oort, F., Hartman, C., Reijneveld, S., Veenstra, R., Vollebergh, W., Buitelaar, J., Verhulst, F. & Oldehinkel, A. 2015. Mental health in Dutch adolescents: a TRAILS report on prevalence, severity, age of onset, continuity and co-morbidity of DSM disorders. Psychological Medicine, 45, 345-360. Passos, I. C., Mwangi, B., Cao, B., Hamilton, J. E., Wu, M.-J., Zhang, X. Y., Zunta- Soares, G. B., Quevedo, J., Kauer-Sant’anna, M. & Kapczinski, F. 2016. Identifying a clinical signature of suicidality among patients with mood disorders: A pilot study using a machine learning approach. Journal of Affective Disorders, 193, 109-116. Patel, V., Flisher, A. J., Hetrick, S. & Mcgorry, P. 2007. Mental health of young people: a global public-health challenge. The Lancet, 369, 1302-1313. Patten, S. B., Williams, J. V., Lavorato, D. H., Modgill, G., Jetté, N. & Eliasziw, M. 2008. Major depression as a risk factor for chronic disease incidence: longitudinal analyses in a general population cohort. General hospital psychiatry, 30, 407-413. Patton, G. C., Coffey, C., Romaniuk, H., Mackinnon, A., Carlin, J. B., Degenhardt, L., Olsson, C. A. & Moran, P. 2014. The prognosis of common mental disorders in adolescents: a 14-year prospective cohort study. The Lancet, 383, 1404-1411. Patton, G. C., Hetrick, S. E. & Mcgorry, P. 2007. Service responses for youth onset mental disorders. Current Opinion in Psychiatry, 20, 319-324. Paus, T., Keshavan, M. & Giedd, J. N. 2008. Why do many psychiatric disorders emerge during adolescence? Nature Reviews Neuroscience, 9, 947-957. Perry, B. D., Pollard, R. A., Blakley, T. L., Baker, W. L. & Vigilante, D. 1995. Childhood trauma, the neurobiology of adaptation, and? use? dependent? development of the brain: How? states? become? traits? Infant mental health journal, 16, 271-291. Piwek, L., Ellis, D. A., Andrews, S. & Joinson, A. 2016. The rise of consumer health wearables: promises and barriers. PLoS Medicine, 13, e1001953. Rehm, J., Mathers, C., Popova, S., Thavorncharoensap, M., Teerawattananon, Y. & Patra, J. 2009. Global burden of disease and injury and economic cost attributable to alcohol use and alcohol-use disorders. The Lancet, 373, 2223-2233. Rickwood, D., Nicholas, A., Mazzer, K., Telford, N., Parker, A., Tanti, C. & Simmons, M. 2015a. Satisfaction with youth mental health services: further scale development and findings from headspace–Australia's National Youth Mental Health Foundation. Early intervention in psychiatry.

137

Rickwood, D., Telford, N. R., Parker, A. G., Tanti, C. J. & Mcgorry, P. D. 2014. headspace—Australia’s innovation in youth mental health: who are the clients and why are they presenting? The Medical Journal of Australia, 200, 108-111. Rickwood, D. J., Mazzer, K. R., Telford, N. R., Parker, A. G., Tanti, C. J. & Mcgorry, P. D. 2015b. Changes in psychological distress and psychosocial functioning in young people visiting headspace centres for mental health problems. The Medical Journal of Australia, 202, 537-542. Rinaldi, M., Killackey, E., Smith, J., Shepherd, G., Singh, S. P. & Craig, T. 2010. First episode psychosis and employment: a review. International Review of Psychiatry, 22, 148-162. Roberts, R. E., Fisher, P. W., Turner, J. B. & Tang, M. 2015. Estimating the burden of psychiatric disorders in adolescence: The impact of subthreshold disorders. Social psychiatry and psychiatric epidemiology, 50, 397-406. Robinson, J., Cotton, S., Conus, P., Graf Schimmelmann, B., Mcgorry, P. & Lambert, M. 2009. Prevalence and predictors of suicide attempt in an incidence cohort of 661 young people with first-episode psychosis. Australian & New Zealand Journal of Psychiatry, 43, 149-157. Rochat, L., Manolov, R. & Billieux, J. 2017. Efficacy of metacognitive therapy in improving mental health: A meta‐analysis of single‐case studies. Journal of clinical psychology. Rudd, M. D., Bryan, C. J., Wertenberger, E. G., Peterson, A. L., Young- Mccaughan, S., Mintz, J., Williams, S. R., Arne, K. A., Breitbach, J. & Delano, K. 2015. Brief cognitive-behavioral therapy effects on post- treatment suicide attempts in a military sample: results of a randomized clinical trial with 2-year follow-up. American Journal of Psychiatry, 172, 441-449. Saffer, B. Y. & Klonsky, E. D. 2018. Do neurocognitive abilities distinguish suicide attempters from suicide ideators? A systematic review of an emerging research area. Clinical Psychology: Science and Practice, 25. Saxena, S., Funk, M. & Chisholm, D. 2013. World health assembly adopts comprehensive mental health action plan 2013–2020. The Lancet, 381, 1970-1971. Schmidt, U., Landau, S., Pombo‐Carril, M. G., Bara‐Carril, N., Reid, Y., Murray, K., Treasure, J. & Katzman, M. 2006. Does personalized feedback improve the outcome of cognitive‐behavioural guided self‐care in bulimia nervosa? A preliminary randomized controlled trial. British Journal of Clinical Psychology, 45, 111-121. Schulenberg, J. E., Sameroff, A. J. & Cicchetti, D. 2004. The transition to adulthood as a critical juncture in the course of psychopathology and mental health. Development and psychopathology, 16, 799-806.

138

Scott, D. & Happell, B. 2011. The high prevalence of poor physical health and unhealthy lifestyle behaviours in individuals with severe mental illness. Issues in mental health nursing, 32, 589-597. Scott, E. M., Hermens, D. F., Glozier, N., Naismith, S. L., Guastella, A. J. & Hickie, I. B. 2012. Targeted primary care-based mental health services for young Australians. Med J Aust, 196, 136-140. Scott, J., Davenport, T. A., Parker, R., Hermens, D. F., Lind, P. A., Medland, S. E., Wright, M. J., Martin, N. G., Gillespie, N. A. & Hickie, I. B. 2017. Pathways to depression by age 16 years: examining trajectories for self- reported psychological and somatic phenotypes across adolescence. Journal of Affective Disorders. Scott, J., Leboyer, M., Hickie, I., Berk, M., Kapczinski, F., Frank, E., Kupfer, D. & Mcgorry, P. 2013. Clinical staging in psychiatry: a cross-cutting model of diagnosis with heuristic and practical value. The British Journal of Psychiatry, 202, 243-245. Scott, J., Scott, E. M., Hermens, D. F., Naismith, S. L., Guastella, A. J., White, D., Whitwell, B., Lagopoulos, J. & Hickie, I. B. 2014. Functional impairment in adolescents and young adults with emerging mood disorders. Br J Psychiatry, 205, 362-8. Scott, K. M., Lim, C., Al-Hamzawi, A., Alonso, J., Bruffaerts, R., Caldas-De- Almeida, J. M., Florescu, S., De Girolamo, G., Hu, C. & De Jonge, P. 2016. Association of mental disorders with subsequent chronic physical conditions: world mental health surveys from 17 countries. JAMA psychiatry, 73, 150-158. Seligman, L. D. & Ollendick, T. H. 1998. Comorbidity of anxiety and depression in children and adolescents: An integrative review. Clinical Child and Family Psychology Review, 1, 125-144. Shanahan, M. J. 2000. Pathways to adulthood in changing societies: Variability and mechanisms in life course perspective. Annual review of sociology, 26, 667-692. Sheline, Y. I., Price, J. L., Yan, Z. & Mintun, M. A. 2010. Resting-state functional MRI in depression unmasks increased connectivity between networks via the dorsal nexus. Proceedings of the National Academy of Sciences, 107, 11020-11025. Shimokawa, K., Lambert, M. J. & Smart, D. W. 2010. Enhancing treatment outcome of patients at risk of treatment failure: meta-analytic and mega- analytic review of a psychotherapy quality assurance system. Journal of consulting and clinical psychology, 78, 298. Slade, T., Johnston, A., Oakley Browne, M. A., Andrews, G. & Whiteford, H. 2009. 2007 National Survey of Mental Health and Wellbeing: methods and key findings. Australian & New Zealand Journal of Psychiatry, 43, 594- 605.

139

Steel, Z., Marnane, C., Iranpour, C., Chey, T., Jackson, J. W., Patel, V. & Silove, D. 2014. The global prevalence of common mental disorders: a systematic review and meta-analysis 1980–2013. International journal of epidemiology, 43, 476-493. Swartz, J. R., Williamson, D. E. & Hariri, A. R. 2015. Developmental change in amygdala reactivity during adolescence: effects of family history of depression and stressful life events. Am J Psychiatry, 172, 276-83. Szanto, K., De Bruin, W. B., Parker, A. M., Hallquist, M. N., Vanyukov, P. M. & Dombrovski, A. Y. 2015. Decision-making competence and attempted suicide. The Journal of clinical psychiatry, 76, e1590. Tarrier, N., Taylor, K. & Gooding, P. 2008. Cognitive-behavioral interventions to reduce suicide behavior: a systematic review and meta-analysis. Behavior modification, 32, 77-108. Taylor-Rodgers, E. & Batterham, P. J. 2014. Evaluation of an online psychoeducation intervention to promote mental health help seeking attitudes and intentions among young adults: randomised controlled trial. J Affect Disord, 168, 65-71. Tylee, A., Haller, D. M., Graham, T., Churchill, R. & Sanci, L. A. 2007. Youth- friendly primary-care services: how are we doing and what more needs to be done? The Lancet, 369, 1565-1573. Uys, J. D., Marais, L., Faure, J., Prevoo, D., Swart, P., Mohammed, A. H., Stein, D. J. & Daniels, W. M. 2006. Developmental trauma is associated with behavioral hyperarousal, altered HPA axis activity, and decreased hippocampal neurotrophin expression in the adult rat. Annals of the New York Academy of Sciences, 1071, 542-546. Van Heeringen, K., Wu, G.-R., Vervaet, M., Vanderhasselt, M.-A. & Baeken, C. 2017. Decreased resting state metabolic activity in frontopolar and parietal brain regions is associated with suicide plans in depressed individuals. Journal of psychiatric research, 84, 243-248. Vigo, D., Thornicroft, G. & Atun, R. 2016. Estimating the true global burden of mental illness. The Lancet Psychiatry, 3, 171-178. Vos, T., Barber, R. M., Bell, B., Bertozzi-Villa, A., Biryukov, S., Bolliger, I., Charlson, F., Davis, A., Degenhardt, L. & Dicker, D. 2015. Global, regional, and national incidence, prevalence, and years lived with disability for 301 acute and chronic diseases and injuries in 188 countries, 1990-2013: a systematic analysis for the Global Burden of Disease Study 2013. The Lancet, 386, 743. Walker, E. F., Sabuwalla, Z. & Huot, R. 2004. Pubertal neuromaturation, stress sensitivity, and psychopathology. Development and psychopathology, 16, 807-824.

140

Walker, S. 2010. A description of the establishment of a new child and adolescent mental health service in the United Kingdom. Journal of child and adolescent mental health, 22, 35-39. Wardenaar, K. J., Van Loo, H. M., Cai, T., Fava, M., Gruber, M. J., Li, J., De Jonge, P., Nierenberg, A. A., Petukhova, M. & Rose, S. 2014. The effects of co-morbidity in defining major depression subtypes associated with long-term course and severity. Psychological medicine, 44, 3289-3302. Webb, M. J., Kauer, S. D., Ozer, E. M., Haller, D. M. & Sanci, L. A. 2016. Does screening for and intervening with multiple health compromising behaviours and mental health disorders amongst young people attending primary care improve health outcomes? A systematic review. BMC family practice, 17, 104. Weisz, J. R., Jensen-Doss, A. & Hawley, K. M. 2006. Evidence-based youth psychotherapies versus usual clinical care: a meta-analysis of direct comparisons. American Psychologist, 61, 671. Weisz, J. R., Jensen, A. L. & Mcleod, B. D. 2005. Development and Dissemination of Child and Adolescent Psychotherapies: Milestones, Methods, and a New Deployment-Focused Model. Weisz, J. R. & Kazdin, A. E. 2010. Evidence-based psychotherapies for children and adolescents, Guilford Press. Westen, D., Novotny, C. M. & Thompson-Brenner, H. 2004. The empirical status of empirically supported psychotherapies: assumptions, findings, and reporting in controlled clinical trials. Psychological bulletin, 130, 631. Whiteford, H. A., Degenhardt, L., Rehm, J., Baxter, A. J., Ferrari, A. J., Erskine, H. E., Charlson, F. J., Norman, R. E., Flaxman, A. D. & Johns, N. 2013. Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010. The Lancet, 382, 1575-1586. Ybarra, M. L. & Eaton, W. W. 2005. Internet-based mental health interventions. Mental health services research, 7, 75-87. Yung, A. R. 2016. Youth services: the need to integrate mental health, physical health and social care. Social psychiatry and psychiatric epidemiology, 51, 327-329. Zachrisson, H. D., Rödje, K. & Mykletun, A. 2006. Utilization of health services in relation to mental health problems in adolescents: a population based survey. BMC public health, 6, 34.

141