The Role of Self-Esteem and Self-Criticism in Suicidal Thoughts

Home , Criticism

THE ROLE OF SELF-ESTEEM AND SELF-CRITICISM

IN SUICIDAL THOUGHTS AND BEHAVIOURS

A thesis submitted to the University of Manchester

for the degree of Doctor of Clinical Psychology

in the faculty of Biology, Medicine and Health

2017

Catherine F. O’Neill

School of Health Sciences Division of Psychology and Mental Health

LIST OF CONTENTS List of Appendices…………………………………………………………………………………………………………..3 List of Tables…………………………………………………………………………………………………………………..3 List of Figures………………………………………………………………………………………………………………….3 Word Count…………………………………………………………………………………………………………………...4 Thesis Abstract……………………………………………………………………………………………………………….5 Declaration…………………………………………………………………………………………………………………….7 Copyright Statement………………………………………………………………………………………………………8 Acknowledgements………………………………………………………………………………………………………..9 Paper 1: Systematic Review

Abstract………………………………………………………………………………………………………………………………………………………11

Introduction………………………………………………………………………………………………………………………………………………..13

Method……………………………………………………………………………………………………………………………………………………….20

Results………………………………………………………………………………………………………………………………………………………...23

Discussion……………………………………………………………………………………………………………………………………………………42

Conclusion……………………………………………………………………………………………………………………………………………………52

References…………………………………………………………………………………………………………………………………………………..52 Paper 2: Experimental Paper

Abstract……………………………………………………………………………………………………………………………………………………….63

Introduction…………………………………………………………………………………………………………………………………………………64

Method………………………………………………………………………………………………………………………………………………………..74

Results………………………………………………………………………………………………………………………………………………………….80

Discussion…………………………………………………………………………………………………………………………………………………….88

Conclusion……………………………………………………………………………………………………………………………………………………94

References……………………………………………………………………………………………………………………………………………………95 Paper 3: Critical Reflection

Introduction………………………………………………………………………………………………………………………………………………….100

Definitions……………………………………………………………………………………………………………………………………………………101

Paper one reflections……………………………………………………………………………………………………………………………………106

Paper two reflections……………………………………………………………………………………………………………………………………114

Overall conclusions……………………………………………………………………………………………………………………………………….124

References…………………………………………………………………………………………………………………………………………………….125

LIST OF APPENDICES

Appendix A: Author submission guidelines – Clinical Psychology Review 131

Appendix B: Author submission checklist – Journal of Abnormal Psychology 143

Appendix C: BECS Study Protocol 146

Appendix D: Protocol for Managing Disclosure of Risk 160

Appendix E: Participant Information Sheet 171

Appendix F: Ethical Approval, University of Manchester 174

Appendix G: Ethical Approval, NHS 176

Appendix H: Ethical Approval Manchester Mental Health and Social Care Trust 183

Appendix I: Data Extraction Sheet for Systematic Review 185

LIST OF TABLES Systematic Review

Table 1: Included Studies and extracted data 26

Table 2: Results of Newcastle-Ottowa scale adapted for cross-sectional studies 31 Experimental Paper

Table 1: Participants by self-reported diagnosis/ mental health condition 81

Table 2: Descriptive statistics 82

Table 3: Bivariate correlations of all variables 82

Table 4: Partial correlations controlling for depression and hopelessness 82

Table 5: Multiple regression predicting suicide probability (inadequate self) 84

Table 6: Multiple regression predicting suicide probability (self-attacking) 86 Reflective Paper

Table 1: Factor analysis of self-esteem scales 112

LIST OF FIGURES

Figure 1 IMV model of Suicide (O’Connor, 2011) 68

Figure 2 PRISMA diagram 25

Figure 3 Mediation analysis – entrapment as mediator- inadequacy 86

Figure 4 Mediation analysis – entrapment as mediator – self-attacking 88

WORD COUNT

Section Word count Thesis abstract 368 Systematic review 9,944 Empirical paper 6,715 Critical reflection 6,272 Total 23,299

THESIS ABSTRACT

This thesis forms part of the examination for the degree of Doctor of Clinical Psychology in the Faculty of Biology, Medicine and Health at the University of Manchester. This thesis has been written by Catherine Frances O’Neill and submitted in December 2017 for examination

January 2018.

The thesis focuses on the relationship between self-esteem, self-criticism, self- reassurance and suicidality. Paper one describes a systematic literature review and meta- analyses of the available research on self-esteem and suicidality. The review aimed to examine the relationship between self-esteem and suicidality, and whether there were any key moderators or mediators of this relationship. The findings indicate that there is an overall moderate negative relationship between global self-esteem and suicidality with an r=-0.42. Where studies examined levels of contingent self-esteem (i.e. self-esteem based upon external factors such as academic achievement and appearance) there was a stronger relationship with suicidal ideation and behaviours. Implicit self-esteem was measured by one study, which showed no relationship to suicide, however the combination of high implicit and low explicit self-esteem was significantly related. The review illustrated the need for further investigation of the relative contingency of self-esteem to suicidality, along with differences in implicit and explicit self-esteem and suicidality. Further longitudinal and experimental studies are also warranted to explore causality.

Paper two describes an investigation into the role of self-criticism, specifically feelings of inadequacy and self-attacking, self-reassurance and suicide risk. A total of 101 participants from mental health inpatient and community mental health settings completed questionnaires for this study. Results indicated that self-attacking was significantly 5 associated with suicide risk, when hopelessness and depression were controlled for in the analysis. Entrapment, hopelessness and self-attacking were significant predictors of suicide risk, above depression, which was non-significant. Inadequacy was not a significant predictor of suicide risk. Self-reassurance, whilst negatively related to both self-attacking and inadequacy, did not moderate the relationship between self-attacking, inadequacy and suicide risk. The theoretical, clinical and research implications, along with limitations are discussed within this paper.

Paper three provides a critical reflection of papers one and two, including an exploration of the challenges of defining and measuring self-esteem, self-criticism and suicide. Personal reflections and recommendations for good practice in future suicide research are included in this paper.

DECLARATION

No portion of the work referred to in the thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning

i. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the “Copyright”) and s/he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes. ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made. iii. The ownership of certain Copyright, patents, designs, trademarks and other intellectual property (the “Intellectual Property”) and any reproductions of copyright works in the thesis, for example graphs and tables (“Reproductions”), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions. iv. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see http://documents.manchester.ac.uk/DocuInfo.aspx?DocID=24420), in any relevant Thesis restriction declarations deposited in the University Library, The University Library’s regulations (see http://www.library.manchester.ac.uk/about/regulations/) and in The University’s policy on Presentation of Theses

ACKNOWLEDGEMENTS

I would like to thank my Supervisors, Professor Gillian Haddock and Dr Daniel Pratt for their support and guidance in developing this thesis. I would also like to acknowledge the unwavering support of my Husband, Stephen Jackson - thank you Steve, for all your time, patience and understanding. Also to my daughter Maia, for being such a wonderful counterpoint to this process, spending time with you has been the best part of these last 2 years. Finally, thank you to the ClinPsyD Mum’s and my BECS trainee colleagues, you are all amazing, inspiring, strong women -your support has been invaluable.

PAPER ONE: SYSTEMATIC REVIEW

The role of self-esteem in suicidal thoughts and behaviour: A systematic review

The following paper has been prepared for submission to the Journal of Clinical Psychology Review.

The guidelines for authors can be found in Appendix A.

Abstract 196

Main Text (excluding 9,748 references)

ABSTRACT

Self-to-self relating has been highlighted as a key feature of explanatory models of suicide over the past two decades. To date there has not been a systematic attempt to synthesise the literature exploring the relationship between suicide and self-esteem in an adult population. This review systematically searched available published literature which explored the relationship between self-esteem and suicidality. Studies of sufficient number, which reported similar statistics were combined in meta-analysis. Eighteen published journal articles were included. A risk of bias assessment was undertaken to allow for findings to be assessed for reliability. Study findings indicated robust, moderate relationships between global self-esteem and suicidality overall (r=-0.42) and similarly when solely examining suicidal ideation (r=-0.42). Although limited in number, studies which examined level of contingency of self-esteem (i.e. that self-esteem is based on external factors such as achievement of personal and interpersonal goals and standards) found a stronger relationship between contingent low self-esteem and suicidal thoughts and behaviour than global self-esteem. Individuals with levels of high implicit and low explicit self-esteem were found to experience high levels of suicidal ideation. Further research is required to explore these relationships, preferably of a longitudinal and/ or experimental nature to delineate causality.

HIGHLIGHTS

• Global Self-Esteem is negatively related to suicidal ideation, behaviour and risk

• Contingent self-esteem is related to increased suicidal ideation and behaviour

• Self-esteem research requires precision in terms of its definition and components

KEYWORDS

SUICIDE SELF-ESTEEM SELF-WORTH SELF-CONCEPT

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

INTRODUCTION

Suicide is the cause of 800,000 deaths globally each year, and is the second highest cause of death for 15-29 year olds (WHO, 2017). Despite the significant efforts of researchers and clinicians within this field suicide rates are increasing in many high-income countries (e.g.

ONS, 2015, WHO, 2014, Curtin - CDC, 2014). In the UK, the suicide rate has decreased overall since 2005, however recent publications suggest that this rate decrease is slowing

(Centre for Mental Health Studies, 2017). Researchers need to continue to find novel ways to tackle this serious public health issue. From an epidemiological perspective, rates of suicidal ideation, plans and behaviours vary considerably by geographic region. Estimated worldwide prevalence rates for suicidal ideation have ranged from around 6.2 % to 9.2%, for active plans 2.3% to 3.1% and non-lethal suicide attempts from 0.9% to 2.7% of the global population (Borges, Angst, Nock, Ruscio & Kessler, 2008; Nock et al, 2008). However, higher rates of ideation, plans and attempts were found in the USA (15.5%, 5.4% and 5% respectively), despite an overall decrease in suicide globally between 2002 and 2012 (WHO,

2014).

Alongside their significant personal and familial impact, completed suicide and suicide behaviours are also nineteenth in the global burden of disease rankings (WHO,

2008), which reflects the years of disability, ill health and loss of life experienced by individuals with these problems. It is therefore imperative that researchers are able to understand and respond to the factors that make up the journey from mental distress to suicidal thoughts and behaviours.

In terms of known sociodemographic characteristics linked to suicidality, males are three times more likely to complete suicide than females (WHO, 2014), though females are more likely to attempt to take their own life. Other risk factors include being of younger age, unmarried and with low levels of education, higher levels of poverty and unemployment and substance misuse (Hawton et al, 2009; Hawton et al, 2012; Nock et al, 2013).

There is a substantial amount of empirical evidence linking suicidality to a variety of diagnosable mental health conditions (Donker et al, 2013; Hawton, Casanas I Comabella,

Haw, & Saunders; 2013, Rotenstein et al, 2016, De Crescenzo et al, 2017; Beyer & Weisler,

2016; Panagioti, Gooding, Pratt & Tarrier, 2015). Many of these studies examine the relationship of diagnoses to suicidal ideation, rather than suicidal behaviours. Recent research has suggested that in fact most mental health diagnoses can be significantly related to increased levels of suicidal ideation (Klonsky, May & Saffer, 2016). However, diagnoses may have little predictive power in terms of future suicidal behaviour as the vast majority of individuals with these diagnoses do not go on to attempt suicide (O’Connor &

Nock, 2014).

In light of this, research has attempted to identify further factors that may be important to consider in the understanding and prevention of suicide. A variety of psychological constructs associated with suicidal thoughts and behaviours have been examined, such as hopelessness, negative self-appraisals, situational appraisals of defeat and entrapment and personality traits such as perfectionism (Johnson, Gooding &Tarrier,

2008; O’Connor, Cleare, Eschle, Wetherall & Kirtley, 2016; Beevers & Miller, 2009, Brezo et al, 2006, Joiner, 2007). The identification of these constructs allows for targeted interventions to be developed by clinicians, which aims to prevent the escalation of suicidal

14 thoughts to behaviour. One area highlighted in the search for suicide risk factors is self- esteem.

Early models of suicidality highlighted the importance of self-to-self relating in terms of risk of suicide (Baumeister, 1990). In Baumeister’s model, suicide attempts represented an escape from aversive self-awareness – having fallen short of standards and expectations and the negative self-attributions which follow. Following this, both the cry of pain model

(Williams, 1997) and theories of arrested flight (Gilbert & Allan, 1998) proposed that evolutionary and cognitive processes created a sense of entrapment which was compounded by negative appraisals and difficulties with problem solving. Negative self- evaluations in this context represented an inescapable stressor which increased feelings of entrapment and the search for escape from ongoing intra-physic pain.

The Schematic Appraisal Model of Suicide (SAMS; Johnson et al, 2008) proposes that a number of cognitive factors are active in the development of suicidal thoughts and behaviours. Key mechanisms within the model include information processing biases, where negative and emotionally laden imagery and thoughts are coded into memory and more frequently accessed during times of vulnerability to suicidality. The model also highlights the role of appraisals, particularly the importance of negative appraisals of the self. The SAMS model suggests that low self-esteem, negative self-perceptions and appraisals of personal attributes contribute to suicidality via a sense of low personal agency, problem solving ability and of having limited social resources to respond to others. These elements contribute to a ‘suicide schema’, which is activated when mood becomes low.

The integrated motivational-volitional model of suicidality (IMV model; O’Connor et al, 2011) describes a pathway from mental distress to suicidal behaviour via a number of

15 predisposing, motivational and volitional factors. These factors include pre-disposing personality characteristics (inclusive of self-esteem), life events and environmental stressors

(in the pre-motivational phase), and a motivational phase with includes experiences of defeat and humiliation, entrapment and alongside threat to self. The volitional phase proposes separate mechanisms to move from suicidal ideation to action, including volitional moderators such as impulsivity, social learning and active intentions. The specific role of self-esteem in this model is described as influencing both the pre-motivational phase, with suicidal ideators/ enactors differing from controls on levels of self-esteem, and also its impact in the motivational phase in terms of its influence on levels of defeat, humiliation and a sense of ‘threat to the self’ (O’Connor, Smyth, Fergusson, Ryan, & Williams, 2013).

Self-esteem is a multi-dimensional concept, frequently cited in both clinical and social psychological literature (Rosenberg, 1979; Crocker 2002; Leary and Baumeister, 2000

& De Ruiter, Van Geert & Kunnen, 2017). Rosenberg’s (1965) seminal work on self-image described self-esteem in terms of a global, stable attitude or feeling about the self. These ideas have been further developed, with self-esteem seen as contingent on performance in specific domains in life, for example academic performance, personal appearance or friendship (Crocker et al, 2002, Cambron, Acitelli, & Pettit 2009, 2010). This idea has been developed further by Heppner & Kernis (2011) who have referred to ‘fragile vs secure self- esteem’, where individuals with higher levels of fragile self-esteem being shown to be sensitive to perceived attacks or threats, and at increased risk of suicidal ideation.

Poor self-esteem has been found to be reliably correlated with a range of difficulties; suicidal ideation and attempts, non-suicidal self-injury, depression, persecutory delusions, obesity and in the case of inflated self-esteem, violence (Emler, 2001; Baumeister,

Campbell, Krueger & Vohns 2003; Forrester, Slater, Jomar, Mitzman & Taylor, 2017; Torres 16

& Fernandez, 1995; Walker & Bright, 2009; Kesting and Lincoln, 2013 and Grifiths, Parsons,

& Hill, 2010). A variety of studies have found a significant relationship between self-esteem and suicidality (Chen, Wu & Bond, 2009; Kleiman and Riskund, 2013; Gooding et al, 2015;

Wilke, 2006; Chatard, Selimbegovic, & Kovan, 2009). The relationship between self-esteem and suicidality has also been found to remain significant when depression, a known correlate of suicidal thoughts and behaviours, is controlled for in the analysis (Gooding et al,

2015). However, it is important to note that much of the literature on self-esteem uses cross-sectional design and correlational results, suggesting that it is just as possible that low self-esteem is an emergent phenomenon that derives from loss of social status/ victimisation and suicidal feelings rather than a causal mechanism.

Given the evidence considered above, self-esteem appears to have a role in suicidal thoughts and behaviours. The strength and exact nature of this relationship is unclear, given that there has never been a comprehensive review on the topic. Thus, this review will investigate the relationship between self-esteem and suicidality. The review will also examine the role of self-esteem as a potential moderator or mediator in the relationship between suicidality and other variables, such as childhood maltreatment, socially prescribed perfectionism and sports participation (Glassman, Weierich, Hooley, Deliberto, & Nock,

2007; Rice, Ashby & Slaney, 1998; Taliaferro, Rienzo, Miller, Pigg & Dodd, 2010).

Understanding the role of self-esteem as a mediator can be helpful when undertaking therapeutic work as it allows client and clinician to prioritise interventions on areas which will have the greatest impact on distressing suicidal thoughts and impulses. Similarly, an examination of mediators between self-esteem and suicidality will also be carried out, including known correlates of suicidality; depression, gender, age, defeat and entrapment

(Victor & Klonsky, 2014, O’Connor et al, 2016). Again, not only is this information clinically

17 valuable, it allows for a picture to emerge around the pathways involved when people move from predisposing personality factors to suicidal behaviours.

As discussed above, there is a lack of consensus in the self-esteem literature as to its precise definition. The measure found most commonly in the literature is the Rosenberg Self

Esteem Scale (1965) which describes self-esteem as a characteristic linked to evaluating one’s global self-worth. As such this definition will be used to decide if papers meet review criteria (for a full discussion please see paper 3). Searches will initially be inclusive of papers with a focus on both trait self-esteem, that is a stable and long-term sense of self-worth, and state self-esteem which may be related to changes to self-appraisals in response to external and internal factors.

Imprecise terms within suicide research have provoked challenges when trying to combine and analyse studies. Researchers may describe suicidal behaviour as including any form of suicidal activity, including non-fatal attempts, and/or suicidal ideation, and/or actions undertaken to complete suicide (Klonsky et al, 2016). Suicidal ideation is defined in this review as including thoughts of wanting to die, alongside specific thoughts of engaging in suicidal behaviour. The term suicidal behaviour is used in this review to refer to potentially self-injurious actions where there is implicit or explicit evidence that the person intended to end their life. Suicidality will be used to describe the combination of suicidal thoughts and behaviour. Papers which examined Non-suicidal self-injury without explicit suicidal intention were excluded from this review, as the function of this behaviour has been shown to be linked to emotion regulation or communicating distress rather than the ending of life (Taylor et al, 2018). However, it should be noted that taxometric analysis of both suicidal self injury (SSI) and non-suicidal self-injury has suggested these behaviours may exist on a continuum (Orlando et al, 2015). 18

There have previously been no systematic reviews which have examined the role of self-esteem in suicidality in an adult population, thus this review will contribute new knowledge to the field.

The questions to be addressed in this review are:

• In an adult population, what is the relationship between self-esteem and suicidality?

• What factors moderate the relationship between self-esteem and suicidality?

• Are there any differences in associations between self-esteem and suicidality in

those who experience suicidal ideation in comparison with individuals who engage in

suicidal behaviours?

METHOD

Pre-registration of review protocol

The study was registered on the PROSPERO site (reference CRD42016033130) prior to commencing searches.

Inclusion criteria of the studies included in the review:

• Published in a peer reviewed journal in the English Language

• The sample included adults, aged 18- 65 years

• Included a validated measure of self-esteem

• Included a validated measure of suicidal ideation or behaviour

• Analysis of the relationship between self-esteem and suicidality was reported

Exclusion criteria comprised:

• Studies which used qualitative methods and case series

• Studies with a specific focus on non-suicidal self-injury

• Samples with mixed populations (both adults and children/ adolescents/older

adults), where adult data was not reported separately

• Studies published before 1990 to ensure a focus upon most recent models of

suicidality.

Search strategy for inclusion in the review

The initial search strategy was deliberately kept broad to minimise the likelihood of missing relevant studies. Electronic searches of EMBASE (1990 to June 2017, Medline (1990-June

2017), PsychInfo (1990-June 2017) were carried out with search terms including: Suicid*,

20 self-esteem, self-worth, self-image, self-regard, self-concept, self-confidence, self- evaluation, self-liking, self-attitude, self-competence and self-appraisal. These terms were devised from existing MESH terms, and existing self-esteem systematic review literature

(see below). Searches were carried out of the full text of articles, rather than abstracts. In the development of appropriate search terms, two other reviews of self-esteem literature were examined (Randall et al, 2015, Kesting and Lincoln, 2013). The search was initially carried out by the first author with twenty percent of full text articles categorised by an independent reviewer. Full agreement was reached between reviewers following discussion.

Eighteen articles were identified for inclusion in the review (see fig 1). A data extraction form was used by the first author to extract data from the included articles. Extracted data included Study ID, country of publication, research design, population studies, recruitment strategy, any follow up, details of setting (clinical vs non-clinical), number of participants and percentage of male/ female, mean age (standard deviation and range), primary diagnoses, outcomes, type of analysis and values reported, measures used, key descriptive outcomes, direction of association between studies variables (see appendix J for a copy of the data extraction sheet).

Risk of bias assessment

The Newcastle-Ottowa scale (Wells et al, 2011), adapted for cross-sectional studies, was used to assess the risk of bias within each study. A higher score indicates a reduced risk of bias. The risk of bias assessment of each study was conducted by the first author and 20% of the papers were independently reviewed by the fifth author. Any discrepancies were discussed until 100% agreement was achieved. Risk of bias assessment covered the

21 following areas; representativeness, sample size, reporting of non-respondents, comparability and control of important factors, assessment of outcome and statistical tests used. Risk of bias assessment allows the strengths and weaknesses of studies to be examined and will inform the strength of any recommendations based upon their findings, with greater emphasis being placed on those studies whose findings came from less biased studies. Within the risk of bias assessment, high quality studies were considered to have controlled for key confounding variables including depression, age and the gender of participants. Depression has been consistently linked to suicidality in the literature

(McGinty, Sayeed & Upthegrove, 2017; Devenish, Berk & Lewis, 2016; Hawton et al, 2013).

Age and gender are considered key demographic variables in terms of suicidality (WHO,

2014; Hawton et al, 2009; Hawton et al, 2012).

Meta-Analysis

Where there were more than 5 papers which used similar methodology and statistical analyses, a meta-analysis was undertaken to examine the relationship between self-esteem and suicidality. An a-priory decision was made to utilise the inverse variance heterogeneity

(IVhet) model in the meta-analysis, as this accounts for the impact of random effects in the data. This decision was made due to the high levels of heterogeneity in the studies in terms of client groups, locations and age and gender mix. The statistical package meta-xl was used to carry out meta-analysis (Barendregt, 2011). Heterogeneity was assessed using the I-

Squared statistic. In response to high i-squared values, which may be expected due to differences in design, measurement and sampling of studies included in the review, the

DerSimonian and Laird (1986) inverse variance heterogeneity (IVhet model) was used.

Correlational data were selected as the effect size for use in the meta-analysis, and studies that did not provide data in this format were not included in the meta-analysis but were included in the narrative synthesis.

RESULTS

Study characteristics

Studies were identified from initial searches (n=6358), leaving nineteen studies that were selected for inclusion in this review. See Figure 2 for PRISMA diagram including reasons for exclusion at the full text stage.

All included studies and their characteristics and data can be found in Table 1. All studies used cross-sectional methods. Ten studies used non-clinical samples (undergraduate students and general population samples) and nine studies were carried out in clinical populations. One study researched psychiatric inpatients, one psychiatric outpatients, one mixed out and inpatients. One study examined prison inpatients, one used a specific group of patients with persecutory delusions and one studied a group of patients attending plastic surgery services for gunshot wounds to the face following a suicide attempt. There was one study which examined adults attending a substance misuse clinic.

In terms of the location in which the research was carried out, six studies came from the USA, one from Canada and three from the UK. Two studies came from China, one from

South Korea, two from Japan and one from Taiwan. Two studies came from Turkey and one from the Netherlands. In terms of the age range of participants, nine out of twenty of the studies had participants of under 25 years, and nine with participants over 25 (one non- reported). Eight of the 19 studies had more than 50% males, and ten had less than 50% males (1 non-reported).

Figure 2 PRISMA flow diagram

Records identified through database searching) N= 6358

dentification I

Records after non-English language removed (n= 5151)

duplicates removed (n = 3833)

Screening Records screened Records excluded

(n = 3833) (n =3619)

Full-text articles assessed Full-text articles excluded, Eligibility for eligibility with reasons below (n = 214) (n = 195)

No validated self-esteem measure 50

Studies included in No validated measure of suicide 46 ideation or behaviour quantitative synthesis (n = 19) No joint analysis of SE/ Suicidality 43 Includes under 18 yrs within sample 28

Includes over 65 yrs within sample 14

Duplicate entry 5

Qualitative design 3

Not English language 2

Not peer reviewed 2

Case series design 1

Primary focus of NSSI 1

Study Study Design Participants Measures (Self-esteem and Reported statistics Other relevant findings suicidality) Bhar, Cross sectional N = 338 Beck Self Esteem Scale (Self and BSE self-scale R² change 0.1 F change Both Self-based and Other Based Ghahramanlou- Adult Outpatients Other based) 5.47 p=0.02* (controlling for depression self-esteem predicted suicidal Holloway, Diagnoses: Mood disorders (74%), Beck Scale for Suicidal Ideation and hopelessness) ideation. Brown & Beck Anxiety Disorders (46%), (BSS) (2008) Substance use (28%), personality BSE other-scale R² change = 0.2, USA disorders (51%) Fchange = 9.27 p=0.03*(controlling for

depression and hopelessness) Age = 36.3 (SD 10.6)

Collette, Pugh, Cross- N = 42 Rosenberg Self Esteem Scale R=-0.77 *** There were significant differences Waite & sectional Individuals with persecutory Beck Scale for Suicidal Ideation between clinical and non-clinical Freeman (2016) delusions (21), age and gender (BSS) groups in mean scores on RSES UK matched non-clinical controls Clinical = 11.95 (5.63) Age = (Clinical) 45.6 (SD 12.1) Non-clinical 21.10 (4.49) (Non-clinical) 41.9 (SD 12.2)

Creemers, Cross- N = 95 RSES (Explicit SE) Implicit SE r=0.18 Low explicit self-esteem with high Scholte, Engels, sectional Undergraduate Students Name-letter task (Implicit SE) Explicit SE r= -0.36** implicit self-esteem showed Prinstein & Age = 21.2 (SD 1.88) Suicidal Ideation Questionnaire increased suicidal ideation Wiers (2012) (SIQ) (b=0.029) The Netherlands De Man & Cross- N= 131 RSES R = -0.5*** No relationship between Gutierrez (2002) sectional Undergraduate students Beck Scale for Suicidal Ideation sR = -0.20* (when depression was instability of self-esteem and Canada Age = 18-24 (specific ages not (BSS) controlled for) suicidal ideation. The relationship reported) between self-esteem level and instability was significant Demirbas, Celk, Cross- N = 70 Coopersmith Self -esteem R=-0.64*** Ilhan & Dogan sectional Male adults attending alcohol and inventory (2003) substance misuse clinic SPS Turkey Age = 41.6 +/- 8.1

Gooding et al Cross- N = 65 Robson Self Concept R=-0.70** The relationship between low (2015) sectional Male prisoners at high risk of Questionnaire Ideation subscale = -0.58** self-esteem and suicide UK suicide Suicide Probability Scale (SPS) Hopelessness subscale = -0.71** probability was driven by the Age = 35.8 (SD 10.95) Negative evaluation ss = -0.49** hopelessness component of the Hostility subscale = -0.30* SPS, but not its hostility, suicidal

Study Study Design Participants Measures (Self-esteem and Reported statistics Other relevant findings suicidality) ideation or negative evaluation subscales.

Jahn, DeVylder Cross- N = 590 RSES Self-esteem and suicidal ideation – R= - Of the mediators, only self- & Hillmire sectional Undergraduate students Columbia Suicide Severity 0.36*** esteem was a significant direct (2016) Age = 18.8 years (SD 1.6) Rating Scale (CSSRS) Self-esteem and suicide attempts – R= - predictor of suicide attempts, b= USA USA 0.18** -0.121, p= 0.013, z =-2.476

Kim et al (2015) Cross- N= 414 RSES R=0.35*** Suicidal ideation group scored South Korea sectional Korean Soldiers BSS higher on measure of self-esteem Age = Clinical 21.03 (+/- 1.36) than healthy controls on one table Non-Clinical 20.82 (+/- 1.10) in the paper. See discussion below.

Lakey, Hirsch, Cross- N = 371 RSES RSES + SBQ-R = r = -0.35** Contingent low self-esteem Nelson & sectional Undergraduate students Contingent Self-Esteem Scale Contingent SE = r = -0.41** related to greater suicidal Nsamenang Age = 19.41 (SD = 2.08) Suicidal Behaviours behaviour, non-contingent high (2014) Questionnaire- Revised (SBQR) self-esteem related to lower USA suicidal behaviour Lin (2015) Cross- N = 18 RSES R=-0.63*** Gratitude had indirect effects on Taiwan sectional Undergraduate Students Positive and Negative Suicidal suicidal ideation via self-esteem Age = 20.13 (1.11) Ideation (PANSI) and depression.

Mitsui et al Cross- N = 45 RSES F= 19.7 (between 3 groups)*** (2014) sectional Adult Japanese University Suicide section on MINI Mean RSES: Japan Students (validated for standalone use) MDE w/ suicidality 19.9 Age = Major Depressive Episode MDE no suicidality 24.8 (MDE) w/ suicide 22.3 +/- 3.3 Health control 34.5 MDE no suicide 21.7 +/- 2.4 Healthy Controls 20.9 +/- 1.9

Nakamura et al Cross- N = 228 RSES Odds ratio 0.975 p=0.0003 *** Risk factor of suicidal depression (2011) sectional Employees at Japanese companies Suicidal Depression Subscale of CI 0.91-0.988 in relation to self-esteem. OR Japan Age = (Male) 43 +/- 9.2 GHQ-30 (has been suggests negative relationship (Female) 36.3 +/- 7.9 independently validated) between self-esteem and suicidal depression

Study Study Design Participants Measures (Self-esteem and Reported statistics Other relevant findings suicidality) Ozturk, Bozkurt, Cross- N = 24 RSES No significant difference overall Constancy of self-respect and Durmus, Deveci sectional Adults attending department of SPS between suicidal group and healthy depressive mood questions on & Sengezer plastic surgery with facial controls on RSES the RSES were higher in suicidal (2006) deformities due to attempted No figures reported group than health controls ** Turkey suicide by gunshot wound (12) and healthy controls (12) Age = n/r Palmer (2004) Cross- N= 116 Self-esteem rating scale Group 1 (Depressed no ideation/ USA sectional Depressed Inpatients x 3 groups Suicide risk scale attempt) scored significantly higher on Age= measures of self-esteem and lower on Group 1 No Ideation – 34 (SD 14) suicide risk measures than group 2 Group 2 Ideation no Attempt– 34 (ideators) (SD 8.6) F=5.94 p=0.5* Group 3 Ideation and Attempt – 34 Or Group 3 (past suicide attempt) (SD 6.3) F=5.94 p=0.1** . Plutchik, Botsis Cross- N = 685 Self Esteem Scale R= -0.29* & Van Praag sectional Psychiatric in-patients Suicide Risk Scale (1995) Age = (M) 31.75 (SD 9.75), (F) 36.2 USA (SD 10.75) Mixed diagnoses; affective disorders (30%), schizophrenic disorders (30%), anxiety (4%), drug abuse (40%), adjustment disorders (16%) and personality disorders (40%) Taliaferro et al Cross- N=450 Rosenberg Self Esteem Scale R= - 0.52*** (2010) sectional Undergraduate students (RSES) USA Age = 19 (SD 1.4) Adult Suicidal Ideation Questionnaire

Tarrier, Cross- N = 59 Self-evaluation and Social Negative Evaluation of Self (NES) and BSS Barrowclough, sectional Adult inpatients and outpatients Support in Schizophrenia Scale r=-0.371** Andrews & Diagnose: Schizophrenia 83%, BSS Gregg (2004) schhizophreniform disorder 12%, Positive Evaluation of Self (PES) and BSS UK schizoaffective disorder 5% r= -0.108

Age = 27.2 (SD 7.6)

Study Study Design Participants Measures (Self-esteem and Reported statistics Other relevant findings suicidality)

Zhang, et al Cross- N = 514 RSES All clients (2010) sectional Chinese adult men and women BSS β= -0.195 t =3.226*** China from 3 correctional facilities Male –β = -0.029 t = 0.415 Age = (M) 30.36 (F) 30.51 Female–β= -0.505 t = 4.628***

Zhang, Norvilitis Cross- N = 626 RSES USA sample B =0.23 (total), 0.14 (M), Low self-esteem is related to & Ingersoll sectional Undergraduate Students Self-report Suicidal Ideation 0.23** (F) greater levels of suicidal ideation (2007) Age = USA 22 (SD 2.7); n= 283 Scale in Chinese women than men. USA Chinese 21.7 (SD 1) Chinese sample B= 0.11 (total), USA n = 343 -0.10 (M), 0.23** (F)

* = p < .05; ** = p < .001; *** = p < .0001

Abbreviations:

GP= Group, MSE = Mean Self-Esteem, BSE= Beck Self-Esteem Scale, RSES = Rosenberg Self-Esteem Scale, SPS = Suicide Probability Scale, BSS = Beck Suicide Scale, SBQR= Suicide Behaviours Questionnaire Revised, MDE = Major Depressive Episode

Table 1 Included Studies and Extracted Data

RISK OF BIAS ASSESSMENT

All studies were cross sectional in design. As studies were largely cross sectional it is not possible to draw any firm conclusions around the causality of variables reported.

There were a number of key issues associated with the methodological quality of the studies. No studies provided a justification for the sample size of their studies, including carrying out a power calculation. A number of studies carried out research on a small population (n<35), for example, Ozturk et al, 2006 (n=24) and Lin et al, 2015 (n=18) and without sufficient power it is possible that these studies may increase their risk of making a type 2 error. The vast majority of studies (k= 14) did not report or make reference to the response rate of participants, or their characteristics. Papers which received the highest ratings provided a good explanation for the representativeness of their sample and described the treatment of non-responders (Taliaferro et al, 2010, Tarrier et al, 2004 and

Zhang et al, 2010). Few studies controlled for all three key confounding variables associated with suicidality (for the purposes of this review these included depression as the key confounding variable, and age/ gender as other known correlates), although some did control for at least two of them (Taliaferro et al, 2010, De Man, 2002, Lakey et al, 2014,

Zhang et al, 2010, Lin et al, 2005 and Palmer et al, 2004). Nine studies controlled for depression in their analysis, using symptom severity scores, nine controlled for age and ten controlled for gender. Four studies did not control for any confounding factors or did not report data on confounders (Zhang et al, 2007, Kim et al, 2015, Creemers et al, 2012 and

Demirbas et al, 2003). All studies used self-report measures. Risk of bias results have been considered in the discussion below, when reflecting on the reliability of results and the strength of conclusions that can be drawn from them.

Table 2 Results of the Newcastle-Ottowa Scale adapted for cross sectional studies.

Study Selection Comparability Outcome Total Representativeness * Control for Assessment of /10* Sample Size * most outcome * Non-respondents * important Statistical test Validated factor * * measurement tool ** Controls for Max 3* Max 5* any other factor * Max 2* Bhar et al (2008) *** ** * 6 Collett et al (2016) ** * ** 5 Creemers et al (2012) *** ** 5 De man et al (2002) *** ** * 6 Demirbas et al (2003) *** ** 5 Gooding et al (2015) *** ** ** 7 Jahn et al (2016) *** * ** 6 Kim et al (2015) ** * 4 Lakey et al (2014) *** ** ** 7 Lin et al (2015) *** ** ** 7 Mitsui et al (2014) ** * * 4 Nakamura et al (2011) *** * ** 7 Ozturk et al (2006) ** * 3 Palmer et al (2004) *** ** * 6 Plutchik et al (1995) *** * ** 6 Taliaferro et al (2010) **** ** ** 8 Tarrier et al (2004) **** * *** 8 Zhang et al (2010) **** ** ** 8 Zhang et al (2007) *** * 4

One study, Kim et al (2015), referenced an adapted version of the Rosenberg Self-Esteem

Scale (Rosenberg, 1965) in their paper, however, the reference that this version was linked to was unavailable, and, when study authors were contacted to try and ascertain the scoring system or validation, no response was given. As there appeared to be anomalies in the reporting of this data (differences between the in-text reporting of statistics and tables) this paper was left out of the analysis of results.

Measures Used

Twelve of the included studies used the Rosenberg Self-Esteem Scale (RSES, Rosenberg,

1965), which measures implicit and explicit self-esteem. For the majority of these studies, the total RSES score was reported, though one study (Creemers et al, 2012) used only the explicit self-esteem subscale, and used the name letter task (Nuttin, 1987) to measure implicit self-esteem. Other measures of self-esteem included, the Beck Self Esteem Scale, n=

2, (Beck, Brown, Steer, Kuyken & Grisham, 2001), Coopersmith Self Esteem Scale (Myhill and

Lorr, 1978), n=1 and Robson’s Self-Concept Questionnaire n=1. In terms of suicidality measures, these included measures of suicidal ideation (n= 10), current suicidality (n = 6) and suicide probability/risk (n=5). Suicidality measures included the Beck Scale for Suicidal

Ideation (BSS), Suicidal Ideation Questionnaire (SIQ), Positive and Negative Suicidal Ideation

Scale (PANSI), Suicide probability scale (SPS), Suicide Risk Scale, Columbia Suicide Severity

Scale, Suicidal Behaviours Questionnaire (SBQ). Suicide subscales for both the MINI and

GHQ-30 were also included as they have been independently validated for use in suicide research.

The relationship between self-esteem and suicidality

This section details the overall findings of the review. Following this, synthesis of the results with regards the specific relationship of self-esteem to suicidal ideation and behaviours will be undertaken. This is to explore recent findings in the literature that suggest differences exist in the relationship between individual psychological characteristics and suicidal ideation vs suicidal behaviour (Nock et al, 2009).

Eighteen studies examined the relationship between self-esteem and suicidality. All included papers showed a statistically significant negative association between self-esteem and suicidality, apart from two – Ozturk et al (2006) mentioned below and Creemers et al

(2012) who found a non-significant positive association between implicit self-esteem and suicidality.

Eleven studies reported correlational data when examining the relationship between different aspects of self-esteem and suicidal thoughts, behaviour and probability. An overall meta-analysis of the correlational data within the review was undertaken using a random effects IVHET model. This was based on data from 11 studies, see Figure 2 below. The overall effect was a pooled correlation of r=-0.42 (CI -0.54 to -0.29) suggesting a moderate relationship between self-esteem and suicidality with high levels of heterogeneity (i² = 82%).

Within the meta-analysis, in studies where more than one aspect of suicidality or self- esteem was reported on, for example global vs contingent self-esteem or suicidal ideation vs behaviours (Jahn et al, 2016, Lakey et al, 2014, Creemers et al, 2012), a single result was chosen to enter. Results were chosen in line with exclusion criteria or those that were closest to the original definition of self-esteem (that of a global measure of self-worth) in an attempt to reduce heterogeneity in the meta-analysis.

Figure 2 Forest Plot showing the pooled results of all correlational studies examining the relationship between suicidality and self-esteem

Three studies used regression analyses to examine multiple factors in relation to suicidal ideation (Bhar et al, 2008, Zhang et al, 2010, 2007) and one for current suicidality

(Nakamura et al, 2011). All four studies found a significant negative relationship between self-esteem and suicidality. Two studies used ANOVA to examine differences between groups (Palmer, et al 2004, Mitsui et al, 2014). Both Palmer and Mitsui examined 3 groups of participants with varying suicide history (healthy controls, ideation and attempts) and found significant differences between groups in their levels of self-esteem and suicidal risk regardless of group membership.

One small study used a t-test to examine differences in self-esteem between suicidal individuals and healthy controls (Ozturk et al, 2006). There were no significant differences between the two groups overall. However, these results should be interpreted with caution

34 as this study was of very low methodological quality, underpowered and there were no controls on confounding variables.

Creemers et al (2012) used the Name-Letter Task (Nuttin, 1985) to measure implicit self-esteem. This test has been criticised by Krizan (2008) with regards respondents’ awareness of the self-referential nature of the test, the implication being that conscious beliefs may have moderated results. A review by Buhrmester, Blanton and Swan (2011) of implicit self-esteem measures found a lack of consistency of evidence to support their use

(specifically the name-letter task and implicit associations test). Given the concerns about the validity of these measures any conclusions drawn from this study’s results must be made with some caution.

Four studies examined the relationship between suicide probability or risk of suicide and self-esteem. This was measured using the suicide probability scale (n=4) and the suicide risk scale (n=1). Measures of suicide risk or probability encompass items that are associated with increased risk of completed suicide, alongside measurement of suicidal ideation and behaviours. For example, the Suicide Probability Scale has sub scales which include hostility, suicidal ideation, negative self-evaluation and hopelessness (Cull and Gill, 1982).

Three studies reported correlational data when examining suicide probability and self-esteem (Demirbas et al, 2003, Gooding et al 2015, and Plutchik et al, 1995), r= -0.64,

-0.70 and -0.29 respectively. One study used a t-test to compare individuals with a history of suicide attempts against healthy controls. Levels of suicide probability were increased in patients with a history of suicidal attempts vs those without a history of suicidal behaviours

(Dermirbas et al, 2013).

The suicide probability scale (SPS) was utilised in papers by Gooding et al (2015) and

Demirbas et al (2013). Both showed high levels of correlation between suicide probability and low self-esteem. The Suicide Risk Scale (Plutchik et al, 1989) was used by Plutchik et al

(1995), and in a highly powered study with n=685 participants, found a relatively low correlation of r=-0.29. The difference between these results raises the question of whether the relationship between suicide probability and self-esteem has been distorted by the relationship of self-esteem to the subscales of ‘negative self-evaluation’ or ‘hopelessness’ on the SPS. Lower levels of self-esteem have been related to increased negative self- evaluation (Barrowclough et al, 2003) and levels of hopelessness (Tarrier et al, 2004), and these have both been linked to higher levels of suicide risk. In terms of future studies it may be useful to consider reporting the specific relationships between self-esteem and subscales on the SPS, or using a different measure.

Ozturk (2006) also examined the relationship between suicide probability and self- esteem in a very small sample of 24 individuals who were seeking plastic surgery following a suicide attempt by gunshot wound to the face. There were no overall significant differences on suicide probability or self-esteem measure between this group (n=12) and healthy controls (n=12). However, as stated above, this study had serious methodological concerns, with regards power, control of variables and reporting of results.

The relationship between self-esteem and suicidal ideation

In total, ten studies examined the association between self-esteem and suicidal ideation specifically. Six studies used correlational analysis, all results showed a negative relationship when considering explicit or global self-esteem and suicidal ideation. Four studies used regression (1 logistic and 3 multiple) in their analysis and did not report correlational data

(Bhar et al, 2008; Nakamura et al, 2011; Zhang et al, 2010; 2007). Correlations ranged from -

0.36 for Jahn et al (2016) and Creemers et al (2012) to -0.77 for Collette et al (2016). The findings by Collette et al (2016) are based upon a small sample size (n=21 for the clinical sample), and this study received a relatively low score on the Newcastle-Ottowa scale, suggesting higher risk of bias (5/10) so its findings should be interpreted with caution.

Similarly, high r values (r=-0.63) were found in another underpowered study reported by Lin et al, 2015 (n=18). In studies where there were greater participant numbers (590 and 450 participants respectively) and higher scores on the Newcastle-Ottowa Scale (suggesting lower risk of bias), findings suggest a low-moderate correlation of between r= -0.36 and r=-

0.52 (Jahn et al, 2016 and Taliaferro et al, 2010), which demonstrates a moderate negative relationship between self-esteem and suicidal ideation. These results are in keeping with a study by Nakamura et al (2011), who found a negative relationship between self-esteem and suicidal depression symptoms in employees at 3 Japanese companies.

Further examination suggested that the relationship between self-esteem and suicidal ideation remained significant (r=-0.5) when controlling for 2 or more confounder variables, including depression symptoms (Taliaferro et al, 2010). A similar finding was reported by Bhar et al (2008) and Zhang et al (2010) where self-esteem and suicidal ideation continued to have a significant relationship when depression, hopelessness and

37 demographic variables were controlled for (significant at the p< 0.05 and 0.001 levels respectively). However, one study reported a reduced correlation once depression was controlled for, moving from r=-0.5 to r=-0.2 (De Man, 2002). Overall, it would appear that self-esteem is related to suicidal ideation, with higher quality studies suggesting a moderate relationship. This relationship continues when controlling for key diagnostic and demographic variables.

Zhang et al (2007) examined differences in the self-esteem and suicidality levels of students in both China and USA. The study found that USA samples showed a significant negative relationship between self-esteem and suicidal ideation, however only the females in the Chinese sample showed a significant negative relationship between self-esteem and suicidal ideation (correlation coefficients not reported). This reflects a previous finding by the same research group where female prisoners in China demonstrated higher levels of suicidal ideation and lower levels of self-esteem than their male counterparts (Zhang et al,

2010). Further to these findings, gender may be an important variable to consider when examining the relationship between self-esteem and suicidality.

Creemers et al (2012) examined explicit and implicit self-esteem separately in relation to suicidal ideation and found differences in the relationship between implicit and explicit self-esteem and suicidality. A non-significant positive correlation was found between implicit self-esteem and suicidal ideation (r=0.18) and a significant negative relationship between explicit self-esteem and suicidal ideation (r=-0.36). An additional finding of note was that an interaction effect was reported, whereby individuals with both low levels of explicit self-esteem and higher implicit self-esteem showed increased suicidal ideation

(b=0.029, se=0.009, p= <0.001). This study did not control for confounders in the analysis,

38 suggesting a possible additional role for another known correlate of suicidality, such as depression, age or gender.

A random effects meta-analysis was carried out on studies which examined the relationship between self-esteem and suicidal ideation, used correlational data, and the

Rosenberg self-esteem scale (1965) – see Figure 3. A moderate negative pooled correlation was found (r=-0.42 [95% CI: -0.24 to -0.58]) with high levels of heterogeneity (i2 = 85%).

Figure 3 Forest Plot showing the pooled results of studies examining suicidal ideation and using the Rosenberg Self-Esteem scale as their primary outcome measure.

The relationship between suicide attempts and self-esteem

In line with Palmer et al (2004) and Dermirbas et al (2013) above, it would appear that a history of suicidal behaviour can be an important factor in the prediction of future risk of suicide. Three studies reported correlational data on the relationship between suicidal behaviour and self-esteem (Jahn et al, 2016, Tarrier et al, 2004 and Lakey et al, 2014). These correlations reported values of r=-0.19, r=-0.37 and r=-0.35 respectively. All three studies utilised lifetime suicide attempts in terms of timescales, however the study by Lakey et al

(2014) used the Suicidal Behaviours Questionnaire- Revised (SBQ-R). The SBQ refers to suicidal ideation and attempts in the same question “Have you ever thought about or attempted to kill yourself?”. It is therefore difficult to accurately differentiate results between ideators and individuals who have carried out suicidal behaviours. This issue speaks to some of the confusion in the suicide literature, with the ‘suicidal behaviours questionnaire’ potentially describing a group of individuals who have not engaged in suicidal behaviours.

Mitsui et al (2014) found significant group differences between healthy controls, individuals with depression but no suicidal behaviours and those with depression and suicidal behaviours on self-esteem measures, with RSES scores of 34.5, 24.8 and 29.9 respectively, suggesting a potential negative relationship between self-esteem and individuals who engage in suicidal behaviours. However, this study controlled for only 1 confounding variable (depression) and had a relatively small sample size n=45 for entire study, thus n=15 in each group, and thus is at increased risk of bias.

Jahn et al (2016) found that reduced self-esteem was negatively associated with individuals who reported a history of suicide attempts (r=-0.18). This study was well powered (n=590), controlled for a range of suicide risk factors such as depression, social anxiety and personality traits (as scored on the ‘big five inventory’- openness, neuroticism, agreeableness, conscientiousness and extraversion) and scored as being of moderate quality.

One study in the review examined the relative contingency of self-esteem in relation to suicidality. Contingent self-esteem is dependent upon the favourable outcome of events, relationships and social approvals deemed of worth to an individual. Lakey (2014) found

40 that both contingent and non-contingent self-esteem were significantly associated with suicidal behaviours, with r=-0.41 and r=-0.35 respectively. Individuals with high scores for contingent self-esteem and low levels of non-contingent or global self-esteem were classed as having ‘fragile self-esteem’ as it was highly contingent. The relative ‘fragility’ of self- esteem was linked to higher levels of suicidal behaviour. This study was well powered (n=

371) and scored highly in terms of quality, suggesting increased confidence in its findings.

The previous sections have examined the relationship between self-esteem and suicidality in individuals experiencing suicidal ideation vs individuals who carry out suicidal behaviours. This difference was explicitly explored by Palmer et al (2004), who examined the relationship between self-esteem and suicide risk with regards to three groups of depressed inpatients; those with no history of suicidal ideation or attempts, those with just ideation and those with both suicidal ideation and attempts. They found significantly higher suicide risk scores and significantly lower self-esteem scores in the group who had experienced both ideation and at least one previous attempt at suicide, in comparison with sole ideators and with those with no ideation or previous attempts. This supports general trends in suicide literature which suggests that history of suicidal attempts or non-suicidal self-injury can increase future suicide risk (O’Connor and Nock, 2014, BPS, 2017).

Moderation and Mediation

A number of studies examined the moderating and mediating effects of self-esteem and a variety of other factors on suicidal ideation and behaviour. De Man (2002) examined the role of stability of self-esteem when examining the relationship between self-esteem and suicidal ideation. Their moderation analysis found that the negative relationship between self-esteem and suicidal ideation was increased in those with unstable self-esteem

(measured by the standard deviation of RSES scores over 5 timepoints). This may have some relevance to other studies in this review (Creemers et al, 2012, Lakey et al, 2014) which examined the contingency of self-esteem as an important factor implicated in levels of suicidal ideation, which will be discussed below.

Gooding et al (2015) carried out a sub-scale analysis of the relationship between self- esteem and the suicide probability scale and found that the relationship between self- esteem and suicidality was significantly related to the hopelessness component of the SPS, above other sub-scales such as suicidal ideation, negative self-evaluation and hostility. They carried out further analyses to examine the moderating effect of self-esteem on key factors associated with suicidality; defeat and entrapment and found no significant effect. These findings were in keeping with Tarrier et al (2004) who used path analysis to describe the relationship of negative self-evaluation to increased suicide risk via hopelessness.

DISCUSSION

The aim of this study was to examine the contributory role of self-esteem in suicidality in an adult population. The findings suggest with some consistency that there is a negative relationship between levels of self-esteem and different forms of suicidal processes overall.

The pooled correlation of 11 studies examining the relationship between self-esteem and suicidality suggested a moderate relationship, with an r=-0.42 (CI r= -0.54 to -0.29). When only suicidal ideation was examined r =-0.42 (with a range of r= -0.36 to -0.77) there was a moderately significant relationship between self-esteem and suicidal ideation. This finding was consistent across participant groups and geographic region. When suicidal behaviour was specifically examined, although results still demonstrated a consistent negative relationship between self-esteem and suicidal behaviour, generally these correlations were lower (range of r=-0.19 to -0.37). This might suggest that self-esteem has a stronger relationship with suicidal ideation than behaviour. One reason for this may involve the different pathways from suicidal ideation to behaviour (Nock, Kessler, & Franklin, 2016,

Klonsky et al, 2016). Some models of suicidality have highlighted the importance of volitional factors (such as social learning, access to means, history of completed suicide) as the key elements that move individuals from thoughts of suicide, to taking steps to attempt suicide (O’Connor et al, 2013). It may be that self-esteem is an important component of the initial phases of suicidality, but becomes less so as volitional factors begin to play more prominent roles.

Due to the cross-sectional design of the majority of studies it is not possible to infer causal relationships between self-esteem and suicidality more broadly. It may be that there are other factors involved in this relationship such as hopelessness (in line with Gooding et

43 al, 2015 and Tarrier et al, 2004), depression, or that there is a bidirectional relationship where self-esteem and suicidality impact on one another.

One study in the review examined the role of implicit self-esteem in suicidality

(Creemers et al 2002). As such limited conclusions can be drawn from the data. However, it is interesting that this result showed no significant relationship between implicit self-esteem and suicidal ideation. What was also interesting about the result was that although implicit self-esteem was not directly related to suicidal ideation, the relative ‘fragility’ of self-esteem

(e.g. high implicit and low explicit self-esteem) was significantly related to suicidal ideation.

Those who have high implicit self-esteem may have high expectations in terms of their levels of success or achievement in the world, and the presence of low explicit self-esteem suggests that they are not meeting their own personal standards. It is hypothesised that it is this tension which is related to increased suicidal ideation and this would be an interesting area for further study.

One mechanism by which tensions between implicit and explicit self-esteem may be linked to suicidality includes the role of positive future thoughts. It is possible that individuals with high implicit self-esteem have high expectations of themselves in terms of future relationships and success. O’Connor, Smyth and Williams (2015) recently undertook a prospective study which investigated positive future thinking and future suicide risk. Their findings suggested that not all positive future thoughts are protective, with individuals reporting high numbers of positive intrapersonaI thoughts about the future (positive thoughts around being less depressed, more confident, healthier) being at increased risk of suicide. The suggestion is that individuals with unrealistic high expectations of themselves may not achieve what they hoped in terms of personal progress, which may be associated

44 with more intense suicidal thoughts and behaviour (O’Connor et al, 2016). Unrealistic high standards are often a feature of clinical levels of perfectionism, an area which has been linked to increased suicidality (O’Connor, 2007; O’Connor & Noyce, 2008). One possible pathway to increased suicidality may be from perfectionistic personality traits, interacting with fragile or highly contingent self-esteem and increased vulnerabilities to suicidal ideation.

The relative fragility of self-esteem relates to work by Heppner and Kernis around fragile self-esteem (2011), where people with high levels of contingent self-esteem were particularly sensitive to perceived threats. Individuals with ‘fragile’ self-esteem face a constant threat of their self-esteem being damaged by external stressors, such as negative feedback, social rejection, negative aspects of the self, and must remain vigilant to defend any potential loss of self-worth (Rosenberg & Owens, 2001). In this respect it would appear that those with ‘fragile’ self-esteem experience their self-worth as highly contingent on external or internal feedback. Contingent self-esteem was examined in one study within the review (Lakey 2014), and stability of self-esteem in another (De Man et al 2002). In the study by Lakey (2014), contingent self-esteem was described as ‘fragile’ in that it was reliant on uncontrollable external factors. Contingent self-esteem showed a stronger correlation with suicidal ideation than global self-esteem (r=0.41 vs 0.35 both at the p=0.01 level).

These findings may also be linked to the literature examining the relationship between mood instability and suicidality, a meta-analysis by Palmier-Claus, Taylor, Varese and Pratt (2012) suggested a moderate relationship between mood instability and suicidality. In relation to this it could be hypothesised that ‘fragile or unstable’ self-esteem and its subsequent fluctuations is linked to changes in mood state, either as an antecedent

45 or following low mood, which in turn increases vulnerability to suicidal thoughts and behaviours.

Theoretical Implications

There are currently a range of theoretical models which aim to explain the pathway from environmental, physical and psychological factors to increasing levels of suicidal thoughts and behaviours (for a comprehensive review see O’Connor and Nock, 2014). The Integrated

Motivational and Volitional model (O’Connor, 2011) reports the mechanisms by which individuals move from external and internal risk factors to increased suicide risk. This is via an increased sensitivity to markers of defeat and entrapment and key motivational and volitional factors. Within the IMV model it is proposed that self-esteem acts both as a predisposing factor to suicidal thoughts and also works in the motivational phase of suicide, interacting with threat to self moderators to increase feelings of social defeat.

Threat to self moderators (TSM), include dysfunctional self-regulation caused by the failure to disengage from a goal which is not attainable, and re-engage with new goals

(O’Connor et al, 2016; Wrosch, Scheier, Miller, Schultz, & Carver, 2003). Self-esteem may have a role to play in this moderation process, and has been considered in relation to persistence in a losing strategy when gambling (Zhang & Baumeister, 2006). Individuals with threatened self-esteem were more likely to invest in losing endeavours, potentially causing increases in entrapment. One of the possible mechanisms of self-esteem’s relationship with suicidality may be that if self-esteem is threatened, which is more likely in those with fragile or highly contingent self-esteem, individuals then experience an inability to disengage from

46 situations which are causing them harm. This may increase feelings of entrapment and go on to increase the likelihood of suicidal behaviour.

The literature on low self-esteem causing a persistence in losing strategies would support the Schematic Appraisal model of Suicide (Johnson et al, 2008). In this model suicide risk is increased in individuals with low self-esteem, defined as a combination of negative self-appraisals, low sense of personal agency and poor problem-solving abilities.

Self-esteem in this context could contribute to suicidality by causing a deficit in both agency and problem solving, leading to increased feelings of defeat and entrapment and increasing levels of suicidality.

Clinical Implications

Given the finding that low self-esteem has shown a consistent negative association with suicidal thoughts, risk and behaviour it may be important for clinicians to consider increased risk of suicide in individuals also displaying low levels of self-esteem. However, self-esteem alone is unlikely to be an accurate predictor of increased risk of suicide, so assessment of levels of defeat, entrapment and hopelessness will be important alongside any assessment of low self-esteem.

As mentioned above within the IMV model of suicide, motivational moderators of suicidal behaviour may include difficulty with disengaging or re-engaging with goals in individuals who are at risk of suicide. Individuals with low self-esteem may be at particular risk of impaired self-regulation via goal disengagement/re-engagement (though this is based on limited evidence at this time). One possible avenue to address this may be for clinicians

47 to utilise problem solving. Problem solving training has been successfully used on a number of occasions with individuals at risk of suicide (Barnes et al, 2017; Pollock and Williams, 2004 and Gyongyi et al, 2012). It may be that interventions aimed at increasing problem-solving ability, particularly social problem-solving ability, may attenuate potential issues with regards those with low self-esteem sticking with a ‘losing strategy’ and experiencing increased feelings of entrapment. Similarly, approaches such as ‘Method of Levels’ (Carey,

2006), which aim to broaden a service users perspective on their difficulties may be particularly useful for individuals at risk of suicide who exhibit low levels of self-esteem.

Recent psychological literature has demonstrated that cognitive behavioural interventions to improve self-esteem have positively impacted on participants low self- esteem, depression and anxiety levels (Waite, McManus & Shafran, 2012; Pack & Condren,

2014). Thus, tailoring psychological interventions to target self-esteem as a suicide prevention treatment may also be a valid option.

Future Directions

From the results described in this review it would appear that there is a negative relationship between self-esteem and suicide probability, ideation and to a lesser extent behaviour. In their comprehensive review of the suicide literature, Franklin et al (2017) examined 50 years of suicide research data. Their findings suggest that in order to move from simple correlations, with the ensuing difficulties around causality, longitudinal research would be required. This would allow associations to be developed into clinically useful risk factors. As this review has found, very few longitudinal studies are associated with the self-esteem/ suicide literature (indeed Franklin et al did not examine self-esteem in

48 their review for this very reason). If further work is required on examining the relationship between self-esteem and suicide then longitudinal studies are essential.

Another key finding of the review is reflected in the self-esteem literature more generally; a lack of consensus over definitions of self-esteem and the most appropriate tools to use in its measurement.

Global self-esteem is broadly considered to be a ‘trait’ construct, relatively stable throughout a person’s life and potentially related to inherited personality characteristics and temperament (Neiss, Sedikides & Stevenson, 2002; Brown & Marshall, 2006). Global self-esteem is a well-researched concept with high face validity; research has linked high levels of global self-esteem to psychological wellbeing, happiness and in some cases job performance (Rosenberg, Schooler, Shoenback & Rosenberg, 1995; Baumeister, Campbell,

Krueger, & Vohs, 2003). Measures of global self-esteem are found throughout the self- esteem literature. Rosenberg’s (1965) self-esteem scale is the most commonly used measure of global self-esteem found in the psychological literature, and within this review

(12/18 studies).

Heatherton and Wyland (2003) updated a review by Blascovich & Tomaka (1991) and reflected concerns with the measurement and conceptualisation of self-esteem. Rather than measures of global self-esteem, they recommended an amended version of the Feelings of

Inadequacy scale (Janis and Field, 1959) which encompasses the social, performance and physical aspects of self-esteem found in the broader literature (Heatherton & Polivy, 1991;

Baumeister et al, 2003).

One future avenue of research may involve consideration of specific aspects of self- esteem, alongside the more widely used ‘global’ self-esteem concept. This might include,

49 measurement of implicit vs explicit or levels of contingency of self-esteem. This may provide a more accurate picture of self-esteems relationship to suicidality, and allow for targeted interventions to be derived. It would also be of interest to examine the relationship of perfectionism to self-esteem and how this trait was related to vulnerabilities to both lowered self-esteem and levels of suicidal ideation.

There are a range of scales used for assessing different elements of suicidal behaviour and thoughts. Some focus on suicidal ideation (for example the Scale for Suicidal

Ideation, Beck, Kovacs & Weissman, 1979), others on suicide risk/ probability (for example the Suicide Probability Scale, Cull and Gill, 1988) and others on specific suicidal behaviours

(such as the SBQ-R, Osman, et al, 2001). Winters, Myers & Proud (2002) in a 10 year review of suicide scales found that suicide questionnaires had clear constructs with regards suicidal ideation, but felt that reliability and validity could diminish over time due to the time and attitude-specific nature of some of the questions. This was supported by the findings in a review by Ghasemi, Shaghaghi & Allahverdipour (2015). Future recommendations would include more precise measurement of the different aspects of suicide – ideation, behaviour and risk -and clear definition of the suicidal processes being investigated. There is some difficulty in combining literature due to inconsistencies in the way suicidal thoughts, behaviours and risk is being measured. More specificity over the timescales being considered would also be useful, for example current suicidal ideation and attempts vs lifetime suicide attempts, or within a fixed historical period.

As the methodological quality of many of the studies reviewed was low-moderate, particularly with a view to sample size and power calculations, one future recommendation would be to ensure studies are sufficiently powered to avoid participant bias.

In a review by Franklin et al (2017) one recommendation was to ensure that risk factors for suicidality should be considered not only in terms of their statistical significance but also clinical significance. In many of the reviewed studies a statistically significant difference between groups might account for only 5 points on a RSES, which may not amount to a clinically significant impact from a participants’ perspective. There is no clear guidance on what constitutes a clinically significant change on the RSES, however, 10-15 points separate the low, moderate and high self-esteem categories respectively.

There have also been reflections that self-esteem measures often have a consistent pattern, with most people rating themselves at a moderate to high level, possibly due to issues around perceptions of social desirability (Heatherton and Wyland, 2003), so emerging literature on the use of implicit measures may provide a useful opportunity to circumnavigate this issue.

Limitations of study

Commonly used scales in self-esteem, such as the Rosenberg Self Esteem Scale

(1965) or the Coopersmith self-esteem scale (1967), generally considered reliable and valid in a clinical sense, are thought to measure enduring, global ‘trait’ self-esteem. However, there is some debate as to the accuracy of explicit measures and their explanation of self- esteem. Indeed, some studies have reported that individuals with narcissism, conceitedness or grandiose beliefs about the self can score highly on explicit self-esteem measures

(Campbell, Rudich & Sedikides, 2002, Heppner and Kernis, 2011). This calls into question some of the results using global self-esteem measures and whether they measure true feelings of worth.

Inconsistencies in the control of confounding variables is also a limitation in terms of the studies within the review. Many studies did not control for some of the most important factors associated with completed suicidality such as gender, age, hopelessness and depression. The impact of these known correlates on the relationship between self-esteem and suicidal ideation is mixed, with some studies reporting significant results when depression and hopelessness were controlled for (Bhar et al, 2008 and Zhang et al, 2010), others reported a drop in the relationship between suicidal ideation and self-esteem when the effect of depression was controlled for (though the relationship was still statistically significant).

The exclusion of papers with a sole focus on non-suicidal self-injury (NSSI) may also be considered as a potential limitation within the study. Recent taxometric research has suggested that suicidal self-injury and NSSI may exist on a continuum (Orlando et al, 2015).

Clinicians working with individuals who self-injure may need to be aware that increased frequency and methods of self-injury have been associated with increased suicide risk

(Whitlock et al, 2013; Csorba, Dinya, Plener, Nag and Pali, 2009). On this basis it may be that exclusion of studies with a specific NSSI focus may have limited the scope of the review, and thus excluded relevant findings.

Historically it has been found that many non-significant studies are not published, a review by Ross et al (2012), of clinicaltrials.com found that only 46% of registered studies had reported results, other figures have ranged from only 22% (Prayl, Hurley & Smyth, 2012) to 71% of large trials (Jones et al, 2013). It may be that this review, in focusing on published literature, has been subjected to publication bias and as such is at increased risk of reporting significant findings.

Conclusion

This is the first review to have examined the role of self-esteem in suicidality, in an adult population. Overall, methodological issues limit the strength of the findings that can be made, however it does appear that there is a consistent negative relationship between self- esteem and suicidal thoughts and behaviour. Issues with definitions and reporting in the literature effect both self-esteem and suicide research. Future recommendations would include developing consistent guidelines around the assessment and reporting of these factors. There is also further research to be carried out exploring a more heterogeneous understanding of self-esteem and its relation to suicidality, with particular reference to its relative contingency and relationship with entrapment.

REFERENCES

Addington, J., Shah, H., Liu, L., & Addington, D. (2014). Reliability and Validity of the Calgary Depression Scale for Schizophrenia (CDSS) in Youth at Clinical High Risk for Psychosis. Schizophrenia Research, 153(0), 64–67. https://doi.org/10.1016/j.schres.2013.12.014

Barendregt, J. J. (2010). The Effect Size in Uncertainty Analysis. Value in Health, 13(4), 388–391. https://doi.org/10.1111/j.1524-4733.2009.00686.x

Barnes, S. M., Monteith, L. L., Gerard, G. R., Hoffberg, A. S., Homaifar, B. Y., & Brenner, L. A. (2017). Problem-Solving Therapy for Suicide Prevention in Veterans With Moderate-to-Severe Traumatic Brain Injury. Rehabilitation Psychology. https://doi.org/10.1037/rep0000154

Baumeister, R. F. (1990). Suicide as escape from self. Psychological Review, 97(1), 90–113.

Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003b). Does High Self-Esteem Cause Better Performance, Interpersonal Success, Happiness, or Healthier Lifestyles? Psychological Science in the Public Interest: A Journal of the American Psychological Society, 4(1), 1–44. https://doi.org/10.1111/1529-1006.01431

Beck, A. T., Brown, G. K., Steer, R. A., Kuyken, W., & Grisham, J. (2001). Psychometric properties of the Beck self-esteem scales. Behaviour Research and Therapy, 39(1), 115–124.

Beck, A. T., Kovacs, M., & Weissman, A. (1979). Assessment of suicidal intention: the Scale for Suicide Ideation. Journal of Consulting and Clinical Psychology, 47(2), 343–352.

Beevers, C. G., & Miller, I. W. (2004). Perfectionism, Cognitive Bias, and Hopelessness as Prospective Predictors of Suicidal Ideation. Suicide and Life-Threatening Behavior, 34(2), 126–137. https://doi.org/10.1521/suli.34.2.126.32791

Beyer, J. L., & Weisler, R. H. (2016). Suicide Behaviors in Bipolar Disorder: A Review and Update for the Clinician. The Psychiatric Clinics of North America, 39(1), 111–123. https://doi.org/10.1016/j.psc.2015.09.002

Bhar, S., Ghahramanlou-Holloway, M., Brown, G., & Beck, A. T. (2008). Self-Esteem and Suicide Ideation in Psychiatric Outpatients. Suicide and Life-Threatening Behavior, 38(5), 511–516.

Blascovich, J., & Tomaka, J. (1991). Measures of self-esteem. In J. P. Robinson, P. R. Shaver, & L. S. Wrightsman (Eds.), Measures of social psychological attitudes, Vol. 1. Measures of personality and social psychological attitudes (pp. 115-160). http://dx.doi.org/10.1016/B978-0-12-590241-0.50008-3

Borges, G., Angst, J., Nock, M. K., Ruscio, A. M., & Kessler, R. C. (2008). Risk factors for the incidence and persistence of suicide-related outcomes: A 10-year follow-up study using the National Comorbidity Surveys. Journal of Affective Disorders, 105(1–3), 25–33. https://doi.org/10.1016/j.jad.2007.01.036

Brezo, J., Paris, J., Tremblay, R., Vitaro, F., Zoccolillo, M., Hébert, M., & Turecki, G. (2006). Personality traits as correlates of suicide attempts and suicidal ideation in young adults. Psychological Medicine, 36(2), 191–202. https://doi.org/10.1017/S0033291705006719

Buhrmester, M. D., Blanton, H., & Swann, W. B. (2011). Implicit self-esteem: nature, measurement, and a new way forward. Journal of Personality and Social Psychology, 100(2), 365–385. https://doi.org/10.1037/a0021341

Cambron, M. J., Acitelli, L. K., & Pettit, J. W. (2009). Explaining Gender Differences in Depression: an Interpersonal Contingent Self-Esteem Perspective. Sex Roles, 61(11–12), 751. https://doi.org/10.1007/s11199-009-9616-6

Cambron, M. J., Acitelli, L. K., & Steinberg, L. (2010a). When Friends Make You Blue: The Role of Friendship Contingent Self-Esteem in Predicting Self-Esteem and Depressive Symptoms. Personality and Social Psychology Bulletin, 36(3), 384–397. https://doi.org/10.1177/0146167209351593

Campbell, W. K., Rudich, E. A., & Sedikides, C. (2002). Narcissism, self-esteem, and the positivity of self- views: Two portraits of self-love. Personality and Social Psychology Bulletin, 28(3), 358–368.

Carey, T. A. (2006). The Method of Levels: How to do psychotherapy without getting in the way. Hayward, CA: Living Control Systems Publishing.

Chatard, A., Selimbegović, L., & Konan, P. N. (2009). Self-esteem and suicide rates in 55 Nations. European Journal of Personality, 23(1), 19–32. https://doi.org/10.1002/per.701

Chen, S. X., Wu, W. C. H., & Bond, M. H. (2009). Linking family dysfunction to suicidal ideation: Mediating roles of self-views and world-views. Asian Journal of Social Psychology, 12(2), 133–144. https://doi.org/10.1111/j.1467-839X.2009.01280.x

Collett, N., Pugh, K., Waite, F., & Freeman, D. (2016). Negative cognitions about the self in patients with persecutory delusions: An empirical study of self-compassion, self-stigma, schematic beliefs, self- esteem, fear of madness, and suicidal ideation. Psychiatry Research, 239, 79–84. https://doi.org/10.1016/j.psychres.2016.02.043

Creemers, D. H. M., Scholte, R. H. J., Engels, R. C. M. E., Prinstein, M. J., & Wiers, R. W. (2012). Implicit and explicit self-esteem as concurrent predictors of suicidal ideation, depressive symptoms, and loneliness. Journal of Behavior Therapy and Experimental Psychiatry, 43(1), 638–646. https://doi.org/10.1016/j.jbtep.2011.09.006

Crocker, J. (2002). Contingencies of self-worth: Implications for self-regulation and psychological vulnerability. Self and Identity, 1, 143–149.

Csorba, J., Dinya, E., Plener, P., Nagy, E., & Pali, E. (2009) Clinical diagnoses, characteristics of risk behavior, differences between suicidal and non-suicidal subgroups of Hungarian adolescent outpatients practising self- injury. European Child Adolescent Psychiatry, 18, 309–320. http://dx.doi.org/10.1007/s00787-008-0733-5

Cull, J. G., & Gill, W. S. (1988). Suicide probability scale (SPS) manual. Los Angeles: Western Psychological Services.

Curtin SC, Warner M, Hedegaard H. (2016) Increase in suicide in the United States, 1999–2014. NCHS data brief, no 241. Hyattsville, MD: National Center for Health Statistics.

De Crescenzo, F., Serra, G., Maisto, F., Uchida, M., Woodworth, H., Casini, M. P., … Vicari, S. (2017). Suicide Attempts in Juvenile Bipolar Versus Major Depressive Disorders: Systematic Review and Meta-Analysis. Journal of the American Academy of Child and Adolescent Psychiatry, 56(10), 825– 831.e3. https://doi.org/10.1016/j.jaac.2017.07.783

De Man, A. F., & Gutierrez, B. I. (2002). The relationship between level of self-esteem and suicidal ideation with stability of self-esteem as moderator. Canadian Journal of Behavioural Science/Revue Canadienne Des Sciences Du Comportement, 34(4), 235.

De Ruiter, N. M. P., Van Geert, P. L. C., & Kunnen, E. S. (2017). Explaining the “how” of self-esteem development: The self-organizing self-esteem model. Review of General Psychology, 21(1), 49–68. https://doi.org/10.1037/gpr0000099

Demirbas, H., Celk, S., Ilhan, I. O., & Dogan, Y. B. (2003). AN EXAMINATION OF SUICIDE PROBABILITY IN ALCOHOLIC IN-PATIENTS. Alcohol and Alcoholism, 38(1), 67–70. https://doi.org/10.1093/alcalc/agg019

DerSimonian, R., & Laird, N. (1986). Meta-analysis in clinical trials. Controlled Clinical Trials, 7(3), 177–188.

Devenish, B., Berk, L., & Lewis, A. J. (2016). The treatment of suicidality in adolescents by psychosocial interventions for depression: A systematic literature review. The Australian and New Zealand Journal of Psychiatry, 50(8), 726–740. https://doi.org/10.1177/0004867415627374

Donker, T., Calear, A., Busby Grant, J., van Spijker, B., Fenton, K., Hehir, K. K., … Christensen, H. (2013). Suicide prevention in schizophrenia spectrum disorders and psychosis: a systematic review. BMC Psychology, 1(1). https://doi.org/10.1186/2050-7283-1-6

Emler, N. (2001). Self-esteem: the costs and causes of low self-worth. Layerthorpe: Publ. for the Joseph Rowntree Foundation by York Publ. Services.

Forrester, R. L., Slater, H., Jomar, K., Mitzman, S., & Taylor, P. J. (2017). Self-esteem and non-suicidal self- injury in adulthood: A systematic review. Journal of Affective Disorders, 221, 172–183. https://doi.org/10.1016/j.jad.2017.06.027

Franklin, J. C., Ribeiro, J. D., Fox, K. R., Bentley, K. H., Kleiman, E. M., Huang, X., … Nock, M. K. (2017). Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research. Psychological Bulletin, 143(2), 187–232. https://doi.org/10.1037/bul0000084

Ghasemi, P., Shaghaghi, A., & Allahverdipour, H. (2015). Measurement Scales of Suicidal Ideation and Attitudes: A Systematic Review Article. Health Promotion Perspectives, 5(3), 156–168. https://doi.org/10.15171/hpp.2015.019

Gilbert, P., & Allan, S. (1998). The role of defeat and entrapment (arrested flight) in depression: an exploration of an evolutionary view. Psychological Medicine, 28(3), 585–598.

Glassman, L. H., Weierich, M. R., Hooley, J. M., Deliberto, T. L., & Nock, M. K. (2007a). Child Maltreatment, Non-Suicidal Self-Injury, and the Mediating Role of Self-Criticism. Behaviour Research and Therapy, 45(10), 2483–2490. https://doi.org/10.1016/j.brat.2007.04.002

Glassman, L. H., Weierich, M. R., Hooley, J. M., Deliberto, T. L., & Nock, M. K. (2007b). Child Maltreatment, Non-Suicidal Self-Injury, and the Mediating Role of Self-Criticism. Behaviour Research and Therapy, 45(10), 2483–2490. https://doi.org/10.1016/j.brat.2007.04.002

Gooding, P., Tarrier, N., Dunn, G., Shaw, J., Awenat, Y., Ulph, F., & Pratt, D. (2015). The moderating effects of coping and self-esteem on the relationship between defeat, entrapment and suicidality in a

sample of prisoners at high risk of suicide. European Psychiatry, 30(8), 988–994. https://doi.org/10.1016/j.eurpsy.2015.09.002

Griffiths, L. J., Parsons, T. J., & Hill, A. J. (2010). Self-esteem and quality of life in obese children and adolescents: a systematic review. International Journal of Pediatric Obesity: IJPO: An Official Journal of the International Association for the Study of Obesity, 5(4), 282–304. https://doi.org/10.3109/17477160903473697

Gyöngyi et al (2012). An effective method of therapy in suicide prevention: problem solving training in the clinical population. Psychiatria Hungarica : A Magyar Pszichiátriai Társaság tudományos folyóirata, 27 (2):92

Hawton, K., Bergen, H., Kapur, N., Cooper, J., Steeg, S., Ness, J., & Waters, K. (2012). Repetition of self- harm and suicide following self-harm in children and adolescents: findings from the Multicentre Study of Self-harm in England. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 53(12), 1212–1219. https://doi.org/10.1111/j.1469-7610.2012.02559.x

Hawton, K., Casañas I Comabella, C., Haw, C., & Saunders, K. (2013). Risk factors for suicide in individuals with depression: a systematic review. Journal of Affective Disorders, 147(1–3), 17–28. https://doi.org/10.1016/j.jad.2013.01.004

Hawton, K., Harriss, L., Casey, D., Simkin, S., Harrison, K., Bray, I., & Blatchley, N. (2009). Self-harm in UK armed forces personnel: descriptive and case–control study of general hospital presentations. The British Journal of Psychiatry, 194(3), 266–272. https://doi.org/10.1192/bjp.bp.107.048595

Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for measuring state self- esteem. Journal of Personality and Social Psychology, 60(6), 895-910. http://dx.doi.org/10.1037/0022-3514.60.6.895

Heatherton, T. F., & Wyland, C. L. (2003). Assessing self-esteem. In S. J. Lopez & C. R. Snyder (Eds.), Positive psychological assessment: A handbook of models and measures (pp. 219-233). http://dx.doi.org/10.1037/10612-014

Heppner, W. L., & Kernis, M. H. (2011). High Self-Esteem: Multiple Forms and Their Outcomes. In S. J. Schwartz, K. Luyckx, & V. L. Vignoles (Eds.), Handbook of Identity Theory and Research (pp. 329–355). Springer Science+Business Media.

Jahn, D. R., DeVylder, J. E., & Hilimire, M. R. (2016). Explanatory risk factors in the relations between schizotypy and indicators of suicide risk. Psychiatry Research, 238, 68–73. https://doi.org/10.1016/j.psychres.2016.02.021

Janis, I. S., & Field, P. B. (1959). A behavioral assessment of persuasibility: Consistence of individual differences. In C. I. Hovland & I. L. Janis (Eds.), Personality and persuasibility. New Haven, CT: Yale University Press.

Johnson, J., Gooding, P., & Tarrier, N. (2008). Suicide risk in schizophrenia: Explanatory models and clinical implications, The Schematic Appraisal Model of Suicide (SAMS). Psychology and Psychotherapy: Theory, Research and Practice, 81(1), 55–77. https://doi.org/10.1348/147608307X244996

Joiner, T. (2007). Why People Die by Suicide (1 edition). Cambridge, Mass.: Harvard University Press.

Jones, C. W., Handler, L., Crowell, K. E., Keil, L. G., Weaver, M. A., & Platts-Mills, T. F. (2013). Non- publication of large randomized clinical trials: cross sectional analysis. BMJ (Clinical Research Ed.), 347, f6104.

Kesting, M.-L., & Lincoln, T. M. (2013). The relevance of self-esteem and self-schemas to persecutory delusions: a systematic review. Comprehensive Psychiatry, 54(7), 766–789. https://doi.org/10.1016/j.comppsych.2013.03.002

Kleiman, E. M., & Riskind, J. H. (2013). Utilized social support and self-esteem mediate the relationship between perceived social support and suicide ideation. A test of a multiple mediator model. Crisis, 34(1), 42–49. https://doi.org/10.1027/0227-5910/a000159

Klonsky, E. D., May, A. M., & Saffer, B. Y. (2016). Suicide, Suicide Attempts, and Suicidal Ideation. Annual Review of Clinical Psychology, 12(1), 307–330. https://doi.org/10.1146/annurev-clinpsy-021815- 093204

Krizan, Z. (2008). What is implicit about implicit self-esteem? Journal of Research in Personality, 42(6), 1635–1640. https://doi.org/10.1016/j.jrp.2008.07.002

Lakey, C. E., Hirsch, J. K., Nelson, L. A., & Nsamenang, S. A. (2014). Effects of Contingent Self-Esteem on Depressive Symptoms and Suicidal Behavior. Death Studies, 38(9), 563–570. https://doi.org/10.1080/07481187.2013.809035

Leary, M. R., & Baumeister, R. F. (2000). The nature and function of self-esteem: Sociometer theory. In Advances in Experimental Social Psychology (Vol. 32, pp. 1–62). Academic Press. https://doi.org/10.1016/S0065-2601(00)80003-9

Lin, C.-C. (2015). The relationships among gratitude, self-esteem, depression, and suicidal ideation among undergraduate students. Scandinavian Journal of Psychology, 56(6), 700–707. https://doi.org/10.1111/sjop.12252

McGinty, J., Sayeed Haque, M., & Upthegrove, R. (2017). Depression during first episode psychosis and subsequent suicide risk: A systematic review and meta-analysis of longitudinal studies. Schizophrenia Research. https://doi.org/10.1016/j.schres.2017.09.040

Mitsui, N., Asakura, S., Shimizu, Y., Fujii, Y., Toyomaki, A., Kako, Y., … Kusumi, I. (2014). The association between suicide risk and self-esteem in Japanese university students with major depressive episodes of major depressive disorder. Neuropsychiatric Disease and Treatment, 811. https://doi.org/10.2147/NDT.S59349

Myhill, J., & Lorr, M. (1978). The Coopersmith Self-Esteem Inventory: analysis and partial validation of a modified adult form. Journal of Clinical Psychology, 34(1), 72–76.

Nock, M. K., Borges, G., Bromet, E. J., Cha, C. B., Kessler, R. C., & Lee, S. (2008). Suicide and Suicidal Behavior. Epidemiologic Reviews, 30(1), 133–154. https://doi.org/10.1093/epirev/mxn002

Nock, M. K., Hwang, I., Sampson, N., Kessler, R. C., Angermeyer, M., Beautrais, A., … Williams, D. R. (2009). Cross- National Analysis of the Associations among Mental Disorders and Suicidal Behavior: Findings from the WHO World Mental Health Surveys. PLOS Medicine, 6(8), e1000123. https://doi.org/10.1371/journal.pmed.1000123

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(3), 300. https://doi.org/10.1001/2013.jamapsychiatry.55

Nock, M. K., Kessler, R. C., & Franklin, J. C. (2016). Risk Factors for Suicide Ideation Differ From Those for the Transition to Suicide Attempt: The Importance of Creativity, Rigor, and Urgency in Suicide Research. Clinical Psychology: Science and Practice, 23(1), 31–34. https://doi.org/10.1111/cpsp.12133

Nuttin, J. M. (1987). Affective consequences of mere ownership: The name letter effect in twelve European languages. European Journal of Social Psychology, 17(4), 381–402. https://doi.org/10.1002/ejsp.2420170402

O’Connor, R. C. (2007). The relations between perfectionism and suicidality: a systematic review. Suicide & Life-Threatening Behavior, 37(6), 698–714. https://doi.org/10.1521/suli.2007.37.6.698

O’Connor, R. C. (2011). Towards an integrated motivational–volitional model of suicidal behaviour. International Handbook of Suicide Prevention: Research, Policy and Practice, 181–198.

O’Connor, R. C., Cleare, S., Eschle, S., Wetherall, K., & Kirtley, O. J. (2016). The Integrated Motivational- Volitional Model of Suicidal Behavior. In R. C. O’Connor & J. Pirkis (Eds.), The International Handbook of Suicide Prevention (pp. 220–240). John Wiley & Sons, Ltd. https://doi.org/10.1002/9781118903223.ch13

O’Connor, R. C., & Nock, M. K. (2014). The psychology of suicidal behaviour. The Lancet Psychiatry, 1(1), 73–85. https://doi.org/10.1016/S2215-0366(14)70222-6

O’Connor, R. C., Smyth, R., Ferguson, E., Ryan, C., & Williams, J. M. G. (2013). Psychological Processes and Repeat Suicidal Behavior: A Four-Year Prospective Study. Journal of Consulting and Clinical Psychology, 81(6), 1137–1143. https://doi.org/10.1037/a0033751

O’Connor, R. C., Smyth, R., & Williams, J. M. G. (2015). Intrapersonal Positive Future Thinking Predicts Repeat Suicide Attempts in Hospital-Treated Suicide Attempters. Journal of Consulting and Clinical Psychology, 83(1), 169–176. https://doi.org/10.1037/a0037846

Office of National Statistics (2015) Suicides in the UK: 2015 registrations. London, ONS.

Orlando, C.M, Broman-Fulks, J, Whitlock, J, Curtin, L & Michael, K.D (2015) Nonsuicidal Self-Injury and Suicidal Self-Injury: A Taxometric Investigation. Behaviour Therapy, Volume 46, Issue 6, November 2015, Pages 824-833

Osman, A., Bagge, C. L., Gutierrez, P. M., Konick, L. C., Kopper, B. A., & Barrios, F. X. (2001). The Suicidal Behaviors Questionnaire-Revised (SBQ-R): validation with clinical and nonclinical samples. Assessment, 8(4), 443–454. https://doi.org/10.1177/107319110100800409

Öztürk, S., Bozkurt, A., Durmus, M., Deveci, M., & Sengezer, M. (2006). Psychiatric analysis of suicide attempt subjects due to maxillofacial gunshot. Journal of Craniofacial Surgery, 17(6), 1072–1075.

Pack, S., & Condren, E. (2014). An evaluation of group cognitive behaviour therapy for low self-esteem in primary care. The Cognitive Behaviour Therapist, 7. https://doi.org/10.1017/S1754470X14000051

Palmier-Claus, J. E., Taylor, P. J., Varese, F., & Pratt, D. (2012). Does unstable mood increase risk of suicide? Theory, research and practice. Journal of Affective Disorders, 143(1), 5–15. https://doi.org/10.1016/j.jad.2012.05.030

Palmer Jr, C. J. (2004). Suicide attempt history, self-esteem, and suicide risk in a sample of 116 depressed voluntary inpatients. Psychological Reports, 95(3_suppl), 1092–1094.

Panagioti, M., Gooding, P. A., Pratt, D., & Tarrier, N. (2015). An empirical investigation of suicide schemas in individuals with Posttraumatic Stress Disorder. Psychiatry Research, 227(2–3), 302–308. https://doi.org/10.1016/j.psychres.2015.02.019

Plutchik, R., Botsis, A. J., & Van Praag, H. M. (1995). Psychopathology, self-esteem, sexual and ego functions as correlates of suicide and violence risk. Archives of Suicide Research, 1(1), 27–38. https://doi.org/10.1080/13811119508258972

Pollock, L. R., & Williams, J. M. G. (2004). Problem-solving in suicide attempters. Psychological Medicine, 34(1), 163–167.

Prayle, A. P., Hurley, M. N., & Smyth, A. R. (2012). Compliance with mandatory reporting of clinical trial results on ClinicalTrials.gov: cross sectional study. BMJ, 344, d7373. https://doi.org/10.1136/bmj.d7373

Kim Lee, N. Y. K., & Ho Lim, M. (2015). Suicidal Idea, ADHD, Depression, Anxiety, Self-esteem and Impulsiveness in Korean Soldiers. Journal of Psychiatry, 18(5). https://doi.org/10.4172/2378- 5756.1000314

Randal, C., Pratt, D., & Bucci, S. (2015). Mindfulness and Self-esteem: A Systematic Review. Mindfulness, 6(6), 1366–1378. https://doi.org/10.1007/s12671-015-0407-6

Rice, K. G., Ashby, J. S., & Slaney, R. B. (1998). Self-esteem as a mediator between perfectionism and depression: A structural equations analysis. Journal of Counseling Psychology, 45, 304–314.

Rosenberg, M. (1979). Conceiving the self. Basic Books.

Rosenberg, M. (2015). Society and the Adolescent Self-Image. Princeton: Princeton University Press.

Rosenberg, M., & Owens, T.J. (2001). Low self-esteem people: A collective portrait. In T.J. Owens. S. Stryker, & N. Goodmanm (Eds.), Extending self-esteem theory and research (pp. 400-436). New York: Cambridge University Press.

Rosenberg, M., Schooler, C., Schoenbach, C., & Rosenberg, F. (1995). Global Self-Esteem and Specific Self- Esteem: Different Concepts, Different Outcomes. American Sociological Review,60(1), 141-156.

Ross, J. S., Tse, T., Zarin, D. A., Xu, H., Zhou, L., & Krumholz, H. M. (2012). Publication of NIH funded trials registered in ClinicalTrials.gov: cross sectional analysis. BMJ, 344, d7292. https://doi.org/10.1136/bmj.d7292

Rotenstein, L. S., Ramos, M. A., Torre, M., Segal, J. B., Peluso, M. J., Guille, C., … Mata, D. A. (2016). Prevalence of Depression, Depressive Symptoms, and Suicidal Ideation Among Medical Students: A Systematic Review and Meta-Analysis. JAMA, 316(21), 2214–2236. https://doi.org/10.1001/jama.2016.17324

Saxena, S., Krug, E. G., Chestnov, O., & World Health Organization (Eds.). (2014). Preventing suicide: a global imperative. Geneva: World Health Organization.

Schwartz, S. J., Luyckx, K., & Vignoles, V. L. (Eds.). (2011). Handbook of Identity Theory and Research. New York, NY: Springer New York. https://doi.org/10.1007/978-1-4419-7988-9 Taliaferro, L. A., Rienzo, B. A., & Donovan, K. A. (2010). Relationships between youth sport participation and selected health risk behaviors from 1999 to 2007. The Journal of School Health, 80(8), 399–410. https://doi.org/10.1111/j.1746-1561.2010.00520.x

Taliaferro, L. A., Rienzo, B. A., Miller, D. M., Pigg, M. R., & Dodd, V. J. (2010). Potential Mediating Pathways Through Which Sports Participation Relates to Reduced Risk of Suicidal Ideation. Research Quarterly for Exercise and Sport, 81(3), 328–339. https://doi.org/10.1080/02701367.2010.10599681

Tarrier, N., Barrowclough, C., Andrews, B., & Gregg, L. (2004). Risk of non-fatal suicide ideation and behaviour in recent onset schizophrenia: The influence of clinical, social, self-esteem and demographic factors. Social Psychiatry and Psychiatric Epidemiology, 39(11), 927–937. https://doi.org/10.1007/s00127-004-0828-3

Taylor, P.J, Jomar, K, Dhingra, K, Forrester, R, Shahmalak, U, Dickson, J (2018) A meta-analysis of the prevalence of different functions of non-suicidal self-injury. Journal of Affective Disorders, February 2018 Volume 227, Pages 759–769

Torres, R., & Fernández, F. (1995). Self-esteem and value of health as determinants of adolescent health behavior. Journal of Adolescent Health, 16(1), 60–63. https://doi.org/10.1016/1054-139X(94)00045- G

Victor, S. E., & Klonsky, E. D. (2014a). Correlates of suicide attempts among self-injurers: a meta-analysis. Clinical Psychology Review, 34(4), 282–297. https://doi.org/10.1016/j.cpr.2014.03.005

Waite, P., McManus, F., & Shafran, R. (2012). Cognitive behaviour therapy for low self-esteem: a preliminary randomized controlled trial in a primary care setting. Journal of Behavior Therapy and Experimental Psychiatry, 43(4), 1049–1057. https://doi.org/10.1016/j.jbtep.2012.04.006

Walker, J. S., & Bright, J. A. (2009). False inflated self-esteem and violence: a systematic review and cognitive model. The Journal of Forensic Psychiatry & Psychology, 20(1), 1–32. https://doi.org/10.1080/14789940701656808

Wells, G., Shea, B., O’Connell, D., Peterson, je, Welch, V., Losos, M., & Tugwell, P. (2000). The Newcastle– Ottawa Scale (NOS) for Assessing the Quality of Non-Randomized Studies in Meta-Analysis. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp.

Whitlock, J. Muehlenkamp, J. Eckenrode, A. Purington, P. Barrera, G. Baral-Abrams, . . ., E. Smith (2013) Non-suicidal self-injury as a gateway to suicide in adolescents and young adults Journal of Adolescent Health, 52 (4) (2013), pp. 486-492

Wilke, D. J. (2004). Predicting Suicide Ideation for Substance Users: The Role of Self-Esteem, Abstinence, and Attendance at 12-Step Meetings. Addiction Research & Theory, 12(3), 231–240. https://doi.org/10.1080/16066350310001653121

Williams, M. (1997). Cry of Pain: Understanding Suicide And Self-Harm. London: Penguin.

Winters, N. C., Myers, K., & Proud, L. (2002). Ten-Year Review of Rating Scales. III: Scales Assessing Suicidality, Cognitive Style, and Self-Esteem. Journal of the American Academy of Child & Adolescent Psychiatry, 41(10), 1150–1181. https://doi.org/10.1097/00004583-200210000-00006

World Health Organisation (2017) World Health Statistics: suicide mortality rate. Geneva. World Health Organisation

World Health Organisation (2009) Global Health Risks: A response to the need for comprehensive, consistent and comparable information on health risks at global and regional level. Geneva, World Health Organisation.

Wrosch, C., Scheier, M. F., Miller, G. E., Schulz, R., & Carver, C. S. (2003). Adaptive Self-Regulation of Unattainable Goals: Goal Disengagement, Goal Reengagement, and Subjective Well-Being. Personality and Social Psychology Bulletin, 29(12), 1494–1508. https://doi.org/10.1177/0146167203256921

Zhang, L., & Baumeister, R. F. (2006). Your money or your self-esteem: threatened egotism promotes costly entrapment in losing endeavors. Personality & Social Psychology Bulletin, 32(7), 881–893. https://doi.org/10.1177/0146167206287120

Zhang, J., Norvilitis, J. M., & Ingersoll, T. S. (2007). Idiocentrism, Allocentrism, Psychological Well Being and Suicidal Ideation: A Cross Cultural Study. OMEGA - Journal of Death and Dying, 55(2), 131–144. https://doi.org/10.2190/OM.55.2.c

Zhang, J., Liang, B., Zhou, Y., & Brame, W. (2010). Prison inmates’ suicidal ideation in China: a study of gender differences and their impact. International Journal of Offender Therapy and Comparative Criminology, 54(6), 959–983. https://doi.org/10.1177/0306624X09348200

PAPER TWO: EXPERIMENTAL PAPER

The Relationship Between Self-criticism, Self-reassurance and Suicide Risk

The following paper has been prepared for submission to the Journal of Abnormal Psychology.

The guidelines for authors can be found in Appendix A.

Abstract 173

Main Text (excluding 6,542 references)

ABSTRACT

The relationship between self-to-self relating and suicide has received attention in explanatory models of suicide. However, exploration of specific types of self-criticism, namely feelings of inadequacy (associated with perfectionism) and self-attacking has not been addressed in a clinical suicidal population. The role of self-reassurance/ self- compassion in reducing self-criticism has received recent attention. The present study assessed the relative contribution of self-criticism to suicide risk, alongside established predictors of suicidal ideation; hopelessness, depression, defeat and entrapment. It also examined the potential moderating effect of self-reassurance on the relationship between self-criticism and suicide risk. Participants completed measures of inadequacy, self- attacking, self-reassurance, defeat, entrapment, depression and hopelessness (N= 101).

Results demonstrated that self-attacking has a direct relationship with suicide risk, alongside established predictors, entrapment and hopelessness. Depression was not found to be a significant predictor of suicide risk in this population. Self-reassurance did not have a moderating influence on the relationship between inadequacy, self-attacking and suicide risk. Addressing particularly hostile forms of self-criticism may be a promising area in terms of future research and clinical practice.

INTRODUCTION

Suicide is a major public health issue, with significant impact on both individuals and their family members. Although suicide rates have been declining in the UK over the past decade

(CMHS, 2017), the rate of decrease has been slowing in recent years. In a recent systematic review of risk factors for completed suicide, Franklin et al (2017) found that reported risk factors rarely predicted suicidal thoughts and behaviours better than chance, and that further work was needed to better identify predictors. Novel research is needed to understand the processes that lead to completed suicide and to develop interventions which can effectively reduce suicide risk.

Established predictors of completed suicide include mood and substance misuse disorders, issues with impulse control and post-traumatic stress disorder (Nock et al, 2009).

In terms of known sociodemographic characteristics linked to suicide, males are three times more likely to complete suicide than females (WHO, 2014), whereas females are more likely to make a suicide attempt. Other risk factors for completed suicide include being of younger age, unmarried and with low levels of education, higher levels of poverty and unemployment and substance misuse (Hawton et al, 2009, Hawton et al, 2012 and Nock et al, 2013).

Recent work has moved towards the examination of psychological, interpersonal and cognitive factors in relation to suicidal thoughts and behaviours, including hopelessness

(Britton et al, 2008, McClean 2008), poor problem solving (Pollock & Williams, 1998) and interpersonal difficulties (Joiner, 2005), for a full review see O’Connor and Nock (2014).

Psychological models of suicide have highlighted the importance of self-criticism in the pathway from suicidal ideation to completed suicide. Suicide has been described as a way of escaping aversive self-awareness, following internalised experiences of failure

(Baumeister, 1990). Evolutionary models such as those by Gilbert & Allen (1998) described suicide in terms of ‘arrested flight’, caused by negative appraisals of the self and problem- solving difficulties. Suicide in these models is viewed as an extreme form of escape behaviour, a flight from unbearable internal pain, and an attempt to finally resolve seemingly unresolvable issues. The ‘Cry of pain model’ (Williams, 1997) highlights dynamic psychological and social factors that increase the risk of suicide. It suggests that ongoing stressors are a major factor in suicidal ideation and attempts, along with appraisal of stressors and their consequences in terms of defeat (or failed struggle), inflexible negative perceptions of self, a sense of arrested flight (entrapment) and perceived absence of rescue.

It is suggested that self-criticism may fulfil the role of an ongoing stressor in this context, but also contribute to feeling of defeat, due to the unrelenting self-attacking that can occur in some clients. The Schematic Appraisals Model of Suicide (Johnson, 2008), highlighted the importance of negative self-appraisals with regards suicidality, and the development of

‘suicide schemas’ that could be activated by increasingly low threshold of mood fluctuations.

This work has been further developed by O’Connor et al (2011), in their integration of a range of historical models of suicide. The Integrated Motivational-Volitional Model

(IMV) of suicide (see Fig 1 below), proposes a sequential process of motivational and volitional factors in the pathway to suicide attempts. The IMV model combines established psychological theories of suicide in terms of their moderating impact on the pathway from distress to suicidal behaviours. Self-criticism in this model may impact on both the pre-

66 motivational phase of suicidality, social defeat and threat to self moderators to move individuals from their experiences of emotional distress to increasing levels of suicidality.

Figure 1 IMV model (O’Connor, 2011)

Pre-motivational Motivational Phase: Volitional Phase: phase: background Behavioural Enaction Ideation/ Intention Formation and triggering factors

Diathesis Defeat and Entrapment Suicidal Suicidal Humiliation ideation & Behaviour

intent

Environment Threat to Motivational Volitional self Moderators Moderators moderators (MM) (VM) (TSM)

e.g. social problem e.g. Thwarted e.g. capability, Life events solving, coping, belongingness, impulsivity, planning, access to means, memory biases, burdensomeness, imitation rumination future thoughts, goals, norms, social support

As described above, many models of suicide incorporate ideas that relate to self- criticism – from negative evaluations of self in both the SAMS (Johnson et al, 2008) and

Baumeister (1990) models, to the desire from escape from ongoing stressors in both the cry of pain (Williams, 1997) and arrested flight models (Gilbert and Allen, 1998).

Self-criticism has been described as an experience whereby a persons’ “own self-evaluations and sense of self (internal world) become critical, hostile and persecuting.” (Gilbert &

Proctor, 2006, p2). It has been linked to self-harm, depression and psychopathology in a number of studies (Gilbert, Clarke, Hempel, Miles & Irons, 2004; Irons, Gilbert, Baldwin,

Baccus & Palmer, 2006; Richter, Gilbert & McEwan, 2009, Gilbert et al, 2010; O’Connor,

2007). It has been associated with a range of issues, including recollections of parental rejection, bullying in adolescence and child maltreatment and has a mediating role between negative early experiences, depressive symptoms and suicidal cognition (Glassman,

Weierich, Hooley, Deliberto, & Nock, 2007, Campos, Besser & Blatt, 2013; Castilho, Pinto-

Gouveia & Duarte, 2017). Self-criticism has also been directly linked to suicide (Falgares et al, 2017, Campos, Besser & Blatt, 2013 and O’Connor and Noyce, 2008). Fazaa & Page (2003) found that self-critical individuals showed increased levels of suicidal intent and completed suicide, and were more likely to attempt suicide due to ‘intra-psychic stressors’, such as failure to achieve a goal or career concerns, with the explicit motivation to escape.

Research into the forms and functions of self-criticism suggests that rather than a univariate construct, self-criticism may be better conceptualised as comprising two factors; feeling the self as inadequate and inferior (linked to perfectionism), and self-attacking behaviours, linked to disgust (Gilbert, Clarke, Hempel, Miles and Irons, 2004).

Self-criticism: Inadequacy and Perfectionism

One form of self-criticism, inadequacy, has been conceptualised as a motivational strategy to avoid loss of social status; driving self-improvement in an attempt to combat feelings of incompetence (Gilbert, 2005; Gilbert et al, 2010). This form of self-criticism has been linked to perfectionistic self-presentation, in that an attempt is being made to buffer against feelings of external shame (perceived judgement from another). However, often this results in increased levels of self-criticism and negative internal experiences (Costa, Hausenblas,

Oliva, Cuzzocrea, Larcan, 2016; Shahar et al, 2015b). Perfectionism is a trait that has been of interest to suicide researchers, and a previous systematic review has suggested that ‘self- critical evaluative concerns perfectionism’ is linked to suicide (O’Connor, 2007).

Much of the literature related to perfectionism and suicide has highlighted the importance of the ‘social’ aspect of perfectionism – with socially prescribed perfectionism being internalised as self-criticism and leading to increased feelings of social disconnection

(O’Connor, 2007; O’Connor & Noyce, 2008). Social disconnection is a key component of psychological models of suicide such as the Integrated motivational-volitional model and the interpersonal model of suicide. It increases risk of completed suicide through its effect on entrapment and thwarted belongingness (O’Connor et al, 2011, Joiner et al, 2009, O’Connor and Nock, 2014). This study seeks to further understand the role feelings of inadequacy have in suicide probability, with particular reference to their impact on defeat and entrapment.

Self-Attacking and Disgust

In contrast to the self-improvement/ shame avoidance motivation underlying self-critical perfectionism, some self-critical individuals have been found to demonstrate high levels of self-attacking and feelings of disgust towards the self. Gilbert et al (2004) and O’Connor

(2011) both make reference to social rank theory in terms of self-attacking as an attempt to avoid loss of social status by self-punishment or literal removal of an unwanted or hated aspect of the self - linked to ‘cutting off’ and dissociating from an aspect of the self, rather than attempting to ‘do better’ in the eyes of another.

Self-attacking has been found to partially mediate the relationship between dysfunctional cognitions and depression (Simpson, Hillman, Crawford & Overton, 2010).

Gilbert et al (2010) found that the self-attacking form of self-criticism was significantly correlated with self-harm, depression and anxiety after controlling for the self-criticism related to inadequacy. Self-attacking has also been linked to suicidal ideation in individuals with schizophrenia (Joiner, Gencoz, Gencoz, Metalsky & Rudd, 2001) and individuals engaging in non-suicidal self-injury (Xavier, Pinto-Gouveia, Cunha, Carvalho, 2016).

Literature to date has not examined self-attacking in a clinical, suicidal population. However, as a history of self-injury or suicide attempts is one of the key risk factors for suicidal behaviour (O’Connor and Nock, 2014), it is likely that the self-attacking form of self-criticism is related to increased suicide risk.

The role of defeat and entrapment in suicidality

Defeat can be defined as a failed struggle related to loss of status or identity (Taylor,

Gooding, Wood & Tarrier, 2011). Entrapment is linked to uncontrollable, unremitting and inescapable situations (Gilbert, 2005; Gilbert & Williams, 2005). Both concepts have been linked to suicidality in a variety of populations (Slade, Edelmann, Worral & Bray, 2014,

Taylor et al, 2010a and Taylor, Wood, Gooding & Tarrier, 2010b). It has been suggested that adverse early experiences can create unrealistic high standards in some people, leading to internalised self-criticism, internal bullying and a sense of social defeat (O’Connor and

Noyce, 2008; Gilbert, 1989; Gilbert et al, 2004; 2007). Social defeat has been identified as a key precipitant for suicidal ideation in a number of theoretical models of suicide (Gilbert &

Allen, 1998; Williams, 1997 and O’Connor et al, 2011). The relationship between internalised self-criticism and social defeat is one potential mechanism whereby self- criticism impacts on suicidality. Similarly, there is emerging literature on the direct relationship of self-criticism to entrapment (Martin, Gilbert, McEwan & Irons, 2006,

Sturman & Mongrain, 2005), an increasingly important precursor to suicidal behaviour;

(O’Connor & Portzky, 2018). For this reason, the present study will examine the potential mediating effects of defeat and entrapment in the relationship between self-criticism and suicide probability.

Factors which may ameliorate self-criticism

One emerging factor within the self-criticism literature is the role of self-compassion, kindness and reassurance as a potential strategy that may mitigate the impact of self- criticism on a range of mental health difficulties. A recent meta-analysis by Kirkby, Tellegen

& Steindl (2017) examined the role of interventions aimed at increasing self-compassion across a range of populations (including those in psychological distress, with anxiety and with depression), a moderate effect size was found. Similarly compassion focused therapy based interventions have been shown to impact on self-criticism levels in individuals with social anxiety, Non-Suicidal Self-Injury (NSSI) and those accessing weight management support (Duarte et al, 2017, Van Vilet &Kalnins, 1011; Boersma, Hakanson, Salomonsson &

Johansson 2015)

Self-compassion/ self-reassurance has demonstrated an impact on self-criticism levels in clinical populations related to suicide (such as individuals with depression and eating disorders and NSSI) and low levels of self-compassion have been related to increased suicidal ideation (Collette, Pugh, Waite & Freeman, 2016). It is therefore hypothesised that self-compassion/ self-reassurance will have an impact on self-criticism in suicidal individuals.

This will be examined in the study, along with any potential moderating effect on self- criticisms relationship to suicide. To date, no studies have looked directly at a self- compassion in a suicidal population, thus, due to the limited available evidence, the role of self-compassion/ self-reassurance and its relationship to suicide risk will be examined as an exploratory hypothesis in this study.

For definitional clarity, Suicidal ideation is defined in this paper as including thoughts of wanting to die, alongside specific thoughts about engaging in suicidal behaviour. The term suicidal behaviour is used to refer to potentially self-injurious actions where there is implicit or explicit evidence that the person intended to end their life. Suicidality will be used to describe the combination of suicidal thoughts and behaviour.

This study aimed to investigate the role of different forms of self-criticism

(inadequacy and self-attacking) and their relationship to suicidality. In order to understand the mechanisms underpinning this relationship, the potential role of defeat and entrapment will also be examined. This is in line with current theoretical models of suicide, where defeat and entrapment are seen to play a key role in the pathway to suicidal thoughts and behaviours. The study also aimed to explore the moderating role of self-reassurance on self- criticism in its relationship with suicide risk. The influence of known correlates of suicide – depression and hopelessness – will be controlled for.

Hypothesis one makes the prediction that inadequacy will be significantly associated with suicide risk. Hypothesis two will examine the potential mediating role of defeat and entrapment on the relationship between inadequacy and suicide risk. Hypothesis three makes the prediction that self-attacking will be significantly associated with suicide risk.

Hypothesis four will again examine the potential mediating role of defeat and entrapment on the relationship between self-attacking and suicide risk.

Exploratory hypothesis one predicts that self-reassurance will moderate the relationship between inadequacy and suicide risk. Exploratory hypothesis two predicts that self- reassurance will moderate the relationship between self-attacking and suicide risk.

METHOD

Inclusion criteria

Participants comprised of adults (over 18 years) who had self-reported suicidal ideation or behaviours in the preceding 12 months. They were also required to be in contact with mental health services at time of recruitment, have sufficient English language proficiency to complete self-report questionnaires and capacity to give informed consent. Exclusion criteria included individuals with organic brain disease or who were heavily intoxicated at the time of assessment. Eligibility around organic brain disease was determined by discussion with keyworker prior to contacting client. Intoxication was determined by the interviewer at the time of interview.

Procedure

Participants were recruited from mental health services across the North West of England.

These services included mental health inpatient wards, community mental health teams and crisis resolution and home treatment teams. Inpatient participants were also taking part in a suicide prevention trial (the RfPB funded INSITE trial, Haddock et al, 2016) and consent to use data collected within this trial was provided by all participants. Care coordinators and nursing staff identified participants who met the inclusion or exclusion requirements of the study for community participants and contacted them to obtain consent for researchers to contact them. Community participants were telephoned and screened by the researcher for historical suicidal ideation or behaviours in the last 12 months. Participants were then given

75 an information sheet and a minimum of 24 hours to consider whether they would like to participate.

Following written consent researchers administered questionnaires and interview schedules as detailed below. Interviews were carried out in one session, but if a participant required additional sessions then this was provided. All participants were given safety cards listing support services at the conclusion of assessment and liaison with services was undertaken in line with agreed risk management protocols. Ethical approval was provided for this study, by the NHS ethics committee in the North West of England - Lancaster.

Measures

Suicide Risk

The Suicide Probability Scale (SPS; Cull and Gill, 1988).

The SPS is a thirty-six item self-report measure that measures future risk of suicide, where participants are asked to rate their experiences on a 4 point Likert scale. Responses focus on the frequency of emotions and behaviours from ‘none or a little of the time’ to ‘most or all of the time’ (questionnaire did not specify a time period). Raw scores were used. The measure has an internal consistency of 0.93 and test-retest reliability of 0.92 (Cull and Gill, 1998).

Studies have reported alpha coefficients of 0.92 and 0.90 (Gutierrez et al, 2001; Tatman,

Greene & Karr, 1993). In the present study internal consistency was α = 0.86.

Depression and Hopelessness

The Calgary Depression Scale for Schizophrenia (CDSS; Addington, Addington & Schissel,

1990).

The CDSS is a nine item interviewer administered scale measuring depression severity in the past 2 weeks. Questions examine various symptoms associated with depression including early wakening, guilty ideas of reference and pathological guilt. It has good internal consistency (α = 0.82) and test-retest validity, 0.83 (Addington, Shah, Liu & Addington,

2014). In the present study internal consistency was α = 0.72.

The Beck Hopelessness Scale (BHS; Beck, Weismann, Lester and Trexter, 1974).

The BHS is a twenty item self-report scale which measures negative beliefs around three areas of hopelessness in the past week; feelings about the future, loss of motivation, future expectations. Beck et al (1974) reported α=0.93. Durham (1982) found that reliability of BHS scores was 0.86 to 0.83 for psychiatric populations. In the present study internal consistency was α= 0.92.

Defeat and Entrapment

The Defeat Scale (Gilbert and Allen, 1998).

The Defeat Scale is a 16 item self-report scale that examines perception of social rank and failed struggle over the past seven days. Participants rate their agreement with statements associated with defeat on a four point Likert scale from never to always. The scale has

77 previously been found to have an internal consistency of α = 0.93-0.94 (Taylor et al, 2011).

In the current study the internal consistency of this scale was α = 0.84.

The Entrapment Scale (Gilbert and Allen, 1998).

The Entrapment scale is a 16 item self-report scale which examines motivation to escape on a five point Likert scale, the scale does not specify a timescale. It has previously been found to have an internal consistency of α = 0.95 (Taylor et al, 2010a). The internal consistency in this study was rated as α = 0.86

Self-Criticism and Self-Reassurance

The Forms of Self-Criticising/ Attacking and Self Reassuring Scale (FSCRS; Gilbert, Clarke,

Hempel, Miles and Irons, 2004)

The FSCRS is a 22 item self-report scale that examines self-criticism and the ability to self- reassure when things go wrong. Participants answer items against a five point Likert scale ranging from 0 ‘not at all like me’ to 4 ‘extremely like me’. Sub scales include ‘inadequate self’ which focuses on a sense of not being good enough or personal disappointment and

‘hated self’ which examines the level of self-attacking or desire to persecute the self. A final factor examines ‘reassured self’ which involves ideas around self-kindness and forgiveness of self. In a clinical sample the α for inadequate self was 0.87-0.89, for self-attacking was

0.83-0.86 and for reassured self was 0.85-0.87 (Baiao, Gilbert, McEwan &Carvalho, 2014). In the current study α for inadequate self was 0.83, hated self was 0.79 and reassured self was

0.86.

Demographic Data

Demographic data for age, gender, ethnicity, marital status and diagnosis (self-report) were collected from participants at the time of interview and was based on self-report.

Data Analyses

A power calculation indicated that the number of participants required would involve a ratio of 10:1, i.e. ten participants per predictor variable, following Cohen (1992). Thus, a minimum of 60 participants was required with five predictor variables (depression, hopelessness, defeat, entrapment and self-criticism variables) plus 10 participants for the mediation/ moderation condition, with the outcome variable of suicide risk. Additional analyses were planned for the self-criticism sub-domains of self-attacking, inadequacy and self-reassurance along with the previous 4 predictor variables. With fifty participants, it was expected that results would have an 80% power to detect simple correlations of 0.4 or more between pairs of measures. It was hypothesised that an effect size of approximately r = 0.4 should be expected (this was calculated by taking the mean of three studies with a similar methodology (Gilbert et al, 2010; Enns, Cox & Clara, 2003 and Gilbert & Miles, 2000).

Additional participants were recruited to allow for exploratory hypotheses to be carried out.

Tests for normality and identification of potential outliers were completed including tests for kurtosis, skewness and box plots to check for outliers. A decision was made to use boot strapping in both correlation and regression analyses using 1000 samples as there was a mixture of normal and non-normally distributed results across the measures. Initial

79 bivariate correlations were carried out, followed by correlations where known correlates of suicidality (depression and hopelessness) were partialled out. Missing data were accounted for by taking an average of scores for that response in accordance with Tabachnick and

Fiddell (2007).

Assumptions necessary for conducting multiple regression modelling were tested, including independence of observations (Durbin-Watson statistic), checking of linear relationships, test for multicollinearity and significant outliers. Variables that had shown a significant relationship with suicide risk at a bivariate correlational level after controlling for depression and hopelessness, were selected for entry into multiple regression analyses for each of the self-criticism variables - inadequacy and self-attacking (hypotheses one and three).

Mediation analysis was then carried out using the Hayes (2013) PROCESS Macro for IBM

SPSS version 23, which uses bootstrapping to generate confidence intervals through random resampling. In the mediation analyses, bootstrapping with 5000 resamples and 95% percentile based confidence intervals were calculated to examine the indirect effects of different forms of self-criticism on suicide risk via entrapment (hypotheses two and four).

A moderation analysis using the PROCESS model (Hayes, 2013) was carried out to examine the moderating effect of self-compassion/self-reassurance on the relationship of inadequacy and self-attacking to suicide risk (exploratory hypotheses one and two).

RESULTS

Study characteristics

Of the 101 participants recruited 42% were male, 57% female, 1% transgender/ other. In terms of ethnicity, 95% of the sample described themselves as white British and 5% as black

British. Self-report diagnoses/mental health problem are shown in Table 1. 67% of participants were single, 9% married and 23% divorced.

Table 1 Participants by self-reported diagnosis/mental health condition

Diagnoses Personality Disorder 41 Depression 35 Psychosis 25 Anxiety 13 Bipolar Disorder 12 Post-Traumatic Stress Disorder 8 Unknown 8 Substance misuse 3 Eating disorder 2

Table 2 details the descriptive statistics of variables, mean suicide probability scores were

96.47, above the clinical cut off of 80 found in other papers (Bagge & Osman, 1998; Liang &

Yang, 2010). Calgary Depression Scale score was 13.04, above the clinical cut off of 6 suggested by Addington, Shah, Liu & Addington (2014). The sample scored in the moderate/severe range on Beck Hopelessness Scale. There are no clinical cut offs for either the defeat or entrapment scales, but a maximum score is 64 and participants scores on average 45.74 and 43.77 respectively.

Table 2 Descriptive Statistics

Mean SD Range

Suicide Risk 96.47 18.17 54-131

Depression 13.04 5.36 1- 26

Hopelessness 12.86 5.92 0. - 20

Defeat 45.74 12.67 1-64

Entrapment 43.77 11.89 8-63

Self-Attacking 13.31 5.37 0 - 20 Inadequacy 28.61 6.28 8-36

Self-Reassurance 9.46 6.36 0-27

Table 3 Bivariate Correlations of all variables

Suicide Depression Hopelessness Defeat Entrapment Inadequacy Self- Self-Hatred Risk reassurance

Suicide risk 1 .42** .62** .54** .65** .40** -.34** .46**

Depression 1 .60** .54** .43** .31** -.32** .34** Hopelessness 1 .74** .59** .34** -.51** .38** Defeat 1 .60** .37** -.47** .42** Entrapment 1 .40** -.28** .45** Inadequacy 1 -.49** .70*** Self- 1 -.62** reassurance Self-Hatred 1 **. Correlation is significant at the 0.01 level (2-tailed). c. Unless otherwise noted, bootstrap results are based on 1000 bootstrap samples

Bivariate correlations between all variables are presented in Table 3 above. All variables were significantly related to suicide risk as measured by the SPS; depression, hopelessness, defeat, entrapment, self-attacking, self-reassurance and inadequacy. The variable that

82 correlated most highly with suicide probability was entrapment (r= 0.65). It was also significantly related to all other variables. Depression was significantly correlated with all variables. The variable depression correlated with most highly was hopelessness (r= 0.60).

Hopelessness showed a significant relationship with all variables. Hopelessness was highly intercorrelated with defeat, r=0.74. Defeat showed a significant relationship with all variables. Self-attacking was related significantly to all variables, with strongest relationships being observed with inadequacy (r=0.70) and a significant negative relationship with self- reassurance (r=-0.62). Similarly, inadequacy was related to all variables, with a moderate negative relationship with self-reassurance (r=-0.49). Both self-attacking and inadequacy were related to suicide risk to a similar degree (r=0.46 and 0.40 respectively).

Further partial correlations were undertaken following initial analysis, where known correlates of suicide (depression and hopelessness) were controlled for (Table 4).

Table 4 Partial Correlations Controlling for Depression and Hopelessness

Suicide Defeat Entrapment Inadequate Reassured Hated Probability Scale Scale self FSCRS Self FSCRS Self Scale FSCRS Suicide 1 .16 0.45*** 0.25* -.04 0.30** Probability Scale Defeat Scale 1 0.29** .17 -.16 0.20* Entrapment 1 0.25* .030 0.29** Scale Inadequate 1 -.39*** 0.65*** self FSCRS Reassured Self 1 -.54*** FSCRS Hated Self 1 FSCRS *. Correlation is significant at the 0.05 level (2-tailed). **. Correlation is significant at the 0.01 level (2-tailed). ***Correlation is significant at the 0.001 level (2 -tailed) c. Unless otherwise noted, bootstrap results are based on 1000 bootstrap samples

Once depression and hopelessness were partialled out, defeat no longer showed a significant relationship with suicide probability, inadequacy and self-reassurance. Once depression and hopelessness were partialled out, only entrapment (r=0.45), self-attacking

(r=0.30) and inadequacy (r=0.25) were significantly associated with suicide risk. Self- reassurance remained significantly negatively related to both self-attacking and inadequacy

(r=-0.54 and -0.39 respectively), but not to suicide risk.

The Role of Inadequacy in Suicide Risk – Hypothesis one

It was hypothesised that the ‘inadequacy’ form of self-criticism would be significantly associated with suicide risk. Inadequacy was associated with suicide risk (r = .40 P < 0.01) at a bivariate correlational level. When known correlates of suicide were controlled for

(depression and hopelessness), a significant relationship remained between inadequacy and suicide risk (r=.25, p <0.05 level).

A multiple regression analysis was carried out to further explore the relationship between levels of inadequacy and suicide risk. Defeat and self-reassurance were not associated with suicide risk and as such were not entered into the regression model. In the model that accounted for 53% of the variance (R²= 0.53, F (4, 94) = 26.26, p<0.001), hopelessness and entrapment were both found to be significant predictors of suicide risk (hopelessness β =

0.342, t = 3.45, p <0.001, entrapment β = 0.397, t= 4.35, p<0.001), whereas depression and inadequacy were not.

Table 5 Multiple Regression Predicting Suicidal Probability (inadequate self)

Unstandardized Standardized 95.0% Coefficients Coefficients Confidence Interval for B

B Std. Beta Lower Upper Error Bound Bound (Constant) 45.89 6.611 32.76 59.01

Calgary .03 .303 .01 -.57 .63 Depression Scale Beck 1.05 .305 .34*** .45 1.66 Hopelessness Scale Entrapment .61 .139 .40*** .33 .88 Scale Inadequate .35 .227 .12 -.10 .80 self FSCRS

*. Correlation is significant at the 0.05 level

***. Correlation is significant at the 0.001 level (2-tailed).

The mediating role of entrapment in the relationship between inadequacy and suicide risk

- Hypothesis Two

Hypothesis two entailed a path analysis of the relationship between inadequacy and suicide risk and the possible mediating effects of defeat and entrapment. As defeat was not significantly associated with suicide risk it was not included in this analysis.

A mediation analysis was undertaken using bootstrapped confidence intervals and the

PROCESS macro for SPSS. This mediation analysis was found to be significant, see figure 3 below. The results suggest that entrapment does have a partial indirect mediating effect on the relationship between inadequacy and suicide risk.

Figure 3 Mediation Analysis – Entrapment as mediator – Inadequacy

The role of Self-attacking in Suicide Risk – Hypothesis Three

Hypothesis three proposed that the ‘self-attacking’ form of self-criticism would be significantly associated with suicide risk. Self-attacking was significantly associated with suicide risk (r = 0.46, p < 0.01). When known correlates of suicide were controlled for

(depression and hopelessness) the relationship between self-attacking and suicide risk remained significant (r=.30, p <0.01 level).

A multiple regression analysis was carried out to further examine the relationship between levels of self-attacking and suicide risk (see table 6). Self-attacking, hopelessness and entrapment were all found to be significant predictors of suicide risk (self-attacking β =

0.16, t = 2.03, p=<0.05; entrapment β = 0.38, t= 4.15, p=0.001; hopelessness β = 0.34, t=3.41, P=<0.001) in the model. Hypothesis three was therefore supported. Depression was not found to be a significant predictor of suicide risk. The results of the regression found that the predictors accounted for 53% of the variance (R²= 0.53, F (4, 95) = 27.13, p<0.001).

Table 6 Multiple regression predicting suicidal probability (self-attacking)

Unstandardized Standardized 95.0% Coefficients Coefficients Confidence Interval for B

B Std. Beta Lower Upper Error Bound Bound (Constant) 50.40 5.14 40.185 60.61

Calgary .01 .30 .02 -.59 .61 Depression Scale Beck 1.03 .30 .34** .43 1.63 Hopelessness Scale Entrapment .58 .14 .38*** .30 .86 Scale Hated Self .55 .27 .16* .01 1.10 FSCRS

The mediating role of entrapment in the relationship between self-attacking and suicide risk

Hypothesis four examined the mediating role of entrapment in the relationship between self-attacking and suicide risk. The relationship between self-attacking and suicide risk was partially mediated by entrapment (β= 0.84, bootstrap CI 0.42 - 1.31), see Figure 3 for results.

As the gap between bootstrapped confidence intervals does not contain zero, that would suggest a significant mediating effect involving entrapment, though self-attacking remained significantly and directly linked to suicide risk.

Figure 4 Mediation Analysis – Entrapment as mediator – self-attacking

Self-Reassurance

Self-reassurance was related to suicide risk at a correlational level (r=-0.34 p< 0.01). In line with exploratory hypothesis one and two a moderation analysis was undertaken to examine whether the relationship between inadequacy or self-attacking and suicide risk was moderated by level of self-reassurance. The interaction effect for hypothesis one; that the relationship between inadequacy and suicide risk would be moderated by self-reassurance was non-significant (β=-0.05, SE 0.67, T=-1.13, P= 0.26. The interaction effect for hypothesis two, that the relationship between self-attacking and suicide risk would be moderated by self-reassurance was also non-significant (β = -0.03, SE 0.67, t = -.48, P= 0.64). This suggests that although self-reassurance has a significant negative relationship with both inadequacy and self-attacking, this does not influence their relationship to suicide risk

DISCUSSION

The present study aimed to test four main hypotheses and two exploratory hypotheses.

First, that feelings of inadequacy would be significantly related to suicide risk. Second that experiences of self-attacking/ self-hatred would be significantly related to suicide risk.

Following this, two mediation hypotheses were proposed, to examine the role of defeat and entrapment in the relationship between inadequacy, self-attacking and suicide risk. The final two exploratory hypotheses aimed to examine whether the ability to be reassuring towards oneself during times of adversity moderated any relationship between inadequacy, self- attacking and suicide risk.

Results suggest that self-attacking is a significant independent predictor of suicide risk, alongside established predictors of suicide; entrapment and hopelessness. The relationship between self-attacking and suicide risk remained significant when considering the mediating role of entrapment, and analysis demonstrated that entrapment had a partial mediating effect on the relationship between self-attacking and suicide risk. The results did not support inadequacy as a significant predictor of suicide risk, above established predictors. Mediation analysis suggested that Inadequacy’s relationship with suicide risk is partially mediated through entrapment.

Overall results of both regression analyses suggest that entrapment is the strongest predictor of suicide risk in a model which includes inadequacy, depression, hopelessness and entrapment. Depression was not a significant predictor of suicide risk in this model. A similar finding was reported in the analysis involving self-attacking. This supports recent publications by O’Connor and Portzky (2018) suggesting entrapment is an important factor in the progression from mental distress to suicidal behaviour. Previous research involving

89 both systematic reviews and meta-analyses has suggested that entrapment is a key predictor of suicidal behaviour in both cross-sectional and prospective studies (Taylor et al,

2011; Siddaway, Taylor, Wood & Schulz, 2015).

Self-reassurance, though negatively related to self-attacking and inadequacy, did not demonstrate a moderating effect on the relationship between self-criticism variables and suicide risk. This may be related to the finding that highly self-critical individuals struggle to generate feelings of self-reassurance and warmth towards the self (Gilbert, Baldwin, Irons,

Baccus & Palmer, 2006). It would appear that self-reassurance alone is not enough to influence the hostile and self-attacking attitudes that relate to increased suicide risk in a clinical suicidal population. However, it could also relate to the scope of the measure used in this study. The present study examined ‘self-reassurance’; however other measures of self- compassion have included broader definitions. One example is the Self Compassion Scale

(Neff, 2003a), which examines self-compassion in terms of three components; kindness towards self (similar to the self-reassurance element of the FSCRS), common humanity and mindful awareness and acceptance of negative emotional states. It is possible that these wider elements of self-compassion may have demonstrated a greater moderating effect in this population.

Research suggests that for highly self-critical individuals, adverse childhood experiences such as emotional and sexual abuse, physical neglect and perceived overprotectiveness of the early primary caregiver, have an impact on levels of internalised self-criticism and shame (Glassman et al, 2007, Campos et al, 2013, Castilho et al, 2016).

This in turn predisposes individuals to experiences of social defeat (Gilbert et al 2004; 2007) and increased sensitivity to signals of threat to self. Self-threat has been operationalised as a

90 state ‘where favourable views about oneself are questioned, contradicted, impugned, mocked, challenged or otherwise put in jeopardy’ (Baumeister, Smart & Boden, 1996 p8).

This definition is remarkably similar to the items on the FSCRS, where self-critical attitudes and behaviours include “I feel beaten down by my own self-critical thoughts”, “I have a sense of disgust with myself” and “I call myself names”. This internalised self-criticism and self-attacking may lead to increased feelings of entrapment as an individual is unable to escape the persecutory aspects of the self (Baumeister et al, 1990, Gilbert et al, 2005).

It has been suggested that increased levels of internalised self-criticism and social perfectionism lead ultimately to social disconnection and experiences of thwarted belongingness (O’Connor and Noyce, 2008). Thwarted belongingness is a well evidenced construct in the interpersonal theory of suicide (Joiner et al, 2009) and involves the experience of alienation from family, friends or other valued groups. It has been hypothesised that experiences of thwarted belongingness moderate experiences of entrapment by isolating an individual and this leads to increased experiences of suicidal ideation and later behaviour (O’Connor, 2011, O’Connor, Cleare, Eschle, Weatherall, Kirtley,

2016). However, this has yet to be supported conclusively by evidence, and a recent study by Forkmann and Teismann (2017) suggested that thwarted belongingness did not moderate the relationship between entrapment and suicidal ideation.

Further prospective studies are required to confirm the relationship of self-criticism to suicide risk and entrapment and the role of potential moderating factors such as thwarted belongingness. Alongside this, experimental manipulation studies, where interventions to address self-critical thoughts are delivered to suicidal individuals and age and gender matched controls, would be recommended.

Theoretical Implications

A number of explanatory models of suicide incorporate ideas that relate to negative evaluations of the self (Johnson et al, 2008), escape from aversive self-awareness

(Baumeister, 1990) and their relationship to increased levels of suicidality. Both the ‘Cry of pain model’ (Williams, 1997) and Gilbert and Allen’s model of ‘arrested flight’ highlight dynamic psychological and social factors that increase the risk of suicide. They suggest that ongoing stressors are a major factor in suicidal ideation and attempts, along with appraisal of stressors and their consequences in terms of defeat (or failed struggle), inflexible negative perceptions of self, a sense of arrested flight (entrapment) and absence of rescue.

In this context the hostile and aggressive internal environment of individuals with high levels of self-attacking may create an inescapable ongoing stressor, which may impact directly on suicidal thoughts, and a sense of ongoing entrapment. This inescapable threat may lead to suicide being seen as the only answer to ongoing internal attacks.

The cry of pain model was integrated with other key models of suicidality by

O’Connor et al (2011), who outlined the Integrated Motivational-Volitional Model (IMV) of suicide. This model again highlighted the key areas of defeat and entrapment in suicidal rumination and behaviour. It proposed that the motivation to escape from defeating circumstances (in this context inescapable hostile self-attacking) drives a search for a solution to end the psychological pain being experienced (Shneidman, 1996). Self-criticism in this model may impact on both the pre-motivational phase of suicidality, increasing levels of social defeat and interacting with threat to self moderators to move individuals from their experiences of emotional distress to increasing levels of suicidality

Clinical Implications

Given the present findings, assessment of self-criticism experienced by clients, alongside their experiences of feeling trapped, hopeless or unable to escape would be recommended in clinical practice. This would allow clinicians to target the specific antecedents of suicidal thoughts and behaviours with a view to preventing escalating levels of suicide risk.

Although the evidence base is still emerging, new approaches such as Compassion

Focused Therapy (CFT) aim to develop healthy self-self relating in highly self-critical individuals, with techniques that can be utilised to activate the attachment system and develop a more nurturing approach to the self (Gilbert et al, 2005, 2010; Welford, 2012;

Irons and Beaumont, 2017). As discussed earlier in the paper compassion focused therapy based interventions have been shown to impact on self-criticism levels in individuals with social anxiety, Non-Suicidal Self-Injury (NSSI) and those accessing weight management support (Duarte et al, 2017, Van Vilet &Kalnins, 1011; Boersma, Hakanson, Salomonsson &

Johansson 2015), and thus may help to reduce suicide risk in individuals with high levels of self-attacking.

Targeting of self-critical cognitive processes is one possible avenue for clinicians to undertake in the reduction of suicide risk in clients. This might involve the integration of models which address self-critical thoughts into existing suicide prevention CBT protocols

(CBT-SP; Mewton and Andrews, 2016; Stanley et al, 2009). CBT-SP have demonstrated effectiveness over condition specific models in the treatment of suicidal thoughts and behaviours. Evidence from the present study suggests that it is not enough to target depressive cognitions, other factors have better predictive power in terms of suicide risk and would make more efficient targets for symptom reduction.

Thwarted belongingness, a key facet of the interpersonal theory of suicide, has been linked to internalised criticism in the literature (O’Connor & Noyce, 2008), thus interventions to address this experience may be helpful in reducing risk of suicide in highly self-critical individuals. Adaptations to interpersonal interventions such as psychodynamic interpersonal therapy (PIT) and interpersonal therapy (IPT) derived from the interpersonal theory of suicide have been implemented with positive results (Brown and Jager-Hyman,

2014; Van Orden, Talbot and King, 2012).

Limitations

One limitation of the study was that it was based on cross-sectional data and as such it is not possible for causal inferences to be made. During the data collection phase, it became increasingly obvious from informal qualitative feedback that many participants had recently experienced a crisis in their personal lives, from relationship breakdown to benefit payment delays or bereavement. Future recommendations would include that a life events measure to be included in the questionnaire battery to capture these experiences formally. The study did not include an ethnically diverse population so the findings may not be generalisable to others from diverse cultural backgrounds. This could be rectified by ensuring a sample representative of the population is obtained. A strength of the study was that it did include a range of individuals from within the mental health population, with a diversity of diagnoses, inpatient and community settings. The study was also of sufficient power and control of key confounding variables of depression and hopelessness in the analysis.

Conclusions

The results of this study suggest that the more severe and hostile form of self-criticism; self- attacking- is significantly and directly related to elevated levels of suicide risk. The results highlight the need for existing explanatory models of suicide to continue to incorporate self- self-relating as a key factor that may increase suicide risk. Future recommendations include longitudinal studies examining the role of self-attacking attitudes on suicidality and an exploration of the effectiveness of therapeutic interventions which target self-criticism in a suicidal population.

REFERENCES

Addington, D., Addington, J., & Schissel, B. (1990). A depression rating scale for schizophrenics. Schizophrenia Research, 3(4), 247–251.

Addington, J., Shah, H., Liu, L., & Addington, D. (2014). Reliability and validity of the Calgary Depression Scale for Schizophrenia (CDSS) in youth at clinical high risk for psychosis. Schizophrenia Research, 153(1), 64–67. https://doi.org/10.1016/j.schres.2013.12.014

Bagge, C., & Osman, A. (1998). The Suicide Probability Scale: Norms and factor structure. Psychological Reports, 83, 637–8. https://doi.org/10.2466/PR0.83.6.637-638

Baião, R., Gilbert, P., McEwan, K., & Carvalho, S. (2015). Forms of Self-Criticising/Attacking & Self-Reassuring Scale: Psychometric properties and normative study. Psychology and Psychotherapy: Theory, Research and Practice, 88(4), 438–452. https://doi.org/10.1111/papt.12049

Baumeister, R. F. (1990). Suicide as escape from self. Psychological Review, 97(1), 90–113.

Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened egotism to violence and aggression: The dark side of high self-esteem. Psychological Review, 103(1), 5-33.

Beck, A. T., Weissman, A., Lester, D., & Trexler, L. (1974). The measurement of pessimism: the hopelessness scale. Journal of Consulting and Clinical Psychology, 42(6), 861–865.

Blatt, S. J., Quinlan, D. M., Chevron, E. S., McDonald, C., & Zuroff, D. (1982). Dependency and self-criticism: psychological dimensions of depression. Journal of Consulting and Clinical Psychology, 50(1), 113–124.

Boersma, K., Håkanson, A., Salomonsson, E., & Johansson, I. (2015). Compassion Focused Therapy to Counteract Shame, Self-Criticism and Isolation. A Replicated Single Case Experimental Study for Individuals With Social Anxiety. Journal of Contemporary Psychotherapy, 45(2), 89–98. https://doi.org/10.1007/s10879-014-9286-8

Britton, P. C., Duberstein, P. R., Conner, K. R., Heisel, M. J., Hirsch, J. K., & Conwell, Y. (2008). Reasons for Living, Hopelessness, and Suicide Ideation Among Depressed Adults 50 Years or Older. The American Journal of Geriatric Psychiatry, 16(9), 736–741. https://doi.org/10.1097/JGP.0b013e31817b609a

Brown, G. K., & Jager-Hyman, S. (2014). Evidence-Based Psychotherapies for Suicide Prevention: Future Directions. American Journal of Preventive Medicine, 47(3, Supplement 2), S186–S194. https://doi.org/10.1016/j.amepre.2014.06.008

Campos, R. C., Besser, A., & Blatt, S. J. (2010). The mediating role of self-criticism and dependency in the association between perceptions of maternal caring and depressive symptoms. Depression and Anxiety, 27(12), 1149– 1157. https://doi.org/10.1002/da.20763

Castilho, P., Pinto-Gouveia, J., & Duarte, J. (2017). Two forms of self-criticism mediate differently the shame– psychopathological symptoms link. Psychology and Psychotherapy: Theory, Research and Practice, 90(1), 44– 54. https://doi.org/10.1111/papt.12094

Cohen, J. (1992). Statistical Power Analysis. Current Directions in Psychological Science, 1(3), 98–101. https://doi.org/10.1111/1467-8721.ep10768783

Collett, N., Pugh, K., Waite, F., & Freeman, D. (2016). Negative cognitions about the self in patients with persecutory delusions: An empirical study of self-compassion, self-stigma, schematic beliefs, self-esteem, fear of madness, and suicidal ideation. Psychiatry Research, 239, 79–84. https://doi.org/10.1016/j.psychres.2016.02.043

Costa, S., Hausenblas, H. A., Oliva, P., Cuzzocrea, F., & Larcan, R. (n.d.). Maladaptive perfectionism as mediator among psychological control, eating disorders, and exercise dependence symptoms in habitual exerciser. Journal of Behavioral Addictions, 5(1), 77–89. https://doi.org/10.1556/2006.5.2016.004

Cull J, Gill W (1982) Suicide probability scale. WPS, Los Angeles, CA

Duarte, C., Stubbs, J., Pinto-Gouveia, J., Matos, M., Gale, C., Morris, L., & Gilbert, P. (2017). The Impact of Self- Criticism and Self-Reassurance on Weight-Related Affect and Well-Being in Participants of a Commercial Weight Management Programme. Obesity Facts, 10(2), 65–75. https://doi.org/10.1159/000454834

Durham, T. W. (1982). Norms, reliability, and item analysis of the hopelessness scale in general psychiatric, forensic psychiatric, and college populations. Journal of Clinical Psychology, 38(3), 597–600. https://doi.org/10.1002/1097-4679(198207)38:3<597::AID-JCLP2270380321>3.0.CO;2-6

Enns, M. W., Cox, B. J., & Clara, I. P. (2005). Perfectionism and Neuroticism: A Longitudinal Study of Specific Vulnerability and Diathesis-Stress Models. Cognitive Therapy and Research, 29(4), 463–478. https://doi.org/10.1007/s10608-005-2843-04

Falgares, G., Marchetti, D., De Santis, S., Carrozzino, D., Kopala-Sibley, D. C., Fulcheri, M., & Verrocchio, M. C. (2017). Attachment Styles and Suicide-Related Behaviors in Adolescence: The Mediating Role of Self-Criticism and Dependency. Frontiers in Psychiatry, 8. https://doi.org/10.3389/fpsyt.2017.00036

Fazaa, N., & Page, S. (2003). Dependency and Self-Criticism as Predictors of Suicidal Behavior. Suicide and Life- Threatening Behavior, 33(2), 172–185. https://doi.org/10.1521/suli.33.2.172.22777

Forkmann, T., & Teismann, T. (2017). Entrapment, perceived burdensomeness and thwarted belongingness as predictors of suicide ideation. Psychiatry Research, 257(Supplement C), 84–86. https://doi.org/10.1016/j.psychres.2017.07.031

Gilbert P. 1989. Human Nature and Suffering. Lawrence Erlbaum Associates: Hove.

Gilbert, P. (Ed.). (2005). Compassion: Conceptualisations, Research and Use in Psychotherapy (New Ed edition). London ; New York: Routledge . Gilbert, P., & Allan, S. (1998). The role of defeat and entrapment (arrested flight) in depression: an exploration of an evolutionary view. Psychological Medicine, 28(3), 585–598.

Gilbert, P., Broomhead, C., Irons, C., McEwan, K., Bellew, R., Mills, A., … Knibb, R. (2007). Development of a striving to avoid inferiority scale. The British Journal of Social Psychology, 46(Pt 3), 633–648. https://doi.org/10.1348/014466606X157789

Gilbert, P., Clarke, M., Hempel, S., Miles, J. N. V., & Irons, C. (2004). Criticizing and reassuring oneself: An exploration of forms, styles and reasons in female students. The British Journal of Clinical Psychology, 43(Pt 1), 31–50. https://doi.org/10.1348/014466504772812959

Gilbert, P., McEwan, K., Irons, C., Bhundia, R., Christie, R., Broomhead, C., & Rockliff, H. (2010). Self-harm in a mixed clinical population: the roles of self-criticism, shame, and social rank. The British Journal of Clinical Psychology, 49(Pt 4), 563–576. https://doi.org/10.1348/014466509X479771

Gilbert, P., & Miles, J. N. V. (2000). Sensitivity to Social Put-Down: it’s relationship to perceptions of social rank, shame, social anxiety, depression, anger and self-other blame. Personality and Individual Differences, 29(4), 757–774. https://doi.org/10.1016/S0191-8869(99)00230-5

Glassman, L. H., Weierich, M. R., Hooley, J. M., Deliberto, T. L., & Nock, M. K. (2007a). Child maltreatment, non- suicidal self-injury, and the mediating role of self-criticism. Behaviour Research and Therapy, 45(10), 2483– 2490. https://doi.org/10.1016/j.brat.2007.04.002

Glassman, L. H., Weierich, M. R., Hooley, J. M., Deliberto, T. L., & Nock, M. K. (2007b). Child maltreatment, non- suicidal self-injury, and the mediating role of self-criticism. Behaviour Research and Therapy, 45(10), 2483– 2490. https://doi.org/10.1016/j.brat.2007.04.002

Gutierrez, P. M., Osman, A., Barrios, F. X., & Kopper, B. A. (2001). Development and initial validation of the Self-harm Behavior Questionnaire. Journal of Personality Assessment, 77(3), 475–490. https://doi.org/10.1207/S15327752JPA7703_08

Haddock, G., Davies, L., Evans, E., Emsley, R., Gooding, P., Heaney, L., … Awenat, Y. (2016). Investigating the feasibility and acceptability of a cognitive behavioural suicide prevention therapy for people in acute psychiatric wards (the ‘INSITE’ trial): study protocol for a randomised controlled trial. Trials, 17, 79. https://doi.org/10.1186/s13063-016-1192-9

Hawton, K., Bergen, H., Kapur, N., Cooper, J., Steeg, S., Ness, J., & Waters, K. (2012). Repetition of self-harm and suicide following self-harm in children and adolescents: findings from the Multicentre Study of Self-harm in England. Journal of Child Psychology and Psychiatry, and Allied Disciplines, 53(12), 1212–1219. https://doi.org/10.1111/j.1469-7610.2012.02559.x

Hawton, K., Witt, K. G., Taylor Salisbury, T. L., Arensman, E., Gunnell, D., Hazell, P., … van Heeringen, K. (2016). Psychosocial interventions for self-harm in adults. In Cochrane Database of Systematic Reviews. John Wiley & Sons, Ltd. https://doi.org/10.1002/14651858.CD012189

Hayes, A (2013) Introduction to Mediation, Moderation and Conditional Process Analysis: A regression-based approach. London, Guildford Press

Irons, C., & Beaumont, E. (2017). The Compassionate Mind Workbook: A step-by-step guide to developing your compassionate self (01 edition). London: Robinson.

Irons, C., Gilbert, P., Baldwin, M. W., Baccus, J. R., & Palmer, M. (2006). Parental recall, attachment relating and self- attacking/self-reassurance: their relationship with depression. The British Journal of Clinical Psychology, 45(Pt 3), 297–308.

Joiner, T. E., Gencoz, F., Gencoz, T., Metalsky, G. I., & Rudd, M. D. (2001). The Relation of Self-Hatred and Suicidality in People With Schizophrenia-Spectrum Symptoms. Journal of Psychopathology and Behavioral Assessment, 23(2), 107–115. https://doi.org/10.1023/A:1010915709011

Joiner, T.E. (2005). Why people die by suicide. Cambridge, MA: Harvard University Press.

Joiner, T. E., Van Orden, K. A., Witte, T. K., Selby, E. A., Ribeiro, J. D., Lewis, R., & Rudd, M. D. (2009). Main Predictions of the Interpersonal-Psychological Theory of Suicidal Behavior: Empirical Tests in Two Samples of Young Adults. Journal of Abnormal Psychology, 118(3), 634–646. https://doi.org/10.1037/a0016500

Kirby, J. N., Tellegen, C. L., & Steindl, S. R. (2017). A Meta-Analysis of Compassion-Based Interventions: Current State of Knowledge and Future Directions. Behavior Therapy, 48(6), 778–792. https://doi.org/10.1016/j.beth.2017.06.003

Liang Y, Yang L. (2010) Study on reliability and validity of the suicide probability scale. China J Health Psychol;2:225- 227

Martin, Y., Gilbert, P. P., McEwan, K., & Irons, C. (2006). The relation of entrapment, shame and guilt to depression, in carers of people with dementia. Aging & Mental Health, 10(2), 101–106. https://doi.org/10.1080/13607860500311953

Mewton, L., & Andrews, G. (2016). Cognitive behavioral therapy for suicidal behaviors: improving patient outcomes. Psychology Research and Behavior Management, 9, 21–29. https://doi.org/10.2147/PRBM.S84589

Neff, K. D. (2003a). Development and validation of a scale to measure self-compassion. Self and Identity, 2, 223–250.

O’Connor, R. C. (2007). The relations between perfectionism and suicidality: a systematic review. Suicide & Life- Threatening Behavior, 37(6), 698–714. https://doi.org/10.1521/suli.2007.37.6.698

O’Connor, R. C. (2011). Towards an integrated motivational–volitional model of suicidal behaviour. International Handbook of Suicide Prevention: Research, Policy and Practice, 181–198.

O’Connor, R. C., Cleare, S., Eschle, S., Wetherall, K., & Kirtley, O. J. (2016). The Integrated Motivational-Volitional Model of Suicidal Behavior. In R. C. O’Connor & J. Pirkis (Eds.), The International Handbook of Suicide Prevention (pp. 220–240). John Wiley & Sons, Ltd. https://doi.org/10.1002/9781118903223.ch13

O’Connor, R. C., & Nock, M. K. (2014). The psychology of suicidal behaviour. The Lancet Psychiatry, 1(1), 73–85. https://doi.org/10.1016/S2215-0366(14)70222-6

O’Connor, R. C., & Portzky, G. (2018). The relationship between entrapment and suicidal behavior through the lens of the integrated motivational–volitional model of suicidal behavior. Current Opinion in Psychology, 22(Supplement C), 12–17. https://doi.org/10.1016/j.copsyc.2017.07.021

Pollock, L. R., & Williams, J. M. G. (1998). Problem Solving and Suicidal Behavior. Suicide and Life-Threatening Behavior, 28(4), 375–387. https://doi.org/10.1111/j.1943-278X.1998.tb00973.x

Richter, A., Gilbert, P., & McEwan, K. (2009). Development of an early memories of warmth and safeness scale and its relationship to psychopathology. Psychology and Psychotherapy, 82(Pt 2), 171–184. https://doi.org/10.1348/147608308X395213

Siddaway, A. P., Taylor, P. J., Wood, A. M., & Schulz, J. (2015). A meta-analysis of perceptions of defeat and entrapment in depression, anxiety problems, posttraumatic stress disorder, and suicidality. Journal of Affective Disorders, 184(Supplement C), 149–159. https://doi.org/10.1016/j.jad.2015.05.046

Shahar, B., Szepsenwol, O., Zilcha-Mano, S., Haim, N., Zamir, O., Levi-Yeshuvi, S., & Levit-Binnun, N. (2015). A Wait- List Randomized Controlled Trial of Loving-Kindness Meditation Programme for Self-Criticism. Clinical Psychology & Psychotherapy, 22(4), 346–356. https://doi.org/10.1002/cpp.1893

Slade, K., Edelmann, R., Worrall, M., & Bray, D. (2014). Applying the Cry of Pain Model as a predictor of deliberate self-harm in an early-stage adult male prison population. Legal and Criminological Psychology, 19(1), 131–146. https://doi.org/10.1111/j.2044-8333.2012.02065.x

Stanley, B., Brown, G., Brent, D. A., Wells, K., Poling, K., Curry, J., … Hughes, J. (2009). Cognitive-Behavioral Therapy for Suicide Prevention (CBT-SP): Treatment Model, Feasibility, and Acceptability. Journal of the American Academy of Child & Adolescent Psychiatry, 48(10), 1005–1013. https://doi.org/10.1097/CHI.0b013e3181b5dbfe

Sturman, E., & Mongrain, M. (2005). Self-criticism and major depression: An evolutionary perspective. British Journal of Clinical Psychology, 44(4), 505–519. https://doi.org/10.1348/014466505X35722

Tabachnick, B. G., & Fidell, L. S. (2007). Using Multivariate Statistics (5th ed.). New York: Allyn and Bacon

Tatman, S. M., Greene, A. L., & Karr, L. C. (1993). Use of the Suicide Probability Scale (SPS) with Adolescents. Suicide and Life-Threatening Behavior, 23(3), 188–203. https://doi.org/10.1111/j.1943-278X.1993.tb00179.x

Taylor, P. J., Gooding, P. A., Wood, A. M., Johnson, J., Pratt, D., & Tarrier, N. (2010). Defeat and entrapment in schizophrenia: The relationship with suicidal ideation and positive psychotic symptoms. Psychiatry Research, 178(2), 244–248. https://doi.org/10.1016/j.psychres.2009.10.015

Taylor, P. J., Gooding, P., Wood, A. M., & Tarrier, N. (2011). The role of defeat and entrapment in depression, anxiety, and suicide. Psychological Bulletin, 137(3), 391–420. https://doi.org/10.1037/a0022935

Taylor, P. J., Wood, A. M., Gooding, P., & Tarrier, N. (2010). Appraisals and Suicidality: The Mediating Role of Defeat and Entrapment. Archives of Suicide Research, 14(3), 236–247. https://doi.org/10.1080/13811118.2010.494138

Van Vliet, K. J., & Kalnins, G. R. C. (2011). A compassion-focused approach to nonsuicidal self-injury. Journal of Mental Health Counseling, 33, 295e311. https://dx.doi.org/10.17744/mehc.33.4.j7540338q223t417.

Welford, M. (2012). The Compassionate Mind Approach to Building Self-Confidence: Series editor, Paul Gilbert (UK ed. edition). Robinson.

Williams, M. (1997). Cry of Pain: Understanding Suicide And Self-Harm. London: Penguin.

Xavier, A., Pinto-Gouveia, J., Cunha, M., & Carvalho, S. (2016). Self-Criticism and Depressive Symptoms Mediate the Relationship Between Emotional Experiences With Family and Peers and Self-Injury in Adolescence. The Journal of Psychology, 150(8), 1046–1061. https://doi.org/10.1080/00223980.2016.1235538

100

PAPER THREE CRITICAL REFLECTION

Choice of research area

Suicide is a major public health issue (ONS, 2016), and increasingly research has focused on understanding the psychological, motivational and volitional factors which lead from experiences of mental distress to suicidal thoughts and behaviours. A range of explanatory models of suicide have been proposed, which reflect our developing understanding of this complex and devastating process.

Self-esteem and self-criticism have been deemed important contributory factors to suicidality in a range of psychological models, which have been described in papers one and two. To date no systematic reviews have been carried out which examined the role of self- esteem in suicide in an adult population, perhaps due to a paucity of prospective or experimental studies in this area. Following initial searches, it was deemed that a review of literature would provide a useful and novel summary of the role of self-esteem in suicidal processes.

Following the review, the clinical application of techniques to improve self-esteem and self-criticism were considered. One area of interest to the author was around novel or third wave cognitive behavioural approaches to the treatment of low self-esteem and self- criticism. Compassion Focused Therapy (Gilbert, 2009, Gilbert & Choden, 2014), proposes a variety of techniques and adaptations to mainstream CBT with a particular focus on individuals who are highly self-critical or prone to shame. Self-criticism in this model pertains to an evolutionary pre-occupation with social rank, and two key processes are identified, one involving strategies to manage feelings of inadequacy (linked to

101 perfectionism) and those of self-attacking (linked to disgust). The model proposes that rather than working on self-esteem directly, with its links to social rank and contingency on external achievement, a focus should be on developing a kind, accepting and non- judgemental approach to the self, alongside taking compassionate action to address key issues. Thus, an empirical investigation was developed to examine the role of self-critical processes in relation to suicide, alongside the potential moderating effect of self- reassurance.

Definitions

One of the themes across both papers one and two were issues with definition of the key concepts within both papers. Literature on self-esteem and related constructs varied widely in definition, measures, and causal factors. Similarly, the literature on self-criticism included varied definitions, described as perfectionistic self-presentation, self-attacking, feelings of inadequacy and judgement towards the self, along with being described as a univariate construct (Luyten et al, 2007, Fazaa, 2003, Satterwhite & Luchner, 2016). Self-criticism has been related to an introjective personality structure, with high levels of striving for achievement, and associated feelings of worthlessness and a feeling of failure to live up to standards (Desmet, Vanheule & Verhaeghe, 2007).

One challenging aspect to consider when undertaking this research was whether both self-esteem and self-criticism represented overlapping or separate constructs. In the review paper the choice of Rosenberg’s (1965) definition of ‘global stable attitude or feeling about the self’, may overlap with the definition of self-criticism as ‘unfavourable or severe judgement of oneself, one's abilities, one's actions’ (Collins, 2016). For the purposes of this

102 thesis both constructs were considered to be separate, with self-criticism being viewed as more of an action, perhaps with different motivational components (i.e. self-improvement/ removal of unwanted aspects of self), rather than a global attitude. However, there is certainly an argument that they could be considered elements of the same self-self-relating process. Indeed, in a recent review by Forrester et al (2017) on the role of self-esteem in non-suicidal self-injury, papers with a focus on self-criticism were included. Definitional issues are not uncommon in the self-esteem literature, and this at times poorly defined construct has been linked to high levels of heterogeneity across study findings (Jordan,

Logel, Spencer, Zanna, & Whitfield, 2009). This is described in more detail below.

Similarly, the literature on suicide has its own issues with definition and measurement, which can make synthesis of results challenging. One of the key issues when identifying risk factors for individuals attempting suicide is that a considerable amount of the literature has conflated suicidal ideation (and associated risk factors), with risk factors for suicidal behaviours (Klonsky & May 2014). Closer examination of the literature suggests that risk factors for these two groups are distinct, something that is reflected in more recent models of suicide (Joiner et al, 2005, O’Connor et al, 2011).

Throughout the suicide literature there are a range of scales used for assessing different elements of suicidal behaviour and thoughts. Some focus on suicidal ideation (for example the Scale for Suicidal Ideation, Beck, 1979), others on suicide risk/ probability (for example the Suicide Probability Scale, Cull and Gill, 1988) and others on specific suicidal behaviours (such as the SBQ-R, Bagge & Osman, 2001). However, even within measures which purport to explicitly measure specific elements of suicidal experience, there can confounding issues, for example the ‘suicide behaviours questionnaire’ (SBQ; Bagge &

103

Osmanl, 2001). The SBQ refers to suicidal ideation and attempts in the same question “Have you ever thought about or attempted to kill yourself?”. It is therefore difficult to accurately differentiate results between ideators and individuals who have carried out suicidal behaviours.

This links to wider definitional issues in the suicide literature, with some considering suicidal behaviours to include both ideation and actions (as described above), rather than attempts. Another related yet distinct descriptor in suicide literature is that of self-harm. In some papers self-harm may encompass any action to self-injure, inclusive of individuals who are attempting suicide and those who are not. Others would see self-harm without explicit suicidal motivation as more accurately described as ‘non-suicidal self-injury’ (Klonsky, May

& Saffer, 2016). For the purposes of this thesis, guidance was followed from the seminal paper by O’Carroll et al (1996), which advised using specific nomenclature for suicidal thoughts and behaviours, along with providing definitions early in a paper, for clarity.

What is self-esteem?

As discussed above, one of the first key questions faced by the author upon commencing the review included establishing inclusion and exclusion criteria. The initial phase of this involved developing an understanding of self-esteem and undertaking sufficient reading to understand the key issues with defining it.

Self-esteem has been described as a multi-dimensional concept, cited frequently in both clinical and social psychological literature (Rosenberg, 1979, Crocker 2002, Leary and

Baumeister, 2000 & De Ruiter, Van Geert & Kunnen, 2017). However, there is still a lack of

104 consensus over its precise definition. The varied definitions and inconsistent empirical findings mean that this concept creates challenges when trying to undertake a systematic and comprehensive review (Heppner and Kernis 2011).

As discussed in paper one the forms and functions of self-esteem have been explained in a variety of different ways. From a ‘global stable attitude or feeling about the self’, the relative contingency of self-esteem on external performance, to how we perceive ourselves as a valued or devalued relational partner (Rosenberg, 1965; Crocker et al, 2002;

Baumeister et al, 2000). Others have referred to self-esteem in term of its self-organising capacity, rather than flowing from an executive agent, making reference to dynamic theory.

The SOSE (Self-Organising Self-Esteem) model described self-esteem as a bidirectional relationship between hierarchically ordered levels (De Ruiter et al, 2017). The challenge with reviewing self-esteem literature was to utilise a definition which would be wide enough to cover the majority of the literature on the topic. Given that Rosenberg’s self-esteem scale is the most frequently used measure of self-esteem, and other commonly used measures are thought to map the same areas (though this is fraught with controversy in itself),

Rosenberg’s (1965) definition was used for the review of a ‘global stable attitude or feeling of self-worth’. It was necessary to define self-esteem in this way, as there are so may varied constructs that have linked or similar meanings. Throughout the screening phase of the review it was challenging to have to consider each construct in turn in relation to the

Rosenberg definition and decide to include and exclude as appropriate.

As discussed above, a number of seemingly self-esteem related concepts such as

‘perceived burdensomeness’, ‘narcissistic wounds’ and ‘self-efficacy’ were examined in papers being considered for inclusion in this review. It was decided that these constructs

105 were sufficiently dissimilar from the definition of self-esteem as a ‘global evaluation of self- worth’ (Rosenberg, 1965) set out prior to the commencement of the review and were thus excluded on a case by case basis. For example, ‘perceived burdensomeness’ was discussed and considered to relate to external factors – our perception of others’ judgements about the self, rather than our own judgement. Similarly, ‘narcissistic wounds’ was discussed, in the literature it was considered a component of ‘mental pain’, and was thought to comprise of both external factors (someone enacts a wound upon you) and judgements of self- worth.

Where inclusion was unclear the reviewer highlighted each paper/ construct, its definition and use within the paper in question and was discussed with an independent party

(research supervisors) in relation to self-esteem as defined by Rosenberg (1965). Constructs such as ‘worthlessness’, ‘defectiveness’ and ‘negative self-evaluation’ were included as they were considered a better fit with the definition of self-esteem as ‘a global self-evaluation’.

106

PAPER ONE CRITICAL REFLECTION

Prior to carrying out the review it was decided that a systematic review would be the most appropriate methodology to use to address the research question. A systematic review would allow results to be replicated, and for a broad search of published literature to be carried out. The synthesis of this literature would provide the first attempt to draw together systematically report the findings of papers on self-esteem and suicide, and thus provide a useful summary of results for future researchers and clinicians.

SEARCH TERMS

When considering the review topic initial thoughts were to explore the literature on relationship between self-criticism and suicide. This would link to the researchers interests in compassion focused therapy (Gilbert, 2009). Initial searches suggested a limited number of papers using the terms SUICID* and SELF-CRITICISM, and when further terms were added

(SELF-ATTACKING, SELF-HATRED, SELF-DISGUST, INADEQUACY AND SELF-LOATHING), the evidence base did not appear sufficient to warrant a review. Further searches were undertaken of self-related factors, including suicide* and ‘SELF-ESTEEM, SELF-CONCEPT,

SELF-IMAGE, SELF-APPRAISAL’. Searching self-esteem alone resulted in a high number of papers (n=2914) and self-esteem has been shown to be an established correlate of suicide, with prominence in both the ‘SAMS’ model (Johnson et al, 2008) and earlier models of suicide such as Baumeister (1990). The relationship between self-esteem and suicide have never been the focus of a systematic review, as such the paper would contribute a valuable summary of key findings.

107

Once the topic was chosen, further synonyms were elicited by mapping the term to subject heading and generating MESH terms, these included ‘SELF-CONCEPT, SELF-

PERCEPTION, SELF-ESTEEM’. A thesaurus was also used to develop further terms, alongside two published papers (Randall et al, 2015, Kesting & Lincoln, 2013) including ‘SELF-WORTH,

SELF-IMAGE, SELF-REGARD, SELF-CONFIDENCE, SELF-EVALUATION, SELF-LIKING, SELF-

ATTITUDE, SELF-COMPETENCE AND SELF-APPRAISAL’.

Inclusion and exclusion criteria

As discussed above, a key element of the inclusion and exclusion criteria related to understanding the definitions and measurement of all key variables of interest. Following this, decisions were made around further inclusion/ exclusion criteria.

As discussed in both papers one and two, the salience of self to self relating in the suicide literature began with a key marker paper by Baumeister (1990). This paper described suicide as a means of escaping aversive self-awareness and negative perceptions of the self, a theme that was taken up in subsequent theories including the cry of pain model (Williams,

1997), the SAMS model (Johnson, 2008) and more recently in the IMV model of suicide

(O’Connor et al, 2011) via its moderating impact on defeat and entrapment. The decision was made to therefore restrict the review to papers from 1990 onwards, to ensure the review was current, and encompassed leading models of suicide and their relationship to attitudes towards the self.

In terms of participant age, at the commencement of the review it was the authors understanding that there were existing literature reviews around self-esteem based on an adolescent population, which included sections on suicide (Emler, 2001). There have been no systematic reviews on the role of self-esteem in suicide in an adult population. Given the

108 pressing concerns around adult male suicide rates, which show males age 45-49 at the highest risk of completed suicide (Samaritans, 2016), it was decided that adults should be the focus of the review. This was to obtain an understanding of the role of self-esteem in adult suicidal ideation and behaviour. Part of the rationale for this was that it was felt by the author that times of transition in terms of age (i.e. from child to adult, adult to older adult) could be a particularly unstable time in terms of self-esteem, with fluctuations being found between relatively short time points (Orth, Trzeniewski & Robins, 2010; Harter and

Whitesell, 2003). Adolescence is a time when the self is still very much forming (Marcia,

1980) and it was felt that a separate review of the relationship of self-esteem to suicide aimed solely at adolescents would be more appropriate. Similarly, when considering the literature on older adults and the factors associated with suicide, initial searches suggested that key issues around perceived burdensomeness, role loss and perceived social support would have unique cohort effects on self-esteem and suicide (Harrison et al, 2010;

Cukrowitz et al, 2011). Thus, the recommendation would be that a separate review of literature involving older adults (65+) would be most appropriate. Based upon these initial findings, a focus on adult participants, age 18-65 was considered to be the most appropriate inclusion criteria.

A decision was made to focus on published peer reviewed literature. It has been suggested that reviews which do not examine grey literature may be open to publication bias (Button et al., 2013). However, it was felt that a) due to the limited time and resources available as a ClinPsyD student the review would be limited to published papers and b) as many sources of grey literature only include an abstract, it would not be possible to consider their comprehensive quality assessment, alongside full text publications.

109

Measurement of self-esteem as a construct

It was decided that given the complexities of defining self-esteem in the literature, the use of a validated measure of self-esteem would be required as an inclusion criterion for each study. However, there are complexities in definition and constituent factors of self-esteem even when validated measures are used. There are also debates around whether self- esteem can even be measured, or whether self-report scales by their nature are open to bias based on an individual’s desire to appear socially desirable.

Commonly used scales in self-esteem research, such as the Rosenberg Self Esteem

Scale (1965) or the Coopersmith self-esteem scale (1967), generally considered reliable and valid in a clinical sense, are thought to measure enduring, global ‘trait’ self-esteem.

However, as described above there is some debate as to the accuracy of explicit measures such as these in the measurement of true self-esteem. Indeed, some studies have reported that individuals with narcissism, conceitedness or grandiose beliefs about the self can score highly on explicit self-esteem measures (Campbell, Rudich & Sedikides, 2002; Heppner &

Kernis, 2011). However, this may be masking crippling experiences of self-loathing or self- hatred which may only present themselves in times of personal crisis (Tanner and Webster,

2003).

One resolution of the issue with distortion of measurements of self-esteem on explicit measures is to use implicit tasks to measure self-esteem. Implicit tasks were found in one study in the review and a number of excluded full text studies. One commonly used tool to test implicit self-esteem is the Name-Letter Test (Nuttin, 1985), where participants are asked to rate the letters of the alphabet, with the hypothesis that people with higher trait self-esteem will rate the letters of their own name with a higher score. This has been

110 criticised by Krizan (2008) with regards respondents’ awareness of the self-referential nature of the test, the implication being that conscious beliefs may have moderated results.

Albers, Rotterveel & Dijksterhuis (2009) have also criticised the test suggesting results derive from an individual’s preference for letters generally, and suggested that the test should be adapted with ‘letter liking’ controlled for in the analysis. Implicit measures of self-esteem such as the name letter test and implicit associations test were criticised in a review by

Buhrmester et al (2011). This area of research is still relatively in its infancy and as such further evidence is required before these types of test can be considered a reliable representation of implicit self-esteem. However, it was interesting to reflect on differences in outcomes when using explicit vs implicit measures in the literature review.

Issues with the assessment and definition of self-esteem are also reflected in the diversity of measures associated with it. Table one comprises of available factor analyses of key measures of global or explicit self-esteem. As the table suggests, despite the significant amount of self-esteem research available, it is not clear whether measures are examining the same construct, which may call into question the reliability and validity of pooled results. This is an issue across the self-esteem literature, rather than just pertaining to this review, and more consistent and specific measurements of self-esteem is required to increase reliability of findings. Thus, it may be appropriate for self-esteem researchers going forwards to either a) seek consensus on the definition of self-esteem, in a similar way to the

O’Carroll (1996) paper in suicide research, or b) being explicit in terms of language and definitions when researching specific factors associated with self-esteem such as academic performance vs global or trait aspects.

111

Table 1 Factor analysis of self-esteem measures

Name of measure Factors within scale Author

Rosenberg Self- Up to 6 proposed factors (limited consensus on esteem Scale (RSES) this) Global Self-esteem Rosenberg (1965) Positive and Negative Self-Image Urban et al Method factors (2014) Self-acceptance and self-assessment Tafarodi and Milne (2002) Coopersmith Self- Anxiety Myhill and Lorr esteem inventory Defensiveness (1978) Negative Social Attitude Rejection of Self Inadequacy of Self Beck Self-esteem Views of self (self scale) Beck et al (2001) scale Views of others’ perceptions of self (other scale) Robson Self-concept Correlates highly with RSES Ghaderi (2005) questionnaire 7 proposed factors Robson (1989) (1) subjective sense of significance, (2) worthiness, (3) appearance and social acceptability, (4) competence, (5) resilience and determination, (6) control over personal destiny, and (7) the value of Existence Self-esteem rating Unidimensional factor – self-esteem Nugent and scale Thomas (1993)

Non-Suicidal Self Injury (NSSI) vs Suicidality

NSSI has been linked to suicidal thoughts and behaviour in a number of studies, particularly in young adults (Whitlock et al, 2013; Andover, Morris, Wren & Bruzzese, 2012). However, some suggest that NSSI is not linked in a linear manner to suicide (Andrewes, Hulbert,

Cotton, Betts, & Chanen, 2017). Emerging evidence suggests NSSI is different in its intent, function and epidemiology to suicidal thoughts and behaviour (Butler and Malone, 2013).

112

For this reason, studies with an explicit focus on NSSI without suicidal intent (specified in their inclusion/ exclusion criteria) were excluded from the review.

Analysis

High levels of heterogeneity were found in the meta-analysis which might be expected given the variety of study locations, populations and variables assessed, thus the inverse variance heterogeneity (IVhet) model was used which accounts for the impact of random effects in the data.

Clinical Implications

As discussed in paper one, there is some evidence for the examination of self-esteem as a risk factor in suicidal thoughts and behaviour. Although the evidence was limited, it also suggests that low levels of self-esteem are related more strongly to suicidal ideation to actual attempts or behaviour. It is hypothesised that this may be due to the staged nature of increasing levels of suicidality – with different motivational or volitional factors gaining salience at each stage (see Klonsky et al, 2014; 2016 for a review). It may be that self- esteem is a more important pre-disposing factor in the earlier stages of suicidality, those involving vulnerabilities to experiences of defeat and consequent experiences of entrapment. As discussed in paper one, it may be useful to consider interventions aimed at increasing self-esteem in individuals at risk of suicide. This may be a particularly important facet of any treatment plans aimed at improving mental health, as some research suggests

113 that pre-existing self-esteem levels predict therapeutic outcomes, with those having higher self-esteem as more likely to improve (Parker, Page and Hooke, 2013).

One intervention that has shown effectiveness in the treatment of self-esteem is cognitive behavioural therapy (CBT). A number of studies have shown that the use of CBT can improve scores on Rosenberg Self-Esteem Scale in populations including patients with dual diagnosis and psychosis (Oestrich, Austin, Lykke & Tarrier, 2007; Gumley et al, 2006;

Hall & Tarrier, 2003).

114

PAPER TWO CRITICAL REFLECTION

Rationale for Joint Project

The data collection phase of the study involved the collection of data by three clinical psychology trainees with an interest in suicidality. Jointly named the BECS study (Beliefs,

Emotion, Criticism in Suicide), this approach allowed for a maximum number of participants to be recruited, whilst also limiting potential participant burden and potential impact on clinicians when seeking referrals. Each trainee undertook a separate and distinct literature review related to their specific area of interest and generated independent research questions. A battery of measures was then decided upon for the empirical investigation, which allowed for a range of variables of interest to be assessed, whilst again minimising the impact of participant burden. In terms of participant burden, the time taken to complete varied considerably between participants. The Trainees had originally timed completion of measures to be around 1 hour, which was felt to be an acceptable time period following consultation with the INSURG, service user reference group. However, a number of individuals took considerably longer than this, requiring 2 or 3 visits. This may have been because many preferred to have the questionnaires read to them, interview style, due to literacy and concentration issues. Many service users wanted to go beyond the simple e.g.

0-7 answers and share something of their experiences in relation to items. Researchers had to strike a balance between ensuring people felt comfortable and validated, along with maintaining a focus on the completion of questionnaires, particularly given the distressing nature of some of the items.

115

The Trainees had links with the INSITE study, an NIHR pilot study examining the impact of suicide prevention CBT on an inpatient population (Haddock et al, 2016). This study was collecting data that matched key measures of interest to the Trainees (including the Suicide Probability Scale, Beck Hopelessness Scale, Calgary Depression Scale and Defeat and Entrapment Scales). The research team kindly allowed the Trainees to add their specific measures to the INSITE battery (for the purposes of this paper this was the Forms of Self-

Criticising/ Attacking and Self-Reassuring Scale; FSCRS). Ethical approval for this was obtained and BECS Trainee’s were involved in supporting the research team in collecting data from the inpatient wards and at follow up.

In terms of individual contributions, it was agreed that trainees would take an equal role in the development of the study protocol and associated documents, along with allocating sections of the IRAS ethics form to complete. Each trainee recruited community participants from individual NHS Trusts. The author recruited from Manchester Mental Health and Social

Care Trust, as they had previously worked as a Clinical Studies Officer in this Trust and had relevant contacts for recruitment (see appendix H). Personal contribution to recruitment was to undertake a number of team presentations in both community mental health teams, three crisis resolution and home treatment teams, and at three primary care mental health teams within the Trust. This was in order to obtain a broad spectrum of severity of suicidality, from relatively mild levels of suicidal ideation to severe and high-risk clients on the verge of inpatient admission.

116

Design

A within subjects cross-sectional design was utilised to answer the research question. In the initial stages of the project this design was felt to allow for the most variables to be considered simultaneously, across a wide population of service users. However, if there had been sufficient time, it would have been useful to invite participants to complete measures at two time points to allow the predictive value of variables to be tested, particularly to see if high scores for suicide probability and self-attacking resulted in increased levels of suicidal behaviours at subsequent time points.

It was also noted while conducting interviews that many clients had not had the opportunity to talk to someone about some of the very dark and distressing experiences they had had. Indeed, it was reported by a number of participants that most people do not want to talk about suicide in any depth or participants were worried about burdening care teams or relatives. As such it would have been helpful to have some way of capturing the experiences of those living with suicidality in a richer way, perhaps using qualitative or mixed methods. Many of the participants in the study reflected that it was ‘nice to be able to use their negative experiences to help others’, and found it useful to speak about their experiences. This reflects findings by Awenat et al (2017) who highlighted the qualitative experiences of service users involved in suicide research – with key themes around ‘being listened to’, ‘mutual respect’ and ‘being part of a change’.

117

Measures

One key factor when deciding on measures to include in the study was participant burden.

Trainees ran trials of the questionnaire battery within their team and with service users to ensure that the time to complete was not overly long. The aim was for questionnaires to be completed in roughly one hour. Researchers were aware of the highly emotive nature of questionnaires around suicide and self-attacking, and that sufficient time would need to be included in the session for a debrief at its conclusion. For this reason the Forms of Self-

Criticism/ Attacking and Reassuring (Gilbert et al; 2004) scale was chosen, as its three subscales allowed for an examination of different facets of self-criticism and self- reassurance without being overly long (21 items). One reflection on the choice of this measure is whether the ‘self-reassurance’ subscale really captures the variable of self- compassion, which the researcher was interested in. There have been some papers which have explored the idea that this scale consists of a single factor, with self-criticism and self- reassurance as separate ends of this continuum (Kupeli, Chilcot, Schmidt, Campbell & Troop,

2012). A broader measure of self-compassion might include the Self-Compassion Scale

(Neff, 2003a; 2003b), which examines self-compassion in terms of three components; kindness towards self (similar to the self-reassurance element of the FSCRS), common humanity in terms of the experience of failure and pain and mindful awareness and acceptance of negative emotional states. If participant burden had not been a factor then including this measure may have resulted in a greater understanding of the potential moderating role of self-compassion on self-criticism and its relationship to suicidality.

During the interviews, one of the key issues noted by many participants was the impact of recent significant life events that preceded periods of increased suicidality. These

118 might have included relationship breakdown, bullying and harassment at work or changes to benefits and financial hardship. If the study were to run again, the researcher would like to include a measure of significant life events such as the Life Events Inventory (Spurgeon,

Jackson & Beech 2001).

Ethics

As mentioned previously, Trainees needed to ensure that participant safety and support was properly considered and built into the study design. Consequently, considerable thought was put into the ethical issues that surrounded discussing suicide with vulnerable individuals. In the initial stages of the study, researchers met with the INSURG (INSITE service user reference group), PAPYRUS (a young suicide charity) and The Sanctuary (a third sector crisis service) to obtain feedback on study topic, design and to consider what supports might need to be in place for participants following the interview sessions. The outcome of these discussions, along with discussion with experienced suicide researchers at the University of Manchester, resulted in the production of the risk management protocol which can be found in appendix D. The research ethics committee reflected that the study had clearly demonstrated a lot of thought and sensitivity when it came to working with this population, and ethical approval was given (see appendix G).

Analysis

Previous research using the FSCRS has also found high levels of intercorrelation with the inadequacy and self-attacking subscales of the FSCRS, between r=0.68 and r=0.80 (Gilbert et

119 al, 2004, Irons et al (2006), Richter et al, 2009 and Gilbert et al, 2010). This has been managed in a variety of ways, from removing one or other of the measures, combining the measures, or examining the constructs separately. With regards the current study, separate regression analyses were carried out for each variable, limiting any potential confounding effects.

Personal reflections

In carrying out the research, the researcher was struck by the high levels of mental distress participants had become habituated to living with – with disclosures of ongoing intense suicidality being part of everyday life for many. In line with this was a reflection on the relative attitudes towards this risk by clinicians working at different levels within the mental health care system. It was interesting to see that many primary care mental health practitioners were very protective of clients and concerned that suicide research may cause acute distress to relatively well participants. This was contrasted with those working with very high levels of risk in the community, such as crisis resolution and home treatment teams, who perceived relatively severe clients as moderate given their ongoing caseload.

Something that might be of interest for future studies in this area is to consider staffing groups’ perceptions of suicide risk and how that impacts on clinical decisions and referral patterns. Anecdotally it appeared that what would be considered severe or high risk in primary care, might be perceived as relatively mild by individuals working in inpatient or crisis settings. Decisions about access to services were made on the basis of individual team consensus and this may lead to certain groups of patients falling through the gap between services (too high risk for one, not high enough for another). Perhaps one resolution to this

120 could involve risk events with different staff teams, where differing perceptions of risk using clinical case studies could be highlighted and consensus gained. Another suggestion could be joint allocation meetings, where individuals needing to be ‘stepped up or down’ in terms of the intensity of their care with regards risk, could be discussed.

Other areas that impacted upon the researcher were the loneliness and isolation experienced by many participants in the study. Many participants reported the relief at being able to speak about ongoing suicidal experiences with someone, particularly someone they did not feel as concerned about being a burden to, given that we were ‘experts’ in suicide and used to hearing such material. As with all research conducted by clinicians, it was hard at times to be strictly research focused – just filling in 0-7 on questionnaires.

People wanted to share the meaning and reflections on questions, but it was important to maintain a research focus, rather than straying into pseudo-counselling outside of the normal therapeutic frame.

There was also the consideration of self-care when discussing and researching the material for this thesis. Not only were interviews distressing at times, with high risk clients disclosing current suicidal plans and actions, but also reading through literature on the relationship between suicide and self-attacking, self-hatred and disgust could at times be quite depressing. Ensuring that supervision was used appropriately, and utilising the support of BECS team mates was an invaluable resource in managing any feelings that came up during the research. A recent BPS (2017) position statement on understanding and preventing suicide, suggests that friends or relatives who have been exposed to suicide

121 deaths are at risk of post-traumatic responses, along with increased risk of suicide.

Increased access to psychological support, along with organisational ‘post-vention’ may be useful in providing support to those effected.

Clinical Implications

The results suggest that the self-attacking form of self-criticism, entrapment and hopelessness all have a significant relationship with suicidality. This is over and above depression, which is considered an established predictor of suicide.

In terms of clinical work, this may mean that increasingly self-attacking self- statements could be considered a risk factor for suicide, particularly when paired with disclosures around hopelessness and increasing volitional factors such as suicide plans and models of suicidal behaviour. Clinicians should consider being alert to suicide risk in client groups where self-attacking thoughts are prevalent, such as individuals living with eating disorders, those who self-harm or present with low self-esteem (Duarte & Pinto-Gouveia

2017; Gilbert et al, 2010).

As discussed in paper two there is an emerging evidence base for interventions that target self-criticism. One area that shows promise is the use of self-compassion, where participants learn to tolerate and accept ‘unacceptable’ parts of themselves, and approaches are used which activate the vagal nerve and attachment pathways of the brain to sooth self-critical thoughts (Falconer, King and Brewin, 2015, Gilbert, 2005; Gilbert and

Choden, 2014). Indeed, recent evidence suggests that health care professionals would also

122 benefit from this approach, using compassion focused therapy training to reduce self- criticism and self-persecution (Beaumont, Irons, Rayner & Dagnall, 2017).

Dissemination

Study results will be developed into a service user guide and disseminated to both individuals who participated in the study, and also those community groups who supported the study in the early stages (Papyrus, the Sanctuary and the INSURG). The researcher will also present results at team meetings for those teams who participated in recruitment.

There are plans to develop a poster to exhibit at the annual Compassionate Mind conference. Both systematic review and experimental study will be submitted for peer review in academic journals.

The process of suicide research

Given the reflections discussed above, a number of key recommendations are proposed for future suicide researchers:

• Become aware of issues around terminology from the start. Read O’Carroll (1996) to

consider definitions of suicidal thoughts, behaviours, attempts

• Consult with relevant patient organisations such as Papyrus and the local Sanctuary

(crisis) service to obtain feedback on topic choice, discuss potential impact on both

service users and researchers and discuss organisational self-care strategies

• Develop a service user reference group where possible; to discuss topic in depth,

along with consideration of participant burden, components of post-interview

debrief session and signposting to sources support for participants

123

• When conducting research, ensure that from the start participants are aware of the

potentially distressing nature of the topic and right to refuse/ withdraw

• However, balance this with information from qualitative and quantitative literature

around the benefits of participation in suicide research, around being heard and

contributing to change (Gould et al, 2005; Smith, Poindexter & Cukrowicz, 2010)

• Prepare for flexibility in terms of time spent interviewing; participants may wish to

share more of their experiences and reflections and this must be balanced with a

strict research focus in interviews

• Have robust risk protocols in place and emergency procedures if a client discloses

active and current suicide risk in session – GP/ care-coordinator details, emergency

contacts list, and obtain support from wider research team to discuss incidents

• Regular review of any risk issues and their management in supervision

• Self-care strategies – speak with colleagues and supervisors for support, tap into

wider research group with interest in this area to discuss (suicide is not a topic many

non-mental health professionals want to speak about). Take regular breaks and

acknowledge that this topic can be distressing.

• On a personal level – find a way to tap into the ‘silly’ side of life. Suicide is a very dark

topic, it can help to spend time having fun and spending time playing with children,

loved ones and remembering the lighter side of life. This was also recommended by

a staff member at ‘The Sanctuary’ Manchester.

124

OVERALL CONCLUSIONS

This thesis examined the role of self-to-self relating in suicide – namely the role of self- esteem and self-criticism in suicidal processes. The systematic review demonstrated a negative relationship between self-esteem and suicide, with potential future avenues of research to consider the relative fragility or contingency of self-esteem as an important additional dimension to this relationship. Paper two looked at specific types of self-criticism and their relationship to suicidality. The results suggest that in a suicidal population, inadequacy/ perfectionism has less predictive power with regards suicidality than self- attacking, entrapment and hopelessness. Also of note is that these factors supersede depression in terms of strength of prediction of suicide. The moderation analysis suggested that self-reassurance, whilst demonstrating a negative relationship with both types of self- criticism, did not moderate their relationship with suicide.

The research overall justifies the continued focus on self-related processes in explanatory models of suicide, and offers novel suggestions for future research in this area

125

REFERENCES

Albers, L., Rotteveel, M., & Dijksterhuis, A. P. (2009). Towards Optimizing the Name Letter Test as a Measure of Implicit Self-esteem. Self and Identity, 8(1), 63–77. https://doi.org/10.1080/15298860802091062

Andrewes, H. E., Hulbert, C., Cotton, S. M., Betts, J., & Chanen, A. M. (2017). Patterns of Non-Suicidal Self-Injury and Their Relationship with Suicide Attempts in Youth with Borderline Personality Disorder. Archives of Suicide Research: Official Journal of the International Academy for Suicide Research, 1–14. https://doi.org/10.1080/13811118.2017.1358226

Awenat, Y. F., Moore, C., Gooding, P. A., Ulph, F., Mirza, A., & Pratt, D. (n.d.). Improving the quality of prison research: A qualitative study of ex-offender service user involvement in prison suicide prevention research. Health Expectations, n/a-n/a. https://doi.org/10.1111/hex.12590

Baumeister, R. F. (1990). Suicide as escape from self. Psychological Review, 97(1), 90–113.

Bagge, C., & Osman, A. (1998). The Suicide Probability Scale: Norms and factor structure. Psychological Reports, 83, 637–8. https://doi.org/10.2466/PR0.83.6.637-638

Beaumont, E., Irons, C., Rayner, G., & Dagnall, N. (2016). Does Compassion-Focused Therapy Training for Health Care Educators and Providers Increase Self-Compassion and Reduce Self-Persecution and Self-Criticism? The Journal of Continuing Education in the Health Professions, 36(1), 4–10. https://doi.org/10.1097/CEH.0000000000000023

Beck, A. T., Kovacs, M., & Weissman, A. (1979). Assessment of suicidal intention: the Scale for Suicide Ideation. Journal of Consulting and Clinical Psychology, 47(2), 343–352.

Beck, A. T., Brown, G. K., Steer, R. A., Kuyken, W., & Grisham, J. (2001). Psychometric properties of the beck self- esteem scales. Behaviour Research and Therapy, 39(1), 115–124.

Butler, A. M., & Malone, K. (2013). Attempted suicide v. non-suicidal self-injury: behaviour, syndrome or diagnosis? The British Journal of Psychiatry, 202(5), 324–325. https://doi.org/10.1192/bjp.bp.112.113506

Campbell, W. K., Rudich, E. A., & Sedikides, C. (2002). Narcissism, self-esteem, and the positivity of self-views: Two portraits of self-love. Personality and Social Psychology Bulletin, 28(3), 358–368.

Collett, N., Pugh, K., Waite, F., & Freeman, D. (2016). Negative cognitions about the self in patients with persecutory delusions: An empirical study of self-compassion, self-stigma, schematic beliefs, self-esteem, fear of madness, and suicidal ideation. Psychiatry Research, 239, 79–84. https://doi.org/10.1016/j.psychres.2016.02.043

Coopersmith, S (1967) The antecedents of self-esteem. SanFrancisco: Freeman

Crocker, J. (2002). The Costs of Seeking Self–Esteem. Journal of Social Issues, 58(3), 597–615. https://doi.org/10.1111/1540-4560.00279

126

Cull, J. G., & Gill, W. S. (1988). Suicide probability scale (SPS) manual. Los Angeles: Western Psychological Services

Cukrowicz, K. C., Cheavens, J. S., Van Orden, K. A., Ragain, R. M., & Cook, R. L. (2011). Perceived Burdensomeness and Suicide Ideation in Older Adults. Psychology and Aging, 26(2), 331–338. https://doi.org/10.1037/a0021836

Desmet, M., Vanheule, S., & Verhaeghe, P. (2006). Dependency, self-criticism, and the symptom specificity hypothesis in a depressed clinical sample. Social Behavior and Personality: An International Journal, 34, 1017– 1026. https://doi.org/10.2224/sbp.2006.34.8.1017

Dictionaries, C. (2016). Collins English Dictionary and Thesaurus Essential edition: All-in-one support for everyday use (01 edition). Collins.

Duarte, C., & Pinto-Gouveia, J. (2017). The impact of early shame memories in Binge Eating Disorder: The mediator effect of current body image shame and cognitive fusion. Psychiatry Research, 258(Supplement C), 511–517. https://doi.org/10.1016/j.psychres.2017.08.086

Emler, Nicholas (2001) Self esteem: the costs and causes of low self worth. . York Publishing Services, York

Falconer, C. J., King, J. A., & Brewin, C. R. (2015). Demonstrating mood repair with a situation‐based measure of self‐ compassion and self‐criticism. Psychology and Psychotherapy, 88(4), 351–365. https://doi.org/10.1111/papt.12056

Fazaa, N & Page, S (2003) Dependency and self-criticism as predictors of suicidal behaviour. Suicide Life Threat Behaviour, Summer; 33(2):172-85.

Forrester, R. L., Slater, H., Jomar, K., Mitzman, S., & Taylor, P. J. (2017). Self-esteem and non-suicidal self-injury in adulthood: A systematic review. Journal of Affective Disorders, 221, 172–183. https://doi.org/10.1016/j.jad.2017.06.027

Ghaderi, A. (2005). Psychometric Properties of the Self-Concept Questionnaire. European Journal of Psychological Assessment, 21, 139–146. https://doi.org/10.1027/1015-5759.21.2.139

Gilbert, P. (2009). Introducing compassion-focused therapy. Advances in Psychiatric Treatment, 15(3), 199–208. https://doi.org/10.1192/apt.bp.107.005264

Gilbert, P. P. (2014). Mindful Compassion: How the Science of Compassion Can Help You Understand Your Emotions, Live in the Present, and Connect Deeply with Others. Oakland, CA: New Harbinger.

Greco, T., Zangrillo, A., Biondi-Zoccai, G., & Landoni, G. (2013). Meta-analysis: pitfalls and hints. Heart, Lung and Vessels, 5(4), 219–225.

Gumley, A., Karatzias, A., Power, K., Reilly, J., McNay, L., & O’Grady, M. (2006). Early intervention for relapse in schizophrenia: Impact of cognitive behavioural therapy on negative beliefs about psychosis and self-esteem. British Journal of Clinical Psychology, 45(2), 247–260. https://doi.org/10.1348/014466505X49925

127

Hall, P. L., & Tarrier, N. (2003). The cognitive-behavioural treatment of low self-esteem in psychotic patients: a pilot study. Behaviour Research and Therapy, 41(3), 317–332. https://doi.org/10.1016/S0005-7967(02)00013-X

Harrison, K. E., Dombrovski, A. Y., Morse, J. Q., Houck, P., Schlernitzauer, M., Reynolds, C. F., & Szanto, K. (2010). ALONE? PERCEIVED SOCIAL SUPPORT AND CHRONIC INTERPERSONAL DIFFICULTIES IN SUICIDAL ELDERS. International Psychogeriatrics / IPA, 22(3), 445–454. https://doi.org/10.1017/S1041610209991463

Harter, S., & Whitesell, N. R. (2003). Beyond the Debate: Why Some Adolescents Report Stable Self-Worth Over Time and Situation, Whereas Others Report Changes in Self-Worth. Journal of Personality, 71(6), 1027–1058. https://doi.org/10.1111/1467-6494.7106006

Heatherton, T. F., & Polivy, J. (1991). Development and validation of a scale for measuring state self-esteem. Journal of Personality and Social Psychology, 60(6), 895-910. http://dx.doi.org/10.1037/0022-3514.60.6.895

Irons, C., & Beaumont, E. (2017). The Compassionate Mind Workbook: A step-by-step guide to developing your compassionate self (01 edition). Place of publication not identified: Robinson.

Joiner, T. (2005). Why People Die by Suicide (1 edition). Cambridge, Mass.: Harvard University Press.

Jordan, C., Logel, C., Spencer, S., Zanna, M., & Whitfield, M. (2009). The heterogeneity of self-esteem: Exploring the interplay between implicit and explicit self-esteem. In Attitudes: Insights from the New Implicit Measures (pp. 251–284). https://doi.org/10.4324/9780203809884

Klonsky, E. D., & May, A. M. (2014). Differentiating suicide attempters from suicide ideators: a critical frontier for suicidology research. Suicide & Life-Threatening Behavior, 44(1), 1–5. https://doi.org/10.1111/sltb.12068

128

Klonsky, E. D., May, A. M., & Saffer, B. Y. (2016). Suicide, Suicide Attempts, and Suicidal Ideation. Annual Review of Clinical Psychology, 12, 307–330. https://doi.org/10.1146/annurev-clinpsy-021815-093204

Krizan, Z. (2008). What is implicit about implicit self-esteem? Journal of Research in Personality, 42(6), 1635–1640. https://doi.org/10.1016/j.jrp.2008.07.002

Kupeli, N., Chilcot, J., Schmidt, U. H., Campbell, I. C., & Troop, N. A. (2013). A confirmatory factor analysis and validation of the forms of self-criticism/reassurance scale. British Journal of Clinical Psychology, 52(1), 12–25. https://doi.org/10.1111/j.2044-8260.2012.02042.x

Luyten, P., Sabbe, B., Blatt, S. J., Meganck, S., Jansen, B., De Grave, C., … Corveleyn, J. (2007). Dependency and self- criticism: relationship with major depressive disorder, severity of depression, and clinical presentation. Depression and Anxiety, 24(8), 586–596. https://doi.org/10.1002/da.20272

Marcia, J. E. (1980). Identity in adolescence. In J. Adelson (Ed.), Handbook of adolescent psychology (pp. 159– 187). New York: Wiley.

Myhill, J., & Lorr, M. (1978). The coopersmith self-esteem inventory: Analysis and partial validation of a modified adult form. Journal of Clinical Psychology, 34(1), 72–76. https://doi.org/10.1002/1097- 4679(197801)34:1<72::AID-JCLP2270340115>3.0.CO;2-T

Neff, K. D. (2003a). Development and validation of a scale to measure self-compassion. Self and Identity, 2, 223–250.

Neff, K. D. (2003b). Self-compassion: An alternative conceptualization of a healthy attitude toward oneself. Self and Identity, 2, 85–102.

Nosek, B. A., Mokrysz, C., Robinson, E. S. J., Ioannidis, J. P. A., Flint, J., Button, K. S., & Munafò, M. R. (2013). Power failure: why small sample size undermines the reliability of neuroscience. Nature Reviews Neuroscience, 14(5), 365. https://doi.org/10.1038/nrn3475

Nugent, W. R., & Thomas, J. W. (1993). Validation of a Clinical Measure of Self-Esteem. Research on Social Work Practice, 3(2), 191–207. https://doi.org/10.1177/104973159300300205

Nuttin, J. M. (1985). Narcissism beyond Gestalt and awareness: The name letter effect. European Journal of Social Psychology, 15(3), 353–361. https://doi.org/10.1002/ejsp.2420150309

O’Carroll, P. W., Berman, A. L., Maris, R. W., Moscicki, E. K., Tanney, B. L., & Silverman, M. M. (1996). Beyond the Tower of Babel: a nomenclature for suicidology. Suicide & Life-Threatening Behavior, 26(3), 237–252.

O’Connor, R. C. (2011). Towards an integrated motivational–volitional model of suicidal behaviour. International Handbook of Suicide Prevention: Research, Policy and Practice, 181–198.

Oestrich, I. H., Austin, S. F., Lykke, J., & Tarrier, N. (2007). The Feasibility of a Cognitive Behavioural Intervention for Low Self-Esteem within a Dual Diagnosis Inpatient Population. Behavioural and Cognitive Psychotherapy, 35(4), 403–408. https://doi.org/10.1017/S1352465807003633

Orth, U., Trzesniewski, K. H., & Robins, R. W. (2010). Self-esteem development from young adulthood to old age: a cohort-sequential longitudinal study. Journal of Personality and Social Psychology, 98(4), 645–658. https://doi.org/10.1037/a0018769

129

Parker, T. J., Page, A. C., & Hooke, G. R. (2013). The influence of individual, group, and relative self-esteem on outcome for patients undergoing group cognitive-behavioural therapy treatment. British Journal of Clinical Psychology, 52(4), 450–463. https://doi.org/10.1111/bjc.12029

Randal, C., Pratt, D., & Bucci, S. (2015). Mindfulness and Self-esteem: A Systematic Review. Mindfulness, 6(6), 1366– 1378. https://doi.org/10.1007/s12671-015-0407-6

Robson, P. (1989). Development of a new self-report questionnaire to measure self esteem. Psychological Medicine, 19(2), 513–518. https://doi.org/10.1017/S003329170001254X

Rosenberg, M. (1979). Conceiving the self. Basic Books.

Rosenberg, M. (1965). Society and the Adolescent Self-Image. Princeton: Princeton University Press.

Satterwhite, A., & Luchner, A. (2016). Exploring the Relationship among Perceived Resilience, Dependency, and Self- Criticism: The Role of Culture and Social Support. North American Journal of Psychology, 18, 71–84.

Schwartz, S. J., Luyckx, K., & Vignoles, V. L. (Eds.). (2011). Handbook of Identity Theory and Research. New York, NY: Springer New York. https://doi.org/10.1007/978-1-4419-7988-9

Spurgeon A, Jackson CA, Beach, JR. The Life Events Inventory: Re-scaling based on an Occupational Sample. Occup Medicine 51(4): 287-293.

Tafarodi, R. W., & Milne, A. B. (2002). Decomposing Global Self-Esteem. Journal of Personality, 70(4), 443–484. https://doi.org/10.1111/1467-6494.05017

Tanner, C. and Webster, P., 2003. Nacissism - A CAT Perspective. Reformulation, Summer, pp.16-18

Taylor, P. J., Awenat, Y., Gooding, P., Johnson, J., Pratt, D., Wood, A., & Tarrier, N. (2010). The subjective experience of participation in schizophrenia research: a practical and ethical issue. The Journal of Nervous and Mental Disease, 198(5), 343–348. https://doi.org/10.1097/NMD.0b013e3181da8545

Urbán, R., Szigeti, R., Kökönyei, G., & Demetrovics, Z. (2014). Global self-esteem and method effects: competing factor structures, longitudinal invariance and response styles in adolescents. Behavior Research Methods, 46(2), 488–498. https://doi.org/10.3758/s13428-013-0391-5

Whitlock, J., Muehlenkamp, J., Eckenrode, J., Purington, A., Baral Abrams, G., Barreira, P., & Kress, V. (2013). Nonsuicidal Self-Injury as a Gateway to Suicide in Young Adults. Journal of Adolescent Health, 52(4), 486–492. https://doi.org/10.1016/j.jadohealth.2012.09.010

Williams, M. (1997). Cry of Pain: Understanding Suicide And Self-Harm. London: Penguin.

130

APPENDICES CONTENTS

Appendix A: Author submission guidelines – Clinical Psychology Review

Appendix B: Author submission checklist – Journal of Abnormal Psychology

Appendix C: BECS Study Protocol

Appendix D: Protocol for Managing Disclosure of Risk

Appendix E: Participant Information Sheet

Appendix F: Ethical Approval, University of Manchester

Appendix G: Ethical Approval, NHS

Appendix H: Ethical Approval Manchester Mental Health and Social Care Trust

Appendix I: Data Extraction Sheet for Systematic Review

131

APPENDIX A CLINICAL PSYCHOLOGY REVIEW AUTHOR GUIDELINES

132

133

134

135

136

137

138

139

140

141

142

143

APPENDIX B JOURNAL OF ABNORMAL PSYCHOLOGY AUTHOR CHECKLIST Journal of Abnormal Psychology Manuscript Checklist

Checklist for Manuscript Submission

Numbers following entries refer to relevant section numbers in the Publication Manual.

Format

• Have you checked the journal’s website for instructions to authors regarding specific formatting requirements for submission (8.03)?

• Is the entire manuscript—including quotations, references, author note, content footnotes, and figure captions—double-spaced (8.03)? Is the manuscript neatly prepared (8.03)?

• Are the margins at least 1 in. (2.54 cm; 8.03)?

• Are the title page, abstract, references, appendices, content footnotes, tables, and figures on separate pages (with only one table or figure per page)? Are the figure captions on the same page as the figures? Are manuscript elements ordered in sequence, with the text pages between the abstract and the references (8.03)?

• Are all pages numbered in sequence, starting with the title page (8.03)?

Title Page and Abstract

• Is the title no more than 12 words (2.01)?

• Does the byline reflect the institution or institutions where the work was conducted (2.02)?

• Does the title page include the running head, article title, byline, and author note (8.03)? (Note, however, that some publishers prefer that you include author identification information only in the cover letter. Check with your publisher and follow the recommended format.)

• Does the abstract range between 150 and 250 words (2.04)? (Note, however, that the abstract word limit changes periodically. Check APA Journals Manuscript Submission Instructions for All Authors for updates to the APA abstract word limit.)

Paragraphs and Headings

• Is each paragraph longer than a single sentence but not longer than one manuscript page (3.08)?

• Do the levels of headings accurately reflect the organization of the paper (3.02–3.03)?

• Do all headings of the same level appear in the same format (3.02–3.03)?

Abbreviations

• Are unnecessary abbreviations eliminated and necessary ones explained (4.22–4.23)?

144

• Are abbreviations in tables and figures explained in the table notes and figure captions or legends (4.23)?

Mathematics and Statistics

• Are Greek letters and all but the most common mathematical symbols identified on the manuscript (4.45, 4.49)?

• Are all non-Greek letters that are used as statistical symbols for algebraic variables in italics (4.45)?

Units of Measurement

• Are metric equivalents for all nonmetric units provided (except measurements of time, which have no metric equivalents; see 4.39)?

• Are all metric and nonmetric units with numeric values (except some measurements of time) abbreviated (4.27, 4.40)?

References

• Are references cited both in text and in the reference list (6.11–6.21)?

• Do the text citations and reference list entries agree both in spelling and in date (6.11–6.21)?

• Are journal titles in the reference list spelled out fully (6.29)?

• Are the references (both in the parenthetical text citations and in the reference list) ordered alphabetically by the authors’ surnames (6.16, 6.25)?

• Are inclusive page numbers for all articles or chapters in books provided in the reference list (7.01, 7.02)?

• Are references to studies included in your meta-analysis preceded by an asterisk (6.26)?

Notes and Footnotes

• Is the departmental affiliation given for each author in the author note (2.03)?

• Does the author note include both the author's current affiliation if it is different from the byline affiliation and a current address for correspondence (2.03)?

• Does the author note disclose special circumstances about the article (student paper as basis for the article, report of a longitudinal study, relationship that may be perceived as a conflict of interest; 2.03)?

• Does the author note provide information about prior dissemination of the data and narrative interpretations of the data/research appearing in the article (e.g., presented at a conference or meeting, presented as part of a colloquia at a university, posted on a listserv, or shared on a website, including academic social networks like ResearchGate, etc.)?

• In the text, are all footnotes indicated, and are footnote numbers correctly located (2.12)?

145

Tables and Figures

• Does every table column, including the stub column, have a heading (5.13, 5.19)?

• Have all vertical table rules been omitted (5.19)?

• Are all tables referred to in text (5.19)?

• Are the elements in the figures large enough to remain legible after the figure has been reduced to the width of a journal column or page (5.22, 5.25)?

• Is lettering in a figure no smaller than 8 points and no larger than 14 points (5.25)?

• Are the figures being submitted in a file format acceptable to the publisher (5.30)?

• Has the figure been prepared at a resolution sufficient to produce a high-quality image (5.25)?

• Are all figures numbered consecutively with Arabic numerals (5.30)?

• Are all figures and tables mentioned in the text and numbered in the order in which they are mentioned (5.05)?

• Is written permission to use previously published text; test; or portions of tests, tables, or figures enclosed with the manuscript (6.10)? See Permissions Alert (PDF, 13KB) for more information.

• Are page or paragraph numbers provided in text for all quotations (6.03, 6.05)?

Submitting the Manuscript

• Is the journal editor’s contact information current (8.03)?

• Is a cover letter included with the manuscript? Does the letter

o include the author’s postal address, e-mail address, telephone number, and fax number for future correspondence?

o state that the manuscript is original, not previously published, and not under concurrent consideration elsewhere?

o inform the journal editor of the existence of any similar published manuscripts written by the author (8.03, Figure 8.1)?

o mention any supplemental material you are submitting for the online version of your article?

146

APPENDIX C BECS STUDY PROTOCOL PROTOCOL

An investigation into the relationship between Emotions, Beliefs, Coping

Strategies and Self-Criticism:

Beliefs, Emotions, Criticism and Suicidality (BECS)

147

Contents

1) RESEARCH TEAM & KEY CONTACTS 149 2) INTRODUCTION 5 3) BACKGROUND 5 4) PROJECT OBJECTIVES 6 5) PROJECT DESIGN & PROTOCOL 7 6) PARTICIPANT IDENTIFICATION 9 7) OUTCOME 9 8) DATA COLLECTION, SOURCE DATA AND CONFIDENTIALITY 10 9) STATISTICAL CONSIDERATIONS 10 10) DATA MONITORING AND QUALITY ASSURANCE 12 11) ETHICAL CONSIDERATIONS 12 12) STATEMENT OF INDEMNITY 12 13) PUBLICATION POLICY 12 14) REFERENCES 13

148

1) RESEARCH TEAM & KEY CONTACTS

Chief Investigator: Co-investigator(s):

Name: Dr Daniel Pratt Name: Professor Gillian Haddock,

Address: School of Psychological Sciences, Address: School of Psychological Sciences, 2nd Floor, Zochonis Building, 2nd Floor, Zochonis Building, University of Manchester University of Manchester Brunswick Street Brunswick Street M13 9PL M13 9PL

Email: [email protected] Email: [email protected]

Telephone: 0161 306 0400 Telephone: 0161 275 8455

Fax: 0161 275 8487 Student: Student:

Name: Kate Ward Name: Catherine O’Neill

Email: [email protected] Email: Catherine.F.O’[email protected]

Telephone: 0161 306 0400 Telephone: 0161 306 0400

Fax: 0161 275 2623 Fax: 0161 275 2623 Student: Sponsor(s):

Name: Meryl Kilshaw Name: Professor Nalin Thakker

Address: School of Psychological Sciences, Address: nd 2 Floor, Zochonis Building, Associate Vice-President (Research Integrity) University of Manchester The University of Manchester Brunswick Street M13 9PL Oxford Road Manchester

Email: [email protected] Email: [email protected]

Telephone: 0161 306 0400 Telephone: 0161 275 9795

Fax: 0161 275 2623

149

Lead R&D Trust contact(s): R&D contact:

Name: Jennifer Higham Name: Lisa Dowell

Address: Greater Manchester West Mental Health Address: Manchester Mental Health and Social Foundation NHS Trust Care Trust R&D Office R&D Office Room 109 Harrop House The Rawnsley Building BSTMHT Manchester Bury New Toad M13 9LP Prestwich Manchester M25 3BL

Email: [email protected] Email: [email protected]

Telephone: 0161 7723954 Telephone: 0161 276 3309

R&D contact : R&D contact:

Name: Pat Mottram Name: Anthony Hodgson

Address: Cheshire & Wirral Partnership NHS Trust Address: 5 Boroughs Partnership NHS Trust Academic Unit, Lower Ground Floor Hollins Park House The Stein Centre Hollins Lane Church Lane Winwick Birkenhead Warrington Wirral WA2 8WA CH24 0LD

Email: [email protected] Email: [email protected]

Telephone: 0151 488 7311 Telephone: 01925 664475

Fax: 0151 488 7337 Fax: 01925 664086

150

2) INTRODUCTION

Suicide is a major public health issue (Bertolote et al, 2003), with devastating impact on individuals, families and carers. Suicide is one of the leading causes of death worldwide. As a world-wide public health issue it is therefore a necessity to explore factors that lead to suicidal thoughts and behaviours in order to find ways to improve resilience to and improve prevention of suicide. It is vital to study reasons that lead to suicide. There is also a value to doing this study for the NHS as the study could help add to therapies being used at present to help people with suicidal thinking or behaviour.

3) BACKGROUND

Suicide is the most serious outcome of mental illness (University of Manchester, 2010), resulting in approximately 6000 deaths annually (Office for National Statistics, 2011). Attempted suicide and feelings of suicidality are recurrent experiences for many individuals (Haw, Bergen, Casey and Hawton, 2007) resulting in considerable personal distress and placing substantial pressure on NHS services. The effect of suicide is wide reaching, which influences family, friends and healthcare staff. The related health economic burden is also significant (Department of Health, 2010).

Suicide prevention is therefore a principal public health issue and NHS priority, as demonstrated in major policy directives across the UK (Department of Health, 2011). Fourteen percent of suicides occur in acute mental health settings (Windfuhr & Kapur 2011). These deaths are the most preventable (Duffy & Ryan, 2004) given the service user’s proximity to services. As a public health issue it is therefore essential to explore factors that contribute to suicidal thoughts and behaviours to find ways to improve resilience to and improve prevention of suicide.

The ‘Beliefs, Emotion, Compassion and Suicide’ study (BECS) will examine a range of factors thought to be associated with increased suicidality. A recent extensive study on suicidal behaviour found that reducing symptoms of depression does not necessarily lead to a decline in suicides (Nock et al., 2013). Therefore work is needed to increase our understanding of the pathways to suicidal behaviours. There are three different work packages within the BECS study which will investigate three different potential pathways to suicidal behaviours. This will contribute to the development of strategies for clinical prediction and prevention of suicide.

3.1 Emotions work package

It is proposed that there are basic emotions which underlie emotional experience and related behaviour (e.g., Ekman, 1982; Oately & Johnson-Laird, 1987). There is some agreement that the basic emotions are anger, disgust, fear, happiness and sadness (Power & Dalgleish, 1997). Anger has long been considered crucial in the development of suicidality (Goldney et al., 1997). Studies that have revealed a predictive effect of depression or anger on suicidality have not controlled for possible comorbid effects of other emotions. The Emotions work package predicts that suicidal ideation is predominantly derived from the combination of sadness, anger and disgust that elevate an individual’s suicidal ideation. A further aim of this work package is to examine the contributory effects of the basic emotions to levels of suicidal thoughts.

3.2 Beliefs and Emotion work package

151

An inability to regulate emotions has been found to continue to be a risk factor for suicide even when controlling for depression and hopelessness (Rajappa, Gallagher & Miranda, 2012). Specifically, a perception of limited access to emotional regulation strategies and non-acceptance of emotions (Rajappa et al., 2012) and negative urgency (defined by acting impulsively in response to eliminate negative emotions) (Whiteside & Lynam, 2001; Weinberg & Klonsky, 2009) have been found to be associated with suicidality (Anestis & Joiner, 2010). Beliefs about the inability to problem solve difficult emotions increases the perception that the situation is inescapable which increases feelings of helplessness and hopelessness (Williams, Barnhofer, Crane & Duggan, 2006).

The aim of the Beliefs and Emotion work package is to investigate the relationship between beliefs about emotion, emotion regulation strategies and suicidality. In particular, the role of fixed beliefs about emotion and suicide is to be investigated. The work package will also seek to identify which specific emotion regulation strategies are most strongly associated with suicide.

3.3 Criticism work package

Research has highlighted the potential role of disgust in the development and maintenance of suicidal ideation and behaviour (Chu, Buchman-Schmitt, Michaels, Ribeiro & Joiner, 2013). Feelings of disgust with the self may be understood as a form of negative self-view such as self-criticism, hatred, and attacking (Chen, Wu, & Bond, 2009). Self- Criticism has been linked to suicidality (Fazaa and Page, 2003; O’Connor and Noyce, 2008). This work package will also examine the relationship between different types of self-criticism to suicide; inadequacy, which has been linked to perfectionism, and self-hatred, which has been related to disgust and contempt (Gilbert et al 2004; Whetton and Greenberg 2005).

People with self-resilience and low self-criticism have been found to be able to exhibit more active coping and positive emotion towards the self which allowed them to defend against harsh self- criticism (Whetton, 2005). Gilbert et al (2006) found that trait self-reassurance was related to lower depression and more favourable social comparisons. This work package will examine the impact of levels of self-reassurance within a clinical population, and whether it is able to moderate the effect of self-criticism on suicidality.

Results from the BECS study can be used to inform and develop suicide prevention interventions.

4) PROJECT OBJECTIVES

4.1 Primary Question/Objective:

What is the underlying psychological architecture for suicidality with specific focus on emotions, beliefs and criticism?

4.2 Secondary Question/Objective:

The three different work packages within the BECS study have different secondary research questions/objectives as follows:

4.2.1 Emotions work package

Which basic emotions contribute to suicidal ideation? The objective is to look at the influence of different kinds of basic emotions on suicidality. A further objective will be to explore the role of other basic emotions such as fear and happiness in exploratory analyses.

152

4.2.2 Beliefs and Emotion work package

What is the relationship between beliefs about emotion, emotion regulation strategies and suicidality? More specifically, the objectives of the study are to investigate the role of fixed beliefs about emotion and suicide and the role of emotion regulation strategies and suicide. The study will also seek to identify which specific emotion regulation strategies are most strongly associated with suicide.

4.2.3 Criticism work package

Is self-criticism, specifically self-hatred and inadequacy, and self-reassurance related to suicidal ideation? What is the relationship of self-criticism and self-reassurance to suicidal ideation in a population who have experienced suicidal ideation or behaviour in the last 12 months?

5) PROJECT DESIGN & PROTOCOL

The study will adopt a cross sectional Within Subjects design, with a sample from across the suicidality continuum to investigate the relationship between suicidality, beliefs about emotion, emotion regulation, basic emotions in suicide and self-criticism.

5.1 Participant Identification

The study will seek to recruit sixty individuals with a range of ages and of both sex who have experienced suicidal ideation in the past twelve months. Participants will be recruited from a community clinical sample population. Potential participants will be identified by a mental health clinician (e.g. Care Co-ordinator, Lead Clinician, Named Nurse, Support worker) who will ask the participant whether they are interested in taking part in the study. Once informed consent has been established by the researcher, the researcher will ask participants a question to establish eligibility for the study. Potential participants will be required to answer yes to ascertain this: (a) Have you had any thoughts about or attempts to kill yourself within the past 12 months?

5.2 Study Measures

Participants will be asked to answer a series of questionnaires in the following order:

The Beck Scale for Suicidal Ideation (BSS; Beck, Kovacs & Weissman, 1979). This is a twenty one item self-report scale assessing suicidal ideation, planning and intent.

The Suicide Probability Scale (SPS; Cull & Gill, 1988). This is a thirty six self-report scale exploring the probability of suicide in the future.

The Calgary Depression Scale for Schizophrenia (CDSS; Addington, Addington & Schissel, 1990). This is a nine item scale measuring depression.

153

The Beck Hopelessness Scale (BHS; Beck & Steer, 1988). This is a twenty item self-report scale which measures negative beliefs around three areas of hopelessness.

The Forms of Self-Criticising/Attacking & Self Reassuring Scale (Gilbert, Clarke, Hempel, Miles and Irons (2004). This is a 22 item scale that measures self-criticism and the ability to self-reassure. It measures different ways people think and feel about themselves when things go wrong for them

The Defeat Scale (Gilbert and Allen, 1998). This is a 16 item scale assessing defeat, failed struggle and low social rank

The Entrapment Scale (Gilbert and Allen, 1998). This is a 16 item scale assessing feelings of being trapped by internal and external events.

Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004). This is a thirty six item self- report questionnaire which assesses six difficulties of emotion regulation in response to stress.

Implicit Beliefs about Emotion Scale (IBES; Tamir, John, Srivastava & Gross, 2007). This is a four item scale with two items measuring fixed beliefs about emotions, and two items measuring incremental beliefs about emotions.

The Basic Emotional States (BES; Power, 2006) state and trait scale. This is a three-part scale constructed from 20 emotional terms which relate to the five basic emotions of anger, disgust, fear, happiness and sadness.

In total this is 200 items, and will take an estimated maximum of one hour and ten minutes to complete. Following this, participants will be given the opportunity to take part in a Debrief with the researcher, to enable the participant to process any difficult feelings that have arisen from completing the questionnaires (this will take approximately 20 minutes). This means that participants will be in the study for one and a half hours.

6) PARTICIPANT IDENTIFICATION

6.1 Inclusion Criteria: i) Adults over 18 years. ii) Mental capacity to provide informed consent to take part in research (This will be established through clinical observations at the time of interview according to British Psychological Society (BPS) guidelines (Assessment of Capacity in Adults: Interim Guidance for Psychologists, 2006).

154 iii) Experienced suicidal ideation within the past 12 months (according to self-report). Potential participants will be asked to answer a question and will answer yes to establish this: (a) Have you had any thoughts about or attempts to kill yourself within the past 12 months? iv) Sufficient English language proficiency. v) People accessing NHS Mental Health services with a named worker. 6.2 Exclusion Criteria: i) Organic brain disease, reported by referring clinicians. ii) Significant drug or alcohol intoxication at the time of interview as assessed by self-report or observed by the interviewer. 6.3 Recruitment: If potential participants agree to take part, then participants will be approached by a researcher who will explain the study, answer questions and provide written information sheets. They will be informed that the purpose of the study is to look at the role of thoughts and feelings related to suicide. They will also be informed that the study will require them to complete a series of questionnaires which will ask them about their general thoughts and beliefs, in addition to their recent experiences of suicide. It will be important to stress that if they do not consent to the study then this will not affect the care they receive from their care team. The participant information sheet and consent form will be left with the eligible participant for a minimum of 24 hours to allow them to consider their participation in the study. 6.4 Patients who withdraw consent:

Participants can withdraw consent at any time without giving any reason, as participation in the research is voluntary, without their care or legal rights being affected.

7) OUTCOME

For all of the BECS work packages, the main outcome measure will be the Beck Scale for Suicidal Ideation (Beck, et al., 1979). The Beliefs and Emotion work package will also use the Suicide Probability Scale (Cull & Gill, 1988) as an outcome measure.

In addition, it is hoped that the information gained from this research will help increase our understanding of the pathways to suicidal behaviours. This is to help find ways to prevent suicide and also to find ways to help people to build up their strength against suicide. There is a value to doing this study for the NHS as the study could help add to therapies being used at present to help people with suicidal thinking or behaviour.

In terms of participant benefit, this research will help to identify future areas to target during treatment, providing an evidence based and effective treatment for individuals experiencing increased suicidality.

8) DATA COLLECTION, SOURCE DATA AND CONFIDENTIALITY

Data will be collected from participants using questionnaires and a demographic information sheet. Prior to completing the questionnaires, participants will be asked to complete a consent form. Participants will be allocated a participant number so that identifiable information (for example, their name on the consent form) is separated from answers to the questionnaires and demographic

155 information. All documents will be transported to the University of Manchester in a secure case. The information from the questionnaires and the demographic information will be entered onto the research database in accordance with ISO/IEC 27002 (Information Technology – Code of Practice for Information Security Management, 2005; 2007). Hard copies will be stored in locked filing cabinets in a secure office at the University of Manchester.

9) STATISTICAL CONSIDERATIONS

9.1 Statistical Analysis Statistical analysis has been separated into each work package of the BECS study.

9.1.1 Emotions work package

Hypotheses (Hypotheses 1 – 5): The relationship between the emotional profile of Sadness, Anger, Disgust (Basic Emotional States (BES)) and current ‘state’ suicide ideation (The Suicide Probability Scale (SPS)) will be examined using a correlational analysis and regression analysis. Hopelessness (The Beck Hopelessness Scale (BHS)), depression (The Calgary Scale for Schizophrenia (CDSS)) and ‘trait suicidality’ (The Suicide Probability Scale (SPS) will be controlled for in the analysis. The correlation will estimate the degree of association and interdependence between the variables. Whilst the regression will investigate the prognostic significance of the basic emotions to suicidality after adjusting for important confounding factors.

Exploratory Hypotheses: The relationship between the emotional profile of fear and happiness (Basic Emotional States (BES)) and current ‘state’ suicide ideation (The Suicide Probability Scale (SPS)) will be examined using a correlational analysis and regression analysis. Hopelessness (The Beck Hopelessness Scale (BHS)), depression (The Calgary Scale for Schizophrenia (CDSS)) and ‘trait suicidality’ (The Suicide Probability Scale (SPS) will be controlled for in the analysis.

9.1.2 Beliefs and Emotion work package

Primary aim (Hypotheses 1-3): The relationship between beliefs about emotions (Implicit Beliefs about Emotion Scale (IBES)), emotion regulation (Difficulties in Emotion Regulation Scale (DERS)) and suicide (The Beck Scale for Suicide Ideation (BSS) and Suicide Probability Scale (SPS)) will be examined using a correlational analysis and regression analysis. Hopelessness (The Beck Hopelessness Scale (BHS)) and depression (The Calgary Scale for Depression (CDSS)) will be controlled for in the analyses. The role of fixed beliefs (IBES) on suicide (BSS and SPS) via emotion regulation strategies (DERS) will be further analysed using regression analysis to establish whether emotional regulation strategies are a mediating variable.

Exploratory aims (Hypotheses 4-6):

The relationship between different emotion regulation strategies (subscales of the DERS) and suicide (BSS and SPS) will also be examined using correlational analysis and a regression analysis. The role of fixed beliefs (IBES) on suicide (BSS and SPS) via emotion regulation strategies (subscales of the DERS) will be further analysed using regression analysis to establish whether emotional regulation strategies are a mediating variable. Data will be analysed using available SPSS statistical software.

9.1.3 Criticism work package

156

Hypothesis one and two The relationship between self-hatred (on the FSCRS) and suicide probability will be examined using a correlational analysis and regression analysis. Hopelessness and depression will be controlled for in the analyses.

The relationship between Inadequacy (on the FSCRS) and suicide probability will be examined using a correlational analysis and regression analysis. Hopelessness and depression will be controlled for in the analyses.

Hypothesis three The relationship between self-reassurance (FSCRS) and suicide will also be examined using correlational analysis and a regression analysis. This relationship will be further analysed using a moderator regression analysis to establish whether self-reassurance is a moderating variable with regards the relationship between self-criticism (self-hated and inadequacy) and suicide probability.

If the study is able to recruit increased numbers of participants this would allow for more correlates to be controlled for in the analysis, including defeat and entrapment along with current suicidality.

9.2 Sample Size:

A power calculation has been based on the conventional 10:1 rule for multiple regression (i.e. at least 10 participants per prognostic measures are required for a reasonably robust regression analysis). Separate analyses will be carried out for each of the BECS work packages. For the Beliefs and Emotion work package there are up to four potential predictors (beliefs about the fixedness of emotions, emotion regulation, depression and hopelessness). This requires a sample size of 40. In the Emotion Work Package of the study there will be five basic prognostic variables (Sadness, Anger, Disgust, Hopelessness and Depression) therefore a minimum number of 50 participants will be required to conduct this analysis. In the Criticism Work package, 60 participants will be required as there are six predictor variables to be included in the analysis; self-hatred, inadequacy, self- reassurance, depression, hopelessness, and an intermediate variable for use in the moderator analysis around self-reassurance. With this number of participants the study will have 80% power to detect simple correlations of 0.45 or more between pairs of measures.

Data will be pooled so that a total minimum number of 60 participants will be required to allow for the separate analyses being undertaken in each arm of the study. Therefore a total number of 60 participants will be recruited to allow to control for additional known correlates for suicidality.

10) DATA MONITORING AND QUALITY ASSURANCE The study will be subject to the audit and monitoring regime of the University of Manchester.

11) ETHICAL CONSIDERATIONS An Application for NHS and University Ethics has been made.

The study will be conducted in full conformance with principles of the “Declaration of Helsinki”, Good Clinical Practice (GCP) and within the laws and regulations of the country in which the research is conducted.

157

12) STATEMENT OF INDEMNITY The University of Manchester will arrange insurance for research involving human subjects that provides compensation for non-negligent harm to research subjects occasioned in circumstances that are under the control of the University of Manchester, subject to policy terms and conditions.

13) PUBLICATION POLICY

Participants will be sent a lay version of the results from the study who opt in at the consent stage. Relevant organisations, for example, Papyrus will also be sent a copy of the research results. A summary report will be given to the manager of all host sites (including all participating inpatient wards and community teams). Presentations for the hosting NHS organisations will be arranged for all interested staff on the findings of the study.

158

14) REFERENCES

Addington, D., Addington, J., & Schissel B. (1990). A depression rating scale for schizophrenics. Schizophrenia Research 3, 247-251.

Anestis,, M. D. & Joiner, T. E. (2012). Behaviourally-indexed distress tolerance and suicidality. Journal of Psychiatric Research, 46(6), 703-707.

Beck, A. T. & Steer, R. A. (1988). Manual for the Beck Hopelessness Scale. San Antonio, TX: Psychological Corporation.

Beck, A. T., Kovacs, M., & Weissman, A. (1979). Assessment of suicidal intention: The Scale for Suicide Ideation. Journal of Consulting and Clinical Psychology, 47(2), 343-352.

Bertolote, J.M., Fleischmann, A., De Leo Jose & Wasserman, D. (2003). Suicide and mental disorders: do we know enough? British Journal of Psychiatry,183, 382 – 283.

Chen, S. X., Wu, W. C. H., & Bond, M. H. (2009). Linking family dysfunction to suicidal ideation: The mediating roles of self-views and world-views. Asian Journal of Social Psychology, 12, 133–144.

Chu, C., Buchman-Schmitt, J. M., Michaels, M. S., Ribeiro D. R., & Joiner, T. (2013). Discussing Disgust: The role of disgust with life in suicide. Journal of Cognitive Therapy, 6, 235-247. Cull, J. G., & Gill, W. S. (1988). Suicide Probability Scale Manual. Los Angeles: Western Psychological Services.

Department of Health (2010). No health without mental health. London. HMSO. Department of Health (2011) Consultation on preventing suicide in England: A cross-government outcomes strategy to save lives. HMSO. London.

Fazaa, N & Page, S (2003) Dependency and self-criticism as predictors of suicidal behaviour. Suicide Life Threat Behaviour, Summer; 33(2):172-85. Gilbert, P., Baldwin, M., Irons, C., Baccus, J. & Clark, M. (2006). Self-criticism and self-warmth: An imagery study exploring their relation to depression. Journal of Cognitive Psychotherapy: An International Quarterly. 20, 183-200

Gilbert, P., Clarke, M., Kempel, S. Miles, J.N.V. & Irons, C. (2004). Criticizing and reassuring oneself: An exploration of forms style and reasons in female students. British Journal of Clinical Psychology 43, 31-50.

Goldney, R., Winefield, A., Saebel, J., Winefield, H., Tiggeman, M. (1997). Anger, suicidal ideation, and attempted suicide: a prospective study. Comprehensive Psychiatry, 38,264–268. Gratz,, K.L. & Roemer, E. Multidimensional Assessment of Emotion Regulation and Dysregulation: Development, Factor Structure, and Initial Validation of the Difficulties in Emotion Regulation Scale. Journal of Psychopathology and Behavioral Assessment, 26, (1) 41-54.

159

Haw, C,. Bergen, H,. Casey, D,. & Hawton K. (2007) Repetition of deliberate self-harm: A study of the characteristics and subsequent deaths in patients presenting to a general hospital according to extent of repetition. Suicide and Life-threatening Behaviour, 37, 379-396.

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 behaviour among adolescents. Results from the National Comorbidity survey replication adolescent supplement. Journal of American Medical Association and Psychiatry, 70 (3), 300-310.

O’Connor, R.C., & Noyce, R. (2008). Personality and cognitive processes: Self-criticism and different types of rumination as predictors of suicidal ideation. Behaviour Research and Therapy, 46, 392-401 Power, M. J. & Dalgleish, T. (1997). Cognition and emotion: From order to disorder. Hove: Psychology Press. Power, M. J. (2006). The structure of emotion: An empirical comparison of six models. Cognition and Emotion, 20, 694–713. Rajappa, K., Gallagher, M. & Miranda, R. (2012). Emotion dysregulation & vulnerability to suicidal ideation and attempts. Cognitive Therapy & Research, 36, 833-839. Tamir, M., John, O.P., Srivastava, S. & Gross, J.J. (2007). Implicit Theories of Emotion: Affective and Social Outcomes Across a Major Life Transition. Journal of Personality and Social Psychology, 92 (4), 731-744.

National Confidential Inquiry into Suicide and Homicide by People with Mental Illness (2010) University of Manchester.

Office for National Statistics (2011). Suicide rates in the UK 2000-2009. London ONS.

Weinberg, A. & Klonsky, E.D. (2009). Measurement of emotion dysregulation in adolescents. Psychological Assessment, 21, 616-621.

Whelton, W. J., & Greenberg, L. S. (2005). Emotion in self-criticism. Personality and Individual Differences, 38 (7), 1583–1595.

Whiteside, S.P. & Lynam, D.R. (2001). The five factor model and impulsivity: using a structural model of personality to understand impulsivity. Personality and Individual Differences, 30, 669-689.

Williams, J.M.G., Barnhofer, T., Crane, C. & Duggan, D.S. (2006). The role of over-general memory in suicidality. In T.E. Ellis (Ed.), Cognition and suicide: Theory, research, and therapy (pp.173-192). Washington DC: American Psychological Association. Windfuhr K., & Kapur N. (2011) International perspectives on the epidemiology and aetiology of suicide and self-harm. In O’Connor R,. Platt S,. Gordon J. International handbook of suicide prevention: Research, policy and practice. Chichester. Wiley-Blackwell.

160

APPENDIX D PROTOCOL FOR THE MANAGEMENT OF DISCLOSURE OF RISK BECS: PROTOCOL FOR MANAGING DISCLOSURE OF RISK

This protocol directs the practice of all non-clinical staff (including students, trainees, and volunteer assistants – all hereafter referred to as staff). Rationale During a session or other contact with BECS research staff a participant may indicate an intention to harm themselves or others. Alternatively they may provide information to the effect that a child or other vulnerable person may be in danger. Any information of this nature must be acted upon. At the beginning of each research interview the participant will be informed that what is discussed is private and confidential except if they indicate any current intention to harm themselves or others, or if they provide information to the effect that a child or other vulnerable person may be in danger. In such situations the staff member has a legal duty to break confidentiality. The particular setting within which risk is disclosed (i.e. hospital ward or community) will determine the specific actions to be taken. Community participants In the case that the individual indicates current intention to harm themselves or others the action taken is to remind the participant of the staff member’s Duty of Care to break confidentiality where risk is identified (as already outlined at the commencement of the interview) and then contact their care co-ordinator and / or psychiatrist or GP. The immediacy of this action will depend upon the time frame involved. If an imminent risk is identified, i.e. the individual reports that they intend to harm themselves within the next 48 hours, immediate action should be taken and the session should immediately change focus to the imminent threat. However if the individual reports that they intend to act on their thoughts in a few days, or longer, action by the worker may involve continuing with the session in light of the information discussed, reviewing how they are feeling at the end of the session and calling the care-coordinator / named worker following this. If the individual indicates that a child / other vulnerable person may be in danger the action taken would be to call the respective Child or Adult Safeguarding Team (see contact details below). In either eventuality the participant will be informed that confidentiality needs to be broken and, if at all possible, will be encouraged to work in collaboration with the staff member to this end. Unless there are circumstances that would contraindicate (e.g. risk to safety of staff), the participant should be informed that this action is to be taken.

161

If this scenario occurs during a face to face contact the individual may be given the option of phoning the care co-ordinator themselves in the presence of the staff member or staying in the room whilst a call is made. Alternatively the individual may choose to wait in a safe place such as an adjoining room. Based upon the telephone discussion the worker will act on any part of the action plan generated that involves action on their part. In the eventuality that the care co-ordinator or named worker are not contactable a call should be made within the hours of 9am – 5pm Monday to Friday to the Duty worker for the appropriate Community Mental Health Team or outside of these hours a call should be made to the Crisis Team or A&E. Details are listed in Appendix 2. Once again the worker will act in accordance with any action plan agreed. This may involve faxing information over to A&E, accompanying the individual to A&E etc. The police will be contacted if the person absconds during this process. If the scenario occurs during a telephone contact the individual will be informed that confidentiality will need to be breached. The same plan as above will be implemented and the individual should be called back to feedback the planned actions. In the eventuality that the individual discloses that a child / vulnerable adult may be in danger the Child / Adult Safeguarding Team should be contacted. If it is outside of 9am – 5pm and there is considered to be imminent risk to a child / vulnerable adult the police should be informed. Inform these staff also. Details of out of hours Child / vulnerable Adult Safeguarding Team services are listed in Appendix 2. If the worker is uncertain as to the appropriate course of action to take they should initially contact the project supervisors (i.e. Professor Gillian Haddock or Dr Daniel Pratt). If the project supervisors are unavailable contact your clinical supervisor. If they are unavailable the flow diagram of contacts found on page 7/8 should be followed. In the unlikely event that all avenues are exhausted the worker should follow the previously outlined plan (commencing with contacting the Care Coordinator). If the client is currently harming him or herself or has done so recently, and there is a need for medical attention, it would be important to negotiate with the client that they attend hospital or that they allow an ambulance to be called. A&E will be contacted to inform them of the reason for attendance and confirm the individual’s arrival. The mental health team or duty psychiatrist would ensure that anyone refusing medical attention was assessed under the Mental Health Act. A decision regarding the need for a compulsory admission to hospital will then be made by an approved social worker in accordance with the Mental Health Act 1983. If the participant disclosing that they have committed a criminal offence, then this should be shared with the police as soon as possible. Ensure that you record all information and actions taken, including telephone calls and discussions with your project lead/clinical supervisor, in the participant’s file.

162

FACTORS TO CONSIDER IF A PARTICIPANT EXPRESSES HARM TO SELF OR OTHERS If a participant you are interviewing expresses ideas of harm to self or others these are important factors to consider and pass on.

• Ideation (frequency, intensity, duration, triggers) • Plans/intent • Access to means to carry out plans • Timeframe • Protective factors • Access to support/isolation • Hopelessness • Drug or alcohol use • Command hallucinations and perceived power or control over voices

Any concerns you have should be discussed with Dr Daniel Pratt or Professor Gillian Haddock.

163

FLOWCHART OF CONTACTS FOR COMMUNITY PARTICIPANTS WITH IDENTIFIED INTENT TO HARM OTHERS

In situations where a Child / vulnerable Adult is at risk the appropriate Safeguarding Team should be contacted.

Participant expresses imminent harm to others

ENSURE OWN SAFETY - LEAVE IF FEEL THREATENED

Call Care Coordinator

If no answer If unsure call Dr Daniel Pratt or Professor Gillian Haddock

Call Duty Worker at Team

If no answer

Call GP

If not available

Call Dr Daniel Pratt or Professor Gillian Haddock

IF UNSURE OF IMMEDIATE SAFETY, PHONE POLICE

In all instances record in clinical notes & within BECS Records

164

FLOWCHART OF CONTACTS FOR COMMUNITY PARTICIPANTS WITH IDENTIFIED IMMINENT SUICIDAL INTENT

Client expresses imminent suicidal intent

Where possible do not leave the client alone

If unsure call Dr Daniel Pratt or Professor Gillian Haddock Call Care Coordinator If no answer

Call Duty Worker at Team

If no answer

Call GP

If not available

Dr Daniel Pratt or Professor Gillian Haddock

If not available

Call Ambulance and / or police

Contact A&E to confirm arrival & inform of reasons for attendance

In all instances record in clinical notes & within BECS Records

165

APPENDIX 1: DISCLOSURE OF RISK DURING AN INTERVIEW

Staff are to follow these guidelines in the event that, when conducting an interview, they identify that a participant might be at risk, or could pose a risk either to themselves or to others. The examples presented here are to be modified according to the situation. ▪ Prior to commencing an interview with a participant, the researcher will carefully explain that, although the interview is going to be confidential, if any risk is identified or disclosed during the interview, then the researcher will have to communicate these concerns to other professionals:

“Before we begin the interview, I just want to explain again that what we will talk about will be confidential, but if I feel that there might be a risk in what you are saying, for example to yourself or to others, I will need to pass this on to other staff members. But if I do this, I will tell you”. ▪ If during the interview a participant’s account indicates that there might be distressed or they disclose some type of suicidality or risk factors, the researcher will reflect the distress they appear to be in and will ask if they want to continue the interview, and/or offer a brief break:

“You seem to be going through a hard time at the moment – do you want to continue with the interview? You know we can take a break at any time or we can stop if you want to”. “It sounds like there have been a few things upsetting you recently – are you okay to continue with the interview or would you prefer to take a bit of a break for a few minutes?” ▪ If during the interview the participant has disclosed a clear risk of suicidality (for instance, a description of plans for self-harming, or explaining that they are in possession of medication to take an overdose), at the end of the interview the researcher will explain the need to communicate this to staff:

“You’ve spoken about wanting to take an overdose with some medication you have, and it sounds like you are quite upset about some of the things we’ve been talking about. What I’m going to do, like we’d talked about at the beginning, is to speak with the nurse on duty or your care coordinator/named worker and tell them how you are feeling so that they know what’s going on for you and so that they can help you” ▪ If during the interview the participant’s account indicates or suggests a possible risk of suicidality (for example, talking about occasional fleeting feelings of wanting to die, or sometimes wishing they could just be gone to end their problems), the researcher will try to ascertain some further information:

“You’ve said that you sometimes wish you could just be gone and end your problems this way, have you recently had this kind of thoughts? Do you mean that you have a plan for this or are they just thoughts?” “You said that you sometimes have felt like you want to die – if you were to feel like this again, do you think you would communicate this to staff?” At the end of the interview, the researcher will talk about this with the participant: “You said that sometimes you have felt like you want to die, although not in the last week – do you mind if I just mention this to the nurse on shift/your care coordinator/named worker, so that they are aware too?” If the participant accepts, this information can be given to staff.

166

If the participant declines, the researcher will contact Dr Daniel Pratt or Professor Gillian Haddock to consult with them, on a case by case basis, the need to report this to staff. ▪ If any risk of suicidality has been disclosed by a participant during an interview and this risk needs to be reported to staff, the researcher will do so verbally to a staff nurse, Clinical Practice Lead or shift leader. Staff will also write an entrance in the clinical notes:

“During a research interview, [participants name and surname] has described plans and intent to take an overdose if access to medication was available. This was communicated to [staff name and role] when the interview ended. [Participant] is aware that I have passed this information on to other staff.”

APPENDIX 2: LIST OF USEFUL CONTACTS

BECS Staff Contact Details:

Gillian Haddock

Dr Daniel Pratt

Catherine O’Neill

Kate Ward

Meryl Kilshaw

167

Community Support Resources For Participants:

Manchester and Salford Address: 72-74 Oxford St, Manchester M1 5NH Samaritans Phone: 0161 236 8000

Address: Zion Community Centre, 339 Stretford Road, Manchester Mind Hulme, Manchester, M15 4ZY Phone: 0161 226 9907

Manchester Assertive Address: PO Box 201 Manchester M21 8WR Outreach Phone: 0161 881 4799

The Samaritans of Address: 46 Arpley Street, Warrington, Cheshire, Warrington, Halton and St WA1 1LX Helens Phone: 01925 235000 Address: The Resource Centre, 30a Widnes Road, Halton Mind Widnes, WA8 6AD Phone: 0151 495 3991

Address: 36 Upper Northgate Street, Chester, The Samaritans of Chester Cheshire, CH1 4EF and District Phone: 01244 377999

168

Child Protection Service Contact Details:

If a child is at immediate risk, contact the police on 999

Manchester 0161 234 5001 (24 hour service) or email [email protected] Salford (8.30 – 4.30) 0161 603 4500 Duty (out of hours) 0161 794 8888 Trafford (8.30 – 4.30) 0161 912 5125 Duty (out of hours) 0161 912 2020 Bolton (9 - 5) 01204 337729 Duty (out of hours) 01204 337777 Halton (Mon – Thurs 9-5 Fri 9 – 4:30pm) 0151 907 8305 Duty (out of hours) 0345 050 0148

Cheshire West & Chester (Mon – Thurs 9-5 Fri 9 – 4:30pm) 01606 275 099 Duty (out of hours) 01244 977 277 Other

• NSPCC Child Protection helpline on 0808 800 5000 (free 24 hour service) • Childline 08001111 (a free 24 hour helpline for children)

169

ACCIDENT AND EMERGENCY NUMBERS

Oxford Road, Oxford Road, Manchester, M13 9WL Manchester Royal Infirmary 0161 276 4147 Delaunays Road, Crumpsall, Manchester, M8 5RB North Manchester General 0161 624 0420 A Block, Wythenshawe Hospital, Southmoor Road, Wythenshawe Hospital Manchester M23 9LT 0161 291 6041 Hope Building, Salford Royal, Stott Lane, Salford, M6 Salford Royal Infirmary 8HD 0161 789 7373 Moorside Road, M41 5SL Trafford General Hospital 0161 748 4022 Minerva Road, Farnworth, Bolton, Lancashire, BL4 0JR Royal Bolton Hospital 01204 390390

Warrington Road, Prescot, Merseyside, L35 5DR Whiston Hospital 0151 426 1600

The Countess Of Chester Health Park, Chester, Countess of Chester Hospital Cheshire, CH2 1HJ 01244 365000

170

CRISIS TEAM CONTACT NUMBERS:

MMHSCT

Crisis Line: 0161 922 3801 This line is in operation from Mon-Thurs 5pm - 9am and from 5pm on a Friday until 9am Monday. The Crisis line is also open on Bank Holidays).

GMW Area Crisisline: 0800 028 8000 GMW service users may contact Crisisline directly. Crisisline operates: Monday to Friday: 5pm – 9am, Saturday, Sunday and bank holidays: open 24 hours

Warrington and Halton

Crisis team: 01925 664811 This line is in operation 8am to 8pm, 7 days a week, 365 days per year

Cheshire and Wirral area

Crisis Line: 01244 397303 (Juniper Ward) This line is in operation from 5pm – 9am, 7 days a week, 365 days per year.

(The community mental health team can help during the day 9am – 5pm, Monday to Friday. Ask for your care co-ordinator or the person on duty on 0151 357 7600)

OTHER USEFUL CONTACTS:

Mind infoline 0300 123 3393 weekdays 9am - 6pm [email protected] Samaritans 08457 90 90 90 open 24 hours a day [email protected] HOPELineUK 0800 068 41 41, email: [email protected] , text: 07786 209697 NHS helpline: 111 (open 24 hours a day)

171

APPENDIX E PARTICIPANT INFORMATION SHEET Participant Information Sheet Beliefs, Emotions, Criticism and Suicide (BECS)

We would like to invite you to participate in a study that looks at the relationship between different types of emotion, how people feel about their emotions and suicide. Please read all of the information on this sheet carefully. If you would like any more information or feel that there is anything you don’t understand, please contact the researchers. Please take your time to think about taking part or not. What is the purpose of the study? Suicide is a leading cause of death world-wide. It is important to study reasons that lead to suicide. This is to help find ways to prevent suicide. It could also help to develop new treatments for people having suicidal thoughts and feelings. We need to discover why people take their own lives and how we can help stop this from happening.

This research will particularly look at whether people who have been suicidal show patterns of particular emotions, or whether feeling particular ways increases their risk of suicide. It will also look at what people think about their emotions increases or decreases their risk of suicide.

Who is organising this research?

This research is being organised by Students on the Clinical Psychology Doctorate programme (under the supervision of qualified Clinical Psychologists at the University of Manchester. Why have I been invited to take part? You have been invited to take part because you may have experienced thoughts of taking your own life in the past 12 months and are receiving services from an NHS mental health team. Do I have to take part? No, it is your choice to take part or not. If you change your mind at any point during the study you can leave at any time. You do not have to give a reason and it will not affect your treatment with the mental health team. What will taking part involve for me? You should receive this participant information sheet from your mental health worker, or by post after giving consent to find out more about the research. You can take as long as you need to consider if you want to take part or not. You can also ask the researchers any questions you may have.

172

If you decide to take part in the research, a researcher will come and meet you in your home or a community venue of your choice. They will ask you to fill in a consent form, some information about yourself and fill in some questionnaires that relate to your feelings and thoughts related to suicide. These questionnaires will not have your name on and will not be seen by your healthcare team. Answering the questions should take around 1 hour and 10 minutes. There will then be a 20 minute debrief period where you can discuss any thoughts or feelings the study has brought up. You will be able to take breaks at any point. Your data will be kept private and confidential, only researchers on the BECS study will be able to access it and it will be stored in a locked cabinet. What are the risks of taking part? We do not expect there to be any risks to people taking part on this study, however speaking with the researcher and filling in the questionnaires may become upsetting. You will be able to take a break whenever you like and can stop the interview at any time. There will also be time at the end of the session for you to talk about any difficult feelings and the researchers will check whether you would like us to contact anyone for you if you need some extra support. What are the benefits of taking part? Some researchers looked at the benefits of taking part in research on suicide. Most people felt the experience was positive and felt glad to have helped others through sharing their experiences. Some felt it was helpful to talk about their experiences and felt good about contributing to research.

Who will know I am taking part?

Your care team will be told that you are taking part in the BECS study. However, any information you provide as part of the study is confidential to the BECS team. We will only share information with your care team if you told us something that we thought put you or someone else in danger. If we needed to do this, we would try to talk to you about it first.

What if there is a problem?

If you are worried about any part of the study, you can speak to one of the researchers who will do their best to answer any questions. If they cannot help you, you can speak to the study lead Investigators, Dr Daniel Pratt or Prof Gill Haddock at the University of Manchester on 0161 306 0400.

We aim to publish the results from this study in scientific papers and to present the results to other scientists. We will also write the results in a non-scientific way. These results will not be available until the study is finished. If you would like a copy of these results please tell a member of the research team when you are being interviewed.

173

If you would like any more information about this study, please contact one of the researchers on this project:

Kate Ward Meryl Kilshaw Cat O’Neill

Tel: INSERT MOBILE TELEPHONE NUMBER HERE

INSERT EMAIL ADDRESS HERE

174

APPENDIX F ETHICAL APPROVAL (UNIVERSITY OF MANCHESTER, RESEARCH SPONSOR)

From: Lynne Macrae > Sent: 04 February 2015 15:26 > To: Meryl Kilshaw; Kate Ward; Catherine O'Neill > Subject: RE: Amendments > > Hi Meryl, Catherine and Kate > > Thanks very much for your patience. I have completed the review and can confirm that the University can act as research governance sponsor for your study. We discussed the issues around the overlap with the INSITE study and how this will have to be outlined on the review form. Other than that, there were only a couple of issues – please see review form attached. You will notice that some of the points are comments rather than requests for changes. If you don’t feel they are appropriate for your study, that is fine, just let me know via the review form. > > ______> Next Steps > > Once you have gone through the comments, made any changes and are happy that you have the final version of the application please send the following: > > > * By email to [email protected]: > 1. A copy of the updated review form > 2. PDF of IRAS form > 3. Any revised documents > > > * In IRAS: > 1. electronic authorisation request(s) in IRAS for REC/R&D form. When sending the authorisation request please use the following email address for the sponsor representative: ([email protected]) > > Letters and Insurance certificates > In anticipation of signing off the IRAS REC and R&D forms, I am attaching a copy of the sponsorship and insurance letters and a copy of the University’s insurance certificates (zip file). All of these documents will have to be submitted as part of your REC/R&D application. > > Please note that I didn’t see an insurance assessment form with the initial application. I have re-attached the form but, please just let me know if I have missed it. > > Sponsor Conditions > Finally, I have attached a copy of the sponsor conditions. All the conditions are relevant however I would like to highlight the following: > > > * All Amendment must be submitted for sponsor approval before being sent to the REC. > > > * Once you have final ethical approval, please send a copy of your IRAS REC form, REC approval letter,

175 approved documents (as listed in the REC approval letter) to [email protected]. This is to ensure that we maintain sponsor oversight but also so that insurance cover is in place for your study. > > BW > Lynne > > Lynne MacRae > Research Practice Coordinator > Faculty of Medical & Human Sciences > University of Manchester > Room 3.53 Simon Building > Brunswick Street > Manchester M13 9PL > > Tel: 0161 275 5436 > (10am-4pm - voicemail outside of these hours) > > Email address for requesting electronic authorisation in IRAS: [email protected] (note no initial ‘K’) > Website: FMHS Research Governance Website > Twitter: @fmhs_ethics

176

APPENDIX G ETHICAL APPROVAL DOCUMENTS (NHS)

177

178

179

180

181

182

183

APPENDIX I ETHICAL APPROVAL MHSC

184

185

APPENDIX J DATA EXTRACTION SHEET

186