Welcome to the Schizophrenia Resource Centre

Welcome, this website is intended for healthcare professionals in EMEA with an interest in the treatment of schizophrenia. By clicking the link below you are declaring and confirming that you are a healthcare professional

You are here

The relationships of suicidal ideation with symptoms, neurocognitive function, and psychological factors in patients with first-episode psychosis

Schizophrenia Research, Volume 157, Issues 1–3, August 2014, Pages 12–18



Individuals with first-episode psychosis (FEP) have markedly elevated risk for suicide. Previous research on suicidality in early psychosis mainly focused on attempted and completed suicide. Data regarding risk factors for suicidal ideation, which is a common antecedent and predictor of suicide attempt, were limited. This study aimed to examine the prevalence of suicidal ideation and its relationships with clinical, neurocognitive and psychological factors in FEP patients.


Eighty-nine Chinese patients aged 15 to 25 years presenting with FEP to specialized early intervention service were recruited. A comprehensive set of assessments examining pre-treatment illness characteristics, symptom severity, neurocognitive function, and psychological factors were administered. Current suicidal ideation and history of suicide attempt were systematically evaluated.


Approximately 42% of patients expressed suicidal ideation after service entry. Univariate regression analyses found that suicidal ideation was significantly associated with past suicide attempt, depressive symptoms, emotion expressivity, hopelessness, future expectation, attentional impulsiveness, internal and external locus of control, and the likelihood of endorsing fear of social approval and survival and coping beliefs as reasons for living. Final multivariate model showed that previous suicide attempt, depression, less severe diminished expression, greater degree of hopelessness and lower level of internal locus of control independently predicted suicidal ideation.


Suicidal ideation was prevalent in FEP patients. Our findings implied that close monitoring and prompt intervention of those potentially modifiable risk factors for suicidal ideation including depression, hopelessness and perceived inadequate personal control may reduce suicide risk in the early course of psychotic illness.

Keywords: Suicidal ideation, First-episode psychosis, Depression, Negative symptoms, Hopelessness.

1. Introduction

Psychotic disorders including schizophrenia are associated with markedly elevated risk for suicide ( Hawton et al., 2005 ). Evidence showed that suicide constitutes a major cause of premature mortality among individuals with schizophrenia ( Saha et al., 2007 ) whose lifetime suicide risk is estimated as 4.9% ( Palmer et al., 2005 ) and 20–40% will attempt suicide during the course of their illness ( Harvey-Friedman, 2006 ). Early phase of psychotic disorder was consistently found to be a particularly vulnerable period for suicide when compared to other stages of the illness (Nordentoft et al, 2004 and Palmer et al, 2005). Recent studies demonstrated that the first 1 to 2 years of treatment for first-episode psychosis (FEP) conferred the highest suicide risk (Nordentoft et al, 2004 and Dutta et al, 2010). A recent meta-analytic review confirmed a high rate of self-harm in FEP populations with 18.4% and 11.4% made at least one attempt before and after treatment, respectively ( Challis et al., 2013 ).

Suicidal ideation is a common antecedent of suicidal behavior (includes attempted and completed suicide). Previous studies showed that up to 40% of patients with schizophrenia reported suicidal ideation (Fenton et al, 1997 and Fialko et al, 2006). Systematic reviews also found that suicidal ideation was associated with increased risks of self-harm and suicide in schizophrenia (Haw et al, 2005 and Hawton et al, 2005). In fact, suicide risk can be conceptualized as a multi-stage process with suicidal ideation being regarded as a critical initial step along a continuum of increasing severity toward attempted and completed suicide (Hawley et al, 1991 and Kontaxakis et al, 2004). Emergence of suicidal ideation thus represents an early indicator of potential suicidal intent which may then result in suicide attempt, and it is suggested that better understanding of risk factors for suicidal ideation would facilitate suicide prediction and prevention ( Mann et al., 1999 ). Nonetheless, despite its significant clinical implications, relatively few studies have been conducted to examine prevalence and correlates of suicidal ideation in FEP patients.

Prior first-episode studies reported that the rate of suicidal ideation at initial presentation ranged between 26.2% and 56.5% (Nordentoft et al, 2002, Tarrier et al, 2007, Bakst et al, 2010, Barrett et al, 2010, Melle et al, 2010, and Upthegrove et al, 2010). Baseline suicidal ideation was found to strongly predict suicide attempt and suicide ideation at follow-up (Young et al, 1998 and Bertelsen et al, 2007). Discrepant findings, however, were observed with respect to the relationship between clinical variables and suicidal ideation in early psychosis, with some studies showing past suicide attempt (Bertelsen et al, 2007 and Bakst et al, 2010), depression (Bertelsen et al, 2007 and Bakst et al, 2010), hopelessness (Nordentoft et al, 2002 and Bakst et al, 2010) and psychotic symptoms (Nordentoft et al, 2002 and Bertelsen et al, 2007) as independent factors associated with suicidal ideation, but not others ( Tarrier et al., 2007 ). Of note, several methodological issues concerning investigation of risk factors for suicidal ideation warrant consideration. First, many studies that have assessed suicidal ideation in FEP used only a single item that is selected from a depression rating instrument (Bakst et al, 2010, Barrett et al, 2010, and Melle et al, 2010) which may compromise an accuracy of estimating the rate of suicidal ideation. Second, as this single-item measure generates a composite rating for overall suicidality combining suicidal ideation and attempt, such measurement overlap may thus obscure potentially significant relationships between various putative risk factors and suicidal ideation. Third, existing literature on prediction of suicidal ideation in FEP primarily focused on demographic and clinical characteristics. There is a lack of systematic evaluation of related neurocognitive and psychological variables including executive functions (Marzuk et al, 2005 and Miranda et al, 2012), impulsivity trait ( Brezo et al., 2006 ), perceived locus of control (Goldney et al, 1991 and Lester et al, 1991) and reasons for living (Hirsch and Ellis, 1996 and Pinto et al, 1998), which have otherwise been linked to suicidal ideation in normal population and patients with other psychiatric diagnoses.

To better understand the risk profiles for suicidal ideation in patients with psychosis, in particular those identifiable factors that emerged in the early stage of illness, is crucial to suicide prevention. In this regard, we present a study conducted in a representative Chinese cohort of young people presenting with FEP to a specialized early intervention program in Hong Kong with an aim to (1) examine the prevalence of suicidal ideation after service entry and (2) to identify clinical, neurocognitive and psychological factors associated with suicidal ideation.

2. Materials and methods

2.1. Participants

Participants were recruited from both outpatient and inpatient psychiatric units of three hospitals in Hong Kong Island and Kowloon West upon their enrolment to a specialized early intervention program for psychosis, namely Early Assessment Service for Young People with Psychosis (EASY). Since 2001, this publicly-funded service has been providing comprehensive assessment and phase-specific intervention for all individuals aged 15–25 years presenting with FEP in Hong Kong ( Tang et al., 2010 ). The program adopts case-management approach and assertively follows up patients for the first three years after their initial episodes. In this study, we excluded patients who had intellectual disability, psychotic disorders due to general medical condition or substance-induced psychosis. A total of 116 patients who were consecutively diagnosed as first-episode DSM-IV ( APA, 1994 ) schizophrenia, schizophreniform disorder, delusional disorder, brief psychotic disorder or psychosis not otherwise specified (NOS) were approached for study recruitment. Of the initial cohort, 25 refused participation and 2 failed to complete study assessments, and 89 patients were thus retained as the study sample. Study assessments were administered to participants by a research staff on average 40 days following service entry. Training on the use of assessments was provided to a research staff prior to participant recruitment. Regular meetings were held throughout the study period to ensure strict adherence to assessment rating manual and to resolve any ambiguity in clinical information during the data collection process. The study was approved by the local institutional review board and patients gave written informed consent before participation.

2.2. Assessments

2.2.1. Suicidal ideation and history of suicide attempt

Suicidal ideation in the past week was assessed using Beck Scale for Suicidal Ideation (BSS) which is a 21-item self-report questionnaire with each item being rated on a 3-point scale (from 0 to 2, higher score indicates higher level of suicidality) ( Beck and Steer, 1991 ). The first 5 items of BSS were used to ascertain the presence of suicidal thoughts of an active or passive nature ( Beck and Steer, 1991 ). Participants who scored > 0 on one or more of these first 5 items were classified as “patients with suicidal ideation”, while those who scored 0 on all of these 5 items were considered as “patients without suicidal ideation”. History of suicide attempt was ascertained by the 20th item of BSS (0 = no previous suicide attempt, 1 = one past suicide attempt in lifetime, 2 = two or more past suicide attempts in lifetime). The Chinese version of BSS was used in this study with previous validation being conducted in early psychosis patients showing high internal consistency (Cronbach's alpha = 0.78).

2.2.2. Clinical assessments

Diagnosis of each participant was ascertained by senior research psychiatrist using all available information including Chinese-bilingual Structured Clinical Interview for DSM-IV (CB-SCID-I/P, So et al., 2003 ), informant histories and medical records. Previous validation study showed that CB-SCID-I/P yielded reliable DSM-IV diagnoses in Chinese patients with psychotic disorders ( So et al., 2003 ). Interview for Retrospective Assessment of the Onset of Schizophrenia (IRAOS, Hafner et al., 1992 ) was used to confirm the first-episode status and to assess duration of untreated psychosis (DUP) which was defined as time interval between onset of psychotic symptoms and service entry. Positive symptoms subscale of Positive and Negative Syndrome Scale (PANSS, Kay et al., 1987 ) was used to assess positive psychotic symptom level. Negative symptoms were measured by Scale of Assessment of Negative Symptoms (SANS, Andreasen, 1982 ). As negative symptom construct was consistently shown to comprise two distinct sub-domains of diminished expression (DE) and avolition–asociality (AA) ( Messinger et al., 2011 ), we thus derived DE and AA domain scores based on the method applied by previous research ( Foussias et al., 2009 ): DE comprised items of Affective flattening subscale (excluding global item) and poverty of speech item of Alogia subscale; and AA consisted of items of Avolition–apathy and Anhedonia–asociality subscales (excluding global items). Montgomery–Asberg Depression Rating Scale (MADRS, Montgomery and Asberg, 1979 ) was employed to assess depressive symptom severity. Total MADRS score excluding item 10, which evaluated suicidality, was computed for analysis. Level of insight was assessed by Birchwood Insight Scale (IS) which comprises three separate dimensions including awareness of illness, attribution of symptoms and need for treatment ( Birchwood et al., 1994 ). The Chinese version of IS, which was validated in local population ( Lam et al., 1996 ), was adopted in this study.

2.2.3. Psychological factors for suicidal ideation

Hopelessness was measured using the Chinese version of Beck Hopelessness Scale (CHOPE, Shek, 1993 ) which was translated and modified from the original Beck Hopelessness Scale ( Beck et al., 1974 ). CHOPE is a self-administrated questionnaire which comprises 20 statements evaluating the extent of negative attitudes about the future on a 6-point Likert scale (instead of true–false statements in the original scale), with higher score representing greater degree of hopelessness. Factor analysis of CHOPE generated three components including hopelessness, certainty about the future and future expectation factors ( Shek, 1993 ). Validation of CHOPE has previously been reported with high internal consistency and reliability ( Shek, 1993 ). Internality, Powerful Others, and Chance Scale (IPC) was used to measure locus of control which can be further divided into internal control, powerful others, and chance control factors ( Levenson, 1981 ). The Chinese version of IPC has been shown to have good psychometric properties in a previous study ( Chan, 2000 ). Barratt Impulsiveness Scale (BIS) was used to assess personality trait of impulsivity and comprises three separate dimensions including attentional, motor and non-planning impulsiveness ( Patton et al., 1995 ). Brief Reasons for Living Inventory for Adolescents (BRFL-A) was employed to assess how participants weighed a range of reasons that may be important for not committing suicidal behavior and comprises five factors including fear of social disapproval, moral objections, survival and coping beliefs, responsibility to family, and fear of suicide ( Osman et al., 1996 ). Chinese versions of BIS and BRFL-A were used and were shown to have satisfactory internal consistency with Cronbach's alpha above 0.7 in a previous validation study on patients with early psychosis.

2.2.4. Neurocognitive assessments

Two executive function tests were administered to evaluate the degree of cognitive inflexibility and impaired control of inhibition. Modified Wisconsin Card Sorting Test (MWCST, Nelson, 1976 ) was used with the number of perseverative errors and categories completed being measured as indices for cognitive flexibility. Hayline Sentence Completion Test (HSCT) Part B ( Burgess and Shallice, 1996 ) was employed to assess participants' ability to suppress inappropriate response which was suggested as a neurocognitive measure of impulsivity. Participant was firstly presented with sentences in which the last word was omitted but was strongly suggested by the rest of the sentence, and was then required to give a word that made no sense in the sentence context. To correctly perform the task, the participant was required to inhibit a strongly cued automatic response and to provide an unrelated answer. Total error score was computed according to the error classification and scoring procedure outlined by Burgess and Shallice (1996) as an index for executive inhibitory control. The present study adopted the Chinese version of HSCT which has previously been applied in a local first-episode schizophrenia population ( Chan et al., 2012 ).

2.3. Statistical analysis

The primary analysis of this study focused on identifying factors predicting the occurrence of suicidal ideation in FEP patients. First, a series of univariate logistic regression analyses were conducted with suicidal ideation status (i.e., participants with vs. without suicidal ideation) as dependent variable, and demographic, pre-treatment, clinical, neurocognitive, and psychological variables as candidate predictors. Second, those variables that showed a p value < 0.10 in preceding analyses were then entered into a multivariate regression model to determine which factors independently predicted suicidal ideation based on Wald statistics. The level of statistical significance was set at p < 0.05.

3. Results

3.1. Characteristics of the sample

Eighty-nine participants were recruited and 48.3% (n = 43) were male. The mean age of the sample at intake was 20.5 years (S.D. = 3.3) and the mean educational level was 11.8 years (S.D. = 2.2). The median DUP of the sample was 109 days (mean = 239.9, S.D. = 364.1). The majority (50.6%) were diagnosed with schizophrenia-spectrum disorder (schizophrenia: n = 40; schizophreniform disorder: n = 5). For other non-affective psychoses, 24.7% (n = 22) of the cohort had brief psychotic disorder, 7.8% (n = 7) had delusional disorder and 16.9% (n = 15) had psychosis NOS. Of the 89 participants, 23 (25.8%) had engaged in pre-treatment suicide attempt and 37 (41.6%) expressed suicidal ideation after service enrolment.

3.2. Univariate associations with suicidal ideation

The associations of suicidal ideation with demographic, pre-treatment, clinical, psychological and neurocognitive factors are shown inTable 1 and Table 2. Patients with suicidal ideation were significantly more likely to have past suicide attempt, greater depressive symptom severity and lower DE score than those without suicidal ideation. Suicidal ideators were also found to have a higher degree of hopelessness, lower CHOPE future expectation factor score, greater attentional impulsiveness, lower level of internal locus of control, higher ratings on IPC Powerful Others and Chance subscales, and to be more likely to endorse fear of social approval but less likely to regard survival and coping beliefs as reasons for living than non-ideators.

Table 1 Univariate analyses of demographic and clinical characteristics of suicidal ideation in patients with first-episode psychosis.

Variables of interest Patients with suicidal ideation (n = 37) Patients without suicidal ideation (n = 52) OR p value
Age at entry, mean (SD), y 20.1 (3.2) 20.8 (3.4) − 0.94 0.312
Male sex, n (%) 15 (40.5) 28 (53.8) − 0.58 0.217
Years of education, mean (SD) 11.7 (1.8) 11.8 (2.5) − 0.99 0.952
Illness characteristics
Previous suicide attempt, n (%) 16 (43.2) 7 (13.5) 4.90 0.002
History of substance abuse, n (%) 3 (8.1) 5 (9.6) 1.21 0.807
Log (DUP + 1), mean, (SD) 1.8 (1.0) 1.8 (0.9) − 0.99 0.965
Psychiatric diagnosis, n (%)     − 0.73 0.463
Schizophrenia-spectrum disorder a 17 (45.9) 28 (53.8)    
Other non-affective psychoses b 20 (54.1) 24 (46.2)    
Symptom severity
PANSS positive symptom score, mean (SD) 17.8 (5.4) 19.1 (5.2) − 0.95 0.286
PANSS general psychopathology score, mean (SD) 31.9 (9.7) 33.4 (10.5) − 0.99 0.541
SANS diminished expression domain score c , mean (SD) 3.0 (3.8) 6.8 (9.3) − 0.92 0.041
SANS avolition–asociality domain score d , mean (SD) 5.9 (7.7) 7.7 (9.2) − 0.98 0.368
MADRS total score e , mean (SD) 14.1 (10.3) 9.8 (8.0) 1.05 0.049
Insight measures
IS Attribution of symptoms domain score, mean (SD) 2.3 (0.7) 2.3 (0.8) − 0.96 0.895
IS Awareness of illness domain score, mean (SD) 2.0 (0.7) 2.1 (0.7) − 0.83 0.553
IS Need for treatment domain score, mean (SD) 2.1 (0.6) 2.2 (0.7) − 0.78 0.458
IS total score, mean (SD) 6.3 (1.4) 6.6 (1.2) − 0.88 0.443

a Schizophrenia-spectrum disorder included schizophrenia, schizophreniform disorder and schizoaffective disorder.

b Other non-affective psychoses included brief psychotic disorder, delusional disorder and psychosis not otherwise specified.

c Diminished expression domain score was computed by summing individual items of Affective Blunting subscale (excluding item of global rating for affective blunting) and poverty of speech item of Alogia subscale of SANS.

d Avolition–asociality domain score was computed by summing individual items of Avolition–apathy subscale and Anhedonia–asociality subscale excluding items of global ratings for both two subscales of SANS.

e MADRS total score was computed by summing individual items of MADRS excluding item 10 which measured suicidal thoughts.

Abbreviations: DUP = Duration of untreated psychosis; IS = Insight Scale; MADRS = Montgomery–Asberg Depression Rating Scale; PANSS = Positive and Negative Syndrome Scale; SANS = Scale for the Assessment of Negative Symptoms.

Table 2 Univariate analyses of psychological and neurocognitive variables of suicidal ideation in patients with first-episode psychosis.

Variables of interest Patients with suicidal ideation (n = 37) Patients without suicidal ideation (n = 52) OR p value
Hopelessness measure
CHOPE Certainty about the future factor, mean (SD) 22.6 (3.9) 23.9 (3.3) − 0.90 0.095
CHOPE Hopelessness factor, mean (SD) 35.3 (5.6) 28.9 (6.4) 1.19 < 0.001
CHOPE Future expectation factor, mean (SD) 17.4 (2.3) 18.8 (3.1) − 0.83 0.029
Impulsivity measure
BIS a Attentional impulsiveness, mean (SD) 19.7 (3.1) 17 (3.1) 1.32 0.001
BIS Non-planning impulsiveness, mean (SD) 27.7 (4.0) 26.6 (4.8) 1.06 0.281
BIS Motor impulsiveness, mean (SD) 22.6 (3.6) 21.4 (3.8) 1.09 0.149
Locus of control
IPC b Internal control factor, mean (SD) 27.8 (4.7) 32.8 (5.5) − 0.82 < 0.001
IPC Powerful others factor, mean (SD) 25.8 (8.1) 21.9 (7.4) 1.07 0.028
IPC Chance factor, mean (SD) 28.7 (7.6) 25.2 (5.3) 1.10 0.016
Reasons for living
BRFL-A c Fear of social disapproval, mean (SD) 4.1 (1.1) 3.4 (1.3) 1.61 0.013
BRFL-A Moral objections, mean (SD) 3.5 (1.3) 3.2 (1.0) 1.28 0.206
BRFL-A Survival and coping beliefs, mean (SD) 4.2 (0.9) 4.8 (0.8) − 0.41 0.003
BRFL-A Responsibility to family, mean (SD) 4.7 (0.9) 4.4 (1.2) 1.27 0.256
BRFL-A Fear of suicide, mean (SD) 3.9 (1.1) 3.6 (1.3) 1.21 0.319
Cognitive inflexibility
MWCST categories completed, mean (SD) 4.5 (2.0) 4.5 (1.9) 1.0 0.999
MWCST perseverative errors, mean (SD) 3.0 (3.2) 3.3 (3.1) − 0.97 0.647
Dyscontrol of executive inhibition
HSCT Part B total error score, mean (SD) 3.8 (3.9) 4.2 (4.0) − 0.97 0.620

a BIS is a self-report instrument consisting of 30 items and each item is rated on a 4-point scale with a higher score indicating greater impulsiveness.

b IPC is a 24-item self-rated questionnaire and an individual item is scored in a 6-point scale with a higher score indicating higher expectation of control by the source designated.

c BRFL-A is a 14-item self-report instrument and each item is rated on a 6-point scale with a higher score indicating greater importance of the item as a reason for staying alive.

Abbreviations: BIS = Barratt Impulsivity Scale; BRFL-A = Brief Reasons for Living Inventory for Adolescents; CHOPE = Chinese version of Beck Hopelessness Scale; HSCT = Hayline Sentence Completion Test; IPC = Internality, Powerful Others and Chance Scale; MWCST = Modified Wisconsin Card Sorting Test.

3.3. Predictors of suicidal ideation in multivariate model

Multivariate regression analysis revealed that previous suicide attempt, depression, less severe DE, hopelessness, and lower level of internal locus of control independently predicted suicidal ideation ( Table 3 ). Overall, 84.6% of the participants were predicted correctly with the model (78.6% and 89.2% for suicidal ideators and non-ideators, respectively), which accounted for 66.3% of the variance.

Table 3 Multivariate logistic regression analysis for predictors of suicidal ideation in patients with first-episode psychosis a .

Variables in the Equation B SE Wald df p value Exp (B) 95% CI
Previous suicide attempt 2.601 1.101 5.583 1 0.018 13.472 1.558–116.494
MADRS total score b 0.174 0.063 7.633 1 0.006 1.137 1.052–1.347
SANS diminished expression domain score c − 0.378 0.128 8.721 1 0.003 0.685 0.534–0.881
CHOPE Hopelessness factor 0.164 0.066 6.185 1 0.013 1.179 1.035–1.342
IPC Internal control factor − 0.169 0.084 4.051 1 0.044 0.845 0.717–0.996
Constant 3.062 4.678 0.428 1 0.513 21.371  
Final model: Nagelkerke R2 = 0.663, χ2 = 44.31, p < 0.0001
Hosmer & Lemeshow test supported the goodness of fit of the model (χ2 = 1.262, df = 6, p = 0.989)

a CHOPE Future expectation factor, IPC Powerful others factor and Chance factor, BIS Attentional impulsivity, BRFL-A Fear of social disapproval and Survival and coping beliefs, which were entered in the stepwise logistic regression model, were excluded as predictors of suicidal ideation after analysis.

b MADRS total score was computed by summing individual items of MADRS excluding item 10 which measured suicidal thoughts.

c Diminished expression domain score was computed by summing individual items of the Affective Blunting subscale (excluding item of global rating for affective blunting) and poverty of speech item of the Alogia subscale of SANS

Abbreviations: BIS = Barratt Impulsiveness Scale; BRFL-A = Brief Reasons for Living Inventory for Adolescents; CHOPE = Chinese version of Beck Hopelessness Scale; IPC = Internality, Powerful Others and Chance Scale; MADRS = Montgomery–Asberg Depression Rating Scale; SANS = Scale for the Assessment of Negative Symptoms.

4. Discussion

Our findings showed that 41.6% of patients expressed suicidal ideation after service entry. This was consistent with the literature which revealed that approximately one-fourth to half of FEP patients reported suicidal ideation at initial presentation (Nordentoft et al, 2002, Tarrier et al, 2007, Bakst et al, 2010, Barrett et al, 2010, Melle et al, 2010, and Upthegrove et al, 2010). In line with previous first-episode studies, our data indicated past suicide attempt as a predictor of suicidal ideation (Bertelsen et al, 2007 and Bakst et al, 2010). We also confirmed results of earlier reports which found that depression and hopelessness were associated with suicidal ideation in first-episode populations (Nordentoft et al, 2002, Bertelsen et al, 2007, and Bakst et al, 2010). In particular, our results demonstrated that hopelessness predicted suicidal ideation independent of depression, thereby highlighting its critical role in determining suicide risk. Alternatively, substantial evidence indicated that depressive symptoms frequently emerged during early illness phase (Romm et al, 2010 and Cotton et al, 2012). Hence, close monitoring and early intervention for depression in FEP are crucial to minimizing suicidal ideation and attempt, particularly in the first few years of illness which is consistently shown to be a period with the highest suicide risk.

Our result that less severe DE was linked to suicidal ideation was in contrary to the majority of literature which suggested a lack of association between negative symptoms and suicidal behavior (Haw et al, 2005 and Hawton et al, 2005). Nonetheless, our findings concurred with some previous studies which revealed that negative symptom severity was inversely related to the risk of completed suicide in schizophrenia ( Fenton et al., 1997 ) and suicide attempt in FEP ( Bertelsen et al., 2007 ). In fact, it has been postulated that patients with prominent negative symptoms, particularly deficits in emotion expressivity, may have significantly impaired capacity to experience emotional distress caused by the illness ( Fenton et al., 1997 ). This may probably reduce the likelihood of developing a sense of hopelessness and suicidal ideation. Recent evidence also indicated that schizophrenia patients with affect flattening, a core feature of DE, was associated with a dysfunction in emotion perception ( Gur et al., 2006 ). Conversely, a null finding on the relationship between negative symptoms and suicidal ideation or attempted suicide as revealed by many previous reports may partly be attributable to the methodological limitation of examining negative symptoms as a unitary construct rather than separate symptom sub-domains of DE and AA ( Messinger et al., 2011 ). Evaluating negative symptoms by single composite rating as adopted by most prior studies may thus mask potential differential relationships of negative symptom sub-domains with suicidal ideation.

The impact of perceived locus of control on suicide risk in psychosis has, thus far, not been adequately investigated. Literature on suicide attempters with non-psychotic conditions showed that individuals having higher level of internal locus of control exhibited better psychosocial adjustment to stress, while those with greater external locus of control were more likely to experience learned helplessness which might in turn trigger depression, suicidal thoughts and attempts (Goldney et al, 1991, Lester et al, 1991, and Beautrais et al, 1999). In this study, we found that suicidal ideation was negatively related to internal locus of control and positively associated with external locus of control. Final multivariate model further confirmed the role of internal locus of control in independently predicting suicidal ideation in our first-episode cohort. Our results may therefore suggest that patients who perceive the illness and its consequence as beyond their personal control may be more prone to develop suicide ideation.

Consistent with most prior FEP (Pompili et al, 2011 and Challis et al, 2013) and schizophrenia research (Haw et al, 2005 and Hawton et al, 2005) on suicidality, we failed to find a significant association between positive symptoms and suicidal ideation. We also failed to demonstrate that insight into illness significantly distinguished suicidal ideators from non-ideators. This was in contrary to some past studies which showed that better insight was related to suicidal ideation and attempted suicide in FEP patients (Crumlish et al, 2005, Robinson et al, 2009, and Barrett et al, 2010). Our findings of lack of association between insight and suicidal ideation were, however, in agreement with all of those very few studies that have examined predictors of suicidal ideation in FEP (Nordentoft et al, 2002, Bertelsen et al, 2007, Tarrier et al, 2007, and Bakst et al, 2010). It should, however, be noted that in general the results on the relationship of insight with suicidality in psychosis were mixed ( Lopez-Morinigo et al., 2013 ). Some investigators reported that insight into specific symptoms rather than general awareness of the illness was related to suicidal ideation and attempts ( Amador et al., 1996 ), while others revealed that the effect of insight on suicide risk was primarily mediated by hopelessness and depression (Kim et al, 2003 and Bourgeois et al, 2004). Some even demonstrated that treatment-related insight improvement was associated with a reduced risk of attempted suicide ( Bourgeois et al., 2004 ). Alternatively, it is plausible that the timing of insight assessment performed in this study may be too early for the majority of patients to have fully developed insight into the illness. Additionally, insight scale applied in our study did not evaluate awareness of illness consequence which has been linked to suicidal ideation and attempt in chronic schizophrenia sample ( Schwartz, 2000 ).

Data regarding the relationship between neurocognition and suicidal ideation or attempt in psychotic disorders were limited with inconsistent results which may partly be attributable to methodological variations across studies, including differences in defining suicidality measure (i.e., completed suicide, suicide attempt, suicidal ideation or a combination of these suicide-related variables) and the choice of neurocognitive assessments administered. Five out of six previous studies that investigated the association of suicidality with neurocognitive functions focused on patients with chronic schizophrenia (De Hert et al, 2001, Kim et al, 2003, and Potkin et al, 2003 Delaney et al., 2012 ). One study found that higher intelligence predicted completed suicide ( De Hert et al., 2001 ), while another study revealed that patients with a history of suicide attempt had better executive function than those lifetime non-attempters ( Nangle et al., 2006 ). Three other studies which used a single composite rating in measuring suicidality (combining suicidal ideation and suicide attempt) demonstrated conflicting findings, with one showing a significant relationship between global cognition and suicidality ( Delaney et al., 2012 ) while the others suggesting lack of independent contributions of neurocognitive functions to suicidality (Kim et al, 2003 and Potkin et al, 2003). Yet, the only study that examined FEP sample failed to find a significant difference on neurocognitive performance between suicide attempters and non-attempters ( Barrett et al., 2011 ). Conversely, recent studies on non-clinical samples and patients with non-psychotic mental disorders suggested that dyscontrol of executive inhibition and cognitive inflexibility may be particularly implicated in the emergence of suicidal ideation (Marzuk et al, 2005 and Miranda et al, 2012) and attempted suicide (Swann et al, 2005 and Wu et al, 2009). However, our results indicated that neither two executive functions were related to suicidal ideation in FEP population.

The study results have to be interpreted considering the following methodological limitations. First, the evaluation of suicidal ideation relied on the participants' self-reports and was thus subject to the problem of under-reporting. Nonetheless, we employed BSS, a questionnaire that specifically and systematically enquired into the presence of suicidal ideation, thereby minimizing an under-estimation bias. Second, the cross-sectional study design precluded us from drawing conclusions about causal relationship between identified risk factors and suicidal ideation. Third, as it is suggested that correlates of suicide attempt and suicidal ideation might be different ( Nordentoft et al., 2002 ), hence caution should be exercised in generalizing our findings to suicide attempt. Fourth, it should be noted that this study did not investigate factors predicting transition from suicidal ideation into an attempt. Although evidence indicated that suicidal ideation is a predictor of attempted and completed suicide, it is also acknowledged that suicidal ideation does not necessarily convert to suicidal behavior. Thus, further longitudinal follow-up studies are required to identify factors determining progression from suicidal ideation toward an attempt.

Our results have several clinical implications. First, a significant proportion of FEP patients reported suicidal ideation after service entry. Thus, systematic suicide risk assessment including the presence of suicidal ideation should be conducted at intake and then on a regular basis during follow-up to facilitate early identification and intervention of high-risk cases. Second, patients with a previous suicide attempt should receive a more intensive monitoring of emergence of suicidal ideation. Third, given that depression, sense of hopelessness and perceived inadequate personal control, which were found to be associated with suicidal ideation, can be routinely assessed and potentially modifiable by treatment such as cognitive therapy, close monitoring and prompt intervention of these risk factors may reduce occurrence and perpetuation of suicidal ideation, and hence enhance suicide prediction and prevention in the early course of psychosis.

In conclusion, our results indicated a high rate of suicidal ideation in young people presenting with FEP. Previous suicide attempt, depression, fewer deficits in emotion expressivity, hopelessness and lower level of internal locus of control were found to be independently associated with suicidal ideation. Further prospective research investigating the longitudinal course of suicidal ideation and its dynamic relationships with clinical, neurocognitive and psychological variables is required to enhance understanding and thus prediction of the transition from suicidal ideation into an actual attempt.

Role of funding source

The study was supported by grant 21500.10202404 from the Research Grants Council, University Grants Committee, Hong Kong.


Author E.Y.H.C. designed the study. Author W.C.C. and E.S.M.Cs. managed literature search and statistical analysis and wrote the first draft of the manuscript. All authors contributed to and have approved the final manuscript.

Conflicts of interest

Author E.Y.H.C. has participated in the paid advisory board for Otsuka, has received educational grant support from Janssen-Cilag, and has received research funding from Astra-Zeneca, Janssen-Cilag, Eli Lilly, Sanofi-Aventis and Otsuka. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.


We thank all the coordinating clinicians and staff from the psychiatric units at Queen Mary Hospital, Pamela Youde Nethersole Eastern Hospital and Kwai Chung Hospital for their kind assistance and support. We are also grateful to the individuals who participated in the study.


  • Amador et al., 1996 X.F. Amador, J. Harkavy-Friedman, C. Kasapis, S.A. Yale, M. Flaum, J.M. Gorman. Suicidal behavior in schizophrenia and its relationship to awareness of illness. Am. J. Psychiatry. 1996;153:1185-1188
  • American Psychiatric Association, 1994 American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders 4th Edition. (American Psychiatric Association, Washington DC, 1994)
  • Andreasen, 1982 N.C. Andreasen. Negative symptoms in schizophrenia. Definitions and reliability. Arch. Gen. Psychiatry. 1982;39:784-788 Crossref
  • Bakst et al., 2010 S. Bakst, J. Rabinowitz, E.J. Bromet. Antecedents and patterns of suicide behavior in first-admission psychosis. Schizophr. Bull.. 2010;36:880-889 Crossref
  • Barrett et al., 2010 E.A. Barrett, K. Sundet, A. Faerden, et al. Suicidality in first episode psychosis is associated with insight and negative beliefs about psychosis. Schizophr. Res.. 2010;123:257-262 Crossref
  • Barrett et al., 2011 E.A. Barrett, K. Sundet, C. Simonsen, et al. Neurocognitive functioning and suicidality in schizophrenia spectrum disorders. Compr. Psychiatry. 2011;52:156-163 Crossref
  • Beautrais et al., 1999 A.L. Beautrais, P.R. Joyce, R.T. Mulder. Personality traits and cognitive styles as risk factors for serious suicide attempts among young people. Suicide Life Threat.. 1999;29:37-47
  • Beck and Steer, 1991 A.T. Beck, R.A. Steer. Manual for Beck Scale for Suicidal Ideation. (Psychological Corporation, San Antonio, TX, 1991)
  • Beck et al., 1974 A.T. Beck, A. Weissman, D. Lester, L. Trexler. The measurement of pessimism: the Hopelessness Scale. J. Consult. Clin. Psychol.. 1974;42:861-865 Crossref
  • Bertelsen et al., 2007 M. Bertelsen, P. Jeppesen, L. Petersen, et al. Suicidal behaviour and mortality in first-episode psychosis: the OPUS trial. Br. J. Psychiatry. 2007;191(Suppl.51):140-146
  • Birchwood et al., 1994 M. Birchwood, J. Smith, V. Drury, J. Healy, F. Macmillan, M. Slade. A self-report Insight Scale for psychosis: reliability, validity and sensitivity to change. Acta Psychiatr. Scand.. 1994;89:62-67 Crossref
  • Bourgeois et al., 2004 M. Bourgeois, J. Swendsen, F. Young, et al. Awareness of disorder and suicide risk in the treatment of schizophrenia: results of the international suicide prevention trial. Am. J. Psychiatry. 2004;161:1494-1496 Crossref
  • Brezo et al., 2006 J. Brezo, J. Paris, G. Turecki. Personality traits as correlates of suicidal ideation, suicide attempts, and suicide completions: a systematic review. Acta Psychiatr. Scand.. 2006;113:180-206 Crossref
  • Burgess and Shallice, 1996 P.W. Burgess, T. Shallice. Response suppression, inhibition and strategy use following frontal lobe lesions. Neuropsychologia. 1996;34:263-273
  • Challis et al., 2013 S. Challis, O. Nielssen, A. Harris, M. Large. Systematic meta-analysis of the risk factors for deliberate self-harm before and after treatment for first-episode psychosis. Acta Psychiatr. Scand.. 2013;127:442-454 Crossref
  • Chan, 2000 R.C.K. Chan. Stress and coping in spouses of persons with spinal cord injuries. Clin. Rehabil.. 2000;14:137-144 Crossref
  • Chan et al., 2012 K.K.S. Chan, J.Q. Xu, K.C.M. Liu, C.L. Hui, G.H. Wong, E.Y. Chen. Executive function in first-episode schizophrenia: a three-year prospective study of the Hayline Sentence Completion Test. Schizophr. Res.. 2012;135:62-67 Crossref
  • Cotton et al., 2012 S.M. Cotton, M. Lambert, B.G. Schimmelmann, et al. Depressive symptoms in first episode schizophrenia spectrum disorder. Schizophr. Res.. 2012;134:20-26 Crossref
  • Crumlish et al., 2005 N. Crumlish, P. Whitty, M. Kamali, et al. Early insight predicts depression and attempted suicide after 4 years in first-episode schizophrenia and schizophreniform disorder. Acta Psychiatr. Scand.. 2005;112:449-455 Crossref
  • De Hert et al., 2001 M. De Hert, K. McKenzie, J. Peuskens. Risk factors for suicide in young people suffering from schizophrenia: a long-term follow-up study. Schizophr. Res.. 2001;47:127-134 Crossref
  • Delaney et al., 2012 C. Delaney, J. McGrane, E. Cummings, et al. Preserved cognitive function is associated with suicidal ideation and single suicide attempts in schizophrenia. Schizophr. Res.. 2012;140:232-236 Crossref
  • Dutta et al., 2010 R. Dutta, R.M. Murray, M. Hotopf, J. Allardyce, P.B. Jones, J. Boydell. Reassessing the long-term risk of suicide after a first episode of psychosis. Arch. Gen. Psychiatry. 2010;67:1230-1237 Crossref
  • Fenton et al., 1997 W.S. Fenton, T.H. McGlashan, B.J. Victor, C.R. Blyler. Symptoms, subtype and suicidality in patients with schizophrenia spectrum disorders. Am. J. Psychiatry. 1997;154:1999-2004
  • Fialko et al., 2006 L. Fialko, D. Freeman, P.E. Bebbington, et al. Understanding suicidal ideation in psychosis: findings from the Psychological Prevention of Relapse in Psychosis (PRP) trial. Acta Psychiatr. Scand.. 2006;114:177-186 Crossref
  • Foussias et al., 2009 G. Foussias, S. Mann, K.K. Zakzanis, R. van Reekum, G. Remington. Motivational deficits as the central link to functioning in schizophrenia: a pilot study. Schizophr. Res.. 2009;115:333-337 Crossref
  • Goldney et al., 1991 R.D. Goldney, S. Smith, A.H. Winefield, M. Tiggeman, H.R. Winefield. Suicide ideation: it's enduring nature and associated morbidity. Acta Psychiatr. Scand.. 1991;83:115-120 Crossref
  • Gur et al., 2006 R.E. Gur, C.G. Kohler, J.D. Ragland, et al. Flat affect in schizophrenia: relation to emotion processing and neurocognitive measures. Schizophr. Bull.. 2006;32:279-287 Crossref
  • Hafner et al., 1992 H. Hafner, A. Riecher-Rossler, M. Hambrecht, et al. IRAOS: an instrument for the assessment of onset and early course of schizophrenia. Schizophr. Res.. 1992;6:209-223 Crossref
  • Harvey-Friedman, 2006 J.M. Harvey-Friedman. Can early detection of psychosis prevent suicidal behavior?. Am. J. Psychiatry. 2006;163:768-770
  • Haw et al., 2005 C. Haw, K. Hawton, L. Sutton, J. Sinclair, J. Deeks. Schizophrenia and deliberate self-harm: a systematic review of risk factors. Suicide Life Threat.. 2005;35:50-62 Crossref
  • Hawley et al., 1991 C.J. Hawley, D.V. James, P.L. Birkett, D.S. Baldwin, M.J. de Ruiter, R.G. Priest. Suicidal ideation as a presenting complaint. Br. J. Psychiatry. 1991;159:232-238 Crossref
  • Hawton et al., 2005 K. Hawton, L. Sutton, C. Haw, J. Sinclair, J.J. Deeks. Schizophrenia and suicide: systematic review of risk factors. Br. J. Psychiatry. 2005;187:9-20 Crossref
  • Hirsch and Ellis, 1996 J.K. Hirsch, J.B. Ellis. Differences in life stress and reasons for living among college suicide ideators and non-ideators. Coll. Stud. J.. 1996;30:377-386
  • Kay et al., 1987 S.R. Kay, L.A. Opler, A. Fiszbein. Positive and Negative Syndrome Scale (PANSS) for schizophrenia. Schizophr. Bull.. 1987;13:261-276 Crossref
  • Kim et al., 2003 C.H. Kim, K. Jayathilake, H.Y. Meltzer. Hopelessness, neurocognitive function, and insight in schizophrenia: relationship to suicidal behavior. Schizophr. Res.. 2003;63:71-80 Crossref
  • Kontaxakis et al., 2004 V. Kontaxakis, B. Havaki-Kontaxaki, M. Margariti, S. Stamouli, C. Kollias, G. Christodoulou. Suicidal ideation in inpatients with acute schizophrenia. Can. J. Psychiatry. 2004;49:476-479
  • Lam et al., 1996 L.C. Lam, C.K. Chan, E.Y. Chen. Insight and general public attitude on psychotic experiences in Hong Kong. Int. J. Soc. Psychiatry. 1996;42:10-17 Crossref
  • Lester et al., 1991 D. Lester, I.J. Castromayor, T. Icli. Locus of control, depression, and suicidal ideation among American, Philippine, and Turkish students. J. Soc. Psychol.. 1991;131:447-449 Crossref
  • Levenson, 1981 H. Levenson. Differentiating among internality, powerful others, and chance. H.M. Lefcourt (Ed.) Research with the Locus of Control Construct in: volume 1 (Academic Press, New York, 1981) 15-63 Crossref
  • Lopez-Morinigo et al., 2013 J.D. Lopez-Morinigo, R. Ramos-Rios, A.S. David, R. Dutta. Insight in schizophrenia and risk of suicide: a systematic update. Compr. Psychiatry. 2013;53:313-322
  • Mann et al., 1999 J.J. Mann, C. Waternaux, G.L. Haas, K.M. Malone. Toward a clinical model of suicidal behavior in psychiatric patients. Am. J. Psychiatry. 1999;156:181-189
  • Marzuk et al., 2005 P.M. Marzuk, N. Hartwell, A.C. Leon, L. Portera. Executive functioning in depressed patients with suicidal ideation. Acta Psychiatr. Scand.. 2005;112:294-301 Crossref
  • Melle et al., 2010 I. Melle, J.O. Johannessen, S. Friis, et al. Course and predictors of suicidality over the first two years of treatment in first-episode schizophrenia spectrum psychosis. Arch. Suicide Res.. 2010;14:158-170 Crossref
  • Messinger et al., 2011 J.W. Messinger, F. Tremeau, D. Antonius. Avolition and expressive deficits capture negative symptom phenomenology: implications for DSM-5 and schizophrenia research. Clin. Psychol. Rev.. 2011;31:161-168 Crossref
  • Miranda et al., 2012 R. Miranda, M. Gallagher, B. Bauchner, R. Vaysman, B. Marroquin. Cognitive inflexibility as a prospective predictor of suicidal ideation among young adults with a suicide attempt history. Depress Anxiety. 2012;29:180-189
  • Montgomery and Asberg, 1979 S.A. Montgomery, M.A. Asberg. A new depression scale designed to be sensitive to change. Br. J. Psychiatry. 1979;134:382-389 Crossref
  • Nangle et al., 2006 J.M. Nangle, S. Clarke, D.W. Morris, et al. Neurocognition and suicidal behaviour in an Irish population with major psychotic disorders. Schizophr. Res.. 2006;85:196-200 Crossref
  • Nelson, 1976 H.E. Nelson. A modified card sorting test sensitive to frontal lobe defects. Cortex. 1976;12:313-324 Crossref
  • Nordentoft et al., 2002 M. Nordentoft, P. Jeppesen, M. Abel, et al. OPUS study: suicidal behaviour, suicidal ideation and hopelessness among patients with first-episode psychosis: one-year follow-up of a randomized controlled trial. Br. J. Psychiatry. 2002;181(Suppl.43):98-106
  • Nordentoft et al., 2004 M. Nordentoft, T.M. Laursen, E. Agerbo, P. Qin, E.H. Hoyer, P.B. Mortensen. Change in suicide rates for patients with schizophrenia in Denmark, 1981–97: nested case–control study. BMJ. 2004;329:261 Crossref
  • Osman et al., 1996 A. Osman, B.A. Kopper, F.X. Barrios, J.R. Osman, T. Besett, M.M. Linehan. The Brief Reasons for Living Inventory for Adolescents (BRFL-A). J. Abnorm. Child Psychol.. 1996;24:433-443 Crossref
  • Palmer et al., 2005 B.A. Palmer, V.S. Pankratz, J.M. Bostwick. The lifetime risk of suicide in schizophrenia: a reexamination. Arch. Gen. Psychiatry. 2005;62:247-253 Crossref
  • Patton et al., 1995 J.H. Patton, M.S. Stanford, E.S. Barratt. Factor structure of the Barratt Impulsiveness Scale. J. Clin. Psychol.. 1995;51:768-774 Crossref
  • Pinto et al., 1998 A. Pinto, M.A. Whisman, Y. Conwell. Reasons for living in a clinical sample of adolescents. J. Adolesc.. 1998;21:397-405 Crossref
  • Pompili et al., 2011 M. Pompili, G. Serafini, M. Innamorati, et al. Suicide risk in first episode psychosis: a selective review of the current literature. Schizophr. Res.. 2011;129:1-11 Crossref
  • Potkin et al., 2003 S.G. Potkin, R. Anand, L. Alphs, K. Fleming. Neurocognitive performance does not correlate with suicidality in schizophrenic and schizoaffective patients at risk for suicide. Schizophr. Res.. 2003;59:59-66 Crossref
  • Robinson et al., 2009 J. Robinson, S. Cotton, P. Conus. Prevalence and predictors of suicide attempt in an incidence cohort of 661 young people with first-episode psychosis. Aust. N. Z. J. Psychiatr.. 2009;43:149-157 Crossref
  • Romm et al., 2010 K.L. Romm, J.I. Rossberg, A.O. Berg, et al. Depression and depressive symptoms in first episode psychosis. J. Nerv. Ment. Dis.. 2010;198:67-71 Crossref
  • Saha et al., 2007 S. Saha, D. Chant, J. McGrath. A systematic review of mortality in schizophrenia: is the differential mortality gap worsening over time?. Arch. Gen. Psychiatry. 2007;64:1123-1131 Crossref
  • Schwartz, 2000 R.C. Schwartz. Insight and suicidality in schizophrenia: a replication study. J. Nerv. Ment. Dis.. 2000;188:235-237 Crossref
  • Shek, 1993 D.T.L. Shek. Measurement of pessimism in Chinese adolescents: the Chinese Hopelessness Scale. Soc. Behav. Personal.. 1993;21:107-120
  • So et al., 2003 E. So, I. Kam, C.M. Leung, D. Chung, Z. Liu, S. Fong. The Chinese-bilingual SCID-I/P project: stage 1: reliability for mood disorders and schizophrenia. HK. J. Psychiatry. 2003;13:7-18
  • Swann et al., 2005 A.C. Swann, D.M. Dougherty, P.J. Pazzaglia, M. Pham, J.L. Steinberg, F.G. Moeller. Increased impulsivity associated with severity of suicide attempt history in patients with bipolar disorder. Am. J. Psychiatry. 2005;162:1680-1687 Crossref
  • Tang et al., 2010 J.Y.M. Tang, G.H.Y. Wong, C.L.M. Hui, et al. Early intervention for psychosis in Hong Kong: the EASY programme. Early Interv. Psychiatry. 2010;4:214-219 Crossref
  • Tarrier et al., 2007 N. Tarrier, S. Khan, J. Cater, A. Picken. The subjective consequences of suffering a first episode psychosis: trauma and suicide behaviour. Soc. Psychiatry Psychiatr. Epidemiol.. 2007;42:29-35 Crossref
  • Upthegrove et al., 2010 R. Upthegrove, M. Birchwood, K. Ross, K. Brunett, R. McCollum, L. Jones. The evolution of depression and suicidality in first episode psychosis. Acta Psychiatr. Scand.. 2010;122:211-218
  • Wu et al., 2009 C.S. Wu, S.C. Liao, K.M. Lin, M.M.C. Tseng, E.C.H. Wu, S.K. Liu. Multidimensional assessments of impulsivity in subjects with history of suicidal attempts. Compr. Psychiatry. 2009;50:315-321 Crossref
  • Young et al., 1998 A.S. Young, K.H. Nuchterlein, J. Mintz, J. Ventura, M. Gitlin, R.P. Liberman. Suicide ideation and suicide attempts in recent-onset schizophrenia. Schizophr. Bull.. 1998;24:629-634 Crossref


a Department of Psychiatry, The University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong

b State Key Laboratory of Brain and Cognitive Sciences, The University of Hong Kong, Hong Kong

lowast Corresponding author: Tel.: + 852 22554486; fax: + 852 28551345.