Critical Issues in Adolescence

ContentServer.pdf

Home >Psychology homework help >Critical Issues in Adolescence

Research Trends

Predicting Suicidal Risk in a Cohort of Depressed Children and Adolescents

Sarah E. Hetrick1, Alexandra G. Parker1, Jo Robinson1, Nicole Hall2, and Alasdair Vance2

1Orygen Youth Health Research Centre, Centre for Youth Mental Health, University of Melbourne, Locked Bag 10, Parkville, Victoria 3052, Australia, 2Academic Child Psychiatry Unit, Department of

Pediatrics, University of Melbourne and Royal Children’s Hospital and Murdoch Childrens Research Institute, Parkville, Victoria, Australia

Abstract. Background: In children and adolescents with a depressive disorder, predicting who will also go on to exhibit suicide-related behaviors (SRBs), including suicide attempt or self-harm, is a key challenge facing clinicians. Aims: To investigate the relative contri- butions of depressive disorder severity, hopelessness, family dysfunction, and perceived social support to the risk of suicide-related behaviors. Methods: This was a cross-sectional study of a group of 10–16-year-olds with major depressive disorders and dysthymic disorder. Results: Child-rated depressive disorder symptom severity emerged as the greatest predictor of risk. Hopelessness and family dysfunction were also significant predictors of SRBs. In combination these variables were strong predictors, accounting for 66% of the variance. This is a cross-sectional study design, rather than longitudinal, therefore risk prediction over time was not possible. Conclusions: Understanding the child and adolescents depressive disorder symptom severity from their perspective, their level of hopelessness, as well as their family context is critical in understanding the risk of SRBs. These findings may help to provide direction for targeted interventions to address these clinical risk factors.

Keywords: suicide risk, depression, child and adolescent, family functioning

Introduction

The rates of death by suicide have increased in young peo- ple since the middle of the last century (Wasserman, Cheng, & Jiang, 2005), and although there has been a decrease in rates since the mid-1990s (Gould, Greenberg, Velting, & Shaffer, 2003), it remains of considerable concern. Delib- erate self-harm (DSH), including that with the intention of dying, and suicidal ideation have also been a growing prob- lem and are common among adolescents (Hawton, Rod- ham, Evans, & Weatherall, 2002). For example, up to 7% of secondary school students reported engaging in DSH in a 12-month period (De Leo & Heller, 2004), and at any point in time 15–25% of young people may be experienc- ing suicidal ideation (Grunbaum et al., 2004).

After a history of self-harm or suicide attempt (Cava- nagh, Carson, Sharpe, & Lawrie, 2003), the presence of a depressive disorder or depressive symptoms is a key risk factor for future suicidal behavior, including suicide com- pletion (Andrews & Lewinsohn, 1992; Beautrais, 1998; Shaffer et al., 1996). However, among those with a depres- sive disorder or symptoms the prediction of who will go on to attempt suicide or self-harm is a key challenge facing

clinicians. While the risk of suicidal behaviors is similar for those with major depressive disorder and dysthymic disorder (Hetrick, Vance, & Hall, 2008), the severity of de- pressive disorder symptoms appears to be closely associat- ed with suicidal behaviors (Asarnow, 1992; Brent et al., 1986; Esposito & Clum, 2002; Hetrick et al., 2008).

There are numerous additional risk factors for suicide. A useful model to conceptualize the relationship between the various risk factors is the stress-diathesis model (Mann, 2003) that has been summarized and simplified in a recent review (Hawton & van Heeringen, 2009). Precursors are said to occur in the context of a diathesis, which may be one or more factors that potentially predispose someone to suicidal behavior. The diathesis is influenced by familial and genetic factors, for example, early and current family environment and interactions (Bridge, Goldstein, & Brent, 2006; Hawton & van Heeringen, 2009). An example of a diathesis is a negative cognitive style, such as hopelessness or a tendency toward negative causal attributions, which may be influenced by poor family functioning and a lack of parental support. The onset of a depressive disorder or a stressful life event in this context is said to increase the risk for death by suicide.

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Of particular interest to those researching suicide has been hopelessness. It is associated with and is a major risk factor for suicidal behaviors in adults, independent of the effect of depressive disorders (Coryell & Young, 2005; Joiner, Brown, & Wingate, 2005; Kingsbury, Hawton, Steinhardt, & James, 1999). However, the association is not yet clearly established in children and adolescents.

A perceived lack of social support, particularly sup- port from the family (Lewinsohn, Rohde, & Seeley, 1993), has been shown to be predictive of suicidal behav- iors in young people (Mazza & Reynolds, 1998), even after the effects of depressive disorders are controlled (Lewinsohn et al., 1993; Mazza & Reynolds, 1998). A sense of belonging to, and support from, a peer group is also important, with studies showing an association be- tween social networks and psychological and general health, as well as mortality outcomes (Giles, Glonek, Lu- szcz, & Andrews, 2005; Jorm, 2005). On the other hand, Greening and Stoppelbein (2002) showed that perceived lack of social support from the family was related to sui- cide risk, while perceived lack of support from friends was not.

Difficult family relationships and interactions, family turmoil, and problems with parents have also been asso- ciated with suicidal behavior (de Wilde, Kienhorst, Diek- stra, & Wolters, 1993, 1994; Kienhorst, de Wilde, Diek- stra, & Wolters, 1992; Kosky, Silburn, & Zubrick, 1986; Lewinsohn et al., 1993). Although several studies have confirmed that this association is independent of the ef- fects of a depressive disorder (Groholt, Ekeberg, Wich- strom, & Haldorsen, 2000; King, Raskin, Godwski, But- kus, & Opipari, 1990), the association overall remains equivocal (Brent, Kolko, Allan, & Brown, 1990). This may be the result of parental versus child accounts of de- pression being used in such studies; for example, studies have shown that parents tend to report less depressive symptoms than do children (Angold, 1988).

To date, the role of perceived social support has not been extensively examined and the findings for hopeless- ness and family dysfunction as potential discriminating risk factors for suicidal behaviors in those who have a depressive disorder are mixed. This study seeks to use the stress-diathesis model to examine the association be- tween depressive disorder symptom severity, hopeless- ness, perceived social support and family dysfunction with suicidal behavior in children and adolescents with major depressive disorder or dysthymic disorder. The findings will contribute to the theoretical discussion of this area and potentially enable clinicians to manage young people at risk of suicide in their practice.

The hypothesis is that young people with depressive disorders who also exhibit suicidal behavior will have higher depressive disorder symptom severity, higher lev- els of hopelessness, lower perceived social support and higher family dysfunction compared to young people with depressive disorders who do not exhibit suicidal behaviors.

Methods

Study Population

This was a cross-sectional study of 59 participants conduct- ed from 2003 to 2005 in specialized depressive disorder clinics set up in Child and Adolescent Mental Health Ser- vices in two metropolitan hospitals in Melbourne, Austra- lia. These services are government funded and assess and manage children and adolescents from primary health care and mental health clinicians within the community. Ethics approval was obtained from the relevant ethics committees. Participants and their families/guardians were required to provide written consent before participation.

Participants had to be between 10 to 16-years-old and meet DSM-IV criteria for major depressive disorder or dys- thymic disorder. Major depressive disorder and dysthymic disorder were defined by a semistructured interview (Sil- verman & Albano, 1996) with the participant and the par- ticipant’s parent(s) and by the participants report of depres- sive disorder symptom severity scores (Achenbach & Edel- brock, 1983) being greater than 1.5 standard deviations above the mean for a given participant’s age and gender.

Exclusion criteria included current medication, an IQ below 70 (as assessed by the WISC-III-R), a diagnosis of pervasive developmental disorder, evidence of a major neurological condition, a psychotic disorder, bipolar disor- der, or ADHD combined type.

Measures

Demographic data relevant to depression and SRBs was collected, including age, gender, a measure of social adver- sity (using items including family income, mother’s educa- tion, single parent status, sibling size, and broken home status) in the Parental Account of Childhood Symptoms (Taylor, Schachar, Thorley, & Wieselberg, 1986), and pa- rental psychopathology (The Hopkins Symptom Checklist [HSCL]; Derogatis, Lipman, Rickels, Uhlenhuth, & Covi, 1974). The psychometric properties of each measure are sound (Derogatis et al., 1974; Taylor, Schachar, Thorley, & Wieselberg, 1986).

The Wechsler Intelligence Scale for Children Revised (WISC-III-R; Wechsler, 1991) provides verbal, perfor- mance, and full-scale IQ. It is suitable for children aged 6 to 16 years. It is a clinician-administered test involving a series of timed and untimed, verbal and nonverbal tasks.

The Anxiety Disorders Interview Schedule for Children (A-DISC; Silverman & Albano, 1996) is a semistructured diagnostic interview schedule based on DSM-IV criteria with child and parent versions. In each version, (1) symptoms and (2) associated impairment in academic, social and family do- mains are rated separately. Research findings support the clinical utility, reliability and validity of the A-DISC.

The Children’s Depression Inventory (CDI; Kovacs, 1992) consists of 27 self-rated items to assess depression.

14 S. E. Hetrick et al.: Predicting Suicidal Risk in a Cohort of Depressed Children and Adolescents

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Each item has a 3-point response set (scored 0, 1, or 2) such that total scores range from 0 to 54, with higher scores reflect- ing more severe depression. It has good psychometric prop- erties with internal consistency ranging from 0.71 to 0.87 and test-retest reliability from 0.82 to 0.87 (Kovacs, 1992).

The Child Behavior Checklist (CBCL; Achenbach & Edelbrock, 1983) consists of 112 behavior problem items that are rated by the parent on a 3-point scale as to how well each describes the child. The Withdrawn/Depressed subscale was used, where higher scores reflect greater par- ent-reported depression severity. The CBCL is well re- searched and has adequate psychometric properties.

The Hopelessness Scale for Children (HSC; Kazdin, Rodgers, & Colbus, 1986) consists of 17 true/false items that assess the respondent’s expectations of the future. Higher scores reflect a greater level of self-reported hope- lessness. It has moderate to high internal consistency (Cronbach’s α = 0.75; Kazdin et al., 1986).

The family and the friend components of the Perceived Social Support Questionnaire (Procidano & Heller, 1983) each consist of 20 items to assess the respondent’s perception of social support, including dimensions related to the extent to which the respondent perceived that their family and friends fulfill their needs for support, feedback, and interac- tion. Higher scores reflect a higher perceived level of social support from friends and family. It has high internal consis- tency (Cronbach’s α = 0.88 to 0.90) and test-retest reliability over one month is 0.83 (Procidano & Heller, 1983).

The Family Assessment Device (Epstein, Baldwin, & Bishop, 1983) was developed from the McMaster Model of Family Functioning and designed to measure the struc- tural and organizational characteristics of a family. Six di- mensions of family functioning are probed: problem-solv- ing, communication, affective responsiveness, affective in- volvement, roles, and behavior control. Each dimension consists of a subscale with an additional general function- ing subscale that measures overall health/dysfunction in the family unit. It consists of 60 items completed by a par- ent/guardian of the child or adolescent; the general func- tioning subscale (FAD-GF) was used in the current analy- sis. Higher scores indicate better levels of general family functioning. It has moderate to high internal consistency (Cronbach’s α = 0.72 to 0.92; Epstein et al., 1983).

A questionnaire, the Suicidal Ideation Device (SID), was designed for this study in order to capture the extent and type of a range of suicide-related behaviors (SRBs), including sui- cidal ideation and suicide attempt(s) that may have occurred in the month preceding study entry. Using the SID, respon- dents indicated the presence or absence of eight behavioral items with total scores ranging from 0 to 8; higher scores indicate more severe SRB.

The SID was based on the screening questions used in the Schedule for Affective Disorders and Schizophrenia for School-Aged Children. These screening questions have been shown to fit a Guttman scale, indicating that these behavioral items exist on a continuum of severity (Lewin- sohn, Rohde, & Seeley, 1996). There are few measures that

capture this continuum, despite research indicating that these behaviors, as well as suicide completion, exist on a continuum, with each type of behavior a marker of severity, increasing the risk of even more severe SRB (Andrews & Lewinsohn, 1992; Brent et al., 1986; Lewinsohn et al., 1996). Psychometric data collected from the current study showed good correlation between this questionnaire and a clinical suicidal ideation and attempt index formed from independent clinical interviews by a child-and-adolescent psychiatrist (Pearson product moment correlation coeffi- cient, r = .82 p < .0005). Similarly, data collected in the study demonstrated construct validity by correlation be- tween the CDI total score and the SID of 0.86; test-retest reliability of 0.90 over one week; interrater reliability 0.89, and high internal consistency (Cronbach’s α was 0.96).

Procedure

Consecutive referrals to the clinic were invited to participate in the study, with over 98% agreeing to participate. Once written consent was obtained, a child-and-adolescent psychi- atrist and a supervised trainee clinical psychologist undertook the assessment procedure. The A-DISC was administered to participants who completed the above self-report measures. Concurrently, each participant’s parent(s)/guardians sepa- rately completed the diagnostic assessment (A-DISC) and the FAD. At a separate session each participant was administered the WISC-III-R by a probationary psychologist, for the pur- pose of screening for exclusion from the study and measuring possible group differences in intelligence.

Data Analysis

Data analysis was performed using the Statistical Package for the Social Sciences (SPSS/PC), version 10.0. An inde- pendent samples t-test, with type I error rate set at 0.05, was completed to examine the differences in factors known to influence SRBs, including age, gender (Lewinsohn et al., 1996), IQ (Voracek, 2004), parental psychopathology, and social disadvantage and adversity (Beautrais, 2000) as well as in the dependent variables of interest between a group with a depressive disorder as well as SRBs, and a group with a depressive disorder and no SRBs. Preliminary checks identified no multivariate outliers or any violation of the assumptions of normality, linearity, multicollinearity or singularity, and homogeneity of variance-covariance matrices (Tabachnick & Fidell, 1996).

Hierarchical regression analysis was conducted to deter- mine the proportion of variance that each of the psychosocial variables contributed to suicide risk above depression severity.

Consistent with Cohen’s 1992 recommendation for in- terpreting effect sizes, any step that accounts for an addi- tional 1–5.9% of the variance was interpreted as a small effect size, between 5.9–13.8% as a medium effect size, and over 13.8% as a large effect size.

S. E. Hetrick et al.: Predicting Suicidal Risk in a Cohort of Depressed Children and Adolescents 15

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Results

Of the original 59 participants who were eligible for the study, 3 were subsequently excluded from the analysis be- cause of missing data, leaving 56 participants. These 56 participants were divided into two groups based on whether or not they reported SRBs on the SID. In the group with a depressive disorder who also had SRBs (n = 43), 24 had major depressive disorder, and 19 had dysthymic disorder. The mean score for depression severity in this group on the CDI was 20.8 (SD = 8.6). In the group with a depressive disorder who did not have SRBs (n = 13) 5 had major de- pressive disorder and 8 had dysthymic disorder. The mean score for depressive severity in this group on the CDI was 11.8 (SD = 7.3), below the clinical cutoff.

There were no statistically significant differences between the groups in age, IQ, parental psychopathology, or social adversity scores (see Table 1). There was no significant dif- ference between groups on gender (see Table 1), however, it is notable that there were more males than females in the non-SRB group compared to the SRB group (see Table 1).

An independent samples t-test yielded a significant differ- ence between the groups on the dependent variables of inter- est except for the variables measuring perceived social sup- port. The group who reported SRBs reported significantly higher depressive disorder symptom severity (CDI), greater hopelessness, and poorer family function (See Table 2).

A hierarchical regression analysis was conducted to ex- amine the unique relationships between SRBs (measured as a continuous variable) and the variables that were sig- nificantly different between the SRB and non-SRB groups, namely depression (both child and parent report; while par- ent report was not significantly different between groups, it approached significance), hopelessness, and family func- tioning. Because of the reported age and sex differences in the onset of SRBs (Lewinsohn et al., 1996), this was con- trolled for by entering these variables in the first step. Both child and parent reports of depression symptoms were en- tered in the first step because of their significant correlation (r = .4, p < .01). The entry of age, gender, self- and parent- reported depressive symptoms (CDI and CBCL-WD, re- spectively) in the first step of the analysis predicted 52.4% of the variance, F (4, 51) = 16.1, p < .001, constituting a large effect size. SRBs were predicted by older age and higher self-reported depressive disorder symptom severity (see Table 3). The entry of the main effect of hopelessness in the second step accounted for an additional 9.7% of the variance in SRBs, constituting a medium effect size, ΔR2 = .097, F change (1, 50) = 14.1, p ≤ .001. The entry of family functioning in the third step accounted for an additional 4.3% of the variance in SRBs, constituting a small effect size, ΔR2 = .043, F change (1, 49) = 7.1, p = .011, with the final model accounting for 66.2% of the total variance. SRBs were predicted by older age, higher child-reported depressive disorder symptom severity, greater hopeless- ness, and poorer family functioning.

Table 1. Demographic characteristics for group with SRBs and group without SRBs

Variable SRB present mean (SD)

SRB not present mean (SD)

Age 14.1 (2.0) 13.2 (2.4)

N (male/female) 22/21 11/2

Full scale IQ 104.3 (14.0) 101.0 (16.6)

Social adversity 6.9 (1.2) 6.6 (1.3)

Parental psychopathology 93.3 (19.9)* 97.9 (28.3)*

Note. For age, gender, full scale IQ and social adversity, parental psy- chopathology, SRB present n = 43 (three cases missing); SRB not present n = 13. *score above the clinical cutoff.

Table 2. Clinical variables related to suicide related behav- iors in depressive disorders

Clinical correlate

SRB presenta SRB not presentb

df t p

Mean (SD) Mean (SD)

n = 43 n = 13

CDI 20.8 (8.6) 11.8 (7.3) 54 3.4 .001

CBCL-WD 65.4 (8.9) 70.3 (11.3) 54 –1.8 .08

HSC 7.0 (3.4) 3.4 (2.5) 54 2.9 .01

FAD-GF 2.0 (0.5) 2.3 (0.4) 54 –2.0 .05

PSS-FA 8.5 (5.8) 9.8 (5.9) 54 –0.7 .48

PSS-FR 11.4 (5.1) 12.1 (4.4) 54 –0.4 .69

Note. SRB = Suicide Related Behaviors, CDI = Children’s Depression Inventory; CBCL-WD = T-score for the withdrawn/depressed sub- scale of the CBCL; HSC = Hopelessness Scale for Children; PSS-FA = Perceived Social Support Questionnaire – Family; PSS-FR = Per- ceived Social Support Questionnaire – Friends; FAD-GF = General Functioning subscale of the Family Assessment Device. aa score of 1 or more on the Suicidal Ideation Device, ba score of 0 on the Suicidal Ideation Device.

Table 3. Standardized regression coefficients for each step of regression analysis predicting suicide-related behaviors from depression severity, hopelessness, and family functioning

Step 1 Step 2 Step 3

Age .254** .254** .264**

Gender .046 .001 .043

Depression Severity Child report (CDI) .728** .675** .695**

Depression Severity Parent report (CBCL-WD)

.136 .128 .131

Hopelessness (HSC) .322** .343**

Family Functioning (FAD-GF) –.214*

Adjusted R2 .524** .621** .662*

Note. **p < .01, *p < .05.

16 S. E. Hetrick et al.: Predicting Suicidal Risk in a Cohort of Depressed Children and Adolescents

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Discussion

The results demonstrate that the presence of SRBs in the context of a current depressive disorder episode (major de- pressive disorder or dysthymic disorder) in 10- to 16-year- old children and adolescents is best predicted by the child report of depressive disorder symptom severity. Child re- port of depressive disorder symptom severity explained the greatest amount of variance in SRBs and was an important indicator of risk over and above the simple presence of a depressive disorder. This dimensional aspect of diagnosis is at least as important as simple categorization based on the presence or absence of depressive disorder (Beardslee et al., 1996; Greening & Stoppelbein, 2002; Klein et al., 1996; Kovacs, Akiskal, Gatsonis, & Phoebe, 1994; Ryan et al., 1987). Previous research has shown that SRBs repre- sent an effort to relieve distress (de Wilde, Kienhorst, & Diekstra, 2001; Hawton, Cole, O’Grady, & Osborn, 1982) and a higher load of symptoms, therefore, provides a great- er impetus toward these behaviors.

While not essential for diagnosis, Beck (1967) highlighted hopelessness as often present in those with depression and it has been the focus of much subsequent research (Beck, 1967). In the present study, self-reported levels of hopeless- ness were a significant predictor of SRBs once child and par- ent reports of depression were accounted for. While this find- ing is in contrast with several studies in similar age groups (Brent et al., 1990; de Wilde et al., 1993; Greening & Stop- plebein, 2002), it is consistent with much of the literature in adults. This supports previous findings that suicide attempts in adolescents are best predicted by longer-standing trait vari- ables, including hopelessness, rather than less enduring vari- ables that are state dependent (Goldston, Reboussin, & Daniel, 2006). As a consequence, in children and adolescents, hopelessness is likely to act as a predisposing factor.

While it has been posited that hopelessness, specifically in relation to a sense of not belonging, increases the risk of SRBs (Joiner, 2002), this study showed that there were no differences between those with SRBs, and those without, in their perceptions of social support from friends and fam- ily. Perhaps, however, for children and adolescents a sense of belonging is secured via good family functioning, such that it is this aspect of a young person’s psychosocial envi- ronment that modifies the risk of SRBs in the context of a depressive disorder. This is consistent with the findings of the present study that family dysfunction was a significant predictor of SRBs once child and parent reports of depres- sion were accounted for. It is interesting to note that while the parental report of child depression symptoms was not a significant predictor of SRBs, there was an inverse rela- tionship between this dependent variable and SRBs. In the context of poorer family functioning, children and adoles- cents may feel less able to communicate their emotions to their parents, or conversely parents may be unable to rec- ognize their child’s emotional world, further impacting on the child’s sense of well-being.

Perhaps, not surprisingly, the study demonstrated that children and adolescents with a depressive disorder and no SRBs were more likely to be younger. This is consistent with literature that shows higher rates of SRBs in females after the onset of puberty (Bridge et al., 2006; Lewinsohn et al., 1996).

Overall, the results generally support the stress-diathesis model with the depressive disorder acting as the stressor, and hopelessness in the context of family discord acting as the diathesis. There are also important clinical implications from the study, in that the results of the study encourage clinicians to define depressive disorder symptom severity, particularly from the child’s perspective. A dimensional measure of depressive disorder symptom severity, over and above a careful defining of the type of depressive disorder, is critical to understanding the risk of SRBs. Understanding a child’s or adolescent’ context in terms of family function- ing is critical. High family dysfunction is an important pre- dictor of SRBs and it is suggested that parents of children and adolescents with depression and SRBs may be less aware of the emotional world and support needs of their child. These findings may help to provide direction for tar- geted interventions to address these clinical risk factors.

Strengths and Limitations

This was a cross-sectional study design, rather than longi- tudinal, therefore, risk prediction over time was not possi- ble. The study findings are in the context of a small number of participants, and it is notable that there were few females in the group who did not have suicidal behaviors. However, while it is a small study, it is of a relatively similar size to many studies in this field and it incorporates detailed as- sessments of the children and adolescents as well as their parents/caregivers. SRBs are rated from the child’s or ad- olescent’s perspective and are not objectively observed. One of the strengths of the study is that a number of psy- chosocial factors, guided by the stress-diathesis model, were tested, which also incorporated potentially protective factors. Nevertheless, the variables tested do not represent all of the significant correlates of adolescent suicide, and future studies should include a larger sample and more psy- chosocial factors to evaluate predictors in high-risk popu- lations.

Conclusion

The results of this study are consistent with the stress- diathesis model of suicide (Mann, 2003), showing that chil- dren and adolescents with depressive disorders who have a high load of symptoms, in the context of hopelessness, and against a background of family dysfunction are at greater risk of suicidal behaviors. It is critical that clinicians

S. E. Hetrick et al.: Predicting Suicidal Risk in a Cohort of Depressed Children and Adolescents 17

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

incorporate a self-rated depression severity measure into their assessment of a child or adolescent who presents with a depressive disorder, as well as ensuring an understanding of the extent to which hopelessness characterizes the child/adolescent’s cognitive style.

Acknowledgments

The authors would like to thank the Alfred and Royal Chil- dren’s Hospital Child and Adolescent Mental Health Ser- vices for their support of this study.

The study reported here was conducted as part of the DPsych research of the first author, at the Academic Child Psychiatry Unit, under the supervision of the fifth author.

References

Achenbach, T. M., & Edelbrock, C. S. (1983). Manual for the Child Behavior Checklist and Behavior Profile. Burlington: University of Vermont.

Andrews, J. A., & Lewinsohn, P. M. (1992). Suicide attempts among older adolescents: Prevalence and co-occurrence with psychiatric disorder. Journal of the American Academy of Child and Adolescent Psychiatry, 31, 655–662.

Angold, A. (1988). Childhood and adolescent depression I: Epi- demiological and etiological aspects. British Journal of Psy- chiatry, 152, 601–617.

Asarnow, J. R. (1992). Suicidal ideation and attempts during mid- dle childhood: Associations with perceived family stress and depression among child psychiatric inpatients. Journal of Clin- ical Child Psychology, 21, 35–40.

Beardslee, W. R., Keller, M. B., Seifer, R., Lavori, P. W., Staley, J., Podorefsky, D., & Shera, D. (1996). Prediction of adoles- cent affective disorder: Effects of prior parental affective dis- orders and child psychopathology. Journal of the American Academy of Child and Adolescent Psychiatry, 35, 279–288.

Beautrais, A. (1998). Psychiatric illness in a New Zealand sample of young people making serious suicide attempts. New Zea- land Medical Journal, 111, 44–48.

Beautrais, A. (2000). Risk factors for suicide and attempted sui- cide among young people. Australian and New Zealand Jour- nal of Psychiatry, 34, 420–436.

Beck, A. T. (1967). Depression: Clinical, experimental and theo- retical aspects. New York: Hober.

Brent, D. A., Kalas, R., Edelbrock, C., Costello, A. J., Dulcan, M. K., & Conover, N. (1986). Psychopathology and its rela- tionship to suicidal ideation in childhood and adolescence. Journal of the American Academy of Child and Adolescent Psychiatry, 25, 666–673.

Brent, D. A., Kolko, D. J., Allan, M. J., & Brown, R. V. (1990). Suicidality in affectively disordered adolescent inpatients. Journal of the American Academy of Child and Adolescent Psychiatry, 29, 596–593.

Bridge, J. A., Goldstein, T. R., & Brent, D. A. (2006). Adolescent suicide and suicidal behavior. Journal of Child Psychology and Psychiatry, 47, 372–394.

Cavanagh, J. T. O., Carson, A. J., Sharpe, M., & Lawrie, S. M.

(2003). Psychological autopsy studies of suicide: A systematic review. Psychological Medicine, 33, 395–405.

Coryell, W., & Young, E. A. (2005). Clinical predictors of suicide in primary major depressive disorder. Journal of Clinical Psy- chiatry, 66, 412–417.

De Leo, D., & Heller, T. S. (2004). Who are the kids who self- harm? An Australian self-report school survey. Medical Jour- nal of Australia, 181, 140–144.

de Wilde, E. J., Kienhorst, I. C. W. M., & Diekstra, R. F. W. (2001). Suicidal behaviors in adolescents. In I. Goodyer (Ed.), The depressed child and adolescent (2nd ed., pp. 267–291). Cambridge: Cambridge University Press.

de Wilde, E. J., Kienhorst, I. C. W. N., Diekstra, R. F. W., & Wol- ters, W. H. G. (1993). The specificity of psychological charac- teristics of adolescent suicide attempters. Journal of the Amer- ican Academy of Child and Adolescent Psychiatry, 32, 51–59.

de Wilde, E. J., Kienhorst, C. W. M., Diekstra, R. F. W., & Wol- ters, W. H. G. (1994). Social support, life events, and behav- ioral characteristics of psychologically distressed adolescents at high risk for attempting suicide. Adolescence, 29(113), 49–60.

Derogatis, L. R., Lipman, R. S., Rickels, K., Uhlenhuth, E. H., & Covi, L. (1974). The Hopkins Symptom Checklist (HSCL): A self-report measure inventory. Behavioral Science, 19, 1–14.

Epstein, N. B., Baldwin, L. M., & Bishop, D. S. (1983). The Mc- Master Family Assessment Device. Journal of Marital and Family Therapy, 9, 171–180.

Esposito, C. L., & Clum, G. A. (2002). Psychiatric symptoms and their relationship to suicidal ideation in a high-risk adolescent community sample. Journal of the American Academy of Child and Adolescent Psychiatry, 41, 44–51.

Giles, L. C., Glonek, G. F. V., Luszcz, M. A., & Andrews, G. R. (2005). Effects of social networks on 10-year survival in very old Australians: The Australian longitudinal study of aging. Journal of Epidemiology and Community Health, 59, 574–579.

Goldston, D. B., Reboussin, B. A., & Daniel, S. S. (2006). Predic- tors of suicide attempts: State and trait components. Journal of Abnormal Psychology, 115, 842–849.

Gould, M. S., Greenberg, T., Velting, D. M., & Shaffer, D. (2003). Youth suicide risk and preventive interventions: A review of the Past 10 years. Journal of the American Academy of Child and Adolescent Psychiatry, 42, 386–405.

Greening, L., & Stoppelbein, L. (2002). Religiosity, attributional style, and social support as psychosocial buffers for African American and White adolescents’ perceived risk for suicide. Suicide and Life Threatening Behavior, 32, 404–417.

Groholt, B., Ekeberg, O., Wichstrom, L., & Haldorsen, T. (2000). Young suicide attempters: A comparison between a clinical and epidemiological sample. Journal of the American Acade- my of Child and Adolescent Psychiatry, 39, 868–875.

Grunbaum, J. A., Kann, L., Kinchen, S., Ross, J., Hawkins, J., & Lowry, R. (2004). Youth risk behavior surveillance – United States 2003. MMWR Surveillance Summaries: Morbidity and Mortality Weekly Report Surveillance Summaries/CDC, 53(2), 1–96.

Hawton, K., Cole, D. A., O’Grady, J., & Osborn, M. (1982). Mo- tivational aspects of deliberate self-poisoning in adolescents. British Journal of Psychiatry, 141, 286–291.

Hawton, K., Rodham, K., Evans, E., & Weatherall, R. (2002).

18 S. E. Hetrick et al.: Predicting Suicidal Risk in a Cohort of Depressed Children and Adolescents

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

Deliberate self-harm in adolescents: Self-report survey in schools in England. British Medical Journal, 325, 1207–1211.

Hawton, K., & van Heeringen, K. (2009). Suicide. Lancet, 373, 1372–1381.

Hetrick, S., Vance, A., & Hall, N. (2008). In young people with a depressive disorder, does diagnostic specificity matter in the prediction of suicidality? Australian e-Journal for the Ad- vancement of Mental Health, 7(3), http://www.ause- inet.com/journal/vol7iss3/hetrick.pdf

Joiner, T. (2002). The trajectory of suicidal behavior over time. Suicide and Life-Threatening Behavior, 32, 33–41.

Joiner, T., Brown, J., & Wingate, L. (2005). The psychology and neurobiology of suicidal behavior. Annual Review of Psychol- ogy, 56, 287–314.

Jorm, A. F. (2005). Social networks and health: Its time for an intervention trial. Journal of Epidemiology and Community Health, 59, 537–538.

Kazdin, A. E., Rodgers, A., & Colbus, D. (1986). The Hopeless- ness Scale for Children: Psychometric characteristics and con- current validity. Journal of Consulting and Clinical Psychol- ogy, 54, 241–245.

Kienhorst, I. C. W. N., de Wilde, E. J., Diekstra, R. F. W., & Wol- ters, W. H. G. (1992). Differences between adolescent attempt- ers and depressed adolescents. Acta Psychiatrica Scandina- vica, 85, 222–228.

King, C. A., Raskin, A., Godwski, C. L., Butkus, M., & Opipari, L. (1990). Psychological factors associated with urban adoles- cent female suicide attempts. Journal of the American Acade- my of Child and Adolescent Psychiatry, 29, 289–294.

Kingsbury, S., Hawton, K., Steinhardt, K., & James, A. (1999). Do adolescents who take overdoses have specific psychologi- cal characteristics? A comparative study with psychiatric and community controls. Journal of the American Academy of Child and Adolescent Psychiatry, 38, 1125–1131.

Klein, D. N., Kocsis, J. H., McCullough, J. P., Holzer, C. E., Hirschfeld, R. M. A., & Keller, M. B. (1996). Symptomatology in dysthymic and major depressive disorder. The Psychiatric Clinics of North America, 19, 41–53.

Kosky, R., Silburn, S., & Zubrick, S. (1986). Symptomatic de- pression and suicidal ideation: A comparative study with 628 children. Journal of Nervous and Mental Disorders, 174, 523–528.

Kovacs, M. (1992). The Children’s Depression Inventory. North Tonawanda, NY: Mental Health Systems.

Kovacs, M., Akiskal, H., Gatsonis, C., & Phoebe, P. (1994). Childhood-onset dysthymic disorder: Clinical features and prospective naturalistic outcomes. Archives of General Psychi- atry, 51, 365–374.

Lewinsohn, P. M., Rohde, P., & Seeley, J. R. (1993). Psychosocial characteristics of adolescents with a history of suicide at- tempts. Journal of the American Academy of Child and Ado- lescent Psychiatry, 32, 60–69.

Lewinsohn, P. M., Rohde, P., & Seeley, J. R. (1996). Adolescent suicidal ideation and attempts: Prevalence, risk factors and clinical implications. Clinical Psychology: Science and Prac- tice, 3, 25–46.

Mann, J. J. (2003). Neurobiology of suicidal behavior. Nature Re- views Neuroscience, 4, 819–828.

Mazza, J. J., & Reynolds, W. M. (1998). A longitudinal investi- gation of depression, hopelessness, social support, and minor

life events and their relation to suicidal ideation in adolescents. Suicide and Life-Threatening Behavior, 28, 358–374.

Procidano, M. E., & Heller, K. (1983). Measures of perceived so- cial support from friends and family: Three validation studies. American Journal of Community Psychology, 11, 1–24.

Ryan, N. D., Puig-Antich, J., Ambrosini, P., Rabinovich, H., Rob- inson, D., Nelson, B., . . . Twomey, J. (1987). The clinical pic- ture of major depression in children and adolescents. Archives of General Psychiatry, 44(854–861).

Shaffer, D., Gould, M. S., Fisher, P., Trautman, P., Moreau, D., Kleinman, M. M. A., & Flory, M. (1996). Psychiatric diagnosis in child and adolescent suicide. Archives of General Psychia- try, 53, 338–348.

Silverman, W. K., & Albano, A. M. (1996). Anxiety Disorders In- terview Schedule for DSM-IV. Texas: Graywind.

Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate sta- tistics (3rd ed.). New York: Harper Collins.

Taylor, E., Schachar, R., Thorley, G., & Wieselberg, M. (1986). Conduct disorder & hyperactivity: I. Separation of hyperactiv- ity and antisocial conduct in British child psychiatric patients. British Journal of Psychiatry, 149, 760–767.

Voracek, M. (2004). National intelligence and suicide rate: An ecological study of 85 countries. Personality and Individual Differences, 37, 543–553.

Wasserman, D., Cheng, Q., & Jiang, G.-X. (2005). Global suicide rates among young people aged 15–19. World Psychiatry 4, 114–120.

Wechsler, D. (1991). Wechsler Intelligence Scale for Children – Third edition. San Antonio: The Psychological Corporation.

Received June 11, 2010 Revision received February 1, 2011 Accepted Feruary 10, 2011 Published online July 8, 2011

About the authors

Sarah Hetrick, DPsych, is a clinical psychologist and Senior Re- search Fellow at Orygen Youth Health Research Centre, Centre of Youth Mental Health, Melbourne University, Australia. She undertakes evidence synthesis and holds an NHMRC training fel- lowship focused on evidence implementation for youth depres- sion. She is a lead investigator on several research projects.

Alexandra Parker, PhD, MPsych, is clinical psychologist and Re- search Fellow at Orygen Youth Health Research Centre, Centre of Youth Mental Health, Melbourne University, Australia. Her research focus is on early intervention in youth mental health. She has also undertaken evidence synthesis regarding interventions across mental health disorders affecting young people.

Jo Robinson, is a Research Fellow at Orygen Youth Health Re- search Centre, Centre of Youth Mental Health, Melbourne Uni- versity, Australia, where she leads a research program specializ- ing in youth suicide prevention. She is also undertaking her PhD (University of Melbourne), examining the effects of a web-based intervention on at-risk youth.

Nicole Hall, DPsych, is a clinical psychologist and research offi-

S. E. Hetrick et al.: Predicting Suicidal Risk in a Cohort of Depressed Children and Adolescents 19

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .

cer at the Academic Child Psychiatry Unit, Royal Women’s Hos- pital, Melbourne, Australia. She has had ongoing involvement in the assessment and treatment of children and adolescents with psychiatric disorders and in research to understand these disor- ders.

Professor Alasdair Vance, MD, leads the Developmental Neuro- psychiatry Research group, Murdoch Childrens Research Insti- tute and is Head, Academic Child Psychiatry, University of Mel- bourne, Australia. He is an Honorary Academic Scholar, Depart- ment of Child and Adolescent Psychiatry, Institute of Psychiatry, University College, and Child Study Center, New York University School of Medicine, USA.

Sarah E Hetrick

Orygen Youth Health Research Centre Centre for Youth Mental Health University of Melbourne Locked Bag 10 Parkville Victoria 3052 Australia Tel. +61 3 8387 2274 Fax +61 3 8387 2466 E-mail shetrick@unimelb.edu.au

20 S. E. Hetrick et al.: Predicting Suicidal Risk in a Cohort of Depressed Children and Adolescents

T hi

s do

cu m

en t i

s co

py ri

gh te

d by

th e

A m

er ic

an P

sy ch

ol og

ic al

A ss

oc ia

tio n

or o

ne o

f i ts

a lli

ed p

ub lis

he rs

. T

hi s

ar tic

le is

in te

nd ed

s ol

el y

fo r t

he p

er so

na l u

se o

f t he

in di

vi du

al u

se r a

nd is

n ot

to b

e di

ss em

in at

ed b

ro ad

ly .