The Role of Sociodemographic, Clinical and Neuropsychological Variables in Suicide Attempts in Depressed Adolescents

Abstract

Introduction:

In this study, we aimed to evaluate the relationship between sociodemographic, clinical and neuropsychological variables and suicide attempts by comparing adolescents with Major Depressive Disorder (MDD) with and without suicide attempt.

Method:

30 adolescents with and without suicide attempt were included in this study. Sociodemographic and Clinical Data Form, Schedule for Affective Disorders and Schizophrenia for School-Age Children--Present and Lifetime Version (K-SADS-PL), The Children Depression Inventory (CDI), Beck Hopelessness Scale (BHS), Barratt Impulsivity Scale (BIS) and neuropsychological tests (Go/NoGo test, Stroop test Wisconsin Card Sorting Test) was applied to participants.

Results:

Depressed adolescents with suicide attempt compared to adolescents without suicide attempt; motor and total impulsivity scores in BIS-11, commission errors in Go/NoGo test which measure motor inhibition, completion time and errors in Stroop 5 which measure interference inhibition and scores of perseveration in Winsconsin Card Sorting Test (WCST) which measure cognitive inhibition were found to be higher. Family history of suicide attempt (OR: 5.87), commission errors (OR: 1.37), perseverative errors (OR: 1.09) ve total impulsivity (OR: 1.05) were remained in the logistic regression model.

Conclusion:

Results of this study suggested that family history of suicide attempt was the most important factor predicting suicide attempt in adolescents with depression. Other factors predicting suicide attempt were executive dysfunction and impulsivity.

INTRODUCTION

Suicide attempt is a serious and preventable public health problem. Some factors seen in adolescent suicides are; previous suicide attempts (1, 2), family history of suicide (2), hopelessness (3), depression (4) and low self-esteem (5). Depressed adolescents may experience suicidal ideation three times more than other adolescents. (6). According to the World Health Organization (WHO), although there has been an increase in options for the treatment of depression , no significant decrease in suicide rates has been observed, the increment of suicide is expected to remain in the next 15 years (7). The factors leading to suicide attempt are still needed to be studied.

To examine the association of depression, the most common psychopathology of suicide(8), with sociodemographic and clinical factors through suicide is critical in the sense of preventive mental health. According to a metaanalysis of adult depression studies, the most prominent clinical factors in suicide are; family psychopathology, history of previous attempts, depression severity and hopelessness (9). As factors in suicide attempts for depressed adolescents; family history of suicide (10), impulsivity (11), hopelessness (12), low socioeconomic status and broken home (13), were reported that should be examined while evaluating suicide risk.

Some publications suggest that dysfunctions in the prefrontal cortex, which grows rapidly during adolescence, may play a role in increasing suicide attempts in adolescence (14). One of the methods of evaluating prefrontal cortex functions is the use of neuropsychological tests that measure executive functions. A number of current research supports that executive dysfunction in depression may have an crucial role in suicidal behavior (15). Executive dysfunction in inhibition and cognitive flexibility is the focus of research on the subject. (3). Inhibition is the ability to respond more slowly and accurately by suppressing fast conditioned responses (16). It has been argued that lack of inhibition contributes to the suicide process with the inability to regulate depressed affect and the repetition of negative ruminations (17). The role of the anterior cingulate cortex and the orbitofrontal cortex in motor and interference inhibition is one of the mechanisms in which neuroanatomic suicide studies were concerned (18). In depression studies, using both stroop test (19) and stop-signal paradigm (20, 21), inhibition-suicide relationship was supported. Cognitive flexibility, the other executive functions, is the ability to adapt cognitive strategies to tackle unexpected situations. The cognitively flexible individual is aware of the possible options of the new situation, experiences positive thoughts and affections in their relationships (14). Suicidal ideators in depression was reported to have low cognitive flexibility (high perseveration) in WCST (16).

The relationship between suicide attempt and executive function was generally examined in adult samples independent of a specific diagnosis. As a result of the literature, reviewing the factors responsible for suicide attempt in adolescents with depression, together with neuropsychological variables, only one study have been reached (22). According to the results of the study, there was no significant difference between the adolescents with and without suicide attempt in terms of socioeconomic factors, family history of suicide attempts, impulsivity, interference inhibition and cognitive flexibility. (22). The limitations of the study were tought as that the groups were not matched in terms of age, gender, duration and severity of depression and that hopelessness levels were not examined. In our study, we aimed to compare sociodemographic, clinical and neuropsychological characteristics of adolescents with and without suicidal attempts and to determine which factors play prevalent role in suicide attempts.

METHOD

Participants and Procedure

Thirty adolescents aged 12-18 years, were diagnosed MDD, admitted to the emergency department of Ondokuz Mayıs University (OMU) Hospital with suicide attempt between June and December 2017, included in this study. During the same time period, 30 adolescents with MDD without suicide attempt who did not show any significant difference in terms of duration and severity of depression and were matched in terms of age and gender with suicide attempters were included. In order to get the most reliable and accurate findings, clinical evaluation, scales and neuropsychological tests were performed within the first 48 hours after suicide attempt.

Exclusion criteria for both groups; taking less than 19 points from CDI, hearing and vision problems, active medical disease that may effect test performance (Neurological, asthma, diabetes mellitus), head trauma, Central Nervous System (CNS) infection and within the last one year history of Electroconvulsive Therapy ( ECT ), according to DSM (Diagnostic and Statistical Manual of Mental Disorders)-IV, diagnosis of psychotic disorder, substance abuse or bipolar disorder. Additionally, exclusion criteria in the group without suicide attempt; history of self-injurious behavior, suicidal ideation and attempt.

Ethics committee approval was taken from the OMU Medical Faculty Clinical Research Ethics Committee and the study was conducted in accordance with the Helsinki Declaration. (B.30.2.ODM.0.20.08/1006). The third author, GG has the certificate and proficiency for neuropsychological tests.

Scales

Sociodemographic and Clinical Data Form

Its prepared by us to record sociodemographic and clinical information. In addition to sociodemographic information, includes comorbid psychiatric disorder, depression duration, psychiatric drug use, history of self-harm, history-number of suicide attempts and age of first suicide.

Schedule for Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version (KSADS–PL)

It was applied to adolescents by questioning both adolescents and one of the parents to confirm the diagnosis of MDD and to identify comorbid psychiatric diagnoses. Turkish validity and reliability study was conducted by Gökler et al. (23).

The Children Depression Inventory (CDI)

It was used to confirm the diagnosis of MDD and to determine the severity of depression in the two groups. Turkish validity and reliability study was conducted by Belma Öy (24).

Barratt Impulsivity Scale (BIS)

30-item self-report scale measuring impulsivity. Turkish validity and reliability study was made by Güleç et al. (25). It have three subscales, attentional, motor and nonplanning. The greater scores on each subscale define high-level impulsivity.

Beck Hopelessness Scale (BHS)

Self-report scale of 20 items measuring the hopelessness. There was a Turkish validity reliability study (26).

Neuropsychological Tests

Porteus Maze Test

It’s adapted by Beylan Togrol (27). Used to exclude mental retardation in participants; those who scored 90 or higher were included in the study.

Go/No Go Test

It measures motor response inhibition and is applied via computer. In the current study, two blocks were carried out with the same type of stimulus. Each block contains 100 trials (100x2). 75 ‘O’ letter stimulants and 25 ‘X’ letter stimulants were assigned. The presentation time of the stimuli is 500 ms (millisecond) and the presentation interval is 1000 ms. At the beginning of each block, instructions for participants were given and a ‘+’ sign (fixation cross) 1000ms was presented in the middle of the screen immediately after the instruction (28).

Stroop Test

The stroop test measures the ability to shift the perceptual set with different demands as well as the inhibition of habituated behavior patterns. As stated in the factor analysis, three factors obtained by Karakaş et al.; interference inhibition ( part 5), reading (part 1 and 2) and color naming (part 3 and 4) (29). In the current study, we use completion time, number of errors and self corrections in Stroop 5 to measure interference inhibition, as higher scores show more impairment in the interference inhibition.

Wisconsin Card Sorting Test (WCST)

The test have two decks of cards that contain four stimulus cards and 64 response cards. Validity and reliability study was carried out by Karakaş et al (30). As reported by the principal components analysis (30), sub-scores are collected under three factors. First factor which measures perseveration, contains WCST 1.-2.-4.-5.-6.-8.scores. When these scores rises, it may be interpereted as cognitive flexibility decreases except WCST 4 which is calculated reversely. In our study, scores in perseveration factor were examined paticularly to measure cognitive flexibility.

Statistical Analysis

Data analysis was carried out using SPSS 21.0. The level of statistical significance was accepted at p <0.05. Normal distribution was checked using Kolmogorov–Smirnov/Shapiro–Wilks tests. The student t-test, Mann-Whitney U-test, Chi-square test and Fisher’s Exact test were used to compare variables between groups. For predicting suicide attempt, logistic regression analysis was applied with the retrospective elimination method to the groups using sociodemographic, clinical and neuropsychological data.

RESULTS

Sociodemographic and Clinical Variables

There was no significant difference in the sociodemographic characteristics between suicide attempters and non-attempters (Table 1). Family history of suicide attempt and comorbid psychiatric disease were higher significantly in suicide attempters. (respectively p=0.044 and p=0.004) (Table 2). The suicide attempters had higher values in BIS motor, non-planning, and total scores (respectively p=0.002, p=0.001, p=0.006). No significant difference was found between the other clinical variables in the two groups (Table 2 and 3).

Table 1.

Sociodemographic variables

	Depressed attempters	Depressed non-attempters	Test Statistics	P
Gender
Male	2(%6.7)	2(%6.7)
Female	28(%93.3)	28(%93.3)
Age	15.8±1.2	15.5±1.2
Education
Secondary school	3(%10)	2(%6.7)
High school	20(%66.7)	24(%80.0)
Dropping-out	7(%23.3)	4(%13.3)		0.360*
Mother education
Primary school	18(%60)	16(%53.3)	Χ2=0.33	0.844
Secondary school	8(%26.7)	10(%33.3)
High school	4(%13.3)	4(%13.3)
Father education
Primary school	14(%46.7)	11(%36.7)	Χ2=1.70	0.636
Secondary school	9(%30)	8(%26.7)
High school	6(%20)	8(%26.7)
Unversity	1(%3.3)	3(%10)
Mother employment	6(%20)	3(%10)	Χ2=1.17	0.472
Father employment	24(%82.8)	26(%86.7)	Χ2=0.48	0.731
Income( monthly)
<1500TL	9(%30)	9(%30)		0.119*
1500-4500TL	21(%70)	18(%60)
>4500TL	0(%0)	3(%10)
Family type			Χ2=3.86	0.144
Extended family	18(%60)	12(%40)
Nuclear family	6(%20)	5(%16.7)
Broken home	6(%20)	13(%43.3)	Χ2=3.77	0.052
Instutional care	2(%6.7)	6(%20)		0.254*

Note. Chi-square test

^*Fisher’s Exact test TL=Turkish Lira

Table 2.

Clinical variables

	Depressed attempters	Depressed non-attempters	Test Statistics	P
Depression duration (month) (min-max)	4(1-24)	4(1-24)	z=0.81	0.413***
CDI score (min-max)	22.5(19-40)	24(19-44)	z=1.00	0.313***
Porteus Maze test score	104.47±6.2	107.27±7.1	t=1.61	0.11**
Psychotropic usage	13(%43.3)	19(%63.3)	Χ2=2.41	0.196
Comorbidity	12(%40)	23(%71)	Χ2=8.29	0.004
Anxiety disorder	2(%6.7)	7(%23.3)		0.145*
PTSD	3(%10)	10(%33.3)		0.060*
CD	1(%3.3)	0(%0)		1.000*
ADHD	6(%20)	6(%20)	Χ2=0.00	1.000*
Family psychopathology	8(%26.6)	12(%40)	Χ2=1.20	0.273
Depression	5(%16.7)	8(%26.7)	Χ2=0.88	0.347
Family history of suicide attempt	6(%20)	1(%3.3)	Χ2=4.04	0.044

Note. Statistically significant values are written in bold type.Chi-square test

^*Fisher test

^**student t-test

^***Mann-Whitney U test CDI= Children’s Depression Inventory; PTSD= Post Traumatic Stress Disorder; CD= Conduct Disorder; ADHD= Attention Deficit Hyperactivity Disorder

Table 3.

BHS and BIS-11 scores

	Depressed attempters	Depressed non-attempters	Test Statistics	P
BHS score	12.7±5	10.7±5.8	t=1.40	0.167
BIS-attentional	19.6±5.2	18.4±6.6	t=0.83	0.423
BIS-motor	25(13-39)	17.5(13-38)	z=3.07	0.002*
BIS-nonplanning	32.5±4.8	27.2±6.2	t=3.70	0.001
BIS-total	77±13.2	66±16.6	t=2.85	0.006

Note. Statistically significant values are written in bold type. Student t-test

^*Mann-Whitney U test BIS= Barratt Impulsivity Scale; BHS= Beck Hopelessness Scale

Neuropsychological Variables

On the Go/NoGo test, commission errors were higher in the suicide attempters (p<0.001). In the Stroop test, the part 5 completion time and errors were higher in the suicide attempters (respectively, p=0.037, p=0.013). In the WCST test ; total number of trials, total number of errors, perseverative responses, number and percent perseverative errors were higher in suicide attempters (p<0.001) (Table 4).

Table 4.

Neuropsychological variables

	Depressed attempters	Depressed non-attempters	Test xstatistics	P
Stroop 5 completion time	34.7±10.6	29.5±8.4	t=2.13	0.037
Stroop 5 number of errors	1.5 (0-5)	0 (0-5)	z=2.47	0.013*
Stroop 5 self-corrections	2 (0-5)	1 (0-5)	z=1.40	0.161*
Go/NoGo errors of commision	17 (5-32)	7 (4-12)	z=4.49	<0.001*
WCST total number of trials	128 (84-128)	95.5 (67-128)	z=3.94	<0.001*
WCST total number of errors	33(16-57)	17.5 (6-65)	z=3.55	<0.001*
WCST perseverative responses	22.5(8-42)	11.5 (5-40)	z=4.33	<0.001*
WCST perseverative errors	19 (8-37)	9.5 (4-35)	z=4.54	<0.001*
WCST number of categories completed	5.5(3-6)	6 (1-6)	z=1.88	0.059*
WCST percent perseverative errors	%16 (7.8-28.9)	%9.8(5.5-40)	z=3.88	<0.001*

Note. Statistically significant values are written in bold type. Student t-test

^*Mann-Whitney U test WCST= Wisconsin Card Sorting Test

Advanced Analysis of Variables for Predicting Suicide Attempts

For predicting suicide attempt, logistic regression analysis was applied with the retrospective elimination method to the groups using sociodemographic, clinical and neuropsychological data. The logistic regression model was statistically significant ( χ²(2):-8,678, p<0,001). The model explained as a whole %48.3 (CoxSnell R square) and correctly classified suicide attempts in %81.7 of the cases. Family history of suicide attempt (OR: 5.87), Go/NoGo test commission errors (OR: 1.37), WCST perseverative errors (OR: 1.09) and BIS-total score (OR: 1.05) remained significant in the model (Table 5).

Table 5.

Logistic regression analysis of socioeconomic, clinical and neuropsychological data

Variables remaining in the model	β	Standart error	Wald	OR	95%CI	P
Family history of suicide	1.771	1.193	2.202	5.876	4.778-6.834	0.045
Go/NoGo commision errors	0.319	0.101	10.054	1.376	1.130-1.676	0.002
WCST perseverative errors	0.090	0.047	3.776	1.095	0.999-1.999	0.048
BIS total score	0.056	0.025	4.845	1.058	1.006-1.112	0.028
Coefficient	-8.678	2.395	13.132			0.000

Note. OR= Odds ratio;CI= Confidence Interval BIS= Barratt Impulsivity Scale ;WCST= Wisconsin Card Sorting Test

DISCUSSION

In the light of our findings; among the sociodemographic and clinical factors of adolescent with suicide attempt; psychiatric comorbidity, family history of suicidal attempt and impulsivity were found to be significantly higher, in propitiously with the literature. (10, 31, 32). Family history of suicide attempt (OR: 5.87) and impulsivity (OR: 1.05) remained significant in the regression model. Conjointly, for studies in which the factors related to the parents were controlled, it was determined that family history of suicide attempt predicted the suicide attempt in the adolescent, so it supported that the main factor is genetic burden (33). İncrement of motor impulsivity in suicide attempters is in line with the literature (34). This finding may support that adolescents with suicide attempt are individuals who react quickly, act without thinking and want their wishes to be happen immediately. Contrary to our hypothesis, the two groups did not differ significantly in terms of hopelessness. In studies where the sample was large and consisted of adolescents with moderate depression resistant to treatment, the hopelessness of the suicide attempters was reported to be higher (35). Besides small sample size and low severity of depression, another factor that might be effective in the fact that the levels of hopelessness didn’t differ significantly between the two groups in the current study was that Post Traumatic Stress Disorder (PTSD) rates were higher in non-attempters (altough, not statistically significant). As stated in the diagnostic criteria, PTSD is a psychiatric disorder that presents with constant and exaggerated negative beliefs and expectations with regard to one’s ownself, others, or the world; this may explain the proximity of the hopelessness of the two groups.

The main finding of our study was that adolescents with suicide attempts have lower executive functions. First and foremost of our data in this topic, Go/NoGo test commission errors was higher in suicide attempters,in other words, motor response inhibition ability was lower. In the go/no go test, it’s instructed to press the button when the stimulus seen on the screen and not to press it when the stimulus is absent. If the button is pressed when there is no stimulus on the screen, a commission error is made. The commission error shows the person’s responses without adequately evaluating the stimulus. More errors indicate that there is a problem in inhibiting the automatic motor response (28). The significance of motor response inhibition continued in regression model (OR:1.37). This finding is in line with the literature, depressed adolescents with suicide attempt may be considered to have trouble stopping their suicidal-themed negative thoughts once triggered (20, 21). As reported by our study findings, suicide attempters’ stroop test performances associated with interference inhibition (Stroop 5) were also low. Resulting from interference inhibition dysfunction, useless-distractive negative information cannot be inhibited, making the appropriate choices becomes difficult so, individuals happen to be vulnerable to suicide (17). Additionally, another executive dysfunction detected in our study, is impairment of cognitive flexibility. Cognitively flexible individuals can evaluate alternative behavioral options and find solutions in difficult situations (36). Thus, the probability of choosing suicide may decrease compared to those who are cognitively inflexible (36). Our findings regarding both cognitive flexibility and interference inhibition; compatible with literature data in adult sample (19, 37-39). However, in the regression analysis it was determined that the number of perseverative errors that measure cognitive flexibility increased the risk of suicide attempts at a low level (OR: 1.09) and interference-related scores of the stroop test could not remain in the model. This findings support metaanalysis findings showing that the effect of impaired cognitive flexibility and interference inhibition on suicide attempt in mood disorders is low (40). On the other hand, it was thought that the low sample size could be responsible for the paucity of significance of the data.

Our study is one of the rare studies in the literature on the adolescent sample related to this topic. Other strengths of our study; matching groups in terms of various features (age-gender, duration and severity of depression) and using gold standard neuropsycological tests ( Go/NoGo test, stroop test, WCST). Small sample size, the most important limitation, may have prevented some variables from remaining significant in regression analysis. The absence of a healthy control group also prevented to examine the mediating role of depression on results. For ethical reasons, it was not possible to follow the sample without medication before neuropsychological tests. The cross-sectional evaluation of variables also prevented us from interpreting the permanence of the measured features.

Results of this study suggested that family history of suicide attempt was the most important factor predicting suicide attempt in depressed adolescents. Executive dysfunctions have been found to effect the risk of suicide attempts to a small extent. It is recommended to plan studies with larger samples related to the subject, in which neuropsychological tests and neuroimaging methods are used together.