Академический Документы
Профессиональный Документы
Культура Документы
net/publication/256478610
CITATIONS READS
126 1,056
3 authors:
Fabrice Jollant
Paris Descartes, CPSC
145 PUBLICATIONS 3,959 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Cognitive and neuroimaging alterations in elderly suicide attempters (SPAD and SPAD-IRM): A Nîmes study View project
All content following this page was uploaded by Marcelo Berlim on 07 March 2014.
Background. Suicidal behavior results from a complex interplay between stressful events and vulnerability factors,
including cognitive deficits. However, it is not clear which cognitive tests may best reveal this vulnerability. The objective
was to identify neuropsychological tests of vulnerability to suicidal acts in patients with mood disorders.
Method. A search was made of Medline, EMBASE and PsycINFO databases, and article references. A total of 25 studies
(2323 participants) met the selection criteria. A total of seven neuropsychological tests [Iowa gambling task (IGT),
Stroop test, trail making test part B, Wisconsin card sorting test, category and semantic verbal fluencies, and continuous
performance test] were used in at least three studies to be analysed.
Results. IGT and category verbal fluency performances were lower in suicide attempters than in patient controls
[respectively, g = –0.47, 95% confidence interval (CI) –0.65 to –0.29 and g = –0.32, 95% CI –0.60 to –0.04] and healthy con-
trols, with no difference between the last two groups. Stroop performance was lower in suicide attempters than in patient
controls (g = 0.37, 95% CI 0.10–0.63) and healthy controls, with patient controls scoring lower than healthy controls.
The four other tests were altered in both patient groups versus healthy controls but did not differ between patient groups.
Conclusions. Deficits in decision-making, category verbal fluency and the Stroop interference test were associated with
histories of suicidal behavior in patients with mood disorders. Altered value-based and cognitive control processes may
be important factors of suicidal vulnerability. These tests may also have the potential of guiding therapeutic interventions
and becoming part of future systematic assessment of suicide risk.
Key words: Cognitive control, decision making, mood disorders, neuropsychology, suicidal behaviour, vulnerability.
(McGirr et al. 2013). Apart from improving our under- (3) compared at least two groups of which one com-
standing of the suicidal vulnerability, these cognitive prised patients with a history of suicide attempt
deficits could be both measurable markers of vulner- (defined as any act carried out with a certain intent
ability and the target of future therapeutic interventions to die and different from non-suicidal self-injury;
aimed at reducing the long-term risk of suicidal acts. Mann, 2003); and (4) included patients who suffered
There is a need to synthesize evidence from cumu- from mood disorders (both unipolar or bipolar) but
lating literature on cognitive deficits in suicide attemp- not schizophrenia or other major psychiatric disorder
ters in order to determine which neuropsychological diagnosed according to the Diagnostic and Statistical
tests reveal impairments more strongly associated with Manual of Mental Disorders, fourth edition, text revi-
a history of suicidal act, i.e. a higher risk of recurrence sion (DSM-IV-TR) criteria. We focused on mood dis-
of suicidal acts and completed suicide. It is notably orders as more studies have been published in this
important to disentangle the cognitive deficits associ- domain and to limit heterogeneity in studied popu-
ated with suicidal behavior from those more closely lations. Full articles were then obtained for final
related to co-morbid disorders such as major de- analyses.
pression. In the current paper, we systematically re- Of the 1830 originally identified abstracts, 25 studies
viewed the literature on the neuropsychology of met the inclusion criteria (Bartfai et al. 1990; Ellis et al.
suicidal behavior, and conducted a meta-analysis 1992; Becker et al. 1999; King et al. 2000; Keilp et al.
to explore putative cognitive markers of suicidal 2001, 2013; Audenaert et al. 2002; Jollant et al. 2005,
vulnerability. 2010; LeGris et al., 2012; Raust et al. 2007; Westheide
et al. 2008; Yen et al. 2008; Malloy-Diniz et al. 2009;
Oldershaw et al. 2009; Cha et al. 2010; Martino et al.
Method
2010; Gilbert et al. 2011; Richard-Devantoy et al. 2011,
Data sources 2012, 2013a; Bridge et al. 2012; McGirr et al. 2012;
Miranda et al. 2012; Gorlyn et al. 2013). Although
An English and French systematic literature search of
eligible, three studies were not included, or only
Medline, EMBASE and PsycINFO databases was per-
partially (for instance, some tests but not others),
formed for human studies published from 1 January
because precise means and standard deviations were
1960 to 31 March 2013. The medical subject heading
not available in papers and could not be obtained
(MeSH) term ‘suicide’ was combined with the MeSH
after contacting the authors (Williams & Broadbent,
terms ‘cognition’, ‘neuropsychology’, ‘neuropsycho-
1986; Dombrovski et al. 2008; Gilbert et al. 2011).
logical tests’, ‘executive function’, ‘decision making’,
Finally, one study which assessed patients with border-
‘problem solving’, ‘prefrontal cortex’, and with the
line personality disorder was also included as levels of
title/abstract (TIAB) terms ‘neuropsychological func-
depression and rates of co-morbid mood disorders
tions’, ‘executive functioning’ and ‘executive perform-
were high (LeGris et al. 2012).
ance’. An iterative process was used to ensure that all
The quality of each study was assessed indepen-
relevant articles were obtained. A further hand search
dently by two reviewers (S.R.-D. and F.J.) using the
of the bibliographical references of the selected papers
Crombie criteria adapted by Petticrew & Roberts
and existing reviews was conducted to identify
(2006).
additional potential studies. References were also
selected from our research group’s online database
(www.bdsuicide.disten.com). Data extraction and analyses
A standardized form was used to extract data, which
Study selection
included authors, date of publication, study design,
Abstract selection was based on the STrengthening the settings, study population, cognitive tests used, defi-
Reporting of OBservational studies in Epidemiology nition of suicidal behavior, and neuropsychological
(STROBE) checklist (Von Elm et al. 2008) which de- scores (means and standard deviations).
scribes items that should be included in reports of Only tasks that were used in at least three separate
cohort studies. Abstracts identified through the litera- studies were included in the analyses. Overall, seven
ture search were independently evaluated by two neuropsychological tests were analysed as they met
reviewers (S.R.-D. and F.J.) and selected by a consensus this criterion: (1) Iowa gambling task (IGT; net score
from all authors. = number of advantageous minus disadvantageous
Studies that met the following inclusion criteria were choices); (2) Stroop test (Stroop; interference score,
included in this meta-analysis: (1) published in an i.e. time to read the color–word interference sheet
English or French language peer-reviewed journal; minus time to name the color of each block sheet);
(2) included at least one neuropsychological task; (3) trail-making test part B (TMTB; time completion);
Neuropsychological markers of vulnerability to suicidal behavior 3
(4) WCST (perseverative errors); (5) FAS semantic 668 healthy controls (mean age 39.2 years, S.D. = 9.5
verbal fluency test (number of words); (6) categorical years; 50.8% males). In 18 studies, patients were
verbal fluency test (Animals; number of words); and under a psychotropic medication; in four studies,
(7) continuous performance test (CPT; commission patients did not have any medication (Keilp et al.
errors). 2001, 2013; Jollant et al. 2005; Gorlyn et al. 2013), and
Analyses were performed using Comprehensive data were not available for three studies even after con-
Meta-Analyses version 2.0 (Biostat, USA), and IBM tacting the authors (Bartfai et al. 1990; Ellis et al. 1992;
SPSS version 20 (IBM Corporation, USA) software. Becker et al. 1999).
Three groups were compared: suicide attempters Table 2 presents the results of the contrasts between
(patients with a history of suicide attempt), patient the three groups for the seven neuropsychological
controls (i.e. patients with no personal history of tests, and Table 3 provides a summary of the main
suicidal act) and healthy controls. When two groups findings. Detailed information on heterogeneity and
of suicide attempters were reported in one study publication bias can be found in the online Sup-
[e.g. low versus high lethality (McGirr et al. 2012, plementary material.
Keilp et al. 2013) or violent versus non-violent (Jollant
et al. 2005)], the combined means and standard devi- Suicide attempters versus patient controls
ations were calculated to obtain a global group, using
Suicide attempters had significantly lower IGT net
the following formula:
scores and Animals scores, and lower Stroop perform-
NX μX + NY μY ance than patient controls (Fig. 1), all representing
μX<Y = , moderate effect sizes. The fail-safe N, i.e. the number
NX + NY
of unpublished or missing null findings that would
be needed to render the results non-significant, was
NX σ 2X + NY σ 2Y NX NY 60 for the IGT, five for the Animals and 29 for the
σ X<Y = + + (μX − μY )2 .
NX + NY (NX + NY )2 Stroop. Mean age, gender, depression level, and the
proportion of unipolar/bipolar disorders did not differ
We performed a meta-analysis of aggregate data between the two groups for any test, thus ruling out
and used a random-effects model as we assumed these variables as confounding factors. Regarding
that the true effect sizes had probably varied between Stroop, results were significant whether assessing the
the included studies (Riley et al. 2011). Pooled Hedges’ traditional version only or including the emotional
g effect sizes for subjects’ neuropsychological scores version.
and depression ratings were computed (Hedges & Heterogeneity exceeded that expected by chance at
Olkin, 1985). Qualitative descriptors of the obtained p < 0.05 only for the Stroop and TMTB, implying that
effect sizes are usually considered small if <0.3, moder- the variance among the effect sizes was greater than
ate if between 0.4–0.8, and large if >0.8 (Egger et al. expected by sampling error (online Supplementary
2001). Table S1). The study by Richard-Devantoy et al.
Heterogeneity was assessed using the Q statistics (2011) was probably responsible for the heterogeneity
and the I2 index (Cooper et al. 2009). Values of p < 0.10 related to the Stroop, and the studies by Ellis et al.
for the former and >35% for the latter were deemed (1992) and Yen et al. (2008) for the TMTB. After exclud-
as indicative of study heterogeneity. Finally, we used ing these studies, the heterogeneity disappeared and
funnel plots, Rosenthal’s fail-safe N (Rosenthal, 1979) the main results remained significant.
and Egger’s regression intercept (Egger et al. 1997) to The funnel plots were reasonably symmetrical for
test for the presence of publication bias (Cooper et al. all neuropsychological tests except for the CPT, sug-
2009). gesting a low risk of publication bias. Nevertheless,
the more conservative Egger’s regression intercept
suggested no publication bias.
Results
Suicide attempters versus healthy controls
A total of 25 studies were included (Table 1) comprising
2323 participants, of whom 831 were suicide attempters Suicide attempters had significantly lower perform-
[mean age = 40.5 years, S.D. = 9.4 years; 43.6% males; ance than healthy controls on all seven neuropsycho-
75.4%, 95% confidence interval (CI) 15–100% unipolar logical tasks, all with moderate to high effect sizes.
disorder; 52.4%, 95% CI 15–100% bipolar disorder], The fail-safe N was 100 for the IGT, 96 for the
824 patient controls (mean age = 41.6 years, S.D. = 9.4 Stroop, 26 for the TMTB, 27 for the WCST, 44 for the
years; 41.0% males; 77.2%, 95% CI 3–100% unipolar dis- FAS, 36 for Animals, and seven for the CPT. Mean
order; 54.9%, 95% CI 17–100% bipolar disorder], and age and gender did not differ between the two groups,
4 S. Richard-Devantoy et al.
Table 1. Studies included in the meta-analysis
Mean Mean
Age, years Males, depression UP, BP, Age, years Males, depression UP, BP, Mean age, Males, Neuropsychological
Study n (S.D.) % score (S.D.)a M % % n (S.D.) % score (S.D.)a M % % n years (S.D.) % tests
Audenaert et al. (2002) 10 31.9 (10.9) 50 27.8 (3.0)b Y 100 0 10 21.8 (2.0) 40 FAS, Animals
Bartfai et al. (1990) 9 33.0 100 N.A. N.A. 22.2 22.2 7 30.0 100 N.A. N .A . N .A . N.A. 8 28.0 100 WCST
Becker et al. (1999) 31 38.4 (16.2) 35.4 20.8 (11.9)c N.A. N.A. N.A. 31 38.2 (14.6) 35.4 10.6 (8.6)c N .A . N .A . N.A. Stroop
Bridge et al. (2012) 40 15.5 (1.4) 25 12.8 (8.3)d Y 75 40 15.6 (1.4) 25 13.5 (13.3)d 40 IGT
Cha et al. (2010) 68 34.1 (10.5) 57.4 N.A. Y 85.3 56 35.2 (13.2) 64.8 N.A. Y 62.5 Stroop
Ellis et al. (1992) 20 30.4 30 17.3 (13.3)c N.A. 55 N.A. 27 40.0 26 17.4 (13.0)c N .A . 85 33 TMT, WCST
Gilbert et al. (2011) 28 43.7 (10.7) 50 8.6 (7.5)b Y 0 100 39 41.1 (12) 59 10.7 (6.9)b Y 0 100 IGT, Stroop (N/A),
TMTB (N/A),
WCST (N/A), FAS
(N/A), Animals (N/A)
Gorlyn et al. (2013) 26 35.0 (11.9) 65 25.8 (6.84)b N 69.4 30.6 46 37.7 (10.7) 32.6 25.6 (7.2)b N 78.8 21.3 42 32.3 (10.6) 61.9 IGT
Jollant et al. (2005) 69 43.0 (11.8) 49 N.A. Y 70.5 26.1 25 40.4 (12.2) 35 N.A. Y 72 28 82 38.8 (9.1) 73.2 IGT
Jollant et al. (2010) 13 40.3 (11.3) 100 2.4 (2.3)b N 100 0 12 43.6 (10.7) 100 2.8 (2.3)b N 100 0 15 32.6 (10.1) 100 IGT
Keilp et al. (2001) 29 38.5 (10.6) 37.8 29.3 (5.2)b N 73.3 0 21 41 (10.9) 57.1 27.5 (6.4)b N 81 0 22 41.2 (16.8) 63.6 Stroop, TMTB, WCST,
FAS, Animals, CPT
Keilp et al. (2013) 72 35.7 (11.6) 36 25.7 (7.3)b N 50 22 80 40.1 (11.9) 47.5 25.6 (7.4)b N 63 17 56 31.5 (11.1) 50 Stroop, TMTB, WCST,
FAS, Animals, CPT
King et al. (2000) 18 66.7 (10.1) 44 28.1 (8.36)b Y 100 0 29 64.2 (10.9) 31 29.5 (8.6)b Y 100 0 30 N.A. N.A. TMT, WCST, FAS
LeGris et al. (2012) 13 32.2 (10.5) 0 N.A. Y – – 13 32.2 (10.5) 0 N/A Y – – 41 31.2 (9.0) 0 IGT, Stroop
Malloy-Diniz et al. 18 41 (13.8) 33.3 9.6 (5.1)c Y 0 100 21 40.8 (12.6) 47.6 8.0 (4.3)c Y 0 100 50 36.9 (9.8) 33.3 IGT, Stroop, WCST, CPT
(2009)
Martino et al. (2010) 22 42.1 (10.6) 21.5 2.4 (1.9)b Y 0 100 63 39.2 (10.8) 39.4 2.0 (2.0)b Y 0 100 34 40 (12.9) 35.3 IGT (N/A), TMTB,
WCST, FAS
McGirr et al. (2012) 34 67.8 (7.84) 50 22.3 (5.2)b Y 100 0 29 70.3 (9.0) 34.4 19.85 (3.6)b Y 100 0 30 69.7 (6.8) 53.3 WCST
Miranda et al. (2012) 13 18.3 (0.6) 8 14.3 (9.1)c Y 15 15 32 18.31 (0.8) 28 14.12 (9.0)c Y 3 0 WCST
Oldershaw et al. (2009) 54 15.8 (1.5) 9.3 N.A. Y 100 0 22 15.7 (1.3) 9.1 N.A. Y 100 0 57 15.8 (1.5) 19.3 IGT
Raust et al. (2007) 30 39.8 (13.6) 40 3.6 (2.8)e Y 50 27 39 44.6 (12.4) 54 Stroop
Richard-Devantoy et al. 10 75.3 (2.3) 30 28.1 (0.6)b Y 100 0 10 72.9 (1.3) 30 27.7 (2)b Y 100 0 Stroop
(2011)
Neuropsychological markers of vulnerability to suicidal behavior 5
except for the TMTB with more men among the healthy
controls compared with the suicide attempters.
Heterogeneity exceeded that expected by chance for
S.D., Standard deviation; M, medication; UP, unipolar disorder; BP, bipolar disorder; Y, yes; FAS, FAS verbal fluency test; Animals, animal verbal fluency test; N.A., data not
all tests except for the FAS and Animals. No specific
FAS, Animals
Stroop, TMTB,
FAS, Animals
available; WCST, Wisconsin card sorting test; IGT, Iowa gambling task; TMT, trail-making test; TMTB, trail-making test part B; CPT, continuous performance test.
(2012) and Malloy-Diniz et al. (2009) were probably
responsible for the heterogeneity related to the
35.1
N.A.
40
39 (10.9)
37
29
100
Y
2.97 (2.58)b
28.3 (2.4)b
Depression scores come from different scales and, therefore, cannot be directly compared across studies:
29
41.4 (12.9)
75.9 (6.7)
33.3 31
87
100
Y
Y
cant anymore.
Heterogeneity exceeded that expected by chance for
27.7 (3.5)b
4.6 (2.6)b
66.7
N.A.
45
34.8 (8.3)
35
29
Richard-Devantoy et al.
e
c
1.7 (*)
1.9 (*)
−5.0***
2.1**
2.8**
CI, Confidence interval; IGT, Iowa gambling task; Animals, animal verbal fluency test; TMTB, trail-making test part B; FAS, FAS verbal fluency test; WCST, Wisconsin card sorting
−1.6
−1.8
the TMTB.
Za
Discussion
2.2**
2.9**
2.3*
Za
−0.8
−1.0
0.1
1.7
Za
WCST
TMTB
CPT
FAS
IGT
IGT, Iowa gambling task; Animals, animal verbal fluency test; TMTB, trail-making test part B; FAS, FAS verbal fluency test;
WCST, Wisconsin card sorting test; CPT, continuous performance test.
* p < 0.05, ** p < 0.01, *** p < 0.001, (*) contrast became significant after excluding studies responsible for heterogeneity.
to suicidal behavior (Mann, 2003), with patients at signals of rejection, which, when they happen, lead
higher risk of suicide showing alterations not found to an intense negative state. Their difficulty to control
in patients at lower risk. this response (possibly associated with a ruminative
At the neurocognitive level, we could hypothesize mode of thinking; O‘Connor & Noyce, 2008) and
that vulnerability to suicidal acts results from a com- with a pre-existing difficulty to envision the long-term
bination of alterations in value-based/motivational/ consequences of some options, may limit the extent
reward-learning processes (supporting decision mak- of their choices and lead them to consider suicide as
ing as measured by the IGT) on one side, and in cogni- the only possible way to escape this painful state. An
tive control processes (as measured by Animals additional hypothesis could be that limited verbal abil-
and Stroop) on the other side (Jollant et al. 2011). ities (as revealed by the animal verbal fluency test)
Cognitive control refers to mechanisms that ‘orches- may reduce the possibilities to solve problems at an
trate thought and action in accordance with internal explicit level and, consequently, increase the risk of
goals’ (Miller & Cohen, 2001) and, therefore, encom- ‘acting out’ negative emotions and suicidal ideas.
pass multiple functions from task switching, response In addition, disadvantageous decision-making may
inhibition, error detection, response conflict and work- also increase the risk of interpersonal difficulties, a
ing memory (Glascher et al. 2012). This schematic dis- classical trigger of suicidal crisis (Jollant et al. 2007).
tinction is generally supported by neuroimaging and This proposed model has yet to be properly tested in
lesion studies (Kouneiher et al. 2009; Stuss, 2011; empirical studies. Of note, this model (and the results
Glascher et al. 2012). In the case of suicide attempters, of this meta-analysis) partly overlaps with Mark
disadvantageous IGT performance has been related Williams’s ‘cry of pain’ cognitive model (Williams &
to deficient encoding of abstract risk in the ventro- Pollock, 2001). This model proposed that vulnerable
lateral orbitofrontal cortex (Jollant et al. 2010), a region people show: (1) a higher sensitivity to signals of
that was also more responsive to social signals of rejec- defeat as revealed by the emotional version of the
tion in this population compared with patient controls Stroop; (2) the feeling to be trapped associated with
(Jollant et al. 2008). However, decreased verbal fluency lower problem-solving abilities and too general auto-
has been linked to more dorsal prefrontal regions, biographic memory; and (3) the feeling of hopelessness
including the anterior cingulate and the dorsolateral correlated with reduced verbal fluency (using a modifi-
prefrontal cortices (Audenaert et al. 2002; Oquendo ed version). The link with the ‘escape from self’ theory
et al. 2003). These latter brain regions also underlie of suicide is less obvious (Baumeister, 1990).
brain processing during Stroop (Alvarez & Emory, The two other implications of our findings are
2006), although specific investigations of suicide more speculative at this stage. The IGT, Animals and
attempters are lacking. Stroop could be helpful tests in the future for assessing
A clinical translation of this neurocognitive model the long-term suicidal risk of patients with mood dis-
could be that vulnerable individuals are more likely orders. These neuropsychological tests could be part
to strongly value particular life events, notably social of a future comprehensive assessment of patients
8 S. Richard-Devantoy et al.
(a)
Study name Statistics for each study Sample size Hedges's g and 95% CI
Hedges's Standard Lower Upper Suicide Patient Relative
g error Variance limit limit Z-Value p-Value Attempters Controls weight
Jollant et al, 2005 -0.725 0.237 0.056 -1.191 -0.260 -3.054 0.002 69 25 13.81
Malloy-Diniz et al, 2009 -0.844 0.329 0.108 -1.489 -0.200 -2.567 0.010 18 21 7.49
Oldershaw et al, 2009 -0.107 0.251 0.063 -0.598 0.384 -0.428 0.669 54 22 12.51
Jollant et al, 2010 -0.927 0.409 0.167 -1.729 -0.126 -2.269 0.023 13 12 4.93
Martino et al, 2010 -0.520 0.249 0.062 -1.008 -0.033 -2.093 0.036 22 63 12.69
Gilbert et al, 2011 -0.431 0.248 0.061 -0.916 0.054 -1.741 0.082 28 39 12.78
Bridge et al, 2012 -0.600 0.226 0.051 -1.044 -0.156 -2.650 0.008 40 40 15.06
Legris et al, 2012 -0.326 0.329 0.109 -0.971 0.320 -0.989 0.323 29 13 7.47
Gorlyn et al, 2013 -0.113 0.243 0.059 -0.589 0.364 -0.463 0.643 26 46 13.25
-0.473 0.092 0.008 -0.654 -0.293 -5.142 0.000 299 281
Results indicate that suicide attempters made significantly more risky than safe choices relative to patient controls.
(b)
Study name Statistics for each study Sample size Hedges's g and 95% CI
Keilp et al, 2001 -0.400 0.285 0.081 -0.958 0.159 -1.403 0.161 29 21 19.39
Richard-Devantoy et al, 2012 -0.567 0.316 0.100 -1.187 0.053 -1.791 0.073 20 20 16.44
Keilp et al, 2013 -0.047 0.162 0.026 -0.364 0.270 -0.290 0.772 72 80 40.41
Richard-Devantoy et al, 2013a -0.550 0.248 0.062 -1.036 -0.063 -2.213 0.027 35 31 23.76
-0.320 0.143 0.020 -0.601 -0.040 -2.236 0.025 156 152
Results indicate that suicide attempters named significantly fewer animals in 1 min than patient controls.
(c)
Study name Statistics for each study Sample size Hedges's g and 95% CI
Hedges's Standard Lower Upper Suicide Patient Relative
g error Variance limit limit Z-Value p-Value Attempters Controls weight
Becker et al, 1999 0.163 0.251 0.063 -0.329 0.656 0.649 0.516 31 31 13.80
Keilp et al, 2001 0.430 0.285 0.081 -0.129 0.989 1.507 0.132 29 21 12.12
Malloy-Diniz et al. 2009 0.282 0.316 0.100 -0.338 0.902 0.891 0.373 18 21 10.78
Cha et al, 2010 0.099 0.179 0.032 -0.253 0.451 0.551 0.582 68 56 18.04
Richard-Devantoy et al, 2011 1.955 0.528 0.279 0.920 2.990 3.701 0.000 10 10 5.22
Legris et al, 2012 0.016 0.327 0.107 -0.626 0.658 0.048 0.962 29 13 10.33
Richard-Devantoy et al, 2012 0.735 0.321 0.103 0.106 1.363 2.291 0.022 20 20 10.60
Keilp et al, 2013 0.354 0.163 0.027 0.035 0.673 2.173 0.030 72 80 19.11
0.372 0.135 0.018 0.108 0.636 2.762 0.006 277 252
Results indicate that suicide attempters took significantly more time than patient controls in conditions of interference.
Fig. 1. Comparison of the Iowa gambling task net scores (a), animal verbal fluency scores (b) and Stroop interference scores
(c) between suicide attempters and patient controls. CI, Confidence interval.
with mood disorders. However, longitudinal studies or with a unipolar depressive disorder (Westheide
will be first necessary to measure the predictive et al. 2008), or a combination of patients with both sub-
value of these cognitive deficits versus the simple col- types of mood disorder (Jollant et al. 2005). Also, some
lection of clinical signs and symptoms and past history. studies were conducted in patients who were acutely
Finally, interventions aiming at remediating cognitive depressed (Abrahams et al. 2009; Richard-Devantoy
deficits revealed by these tests should be developed et al. 2012), while others focused on those in remission
and their overall clinical utility assessed. (Jollant et al. 2005). Also, participants in some studies
were on medication while in other studies they were
not. Moreover, it was not possible to separate neuro-
Limitations
psychological targets of low from high lethality suicide
First, studies included in this review examined various attempters, and violent and non-violent suicide
populations. For example, some enrolled only adoles- attempters, because too few studies distinguished
cents (Bridge et al. 2012), middle-aged (Keilp et al. both groups. Personality disorders were not formally
2013) or elderly (Richard-Devantoy et al. 2012) partici- assessed or reported in too many studies to be taken
pants. In addition, some studies included only patients in to account in analyses. Finally, some studies were
with bipolar depression (Malloy-Diniz et al. 2009), only conducted in males (Jollant et al. 2010) while
Neuropsychological markers of vulnerability to suicidal behavior 9
cognitive control and value-based decision making in the borderline personality disorder. Psychiatry Research 196,
prefrontal cortex. Proceedings of the National Academy of 101–108.
Sciences USA 109, 14681–14686. LeGris J, van Reekum R (2006). The neuropsychological
Gorlyn M, Keilp JG, Oquendo MA, Burke AK, Mann JJ correlates of borderline personality disorder and suicidal
(2013). Iowa Gambling Task performance in currently behaviour. Canadian Journal of Psychiatry 51, 131–142.
depressed suicide attempters. Psychiatry Research 207, MacLeod CM (1991). Half a century of research on the
150–157. Stroop effect: an integrative review. Psychological Bulletin
Guillaume S, Perroud N, Jollant F, Jaussent I, Olié E, 109, 163–203.
Malafosse A, Courtet P (2012). HPA axis genes may Malloy-Diniz LF, Neves FS, Abrantes SS, Fuentes D,
modulate the effect of childhood adversities on Correa H (2009). Suicide behavior and neuropsychological
decision-making in suicide attempters. Journal of Psychiatric assessment of type I bipolar patients. Journal of Affective
Research 47, 259–265. Disorders 112, 231–236.
Harrison JE, Buxton P, Husain M, Wise R (2000). Short test of Mann JJ (2003). Neurobiology of suicidal behaviour.
semantic and phonological fluency: normal performance, Nature Reviews Neuroscience 4, 819–828.
validity and test–retest reliability. British Journal of Clinical Martino DJ, Strejilevich SA, Torralva T, Manes F (2010).
Psychology 39, 181–191. Decision making in euthymic bipolar I and bipolar II
Hedges LV, Olkin I (1985). Statistical Methods for disorders. Psychological Medicine 41, 1319–1327.
Meta-Analysis. Academic Press: London. McGirr A, Dombrovski AY, Butters MA, Clark L, Szanto K
Jollant F, Bellivier F, Leboyer M, Astruc B, Torres S, (2012). Deterministic learning and attempted suicide among
Verdier R, Castelnau D, Malafosse A, Courtet P (2005). older depressed individuals: cognitive assessment using the
Impaired decision making in suicide attempters. American Wisconsin Card Sorting Task. Journal of Psychiatric Research
Journal of Psychiatry 162, 304–310. 46, 226–232.
Jollant F, Guillaume S, Jaussent I, Castelnau D, McGirr A, Jollant F, Turecki G (2013). Neurocognitive
Malafosse A, Courtet P (2007). Impaired decision-making alterations in first degree relatives of suicide completers.
in suicide attempters may increase the risk of problems in Journal of Affective Disorders 145, 264–269.
affective relationships. Journal Affective Disorders 99, 59–62. Miller EK, Cohen JD (2001). An integrative theory of
Jollant F, Lawrence NL, Olié E, Guillaume S, Courtet P prefrontal cortex function. Annual Review of Neuroscience
(2011). The suicidal mind and brain: a review of 24, 167–202.
neuropsychological and neuroimaging studies. World Miranda R, Gallagher M, Bauchner B, Vaysman R,
Journal of Biological Psychiatry 12, 319–339. Marroquin B (2012). Cognitive inflexibility as a prospective
Jollant F, Lawrence NS, Giampietro V, Brammer MJ, predictor of suicidal ideation among young adults with a
Fullana MA, Drapier D, Courtet P, Phillips ML (2008). suicide attempt history. Depression and Anxiety 29, 180–186.
Orbitofrontal cortex response to angry faces in men with O’Connor RC, Noyce R (2008). Personality and cognitive
histories of suicide attempts. American Journal of Psychiatry processes: self-criticism and different types of rumination
165, 740–748. as predictors of suicidal ideation. Behaviour Research and
Jollant F, Lawrence NS, Olié E, O’Daly O, Malafosse A, Therapy 46, 392–401.
Courtet P, Phillips ML (2010). Decreased activation of Oldershaw A, Grima E, Jollant F, Richards C, Simic M,
lateral orbitofrontal cortex during risky choices under Taylor L, Schmidt U (2009). Decision making and problem
uncertainty is associated with disadvantageous solving in adolescents who deliberately self-harm.
decision-making and suicidal behavior. Neuroimage 51, Psychological Medicine 39, 95–104.
1275–1281. Oquendo MA, Galfalvy H, Russo S, Ellis SP,
Keilp JG, Gorlyn M, Russell M, Oquendo MA, Burke AK, Grunebaum MF, Burke A, Mann JJ (2004). Prospective
Harkavy-Friedman J, Mann JJ (2013). Neuropsychological study of clinical predictors of suicidal acts after a major
function and suicidal behavior: attention control, memory depressive episode in patients with major depressive
and executive dysfunction in suicide attempt. Psychological disorder or bipolar disorder. American Journal of Psychiatry
Medicine 43, 539–551. 161, 1433–1441.
Keilp JG, Sackeim HA, Brodsky BS, Oquendo MA, Oquendo MA, Placidi GP, Malone KM, Campbell C,
Malone KM, Mann JJ (2001). Neuropsychological Keilp J, Brodsky B, Kegeles LS, Cooper TB, Parsey RV,
dysfunction in depressed suicide attempters. American Van Heertum RL, Mann JJ (2003). Positron emission
Journal of Psychiatry 158, 735–741. tomography of regional brain metabolic responses to a
King DA, Conwell Y, Cox C, Henderson RE, Denning DG, serotonergic challenge and lethality of suicide attempts in
Caine ED (2000). A neuropsychological comparison of major depression. Archive of General Psychiatry 60, 14–22.
depressed suicide attempters and nonattempters. Journal Petticrew M, Roberts H (2006). Systematic Reviews in the Social
of Neuropsychiatry and Clinical Neurosciences 12, 64–70. Sciences: a Practical Guide. Blackwell: Oxford.
Kouneiher F, Charron S, Koechlin E (2009). Motivation Raust A, Slama F, Mathieu F, Roy I, Chenu A, Koncke D,
and cognitive control in the human prefrontal cortex. Nature Fouques D, Jollant F, Jouvent E, Courtet P, Leboyer M,
Neuroscience 12, 939–945. Bellivier F (2007). Prefrontal cortex dysfunction in
LeGris J, Links PS, Van Reekum R, Tannock R, Toplak M patients with suicidal behavior. Psychological Medicine
(2012). Executive function and suicidal risk in women with 37, 411–419.
Neuropsychological markers of vulnerability to suicidal behavior 11
Richard-Devantoy S, Annweiler C, Le Gall D, Garre JB, Stuss DT (2011). Functions of the frontal lobes: relation
Olié JP, Beauchet O (2011). Cognitive inhibition in suicidal to executive functions. Journal of the International
depressed elderly: a case–control pilot study. Journal of Neuropsychological Society 17, 759–765.
Clinical Psychiatry 72, 871–872. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC,
Richard-Devantoy S, Guillaume S, Olié E, Courtet P, Vandenbroucke JP; STROBE Initiative (2008). The
Jollant F (2013a). Altered explicit recognition of Strengthening the Reporting of Observational Studies
facial disgust associated with predisposition to suicidal in Epidemiology [STROBE] statement: guidelines for
behavior but not depression. Journal of Affective Disorders. reporting observational studies. Journal of Clinical
Published online 9 March 2013. doi: 10.1016.j. Epidemiology 61, 344–349.
jad.2013.01.049. Westheide J, Quednow BB, Kuhn KU, Hoppe C,
Richard-Devantoy S, Jollant F, Kefi Z, Turecki G, Olié JP, Cooper-Mahkorn D, Hawellek B, Eichler P, Maier W,
Annweiler C, Beauchet O, Le Gall D (2012). Deficit of Wagner M (2008). Executive performance of depressed
cognitive inhibition in depressed elderly: a neurocognitive suicide attempters: the role of suicidal ideation. European
marker of suicidal risk. Journal of Affective Disorders 140, Archives of Psychiatry and Clinical Neurosciences 258, 414–421.
193–199. Williams JM, Broadbent K (1986). Distraction by emotional
Richard-Devantoy S, Olié E, Guillaume S, Bechara A, stimuli: use of a Stroop task with suicide attempters. British
Courtet P, Jollant F (2013b). Distinct alterations in Journal of Clinical Psychology 25, 101–110.
value-based decision-making and cognitive control in Williams JMG, Pollock LR (2001). Psychological aspects
suicide attempters: toward a dual neurocognitive model. of the suicidal process. In Understanding Suicidal Behaviour
Journal of Affective Disorders. Published online 19 July 2013. (ed. K. van Heeringen), pp. 76–93. John Wiley and Sons:
doi: 10.1016.j.jad.2013.06.052. Chicester.
Riley RD, Higgins JP, Deeks JJ (2011). Interpretation of Yen CF, Cheng CP, Ko CH, Yen JY, Huang CF, Chen CS
random effects meta-analyses. British Medical Journal (2008). Suicidality and its association with insight and
342, d549. neurocognition in Taiwanese patients with bipolar I
Rosenthal R (1979). The file drawer problem and tolerance disorder in remission. Journal of Nervous and Mental
for null results. Psychological Bulletin 86, 638–641. Disease 196, 462–467.