Content uploaded by Li Haolun
Author content
All content in this area was uploaded by Li Haolun on May 05, 2023
Content may be subject to copyright.
Suicide-Relevant Information Processing in Unipolar and Bipolar
Depression: An Eye-Tracking Study
Haolun Li
1, 2, 3
, Zhijun Li
1, 2
, Guanyi Lyu
1, 2
, Mi Wang
1, 2
, Bangshan Liu
1, 2
,
Yan Zhang
1, 2
, Lingjiang Li
1, 2
, and Greg J. Siegle
3
1
National Clinical Research Center for Mental Disorders, Department of Psychiatry,
The Second Xiangya Hospital of Central South University
2
Hunan Key Laboratory of Psychiatry and Mental Health, Institute of Mental Health,
Hunan Medical Center of Mental Health, China National Technology Institute on Mental Disorders
3
University of Pittsburgh School of Medicine
Suicide-relevant attentional biases are found in suicide attempters (SAs) with depression. Wenzel and Beck
provide a theoretical framework that suggests suicide-related attention biases confer vulnerability to suicide.
In this study, we integrated eye-tracking dynamics of suicide-related attentional biases with self-report mea-
sures to test their model. A free-viewing eye-tracking paradigm, which simultaneously presented four
images with different valences (suicide-related, negative, positive, neutral), was examined in 76 SAs with
unipolar or bipolar depression, 66 nonsuicidal depressive participants (ND), and 105 healthy never-
depressed healthy control participants (HC). Structural equation modeling (SEM) was used for the theory
testing. SA gazed more at suicide-relevant stimuli throughout the 25-s trial compared with ND. SA and
ND initially detected suicide-related stimuli faster than HC. Groups did not differ on how often theyinitially
gazed at suicide images or how fast they disengaged away from them. Eye-tracking indices of attentional
biases, together with self-reported hopelessness, adequately fit an SEM consistent with Wenzel and
Beck’s cognitive theory of suicide-related information processing. Potentially, suicide-related attention
biases could increase vulnerability to suicidal ideation and eventual suicidal behaviors.
General Scientific Summary
This study confirmed, using eye-tracking, that suicide attempters process suicide-relevant and other
emotional information differently from nonsuicide-attempting depressed and health-control individuals.
Results were consistent with Wenzel and Beck’s model, which suggests that suicide-specific attentional
bias and hopelessness could lead to suicide ideation and behaviors.
Keywords: suicide, depression, eye-tracking, structural equation model, information process
Supplemental materials: https://doi.org/10.1037/abn0000807.supp
Wenzel and Beck (2008) hypothesize that early attentional orienting
toward suicide cues (1–2 s following exposure to a stimulus) and later
sustained attention to suicide-relevant information (minutes to hours)
yield preoccupation with and subsequent vulnerability to suicide.
Indeed, suicide attempters (SAs) display biased attention to
suicide-relevant information (Becker et al., 1999;Cha et al., 2010;
Richard-Devantoy et al., 2016) and higher attention-related brain poten-
tials when processing suicide-relevant words compared to nonsuicide
Haolun Li https://orcid.org/0000-0001-9472-9764
Lingjiang Li https://orcid.org/0000-0002-8775-7852
This study was funded by the National Natural Science Foundation of China
(81171286, to Lingjiang Li), the National Key Research and Development
Program of China (2019YFA0706200 to Lingjiang Li). Mr. Haolun Li received
financial support for a joint-PhD program between Central South University and
the University of Pittsburgh from the China Scholarship Council (No.
201906370221 to Haolun Li). These funding agencies had no role in the design
and conduct of the study, the data collection and analysis, the interpretation of
the results, and the submission of the manuscript for publication.
Greg J. Siegle receives royalty payments on a patent regarding a novel
depression intervention licensed to Apollo Neurosciences, which is not rele-
vant to this article, and consults for Johnson and Johnson on novel
pharmacology unrelated to this project. All other authors report no biomed-
ical financial interests or potential conflicts of interest.
Lingjiang Li served as lead for funding acquisition and supervision,
contributed equally to software, and served in a supporting role for resources.
Haolun Li contributed toward conceptualization, original draft, data analysis,
reviewing, and editing. Guanyi Lyu and Zhijun Li contributed toward data col-
lecting. Lingjiang Li contributed toward reviewing and conceptualization. Greg
J. Siegle contributed toward editing, modeling, and reviewing. Mi Wang,
Bangshan Liu, and Yan Zhang contributed toward clinical data collection.
Correspondence concerning this article should be addressed to
Lingjiang Li, National Clinical Research Center for Mental Disorders,
Department of Psychiatry, The Second Xiangya Hospital, Central South
University, No. 139 Renmin Middle Road, Changsha, Hunan, 410011,
China. Email: LLJ2920@csu.edu.cn
Journal of Psychopathology and Clinical Science
© 2023 American Psychological Association 2023, Vol. 132, No. 4, 361–371
ISSN: 2769-7541 https://doi.org/10.1037/abn0000807
361
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
negative words (Baik et al., 2018). Biased attention toward
suicide-relevant, more than negative information, predicts future suicide
behaviors (Cha et al., 2010). Here, we consider the differential contribu-
tions of earlier and later attention to suicide-relevant stimuli, by moving
away from point estimates afforded by reaction times, as recommended
by Bar-Haim (2010), Kellough et al. (2008), and Shechner et al. (2013),
and by employing simultaneous stimuli with more than two valences to
approximate ecological attentional challenges, for example, as in
Hermans et al. (1999)andSoleymanietal.(2020).
We use the continuous eye-tracking index to differentiate between
early and sustained visual processing of emotional information, as
has successfully been used in other similar studies (Armstrong &
Olatunji, 2012). For example, depressed participants show sustained
eye-tracking biases to negative stimuli compared with nondysphoric
volunteers (Eizenman et al., 2003). Conversely, people with gener-
alized anxiety disorder look faster and more frequently toward threat
stimuli despite showing no significant differences in reaction times
(Mogg et al., 2000). These data are in keeping with earlier literature
using other methods showing early attention biases in anxiety and
more sustained information processing in depression (MacLeod et
al., 1986;MacLeod & Mathews, 1991;Williams et al., 1996).
A few eye-tracking studies have considered suicide. Children with
a history of suicidal ideation display sustained eye-tracking toward
socially threatening expressions (Tsypes et al., 2017). In college stu-
dents, fixation time to self-death words and scores on the Suicide
Implicit Attitudes Test were minimally correlated (r=.068;
Vannoy et al., 2016). However, few suicide ideators attempt suicide,
especially in nonclinical populations (Klonsky & May, 2014;ten
Have et al., 2009). It is possible that stronger effects would be
observed in SAs from the clinical population.
We thus investigated suicide-related attentional biases in clinically
depressed participants using a free-viewing paradigm to index early
and sustained attention (Eizenman et al., 2003;Kellough et al.,
2008;Wells et al., 2014). This work could ideally reveal cognitive tar-
gets for suicide prevention and, upon longitudinal replication, neuro-
psychological markers of suicide risk, could provide preliminary
proof and parameters to suicide-specific attention training in the
next step (Price et al., 2016).
Our primary hypothesis, based on Wenzel and Beck’s theory, was
that SAs would demonstrate early and sustained attentional biases to
suicide-relevant information assessed via eye-tracking indices. We
plan to investigate the temporal feature of this suicide-specific atten-
tional bias via time-course analysis. Second, as exploratory, we fur-
ther hypothesized that these patterns of gaze trajectories would be
suicide-specific and thus compared attentional biases across the
suicide-related and other nonsuicide-emotional information (nega-
tive, positive, and neutral). Finally, we tested Wenzel and Beck’s
theory by integrating our eye-tracking indices and clinical measure-
ments in a structural equation model (SEM).
Method
Sample
We recruited a wide age range (ages 12–48) to allow a life-span
developmental perspective with the intent that if there were age-related
differences in primary variables, we would analyze the groups sepa-
rately. There were 247 subjects including out- and in-patient clinical
participants (SA: n=76 and nonsuicidal depressive participants
[ND]: n=66) recruited from a local psychiatric hospital and healthy
volunteers (healthy control participants [HC]: n=105) from the
community. Power analysis using G*Power (Version 3.1.9.7; Faul
et al., 2007)atα=0.05, power =0.8, suggested that for a sample
of n=247, one-way three-group ANOVA effect sizes as small as
f=0.2 (partial eta squared ,0.04), and using empirical estimates
for within-group correlations (r=.5), Three-Group ×Four-Condition
mixed-plot ANOVA effect sizes as small as f=0.13 ( partial eta
squared ,0.017) could be detected. SA and ND participants met the
diagnosis of current DSM-IV major depressive disorder or bipolar II
disorder in a depressive episode and had Beck Depression Inventory
II (BDI) scores .18. SA participants had at least one suicide attempt
in the past 12 months. ND participants had no history of suicide
attempts or nonsuicidal self-injurious behavior (NSSI). HC participants
had BDI scores ,13 and no history of any psychiatric disorder, suicide
attempts, nor NSSI. Exclusion criteria for all participants included neu-
rological disorders, a present or past DSM-IV diagnosis of substance
abuse, psychotic disorder, obsessive-compulsive disorder, PTSD, his-
tory of electroconvulsive therapy in the past 6 months, color blindness,
clinical history of developmental disorders (e.g., autism, ADHD), and
inability to complete the experiment. Detailed descriptions about
excluded participants and participants selection are provided in supple-
mental materials (Figure S1 in the online supplemental materials). All
participants and parents who were used to obtain collateral/confirma-
tory information gave written informed consent, and all procedures
of this study were approved by the Medical Ethics Center of the univer-
sity institutional review board. For the clinical participants of inpatients
and outpatients, all the procedures of this study (clinical interviewing,
self-report and physiological data collecting, rules of informed consent)
were the same.
Assessments
Diagnoses were confirmed with patient medical records and a struc-
tured diagnostic interview (MINI-International Neuropsychiatric
Interview, 6.0; Si et al., 2009). Exclusions (e.g., developmental and
psychotic disorders) were screened by the MINI and by checking
medical records. Current depressive symptoms and severity were
measured with the Beck Depression Inventory (BDI-II), a widely
used 21-item self-report scale (Beck et al., 1996;Chan, 1991;
Z. Wang et al., 2011). State hopelessness was assessed via item-2
of the BDI (Woosley et al., 2014). We used the Beck Suicide
Ideation scale (BSI) to measure suicidal ideation within the past
week (Beck et al., 1988). History of suicide attempt was confirmed
by a psychiatrist (L or Z) via structured interview (MINI-Suicidality
section, coded as moderate or high risk; Si et al., 2009), any mention
of suicidality from the medical records, subjects’self-report, and col-
lateral information from family members and friends, all of whom
were consented for this purpose.
Free-Viewing Paradigm
Stimuli
The task included suicide-related, positive, negative, and neutral
images. Nonsuicide images were from the International Affective
Picture System (IAPS; The Center for the Study of Emotion and
Attention, 1999) with comparable suicide-relevant images generated
via a pilot study (Details about pilot study and criteria of image
selection can be seen in the online supplemental materials.).
LI ET AL.362
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Task
Twenty 25-s trials were presented via Experiment Builder (SR
Research). In 16 study trials, four images (suicide-relevant, negative,
positive, and neutral) were presented simultaneously. Four filler tri-
als presented four neutral images to reduce valence-based biased
visual search strategies. Images within each slide were cut using
Photoshop CS6 (Adobe) to be 16.7 ×12.5 cm
2
, subtending
13.5 hr ×10.0 V degrees of visual angle. The spatial location of
the four valences within trials and the trial order were randomized.
For each trial, areas of interest (AOIs) were defined as rectangles
encompassing each of the different images. Each trial began with
a 2-s period during which participants fixated on a cross in the center
of the screen. Participants were instructed to watch the screen as nat-
urally as possible after the fixation period, as if this was watching
television or looking at images on their phones.
Eye-Tracking System and Quantification of Eye-Tracking
Metrics
Eye movements were recorded by an Eyelink-1000 eye-tracking
system (SR Research) at a sampling rate of 1,000 Hz. Participants
sat, with their heads immobilized by a chin rest 70 cm from the dis-
play screen, the total viewing portion of which subtended 41.6 hr ×
24.1 V degrees of visual angle. Stimuli were presented on a BenQ
XL2411Z 24-in. (61 cm) flat screen. Before starting the paradigm,
eye calibration was performed on a 3 ×3-Point fixation matrix.
Calibration was repeated until the mean visual angle of gaze on
each point was less than 1°. A fixation was defined as remaining
within 1° visual angle for 100 ms.
Eye-Tracking Index
Six eye-tracking measures were selected in this study (Table 1).
The first one-tenth of the trial period (2.5 s) was considered
to represent “early stage”attention (Rinck & Becker, 2006)
with the remaining trial period considered “sustained”attention.
Individuals with suicide risk display a more sustained attentional
bias toward threatening stimuli (gaze duration) rather than initial
orienting attention (initial allocation; Tsypes et al., 2017). Thus,
gaze-duration-related measurements (e.g., initial gaze duration,
IGD; total dwell time, DT; proportional dwell time, PDT), which
were considered to mainly reflect overall attentional bias, were
adopted as primary measures in this study (IGD for the early
stage; DT, PDT for the later stage). Other eye-tracking parameters,
which have been associated with depression and anxiety
(Armstrong & Olatunji, 2012;Mogg et al., 2000;Sanchez et al.,
2013), were considered secondary outcomes: The percentage of
first fixation locating (PFF) and the time to initially detect a stimulus
(latency of first fixation; LFF) were reported as indicators of vigi-
lance and attentional orienting. Disengagement latency (LD) indi-
cates the processing of attentional control on emotional
information (Nummenmaa et al., 2006). We prioritized group con-
trasts on suicide-relevant stimuli and analyzed whether these scan-
path features were moderated by other emotional information (neg-
ative, positive, and neutral) as exploratory analyses of this study.
Data Analysis Strategy
One-way analysis of variances (ANOVA) was applied for compar-
ing group difference in suicide-specific eye-tracking performance.
Then, we conducted Three Between-Subject (group: SA, DC,
HC) ×Four Within-Subject (valence: suicide-relevant, negative, pos-
itive, neutral) mixed-design ANOVAs to explore whether group dif-
ferences were associated with different types of stimuli. Simple
effects within valences were analyzed via one-way ANOVAs subject
to Bonferroni correction for follow-up pairwise comparisons.
Analyses were via SPSS 25.0 (SPSS, IBM). All tests were two-tailed
and the level of statistical significance was at p,.05.
Dwell time (DT) was selected as a continuous index of attention
bias. We divided the 25-s trial into 1-s epochs and used Guthrie and
Buchwald’s method to control type I error at p=.05 level across
tests at each epoch (Guthrie & Buchwald, 1991). This strategy
calculated the number of consecutive significant between-groups
tests necessary to infer statistical significance given the autocorrelation
of the time series. Simulations suggested at least three consecutive
second-wise significant differences were necessary to infer a signifi-
cant window of group differences (Guthrie & Buchwald, 1991).
Structural equation model was evaluated using Amos 21.0 (IBM
Corporation). Full information maximum likelihood was used to
impute missing data (Enders & Bandalos, 2001). According to
Wenzel and Beck’s theory, suicide-related early attention (SEA)
and suicide-related sustained attention (SSA) were formulated as
latent variables and indicated by suicide-specific eye-tracking out-
comes. Chi-square (χ
2
,p..05), comparative fit index (CFI .
0.90), root mean square error of approximation (RMSEA ,0.08),
standardized root mean square residual (SRMR ,0.08), goodness
Table 1
Main Eye-Tracking Measurements
Measures Abbreviation (units) Definition
Early attention
Initial gaze duration IGD (ms) The sum of time spent on fixations on S-AOI within the first 1/10th of the trial period (2.5 s)
Latency of first fixation LFF (ms) The time it took for individuals to initially fixate on the S-AOI across the whole trial (25 s)
Percentage of first fixations PFF (%) The percentage of the first fixation of each trial locates within the Suicide items.
Sustained attention
Dwell time DT (ms) The sum of time spent during fixations on S-AOI across each trial
Latency of disengagement LD (ms) The average latency to saccade away from the S-AOI after a previous fixation on it.
Percentage of dwell time PDT (%) It was calculated as the DT within the S-AOI divided by the total DT for all four AOIs
Note. AOI =areas of interest; DT =dwell time; LD =latency of disengagement; LFF =latency of first fixation; PDT =percentage of dwell time; PFF =
percentage of first fixation; S-AOI =suicide-AOI.
SUICIDAL INFORMATION PROCESSING IN DEPRESSION 363
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
of fit (GFI .0.90), and Tucker Lewis index (TLI .0.90) were
selected to assess model fit as recommended (Bentler, 1990;
Kline, 2005). As Wenzel and Beck’s theory regards depression,
we selected only our depressed participants (SA and ND groups)
for model testing (n=142). Using the R package semPower
(Version 1.3.0), at df =25, 142 samples can provide at least 69%
power corresponding to RMSEA =0.073 on α=0.05 (Moshagen
& Erdfelder, 2016). There were no missing data in the clinical
sample.
We examined potential moderators and confounding variables,
including demographic characteristics (age, gender, education
years) and clinical features (e.g., bipolar and unipolar depression sta-
tus, source of in- and out-patient recruitment, depressive severity)
using multiple regression, interaction analysis, separated group com-
parison, and matched-subgroup comparisons. Clinical features
including the type and severity of depression and the recruitment
source were only examined within clinical groups.
Transparency and Openness
We report details about how we determined our data collection,
sample size, experiment materials selection, devices, operation,
and all variables in this study above. The hypothesis and design of
this study were not preregistered. Our IRB restricts us from providing
raw eye-tracking and clinical data, but aggregate indices and clinical
scale scores per participant are available upon request.
The code for eye-tracking data processing can be found through
the public website (https://github.com/Nastasii/FV_suicide.git).
The nonsuicidal stimulus materials used in this paradigm can be
requested through (https://csea.phhp.ufl.edu/media/iapsmessage
.html;P. J. Lang et al., 2005). The details of suicidal experiment
materials can be seen in the online supplemental materials.
Results
Group Characteristics
Demographic and clinical descriptions of the three groups are pro-
vided in Table 2. SA were younger, had fewer years of education,
and more females than the other groups. SA had higher depressive
severity and more bipolar participants than the ND group.
Suicide-Specific Eye-Tracking Performance
Early Attention
Results of one-way ANOVA indicated that SA gazed longer
(IGD) on suicide-relevant stimuli in the first 2,500 ms compared
with ND and HC participants, and no significant difference was
found in initial gaze duration (IGD) between ND and HC
(Table 3). SA and ND participants initially detected suicide-related
items faster (LFF) than HC participants, though there was no differ-
ence between SA and ND (Table 3). However, the first fixation loca-
tion (PFF) did not significantly differ in the four valences (Table 3).
Further analysis and details about pairwise comparison can be seen
in Table S5-A in the online supplemental materials.
Sustained Attention
As seen in Table 3, groups differences in disengagement latency
(LD) were not significant. Dwell time (DT) differed across groups:
SAs gazed significantly longer time on suicide stimulus than ND
and HC participants (Table 3). Similarly, the SAs spent a greater per-
centage of time dwelling (PDT) toward suicide-related stimulus than
ND and HC groups (Table 3;Tables S5-A, S6-A in the online sup-
plemental materials).
Scan-Path Patterns Across Other Valences
A three (group) by four (valence) mixed-design ANOVA was
conducted to examine the group differences in attentional biases
toward different valences of information. The interaction effects
were significant in measures of initial gaze duration (IGD), latency
to first fixation (LFF), dwell time (DT), and proportional dwell
time (PDT) (Table 3).
Early Attention
For initial gaze duration, there was no significant group simple
effect for negative stimuli, but there was for positive (SA ,ND ,
HC) and neutral (SA ,HC) stimuli (Table 3). There were no signif-
icant effects from the latency of first fixation (LFF) toward negative,
positive, or neutral stimuli (Table 3). There was a significant main
effect of valence for the place of first fixation (PFF), F(2.83,
691.71) =21.94, p,.001, η
2
p
=0.08. Compared with other cate-
gories, people had a lower proportion of initially gazing toward neg-
ative images (Table S5-BCD in the online supplemental materials).
Sustained Attention
The simple effect of group comparisons on each valence revealed
that SAs gazed longer time (DT) with a greater percentage of total
dwell time (PDT) to negative information, with the opposite
effects for positive images (SA ,ND ,HC), and neutral images
(SA ,ND, HC) (Figure 2;Table 3). The main effect of valence
on disengaging latency (LD) was significant, F(2.96, 722.49) =
12.96, p,.001, η
2
p
=0.05. Follow-up comparison between
valences indicated that SAs use significantly longer time in distract-
ing away from suicide images compared with other types of stimulus
(Table S5-BCD in the online supplemental materials).
Time-Course Analysis
The time by group interaction for total gazing time (DT) on sui-
cide images was not significant, F(11.8, 2,899.4) =1.2, p=0.24,
ηp2=0.01. There was a significant main effect of time for DT on
suicide images, F(11.8, 2,899.4) =22.47, p,.001, ηp2=0.084.
A priori sample-wise ANOVAs showed significant (.3satp,.05)
group differences from 3 to 6 s, F(2,245) =8.62, p,.05, η
2
p
=
0.07, and 10–25 s, F(2, 245) =20.506, p,.05, η
2
p
=0.14, follow-
ing stimulus onset, with SA .ND from 11 to 25 s following stimu-
lus onset (Figure 1).
Model Testing
Our collected eye-tracking and self-report data allowed us to eval-
uate a structural equation model constructed (Figure 3A)toreflect
Wenzel and Beck’s theoretical model (Figure 3B). Suicide act status
was trichotomized (none: n=66 (46%), single, n=31 (22%), and
multiple, n=45 (32%) attempts; Boisseau et al., 2013).
LI ET AL.364
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
As the χ
2
statistic is sensitive to sample size and distribution (Bentler,
1990), we give it less emphasis than our other reported model fitindi-
ces. Our model exhibited a good fit to the data using the other a priori
cutoffs: χ
2
=43.529, df =25, p=.012 (cutoff: p..05), CFI =0.963
(.0.90), RMSEA =0.073 (,0.08), SRMR =0.0753 (,0.08),
GFI =0.938 (.0.90), TLI =0.947 (.0.90).
All path loadings were statistically significant, p,.05 (Figure 3;
Table S8 in the online supplemental materials). The six eye-tracking
outcomes were significantly indicative of suicide-relevant cognition
bias. As hypothesized (Wenzel & Beck, 2008), the coefficients from
suicide ideation to both state hopelessness (β=0.27, p,.001) and
suicide-related sustained attention (SSA; β=0.45, p,.001) were
significant with effect sizes (R
2
=0.35) typically considered to be
large (Cohen, 1988).
Sensitivity and Interaction Analysis
We verified our main group effects were not moderated by dif-
ferences in demographic and clinical factors through sensitivity
and subgroup analysis (details in Tables S2–S4 and Figures S3,
S4 in the online supplemental materials). A gender by group inter-
action in the suicide-related percentage of the first fixation was
found. Male SAs gazed more frequently toward suicide images
initially than females. In particular, there were not salient differ-
ences in findings between adolescents and adults in this sample
(Table S3-A in the online supplemental materials). There was a
significant group (SA, ND) by source (in-/out-patients) interaction
effect in suicide-specific dwelling time and proportional dwelling
time (Table S3-A, E; Figure S3-B1, B2 in the online supplemental
materials). Separated comparisons showed that participants shared
similar demographic and clinical features even under different
subgroup (male vs. female, children vs. adults; Table S7-B, C in
the online supplemental materials).
Discussion
As hypothesized, SA showed attentional biases toward
suicide-relevant stimuli in both early and sustained stages of infor-
mation processing compared with ND and HC participants. Based
on neurobehavioral measurements, model testing supported
Wenzel and Beck’s theory that information processing is associated
with hopelessness, which can lead to suicidal behaviors.
Early Attention
Primary outcome indicated that SA showed early attentional bias
toward suicide stimulus by gazing more time and secondary out-
comes indicated that the suicide-related vigilance was not
distinguished between SAs and nonsuicide-depressive patients.
Together, outcomes for the three scan-path metrics associated with
early-stage attention in this study can be seen as a continuum from
stimulus-driven to goal-driven information processing. We found
that there was a gradually increasing effect of elaborative and con-
scious information processing on suicide stimuli in early attention
in SAs, as in the study of Maljkovic and Nakayama (1994) whereby
both depression groups had decreased time to detect suicide cues
(LFF), the SA group but not the ND group had increased initial
gaze frequency, and only the HC group had latency to first fixation
after the “early stage attention”(.2,500 ms) cutoff. Similar patterns
have been observed in depressed individuals in response to negative
Table 2
Demographic and Clinical Characteristics by Group
Variable SA (n=76) ND (n=66) HC (n=105) pdfF/Welch’sF/χ
2
value Effect size Post hoc contrast
Age (years) 18.70 (3.74) 21.95 (4.91) 20.34 (5.28) ,.001*** 2,244 8.30 w
2
=0.064 SA＜ND
Education (years) 11.95 (2.35) 13.69 (2.66) 13.51 (2.92) ,.001*** 2,153 11.31 w
2
=0.066 SA＜HC, ND
Gender (m/f)/F% 13/63 (83%) 24/42 (64%) 35/70 (67%) 0.016* 2 8.26 V=0.18 ND, HC ,SA
Disorders (bipolar/unipolar) U% 27/49 (64%) 13/53 (80%) —.036* 1 4.37 f=−0.18 SA ,ND
BDI-II score 35.51 (8.55) 29.80 (7.81) 5.01 (3.44) ,.001*** 2,112 657.01 w
2
=0.82 HC ,ND ,SA
Hopeless (item2) 2.22 (0.79) 1.55 (0.99) 0.09 (0.28) ,.001*** 2,105 303.37 w
2
=0.59 HC ,ND ,SA
BSI score 20.72 (7.72) 9.77 (7.04) 0.35 (0.88) ,.001*** 2,95 316.10 w
2
=0.70 HC ,ND ,SA
Source (in/out-patient), inpatient% 38/76 (50%) 25/66 (37.9) 0.147 1 2.103 f=0.12
Suicide act Single 31 Multiple 45
Note. BDI =Beck Depression Inventory-II; BSI =Beck Suicide Ideation Scale, state hopelessness is measured by the score of item 2 of BDI-II; SA =suicide attempters, ND =group of nonsuicide
depression participants; HC =healthy comparison. p-values for contrasts are from post hoc Tukey HSD tests based on the omnibus test error term.
*p,.05. ***p,.001.
SUICIDAL INFORMATION PROCESSING IN DEPRESSION 365
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
stimuli (Armstrong & Olatunji, 2012;Maljkovic & Nakayama,
1994;Mogg & Bradley, 2005). This pattern could help explain the
historical variability in observing early attention biases in depression
using less temporally precise methods (Armstrong & Olatunji, 2012;
Mogg & Bradley, 2005;Peckham et al., 2010).
Sustained Attention
Consistent with Wenzel and Beck’s hypothesis that individuals at
risk for suicide become fixated on and occupied by suicide cues
(Wenzel & Beck, 2008), SA exhibited a longer dwell time and a
Table 3
Eye-Tracking Results for Groups of Suicide Attempters (SA), Nonsuicide Depression (ND), and Health Control (HC)
Valence Eye measures SA (n=76) M(SD)ND(n= 66) M(SD)HC(n= 105) M(SD)pdfFEffect size η
2
p
Post hoc contrast
Suicide IGD (ms) 473.4 (119.9) 405.3 (121.3) 370.1 (112.9) ,.001*** 2,244 17.07 0.123 HC, ND ,SA
LFF (ms) 2,067.9 (777.5) 2,315.9 (916.7) 2,722.0 (1,138.0) ,.001*** 2,244 10.50 0.070 SA, ND ,HC
PFF (%) 28.1 (9.1) 26.99 (7.8) 27.44 (7.9) .710 2,244 0.34 0.003
DT (ms) 6,061.23 (2,102.30) 3,958.91 (1,587.29) 3,017.51 (1,211.82) ,.001*** 2,244 77.94 0.390 HC ,ND ,SA
LD (ms) 218.8 (52.6) 201.9 (49.1) 213.0 (40.3) .097 2,244 2.36 0.010
PDT (%) 39.4 (13.4) 25.8 (9) 19.4 (7.5) ,.001*** 2,244 89.22 0.422 HC ,ND ,SA
Negative IGD (ms) 316.5 (92.1) 323.1 (98.2) 315.7 (97.3) .877 2,244 0.13 0.001 —
LFF (ms) 2,605 (914) 2,643 (1,008) 2,915 (1,248) .115 2,244 2.18 0.018 —
PFF (%) 20.0 (6.7) 21.1 (8.3) 22.0 (8.1) .219 2,244 1.53 0.012 —
DT (ms) 3,958 (1,139) 3,718 (1,057) 3,324 (967) ＜.001*** 2,244 8.46 .065 HC,SA
LD (ms) 203.1 (40.3) 199.9 (45.9) 209.2 (38.2) .321 2,244 1.14 .009
PDT (%) 25.4 (6.0) 24.4 (6.0) 21.4 (6.0) ,.001*** 2,244 11.22 0.084 HC ,ND, SA
Positive IGD (ms) 251.2 (78.8) 298.9 (89.7) 339.3 (100.8) ,.001*** 2,244 20.46 0.144 SA ,ND ,HC
LFF (ms) 3,008 (1,288) 2,656 (1,299) 2,788 (1,205) .236 2,244 1.07 0.012 —
PFF (%) 26.3 (8.4) 28.0 (8.3) 27.1 (8.8) .492 2,244 .071 0.006 —
DT (ms) 2,706 (1,157) 3,882 (1,689) 5,223 (1,922) ,.001*** 2,244 51.58 0.297 SA ,ND ,HC
LD (ms) 192.0 (36.3) 186.1 (32.9) 200.3 (35.9) .034* 2,244 3.42 0.027 ND ,HC
PDT (%) 17.4 (7.0) 25.6 (10.5) 33.3 (11.3) ,.001*** 2,244 56.30 0.316 SA ,ND ,HC
Neutral IGD (ms) 230.7 (98.5) 257.8 (85.1) 281.7 (107.6) .003* 2,244 5.82 0.046 SA ,HC
LFF (ms) 3,192 (1,138) 2,968 (1,141) 3,242 (1,331) .344 2,244 1.07 0.009 —
PFF (%) 25.6 (10.5) 23.9 (8.8) 23.4 (9.6) .310 2,244 1.18 0.010 —
DT (ms) 2,772 (1,197) 3,669 (1,042) 4,061 (1,069) ,.001*** 2,244 20.55 0.200 SA ,ND, HC
LD (ms) 201.9 (40.4) 201.2 (39.9) 203.7 (35.1) .901 2,244 0.10 0.001
PDT (%) 17.7 (6.9) 24.2 (6.5) 26.0 (5.9) ,.001*** 2,244 38.05 0.238 SA ,ND, HC
Interact effect: group ×valence IGD (ms) ,.001*** 5.3, 661.9 13.2 0.098
LFF (ms) ,.001*** 5.6, 684.7 6.71 0.052
PFF (%) .46 5.7, 706.4 0.94 0.008
DT (ms) ,.001*** 3.9, 478.9 57.79 0.321
LD (ms) .217 6, 732 1.39 0.011
PDT (%) ,.001*** 3.9, 480.2 59.69 0.329
Note. SA =suicide attempters; ND =group of nonsuicide depression participants; HC =healthy comparison; IGD =initial gaze duration; LFF =latency of
first fixation; PFF =percentage of first fixation; DT =dwell time; LD =latency of disengagement; PDT =percentage of dwell time.
*p,.05, ***p,.001.
Figure 1
Suicide-Stimulus Dwell Time by Each Second, in Three Groups
Note. SA =suicide attempters; ND =group of nonsuicide depression participants; HC =healthy compar-
ison. See the online article for the color version of this figure.
LI ET AL.
366
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
higher percentage of dwelling time on suicide-relevant stimuli than
ND and HC participants through the 25 s trials (Table 2A,B;
Figure 1). Simple effect analyses revealed that the suicide-specific
proportion of gazing time was also moderated by patient source.
Hospitalized SAs were more sensitive in gazing toward suicide
images in the SA group, which is a line with studies about the rela-
tionship between suicidality and the source of patients (Pfeffer et al.,
1986; Table S3-E and Figure S3-B1, B2 in the online supplemental
materials). Group differences were not significant for disengagement
latency for suicide images, which was not expected based on previ-
ous depression and anxiety research (De Raedt, 2013;Ellenbogen &
Schwartzman, 2009;Sanchez et al., 2013). One reason is that we
were using the average latency of the whole trial instead of the
latency of initially disengaging away from suicide-AOI, to reflect
the ability of elaborative attention control. Wenzel and Beck suggest
the problem is not that individuals cannot disengage, but that they
re-engage, perhaps in a more perseverative way.
Our data are thus consistent with attention biases in SAs being a
continuous and accumulative process from early to sustained stages
of information processing, particularly as illustrated in Figure 1. This
formulation could suggest that suicide-relevant processes share fea-
tures with other sustained cognitive processes such as ruminative
thinking (De Lissnyder et al., 2011;Disner et al., 2011;Tsypes et
al., 2017).
Selective Attention to Other Valences
Consistent with previous eye-tracking studies (Armstrong &
Olatunji, 2012;Eizenman et al., 2003;Wells et al., 2014), our results
suggested that ND and SA participants showed reduced attentional
biases toward positive information compared with HC participants
(Figure 2), potentially as a dysfunctional reactivity in reward-responsive
(e.g., striatal) brain areas, as observed in depression (Pizzagalli et al.,
2008,2009). Indeed, striatal reactivity has been associated with eye-
tracking to pleasant stimuli (Kafkas & Montaldi, 2015). Yet, no differ-
ence was found in non-suicide-negative attentional biases between ND
and SA participants. These results suggest that suicide-specific cogni-
tive factors, rather than nonsuicide-negative information, are a more
salient marker for suicide attempts. Indeed, many studies indicate that
information processing biases are increased in response to personally
relevant information for specific populations (Jones et al., 2011;
Siegle et al., 2003). Assessments based on suicide-specific cognitive
processing may enhance the ability to identify suicide risk (Rogers et
al., 2022).
Model Testing
Scan-path measurements served as manifest indicators of cogni-
tive processing in our structural equation model (Figure 3A;
Tables S8 and S9 in the online supplemental materials). As predicted
by Wenzel and Beck (2008), early attention bias indicators were pos-
itively associated with more sustained attention indices. This result
supports the idea that people may focus more on suicide-relevant
information when they easily detect it. Integration across eye-
tracking and clinical indices was also supportive of Wenzel and
Beck’s model, which suggests that suicide-relevant attentional
bias, state hopelessness, suicide ideation, and suicide acts are related.
Our finding that sustained attention on suicide cues was specifically
associated with state hopelessness is consistent with the notion of a
cognitive-affective spiral (Teasdale & Dent, 1987), by which, over
time, attention to suicide-relevant information increases cognitive
processes historically linked to suicide ideation and eventually, the
possibility of suicide acts (Wenzel & Beck, 2008).
One of the challenges in studies of suicide ideation and suicidal
behavior is the reliability of suicide risk evaluation based on self-
report assessment (Busch et al., 2003;Wilson, 2009). This study
suggests that suicide-specific attentional biases could be a candidate
biobehavioral marker for inclusion in a comprehensive suicide risk
assessment. We have specifically added to the current literature by
showing that suicide-specific attentional biases are more likely to
occur in sustained stages of information processing (seconds after
stimulus onset). The present study also extends the current thinking
on suicide-specific attention by showing that SAs may not have dif-
ficulty in disengaging from suicide-related information. This pre-
served capacity may be useful to capitalize on in clinical
interventions.
Limitations and Future Directions
There are some limitations in this study. Our sample consisted of
both adolescents and adults, though age did not have moderate
effects. Most of our adults were young and middle-aged, limiting
generalizability to older individuals. The sample size (n=142)
and observed power (0.69) were relatively small for parameter esti-
mation in SEM (Y. A. Wang & Rhemtulla, 2021). Our test of Beck’s
model used item 2 of the BDI-II as a proxy for hopelessness. Though
scores on this item correlate strongly with the Beck Hopelessness
Scale, a fuller assessment of hopelessness could provide more infor-
mation about hopelessness than our single-item measure (Beck et al.,
1974). The assessments of previous suicidal behavior based on clin-
ical records were subject to chance and subjective variability. In the
current cross-sectional data set, we could not evaluate the extent to
which attention biases were predictive of future risk for suicidal
acts or how they change across the course of disease. Longitudinal
data are being collected and will be analyzed in a separate
Figure 2
Dwell Time by Each Valences of Stimulus in Three Groups
Note. SA =suicide attempters, ND =group of nonsuicide depression par-
ticipants, HC =healthy comparison. See the online article for the color ver-
sion of this figure.
SUICIDAL INFORMATION PROCESSING IN DEPRESSION 367
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
manuscript. The negative images selected in the task had higher
mean arousal than the suicide items (negative: 5.37, suicide: 5.17,
Table S1 in the online supplemental materials). Though the main
eye-tracking outcomes of our study did indicate a significant atten-
tional bias toward suicide-related images in SA, this pattern might
have been even stronger had the images been matched for arousal
Figure 3
Cognitive Theory of Suicide
Note. (A) LD =latency of disengagement; PDT =percentage of dwell time; DT =dwell time; PFF =per-
centage of first fixation on suicide; LFF =latency of first fixation; IGD =initial gaze duration; SI =suicide
ideation; S-SA =suicide-related sustained attention; S-EA =suicide-related early attention
rectangle-observed variable; circle-latent variable. (B) Beck’s suicide-relevant cognitive processing theory.
*p,.05. ***p,.001.
LI ET AL.
368
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
across valences. Although we defined suicide attempts as having
occurred within 1 year, the period between the time point of the
last suicide attempt and data collection was variable and might
have influenced suicide-related performance (Baik et al., 2018;
Cha et al., 2010;Richard-Devantoy et al., 2016). Future work quan-
tifying time since suicide attempt and ideally, stratifying participants
on event recency could help to address this concern. All participants
were recruited in China. Suicide epidemiological characteristics in
China are different from other regions. There are also cultural differ-
ences in attitudes toward death between China and Western nations
(Cao et al., 2015). The generalizability of our findings to other cul-
tures may thus be influenced by these factors, future study could
compare with suicide-information processing in people from differ-
ent cultures. Group may be confounded with age range such that
very young subjects in the nonsuicidal but depressed group may
eventually attempt suicide. As we selected only suicidal individuals
who were depressed, there are SAs without depression to whom the
current work does not generalize. Evidence indicates they may be
qualitatively different from suicidal individuals who are depressed,
for example, characterized by more inward aggression or violence
(Pitchot et al., 2001).
Conclusions
Consistent with Wenzel and Beck’s (2008) cognitive model of
suicide, we found that individuals who are prone to suicide show
attentional biases toward suicide-relevant stimuli at both early and
later stages of information processing. They differed on some indices
from other individuals with depression but without a history of sui-
cide attempt. Both early- and later-stage biases were related to clin-
ical features, which could suggest that they are active in exacerbating
suicide ideation and potentiate suicide acts when people are in a state
of hopelessness. Differential results for suicide-relevant versus
nonsuicide-negative information indicate that clinically, it may be
important to target suicide-relevant cognitions rather than exclu-
sively targeting general negative attention biases for depressed
patients in working to address suicide risk.
References
Armstrong, T., & Olatunji, B. O. (2012). Eye tracking of attention in the
affective disorders: A meta-analytic review and synthesis. Clinical
Psychology Review,32(8), 704–723. https://doi.org/10.1016/j.cpr.2012
.09.004
Baik, S. Y., Jeong, M., Kim, H. S., & Lee, S.-H. J. (2018). ERP Investigation
of attentional disengagement from suicide-relevant information in patients
with major depressive disorder. Journal of Affective Disorders,225, 357–
364. https://doi.org/10.1016/j.jad.2017.08.046
Bar-Haim, Y. (2010). Research review: Attention bias modification (ABM):
A novel treatment for anxiety disorders. Journal of Child Psychology and
Psychiatry,51(8), 859–870. https://doi.org/10.1111/j.1469-7610.2010
.02251.x
Beck, A. T., Steer, R. A., Ball, R., & Ranieri, W. F. (1996). Comparison of
Beck Depression Inventories-IA and -II in psychiatric outpatients. Journal
of Personality Assessment,67(3), 588–597. https://doi.org/10.1207/
s15327752jpa6703_13
Beck, A. T., Steer, R. A., & Ranieri, W. F. (1988). Scale for suicide ideation:
Psychometric properties of a self-report version. Journal of Clinical
Psychology,44(4), 499–505. https://doi.org/10.1002/1097-4679(198807)
44:4,499::AID-JCLP2270440404.3.0.CO;2-6
Beck, A. T., Weissman, A., Lester, D., & Trexler, L. (1974). The measure-
ment of pessimism: The hopelessness scale. Journal of Consulting and
Clinical Psychology,42(6), 861–865. https://doi.org/10.1037/h0037562
Becker, E. S., Strohbach, D., & Rinck, M. (1999). A specific attentional bias
in suicide attempters. The Journal of Nervous and Mental Disease,
187(12), 730–735. https://doi.org/10.1097/00005053-199912000-00004
Bentler, P. M. (1990). Comparative fit indexes in structural models.
Psychological Bulletin,107(2), 238–246. https://doi.org/10.1037/0033-
2909.107.2.238
Boisseau,C.L.,Yen,S.,Markowitz,J.C.,Grilo,C.M.,Sanislow,C.A.,Shea,
M.T.,Zanarini,M.C.,Skodol,A.E.,Gunderson,J.G.,Morey,L.C.,&
McGlashan, T. H. (2013). Individuals with single versus multiple suicide
attempts over 10 years of prospective follow-up. Comprehensive Psychiatry,
54(3), 238–242. https://doi.org/10.1016/j.comppsych.2012.07.062
Burnham, K. P., & Anderson, D. R. (2004). Multimodel inference:
Understanding AIC and BIC in model selection. Sociological Methods
and Research,33(2), 261–304. https://doi.org/10.1177/0049124104268644
Burton, L., Pfaff, D., Bolt, N., Hadjikyriacou, D., Silton, N., Kilgallen, C.,
Cofer, J., & Allimant, J. (2010). Effects of gender and personality on
the Conners Continuous Performance Test. Journal of Clinical and
Experimental Neuropsychology,32(1), 66–70. https://doi.org/10.1080/
13803390902806568
Busch, K. A., Fawcett, J., & Jacobs, D. G. (2003). Clinical correlates of inpa-
tient suicide. The Journal of Clinical Psychiatry,64(1), 14–19. https://
doi.org/10.4088/JCP.v64n0105
Cao, X. L., Zhong, B. L., Xiang, Y. T., Ungvari,G. S., Lai, K. Y., Chiu, H. F.,
& Caine, E. D. (2015). Prevalence of suicidal ideation and suicide attempts
in the general population of China: A meta-analysis. The International
Journal of Psychiatry in Medicine,49(4), 296–308. https://doi.org/10
.1177/0091217415589306
Cha, C. B., Najmi, S., Park, J. M., Finn, C. T., & Nock, M. K. (2010).
Attentional bias toward suicide-related stimuli predicts suicidal behavior.
Journal of Abnormal Psychology,119(3), 616–622. https://doi.org/10
.1037/a0019710
Chan, D. W. (1991). Depressive symptoms and depressed mood among
Chinese medical students in Hong Kong. Comprehensive Psychiatry,
32(2), 170–180. https://doi.org/10.1016/0010-440X(91)90010-A
Cohen, J. (1988). Statistical power analysis for the behavioral sciences.
Erbaum Press.
Cohen, J. (2013). Statistical power analysis for the behavioral sciences.
Academic Press.
De Lissnyder, E., Derakshan, N., De Raedt, R., & Koster, E. H. (2011).
Depressive symptoms and cognitive control in a mixed antisaccade task:
Specific effects of depressive rumination. Cognition and Emotion,25(5),
886–897. https://doi.org/10.1080/02699931.2010.514711
De Raedt, R. (2013). Activation of implicit self-schemas and a difficulty to
disengage from negative cognitions in depression: An experimental psy-
chopathology approach. Psychologica Belgica,46(1–2), 117–130.
https://doi.org/10.5334/pb-46-1-2-117
Disner, S. G., Beevers, C. G., Haigh, E. A., & Beck, A. T. (2011). Neural
mechanisms of the cognitive model of depression. Nature Reviews
Neuroscience,12(8), 467–477. https://doi.org/10.1038/nrn3027
Donoho, D., & Jin, J. (2008). Higher criticism thresholding: Optimal feature
selection when useful features are rare and weak. Physical Sciences,
105(39), 14790–14795. https://doi.org/10.1073/pnas.0807471105
Eizenman, M., Yu, L. H., Grupp, L., Eizenman, E., Ellenbogen, M., Gemar,
M., & Levitan, R. D. (2003). A naturalistic visual scanning approach to
assess selective attention in major depressive disorder. Psychiatry
Research,118(2), 117–128. https://doi.org/10.1016/s0165-1781(03)
00068-4
Ellenbogen, M. A., & Schwartzman, A. E. (2009). Selective attention and
avoidance on a pictorial cueing task during stress in clinically anxious
and depressed participants. Behaviour Research and Therapy,47(2),
128–138. https://doi.org/10.1016/j.brat.2008.10.021
SUICIDAL INFORMATION PROCESSING IN DEPRESSION 369
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Enders, C. K., & Bandalos, D. L. (2001). The relative performance of full
information maximum likelihood estimation for missing data in structural
equation models. Structural Equation Modeling: A Multidisciplinary
Journal,8(3), 430–457. https://doi.org/10.1207/S15328007SEM0803_5
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power 3: A
flexible statistical power analysis program for the social, behavioral, and
biomedical sciences. Behavior Research Methods,39(2), 175–191.
https://doi.org/10.3758/bf03193146
Guthrie, D., & Buchwald, J. S. (1991). Significance testing of difference
potentials. Psychophysiology,28(2), 240–244. https://doi.org/10.1111/j
.1469-8986.1991.tb00417.x
Hardy, L. H., Rand, G., & Rittler, M. C. (1946). Tests for detection and anal-
ysis of color blindness; comparison of editions of the Ishihara test.
Archives of Ophthalmology,35(2), 109–119. https://doi.org/10.1001/
archopht.1946.00890200114004
Hermans, D., Vansteenwegen, D., & Eelen, P. (1999). Eye movement regis-
tration as a continuous index of attention deployment: Data from a group of
spider anxious students. Cognition and Emotion,13(4), 419–434. https://
doi.org/10.1080/026999399379249
Jones, N. P., Siegle, G. J., Proud, L., Silk, J. S., Hardy, D., Keljo, D. J., Dahl,
R. E., Szigethy, E. (2011). Impact of inflammatory bowel disease and high-
dose steroid exposure on pupillary responses to negative information in
pediatric depression. Psychosomatic Medicine,73(2), 151–157. https://
doi.org/10.1097/PSY.0b013e318207ffea
Kafkas, A., & Montaldi, D. (2015). Striatal and midbrain connectivity with
the hippocampus selectively boosts memory for contextual novelty.
Hippocampus,25(11), 1262–1273. https://doi.org/10.1002/hipo.22434
Kellough, J. L., Beevers, C. G., Ellis, A. J., & Wells, T. T. (2008). Time
course of selective attention in clinically depressed young adults: An eye
tracking study. Behaviour Research and Therapy,46(11), 1238–1243.
https://doi.org/10.1016/j.brat.2008.07.004
Kline, R. B. (2005). Principles and practice of structural equation modeling
(2nd ed.). Guilford Press.
Kline, R. B. (2015). Principles and practice of structural equation modeling.
Guilford Press.
Klonsky, E. D., & May, A. M. (2014). Differentiatingsuicide attempters from
suicide ideators: A critical frontier for suicidology research. Suicide and
Life-Threatening Behavior,44(1), 1–5. https://doi.org/10.1111/sltb.12068
Kõlves, K., McDonough, M., Crompton, D., & de Leo, D. (2018). Choice of
a suicide method: Trends and characteristicsPsychiatry Research,260,67–
74https://doi.org/10.1016/j.psychres.2017.11.035
Lakens, D. (2013). Calculating and reporting effect sizes to facilitate cumu-
lative science: A practical primer for t-tests and ANOVAs. Frontiers in
Psychology,4, Article 863. https://doi.org/10.3389/fpsyg.2013.00863
Lang, P. (1980). Self-assessment manikin. The Center for Research in
Psychophysiology, University of Florida.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2005). International affective
picture system (IAPS): Affective ratings of pictures and instruction man-
ual. NIMH, Center for the Study of Emotion and Attention.
MacLeod, C., & Mathews, A. M. (1991). Cognitive experimental approaches
to the emotional disorders. In P. R. Martin (Ed.), Handbook of behavior
therapy and psychological science an integrative approach (Vol. 16,
pp. 116–150). Allyn & Bacon.
MacLeod, C., Mathews, A. M., & Tata, P. (1986). Attentional bias in emo-
tional disorder. Journal of Abnormal Psychology,95(1), 15–20. https://
doi.org/10.1037/0021-843X.95.1.15
Maljkovic, V., & Nakayama, K. (1994). Priming of pop-out: I. Role of fea-
tures. Memory and Cognition,22(6), 657–672. https://doi.org/10.3758/
bf03209251
Mogg, K., & Bradley, B. P. (2005). Attentional bias in generalized anxiety
disorder versus depressive disorder. Cognitive Therapy and Research,
29(1), 29–45. https://doi.org/10.1007/s10608-005-1646-y
Mogg, K., Millar, N., & Bradley, B. P. (2000). Biases in eye movements to
threatening facial expressions in generalized anxiety disorder and
depressive disorder. Journal of Abnormal Psychology,109(4), 695–704.
https://doi.org/10.1037/0021-843X.109.4.695
Moshagen, M., & Erdfelder, E. (2016). A new strategy for testing structural
equation models. Structural Equation Modeling: A Multidisciplinary
Journal,23(1), 54–60. https://doi.org/10.1080/10705511.2014.950896
Nock, M. K., Joiner, T. E., Gordon, K. H., Lloyd-Richardson, E., & Prinstein,
M. J. (2006). Non-suicidal self-injury among adolescents: Diagnostic cor-
relates and relation to suicide attempts. Psychiatry Research,144(1), 65–
72. https://doi.org/10.1016/j.psychres.2006.05.010
Nummenmaa, L., Hyönä, J., & Calvo, M. G. (2006). Eye movement assess-
ment of selective attentional capture by emotional pictures. Emotion,6(2),
257–268. https://doi.org/10.1037/1528-3542.6.2.257
Peckham, A. D., McHugh, R. K., & Otto, M. W. (2010). A meta-analysis of
the magnitude of biased attention in depression. Depression and Anxiety,
27(12), 1135–1142. https://doi.org/10.1002/da.20755
Pfeffer, C. R., Plutchik, R., Mizruchi, M. S., & Lipkins, R. (1986). Suicidal
behavior in child psychiatric inpatients and outpatients and in nonpatients.
American Journal of Psychiatry,143(6), 733–738. https://doi.org/10
.1176/ajp.143.6.733
Pitchot, W., Hansenne, M., & Ansseau, M. (2001). Role of dopamine in non-
depressed patients with a history of suicide attempts. Eur Psychiatry,
16(7), 424–427. https://doi.org/10.1016/S0924-9338(01)00601-0
Pizzagalli, D. A., Holmes, A. J., Dillon, D. G., Goetz, E. L., Birk, J. L.,
Bogdan, R., Dougherty, D. D., Iosifescu, D. V., Rauch, S. L., & Fava,
M. (2009). Reduced caudate and nucleus accumbens response to rewards
in unmedicated individuals with major depressive disorder. American
Journal of Psychiatry,166(6), 702–710. https://doi.org/10.1176/appi.ajp
.2008.08081201
Pizzagalli, D. A., Iosifescu, D., Hallett, L. A., Ratner, K. G., & Fava, M.
(2008). Reduced hedonic capacity in major depressive disorder:
Evidence from a probabilistic reward task. Journal of Psychiatric
Research,43(1), 76–87. https://doi.org/10.1016/j.jpsychires.2008.03.001
Price, R. B., Greven, I. M., Siegle, G. J., Koster, E. H., & De Raedt, R.
(2016). A novel attention training paradigm based on operant conditioning
of eye gaze: Preliminary findings. Emotion,16(1), 110–116. https://
doi.org/10.1037/emo0000093
Richard-Devantoy, S., Ding, Y., Turecki, G., & Jollant, F. (2016). Attentional
bias toward suicide-relevant information in suicide attempters: A cross-
sectional study and a meta-analysis. Journal of Affective Disorders,196,
101–108. https://doi.org/10.1016/j.jad.2016.02.046
Rinck, M., & Becker, E. S. (2006). Spider fearful individuals attend to threat,
then quickly avoid it: Evidence from eye movements. Journal of Abnormal
Psychology,115(2), 231–238. https://doi.org/10.1037/0021-843X.115.2
.231
Rogers, M. L., Law, K. C., Houtsma, C., Tucker, R. P., Anestis, M. D., &
Joiner, T. E. (2022). Development and initial validation of a scale assessing
suicide-specific rumination: The Suicide Rumination Scale. Assessment,
29(8), 1777–1794. https://doi.org/10.1177/10731911211033897
Sanchez, A., Vazquez, C., Marker, C., LeMoult, J., & Joormann, J. (2013).
Attentional disengagement predicts stress recovery in depression: An eye-
tracking study. Journal of Abnormal Psychology,122(2), 303–313. https://
doi.org/10.1037/a0031529
Shechner, T., Jarcho, J. M., Britton, J. C., Leibenluft, E., Pine, D. S., &
Nelson, E. E. (2013). Attention bias of anxious youth during extended
exposure of emotional face pairs: An eye-tracking study. Depression and
Anxiety,30(1), 14–21. https://doi.org/10.1002/da.21986
Si, T.-M., Shu, L., Dang, W.-M., Se, Y.-A., Chen, J.-X., Dong, W.-T., Kong,
Q.-M., & Zhang, W.-H. (2009). Evaluation of the reliability and validity of
Chinese version of the Mini-International Neuropsychiatric Interview in
patients with mental disorders. Chinese Mental Health Journal,23(7),
493–503.
Siegle, G. J., Steinhauer, S. R., Carter, C. S., Ramel, W., & Thase, M. E.
(2003). Do the seconds turn into hours? Relationships between sustained
pupil dilation in response to emotional information and self-reported
LI ET AL.
370
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
rumination. Cognitive Therapy and Research,27(3), 365–382. https://
doi.org/10.1023/A:1023974602357
Soleymani, A., Ivanov, Y., Mathot, S., & de Jong, P. J. (2020). Free-viewing
multi-stimulus eye tracking task to index attention bias for alcohol versus
soda cues: Satisfactory reliability and criterion validity. Addictive Behaviors,
100, Article 106117. https://doi.org/10.1016/j.addbeh.2019.106-117
Teasdale, J. D., & Dent, J. (1987). Cognitive vulnerability to depression: An
investigation of two hypotheses. British Journal of Clinical Psychology,
26(2), 113–126. https://doi.org/10.1111/j.2044-8260.1987.tb00737.x
ten Have, M., de Graaf, R., van Dorsselaer, S., Verdurmen, J., van ’t Land,
H., Vollebergh, W., & Beekman, A. (2009). Incidence and course of sui-
cidal ideation and suicide attempts in the general population. The
Canadian Journal of Psychiatry,54(12), 824–833. https://doi.org/10
.1177/070674370905401205
The Center for the Study of Emotion and Attention. (1999). The international
affective picture system: Digitized photographs. The Center for Research
in Psychophysiology, University of Florida Gainesville.
Tsypes, A., Owens, M., & Gibb, B. E. (2017). Suicidal ideation and atten-
tional biases in children: An eye-tracking study. Journal of Affective
Disorders,222, 133–137. https://doi.org/10.1016/j.jad.2017.07.012
Vannoy, S., Gable, S., Brodt, M., Cadet, M., Andrews, B., & Maloney, M.
(2016). An application of eye tracking technology to detect attention bias
for suicide related stimuli [Unpublished manuscript]. Department of
Counseling & School Psychology, University of. Massachusetts Boston.
Wang, Y. A., & Rhemtulla, M. (2021). Power analysis for parameter estima-
tion in structural equation modeling: A discussion and tutorial. Advances
in Methods and Practices in Psychological Science,4(1), Article
2515245920918253. https://doi.org/10.1177/2515245920918253
Wang, Z., Yuan, C.-M., Huang, J., Li, Z.-Z., Chen, J., Zhang, H.-Y., Fang,
Y.-R., & Xiao, Z.-P. (2011). Reliability and validity of the Chinese version
of Beck Depression Inventory-II among depression patients. Chinese
Mental Health Journal,25(6), 476–480.
Wells, T. T., Clerkin, E. M., Ellis, A. J., & Beevers, C. G. (2014). Effect of
antidepressant medication use on emotional information processing in
major depression. American Journal of Psychiatry,171(2), 195–200.
https://doi.org/10.1176/appi.ajp.2013.12091243
Wenzel, A., & Beck, A. T. (2008). A cognitive model of suicidal behavior:
Theory and treatment. Applied and Preventive Psychology,12(4), 189–
201. https://doi.org/10.1016/j.appsy.2008.05.001
Williams, J. M. G., Mathews, A., & MacLeod, C. (1996). The emotional
Stroop task and psychopathology. Psychological Bulletin,120(1), 3–24.
https://doi.org/10.1037/0033-2909.120.1.3
Wilson, T. D. (2009). Know thyself. Perspectives on Psychological Science,
4(4), 384–389. https://doi.org/10.1111/j.1745-6924.2009.01143.x
Woosley, J. A., Lichstein, K. L., Taylor, D. J., Riedel, B. W., & Bush, A. J.
(2014). Hopelessness mediates the relation between insomnia and suicidal
ideation. Journal of Clinical Sleep Medicine,10(11), 1223–1230. https://
doi.org/10.5664/jcsm.4208
Wu, K. C., Chen, Y. Y., & Yip, P. S. (2012). Suicide methods in Asia:
Implications in suicide prevention. International Journal of
Environmental Research and Public Health,9(4), 1135–1158. https://
doi.org/10.3390/ijerph9041135
Received May 11, 2022
Revision received November 2, 2022
Accepted November 7, 2022 ▪
SUICIDAL INFORMATION PROCESSING IN DEPRESSION 371
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
A preview of this full-text is provided by American Psychological Association.
Content available from Journal of Psychopathology and Clinical Science
This content is subject to copyright. Terms and conditions apply.
