FTP
FTP
FTP
AND
Research Article
NEGATIVE BIASES AND RISK FOR DEPRESSION; INTEGRATING SELF-REPORT AND EMOTION TASK MARKERS
Anna J. Watters, B.Psyc. (Hons) and Leanne M. Williams, Ph.D.
Background: Negativity biases and their impact on reactivity to negative emotion are implicated in the mechanisms of risk for depression. The aim of this study was to determine whether self-reported negativity bias is related to objective cognitive measures of emotional reactivity. Methods: A previously established Web selfreport measure of negativity bias was used to assess 1,080 volunteers from the Brain Resource International Database (overseen by the nonprofit BRAINnet Foundation). We identified matched subgroups of High Risk (n 5 216) and Low Risk (n 5 216) participants using a psychometric high-risk method, which classified High Risk as the samples top 30% of negativity bias scores and Low Risk as the bottom 30%. These subsamples also completed the WebNeuro cognitive tasks for assessing both conscious and nonconscious reactions to facial emotions. Task performance was quantified by accuracy, reaction time, and misidentification errors. Results: The High Risk (high negativity bias) subgroup was distinguished by greater reactivity to negative emotion in both conscious and nonconscious processing. The High Risk profile was reflected in higher accuracy for sadness (nonconsciously) and disgust (consciously), and more frequent misidentification of neutral as anger (consciously). Conclusions: These results are consistent with seminal theories that a systematic cognitive negativity bias produces a hyperreactivity to negative emotion, which can impact nonconscious as well as conscious processing. The results provide a step toward objective markers of risk for depression that would help the community act regarding preventative programs. Replication in patient samples is warranted. Depression and Anxiety 28:703718, 2011. r 2011 Wiley-Liss, Inc. Key words: risk markers; depression; emotion; facial expressions of emotion; negativity bias; Brain Research and Integrative Neuroscience Network (BRAINnet)
The authors disclose the following financial relationships within the past 3 years: Contract grant sponsor: ARC-Discovery; Contract grant number: DP0773994; Contract grant sponsor: Pfizer.
Correspondence to: Anna J. Watters, Sydney Medical School, Westmead Millennium Institute, The Brain Dynamics Center, Acacia House, Hawkesbury Rd., Westmead Hospital, Westmead, Sydney, NSW 2145, Australia. E-mail: [email protected]
15%, but substantially more people (up to 30%) are at risk.[1] The World Health Organization ranks depression second only to ischemic heart disease in terms of social and economic burden. Despite its impact, there are no agreed-upon objective markers of risk for depression. Identifying such objective markers would aid in developing prevention and support strategies.
Sydney Medical School, Westmead Millennium Institute, The Brain Dynamics Center, Westmead Hospital, Westmead, Sydney, NSW, Australia
INTRODUCTION
L.M.W. has received consulting fees (Brain Resource Ltd.). L.M.W. is a stockholder and has stock options in Brain Resource Ltd. She has received Advisory Board fees from Pfizer. Received for publication 20 December 2010; Revised 18 May 2011; Accepted 20 May 2011 DOI 10.1002/da.20854 Published online in Wiley Online Library (wileyonlinelibrary.com).
704
In the swiftly emerging integration of psychiatry and neuroscience, the search for endophenotypic markers of risk of depression is escalating rapidly.[26] Seminal theories of depression highlight the cascade of negativity that contributes to the unfolding of depression.[2,7,8] For instance, Beck proposed that a systematic cognitive bias is linked to genetic susceptibility.[2] This results in a greater-than-normal response to negative emotion (especially nonconsciously), both behaviorally and physiologically in the brain. In this study, we extend these theories into an integrative neuroscience framework that focuses on the brains underpinnings in the link between negativity bias and reactivity to negative emotion. The self-reported Negativity Bias measure meets a number of the criteria specified for risk markers[4,9] for depression. Negativity Bias accurately identifies those with Major depression, and more specifically identifies depression as compared to other mental illnesses (Williams et al., In submission). The endogenous nature of Negativity bias, and therefore its relative stability, is evidenced in its correlation with specific genetic variations (e.g. 5HTT-LPR Short allele[1012]), and brain[10,11] and physiological markers.[10,12] Further, these variations and markers are themselves strong candidates for risk of depression. The participants in these cited studies had no current or past diagnosis of depression, which demonstrates that Negativity Bias is present without the onset of depression. The proposed mechanism by which Negativity bias confers risk of depression begins as a genetic predisposition, which interacts with stress to cause an individual to automatically appraise situations in a negative way. This, then, leads to observed behavioural symptoms of depression.[2,10,13] Other studies have found similar constructs for endogenous markers of cognitive bias.[2,14] Results from cognitive emotion tasks have potential as candidate markers for risk of depression. These tasks
provide a more objective indication of automatic bias in appraisal at the level of the emotional brain as compared to reported thoughts. Facial expressions of emotion activate hardwired systems within the emotional brain.[15] These systems are active from infancy[16] and are consistent across cultures.[17] In this study, we address two issues concerning the link between self-reported negativity bias and objective measures of altered reactivity to emotion in risk for depression. First, we review the current findings for objective cognitive emotion tasks in both participants with depression and participants defined as at risk of depression. Second, we present a new empirical study that links cognitive measures of emotion reactivity to risk for depression as defined by negativity bias. In our review of past research, we distinguished studies that used explicit tasks to assess conscious identification of facial emotion from those using implicit tasks that assess more nonconscious recognition of facial emotion. We used the construct of implicit priming to assess the nonconscious influence of emotion. Prior exposure to emotion stimuli has been found to affect responses on later nonemotion tasks like recognition memory.[18] An illustrative example is as follows: participants are first exposed to facial emotion in an explicit way (e.g. asked to identify an emotional facial expression). Later, they complete an implicit task (e.g. recognizing which faces they have seen before from among newly added distractor faces). If there is a nonconscious bias for negative emotion, we may expect prior exposure to negative expressions to speed recognition of those faces in the later recognition task, without people being aware of this influence. In our review, we identified three main trends, which we summarize first for Major depression and second for high-risk samples. Most of the research has focused on conscious emotion reactivity in patients with Major depression (Table 1 for summary; Appendix for details). In these studies, the most robust finding was that
TABLE 1. The number of studies reporting increased, decreased, and no difference in performance of explicit identification of facial emotion in patients with Major depression (MD) compared to Controls (C)a
Measure Accuracy Outcome for MD versus C MD 5 C MDoC MD4C MD 5 C MDoC MD4C MD 5 C MD4C MDoC Total studies 12 19 0 7 4 0 n/a n/a n/a Studies reporting emotions separately 6 14 0 3 3 0 7 8 2 Sad Anger Disgust n/a Fear Neutral Happy
Reaction time
n/a
Bias
n/a
a Symbols o and 4 indicate studies finding patients with MD impaired or facilitated, respectively, relative to controls, for at least one result from Accuracy and Reaction Time measures. For Bias, these symbols indicate the over interpretation (4) and under interpretation (o) of emotions by MD relative to controls. 5 indicates no difference between groups. In studies reporting results for each emotion, the emotions that stood out are indicated by depth of shading (# significant results when that emotion is studied separately / # times emotion studied separately). Note that only three studies ([26,29,30][for bias measures]) fell into more than one outcome category for each measure.
705
patients with depression are less accurate at identifying all emotions (positive, negative, and neutral);[19,20] especially happiness,[2124] sadness,[2325] and neutral.[26,27] One study found that patients with depression had slower reaction times to emotions,[28] with the majority of findings for sadness.[29,30] Few studies have examined nonconscious processing of facial emotion in the context of recognizing facial emotions that the patient previously identified. In one adult study, patients with depression were superior at recognizing sad faces and worse at identifying happy faces.[31] This pattern is consistent with a nonconscious hyper-reactivity to negative emotions which are consistent with the patients mood, and a loss of reaction to positive emotion. A study of adolescent depression suggests that age may play a role in implicit emotion disturbances for fear.[32] These findings raise the question of whether or not there is a bias in facial emotion processing in depression, and in particular whether that bias is consistent with theories for a hyper-reactivity to negative emotions in patients with depression.[2,33] One way of measuring bias has been to examine the pattern of errors (misidentifications) made. Studies have found that in patients with depression, pattern of errors is consistent with a bias toward misidentifying more negative emotion in expressions (particularly more sadness) and less misidentification of happiness[29,34,35] (Table 1; Appendix). While less research has been carried out for patients at high-risk of depression, there is already evidence of altered reactions to emotion that suggest an objective marker for the brain-related risk of developing depression (Table 2). For conscious tasks, high-risk individuals with subclinical depressive symptoms have been reported as being less accurate[36] and quicker[37] at identifying negative and neutral emotions, and other research shows them to be less accurate in identifying happiness.[38] Misidentification rates show that highrisk individuals also have a greater tendency to perceive negative emotion.[36,39] To date, no research has used nonconscious tasks of emotion in high-risk participants. Our empirical study extended the research in high-risk participants. It was the first to include a nonconscious as well as conscious cognitive task of facial emotion, and to assess risk for depression according to the level of Negativity Bias. The cognitive tasks for emotion drew on our theoretical framework known as the INTEGRATE model.[6,60] This framework highlights the operation of two parallel pathways for processing facial emotion,[6,61] one conscious pathway that processes explicit cues of emotion, and one nonconscious pathway that occurs rapidly based on low-level implicit cues of emotion. These parallel pathways relate to different modes of functional brain connectivity.[61,62] Depression has been associated with altered brain activity when facial emotions are presented both consciously[63] and nonconsciously.[64] The current research investigates whether an emotion bias, involving heightened reactivity toward negative
stimuli, is demonstrated in individuals at higher risk for depression (as determined by self-reported Negativity Bias), thereby establishing its potential as an objective indicator of risk for depression. Based on the past findings, we expected the High Risk group to be less accurate and slower in the explicit identification of negative expressions of emotion. We predicted that negative emotion would have the opposing influence on implicit recognition of emotion: an increased accuracy for negative emotion coupled with a complementary faster response time. In addition, we explored whether the High Risk group would show biases in misidentifying one emotion from another.
NEGATIVITY BIAS
Negativity bias captures a predisposition toward hypersensitivity to stress and the expectation of negative outcomes, which elevates risk for depression (Williams et al., In submission).[60,71] The questionnaire called BRISC measures self-reported Negativity Bias across settings and populations. This Negativity Bias measure has been developed using factor analytic techniques and normed across nine decades.[72,73] It has 84.9% sensitivity, 87.6% specificity, 70.7% positive predictive power, and 94.3% negative predictive power for identifying individuals with mental illness (Williams et al., In submission). Sensitivity is highest for depressive disorders. As part of its development, the BRISC has been shown to correlate strongly with scales for depressive symptoms (such as the DASS Depression scale[74]) and with biological indicators of risk for depression.[10,11,60,73] High and Low Risk groups were selected using the BRISC scores of 1,080 individuals from the BRID; the highest 30% of scores made up the High Risk group, and the lowest 30% made up the Low Risk group. These ranges correspond to the optimal margins differentiating high from low risk (Williams et al., In submission). The DASS Depression and Anxiety Scales[74] were included as secondary measures to confirm that groups differed specifically on these symptoms of depression and associated anxiety.
706
TABLE 2. Patterns of accuracy, reaction time and bias in conscious identification (Consc.) and nonconscious recognition (Nonconsc.) of facial emotion tasks by emotion, in participants with Major Depression (MD) versus Controls (C) and High Risk (HR) versus Low Risk (LR) groupsa
Accuracy Emotional stimulus Group/task MD C/Consc. Study S . . . . . . . A . D . . F . N . . . H . . . . . . . . S Reaction time Emotional stimulus A D F N H S Bias Emotional stimulus A D F N H
(41) (42) (22) (24) (28) (27) (32) (21) (25) (26) (43) (44) (45) (46) (47) (23) (48) (30) (49) (50) (51) (31) (20) (29) (52) (53) (54) (55) (56) (57) (58) (59) (60) (35) (36) MD C/Nonconc. (32) (33) HR LR/Consc. (37) (38) (39) (40)
a
. mS mS kH mN mN mS mSkH
. . .
. .
kPOS kPOS mS mS
mS
mS
. . . . . . . . . . . . . . . . . . . . . . . . . . .
. .
. .
mS
mA
mF
kH
mS kH kH
kH
kH
kH
| . . . . .
mS kH
mS mS kH kH mSADF
mS kH
. . .
kH
kH kH
kH
Abbreviations for emotions are; S, Sad; A, Anger; D, Disgust; F, Fear; N, Neutral; H, Happy; NEG, Negative emotions; POS, Positive emotions. Results are presented as the performance of the clinical group relative to the respective control group. For Accuracy and Reaction Time results: , no significant difference; ., a deficit; |, an enhancement. For Bias, results indicate the emotion/s perceived for each emotional stimulus: m(emotion X), increased perception of emotion X; k(emotion X), decreased perception of emotion X; , no difference in perception of emotion. Results based across emotions are indicated by . Where a study reports results on both levels, preference is given to significant results and results given for emotions separately.
battery.[6,75] A total of 96 photographs, consisting of 4 male and 4 female identities showing evoked facial expressions (neutral, happy, sad, fear, anger, or disgust), were selected from a previously established set.[76] These photographs have been standardized according to size, luminance, and central alignment of the eyes, and norms for ages 692 Depression and Anxiety
years have been established.[6,75] In the explicit emotion identification task, the photographs (stimuli) were presented on a computer monitor in a pseudorandom order for 2 sec each. Participants were instructed to identify the emotion depicted on each face by selecting the corresponding emotion label from six options (using a mouse click).
707
The explicit emotion identification task was used as the study phase of the implicit emotion recognition task. The second phase was given after the participant spent 20 min on various other set tasks. Twenty-four of the photographs from the study phase were randomly selected for each participant (two male and two female faces depicting each of the six expressions). These were presented a second time in a new pseudorandom order. This time, each photograph was shown beside one of the 24 new photographs. In each pairing, the new photograph showed an identity different from the repeating photograph, with the new face being of the same gender and depicting the same emotion as the repeating photograph. For each pairing, participants indicated (with a mouse click) which of the two faces they had seen in the previous task. Measures taken were percentage of correct responses and reaction time for both tasks, and misclassifications made for each emotion in the explicit identification task. The High and Low Risk groups were compared using two-tailed t-tests (a 5 .05). In the explicit task, correct identification of emotion was quantified by percentage accuracy and reaction time (for correct answers) for each kind of emotional stimulus. Bias was quantified by incorrect identification. For example, we calculated the percentage of times neutral was misidentified as sad, disgust, fear, anger, or happiness. The family-wise corrected a level was .01. In the implicit task, correct recognition of faces was quantified by percentage accuracy and reaction times (for correct answers) for each kind of emotional stimulus, and grouped according to what type of emotion was influencing recognition. Note that, given there were six response options in the explicit task, the chance of giving the correct response in this task by choosing at random was 16.7%.
time for any emotions (Table 4). For the implicit recognition task, the High Risk group was more accurate at recognizing sad faces (t[430] 5 2.09, P 5 0.037) (Table 5), yet slower at responding to them (t[430] 5 2.22, P 5 0.027) (Table 6). Bias was examined by looking at the average percentage of each kind of erroneous response to each kind of emotional stimulus (e.g. percentage of presentations for which neutral was labeled as sad). The
TABLE 4. Reaction Time (ms7SD) of explicit identification of facial expressions in participants at High Risk and Low Risk for depression
Group High risk Sad Anger Disgust Fear Neutral Happy 2,369.57735.8 2,577.07846.0 2,906.971,366.5 3,117.671,028.9 1,974.27662.9 1,572.87359.4 Low risk Significance tdf 430 P 0.057 0.870 0.723 0.511 0.059 0.474 Effect size r .090 .008 .017 .032 .090 .034
2,532.871,044.8 1.91 2,563.37893.5 0.16 2,950.271,169.1 0.35 3,184.971,099.7 0.66 1,864.97536.1 1.89 1,596.87335.5 0.72
RESULTS
High (M 5 1.22, SD 5 0.91) and Low Risk (M 5 0.91, SD 5 0.20) groups were significantly different on Negativity bias Z-scores (t[430] 5 33.43, Po0.001). Groups also differed on DASS Depression (t[430] 5 19.57, Po0.001) and Anxiety (t[430] 5 16.40, Po0.001) scores; we report means for High (Depression: M 5 7.01, SD 5 4.64; Anxiety: M 5 4.63, SD 5 3.55) and Low risk (Depression: M 5 0.71, SD 5 0.96; Anxiety: M 5 0.55, SD 5 0.89) groups. For explicit identification, the High Risk group was more accurate at recognizing disgust (t[430] 5 1.97, P 5 0.049) and less accurate at recognizing neutral (t[430] 5 2.57, P 5 0.011) (Table 3). There were no differences between groups in reaction
TABLE 3. Accuracy (%7SD) of explicit identification of facial expressions in participants at High Risk and Low Risk for depression
Group High risk Sad Anger Disgust Fear Neutral Happy 70.1722.5 58.7716.6 46.5717.8 81.6718.1 77.6718.4 98.674.3 Low risk 68.6723.5 57.9715.1 43.2716.4 80.0718.1 81.8715.7 98.674.5 Significance tdf 430 0.69 0.53 1.97 0.94 2.57 0.00 P 0.493 0.596 0.049 0.348 0.011 1.000 Effect Size r .033 .025 .096 .044 .122 .000
TABLE 5. Accuracy (%7SD) of implicit identification of facial expressions in participants at High Risk and Low Risk for depression
Group High risk Sad Anger Disgust Fear Neutral Happy 98.776.5 96.679.8 99.374.1 100.0700.0 95.8710.2 93.7710.9 Low risk 97.178.7 96.679.5 98.876.2 99.772.9 96.179.7 93.3711.7 Significance tdf 430 2.09 0.00 0.91 1.74 0.32 0.354 P 0.037 1.000 0.370 0.083 0.748 0.724 Effect size r .104 .000 .048 .073 .015 .018
TABLE 6. Response Time (ms7SD) of implicit identification of facial expressions in participants at High Risk and Low Risk for depression
Group High risk Sad Anger Disgust Fear Neutral Happy 1,506.67378.5 1,439.47328.8 1,354.87311.3 1,304.47277.4 1,559.97363.1 1,552.17420.8 Low risk 1,433.47301.2 1,447.27336.3 1,342.67306.3 1,273.27257.5 1,554.57359.4 1,515.97368.0 Significance tdf 430 2.22 0.24 0.41 1.21 0.16 0.95 P 0.027 0.808 0.681 0.226 0.876 0.342 Effect size r .106 .057 .020 .058 .007 .046
708
Figure 1. Explicit labeling responses to neutral stimuli by High (HR) and Low Risk (LR) groups by emotion label. (significant difference between groups at the .05 a level).
means of High and Low Risk groups were compared using t-tests (two-tailed, a 5 .023). The High Risk group misidentified neutral faces as angry more often than the Low Risk group (t[430] 5 2.81, P 5 0.005; effect-size r 5 .13) (Fig. 1). An alternative way to calculate accuracy and bias draws on signal detection theory (SDT).[24,32] Applied to the current paradigm, these calculations would improve mathematical interpretation of the constructs by additionally taking into account how well the target expression is distinguished from other expressions. We did not use them here as our paradigm involves a multiple category decision which we presume involves qualitatively different decision criteria and a correct hit and false alarm rate that do not hold all the response information, as is assumed in SDT. Instead, we used percentages of correct and incorrect responses to describe the data in accordance with the majority of studies,[25,31] with a view toward developing a more suitable formula.
DISCUSSION
The current study examined whether individuals at greater risk for depression (as determined by a high Negativity Bias) show increased reactivity in conscious and nonconscious emotion processing. High Risk individuals were more accurate at consciously identifying
Depression and Anxiety
disgust and nonconsciously recognizing sad faces compared to Low Risk individuals, and they had a greater tendency toward interpreting neutral faces as anger. The outstanding trends in the literature on conscious identification in patients with depression and in those at high risk of depression show overall deficits; specifically for happiness, sadness, and neutral; and a negative emotion bias (i.e. a bias toward perceiving more sadness and negative emotion in faces, and less happiness). The High Risk group was defined by a distinctive profile of responses to negative emotion that operated both explicitly (consciously) and implicitly (nonconsciously). Explicitly, the High Risk group had a greater accuracy for identifying disgust, and a greater response bias in favor of identifying neutral as anger. These results are in agreement with our definition of explicit biases to negative emotion. Implicitly, in the same High Risk group, prior exposure to sad faces slowed down their later recognition of these faces. This effect is defined as an implicit bias that gauges the nonconscious influence of emotion. Taken together, explicit effects for disgust and anger are consistent with a bias for threat-related negative emotion. The bias regarding anger was small in effect size; however, this may contribute some meaningful effects regarding states of risk in depressive conditions. One possibility is that threat-related biases reflect anxiety and a hypervigilance to fear cues. A bias in favor of anger would be consistent with an individuals mood when experiencing anxiety. The High Risk group in this study had elevated levels of anxiety on the DASS. Anxiety is an established risk factor for depressive disorder, and is highly comorbid with it.[77] Implicitly, we saw a different kind of bias to sad emotion. Previous studies of depressive disorder have highlighted a bias in favor of explicit identification of sad emotion, which is consistent with the defining mood symptoms of depression.[27,36] In our High Risk group, prior exposure to sad faces had the implicit effect of slowing down responses regarding the later recognition of these faces. This slowing of recognition was in the absence of a bias toward explicit identification of sad. It is possible that in high-risk states, hypersensitivity toward sadness remains nonconscious and disrupts the task at hand. A more explicit bias to identifying sad may occur with the onset of diagnosable depressive symptoms. Future research is needed to not only test the generalizability of this proposal but also to explore whether different biases to negative emotion characterize the unfolding of risk into overt illness. In the general population, poorer accuracy on emotion tasks is typically associated with slower reaction time.[6,27] In this study, the High Risk group showed a dissociation of higher accuracy with a slower reaction time in the implicit recognition of sad, which we take to suggest a specific emotion bias.
709
These results are consistent with the same dissociation in the implicit recognition of sad for patients with Major depression, representing a mood-congruent bias.[31] Investigations using other brain measures may show whether slower reaction time is related to greater activation. Our research contributes new findings regarding emotion bias as a risk marker for depression. The differences characterizing High Risk individuals were consistent with a heightened sensitivity to negative emotion. Importantly, these High Risk individuals had not experienced Major depression previously, which indicates that any differences found existed prior to the patient reaching the threshold for Major depression. Since High and Low Risk patients also differ on Depression and Anxiety symptoms, it is possible that even current subclinical symptoms contribute to the emotion differences found between the groups. This, however, does not undermine the influence of Negativity Bias, a predispositional factor, on both current symptoms and emotion bias. It may have a more long-term effect. Further research is underway to investigate the relationship between emotion bias and cognitive functioning, as well as the role early life stress has in increasing risk and its prevalence in those at familial risk. In the INTEGRATE framework, emotion bias is proposed to be an early stage along the pathway to depression.[78] This framework integrates evidence that individuals differ in genetically based predisposition for Negativity bias in self-regulation, which can be exacerbated by stress,[1012,79] and proposes that this leads to changes in the brain that cause automatic biases to emotionally important stimuli.[80] These changes become exacerbated in depression and further problems with cognitive processes are developed. Changes in the emotional brain are central to depression, which is defined by lowered mood, anhedonia, and associated anxiety. Current research is elucidating the interconnections between brain mechanisms for depression and emotion.[15,64,81] This study is the first step in exploring emotion bias as a vulnerability marker for depression. The clinical implications of this studys results support the possibility of biases in fundamental emotion pathways congruent with depression and anxiety. It may be beneficial to understand these changes in the emotion pathways that accompany cognitions. Biases in favor of specific emotions may be one specific shift that determines risk for a depressive illness. These biases, when added to more universal factors that have larger effect sizes (e.g. age, stress exposure) may operate to increase the likelihood of developing the overt illness. The value of these biases as markers is that they can supplement a diagnostic approach based on the assessment of symptoms.[2,8183] They can also be used to identify individuals who are most vulnerable to the disorder before their symptoms develop, and to engineer interventions that target underlying emotion pathways.
The effect sizes of the differences found in the High Risk group were small and we interpret them as indicative results. Our context is that small effects may still have large contributions in terms of acting as susceptibility factors in contributing to the development of the overt illness. One limitation of this study is the high accuracy shown by both groups for the identification of happiness and recognition of fear. This would have decreased sensitivity to detecting a difference for positive emotion. A limitation with any risk study is the lack of confirmation regarding the likelihood of High Risk individuals developing the illness. A follow-up stage to assess subsequent depressive symptoms would be of benefit. Future studies are needed to consolidate our findings of emotion bias in high-risk patients. The use of concurrent measures such as fMRI and EEG in the same participants would help to elucidate the mechanisms involved. The addition of cognitive measures would also test whether markers were unique to emotion processes. Future studies may also identify High Risk participants according to multiple factors, including genetic, Negativity Bias, and family history; to further understand the impact of these factors on emotion bias and risk for depression.
CONCLUSION
This study investigated emotion processing in risk for depression. Individuals with a dispositional tendency toward poor self-regulation, and without a history of depression, showed characteristic changes in conscious and nonconscious processing of particular negative emotions. These changes are interpreted in the context of the INTEGRATE framework as heightened cognitive sensitivity to negative stimuli, characterizing a vulnerability to depression. Acknowledgments. This project is supported by an ARC-Discovery grant (DP0773994). L.M.W. was supported by a competitive, peer reviewed Pfizer senior research fellowship. The authors acknowledge use of data from the Brain Resource International Database (BRID), under the auspices of the Brain Resource Co. Access to BRID data for scientific purposes is administered independently via the scientific network, BRAINnet (www.BRAINnet.net) and is not subject to constraint. We thank the individuals who gave their time to take part in the study. We also acknowledge the editorial support of Jon Kilner, MS, MA (Pittsburgh, Pennsylvania).
APPENDIX
Detailed summaries of research studies on conscious identification and nonconscious recognition of facial emotion in major depression and risk for depression is given in Table A1.
Depression and Anxiety
710
TABLE A1. Detailed summaries of research studies on conscious identification and nonconscious recognition of facial emotion in major depression and risk for depression
Participants Emotions: Expressions used in stimuli; Sad (S), Anger (A), Disgust (D), Fear (F), Happiness (H), Neutral (N), Surprise (Sur), Interest (Int)(expressions not included in the analysis appear in parenthesis) Stimuli: Total number of stimuli (Sourcesee reference article for full details); Number of identities, emotions, presentations, intensities, gender; Presentation time Task: What participants were asked to do (identify, match, rate) Methods Results Reported in terms of clinical group versus controls. Accuracy (Ac): Percentage or number correct/discrimination accuracy/ intensity threshold Reaction Time (RT): Bias: Misidentifications/judgment ratings Results for all emotions together are indicated by "overall"
Reference
Depression and Anxiety Ac: No differences for any emotion Emotions: S, A, D, F, H, Sur Stimuli: Sixty pairs of posed, black and white photographs (Ekman & Friesen, 1976); 10 identities 6 emotions 2 presentations (target/distracter) Task: Asked which face expresses a named emotion Emotions: S, A, F, H, N Stimuli: Thirty-two pairs from 48 posed, black and white photographs[52]; pairs consisted of the same and different emotional expressions; presentation time 5 1 sec Task: Judge whether emotions are the same or different Ac: Depressed participants were less accurate overall Ac: No differences for any expression Emotions: D, F, H, (N) Stimuli: One hundred and thirty-two color photographs; 4 identities (2F12M) 3 emotions 11 levels of intensity (neutral to 100%); presentation time 5 400 msec. Task: Identify which of the four expressions is depicted Emotions: S, A, D, F, H, Sur, N Stimuli: One hundred and twenty pairs from 60 posed, black and white photographs (Ekman, 1976); 10 identities (5F12M) 6 emotions; pairs of same and different emotions. Task: Judge whether emotions are the same or different Ac: No difference overall RT: Depressed participants were slower overall Emotions: S, A, D, F, H, Sur, N Stimuli: Thirty-five color presentations of a virtual human face (Kiss et al., 2004); 6 emotions 5 intensity levels, 15 neutral expressions Task: Identify which of the seven emotions is depicted Emotions: S, A, D, F, H, Sur, N Stimuli: Twenty-one posed, black and white photographs (Ekman & Friesen, 1976); 3 Ac: Depressed participants were less accurate overall, specifically for Neutral, and for Sadness and Disgust at low intensities Ac: Depressed participants were less accurate overall
Study reference
Depression group (D): Number of participants (number of female participants), mean age (standard deviation) Measures of depression used and scores (SD/range) Number of medicated participants Diagnosis information Control group (C): High Risk (HR): Low Risk (LR): Major depression versus controls/conscious identification Archer et al.[47] D: 12(9), 56.19 (half over 60 years) DSM-III[84] All on antidepressants C: 12(6), 48.04
Asthana et al.[51]
Bediou et al.[49]
D: 25(0), 29.08(7.25) DSM-IV[85], BDI (Hindi version)[86] 33.4 Medication not reported C: 25(0), 25.65(6.64) BDI 9.5 D: 20(7), 38.6(10.3) DSM-IV[85], MADR[87] 23.53(8.61) Nineteen on antidepressants C: 20(7), 25(9.47)
Csukly et al.[36]
Feinberg et al.[19]
D: 21(15), 24.67(8.14) BDI[88] 19.95(5.82), MMPI D[89] 79.05(11.42) 0 on antidepressants C: 25(16), 22.16(6.28) BDI 3.50(2.87), MMPI D 47.84(5.63) D: 23(14), 48.1(13.0) All medicated DSM-IV[85], BDI[88] 25.6(8.0) C: 23(14), 48.6(11.9) BDI 3.6(2.6) D: 20(9), 36.7 Half medicated RDC[90] C: 20(9), 29.8
D: 12(7), 43.3(11.2) Eight on antidepressants DSM-IV[85], HDRS[91] 17.5(4.4) C: 12(7), 41.6(12.3) Ac: No difference overall
Gessler et al.[57]
Ac: No difference overall Bias: No difference in bias scores towards Happiness or Sadness
Gilboa-Schechtman et al.[44]
identities 7 expressions; presentation time 5 500 msec. Task: Identify which of the seven expressions is depicted Emotions: S, A Stimuli: Twenty-four presentations of a target and two alternative matches; photographs (Ekman, 2003); presentation time 5 5.3 sec. Task: Match target to one of two alternatives Emotions: S, A, D, F, H, Sur Stimuli: Twelve still videos; 2 identities (1M, 1F) 6 emotions; photographs (Ekman and Friesen, 1976); presentation time 5 8 sec each Task: Identify which of the six emotions is depicted Emotions: S, H Stimuli: Forty photographs from newspapers, reliably rated as happy or sad; each presented for 5 sec. Task: Identify which of the two emotions is depicted Emotions: Negative (S, A), positive (H) Stimuli: Five hundred and seventy-six trials using 72 photographs (Halberstadt and Niedenthal, 1997); 4 identities 3 emotions 6 intensity levels; The inter-trial interval was 750 msec. Task: Identify expressions as positive, negative or neutral
Gollan et al.[29]
Emotions: S, Harsh(A, D, F), H, Sur and (N) Stimuli: One hundred and ten photographs of facial expressions (Ekman and Friesen, 1976); 17 angry, 15 disgust, 15 fear, 18 happy, 14 neutral, 17 sad, 14 surprise; presented for 3,000 msec Task: Identify which of the six emotions is depicted Emotions: S, H, N Stimuli: Approximately, one hundred and twenty-seven presentations in each of two tasks (Erwin et al., 1992); photographs using 25M and 25F identities; Task 1: Sixty-three sad and 64 neutral; Task 2: Sixty-three happy and 64 neutral; presentation time is 7 sec Task: judge each face on a 17 scale of Neutral to Sad or Neutral to Happy
Ac: Depressed participants were less sensitive to Anger RT(correct): No differences for any emotion (data for Sadness not included) Bias: Depressed participants misidentified low intensities of Sadness and Anger as negative more often Ac: No differences for any emotion RT: Depressed participants were slower for Sadness Bias: Depressed participants misidentified neutral as sad more often and as happy less often Ac: No differences for any emotion Bias: Depressed participants tended to judge Neutral faces as Sad, and Happy faces as Neutral
Hale[34]
Emotions: Negative (S, A, D, F, Rejection) Positive (H, Invitation) Stimuli: Twelve schematic faces; various nonspecified emotions including three ambiguous faces (Bouhuys et al., 1995, 1996) Task: Judge each face for each emotion on a 15 scale Emotions: Negative (S, A, D, F, Rejection) Positive (H, Invitation) Stimuli: Twelve schematic faces; various nonspecified emotions including three ambiguous faces (Bouhuys et al., 1995, 1996) Task: Judge each face for each emotion on a 15 scale
Bias: Depressed participants tended to judge all expressions as Sad, and ambiguous expressions as negative
Hale et al.[35]
D: 21(9), 39.0 Unmedicated RDC[90] C: 15(6), 31.0 D: 20(14), 43.8(10.1) Medication not reported RDC[90] C: 20(10), 29.7(8.4) D: 40(24), 29.48(8.84) Medication not reported Depressed or dysthymic SCID[92] (21 had co-morbid disorders), BDI[88] 22.11(8.07) C: 65(34), 27.64(6.28) SCID, BDI 4.25(3.55) D: 37(18), 1854 Unmedicated SCID[92], BDI-II[93] 32.7(8.65), HRSD[91] 20.7(5.14) C: 29(14), 1854 BDI o5 D: 14(12), 44.5(12.9) Twelve medicated Nine with bipolar/psychotic depression DSM-III-R[94], HRSD[91] 25.5(4.7) 1835, BDI[88] 26.4(8.9) C: 14(12), 36.6(16.32) BDI 4.1(1.1) D: 48(32), 38(12.5) Twenty-eight medicated Outpatients, DSM-IV[85], BDI[88] 27.8(7.1) C: 47(31) 41(12.3) BDI 2.7(2.6) D: 20(11), 41(10.9) Fourteen medicated DSM-IV[85], BDI[88] 28(5.6) C: 19(9), 40 (11.3) BDI 2.6(3.1)
Bias: Depressed participants tended to judge; all expressions less positively; less Invitation in nonambiguous expressions; and less Happiness in ambiguous expressions
711
712
Reference
Depression and Anxiety Ac: Depressed participants were less accurate for Happiness at low intensities Ac: No differences overall or specifically Emotions: S, A, H (F) Stimuli: Forty sequences of a neutral face merging to an expression (Young et al., 2002) in 2% increments; 2 identities(1M11F) 4 emotions 5 presentations; presentation time 5 500 msec Task: Identify which of four emotions is expressed Emotions: S, A, D, F, H, Sur Stimuli: Video footage of actors depicting six emotive facial expressions in neutralemotionneutral changes. Emotional expressions lasted for 2 sec. Task: Identify which of six emotions is depicted Emotions: S, A, F, H, (N) Stimuli: Photographs (Ekman and Friesen, 1976) and some chimeric faces (Bouhuys et al., 1999), presentation time 5 300 msec Task: Identify which of the four emotions is expressed. Emotions: S, H, N Stimuli: Ninety-six photographs (Ekman and Friesen, 1976); 8 identities (4M14F) 3 emotions 3 presentations; presented time 5 200 msec Task: Identify which of three expressions is depicted Emotions: S, H, (N) Stimuli: Fourteen posed photographs; 2 emotions 3 intensity levels 2 gender12 neutral Task: Rate expressions along an 11-point scale from sadneutralhappy Ac: Depressed participants were less accurate overall RT: No differences overall Bias: Depressed participants misidentified emotions as Happy more often. No differences in kinds of responses to neutral. Ac: Depressed participants were less accurate for Neutral RT: Depressed participants were slower overall Bias: Depressed participants misidentified Neutral as Sad more often. Bias: No difference in ratings of each expression Emotions: S, A, D, F, H, Sur Stimuli: Twenty-four posed photographs; 6 emotions 2 identities 2 presentations Task: Identify which of the six emotions is expressed Emotions: S, A, D, F, H, Sur Stimuli: Six photographs (Ekman and Friesen, 1976), one of each emotion Task: Identify which of the six emotions is depicted Ac: Depressed participants were less accurate overall Ac: Depressed participants were less accurate at Happiness and Surprise Bias: Depressed participants misidentified emotions as Sad more often
Harmer et al.[21]
Kan et al.[48]
D: 33(18), 37(12.0) Unmedicated SCID[92], BDI[88]; 24.9(6.7), HAM-D[91]; 18.5(3.7) C: 21(6), 37(12.5) BDI; 1.3(1.9), HAM-D 0.3(0.8) D: 21(18), 33.8(10.70) 11 medicated SCID[92], BDI[88] 30.44(10.63) C: 25(17), 31.64(8.67) BDI 2.8(4.25) D: 16(7), 50.9(12.35) All medicated DSM-IV[85], HDRS[91] 18.3(8.64) C: 17(7), 59.0(14.7)
Langenecker et al.[55]
D: 21(21), 32.2(9.7) Seventeen medicated BDI[88]; 19.0(11.8), HDRS[91] 13.3(5.3) C: 20(20), 29.5(8.6) BDI; 1.7(1.8), HDRS; 0.9(1.2)
Leppanen et al.[27]
Loeb et al.[50]
Mandal[52]
D: 18(11), 45.1 (9.9) All medicated ICD-10[85]; BDI[88] 36.8(9.6) C: 18(11), 44.7 (9.9) BDI 11.1(8.4) Psychiatric patients (predominantly schizophrenic): Medication not reported BDI[88] D: 20(0) Non-D: 22(0) D: 40(18), 29 All medicated DSM-III[84] C: 50(20), 24.6 D: 25(10), 26.4 Medication not reported : 25(11), 25.3
Matthews et al.[58]
Emotions: S, F, H Stimuli: Nine photographs; 3 emotions 3 lighting effects (red, blue, color) Task: Comment freely on each stimulus Emotions: A, F, H Stimuli: Three blocks of six trials for each of 3 emotions; a target photograph and two alternatives; presentation time 5 5 sec Task: Match target to one of two alternatives
Ac: Depressed participants were less impaired at producing words in response to Sad Ac: No difference overall RT: No difference overall
Mendlewicz et al.[43]
Ac: Depressed participants were less accurate for Anger Bias: No differences in intensity ratings of emotions
Mikhailova et al.[23]
Ac: Depressed participants were less accurate for Happiness and Sadness
Nandi et al.[42]
D: 30(12), 30.4 Medication not reported C: 30(15), 32.0 D: 12(3), 24.5(5.5) Unmedicated DSM-IV[85], BDI-II[93] Z15(1543) C: 10(6), 24.3(5.0) BDI-II r5(02) D: 21(21), 17.3(2.2) Eighteen medicated HDRS[91] 20.9(3.3) C: 32(32), 21.2(2.1) HDRS 1.3(1.9) D: 18(0), 21.3(1.7) Unmedicated DSM-III-R[94]; HDRS[91] 22.86(2.17); BDI[88] 20.80(2.18) C: 16(0), 25.4(2.2) D: 20(10), 30.7(6.8) Unmedicated RDC[90] Uni and bipolar C: 20(20), 26.7(6.2) Ac: No differences for any emotion, (but depressed participants were less accurate at determining the intensity of sadness)
Emotions: S, A, F, D, H, (N) Stimuli: Forty photographs (Hess and Blairy, 1995; Matsumoto and Ekman, 1998); 5 emotions 2 identities 4 intensities (neutral100%) Task: Identify which of the five emotions is depicted and judge the intensity on a scale from 1 to 7 Emotions: S, H, N Stimuli: Sixty trials of drawings of facial expressions of five identities; (2 emotions 2 intensities1 neutral) 12 presentations; presentation time 5 80 msec Task: Identify which of the three expressions is depicted Emotions: S, A, D, F, H Stimuli: Photographs of five emotions at 5 intensity levels (Saha, 1973); the five photographs for each emotion were paired with each other to give 10 pairs, and 50 in total Task: Identify which of the five emotions is expressed in the 25 stimuli and judge which is more intense in a pair Emotions: S, A, D, F, H, Surprise, Indifference Stimuli: Fourteen photographs (Ekman, 1976); 7 emotions 2 identities (1M1 1F); no time limit specified Task: Identify which of the seven emotions is depicted
Ac: Both depressed groups were less accurate than the student controls, but not the patient controls, for each emotion
Ridout et al.[31]
Emotions: S, H, N Stimuli: Twenty-one photographs (LeGal & Bruce, 1999); 5 Happy, 5 Sad and 11 Neutral; presentation time 5 500 msec Task: Identify which of the three expressions is depicted Emotions: S, A, D, F, H, Surprise, Interest Stimuli: Forty-eight photographs of seven different facial expressions (Kolb & Taylor, 1981) Task: Identify which of the seven emotions is expressed by matching the photograph to one of seven key photographs Emotions: S, H, (N) Stimuli: Photographs using 10 identities (Young et al., 2002); (2 emotions 2 intensities110 Neutral) 2 presentation durations (100 msec, 2,000 msec) Task: Identify which of the three expressions is depicted
Ac: No differences for any expression RT: No differences for any expression Bias: No differences in misidentifications for any expression Ac: Depressed participants were less accurate overall and specifically for Sadness and Interest Bias: No differences in the frequency of the selection of each emotion
Surguladze et al.[24]
Walker et al.[56]
STUDENTS (1853 years): D: 16(16) BDI[88] 16.19(1026) C: 16(16) BDI 3.75(18) PSYCHIATRIC INPATIENTS (1855 years): D: 16(16), BDI Z22 C: 11(11), BDI r21 D: 16(9), 38.1(11.6) All medicated ICD-10[95]; BDI[88] 18.8(7.18) C: 16(9), 38.5(11.5) BDI 1.69(1.45) D: 17(15), 39(10.9) Unmedicated DSM-III[94]; RDC[90]; BHDS[96] 7 Seven bipolar I, 5 bipolar II, 5 unipolar C: 31(20), 31(10.7) D: 27(15), 46.9(11.2) All medicated DSM-IV[85], BDI[88] 33.0(9.9); HDRS[91] 16.9(5.5) C: 29(17), 43.0(11.9) BDI 3.1(3.5) D(Affective disorder): 14(6), 31(10.2) 11 medicated RDC[90] 7 schizoaffective, 7
Emotions: S, A, F, Shame, H, Sur Stimuli: Sixteen photographs (Izard, 1971, 1977); 3(A, Shame,
Ac: Depressed participants were less accurate for Happiness and Sadness; at 100-msec duration Bias: Depressed participants identified Happy and Neutral as Happy less oftenat 50% intensity/ 2,000 msec Ac: No differences overall RT: No differences overall
713
714
Reference
Depression and Anxiety Ac: Depressed participants were less accurate overall and for each expression H, Surprise)12(S, F) Task: Identify which of the six emotions is depicted Emotions: S, A, D, F, H, Surprise, N Stimuli: Sixteen photographs (Ekman & Friesen, 1976); 2 identities (1M, 1F) 8 expressions (including two intensities of Happiness) Task: To sort into any kind of groups Emotions: S, A, F, H, (N) Stimuli: Ten photographs (Young et al., 2002); 2 identities [1M, 1F] 5 expressions; two kinds of pairs using the same identityneutral/emotion and happy/negative emotion Task: Judge the more intense emotion RT: Depressed participants were slower for Sadness in Neutral-Emotion pairs. Bias: Depressed participants tended to rate Negative emotions as more intense than Happiness, and Happiness as less intense relative to Neutral Ac: Depressed participants were less accurate overall Bias: No differences in interpretations of positive or negative emotions Emotions: Negative (S, A, D, F, Shame) Positive/Neutral (H, Surprise, Interest) Stimuli: Thirty-two photographs (Izard, 1971); 4 identities (M and F) 8 emotions; medium intensity Task: Identify which of the eight emotions is depicted, Ac: Depressed participants were less accurate for Fear Emotions: A, F, H Stimuli: Twenty-four photographs at encoding (3 emotions 8 of each emotion) each of unique identities. Forty-eight photographs at recall: 24 identities at encoding1 24 new identities, all with neutral expressions Task: Answer one of three evaluative questions. Memory recall; indicate whether the picture has been seen before Emotions: S, H, N Stimuli: Twenty-one photographs (LeGal & Bruce, 1999); 5 Happy, 5 Sad and 11 Neutral. Recall phase: The first set plus a second set using different identities Task: Identify which of the three expressions is depicted. Memory recall; indicate whether the picture has been seen before Emotions: S, F, H, (N) Stimuli: Two hundred and ten trials: Three binary response trials (H/S, H/F, S/F) 7 stimuli (2 emotions corresponding to trial type 3 intensities1 neutral) 10 sets (3 emotions 3 intensities1neutral); presentation time: 2,500 msec Task: Identify which of the two response options is depicted Ac: Depressed participants were more accurate at Sad but less accurate for Happiness RT: Depressed participants were slower for all expressions RT: No difference for any expression (and so not reported) Bias: HR participants selected the Happy response less often in H/S and H/F trials
Weniger et al.[40]
major C: 14(7), 25(6.9) D: 21(21), 35(9) All medicated SCID[97], HRSD[91] 22(5); BDI[88] 26(11) C: 30(15), 39(11)
Yoon et al.[30]
D: 21(14), 36.63(12.11) Medication not reported DSM-IV[85]; BDI-II[93] 28.05(13.67) C: 22(16), 30.35(8.30) BDI-II 2.7(4.43)
D: 15(15), 37 14 Medicated DSM-III[84]; BDI(short)[98] 14.3, Z9; D-30[99] Z70 Seven dysthymia, 5 MDD, 3 adjustment disorder with depressed mood C: 15(15), 37 BDI: 1.8, D-30: 44.8(o70) Major depression versus controls/nonconscious recognition Pine et al.[32] D:19(13), 16.4(2.4) C:133(71), 14.8(2.9) Parent As Respondent Informant Schedule (PARIS)
Ridout et al.[31]
D: 16(9), 38.1(11.6) All medicated ICD-10[95]; BDI[88] 18.8(7.18) C: 16(9), 38.5(11.5) BDI 1.69(1.45)
High risk versus low risk for depression/conscious recognition Arce et al.[39] RESILIANCE: CD-RISC[100] Low resilience: 22(19), 18(1.9) BDI 16.6(8.5); Neuroticism 56.3(8.2) High resilience: 15(10), 18(0.5) BDI 31.7(4.8); Neuroticism 42.3(8)
Chan et al.[38] Emotions: S, A, D, F, H, Sur, (N) Stimuli: Two hundred and fifty photographs of ten individuals (Young et al., 1997); 6 emotions 4 examples 10 intensities 1(10 identities N); presentation time 5 500 msec. Task: Identify which of the six emotions is depicted Emotions: S, A, D, F, H, Sur, N Stimuli: Fourteen computer generated faces; 7 expressions 2 presentations; presentation time 5 30 sec Task: Identify which of the seven expressions is depicted
Csukly et al.[36]
Ac: No difference overall. HR participants were less accurate at Happiness at low intensities RT: No difference for each emotion Bias: No difference in misidentifications of expressions Ac: HR participants were less accurate overall and for Neutral Bias: HR participants misidentified Neutral as a Negative emotion more often
Le Masurier et al.[37] Emotions: S, A, D, F, H, Sur, N Stimuli: Two hundred and fifty photographs of ten individuals (Young et al., 1997); 6 emotions 4 examples 10 intensities 1(10 identities N); presentation time 5 500 msec Task: Identify which of the six emotions is depicted
NEUROTICISM (N): EPQ[101] High N: 33(22), 18.82(0.98) EPQ 9.58(812); BDI[88] 8.33(5.97) Low N: 42(18), 19.06(0.88) EPQ 1.25(03); BDI 2.59(2.67) DEPRESSIVE SYMPTOMS: Thirteen depression items from SCL-90[102] Healthy participants: 117(68), 25.3(11.4) SCL-90(Dep) 0.78(0.67) FIRST DEGREE RELATIVES of patients with depression: FHRDC[103] Relatives: 25(25), 39.8(13.9) BDI[88] 3.8(4.4); EPQNeuroticism[101] 6.8(3.7) C: 21(21), 39.1(13.5) BDI 1.6(1.8); EPQ-Neuroticism 6.4(4.8) Ac: No differences overall or specifically. RT: HR participants were faster for Fear
716
REFERENCES
1. Blazer DG, Kessler RC, McGonagle KA, Swartz MS. The prevalence and distribution of major depression in a national community sample: the National Comorbidity Survey. Am J Psychiatry 1994;151:979986. 2. Beck AT. The evolution of the cognitive model of depression and its neurobiological correlates. Am J Psychiatry 2008;165:969977. 3. Gottesman II, Gould TD. The endophenotype concept in psychiatry: etymology and strategic intentions. Am J Psychiatry 2003;160:636645. 4. Hasler G, Drevets WC, Manji HK, Charney DS. Discovering endophenotypes for major depression. Neuropsychopharmacology 2004;29:17651781. 5. Venn HR, Watson S, Gallagher P, Young AH. Facial expression perception: an objective outcome measure for treatment studies in mood disorders? Int J Neuropsychopharmacol 2006;9: 229245. 6. Williams LM, Mathersul D, Palmer DM, et al. Explicit identification and implicit recognition of facial emotions: I. Age effects in males and females across 10 decades. J Clin Exp Neuropsychol 2009;31:257277. 7. Abramson LY, Seligman ME, Teasdale JD. Learned helplessness in humans: critique and reformulation. J Abnorm Psychol 1978; 87:4974. 8. Gotlib IH, Joormann J. Cognition and depression: current status and future directions. Annu Rev Clin Psychol 2010;6:285312. 9. Ingram RE, Siegle GJ. Contemporary methodological issues in the study of depression: not your fathers Oldsmobile. In: Gotlib IH, Hammen CL, editors. Handbook of Depression. New York: Guilford Press; 2002:86114. 10. Williams LM, Gatt JM, Schofield PR, et al. Negativity bias in risk for depression and anxiety: brain-body fear circuitry correlates, 5-HTT-LPR and early life stress. Neuroimage 2009;47:804814. 11. Williams LM, Gatt JM, Grieve SM, et al. COMT Val(108/ 158)Met polymorphism effects on emotional brain function and negativity bias. Neuroimage 2010;53:918925. 12. Gatt JM, Nemeroff CB, Schofield PR, et al. Early life stress combined with serotonin 3A receptor and brain-derived neurotrophic factor valine 66 to methionine genotypes impacts emotional brain and arousal correlates of risk for depression. Biol Psychiatry 2010;68:818824. 13. Gatt JM, Clark CR, Kemp AH, et al. A genotype-endophenotype-phenotype path model of depressed mood: integrating cognitive and emotional markers. J Integr Neurosci 2007;6: 75104. 14. Wichers M, Myin-Germeys I, Jacobs N, et al. Genetic risk of depression and stress-induced negative affect in daily life. Br J Psychiatry 2007;191:218223. 15. Williams LM, Gordon E. Dynamic organization of the emotional brain: responsivity, stability, and instability. Neuroscientist 2007;13:349370. 16. Grossmann T. The development of emotion perception in face and voice during infancy. Restor Neurol Neurosci 2010;28:219236. 17. Ekman P, Friesen WV, OSullivan M, et al. Universals and cultural differences in the judgments of facial expressions of emotion. J Pers Soc Psychol 1987;53:712717. 18. Burton LA, Rabin L, Wyatt G, et al. Priming effects for affective vs. neutral faces. Brain Cogn 2005;59:322329. 19. Feinberg T, Rifkin A, Schaffer C, Walker E. Facial discrimination and emotional recognition in schizophrenia and affective disorders. Arch Gen Psychiat 1986;43:276279. Depression and Anxiety
20. Persad SM, Polivy J. Differences between depressed and nondepressed individuals in the recognition of and response to facial emotional cues. J Abnorm Psychol 1993;102:358368. 21. Harmer CJ, OSullivan U, Favaron E, et al. Effect of acute antidepressant administration on negative affective bias in depressed patients. Am J Psychiatry 2009;166:11781184. 22. Joormann J, Gotlib IH. Is this happiness I see? Biases in the identification of emotional facial expressions in depression and social phobia. J Abnorm Psychol 2006;115:705714. 23. Mikhailova ES, Vladimirova TV, Iznak AF, et al. Abnormal recognition of facial expression of emotions in depressed patients with major depression disorder and schizotypal personality disorder. Biol Psychiatry 1996;40:697705. 24. Surguladze SA, Young AW, Senior C, et al. Recognition accuracy and response bias to happy and sad facial expressions in patients with major depression. Neuropsychology 2004;18:212218. 25. Rubinow DR, Post RM. Impaired recognition of affect in facial expression in depressed patients. Biol Psychiatry 1992;31: 947953. 26. Gur RC, Erwin RJ, Gur RE, et al. Facial emotion discrimination: II. Behavioral findings in depression. Psychiatry Res 1992;42:241251. 27. Leppanen JM, Milders M, Bell JS, et al. Depression biases the recognition of emotionally neutral faces. Psychiatry Res 2004; 128:123133. 28. Cooley EL, Nowicki Jr S. Discrimination of facial expressions of emotion by depressed subjects. Genet Soc Gen Psychol Monogr 1989;115:449465. 29. Gollan JK, Pane HT, McCloskey MS, Coccaro EF. Identifying differences in biased affective information processing in major depression. Psychiatry Res 2008;159:1824. 30. Yoon KL, Joormann J, Gotlib IH. Judging the intensity of facial expressions of emotion: depression-related biases in the processing of positive affect. J Abnorm Psychol 2009;118:223228. 31. Ridout N, Astell AJ, Reid IC, et al. Memory bias for emotional facial expressions in major depression. Cogn Emotion 2003;17: 101122. 32. Pine DS, Lissek S, Klein RG, et al. Face-memory and emotion: associations with major depression in children and adolescents. J Child Psychol Psychiatry 2004;45:11991208. 33. Beck AT, Brown G, Steer RA, et al. Differentiating anxiety and depression: a test of the cognitive content-specificity hypothesis. J Abnorm Psychol 1987;96:179183. 34. Hale 3rd WW. Judgment of facial expressions and depression persistence. Psychiatry Res 1998;80:265274. 35. Hale 3rd WW, Jansen JH, Bouhuys AL, van den Hoofdakker RH. The judgement of facial expressions by depressed patients, their partners and controls. J Affect Disord 1998;47:6370. 36. Csukly G, Czobor P, Simon L, Takacs B. Basic emotions and psychological distress: association between recognition of facial expressions and Symptom Checklist-90 subscales. Compr Psychiatry 2008;49:177183. 37. Le Masurier M, Cowen PJ, Harmer CJ. Emotional bias and waking salivary cortisol in relatives of patients with major depression. Psychol Med 2007;37:403410. 38. Chan SW, Goodwin GM, Harmer CJ. Highly neurotic neverdepressed students have negative biases in information processing. Psychol Med 2007;37:12811291. 39. Arce E, Simmons AN, Stein MB, et al. Association between individual differences in self-reported emotional resilience and the affective perception of neutral faces. J Affect Disord 2009; 114:286293. 40. Weniger G, Lange C, Ruther E, Irle E. Differential impairments of facial affect recognition in schizophrenia subtypes and major depression. Psychiatry Res 2004;128:135146.
717
41. Csukly G, Czobor P, Szily E, et al. Facial expression recognition in depressed subjects: the impact of intensity level and arousal dimension. J Nerv Ment Dis 2009;197:98103. 42. Nandi DN, Saha GB, Bhattacharya BB, Mandal MK. A study on perception of facial expressions of emotions in depression. Ind J Psychiatry 1982;24:284290. 43. Mendlewicz L, Linkowski P, Bazelmans C, Philippot P. Decoding emotional facial expressions in depressed and anorexic patients. J Affect Disord 2005;89:195199. 44. Gilboa-Schechtman E, Foa E, Vaknin Y, et al. Interpersonal sensitivity and response bias in social phobia and depression: Labeling emotional expressions. Cogn Ther Res 2008;32: 605618. 45. Mandal MK, Palchoudhury S. Responses to facial expression of emotion in depression. Psychol Rep 1985;56:653654. 46. Mandal MK, Bhattacharya BB. Recognition of facial affect in depression. Percept Mot Skills 1985;61:1314. 47. Archer J, Hay DC, Young AW. Face processing in psychiatric conditions. Br J Clin Psychol 1992;31:4561. 48. Kan Y, Mimura M, Kamijima K, Kawamura M. Recognition of emotion from moving facial and prosodic stimuli in depressed patients. J Neurol Neurosurg Psychiatry 2004;75:16671671. 49. Bediou B, Krolak-Salmon P, Saoud M, et al. Facial expression and sex recognition in schizophrenia and depression. Can J Psychiatry 2005;50:525533. 50. Loeb A, Beck AT, Feshbach S, Wolf A. Some effects of reward upon the social perception and motivation of psychiatric patients varying in depression. J Abnorm Psychol 1964;68:609616. 51. Asthana HS, Mandal MK, Khurana H, Haque-Nizamie S. Visuospatial and affect recognition deficit in depression. J Affect Disord 1998;48:5762. 52. Mandal MK. Decoding of facial emotions, in terms of expressiveness, by schizophrenics and depressives. Psychiatry 1987;50: 371376. 53. Zuroff DC, Colussy SA. Emotion recognition in schizophrenic and depressed inpatients. J Clin Psychol 1986;42:411417. 54. Gaebel W, Wolwer W. Facial expression and emotional face recognition in schizophrenia and depression. Eur Arch Psychiatry Clin Neurosci 1992;242:4652. 55. Langenecker SA, Bieliauskas LA, Rapport LJ, et al. Face emotion perception and executive functioning deficits in depression. J Clin Exp Neuropsychol 2005;27:320333. 56. Walker E, McGuire M, Bettes B. Recognition and identification of facial stimuli by schizophrenics and patients with affective disorders. Br J Clin Psychol 1984;23:3744. 57. Gessler S, Cutting J, Frith CD, Weinman J. Schizophrenic inability to judge facial emotion: a controlled study. Br J Clin Psychol 1989;28:1929. 58. Matthews SC, Strigo IA, Simmons AN, et al. Decreased functional coupling of the amygdala and supragenual cingulate is related to increased depression in unmedicated individuals with current major depressive disorder. J Affect Disord 2008;111: 1320. 59. Frodl T, Scheuerecker J, Albrecht J, et al. Neuronal correlates of emotional processing in patients with major depression. World J Biol Psychiatry 2009;10:202208. 60. Gordon E, Barnett KJ, Cooper NJ, et al. An integrative neuroscience platform: application to profiles of negativity and positivity bias. J Integr Neurosci 2008;7:345366. 61. Habel U, Windischberger C, Derntl B, et al. Amygdala activation and facial expressions: explicit emotion discrimination versus implicit emotion processing. Neuropsychologia 2007;45: 23692377.
62. Knyazev GG, Slobodskoj-Plusnin JY, Bocharov AV. Eventrelated delta and theta synchronization during explicit and implicit emotion processing. Neuroscience 2009;164:15881600. 63. Williams LM, Palmer D, Liddell BJ, et al. The when and where of perceiving signals of threat versus non-threat. Neuroimage 2006;31:458467. 64. Sheline YI, Barch DM, Donnelly JM, et al. Increased amygdala response to masked emotional faces in depressed subjects resolves with antidepressant treatment: an fMRI study. Biol Psychiatry 2001;50:651658. 65. Gordon E. Integrative neuroscience. Neuropsychopharmacology 2003;28:S2S8. 66. Gordon E, Cooper N, Rennie C, et al. Integrative neuroscience: the role of a standardized database. Clin EEG Neurosci 2005;36: 6475. 67. Trzepacz PT, Baker RW. The Psychiatric Mental Status Examination. New York: Oxford University Press; 1993. 68. Bush K, Kivlahan DR, McDonell MB, et al. The AUDIT alcohol consumption questions (AUDIT-C): an effective brief screening test for problem drinking. Ambulatory Care Quality Improvement Project (ACQUIP). Alcohol use disorders identification test. Arch Intern Med 1998;158: 17891795. 69. Heatherton TF, Kozlowski LT, Frecker RC, Fagerstrom KO. The Fagerstrom test for nicotine dependence: a revision of the Fagerstrom Tolerance Questionnaire. Br J Addict 1991;86: 11191127. 70. Breslau N, Kessler RC. The stressor criterion in DSM-IV posttraumatic stress disorder: an empirical investigation. Biol Psychiatry 2001;50:699704. 71. Williams LM, Whitford TJ, Flynn G, et al. General and social cognition in first episode schizophrenia: identification of separable factors and prediction of functional outcome using the IntegNeuro test battery. Schizophr Res 2008;99:182191. 72. Rowe DL, Cooper NJ, Liddell BJ, et al. Brain structure and function correlates of general and social cognition. J Integr Neurosci 2007;6:3574. 73. Williams LM, Gatt JM, Hatch A, et al. The integrate model of emotion, thinking and self regulation: an application to the "paradox of aging". J Integr Neurosci 2008;7:367404. 74. Lovibond PF, Lovibond SH. Manual for the Depression Anxiety Stress Scale. 2nd ed. Sydney: Psychological Foundation; 1995. 75. Mathersul D, Palmer DM, Gur RC, et al. Explicit identification and implicit recognition of facial emotions: II. Core domains and relationships with general cognition. J Clin Exp Neuropsychol 2009;31:278291. 76. Gur RC, Sara R, Hagendoorn M, et al. A method for obtaining 3-dimensional facial expressions and its standardization for use in neurocognitive studies. J Neurosci Methods 2002;115: 137143. 77. Mineka S, Watson D, Clark LA. Comorbidity of anxiety and unipolar mood disorders. Annu Rev Psychol 1998;49: 377412. 78. Gordon E, Williams LM. The Role of Standardized Databases in Personalized Medicine. Personalized Medicine, Healthcare and Integrative Neuroscience. New York: Oxford University Press; 2011. 79. Gatt JM, Williams LM, Schofield PR, et al. Impact of the HTR3A gene with early life trauma on emotional brain networks and depressed mood. Depress Anxiety 2010;27: 752759. Depression and Anxiety
718
80. Williams LM, Kemp AH, Felmingham K, et al. Neural biases to covert and overt signals of fear: dissociation by trait anxiety and depression. J Cogn Neurosci 2007;19:15951608. 81. Drevets WC, Frank E, Price JC, et al. PET imaging of serotonin 1A receptor binding in depression. Biol Psychiatry 1999;46: 13751387. 82. Nemeroff CB. The burden of severe depression: a review of diagnostic challenges and treatment alternatives. J Psychiatr Res 2007;41:189206. 83. Fernandez A, Pinto-Meza A, Bellon JA, et al. Is major depression adequately diagnosed and treated by general practitioners? Results from an epidemiological study. Gen Hosp Psychiatry 2010;32:201209. 84. American Psychiatric Association. Diagnostic and Statistic Manual of Mental Disorders. 3rd ed. Washington, DC: APA; 1980. 85. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders (IV). Washington, DC: APA; 1994. 86. Prashant A. Psychological stress in adolescents: a study of depression, self-esteem, and physical symptoms. Unpublished Doctoral Dissertation, Banaras Hindu University, India; 1988. 87. Montgomery S, Ashberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry 1989;134:382389. 88. Beck AT, Ward CH, Mendelson M, et al. An inventory for measuring depression. Arch Gen Psychiatry 1961;4:561571. 89. Hathaway SD, McKinley JC. A multiphasic personality schedule (Minnesota): III. The measurement of symptomatic depression. J Psychol 1942;14:7384. 90. Spitzer RL, Endicott J, eds. Research Diagnostic Criteria. New York: Biometrics Research, New York State Psychiatric Institute; 1975. 91. Hamilton M. Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 1967;6:278296.
92. Spitzer RL, Williams JBW, Gibbon M, et al. Structured Clinical Interview for DSM-IV (SCID). New York: New York State Psychiatric Institute, Biomedics Research; 1995. 93. Beck AT, Steer RA, Brown GK. Manual for the BDI-II. San Antonio, TX: The Psychological Corporation; 1996. 94. American Psychiatric Association. Diagnostic and Statistic Manual of Mental Disorders. 3rd rev ed. Washington, DC: APA; 1987. 95. World Health Organisation. The ICD-10 Classification of Mental and Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. Geneva: WHO; 1992. 96. Bunney Jr. WE, Hamburg DA. Methods for reliable longitudinal observation of behavior. Arch Gen Psychiatry, 1963;9:280294. 97. First MB, Spitzer RL, Gibbon M, et al. Structured Clinical Interview for DSM-IV Axis I disorders Patient Edition. New York: Biomedic Research Department; 1995. 98. Beck AT, Rial WY, Rickels K. Short form of depression inventory: cross-validation. Psychol Rep 1974;34:11841186. 99. Dempsey PA. Unidimensional depression scale for the MMPI. J Consult Psychol 1964;28:364370. 100. Campbell-Sills L, Stein MB. Psychometric analysis and refinement of the Connor-davidson Resilience Scale (CD-RISC): validation of a 10-item measure of resilience. J Trauma Stress 2007;20:10191028. 101. Eysenck SBG, Eysenck HJ, Barrett P. A revised version of the psychoticism scale. Personality and individual differences, 1985; 6:2129. 102. Derogatis LR, Cleary PA. Factorial invariance across gender for the primary symptom dimensions of the SCL-90. Br J Soc Clin Psychol 1977;16:347356. 103. Andreasen NC, Endicott J, Spitzer RL, et al. The family history method using diagnostic criteria; reliability and validity. Arch Gen Psychiatry 1977;34:12291235.