Downloaded by [Northeastern University] at 06:48 28 March 2016 E-mail: +972-3-567-3891. Fax: +972-50-9713517. Tel: Israel. 3498838, Haifa Carmel, Mount Haifa, of University Levy-Gigi, Einat Correspondence: Introduction 1 Einat The ORIGINAL Haifa, Neuroimaging Abstract 06 Daa t l, 02 Dcesn Ecebu, 2010). Eichenbaum, & Dickerson 2012; al., et Diana (Davachi,2006; information context-related and object integrates unclear. in still is symptoms individuals trauma-exposed on PTSD to deficit relations its hippocampal and functions of cognitive effect the However reduction. volume hippocampal with trauma associated be PTSD, may itself of exposure independent that suggest findings These 2010). al., et Woon 2006; al., et Karl see analysis, meta (for reduced controls trauma-unexposed a to compared volumehippocampal have PTSD without individuals trauma-exposed only not individuals with posttraumatic stress disorder (PTSD) but also that shown have studies neuroimaging Numerous part repeatedly individuals the nml n hmn oes f TD ugs ta a that associations suggest impaired PTSD in result of may models deficit hippocampal human and Animal with information. deficit t l, 02 Goes 21; osaa t l, 2013; al., et Moustafa 2011; Goosens, 2012; al., et aversiveinformationand contextualbetween events(Acheson between suggest previously to found of traumatic exposure. way their to correlate The individuals negative acquire h ie-ncnet oe age ta te hippocampus the that argues model item-in-context The performance highly of Institute PTSD appropriate Israel that Levy-Gigi in [email protected] their hidden that contextual with exposure, this We negative while have exposed without exposed and and The for occupational used hippocampal with levels non-clinical studies RESEARCH retain a positive the both of context. 1 aim selective PTSD, a context posttraumatic active-duty which to information and of individuals Study novel trauma-exposed traumatic have PTSD, stimulus-outcome of price outcomes highly Gal is of this population This cue–context routine is impairment deficit not demonstrated Affective anxiety later Richter-Levin firefighters study impairment interpret REPORT and exposed necessarily events stress associated of may and of aversive or Neuroscience, was cue are firefighters reversal repeated display depressive civilian in result disorder individuals and associations, reversing who still and reduced associated to may events. with react ! 1,2,3 test in unclear. similar context-related matched-controls are paradigm, 2014 ISSN: reflect (PTSD). and impaired a symptoms. to frequently positive 2 whether hippocampal fail Previous the Department Informa their with unexposed 1025-3890 impairment. http://informahealthcare.com/sts firefighters to Animal a negative However, which possible traumatic associate PTSD environment. Stress, UK learning outcome. study non-PTSD exposed The Ltd. information. and diagnosis, (print), of 2014; separately matched-controls outcome with volume DOI: results To the struggled has Psychology, hidden traumatic human and This 17(4): that 10.3109/10253890.2014.923397 shown to implications no 1607-8888 individuals ipcma fntos Lv-ii t l, 01 ad volume and 2011) al., et (Levy-Gigi functions hippocampal associations (A associations al., et stimulus-outcome Murnane learn participants 1992; paradigm, our al., In 1999). et (Mayes context specific a in association occurs that cue a usuallycontains stimulus a that account into stimulus-outcome paradigm, reversal(S stimulus same a common a (S acquire In 2014). participants 2011, al., et (Levy-Gigi innovativeparadigm reversalcue–contextan used we deficits, similar experience trauma, to exposure repeated outcome. negative a with shops coffee Such all associate may 2013). shop, coffee a in attack al., terror a to exposed et Maren see may impairment explain,for example, who why person, was a review, for 2009, Rudy, ae ve nw soitos wih eur rvrig the reversing require of which outcome associations, new view later background background related information. related context- and in cue of outcomes negative and impairments positive reversing selective detect to us enables manipulation suggest use traumatic impairment in end, history and 343–351 of price evaluates ! As Performance on our paradigm significantly correlated with correlated significantly paradigm our on Performance n re t ts wehr o-TD niiul, with individuals, non-PTSD whether test to order In models trauma-exposed to that context-related of outcomes exposure and may oiie ad ae ne t rvre h otoe f the of outcome the reverse to need later and Positive) predicted, we learn associations of that (electronic) of individuals compared were 3 of affect events repeated who Sagol of trauma- reversal did ! ! such similar that PTSD with able either Negative) of the acquired stimuli. This unique This stimuli. acquired the of Negative) eaie o the or Negative) not the are we Department hat as a a ! on an Negative).takenot does paradigm Sucha Keywords Context, Published Received History Accepted Revised the hippocampus, trauma, orange of cue 29 Neurobiology, firefighters, 23 3 online reversal-learning April May December (A context background 2014 2014 10 repeated phone June first-responders, University (A 2013 2014 exposure n an on ! hat Positive) and of n a on Haifa, to orange grey 344 E. Levy-Gigi & G. Richter-Levin Stress, 2014; 17(4): 343–351

reduction (Levy-Gigi et al., 2014). Specifically, we found that study due to past exposure to potential traumatic event. individuals with PTSD showed a selective deficit in reversing Individuals in both groups showed high rates of consent; the outcome of negative context; after they learned that a hence, approximately 95% of the people we sampled agreed to specific context is associated with a negative outcome, they participate in the study. All participants were interviewed struggled to learn that the same context predicts a positive using the Structured Clinical Interview for Diagnostic and outcome when presented later with a new cue. Statistical Manual for Mental Disorders-Forth Edition The aim of this study was to test whether non-PTSD (DSM–IV) Axis I Disorders (SCID-CV) (First et al., 1996). individuals with repeated traumatic exposure would show a Exclusion criteria included any current DSM-IV psychopath­ deficit in reversing the outcome of negative context similar to ology including PTSD, and any history of psychiatric or what we recently found in individuals with PTSD. To that neurological disorders, alcohol abuse or dependence. Two end, we concentrated on a unique population of active-duty firefighters were excluded from the sample due to a clear firefighters and compared them to trauma-unexposed matched diagnosis of PTSD. The rest thirty-two Non-PTSD fire­ controls. fighters were also interviewed using the SCID Non-Patient We postulated that both groups would equally learn and PTSD module interview (Spitzer et al., 1990) to assess the retain positive and negative stimulus–outcome associations. levels of subclinical PTSD symptoms. All interviews were However, we expected that similar to previous findings in conducted by a well-trained and regularly supervised clinical individuals with PTSD, non-PTSD highly exposed individuals psychologist. The experiment was done in accordance with would show a selective impairment in reversing the outcome the Declaration of Helsinki for the protection of human of negative context compared to trauma-unexposed participants. All participants provided a written informed individuals. consent at the beginning of the experiment.

Methods and materials Tools Participants Cue and context reversal paradigm Thirty-two active-duty firefighters who are repeatedly In this paradigm, participants view a series of boxes on a exposed to trauma as part of their daily routine and thirty- computer screen (Figure 1). On each box, there is a picture of one unexposed controls matched for age, gender and years of a cue (one of various objects, e.g. a hat) presented against a education volunteered to participate in the study (see Table 1 specific context (different background colors, e.g. orange) for a detailed description of the sample). Firefighters were (see Hockley, 2008; Isarida & Isarin, 2007; Lang et al., 2009; randomly recruited from five different fire stations in Macken, 2002; Rutherford, 2004 for studies that manipulated southern Israel, which are all located in a similar setting context in a similar way). When opened, each box is within a radius of 40 miles. All firefighters reported multiple associated with a specific outcome (positive or negative). exposures to Diagnostic and Statistical Manual for Mental Participants receive the following instructions: ‘‘In this Disorders-Fifth Edition (DSM-V) Criterion A events. In order experiment you will be shown various boxes. For each box to further validate the firefighter’s exposure to traumatic you have the option to open it or to leave it closed. If you open events, we used the fire and rescue department archive to a box you will either win or lose 25 points (see Figure 2 for collect data on potential traumatic events that were encoun­ tered by firefighters from the five studied fire stations during Table 2. Mean number of exposures to different potential traumatic the past 10 years (see Table 2). Participants in the unexposed events per year in the past 10 years in Israel southern fire and rescue control group were civilians who work in an industrial stations. factory. They were recruited by a clinical psychologist who

Downloaded by [Northeastern University] at 06:48 28 March 2016 Mean number of interviewed them to ensure no past exposure to DSM-V Type of event potential traumatic events criteria A events. Three participants were excluded from the Car fires 179 Building fires 246 Table 1. Demographic characteristics of trauma exposed firefighters and Factory fires 10 trauma-unexposed matched controls. Bush fires 1548 Car accidents 116 Firefighters Controls Spilling of toxic/combustion 7 (N ¼ 32) (N ¼ 31) substances Gas leak 109 Age (years) 36.47 (8.5) 38.6 (8.07) Breaking and entering due to 249 Male/female 27/5 26/5 fear of a lost life Education (years) 12.44 (0.88) 12.42 (0.81) Missile attacks 290–1096* Medications (N)* 4/32 2/31 Attempted suicide 17 SCID-NP-PTSD 24.28 (6.62) N/A Animal rescue mission 18 Time in fire and rescue service (years) 10.47 (9.39) N/A Rescuing trapped people 32

SCID-NP-PTSD: Structured Clinical Interview for Diagnostic and *There were no significant differences between the numbers of potential Statistical Manual for Mental Disorders-Forth Edition (DSM–IV), traumatic events across the years in all types of traumatic events but Non-Patients PTSD module. missile attacks. Data on this section refers to the past seven years only. *Firefighters: one participant received non-selective beta-blockers and The range of events is due to significant differences between three received other supplementary medications such as benzodiazep­ quiet years (3 of 7 years) and years of emergency circumstances ines; control group: two received other supplementary medications (4 of 7 years). All participants in the study experienced at least one year such as benzodiazepine. of extensive missile attacks. DOI: 10.3109/10253890.2014.923397 Hidden price of repeated traumatic exposure 345 Figure 1. Example of the stimuli in the two phases of the Cue–Context Reversal Task.

example of the different trials). If you do not open the box you job! There is gold inside’’. In the following screen, they see Downloaded by [Northeastern University] at 06:48 28 March 2016 will not win or lose any points. Your job is to earn as many the same reward box, with the following text: ‘‘Now suppose points as possible. Through trial and error you will learn to you see the same box again. You just learned there is gold open the boxes that earn you points and not open the boxes inside. You should open it’’. After opening the box again, they that cost you points. Note that in order to learn whether a box see an open box with gold inside of it a smiley face and a earns or costs you points, you should open each box in the numeric indication that they earned 25 points, and receive the first time you see it’’. The experimenter verifies then that the following feedback. ‘‘Very good. You won gold’’. Later, they participants understand the instructions. Afterward, partici­ see a screen with a new box that has a different object pants take part in a practice phase under close supervision of presented against a different background color on it. ‘‘Next the experimenter. This phase demonstrates the task of using suppose you see another new box. You should open it’’. After two boxes; one associated with a positive outcome and the opening the box, participants see an open box with a bomb other associated with a negative outcome. They see a closed inside of it (negative box) accompanied with a matching box, with a picture of an object presented against a voice, a frown face and a numeric indication that they lost background color, and receive the following instructions: 25 points. ‘‘Oops, there is a bomb inside’’. In the following ‘‘Suppose you see a box for the first time. You should open screen, they see the same negative box, with the following it’’. After opening the box, participants see gold inside of it text: ‘‘Now, suppose you see the same box again. You just (positive box) accompanied with a matching voice, a smiley learned that there is a bomb inside. You should decide not to face and a numeric indication that they earned 25 points. open it’’. After choosing the ‘‘Do not open’’ option, These points are added to the participants’ total amount of participants receive the following feedback: ‘‘You were points indicated at the side of the screen (Figure 2). ‘‘Great right not to open it. There is a bomb inside’’. The experiment 346 E. Levy-Gigi & G. Richter-Levin Stress, 2014; 17(4): 343–351

Figure 2. Example of experimental trials in which participants chose to (a) open a positive-outcome box and (b) open a negative-outcome box.

starts at the end of the practice phase. We created new boxes with the hat on the orange background has gold inside, then for the experiment, different from those presented in the the boxes with the hat on a grey background and a phone on practice phase, using eight cue objects and eight distinctive the orange background will have bomb inside and vice versa). context colors (for a schematic description see Table 3). Therefore, in order to successfully learn these new associ­ Boxes were 400 x 300 size, presented on a 1300 screen. The ations, participants need to reverse the association rule of outcome of each box was counterbalanced across participants. either the original cue or the original context. Boxes in this The paradigm has two phases. In the acquisition phase, phase are presented in 10 blocks of 12 boxes each (two boxes participants learn by trial and error to predict the outcome of from each of the following conditions: positive/negative four different boxes (i.e. open the two positive boxes and skip retention, positive/negative cue reversal and positive/negative the two negative boxes). Each box has a unique cue and context reversal). These sums up to a total of 120 trials; 20 context (i.e. a box with a hat on an orange background has trials per condition. At the end of the task, participants see gold inside while a box with a car on a yellow background has their total earned points; however, the experiment includes no bomb inside). The acquisition phase contains a minimum of actual payment. Downloaded by [Northeastern University] at 06:48 28 March 2016 40 trials. However, in order to ensure learning of the stimulus–outcome associations in this phase, participants Self-report questionnaires and cognitive assessment have to reach a criterion of six consecutive correct responses All participants completed self-report questionnaires in order before they move on to the next phase. Participants who do to control for possible effects of depression and anxiety not reach this criterion within 64 trials are automatically opt- symptoms. Depressive symptoms over the previous two out from the experiment. Correct responses refer to conditions weeks were assessed using the revised version of the Beck in which participants open positive boxes or leave negative Depression Inventory-II (BDI-II; Beck et al., 1996). General boxes closed. Similarly, incorrect responses refer to condi­ anxiety was measured using the State–Trait Anxiety Inventory tions in which participants open negative boxes or leave (STAI; Spielberger et al., 1983) questionnaire. Finally, we positive boxes closed. A subsequent retention and reversal used the scaled scores of the Wechsler Adult Intelligence phase starts immediately after the acquisition phase without Scale III (WAIS-III) vocabulary subtest to estimate IQ levels any signaled switch or delay. In this phase, participants (Wechsler, 1997). Previous studies showed that scores from receive retention trials with the original boxes that keep the this subtest are the best predictor of full IQ scale scores same learned outcome (e.g. a hat on an orange background (Spreen, 1998). has gold inside) in addition to two new types of boxes that share either the cue (e.g. a hat on a gray background) or the Data analysis context (e.g. a phone on an orange background) with an original box (Figure 1). The new boxes are associated with the We used SPSS (version 19) software (SPSS Inc., Chicago, IL) opposite outcome relative to the original boxes (i.e. if the box to analyze the data. All data were checked for normality of DOI: 10.3109/10253890.2014.923397 Hidden price of repeated traumatic exposure 347

Table 3. Schematic description of the Cue–Context Reversal Task.

Acquisition Retention and Reversal A(1)�Positive A(1)�Positive A(5)�Negative E(1) �Negative

B(2)�Positive B(2)�Positive B(6)�Negative F(2)�Negative

C(3) �Negative C(3)�Negative Figure 3. Percentage of correct responses to the four original boxes as a function of Phase (Acquisition vs. Retention), Outcome (Positive C(7)�Positive vs. Negative) and Experimental Group (Trauma Exposed Firefighters vs. Trauma-Unexposed Controls). G(3)�Positive

D(4)�Negative D(4)�Negative D(8)�Positive H(4) �Positive

A–H represent eight different types of cue (hat, phone, car, ball, television, chair, bird and pot). 1–8 represent eight different types of context (orange, grey, yellow, purple, green, pink, blue and red, respectively). In both the acquisition and retention-reversal phases, each stimulus was presented 10 times. This constitutes a total of minimum 40 acquisition trials, 40 retention trials and 80 reversal trials.

distribution using Kolmogorov–Smirnov tests. Since partici­ pants are instructed to open boxes when they first see it, in our Figure 4. Percentage of correct responses for the new associations as a function of Reversal Type (Cue vs. Context), Outcome (Reversal from analyses, we did not include the first response to each new Positive to Negative vs. Reversal from Negative to Positive) and box in the acquisition and reversal trials (note that retention Experimental Group (Trauma Exposed Firefighters vs. Trauma- trials include only old boxes, and therefore all trials are Unexposed Controls). Cue reversal refers to conditions of old cue, analyzed). This was done in order to avoid artificial errors which is presented against a new context; Context reversal refers to Downloaded by [Northeastern University] at 06:48 28 March 2016 conditions of new cue, which is presented against an old context. (i.e. when participants open a negative box for the first time) and possible effects of task compliancy. ANOVA revealed no significant main effects of Group F p Results ( (1,61) ¼ 0.06, ¼ 0.81) and no significant interactions of Acquisition by Group (F(1,61) ¼ 1.26, p ¼ 0.27) Retention Acquisition and retention of stimulus–outcome by Group (F(1,61) ¼ 1.18, p ¼ 0.28) nor Acquisition by associations Retention by Group (F(1,61) ¼ 0.18, p ¼ 0.68). These results The vast majority of the participants (60 of 63) acquired the indicate that there were no significant differences in per­ stimulus–outcome associations within the minimum of 40 formance between acquisition and retention trials. In addition, trials. One trauma-exposed participant and two unexposed it shows that both firefighters and unexposed matched matched controls needed 1–2 additional blocks in order to controls are equally able to learn and retain positive and reach a criterion of six consecutive correct responses. negative stimulus–outcome associations. We conducted a Group (trauma-exposed firefighters vs. Cue and context reversal trauma-unexposed controls) by Acquisition (positive vs. negative stimuli) by Retention (positive vs. negative stimuli) We conducted a Group (trauma exposed firefighters mixed model ANOVA on the percentage of correct responses. vs. trauma-unexposed controls) by Reversal Type In this model, Group was the between-subjects factor, while (cue vs. context) by Outcome (reversal from positive to Acquisition and Retention were the within-subjects factors. negative vs. reversal from negative to positive) mixed model The results are depicted in Figure 3. As predicted, the ANOVA on the percentage of correct responses. In this Downloaded by [Northeastern University] at 06:48 28 March 2016 cin f ru b Rvra Tp i negative-to-positive in ( Type reversals Reversal by Group of action oiiet-eaie eesl ( reversals positive-to-negative results are depicted in Figure 4. There were no significant significant ( no Outcome or TypeReversal were The Group, of main-effects There factor. 4. Figure within-subjects in depicted the are results were Outcome and Type model, Group was the between-subjects factor, while Reversal 348 oee, e on a infcn til itrcin between Group, ReversalType ( and Outcome interaction triple significant a found we However, n uepsd otos n h BII (ek t l, 1996), al., et (Beck BDI-II the on controls unexposed and Table 4 depicts the of comparison trauma-exposed firefighters test (Levene’s firefighters F exposed trauma correct among of scores distribution was the controls to compared unexposed lower for significantly scores correct of distribution ace cnrl ( controls matched unexposed of number the with compared higher significantly the in firefighters second group (number of correct trauma-exposedresponses below median) was of In number controls. the matched contrast, unexposed of number the than lower aboveresponses wasmedian) significantlycorrect of (number group first the in firefighters trauma-exposed of number the that revealed test Chi-square performance. their to according groups two into participants the divide to context negative of respectively). SD ( information context-related of outcomes negative reversing in impaired significantly were firefighters ( correction Bonferroni with comparisons pairwise Follow-up Self-report edny o pn e rvra bxs ( boxes results reversal new The open to them). tendency saw first the in groups the theybetween differences significant no revealed when boxes new two new the opened the they (i.e. two of to them) anysaw first they open when boxes not did a they (i.e. receive zero between could score participants therefore, condition; reversal each in boxes new two are There variable. dependent the as boxesof number opened the with reversalfourconditions, the sample independent conducted we presented, in tendency the to open new reversalboxes when they first are in seen be well. equally preformed can groups both conditions, As reversal other cue. three the is in different 4, Figure it a when with outcome later positive presented a with associated is context same the that learn to struggle they outcome, negative a with context that a specific learn is associated firefighters that after cue-related of outcomes ( negative information reversing in groups the p 7 ( controls matched unexposed of tendency the from differ not did them saw first they boxeswhen these open to participants context same trauma-exposed of tendency boxes,the negative original the with shared boxes reversal when even Hence, a t p 2 (61) ¼ ¼ ¼ ¼ We used the median number of correct responses in reversal In order to test whether there are group-related differences group-related are there whether test to order In ¼ .0) Tee ee o infcn dfeecs between differences significant no were There 0.000). 4.82, .7. olwu aayi rvae a infcn inter- significant a revealed analysis Follow-up 0.07). .1 soe ta, s rdce, eaie o controls to relative predicted, as that, showed 0.01) 0.25, for trauma exposed and unexposed participants, participants, unexposed and exposed trauma for 0.25, ¼ E. 0.81, Levy-Gigi p 5 F questionnaires (1,61) t 0.05) (Figure 5). (Figure 0.05) (57) p ¼ ¼- ¼ .2 M 0.42; & X 4.69, 2 G. (1) 0.73, Richter-Levin ¼ p 17.31, p and 5 ¼ ¼ .8 SD 1.88, 0.05, .7. hs eut indicate results These 0.47). cognitive F (1,61) p F 5 7 (1,61) p 2 .01. ial, the Finally, 0.0001). t ¼ s ¼ ¼ 5 .7 bt o in not but 0.07) 0.11, .4 M 0.34; 0.82; assessment ¼ t -test in each of each in -test 4.44, t (57) p p p s p 4 ¼ s ¼- 5 ¼ 4 0.74). 0.41). 1.94, 0.05, 0.1). 3.7, p ( anxiety of symptoms total and trait state, of levels and symptoms PTSD between there (e.g. 2012), findings al., past et with Levy-Gigi accordance in Finally, anxiety. or correlations reversalbetween depression of PTSD, symptoms learning and significant no were there addition, In controls. unexposed the scores and IQ firefighters significant trauma-exposed and the anxiety between no depression, of were levels in There differences 1997). Wechsler, and (WAIS-III, 1983) al., et (Spielberger STAI the blw n aoe ein i rvrig h ngtv otoe of outcome (Trauma negative group Controls). experimental Trauma-Unexposed the vs. Firefighters of Exposed function reversing a in responses as information correct median) contextual of above percentage and in (below differences Individual 5. Figure ucm i ascae wt te poie ucm when outcome opposite the with associated is negative outcome or positive with associated first was which object, an previous that learn to able were with they learning; reversal cue spared accordance in displayedgroups both 2014), (Levy-Gigi 2011, al., et findings addition, stimulus–outcome negative In and associations. positive retain and learn to able equally were groups both that found we predicted, As paradigm. reversal cue–context novel trauma-unexposeda on controls and matched PTSD without firefighters exposed trauma- highly of performance the information. compared we end, that context-related To and and positive repeated reverse cue- of of to effect outcomes ability negative the the test on to exposure was traumatic study this of aim The Discussion 5 0.05; BDI-II controls. matched unexposed trauma and firefighters exposed trauma of deviation) standard and (means assessment cognitive and Questionnaires 4. Table IQ scores as measured by the WAIS-III vocabulary subtest. al., et (Spielberger Inventory Anxiety State–Trait STAI: 1996). (Beck, Inventory Depression Beck The BDI-II: score IQ STAI-trait STAI-state 1983). r (32) ¼ 0.37, p 5 r 10.72 (1.69) 10.72 (8.48) 26.72 (8.17) 27.38 Firefighters Firefighters (32) 3.72 (4.46) 3.72 0.05, respectively). 0.05, ¼ were significant correlations correlations significant were 0.37, p 5 Stress, 2014; 17(4): 343–351 17(4): 2014; Stress, 0.05; n Q assessment IQ on 10.35 (1.28) 10.35 (4.62) 25.65 (3.92) 24.74 4.16 (3.56) 4.16 Controls r (32) ¼ 0.36, DOI: 10.3109/10253890.2014.923397 Hidden price of repeated traumatic exposure 349

presented later in a different context (e.g. a hat on an orange the gold/bomb inside). The results from such a task may help background is positive while a hat on a gray background is to better understand the mechanisms beyond the impaired negative and vice versa). However, similar to previous ability of individuals with repeated traumatic exposure to findings in individuals with PTSD, firefighters who reverse the negative outcome of contextual information. experience repeated traumatic exposure showed a selective In a previous study, we reported that a deficit in reversing deficit in reversing negative context; after they learned the negative outcome of contextual information was asso­ that a specific context is associated with a negative outcome ciated with reduced hippocampal volume (Levy-Gigi et al., (e.g. a car on a yellow background is negative) they could 2014). Therefore, the results of this study may reflect a not learn that it predicts a positive outcome when presented reduction in hippocampal volume among individuals with later with a new object (e.g. a football on a yellow background repeated traumatic exposure and provide further support for is positive). Moreover, the magnitude of the effect in this imaging studies that described similar structural abnormal­ group was similar to the one we previously observed in fully ities in trauma-exposed individuals independent of PTSD PTSD-diagnosed people (Levy-Gigi et al., 2014). diagnosis (for meta analyses, see Karl et al., 2006; Kitayama This study is the first to show associations between et al., 2005; Smith, 2005; Woon et al., 2010). repeated traumatic exposure and impairment in reversing the Although intuitively it seems that a deficit in reversing the negative outcome of context-related information in non-PTSD negative outcome of contextual information may contribute to individuals. There are several possible ways to interpret the the development of PTSD symptoms, the results revealed no current results. First, the results may suggest that individuals significant correlations between these variables. Leaning on with repeated traumatic exposure fail to associate traumatic this set of data as proof of concept, future cross-sectional outcomes with their appropriate context. Therefore, they may studies may aim to test a larger sample of individuals with experience difficulty to recognize and differentiate novel repeated traumatic exposure in order to further understand the conditions from other negative conditions, which share the link between PTSD symptoms and negative and positive same context. Similar to findings in PTSD individuals (Brown reversal learning. Moreover, a larger sample may allow et al., 2013; Levy-Gigi & Ke´ri, 2012; Levy-Gigi et al., 2012, further testing of individual differences within this group 2014), such impairment may lead to inappropriate general­ (see Figure 5) and enable looking at associations between ization of the negative outcome to the novel conditions. specific response patterns (e.g. intact performance, slower Alternatively, it is possible that like the stronger fear learning or impaired overall performance) and different types conditioning observed in stressed animals (e.g. Giachero of PTSD symptoms. et al., 2013; Rau & Fanselow, 2009; Rau et al., 2005 but see Similar to our previous findings in individuals with PTSD Tsoory et al., 2010), individuals with repeated exposure to (Levy-Gigi et al., 2014), the impairment of individuals with trauma make stronger context–outcome associations when repeated traumatic exposure was selective to conditions of negative outcomes are involved. These stronger associations reversing negative, but not positive outcome of context- may then be more difficult to reverse. Therefore, they struggle related information. These results may suggest that the to learn that a previously negative context becomes positive. hippocampus–amygdala connectivity in individuals who Finally, it is possible that individuals with repeated traumatic repeatedly exposed to trauma facilitates learning in conditions exposure have an inherent bias to associate the context, but of negative feedback (LaBar & Cabeza, 2006). Specifically, not the cue with behavioral outcomes. Therefore, when they although they struggle to learn when negative context see a new cue on a context previously paired with a positive becomes positive, they can successfully learn that a previ­ outcome (e.g. a phone presented against an orange back­ ously positive context becomes negative. Support for such ground), their bias to open the box allows modifying the claim can be found in neuroimaging studies, which observed behavior accordingly (i.e. the participants see a bomb inside enhanced amygdala response in threatening and aversive Downloaded by [Northeastern University] at 06:48 28 March 2016 and learn to skip this box in the future). In contrast, when they contextual conditions (Buchel et al., 1999; Phelps et al., 2001; see a new object on a context previously paired with a Smith et al., 2004, 2006; Stevens et al., 2013) and advantage negative outcome (soccer ball presented against a yellow in attending and processing aversive stimuli in trauma- background), their bias to leave the box closed does not allow exposed individuals (Fani et al., 2012; Kleim et al., 2012; learning (e.g. the participants receive no feedback and do not Vythilingam et al., 2007; Wald et al., 2013). Future fMRI know that their choice was ‘‘wrong’’) and therefore they study, which assesses hippocampus–amygdala connectivity in continue to leave the box closed. highly exposed individuals during context reversal-learning, is Although all these alternatives are plausible explanations needed in order to clarify this point. of the current data, it is important to note that individuals Finally, the results may shed new light on recent studies of from both groups did not differ in their tendency to open new PTSD in first responders. A large number of these studies reversal boxes when they first presented. This fact may reported relatively low PTSD prevalence in firefighters suggest that individuals with repeated exposure to trauma (e.g. Chang et al., 2008; Del Ben et al., 2006; Fushimi, recognize new boxes, even if they share context with a 2012; Meyer et al., 2012; Soo et al., 2011). Furthermore, a negative box, and have an opportunity to learn it predicts number of prospective studies, which aimed to predict PTSD positive outcome. Yet, they struggle to reverse the negative symptoms in active-duty firefighters and police after exposure outcome of these boxes compared to unexposed controls. to traumatic events, revealed low rates of PTSD symptoms Future studies may aim to use a revised task, in which (Guthrie & Bryant, 2006; Orr et al., 2012; Pole et al., 2009). participants get feedback even if they leave a box closed This study highlights the importance of behavioral measures, (e.g. by showing a transparent image of the closed box with showing that repeated traumatic exposure has a hidden price 350 E. Levy-Gigi & G. Richter-Levin Stress, 2014; 17(4): 343–351

even in non-PTSD individuals, which may affect the way the negative outcome of contextual information. This impair­ these individuals interpret and react to their environment. ment is not restricted to trauma-related situations and may Moreover, the fact that our cue–context reversal paradigm affect the way these individuals interpret and react to their uses neutral stimuli suggests that such price is not limited to environment. trauma-related conditions and might reflect a more general impairment. Acknowledgements A possible limitation of this study may relate to the nature of the cue–context reversal paradigm. The basic assumption We wish to thank Szabolcs Keri for his thoughtful comments in this and other similar paradigms (e.g. Fellows & Farah, and contribution for the design of this study. 2003; Foerde & Shohamy, 2011; Rogers et al., 2000) is that the participants are rational learners. However, it is possible Declaration of interest that decision makers have expectancies and inner values and representations on acts, outcomes and contingencies (Tversky The authors report no conflicts of interest. The authors alone & Kehneman, 1981). Therefore, decisions are often guided by are responsible for the content and writing of this article. biases and heuristics rather than stimulus–response mechan­ isms. Accordingly, it may be claimed that factors such as References expectations, risk taking and loss aversion would affect the Acheson DT, Gresack JE, Risbrough VB. (2012). Hippocampal performance on the cue–context reversal paradigm. If this dysfunction effects on context memory: possible etiology for were the case, we would expect to see a robust effect of post-traumatic stress disorder. Neuropharmacology 62:674–85. negative or positive outcome. For example, participants who Beck AT, Steer RA, Brown GK. (1996). Manual for the Beck Depression avoid risk would struggle to learn that a previously negative Inventory-II. San Antonio, TX: Psychological Corporation. Brown AD, Root JC, Romano TA, Chang LJ, Bryant RA, Hirst W. stimulus becomes positive in conditions of both cue and (2013). Overgeneralized autobiographical memory and future thinking context reversal. Moreover, since this tendency represents in combat veterans with post-traumatic stress disorder. J Behav Ther inner values and expectations, and is not necessarily a result Exp Psychiatry 44:129–34. of traumatic exposure, such effects would be expected in both Buchel C, Dolan RJ, Armony JL, Friston KJ. (1999). Amygdala­ hippocampal involvement in human aversive trace conditioning trauma exposed and unexposed groups. However, the results revealed through event-related fMRI. J Neurosci 19:10869–76. show that only trauma-exposed individuals have impaired Chang CM, Lee LC, Connor KM, Davidson JR, Lai TJ. (2008). learning. This impairment is unique to reversal trials and was Modification effects of coping on post-traumatic morbidity among not observed during positive and negative acquisition trials. earthquake rescuers. Psychiatry Res 158:164–71. Davachi L. (2006). Item, context and relational episodic encoding in Furthermore, it was observed exclusively in conditions of humans. Curr Opin Neurobiol 16:693–700. negative context (but not negative cue) reversal trials. Del Ben KS, Scotti JR, Chen Y, Fortson BL. (2006). Prevalence of Although the selectivity of the observed effect support a posttraumatic stress disorder symptoms in firefighters. Work Stress dominant effect of traumatic exposure, future studies may aim 20:37–48. Diana RA, Yonelinas AP, Ranganath C. (2012). Adaptation to cognitive to test whether expectancies and different attitudes toward context and item information in the medial temporal lobes. reward and punishment mediate individual differences in Neuropsychologia 50:3062–9. reversal learning within each group. Dickerson BC, Eichenbaum H. (2010). The episodic memory system: Another possible limitation is that we tested only fire­ neurocircuitry and disorders. Neuropsychopharmacology 35:86–104. Fani N, Tone EB, Phifer J, Norrholm SD, Bradley B, Ressler KJ, fighters without comparing them to other first responders. Kamkwalala A, Jovanovic T. (2012). Attention bias toward threat is It can be claimed that since firefighters are trained to focus and associated with exaggerated fear expression and impaired extinction in react to aversive environmental conditions, they center their PTSD. Psychol Med 42:533–43. attention on the context and ignore other elements, and Fellows LK, Farah MJ. (2003). Ventromedial frontal cortex mediates Downloaded by [Northeastern University] at 06:48 28 March 2016 affective shifting in humans: evidence from a reversal learning therefore display impaired reversal of negative context. paradigm. Brain 126:1830–7. One way to test this claim is by evaluating cue–context First MB, Gibbon M, Spitzer RL, Williams JBW. (1996). User’s Guide reversal learning of firefighters at the end of their training for the Structured Clinical Interview for DSM-IV Axis I Disorders- course and before trauma exposure. In addition, it might be Research Version. New York: Biometrics Research Department, New York State Psychiatric. informative to compare cue–context reversal learning of first Foerde K, Shohamy D. (2011). Feedback timing modulates brain responders from different occupations, for example, fire­ systems for learning in humans. J Neurosci 31:13157–67. fighters who are trained to attend the general context and Fushimi M. (2012). Posttraumatic stress in professional firefighters in criminal scene investigators who are trained to look for japan: rescue efforts after the Great East Japan Earthquake (higashi nihon dai-shinsai). Prehosp Disaster Med 27:416–18. evidences and therefore may focus their attention on different Giachero M, Calfa GD, Molina VA. (2013). Hippocampal structural cues in the environment. plasticity accompanies the resulting contextual fear memory following stress and fear conditioning. Learn Mem 20:611–16. Goosens KA. (2011). Hippocampal regulation of aversive memories. Conclusions Curr Opin Neurobiol 21:460–6. In conclusion, this study showed that repeated traumatic Guthrie RM, Bryant RA. (2006). Extinction learning before trauma and subsequent posttraumatic stress. Psychosom Med 68:307–11. exposure might have a hidden price independent of PTSD Hockley WE. (2008). The effects of environmental context on recogni­ symptoms and other psychiatric diagnosis. Specifically, tion memory and claims of remembering. J Exp Psychol Learn Mem firefighters who are repeatedly exposed to traumatic events Cogn 34:1412–29. as part of their daily routine are impaired in reversing Isarida T, Isarin TK. (2007). Environmental context effects of back­ ground color in free recall. Mem Cognit 35:1620–9. DOI: 10.3109/10253890.2014.923397 Hidden price of repeated traumatic exposure 351

Karl A, Schaefer M, Malta LS, Dorfel D, Rohleder N, Werner A. (2006). symptoms using fear potentiated auditory startle responses. Biol A meta analysis of structural brain abnormalities in PTSD. Neurosci Psychiatry 65:235–40. Biobehav Rev 30:1004–31. Rogers RD, Andrews TC, Grasby PM, Brooks DJ, Robbins TW. (2000). Kitayama N, Vaccarino V, Kutner M, Weiss P, Bremner JD. (2005). Contrasting cortical and subcortical activations produced by atten­ Magnetic resonance imaging (MRI) measurement of hippocampal tional-set shifting and reversal learning in humans. J Cogn Neurosci volume in posttraumatic stress disorder: a meta-analysis. J Affect 12:142–62. Disord 88:79–86. Rau V, DeCola JP, Fanselow MS. (2005). Stress-induced enhancement of Kleim B, Ehring T, Ehlers A. (2012). Perceptual processing advantages fear learning: an animal model of posttraumatic stress disorder. for trauma-related visual cues in post-traumatic stress disorder. Neurosci Biobehav Rev 29:1207–23. Psychol Med 42:173–81. Rau V, Fanselow MS. (2009). Exposure to a stressor produces a long LaBar KS, Cabeza R. (2006). Cognitive neuroscience of emotional lasting enhancement of fear learning in rats. Stress 12:125–33. memory. Nat Rev Neurosci 7:54–64. Rudy JW. (2009). Context representations, context functions, and the Lang S, Kroll A, Lipinski SJ, Wessa M, Ridder S, Christmann C, parahippocampale-hippocampal system. Learn Mem 16:573–85. Schad LR, Flor H. (2009). Context conditioning and extinction in Rutherford A. (2004). Environmental context-dependent recognition humans: differential contribution of the hippocampus, amygdala and memory effects: an examination of ICE model and cue-overload prefrontal cortex. Eur J Neurosci 29:823–32. hypotheses. Q J Exp Psychol A 57:107–27. Levy-Gigi E, Kelemen O, Gluck MA, Ke´ri S. (2011). Impaired context Smith AP, Henson RN, Dolan RJ, Rugg MD. (2004). fMRI correlates of reversal learning, but not cue reversal learning, in patients with the episodic retrieval of emotional contexts. Neuroimage 22:868–78. amnestic mild cognitive impairment. Neuropsychologia 49:3320–6. Smith AP, Stephan KE, Rugg MD, Dolan RJ. (2006). Task and content Levy-Gigi E, Ke´ri S. (2012). Falling out of time: enhanced memory for modulate amygdala-hippocampal connectivity in emotional retrieval. scenes presented at behaviorally irrelevant points in time in Neuron 49:631–8. Posttraumatic Stress Disorder (PTSD). PLoS One 7:e42502. Smith ME. (2005). Bilateral hippocampal volume reduction in adults Levy-Gigi E, Ke´ri S, Myers CE, Lencovsky Z, Sharvit-Benbaji H, with posttraumatic stress disorder: a meta-analysis of structural MRI Orr SP, Gilbertson MW, et al. (2012). Individuals with post-traumatic studies. Hippocampus 15:798–807. stress disorder show a selective deficit in generalization of associative Soo J, Webber MP, Gustave J, Lee R, Hall CB, Cohen HW, Kelly KJ, Prezant DJ. (2011). Trends in probable PTSD in firefighters exposed learning. Neuropsychology 26:758–67. to the World Trade Center disaster, 2001–2010. Disaster Med Public Levy-Gigi E, Szabo C, Richter-Levin G, Ke´ri S. (2014). Reduced Health Prep 2:197–203. hippocampal volume is associated with impaired generalization of Spielberger CD, Gorsuch RL, Lushene RE, Vagg PR. (1983). State-Trait negative context in individuals with PTSD. [Epub ahead of print]. Anxiety Inventory (STAI). Bib 2010:180. Macken WJ. (2002). Environmental context and recognition: the role of Spitzer RL, Williams JB, Gibbon M, First MB. (1990). User’s guide recollection and familiarity. J Exp Psychol Learn Memory Cogn 28: for the structured clinical interview for DSM-III-R: SCID. 153–61. Washington, DC: American Psychiatric Press. Maren S, Phan KL, Liberzon I. (2013). The contextual brain: implica­ Spreen OA. Compendium of neuropsychological tests: administration, tions for fear conditioning, extinction and psychopathology. Nat Rev norms, and commentary. NY: Oxford University Press; 1998. Neurosci 14:417–28. Stevens JS, Jovanovic T, Fani N, Ely TD, Glover EM, Bradley B, Mayes AR, MacDonald C, Donlan L, Pears J, Meudell PR. (1992). Ressler KJ. (2013). Disrupted amygdala-prefrontal functional Amnesics have a disproportionately severe memory deficit for connectivity in civilian women with posttraumatic stress disorder. interactive context. Q J Exp Psychol 45:265–97. J Psychiatr Res 47:1469–78. Meyer EC, Zimering R, Daly E, Knight J, Kamholz BW, Gulliver SB. Tsoory MM, Guterman A, Richter-Levin G. (2010). ‘‘Juvenile stress’’ (2012). Predictors of posttraumatic stress disorder and other psycho­ alters maturation-related changes in expression of the neural cell logical symptoms in trauma-exposed firefighters. Psychol Serv 9: adhesion molecule L1 in the limbic system: relevance for stress- 1–15. related psychopathologies. J Neurosci Res 88:369–80. Moustafa AA, Gilbertson MW, Orr SP, Herzallah MM, Servatius RJ, Tversky A, Kahneman D, Choice R. (1981). The framing of decisions. Myers CE. (2013). A model of amygdala-hippocampal-prefrontal Science 211:453–8. interaction in fear conditioning and extinction in animals. Brain Cogn Vythilingam M, Blair KS, McCaffrey D, Scaramozza M, Jones M, 81:29–43. Nakic M, Mondillo K, et al. (2007). Biased emotional attention in post Murnane K, Phelps M, Malmberg K. (1999). Context-dependent traumatic stress disorder: a help as well as a hindrance? Psychol Med recognition memory: the ICE theory. J Exp Psychol 128:403–15. 37:1445–55. Orr SP, Lasko NB, Macklin ML, Pineles SL, Chang Y, Pitman RK. Wald I, Degnan KA, Gorodetsky E, Charney DS, Fox NA, Fruchter E, (2012). Predicting post-trauma stress symptoms from pre-trauma Goldman D, et al. (2013). Attention to threats and combat-related psychophysiologic reactivity, personality traits and measures of posttraumatic stress symptoms: prospective associations and moder­ Downloaded by [Northeastern University] at 06:48 28 March 2016 psychopathology. Biol Mood Anxiety Disord 2:8. doi: 10.1186/ ation by the serotonin transporter gene. JAMA Psychiatry 70:401–8. 2045-5380-2-8. Wechsler D. (1997). Wechsler Adult Intelligence Scale – Third addition. Phelps EA, O’Connor KJ, Gatenby JC, Gore JC, Grillon C, Davis M. San Antonio: The Psychological Corporation. (2001). Activation of the left amygdala to a cognitive representation of Woon FL, Sood S, Hedges DW. (2010). Hippocampal volume deficits fear. Nat Neurosci 4:437–41. associated with exposure to psychological trauma and post-traumatic Pole N, Neylan TC, Otte C, Henn-Hasse C, Metzler TJ, Marmar CR. stress disorder in adults: a meta-analysis. Prog Neuropsychopharmacol (2009). Prospective prediction of posttraumatic stress disorder Biol Psychiatry 34:1181–8.