|Home | About | Journals | Submit | Contact Us | Français|
Childhood sexual abuse (CSA) has been associated with psychopathology, particularly major depressive disorder (MDD), and high-risk behaviors. Despite grave epidemiological data, the mechanisms underlying these maladaptive outcomes remain poorly understood.
We examined whether CSA history, particularly in conjunction with past MDD, is associated with behavioral and neural dysfunction in reinforcement learning, and whether such dysfunction is linked to maladaptive behavior.
Participants completed a clinical evaluation and a probabilistic reinforcement task while 128-channel event-related potentials were recorded.
Academic setting; participants recruited from the community.
Fifteen remitted depressed females with CSA history (CSA+rMDD), 16 remitted depressed females without CSA history (rMDD), and 18 healthy females.
Participants’ preference for choosing the most rewarded stimulus and avoiding the most punished stimulus was evaluated. The feedback-related negativity (FRN) and error-related negativity (ERN)–hypothesized to reflect activation in the anterior cingulate cortex–were used as electrophysiological indices of reinforcement learning.
No group differences emerged in the acquisition of reinforcement contingencies. In trials requiring to rely partially or exclusively on previously rewarded information, the CSA+rMDD group showed (1) lower accuracy (relative to both controls and rMDD), (2) blunted electrophysiological differentiation between correct and incorrect responses (relative to controls), and (3) increased activation in the subgenual anterior cingulate cortex (relative to rMDD). CSA history was not associated with impairments in avoiding the most punished stimulus. Self-harm and suicidal behaviors correlated with poorer performance of previously rewarded–but not previously punished–trials.
Irrespective of past MDD, women with CSA histories showed neural and behavioral deficits in utilizing previous reinforcement to optimize decision-making in the absence of feedback (blunted “Go learning”). While the current study provides initial evidence for reward-specific deficits associated with CSA, future research is warranted to determine if disrupted positive reinforcement learning predicts high-risk behavior following CSA.
According to the US Department of Health and Human Services1, in 2008 alone, over 69,000 children experienced childhood sexual abuse (CSA) in the US. The National Comorbidity Survey showed that severe childhood adversity accounts for nearly 32% of psychiatric disorders2. Although CSA sequelae are heterogeneous, affective disorders are the most common outcomes in adulthood3. For example, in a birth cohort of 1,000 children, CSA involving sexual intercourse was associated with an increased odds ratio of 8.1 of developing major depressive disorder (MDD)4,5. Similarly, in a sample of adults with a history of CSA, 62% met DSM-IV criteria for lifetime MDD compared to 28% with lifetime posttraumatic stress disorder6.
CSA has also been linked to higher rates of maladaptive behaviors, including self-harm, unsafe sexual behavior, and substance abuse7,8. Although maladaptive behaviors can provide momentary relief from distress, they can have detrimental long-term implications, including increased risk of sexual revictimization9,10. Unfortunately, research examining the functional and neural mechanisms underlying maladaptive behaviors related to CSA is sparse.
Neurobiological studies have emphasized the impact of chronic stress on brain development. In particular, prolonged stress has been linked to dysregulation of the hypothalamic-pituitary adrenal (HPA) axis, leading to increased glucocorticoid release11. Excessive glucocorticoid release, in turn, has been hypothesized to impair neural plasticity in brain regions with prolonged postnatal development and/or high concentration of glucocorticoid receptors12. The anterior cingulate cortex (ACC) reaches peak volume at 10.5 years13–15 and such protracted postnatal development16 might leave it vulnerable to the neurotoxic effects of glucocorticoids11,17–18. In fact, adults reporting childhood adversities showed smaller ACC volumes compared to adults without adversity19,20. Given the role of the ACC in reinforcement learning21, ACC abnormalities following CSA might disrupt the ability to learn from positive and negative outcomes, which might underlie maladaptive decision-making. Our aim was to test these novel hypotheses.
Specifically, we investigated reinforcement learning and putative ACC abnormalities in females with a history of childhood sexual abuse occurring between the ages of 7 to 12 (see Supplement for a rationale of sample selection). A reinforcement task was used in conjunction with 128-channel event-related potentials (ERPs), which allowed the examination of electrophysiological indices of internal (error-related negativity, ERN) and external (feedback-related negativity, FRN) performance feedback. Both waveforms are thought to reflect dopamine (DA)-modulated ACC activity following negative feedback (FRN) and incorrect responses (ERN) critically implicated in reinforcement learning21. Blunted ERN/FRN may suggest decreased sensitivity to task-relevant outcomes, which might lead to deficits in reinforcement learning, including the acquisition of reinforced contingencies and the utilization of these contingencies in order to optimize decision-making in a novel context (incentive-based decision making).
Given that CSA has been strongly linked to MDD4–6, females with a history of CSA and MDD were compared to both females with past history of MDD but no CSA as well as healthy females. We hypothesized that, relative to healthy controls and a psychiatric control group, females who experienced CSA would show behavioral and electrophysiological indices of disrupted reinforcement learning. Moreover, we hypothesized that such abnormalities would be associated with both ACC dysfunction and higher rates of maladaptive behaviors.
Seventy-two participants were recruited through online and printed advertisements. Following the first screen, 56 women were eligible (see Supplement). Seven participants were excluded due to artifacts (N=6) or task non-compliance (N=1). The final sample consisted of three groups: (1) females with a history of CSA and remitted MDD (CSA+rMDD; N=15), (2) females with remitted MDD but no trauma (rMDD; N=16), and (3) females without history of psychopathology or trauma (Controls; N=18). Participants were right-handed with no significant medical or neurological conditions, and were excluded if they reported current mood disorders, or current/past psychotic symptoms, somatoform disorders, personality disorders, lifetime substance dependence, substance abuse within the past six months, seizures, or use of antidepressant medication in the past two months.
Inclusion criteria for the CSA+rMDD group included three or more episodes of coerced sexual contact by at least one male perpetrator between the ages of 7 to 12 years (mean duration: 3.00±2.20 years; see Supplement). CSA+rMDD individuals could not report concurrent physical or emotional abuse during childhood or adolescence on the Traumatic Antecedents Questionnaire (TAQ)22. Groups did not differ in frequency of being disciplined [x2(8)=14.36, P>.07] or exposure to family violence in childhood [x2(4)=7.53, P>.11].
Both the CSA+rMDD and rMDD group met criteria for past MDD, as assessed by the Structured Clinical Interview for DSM disorders (SCID)23. CSA+rMDD and rMDD groups were matched for the number of past MDD episodes, time elapsed since last episode, prior psychological or pharmacological treatment, and comorbidity (Table 1). Based on the SCID and TAQ, controls did not have any current or past psychiatric disorder or lifetime trauma. The study was approved by the Harvard University IRB. Volunteers provided written informed consent and were reimbursed $75.
In a first session, participants completed the TAQ, SCID, Beck Depression Inventory-II (BDI-II)24, Snaith-Hamilton Pleasure Scale (SHAPS)25, and Perceived Stress Scale (PSS)26. The Youth Risk Behavior Survey-Adult version (YRBS)27 was administered to assess frequency of self-harm, risk-taking, violent behavior, unsafe sexual activity, and dysfunctional eating habits. The Coping Inventory Stressful Situation (CISS)28 probed adaptive and maladaptive coping strategies.
On a separate day, EEG data were collected. The state versions of the Spielberger’s State-Trait Anxiety Inventory (STAIX-2)29 and Positive and Negative Affect Schedule (PANAS)30 were administered immediately before and after the EEG recording. The Digit Span Task31 was administered to assess working memory capacity.
During EEG, participants completed the Probabilistic Stimulus Selection Task (PSST)32 to probe reinforcement learning. The PSST consists of a learning phase with two to six training blocks (60 trials/block) to examine explicit learning from positive and negative feedback, and a test phase with a single block (120 trials) to assess decision-making based on previously rewarded or punished contingencies.
In the learning phase, participants were randomly presented in each trial with one of three different stimuli pairs (A–B, C–D, E–F) of Snodgrass images on a computer screen33. Images (1200 ms) were preceded by a fixation cross (1000 ms) and followed by a blank screen (jittered 350, 450, 550 ms). Participants were instructed to press a key to the image that had the highest chance of being correct as quickly and accurately as possible. After each response, feedback (600 ms) was given to indicate correct (“Correct! Well done!” in blue font) or incorrect responses (“Incorrect! Concentrate!” in red font) followed by a jittered inter-trial interval (300–700 ms, in 100-ms increments). Feedback was probabilistic: for the most reliably rewarded AB trials, choosing A led to 80% positive and 20% negative feedback, while choosing B yielded 20% positive and 80% negative outcomes. For CD trials, choosing C led to 70% positive and 30% negative feedback and choosing D to 30% positive and 70% negative feedback. Contingencies for the least reliable stimulus type EF were 60:40%. During this phase, participants learned to choose stimuli A, C, and E more frequently than B, D, or F. Favoring A over B can be achieved either by learning that stimulus A usually leads to positive feedback (“Choose A” = learning from reward), stimulus B usually leads to negative feedback (“Avoid B” = learning from punishment), or both. The learning phase was completed when participants reached the performance criterion of 65% accuracy for AB, 60% for CD, and 50% for EF. If performance criteria were not met, participants completed all six blocks before transitioning to the test phase. To ensure acquisition of learned contingencies, participants were excluded if they achieved less than 50% of correct AB choices in half of the training blocks (see Supplement).
In the test phase, the three previously learned or “familiar” pairs (AB, CD, EF) were intermixed with 12 “novel” combinations of all possible stimuli pairs. No feedback was given, as the test phase examined incentive-related decision making. “Go learning” is measured by the choice of the most rewarded stimulus A in AC, AD, AE, and AF trials. “NoGo” learning is measured by the avoidance of the most punished stimulus B in BC, BD, BE, and BF trials. The test phase (fixation: 1000 ms; stimulus display: 3000 ms; inter-trial-interval jittered: 900–1300 ms, in 100-ms increments) consisted of a single block with 120 trials.
The PSST was presented on a Dell PC using E-Prime 1.1. (Psychology Software Tools, Inc, Pittsburgh, PA). EEG data were recorded using a 128-channel EGI (Electrical Geodesics Inc, Eugene, OR) system within an electrically and acoustically shielded room using a 250-Hz sampling rate (0.1–100 Hz bandpass filter) and referenced to Cz. Impedances were <100 kΩ.
Groups differed in BDI and PSS scores. As both measures were highly correlated (r=.79, P<.001), main analyses entered BDI-II scores as covariates to avoid collinearity. Analyses entering PSS as a covariate yielded comparable results (available upon request). Greenhouse-Geisser correction was used when appropriate; significant ANCOVA findings were followed-up with Fisher’s Least Significant Difference (LSD).
For the training phase, mixed ANCOVAs (covariate: BDI scores) with Group (CSA+rMDD, rMDD, controls) and Condition (AB, CD, EF) as factors were run separately for accuracy and reaction time (RT). [For RTs, a log-transformation was applied to normalize the distribution, and analyses were performed on log-transformed data; untransformed data are presented for simplicity]. In the test phase, two sets of analysis evaluated whether participants relied on learned positive or negative reinforcement to optimize outcomes in the absence of explicit feedback. First, univariate ANCOVAs evaluated group differences in accuracy and RT on AB trials, which represent the most distinctly reinforced stimuli. However, performance in AB trials cannot be unambiguously linked to positive or negative reinforcement learning. Therefore, in a second set of analyses, Group x Condition ANCOVAs were performed on performance from trials including stimulus A or B paired with all other possible stimuli (hereafter referred to as “A Novel” and “B Novel”) as condition.
ERP analyses were conducted using established procedures34,35 (see Supplement). ERPs were computed time-locked to positive or negative feedback (FRN) for the learning phase and time-locked to responses (ERN) in the test phase.
For the learning phase, the FRN was evaluated at frontocentral electrodes (Fz, FCz, Cz) as the most negative peak between 200–400ms following feedback36,37, which was subtracted from the directly preceding positive peak (0–400ms). Accordingly, larger positive values indicate a larger (i.e., more negative) FRN. A mixed Group x Feedback (correct, incorrect) x Electrode (Fz, FCz, Cz) ANCOVAs was performed on FRN amplitude.
For the test phase, only the ERN (and its counterpart elicited by correct responses, the correct-response negativity, CRN) were computed as no feedback was given. The ERN (and CRN) was defined as the most negative deflection 40–80ms post-response at frontocentral electrodes (Fz, FCz, Cz, Pz)38,39,40. Peak-to-peak amplitudes were determined by subtracting the amplitude of the most negative peak in 40–80ms post-response from the amplitude of the directly preceding positive peak (0–80ms). Larger positive values indicate a larger (i.e., more negative) ERN and CRN. Similar to the behavioral analyses, ERN/CRN responses to AB familiar trials were first evaluated in a Group x Response (ERN, CRN) x Electrode (FCz, Fz, Cz, Pz) ANCOVA. Then, mixed Group x Condition (A Novel, B Novel) x Response x Electrode ANCOVA were conducted.
Low Resolution Electromagnetic Tomography (LORETA)41 was used to estimate intracerebral sources. To test a priori hypotheses, current density was extracted (training: 200–400ms; test: 40–80ms) from structurally defined regions-of-interest for cognitive and affective subdivisions of the ACC (see Supplement). Current density was averaged within the respective time frame, intensity-normalized to unity, and log-transformed. Values were then entered in Group x Response x ACC Cognitive Subdivision (Brodmann area 24′ and 32′) and Group x Response x ACC Affective Subdivision (Brodmann area 24, 25, 32) ANCOVAs.
Table 1 summarizes demographics and clinical measures. Participants were on average 29 years old, single, completed high school or college, and reported an average annual income of ≤$50,000. Groups did not differ in age, marital status, or income, but the CSA+rMDD group included a smaller percentage of Caucasians (assessed by self-report) and tended to have fewer participants who completed college (Table 1). CSA+rMDD and rMDD were matched for past number of MDD episodes, time elapsed since last MDD episode, and comorbidity.
Relative to both controls and rMDD, CSA+rMDD reported significantly higher levels of recent stress (PSS) and depressive symptoms (BDI) (Ps<.03), whereas rMDD and controls did not differ (P>.15). No within- or between-group differences were found in pre-/post-EEG state anxiety or positive affect. Overall, participants experienced a decrease in negative affect over time (see eTable 1 in Supplement). Groups did not differ in working memory capacity (Table 1).
Relative to controls, the CSA+rMDD group reported significantly higher use of maladaptive behaviors and emotion-oriented coping strategies, including self-harm and suicidal ideation/attempts, perpetrating violence, unsafe and high-risk sexual behaviors and dysfunctional eating patterns associated with drastic changes in body weight (Ps<.02) (Table 1). Importantly, these behaviors were also significantly more common in CSA+rMDD compared to rMDD (self-harm/suicide, violence-related behavior, unsafe sexual behavior, body weight; Ps<.04). No differences between controls and rMDD emerged (Ps>.50). Across the entire sample (N=49), higher level of emotion-oriented coping was related to violence-related behavior in past 12 months (r=0.29, P=.04) and dysfunctional eating (r=0.30, P=.04).
Table 2 summarizes the average number of blocks completed, as well as accuracy/RT scores in the learning phase. On average, participants completed three training blocks with no differences between groups (F(2,46)=.13, P=.88; see Supplement).
When considering accuracy, an ANCOVA (covariate: BDI) on percentage accuracy revealed Condition (F(2,90)=5.06, P=.008) and Group x Condition (F(4,90)=2.45, P=.05) effects. Consistent with the probabilistic reinforcement schedule, post-hoc tests confirmed that EF trials (.62±.20) had lower accuracy than AB (.76±.17) or CD (.70±.21) trials (Ps<.01). A trend for higher AB than CD accuracy was also seen (P=.06). Follow-up tests for the Group x Condition interaction were not significant (Ps>.13).
An analogous Group x Condition ANCOVA on log-transformed RT data revealed a main effect of Condition (F(2,90)=3.72, P=.03), due to slower responses on EF relative to both AB and CD trials (Ps=.02), and a main effect of Group (F(2,45)=3.27, P=.05). Post-hoc tests revealed slower RT in the CSA+rMDD group compared to rMDD (P=.02) but no other differences (Ps>.16). In sum, groups reached similar learning accuracy, although CSA+rMDD participants were generally slower than rMDD.
An ANCOVA for accuracy on AB trials revealed a significant Group effect (F(2,45)=3.51, P=.04), with CSA+rMDD showing lower accuracy on these familiar, most distinctly reinforced trials relative to both rMDD (P<.02) and controls (P<.05). Similarly, an ANCOVA for RT on AB trials showed a significant Group effect (F(2,45)=6.12, P=.004) with CSA+rMDD responding slower than rMDD (P<.003). No other differences emerged (P>.12).
A Group x Condition (A Novel, B Novel) ANCOVA for accuracy revealed no significant effects (Ps>.11). When examining reward and punishment trials separately, analyses revealed, however, group differences on trials requiring to rely on previously rewarded information (A Novel: F(2,45)=4.02, P=.03) but not previously punished information (B Novel: F(2,45)=0.17, P>.85; Figure 1). Post-hoc tests revealed that CSA+rMDD showed significantly lower accuracy on A Novel trials compared to both controls (P=.05) and rMDD (P=.007), with no differences between the latter (P>.37). When considering RT, an analogous Group x Condition (A Novel, B Novel) ANCOVA showed only a main effect of Condition (F(1,45)=19.88, P<.001), due to shorter RT for A Novel than B novel stimuli (P<.001).
Across the CSA+rMDD and rMDD group (N=31), more frequent engagement in self-harm and suicidal behavior correlated with slower responses on trials requiring to rely on familiar (AB: r=0.38, P=.03) and previously rewarded trials (A Novel: r=0.40, P=.03). Both correlations were confirmed when using non-parametric (Spearman) correlations (AB: Spearman r=0.48, P=.006; A Novel: Spearman r=0.44, P=.02). Findings did not survive a Bonferroni correction (P=.05/16= .003; see Supplement) and thus should be considered preliminary.
The Group x Feedback x Electrode ANCOVA yielded a main effect of Feedback (F(1,42)=7.09, P=.01) with the expected larger (i.e., more negative) FRN following incorrect than correct responses (P=.004). No other effects emerged (Ps>.12).
A Group x Response x Electrode ANCOVA showed a significant Group x Response effect (F(2,26)=4.05, P=.03). Similar to previous research42 and as no effect of Electrode emerged, follow-up analyses focused on Cz. Significant group differences emerged on correct (F(2,26)=9.74, P=.001) but not incorrect (F(2,26)=2.48, P=.10) AB trials. On correct AB trials, controls showed smaller (i.e., less negative) CRN amplitudes than CSA+rMDD (P=.004) and rMDD (P<.001); CSA+rMDD and rMDD did not differ (P=.54) (Figure 2).
An analogous Group x Condition (A Novel, B Novel) x Response x Electrode ANCOVA showed no significant effects. [It should be noted that ERN/CRN are indices of error monitoring and thus are expected to be more pronounced in response to familiar than novel pairings.]
Similar to the behavioral and scalp data, no group differences emerged in ACC activity during the learning phase.
ANCOVAs for AB trials did not yield group differences.
A Group x Response x ACC Affective Subdivision for A Novel trials revealed a main effect of ACC Affective Subdivision (F(2,66)=46.45, P<.001), Group x Response (F(2,33)=6.73, P=.004), and Group x Response x Subdivision (F(4,66)=4.54, P=.014) interactions. Follow-up analyses of the triple interaction revealed a Group x Response interaction for the subgenual ACC (BA 25; F(2,33)=7.87, P=.002) and rostral ACC (BA24; F(2,33)=5.88, P=.007). Groups differed in subgenual activation on correct A Novel trials (F(2,33)=3.93, P=.03) but not incorrect A Novel trials (F(2,33)=1.79, P=.18). Relative to rMDD, CSA+rMDD had significantly higher activation in the subgenual ACC (P=.01); they also tended to show higher activation than controls (P=.07; Figure 3). No differences were found between controls and rMDD (P=.33). Further analyses of the rostral ACC did not yield between-group differences.
A Group x Response x ACC Cognitive Subdivision showed a main effect of ACC Cognitive Subdivision (F(1,33)=13.00, P=.001) and Group x ACC Cognitive Subdivision interaction (F(2, 33)=3.84, P=.03). Follow-up analysis did not yield group differences. No group differences in ACC affective or cognitive subdivisions emerged for B Novel.
The goal of this study was to investigate putative disruption in positive and negative reinforcement learning in women with a history of CSA, and whether such dysfunctions are related to maladaptive behavior. Several novel findings emerged. First, groups did not differ in their ability to acquire the probabilistic reinforcement schedule. Results from the test phase revealed, however, that women in CSA+rMDD group had lower accuracy and slower RT on familiar AB trials compared to both rMDD and controls. Although familiar AB trials do not allow us to disentangle whether participants made choices guided primarily by their positive or negative history of reinforcement, these distinctly reinforced trials provide a critical test of the utilization of reinforced information in the absence of explicit feedback. Importantly, additional analyses clarified that women with CSA histories choose less reliably the most positive stimulus A in the test phase, indicating impaired performance in trials requiring reliance on previously rewarded information (A Novel). Notably, groups did not differ in their avoidance of the most negative stimulus B, suggesting that the CSA group was not affected in trials requiring reliance on previously punished information (B Novel). Highlighting the clinical relevance of these findings, slower RTs in A Novel (and AB) trials correlated with self-harm and suicidal behavior. As correlations did not survive a Bonferroni correction, these latter findings should be considered preliminary. Collectively, these findings confirm our first hypothesis that women with a history of CSA demonstrate disrupted reinforcement learning, and highlight that these impairments are specific to trials that require reward-based reinforcement learning. Of note, lack of group differences (at both the behavioral and neural level) in (1) B novel trials, (2) the acquisition of the reinforcement contingencies, and (3) working memory performance indicate that blunted Go learning in the CSA group was not due to global cognitive impairments. Finally, CSA+rMDD and rMDD groups were matched for the number of prior depressive episodes and length since last depressive episode, and analyses entered BDI-II (and PSS) scores as covariates, suggesting that current depressive symptoms or MDD history did not influence outcomes.
Our second hypothesis focused on neural indices of reinforcement learning. As ERN and CRN index correct and incorrect responses, respectively, to known stimuli, largest (i.e., most negative) amplitudes were expected on incorrect familiar trials. Compared to controls, women with CSA histories showed more negative amplitudes in response to correct AB trials suggesting a more error-like response following correct answers. Although intriguing, it is important to note that the rMDD group showed a similar pattern, suggesting that this electrophysiological marker of disrupted reinforcement learning might not be specific to CSA.
More specificity with respect to CSA emerged from the source localization analyses, in which the CSA group showed increased activation in the affective but not cognitive subdivision of the ACC during “Go” learning trial relative to both rMDD, and, to a lesser extent, controls. Specifically, women with a history of CSA demonstrated increased subgenual ACC activation during correct responses on trials that required reward-based decision making (A Novel). The affective ACC subdivision has extensive connections to limbic and paralimbic structures (e.g., amygdala, nucleus accumbens, orbitofrontal cortex) and is thought to play a key role in stress responsivity, emotional responding, and evaluation of feedback salience43,44. Thus, we speculate that CSA+rMDD may experience higher level of emotional arousal on trials requiring reliance on previously rewarded information which, in turn, may interfere with adaptive decision-making. Notably, the onset of CSA in this sample (7–12 years) coincides with a time period in which the ACC undergoes significant change14,15. Prolonged postnatal development might thus leave the ACC vulnerable to the effects of glucocorticoids18,19. Consistent with these arguments, maltreated adults with MDD exhibited reduced volume in the affective ACC subdivision, and such volume reduction correlated with maltreatment severity and cortisol levels20. Thus, the present findings add to emerging evidence indicating that the affective ACC subdivision may be affected by early life stress.
Although group differences emerged in the utilization of previously reinforced contingencies, groups did not differ on the behavioral, scalp and brain level in their ability to initially acquire the probabilistic reinforcement schedule. During training, participants showed similar accuracy levels and required an equal number of trials before transitioning to the test phase. As expected, FRN amplitude was increased (i.e., more negative) following incorrect than correct feedback, indicating that participants displayed similar reward and punishment responsivity. This was also reflected in similar ACC activation in response to explicit feedback across groups. No association was found between maladaptive behaviors and any of the behavioral or neural measures during training. It can be concluded that participants successfully acquired reinforcement contingencies, and that a CSA history does not affect the ability to learn from explicit positive and negative feedback.
Across level of analyses, the current findings suggest that early sexual abuse experiences are related to deficits in incentive-based decision making and, in particular, reduced “Go” learning. Results fit prior evidence suggesting disrupted reward processing following maltreatment. For example, Guyer and colleagues45 found that children with a history of maltreatment did not modulate reaction time as a function of the likelihood of receiving reward. In a longitudinal sample, adults with childhood maltreatment rated reward-predicting cues less positively and showed reduced anticipatory reward activity in the left pallidus, a brain region implicated in goal-directed behavior46.
In our study, trauma history did not affect the use of previously learned punishment information to a novel context (B Novel trials). In addition, while reduced sensitivity to punishment (as indexed by smaller ERN amplitude) emerged in individuals with high impulsivity47, risk-taking48, and externalizing behavior49, such outcomes were not associated with maladaptive CSA sequelae. Instead, frequent self-harm and suicidal behavior was related to slower RT on trials requiring incentive-based decision-making, including integrating information of previously rewarded trials.
Consistent with clinical evidence that adults with CSA frequently engage in maladaptive behaviors despite negative outcomes7,9, the current CSA+rMDD sample reported elevated levels of self-harm, violence-related behavior, unsafe sexual behavior, and dysfunctional eating. Moreover, violent behavior and dysfunctional eating were more frequently used when individuals adopted an emotion-oriented maladaptive coping style. Although maladaptive behaviors can provide initial relief from CSA-related distress, they also form a primary predictor of future sexual victimization9,10. The pathways linking CSA to later revictimization are a growing concern as 30–50% of individuals with CSA experience sexual violence later in life51,52. Our aim was to examine if high-risk behaviors commonly seen in adults with CSA histories were associated with disrupted reinforcement learning. Frequency of self-harm/suicidal behaviors was significantly related to slower responses in trials that required incentive-based decision making (AB) as well as Go learning (A Novel trials). As maladaptive behaviors were assessed retrospectively, future studies are needed to examine whether disrupted positive reinforcement learning predicts future high-risk behaviors and revictimization.
Some limitations should be acknowledged. First, CSA experiences were based on retrospective reports, and were not externally validated by police or court reports. However, careful assessments of adverse childhood events were conducted as part of the initial clinical assessment. All participants were able to recall central details of the abuse (e.g., age at onset, frequency). Second, although this study excluded participants with other childhood adversities, no causal inference between CSA and disrupted reinforcement learning can be made. The ACC develops over an extended period of time and may therefore be vulnerable to other environmental insults. Third, although follow-up analyses guided by our a priori hypotheses revealed that the CSA+rMDD group had lower accuracy on A Novel trials – but not B Novel trials – compared to both controls and the rMDD group, it is important to emphasize that the Group x Condition interaction was not significant. Thus, the specificity of this behavioral finding is limited. Finally, although MDD is a common outcome of CSA, our findings cannot be generalized to women with CSA histories in general; in addition, in the current study, lifetime somatoform or personality disorders represented exclusion criteria. Future studies should include a CSA group without psychopathology to further investigate CSA-specific effects in reinforcement learning.
Behavioral and source localization results provide preliminary evidence for deficits in relying on previously reinforced information to optimize decision making following CSA. Performance on trials involving incentive-based decision making, including learned reward contingencies, was associated with more frequent engagement in self-harm/suicidal behavior. No group differences emerged when participants needed to rely on negatively reinforced information, suggesting that maladaptive behavior may not be related to difficulties in using punishments to guide decision making. Future studies will need to confirm the role of disrupted positive reinforcement learning and further explore neurobiological dysfunctions as potential mechanisms implicated in high-risk behavior.
Disclosures: Dr. Pechtel reports no biomedical financial interests or potential conflicts of interest. Dr. Pizzagalli has received consulting fees from ANT North America Inc. (Advanced Neuro Technology), AstraZeneca, Shire, and Ono Pharma USA, as well as honoraria from AstraZeneca for projects unrelated to this study. Preliminary findings from this study have been presented in poster form at the Annual Meeting of the Association for Cognitive and Behavioral Therapies, November 15th, 2010, San Francisco. Funding for this study was provided by a research fellowship by the German Research Foundation (Deutsche Forschungsgemeinschaft) to PP and grants from NIMH (R01 MH68376, R21 MH078979) to DAP. Dr. Pechtel had full access to all of the data in the study, performed the statistical analyses, and takes responsibility for the integrity of the data and the accuracy of the data analysis. Funding agencies were not involved in the design and conduct of the study; in the collection, management, analysis, and interpretation of the data; and in the preparation, review, or approval of the manuscript.