Resilience is a dynamic process of recovery after trauma, but in most studies it is conceptualized as the absence of specific psychopathology following trauma. Using the large emergency department AURORA study (n=1,865, 63% women), we took a longitudinal, dynamic and transdiagnostic approach to define a static resilience (r) factor, that could explain >50% of variance in mental wellbeing 6-months following trauma, and a dynamic r-factor, which represented recovery from initial symptoms. We assessed its neurobiological profile across threat, inhibition, and reward processes using fMRI collected 2-weeks post-trauma. Our neuroimaging results suggest that resilience is promoted by activation of higher-level cognitive functioning and salience network regions in response to reward, whereas resilience is hampered by top-down attention and default mode network activation to threat and reward. These findings serve to redefine the concept of resilience as both dynamic and multifaceted, and identify mechanisms that promote resilience after trauma for early interventions.
Article
Rethinking resilience using the r-factor: Neurobiological mechanisms of static and dynamic components of resilience following trauma
https://doi.org/10.21203/rs.3.rs-3115922/v1
This work is licensed under a CC BY 4.0 License
Journal Publication
published 22 Apr, 2024
Version 1
posted
You are reading this latest preprint version
In recent years with high global levels of stress and trauma, the concept of resilience has accrued increasing interest, among scientists and the public alike, as a positive outlook after hardship or the ability to quickly recover from difficulties. Resilience is conceptualized in different ways and its scientific definition debated. It is defined in Webster as “the capability of a strained body to recover its size and shape after deformation caused especially by compressive stress”,1 and is thought to require a set of complex and dynamic processes that allow individuals to maintain psychological wellbeing in the face of adversity2,3. However, neurobiological research in the field of trauma, including our own work, has predominantly studied resilience as the absence of posttraumatic stress disorder (PTSD) or other trauma-related disorders in the aftermath of trauma, or have used a trait-like resilience measure, such as the Connor-Davidson resilience scale (CD-RISC)4–7. While this prior work has been essential in understanding trauma resilience and its underlying neurobiological mechanisms, its unidimensional approach does not capture the complete picture of resilience and it is unclear whether the identified mechanisms are related to protection from one specific disorder or more general wellbeing. Moreover, the use of a single time point after trauma cannot account for dynamic processes of resilience, and the need for longitudinal studies on resilience as a dynamic process of effective adjustment has been highlighted8. Furthermore, a more complete understanding of the neurobiology of resilience across domains and its dynamics is needed for early interventions aiming to boost resilience in the aftermath of trauma.
The neurobiology of resilience has been studied and neural contributors to trauma resilience with a longitudinal element as defining factor were recently reviewed3. With the caveat that most studies were small, a general conclusion from this review was that lower threat reactivity in the amygdala and greater activation in prefrontal and hippocampal regions were associated with greater levels of resilience. Furthermore, the predictive value of activation in the salience network, including the dorsal anterior cingulate cortex (dACC) and insula, depended on whether imaging data were collected prior to or following trauma: specifically, resilience was predicted by lower salience network activation prior to trauma and greater activation following trauma3. When considering early interventions, using the early period following trauma may provide the best window of opportunity for early interventions in a civilian population. However, in a sample of individuals at high risk of trauma exposure, for example, because of their profession, pre-trauma markers may have more clinical value and some true prospective studies on the neurobiology of resilience in the military or police force have been conducted9–11. One notable neuroimaging study included 185 police academy recruits before they experienced primary trauma as part of their duty. Using an approach-avoidance task, greater prefrontal emotional regulation (dorsal and medial frontal pole activity and anterior prefronal cortex; PFC) predicted lower PTSD symptoms4. This study, as well as most other studies to date, used threat- or inhibition-related paradigms3. Because few studies in the early aftermath of trauma have investigated reward-related processing12,13,14, there is a clear need to include reward in addition to threat and inhibition for a more complete picture in future studies of resilience. A second limitation of prior research includes small sample sizes and limited power for unbiased (whole brain) analyses. Indeed, most studies used region of interest analyses and focused on brain regions of the threat neurocircuitry (including the amygdala, hippocampus, ventromedial prefrontal cortex). Together, these limitations of prior work suggest the need for well-powered studies for unbiased whole brain analyses across task constructs.
Another major limitation of most prior work on resilience is its conceptualization as the absence of specific psychopathology (usually PTSD) in the aftermath of trauma. A common factor for resilience across multiple domains of adverse post-trauma sequalae has not been explored so far, but could accelerate research by providing important insights into protective mechanisms across such sequalae and promote the development of potential early interventions for individuals at risk. The idea of using one factor to explain variance across domains of psychopathology is not new. Following the g-factor for general intelligence15, the general factor of psychopathology or p-factor was defined to explain cross-domain risks for psychopathology16. This transdiagnostic approach was developed to overcome challenges with identifying origins, biological markers, or effective treatments for specific psychiatric disorders. The p-factor explains risk, severity and persistence of psychopathology, and is related to family psychiatric history, childhood history, and day-to-day impairment17. Neuroimaging studies have also demonstrated structural alterations associated with p, such as reduced PFC gray matter and lower white matter integrity across tracts [for example18,19 and reviewed in20]. Task-based functional Magnetic Resonance Imaging (MRI) studies in youth showed lower prefrontal/executive region control and greater salience network activation21,22. These studies show that a common transdiagnostic factor that is related to a variety of psychiatric disorders and biological correlates could be defined and widely applied, and suggest the relevance of investigating a common factor for resilience.
Following the rationale for the p-factor, we aimed to investigate resilience across multiple domains of mental wellbeing in the first six months following exposure to a traumatic experience (schematic overview; Fig. 1). Rather than defining a set of transdiagnostic features that contribute to psychopathology, we instead sought to understand transdiagnostic features associated with resilience in the face of a major stressor. The resilience factor, or r-factor, we sought to define is unique in that it encompasses the early aftermath of trauma and includes two components to account for the change following trauma: 1. static r-factor, representing mental wellbeing 6-months post-trauma and corresponds to definitions of absence of unidimensional symptoms in prior studies. 2. dynamic r-factor to account for the dynamic nature of resilience, which represents recovery from initial symptoms early post-trauma period (2-weeks) relative to 6-months post-trauma when resilient and high-risk groups diverge23. After isolating these shared features of resilience, we aimed to identify unique profiles associated with resilience to some chronic outcomes but not others, and brain-based predictors of these more circumscribed components of resilience. In addition to the main r-factor, we explored the possibility that there may be secondary factors that explain significant variance in the aftermath of trauma, potentially representing different types of resilience.
We used data collected as part of the multisite emergency department (ED) study AURORA24, the largest post-trauma study to date, which followed 2,943 civilians brought to EDs post-trauma. Resilience factors were defined based on item-level data on mental wellbeing across domains of anxiety, depression, PTSD, sleep, impulsiveness, alcohol and nicotine use collected 2-weeks and 6-months following trauma. We focused on resilience, rather than the individual symptoms, because the majority of individuals show few symptoms or a quick natural symptom recovery following trauma, indicating a strong “bounce back.” If from this research we can understand the neural underpinnings of this return to adaptive functioning, we can then identify targets for interventions to boost resilience in the early post-stressor time window. In this investigation, we considered the dynamic nature of resilience by defining both static (timepoint: 6-months) and dynamic (change from 2-weeks to 6-months) resilience. This time window was chosen because 6-months post-trauma is an inflection point by which resilient and high-risk groups are known to diverge from one another23. We hypothesized that we would identify one main r-factor explaining resilience to global stress. We investigated the r-factor scores in relation to age, sex, childhood trauma and trait resilience, and expected higher resilience scores in older age, male, individuals with lower childhood abuse and neglect and greater scores on trait resilience.
We then assessed the functional neurobiological profile of the resilience factor across inhibition, threat, and reward-related processes using magnetic resonance imaging (MRI) scans which were collected for a subset of individuals (n=445) at 2-weeks post-trauma. We expected that r would be related to activation in brain regions previously implicated in resilience3 or post-trauma sequalae during inhibition25, threat26, and reward processing27 and selected as regions of interest (ROIs) in the AURORA study28. We tested the hypotheses that greater inhibition-related activation in the hippocampus and vmPFC, lower threat-reactivity in the amygdala and hippocampus, greater threat-reactivity in BA25 and BA32, and greater reward-related activation in the nucleus accumbens, amygdala, and orbitofrontal cortex (OFC) were related to greater static and dynamic r-factor scores. Finally, unbiased whole brain analyses were conducted to identify other brain regions implicated in static and dynamic resilience. Secondary resilience factors (eigenvalues >2) were also created and the same ROI and whole brain correlation analyses were performed.
Participants
Data for the current analyses were collected as part of the multisite ED AURORA study. Trauma-exposed civilians brought to one of 29 participating EDs across the United States were recruited for this large, longitudinal study (details in24,28). This investigation included n=2,772 AURORA participants with clinical item-level data at 2-weeks or 6-months recruited from 09/2017 to 12/2020 (Final freeze 4 Psychometric release). Missing data values were excluded listwise, resulting in n=1,835 (1,175 women; Table 1) participants with complete 6-month item level clinical data usable for analyses.
Participants recruited for AURORA at one of the 22 ED sites that funneled participants to one of the five “deep phenotyping” sites were invited to undergo an MRI scan. Functional MRI (fMRI) data were collected for n=445; the fMRI scan included a response inhibition paradigm (n=428), a fearful faces “social threat” task (n=431), and a monetary reward processing task (n=427). Functional data were excluded from analyses for anatomical concerns (n=7), lack of expected behavioral responses or available data (inhibition, n=45; threat, n=14; reward, n=26), excessive motion (any run with >15% of volumes exceeding 1mm FD; inhibition, n=33; threat, n=25; reward, n=45), technical issues during the scan (inhibition, n=14; threat, n=15; reward, n=24). Final data were available for n=385 for at least one of the tasks (n=329 for inhibition, n=370 for threat processing, and n=325 for reward processing). Of these, associations with resilience factor scores were investigated for n=260 individuals who had scores available (Table 1; n=215 for inhibition, n=249 for threat processing, and n=214 for reward processing). Dynamic resilience factor scores (both complete 2-week and 6-month clinical data) were investigated for respectively n=189, n=221, and n=189.
Principal component analyses defining resilience factor
Using principal component analyses (PCA) of self-reports from 6-months post-trauma, three factors were extracted based on eigenvalues greater than 2 (Table 2). The main resilience factor for global stress, or r-factor reflected general mental wellbeing following trauma (eigenvalue=23.52). Two additional factors were identified, one labeled as reminder acceptance (eigenvalue=2.39), which represents low levels of re-experiencing and avoidance of reminders of the traumatic event, and a second labeled behavioral control (eigenvalue=2.17), which represents low levels of the impulsivity, risk taking, loss of control and feeling rejected, that often follow a major stressor. Notably, the highest loadings for each of the three factors bridged multiple instruments and domains, suggesting that the data support a transdiagnostic approach.
Evaluating demographic characteristics of the factor scores (Table 3) showed that men had greater mean scores of reminder acceptance, whereas women had greater behavioral control scores. Older age correlated with r-factor scores and behavioral control, but less reminder acceptance. Because age and sex contributed to variance of interest in the factor scores and indeed showed associations with the factor scores, they were not again included as covariates in the neuroimaging analyses. Greater childhood trauma, as measured with an abbreviated version of the Childhood Trauma Questionnaire (CTQ), correlated with lower scores for r, reminder acceptance, and behavioral control (Table 3). Correlation values for reminder acceptance were small, and not statistically significant for childhood neglect. Greater levels of trait resilience assessed by the Connor-Davidson Resilience Scale (CD-RISC) positively correlated with r, and (modestly) with behavioral control, but not with reminder acceptance (Table 3).
Dynamic resilience was calculated by estimating factor scores for the 2-week data based on the 6-month data and computing difference scores. There was a statistically significant increase in the score over time (2-weeks to 6-months) for the r-factor and reminder acceptance, but a decrease for behavioral control (Supplementary Table S1). The static and dynamic resilience scores were all statistically significantly correlated, but shared only 19% to 27% of variance (r-factor, r=0.44; reminder acceptance, r=0.52; behavioral control, r=0.46). Finally, the static and dynamic resilience scores did not significantly differ statistically between participants who were included in the neuroimaging analyses and those not (Supplementary Table S2).
ROI analyses for static and dynamic resilience
r-factor
Greater OFC reactivity to monetary reward 2-weeks post-trauma predicted resilience 6-months post-trauma (static r-factor; r=0.14, p=0.042; Fig. 2) and a greater increase in resilience from 2-weeks to 6-months post-trauma (dynamic r-factor; r=0.27, p<0.001*; Fig. 3). As indicated by the asterisk (*), the dynamic r-factor correlation survives additional correction for multiple comparisons across all nine ROIs and for the two types of resilience (18 tests, p<0.0028). There were no statistically significant associations between the static or dynamic r-factor and ROIs from the threat or inhibition tasks.
Reminder acceptance
Lower 2-week hippocampal reactivity to social threat cues predicted reminder acceptance at 6-months (static reminder acceptance; r=-0.20, p=0.002*; Fig 2) and a greater increase of reminder acceptance from 2-weeks to 6-months post-trauma (dynamic reminder acceptance; r=-0.17, p=0.012; Fig 2). There were not statistically significant predictors of static or dynamic reminder acceptance in the reward or inhibition tasks.
Behavioral control
No statistically significant correlations with static or dynamic behavioral control were observed within the a priori ROIs for any of the three fMRI paradigms.
Whole brain analyses for static and dynamic resilience
r-factor
Figure 3 and Table 4 show statistically significant whole brain correlations with static and dynamic r-factor scores across tasks. The static r-factor scores did not correlate with whole brain activation in the inhibition or threat tasks. During reward processing, the static r-factor positively correlated with activation in the right superior temporal gyrus (STG) and insula, such that greater reward-related activation in these regions 2 weeks post-trauma was related to greater levels of resilience 6-months post-trauma. Additionally, the static r-factor score negatively correlated with activation in the bilateral precuneus* and left inferior parietal lobe (IPL)* reactivity, such that less reward-related activation was related to greater resilience. As indicated by an asterisk (*), the negative correlations were robust to additional study-wide Bonferroni corrections for all whole-brain analyses, across three tasks, three factors, and static and dynamic resilience (18 tests, p<0.0028).
The dynamic r-factor showed associations with the reward and threat tasks, but no associations were observed with the inhibition task. During reward processing, the dynamic r-factor positively correlated with activation in the bilateral superior frontal gyrus (SFG)*, bilateral insula*, left superior medial gyrus (SMG)*, and dACC*, such that greater reward-related activation 2-weeks post-trauma predicted a greater increase in the r-factor between 2-weeks and 6-months post-trauma. During threat processing, the r-factor was negatively correlated with activation in the bilateral IPL* such that lower threat reactivity 2-weeks post-trauma was associated with a greater increase in r.
Reminder acceptance
There were no statistically significant results for the whole brain correlations between static or dynamic reminder acceptance and the inhibition, threat, or reward contrasts (Table 4).
Behavioral control
Figure 4 and Table 4 show statistically significant whole brain correlations with behavioral control across tasks. No correlations were observed between static behavioral control scores and neural responses to threat. During inhibition, the static behavioral control score was positively correlated with activation in the right inferior frontal gyrus (rIFG), such that greater rIFG reactivity 2-weeks post-trauma was related to greater behavioral control 6-months post-trauma. During reward processing, the static behavioral control score was negatively correlated with left pre/postcentral gyrus activation, such that less activation was related to greater behavioral control six months post-trauma.
Dynamic behavioral control analyses showed positive correlations for both reward and threat processing, but no associations with the inhibition task were observed. During reward processing, the dynamic behavioral control score positively correlated with activation in the right IPL* and right insula, such that greater activation in these regions was associated with greater increase in behavioral control from 2-weeks to 6-months post-trauma. Similarly, during threat processing, the dynamic resilience score positively correlated with activation in the bilateral IPL* and bilateral postcentral gyrus* and left precentral gyrus, such that greater threat reactivity in these regions was associated with a greater increase in behavioral control.
This is the first study to investigate a common factor for resilience that encompasses the early aftermath of trauma, and assess its functional neurobiological profile in a large longitudinal study with recently traumatized civilians. We identified a main resilience factor (r-factor) that explained more than half of the variance across domains of mental wellbeing 6-months following trauma when resilient and high-groups diverge. A dynamic r-factor was defined to account for the dynamic nature of resilience and represents recovery from initial symptoms after trauma. Like the p-factor16, our factor includes components related to externalizing (e.g., impulse control, risk taking) and internalizing behavior (anxiety, fears, depression symptoms), though no items related to thought disorders were expected to emerge after trauma and were not included in the study. A positive correlation between the r-factor and early post-trauma reward-related OFC activation was demonstrated, both for static and dynamic resilience. Whole brain analyses showed strong correlations between the r-factor and brain activation during reward processing including greater insula and superior and medial frontal gyri, and lower reward- or threat related reactivity in the IPL and precuneus. Two additional unique resilience factors were identified, one for reminder acceptance and one for behavioral control. These factors represent small but important variance in resilience to transdiagnostic symptoms and behaviors. Reminder acceptance was associated with lower threat-related hippocampal activation, whereas brain regions implicated in inhibitory control across response inhibition, reward, and threat processing paradigms predicted behavioral control. This study is not conclusive in the picture of resilience, but highlights that different brain mechanisms may contribute to different forms of resilience. The findings highlight both a primary “bounce back” factor primarily associated with individual differences in early post-stress reward processing, and additional unique aspects of resilience each subserved by a different profile of neural activity. The results serve to redefine the concept of resilience as both dynamic and multifaceted, using the large and unique post-trauma dataset we collected as part of the AURORA study. The discoveries from this investigation now generate new hypotheses and suggest novel directions for basic and clinical research.
The resilience or r-factor we identified represents resilience across clinical domains in the early aftermath of trauma. Most individuals had a positive r-factor score in line with the data that most people show resilience after a traumatic experience, and a positive correlation with trait resilience and negative impact of childhood trauma on the resilience score further validate the factor as a resilience-related phenotype. The assessment of early post-trauma neuroimaging predictors of the resilience factor across inhibition, threat, and reward-related domains showed a dominant role of reward processing in explaining global stress resilience. The response to monetary reward within the OFC, a regulatory structure implicated in encoding value and central to reward processing29,30, was positively related to static and dynamic resilience. Using unbiased whole brain analyses, a number of brain regions were positively associated with greater resilience. Greater reward-related activation in the insula predicted both greater static and dynamic resilience. The role of the insula in interoceptive awareness and subjective emotional experiences is key in the salience network for directing attention to salient stimuli31. Positive correlations were also observed with the dACC and bilateral SFG, regions for higher cognitive functioning and control, and decision making based on social or reward-related information32–36. These neuroimaging findings suggest that greater prefrontal guidance in the use of emotional and salient information for decision making and interoceptive awareness and attention to reward (vs loss) promotes global stress resilience. Our findings correspond with meta-analyses on common transdiagnostic neuroimaging markers. Gray matter decrease in areas of the salience network, i.e., bilateral insula and dACC37, and disrupted activation in the insula and dACC, as well as the left PFC, right vlPFC, intraparietal sulcus and midcingulate/presupplementary motor area38 were observed in transdiagnostic meta-analyses. Our findings represent a validation of these findings suggesting that the (functional) converse of transdiagnostic psychopathology findings predict recovery after a major trauma even though PTSD or trauma were not considered in these meta-analyses.
Conversely, lower activation in the bilateral precuneus and left IPL during reward processing was associated with greater resilience. Similarly, lower bilateral IPL reactivity during threat processing was negatively associated with the dynamic r-factor, such that lower IPL reactivity to threat predicted greater recovery. Precuneus and IPL are key nodes of the (posterior) default mode network and implicated in internally generated or top-down attention and critical regions for spatial attention39–42. It is postulated that attention from these top-down sources competes with bottom-up sources using (reward) salience (i.e., insula, ACC, OFC) to affect decision making43. Prior studies show that the engagement of similar top-down attention neurocircuitry slows down the search for salient targets44, and our findings imply that resilience is hampered by the engagement of regions within the top-down attention or default mode network to reward or threat cues, but benefits from attention to reward salience and regulatory control of higher order brain regions.
In contrast to our hypotheses, no associations were observed between global stress resilience and inhibition-related reactivity, and no prefrontal or subcortical regions from the threat neurocircuitry during threat processing were related to global stress resilience. While trauma-related resilience research, including our own work, has predominantly focused on the threat/inhibition neurocircuitry in response to threat cues, our work shows strong evidence for a globally-adaptive role of reward valuation and attention to reward cues in the early aftermath of trauma, which facilitates resilience in a transdiagnostic manner that includes both internalizing and externalizing symptoms and behaviors. This result is critical in the definition of new avenues for research into early interventions. The findings on greater dynamic resilience -which suggest recovery over time- particularly expose new potential pathways for early interventions aiming to boost resilience in the early aftermath of trauma. Potential therapeutic interventions could include attention bias training focusing on positive, rewarding cues to guide behavior. Indeed, positive attention bias training improved mental health in children and adolescents 45, supporting the notion that attentional bias to positive and not to threat-related stimuli promotes resilience. Targeted brain stimulation to engage the superior frontal/medial gyrus in individuals at risk could be another interesting approach to consider after further study of this association, though more precision is needed before this can be operationalized.
Two secondary resilience factors were identified in this study. The reminder acceptance factor explained 5.3% of the variance and was specifically associated with low levels of avoidance of trauma reminders, experiencing of disturbing memories, reliving the traumatic event, feeling upset and having strong emotional reactions when being confronted with trauma reminders. Moreover, the CD-RISC score did not correlate with reminder acceptance, further suggesting we captured a unique phenotype that warrants further examination. Lower threat-related hippocampal activation correlated with greater (increase of) resilience to trauma reminders. The hippocampus has consistently been implicated in PTSD and (trait) resilience. Greater hippocampal activation measured during response or contextual fear inhibition was associated with resilience or protective of development of PTSD6,7,25, while measured during habituation to threatening faces46 or fear renewal in a novel environment47 greater hippocampal reactivity was observed in patients with PTSD. These findings suggest that the type of paradigm is critical, distinguishing the facilitation of response by providing contextual information from encoding and retrieval of threatening stimuli. Additionally, type of symptoms could differentiate hippocampal involvement. Re-experiencing symptoms have been associated with overengagement of the hippocampus and deeper threat learning as demonstrated in a cross-sectional on emotional memory encoding48. In contrast, more distress-related symptoms are associated with poorer engagement of hippocampal memory, which is in line with our earlier publication from AURORA study using the same threatening faces task showing lower hippocampal activation in participants with PTSD symptoms at two weeks49. Our results therefore show that the trauma reminder resilience score does not equate to the absence of PTSD but is specifically associated with avoidance and re-experiencing underscoring the importance of exploring multiple dimensions of resilience that could explain prior disagreements in the research for which dimensions were mixed.
The behavioral control factor explained low levels of impulsive behaviors, such as acting without thinking and losing control, and minimal risk-taking behavior and feelings of rejection, and explained 4.8% of the total variance. This factor was not well-captured by the a priori ROIs, but instead showed strong associations with a variety of inhibition-related regions across all three fMRI tasks. The most robust finding was greater right IPL reactivity to threat and reward predicted greater recovery of behavioral control over time. The IPL is a known area for inhibitory control50, and prior studies showed that greater IPL reactivity predicted greater recovery from PTSD51, whereas lower IPL activation was associated with greater impulsive behavior in problem gamers52 and the impulsive subtype of attention-deficit/hyperactivity disorder53. We similarly demonstrated that greater inhibition-related rIFG activation during the inhibition task was associated with greater behavioral control at 6-months. The rIFG is another key region for inhibition and impulse control, and though this finding was not robust to the additional Bonferroni correction for multiple testing, it supports earlier findings of lower inhibition-related rIFG activation in PTSD54,55, alcohol craving and drinking56, and change in response with age57. Surprisingly, no other Bonferroni-corrected findings were observed for the inhibition task, though the pre/postcentral gyrus (motor cortex) was consistently implicated in behavioral control: Lower motor cortex reactivity to reward was related to static resilience and more motor cortex reactivity to threat was associated with a greater increase or more recovery of impulsive control from 2-weeks to 6-months. These findings could imply behavioral control can be measured earlier as motor control or less jumpiness, though more research into this area is needed to allow for better interpretation of these relations, highlights interesting new directions for study.
Limitations and future directions
This study has several limitations. One important limitation is that scans were collected 2-weeks post-trauma, and brain responses possibly or partly represent adaptations in this early post-trauma period. We addressed this by including 2-week data in the dynamic analyses and we are interested in mechanisms of change from this 2-week timepoint. Moreover, 2 weeks post-trauma provides an ideal window for intervention for those at risk after experiencing a traumatic event and therefore understanding the neurobiology of resilience early after trauma may have great clinical relevance. It is also important to note that our dynamic r-factor captures the change from early symptoms following trauma, but it does not capture dynamics in resilience that could change with repeated trauma or other forms of stress. Another limitation is that the results are not yet externally replicated. It would be important to see if similar patterns arise in a different dataset before, for example, moving to targeted neuromodulation of regions implicated here.
In this largest study on recently traumatized civilians, we identified a common factor for resilience, the r-factor, that explains more than half the variance of mental wellbeing in the first six months following trauma exposure. Our definition of the r-factor and its neural correlates is not conclusive, but instead provides important novel insights into the structure of mental wellbeing after trauma exposure, and may accelerate research on resilience and novel early interventions following trauma. Our data suggests that global stress resilience is promoted by externally-focused attention to reward, as evidenced by greater reward-related reactivity in higher order regions for salience detection and value-based decision making. Conversely, top-down internally-generated attention to reward or threat as indicated by recruitment of posterior default mode network regions may hamper resilience. These findings provide interesting new directions for resilience research including the use of a common factor across domains and investigating reward processing as a central construct for resiliency. The results from this study suggests that attentional bias training, mindfulness, brain stimulation or neurofeedback aiming to enhance reward-focused salience detection and value-based decision making are promising avenues to explore to mitigate the adverse consequences of trauma by boosting general stress resilience.
Participants
Participants were recruited as part of the multisite ED AURORA study (details in24,28. Participants with trauma exposure, defined for this study as a traumatic event requiring evaluation at an ED, were approached within 72 hours of their event. Participants who experienced certain traumatic events, including a motor vehicle collision, high fall (>10 feet), physical assault, sexual assault, or mass casualty incidents automatically qualified for study inclusion. Other participantsqualified for study inclusion after experiencing other types of trauma if: a) the individual responded to a screener question that they experienced the traumatic exposure as involving actual or threatened serious injury, sexual violence, or death, either by direct exposure, witnessing, or learning about the trauma and b) the research assistant agreed that the traumatic exposure was a plausible qualifying event.
Demographic and clinical data collection
Demographic data was collected after enrollment in the ED and included race/Hispanic ethnicity, sex assigned at birth, marital status, income, education level and employment. Additionally, childhood abuse and neglect was assessed with the Childhood Trauma Quesionnaire-Short Form (CTQ-SF) and trait resilience was assessed with the Connor-Davidson Resilience Scale (CD-RISC).
Item-level clinical data were collected with flash surveys at 2-weeks and 6-months post-trauma, and raw scores for 45 items were used for the analyses (items listed in Table 2). From the Patient-Reported Outcomes Measurement Information System (PROMIS)58,59, nine items were included for depression, four items for anxiety, and two for sleep. PROMIS items were rated on a scale from 1 (none of the time) to 5 (all or almost all the time). Twenty items were included from the PTSD checklist for DSM-5 (PCL-5)60, a well-validated questionnaire on PTSD symptom presence and severity where participants are asked to rate their symptoms on a scale from 0 (not at all) to 4 (extremely). Two items were included for number of days of alcohol use and nicotine using the PhenX Toolkit61. Finally, eight items from the Impulsive Behavior Scale – Short form (SUPPS-P)62 were included to measure impulsive behavior using using a scale from 1 (never) to 5 (very often) .
Principal component analyses for resilience factors
Principal component analyses (PCA) using the item-level clinical data were performed using SPSS v28.0. Clinical data was available for n=2,772 at 6-months. Participants with missing values for variables included in the analyses for this investigation were excluded,resulting in n=1,835 (1,175 women; Table 1) participants with complete 6-month item level clinical data. Varimax rotation with a maximum of 25 iterations for convergence was used. Factors were extracted based on eigenvalues greater than 2. The scores were saved as variables to create resilience factor scores for 6-months (static resilience factor scores). To create dynamic resilience factor scores we estimated the 2-week resilience scores based on the 6-month factors created with the PCA. Time of assessment (2-weeks versus 6-months) was used as a selection variable, such that the PCA was based on 6-month data and estimated factor scores were computed for the 2-week data using the same factors. The 2-week factor scores were subtracted from the 6-month/static factor scores to calculate dynamic resilience factor scores. Table 2 shows the varimax rotation loadings for each of the items for each factor. These participant-level static and dynamic resilience factor scores were used for the neuroimaging analyses.
Functional MRI
High resolution T1-weighted structural scans were collected using multi-echo magnetization prepared rapid acquisition gradient echo (ME-MPRAGE) using consistent parameters across the five imaging sites at a 1mm isotropic resolution (Supplementary Table S3). One imaging site collected a standard 1mm isotropic MPRAGE sequence, as it was unable to collect ME-MPRAGE scans. Functional MRI data was collected using identical parameters to measure blood-oxygen-level-dependent (BOLD) signals across sites, however, scan time slightly different between sites.
Functional MRI was collected during presentation of three functional tasks to assess inhibition, threat, and reward processes, described in more detail in a prior publication28. The three task paradigms were selected for the AURORA study because they have consistently been shown to activate neural substrates of interest and represent the NIH RDoC constructs. Inhibition was measured using the Go/NoGo task63. Participants were shown either an X or an O and were instructed to press 1 for X and 2 for O (Go trials), but not press a button when a right box appeared behind the X or the O (NoGo trial). The contrast Go>NoGo was used to measure response inhibition. Threat processing was assessed with the fearful faces task64. Participants viewed blocks of fearful and neutral faces from the Ekman and Friesen faces library, and the fearful>neutral faces contrast was used to measure threat processing. For reward processing, a short version of Delgado’s monetary reward task65 was used. Participants viewed a card with a question mark and were asked to indicate by a button press whether they guess the card’s value would be higher or lower than 5 before the real value was revealed. Participants were told they would win $1 for each correct guess and lose $0.50 for each incorrect guess and were told they would receive this money. The contrast gain>loss was used to measure reward processing. Unknown to the participants, the outcome was predetermined to create 10 wins and 10 losses resulting in $10.
Preprocessing information is reported in prior studies28. Briefly, functional images were preprocessed with fMRIPrep, version 1.2.266. Echo-planar imaging scans were coregistered to the T1-weighted images, Scans were spatially realigned, slice-time corrected, and normalized to the 2009 ICBM-152 template. ICA-AROMA was used to correct for volume-wise motion and other sources of artifact67. An overall motion threshold was implemented for any run with >15% of volumes with more than 1mm framewise displacement in order to handle cases in which motion was likely too high for effective ICA correction. Finally, mages were then smoothed with a 6-mm kernel.
ROI analyses
Contrast estimates were extracted for our predefined ROIs, also described in our earlier papers28,49. For inhibition (NoGo>Go contrast), the bilateral hippocampus, anatomically-defined based on the Hammers atlas, and the ventromedial prefrontal cortex, functionally defined based on our prior study findings (6mm sphere around –4, 44, -4; 63), were used. For threat processing (fearful>neutral faces contrast), the anatomically defined amygdala based on the CITI168 subcortical atlas, the bilateral hippocampus, anatomically defined based on the Hammers atlas, and the BA25 and BA32, anatomically defined based on the map of the BA from the WFUPickAtlas were used. For reward processing (Gain>Loss), the nucleus accumbens and bilateral amygdala, both anatomically defined based on the CITI168 subcortical atlas, and OFC, anatomically defined based on the Harvard/Oxford atlas.
To test our hypotheses that greater ROI contrast values were related to greater static and dynamic resilience scores, partial correlation analyses between each ROI and static and dynamic resilience scores were performed while correcting for site using dummy variables. Outliers with contrast values that deviated more than 3SD from the mean were assessed and excluded separately per task (inhibition, n=2; threat, n=5; reward, n=2). Additional Bonferroni correction for multiple comparisons (9 ROIs; 6-month and dynamic resilience; 18 tests, p<0.0028) was applied and correlations that survived this strict correction were indicated in the results section by an asterisk (*).
Whole brain analyses
Whole brain group level maps were created for inhibition, threat, and reward processes and included dummy variables for site. Separate models were created for each factor and static and dynamic resilience. A primary threshold of p<0.005 combined with a Family Wise Error (FWE) cluster level correction was applied to correct for multiple comparisons using the default in SPM12.
Unbiased whole brain correlation analyses with the continuous resilience factor scores were conducted to test the general hypothesis that inhibition-, threat-, and reward-related reactivity are related to resilience. Clusters surviving the FWE-corrected threshold were reported in Table 4 and means were extracted and displayed in Figures 4 and 5. Additional Bonferroni correction was applied to correct for the three task constructs and static and dynamic resilience (18 tests; p<0.0028; indicated by an asterisk (*) in Table 4).
Disclosures
- Dr. Neylan has received research support from NIH, VA, and Rainwater Charitable Foundation, and consulting income from Jazz Pharamaceuticals.
- In the last three years, Dr. Clifford has received research funding from the NSF, NIH, and Lifebell AI, and unrestricted donations from AliveCor Inc, Amazon Research, the Center for Discovery, the Gates Foundation, Google, the Gordon and Betty Moore Foundation, MAthworks, Microsoft Research, Nextsense Inc, One Mind Foundation, the Rett Research Foundation, and Samsun Research. Dr. Clifford has financial interest in AliveCor Inc and Nextsense Inc. He also is the CTO of MindChild Medical and CSO of LifeBell AI and has ownership in oth companies. These relationships are unconnected to the current work.
- Dr. Germine receives funding from the National Institute of Mental Health (R01 MH121617) and am on the board of the Many Brains Project. My family also has equity in Intererad Medical Systems, Inc.
- Dr. Rauch reports grants from NIH during the conduct of the study; personal fees from SOB (Society of Biological Psychiatry) paid roles as secretary, other from Oxford University Press reoyalties, other from APP (American Psychiatric Publishing Inc.) royalties, others from VA (Vetereans Administration) per diem for oversigh committee, and other from Community Psychiatry/Mindpath Health paid boad service, including equity outside the submitted work; other from National Association of Behavioral Healthcare for paid Board service; other from Springer Publshing royalties; and Leadership roles on Board or Council for SOBP, ADAA (Anxiety and Depression Association of America), and NNDC (National Network of Depression Centers).
- Dr. Sheikh has received funding from the Florida Medical Malpractice Joint Underwriter’s Association Dr. Alvin E. Smith Safety of Healthcare Services Grant; Allergan Foundation; the NIH/NIA-funded Jacksonville Aging Studies Center (JAX-ASCENT; R33AG05654); and the Substance Abuse and Mental health Services Administration (1H79TI083101-01); and the Florida Blue Foundation.
- Dr. Jones has no competing interest related to this work, thogh he has been an investigator on studies funded by AstraZeneca, Vapotherm, Abbott, and Ophinrex.
- Dr. Joormann receives consulting payments from Janssen Pharmaceuticals.
- Over the past three years, Dr. Pizzagalli has received consulting fees from Albright Stonebridge Group, Boehringer Ingelheim, Compass Pathways, Concert Pharmaceuticals, Engrail Therapeutics, Neumora Therapeutics (former BlackThorn Therapeutics), Neurocrine Biosciences, Neuroscience Software, Otsuka Pharmaceuticals, and Takeda Pharmaceuticals; honoraria from the Psychonomic Society (for editorial work) and Alkermes, and research funding from NIMH, Dana Foundation, Brain and Behavior Research Foundation, and Millennium Pharmaceuticals. In addition, he has received stock option from Neumora Therapeutics (former BlackThorn Therapeutics), Compass Pathways, Engrail Therapeutics, and Neuroscience Software.
- Dr. Harte has no competing interest related to this work, though in the last three years he has received research funding from Aptinyx and Arbor Medical Innovations, and consulting payments from Aptinyx, heron Therapeutics, and Eli Lilly.
- In the past three years, Dr. Kessler was a consultant for Cambridge Health Alliance, Canandaigua VA Medical Center, Holmusk, Partners Healthcare, Inc., RallyPoint Networks, Inc., and Sage Therapeutics. He has stock options in Cerebral Inc., Mirah, PYM, and Roga Sciences.
- Dr. Koenen’s research has been supported by the Robert Wood Johnson Foundation, the Kaiser Family Foundation, the Harvard Center on the Developing Child, Stanley Center for Psychiatric Research at the Broad Insititute of MIT and Harvard, the National Institutes of Health, One Mind, The Anonymous Foundation, and Cohen Veterans Bioscience. She has been a paid consultant for Baker Hostetler, Discovery Vitality, and the Department of Justice. She has been a paid external reviewer for the Chan Zuckerberg Foundation, the University of Cape Town, and Capita Ireland. She has had paid speaking engagemnts in the last three years with the American Psychological Association, European Central Bank. Sigmund Freud University- Milan, Cambridhe Health Alliance, and Coverys. She receives royalties from Guilford Press and Oxford University Press.
- Dr. McLean served as a consultant for Walter Reed Army Institute for Research and for Arbor Medical Innovations.
- Dr. Ressler has performed scientific consultation for Bioxcel, Bionomics, Acer, and Jazz Pharam; serves on Scientific Advisory Boards for Sage, Boehringer Ingelheim, Senseye, and the Brain Research Foundation, and has received sponsored research support from Alto Neuroscience.
Acknowledgements
The investigators wish to thank the trauma survivors participating in the AURORA Study. Their time and effort during a challenging period of their lives make our efforts to improve recovery for future trauma survivors possible. This project was supported by NIMH under U01MH110925, K01MH121653, the US Army MRMC, One Mind, and The Mayday Fund. The content is solely responsibility of the authors and does not necessarily represent the official views of any of the funders. Verily Life Sciences and Mindstrong Health provided some of the hardware and software used to perform study assessments. The Many Brains Project provided software for neurocognitive assessments. Data and/or research tools used in the preparation of this manuscript were obtained from the National Institute of Mental Health (NIMH) Data Archive (NDA). NDA is a collaborative informatics system created by the National Institutes of Health to provide a national resource to support and accelerate research in mental health. Dataset identifier(s): NIMH Data Archive Digital Object Identifier (DOI) 10.15154/zwyn-rb26. This manuscript reflects the views of the authors and may not reflect the opinions or views of the NIH or of the Submitters submitting original data to NDA.
- Resilience Definition & Meaning - Merriam-Webster. https://www.merriam-webster.com/dictionary/resilience.
- Kalisch, R. et al. Deconstructing and Reconstructing Resilience: A Dynamic Network Approach. Perspect. Psychol. Sci. J. Assoc. Psychol. Sci. 14, 765–777 (2019).
- Roeckner, A. R., Oliver, K. I., Lebois, L. A. M., van Rooij, S. J. H. & Stevens, J. S. Neural contributors to trauma resilience: a review of longitudinal neuroimaging studies. Transl. Psychiatry 11, 508 (2021).
- Kaldewaij, R. et al. Anterior prefrontal brain activity during emotion control predicts resilience to post-traumatic stress symptoms. Nat. Hum. Behav. 5, 1055–1064 (2021).
- Richter, A., Krämer, B., Diekhof, E. K. & Gruber, O. Resilience to adversity is associated with increased activity and connectivity in the VTA and hippocampus. NeuroImage Clin. 23, 101920 (2019).
- van Rooij, S. J. H. et al. Childhood Trauma and COMT Genotype Interact to Increase Hippocampal Activation in Resilient Individuals. Front. Psychiatry 7, 156 (2016).
- van Rooij, S. J. H. et al. Hippocampal activation during contextual fear inhibition related to resilience in the early aftermath of trauma. Behav. Brain Res. 408, 113282 (2021).
- Kalisch, R. et al. The resilience framework as a strategy to combat stress-related disorders. Nat. Hum. Behav. 1, 784–790 (2017).
- Admon, R. et al. Human vulnerability to stress depends on amygdala’s predisposition and hippocampal plasticity. Proc. Natl. Acad. Sci. U. S. A. 106, 14120–14125 (2009).
- Gilam, G., Lin, T., Fruchter, E. & Hendler, T. Neural indicators of interpersonal anger as cause and consequence of combat training stress symptoms. Psychol. Med. 47, 1561–1572 (2017).
- van Wingen, G. A., Geuze, E., Vermetten, E. & Fernández, G. Perceived threat predicts the neural sequelae of combat stress. Mol. Psychiatry 16, 664–671 (2011).
- Admon, R. et al. Imbalanced neural responsivity to risk and reward indicates stress vulnerability in humans. Cereb. Cortex N. Y. N 1991 23, 28–35 (2013).
- Vythilingam, M. et al. Reward circuitry in resilience to severe trauma: an fMRI investigation of resilient special forces soldiers. Psychiatry Res. 172, 75–77 (2009).
- Ben-Zion, Z. et al. Neural Responsivity to Reward Versus Punishment Shortly After Trauma Predicts Long-Term Development of Posttraumatic Stress Symptoms. Biol. Psychiatry Cogn. Neurosci. Neuroimaging 7, 150–161 (2022).
- Wallbrown, F. H., Blaha, J., Wallbrown, J. D. & Engin, A. W. The Hierarchical Factor Structure of the Wechsler Intelligence Scale for Children-Revised. J. Psychol. 89, 223–235 (1975).
- Caspi, A. et al. The p Factor: One General Psychopathology Factor in the Structure of Psychiatric Disorders? Clin. Psychol. Sci. J. Assoc. Psychol. Sci. 2, 119–137 (2014).
- Caspi, A. & Moffitt, T. E. All for One and One for All: Mental Disorders in One Dimension. Am. J. Psychiatry 175, 831–844 (2018).
- Romer, A. L. et al. Structural alterations within cerebellar circuitry are associated with general liability for common mental disorders. Mol. Psychiatry 23, 1084–1090 (2018).
- Romer, A. L. et al. Pervasively Thinner Neocortex as a Transdiagnostic Feature of General Psychopathology. Am. J. Psychiatry 178, 174–182 (2021).
- Vanes, L. D. & Dolan, R. J. Transdiagnostic neuroimaging markers of psychiatric risk: A narrative review. NeuroImage Clin. 30, 102634 (2021).
- Jenness, J. L. et al. Alterations in neural circuits underlying emotion regulation following child maltreatment: a mechanism underlying trauma-related psychopathology. Psychol. Med. 51, 1880–1889 (2021).
- Shanmugan, S. et al. Common and Dissociable Mechanisms of Executive System Dysfunction Across Psychiatric Disorders in Youth. Am. J. Psychiatry 173, 517–526 (2016).
- Galatzer-Levy, I. R. et al. Early PTSD Symptom Trajectories: Persistence, Recovery, and Response to Treatment: Results from the Jerusalem Trauma Outreach and Prevention Study (J-TOPS). PLOS ONE 8, e70084 (2013).
- McLean, S. A. et al. The AURORA Study: a longitudinal, multimodal library of brain biology and function after traumatic stress exposure. Mol. Psychiatry 25, 283–296 (2020).
- van Rooij, S. J. H. et al. The Role of the Hippocampus in Predicting Future Posttraumatic Stress Disorder Symptoms in Recently Traumatized Civilians. Biol. Psychiatry 84, 106–115 (2018).
- Stevens, J. S. et al. Amygdala Reactivity and Anterior Cingulate Habituation Predict Posttraumatic Stress Disorder Symptom Maintenance After Acute Civilian Trauma. Biol. Psychiatry 81, 1023–1029 (2017).
- Lokshina, Y., Nickelsen, T. & Liberzon, I. Reward Processing and Circuit Dysregulation in Posttraumatic Stress Disorder. Front. Psychiatry 12, 559401 (2021).
- Stevens, J. S. et al. Brain-Based Biotypes of Psychiatric Vulnerability in the Acute Aftermath of Trauma. Am. J. Psychiatry 178, 1037–1049 (2021).
- O’Doherty, J., Kringelbach, M. L., Rolls, E. T., Hornak, J. & Andrews, C. Abstract reward and punishment representations in the human orbitofrontal cortex. Nat. Neurosci. 4, 95–102 (2001).
- Wallis, J. D. & Kennerley, S. W. Contrasting reward signals in the orbitofrontal cortex and anterior cingulate cortex. Ann. N. Y. Acad. Sci. 1239, 33–42 (2011).
- Menon, V. & Uddin, L. Q. Saliency, switching, attention and control: a network model of insula function. Brain Struct. Funct. 214, 655–667 (2010).
- Hu, S., Ide, J. S., Zhang, S. & Li, C.-S. R. The Right Superior Frontal Gyrus and Individual Variation in Proactive Control of Impulsive Response. J. Neurosci. Off. J. Soc. Neurosci. 36, 12688–12696 (2016).
- du Boisgueheneuc, F. et al. Functions of the left superior frontal gyrus in humans: a lesion study. Brain J. Neurol. 129, 3315–3328 (2006).
- Brockett, A. T., Tennyson, S. S., deBettencourt, C. A., Gaye, F. & Roesch, M. R. Anterior cingulate cortex is necessary for adaptation of action plans. Proc. Natl. Acad. Sci. U. S. A. 117, 6196–6204 (2020).
- Doya, K. Modulators of decision making. Nat. Neurosci. 11, 410–416 (2008).
- Behrens, T. E. J., Hunt, L. T., Woolrich, M. W. & Rushworth, M. F. S. Associative learning of social value. Nature 456, 245–249 (2008).
- Goodkind, M. et al. Identification of a Common Neurobiological Substrate for Mental Illness. JAMA Psychiatry 72, 305–315 (2015).
- McTeague, L. M. et al. Identification of Common Neural Circuit Disruptions in Emotional Processing Across Psychiatric Disorders. Am. J. Psychiatry 177, 411–421 (2020).
- Mahayana, I. T., Tcheang, L., Chen, C.-Y., Juan, C.-H. & Muggleton, N. G. The precuneus and visuospatial attention in near and far space: a transcranial magnetic stimulation study. Brain Stimulat. 7, 673–679 (2014).
- Lee, J. et al. rTMS over bilateral inferior parietal cortex induces decrement of spatial sustained attention. Front. Hum. Neurosci. 7, 26 (2013).
- Hommet, C. et al. Unilateral spatial neglect following right inferior parietal cortectomy. Epilepsy Behav. EB 5, 416–419 (2004).
- Smallwood, J. et al. The default mode network in cognition: a topographical perspective. Nat. Rev. Neurosci. 22, 503–513 (2021).
- Zhang, W., Xie, Y. & Yang, T. Reward salience but not spatial attention dominates the value representation in the orbitofrontal cortex. Nat. Commun. 13, 6306 (2022).
- Anderson, B. A., Laurent, P. A. & Yantis, S. Value-driven attentional capture. Proc. Natl. Acad. Sci. 108, 10367–10371 (2011).
- David, O. A. & Magurean, S. Positive Attention Bias Trained during the Rethink Therapeutic Online Game and Related Improvements in Children and Adolescents’ Mental Health. Children 9, 1600 (2022).
- Kim, Y. J. et al. Association between posttraumatic stress disorder severity and amygdala habituation to fearful stimuli. Depress. Anxiety 36, 647–658 (2019).
- Wicking, M. et al. Deficient fear extinction memory in posttraumatic stress disorder. Neurobiol. Learn. Mem. 136, 116–126 (2016).
- Stevens, J. S. et al. Episodic memory after trauma exposure: Medial temporal lobe function is positively related to re-experiencing and inversely related to negative affect symptoms. NeuroImage Clin. 17, 650–658 (2018).
- Tanriverdi, B. et al. Hippocampal Threat Reactivity Interacts with Physiological Arousal to Predict PTSD Symptoms. J. Neurosci. Off. J. Soc. Neurosci. 42, 6593–6604 (2022).
- Garavan, H., Ross, T. J. & Stein, E. A. Right hemispheric dominance of inhibitory control: an event-related functional MRI study. Proc. Natl. Acad. Sci. U. S. A. 96, 8301–8306 (1999).
- van Rooij, S. J. H., Geuze, E., Kennis, M., Rademaker, A. R. & Vink, M. Neural correlates of inhibition and contextual cue processing related to treatment response in PTSD. Neuropsychopharmacol. Off. Publ. Am. Coll. Neuropsychopharmacol. 40, 667–675 (2015).
- Luijten, M., Meerkerk, G.-J., Franken, I. H. A., van de Wetering, B. J. M. & Schoenmakers, T. M. An fMRI study of cognitive control in problem gamers. Psychiatry Res. 231, 262–268 (2015).
- Solanto, M. V., Schulz, K. P., Fan, J., Tang, C. Y. & Newcorn, J. H. Event-related FMRI of inhibitory control in the predominantly inattentive and combined subtypes of ADHD. J. Neuroimaging Off. J. Am. Soc. Neuroimaging 19, 205–212 (2009).
- Powers, A. et al. Right inferior frontal gyrus and ventromedial prefrontal activation during response inhibition is implicated in the development of PTSD symptoms. Eur. J. Psychotraumatology 13, 2059993 (2022).
- van Rooij, S. J. H. et al. Impaired right inferior frontal gyrus response to contextual cues in male veterans with PTSD during response inhibition. J. Psychiatry Neurosci. JPN 39, 330–338 (2014).
- Grieder, M. et al. Right Inferior Frontal Activation During Alcohol-Specific Inhibition Increases With Craving and Predicts Drinking Outcome in Alcohol Use Disorder. Front. Psychiatry 13, 909992 (2022).
- Kleerekooper, I. et al. The effect of aging on fronto-striatal reactive and proactive inhibitory control. NeuroImage 132, 51–58 (2016).
- Pilkonis, P. A. et al. Item banks for measuring emotional distress from the Patient-Reported Outcomes Measurement Information System (PROMIS®): depression, anxiety, and anger. Assessment 18, 263–283 (2011).
- PROMIS Scoring Manuals. https://www.healthmeasures.net/index.php?option=com_content&view=article&id=180&Itemid=994.
- Blevins, C. A., Weathers, F. W., Davis, M. T., Witte, T. K. & Domino, J. L. The Posttraumatic Stress Disorder Checklist for DSM-5 (PCL-5): Development and Initial Psychometric Evaluation. J. Trauma. Stress 28, 489–498 (2015).
- Hamilton, C. M. et al. The PhenX Toolkit: get the most from your measures. Am. J. Epidemiol. 174, 253–260 (2011).
- Cyders, M. A., Littlefield, A. K., Coffey, S. & Karyadi, K. A. Examination of a short English version of the UPPS-P Impulsive Behavior Scale. Addict. Behav. 39, 1372–1376 (2014).
- Jovanovic, T. et al. Reduced neural activation during an inhibition task is associated with impaired fear inhibition in a traumatized civilian sample. Cortex J. Devoted Study Nerv. Syst. Behav. 49, 1884–1891 (2013).
- Stevens, J. S. et al. Disrupted amygdala-prefrontal functional connectivity in civilian women with posttraumatic stress disorder. J. Psychiatr. Res. 47, 1469–1478 (2013).
- Delgado, M. R., Nystrom, L. E., Fissell, C., Noll, D. C. & Fiez, J. A. Tracking the hemodynamic responses to reward and punishment in the striatum. J. Neurophysiol. 84, 3072–3077 (2000).
- Esteban, O. et al. fMRIPrep: a robust preprocessing pipeline for functional MRI. Nat. Methods 16, 111–116 (2019).
- Pruim, R. H. R., Mennes, M., Buitelaar, J. K. & Beckmann, C. F. Evaluation of ICA-AROMA and alternative strategies for motion artifact removal in resting state fMRI. NeuroImage 112, 278–287 (2015).
Table 1. Demographic data
|
|
Total sample |
Imaging sample |
|
N=1835 |
N=260 |
|
|
|
M (SD) or N (%) |
M (SD) or N (%) |
|
Age |
37.4 (13.6) |
35.4 (13.0) |
|
Sex at birth |
||
|
Male |
660 (36%) |
93 (36%) |
|
Female |
1175 (64%) |
167 (64%) |
|
Race |
||
|
Hispanic White |
185 (10%) |
34 (13%) |
|
Non-hispanic White |
662 (36%) |
93 (36%) |
|
Non-hispanic Black |
916 (50%) |
118 (45%) |
|
Non-hispanic other |
65 (4%) |
13 (5%) |
|
Marital History |
||
|
Current/Previous Marriage |
760 (42%) |
85 (33%) |
|
Never Married |
1063 (58%) |
174 (67%) |
|
Missing |
12 (1%) |
1 (0.4%) |
|
Education |
||
|
High School or less |
209 (11%) |
19 (7%) |
|
Some college or more |
1620 (88%) |
241 (93%) |
|
Missing |
6 (0.3%) |
0 |
|
Currently Employed |
||
|
No |
476 (26%) |
73 (28%) |
|
Yes |
1255 (68%) |
173 (67%) |
|
Missing |
104 (6%) |
41 (5%) |
|
Income |
||
|
<=$35K Yearly |
1089 (59%) |
143 (55%) |
|
>$35K Yearly |
635 (35%) |
101 (39%) |
|
Missing |
111 (6%) |
16 (6%) |
|
Trauma type |
|
|
|
Motor vehicle accident |
1349 (73.5%) |
182 (70.0%) |
|
Physical assault |
164 (8.9%) |
27 (10.4%) |
|
Sexual assault |
9 (0.5%) |
1 (0.4%) |
|
Fall >= 10ft |
32 (1.7%) |
6 (2.3%) |
|
Other (including poisoning, burns, animal-related, fall<10ft) |
281 (15.3%)
|
44 (16.9%)
|
Table 2. Varimax rotation component loadings for the r-factor, reminder acceptance and behavioral control factors (n=1,835)
Table 3. Correlations static factors with age, sex, childhood trauma and trait resilience (N=1,835)
|
|
Sex at birth |
Age |
CTQ total
|
CTQ abuse |
CTQ neglect
|
CD-RISC (6mo) |
|
r-factor |
n.s. |
Older > Younger |
r=-0.23** |
r=-0.22** |
r=-0.15** |
r=0.19** |
|
r=0.08** |
|
|
|
|
||
|
Reminder acceptance |
Men>Women |
Older < Younger |
r=-0.12** |
r=-0.14** |
n.s. |
n.s |
|
t=6.03** |
r=-0.13** |
|
|
|
|
|
|
Behavioral control |
Men < Women |
Older > Younger |
r=-0.22** |
r=-0.19** |
r=-0.18** |
r=0.10** |
|
t=-3.92** |
r=0.13** |
|
|
|
|
CTQ, childhood trauma questionnaire; CD-RISC, Connor-Davidson Resilience Score
**p<0.001; n.s, not significant (p>0.05)
Table 4. Correlations between resilience factors and whole brain responses during inhibition (n=215), threat (n=249), and reward processing (n=214)
|
FACTOR |
COMPONENT |
DIRECTION |
CONSTRUCT |
Statistics# |
MNI coordinates peak |
||||||
|
|
|
|
Inhibition |
Threat |
Reward |
p-value |
cluster size |
z-score |
x |
y |
z |
|
r-factor |
static |
positive |
- |
- |
right superior temporal gyrus (STG) |
0.040 |
157 |
4.22 |
44 |
-18 |
-2 |
|
|
|
|
|
|
right insula |
0.007 |
151 |
4.06 |
38 |
26 |
6 |
|
|
|
negative |
- |
- |
bilateral precuneus |
<0.001* |
376 |
5.09 |
12 |
-68 |
50 |
|
|
|
|
|
|
left inferior parietal lobe (IPL) |
0.001* |
307 |
4.73 |
-38 |
-56 |
50 |
|
dynamic |
positive |
- |
- |
right superior frontal gyrus (SFG) |
<0.001* |
896 |
4.99 |
42 |
-18 |
0 |
|
|
|
|
|
|
|
left superior medial gyrus (SMG) |
<0.001* |
353 |
4.89 |
-6 |
64 |
8 |
|
|
|
|
|
|
left SFG |
<0.001* |
327 |
4.31 |
-46 |
-2 |
-2 |
|
|
|
|
|
|
bilateral dACC |
0.041 |
159 |
3.91 |
2 |
14 |
34 |
|
|
|
negative |
- |
left IPL |
- |
<0.001* |
479 |
5.31 |
-34 |
-72 |
50 |
|
|
|
|
|
right IPL |
|
0.001* |
341 |
4.44 |
44 |
-66 |
46 |
|
Reminder acceptance |
static/dynamic |
positive |
- |
- |
- |
|
|
|
|
|
|
|
|
|
negative |
- |
- |
- |
|
|
|
|
|
|
|
Behavioral control |
static |
positive |
right inferior frontal gyrus (rIFG) |
- |
- |
0.031 |
179 |
4.28 |
40 |
12 |
36 |
|
|
|
negative |
- |
- |
left pre/postcentral gyrus |
0.003 |
246 |
3.87 |
-68 |
-14 |
22 |
|
|
dynamic |
positive |
- |
right postcentral gyrus |
|
0.001* |
357 |
4.37 |
56 |
-26 |
46 |
|
|
|
|
|
left precentral gyrus |
|
0.003 |
293 |
4.57 |
-56 |
14 |
30 |
|
|
|
|
|
left IPL |
|
0.005 |
268 |
3.76 |
-56 |
-22 |
40 |
|
|
|
|
|
|
right IPL |
<0.001* |
471 |
4.08 |
44 |
-46 |
44 |
|
|
|
|
|
|
right insula |
0.010 |
205 |
4.27 |
38 |
24 |
-2 |
|
|
|
negative |
- |
- |
- |
|
|
|
|
|
|
|
# whole brain level correlation analyses were conducted using a primary threshold p<0.005 and cluster-level FWE-corrected threshold. The cluster-level FWE-corrected p-values and corresponding cluster size are reported. The z-score for the peak voxel and corresponding Montreal Neurological Institute (MNI) coordinates are reported. Starred (*) p-values and bolded regions indicate clusters that survived additional Bonferroni correction for three tasks and three constructs for both static and dynamic resilience (18 tests, p<0.0028). |
|||||||||||
Yes there is potential Competing Interest.