Investigating the bed nucleus of the stria terminalis as a predictor of posttraumatic stress disorder in Black Americans and the moderating effects of racial discrimination

Altered functioning of the bed nucleus of the stria terminalis (BNST) may play a critical role in the etiology of posttraumatic stress disorder (PTSD). Chronic stressors such as racial

discrimination and lifetime trauma are associated with an increased risk for PTSD, but it is unknown whether they influence the relationship between BNST functioning and PTSD. We

investigated acute post-trauma BNST resting-state functional connectivity (rsFC) as a predictor of future PTSD symptoms in Black trauma survivors. We also examined whether racial

discrimination and lifetime trauma moderated the relationship between BNST rsFC and PTSD symptoms. Black adults (N = 95; 54.7% female; mean age = 34.04) were recruited from an emergency

department after experiencing a traumatic injury (72.6% were motor vehicle accidents). Two-weeks post-injury, participants underwent a resting-state fMRI scan and completed questionnaires

evaluating their PTSD symptoms as well as lifetime exposure to racial discrimination and trauma. Six-months post-injury, PTSD symptoms were reassessed. Whole brain seed-to-voxel analyses

were conducted to examine BNST rsFC patterns. Greater rsFC between the BNST and the posterior cingulate cortex, precuneus, left angular gyrus, and hippocampus prospectively predicted

six-month PTSD symptoms after adjusting for sex, age, education, and baseline PTSD symptoms. Acute BNST rsFC was a stronger predictor of PTSD symptoms in individuals who experienced more

racial discrimination and lifetime trauma. Thus, in the acute aftermath of a traumatic event, the BNST could be a key biomarker of risk for PTSD in Black Americans, particularly for

individuals with a greater history of racial discrimination or previous trauma exposure.

Posttraumatic stress disorder (PTSD) is a pervasive mental illness that impacts up to 14% of Black Americans in their lifetime [1]. PTSD is a heterogenous disorder that is broadly

characterized by recurring intrusive memories of a traumatic event, hyperarousal, avoidance of trauma-related stimuli, and negative alterations in cognition and mood [2]. Identifying early

markers of risk for PTSD is crucial for enhancing early intervention efforts that attempt to thwart PTSD development [3]. While psychosocial factors such as racial discrimination and other

forms of stress/trauma exposure (e.g., events that qualify as “criterion A” stressors within the Diagnostic Statistical Manual of Mental Disorders 5th edition [DSM-5] definition of PTSD2)

confer risk for PTSD [1, 4], recent research utilizing machine learning techniques indicate the strongest predictive models of PTSD include both neurobiological and psychosocial markers of

risk [5]. Thus, it is imperative to investigate the neural underpinnings of PTSD development in the context of psychosocial risk factors. Functional magnetic resonance imaging (fMRI) studies

have identified key markers of risk for PTSD within threat processing neurocircuitry [6], and growing research suggests altered functioning of the bed nucleus of the stria terminalis (BNST)

may play a critical role in the etiology of PTSD [7,8,9].

The BNST is a subcortical structure located in the medial basal forebrain, and it has dense structural and functional connections with the limbic system, basal ganglia, and hypothalamus

[10,11,12,13]. The BNST is also composed of a high volume of neuropeptide receptors that modulate the stress response [14, 15]. Functionally, the BNST is responsible for detecting

environmental cues that signal uncertain threat, as well as initiating and prolonging states of anxiety (i.e. tonic arousal) in response to the ambiguous threat [16, 17]. When there is

potential to encounter a threat, BNST activation is essential for early initiation of the fear response, which allows one to proactively cope with the threat. However, when BNST activity is

consistently heightened in contextually safe situations (i.e., no threat present), it can lead to persistent states of hyperarousal and hypervigilance [17,18,19], which may increase risk for

PTSD by interfering with processes that help individuals cope with posttraumatic stress (e.g., fear extinction recall) [20,21,22].

Accordingly, neuroimaging research has identified patterns of heightened BNST activity among individuals who have PTSD. Findings from task-based fMRI studies show individuals with PTSD

display greater BNST activity when they are awaiting a future threat (i.e., hypervigilant threat monitoring) [8] and when they are exposed to trauma-related stimuli compared to healthy

controls [7]. Notably, similar patterns of BNST activity are observed when individuals with PTSD are at rest (i.e., not engaged in a task). Relative to healthy controls, individuals with

PTSD exhibit greater BNST resting-state functional connectivity (rsFC) with regions of the brain that are implicated in arousal and vigilance, which is theorized to underlie chronic

anticipatory anxiety and hyperarousal symptoms of PTSD [9]. Considering a diminished ability to inhibit one's fear response in the absence of a threat is a hallmark trait of PTSD [21, 23],

BNST rsFC may be a particularly salient neural marker of risk for future PTSD development.

Black Americans continue to experience race-related stressors, including racial discrimination, across a variety of settings, which can elicit psychological symptoms and neurobiological

alterations that mirror PTSD [24,25,26,27,28,29,30,31]. For instance, greater exposure to racial discrimination is associated with heightened functional connectivity between threat

processing regions of the brain [28,29,30] as well as persistent states of arousal and vigilance [25, 26, 31]. Notably, repeated exposure to DSM-5 defined traumatic events (i.e., exposure to

actual or threatened death, serious injury, or sexual violence) can also elicit similar changes in threat neurocircuitry and threat monitoring behavior [32, 33]. Elevated BNST resting-state

activity may contribute to these posttraumatic alterations in neuropsychological functioning, as the BNST appears to be sensitive to the effects of chronic stressors[14, 34].

In general, repeated stressor exposure can upregulate and alter a variety of neurobiological functions related to threat monitoring and the fear response in humans [32]. Preclinical models

show chronic stress increases the expression of neurochemicals within the BNST that modulate the stress response and upregulate BNST activity [14, 34]. Increasing BNST activity is also

associated with an increase in anxiety-related behaviors. Thus, in the acute aftermath of a traumatic event, BNST connectivity may be an early marker of risk for PTSD, particularly among

individuals who have a greater history of stress and trauma exposure (e.g., racial discrimination, “criterion A” stressors).

Previous research conducted by Webb et al. (2022), Harb et al. (2023), and Bird et al. (2021) identified racial discrimination as a unique predictor of PTSD symptoms [4], tested potential

mediators of the relationship [35], and investigated the relationship between discrimination and activity in threat-processing regions [30] in the same sample as the present study.

Considering greater lifetime exposure to DSM-5 “criterion A” stressors and racial discrimination are associated with an increased risk for PTSD following subsequent trauma exposure [36], and

both experiences elicit similar neural and psychological alterations, it is possible racial discrimination and “criterion A” stressor exposure similarly impact BNST connectivity as a

biomarker of PTSD.

The present study builds upon previous findings from our group [4, 30, 35] by using a prospective longitudinal design to investigate two-week post-trauma (T1) BNST rsFC as a predictor of

future PTSD symptoms (six-months post-trauma; T2) among Black trauma survivors. We also examined whether lifetime exposure to racial discrimination and “criterion A” stressors moderated the

relationship between acute BNST rsFC and future PTSD symptoms. Considering previous research implicates greater BNST activity with PTSD symptomatology [7,8,9], we hypothesized heightened T1

BNST rsFC would predict T2 PTSD symptoms. Additionally, given the effects of chronic stress on BNST functioning and risk for PTSD [14, 15], we hypothesized racial discrimination and prior

trauma exposure would independently moderate the relationship between T1 BNST connectivity and T2 PTSD symptoms.

Two-hundred and fifteen adults (Black/African American, n = 124) were recruited from an emergency department (ED) in the Midwest. Because the present study investigated BNST rsFC as a

biomarker for PTSD among Black trauma survivors and the moderating effects of racial discrimination and exposure to “criterion A” stressors, only individuals who identified as Black or

African American were included in the analyses. Individuals were either screened directly in the ED or via telephone shortly after discharge. Individuals were eligible for study

participation if they met the following criteria: were seen in the ED due to a traumatic injury (as defined by the DSM-52), right-handed, spoke English, between the ages of 18–65, and were

able to schedule their first study visit within two weeks of the trauma. Individuals were excluded for contraindications of MRI scanning (e.g. metal objects in body, current or planned

pregnancy in the next 6 months), if they suffered a head injury more severe than a mild traumatic brain injury (mTBI) as measured by the Glasgow Coma Scale [37], or if the injury resulted in

a loss of consciousness. Additional exclusion criteria included injuries resulting from self-inflicted harm, individuals with severe vision or hearing impairments, a history of psychotic or

manic symptoms, antipsychotic medication prescription, or if they were on police hold.

Eligible participants provided written informed consent prior to the study and were financially compensated for each visit. The study involving human subjects was approved by the Medical

College of Wisconsin Institutional Review Board and was conducted in accordance with the Declaration of Helsinki.

The self-report PTSD Checklist for the DSM-5 (PCL-5) [38] was administered to assess PTSD symptoms at T1. The PCL-5 is a 20-item self-report questionnaire (current sample Cronbach α = 0.94)

that assesses the frequency and severity of PTSD symptoms one has experienced within the past month, rated on a 5-point Likert scale from 0 (not at all) to 4 (extremely). All questions

referred to the index trauma responsible for the ED visit, and scores ranged from 0–80, with higher scores indicating greater symptom severity.

At T2, PTSD symptoms were assessed by a team member trained to administer the Clinician Administered PTSD scale for the DSM-5 (CAPS-5) [39]. The CAPS-5 is a 20-item structured interview

(current sample Cronbach α = 0.92) that measures the frequency and severity of PTSD symptoms one has experienced within the past month. All questions referred to the same index trauma

related to the ED visit, and the interviewer scored each item on the questionnaire from 0 (never/no effort to avoid) to 4 (daily or almost daily/extreme efforts to avoid) based on the

information provided by the participant. Total scores ranged from 0 to 80, with higher scores indicating greater symptom severity. Both the PCL-5 and CAPS-5 are well-established, empirically

validated methods of PTSD assessment [40, 41].

The Perceived Ethnic Discrimination Questionnaire – Community Version (PEDQ-CV) [42] was administered to evaluate lifetime exposure to racial discrimination. The PEDQ-CV consists of 17 items

(current sample Cronbach α = .93) that assess participants’ prior experiences with racial discrimination across various settings. Participants respond on a scale from 1 (never) to 5 (very

often). The scores across all items are averaged to create a total score, with higher scores indicating greater discrimination.

The Life Events Checklist for the DSM-5 (LEC-5) [43] was administered to evaluate lifetime exposure to traumatic events, a known risk factor for PTSD following subsequent trauma exposure.

Participants rated their experience (i.e., 0 = does not apply, 1 = happened to them, 2 = witnessed the event, 3 = learned about the event) with 17 different traumatic events (current sample

Cronbach α = 0.83). The scoring method developed by Weis et al. (2022) [44] was implemented. Total scores were weighted according to one’s proximity to the traumatic event (e.g., happened to

them vs learned about the event). Total scores range from 0–102.

Images were collected on a Discovery MR750 3.0 Tesla scanner, using a GE 32-channel head-coil. High resolution T1-weighted images were acquired with the following parameters: field of view

(FOV), 240 mm; matrix, 256 × 224; slice thickness, 1 mm; 150 slices; repetition time (TR)/echo time (TE), 8.2/3.2 s, flip angle, 12°; voxel size, 1 × 0.938 × 0.938 mm. At T1, participants

underwent an eight-minute resting-state fMRI (rs-fMRI) scan where they were instructed to stare at a fixation cross on a black screen; 240 volumes were acquired using the following

parameters: FOV, 22.4 mm; matrix, 64 × 64; slice thickness, 3.5 mm; 41 sagittal slices; TR/TE, 2000/25 milliseconds; flip angle, 77°; voxel size, 3.5 × 3.5 × 3.5 mm.

Structural and resting-state images were preprocessed using the default pipeline in the CONN Toolbox 20, with SPM 12 and MatLab 2019b [45]. Of the 124 Black Americans in the sample, 111

completed a baseline resting-state scan (no scan, n = 13). The first 3 TRs were discarded to account for initial instability of MR environment. All remaining images were motion-corrected

using a 6-parameter linear transformation, normalized to Montreal Neurological Institute template (MNI 152), and then spatially smoothed using a 4-mm full-width-at-half-maximum kernel. BNST

seed activity was extracted before smoothing. During the first-level analyses, head motion parameters (along with their first-order derivatives), white matter signal, and cerebrospinal fluid

signal were regressed out. Volumes with framewise displacement > 0.3 mm were removed from analyses (“scrubbed). There was no relationship between the number of discarded volumes and PTSD

symptom severity, r(93) = 0.04, p = 0.713. If more than 20% of the resting-state volumes were scrubbed or the scan quality was deemed poor after visual inspection, the participant was

excluded from the analysis. Of the 111 Black Americans who underwent rs-fMRI, 4 were removed from the analysis (n = 2 removed after visual quality checks and n = 2 exceeding 20% of volumes

discarded) based on these criteria. An additional 12 participants were removed from the analyses because they either did not complete a baseline PCL-5 (n = 2), or the six-month CAPS-5 (n = 

10), leaving an n = 95 for the current analysis.

A seed-to-voxel whole brain analysis, correlating the mean BOLD signal from the BNST with all other voxels in the brain, was conducted in CONN [46]. An established BNST seed for 3T fMRI

images [47] was used, and each participant was visually inspected to ensure proper placement of the BNST seed. In CONN, a group-level analysis examined T1 BNST connectivity as a predictor of

T2 PTSD symptoms after adjusting for other known self-report predictors of PTSD, including sex, age, education, and T1 PTSD symptoms (i.e., PCL-5 total scores) [48]. The threshold for

statistical significance was set at two-tailed p