Neuroticism/negative emotionality is associated with increased reactivity to uncertain threat in the bed nucleus of the stria terminalis, not the amygdala

Description: ### UT = temporally uncertain threat anticipation; CT = temporally certain threat anticipation; Baseline = implicit baseline = temporally certain safety anticipation; TmS = Threat minus Safety anticipation contrast ###. Shannon E. Grogans, Juyoen Hur, Matthew G. Barstead, Allegra S. Anderson, Samiha Islam, Hyung Cho Kim, Manuel Kuhn, Rachael M. Tillman, Andrew S. Fox, Jason F. Smith, Kathryn A. DeYoung, & Alexander J. Shackman ### Here we used fMRI to quantify individual differences in EAc reactivity to a well-established anxiety-provocation ("Maryland Threat Countdown") paradigm in 220 young adults selectively recruited from a pool of 6,594 individuals to span a broad spectrum of threat reactivity (i.e., trait neuroticism/negative emotionality, N/NE; trait anxiety). We performed parallel analyses using data from a subset of 213 subjects who also completed an emotional-faces fMRI paradigm. Variants of the emotional-faces paradigm are widely used as probes of amygdala function—often in the guise of quantifying variation in Research Domain Criteria (RDoC) ‘Negative Valence Systems’—and have been incorporated into many prominent biobank studies (e.g. ABCD, UK Biobank). Although photographs of models posing ‘threat-related’ (fearful/angry) facial expressions robustly recruit the amygdala47, they do not elicit distress in typical adults and, as such, are better conceptualized as a probe of threat perception, rather than the experience or expression of negative emotion. Overview As part of an on-going prospective-longitudinal study focused on individuals at risk for the development of internalizing disorders, we used a well-established psychometric measure of N/NE to screen 6,594 first-year university students (57.1% female; 59.0% White, 19.0% Asian, 9.9% African American, 6.3% Hispanic, 5.8% Multiracial/Other; M=19.2 years, SD=1.1 years) 3. Screening data were stratified into quartiles (top quartile, middle quartiles, bottom quartile), separately for males and females. Individuals who met preliminary inclusion criteria were independently and randomly recruited via email from each of the resulting six strata. Because of our focus on psychiatric risk, approximately half the participants were recruited from the top quartile, with the remainder split between the middle and bottom quartiles (i.e., 50% high, 25% medium, and 25% low). This enabled us to sample a broad spectrum of N/NE without gaps or discontinuities, while balancing the inclusion of men and women (Fig. 1 in main report). Simulation work suggests that this over-sampling (‘enrichment’) approach does not bias statistical tests to a degree that would compromise their validity 4. All subjects had normal or corrected-to-normal color vision, and reported the absence of lifetime neurological symptoms, pervasive developmental disorder, very premature birth, medical conditions that would contraindicate MRI, and prior experience with noxious electrical stimulation. All subjects were free from a lifetime history of psychotic and bipolar disorders; a current diagnosis of a mood, anxiety, or trauma disorder (past 2 months); severe substance abuse; active suicidality; and on-going psychiatric treatment as determined by an experienced, masters-level diagnostician using the Structured Clinical Interview for DSM-5 5. Subjects provided informed written consent and all procedures were approved by the Institutional Review Board at the University of Maryland, College Park. Data from this study were featured in prior work focused on the development and validation of the threat-anticipation paradigm, relations between social anxiety and momentary mood, an relations between threat-related brain activity and momentary mood dynamics 8, but have never been used to address the present aims. Participants A total of 241 subjects were recruited and scanned. Of these, 6 withdrew due to excess distress in the scanner, 1 withdrew from the study after the imaging session, and 4 were excluded due to incidental neurological findings. Threat-anticipation task. One subject was excluded from fMRI analyses due to gross susceptibility artifacts in the echoplanar imaging (EPI) data, 2 were excluded due to insufficient usable data (<2 usable scans; see below), 6 were excluded due to excess motion artifact (i.e., the variance of the volume-to-volume displacement of a selected voxel at the center of the brain was >2 SDs above the mean), and 1 was excluded due to task timing issues, yielding a racially diverse final sample of 220 subjects (49.5% female; 61.4% White, 18.2% Asian, 8.6% African American, 4.1% Hispanic, 7.3% Multiracial/Other; M = 18.8 years, SD = 0.4 years). Of these, 2 individuals were excluded from skin conductance analyses due to insufficient usable data (<2 usable ‘scans’; see below). Emotional faces task. Three subjects were excluded due to gross susceptibility artifacts in the EPI data, 1 was excluded due to insufficient usable data (<2 scans), 7 were excluded due to excessive motion artifact, and 6 subjects for inadequate behavioral performance (i.e., accuracy <2 SD for both scans), yielding a final sample of 213 subjects (49.3% female; 61.0% White, 17.8% Asian, 8.5% African American, 4.2% Hispanic, 7.0% Multiracial/Other; M = 18.8 years, SD = 0.3 years). Threat-Anticipation Paradigm Paradigm Structure and Design Considerations. The Maryland Threat Countdown paradigm is a well-established, fMRI-optimized variant of temporally uncertain-threat assays that have been validated using fear-potentiated startle and acute anxiolytic administration (e.g., benzodiazepine) in mice, rats, and humans. The paradigm has been successfully used in several prior fMRI studies. The MTC paradigm takes the form of a 2 (Valence: Threat/Safety) × 2 (Temporal Certainty: Uncertain/Certain) randomized event-related design (3 scans; 6 trials/condition/scan). Simulations were used to optimize the detection and deconvolution of task-related hemodynamic signals (variance inflation factors <1.54). Stimulus presentation and ratings acquisition were controlled using Presentation software (version 19.0, Neurobehavioral Systems, Berkeley, CA). On Certain Threat trials, subjects saw a descending stream of integers (‘count-down;’ e.g., 30, 29, 28...3, 2, 1) for 18.75 s. To ensure robust distress, this anticipatory epoch culminated with the delivery of a noxious electric shock, unpleasant photographic image (e.g., mutilated body), and thematically related audio clip (e.g., scream, gunshot). Uncertain Threat trials were similar, but the integer stream was randomized and presented for an uncertain and variable duration (8.75-30.00 s; M=18.75 s). Subjects knew that something aversive was going to occur, but had no way of knowing precisely when. Consistent with recent recommendations 20, the average duration of the anticipation epoch was identical across conditions, ensuring an equal number of measurements (TRs/condition). The specific mean duration was chosen to enhance detection of task-related differences in the blood oxygen level-dependent (BOLD) signal (‘activation’) and to allow sufficient time for genuinely sustained responses to become evident. Safety trials were similar, but terminated with the delivery of benign reinforcers (see below). Valence was continuously signaled during the anticipatory epoch by the background color of the display. Temporal certainty was signaled by the nature of the integer stream. Certain trials always began with the presentation of the number 30. On Uncertain trials, integers were randomly drawn from a near-uniform distribution ranging from 1 to 45 to reinforce the impression that they could be much shorter or longer than Certain trials and to minimize incidental temporal learning (‘time-keeping’). To concretely demonstrate the variable duration of Uncertain trials, during scanning, the first three Uncertain trials featured short (8.75 s), medium (15.00 s), and long (28.75 s) anticipation epochs. To mitigate potential confusion and eliminate mnemonic demands, a lower-case ‘c’ or ‘u’ was presented at the lower edge of the display throughout the anticipatory epoch. White-noise visual masks (3.2 s) were presented between trials to minimize the persistence of visual reinforcers in iconic memory. Subjects provided ratings of anticipatory fear/anxiety for each condition during each scan using an MRI-compatible response pad (MRA, Washington, PA). Subjects were instructed to rate the intensity of the fear/anxiety experienced during the prior anticipation (‘countdown’) epoch using a 1 (minimal) to 4 (maximal) scale. Subjects were prompted to rate each trial type once per scan. Premature ratings (<300 ms) were censored. All subjects provided at least 6 usable ratings and rated each condition at least once. A total of 6 additional echo-planar imaging (EPI) volumes were acquired at the beginning and end of each scan. Procedures. Prior to scanning, subjects practiced an abbreviated version of the paradigm—without electrical stimulation—until they indicated and staff confirmed that they understood the task. Benign and aversive electrical stimulation levels were individually titrated. Benign Stimulation. Subjects were asked whether they could “reliably detect” a 20 V stimulus and whether it was “at all unpleasant.” If the subject could not detect the stimulus, the voltage was increased by 4 V and the process repeated. If the subject indicated that the stimulus was unpleasant, the voltage was reduced by 4V and the process was repeated. The final level chosen served as the benign electrical stimulation during the imaging assessment (M=21.06 V, SD=4.98 V). Aversive Stimulation. Subjects received a 100 V stimulus and were asked whether it was “as unpleasant as you are willing to tolerate.” If the subject indicated that they were willing to tolerate more intense stimulation, the voltage was increased by 10 V and the process repeated. If the subject indicated that the stimulus was too intense, the voltage was reduced by 5 V and the process repeated. The final level chosen served as the aversive electrical stimulation during the imaging assessment (M=118.02 V, SD=26.09). Following each scan, staff re-assessed whether stimulation was sufficiently intense and increased the level as necessary. Electrical Stimuli. Electrical stimuli (100 ms; 2 ms pulses every 10 ms) were generated using an MRI-compatible constant-voltage stimulator system (STMEPM-MRI; Biopac Systems, Inc., Goleta, CA). Stimuli were delivered using MRI-compatible, disposable carbon electrodes (Biopac) attached to the fourth and fifth digits of the non-dominant hand. Visual Stimuli. Visual stimuli (1.8 s) were digitally back-projected (Powerlite Pro G5550, Epson America, Inc., Long Beach, CA) onto a semi-opaque screen mounted at the head-end of the scanner bore and viewed using a mirror mounted on the head-coil. A total of 72 aversive and benign photographs were selected from the International Affective Picture System. Auditory Stimuli. Auditory stimuli (0.8 s) were delivered using an amplifier (PA-1 Whirlwind) with in-line noise-reducing filters and ear buds (S14; Sensimetrics, Gloucester, MA) fitted with noise-reducing ear plugs (Hearing Components, Inc., St. Paul, MN). A total of 72 aversive and benign auditory stimuli were adapted from open-access online sources. Emotional Faces Paradigm The emotional-faces paradigm took the form of a pseudo-randomized block design and was administered in 2 scans, with a short break between scans. During each scan, subjects viewed standardized photographs of adult models (half female) depicting prototypical angry faces, fearful faces, happy faces, or places (i.e., emotionally neutral natural scenes; 7 blocks/condition/scan). To maximize signal strength and homogeneity and mitigate potential habituation 21,23,24, blocks consisted of 10 briefly presented photographs of faces or places (1.6 s) separated by fixation crosses (0.4 s). To further minimize potential habituation, each photograph was only presented once or twice. To ensure engagement, subjects judged whether the current photograph matched that presented on the prior trial (i.e., a ‘1-back’ continuous performance task). Matches occurred 37.1% of the time. MRI Data Acquisition MRI data were acquired using a Siemens Magnetom TIM Trio 3 Tesla scanner (32-channel head-coil). Foam inserts were used to mitigate potential motion artifact. Subjects were continuously monitored using an MRI-compatible eye-tracker (Eyelink 1000; SR Research, Ottawa, Ontario, Canada). Head motion was monitored using the AFNI real-time plugin 32. Sagittal T1-weighted anatomical images were acquired using a magnetization prepared rapid acquisition gradient echo sequence (TR=2,400 ms; TE=2.01 ms; inversion time=1,060 ms; flip=8°; slice thickness=0.8 mm; in-plane=0.8 × 0.8 mm; matrix=300 × 320; field-of-view=240 × 256). A T2-weighted image was collected co-planar to the T1-weighted image (TR=3,200 ms; TE=564 ms; flip angle=120°). To enhance resolution, a multi-band sequence was used to collect oblique-axial EPI volumes (multiband acceleration=6; TR=1,250 ms; TE=39.4 ms; flip=36.4°; slice thickness=2.2 mm, number of slices=60; in-plane resolution=2.1875 × 2.1875 mm; matrix=96 × 96). Images were collected in the oblique-axial plane (approximately −20° relative to the AC-PC plane) to minimize potential susceptibility artifacts. For the threat-anticipation task, three 478-volume EPI scans were acquired. For the emotional-faces task, two 454-volume EPI scans were acquired. The scanner automatically discarded 7 volumes prior to the first recorded volume. To enable fieldmap correction, two oblique-axial spin echo (SE) images were collected in opposing phase-encoding directions (rostral-to-caudal and caudal-to-rostral) at the same location and resolution as the functional volumes (i.e., co-planar; TR=7,220 ms; TE=73 ms). Measures of respiration and breathing were continuously acquired during scanning using a respiration belt and photo-plethysmograph affixed to the first digit of the non-dominant hand. Following the last scan, subjects were removed from the scanner, debriefed, compensated, and discharged. MRI Data Processing Pipeline Methods were optimized to minimize spatial normalization error and other potential sources of noise. Data were visually inspected before and after processing for quality assurance. Anatomical Data Processing. Methods are similar to those described in other recent reports by our group 6,8. T1-weighted images were inhomogeneity corrected using N4 36 and denoised using ANTS 37. The brain was then extracted using BEaST 38 and brain-extracted and normalized reference brains from IXI 39. Brain-extracted T1 images were normalized to a version of the brain-extracted 1-mm T1-weighted MNI152 (version 6) template 40 modified to remove extracerebral tissue. Normalization was performed using the diffeomorphic approach implemented in SyN (version 2.3.4) 37. T2-weighted images were rigidly co-registered with the corresponding T1 prior to normalization. The brain extraction mask from the T1 was then applied. Tissue priors were unwarped to native space using the inverse of the diffeomorphic transformation 41. Brain-extracted T1 and T2 images were segmented—using native-space priors generated in FAST (version 6.0.4) 42—for subsequent use in T1-EPI co-registration (see below). Fieldmap Data Processing. SE images and topup were used to create fieldmaps. Fieldmaps were converted to radians, median-filtered, and smoothed (2-mm). The average of the distortion-corrected SE images was inhomogeneity corrected using N4 and masked to remove extracerebral voxels using 3dSkullStrip (version 19.1.00). Functional Data Processing. EPI files were de-spiked using 3dDespike, slice-time corrected to the TR center using 3dTshift, and motion corrected to the first volume and inhomogeneity corrected using ANTS (12-parameter affine). Transformations were saved in ITK-compatible format for subsequent use 43. The first volume was extracted for EPI-T1 coregistration. The reference EPI volume was simultaneously co-registered with the corresponding T1-weighted image in native space and corrected for geometric distortions using boundary-based registration 42. This step incorporated the previously created fieldmap, undistorted SE, T1, white matter (WM) image, and masks. The spatial transformations necessary to transform each EPI volume from native space to the reference EPI, from the reference EPI to the T1, and from the T1 to the template were concatenated and applied to the processed EPI data in a single step to minimize incidental spatial blurring. Normalized EPI data were resampled (2 mm3) using fifth-order b-splines. Hypothesis testing focused on anatomically defined EAc regions of interest (ROIs), as detailed below. To maximize anatomical resolution, no additional spatial filters were applied, consistent with recent recommendations 44. Whole-brain exploratory analyses employed data that were spatially smoothed (6-mm) using 3DblurInMask. fMRI Data Exclusions and Modeling Data Exclusions. Volume-to-volume displacement (>0.5 mm) was used to assess residual motion artifact. Scans with excessively frequent artifacts (>2 SD) were discarded. Subjects with who lacked sufficient usable fMRI data (<2 scans of the threat-anticipation task or <1 scan of the emotional-faces task) or showed poor performance on the emotional-faces task (see above; accuracy <2 SD) were excluded from the relevant brain-behavior analyses. Canonical First-Level (Single-Subject) fMRI Modeling. For each subject, first-level modeling was performed using GLMs implemented in SPM12 (version 7771), with the default autoregressive model and the temporal band-pass filter set to the hemodynamic response function (HRF) and 128 s 45. Regressors were convolved with a canonical HRF and its temporal derivative. EPI volumes acquired before the first trial, during intertrial intervals, and following the final trial were unmodeled and contributed to the baseline estimate. Threat-Anticipation Task. Hemodynamic reactivity to the task was modeled using variable-duration rectangular (‘boxcar’) regressors time-locked to the anticipation (‘countdown’) epochs of the Uncertain Threat, Certain Threat, and Uncertain Safety trials. To maximize design efficiency, Certain Safety anticipation—which is psychologically similar to a conventional inter-trial interval—served as the reference condition and contributed to the baseline estimate. The periods corresponding to the presentation of the four types of reinforcers (i.e., Aversive/Benign × Certain/Uncertain), visual masks, and rating prompts were simultaneously modeled using the same approach. Consistent with prior work using the Maryland Threat Countdown paradigm 6,8, nuisance variates included estimates of volume-to-volume displacement, motion (6 parameters × 3 lags), cerebrospinal fluid (CSF) signal, instantaneous pulse and respiration rates, and ICA-derived nuisance signals (e.g. brain edge, CSF edge, global motion, white matter) 46. Volumes with excessive volume-to-volume displacement (>0.5 mm) and those during and immediately following reinforcer delivery were censored. Emotional-Faces Task. Hemodynamic reactivity to blocks of each emotional expression (angry, fearful, and happy) was modeled using rectangular regressors. Place blocks served as the reference condition and contributed to the baseline estimate.