Neuroimaging-based classification of PTSD using data-driven computational approaches: A multisite big data study from the ENIGMA-PGC PTSD consortium.

BACKGROUND: Recent advances in data-driven computational approaches have been helpful in devising tools to objectively diagnose psychiatric disorders. However, current machine learning studies limited to small homogeneous samples, different methodologies, and different imaging collection protocols, limit the ability to directly compare and generalize their results. Here we aimed to classify individuals with PTSD versus controls and assess the generalizability using a large heterogeneous brain datasets from the ENIGMA-PGC PTSD Working group. METHODS: We analyzed brain MRI data from 3,477 structural-MRI; 2,495 resting state-fMRI; and 1,952 diffusion-MRI. First, we identified the brain features that best distinguish individuals with PTSD from controls using traditional machine learning methods. Second, we assessed the utility of the denoising variational autoencoder (DVAE) and evaluated its classification performance. Third, we assessed the generalizability and reproducibility of both models using leave-one-site-out cross-validation procedure for each modality. RESULTS: We found lower performance in classifying PTSD vs. controls with data from over 20 sites (60 % test AUC for s-MRI, 59 % for rs-fMRI and 56 % for d-MRI), as compared to other studies run on single-site data. The performance increased when classifying PTSD from HC without trauma history in each modality (75 % AUC). The classification performance remained intact when applying the DVAE framework, which reduced the number of features. Finally, we found that the DVAE framework achieved better generalization to unseen datasets compared with the traditional machine learning frameworks, albeit performance was slightly above chance. CONCLUSION: These results have the potential to provide a baseline classification performance for PTSD when using large scale neuroimaging datasets. Our findings show that the control group used can heavily affect classification performance. The DVAE framework provided better generalizability for the multi-site data. This may be more significant in clinical practice since the neuroimaging-based diagnostic DVAE classification models are much less site-specific, rendering them more generalizable.

A comparison of methods to harmonize cortical thickness measurements across scanners and sites.

Results of neuroimaging datasets aggregated from multiple sites may be biased by site-specific profiles in participants' demographic and clinical characteristics, as well as MRI acquisition protocols and scanning platforms. We compared the impact of four different harmonization methods on results obtained from analyses of cortical thickness data: (1) linear mixed-effects model (LME) that models site-specific random intercepts (LMEINT), (2) LME that models both site-specific random intercepts and age-related random slopes (LMEINT+SLP), (3) ComBat, and (4) ComBat with a generalized additive model (ComBat-GAM). Our test case for comparing harmonization methods was cortical thickness data aggregated from 29 sites, which included 1,340 cases with posttraumatic stress disorder (PTSD) (6.2-81.8 years old) and 2,057 trauma-exposed controls without PTSD (6.3-85.2 years old). We found that, compared to the other data harmonization methods, data processed with ComBat-GAM was more sensitive to the detection of significant case-control differences (Χ2(3) = 63.704, p < 0.001) as well as case-control differences in age-related cortical thinning (Χ2(3) = 12.082, p = 0.007). Both ComBat and ComBat-GAM outperformed LME methods in detecting sex differences (Χ2(3) = 9.114, p = 0.028) in regional cortical thickness. ComBat-GAM also led to stronger estimates of age-related declines in cortical thickness (corrected p-values < 0.001), stronger estimates of case-related cortical thickness reduction (corrected p-values < 0.001), weaker estimates of age-related declines in cortical thickness in cases than controls (corrected p-values < 0.001), stronger estimates of cortical thickness reduction in females than males (corrected p-values < 0.001), and stronger estimates of cortical thickness reduction in females relative to males in cases than controls (corrected p-values < 0.001). Our results support the use of ComBat-GAM to minimize confounds and increase statistical power when harmonizing data with non-linear effects, and the use of either ComBat or ComBat-GAM for harmonizing data with linear effects.

Altered white matter microstructural organization in posttraumatic stress disorder across 3047 adults: results from the PGC-ENIGMA PTSD consortium.

A growing number of studies have examined alterations in white matter organization in people with posttraumatic stress disorder (PTSD) using diffusion MRI (dMRI), but the results have been mixed which may be partially due to relatively small sample sizes among studies. Altered structural connectivity may be both a neurobiological vulnerability for, and a result of, PTSD. In an effort to find reliable effects, we present a multi-cohort analysis of dMRI metrics across 3047 individuals from 28 cohorts currently participating in the PGC-ENIGMA PTSD working group (a joint partnership between the Psychiatric Genomics Consortium and the Enhancing NeuroImaging Genetics through Meta-Analysis consortium). Comparing regional white matter metrics across the full brain in 1426 individuals with PTSD and 1621 controls (2174 males/873 females) between ages 18-83, 92% of whom were trauma-exposed, we report associations between PTSD and disrupted white matter organization measured by lower fractional anisotropy (FA) in the tapetum region of the corpus callosum (Cohen's d = -0.11, p = 0.0055). The tapetum connects the left and right hippocampus, for which structure and function have been consistently implicated in PTSD. Results were consistent even after accounting for the effects of multiple potentially confounding variables: childhood trauma exposure, comorbid depression, history of traumatic brain injury, current alcohol abuse or dependence, and current use of psychotropic medications. Our results show that PTSD may be associated with alterations in the broader hippocampal network.

Cortical volume abnormalities in posttraumatic stress disorder: an ENIGMA-psychiatric genomics consortium PTSD workgroup mega-analysis.

Studies of posttraumatic stress disorder (PTSD) report volume abnormalities in multiple regions of the cerebral cortex. However, findings for many regions, particularly regions outside commonly studied emotion-related prefrontal, insular, and limbic regions, are inconsistent and tentative. Also, few studies address the possibility that PTSD abnormalities may be confounded by comorbid depression. A mega-analysis investigating all cortical regions in a large sample of PTSD and control subjects can potentially provide new insight into these issues. Given this perspective, our group aggregated regional volumes data of 68 cortical regions across both hemispheres from 1379 PTSD patients to 2192 controls without PTSD after data were processed by 32 international laboratories using ENIGMA standardized procedures. We examined whether regional cortical volumes were different in PTSD vs. controls, were associated with posttraumatic stress symptom (PTSS) severity, or were affected by comorbid depression. Volumes of left and right lateral orbitofrontal gyri (LOFG), left superior temporal gyrus, and right insular, lingual and superior parietal gyri were significantly smaller, on average, in PTSD patients than controls (standardized coefficients = -0.111 to -0.068, FDR corrected P values < 0.039) and were significantly negatively correlated with PTSS severity. After adjusting for depression symptoms, the PTSD findings in left and right LOFG remained significant. These findings indicate that cortical volumes in PTSD patients are smaller in prefrontal regulatory regions, as well as in broader emotion and sensory processing cortical regions.

Developing trauma resilient communities through community capacity-building.

BACKGROUND: Trauma is a significant public health issue, negatively impacting a range of health outcomes. Providers and administrators in public mental health systems recognize the widespread experience of trauma, as well as their limited ability to address trauma within their communities. In response, the Los Angeles County Department of Mental Health funded nine regionally based community partnerships to build capacity to address trauma. We describe partnership and community capacity-building efforts and examine community impact, defined as successful linkages to resources and changes in stress tolerance capacities among community members. METHODS: We conceptualized community capacity-building as dissemination of trauma-informed education and training, community outreach and engagement, and linkage of community members to resources. We measured trauma-informed trainings among partnership members (N = 332) using the Trauma-Informed Organizational Toolkit. Outreach, engagement and linkages were documented using Event and Linkage Trackers. We examined changes in the type of successful linkage after the issuance of statewide mandatory restrictions in response to COVID-19. We examined changes in stress tolerance capacities among community members (N = 699) who were engaged in ongoing partnership activities using the 10-item Conner-Davidson Resilience Scale; the 28-item Coping Orientation to Problems; and the pictorial Inclusion of Community in Self Scale. RESULTS: Training and education opportunities were widespread: 66% of members reported opportunities for training in 13 or more trauma-informed practices. Partnerships conducted over 7800 community capacity-building events with over 250,000 attendees. Nearly 14,000 successful linkages were made for a wide range of resources, with consistent linkage success prior to (85%) and during (87%) the pandemic. In response to COVID-19, linkage type significantly shifted from basic services and health care to food distribution (p < .01). Small but significant improvements occurred in coping through emotional and instrumental support; and sense of community connectedness (p < .05 each). CONCLUSIONS: Community-based partnerships demonstrated effective capacity-building strategies. Despite the pandemic, community members did not report reduced stress tolerance, instead demonstrating gains in external help-seeking (use of emotional and instrumental supports) and perception of community connectedness. Future work will use qualitative methods to examine the impact of community capacity-building and the sustainability of this approach for addressing the impact of trauma within communities.

Health Care Usage and Suicide Risk Screening within 1 Year of Suicide Death.

BACKGROUND: Research indicates patients often seek medical care within 1 year of suicide. Health care encounters are a crucial opportunity for health professionals to identify patients at highest risk and provide preventative services. OBJECTIVE: Study aims were to determine the characteristics of persons seeking health care within 12 months of suicide death and evaluate suicide risk screening (SRS) frequency in the emergency department (ED) vs. clinic settings. METHODS: Medical examiner and hospital data of patients who died by suicide from 2007 to 2013 were evaluated. Descriptive analyses included demographics and frequency of ED vs. clinic visits. We also compared SRS before and after implementation of The Joint Commission's recommendation to assess suicide risk. RESULTS: The 224 deceased patients were primarily single white males (mean age 67 years). Mental health issues, substance abuse, and prior suicide attempts were present alone or in combination in 74%. Visits were primarily behavioral health or substance abuse problems in the ED, and medical issues in the clinic. After implementation of universal SRS in the ED, screening increased from 39% to 92%. Among patients screened in the ED, 73% (37 of 51) screened negative for suicide risk. CONCLUSIONS: Universal SRS increased the number of people screened in the ED. However, negative SRS may not equate to reduced risk for future suicide within 1 year. Future studies might investigate targeted screening of individuals with known suicide risk factors, as well as alternatives to patient self-report of intent to self-harm for patients with mental health or substance abuse problems.

Gender differences in adapting driving behavior to accommodate visual health limitations.

This study investigated whether men and women are equally likely to adapt their driving behaviors in response to visual limitations. Participants were 376 (222 women and 154 men) pre-surgical cataract patients from the Shiley Eye Center in La Jolla, California. All participants completed the National Eye Institute Visual Functioning Questionnaire, which assesses self-reported visual symptoms, functional limitations, and behaviors including driving during the day, at night, or in difficult conditions. Visual acuity was assessed using the log of the minimal angle of resolution (LogMAR) scale. There were no significant differences in LogMAR visual acuity between men and women who reported either that they stopped driving at night because of visual impairment or reported having no difficulty driving at night. Of participants who reported having difficulty driving at night, mean weighted LogMAR scores indicated significantly better visual acuity for women than men. There were no significant differences in LogMAR visual acuity between women and men in any of the difficult driving condition categories. Significantly more women than men reported that they stopped driving in difficult conditions because of eyesight, despite the lack of gender differences in visual acuity for this sample. We found no evidence that cataract disease had different effects on the visual acuity of older adult men and women. However, there was a significant difference between genders in self-reported driving behavior. It is possible that some women are more cautious or have less need to drive. However, failing to adapt driving behaviors to accommodate visual limitations may represent a potential behavioral public health risk for men.

Remodeling of the Cortical Structural Connectome in Posttraumatic Stress Disorder: Results From the ENIGMA-PGC Posttraumatic Stress Disorder Consortium.

BACKGROUND: Posttraumatic stress disorder (PTSD) is accompanied by disrupted cortical neuroanatomy. We investigated alteration in covariance of structural networks associated with PTSD in regions that demonstrate the case-control differences in cortical thickness (CT) and surface area (SA). METHODS: Neuroimaging and clinical data were aggregated from 29 research sites in >1300 PTSD cases and >2000 trauma-exposed control subjects (ages 6.2-85.2 years) by the ENIGMA-PGC (Enhancing Neuro Imaging Genetics through Meta Analysis-Psychiatric Genomics Consortium) PTSD working group. Cortical regions in the network were rank ordered by the effect size of PTSD-related cortical differences in CT and SA. The top-n (n = 2-148) regions with the largest effect size for PTSD > non-PTSD formed hypertrophic networks, the largest effect size for PTSD < non-PTSD formed atrophic networks, and the smallest effect size of between-group differences formed stable networks. The mean structural covariance (SC) of a given n-region network was the average of all positive pairwise correlations and was compared with the mean SC of 5000 randomly generated n-region networks. RESULTS: Patients with PTSD, relative to non-PTSD control subjects, exhibited lower mean SC in CT-based and SA-based atrophic networks. Comorbid depression, sex, and age modulated covariance differences of PTSD-related structural networks. CONCLUSIONS: Covariance of structural networks based on CT and cortical SA are affected by PTSD and further modulated by comorbid depression, sex, and age. The SC networks that are perturbed in PTSD comport with converging evidence from resting-state functional connectivity networks and networks affected by inflammatory processes and stress hormones in PTSD.