PTSD Symptom Networks Among Women with IPV

Featured Article

Journal of Counseling Psychology | 2025, Vol. 72, No. 1, p. 1-14

Article Title

PTSD Symptom Networks During Treatment Among Residents in Domestic Violence Shelters

Authors

Nuha Alshabani; Chobanian and Avedisian School of Medicine, Boston University; Department of Psychiatry, Boston Medical Center of Boston University

James K. Haws; Department of Psychiatry, University of Colorado Anschutz

Caron Zlotnick; Department of Psychiatry and Human Behavior, Warren Alpert Medical School of Brown University; Women and Infants Hospital, Providence, Rhode Island, United States; Department of Psychiatry and Mental Health, University of Cape Town

Dawn M. Johnson; Department of Psychology, The University of Akron

Abstract

Little is known about how the interrelationships among posttraumatic stress disorder (PTSD) symptoms change throughout the treatment. We examined changes in PTSD symptoms among women who experienced intimate partner violence and received one of two evidence-based interventions. We conducted a secondary analysis of a completed randomized, double-blind clinical trial, which demonstrated efficacy in reducing PTSD symptoms. We used cross-lagged panel network analysis to identify the influential PTSD symptoms among women who had completed either the Helping to Overcome PTSD through Empowerment or an adapted version of person-centered therapy. We examined if a symptom’s expected influence, a metric of interconnectedness, would predict overall PTSD symptom reductions at baseline, post-shelter, post-treatment, and 6- and 12-months post-treatment. Women who showed more significant decreases in feeling upset and avoidance also demonstrated greater decreases in their overall PTSD symptoms at post-shelter, post-treatment, and 6 and 12 months post-treatment. Findings indicate that changes in symptoms with high centrality result in larger PTSD network changes observed at both adjacent and future time points. Identifying and targeting symptoms with influential associations produces therapeutic cascades, resulting in symptom reductions.

Keywords

Posttraumatic stress disorder, network analysis, intimate partner violence, shelter, treatment

Summary of Research

“Globally, violence against women is a serious public health and human rights issue, with over a billion women estimated to have experienced intimate partner violence (IPV) in their lifetime. IPV refers to various forms of violence or abuse (e.g., sexual, emotional, physical, or terrorizing) that range in acuity perpetrated by an intimate partner (e.g., spouses, dating partners). Violence against women is associated with significant impacts on well-being, including physical and mental health, economic burden, and loss of personal and social resources… Compared with nonshelter residents, women with IPV (W-IPV) in shelters tend to present with more severe abuse histories (Bargai et al., 2007) and posttraumatic stress disorder (PTSD) rates… While several evidence-based interventions exist to treat IPV related PTSD (e.g., cognitive trauma therapy for battered women; Recovering from IPV through Strengths and Empowerment), Helping to Overcome PTSD through Empowerment (HOPE) and an adapted version of present-centered therapy (PCT+) are the only research-supported treatments for IPV-related PTSD for women with recent or ongoing abuse and residing in domestic violence shelters” (p. 2).

“Instead of viewing symptoms as reflections of an underlying latent disease, the network perspective conceptualizes psychopathology as an emergent property of a dynamical system, resulting from the causal interplay among symptoms. A network is a set of nodes that are connected through edges. A node can represent various psychological variables (i.e., symptoms, behaviors, stressors, beliefs). Edges represent the statistical associations among nodes; accordingly, nodes influence each other. If one node in a network becomes activated, this increases the probability that a connected node also becomes activated. As network activation spreads, nodes become more connected until the network becomes self-sustaining (i.e., a mental disorder). While all nodes in the network are connected, nodes with more or stronger connections with other nodes are critical for network activation and sustainment. The network perspective proposes that central nodes might be targets of intervention or early risk detection. If true, the network perspective may revolutionize clinical treatment by identifying and intervening on influential nodes that should have a downstream effect on other parts of the system” (p. 2-3).

“To our knowledge, no study has used a network analysis approach to examine the impact of these IPV-specific interventions on PTSD for W-IPV. The goal of this study was to compare PTSD symptom networks across five stages of IPV-specific treatment. Leveraging network analyses, this study aimed to identify influential symptoms of PTSD and examine whether changes in the influential symptoms are associated with cascading treatment effects (i.e., overall reductions in PTSD symptoms following treatment). We conducted a secondary analysis from a completed randomized, controlled, population-based trial (RCT), which demonstrated efficacy in reducing PTSD symptoms (Johnson et al., 2020). Specifically, these analyses aim to test if a node’s centrality would predict the overall network changes at adjacent and future time points. In other words, our goal was to identify the most central PTSD network throughout treatment and determine whether changes in influential PTSD symptoms result in larger network changes” (p. 3-4).

“This study examined PTSD symptom networks throughout five time points for W-IPV. Throughout treatment, a symptom’s expected influence—a metric of interconnectedness—was a significant predictor of PTSD symptom network change. Women who showed more significant decreases in highly central symptoms (i.e., feeling upset and avoidance) also demonstrated greater decreases in their overall PTSD symptoms. Although there is some doubt about the suitability of centrality indices for psychopathological networks, our findings align with previous work confirming the clinical utility of network metrics. Our study demonstrated that expected influence predicted PTSD symptom change while controlling for a symptom’s severity” (p. 9).

“Results from this study extend upon previous research, indicating that feeling upset and avoidance are influential nodes by identifying their longitudinal influence in PTSD networks. Specifically, our findings indicate that changes in B4: feeling upset when reminded of the event and C1: avoiding thinking about the event resulted in larger PTSD network changes observed at adjacent and future time points. In other words, women who experienced changes in these symptoms also experienced significant reductions in their overall PTSD at PS and at each subsequent time point, highlighting the longitudinal cascading influence of these symptoms. These findings indicate that changes in negative emotional states and avoidance symptoms at BL and through each time point strongly predict the overall PTSD network change. Our study provides evidence for a future line of research to explore the effectiveness of treatments and treatment components that target these influential symptoms” (p. 9).

Translating Research into Practice

“The present study tested the clinical utility of the network perspective for PTSD among W-IPV. Findings from the present study have important implications for informing interventions for W-IPV, including the unique relationship among PTSD symptoms for this population over time. The interventions in this study both targeted symptoms of PTSD through present-centered techniques that focused on empowerment. At BL, the influential nodes— avoidance and feeling upset when reminded of the event—may have responded to specific components of HOPE and/or PCT. For instance, the HOPE intervention featured a focus on safety, self-care, and empowerment, which is an effective component for alleviating symptoms of PTSD. Increased knowledge of how treatment produces individual PTSD symptom change may be particularly impactful for implementing and sequencing treatment with IPV due to their complex clinical presentations, severe PTSD symptomology, and ongoing stressors. Results from this study point toward influential symptoms that, if disrupted, should lead to a positive therapeutic cascade, which may be useful for tailoring interventions to fit the needs and characteristics of each individual” (p. 12).

Other Interesting Tidbits for Researchers and Clinicians

“Limitations of this study included the sample size. We conducted secondary data analysis, and at the start of data collection, the DSM-5 had not been released, and thus, DSM-IV criteria were captured. As the original trial was not designed to determine the specific timing of symptom change, the timing of these changes remains to be determined. Changes may be happening in a timeframe outside of our measurement windows. Thus, future work is needed to better characterize the timing of symptom change. In addition, a limitation of doing research in a shelter setting is that the length of shelter stay varies from person to person. Thus, our findings may not generalize to other, more controlled treatment settings. Importantly, in Johnson et al.’s (2020) study, the number of sessions completed in the shelter was not associated with PTSD symptoms. Because of the sample size, we did not include other constructs in the network, such as comorbid disorders or other mechanisms associated with well-being for W-IPV (e.g., empowerment). Nevertheless, the present study is one of the few studies to analyze change in PTSD networks over time and the only network study to include W-IPV who resided in shelters—a population who experiences severe IPV and has increased rates of PTSD. The study presents the strengths of testing the network perspective in a sample of W-IPV residing in a shelter, which has the potential to generalize to other community samples of W-IPV. Additionally, findings can help identify and refine interventions for this vulnerable population of W-IPV. Future research is needed to extend the network analysis to other treatments for W-IPV, to compare the effectiveness of these interventions on PTSD symptoms and network change. Such studies may further elucidate mechanisms of change and inform implementation research to address common barriers to treatment, including high attrition. Finally, future research can continue to assess the clinical utility of the network perspective” (p. 12).