Examining the Associations Between PTSD Symptoms and Aspects of Emotion Dysregulation Through Network Analysis

Abstract

Despite the clearly established link between posttraumatic stress disorder (PTSD) and emotion dysregulation, little is known about how individual symptoms of PTSD and aspects of emotion dysregulation interrelate. The network approach to mental health disorders provides a novel framework for conceptualizing the association between PTSD and emotion dysregulation as a system of interacting nodes. In this study, we estimated the structural relations among PTSD symptoms and aspects of emotion dysregulation within a large sample of women who participated in a multi-site study of sexual revictimization (N = 463). We estimated expected influence to reveal differential associations among PTSD symptoms and aspects of emotion dysregulation. Further, we estimated bridge expected influence to identify influential nodes connecting PTSD symptoms and aspects of emotion dysregulation. Results highlighted the key role of concentration difficulties in expected influence and bridge expected influence. Findings highlight several PTSD symptoms and aspects of emotion dysregulation that may be targets for future intervention.

Graphical Abstract

1. Examining the Associations Between PTSD Symptoms and Aspects of Emotion Dysregulation Through Network Analysis

Lifetime trauma exposure is common among women and is associated with negative outcomes (Reeves et al., 2017). Among women who have experienced a traumatic event, posttraumatic stress disorder (PTSD) and emotion dysregulation commonly co-occur (e.g., Badour et al., 2012; Tull et al., 2007). Despite the robust association between PTSD and emotion dysregulation, the question of how the two are related and the appropriate conceptual models for understanding their linkage remains unclear. The present study applies a network analytic approach to advance the conceptual understanding of the co-occurrence between PTSD and emotion dysregulation. Specifically, we use network analysis methodology to estimate the differential associations among symptoms of PTSD and aspects of emotion dysregulation in a sample of trauma-exposed women.

1.2. Emotion Dysregulation and Posttraumatic Stress Disorder

Emotion regulation is a motivated process involving the activation of emotion goals (Tamir et al., 2019) and the selection and implementation of emotion regulation strategies (Gross, 2015) to address those goals. Beyond simply the selection and implementation of emotion regulation strategies, adaptive emotion regulation can be conceptualized more broadly as a multi-dimensional phenomenon involving the abilities to identify, understand, and accept emotions, engage in goal-directed behavior and inhibit impulsive behavior when experiencing negative emotions, and use effective (non-harmful) strategies to modulate emotional responses in the pursuit of desired goals (e.g., Gratz & Roemer, 2004; Gross, 2015; Tamir et al., 2019). Deficits in any one of these aspects reflect the presence of emotion dysregulation (Gratz & Roemer, 2004; Millgram et al., 2020).

Emotion dysregulation is a transdiagnostic mechanism implicated as a maintenance factor in the severity and course of several psychiatric disorders, including PTSD (Aldao et al., 2010; Badour et al., 2012; Kring, 2008; Tull et al., 2007). Individuals meeting diagnostic criteria for PTSD experience difficulty identifying and labeling their emotions (Vine & Aldao, 2014) and recognizing that different regulatory goals require different strategies (Ehring & Quack, 2010). For example, in a study with firefighters, Levy-Gigi and colleagues (2016) found that regulatory choice flexibility—the ability to choose between different strategies depending on contextual demands—moderated the association between traumatic event exposure and PTSD symptoms, such that traumatic exposure was more strongly associated with PTSD symptoms among individuals with poor regulatory choice flexibility.

Evidence also suggests that aspects of emotion dysregulation are related to the development, severity, and maintenance of PTSD (e.g., Kring, 2008; Powers et al., 2015). First, emotion dysregulation, prior to trauma, is associated with a greater chance of developing PTSD and more severe PTSD symptoms after later trauma exposure (Bardeen et al., 2013; Forbes et al., 2020; Seligowski et al., 2015). Further, adaptive emotion regulation in the aftermath of a traumatic event can mitigate against the onset of PTSD symptoms and facilitate recovery (e.g., Bryan et al., 2015; Feldner et al., 2007; Nitzan-Assayag et al., 2015). It is possible that individuals with existing emotion dysregulation may be more likely to use maladaptive emotion regulation strategies in response to the increased distress associated with trauma exposure or when faced with the intense and frequent negative affect that accompanies PTSD symptoms (e.g., Lee et al., 2015; Tull et al., 2007). Indeed, greater use of maladaptive strategies, such as avoidance, has been linked to greater PTSD severity among trauma survivors (Badour et al., 2012; Kumpula et al., 2011; Rauch & Foa, 2006).

In addition to the theorized relation of emotion dysregulation to PTSD severity, specific PTSD symptoms may be linked to increased emotion dysregulation. PTSD is characterized by the inability to effectively down-regulate emotional arousal (Rauch & Foa, 2006; Resick et al., 2012). Trauma survivors often re-experience the traumatic event through flashbacks or nightmares and report heightened emotional and physiological responses to reminders of the traumatic event (Bardeen et al., 2013; Tull et al., 2020), placing increased demands on effective emotion regulation. Although these heightened emotional and physiological responses diminish with time for many people, others learn to fear and avoid situations where negative emotions may become activated, preventing functional exposure to trauma-relevant cues that could lead to natural recovery (Foa & Kozak, 1986; Rauch & Foa, 2006; Tull et al., 2007). Thus, PTSD symptoms may interfere with trauma survivors’ perceived ability to regulate negative emotions.

1.4. The Network Perspective on Psychopathology

Although prior research and theory suggests that PTSD symptoms and emotion dysregulation are strongly related, limited work has examined how these associations may be occurring on a more specific level. The network approach to psychopathology offers a unique opportunity to understand the interplay between specific PTSD symptoms and aspects of emotion dysregulation. The network approach to psychopathology postulates that comorbidity or co-occurrence is generated through a system of relations among interacting nodes (Borsboom & Cramer, 2013; Fried et al., 2017). In psychological networks, nodes can represent various psychological variables (i.e., symptoms, behaviors, features of psychological processes), whereas edges represent the statistical relations (i.e., partial correlations, predictive relations) that can be estimated from the data. According to this approach, nodes from each psychological construct drive nodes in the other (Borsboom, 2017). Thus, the network approach suggests that specific PTSD symptoms and aspects of emotion dysregulation may interact to produce their co-occurrence. According to the network approach, interactions among psychological constructs may be grounded in biological, societal, traumatic, or psychological processes (Borsboom, 2017; McNally, 2016). Regardless, if one node in a network becomes activated, this increases the probability a connected node also becomes activated. This conceptualization has been applied to model a range of psychiatric disorders, including depression and anxiety disorders (Beard et al., 2016), eating disorders (Forbush, Siew, & Vitevitch, 2016), obsessive-compulsive disorder (Jones et al., 2018), personality disorders (Southward & Cheavens, 2018), and PTSD (e.g., Fried et al., 2018; McNally et al., 2015), as well as comorbidity between disorders (Heeren et al., 2018; Langer et al., 2018).

An ever-growing number of studies have identified influential nodes that are central to sustaining PTSD. Notably, a recent systematic review identified intrusive thoughts, negative emotional states, and strong physical sensations as central symptoms within the broader construct of PTSD (Birkeland et al., 2020). Supporting these findings, a study examining trauma-exposed children and adolescents found that negative trauma-related cognitions, psychological distress, and persistent negative emotional states were the nodes with high strength centrality, suggesting that these symptoms might generalize to other developmental periods. In addition to these findings from cross-sectional studies, results from a recent longitudinal network analysis examining within-person changes in PTSD symptoms revealed that negative cognitions and negative emotions had the highest strength centrality, meaning they were the strongest predictors of other nodes at future time points (Reeves & Fisher, 2020). In summary, findings suggest that trauma-related cognitions and emotions are central to the development and maintenance of PTSD following traumatic exposure (e.g., Bartels et al., 2019; Birkeland et al., 2020; Reeves & Fisher, 2020; Weems, 2020).

Although findings from prior network studies have been largely consistent, there are several limitations of the extant literature that interfere with generalizability. First, few studies have examined the network structure of PTSD within community samples of women exposed to traumatic events, despite the high prevalence of lifetime trauma exposure and PTSD symptoms among women (Reeves et al., 2017). This gap precludes our ability to determine whether the key symptoms that have been identified in past studies generalize to other, predominantly female samples of trauma survivors. Second, although most prior work has examined PTSD in isolation, PTSD commonly co-occurs with other forms of psychopathology and dysregulation (Seligowski et al., 2015), which could influence how PTSD networks are expressed. Initial studies have begun to examine PTSD and frequently co-occurring disorders, such as depression (e.g., Lazarov et al., 2019) and eating disorders (Vanzhula et al., 2019), in order to highlight symptoms that may underlie comorbidity.

Despite the known co-occurrence between PTSD and emotion dysregulation, to date there has only been one study examining the node-to-node connections between PTSD and emotion dysregulation using network analytic methodology (Weiss et al., 2020). Weiss and colleagues (2020) implemented network analyses to investigate the associations between broader PTSD symptoms clusters (i.e., intrusions, avoidance, negative alterations in cognition and mood, and alterations in arousal and reactivity) and aspects of positive emotion dysregulation (i.e., difficulties regulating positive emotions) in a trauma-exposed community sample of men and women. Their analysis revealed that the alterations in arousal and reactivity (AAR) PTSD symptom cluster had the highest bridge association with positive aspects of emotion dysregulation, suggesting that arousal may be associated with positive emotion dysregulation (Weiss et al., 2018). In addition to identifying AAR as an influential node, nonacceptance of positive emotions and difficulties engaging in goal-directed behavior in the context of positive emotions had the highest bridge association with PTSD symptom clusters. However, this study did not assess the specific symptoms of PTSD or aspects of emotion dysregulation that were responsible for the observed associations or account for other aspects of emotion dysregulation that may be particularly relevant to PTSD (e.g., difficulties modulating negative affect). Moreover, given evidence that different symptoms of PTSD operate differently and evidence distinct relations to other relevant constructs (e.g., Beevers et al., 2019; Boschloo et al., 2019; Fried et al., 2014), different PTSD symptoms may be differentially associated with specific aspects of emotion dysregulation.

Indeed, given the strong association between PTSD and emotion dysregulation, identifying specific difficulties that link the two constructs may lead to the development of targeted interventions focused on minimizing the potential mutually reinforcing relations between the two phenomena. Network models permit the identification of “bridge nodes” that transcend the psychological constructs of interest and connect theoretically independent constructs, such as those that link PTSD and emotion dysregulation (Jones et al., 2019). Identifying bridge nodes has implications for understanding why certain constructs co-occur and how this co-occurrence is maintained (i.e., by identifying channels through which constructs may be mutually activated). Because bridge nodes are crucial to maintaining the co-occurrence of psychological difficulties, they are especially appealing targets of intervention (Cramer et al., 2010). Likewise, identifying potential directionality among nodes facilitates the identification of nodes that may have a strong influence on other nodes in the network. Identifying and targeting such nodes is thought to be especially valuable for intervention efforts designed to interrupt cascades to other nodes in the network (McNally, 2016).

1.5. The Current Study

The current study addresses key gaps in the literature by identifying the differential associations among symptoms of PTSD and aspects of emotion dysregulation in a large sample of women with traumatic event exposure. First, we computed a PTSD symptom network and estimated each symptoms’ expected influence (EI), enabling us to compare the network structure and central symptoms with previous PTSD network studies. Second, we computed an emotion dysregulation network to identify the structure and estimated EI for each aspect of emotion dysregulation. Third, we computed a combined network with PTSD symptoms and aspects of emotion dysregulation to identify influential bridge nodes connecting the two constructs. To ensure the robustness of our findings, the networks’ stability was evaluated using the correlation stability coefficient (Epskamp & Fried, 2018). Given that no previous analysis has incorporated a network analysis to examine the individual relations between symptoms of PTSD and aspects of emotion dysregulation, these analyses were mostly exploratory.

2. Method

2.1. Participants

Participants were 463 community women who were recruited at four sites (Jackson, Mississippi, Lincoln and Omaha, Nebraska; Oxford, Ohio) to participate in a study on women’s sexual revictimization. The sample for the current study was drawn from the larger sample of 491 women who completed Wave 1 of the longitudinal study. Eligibility criteria for the larger study required participants to be women between the ages of 18 and 25. To be included in the current analyses, participants had to report a potentially traumatic event on the Life Events Checklist, complete the PTSD Checklist-Civilian Version in relation to the Criterion A event, and complete the Difficulties in Emotion Regulation Scale. Participants’ ages at baseline ranged from 18 to 25 years (M = 21.8, SD = 2.23). The racial/ethnic composition of the sample was 50.1% White, 30.3% Black or African American, 5.3% Latina or Hispanic, 2.1% Asian, 1.9% American Indian or Alaskan Native, and 10.3% multiracial. Participants primarily identified as heterosexual (83.7%), with remaining participants identifying as bisexual (9.7%), lesbian or gay (4.1%), or “something else/don’t know” (2.4%).

2.2. Procedures

The Institutional Review Boards for all participating sites approved of the study procedures. A list of potential participants who fulfilled the eligibility criteria, as stated above, were identified through Survey Sampling International. Recruitment letters were sent to randomly selected women from this list to participate in a study on life experiences and adjustment among young adult women, without any mention of a focus on sexual victimization. Participants were also recruited via community advertisements. After providing written informed consent, participants completed the baseline assessment, consisting of a diagnostic interview and a series of self-report questionaries. All questionnaires were administered online and completed on a computer in the laboratory. Participants received $75 for completing the baseline assessment.

2.3. Measures

2.3.1. Life Events Checklist (LEC; Gray, Litz, Hsu, & Lombardo, 2004).

Participants completed the LEC, a 16-item measure that assesses the occurrence of potentially traumatic events (e.g., transportation accident, sexual assault, combat or exposure to a warzone). Participants reported whether they experienced each event using yes or no response options. The LEC has demonstrated adequate test-retest reliability and convergent validity (Gray et al., 2004). In addition to the 16 listed events, participants were given the option to report and specify any other potentially traumatic event not listed.

2.3.2. PTSD Checklist-Civilian Version (PCL-C; Weathers, Litz, Herman, Huska, & Keane, 1993).

Participants completed the PCL-C, a 17-item measure of DSM-IV-TR (American Psychiatric Association, 2000) PTSD symptoms. Participants indicated how much they had experienced each of these symptoms in the past month in relation to their identified Criterion A index traumatic event. The severity of each symptom was rated on a 5-point scale from 1 (Not at all) to 5 (Extremely). The PCL-C has demonstrated good test-retest reliability, internal consistency, and ability to distinguish between individuals with and without a diagnosis of PTSD (Wilkins, Lang, & Norman, 2011). Internal consistency in the current study was acceptable (α = .95).

2.3.3. Difficulties in Emotion Regulation Scale (DERS-16; Bjureberg et al., 2016).

Participants’ emotion dysregulation was assessed using the DERS-16, a brief version of the original 36-item DERS (Gratz & Roemer, 2004). The DERS-16 assesses difficulties in emotion regulation across the following domains: nonacceptance of negative emotional responses, difficulties engaging in goal-directed behaviors, controlling impulsive behaviors when distressed, limited access to effective emotion regulation strategies, and lack of emotional clarity. Participants responded to each item on a 5-point scale from 1 (Almost never) to 5 (Almost always). Total scores on the DERS-16 can range from 16 to 80, with higher scores reflecting greater emotion dysregulation. The DERS-16 has been found to have excellent internal consistency, good test-retest reliability, and good convergent and discriminant validity (Bjureberg et al., 2016). Internal consistency in this sample was acceptable (α =.94).

2.4. Data analysis plan

2.4.1. Polychoric partial correlation networks.

First, we computed descriptive statistics of demographic variables for our sample (see https://osf.io/5n2p6/ for data, polychoric correlations between individual items and the polyserial correlation between items and the total score on the diagonal, and R code for all main and supplemental analyses). Following established guidelines (Jones et al., 2018), we empirically assessed if any nodes were overlapping using the networktools package (Jones, 2018) in R and determined that there were not significant redundancies between the nodes (p > .05). Following analysis of descriptive statistics, we calculated polychoric partial correlation networks for both PTSD symptoms and aspects of emotion dysregulation separately, followed by a combined PTSD and emotion dysregulation polychoric partial correlation network (see supplemental materials for correlation matrix). Following recent recommendations for reducing bias, networks were estimated without regularization using polychoric correlations (Fried et al., 2020; Williams et al., 2020). All networks used a Fruchterman-Reingold algorithm, resulting in a network structure where nodes do not overlap, and edges have approximately the same length. As a result, visual inspection of the network structure cannot be used to draw inferences (Jones et al., 2018). Following recent publications to evaluate the extent to which centrality indices are driven by measurement characteristics (Hereen et al., 2018; Terluin et al., 2016), we examined the association between node variance and EI to examine the role of differential variability, which occurs when differences in item’s variability impacts the network structure and biases interpretation of findings (Terluin et al., 2016; Wigman et al., 2016). In other words, if the correlation between a node’s standard deviation and centrality is high then differential variability may be driving the centrality of the nodes.

2.4.2. Expected influence.

To quantify how well a node is directly connected to other nodes in the network and because associations between PTSD symptoms and aspects of emotion dysregulation can be either positive or negative, we investigated the centrality measure EI (Fonseca-Pedrero et al., 2018; Robinaugh et al., 2016). EI includes the absolute sum of all the edges, while considering the presence of negative associations. We used the R (Version 3.6; R Core Team, 2019) package qgraph (Epskamp et al., 2012) to estimate EI for each node in the PTSD and emotion dysregulation network.

2.4.3. Bridge expected influence.

To identify bridge nodes connecting PTSD and emotion dysregulation, we calculated bridge expected influence (BEI) using the R package networktools (Jones, 2018). BEI calculates the sum of edge weights from a node in one community and its association with nodes from the other community (Jones et al., 2019). An edge with a positive value indicates than an increase in activation of one node is associated with an increase in activation of the node connected to it, whereas a node with a negative edge indicates that an increase in the first node is associated with a decrease in the second node. In this analysis, we set the 17 PTSD symptoms to reflect one community and the 16 aspects of emotion dysregulation to form the other. In addition to examining the bridge nodes connecting communities, we examined the correlations between individual item scores and totals scores on PTSD and emotion dysregulation, respectively.

2.4.4. Stability of network models.

To estimate the stability of the PTSD and emotion dysregulation networks, we performed 5000 case-dropping bootstraps with the bootnet package (Epskamp et al., 2018). The bootstrapping procedure returns a correlation stability (CS) coefficient, which indicates the proportion of cases that could be eliminated from the analysis while still retaining a correlation of at least .7 with the original estimates within a 95% confidence interval. In other words, the CS coefficient is an estimate of stability of estimated values relative to one another across bootstrapped samples. We estimated CS coefficients for EI and BEI. Based on current conventions for network inference (Epskamp & Fried, 2018), the CS-coefficient should be above .50, and when the CS-coefficient is below 0.25 the stability of centrality indices is low.

3. Results

3.1. Sample Description.

In the current sample, the most common potentially traumatic events were “rape or other unwanted sexual experiences” (31%), “sudden unexpected death” (19%), “transportation accident” (11%), and “physical assault” (7%). In total, 42% of the sample was above the clinical cutoff for PTSD (Weathers et al., 1993), as indicated by a score equal to or greater than 31 on the PCL-C (M = 32.23, SD = 15.7). The average age at which participants reported experiencing their index traumatic event was 16 (M_age = 16.3, SD = 4.9). All 33 items are positively skewed (mean skewness = 1.18, range 0.34 to 1.85; mean kurtosis = 0.59, range −.94 to 2.45) with an average item mean of 1.95 (range 1.60 to 2.79) on a 5-point scale from 1 to 5. In examining the association between node variance and EI, the bootstrapped correlation was significant (r = .17, 95% CI [.06, .26]), indicating that differences in centrality were weakly positively related to differences in item variability.

3.2. PTSD network

Edge weights of the PTSD symptom network were stable, CS (correlation [cor] = .7) = 0.75, as were EI values, CS (cor = .7) = 0.75 (Epskamp & Fried, 2018). The PTSD network and the EI values for each node appear in Figure 1. Most associations in this network were positive. The node disturbing memories (EI = 1.52) was the most central symptom in the PTSD network, followed by having difficulty concentrating (EI = 1.11), having physical reactions when reminded of the event (EI = 1.07), and feeling distant or cut off from other people (EI = 1.06). Trouble remembering important parts of the traumatic event (EI = .55) was the least central symptom. Strong edges emerged between being super-alert and easily startled (D4—D5 = .55), as well as between feeling distant or cut off from other people and feeling emotionally numb (C5—C6 = .42). Edges between loss of interest in activities and feeling distant or cut off from other people (C4—C5 = .36), as well as disturbing memories and disturbing dreams (B1—B2 = .34) also had a high magnitude.

Figure 1.

PTSD partial correlation network and expected influence (EI) values for each node. The partial correlations between two nodes are represented by an edge, and edge thickness represents the strength of the association. Blue lines represent a positive association, while red lines represent a negative association. Disturbing memories (B1) and difficulty concentrating (D3) had the highest EI values.

3.3. Emotion dysregulation network

The emotion dysregulation network was stable as indicated by CS coefficients for edge weights, CS (cor =.7) = .75, and for EI values, CS (cor = .7) = .67. The emotion dysregulation network and EI values for each node appear in Figure 2. Feeling bad about oneself for feeling upset (EI = 1.15), feeling out of control when upset (EI = 1.11), and feeling irritated with oneself for feeling upset (EI = 1.05) emerged as the most central nodes in this network. Confused about feelings (EI = .81) was the least central node. The strongest edges in the emotion dysregulation network emerged between difficulty making sense of feelings and confused about feelings (ED1—ED2 = .46), difficulty getting work done when upset and difficulty focusing on other things when upset (ED3—ED7 = .47), and becoming out of control when upset and feeling out of control when upset (ED4—ED8 = .46).

Figure 2.

Emotion dysregulation partial correlation network and expected influence (EI) values for each node. The partial correlations between two nodes are represented by an edge, and edge thickness represents the strength of the association. Blue lines represent a positive association, while red lines represent a negative association. Feeling bad about oneself for feeling upset (ED14), feeling out of control when upset (ED8), and feeling irritated with oneself for feeling upset (ED13) were the nodes with the highest EI.

3.3. Combined PTSD and emotion dysregulation network.

Based on the CS coefficients, the edge weights in the combined PTSD and emotion dysregulation network were stable, CS (cor =.7) = .0.75, as were the BEI values, CS (cor = .7) = .52. The PTSD and emotion dysregulation networks and BEI for each node appear in Figure 3. In the combined network, having difficulty concentrating (BEI = .14), believing that distress will turn into depression (BEI = .11), difficulty making sense of feelings (BEI = .10), and emotions feel overwhelming (BEI = .10) were the nodes with the highest magnitude BEI. For individual PTSD symptoms, there was a significant association between all individual items and the total emotion dysregulation score (p < .05). Regarding aspects of emotion dysregulation, there was a significant association between all individual items and the total PTSD score (p < .05). See Table 2 for correlations between individual items and the total scores.

Figure 3.

Combined partial correlation PTSD and emotion dysregulation network and bridge expected influence (BEI) values for each node. The partial correlations between two nodes are represented by an edge, and edge thickness represents the strength of the association. Blue lines represent a positive association, while red lines represent a negative association. Grey nodes correspond to PTSD nodes and white nodes reflect emotion dysregulation nodes. Having difficulty concentrating (D3), believing distress will turn into depression (ED6), and difficulty making sense of feelings (ED1) were the nodes with the highest BEI values.

Table 2.

Node Network Psychometrics

Item	Mean & SD	Skew	Kurtosis	Expected Influence	Bridge Expected Influence	Correlation with ED Total Score
B1: Disturbing memories	2.14 (123)	.99	−0.07	1.15	0.009	0.376*
B2: Disturbing dreams	1.70 (109)	1.73	2.08	0.80	0.003	0.295*
B3: Acting or feeling as if the experience were happening again	1.72 (111)	1.57	1.54	0.95	0.007	0.354*
B4: Feeling very upset when reminded of the event	2.33 (132)	0.74	−0.65	1.01	0.060	0.440*
B5: Having physical reactions when reminded of the event	1.89 (125)	1.30	0.50	1.07	0.000	0.366*
C1: Avoid thinking, talking, or feelings about the event	2.43 (1.41)	0.77	−0.74	0.81	0.011	0.334*
C2: Avoid activities or situations because reminded of the event	1.98 (1.27)	1.30	0.57	0.94	0.018	0.294*
C3: Trouble remembering important parts	1.78 (1.21)	1.80	2.25	0.56	0.079	0.282*
C4: Loss of interest in activities	1.59 (105)	1.78	2.20	1.00	−0.007	0.294*
C5: Feeling distant or cut off from other people	1.77 (117)	1.31	0.48	1.06	0.014	0.358*
C6: Feeling emotionally numb or being unable to have loving feelings	1.66 (115)	1.68	1.68	1.01	0.017	0.339*
C7: Feeling as if your future will somehow be cut short	1.59 (107)	1.85	2.41	0.75	0.084	0.331*
D1: Trouble falling or staying asleep	2.00 (1.34)	1.10	−0.16	0.89	0.002	0.364*
D2: Feeling irritable or having angry outbursts	1.78 (1.12)	1.35	0.70	0.96	0.096	0.400*
D3: Having difficulty concentrating	1.96 (126)	1.10	−0.03	1.10	0.144	0.451*
D4: Being ‘super-alert’ or watchful or on guard	2.11 (144)	1.21	0.09	0.89	0.000	0.312*
D5: Feeling jumpy or easily startled	1.85 (130)	1.50	1.10	0.97	0.029	0.367*
						Correlation with PCL Total Score
ED1: I have difficulty making sense out of my feelings	1.97 (0.99)	1.02	0.63	0.81	0.101	0.400*
ED2: I am confused about how I feel	1.91 (0.90)	1.10	1.31	0.75	0.064	0.326*
ED3: When I am upset, I have difficulty getting work done	2.72 (1.27)	0.34	−0.94	0.89	0.022	0.295*
ED4: When I am upset, I become out of control	1.56 (0.93)	1.69	2.45	0.97	0.056	0.368*
ED5: When I am upset, I believe I will remain that way for a long time	1.77 (105)	1.30	0.94	0.90	−0.006	0.366*
ED6: When I am upset, I believe that I’ll end up feeling very depressed	1.88 (117)	1.13	0.27	1.00	0.110	0.454*
ED7: When I am upset, I have difficulty focusing on other things	2.88 (122)	0.28	−0.89	0.91	0.035	0.273*
ED8: When I am upset, I feel out of control	1.74 (103)	1.45	1.42	1.11	−0.003	0.321*
ED9: When I am upset, I feel ashamed with myself for feeling that way	1.97 (115)	1.08	0.23	0.92	0.027	0.330*
ED10: When I am upset, I feel like I am weak	2.32 (132)	0.73	−0.58	0.84	0.032	0.363*
ED11: When I am upset, I have difficulty controlling my behaviors	1.71 (0.93)	1.34	1.24	0.90	0.002	0.369*
ED12: When I am upset, I believe that there is nothing I can do to make myself feel better	1.70 (102)	1.46	1.49	0.86	−0.003	0.302*
ED13: When I am upset, I become irritated with myself for feeling that way	2.13 (123)	0.88	−0.19	1.05	0.004	0.324*
ED14: When I am upset, I start to feel very bad about myself	2.14 (1.30)	0.91	−0.34	1.15	0.024	0.387*
ED15: When I am upset, I have difficulty thinking about anything else	2.58 (1.19)	0.54	−0.66	1.03	0.000	0.302*
ED16: When I am upset, my emotions feel overwhelming	2.45 (1.25)	0.66	−0.63	1.01	0.099	0.407*

^*Correlation is significant at the 0.05 level

4. Discussion

The goal of the present study was to extend extant literature by identifying the differential associations among symptoms of PTSD and aspects of emotions dysregulation in a large sample of women with traumatic exposure. Using network analysis methodology, our findings advance understanding of the structure and interrelations of PTSD and aspects of emotion dysregulation by identifying influential nodes between each construct. In addition to investigating broad relations between PTSD symptoms and aspects of emotion dysregulation, we examined key nodes within each construct.

4.1. PTSD Networks

Results of analyses investigating the PTSD symptom network reveal the importance of disturbing memories, concentration difficulties, experiencing physical reactions when reminded of the event, and feeling distant or cut off from other people to PTSD. These results align with those of a recent systematic review of PTSD network analyses, which found that the disturbing or recurrent thoughts and memories node had strong centrality (Birkeland et al., 2020). Results from this study are also consistent with a large PTSD network analysis of four independent trauma exposed populations that differed in cultural background, trauma type, and severity, which found concentration difficulties, disturbing memories, and sleep difficulties to be the most influential symptoms across the four samples (Fried et al., 2018). Intrusive and disturbing traumatic memories may make it difficult for trauma-exposed individuals to focus attention or complete tasks. Further, consistent with the systematic review and other PTSD network studies (Armour et al., 2017; Fried et al., 2018; Greene et al., 2018; Moshier et al., 2018; Peters et al., 2021; von Stockert et al., 2018), the strongest edge within PTSD occurred between hypervigilance and exaggerated startle response. The strong associations between hypervigilance and exaggerated startle response are consistent with the suggestion that these two symptoms may bidirectionally contribute to the maintenance of PTSD (e.g., Kimble et al., 2014; Rauch & Foa, 2006). Overall, the current results mirror prior PTSD network analyses and add to a growing consensus regarding the replicability of PTSD network models (e.g., Birkeland et al., 2020).

Regarding the assessment of PTSD, these findings contrast with the recently proposed criteria for PTSD in the 11^th edition of the International Classification of Disease (ICD) in several keyways. The new ICD conceptualization of PTSD only includes 6 of the 20 symptoms currently in DSM-5. The six symptoms are distressing dreams, intrusions, efforts to avoid distressing thoughts, efforts to avoid external reminders, hypervigilance, and exaggerated startle response, as they were assumed to be core symptoms of PTSD that are not shared with other disorders (Cloitre et al., 2013; Maercker et al., 2013). However, of the six symptoms with the highest EI, only one of the six proposed ICD-11 PTSD symptoms, disturbing memories, demonstrated high EI in our analyses. Moreover, several of the key symptoms identified here reflect symptoms frequently shared across internalizing disorders, such as depression (e.g., difficulty concentrating, anhedonia). These results are consistent with previous analyses documenting that the narrow ICD-11 approach to diagnosing PTSD may miss critical information and provide additional support for the importance of assessing nonspecific symptoms as indicators of PTSD (e.g., Mitchell et al., 2017; O’Donnell et al., 2014; Stein et al., 2014).

4.2. Emotion dysregulation network

Feeling bad about oneself for feeling upset was the most central aspect of emotion dysregulation in our networks. Although this study is among the first to examine emotion dysregulation using network methodology, these results are consistent with a large body of literature emphasizing the benefits of emotional acceptance and negative consequences of emotional nonacceptance (e.g., Ford et al., 2018; Shallcross et al., 2010). Several treatments, including Dialectical Behavior Therapy (DBT; Linehan, 1993), Acceptance and Commitment Therapy (ACT; Hayes et al., 2006), and Mindfulness-Based Cognitive Therapy (MBCT; Segal et al., 2002), encourage clients to observe their emotions as they occur in the moment. The process of observing emotions facilitates emotional acceptance and the decoupling of emotions from behaviors, ultimately resulting in increased emotional awareness, clarity, and adaptive emotion regulation (e.g., Blackledge & Hayes, 2001; Gratz, 2007; Ford & Gross, 2019). Thus, our findings reinforce the importance of accepting one’s internal experience, as well as being compassionate toward oneself when experiencing negative emotions.

In addition to identifying the importance of feeling bad about oneself for feeling upset in the emotion dysregulation network, feeling out of control when upset was an important aspect of emotion dysregulation and had strong associations with becoming out of control when upset and difficulty controlling behaviors when upset. These finding are consistent with research on the role of attentional control in effective emotion regulation (Bardeen et al., 2014; Bardeen et al., 2015). Attentional control is defined as the ability to strategically deploy higher-order executive attention in regulating bottom-up emotional processes and is theorized to be central to effective emotion regulation (Bardeen et al, 2014; Gross, 2015). Results of this study suggest that difficulties focusing attention when distressed may relate strongly to difficulties controlling behavior when distressed.

4.3. Combined PTSD and emotion dysregulation network

In our combined network with PTSD symptoms and aspects of emotion dysregulation, the PTSD node concentration difficulties had the strongest association with aspects of emotion dysregulation, whereas believing that distress will turn into depression was the aspect of emotion dysregulation that had the strongest association with PTSD symptoms. A previous network investigation of PTSD and emotion dysregulation (Weiss et al., 2020) revealed that the alterations in arousal and reactivity (AAR) PTSD symptom cluster had the highest bridge association with positive aspects of emotion dysregulation, whereas nonacceptance of positive emotions and difficulties engaging in goal-directed behavior when experiencing positive emotions had the highest bridge associations with the PTSD symptom clusters. Results from our analysis complement and contrast with these findings in several keys ways. First, consistent with their findings (Weiss et al., 2020), our results identified the node difficulty concentrating, which forms part of the AAR PTSD symptom cluster based on DSM criteria, as having the strongest bridge association with aspects of emotion dysregulation. This finding supports the suggestion that arousal may underlie emotion dysregulation (albeit of positive emotions) among individuals with PTSD (Weiss et al., 2018), and extends that finding by pointing to specific nodes that may be responsible for similar associations between difficulties regulating negative emotions and PTSD. Moreover, our finding is in line with results from a dynamic network analysis indicating that arousal drives negative emotions (Greene et al., 2019).

In contrast, our analysis revealed that the node believing that distress will turn into depression, an aspect of emotion dysregulation reflecting difficulties in the effective regulation of emotional arousal, had the strongest bridge association with PTSD symptoms, whereas Weiss and colleagues (2020) identified nonacceptance of positive emotions and difficulties engaging in goal directed behavior as aspects of emotion dysregulation with high bridge associations. These differences in findings between the two studies could be due to several factors, including differences in the number and content of the nodes in the networks (7 vs. 33), examining different dimensions of emotion dysregulation (positive vs. negative), measuring PTSD and emotion dysregulation differently (clusters vs. symptoms), and sample characteristics (all women vs. mixed community sample). Thus, future work is needed to explore the factors that may account for these differences and work toward reaching a consensus on the key network associations between PTSD and emotion dysregulation.

4.4. Clinical Implications

The node difficulty concentrating had high EI and BEI values, thereby making it an appealing target for clinical intervention. Consistent with previous PTSD symptom networks (Birkeland et al., 2020; Fried et al., 2018; McNally et al., 2015; Peters et al., 2021; von Stockert et al., 2018), difficulty concentrating emerged as a highly central symptom, and had strong associations with sleep difficulties and irritability. Consistent with this finding, McNally and colleagues (2015) propose that sleep difficulties may contribute to irritability, which, in turn, may contribute to concentration difficulties. One strategy for improving concentration difficulties and producing a therapeutic cascade could be mindfulness meditation. The proposed mechanism of action through which mindfulness meditation works is by enhancing concentration (e.g., Hölzel et al., 2011; Keng et al., 2011). Moreover, mindfulness meditation has been found to be an effective intervention for improving PTSD symptoms (e.g., Hilton et al., 2017) and emotion dysregulation (e.g., Mitchell et al., 2017). Theoretically, improving concentration difficulties through mindfulness meditation should lead to a therapeutic cascade in a PTSD and emotion dysregulation psychological network. Targeting concentration through connected nodes may be an alternative clinical approach. For instance, sleep difficulties shared a strong edge with concentration difficulties, suggesting that improving sleep could also improve concentration. There is increasing evidence that sleep disturbance plays a central mechanistic role in the development and maintenance of PTSD (Colvonen et al., 2019) and may bidirectionally reinforce emotion dysregulation (e.g., Goldstein & Walker, 2014; Semplonius & Willoughby, 2018). Thus, targeting concentration difficulties directly through mindfulness meditation or via a connected node may be a clinically useful application; however, future research is needed to test these hypotheses.

4.5. Limitations and Future Research

Results from this study should be considered in light of several limitations. First, this study is cross-sectional, it cannot provide information regarding the direction of the associations between PTSD and emotion dysregulation. Prospective and longitudinal investigations of the differential associations between PTSD and emotion dysregulation needed.

Second, a limitation of network analyses is the possibility of omitting a critical variable from the network, which can undermine inferences regarding centrality. For example, the diagnostic criteria for PTSD in DSM-5 (which were published after the current data were collected) now include negative alterations in cognition and mood (NACM). Even though symptoms from the NACM are present in the PCL-C, using the DSM-IV PTSD symptoms may have influenced the way symptoms grouped together within the PTSD cluster. Further, given that NACM may overlap with emotion dysregulation, this particular set of symptoms may be a potential bridge between PTSD and emotion dysregulation clusters. Indeed, three recent DSM-5 PTSD network studies found that individual symptoms from the NACM had high centrality (Armour et al., 2017; Bartels et al., 2019; Mitchell et al., 2017); however, the three previous studies examined only PTSD symptoms and did not consider aspects of emotion dysregulation. Future studies are needed to incorporate the symptoms reflected in the NACM cluster to determine whether these additional symptoms evidence a distinct relation to emotion dysregulation. Future studies should also examine whether symptoms of other forms of psychopathology commonly associated with both PTSD and emotion dysregulation (including borderline personality disorder) help explain the high co-occurrence of PTSD and emotion dysregulation.

Third, the exclusive focus on young adult women within this study – most of whom reported sexual victimization as their index traumatic event – may limit the generalizability of these findings. Although the high prevalence of sexual violence suggests these findings may have relevance for a large subset of women, these findings may not generalize to men who have experienced sexual trauma or individuals who have experienced other traumatic events. In addition, PTSD and emotion dysregulation network may also differ as a function of gender, given that the factor structure and expression of PTSD symptoms is different between males and females (Armour et al., 2011; Hall et al., 2012; Hourani et al., 2015).

Fourth, using self-report measures to assess both PTSD symptoms and aspects of emotion dysregulation is an additional limitation, as retrospective self-reports of these constructs may reflect worst day symptoms or difficulties, rather than overall levels (e.g., Schuler et al., 2021). Futures studies investigating the association between PTSD and emotion dysregulation should consider implementing micro-longitudinal designs, such as ecological momentary assessment, with objective indicators (i.e., heart rate variability, galvanic skin conductance) to reduce recall bias and limitations posed by self-report assessments.

One direction for future research is highlighted by evidence that PTSD and emotion dysregulation are not temporally stable constructs, but rather fluctuate in severity and course over time (e.g., Cole et al., 2019; Greene et al., 2019). This suggests that networks may also change over time in response to environmental and individual experiences. Future studies should investigate temporal changes in PTSD and emotion dysregulation and their relations to reveal how their interrelations may unfold over time (Bringmann et al., 2016). For example, studies could follow participants after initial trauma exposure to examine how the trajectory of PTSD symptoms and aspects of emotion dysregulation develop and change prospectively. Intensive longitudinal designs would also facilitate examination of whether highly influential nodes temporally predict increases in other PTSD symptoms or emotion dysregulation aspects and, thus, may be useful targets for intervention.

To date, most research examining associations between PTSD and emotion dysregulation has focused exclusively on groups of individuals, rather than the nature of these associations within a specific individual. Yet, findings from these nomothetic approaches do not account for individual differences and, thus, may not generalize to every person (Hamaker, 2012; Molenaar, 2004). For this reason, there has been a call for more person-level studies (i.e., idiographic approaches; Fisher et al., 2019; Schork, 2015). Applied in the present context, it is possible that specific individuals may have bridge and high centrality nodes that differ from those identified within aggregated, group-level data. This possibility suggests the need to conduct idiographic network analyses (Fisher et al., 2019) to examine person-level relations between PTSD and emotion dysregulation. Such research may be useful in elucidating person-specific targets for treatment – consistent with similar strategies used to model psychotherapy progress (Kaiser & Laireiter, 2018). Indeed, although there are efficacious treatments for PTSD, many individuals continue to meet criteria for PTSD following treatment (Steenkamp et al., 2015). Thus, there may be utility in supplementing those interventions with idiographic data on the individual’s most salient and central PTSD symptoms and aspects of emotion dysregulation.

4.6. Conclusion

This study represents the first application of the network analysis approach to elucidate the precise nature of the relations among specific PTSD symptoms and aspects of emotion dysregulation. Within the PTSD network, disturbing memories and concentration difficulties were central nodes, whereas feeling bad about oneself for feeling upset was the most central aspect of emotion dysregulation. Findings suggest that specific PTSD symptoms and aspects of emotion dysregulation are differentially associated through particular nodes, including difficulties concentrating and believing that distress will turn into depression. The PTSD symptoms and aspects of emotion dysregulation identified here may be viable targets for interventions for women who have been exposed to trauma, as well as future research on transdiagnostic relations of PTSD with related psychopathology that shares an underlying emotion dysregulation (e.g., borderline personality disorder). Indeed, consistent with other therapeutic interventions that successfully weaken or eliminate node-to-node relations (Craske et al., 2008), targeting highly influential and bridging nodes that contribute to the high co-occurrence of PTSD and emotion dysregulation may be useful for treatment development and refinement. Overall, our results shed light on some of these associations by identifying specific PTSD symptoms, namely concentration difficulties, and aspects of emotion dysregulation that may be acting as a bridge between the two constructs.

Table 1.

Descriptive Information of Demographic and Study Variables

	N or M	% or SD
Age (Years)	21.74	2.23
Education
Some High School	20	4.3%
GED	18	3.8%
High School Graduate	70	15.1%
Business or Technical Training	8	1.7%
Some College	246	53.1%
College Graduate	64	13.8%
Graduate Degree	37	7.9%
Student Status
Full-time	256	52.1%
Part-time	47	10%
Not a student	187	38.1%
Employment Status
Employed	265	53.97%
Unemployed	226	46.02.%
Ethnicity/Race
White	232	50.1%
African American	139	30.3%
American Indian/Indigenous	9	1.9%
Hispanic/Latino/Spanish origin	25	5.3%
Asian American	10	2.1%
Multiracial	48	10.3%
PTSD Checklist-Civilian Version	32.25	15.52
DERS-16	33.33	13.40
Life Events Checklist
Natural disaster	20	4.31%
Fire or explosion	6	1.29%
Transportation accident	45	9.72%
Serious Accident	8	1.72%
Exposure to toxic substance	-	-
Physical assault	33	7.12%
Assault with a weapon	7	1.51%
Rape	50	10.8%
Other unwanted sexual experience	90	19.4%
Combat	1	.215%
Captivity	5	1.08%
Illness or injury	2	.432%
Severe human suffering	20	4.32%
Sudden, violent death	92	19.87%
Sudden, unexpected death	6	1.29%
Other event	48	10.38%

Highlights.

• The network perspective on psychopathology was used to understand the differential associations between PTSD symptoms and aspects of emotion dysregulation.

• Concentration difficulties emerged as an influential node in the PTSD network and the combined PTSD and emotion dysregulation network.

• Targeting concentration difficulties through mindfulness meditation may be one way to bring about a therapeutic cascade in a PTSD and emotion dysregulation network.

• Targeting concentration difficulties by way of neighboring nodes may be an alternative approach to bring about a therapeutic cascade.