Daily Reciprocal Relations between Negative and Positive Emotion Dysregulation and Posttraumatic Stress Symptoms among Individuals with a History of Sexual Assault

1. Introduction

Sexual assault is alarmingly common in the United States, with national lifetime estimates approaching one third (30%) of the population (Kilpatrick et al., 2013). Individuals with a history of sexual assault demonstrate higher rates and more severe profiles of posttraumatic stress symptoms (PTSS) relative to other trauma-exposed populations (Kilpatrick et al., 2013; Smith et al., 2016). PTSS is characterized by intrusive trauma memories, avoidance of trauma reminders, negative alterations in cognition and mood, and alterations in arousal and reactivity (American Psychiatric Association [APA], 2013). PTSS can be debilitating, resulting in poor physical (Campbell et al., 2008) and mental (Filpas et al., 2006) health outcomes. While some PTSS is expected immediately following sexual assault, a subset of individuals experience chronic PTSS (Santiago et al., 2013; Steenkamp et al., 2012). Therefore, there is an urgent need to better understand malleable factors underlying the maintenance of PTSS among those who have experienced sexual assault.

Central to PTSS is emotion dysregulation. Gratz and Roemer (2004) define emotion dysregulation as a multidimentional construct that reflects difficulties understanding and accepting emotional experiences; deficits controlling impulsive behavior and persisting with goal-directed behavior in the context of emotions; and inflexible use of situationally appropriate strategies for altering emotional responses. Parallel to neuroimaging studies that identify overlapping yet distinct pathways for the regulation of negative versus positive emotions (Kim & Hamann, 2007), self-reported dysregulation occurs for both negative and positive emotions (Weiss et al., 2015). A long line of research has focused on the role of negative emotion dysregulation in the context of PTSS. Findings consistently demonstrate that individuals with greater negative emotion dysregulation experience increased PTSS (Tull et al., 2007; Weiss et al., 2013), including among those who have experienced sexual assault (Ullman et al., 2014). Similarly, recent work has detected a link between positive emotion dysregulation and PTSS, such that individuals with elevated positive emotion dysregulation experience heightened PTSS (Weiss et al., 2018; 2020), even after statistically accounting for the effects of negative emotion dysregulation (Weiss et al., 2019) and including among those who have experienced sexual assault (Forkus et al., 2020). Taken together, both negative and positive emotion dysregulation appear to be important factors related to PTSS that warrant further study.

In particular, theoretical and empirical evidence implicates emotion dysregulation as a risk factor for subsequent PTSS among those with a history of sexual assault. Individuals with elevated negative emotion dysregulation may be more likely to experience difficulties modulating intense negative emotional reactions to trauma reminders; as an attempt to cope, these individuals may be more likely to rely on maladaptive emotion regulation strategies, such as avoidance, that preclude processing of information needed to correct cognitive fear structures, thereby increasing PTSS (Foa & Kozak, 1986). Indeed, findings have found that negative emotion dysregulation prospectively predicts PTSS (Bardeen et al., 2013; Forbes et al., 2020), and treatments aimed at reducing negative emotion dysregulation also alleviate PTSS (Cloitre et al., 2002). Relatedly, individuals with elevated positive emotion dysregulation may be more likely to experience difficulties modulating the heightened anxious arousal accompanying trauma reminders (APA, 2013) given the similarity in physiological responding (e.g., racing heart) between positive emotions and anxious arousal (Raudales et al., 2021; Weiss et al., 2018); this would subsequently increase PTSS. Alternatively, positive emotion dysregulation may lead individuals to engage in maladaptive responses to positive emotions, such as dampening and reduced positive rumination (Wolkenstein et al., 2022), which may lead to characteristic PTSS such as emotional numbing and diminished positive affect (APA, 2013). Yet, much less is known about how positive emotion dysregulation may contribute to PTSS over time. Furthermore, it is possible that PTSS, in turn, exacerbates emotion dysregulation. Lanius and colleagues (2010) propose that PTSS, which involves re-experiencing the trauma (APA, 2013), lead to sensitization of the fear response and kindling of reactivity that overwhelms general emotion regulatory systems. However, this theory largely remains untested.

Despite well-established relations between both negative and positive emotion dysregulation and PTSS, there is a lack of research utilizing dynamic, micro-longitudinal methods to better understand the nature and direction of these associations. Micro-longitudinal design entails repeated sampling of experiences in near real-time and within natural environments, facilitating examinations of both within- and between-individual variations (Bolger & Laurenceau, 2013; Tennen et al., 2000). Compared to traditional cross-sectional or longitudinal research, micro-longitudinal studies improve ecological validity and reduce recall bias. Micro-longitudinal methods will also facilitate the capture of proximal and directional associations between emotion dysregulation and PTSS. To test these relations, it is ideal to use Dynamic Structural Equation Modeling (DSEM; Asparouhov et al., 2018), which is a cutting-edge statistical technique that extends conventional methods by uniquely modeling time-varying processes. An understanding of the dynamic emotion dysregulation context of PTSS can inform micro-interventions (e.g., just-in-time adaptive interventions; Nahum-Shani et al., 2018) to provide brief treatments tailored to emotional states as they occur within a real-world setting.

The current study used a micro-longitudinal approach to test the dynamic relations between both negative and positive emotion dysregulation and PTSS among individuals who have experienced sexual assault. Importantly, this study aimed to address vital limitations of extant research that is primarily cross-sectional and focused on negative emotion dysregulation. We hypothesized reciprocal associations, such that both negative and positive emotion dysregulation would be associated with increased next-day PTSS, and also, PTSS would be associated with increased next-day negative and positive emotion dysregulation.

2. Methods

2.1. Procedures and Participants

All procedures were reviewed and approved by the [redacted] Institutional Review Board. The current study utilized data from a larger study examining the relations among sexual assault and risky behaviors. Participants were recruited from Amazon’s Mechanical Turk (MTurk), an internet-based crowdsourcing platform that generates reliable data (Buhrmester et al., 2011; Shapiro et al., 2013), and represents the general population in terms of demographics (Mischra & Carleton, 2017) and prevalence of mental health problems (Shapiro et al., 2013), including PTSS (van Stolk‐Cooke et al., 2018).

Inclusionary criteria consisted of: (a) living in North America, (b) being at least 18 years of age, (c) having working knowledge of the English language, and (d) reporting a lifetime history of sexual assault. Eligible participants provided informed consent prior to completing surveys on REDCap, a secure data collection platform. Participants first completed a baseline survey. To optimize data quality, validity checks were inserted in the baseline survey to assess for attentive responding and comprehension (4 items; e.g., “I have never brushed my teeth;” Aust et al., 2013; Meade & Craig, 2012; Oppenheimer et al., 2009; Thomas & Clifford, 2017).

Following completion of the baseline survey, participants who indicated interest in participating in seven consecutive days of micro-longitudinal assessment were asked to provide an email address. The following day, these individuals were emailed a REDCap link to a brief (approximately five minute) survey. Surveys were sent once daily at 5pm for seven consecutive days. Survey responses were time stamped and reviewed to ensure that they were completed within seven hours of the initial email.

Participants were compensated $2.00 through MTurk for completing the baseline survey. For the micro-longitudinal portion of the study, participants received $1.00 per survey and a bonus of $5.00 for completing at least six out of the seven surveys, sent in the form of an Amazon gift card at the end of the seven days of surveys.

Participants (N = 121) ranged in age from 18 to 68 (M = 35.27, SD = 11.17). Most of the sample identified as women (70.2%) and white (81.0%). Additional demographics are provided in Table 1.

Table 1.

Descriptive Characteristics of the Sample

Variables	Full Sample (N = 121)
	Mean (SD)
Age	35.27 (11.17)
Trauma Load	9.41 (5.59)
	n (%)
Gender
Woman	85 (70.2%)
Man	33 (27.3%)
Not Listed	3 (2.5%)
Race (could endorse more than one category)
White	98 (81.0%)
African American/Black	12 (9.9%)
Asian	11 (9.1%)
American Indian/Alaska Native	2 (1.7%)
Native Hawaiian/Other Pacific Islander	1 (0.8%)
Not Listed	1 (0.8%)
Ethnicity
Hispanic or Latino/a	10 (8.4%)
Not Hispanic or Latino/a	109 (91.6%)
Employment Status
Employed Part-time	18 (14.9%)
Employed Full-time	86 (71.1%)
Not in Labor Force (student, homemaker)	8 (6.6%)
Unemployed	9 (7.4%)
Family Annual Income
< $15,000	79 (65.3%)
$15,000 – $24,999	2 (1.7%)
$25,000 – $34,999	5 (4.1%)
$35,000 – $49,999	6 (5.0%)
$50,000 – $64,999	3 (2.5%)
$65,000 – $79,999	15 (12.4%)
$80,000 or higher	11 (9.1%)

2.2. Exclusions and Missing Data

See Figure 1 for recruitment flow. Of the obtained 3,157 responses, only 1,081 participants met for all inclusionary criteria. We then excluded 427 participants who failed to pass any of the four validity checks and 40 participants who attempted the survey more than once. Using data obtained from the Life Event Checklist for DSM-5 (LEC-5 items 8 or 9; Weathers et al., 2013), we excluded 128 participants who did not endorse experiencing a sexual assault. Of these 486 eligible participants at baseline, only 155 chose to participate in the micro-longitudinal assessment. We excluded 10 participants who were missing all seven days of surveys for primary study variables, and an additional 24 participants who only completed one of the seven days of surveys for primary study variables. Thus, the final sample consisted of 121 participants. We did not calculate composite scores for measures missing any item-level data, which resulted in less than 20% missing composite scores (see Analytic Rationale).

Figure 1.

Recruitment Flow via Amazon’s Mechanical Turk

2.3. Baseline Measures

Demographic information.

Information regarding age, gender, ethnicity, race, income, educational level, employment status, and relationship status was collected through the baseline survey.

Traumatic exposure.

The Life Events Checklist for DSM-5 (LEC-5; Weathers et al., 2013) is a 17-item self-report measure of lifetime trauma that was used to confirm sexual assault history at baseline (items 8 “sexual assault [rape, attempted rape, made to perform any type of sexual act through force or threat of harm]” or 9 “other unwanted or uncomfortable sexual experience”). Participants indicate their exposure to each event on a scale with six categories: happened to me, witnessed it, learned about it, part of my job, not sure, and does not apply. Sexual assault exposure – consistent with the DSM-5 Criterion A – was based on the endorsement of any of the first four response options (APA, 2013). The LEC-5 has demonstrated strong convergent validity with measures assessing traumatic exposure and psychopathology known to relate to traumatic exposure (Gray et al., 2004).

2.4. Daily Measures

Participants were asked to report on behaviors since the previous scheduled survey (i.e., the past 24 hours).

Negative Emotion Dysregulation.

The Momentary Difficulties in Emotion Regulation Scale (mDERS; Weiss et al., under review), an abbreviated, six-item momentary version of the Difficulties in Emotion Regulation Scale (DERS; Gratz & Roemer, 2004), was utilized to assess negative emotion dysregulation across six domains: nonacceptance of negative emotions, difficulties engaging in goal-directed behaviors when experiencing negative emotions, difficulties controlling impulsive behaviors when experiencing negative emotions, limited access to emotion regulation strategies perceived as effective in the context of negative emotions, lack of negative emotional awareness, and lack of negative emotional clarity. The DERS is a widely used measure with sound psychometric properties, demonstrating good test-retest reliability, convergent validity, and discriminant validity (Gratz & Roemer, 2004). Participants rate the extent to which items apply to them using a 4-point Likert-type scale ranging from 1 (not true at all) to 4 (very true). Items are summed with higher scores indicating greater momentary difficulties regulating negative emotions. Multilevel reliabilities (Geldhof et al., 2014) in the current sample were good (within-person ω = 0.72, between-person ω = 0.94).

Positive Emotion Dysregulation.

The Momentary Difficulties in Emotion Regulation Scale – Positive (mDERS-P; Weiss et al., under review), an abbreviated, six-item momentary version of the Difficulties in Emotion Regulation Scale – Positive (DERS-P; Weiss, Gratz, et al., 2015), was utilized to assess positive emotion dysregulation across six domains: nonacceptance of positive emotions, difficulties engaging in goal-directed behaviors when experiencing positive emotions, difficulties controlling impulsive behaviors when experiencing positive emotions, limited access to emotion regulation strategies perceived as effective in the context of positive emotions, lack of positive emotional awareness, and lack of positive emotional clarity. The DERS-P has demonstrated reliability and validity (Weiss et al., 2019; Weiss et al., 2015). Participants rate the extent to which items apply to them using a 4-point Likert-type scale ranging from 1 (not true at all) to 4 (very true). Items are summed, with higher scores indicating greater momentary difficulties regulating positive emotions. Multilevel reliabilities (Geldhof et al., 2014) in the current sample were acceptable (within-person ω = 0.62, between-person ω = 0.92).

Posttraumatic Stress Symptoms.

A momentary version of the PTSD Checklist for the DSM-5 (PCL-5; Weathers et al., 2013) was used to assess PTSS corresponding to the DSM-5 criteria for posttraumatic stress disorder (APA, 2013). The PCL-5 has excellent psychometric properties (Bovin et al., 2016; Wortmann et al., 2016). The momentary version includes seven dichotomized (i.e., yes/no) items (e.g., “Did you try to avoid things that reminded you of the experience, such as specific thoughts, feelings, people, places, activities, or situations?”). These items correspond to the 7-factor Hybrid model of PTSD, which is the most empirically supported latent structure of PTSD (Armour et al., 2015). Specifically, participants are instructed to respond based on their worst trauma and indicate whether or not they have been bothered by each symptom cluster (i.e., re-experiencing, avoidance, negative affect, anhedonia, externalizing behaviors, anxious arousal, and dysphoric arousal). Items are summed, with higher scores indicative of greater PTSS. Multilevel reliabilities (Geldhof et al., 2014) in the current sample were acceptable (within-person ω = 0.62, between-person ω = 0.93).

2.5. Analytic Rationale

This study used data from seven consecutive days of surveys to capture day-level relations between emotion dysregulation and PTSS. Previous work has suggested that a span of seven days is sufficient to capture both negative and positive emotional processes (Brans et al., 2013) and PTSS (Simpson et al., 2014). Preliminary analyses were conducted using IBM SPSS Statistics Version 27.0 (IBM Corporation, 2016) and primary analyses were conducted using Mplus Version 8.0 (Muthén & Muthén, 2009). To examine the daily reciprocal relations between negative and positive emotion dysregulation and PTSS during the seven-day micro-longitudinal period, we used Dynamic Structural Equation Modeling (DSEM; Asparouhov et al., 2018), a cutting-edge statistical approach that combines time-series modeling with multilevel structural equation modeling that is ideal for micro-longitudinal data. Compared to traditional multilevel approaches, DSEM allows for the modeling of random effects for multiple outcome variables and examines multivariate time-lagged associations (Zhou et al., 2019). Moreover, DSEM does not rely on listwise deletion, and as such, can yield valid results based on all available data without the removal of cases, assuming data are missing at random (Asparouhov et al., 2018). Missing data are sampled from their conditional posterior, which accounts for the autocorrelation structure of an individual’s data (Hamaker et al., 2018). We utilized Monte Carlo Markov Chain (MCMC) Bayesian estimation with two chains, a maximum of 50,000 iterations, and thinning of 10 to ensure stable model convergence, as determined by the potential scale reduction convergence criteria and inspection of trace plots for irregularities (Asparouhov & Muthén, 2010).

A key feature of DSEM is that it decomposes observed scores into separate within- and between-individual components. This method allows for the estimation of within-person relations among emotion dysregulation and PTSS on a daily level while considering how these effects vary by individual and accounting for the autocorrelation structure of repeated assessments within an individual. Daily reports comprise level-1 variables nested within person at level-2.

We tested two separate models. First, we examined the daily reciprocal relations between negative emotion dysregulation and PTSS during the seven-day micro-longitudinal period. Next, we examined the daily reciprocal relations between positive emotion dysregulation and PTSS during the seven-day micro-longitudinal period. We used 95% Bayesian credible intervals (BCI) for describing estimates of reciprocal relations. Of note, gender and trauma load (i.e., number of traumas experienced obtained via the LEC-5) were not associated (ps > .35) with any variables of interest when included as a covariate, and thus, for the sake of parsimony, we report results without gender or trauma load in the analyses. All daily measures were centered at each individual’s latent mean level.

3. Results

3.1. Preliminary Analyses

There was a compliance rate of 85.2% for participants completing daily surveys. Approximately half of the sample (n = 67; 55.4%) met criteria for probable posttraumatic stress disorder using a cutoff total score of 33 on the PCL-5 (Bovin et al., 2016). Descriptive information and zero-order correlations for the primary study variables are detailed in Table 2.

Table 2.

Descriptive Statistics and Zero Order Correlations for Primary Study Variables

Variable	Mean	Standard Deviation	Range	Skewness	Kurtosis	1
1. Negative Emotion Dysregulation	10.96	4.87	6–24	0.74	−0.55	–
2. Positive Emotion Dysregulation	9.77	4.12	6–22	1.16	0.42	.64
3. Posttraumatic Stress Symptoms	1.45	1.95	0–7	1.33	0.78	.46

Note. All zero-order correlations are significant at p < .001.

3.2. Reciprocal Relations for Negative Emotion Dysregulation and PTSS

There were four fixed effects in the model to examine the overall associations between negative emotion dysregulation and PTSS over time. The within-individual standardized effects averaged across all the individuals in the sample are presented in Table 3. Elevated negative emotion dysregulation was significantly associated with increased next-day negative emotion dysregulation. Similarly, elevated PTSS was significantly associated with increased next-day PTSS. There was a significant positive relation between negative emotion dysregulation and next-day PTSS. The relation between PTSS and next-day negative emotion dysregulation was not significant.

Table 3.

Negative Emotion Dysregulation Model: Within-Individual Standardized Effects

Effect	Estimate	PSD	BCI
${\underset{\_}{\phi }}_{\mathrm{nn}}$	0.27	0.06	(0.15, 0.38)
${\underset{\_}{\phi }}_{\mathrm{tt}}$	0.33	0.05	(0.23, 0.32)
${\underset{\_}{\phi }}_{\mathrm{nt}}$	0.12	0.06	(−0.01, 0.24)
${\underset{\_}{\phi }}_{\mathrm{tn}}$	0.09	0.05	(0.001, 0.19)

Note. ${\underset{\_}{\phi }}_{\mathrm{nn}}$ = average negative emotion dysregulation autoregression; ${\underset{\_}{\phi }}_{\mathrm{tt}}$ = average posttraumatic stress symptoms autoregression; ${\underset{\_}{\phi }}_{\mathrm{nt}}$ = average cross-lagged effect, posttraumatic stress symptoms predicting negative emotion dysregulation; ${\underset{\_}{\phi }}_{\mathrm{tn}}$ = average cross-lagged effect, negative emotion dysregulation predicting posttraumatic stress symptoms. PSD = the standard deviation of the posterior distribution. Bold face indicates significant effects based on 95% Bayesian credible intervals (BCI).

The DSEM treated the mean levels of and the time-lagged effects between emotion dysregulation and PTSS as random effects that differ across individuals. There were six random effects in the model. The between-individual variances and correlations are presented in Table 4. When examining between-individual variance, there was significant and substantial variability for all effects of interest, though limited variability for the cross-lagged effect in which negative emotion dysregulation was positively associated with next-day PTSS. The person-level mean values for negative emotion dysregulation (${\mathrm{\mu }}_{\mathrm{n}}$) and PTSS (${\mathrm{\mu }}_{\mathrm{t}}$) were significantly positively correlated. The autoregressive effect for negative emotion dysregulation (${\underset{\_}{\phi }}_{\mathrm{nn}}$) was significantly positively associated with the cross-lagged effect for negative emotion dysregulation predicting next-day PTSS (${\underset{\_}{\phi }}_{\text{tn}}$), and negatively associated with the person-level mean value for negative emotion dysregulation (${\mathrm{\mu }}_{\mathrm{n}}$). The autoregressive effect for PTSS (${\underset{\_}{\phi }}_{\text{tt}}$) was significantly negatively associated with the cross-lagged effect for negative emotion dysregulation predicting next-day PTSS (${\underset{\_}{\phi }}_{\text{tn}}$). The cross-lagged effect for PTSS predicting next-day negative emotion dysregulation (${\underset{\_}{\phi }}_{\text{nt}}$) was significantly negatively associated with the person-level mean value for PTSS (${\mathrm{\mu }}_{\mathrm{t}}$), whereas the cross-lagged effect for negative emotion dysregulation predicting next-day PTSS (${\underset{\_}{\phi }}_{\text{tn}}$) was significantly negatively associated with the person-level mean value for negative emotion dysregulation (${\mathrm{\mu }}_{\mathrm{n}}$).

Table 4.

Negative Emotion Dysregulation Model: Variances and Intercorrelations for Random Effects

Effect	Variance (BCI)	1	2	3	4	5	6
1. ${\phi }_{\mathrm{nn}}$	0.19 (0.11, 0.30)	–
2. ${\phi }_{\mathit{tt}}$	0.20 (0.12, 0.30)	−0.38	–
3. ${\phi }_{\mathrm{nt}}$	0.65 (0.24, 1.36)	−0.26	0.47	–
4. ${\phi }_{\mathrm{tn}}$	0.04 (0.02, 0.06)	0.54	−0.44	0.01	–
5. ${\mathrm{\mu }}_{\mathrm{n}}$	19.26 (12.55, 30.59)	−0.62	0.42	−0.04	−0.43	–
6. ${\mathrm{\mu }}_{\mathrm{t}}$	1.39 (0.82, 2.38)	−0.32	−0.36	−0.48	0.07	0.46	–

Note. ${\phi }_{\mathrm{nn}}$ = person-level negative emotion dysregulation autoregression; ${\phi }_{\mathrm{tt}}$= person-level posttraumatic stress symptoms autoregression; ${\phi }_{\mathrm{nt}}$= person-level cross-lagged effect, posttraumatic stress symptoms predicting negative emotion dysregulation; ${\phi }_{\mathrm{tn}}$= person-level cross-lagged effect, negative emotion dysregulation predicting posttraumatic stress symptoms; ${\mathrm{\mu }}_{\mathrm{e}}$ = person-level negative emotion dysregulation mean (random intercept); ${\mathrm{\mu }}_{\mathrm{t}}$ = person-level posttraumatic stress symptoms mean (random intercept). Bold face indicates significant effects based on 95% Bayesian credible intervals (BCI).

3.3. Reciprocal Relations for Positive Emotion Dysregulation and PTSS

There were four fixed effects in the model to examine the overall associations between positive emotion dysregulation and PTSS over time. The within-individual standardized effects averaged across all the individuals in the sample are presented in Table 5. Elevated positive emotion dysregulation was significantly associated with increased next-day positive emotion dysregulation. Similarly, elevated PTSS was significantly associated with increased next-day PTSS. There was a significant positive association between positive emotion dysregulation and next-day PTSS. However, there was no significant association between PTSS and next-day positive emotion dysregulation.

Table 5.

Positive Emotion Dysregulation Model: Within-Individual Standardized Effects

Effect	β	PSD	BCI
${\underset{\_}{\phi }}_{\mathrm{pp}}$	0.34	0.06	(0.23, 0.47)
${\underset{\_}{\phi }}_{\mathrm{tt}}$	0.25	0.05	(0.15, 0.34)
${\underset{\_}{\phi }}_{\mathrm{pt}}$	0.09	0.06	(−0.03, 0.20)
${\underset{\_}{\phi }}_{\mathrm{tp}}$	0.20	0.05	(0.10, 0.30)

Note. ${\underset{\_}{\phi }}_{\mathrm{pp}}$ = average positive emotion dysregulation autoregression; ${\underset{\_}{\phi }}_{\mathrm{tt}}$ = average posttraumatic stress symptoms autoregression; ${\underset{\_}{\phi }}_{\mathrm{pt}}$ = average cross-lagged effect, posttraumatic stress symptoms predicting positive emotion dysregulation; ${\underset{\_}{\phi }}_{\mathrm{tp}}$ = average cross-lagged effect, positive emotion dysregulation predicting posttraumatic stress symptoms. PSD = the standard deviation of the posterior distribution. Bold face indicates significant effects based on 95% Bayesian credible intervals (BCI).

There were six random effects in the model to examine individual differences in the mean levels of and the time-lagged associations between positive emotion dysregulation and PTSS. The between-individual variances and correlations are presented in Table 6. When examining between-individual variance, there was significant and substantial variability for all effects of interest, though limited variability for the cross-lagged effect in which positive emotion dysregulation was positively associated with next-day PTSS. The person-level mean values for positive emotion dysregulation (${\mathrm{\mu }}_{\mathrm{p}}$) and PTSS (${\mathrm{\mu }}_{\mathrm{t}}$) were not significantly correlated. The autoregressive effect for positive emotion dysregulation (${\underset{\_}{\phi }}_{\text{pp}}$) was significantly negatively associated with the autoregressive effect for PTSS (${\underset{\_}{\phi }}_{\text{tt}}$) and the person-level mean value for positive emotion dysregulation (${\mathrm{\mu }}_{\mathrm{p}}$), as well as significantly positively associated with the cross-lagged effect for positive emotion dysregulation predicting next-day PTSS (${\underset{\_}{\phi }}_{\text{tp}}$). The autoregressive effect for PTSS (${\underset{\_}{\phi }}_{\text{tt}}$) was significantly positively associated with the cross-lagged effect for PTSS predicting next-day positive emotion dysregulation (${\underset{\_}{\phi }}_{\text{pt}}$), while negatively associated with the cross-lagged effect for positive emotion dysregulation predicting next-day PTSS (${\underset{\_}{\phi }}_{\text{tp}}$) and the person-level mean value for PTSS (${\mathrm{\mu }}_{\mathrm{t}}$). The cross-lagged effect for positive emotion dysregulation predicting next-day PTSS (${\underset{\_}{\phi }}_{\text{tp}}$) was significantly negatively associated with the person-level mean value for positive emotion dysregulation (${\mathrm{\mu }}_{\mathrm{p}}$).

Table 6.

Positive Emotion Dysregulation Model: Variances and Intercorrelations for Random Effects

Effect	Variance (BCI)	1	2	3	4	5	6
1. ${\phi }_{\mathrm{pp}}$	0.19 (0.12, 0.29)	–
2. ${\phi }_{\mathrm{tt}}$	0.20 (0.13, 0.30)	−0.52	–
3. ${\phi }_{\mathrm{pt}}$	0.14 (0.04, 0.41)	−0.29	0.50	–
4. ${\phi }_{\mathrm{tp}}$	0.08 (0.04, 0.14)	0.43	−0.39	−0.003	–
5. ${\mathrm{\mu }}_{\mathrm{p}}$	16.81 (10.95, 25.78)	−0.59	0.14	−0.14	−0.42	–
6. ${\mathrm{\mu }}_{\mathrm{t}}$	1.25 (0.63, 2.41)	0.06	−0.59	−0.45	−0.03	0.26	–

Note. ${\phi }_{\mathrm{pp}}$= person-level positive emotion dysregulation autoregression; ${\phi }_{\mathrm{tt}}$= person-level posttraumatic stress symptoms autoregression; ${\phi }_{\mathrm{pt}}$= person-level cross-lagged effect, posttraumatic stress symptoms predicting positive emotion dysregulation; ${\phi }_{\mathrm{tp}}$= person-level cross-lagged effect, positive emotion dysregulation predicting posttraumatic stress symptoms; ${\mathrm{\mu }}_{\mathrm{p}}$ = person-level positive emotion dysregulation mean (random intercept); ${\mathrm{\mu }}_{\mathrm{t}}$ = person-level posttraumatic stress symptoms mean (random intercept). Bold face indicates significant effects based on 95% Bayesian credible intervals (BCI).

4. Discussion

This study leverages innovative statistical modeling to advance our understanding of how emotion dysregulation may relate to the maintenance of PTSS in the daily lives of individuals with a history of sexual assault. This sample represents a trauma-exposed population at high risk for both emotion dysregulation (Raudales et al., 2019) and PTSS (Ehring & Quack, 2010). The present study reveals important nuances of within- and between-individual relations, with similar patterns emerging for emotion dysregulation across valence for the within-individual relations, but different patterns for the between-individual associations. Specifically, findings highlight the effects of both negative and positive emotion dysregulation on next-day PTSS. Addressing a critical limitation of prior literature that has focused almost exclusively on the dysregulation of negative emotions, these findings underscore the need to consider the roles of both negative and positive emotion dysregulation in PTSS among those with a history of sexual assault.

Contrary to our hypothesis, we did not find evidence of a reciprocal relation between emotion dysregulation and PTSS. Instead, we found that both negative and positive emotion dysregulation were associated with next-day PTSS, but that PTSS was not associated with next-day negative and positive emotion dysregulation. Notably, this study is the first to our knowledge to demonstrate the potential role of positive emotion dysregulation in the maintenance of PTSS over time. Our findings extend previous research implicating emotion dysregulation as an important risk factor for PTSS (Bardeen et al., 2013; Forbes et al., 2020). Negative emotion dysregulation may lead individuals to experience difficulties modulating intense negative distress in response to trauma reminders, thereby increasing PTSS. Positive emotion dysregulation might cause individuals to take an evaluative stance towards anxious arousal accompanying trauma reminders given the similarities in physiological experiences (e.g., racing heart) for both positive emotions and anxiety. Positive emotion dysregulation could also entail ineffective strategies for modulating positive emotions, such as dampening and diminished positive rumination (Wolkenstein et al., 2022), which can precipitate defining PTSS including emotional numbing and deficits with experiencing positive affect (APA, 2013). This response, in turn, would exacerbate PTSS. Additional research is warranted to better explicate mechanisms through which negative and positive emotion dysregulation may differentially contribute to PTSS.

Of note, we also found that negative and positive emotion dysregulation were associated with increased next-day negative and positive emotion dysregulation. Similarly, PTSS were associated with elevated next-day PTSS. These patterns align with current conceptualizations of negative and positive emotion dysregulation (Gratz & Roemer, 2004; Weiss et al., 2015) and PTSS (APA, 2013) as dispositional or chronic concerns, respectively. Furthermore, although not a primary aim of this study, results shed light on between-individual differences. There was relatively low variability (i.e., high consistency) across individuals for both positive and negative emotion dysregulation as it related to next-day PTSS. These findings call attention to the salient role that emotion dysregulation appears to play in predicting PTSS. Consistent with this notion, we found that individuals characterized by persistent negative and positive emotion dysregulation were also more likely to have emotion dysregulation that was associated with next-day PTSS. Future studies should seek to explore underlying processes in the maintenance of emotion dysregulation and PTSS over time.

This study is not without limitations. Despite using a micro-longitudinal design, which is known to enhance ecological validity and minimize recall errors, these findings are limited by self-reported emotional experiences and PTSS. Although measures were selected for their excellent psychometric properties, it may be that we captured perceived self-efficacy in emotion regulation ability rather than effective responses to their emotions (John & Eng, 2014). Additional research is needed that utilizes multimodal assessments such as wearable biosensors to monitor physiological indices of emotions in real time (Colombo et al., 2019). Furthermore, the generalizability of these findings remains unclear, particularly as it relates to more diverse samples and other trauma-exposed populations (e.g., treatment-seeking, military veterans).

Taken together, the present study extends prior cross-sectional work by offering valuable insights into the dynamic relations between both negative and positive emotion dysregulation and ensuing PTSS among those who have experienced sexual assault. Findings have implications for the development and refinement of interventions for PTSS. While a growing number of studies have demonstrated the efficacy of targeting negative emotion dysregulation to reduce PTSS (e.g., Cloitre et al., 2002), results from the current study suggest a potential benefit of also focusing on skills training to decrease positive emotion dysregulation among those who have experienced sexual assault. Additionally, if replicated, information from this study may be mobilized via micro-interventions (e.g., just-in-time adaptive interventions), which utilize novel technology to adapt to individuals’ dynamic internal/contextual states and provide the appropriate type and amount of support when an individual needs it most (Nahum-Shani et al., 2018). As such, intervening on increases in an individual’s dysregulation of negative and positive emotions in a real-time, naturalistic setting may be key for alleviating subsequent episodes of heightened PTSS.