The PTSD Symptom Presentation and the Effect of Polytrauma on PTSD Symptom Clusters Among Young People Who Have Experienced Commercial Sexual Exploitation and Trafficking

Abstract

Purpose

The purpose of this study was to describe the PTSD symptom presentation (including dissociative symptoms) of PTSD using the Diagnostic and Statistical Manual of Mental Disorders 5^th Edition diagnostic criteria and explore associations between the symptom severity for each of the four PTSD symptom clusters and polytrauma, defined as multiple exposures to different categories of potentially traumatic events.

Methods

This is a secondary analysis of cross-sectional program evaluation data among 95 young people (aged 11–19) at therapy initiation in a southeastern state in the U.S. We used descriptive statistics and multivariable linear regression to test study objectives.

Results

Eighty-one respondents (90.0%) experienced a potentially traumatic event in ≥ 2 trauma categories, in addition to experiencing CSE/T. Approximately two-thirds of respondents experienced clinically significant PTSD symptoms for each symptom cluster. Of the 31 young people who met full criteria for PTSD, 9 met criteria for the standard PTSD diagnosis, while 22 met criteria for the dissociative subtype of PTSD. On average, experiencing additional trauma categories was associated with substantively higher PTSD symptom cluster scores for each cluster.

Conclusions

These findings support the need for a comprehensive assessment of trauma symptoms that includes cluster-specific PTSD symptoms. They also underscore the need to assess the full breadth and chronicity of trauma experiences to guide treatment planning and delivery, targeting specific domains of trauma impact. These findings can also inform the tailoring and adaptation of evidence-based interventions and strategies to better meet the needs of young people who have experienced CSE/T.

The commercial sexual exploitation and trafficking (CSE/T) of children and adolescents is a severe form of child maltreatment and significant public health problem (Greenbaum, 2020; International Labor Organization, n.d.; Zimmerman & Kiss, 2017). For purposes of the current study, we will use the term, “commercial sexual exploitation and trafficking” to describe exploitative acts based on both the legal definition for sex trafficking (Trafficking Victims Protection Reauthorization Act, 2013) and the broader non-legal term, commercial sexual exploitation, collectively, as experiencing either of these forms of violence can lead to serious mental health outcomes.

Young people who have experienced CSE/T often experience a range of traumatic events and other adversities prior to and during their trafficking/exploitation, in addition to the CSE/T experience itself (Cole et al., 2016; Hickle & Roe-Sepowitz, 2018; Hopper, 2017; Kenny et al., 2019; Self-Brown et al., 2021). Recent findings suggest that, compared to those who have not experienced CSE/T or those are at risk, adolescents who experience CSE/T are more likely to have experienced child sexual abuse, dating violence, physical abuse, gang affiliation, having run away from home, being kidnapped, and child protective services-involvement (Hickle & Roe-Sepowitz, 2018; Kenny et al., 2019; Varma et al., 2015).

Research has documented high rates of severe psychiatric diagnoses, such as posttraumatic stress disorder (PTSD) among young people who have experienced CSE/T (Cole et al., 2016; Hossain et al., 2010). Other scholars have documented posttraumatic stress (PTS) symptoms (i.e., symptoms of PTSD that may not meet diagnostic criteria), posttraumatic cognitions, depression, anxiety, disruptive/impulse control and conduct disorder, emotional distress, suicidal ideation and self-harm, substance use disorder, adjustment disorder, and comorbid mental health diagnoses among young people who have experienced CSE/T (Basson & Fernando, 2012; Chapple & Crawford, 2019; Frey et al., 2019; Goldberg et al., 2017; Hickle & Roe-Sepowitz, 2018; Hopper & Gonzalez, 2018; Hossain et al., 2010; Kenny et al., 2019; Kiss et al., 2015; Lederer & Wetzel, 2014; Perry et al., 2022; Rimal & Papadopoulos, 2016; Self-Brown et al., 2021).

Exposure to cumulative potentially traumatic events, in addition to the CSE/T experience, appear related to increased risk of clinically significant mental health outcomes among this population. Prior findings among help-seeking young people aged 11–19 with CSE/T experiences have documented associations between experiencing more categories of potentially traumatic events and PTSD diagnosis, clinically significant emotional distress (Self-Brown et al., 2021), and clinically significant posttraumatic cognitions (Perry et al., 2022). The CSE/T experience itself may lead to a greater severity of trauma symptomatology more severe than that experienced by trauma-exposed young people who had not experienced CSE/T (Chapple & Crawford, 2019; Cole et al., 2016). For example, a longitudinal study that followed juvenile justice-involved adolescents aged 14–16 (a population with high rates of trauma exposure; Dierkhising et al., 2013; Ford et al., 2013) found that, compared to those who had not experienced CSE/T, those who had were at higher odds of being diagnosed with depression, obsessive compulsive disorder, hostility, paranoia, and psychoticism (Chapple & Crawford, 2019).

Posttraumatic Stress Disorder, PTSD Symptom Clusters, and Dissociative Symptoms

In 2013, the diagnostic criteria for PTSD were updated in the Diagnostic and Statistical Manual 5^th Edition (DSM-5) to broaden the types of events that were considered a potentially traumatic event and expand and reorganize the symptom clusters (American Psychiatric Association, 2000, 2013; Friedman, 2013). Two PTSD subtypes were also added: dissociative and preschool subtype (for children 6 years and younger). The PTSD diagnostic criteria are: 1) the presence of a traumatic event(s), 2) the presence of all four symptom clusters (i.e., intrusion; avoidance; negative alterations in cognitions and mood; and arousal or reactivity), 3) duration of the symptoms lasting longer than one month, 4) the presence of distress or impairment, and 5) the disturbance must not be attributable to a substance or other medical condition. The DSM-5 also allows clinicians to specify if dissociative symptoms (e.g., persistent, or recurrent experiences of feeling detached from one’s body or mental processes, or experiencing surroundings as if they’re unreal, dreamlike, distant, or distorted) are present. If a person meets full diagnostic criteria for PTSD and experiences at least one dissociative symptom, they are considered to have met criteria for the dissociative subtype of PTSD (D-PTSD).

Associations between trauma exposure and dissociative symptoms and PTSD are well documented (Jowett et al., 2021; Vonderlin et al., 2018). In prior research exploring the DSM-5 PTSD criteria, a sample of juvenile justice-involved young people aged 12–19 years who experienced polyvictimization, defined as multiple types of interpersonal and non-interpersonal trauma, had higher PTSD symptom cluster scores than young people who experienced mixed adversities (e.g., having a parent arrested, severe injury or illness) or lived in violent environments only (e.g., living in a war zone, natural disasters, witnessing physical violence; Charak et al., 2019). In a recent meta-analysis including 65 studies, dissociative symptoms were greater among those who experienced abuse or neglect under the age of 18 years (Vonderlin et al., 2018). The authors also found across studies that higher dissociation scores were associated with experiencing physical or sexual abuse, earlier age of onset of abuse, longer duration of abuse, and parental abuse (Vonderlin et al., 2018). Other findings suggest that dissociative symptoms are common among adults diagnosed with PTSD (Foote et al., 2008; Lyssenko et al., 2018) and those who met diagnostic criteria for a dissociative disorder were more likely to be diagnosed with major depressive disorder, drug abuse or dependence and have a history of suicidal ideation, self-injurious behavior, and multiple suicide attempts than those who were not (Foote et al., 2008). Further, those with D-PTSD experience greater symptom severity, chronicity, functional impairment, and suicidality compared to individuals with PTSD but no dissociative symptoms (Stein et al., 2014).

PTSD Symptom Presentation and Commercial Sexual Exploitation and Trafficking

It is well established that young people who experience commercial sexual exploitation are at high risk of experiencing PTSD (Cole et al., 2016; Hossain et al., 2010). Further, research has also documented associations between various types of trauma and dissociative symptoms (Jowett et al., 2021; Vonderlin et al., 2018). Yet, few studies have described symptom patterns for PTSD, and to our knowledge, none have done so using the DSM-5 diagnostic criteria for PTSD and D-PTSD among young people who have experienced CSE/T. Examining PTSD symptom presentation within and across clusters among a unique and highly traumatized population may provide insight into patterns of trauma symptom presentation, which could inform the tailoring and adaptation of evidence-based interventions and strategies to better meet the needs of youth who have experienced CSE/T.

A few emerging studies have begun to explore PTSD symptom clusters for CSE/T youth. For instance, a study of female adolescents (aged 12 to 18) found that participants who had experienced CSE/T reported greater symptoms of anxious arousal, intrusive memories (i.e., nightmares and flashbacks), attempts to avoid intrusive thoughts or memories, and dissociation compared to participants who had not (Lanctôt et al., 2020). However, it appears that PTSD symptoms were based on DSM-IV diagnostic criteria in this study as alterations in cognitions and mood were not compared. Cole et al. (2016) found that young people (aged 10–20 years) who experienced commercial sexual exploitation had higher rates of avoidance, hyperarousal, and dissociation compared to those who only experienced sexual abuse or assault (Cole et al., 2016). However, this study compared also used DSM-IV diagnostic criteria. Perry et al. (2022), using data from the same sample as the current study, found that more than half of young people who experienced CSE/T also experienced clinically significant post-traumatic cognitions, which were associated with experiencing a greater number of traumatic events and greater PTS symptom severity. However, they only reported post-traumatic cognitions and PTS symptoms as a summary score (Perry et al., 2022).

Purpose

There is a dearth of literature documenting cluster-specific PTSD symptoms described in the DSM-5 and the dissociative subtype of PTSD among young people who have experienced CSE/T. Given this gap in the literature and the urgent clinical importance of addressing trauma impacts among this highly traumatized population, the purpose of this study was to 1) describe the PTSD symptom presentation (including dissociative symptoms), using DSM-5 diagnostic criteria, among young people aged 11–19 years who have experienced CSE/T and 2) explore the associations between the PTSD symptom clusters and cumulative trauma history. To address our first aim of describing the PTSD symptom presentation among this population, we determined what percent of young people in our sample a) met criteria for each DSM-5 diagnostic criteria (i.e., presence of a potentially traumatic event, the four symptom clusters, symptom duration, clinically significant distress and impairment, dissociative symptoms) and b) met full criteria for the standard PTSD diagnosis and the dissociative subtype of PTSD. Our second aim was to explore how polytrauma, defined as experiencing ≥ 2 different categories of potentially traumatic events, is associated with greater symptom severity for each PTSD symptom cluster (i.e., intrusion, hyperarousal, avoidance, and negative mood and cognitions). Findings from this study can guide clinicians in their clinical assessment and therapeutic treatment planning for their clients who have experienced CSE/T. This is especially relevant for clinicians who are seeing clients who have experienced multiple potentially traumatic events in addition to the CSE/T experience.

Methods

Participants and Setting

Data were derived from baseline program evaluation data from Project Intersect collected from August 1, 2013, through March 30, 2020. Project Intersect (described elsewhere; Kinnish et al., 2020; Perry et al., 2022; Self-Brown et al., 2021) is a federally funded service grant in a southeastern state in the U.S., the goal of which is to improve the mental health service response with young people (aged 11–19) who experience CSE/T, by training therapists (licensed or supervised and seeking licensure) across the state in CSE/T applications of Trauma-Focused Cognitive Behavioral Therapy (TF-CBT). English-speaking young people aged 11–19 years who were referred to a Project Intersect therapist had the option to participate in the program evaluation, administered at three time points (intake, mid-point, completion) and received a $20 gift card for completing each survey. The DSM-5 criteria for PTSD (American Psychiatric Association, 2013) was updated during the course of the funding period; thus, we restricted the sample for the current study to include surveys completed from October 1, 2014 through March 31, 2020 (n = 95) to include only those who completed measures reflecting these changes. Informed consent was not warranted, given data was collected in the context of clinical services for the purposes of program evaluation. Approval to conduct this secondary data analysis was obtained by the Institutional Review Board at the affiliated university.

Measures

Outcome Variables

PTSD Symptom Clusters

PTSD symptom scores for each symptom cluster (i.e., Cluster B-intrusion symptoms, Cluster C-avoidance symptoms, Cluster D-negative alterations in cognitions and mood, Cluster E-arousal and reactivity) were measured using the University of California Los Angeles PTSD-Reaction Index (UCLA PTSD-RI; Steinberg et al., 2004) based on the DSM-5 criteria (American Psychiatric Association, 2013). The UCLA PTSD-RI is a validated scale and is one of the most widely used tools to measure PTSD among children and young people (Kaplow et al., 2020; Steinberg et al., 2004). For each of the 27 items measuring PTSD symptoms, participants were asked to rate how many days during the last month they had experienced symptoms, for example, “I have trouble feeling happiness or love.” Response options were none (0), little (1), some (2), much (3), and most (4). For all items except three, the symptom was deemed present with a score of 3 or 4. For the three others, a score of 2 or more indicated symptom presence. Intrusion symptoms were measured using 5 items (possible range 0 to 20) and the criterion was considered met if ≥ 1 items were endorsed above the threshold score (Cronbach’s α=0.93). Avoidance symptoms were measured using 2 items (possible range 0 to 8) and the criterion was considered met if ≥ 1 items were endorsed above the threshold score (Cronbach’s α=0.53). Negative alterations in cognitions or mood were measured using 13 items (possible range 0 to 28) and the criterion was considered met if ≥ 2 items were endorsed above the threshold score. Marked alterations in arousal and reactivity were measured using 7 items (possible range 0 to 24) and the criterion was considered met if ≥ 2 items were endorsed above the threshold score (Cronbach’s α=0.83).

Independent Variables

Polytrauma

To capture the breadth of victimization among this population, we used 13 of the 14 items from the trauma history screen of the UCLA PTSD-RI (Steinberg et al., 2004; see Table 1) and excluded one item that assessed CSE/T (as participants had substantiated experiences of CSE/T) to create two polytrauma indicators. From the 13 items included in the current analysis, we identified 7 categories (outlined in Table 1) based on prior literature (de Haan et al., 2020; Perry et al., 2022; Self-Brown et al., 2021). We included direct physical and sexual violence in the direct violence category and collapsed all items related to seeing a violent event into the witnessing violence category. We used a thematic analysis to code the one open-ended item and the theme of loss/separation emerged, so we included responses to this item in the loss/separation category. We defined polytrauma as experiencing a potentially traumatic event in two or more of 7 trauma categories. Using these 7 categories, we created a polytrauma summary score, based on our operational definition of polytrauma, to indicate the level of polytrauma that a young person experienced; a score of 0 indicated that the young person reported experiencing trauma in ≤ 1 category, a score of 1 indicated that the respondent reported experiencing trauma in two categories, a score of 2 indicated 3 categories were experienced, and so on (possible range, 0–7). We also created a dichotomous polytrauma variable; a young person was considered to have experienced polytrauma if they had a polytrauma summary score of ≥ 1.

Table 1.

Reported trauma among young people aged 11–19 years who experienced commercial sexual exploitation and trafficking (CSE/T) and participated in an evaluation of Project Intersect in a state in the southeastern United States (n = 95); data collected from October 1, 2014, through March 31, 2020

Trauma category	Item	No. (%) of respondents who answered yesa
Natural disaster	Were you in a disaster, like an earthquake, wildfire, hurricane, tornado, or flood?	16 (17.78)
Accident	Were you in a bad accident, like a serious car accident or fall?	30 (33.33)
War	Were you in a place where a war was going on around you?	10 (11.11)
Direct violence	Were you hit, punched, or kicked very hard at home? (DO NOT INCLUDE play fighting between brothers or sisters.)	40 (44.44)
	Were you beaten up, shot at, or threatened to be hurt badly in your school, neighborhood, or town?	48 (53.33)
	Did someone touch your private parts when you did not want them to? (DO NOT INCLUDE visits to the doctor.)	62 (68.89)
	Were you ever forced to have sex with someone against your will?	60 (66.67)
Witnessing violence	Did you see a family member hit, punched, or kicked very hard at home? (DO NOT INCLUDE play fighting between brothers or sisters.)b	43 (48.31)
	Did you see someone who was beaten up, shot at, or killed?	69 (76.67)
	Did you see a dead body (do not include funerals)?	28 (31.11)
	Did you see or hear about the violent death or serious injury of a loved one or friend?	64 (71.11)
Medical trauma	Did you have a painful or scary medical treatment when you were very sick or badly injured?	16 (17.78)
Loss/separation	Has anyone close to you died?	78 (86.67)
	Other than the scary things described above, has ANYTHING ELSE ever happened to you that was REALLY SCARY OR UPSETTING?c, d	34 (39.53)

Trauma categories and items were measured using the University of California, Los Angeles Post-Traumatic Stress Disorder–Reaction Index Trauma Screen (Steinberg et al., 2004)

^a Not all participants answered all questions. The denominator is 90 for all items above unless otherwise indicated

^b n = 89 participants who responded to this item; n = 6 missing

^c A review of the open-ended responses to this item was conducted, and the primary theme that emerged was separation from loved ones

^d n = 86 total participants who responded to this item; n = 9 missing

Control and auxiliary variables

Demographic characteristics included age, gender (female, male, transgender), ethnicity (Hispanic or Latinx), race, and living situation (Table 2). Hispanic or Latinx participants could further specify which Hispanic/Latinx ethnic group(s) they identified with (Central American, Cuban, Dominican, Mexican, Puerto Rican, South American, and Other); however, due to small cell counts, we only reported these descriptively. Response options were determined by the grant funder and included Black or African American, Asian, Native Hawaiian or Other Pacific Islander, American Indian, White, or Other and we acknowledge the challenges with measuring ethnicity and race using these limited categories (Ford & Harawa, 2010). Based on the distribution of the data, we created a race variable for use in the regression models with the following categories: Black or African American, White, or Other Race. Other Race included Asian, Native Hawaiian or Other Pacific Islander, American Indian, more than one racial category, or the Other category. We collapsed living situation in the past 30 days into five categories (from 12 total) based on the distribution of the data. A dummy variable for living situation (with a parent/caregiver or other setting) was created for use in the regression models.

Table 2.

Demographic characteristics of young people aged 11–19 years (n = 95) who have experienced commercial sexual exploitation and trafficking (CSE/T) and participated in an evaluation of Project Intersect in a state in the southeastern United States; data collected from October 1, 2014, through March 31, 2020

Variable	Total Samplea
Gender	90 (94.74)
Female	90 (94.74)
Male	4 (4.21)
Transgender	1 (1.05)
Total no. of respondents	95
Race
Black	46 (54.12)
White	21 (24.71)
Other raceb	18 (21.18)
Total no. of respondents	85
Ethnicity
Hispanic/Latinx	11 (12.09)
Non-Hispanic/Latinx	80 (87.91)
Total no. of respondents	91
Specific Hispanic/Latinx ethnic group
Central American	0 (0.00)
Cuban	0 (0.00)
Dominican	2 (22.22)
Mexican	1 (1.11)
Puerto Rican	4 (4.44)
South American	1 (1.11)
Other	1 (1.11)
Total no. of respondents	9
Age, Mean (SD), y	15.72 (1.47)
Total no. of respondents	88
Living Situation
With a caregiver/parent	28 (29.47)
Alone, with friends, or at a homeless shelter	10 (10.53)
In a group home or other non-juvenile detention setting	31 (32.63)
Foster home or temporary placement	7 (7.37)
Other setting (including DJJ)	19 (20.00)
Total no. of respondents	95
Total trauma summary score, Mean (SD)	6.64 (3.26)
Total no. of respondents	90
Polytrauma
Summary Score Mean (SD)c	3.39 (1.45)
Experienced Polytrauma (yes)	81 (90.00%)
Total no. of respondents	90
Intrusion symptoms (Cluster B)
Summary Score, Mean (SD)d	8.84 (6.81)
Total no. of respondents	79
Symptoms present (yes)	52 (61.90)
Total no. of respondents	84
Avoidance symptoms (Cluster C)
Summary Score, Mean (SD)e	4.24 (2.45)
Total no. of respondents	83
Symptoms present (yes)	55 (63.95)
Total no. of respondents	86
Negative alterations in mood and cognitions (Cluster D)
Summary Score, Mean (SD)f	13.44 (7.32)
Total no. of respondents	82
Symptoms present (yes)	56 (64.37)
Total no. of respondents	87
Arousal and reactivity symptoms (Cluster E)
Summary Score, Mean (SD)g	10.87 (6.69)
Total no. of respondents	83
Symptoms present (yes)	54 (62.07)
Total no. of respondents	87
Distress or impairment present
Yes	70 (83.33)
No	14 (16.67)
Total no. of respondents	84
Dissociative symptoms present
Yes	32 (38.10)
No	52 (61.90)
Total no. of respondents	84
PTSD diagnosis with dissociative symptoms
Yes	22 (26.83)
No	60 (73.17)
Total no. of respondents	82
PTSD diagnosis without dissociative symptoms
Yes	9 (10.98)
No	73 (89.02)
Total no. of respondents	82

DJJ Department of Juvenile Justice, PTSD Post Traumatic Stress Disorder. Posttraumatic Stress Disorder symptoms, other PTSD diagnostic criteria, and polytrauma were measured using the University of California, Los Angeles Post-Traumatic Stress Disorder–Reaction Index (Steinberg et al., 2004)

^a All values are number (percentage) unless otherwise indicated

^b Includes Asian, Native Hawaiian/Other Pacific Islander, American Indian, other, or > 1 race (i.e., multiracial)

^c Observed range: 0–6 trauma category, maximum possible polytrauma score = 7. A polytrauma score of 0 indicates that the young person reported experiencing trauma in only 1 category, a score of 1 indicates that the young person reported experiencing trauma in 2 categories, a score of 2 indicates that the young person reported experiencing trauma in 3 categories, and so on

^d Observed range for Criterion B summary score = 0–20

^e Observed range for Criterion C summary score = 0–8

^f Observed range for Criterion D summary score = 0–26

^g Observed range for Criterion E summary score = 0–24

Items from the Child Post-Traumatic Cognitions Inventory (CPTCI) were used as auxiliary variables in the imputation models (see description below). The CPTCI is a validated, 33-item self-report questionnaire used to measure maladaptive cognitions or appraisals among trauma-exposed children and adolescents aged 6–18 years (Lobo et al., 2015; McKinnon et al., 2016; Meiser-Stedman et al., 2009).

Other PTSD Symptom Presentation Variables

Clinically significant distress, symptom duration, dissociative symptoms, and PTSD without dissociative symptoms, and D-PTSD were measured using the UCLA PTSD-RI and were calculated using the relevant scoring described in the UCLA PTSD-RI scoring manual (Steinberg et al., 2004).

Clinically Significant Distress and Symptom Duration

Clinically significant distress in life domains including family, school, and friendships was measured using 8-items. Respondents could answer (yes or no) to items including, “do these thoughts and feelings get you in trouble at home?” If a young person endorsed one or more items, they were experiencing clinically significant distress or impairment. Symptom duration was measured using the item “how long have you had these symptoms?” Response options included < one month and ≥ one month.

Dissociative Symptoms and D-PTSD

Four items were used to measure dissociative symptoms (e.g., “I feel not connected to my body, like I’m not really there inside.”). Response options included none (0), little (1), some (2), much (3), or most (4); a score of 3 or 4 indicated that that symptom was present. Dissociative symptoms were present if ≥ 1 item was endorsed. If PTSD diagnostic criteria were met and the respondent was experiencing dissociative symptoms, they were considered to have D-PTSD.

Data Analysis

We removed one observation due to missing data on all items in the survey (n = 95). We used descriptive statistics to address our first aim of describing the PTSD symptom presentation using DSM-5 criteria among young people who have experienced CSE/T. To address our second aim of exploring the associations between PTSD symptom clusters and polytrauma, we conducted four separate multivariable linear regression models with the summary score for each PTSD symptom cluster (intrusion, avoidance, negative alterations in cognitions and mood, arousal and reactivity, respectively) as the dependent variable. Each regression model included the continuous polytrauma summary score as the primary independent variable of interest, with age, ethnicity, race, and living situation added to the model as control variables.

Based on the results of the omnibus test, initial fit for all four models was poor. We addressed poor model fit in each of the four linear regression models by removing 3 atypical, influential observations (based on Cook’s d > 4/N). The 3 observations were 1) high trauma with no PTSD symptoms, 2) no trauma with high PTSD symptoms, and 3) high trauma with very low PTSD symptoms. Removing these influential observations resulted in good model fit for all four linear regression models.

We used multiple imputation to address missing data using PROC MI and PROC MIANALYZE and data were assumed to be missing at random. To determine the number of imputations, we used the largest Fraction of Missing Information (FMI) from a pilot phase imputation (0.46 for intrusion symptom summary; von Hippel, 2020) to calculate the final number of imputations for use in imputation phase. We determined the final number of imputations (n = 46) based on White et al. (2011), such that FMI divided by the number of imputations was equal to 0.01 (i.e., 1% loss in precision; White et al., 2011).

The intrusion symptom summary had the greatest number of observations missing (n = 16, 16.8%), followed by the summary of negative cognitions and mood symptoms (n = 13, 13.7%), the symptom summaries for avoidance and arousal (n = 12, 12.6%, respectively), and the two dummy variables for race (Black or African American and Other Race; n = 10, 10.5%, respectively). We used auxiliary variables in the imputation models to improve the precision of our estimates (Dong & Peng, 2013; Hardt et al., 2012). We selected items from the CPTCI for consideration for auxiliary variables based on prior literature reporting associations between posttraumatic cognitions and PTS symptoms (Liu & Chen, 2015; Lobo et al., 2015; Nixon et al., 2010; Perry et al., 2022; Ponnamperuma & Nicolson, 2016), and a correlation greater than |.4| and statistically significant t-tests results (α = 0.05) with variables with the greatest amount of missing data (i.e., symptom B, C, D, and E clusters; Dong & Peng, 2013; Enders, 2010). We included four auxiliary variables from the CPTCI based on the criteria above. All analyses were conducted in SAS 9.4 (SAS Institute, Inc), and results were considered statistically significant at the α = 0.05 level.

Results

Most of the young people (Mean_age = 15.7, Standard Deviation_age = 1.47) participating in Project Intersect program evaluation identified as female (94.7%, n = 90) and Black or African American (54.1%, n = 46; Table 2). Nearly all (90.0%; 81 of 90) young people experienced ≥ 2 of 7 categories of potentially traumatic experiences, in addition to CSE/T. Close to two-thirds of the sample reported experiencing each symptom cluster: intrusion symptoms, avoidance symptoms, altered cognitions, and arousal and reactivity in the past 30 days related to the same traumatic event. Interestingly, the number of young people in our sample that met criteria for D-PTSD (22 of 82, 26.8%) was more than double the number that were diagnosed with PTSD without dissociative symptoms (9 of 82, 10.9%).

Polytrauma and PTSD Symptom Clusters

The four adjusted linear regression models indicated that, on average, polytrauma was statistically significantly associated with the summaries for each of the four PTSD symptom clusters (i.e., intrusion, avoidance, negative alterations in cognitions and mood, arousal and reactivity; Tables 3, 4, 5 and 6, respectively), controlling for age, race, ethnicity, and living situation. Regarding intrusion symptoms (Table 3), on average, each additional trauma category experienced corresponded to a 2.58-point difference in the intrusion symptom summary (95% Confidence Interval [CI] [1.59, 3.58; p < 0.0001]), controlling for the other variables in the model. Increased age was also independently associated with the intrusion B summary (β = 1.35, 95% CI [0.37, 2.32]; p = 0.01). Statistically significant associations between the intrusion symptom summary score and race, ethnicity, and living situation did not emerge.

Table 3.

Adjusted linear regression model examining polytrauma and the intrusion symptom summary among young people (aged 11–19 years) participating in Project Intersect program evaluation in a state in the Southeastern U.S., collected from October 1, 2014 to March 31, 2020 (n = 92)^a

Variable	Unstandardized parameter estimate (95% confidence interval)	se	p-value
Polytrauma	2.58 [1.59, 3.58]	0.51	< .0001
Race
White	Ref
Black	-0.27 [-3.42, 2.87]	1.60	.865
Other race	-3.20 [-7.35, 0.95]	2.11	.130
Ethnicity
Non-Hispanic/Latinx	Ref
Hispanic/Latinx	3.21 [-1.16, 7.59]	2.23	.150
Living Situation
Other setting	Ref
With a caregiver/parent	2.15 [-0.74, 5.04]	1.47	.145
Age	1.35 [0.37, 2.32]	0.50	.007

se standard error, Ref reference category. Intrusion symptoms and polytrauma were measured using the University of California, Los Angeles Post-Traumatic Stress Disorder–Reaction Index (Steinberg et al., 2004)

^a The final adjusted model included polytrauma, race, ethnicity, living situation, and age. Analysis consisted of 92 participants. Multiple imputation used (n = 46). P values were considered significant at the α = .05 level

Table 4.

Adjusted linear regression model examining polytrauma and the avoidance symptom summary among young people (aged 11–19 years) participating in Project Intersect program evaluation in a state in the Southeastern U.S., collected from October 1, 2014 to March 31, 2020 (n = 92)^a

Variable	Unstandardized parameter estimate (95% confidence interval)	se	p-value
Polytrauma	0.85 [0.47, 1.23]	0.20	< .0001
Race
White	Ref
Black	0.17 [-1.02, 1.36]	0.61	.782
Other race	-0.81 [-2.31, 0.70]	0.77	.294
Ethnicity
Non-Hispanic/Latinx	Ref
Hispanic/Latinx	0.64 [-0.96, 2.24]	0.82	.432
Living Situation
Other setting	Ref
With a caregiver/parent	0.69 [-0.38, 1.76]	0.55	.208
Age	0.31 [-0.07, 0.69]	0.19	.106

se standard error, Ref reference category. Avoidance symptoms and polytrauma were measured using the University of California, Los Angeles Post-Traumatic Stress Disorder–Reaction Index (Steinberg et al., 2004)

^a The final adjusted model included polytrauma, race, ethnicity, living situation, and age. Analysis consisted of 92 participants. Multiple imputation used (n = 46). P values were considered significant at the α = .05 level

Table 5.

Adjusted linear regression model examining polytrauma and the summary for negative alternations in cognitions and mood among young people (aged 11–19 years) participating in Project Intersect program evaluation in a state in the Southeastern U.S., collected from October 1, 2014 to March 31, 2020 (n = 92)^a

Variable	Unstandardized parameter estimate (95% confidence interval)	se	p-value
Polytrauma	2.96 [1.92, 4.01]	0.53	< .0001
Race
White	Ref
Black	-1.61 [-5.03, 1.81]	1.74	.356
Other race	-2.61 [-6.79, 1.57]	2.13	.222
Ethnicity
Non-Hispanic/Latinx	Ref
Hispanic/Latinx	2.35 [-2.17, 6.88]	2.31	.308
Living Situation
Other setting	Ref
With a caregiver/parent	0.06 [-2.86, 2.98]	1.49	.966
Age	0.38 [-0.70, 1.47]	0.55	.488

se standard error, Ref reference category. Negative alterations in cognitions and mood and polytrauma were measured using the University of California, Los Angeles Post-Traumatic Stress Disorder–Reaction Index (Steinberg et al., 2004)

^a The final adjusted model included polytrauma, race, ethnicity, living situation, and age. Analysis consisted of 92 participants. Multiple imputation used (n = 46). P values were considered significant at the α = .05 level

Table 6.

Adjusted linear regression model examining polytrauma and the arousal and reactivity symptom summary among young people (aged 11–19 years) participating in Project Intersect program evaluation in a state in the Southeastern U.S., collected from October 1, 2014 to March 31, 2020 (n = 92)^a

Variable	Unstandardized parameter estimate [95% confidence interval]	se	p-value
Polytrauma	2.57 [1.54, 3.61]	0.53	< .0001
Race
White	Ref
Black	-1.32 [-4.63, 1.99]	1.69	.435
Other race	-1.22 [-5.23, 2.79]	2.05	.550
Ethnicity
Non-Hispanic/Latinx	Ref
Hispanic/Latinx	2.48 [-1.78, 6.74]	2.17	.253
Living Situation
Other setting	Ref
With a caregiver/parent	0.18 [-2.66, 3.02]	1.45	.902
Age	0.13 [-0.85, 1.12]	0.50	.791

se standard error, Ref reference category. Arousal and reactivity symptoms and polytrauma were measured using the University of California, Los Angeles Post-Traumatic Stress Disorder–Reaction Index (Steinberg et al., 2004)

^a The final adjusted model included polytrauma, race, ethnicity, living situation, and age. Analysis consisted of 92 participants. Multiple imputation used (n = 46). P values were considered significant at the α = .05 level

After controlling for the other variables in the model, polytrauma was the only predictor that was significantly associated with the symptom summaries for avoidance (Table 4), negative alterations in cognitions and mood (Table 5), and arousal and reactivity (Table 6). Controlling for the other variables in the model, each additional trauma category experienced corresponded to an average 0.85-point difference in the avoidance symptom summary score (95% CI [0.47, 1.23]; p < 0.0001; Table 4). On average, each additional trauma category experienced corresponded to a 2.96-point difference in the summary score for negative alterations in mood and cognitions (95% CI [1.92, 4.01]; p < 0.0001; Table 5), controlling for the other variables in the model. Lastly, each additional trauma category experienced corresponded to an average 2.57-point difference in the arousal and reactivity symptom summary score (95% CI [1.54, 3.61]; p < 0.0001; Table 6), adjusting for the other variables in the model.

Discussion

Young people who experience commercial sexual exploitation and trafficking have high rates of trauma exposure (Cole et al., 2016; Hickle & Roe-Sepowitz, 2018; Hopper, 2017; Kenny et al., 2019; Perry et al., 2022; Self-Brown et al., 2021) and severe trauma symptomatology (Cole et al., 2016; Hossain et al., 2010), which may play a role in revictimization risk, broadly (Jaffe et al., 2019). Understanding a client’s trauma complete trauma history as well as their trauma symptomatology (i.e., both the severity of symptoms and prominent symptom clusters) could inform the therapeutic treatment plan for those who have experienced CSE/T, regardless of whether the client meets the criteria for a PTSD diagnosis. Learning more about the clinical presentation of specific trauma symptoms could also lead to the tailoring and adaptation of evidence-based interventions and strategies to address the needs of this uniquely traumatized population more effectively.

Consistent with prior research (Cole et al., 2016; Hickle & Roe-Sepowitz, 2018; Hopper, 2017; Kenny et al., 2019), the young people in our sample experienced multiple potentially traumatic events. Approximately two out of three young people in our sample had experienced PTSD symptoms in the past 30 days in each of the four symptom clusters. These findings support prior research that has found PTSD symptom severity across the four symptom clusters among this population (Cole et al., 2016; Lanctôt et al., 2020; Perry et al., 2022) and other child and adolescent populations that have experienced significant trauma (Charak et al., 2019). Over 8 of 10 respondents experienced clinically significant distress or impairment and nearly 40% were experiencing dissociative symptoms.

While it is not surprising that a large proportion of young people in the current study experienced dissociative symptoms (Vonderlin et al., 2018), it is a novel and clinically impactful finding that nearly 70% of those who experienced dissociative symptoms met criteria for D-PTSD, and the proportion of young people with D-PTSD in the current sample (26.8%) is slightly higher, descriptively, than what has been previously reported in another sample of CSE/T youth (20.0%; Kenny et al., 2019). Dissociative symptoms that develop in the aftermath of experiencing trauma, especially during childhood, may be a means of coping with the event (Vonderlin et al., 2018) and have been found to mediate the relationship between prior trauma, especially trauma experienced during childhood, and PTSD diagnosis (Jowett et al., 2021). Together with prior research, the findings from the current study suggest that dissociative symptoms feature prominently in the trauma symptomatology of young people who have experienced CSE/T. Future research should explore how prior trauma types, cumulative trauma, and the CSE/T experience itself differentially impact dissociative symptoms and D-PTSD among children and youth.

Experiencing a greater number of trauma categories (i.e., polytrauma) was associated with a substantive positive difference in symptom severity for each of the four PTSD symptom clusters. For example, experiencing three additional trauma categories, on average, is associated with an intrusion symptom summary score difference of 7.74 points, holding the other variables constant, which is a substantive difference, given the maximum score across the 5 items included in the intrusion symptom summary is 20 points. Similarly, experiencing an additional three trauma categories for the avoidance symptom summary is associated with a positive score difference of 2.55 out of a total score of 8. Experiencing an additional three trauma categories for the altered mood and cognitions and arousal symptom summaries corresponded to an additional 8.88 and 7.71 points out of total scores of 26 and 24, respectively. Future research is needed to compare the effects of cumulative trauma on each PTSD symptom cluster to understand if prior trauma has differential effects on one symptom cluster versus another.

Interestingly, age was the only control variable to emerge as independently associated with the intrusion symptom summary, with a one-year increase in age being associated with greater intrusion symptom severity. No associations emerged between age and summaries for the other PTSD symptom clusters (i.e., avoidance, negative alterations in mood and cognitions, and arousal and reactivity). Prior research has found inconsistent effects of age on trauma symptoms and trauma and stress disorders (Bal et al., 2005; Gunaratnam & Alisic, 2017; Lobo et al., 2015; Meiser-Stedman et al., 2019; Perry et al., 2022; Self-Brown et al., 2021; Trickey et al., 2012) among traumatized populations, despite notable findings documenting that exposure to violence increases with age (Finkelhor et al., 2015). Similarly, no associations emerged between ethnicity, race, and living situation and any of the symptom clusters. While associations were not detected in this study, Black and African American young people disproportionately experience CSE/T (Clarke et al., 2012; Cook et al., 2018; Murphy et al., 2015; Naramore et al., 2017; National Center for Missing & Exploited Children, 2021; Reid et al., 2017). Future research should explore how race-based traumatic stress, including racism, discrimination, and marginalization affect entry into CSE/T and trauma symptomatology among young people of color (Carter & Pieterse, 2020; Carter et al., 2020; Polanco-Roman et al., 2016; Roberson & Carter, 2021).

Baseline trauma symptoms may impact treatment effects among traumatized populations. Research among veterans and adult females has found differential treatment effects between people with D-PTSD and those with a standard PTSD diagnosis (Bae et al., 2016; Kleindienst et al., 2016; Resick et al., 2012; Wolf et al., 2016). Some studies have reported that dissociative symptoms at the start of therapy negatively predict trauma treatment effects (Bae et al., 2016; Kleindienst et al., 2016; Wolf et al., 2016); however, other findings are mixed (Cloitre et al., 2012; Hagenaars et al., 2010; Halvorsen et al., 2014; Hoeboer et al., 2020; Resick et al., 2012; Zoet et al., 2018). Other research supports the utility of trauma therapies that include a narrative exposure and cognitive processing element for people with trauma-related dissociative symptoms (Resick et al., 2012).

While data collection is ongoing, young people who take part in Project Intersect program evaluation receive Trauma-Focused Cognitive Behavioral Therapy, an evidence-based trauma treatment that includes a trauma narrative and cognitive processing components (Cohen et al., 2017). TF-CBT has been shown to reduce PTSD symptoms and dissociation in trauma-exposed adolescents (Calleja, 2020). Other research has found that despite similar reductions in trauma symptoms among children and adolescents with PTSD and complex PTSD (i.e., which can include dissociative states; Maercker et al., 2013), at the end of therapy those with complex PTSD still experienced statistically significantly greater PTSD symptoms than those with a standard PTSD diagnosis (Sachser et al., 2017). Together with the current findings, we extend recommendations, initially suggested by Cohen et al. (2012), to include a comprehensive clinical assessment of trauma history, PTSD symptoms by symptom cluster, and dissociative symptoms throughout treatment, regardless of whether a client meets criteria for a PTSD diagnosis (Cohen et al., 2012). This information can be used to inform the therapeutic treatment plan and case planning, and may be especially relevant for young people who have experienced CSE/T (Pascual-Leone et al., 2017).

Limitations

Several limitations should be considered while interpreting these findings. First, the data were collected for the purposes of program evaluation among a sample of young people who had substantiated experiences of CSE/T and were referred for mental health treatment services. Therefore, these findings only reflect the trauma symptoms and histories of young people referred for mental health treatment services who have experienced CSE/T, not those who have not been identified or referred to services nor those who do not self-select into the program evaluation. Factors that further limit the generalizability of these findings include the small sample size and lack of variability in the data. Given the relatively small sample size, we were unable to compare the associations between polyvictimization and PTSD symptoms across the four symptom clusters. As larger datasets become available, multivariate models should be used to explore these associations. Further, the majority of the sample identified as female, which may have been due to system-level challenges in identifying male and non-cisgender individuals who have experienced CSE/T (Curtis et al., 2008; ECPAT-USA, 2013; Swaner et al., 2016). It is imperative that research explores ways to better identify male and non-cisgender young people who have experienced CSE/T in order to provide them with the necessary trauma-informed mental health treatment and other needed resources.

Third, we were limited by the measures collected in the program evaluation. Therefore, we cannot ascertain any information about unmeasured potentially traumatic events. Further, the trauma screen did not assess repeated experiences of trauma nor the timing of these in relation to the CSE/T experience. The items assessing PTSD symptoms only assessed trauma symptoms related to one traumatic event that is the most bothersome which, for a population with significant trauma history, may not accurately capture trauma symptoms. Fourth, there were several observations with missing data, especially for the items measuring PTSD symptoms; this may have been due to survey fatigue as these items were placed near the end of the survey. However, we addressed missing data using multiple imputation to reduce bias.

Conclusions

This study supports the importance of the inclusion of a comprehensive trauma history and symptoms in initial assessment and repeated measurement at multiple timepoints throughout the course of treatment. Specifically, these findings support the inclusion of a comprehensive trauma history that captures the full breadth and chronicity of trauma experiences, each PTSD symptom cluster, and dissociative symptoms in the comprehensive clinical assessment, which can be used to inform the therapeutic treatment plan of young people who have experienced CSE/T. Randomized controlled trials are needed to understand the treatment effects of evidence-based therapies, such as TF-CBT, among young people who have experienced CSE/T and explore any differential treatment effects of TF-CBT based on specific trauma histories and trauma symptoms at baseline. Further, in light of research suggesting significant differences in PTSD symptoms between adolescents with PTSD and those with complex PTSD at the end of TF-CBT (Sachser et al., 2017), future research should develop and test adaptations to TF-CBT using innovative research designs, such as a Sequential Multiple Assignment Randomized Trial, to understand if an adaptive intervention is needed for young people who have experienced CSE/T who begin therapy with pronounced trauma and dissociative symptoms (Almirall et al., 2014).

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Program evaluation findings reported in this article were from Project Intersect, a service grant funded by the Substance Abuse and Mental Health Services Administration (SAMHSA) as part of the National Child Traumatic Stress Network (SM061107-03; to principal investigator K.K.). The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of SAMHSA.

Declarations

Conflict of Interest

We have no known conflicts of interest to disclose.