Assessing the Validity and Clinical Utility of a Developmental Trauma Diagnosis in Ethnic Minority Adolescents

Abstract

Developmental Trauma Disorder (DTD), a proposed diagnostic construct designed to reflect symptoms common among multiply-traumatized youth, was examined in a population of primarily female, predominantly Hispanic and African-American adolescents seeking psychiatric treatment (N = 53). The study uses a mix of interview and self-report data to test the prevalence of DTD symptoms relative to PTSD symptoms in this sample as well as to differentiate the DTD symptoms from DSM-IV and DSM-5 PTSD. DTD symptomatology was found to be as prevalent in the sample as PTSD symptomatology and further, DTD was found to be distinct from PTSD These results add to the literature suggesting the utility and need for the addition of a DTD diagnosis to the DSM, as well as offer insight into a population that has not been heavily scrutinized under a DTD framework.

Complex trauma is the experience of chronic, prolonged, and developmentally adverse traumatic events, most often of an interpersonal nature (Courtois 2004). These extreme forms of traumatic stressors are typically experienced at developmentally vulnerable times in an individual’s life, such as early childhood or adolescence, when self-definition and self-regulation are formed and consolidated (Storr et al. 2007). Complex trauma also typically occurs within the child’s caregiving system, involving harm, abandonment, or betrayal by caregivers, resulting in a distortion of fundamental attachment security (Ford and Courtois 2009; Spinazzola et al. 2005). Often, affect dysregulation and psychological impairment among family members may also be traumatizing for children (Ozurk and Sar 2005). Childhood abuse (sexual, emotional and physical) and neglect (physical and emotional) constitute typical forms of complex trauma.

Childhood exposure to interpersonal trauma is a prevalent public health problem. Worldwide, approximately one third of children are estimated to experience physical abuse and approximately one in four girls and one in five boys experience sexual victimization during childhood (Anda et al. 1999; Putnam 2003; Covell and Becker 2011). In the U.S., Kilpatrick et al. (1998) estimated the lifetime prevalence of exposure to sexual assault, physical assault, and witnessing violence as 8, 17, and 39%, respectively. Each year more than 3 million children are reported to authorities for abuse and neglect in the U.S., and approximately 1.5 million of those cases are substantiated (U.S. Department of Health and Human Services et al. 2016). Some estimates place the fiscal cost of childhood abuse and neglect in 2007 at $103.8 billion (Wang and Holton 2007), including foster care and residential treatment. There are significant mental health consequences for those whom are victimized, and numerous studies have shown that exposure to interpersonal trauma can chronically and pervasively alter social, psychological, cognitive, and biological development (Burns et al. 1998; Cook et al. 2005; Spinazzola et al. 2005).

While the prevalence and impact of childhood interpersonal trauma have been well documented (Finkelhor et al. 2005, 2009), studies suggest that the symptom clusters of PTSD do not adequately capture the profound developmental disturbances or the traumatic families of origin of individuals who have experienced complex trauma (see Cook et al. 2005; Ford and Courtois 2009; Ford and Kidd 1998; Herman 1992b; McLean and Gallop 2003; Pearlman and Courtois 2005). In the DSM (American Psychiatric Association 2000, 2013), no single current psychiatric diagnosis could account for the myriad symptoms that research has shown frequently to occur in children exposed to interpersonal trauma (Fleming and Resick 2015). For example, in one study of 364 abused and neglected children (Ackerman et al. 1998), the most common diagnoses in order of frequency were separation anxiety, phobic disorder, PTSD, attentional deficit hyperactivity disorder and oppositional defiant disorder. Similarly, a survey of 1699 children receiving trauma-focused treatment showed that the vast majority (78%) had been exposed to multiple and/or prolonged interpersonal traumas (with a modal three trauma exposure types); less than 1/4, however, met diagnostic criteria for PTSD (Pynoos et al. 2008). Therefore, many children and adolescents who are severely impacted by violence may not meet criteria for PTSD (Cloitre et al. 2009; Fleming and Resick 2015). In fact, many forms of interpersonal trauma (e.g., emotional abuse, neglect, separation from caregivers, traumatic loss, and inappropriate sexualized behavior of caregivers) do not necessarily meet PTSD Criterion A definitions for a traumatic event in the DSM-5, despite their resultant impairment. As a result, a majority of treatment-presenting traumatized children and adolescents are diagnosed with multiple comorbid diagnoses in order to adequately describe their symptoms (Putnam et al. 2008). Because many child survivors of maltreatment-related trauma present with diverse and wide-ranging symptoms, the current diagnostic system may be in danger of misdiagnosing or overlooking the impact of developmental trauma, potentially contributing to over- or under-treatment that fails to address complex trauma (D'Andrea et al. 2012).

Two proposed diagnostic categories that address symptomatology specific to childhood chronic trauma include Disorders of Extreme Stress Not Otherwise Specified (DESNOS; van der Kolk et al. 2005), formerly known as Complex Post-Traumatic Stress Disorder (C-PTSD; Herman 1992a; van der Kolk et al. 2005), and Developmental Trauma Disorder (DTD; van der Kolk 2005). Each of these constructs describes the impairments and distress exhibited by those who have experienced complex trauma. DTD, in particular, is intended to capture symptoms that are frequently described in trauma-exposed child clinical samples: dysregulation of affect and physiology, of attention and behavior, and of self and relationship.

To date, few studies have attempted to validate the proposed DTD criteria relative to DSM-IV and DSM-5 criteria. Klasen et al. (2013) found that 33% of Ugandan former child soldiers met criteria for PTSD while 78.2% met criteria for DTD, suggesting that the DTD diagnosis was describing the symptoms of these highly traumatized children with a high degree of accuracy. Moreover, only 1% met criteria for PTSD alone, suggesting the PTSD criteria did not adequately capture their diverse symptoms. Stolbach et al. (2013) recently conducted a field trial of the proposed criteria for DTD in a sample of urban children (N = 214). Their results showed that children with complex trauma histories, as defined in DTD Criterion A, were much more likely to meet the proposed DTD symptom criteria than children who did not meet the exposure criterion, further supporting the DTD construct as a distinct and valid diagnostic category.

In order to account for the developmental differences in the expressions of PTSD in different age groups, the DSM-5 included a new developmental subtype of PTSD called posttraumatic stress disorder in preschool children. This marks a significant advance for DSM taxonomy, as it is the first developmental subtype of an existing disorder. However, this subtype can only be diagnosed in children ages 6 years and younger, leaving children ages seven and older as well as adolescents without a developmentally sensitive set of criteria to diagnose PTSD. Furthermore, the traumatic antecedents as currently defined do not account for forms of disrupted caregiving which may adversely impact children, including emotional abuse, neglect, and multiple changes in caregiving. Though the diagnoses of Reactive Attachment Disorder and the new Disinhibited Social Engagement Disorder do acknowledge the developmental impact of neglect or changes in caregiving, their symptom profiles are limited to social withdrawal and indiscriminate social engagement, respectively. Additionally, several symptoms that would meet criteria for the adult diagnosis are absent from those available to meet criteria for children (e.g. self-harm, negative self-image). Thus, the full scope of cognitive, affective, behavioral, and relational symptoms that may burden traumatized adolescents are still not clearly connected in the diagnostic picture to traumatic antecedents. Furthermore, the question of whether early-childhood PTSD is applicable to older children and adolescents remains to be evaluated.

In sum, while modifications to the diagnostic symptom have, in some ways, broadened the symptom set for posttraumatic reactions, it remains unclear whether the diagnoses available to traumatized youth are describing their symptoms. Utilizing clinical intake data collected from a highly trauma-exposed sample of urban youth, the current study aimed to determine the clinical utility and specificity of DTD and whether these symptoms are accounted for by any existing DSM-IV or DSM-5 disorders, including PTSD. Specifically, the current study examined the following topics:

1. The prevalence of DTD and DSM-IV PTSD, DSM-5 PTSD, and PTSD for children under 6.

2. Whether DTD would manifest as discrete from PTSD.

Method

Participants

All participants were admitted to the Adolescent Depression and Suicide Prevention Program (ADSP) of Montefiore Medical Center/Albert Einstein College of Medicine. Located in Bronx, NY, ADSP is an outpatient clinic for adolescents exhibiting depressive or suicidal features. Sample information was attained through archival clinical admissions data provided by the ADSP.

Measures

Developmental Trauma Disorder Symptoms

The Structured Interview for Disorders of Extreme Stress-Adolescent (SIDES-A; Pelcovitz 2004) is a structured interview specifically designed to asses high-trauma exposure in adolescents. The interview consists of six subscales (63 items in all) intended to evaluate alterations in the following areas: 1) affect and impulse regulation; (2) attention or consciousness; (3) self-perception; (4) perception of the perpetrator (5) relations with others; (6) somatization; and (7) systems of meaning. Each item is scored on lifetime occurrence and current level of distress (on a 0–3 scale). Anchors are provided for each rating choice, but the interviewer has the opportunity to discuss each question openly, allowing for collaboration between the youth and the interviewer to determine the best score for each question (see Pelcovitz 2004, for a full description of the interview structure). Items assessing the perceptions of the perpetrator were omitted from the ADSP’s intake evaluations and thus are not included in the current study.

Although the scale originally was intended to ask adolescents about particular experiences since the time that they were exposed to a severe traumatic event, all questions were framed as referring to experiences they may have had any time throughout their lives without reference to particular traumatic or stressful events. This approach was implemented because the scale as originally designed would have been untenable for the current population given (a) the probable complexity of this population’s traumatic histories and (b) the implied requirement that the adolescents have sufficient cognitive and emotional resources to report on subjective experiences mapped over an internal timeline of traumatic and stressful events. Trauma exposure and duration questions were used to determine DTD Criterion A.

It is worth noting that the SIDES-A is currently in revision form and thus has no current norms or empirical validation. Although DTD items were derived retrospectively using proxy measures from this archival dataset, empirically based symptom cluster and overall diagnostic scoring was employed (Ford et al. 2018) using the scoring criteria of three cluster B symptoms, two cluster C symptoms, and two cluster D symptoms. It was based on the original SIDES adult version, which has evidence of adequate inter-rater reliability and internal consistency for the SIDES total score and subscale scores in community and clinical samples (Pelcovitz et al. 1997). It should also be noted that the SIDES-A was developed based on the aforementioned construct of complex trauma, and as such it does not specifically capture DTD. Ford and the Developmental Trauma Disorder Work Group (2014) developed the Developmental Trauma Disorder Structured Interview for Child (DTDSI-C) during the years since the present data were collected. Since there was no developed DTD measure at the time of the current study, items from the SIDES-A and The Schedule for Affective Disorders and Schizophrenia for School-Aged Children-Present and Lifetime Version (K-SADS-PL; Kauffman et al. 1997) were used as a proxy as they contain items that directly map onto DTD symptoms (see Table 1 for how SIDES-A and KSADS items were matched to DTD symptoms). A DTD symptom was determined “present” if any one of the corresponding SIDES-A or KSADS items were endorsed. DTD symptomatology was determined by the proposed criteria outlined by NCTSN and currently being used in DTD field trials.

Table 1.

DTD Sub-Criteria and Corresponding Items

Sub-criterion	Items
B1: Inability to modulate, tolerate, or recover from extreme affect states	SIDES-A 1: Anger sensitivity SIDES-A 2: Preoccupation with Anger SIDES-A 7: Worried about anger
B2: Disturbances in regulation in bodily functions	SIDES-A 13: Avoids physical contact SIDES-A 36: Digestive problems SIDES-A 37: Chronic Pain SIDES-A 38: Cardiopulmonary problems SIDES-A 39: Conversion Symptoms SIDES-A 40: Sexual Dysfunction
B3: Diminished awareness of sensations, emotions and bodily states	SIDES-A 21: Lose track of time SIDES-A 22: Dissociation SIDES-A 23: Derealization
B4: Impaired capacity to describe emotions or bodily states	SIDES-A 24: Difficulty labeling and expressing feelings and internal experiences
C1: Preoccupation with threat, or impaired capacity to perceive threat	SIDES-A 09: Careless about safety
C2: Impaired capacity for self-protection	SIDES-A 30: Minimizing danger SIDES-A 18: Careless sexual behavior SIDES-A 19: Excessive risk taking
C3: Maladaptive attempts at self-soothing	SIDES-A 3: Inability to self-sooth SIDES-A 5: Violent Ideation
C4: Habitual or reactive self-harm	SIDES-A 10: Deliberate self-harm
C5: Inability to initiate or sustain goal-directed behavior	KSADS: ADHD diagnosis
D1: Intense preoccupation with safety of the caregiver or other loved ones	Not assessed
D2: Persistent negative sense of self	SIDES-A: 25 Perceives self as ineffectiveness SIDES-A: 26 Views self as permanently damaged
D3: Extreme and persistent distrust, defiance or lack of reciprocal behavior in close relationships with adults or peers	SIDES-A: 31 Distrust in others SIDES-A: 32 Avoid forming relationships
D4: Reactive physical or verbal aggression toward peers, caregivers, or other adults	SIDES-A: 6 Trouble controlling anger
D5: Impaired Psychological Boundaries	SIDES-A: 17 Attempts age-inappropriate sexual behavior with other children or adults
D6. Impaired interpersonal empathy	SIDES-A: 33 Difficulty with conflict

Psychiatric Diagnoses

The Schedule for Affective Disorders and Schizophrenia for School-Aged Children-Present and Lifetime Version (K-SADS-PL; Kauffman et al. 1997), a semi-structured interview designed to categorically diagnose current DSM-IV Axis I disorders among children and adolescents, was used to assess mood, anxiety, and personality disorders in the present study. In order to assess a possible DSM-5 diagnosis of PTSD, because the K-SADS for DSM-5 is not yet available, items were drawn from the SIDES-A and K-SADS to create Criterion B, C, D, and E symptoms of PTSD. A list of questions for each symptom is listed in Tables 2 and 3. Multiple items on the SIDES may address each of the new PTSD symptoms; a symptom was determined “present” if any one of the potential contributing questions was ranked as significantly present. In creating continuous symptom severity, the maximum severity for a given symptom was utilized. Similarly, according to DSM criteria, a version of PTSD for children under six was also computed, to investigate the possibility that extending the age range for these criteria would better capture clinical need. Thus, three versions of PTSD were tallied: DSM-IV, DSM-5 for adults, and DSM-5 for children.

Table 2.

DSM-IV PTSD Criteria and Corresponding KSADS Items

DSM-IV PTSD Criterion	KSADS PTSD Items
B. Presence of at least one of the following Re-Experience items:	Screener 1: Recurrent Thoughts or Images of Event, Supplemental 1: Repetitive Play, Screener 3: Nightmares, Supplemental 2: Dissociative Episodes, Illusions, or Hallucinations, Supplemental 3: Distress Elicited to Exposure to Stimuli, or Supplemental 13: Physiologic Reactivity
C. Presence of at least three of the Persistent Avoidance items:	Screener 2: Avoid Thoughts or Feelings, Supplemental 4: Avoid Activities, Supplemental 5: Inability to Recall, Supplemental 6: Diminished Interest, Supplemental 7: Feelings of Detachment, Supplemental 8: Restricted Affect, or Supplemental 9: Foreshortened Future
D. At least two of the Increased Arousal items	Screener 4: Insomnia, Screener 5: Irritability, Supplemental 10: Difficulty Concentrating, Supplemental 11: Hypervigilance, or Supplemental 12: Exaggerated Startle Response
E. Duration of at least one month; and	Supplemental 15: Duration
F. Evidence of functional impairment	Supplemental 14: Impairment

Table 3.

DSM-5 PTSD Criteria and Corresponding KSADS Items

DSM-V PTSD Criterion	KSADS PTSD Items
B. Presence of at least one of the following Intrusion items:	Screener 1: Recurrent Thoughts or Images of Event, Supplemental 1: Repetitive Play, Screener 3: Nightmares, Supplemental 2: Dissociative Episodes, Illusions, or Hallucinations Supplemental 3: Distress Elicited to Exposure to Stimuli, or Supplemental 13: Physiologic Reactivity
C. Presence of at least one of the following Avoidance items:	Screener 2: Avoid Thoughts or Feelings, or Supplemental 4: Avoid Activities,
D. Presence of at least two of the following Negative Alterations in Cognition or Mood items:	Supplemental 5: Inability to Recall, Supplemental 6: Diminished Interest, Supplemental 7: Feelings of Detachment, Supplemental 8: Restricted Affect, or Supplemental 9: Foreshortened Future
E. Presence of at least two of the following Alterations in Arousal and Reactivity items:	Screener 4: Insomnia, Screener 5: Irritability, Supplemental 10: Difficulty Concentrating, Supplemental 11: Hypervigilance, or Supplemental 12: Exaggerated Startle Response
F. Duration of at least one month; and	Supplemental 15: Duration
G. Evidence of functional impairment	Supplemental 14: Impairment

The K-SADS is one of the most widely used assessment tools among children and adolescents and has been found to be psychometrically sound. The instrument has high inter-rater reliability, ranging from 93 to 100% agreement, as well as 18-day test-retest reliability ranging from r = .77–1.00 for a host of diagnoses (major depressive disorder, bipolar disorder, generalized anxiety disorder, conduct disorder, and oppositional defiant disorder), indicating good reliability (Ambrosini 2000; Kauffman et al. 1997). Additionally, the instrument evidences concurrent and predictive validity (Kauffman et al. 1997; Lewinsohn et al. 1999). Due to the format of the ADSP’s intake process, inter-rater reliability was not available for this sample. However, all clinicians administered the K-SADS under supervision and worked with their supervisor in order to reach a consensus on diagnostic conclusions.

Trauma Exposure

The Montefiore Exposure to Trauma Screener (METS; Wren and Miller 2010) is a 38-item structured interview used to evaluate trauma exposure in adolescents. To create a more complete picture of possible traumatic antecedents, eleven traumatic events were augmented from existing measures to use with the K-SADS PTSD module. The measure includes seven items from the Traumatic Events Screening Inventory-Child (TESI-C; Ghosh-Ippen et al. 2002), 14 from the Traumatic Events Screening Inventory-Child Self-Report (TESI-R; Ghosh-Ippen et al. 2002), and five from the North Shore-Long Island Jewish Health System (NSLIJHS) Trauma History Checklist (2006).

The METS assesses complex trauma exposure across 12 different areas: separation/loss/neglect (6-items), witness to violence (6-items), victim of violence (4-items), emotional abuse (2-items), accidents (5-items), witness to family/domestic violence (4-items), natural disaster (3-items), sexual abuse/exposure (3-items), terrorist attacks/war (2-items), traumatic news (1-item), physical abuse (1-item), and other (1-item) (i.e., giving the option for additional self-report). Interpersonal trauma exposure was determined by endorsement of interpersonal METS items (e.g., sexual/physical/emotional abuse, neglect,). Conversely, non-interpersonal trauma exposure was determined by endorsement of non-interpersonal METS items (e.g., natural disaster, accident).

Data Analysis

The study questions are primarily descriptive in nature and are not testing variable associations. A chi-square test of independence was used to analyze the prevalence and co-occurrence of PTSD and DTD in this population. All expected cell sizes were > = 5, but in cases where they were small, we also report the significance of Fisher’s exact test.

Results

Demographics

The sample (N = 48) was comprised of treatment-seeking adolescents with a mean age of 14.9 years, 77% (n = 36) of whom were female. Fifty-seven percent (n = 23) identified as Hispanic, 30% (n = 14) as Black, 8% (n = 4) as White, and 6% (n = 3) as Asian. Socioeconomic status was not assessed per family, but the agency is in the 16th Congressional district, which ranks first in proportion of children living below the poverty line in the country: 38% of the population lives below the poverty line1 (Sisk 2010).

Psychiatric Diagnosis

Adolescents in the sample were diagnosed with an average of three psychiatric disorders. Aside from the disorders targeted for this paper, major depressive disorder (59%, n = 23), generalized anxiety disorder (23%, n = 11), and dysthymic disorder (19%, n = 9) were the most prevalent DSM-IV diagnoses met by the sample. There were also high rates of social phobia (15%, n = 7), depressive disorder NOS (15%, n = 7), attention deficit hyperactivity disorder (13%, n = 6), and substance abuse (13%, n = 6) found in the sample.

Trauma Exposure

All of the adolescents in the sample reported experiencing a traumatic event in their lifetime, with an average of 12.34 (SD = 5.02) types; the entire sample endorsed personal experience of chronic interpersonal trauma, and most of the sample (98%, n = 46) endorsed additional non-interpersonal acute trauma exposure. All of the adolescents who identified as Black, Asian, and White, as well as 94% of those identifying as Hispanic, met for DTD criterion A1 (direct experience or witnessing of repeated and severe episodes of interpersonal violence). Conversely, none of those who identified as Black, Asian, or White and only 3.4% of adolescents identifying as Hispanic met for criterion A2 (significant disruptions of protective caregiving as the result of repeated changes in primary caregiver; repeated separation from primary caregiver; or exposure to severe and persistent emotional abuse).

Question 1: How Prevalent Are DTD and PTSD Symptoms in this Treatment Sample?

The prevalence rates of PTSD and DTD are as follows: DSM-IV PTSD, 19% (n = 9); DSM-5 PTSD, adult criteria, 21% (n = 11); DSM-5 PTSD, child criteria, 30% (n = 16); and DTD, 48% (n = 25).

For those that met for a DSM-IV PTSD diagnosis, the distribution among ethnic groups was: 27% Black, 73% Hispanic, 0% White, and 0% Asian. Regarding a DSM-5 PTSD-adult diagnosis, the distribution among ethnic groups was: 18% Black, 82% Hispanic, 0% White, and 0% Asian. In terms of meeting for the proposed criteria for DTD, the distribution within ethnic groups was: 30% White, 23% Hispanic, 20% among Black, and 0% among Asian.

At the level of each symptom cluster, DSM-IV PTSD prevalence was: intrusion, 53% (n = 25); avoidance, 27% (n = 13); hyperarousal, 30% (n = 14). At the level of each symptom cluster, DSM-5 PTSD-adult prevalence was: intrusion, 38% (n = 18); avoidance, 48% (n = 21); affect/cognition, 38% (n = 18); and hyperarousal, 32% (n = 15). At the level of each symptom cluster, DSM-5 PTSD-child prevalence was as follows: intrusion, 57% (n = 27); avoidance or alterations in affect, 57% (n = 27); hyperarousal, 30% (n = 14).

At the level of each symptom cluster, DTD prevalence was as follows: affect/physiological dysregulation, 83% (n = 39); attentional/behavioral dysregulation, 55% (n = 26); self/relational dysregulation, 72% (n = 34).

Question 2: Is there any Evidence that DTD Exists as Discrete from PTSD?

In order to measure the convergence and divergence of PTSD and DTD diagnoses in the sample, cross-tabulations were performed. Results showed a significant association between DTD diagnosis and DSM-IV PTSD, x² (1) = 6.01, p = .009, such that DTD diagnosis was associated with lower likelihood of DSM-IV PTSD diagnosis. Of participants who met criteria for DTD, 36% (n = 9) met for DSM-IV PTSD whereas 64% (n = 16) did not. From the point of view of PTSD, of participants who met for DSM-IV PTSD, 90% (n = 9) met criteria for DTD, whereas 10% (n = 1) did not. Fisher’s exact test statistic was also significant, p = .012. The findings were similar for DSM-5 PTSD-adult, x²(1) = .81, p = .368. Here, there was no change in likelihood of DSM-5 PTSD-adult based on having DTD; 19 (76%) out of 26 youth with DTD did not have PTSD, and youth with DSM-5 PTSD-adult were more likely to have DTD than to not (6 of 9, or 66%, with DSM-5 PTSD-adult also had DTD). In applying the DSM-5 PTSD-child criteria, there was no association between DTD and DSM-5 PTSD-child, x² (1) = .12, p = .724. Eight out of 25 (n = 32%) youth meeting criteria for DTD met for DSM-5 PTSD-child, whereas 17 (68%) did not. Similarly, 16 out of 22 youth (73%) with DSM-5 PTSD-child did not meet for DTD, whereas 6 (27%) did.

We also examined other potential PTSD comorbidities in response to the argument that DTD may be comorbid PTSD and Borderline Personality Disorder (BPD). In fact, the majority of children with DTD (21/25) who met criteria for DTD did not meet criteria for comorbid PTSD and BPD. Similarly, in examining BPD alone (though its diagnosis is controversial in adolescents; see Kaess et al. 2014 for review), youth with symptoms consistent with BPD were likely to have DTD (13/14), but youth with DTD were equally likely to have or not have symptoms consistent with BPD (13/25), x² (1) = 12.60 p < .001.

Discussion

The present study aimed to expand upon the existing research literature on the consequences of childhood maltreatment by testing the validity and clinical utility of a developmentally sensitive trauma diagnosis (DTD) among a sample of ethnic minority adolescents. This study joins a growing body of research examining the sequelae of complex trauma exposure in clinical samples of ethnic minority adolescents (Ford 2012; Stolbach et al. 2013). The present data demonstrate a high exposure rate for direct and vicarious victimization in this sample, with the vast majority of participants reporting that they have witnessed or been the victim of violence in their community or in their home. Moreover, the available evidence suggests that DTD symptomatology is as prevalent in the sample as PTSD symptomatology, regardless of the PTSD conceptualization utilized, and that its symptoms exist independent of symptoms of PTSD, such that youth meeting DTD criteria thresholds may not also meet PTSD thresholds. This finding undermines the notion that DTD is simply a more severe form of PTSD. Taken together, these findings suggest that additional clinical need may be captured by DTD, beyond existing PTSD.

Even though every adolescent in this clinical sample endorsed experiencing exposure to a PTSD criterion A1 traumatic event during their lifetime, a minority met criteria for PTSD. This finding is in line with previous research from Pynoos et al. (2008) who found that less than a quarter of the children in treatment for trauma-related psychopathology with the National Child Traumatic Stress Network (NCTSN) met criteria for PTSD. One explanation for this discrepancy is that resilience and vulnerability factors may mediate the development of trauma-related symptomology. However, this was a sample with significant clinical impairment, including self-harm as a primary referral cause. Another explanation is that PTSD may not fully capture the range of posttraumatic impairments associated with exposure to childhood maltreatment. Many forms of interpersonal trauma, in particular psychological maltreatment (e.g., neglect, separation from caregivers, traumatic loss and inappropriate sexual behavior), do not necessarily meet PTSD criterion A1 definitions for a traumatic event in the DSM-IV (van der Kolk 2005). Further, in the criteria for the DSM-5, PTSD criterion A2, which includes the subjective response of intense fear, helplessness, or horror is retracted. Although there is controversy about the A2 criteria, there are several studies among school-age children and adolescents that indicate its utility. A2 has demonstrated incremental explanatory power in dose of exposure studies over and above A1 in predicting variance in child and adolescent PTSD within and across exposure groups (Giannopoulou et al. 2006; Roussos et al. 2005; Zatzick et al. 2008). These results highlight the weakness of PTSD criteria in adequately reflecting both the experiences and symptoms of complexly traumatized youth and the need for broadening current diagnostic nosology.

The prevalence of DSM-5 PTSD and DTD were, together, approximately 40%. Given that the sample is in significant clinical distress with clear functional impairment, it is open to interpretation whether this rate is high or low. One interpretation of these data is that DTD requires too stringent a threshold for diagnosis. Currently, eight symptoms are required, distributed across three clusters; this configuration is a higher bar than PTSD, which requires six symptoms across four clusters. Though it has been argued that DTD may simply represent more severe PTSD or a conglomerate of comorbidities, low comorbidity with PTSD (relative to DTD alone) speaks to DTD as a distinct, albeit clearly related, diagnostic entity. The overlap between PTSD and DTD was similar to the overlap between any two mood, anxiety, and trauma-related disorders (Cloitre et al. 2014); for example, the rate of depression among PTSD patients is approximately 44.5% (Shalev et al. 2014). These findings still held for DSM-5 PTSD, which has added symptoms of negative mood, negative schemas, and risk-taking, all of which are represented in DTD.

Further, the high prevalence of comorbid diagnoses found in this sample suggests that in the absence of a developmentally sensitive trauma-specific diagnosis for adolescents, issues of comorbidity arise. In this study, adolescents were diagnosed with an average of three Axis I disorders with PTSD estimated as the fifth most prevalent diagnosis. These findings converge with previous longitudinal research by Putnam et al. (2008) on the comorbid effects of child abuse as well as with research by Ackerman et al. (1998) and Copeland et al. (2007), who reported PTSD to be the 5th and 10th most common disorder in childhood following exposure to traumatic stressors, respectively. Despite expected overlap, DTD was found to occur in the absence of PTSD symptomatology, suggesting that the two conditions had significantly different symptomatic and functional impairment features. These results lend additional credence to the growing body of literature that supports the specificity and utility of a developmental trauma diagnosis (D'Andrea et al. 2012; Stolbach et al. 2013; Ford et al. 2013). Furthermore, these findings support the clinical utility and practicality of the DTD diagnosis when conceptualizing the constellation of symptoms unique to adolescents who present with complex trauma histories.

Limitations

Though the results of the current study may have important theoretical and clinical implications for understanding the distinction between symptom sets, there are several limitations that must be considered. For example, there are well-known limitations inherent in retrospective self-report data, such as error due to forgetting and biased recall and reliance. Even interview data, utilized here, contains an element of self-report. The adolescents in this sample were not as temporally removed from the reported events as adult participants; however, recent systematic reviews suggest that there is a significant tendency to under-report instances of maltreatment in retrospective recall of childhood events (see Brewin et al. 1993; Hardt and Rutter 2004), which may reduce the diagnosis rates in this study. This temporal proximity to the trauma may be related to the low reporting of emotional abuse in the sample, which is in contrast with other population-wide studies. Additionally, the current study does not examine the exact timing of trauma. Given evidence that the timing of trauma in infancy (Scheeringa and Zeanah 2001) and throughout childhood and adolescence (Thornberry et al. 2001) may affect critical outcomes, future studies should take trauma timing into account.

Finally, the generalizability of the results of this study may be limited by the small sample size as well as by sample characteristics, as the participants were outpatients at a depression and suicide prevention clinic with base rates of symptoms and prevalence of comorbidities higher than the general population but lower than the inpatient groups used in previous adolescent research. Accordingly, the findings may not be applicable to less acutely distressed adolescents, adolescents who are more socioeconomically advantaged, or to adults.

Clinical Implications and Future Directions

The results of the current study strengthen a growing trend in the literature that supports the validity and clinical utility of a developmental trauma diagnosis to better capture the clinical picture associated with exposure to complex trauma in adolescents. This classification is important particularly because of its vast theoretical and clinical implications: psychiatric diagnoses guide the development of treatment interventions, insurance reimbursement, and scientific inquiry.

In the present study, approximately half of the sample met criteria for DTD even though 97% reported exposure to DTD criterion A1 and there was an overall high prevalence of DTD symptomatology across the sample. However, to date, few treatments have evolved that explicitly address complex/developmental trauma exposure in this age group. Exceptions include STAIR (Cloitre et al. 2002), TARGET (Ford and Russo 2006), TF-CBT (Cohen et al. 2012), ARC (Arvidson et al. 2011), SPARCS (Cook et al. 2005), and Real Life Heroes (Kagan et al. 2008). In the absence of formal diagnostic recognition, the majority of treatment research has been devoted to PTSD, which may not fully capture treatment needs or translate to a more complexly-symptomatic sample (Cloitre et al. 2011; Cloitre et al. 2012).

While it is widely recognized that many children and adolescents who experience complex trauma can be highly resilient, the mechanisms of said resilience and the factors that confer it have not been sufficiently or specifically studied. It would be beneficial for future research to understand how these variables may serve as risk and protective characteristics to developing PTSD or DTD as well as how prior experience, stressor-related features, and cognitive attributes contribute to the clinical presentation. Moreover, identifying these factors can aid in establishing early and possibly preventive treatments (Cicero et al. 2011). Alternatively, because an overwhelming majority of the sample reported trauma but did not meet criteria for either PTSD diagnosis or DTD diagnosis, it is plausible that the proposed DTD diagnosis could better capture adolescent trauma-related psychopathology with fewer required symptoms. Future studies would benefit from delineating the thresholds that confer clinical utility, sensitivity, and specificity for each of the proposed DTD symptom clusters in order to more accurately capture the clinical impressions of adolescents exposed to chronic interpersonal trauma.

Compliance with Ethical Standards

Disclosure of Interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Ethical Standards and Informed Consent

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation [institutional and national] and with the Helsinki Declaration of 1975, as revised in 2000. All information was collected as part of routine clinical care and data were culled from charts.