Trauma Exposure and PTSD Among Older Adolescents in Foster Care

Abstract

Purpose

Youth in foster care represent a highly traumatized population. However, trauma research on this population has focused primarily on maltreatment rather than the full spectrum of trauma experiences identified within the DSM-IV. The current study aims to fill this gap by reporting the prevalence of exposure to specific types of traumatic events for a large sample of youth with foster care experience. The study also reports the likelihood of lifetime PTSD diagnoses associated with each specific type of trauma.

Method

Data are from a longitudinal panel study of 732 adolescents aged 17 and 18 who were in foster care. Lifetime trauma exposure and PTSD diagnosis were assessed using the Composite International Diagnostic Interview. Statistical comparisons were made using logistic regressions.

Results

The majority of respondents had experienced at least one trauma in their lifetime. While overall trauma prevalence did not differ by gender, males were more likely to experience Interpersonal Violence and Environmental Trauma, while females were more likely to experience Sexual Trauma. Caucasian participants reported higher rates of trauma exposure than African-American participants. The types of trauma associated with the highest probability of a lifetime PTSD diagnosis were rape, being tortured or a victim of terrorists, and molestation.

Conclusions

Youth in foster care are a highly traumatized population and meet diagnostic criteria for PTSD at higher rates than general youth populations. The ongoing impact of trauma may be particularly problematic for these young people given their abrupt transition to independence.

Trauma Exposure and PTSD Among Older Adolescents in Foster Care

Children and youth are taken into state custody and placed into foster care primarily on the basis of substantiated physical abuse, sexual abuse, or neglect [1]. Youth with child welfare experience are thus likely to have elevated risk for symptoms associated with trauma exposure. A study assessing foster care alumni found that 30% of respondents met lifetime diagnostic criteria for PTSD compared with 7.6% of a general population sample with similar demographics [2]. Likewise, two studies focusing on adolescents, aged 17–18 years, preparing to exit the child welfare system reported lifetime prevalence rates for PTSD of 14% and 16% based on structured diagnostic interviews [3, 4]. In contrast, the lifetime prevalence of PTSD among older youth in the general population, also aged 17–18, was found in two studies to be 6 to 7% [5, 6]. A similar lifetime prevalence of 7% was reported for a large sample of lower income urban adults aged 19–23 [7]. Thus, the lifetime prevalence of PTSD for transition-age youth with foster care experience appears to be roughly twice that of same-age counterparts in the general population.

A necessary prerequisite for a diagnosis of PTSD is the experience of a traumatic event. Some prior studies have estimated the nature and extent of trauma experienced by older adolescents and young adults. In a community sample of 18-year olds [5], approximately 43% had experienced a DSM-valid trauma, the most common being news of another’s sudden death or accident (13%), seeing someone hurt or killed (13%), and experiencing a sudden injury or accident (10%). In a slightly older sample of urban youth, Breslau et al [7] found a much higher trauma exposure rate of 82.5%. Nevertheless, the most commonly reported traumas were similar, including learning about an unexpected death (51.9%), witnessing someone being killed or seriously injured (35.9%), and being mugged or threatened with a weapon (35.6%). Additionally, a study by Copeland et al [8] of a representative sample of 16-year olds found 68.2% had at least one traumatic experience. The most common traumas reflected those in the other two studies, including witnessing a life event (23.7%), learning about a life event (21.4%), and violent death of a sibling or peer (14.5%). Research also indicates trauma experiences are associated with demographic characteristics. Roberts and colleagues [9] reported a variety of racial differences in trauma exposure in a large nationally representative sample of adults. In terms of gender, both Breslau et al [7] and Giaconia et al [5] found that young adult females were more likely to experience sexual trauma. These findings are consistent with a meta-analysis of gender differences in the trauma experiences of adults [10].

Specific types of trauma exposure have been linked to the likelihood of PTSD diagnosis for youth. Giaconia et al [5] found that 14.5% of youth experiencing any type of trauma met lifetime diagnostic criteria for PTSD, while Breslau et al [7] found this to be the case for 8.8% of youth exposed to trauma. In all studies reviewed, traumas involving sexual violence (rape, assault, abuse) were most predictive of lifetime diagnosis of PTSD [5, 7, 11, 12]. For example, two studies reported that approximately half of participants who experienced rape met diagnostic criteria for PTSD at some time in their lives [5, 7].

Although evidence is emerging about the nature and consequences of trauma experienced by older youth in the general population, much less is known about the trauma experiences of youth in foster care. To date, the literature addressing trauma exposure among youth in the child welfare system has focused on abuse and neglect rather than the broader range of events that constitute trauma. Three-fourths of the McMillen et al [4] sample of 17-year olds in foster care reported maltreatment at some point in their lives. In addition to experiencing maltreatment prior to entering the child welfare system, youth sometimes are maltreated while in foster care. The Casey Family National Foster Care Alumni Study found that over 90% of alumni had histories of maltreatment, and 21% reported maltreatment that occurred while they were in foster care [13]. Some DSM trauma categories are reflected in measures of maltreatment. However, evaluating the trauma experiences of youth in foster care only through maltreatment measures yields an incomplete picture of the trauma experienced by these youth. Furthermore, there is limited knowledge of how trauma relates to PTSD symptoms for this population. A study of foster youth aged 8–19 found that 64% of sexually abused participants compared to 42% of physically abused participants met criteria for PTSD, whereas only 18% of those experiencing a trauma other than physical or sexual abuse met criteria for PTSD [14]. However, no studies have investigated associations between particular types of trauma and the development of PTSD among youth in child welfare for the full range of DSM traumas.

The present study reports rates of exposure to specific types of traumatic events based on structured clinical interviews with a sample of older adolescents in the child welfare system. Potential differences in exposure associated with gender and race are investigated due to differences observed in the general population. Finally, the study examines which types of trauma are most predictive of meeting lifetime diagnostic criteria for PTSD.

Method

The data used in the current study are from baseline interviews of a longitudinal panel study tracking a cohort of adolescents exiting the public child welfare system to investigate the transition to independent living [15]. A representative sample was obtained using a systematic sampling procedure [16]. The sampling frame included adolescents who: were in out-of-home care supervised by public child welfare agencies in three Midwestern states (Illinois, Iowa, Wisconsin) between April 2002 and June 2002; were 17 years or older at time of recruitment; and had been in out-of-home care for at least one year prior to recruitment. All eligible youth in Iowa and Wisconsin and a random selection of 67% of eligible youth in Illinois were included in the sampling frame. The only exclusion criteria were inability to participate due to developmental disability, language barriers (non-English speaking), in-patient psychiatric institutionalization, or incarceration. Of 770 eligible cases, 732 (95%) consented to participate and completed in-person baseline interviews according to IRB-approved protocols.

The sample was evenly divided among females (51.5%) and males (48.5%). All respondents were 17 (59.0%) or 18 (41.0%) years old (Mean=17.4, SD=.5). The mean age of entry into the child welfare system was 10.8 years (SD = 4.0). A majority of the sample identified as African American (57.3%), followed by Caucasian (31.0%), mixed race (9.8%), American Indian/Native Alaskan (1.4%), and Asian/Pacific Islander (0.5%). Only 8.6% of the sample identified Hispanic ethnicity. At the time of the baseline interview, 35.8% were in foster homes with non-relatives, 30.5% in kinship foster homes with relatives, 18.1% in group care or residential treatment facilities, 8.6% in independent living arrangements, 0.7% in adoptive homes (pre-finalization), and 6.3% in some other setting.

Participants were assessed using the Composite International Diagnostic Interview (CIDI) [17]. Information regarding lifetime exposure to trauma was obtained through self-reported responses to the initial trauma screener of the assessment for PTSD, and a lifetime diagnosis of PTSD was derived from the CIDI algorithm based on the criteria of the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV). Statistical comparisons were made using logistic regressions. Independent variables used for comparisons of trauma exposure include gender and race. Race comparisons were limited to participants reporting race as African-American (n=417) or Caucasian (n=226).

Results

Lifetime Prevalence of Trauma Exposure

Lifetime prevalence rates for specified traumatic events and more general trauma categories are reported in Table 1. A majority of respondents (80.3%) had experienced at least one DSM-qualifying trauma in their lifetime. Almost two-thirds (61.7%) had experienced two or more DSM-qualifying events. The most common categories of traumatic events experienced were Indirect Traumas and Interpersonal Violence, experienced by 54.5% and 50.1% of the respondents, respectively. The most commonly reported specific traumas included witnessing someone being injured or killed (40.4%); being physically attacked or assaulted (30.3%); being molested (27.2%); and being threatened with a weapon, kidnapped, or held captive (26.5%).

Table 1.

Prevalence of Trauma Exposure Overall and by Gender and Race

	Total, %	Female	Male		Afr. Amer.	Caucasian
Event	(N=732)	(n=377)	(n=355)	OR (95% CI)	(n=417)	(n=226)	OR (95% CI)
Interpersonal Violence	50.1	42.4	58.3	1.9 (1.4–2.5)***	43.9	57.1	1.7 (1.2–2.4)**
Direct Combat	22.7	18.3	27.3	1.7 (1.2–2.4)**	24.5	19.0	0.7 (0.5–1.1)
Physically attacked/assaulted	30.3	26.5	34.4	1.5 (1.1–2.0)*	23.7	38.9	2.1 (1.4–2.9)***
Threatened with weapon/ kidnapped/ captive	26.5	18.6	34.9	2.3 (1.7–3.3)***	21.1	31.9	1.8 (1.2–2.5)**
Tortured/victim of terrorists	4.4	4.0	4.8	1.2 (0.6–2.5)	3.1	5.8	1.9 (0.9–4.2)
Sexual Trauma	29.9	44.8	14.1	0.2 (0.1–0.3)***	20.6	46.0	3.3 (2.3–4.7)***
Raped	17.5	28.6	5.6	0.1 (0.1–0.2)***	13.7	23.5	2.0 (1.3–3.0)**
Molested	27.2	40.3	13.2	0.2 (0.2–0.3)***	18.0	42.9	3.4 (2.4–5.0)***
Environmental Trauma	35.7	30.5	41.1	1.6 (1.2–2.2)**	30.9	42.5	1.6 (1.2–2.3)**
Life-threatening Accident	23.9	19.1	29.0	1.7 (1.2–2.5)**	22.1	24.8	1.2 (0.8–1.7)
Natural Disaster	21.0	18.8	23.4	1.3 (0.9–1.9)	17.7	27.9	1.8 (1.2–2.6)**
Indirect Trauma	54.5	50.7	58.6	1.4 (1.0–1.8)*	55.6	50.4	0.8 (0.6–1.1)
Witnessed someone injured/killed	40.4	32.6	48.7	2.0 (1.5–2.7)***	43.4	35.8	0.7 (0.5–1.0)
Shock because happened to someone close – DSM valid	20.6	24.7	16.3	0.6 (0.4–0.9)**	19.4	21.2	1.1 (0.8–1.7)
Shock because happened to someone close – sub-DSM	11.7	11.1	12.4	1.1 (0.7–1.8)	13.9	6.6	0.4 (0.2–0.8)**
Other Trauma	20.8	22.0	19.4	0.9 (0.6–1.2)	13.2	29.6	2.8 (1.9–4.2)***
Other - DSM valid	14.8	17.5	11.8	0.6 (0.4–1.0)*	9.6	20.8	2.5 (1.6–3.9)***
Other - sub-DSM	6.0	4.5	7.6	1.7 (0.9–3.3)	3.6	8.8	2.6 (1.3–5.2)**
Any traumatic event (not incl. Sub-DSM)	80.3	80.6	80.0	1.0 (0.7–1.4)	76.5	85.0	1.7 (1.1–2.7)*
Only one traumatic event (not incl. Sub-DSM)	18.6	20.2	16.9	0.8 (0.6–1.2)	20.1	15.9	0.8 (0.5–1.2)
2 or more traumatic events (not incl. Sub-DSM)	61.7	60.5	63.1	1.1 (0.8–1.5)	56.4	69.0	1.7 (1.2–2.4)**
Any traumatic event (incl. sub-DSM)	82.0	82.8	81.1	0.9 (0.6–1.3)	78.2	87.2	1.9 (1.2–3.0)**
Only one traumatic event (incl. sub-DSM)	16.8	19.1	14.4	0.7 (0.5–1.1)	18.0	15.9	0.9 (0.6–1.3)
2 or more traumatic events (incl. sub-DSM)	65.2	63.7	66.8	1.1 (0.8–1.6)	60.2	71.2	1.6 (1.2–2.3)**
Meets lifetime criteria for PTSD diagnosis	15.1	21.9	7.9	0.3 (0.2–0.5)***	13.3	16.1	1.3 (0.8–2.0)

^*p<.05.

^**p<.01.

^***p<.001

Gender comparisons

Gender and race comparisons are also reported in Table 1. Overall trauma prevalence did not differ markedly by gender. However, statistically significant gender differences were observed in types of traumatic events experienced. Males were 1.4 times more likely than females to experience Interpersonal Violence, and more likely to have experienced all of the singular events within that category, except being tortured or a victim of terrorists. Similarly, males were 1.3 times more likely than females to report Environmental Trauma, although only one specific event in this category, experiencing a life-threatening accident, differed significantly. Regarding specific Indirect Trauma events, males were 1.5 times more likely than females to witness someone being injured or killed, while females were 1.5 times more likely to experience a DSM-qualifying shock from learning of events happening to someone else. Females were 3.2 times more likely than males to experience Sexual Trauma, with females 3.1 times more likely to be molested and 5.1 times more likely to be raped.

Racial comparisons

Caucasian participants had higher rates of trauma exposure than African-American participants (85% vs. 76.5%) and also were more likely to experience two or more DSM-qualifying traumas (69% Caucasian vs. 56% African American). In fact, Caucasian participants reported higher trauma exposure than African-American participants in all categories except Indirect Trauma. The largest differentials involved Sexual Trauma and Other Trauma. The only traumatic event experienced more often by African Americans than Caucasians was a sub-DSM level shock resulting from something happening to someone close, which African American respondents were twice as likely to report.

Prediction of PTSD

PTSD prevalence by type of trauma exposure is reported in Table 2. The types of trauma associated with the highest probability of a lifetime PTSD diagnosis were rape (39.7% met PTSD criteria; OR, 5.9), being tortured or a victim of terrorists (45.2% met criteria; OR, 5.1), and molestation (32.8% met criteria; OR, 5.1). The specific trauma least likely to be associated with a lifetime PTSD diagnosis was direct combat (16.9% met criteria). Those experiencing two or more traumatic events were much more likely to meet PTSD diagnostic criteria (22.5%) than those who had experienced only one trauma (6.7%).

Table 2.

PTSD Prevalence based on Trauma Exposure

	PTSD, %
Traumatic Event	Exposed	Not Exposed	OR (95% CI)
Interpersonal Violence	20.0	10.2	2.2 (1.4–3.4)***
Direct Combat	16.9	14.6	1.2 (0.7–1.9)*
Physically attacked/assaulted	23.1	11.7	2.3 (1.5–3.4)***
Threatened with weapon/ kidnapped/ captive	24.2	11.9	2.4 (1.6–3.6)***
Tortured/victim of terrorists	45.2	13.8	5.1 (2.4–10.7)***
Sexual Trauma	33.5	7.4	6.3 (4.1–9.7)***
Raped	39.7	10.1	5.9 (3.8–9.2)***
Molested	32.8	8.7	5.1 (2.4–10.7)***
Environmental Trauma	21.2	11.8	2.0 (1.3–3.0)**
Life-threatening Accident	20.6	13.4	1.7 (1.1–2.6)*
Natural Disaster	22.2	13.2	1.9 (1.2–2.9)**
Indirect Trauma (not incl. sub-DSM)	20.0	10.3	2.2 (1.4–3.3)***
Witnessed someone injured/killed	21.4	10.9	2.2 (1.5–3.3)***
Shock because happened to someone close	23.8	12.9	2.1 (1.4–3.3)**
Other DSM-valid Trauma	28.0	12.9	2.6 (1.6–4.3)***
Any traumatic event (not incl. Sub-DSM)	18.8	0.0	-
Only one traumatic event (not incl. Sub-DSM)	6.7	0.0	-
2 or more traumatic events (not incl. Sub-DSM)	22.5	3.2	8.7 (4.3–17.5)***

^*p<.05.

^**p<.01.

^***p<.001

Discussion

This study investigated the lifetime exposure of older youth in the child welfare system to various traumas as well as the prevalence of PTSD diagnosis corresponding to that exposure. To place the findings in context, the results are compared in Table 3 with those of the two most comparable age cohort samples [5, 7]. The overall trauma exposure rate for youth in foster care was double that found in the youth sample from the general population and very similar to that of the sample of urban young adults. Both the current sample and Breslau et al [7] sample experienced high levels of interpersonal violence and indirect traumas; however, the current sample experienced substantially higher rates of sexual trauma than the other groups. While the current sample and the Breslau et al [7] sample experienced similar rates of trauma overall, the youth in foster care experienced substantially higher lifetime PTSD diagnosis rates, with 18.8% of traumatized youth meeting diagnostic criteria compared with only 8.8% of trauma-experienced urban young adults. While the Giaconia et al [5] sample experienced less trauma than the current sample, the rates of PTSD in these samples were much more similar (14.5% of traumatized young adults in Giaconia et al sample met criteria for PTSD). For all three studies, rape is the trauma most highly associated with meeting diagnostic criteria for PTSD. The current sample and Breslau et al [7] sample also had high PTSD likelihood for experiences of torture or being a victim of terrorists. In all three studies, experiencing a natural disaster had low associations with PTSD compared with other traumas.

Table 3.

Comparisons of Trauma Exposure and PTSD Across Three Similarly Aged Samples.

Trauma Type	Current Study Exposure % / PTSD %	Giaconia et al (1995) Exposure % / PTSD %	Breslau et al (2004) Exposure % / PTSD %
Interpersonal Violence	Direct combat: 22.7% / 16.9% Physically attacked/assaulted: 30.3% / 23.1% Threatened with weapon/ kidnapped/ captive: 26.5% / 24.2% Tortured/ victim of terrorists: 4.4% / 45.2%	Physically attacked/ assaulted: 6.5% / 12% Threat: 2.1% / 0%	Badly beaten: 10.1% / 13.3% Shot/stabbed: 14.2% / 9.4% Held captive/ tortured/ kidnapped: 2.1% / 20% Mugged/ threatened with weapon: 35.9% / 4.1%
Sexual Trauma	Rape: 17.5% / 39.7% Molestation: 27.2% / 32.8%	Rape: 2.1% / 50%	Rape: 5.6% / 46.2% Other sexual assault: 6.4% / 29%
Environmental Trauma	Life-threatening accident: 23.9% / 20.6% Natural disaster: 21% / 22.2%	Natural disaster: 1.3% / 0%	Serious car accident: 14.1% / 10% Other serious accident: 6.7% / 5.9% Natural disaster: 8.6% / 0% Life-threatening illness: 2.9% / 23.1% Child's life-threatening illness: 1.5% / 5.3%
Indirect Trauma	Witnessed someone injured/killed: 40.4% / 21.4% Shock because happened to someone close: 20.6% / 23.8%	Witnessed someone injured/killed: 12.8% / 8.2% Shock because happened to someone close: 13% / 14% Others' experience: 2.1% / 12.5%	Witnessed someone injured/ killed: 35.9% / 5.4% Discovered a dead body: 7.5% / 5.3%
Other Trauma	Other trauma: 14.8% / 28%	Other trauma: 1% / 50% Narrow escape: 1.6% / 0% Sudden injury or accident: 10.4% / 12.5%
Any traumatic event	Any traumatic event: 80.3% / 18.8%	Any traumatic event: 43% / 14.5%	Any traumatic event: 82.5% / 8.8%

Although there was no gender difference in overall trauma exposure, females were much more likely than males to experience sexual trauma – the trauma category most strongly associated with PTSD. This finding may explain why females exiting care have a higher rate of PTSD than males [3]. A meta-analysis incorporating studies of multiple populations similarly found females to be approximately twice as likely as males to meet diagnostic criteria for PTSD, more likely to experience sexual trauma in both childhood and adulthood, and less likely than males to experience all other types of trauma except non-sexual child maltreatment, for which there was no difference [10].

The current study did observe differential trauma exposure based on race, with higher rates among Caucasian than African American participants. This finding is consistent with a study of child welfare system youth indicating that African American youth reported maltreatment at lower rates than other racial/ethnic groups [18]. It appears that Caucasian youth in the child welfare system have more violent or traumatic histories than their African American counterparts. One possibility is that more African American children are being placed into out-of-home care for non-abuse related reasons, such as neglect. Another possibility is that African American families are under higher surveillance in general, which may result in child removal under less serious circumstances. Indeed, these potential explanations are supported by follow-up analyses of administrative data from the Illinois child welfare agency available for 361 youth in the current study from that state. Trend level differences indicate higher rates of verified allegations for subsistence related neglect, such as inadequate food (p=.07) and inadequate shelter (p=.06), among African American youth. In contrast, verified allegations involving physical injury were substantially more likely for Caucasian youth (p<.01).

The reported findings should be interpreted with awareness of certain study limitations. For example, retrospective assessment of lifetime prevalence for trauma and PTSD is subject to errors of memory. Furthermore, although the large, tri-state sample was derived with an epidemiologically-oriented, population-based sampling approach, the sample is not entirely representative of youth in the child welfare system due to the exclusion criteria employed. In particular, trauma prevalence may be underestimated given the omission of youth in inpatient psychiatric wards and correctional facilities. Finally, the study investigated only one potential consequence of trauma, PTSD, although trauma is associated with other mental health conditions such as depression, alcohol abuse, and other anxiety disorders [19, 20].

In summary, youth in foster care are a highly traumatized population and meet diagnostic criteria for PTSD at higher rates than general youth populations. For youth aging out of care, the ongoing impact of trauma may be particularly problematic given their sudden transition from intense system dependence to rugged independence, a challenge not typically expected of youth in the general population [21, 22, 23]. Approximately 30% of respondents in the current study reported experiencing their worst trauma at or after age 16. In most states, youth must exit the foster care system at the age of 18. This means that youth are being released from care while they are still attempting to process highly traumatic experiences. Furthermore, access to health insurance and receipt of mental health counseling is dramatically reduced when youth leave the child welfare system [24]. The current findings suggest that child welfare policy and practice should account for the potential needs of these youth for trauma-related services, specifically evidence-based and age-appropriate treatments for PTSD.