Chat with us, powered by LiveChat Treatment options for children and adolescent trauma survivors can include cognitive behavioral th | Max paper


Treatment options for children and adolescent trauma survivors can include cognitive behavioral therapy (CBT) and crisis management to reduce anxiety, worry, and fear of repeated trauma. Play therapy is an effective method often used with young children with posttraumatic stress disorder because they often have difficulty dealing with trauma directly. Cases in which a child or adolescent is acting out sexually in response to a sexual trauma or in which he or she may be using drugs or alcohol as a coping mechanism as a result of trauma often require additional treatment modalities.

For this Assignment, review the media program Trauma and consider the symptomology of PTSD and how trauma can affect children and adolescents. Then, select a different type of childhood or adolescent trauma from the one you selected for the Discussion. Consider the posttraumatic stress symptoms that are likely to occur and what type of treatment interventions you might use to treat the child or adolescent. Also, think about how you might support or educate parents or guardians as they attempt to support their child or adolescent.

The Assignment (2–3 pages):

  • Describe a major trauma or event that may occur to children and/or adolescents.
  • Describe three potential symptoms of posttraumatic stress disorder that may occur as a result of the major trauma or event, and explain why these symptoms may occur.
  • Describe one intervention you might use in treating this type of trauma. Justify the selection of your intervention using the week’s resources and current literature.
  • Explain two ways you might educate or support the parents/guardians as they help their child or adolescent through the trauma. Be specific.


How Mental Health Interviews Conducted Alone, in the Presence
of an Adult, a Child or Both Affects Adolescents’ Reporting
of Psychological Symptoms and Risky Behaviors

Aubrey V. Herrera1 • Corina Benjet1 • Enrique Méndez1 • Leticia Casanova1 •

Maria Elena Medina-Mora1

Received: 26 November 2015 / Accepted: 6 January 2016 / Published online: 20 January 2016

� Springer Science+Business Media New York 2016

Abstract The normative process of autonomy develop-

ment in adolescence involves changes in adolescents’

information management typically characterized by

decreasing disclosure and increasing concealment. These

changes may have an important impact on the early detection

and timely treatment of mental health conditions and risky

behavior. Therefore, the objective was to extend our under-

standing of how these developmental changes in adolescent

disclosure might impact adolescent mental health interviews.

Specifically, we estimated the effects of third party presence

and type of third party presence (adult, child, or both) on

adolescents’ reports of psychiatric symptoms, substance use,

suicidal behavior, and childhood adversity. In this represen-

tative sample of 3005 adolescents from Mexico City (52.1 %

female), administered the World Mental Health Composite

International Diagnostic Interview (WMH-CIDI-A), adult

presence influenced reporting the most; in their presence,

adolescents reported more ADHD, parental mental illness

and economic adversity, but less panic disorder, PTSD, drug

use and disorder, and suicidal behavior. The presence of

children was associated with increased odds of reporting

conduct disorder, opportunity for drug use, parental criminal

behavior, neglect, and the death of a parent. While adolescent

information management strategies are normative and even

desirable as a means of gaining emotional autonomy, they

may also interfere with timely detection and treatment or

intervention for mental health conditions and risky behaviors.

Research and practical implications of these findings are


Keywords Adolescence � Privacy � Information
management � Social desirability � Mental health � Mexico


The increasing autonomy of adolescence has important

implications for adolescents’ information management; in

other words, how much, what, to whom and under what

conditions adolescents disclose or conceal information

(Campione-Barr et al. 2015; Tilton-Weaver 2014). Auton-

omy development through the realignment of parent–child

relationships in adolescence is achieved through a process of

lessening parental control, and reduced parental knowledge

and adolescent willingness to disclose (Keijsers and Poulin

2013). This is important because the developmental decrease

in disclosure and increase in concealment specifically rela-

ted to risky behaviors and psychological distress might limit

timely detection and treatment of mental disorders during a

stage of development with high risk for the onset of psy-

chiatric disorder.

Disclosure by adolescents is mediated by adolescents’

attitudes regarding what parents have a right to know (Chan

et al. 2015). U.S. adolescents have been found to feel more

obligated to disclose to parents prudential information (de-

fined as information that pertains to one’s comfort, safety

and health such as risky behaviors) than personal (defined as

that pertaining to privacy, preferences, and control over

one’s body) or conventional information (that pertaining to

social norms) (Smetana et al. 2006). However, adolescents

who engage in risky behavior tend to disclose less about

prudential behaviors than personal or multifaceted ones.

& Corina Benjet
[email protected]

Dirección de Investigaciones Epidemiológicas y

Psicosociales, Instituto Nacional de Psiquiatrı́a Ramón de la

Fuente, Calzada México-Xochimilco 101, San Lorenzo

Huipulco, 14400 México, D.F., Mexico


J Youth Adolescence (2017) 46:417–428

DOI 10.1007/s10964-016-0418-1

Adolescents fail to disclose prudential behaviors for fear of

disapproval or punishment (Smetana et al. 2009) whereas

they fail to disclosure personal information because they

consider such information legitimately under their personal

control and they do not feel obliged to share this domain of

information with parents (Smetana and Asquith 1994).

Additionally prior research has found that U.S. adolescents

disclose more to mothers than to fathers (Smetana et al.

2006) and more to mothers than to siblings (at least in early

adolescence, although this gap between mothers and siblings

narrows toward emerging adulthood; Campione-Barr et al.

2015). Both depressive symptomatology and risky behaviors

have been negatively associated to disclosure with both

parents and siblings (Campione-Barr et al. 2015; Kerr and

Stattin 2000; Laird et al. 2013).

Because adolescents do not seek treatment for them-

selves, detection and treatment of mental disorders and

risky health behaviors in adolescents requires detection by

an adult third party such as parent or teacher and thus

largely depends on adolescent disclosure and information

management. To evaluate sensitive topics with adolescents

in clinical settings and in research, including health risk

behavior, mental disorders, and adverse experiences, self-

reporting is commonly used to gather this information.

However, questions regarding adolescent self-report

stem from numerous findings of commonly discrepant

reports between adolescents and their parents (e.g., Jensen

et al. 1999; Rescorla et al. 2013; Salbach-Andrae et al.

2009; Seiffge-Krenke and Kollmar 1998). Most of these

aforementioned studies find that parents report more

externalizing symptoms, that adolescents report more

internalizing symptoms and that there are greater discrep-

ancies for externalizing than internalizing symptoms.

Conclusions from this research have suggested that neither

adolescent nor parent report should be considered true or

untrue, but rather that both informants likely contribute

useful albeit discrepant information related to clinically

meaningful conditions and associated impairment. The

possible exceptions are ADHD and oppositional defiant

disorder for which adolescents’ reports alone are found to

have less clinical validity (Jensen et al. 1999).

Why might adolescents or persons of any age use infor-

mation management strategies involving incomplete dis-

closure or concealment? When asked sensitive questions,

people might misreport information because they view the

questions as intrusive, worry that disclosure of the infor-

mation might bring negative consequences, or want to pre-

sent themselves in a more favorable light by responding in a

socially desirable manner (Holtgraves 2004; Johnson and

Van de Vijver 2002; Krumpal 2013; Tourangeau and Yan

2007). This misreporting or concealing of information for

sensitive topics is a notable source of error attributed to bias

in survey estimates, may also be relevant in clinical settings,

and is largely dependent on the situation (Krumpal 2013;

Tourangeau and Yan 2007). When filtering responses to be

more desirable or acceptable, respondents aim to avoid

embarrassment when speaking with either the interviewer or

a third party bystander who may be present (Holtgraves

2004; Krumpal 2013; Mneimneh 2012; Tourangeau and Yan

2007). If the bystander is already aware of the information

being asked and can serve as a ‘‘truth control,’’ the

respondent will be less inclined to report dishonestly

(Mneimneh 2012; Tourangeau and Yan 2007). Anonymous

reporting is an option in some cases, and several studies

suggest that anonymity increases truthful responses for

sensitive information (Krumpal 2013; Ong and Weiss 2000;

Tourangeau and Yan 2007). Contrasting this, anonymity

cannot be assumed to automatically increase the quality of

reporting, because confidential questionnaires may give the

same results as anonymous questionnaires (van de Looij-

Jansen et al. 2006). Additionally, a 2002 study by Newman

et al. reports that taking the human interviewer out of the

picture may reduce the chance that respondents will report

psychological symptoms due to the impersonal nature.

Another influence on self-reports is the respondent’s

culture. It dictates both what is deemed a desirable

response and the privacy, or lack thereof, of the interview

setting (Diop et al. 2015; Hofstede et al. 2010; Mneimneh

2012; Pollner and Adams 1994). For example, it is more

common in a collectivist culture for a family member or

friend to invade privacy, compared to an individualistic

culture, where privacy is more prioritized (Hofstede et al.

2010). Cultural norms may dictate how easy or difficult it

is for an interviewer to ask for privacy during the interview

(Mneimneh 2012). While Tilton-Weaver (2014) found that

relationship dynamics and adolescent delinquent behaviors

play a role in adolescents’ information management, she

suggested that future research should sample more diver-

gent cultures where strong expectations to conform to the

rules and expectations of parents and other adults might

affect this relationship. In one study with Japanese ado-

lescents, levels of adolescent disclosure to parents, and

perceptions of the obligation to disclose to parents were

less than typically found in U.S. adolescents (Nucci et al.

2014). These authors interpreted these results to reflect a

cultural tendency in Japan to view adolescents’ actions as

matters of personal responsibility. One of the few studies of

a Latino population, Puerto Rican adolescents in the U.S.,

showed that greater adherence to Latino family values was

associated with more disclosure and less lying to mothers

(not so with fathers) regarding prudential issues, but not

peer-multifaceted issues (Villalobos and Smetana 2012). A

comparison of U.S. adolescents from Mexican, Chinese

and European backgrounds found Mexican–American

youth to disclose risky prudential behaviors less to mothers

than European American teens (Yau et al. 2009). The

418 J Youth Adolescence (2017) 46:417–428


authors concluded that, since Mexican culture emphasizes

more conformity to external standards, that these adoles-

cents may have felt that they had more at stake than other

adolescents when violating parental rules.

Interview settings with the presence of a third party have

shown an association with how interviewees respond to

sensitive topics, especially underreporting or over reporting

of sensitive information (Diop et al. 2015; Krumpal 2013;

Taietz 1962; Tourangeau and Yan 2007). Information man-

agement in adolescence is likely to be more pronounced than

in other stages of life given the developmental process

toward autonomy on the one hand and the power of adults

over adolescents to punish on the other. Adolescents inter-

viewed at home, at school, or in a clinical setting, may have

varying willingness to disclose sensitive information

depending on the context. At home, typically in the presence

of third parties who may have the power to punish the ado-

lescent or disprove of certain behaviors, typically a parent,

adolescents tend to underreport risky health behaviors, such

as alcohol and substance use (Aquilino et al. 2000; Brener

et al. 2006; Hoyt and Chaloupka 1994; Kann et al. 2002;

Krumpal 2013). In Aquilino et al.’s 2000 study, the presence

of a sibling showed fewer effects on reporting. In these

general population household surveys, interviewing the par-

ticipants privately is often not possible. This problem may be

even more pronounced in developing countries where the

number of family members in the household may be larger,

including extended family members, while at the same time

the size of the home smaller with fewer private spaces

(Aquilino 1997; Mneimneh 2012). While reporting infor-

mation at school, in the presence of peers or friends, the

literature shows an over reporting of these risky health

behaviors, in order to gain acceptance or seem ‘‘cool’’

(Brener et al. 2006; Gfroerer et al. 1997; Hoyt and Chaloupka

1994; Kann et al. 2002). However, Davis et al.’s 2010 study

reports the opposite and finds that adolescents understate

alcohol consumption, in order to present themselves more

favorably. In clinical settings, it is usually possible to inter-

view the adolescent in private, where more confidentiality is

reassured, and thus, they may report more openly (Gans and

Brindis 1995).

Like risky health behaviors, mental disorders are a sen-

sitive, and often stigmatized, issue. For mental disorders,

there are no reliable biomarkers to diagnose disorders and,

thus, the use of standardized diagnostic interviews is the

main form of diagnosis (Kessler and Ustun 2004). With such

interviews, respondents may not want to report symptoms or

feel embarrassed to report them in the presence of a third

party (Epstein et al. 2001). However, Pollner and Adams

(1994) reported that the presence of a third party was not

related to reporting symptoms of possible mental disorders,

and responses likely reflected methodological, situational,

and cultural influences. Overall, there is scant literature

available concerning the impact of a third party presence,

such as a parent, a sibling or other family member, during

adolescent mental health interviews and none to our

knowledge in the adolescent population of Mexico City.

The Current Study

The objective of this article is to build on research in the area

of adolescent information management, extending our

understanding of how these developmental changes in ado-

lescent disclosure might impact adolescent mental health

interviews. More specifically, we estimate the effects of third

party presence on adolescents’ reports of psychiatric symp-

toms meeting diagnostic criteria, substance use, suicidal

behavior, and childhood adversity. Additionally, the effect is

examined by type of third party present (adult, child, or both).

Because the literature shows that adolescents often fail

to disclosure prudential information that they fear will get

them in trouble, we hypothesize that rates of psychiatric

disorders, substance use, and suicidality will be lower in

the presence of an adult, that rates of conduct problems and

substance use, will be higher when only children are pre-

sent, due to adolescents’ desire to show off in front of

peers, and that childhood adversities will be higher when

an adult is present, as parents are likely to already possess

this information and thus may serve as a ‘‘truth control’’.



Data were collected as part of the 2005 Mexican Adoles-

cent Mental Health survey, with 3005 adolescent partici-

pants, aged 12–17 years old, representative of the almost

two million adolescents living in Mexico City. They were

chosen from a stratified multistage area probability sample.

Census count areas, cartographically defined and updated

for the XII Population and Housing Census, were the pri-

mary sampling units for all strata. City blocks with prob-

ability proportional to size were chosen as the secondary

sampling units. The Kish method of random number charts

was used to randomly select one eligible adolescents from

each household, and the response rate was 71 %. As

described previously, the sociodemographic distribution

was similar to the distribution of the adolescent population

in Mexico City: one half female, nearly 80 % students,

two-thirds residing with both parents, and the socioeco-

nomic level of the parents typically being low (Benjet et al.


J Youth Adolescence (2017) 46:417–428 419



Adolescents and their parents were given both a verbal and

written explanation of the study, followed by the parent or

guardian providing signed informed consent and the ado-

lescent providing assent. Adolescent interviews were con-

ducted in their home by extensively trained lay

interviewers and each participant and their family were

offered information on local mental health services. The

Internal Review Board of the National Institute of Psy-

chiatry Ramon de la Fuente approved this project.


The computer-assisted, fully structured adolescent version of

the World Mental Health Composite International Diag-

nostic Interview (WMH-CIDI-A; Kessler et al. 2009;

Merikangas et al. 2009) was administered to participants.

This research tool collects sociodemographic and other

information to generate diagnoses of DSM-IV disorders

(anxiety, mood, substance and disruptive behavioral disor-

ders), other risk behaviors such as substance use and suicidal

behavior, and risk factors such as childhood adversity. The

interviewer read each question aloud to the participant and

input responses directly into the computer. This computer-

assisted version includes complex logical skip patterns such

that each particular question posed was chosen by the

computer based on the previous responses of the participant.

Consistency check systems were in place to ensure that

inconsistent responses were probed and corrected.

Socio-Demographic Variables

The socio-demographic variables examined in this study were

asked about in the WMH-CIDI-A. This includes the sex of the

participants, age, family constellation, student status, parental

education and parental income. Family constellation was cat-

egorized as living with both biological parents or not living

with both biological parents (as reported by the adolescent).

Participants were considered students if currently enrolled as a

student. The adolescents were asked about the educational

attainment of each of their parents which was then categorized

as none/primary (six or less years of education), secondary

(7–9 years of education), high school (10–12 years of educa-

tion) or college (thirteen or more years of education); the score

of the parent with the highest level of education was used.

Parents reported family income was categorized into tertiles.

Psychiatric Disorder

The WMH-CIDI-A begins with a screening section which

includes a few screening questions for each disorder. Ado-

lescents who respond having ever experienced the symptoms

in the screening section then are asked more in depth ques-

tions for each disorder for which they screened positive. This

article reports the psychiatric disorders for which participants

met diagnostic criteria according to the diagnostic and sta-

tistical manual of mental disorders, fourth edition (DSM-IV;

American Psychiatric Association 1994) in the prior

12 months. All disorders used organic exclusion rules as well

as hierarchy definitions in order to avoid double counting of

disorders in the same person. The disorders are grouped as

follows: anxiety disorders (panic disorder, generalized anx-

iety disorder, agoraphobia, social phobia, specific phobia,

separation anxiety disorder and posttraumatic stress disorder

(PTSD)), mood disorders (major depressive disorder and

bipolar disorder (I and II)), disruptive behavior disorders

(intermittent explosive disorder, oppositional-defiant disor-

der, conduct disorder and attention-deficit/hyperactivity dis-

order) and substance use disorders (alcohol abuse or

dependence and drug abuse or dependence). A comparison of

diagnosis based on the adolescent CIDI with blinded clinical

interviews using the K-SADS showed generally good diag-

nostic concordance (Kessler et al. 2009).

Substance Use

The substance use section of the WMH-CIDI-A asks ado-

lescents about their lifetime use (defined as consumption of

the substance at least once at any time in one’s life) of

alcohol and illicit drugs, opportunity to use alcohol and

drugs, and substance abuse and dependence. The illicit drugs

included marijuana, cocaine in any of its forms, tranquilizers

or stimulants used without a medical prescription such as

methamphetamine, and other substances (e.g., heroin, inha-

lants, LSD, etc.) that were grouped as ‘‘other drugs.’’ Par-

ticipants were asked about each category of drugs openly and

then presented with a list of numerous different street names

for these drugs. The questions regarding opportunity to use

alcohol and drugs were posed after the questions about

alcohol and drug use such that opportunities to use alcohol

and drugs referred to all the alcohol and drugs previously

presented to the respondents. All participants were asked

about opportunities regardless of whether they had previ-

ously endorsed consuming any substance. Opportunity to use

alcohol and opportunity to use drugs were asked about sep-

arately and defined as having the opportunity to use any

substance regardless of whether or not the respondent did so.

They were given as an example that someone offered them

drugs or that they were present when others were using and

could have done so if he or she chose to.

Suidicial Behavior

The suicidal behavior section had a somewhat different

format than the other sections of the WMH-CIDI-A such

420 J Youth Adolescence (2017) 46:417–428


that the interviewer did not read the questions aloud to the

adolescent, but rather asked the respondent to report

whether they had experienced any of three experiences

which were printed in a booklet and referred to by letter.

These experiences were ‘‘Have you ever seriously thought

about committing suicide?,’’ ‘‘Have you ever made a plan

for committing suicide?,’’ and ‘‘Have you ever attempted


Childhood Adversity

Twelve different types of childhood adversity were evalu-

ated and classified as present or not present using the same

criteria as the World Mental Health Survey Initiative (Green

et al. 2010). These included parental mental illness, parental

substance problems, parental criminal behavior, witnessing

family violence, physical abuse, sexual abuse, neglect, death

of a parent, parental divorce, other parent loss, life-threat-

ening physical illness, and economic adversity. Parental

pathology used questions from the family history research

diagnostic criteria interview and included parental mental

illness, substance problems, and criminal behavior (Endicott

et al. 1978). Physical abuse and witnessing family violence

were evaluated with a modified version of the Conflict

Tactics Scale (Strauss 1979). Neglect was assessed with

questions often used in child welfare studies (Courtney et al.

1998). Sexual abuse was determined by reading a definition

of rape and a definition of other forms of abuse or

molestation. To be consistent with the other World Mental

Health Surveys, chronic sexual abuse was defined as

reporting at least three episodes of sexual abuse thus rep-

resenting chronic sexual abuse as opposed to acute one time

trauma (Kessler et al. 2010). To assess parental loss, the

adolescents were asked whether they lived with both of their

parents all of their lives. Those who did not were asked

whether this was because their parents had separated or

divorced, a parent had died or some other reason. Those

endorsing separations of 6 months or more from either

parent for some other reason were classified as other par-

ental loss, with reasons ranging from having gone to

boarding school, having left home, or that their parent was in

prison. Physical illness is based on the adolescent’s report of

having experienced a life-threatening physical illness.

Family economic adversity was determined by the family

ever receiving money from a government assistance pro-

gram for poor families or by lack of parental employment

most or all of the time during the participant’s childhood.

Third Party Presence

An interviewer observation section at the end of the WMH-

CIDI-A, required the interviewer to register if anyone was

present during the interview aside from the adolescent, and

if so who. Those responses were categorized as follows: (1)

no one, (2) only children, (3) children and adults, or (4)

adults only.

Statistical Analysis

Data were weighted to adjust for non-response, differential

probabilities of selection, and post-stratification to the

adolescent population of Mexico City according to the year

2000 Census in the target sex and age range. The preva-

lence of reported DSM-IV diagnoses, substance use, sui-

cidal behavior and childhood adversity were calculated by

privacy category. Then, age- and sex-adjusted odds ratios

were calculated in SUDAAN software (Research Triangle

Institute 2009) for the association of privacy with the DSM-

IV diagnoses, substance use, suicidal behavior and child-

hood adversity.


Only 42.7 % of interviews were conducted with the adoles-

cent alone, whereas children only were present in 3.4 % of

interviews, both children and adults were present in 4.8 %,

and at least one adult was present in 49.1 % of interviews. Of

the adults that were present 84.6 % was a parent. Sociode-

mographic characteristics are depicted in Table 1. Third

party presence varied by sex, age, and parental education. A

greater proportion of interviews with females, versus males,

had only children present; families with lower parental

education tended to have children present, as well. More

interviews were conducted alone when the interviewee was

an older adolescent versus a younger adolescent.

The association of third party presence with DSM-IV

diagnosis adjusted by age and sex is shown in Table 2.

Having only children present increases the probability of

reporting any disruptive behavior disorder (OR 1.74; 95 %

CI 1.21–2.51), which is explained primarily by the threefold

odds of reporting a conduct disorder in particular (CI

1.17–7.65). The presence of both children and adults also

increases the probability of reporting any disruptive behav-

ior disorder (OR 1.80; 95 % CI 1.05–3.10), but is explained

by the increased odds of reporting intermittent explosive

disorder (OR 2.35; 95 % CI 1.30–4.22). Having only adults

present increases the odds of reporting attention-deficit/hy-

peractivity disorder (ADHD) by 83 % (95 % CI 1.08–3.10)

but reduces the odds of reporting panic disorder (OR 0.62;

95 % CI 0.40–0.95), PTSD (OR 0.55; 95 % CI 0.35–0.86),

and a drug use disorder (OR 0.17; 95 % CI 0.07–0.38).

Table 3 presents the association of substance use and

third party presence, adjusted by age and sex. With only

children present, there are increased odds of reporting a

drug opportunity (OR 1.73; 95 % CI 1.10–2.70). Having

J Youth Adolescence (2017) 46:417–428 421


both children and adults present is unrelated to reporting of

substance use. However, when only adults are present,

odds are reduced by roughly half for reporting drug use

(OR 0.47; 95 % CI 0.31–0.73).

The association between suicidal behavior and third

party presence is shown in Table 4. Having children only

or children and adults present was unrelated to reporting of

suicidal behavior. However, having only adults present

reduces the odds of reporting suicidal behavior with ORs

ranging from 0.48 for suicide attempts to 0.68 for suicidal


Table 5 depicts the association between childhood

adversity and third party presence. With only children

present, there is a greater probability of reporting the fol-

lowing: parental criminal behavior, neglect, and parent

death with ORs ranging from 1.88 for death of a parent to

2.44 for parental criminal behavior. When children and

adults are present, there is a greater probability of reporting

parental mental illness (OR 2.14; 95 % CI 1.34–3.41).

With only adults present, there is a greater probability of

reporting both parental mental illness (OR 1.39; 95 % CI

1.13–1.71) and economic adversity (OR 1.22; 95 % CI



The normative process of autonomy development in ado-

lescence involves changes in adolescents’ information

management typically characterized by decreasing disclo-

sure and increasing concealment of information (Campi-

one-Barr et al. 2015; Tilton-Weaver 2014). These changes

may have an important impact on the early detection and

timely treatment of mental health conditions and risky

behavior. These developmental changes in information

management depend on various factors including the

domain of the information managed and the person to

whom one discloses (mother, father, sibling, peer, other

adult) (Smetana et al. 2006; Campione-Barr et al. 2015).

While previous studies have found that less disclosure and

increased concealment is related to delinquent behavior

(Campione-Barr et al. 2015; Kerr and Stattin 2000; Laird

et al. 2013) and that there are important discrepancies

between children’s and parental reports of psychological

distress and risky behavior (Jensen et al. 1999; Rescorla

et al. 2013; Salbach-Andrae et al. 2009: Seiffge-Krenke

and Kollmar 1998), the purpose of this investigation was to

understand how the presence of third parties during

Table 1 Sociodemographic …

Profiles of Youths With PTSD and Addiction

Yvonne M. Chasser

Massachusetts General Hospital, Boston, MA, USA

Objective: Understanding the relationship between PTSD and addiction in adolescents may
dramatically improve evidence-based practice in child psychiatry. We hypothesized that in a sample
of substance addicted youth, PTSD would correlate with (1) female gender and racial minority
status, (2) preference for anxiolytic substances, (3) higher burden of self injury and suicide attempts,
and (4) earlier age of first use. Methods: One-hundred and ninety-five adolescents (52% female, ages
14–18) were court-referred to residential treatment and assessed at intake. Multi-informant data
regarding Axis I diagnostic status and other clinical variables were collected via rater-administered,
semi-structured interviews; medical chart review; and youth, parent, and clinician reports.
Differences between subjects were evaluated using Fisher’s exact test for binary variables or
Kruskal-Wallis Chi-Square Test. Results: Substance dependent youth with comorbid PTSD were
significantly more likely to be female and Latino. PTSD was correlated with preference for alcohol,
narcotics, tranquilizers and inhalants. In addition, PTSD was correlated with higher burden of self-
injury and suicidal behavior. Youth with PTSD were also more likely to have begun using before age
13 and reported that PTSD symptoms preceded first use. Conclusions: Girls and Latinos may be
particularly vulnerable in developing comorbid substance dependence and PTSD. Drug preferences
among youth with PTSD may reflect tendencies to target symptoms such as hypervigilance and
anxiety. The recognition and early intervention among youth with trauma could prevent early first
use and eventual substance dependency.

addiction; adolescent; post-
traumatic stress disorder;
substance use disorder


In the United States, nearly 5% of youths ages 12 to 17
meet diagnostic criteria for illicit drug abuse or depen-
dence (Substance Abuse and Mental Health Services
Administration [SAMHSA], 2011). Youths ages 12 to 17
with substance use disorders (SUDs) exhibit increased
rates of risky behaviors, poor performance in school,
poor overall health, increased likelihood of continuing to
abuse alcohol and other drugs (AOD) into adulthood,
and early mortality (Battin-Pearson et al., 2000; Feigel-
man & Gorman, 2010; Thompson & Auslander, 2011).
Post-Traumatic Stress Disorder (PTSD) is also common
among youths and further complicates the problem of
SUD. While epidemiological studies indicate that 5% of
adolescents have met the criteria of PTSD in their life-
time (Merikangas et al., 2010), the rate of PTSD is as
much as five times higher among adolescents in treat-
ment for SUD, and is associated with more costly AOD
treatment and a worse prognosis (Brady, Killeen, Saladin,
Dansky, & Becker, 1994; Brown, Stout, & Mueller, 1999;
Funk, McDermeit, Godley, & Adams, 2003).

The emerging field of adolescent addiction psychia-
try has limited data regarding the clinical profile of
youths with comorbid SUD and PTSD. Early research
shows female gender is correlated with comorbid
PTSD and SUD (Montoya, Covarrubias, Patek, &
Graves, 2003). Research in adult women with SUD
has shown higher rates (30% to 59%) of comorbid
PTSD compared to men due to higher burden of
childhood physical or sexual abuse (Blume & Zilber-
man, 2004). This pattern has been shown to replicate
among adolescents with addiction; substance-depen-
dent girls have higher rates of PTSD than boys with
SUD due to higher rates of traumatic events, includ-
ing sexual abuse and rape (Najavits, Weiss, & Shaw,
1997). Race and ethnicity may also differentiate indi-
viduals with comorbid PTSD and SUD. Latino and
African-American youths have been shown to have
significantly higher rates of current and lifetime
PTSD (Kilpatrick & Saunders, 1997). Studies with
substance-dependent youths have found higher rates
of internalizing mental disorders such as PTSD and

CONTACT Yvonne M. Chasser [email protected] Massachusetts General Hospital, 55 Fruit Street, Department of Psychiatry, Wang 8, Boston,
MA 02135, USA.

Color versions of one or more of the figures in the article can be found online at
© 2016 Taylor & Francis Group, LLC

2016, VOL. 25, NO. 5, 448–454

depression among Latino, African-American, and
mixed-race youths (Chisolm, Mulatu, & Brown,

In addition, little is known about the drug use patterns
and self-injurious behaviors in which youths with
comorbid PTSD and SUD engage. Studies of adolescents
without addiction show PTSD alone is associated with
higher rates of self-injurious and suicidal behaviors com-
pared to control groups (Lubman, Allen, Rogers, Cemen-
ton, & Bonomo, 2007; Nooner et al., 2012). Moreover,
active substance abuse is an independent risk factor for
suicide in adolescents (Bukstein et al., 1993). However,
no study to date has determined the overlay of PTSD in
risk of self-injurious and suicidal behaviors among
youths with addiction.

Given the complexity of studying SUDs and PTSD, no
unifying theory would be sufficient to suggest an expla-
nation for the relationship between these multifaceted
issues. The most popular hypothesis currently is the self-
medication theory (Khantzian & Albanese, 2008), which
proposes that addiction is the resultant attempt to reduce
psychological distress caused by the primary underlying
psychiatric cause, such as PTSD. Other theories point to
the environment of drugs and drug culture as the pri-
mary cause by predisposing individuals to exposure to
trauma when in a more vulnerable state.

This study explores the demographic characteristics,
drug use patterns, and behaviors that distinguish youths
with comorbid PTSD and SUD in a gender-balanced
sample of 195 juveniles, all diagnosed with addiction.
Based on prior research, we hypothesize that PTSD will
be associated with female gender, preference for anxiety-
reducing substances such as alcohol and benzodiaze-
pines, a history of physical and sexual abuse, a history of
increased suicidal and self-injurious behavior, and earlier
age of first use. This information can hasten recognition
of comorbid PTSD and SUD, inform relapse prevention
approaches of triggers associated with preferred drugs,
and better tailor treatment resources for improved


Our study aims are to characterize (a) the demographic
features of dually diagnosed adolescents with PTSD and
SUD compared to SUD alone, (b) the impact of PTSD
on drug preferences among adolescents, and (c) the risk
for self-injurious behaviors associated with dual diagno-
sis compared to SUD alone.

We hypothesize that female gender and belonging to a
minority racial/ethnic group would correlate with dual
diagnosis based on prior research. We also hypothesized
that PTSD would correlate with preference for anxiolytic

drugs such as alcohol and benzodiazepines based on the
known pharmacologic effects of these drugs. Finally, we
hypothesized that PTSD would correlate with earlier age
of first use and a history of physical and sexual abuse, as
well as higher rates of self-injury and suicide attempts.

Material and methods

This was a retrospective cross-sectional study of 195
youths, all who met the Diagnostic and Statistical Man-
ual of Mental Disorders, fourth edition (DSM-IV) diag-
nosis of substance addiction, which today has been
reclassified under Substance Use Disorder (SUD) in the
DSM-5. Therefore, any mention in our study of sub-
stance addiction shall be referred to as a SUD.


Recruitment for this study was conducted from February
2007 to August 2009 at a residential treatment facility in
northeast Ohio for substance-dependent adolescents. All
participants were ages 14 to 18 years, English speaking,
had a stable address and telephone, met the DSM-IV
diagnosis of addiction (which has been reclassified as
SUD under DSM-5), and were medically stable. Exclu-
sion criteria included a major chronic health problem
other than substance use likely to require hospitalization,
currently suicidal or homicidal, and expected incarcera-
tion in the subsequent 12 months. Eligible subjects
signed statements of informed consent/assent after
receiving an invitation to participate along with an infor-
mation packet. Ninety-minute baseline interviews were
conducted within the initial 10 days of treatment and
repeated at discharge after an average of 2.2 months of
residential treatment. Subjects were paid $25 for com-
pleted assessments. All procedures of this study were
approved by the University Hospitals/Case Medical Cen-
ter Institutional Review Board for human investigation,
and a Certificate of Confidentiality from the National
Institute on Alcohol Abuse and Alcoholism was
obtained. A more detailed description of study methods
can be found in Kelly, Stout, Pagano, and Johnson
(2011) and methods are also discussed briefly here.

The residential treatment facility provides a range of
evidence-based therapeutic modalities for adolescents
struggling with an SUD, including cognitive behavioral
therapy, motivational enhancement therapy, reality ther-
apy, adolescent community reinforcement approaches,
gender-specific treatment, medication-assisted treat-
ment, relapse prevention, family, individual, and group
therapies, and assertive continuing care (aftercare). Typi-
cally, residents spend approximately 20 hours per week
in therapeutic activities.



Data were collected via rater-administered, semi-struc-
tured interviews, medical chart review, and youth, par-
ent, and clinician reports. Semi-structured interviews
were conducted in person by experienced clinical inter-
viewers whose training and certification ranged from
bachelor’s level to doctor of medicine. All interviewers
received rigorous training and met regularly for updates
and retraining on diagnostic assessments using the Mini
International Neuropsychiatric Interview Plus (MINI-
Plus). All individuals involved in data collection from
subjects completed National Institutes of Health-
required courses on human subjects’ protection. Demo-
graphic characteristics, SUDs, age of first use, psychiatric
comorbidities, and clinical variables were assessed at
baseline. Depending on the type of variables (continuous
or discrete), the Fisher’s exact test for binary variables or
Kruskal-Wallis chi-square test for continuous variables
was performed to evaluate differences between subjects.

Demographic characteristics

Demographic characteristics including age, gender, race,
ethnicity, single-parent household status, and parental
education attainment were gathered. These results are
reported in Table 1.

Axis I diagnoses

The variables relating to Axis 1 diagnoses were assessed
as follows. The existence of SUD and type of SUD are
reported in Table 2. These variables were measured using
the rater-administered MINI International Neuropsychi-
atric Interview Plus. The MINI-Plus is a comprehensive
diagnostic semi-structured interview that covers a wide
range of diagnoses, including SUDs and PTSD. The
MINI-Plus provides continuous measures such as age of
first use and age of onset of SUD, as well as severity
measures based on symptom counts for disorders such
as PTSD and substance dependence. The MINI has

shown 85% agreement with expert psychiatric opinion
and has demonstrated high concordance with alcohol
and drug consumption as measured by the Alcohol,
Smoking, and Substance Involvement Screening Test
(ASSIST; Sheehan et al., 1998; WHO ASSIST Working
Group, 2002).

Clinical variables

Five clinical variables were assessed: history of suicide
attempt, deliberate self-harm, sexual abuse, physical
abuse, and preteen first use. Two items from the Sched-
ule for Nonadaptive and Adaptive Personality were used
to measure youth history of attempted suicide and delib-
erate self-harm. The Schedule for Nonadaptive and
Adaptive Personality is a self-report questionnaire com-
posed of 375 true/false items designed to assess diagnos-
tic, temperament, and pathological and nonpathological
traits that has shown good psychometric properties with
young adult populations (Klonsky, Oltmanns, & Tur-
kheimer, 2003). The two items were (a) “I have tried to
commit suicide” and (b) “I have hurt myself on purpose
several times.” Participants who endorsed the first item
were considered to have a history of attempted suicide.
Those who endorsed the second item were considered to
have a history of intentional self-harm. In the current
study, physical abuse referred to any behavior that is

Table 1. Sample demographics at intake.

Total PTSD

Demographic Characteristic 195 (100%) NO 158 (81%) YES 37 (19%)

Male Yes 93 (48%) 85 (54%) 8 (22%)���

Age M, SD 16.2 (1.1) 16.2 (1.0) 16.3 (1.2)
Racial Minority Yes 60 (31%) 50 (32%) 10 (27%)
Latino Yes 15 (8%) 9 (6%) 6 (16%)�

Single-Parent Household Yes 98 (50%) 78 (49%) 20 (54%)
Parental Education HS diploma or less 87 (45%) 72 (45%) 15 (40%)

Some college 55 (28%) 47 (30%) 8 (22%)
BAC 53 (27%) 39 (25%) 14 (38%)

�p < 0.05. ���p < 0.001.

Table 2. PTSD and drug dependence.


Total No Yes
Drug Dependence 195 (100%) 158 (81%) 37 (19%)

Alcohol Yes 118 (61%) 90 (57%) 28 (76%)�

Nicotine Yes 193 (99%) 156 (99%) 37 (100%)
Cannabis Yes 179 (92%) 146 (92%) 33 (89%)
Stimulant Yes 49 (25%) 37 (23%) 12 (32%)
Cocaine Yes 50 (26%) 38 (24%) 12 (32%)
Narcotics Yes 58 (30%) 41 (26%) 17 (46%)�

Inhalants Yes 11 (6%) 6 (4%) 5 (14%)�

Tranquilizer Yes 39 (20%) 25 (16%) 14 (38%)��

Hallucinogen Yes 57 (29%) 43 (27%) 14 (38%)

�p < 0.05. ��p < 0.01.


intended to hurt an individual directly using physical
force. Sexual abuse referred to the use of force or threats
to compel a person to engage in any type of sexual activ-
ity against their will or without their understanding,
whether the act is completed or not. Physical and sexual
abuse were assessed by a trained chemical dependency
counselor by asking, “Have you ever experienced sexual
or physical abuse?” with standardized follow-up ques-
tions for positive responses. Preteen first use refers to age
of first use of the drug of choice before age 13.


Demographic characteristics

Table 1 displays the baseline demographic characteristics
of our sample of 195 substance-dependent youths. We
have then further divided each category in the demo-
graphic breakdown on the basis of comorbid PTSD

Our sample of 195 youths contained 37 individuals
with PTSD and 158 without. The sample contained
nearly equal proportions of male (48%) and female
(52%) with an average age of 16.2 years. Female gender
was significantly correlated with a dual diagnosis of
addiction and PTSD. Regarding our sample’s representa-
tion of racial minorities, 31% of subjects belonged to a
racial minority. Our analysis showed that Latino ethnic-
ity was significantly correlated with a dual diagnosis of
addiction and PTSD.

Substance use disorders

Table 2 shows the different types of SUDs in this sample
as assessed in the MINI. Again, within each category of
drug preference, we distinguish individuals with comor-
bid PTSD versus those without. For the entire sample,
the most common type of SUD was nicotine dependency
(99%). and then marijuana dependency (92%) followed
by alcohol dependency (61%). PTSD was significantly
correlated with addiction to alcohol, tranquilizers, inha-
lants, and narcotics. For the other substances, no signifi-
cant difference was found between the proportion of
addicted youths based on their PTSD status.

Clinical variables

Table 3 documents the prevalence of suicide attempts,
self-injury, physical abuse, sexual abuse, and preteen first
use among substance-addicted youths with and without
comorbid PTSD. PTSD was significantly correlated with
all of these variables.

Onset ordering

Figure 1 shows that among substance-dependent youths
with PTSD symptoms, 59% had PTSD symptoms pre-
cede their first use, 11% had simultaneous onset, and the
remainder had PSTD symptoms following their first use.


This was a retrospective cross-sectional study, which
compared the characteristics of 195 substance-addicted
youths, based on comorbid PTSD status. We found that
among substance-addicted youths, comorbid PTSD was
associated with the following:
� female gender,
� Latino ethnicity,
� preference for alcohol, tranquilizers, inhalants, and

� history of physical abuse, sexual abuse, self-harm,
suicide attempts, and

� preteen first use.
These findings offer valuable insights into a mini-

mally studied area. Understanding the profile of
dually diagnosed youths can assist with developing
more tailored approaches to prevention, recognition,
and treatment.

Our finding that dual diagnosis is associated with
female gender is congruent with preliminary data, which
have shown higher rates of PTSD among substance-
addicted girls compared to boys. The same previous
research concluded that this difference results from
higher rates of sexual trauma and rape among female
youths, leading to substance use and eventually addic-
tion, which may be in order to self-medicate anxiety and
negative affect (Blume & Zilberman, 2004). Another
explanation may be that if young females are using drugs
when they experience a trauma, they are more likely
than their male counterparts to develop PTSD. Future
research on substance-addicted youths is needed to dis-
cern whether resiliency characteristics of the individual
or environmental factors determine the concomitant
development of PTSD.

Table 3. PTSD and clinical variables.


Total No Yes
Clinical Variable 195 (100%) 158 (81%) 37 (19%)

Suicidal Attempt(s) Yes 47 (24%) 33 (21%) 14 (38%)�

Self-Injury Yes 69 (35%) 50 (32%) 19 (51%)�

Physical Abuse Yes 45 (23%) 29 (18%) 16 (43%)��

Sexual Abuse Yes 50 (26%) 31 (20%) 19 (51%)���

Preteen First Use Yes 67 (34%) 46 (29%) 21 (57%)��

�p < 0.05. ��p < 0.01. ���p < 0.001.


The association between Latino ethnicity and dual
diagnosis is a particularly interesting finding and adds to
the nascent body of research demonstrating that Latinos
are at higher risk for PTSD than any other racial or eth-
nic group in the United States (Pole, Best, Metzler, &
Marmar, 2005). There was a significantly higher propor-
tion of Latinos in our sample with PTSD (16%) com-
pared to the proportion of Latinos in the sample without
PTSD (6%). Suggested explanations for the higher repre-
sentation of Latinos with dual diagnosis include greater
environmental exposure to trauma, increased peri-
trauma dissociation, and culturally informed maladap-
tive coping mechanisms such as fatalismo.(Fatalismo is a
philosophical doctrine holding that all events are prede-
termined in advance for all time and human beings are
powerless to change them.) Although the underlying rea-
son remains to be further explored, our data support the
data concerning the particular burden of PTSD among
Latinos and warrant further study to understand this

Our data show that having a diagnosis of PTSD is
associated with addiction to alcohol, tranquilizers, inha-
lants, and narcotics. It is interesting to note that among
the drug classes in our study, the only three that uniquely
share similar effects including decreased anxiety, aware-
ness, and sensitivity to pain, correlated with a diagnosis
of PTSD. This suggests that teens with PTSD are prefer-
entially using alcohol, tranquilizers, and narcotics to self-
medicate symptoms of PTSD such as hyperarousal, anxi-
ety, and pain resultant from traumatic experiences. The
association between PTSD and addiction to inhalants is
more difficult to classify, as this is a broad category of
substances with effects that can range from excited to
depressed states and can include hallucinatory states as
well. Understanding which substances youths use is

essential to tailoring recovery programs, as the reasons
for use and triggers to relapse differ greatly.

PTSD was significantly associated with a history of
suicide attempt and a history of self-injurious behavior,
which agrees with the current conception that both
addiction and PTSD independently indicate increased
risk for self-harm. The final clinical variable assessed was
preteen use, which was also found to be associated with
PTSD. In the majority of cases, adolescents with PTSD
reported that their PTSD symptoms preceded first use or
occurred concomitantly. These results imply that early
first use would not be an appropriate point of detection
and intervention in these cases, as it appears only after a
significant amount of trauma has already been done.
Rather, early detection of trauma or signs of PTSD could
prevent preteen first use, as well as the devastating conse-
quences on physical and mental health.

This study is the largest gender-balanced study of sub-
stance-addicted youths to date and demonstrates several
important trends within this population relating to
comorbid psychiatric problems. Although there have
been many studies of substance use in youths, there is a
dearth of work looking at the problem of early use in
preteens and psychiatric comorbidities. In addition, very
few studies have examined the problem of addiction in
youths and have instead documented use of alcohol and
illicit drugs. Because this is a clinical treatment-seeking
population rather than a nonclinical community sample,
we are able to draw stronger conclusions about the rela-
tionships between SUDs and psychiatric concerns due to
the more concentrated pathology in our sample. In addi-
tion, our study uniquely provides the breakdown of
which substances youths were addicted to, whereas the
majority of other studies do not make this categorical

Figure 1. PTSD symptoms versus first use.



Our study was a retrospective cross-sectional design,
examining clinical differences among a sample of addic-
tion treatment-seeking adolescents with a median age of
16.2. Because we surveyed these youths several years after
the onset of their PTSD symptoms and first use, it is pos-
sible our data was affected by recall bias. Youths with a
history of trauma and substance dependence may indeed
have difficulty remembering the order of events, which
occurred several years prior to seeking treatment. Youths
with a history of trauma may also be more likely to recall
the traumatic event with increased perception of causal-
ity. Moreover, youths with a history of trauma preceding
first use may be more likely to present for treatment due
to greater impairment in psychosocial functioning and
consequently be overrepresented in our sample.


Although complex, understanding the relationship between
PTSD and addiction in youths is essential to improving
adolescent mental health. It appears that in substance-
addicted youth populations, PTSD is associated with female
gender and Latino ethnicity. Previous work has attributed
these trends to environmental factors such as increased
rates of sexual trauma for girls and individual factors such
as culturally influenced coping mechanisms for Latinos.
The drug preferences among substance-addicted youths
with PTSD reflect a self-medication pattern and perhaps
socioeconomic influences as well. Recognizing the pro-
file of dual-diagnosis youths may lead to earlier inter-
vention and understanding the type of substance used
will help tailor treatments for more successful recovery.
According to our findings, PTSD signifies a higher over-
all psychological burden within an already distressed
population, including earlier use of drugs and higher
rates of attempted suicide and self-injury. These obser-
vations have important implications for researchers and
clinicians working to better understand the individuals
they care for and develop programs for prevention and


Grateful acknowledgment is given to Dr. Nina Rytwinski for
her dedication in editing several drafts of this work as well as
redirecting the central thesis, and to Dr. Maria Pagano for her
guidance and commitment to mentorship in research.


Yvonne M. Chasser


Battin-Pearson, S., Newcomb, M., Abbott, R., Hill, K. G., Cata-
lano, R. F., & Hawkins, J. D. (2000). Predictors of early high
school dropout: A test of five theories. Journal of Educa-
tional Psychology, 92, 568–582.

Blume, S. B., & Zilberman, M. L. (2004). Addiction in
women. In M. Galenter & H. D. Kleber (Eds.), The
American Psychiatric Publishing textbook of substance
abuse treatment (3rd ed., pp. 539–546) Washington,
DC: American Psychiatric Publishing.

Brady, K. T., Killeen, T., Saladin, M. E., Dansky, B., & Becker,
S. (1994). Comorbid substance abuse and posttraumatic
stress disorder. American Journal on Addictions, 3, 160–164.

Brown, P., Stout, R., & Mueller, T. (1999). Substance use disor-
der and posttraumatic stress disorder comorbidity: Addic-
tion and psychiatric treatment rates. Psychology of Addictive
Behaviors, 13, 115–122.

Bukstein, O. G., Brent, D. A., Perper, J. A., Moritz, G., Baugher,
M., Schweers, J.,… Balach, L. (1993). Risk factors for com-
pleted suicide among adolescents with a lifetime history of
substance abuse: A case-control study. Acta Psychiatrica
Scandinavica, 88(6), 403–408.

Chisolm, D. J., Mulatu, M. S., & Brown, J. R. (2009). Racial/eth-
nic disparities in the patterns of co-occurring mental health
problems in adolescents in substance abuse treatment. Jour-
nal of Substance Abuse Treatment, 37(2), 203–210.

Feigelman,W., & Gorman, B. S. (2010). Prospective predictors
of premature death: Evidence from the National Longitudi-
nal Study of Adolescent Health. Journal of Psychoactive
Drugs, 42, 353–361.

Funk, R. R., McDermeit, M., Godley, S. H., & Adams, L.
(2003). Maltreatment issues by level of adolescent substance
abuse treatment: The extent of the problem at intake and
relationship to early outcomes. Child Maltreatment, 8(1),

Kelly, J. F., Pagano, M. E., Stout, R. L., & Johnson, S. M. (2011).
Influence of religiosity on 12-Step participation and treat-
ment response among substance-dependent adolescents.
Journal of Studies on Alcohol and Drugs, 72, 1000–1011.

Khantzian, E. J., & Albanese, M. J. (2008). Understanding
addiction as self medication: Finding hope behind the pain.
Lanham, MD: Rowman & Littlefield Publishers.

Kilpatrick, D. G., & Saunders, B. E. (1997). Prevalence and
consequences of child victimization: Results from the
National Survey of Adolescents, final report. Washing-
ton, DC: US Department of Justice, Office of Justice

Klonsky, D., Oltmanns, T., & Turkheimer, E. (2003). Deliber-
ate self-harm in a nonclinical population: Prevalence and
psychological correlates. The American Journal of Psychia-
try, 160, 1501–1508.

Lubman, D. I., Allen N. B., Rogers N., Cementon E., &
Bonomo, Y. (2007). The impact of co-occurring mood and
anxiety disorders among substance-abusing youth. Journal
of Affective Disorders, 103(1–3), 105–112.

Merikangas, K. R., He, J. P., Burstein, M., Swanson, S. A., Ave-
nevoli, S., Cui, L., … & Swendsen, J. (2010). Lifetime preva-
lence of mental disorders in the U.S. Adolescent
Comorbidity Survey Replication-Adolescent Sample. Jour-
nal of the American Academy of Child and Adolescent Psy-
chiatry, 49, 980–988.


Montoya, I. D., Covarrubias, L. D., Patek, J. A., & Graves, J. A.
(2003). Posttraumatic stress disorder among Latino and
African-American drug users. The American Journal of
Drug and Alcohol Abuse, 29(4), 729–741.

Najavits, L. M., Weiss, R. D., & Shaw, S. R. (1997). The
link between substance abuse and posttraumatic stress
disorder in women. The American Journal on Addic-
tions, 6, 273–283.

Nooner, K. B., Linares, L. O., Batinjane, J., Kramer, R. A., Silva,
R., & Cloitre, M. (2012). Factors related to posttraumatic
stress disorder in adolescence. Trauma, Violence, & Abuse,
13(3), 153–166.

Pole, N., Best, S. R., Metzler, T., & Marmar, C. R. (2005). Why
are Latinos at greater risk for PTSD? Cultural Diversity and
Ethnic Minority Psychology, 11(2), 144.

Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P.,
Janavs, J., Weiller, E., … & Dunbar, G. C. (1998). The Mini-

International Neuropsychiatric Interview (MINI): The devel-
opment and validation of a structured diagnostic psychiatric
interview for DSM-IV and ICD-10. Journal of Clinical Psychi-
atry, 59(Suppl 20), 22–33.

Substance Abuse and Mental Health Services Administra-
tion. (2011). 2010–2011 National Survey on Drug Use
and Health model-based estimates. Rockville, MD:
Department of Health and Human Services. Retrieved

Thompson, R. G., & Auslander, W. F. (2011). Substance use
and mental health problems as predictors of HIV sexual
risk …


Program Transcript


NARRATOR: Trauma is not restricted to age groups or gender. However, helping
children and adolescent clients who experience trauma is very different than
helping adults. Doctors John Sommers-Flanagan and Eliana Gil explain the
difference in working with children and adolescent clients who have experienced
trauma, how to assess them, and what techniques they have used to address the

JOHN SOMMERS-FLANAGAN: The Diagnostic and Statistical Manual refers to
the core symptoms of PTSD, or post-traumatic stress disorder, as exposure to a
traumatic event and intrusive recollection of that event and numbing and
avoidance experienced by the individual and kind of a hyper-arousal. And so I’m
wondering about that presentation as it initiates the referral process.

And in your work, Eliana, because I know you work a lot with traumatic problems
in youth, I wonder how you see those symptoms as they present to you within an
initial referral.

ELIANA GIL: It’s an interesting constellation of symptoms, because often you
have the hyper-arousal. Kids are actually having intrusive thoughts or nightmares
of very specific pictures of things that have happened to them and scared them.
Kids also are able to do, through play, some of the reenactments of things that
have occurred. And so looking at it from far way, you look at the play, and you
think, wow, what is that about?

Is that pretend? Is that fantasy? Or is that something that the child is actually
experiencing? But the play is very unusual. So there’s actually a lot of literature
on what’s called post-trauma play, which tends to be very different than generic
play, in that it’s very literal. It’s very robotic. Kids are really engaged in the play,
as they usually are, like with pretend talk or role-play or something like that.

And it’s play that’s very repetitive. And as kids do it, their affect is very guarded.
And when you encounter post-trauma play, you know that something is very
different here and that this is really a way that kids begin to show that they’re
living in the climate of the trauma. So definitely we see kids who come in
because there’s intrusive thoughts or memories, through nightmares in particular.
They’re waking up with night terrors.

And we see the post-trauma play that the kids are doing at home or sometimes in
a school setting. And then also, there is the child who appears with this very, very
flat affect, where they’re disengaged, they don’t do the regular things that kids do,
they don’t play, they’re not spontaneous, their social interactions are very

© 2016 Laureate Education, Inc. 1


unusual. I think that one of the insidious lessons of child abuse, talking about
interpersonal trauma, is that people who love you hurt you.

And so I think that what kids develop is this expectation that the world will not be
safe and that this other kid at school is probably going to hurt you, or that adults
in your life are going to continue to hurt you. And that produces in them both a
feeling of wanting to fight first—so the best defense is offense. But the other
possibility is that they just simply withdraw from any kind of interaction, because
it’s just not safe enough.

And so all of these behaviors can bring kids in or can get kids referred to us. And
probably the most typical is after a trauma is disclosed and someone becomes
aware that the child has had some traumatic experience. Then we have the
referral to really rule out post-traumatic stress disorder. And so that’s a very
common presentation as well.

JOHN SOMMERS-FLANAGAN: As you were talking, it made me think about the
whole process of how often, social, interpersonal interactions will develop into
kind of a psychological internal working model. And when you referred to the
whole concept of young people then expecting things to not be safe. And so that
leads me to ask you, how do you intervene with that? How do you help the
children develop new working models so that they might see the world as more

ELIANA GIL: And I do think that with internal working models, we can make a
contribution in counseling and in therapy. And that process is really important.
But even more important is for that to be duplicated outside our weekly
psychotherapy, which sometimes really is not sufficient. But as an example, I
worked with a little girl who was six years old.

And I worked with her for a number of months. And she came in one day, and
she brought me a Ping-Pong paddle. And she said to me, “Here, this is for you.”
And I said, “Oh, what’s it for?”

“For you to hit me.”

And I looked at, and I said, “Now, why would I hit you?” And she said, “Well, you
like me, don’t you?” And it was clear in her mind that as soon as I liked her or she
felt liked by me, the next thing that would happen is that I would injure her.

And that’s a very interesting dynamic, because she’s also bringing me something
to do it. Now, I look at that, and I say, wow, that’s a child who really has figured
out a way to decrease the anticipatory anxiety about getting hurt by just saying,
here, here’s the thing, do it, let’s get it over with. And then we can move on.

© 2016 Laureate Education, Inc. 2


Some people, unfortunately, will look at that, whether it’s a paddle or whether it’s
a child that’s being provocative, as a way that the child is saying, basically I
deserve this, or do this, or provokes them, because I think that sometimes kids
who are pushing a lot and push our buttons, as it were, sometimes as those
working with children, we do have these responses that they’re pulling for. So in
terms of helping that particular child, first and foremost, patience, and secondly,
consistency, and third, really trying to up the therapy experience for her so that it
wasn’t just a weekly situation, but I could see here two or three times a week,
and then the engagement of her foster care system, in this case, to provide the
same kinds of messages and responses that I was doing.

What I found, very interesting, that initially I would try to be very supportive and
warm. That scared kids. And so I found them actually withdrawing. So my
fantasies—this was when I was very young and first starting out in this work—my
fantasy was, I would sit and rock these children who had been injured in a chair
or do something affectionate and warm. And when I came to find was that that
actually increased their anxiety so much, because it was so unfamiliar. And
somehow the familiar interaction, the, you’re going to hurt me, was the one they
expected, felt more comfortable with in a way, tolerated better, and definitely
pulled for.

So I had to go into neutral mode. And so when I work with kids, often it’s the
neutral, non-directive play therapy approach, where you basically are doing
empathic listening. And you’re giving them feedback from time to time about what
you notice them to be doing, but not a lot of positive validation or my intuitive
responses about trying to be warm and much more positive in terms of validation
with them, because they need to develop, stretch their comfort zone around
these new behaviors that they’ve never encountered, and get past the anxiety
that that provokes, and then also keep testing it constantly, because that little girl,
I think I worked with her for another year before she really believed that I would
not hurt her.

It took that long. And so one of the things that I always keep in mind is, repetition
is so valuable and so critical that it’s not enough to do it once. It’s not enough to
do it 100 times. You just have to be really patient, not allow for the pulling of the
children to guide responses that are impatient, perhaps harsh, or anything like
that that we just have to be so careful. And the relationship gets built up, I think,
in a very careful way, because it’s a fragile system at that particular juncture.

But again, I emphasize that without an external caretaker, someone who’s
invested in the child, a relationship I can promote outside the therapy, I think that
again, these efforts in our therapy would not be sufficient.

JOHN SOMMERS-FLANAGAN: When I here you use the “patience,” I think
you’re also talking about for us, as counselors, our expectations. I know we live in
a culture that expects quick change. But I remember just reading recently a

© 2016 Laureate Education, Inc. 3


research study by Michael Lambert, where he said that, contrary to the short
term four to eight session EAP model, in order for individuals to experience—
50% percent of clients who come to counseling to experience significant benefits,
we need a model that is 20 to 25 sessions. And although I know a lot of times,
we can’t work within that bigger, more expanded model, I think it’s really
important for us to keep our expectations in check so that we’re not thinking, oh,
yes, I can just create that safe environment, and that young traumatized person
will experience it.

Of course, as you’re saying, it’s not just the office safe environment. You need to
start building safe environment outside of the office. Otherwise you will be really
confusing the child in terms of whether things are safe and not safe.

ELIANA GIL: I think one of the most optimistic movements has been the
movement towards the understanding of the development of the brain and the
neuroscience of interpersonal trauma and how that affects the child at different
developmental junctures. And what’s so optimistic about that is, thinking about
perhaps the brain having more plasticity that there is no quick fix, but that also at
the basic premise of all this is the relationship and how important the relationship
is to the establishment of these interventions that are designed to do some good
and to stimulate parts of the brain and the child that may be haven’t had an
opportunity to grow.

So I feel good about that, and I think it also challenges a little bit this model of,
let’s do this in four to six sessions. I get concerned that we get economically
driven sometimes and that when we’re talking about kids who’ve had histories of
trauma, severe neglect, severe sexual abuse, physical abuse, just general
maltreatment, that what they come in with is really a distrust of adults and
caretaking of them, kind of a lack of grounding and anchoring in possibilities
around the development of relationships and that there are often hyper-aroused
by things that we’re not even aware of it.

So I can wear, for example, a particular color, and that could trigger a child who’s
been traumatized to have a very active re-experiencing of fear and anxiety that I
may not even understand. Sometimes I’m talking to schoolteachers, and they
say, well, I was talking loudly, but I wasn’t yelling. And I don’t understand why
that child would suddenly have to leave the room. And they don’t understand that
it’s possible that that cues the child that we’re about to have a violent episode
here. I better get out of here, because I need to stay safe.

JOHN SOMMERS-FLANAGAN: What’s coming next?

ELIANA GIL: What’s coming next? So we just have to be so careful to
understand that this process takes time. At least for the people that I’ve worked
with, it’s never a quick fix. Now, within the whole context of trauma, yes, there are
some kids who fare better than others. Trauma, I think is very phenomenological.

© 2016 Laureate Education, Inc. 4


The experience of the child and how that child experiences power and control, if
any, during a traumatic event or what defenses they use, this will set the stage
for basically how receptive they may be to interventions that come along. But it
also is in an important precursor to what kind of symptoms they’re going to
develop and how they’re going to fare. So I always think it’s really important to
look at assessing the traumatic impact, because we can’t just assume that every
child is going to react the same.

And honestly, some of them have internal resources that they use. Some of them
have even things like having a pet that they can really talk with, that they can
hold, and that they sleep under the covers with. That can make a huge difference
in their perception of how safe or how nurtured or how connected they are to
something else that then impacts their other responses to these events that are
happening that may be obviously serious stressors and traumatic stressors. So I
find this assessment piece a very important piece of the puzzle in terms of trying
to understand how we approach children.

I’m amazed at natural reparative healing systems and how some kids can
engage, even in post-trauma play, and be able to, in some ways, do their own
gradual exposure. We talk a lot about cognitive behavioral therapy, this
desensitization that often is provided as a therapeutic intervention. This is what I
think children do in post-trauma play.

When they’re repeating the trauma externally, they’re exposing themselves to
this play, and they’re interacting with it. And they’re in some ways managing it
and resolving things and answering their own questions. And it’s a beautiful thing
to watch when it works towards its proposed goal. Every now and then, I
encounter kids who do that, and they get stuck in it.

And so the actual gradual exposure isn’t as effective, because they just keep re-
experiencing the traumatic memory the same way, and nothing changes. And in
those cases, I have to be more actively involved.

JOHN SOMMERS-FLANAGAN: I’m thinking of the desensitization model. And
that’s going to link us to evidence-based treatments. But I’m also reminded of a
case I’m supervising of a graduate student, who’s working with a 17-year-old
young man. The young man just can’t even talk in this session.

And so I said, “Lower your expectations,” to the graduate student. “Just bring
some games.” And so he took in backgammon. They had contracted to three
sessions, where they would just work together. And then the young man, the
client could decide whether he wanted to continue.

And they mostly played backgammon for three sessions but talked a little bit
while they’re playing. In the third session, the young man says to the therapist, to

© 2016 Laureate Education, Inc. 5


the counselor, well, this is our third session. I guess we have to decide whether
we’re going to continue working together.

And the counselor said, “Yeah, what do you think?” And the client said—who had
been completely opposed to this process, but after three sessions of playing
backgammon, he says, “Well, whatever you think,” kind of giving over the choice,
which I saw as a very clever way to avoid rejection. He doesn’t have to say, “Yes,
I want counseling,” and then have the counselor reject him.

And so the counselor said, “Well, how about if we keep meeting then?” And
session after session, just like in a desensitization model, this 17-year-old who
couldn’t speak about his own personal experiences with another person gets
better and better and more and more able to speak. And I found that process to
be just very, very kind of joyful for me to watch the development of that
relationship and the trust build.

ELIANA GIL: It’s interesting when you say that example, because sometimes
people look at that and they say, well, they just played. There’s nothing going on.
Or someone, a graduate student, might say, well, how do I document that in my
notes? And is that legitimate?

Can I do this and call it therapy? And I think there’s so much to playing a game
together; because it is something you do with another. And it has rules, and it
has a structure, and you get to experience this person without any demands.

There’s so much going on when kids are playing games. But especially the older
kids really seem to enjoy it. And it’s a way to begin to get their feet wet into this
new kind of environment, where they don’t have a lot of control. So I think it’s a
beautiful example of how valuable that can be and how much we have to take
our time.

JOHN SOMMERS-FLANAGAN: And I would say we could document that as
desensitization and social skills, training, and there’s all that going on in addition
to just the building of trust in the relationship. So let’s talk for a moment about
evidence-based strategies. I know with PTSD, post-traumatic stress disorder,
there are a number of evidence-based strategies.

One of the challenges in counseling is, how do we transform or translate the
information from the scientific research into our clinical practice? And so I’m
wondering how you do that, what kinds of evidence-based information you find
useful in your practice.

ELIANA GIL: I welcome the evidence-based practices, and I’ve been very
interested in learning whatever I can. What I think of now is integrating evidence-
based principles and practices into my clinical practice, often because now,
there’s a movement towards, we have to do this in order to get reimbursed. So

© 2016 Laureate Education, Inc. 6


there’s a movement towards, you won’t get paid if you’re not doing some
evidence-based practices.

This is a starting in California and certainly moving all the way across the
country. And that’s good. It’s an accountability. I think that that’s an important

Where I end up a little bit concerned is the model of one size fits all. And that’s
where I have the concern. I value what people do in research. And I value some
of the outcomes, but I also understand that because one particular method is
proven, someone’s had the money and the environment in which to conduct
research, it doesn’t mean that what other people are doing isn’t equally valuable.

It just hasn’t been proven yet. So in the area that I work in, which has to do really,
with children who have been abused, the evidence-based model is trauma-
focused cognitive behavioral therapy. And it is a model that I’ve obviously taken a
lot of training in. And it’s a model that we’ve implemented.

We actually even did a small research study comparing TFCBT to what we do,
which is trauma-focused integrated play therapy. And my feeling is that they’re
both effective. I think that the rigid application of anything is problematic, and with
TFCBT, really, there isn’t a demand for fidelity.

So the original research was done quite a while ago now. I think it’s been at least
10 years. And now where we are is that there’s a hybrid model. And I went to a
recent training by a certified TFCBT trainer, and that’s what she said. So this new
hybrid model, even though it’s evidence-based, is really a new development of
incorporating the feedback from the world out there as people began to present
this to their clients.

And what they found is that it didn’t fit everybody. Children are children. And
sometimes the cognitive behavioral strategies are not as inviting to them. It may
feel more like school to them, and if we’re talking about four and five and six-
year-olds, it kind of falls flat.

Now, some other people have started looking at CBT in a playful way. So they’ve
actually combined the two and started saying, we can teach this in this kind of
fun way. And that engages the kids a little bit more. So we may not do TFCBT in
terms of the actual way it was designed way back in the research, where you do
this in the first few sessions, and then you move to this and you move to this.

But obviously we incorporate the basic principles of it, which is a real focus on a
direct movement of understanding of the trauma and that the narratives are
important. And the narrative we do may not be verbal. But we may have the kids
draw things out or play things out or do things in the sand tray. And that all

© 2016 Laureate Education, Inc. 7


works, as long as for that person there’s an understanding of what was, what this
trauma was.

Some of the compartmentalized feelings and thoughts and reactions are
explored. They feel a restoration of power and control. And they have had the
opportunity to release affect. And they have a good support system.

And as long as all of that is happening, I think the incorporation of the evidence-
based principles might be the most effective way to go. There are also some
other evidence-based practices. One of them is called parent-child
psychotherapy, which was done by Lieberman and Van Horn in San Francisco,
with children who witness domestic violence.

That’s an interesting program, because that has the parents and the kids doing
play therapy together. But in addition to that, there’s a co-construction of a
narrative between a parent and a child. Again, the focus is the restoration of
power and the perception of each other in a different way. Especially in domestic
violence, the kids perceive their parents as having not a lot of power and being
helpless or something like that—so the restoration of a different perception of
that parent.

And there’s also PCIT, parent child interaction therapy. That model is probably, of
all of them, the least accessible right now, because it’s a very expensive training
program. And then there’s a lot of fidelity requirements that a lot of agencies will
have trouble implementing. I understand from some of the people who work with
that that they’re looking to soften the guidelines a little bit so that more programs
can implement it.

And that’s a good model. And then we have something called child-parent
relationship therapy, which is a play therapy base model based on filial therapy,
which has been very well researched. So it’s kind of like an explosion of these
models that are appearing on the scene. And I think there’s a common ground
among all of us who have been doing trauma work for a while.

And I like right now to call what I do evidence informed and continue to have a
model that’s integrated so that whenever possible, we will utilize the evidence-
based models. And again, in your private practices or in your agencies, there’s
going to have to be an implementation process. So in the agencies that we work,
there’s going to have to be the implementation of and adaptation of these models
so that it makes sense to the people providing the services.

And then there’s, of course, the whole other issue of education and training. So
we’re telling counselors and therapists, you need these basic things to graduate.
And then suddenly, what happens is that they are told, well, actually, no, you’re
going to have to go get some certifications now. TFCBT is about to launch a
certification process.

© 2016 Laureate Education, Inc. 8


So it’s going to require people to go back and take training and then consult and
so forth and so on. So I think it’s interesting. And I think for all of us, it’s a
challenge to figure out how much of it we can integrate and in what ways and
what fits best for the families that we’re working with.

JOHN SOMMERS-FLANAGAN: I love the integration of the play with the CBT,
because I think they fit well together. A couple specific questions—what are your
thoughts on EMDR and/or medications in the treatment of children and
adolescents with trauma

ELIANA GIL: I think over the last 20 years or so, what we’ve seen the emergence
of the MDR for the treatment of PTSD. And I think that anyone doing any trauma
work needs to be trained in it, to be honest. And it’s a very powerful training. I
think it gives you another set of tools that I think can be used in the treatment
process that you have.

The evidence is pretty clear that it works. The evidence on it working as well with
children is now growing. So I think again, if you’re working with kids and/or adults
and there’s PTSD trauma backgrounds, you have to be able to know that and be
conversant with it and use it. I’ve taken the training.

There are some clients who respond very well to it. It’s again, one of those
questions that you ask yourself about, does it lead the way in the practice that
you do? Like, there are some people who say, I do EMDR and a therapist might
refer to that clinician to do 10 or 12 sessions of EMDR. Or do you incorporate it
into your practice and use it as you believe it’s indicated?

So I tend to do it in that way rather than referring out for it. And I think it’s a very
valuable tool. And again, that’s evidence-based. And also, we can’t ignore it. It’s
been shown over and over and over again to work. Nobody seems to quite
understand why.



Trauma, Stress, and Adjustment

The DSM-IV described adjustment disorders as a single classification. These are now
recognized as a heterogeneous group of disorders closely associated with stress, both
traumatic and non- traumatic. As such, adjustment disorders are classified in the DSM-5
along with trauma (including posttraumatic stress disorder, formerly included in DSM-IV
“Anxiety Disorders”) and reactive attachment disorder (formerly included in DSM-IV
“Disorders Usually First Diagnosed in Infancy, Childhood, or Adolescence”). The new
inclusive DSM-5 classification is called “Trauma- and Stress-Related Disorders.” The
common criteria across these disorders are exposure to a traumatic or stress-inducing

Reactive Attachment Disorder

The DSM-IV included two subtypes of this disorder, representing distinct behavioral
expressions. The DSM-5 separates these subtypes into different diagnoses (though
they retain a single numerical identifier). Reactive attachment disorder (formerly the
withdrawn/inhibited subtype) results from neglect or other influences early in
development that negatively impact a child’s ability to form firm attachments. It is
characterized by a pattern of emotional withdrawal, limited or absence emotional
responsiveness, and limited positive affect. There may also be observed negative affect
incongruent with circumstance. The disorder generally appears during 9 months to 5
years of age. Diagnoses over the age of 5 years should be made cautiously, as little
research supports or describes the manifestation of this disorder in older children or

Disinhibited Social Engagement Disorder

In contrast to reactive attachment disorder, this new diagnosis—formerly the
indiscriminately social/disinhibited subtype of reactive attachment disorder—is also
linked to significant deficits in caregiving at very young ages. However, unlike reactive
attachment disorder, children with this diagnosis may have well-formed attachments.
This disorder is characterized by a pattern of indiscriminate interaction with adults,
such as a willingness to approach and interact with those who are unfamiliar.
Behaviors may also include excessively verbal or physical that is inconsistent with
cultural or age-appropriate norms. This disorder is also associated with attention-
seeking behaviors, overfamiliarity, and inauthentic expression of emotion. Persistence
through adolescence is often accompanied by increased peer conflict.

Posttraumatic Stress Disorder (PTSD)

The DSM-5 includes several changes to this diagnosis. The revised PTSD diagnosis
can be used with adults, as well as with adolescents and children over the age of 6.
Criterion A, which pertains to the manner in which the traumatic event was experienced,
has been significantly revised to more specifically describe direct, indirect, and
witnessing experiences. Criterion A2 from the DSM-IV has been eliminated, thus
removing interpretations of subjective response.

Criterion B now includes more descriptive wording and is described as “intrusion


symptoms.” Criterion C from the DSM-IV has been separated into two symptom clusters:
persistent avoidance of associated stimuli (Criterion C in the DSM-5) and negative
alterations in cognitions and mood (Criterion D in the DSM-5). The criterion cluster
association with alterations in arousal and reactivity has expanded to include verbally or
physically aggressive behavior, recklessness, and self-destructive behavior. Another
important change in this diagnosis is the addition of specific criteria for children ages 6 or
younger. These criteria are founded in the criteria applicable to adolescents and adults;
however, they also include important age-specific variations. The DSM-5 also includes
important information regarding most common comorbidity differences between children
and adults diagnosed with this disorder. Oppositional defiant disorder and separation
anxiety disorder most commonly occur with this diagnosis in children.

Acute Stress Disorder

As with PTSD, the specific wording of Criterion A has been revised to more clearly
identify the manner in which the trauma was experienced, with the former criterion A2
from the DSM-IV eliminated entirely. Additional symptomology has been regrouped into
five main categories (intrusion, negative mood, dissociation, avoidance, and arousal)
with a total of 14 symptoms; individuals need to have 9 of the 14 symptoms present in
order to meet Criterion B. Onset and duration have been revised as well, noting the
presence of Criterion B symptoms to be present 3 days to 1 month after exposure to the
traumatic event.

Two additional new diagnoses are also part of this classification: other specified
trauma- and stressor-related disorder, and unspecified trauma- and stressor-
related disorder. Both of these diagnoses represent significant clinical distress or
impairment based on diagnostic criteria common to this classification, but do not meet
full criteria for a specific diagnosis. Clinicians should use other specified trauma- and
stressor-related disorder and add the specific reason for the more general diagnosis
(e.g., delayed onset of more than 3 months, or culturally-associated concepts). The
latter diagnosis—unspecified trauma- and stressor-related disorder—is used when
clinicians cannot (or choose not to) identify reasons for the inability to make a more
specific diagnosis, yet clearly observe multiple criteria from the trauma- and stressor-
related disorders classification.


• American Psychiatric Association (2013). Highlights of changes from DSM-IV-TR
to DSM-5. Retrieved from

Agreement of Parent and Child Reports of Trauma Exposure and
Symptoms in the Early Aftermath of a Traumatic Event

Carla Smith Stover and Hilary Hahn
Yale University Child Study Center

Jamie J. Y. Im
Boston University

Steven Berkowitz
University of Pennsylvania

Exposure to violence and potentially traumatic events (PTEs) is a common experience among children
and youth. The assessment of necessary intervention relies upon parental acknowledgment of exposure
and recognition of their child’s distress. Early interventions and treatment are most effective when
parents are aware of the nature of the traumatic exposure, understand their child’s symptomatic response,
and are intimately involved in the treatment process. The present study investigated concordance between
parents and exposed children on child trauma history, the subjective report of the impact of the traumas
experienced, and presence of posttraumatic stress disorder (PTSD) symptoms. Agreement between parent
and child report of traumas experienced was nonsignificant for serious accidents, separation from
significant others, and physical assaults. Nonsignificant agreement was also found for avoidance and
hyperarousal symptoms of PTSD. Correlations were not significant between parent and child report of
the impact of traumas both at the time of the incident and at the time of the interview. Recommendations
are suggested for helping parents improve their capacity to understand the potential impact of exposure
on the child’s psychological functioning.

Keywords: parent-child concordance, PTSD, posttraumatic symptoms, trauma history, peritraumatic

For children to receive intervention or treatment for traumatic
stress reactions, parents or adult primary caregivers typically must
identify or acknowledge their children’s distress following injury
or exposure to violence. Particularly in the acute aftermath of an
accident or other potentially traumatic event, first responders, child
protective services workers, health care professionals, and crisis
workers often refer to parents, rather than the child, for informa-
tion about the child’s reactions, including presence and severity of
symptoms. In addition to many adults’ discomfort with interview-
ing children, there are multiple reasons for which adult caregivers’
are deferred to over their children: (1) medical treatment that may
render the child unavailable, (2) adult attempts to protect their
children from thinking about the event, or (3) the child’s involve-
ment in an ongoing investigation (e.g., participating in a forensic
interview for sexual abuse which precludes speaking about the
event or associated symptoms except with the interviewer assigned

to the investigation). Even when children are included as part of an
assessment, clinicians may favor parent reports in forming an
assessment of the child’s psychiatric condition in the belief that
adults are more accurate reporters (Grills & Ollendick, 2003); this
may be particularly true for younger school-aged children (�9
years old) (Rapee, Barrett, Dadds, & Evans, 1994). However,
discrepancies between parent and child reports both of the child’s
exposure to potentially traumatic events, and of the child’s result-
ing symptoms, must be taken into account by clinicians and
researchers when considering the degree of dependence on care-
givers for evaluation of child reactions. The current study inves-
tigates the concordance between adult caregiver reports and child
reports of current and prior trauma exposure, the impact of prior
traumas on the child, and of symptoms related to those exposures
within the peritraumatic period (within a month of exposure when
posttraumatic stress disorder [PTSD] cannot be diagnosed) after a
child is exposed to a potentially traumatic event (PTE).

Parent–Child Agreement About
Trauma Exposure History

Significant discrepancies have been found between parent and
child reports of the number of traumatic events previously expe-
rienced by the symptomatic child (Schreier, Ladakakos, Morabito,
Chapman, & Knudson, 2005). Parent reports of children’s expo-
sure to violence have repeatedly been shown to underestimate the
child’s level of exposure (Ceballo, Dahl, Aretakis, & Ramirez,
2001; Richters & Martinez, 1993; Selner-O’Hagan, Kindlon,

Carla Smith Stover and Hilary Hahn, Yale University Child Study
Center; Steven Berkowitz, Department of Psychiatry, University of Penn-
sylvania; Jamie J. Y. Im, Department of Psychology, Boston University.

This study was supported by the Substance Abuse and Mental Health
Administration (SAMSHA) National Child Traumatic Stress Network. The
authors would like to acknowledge Arthur Roy and Gina Poole for their
work on this project. This study was supported by NIDA research Grant
K23 Da023334.

Correspondence concerning this article should be addressed to Carla Smith
Stover, Yale University Child Study Center, 230 South Frontage Road, New
Haven, CT 06520. E-mail: [email protected]


Psychological Trauma: Theory, Research, Practice, and Policy © 2010 American Psychological Association
2010, Vol. 2, No. 3, 159 –168 1942-9681/10/$12.00 DOI: 10.1037/a0019185





t i

















f i













r t



l u


f t




r a










Buka, Raudenbush, & Earls, 1998), particularly for boys (Kuo,
Mohler, Raudenbush, & Earls, 2000). Children have been found to
report exposure to violence more often in their neighborhood or at
school, while caretakers report more events near or at home (Raviv
et al., 2001; Thomson, Roberts, Curran, Ryan, & Wright, 2002).
However, in the case of domestic violence, many parents deny or
minimize the presence of children during incidents of violence by
suggesting that the children were asleep, watching television, or
playing outdoors (Jaffe, Wolfe, & Wilson, 1990). Studies have
shown that despite mothers’ efforts to shield their children from
violence, 68% to 87% of incidents of partner abuse are, in fact,
witnessed by children (Jaffe et al., 1990). Although a number of
studies have looked at agreement between parent and child reports
of type and level of trauma experienced by children, to our
knowledge none have evaluated parent– child agreement about the
impact of previous potentially traumatic experiences or how much
previous events currently impact the child.

Parent–Child Agreement About
Symptoms and Diagnosis

When parents and children report independently on child symp-
toms, they provide notably different information. A number of
studies support the finding that parents underestimate their child’s
PTSD symptoms that result from exposure to community violence
(Ceballo et al., 2001), chronic medical conditions (Shemesh et al.,
2005), and injury (Meiser-Stedman, Smith, Glucksman, Yule, &
Dalgleish, 2007, 2008). Some studies have found weak parent–
child agreement regarding the presence of internalizing symptoms,
but stronger agreement when reporting externalizing symptoms
(Ladakakos, 2000; Perlstein, 2004) and a meta-analysis of 119
studies examining concordance among multiple informants found
that agreement between parents and children was significantly
lower for internalizing than externalizing disorders (Achenbach,
McConaughy, & Howell, 1987). However, other studies have
indicated that parents may tend to overreport externalizing symp-
toms (Kolko & Kazdin, 1993).

Poor concordance has been shown for the major childhood
anxiety disorders presenting as either the principal diagnosis (� �
.22–.31) or as part of the diagnostic profile (� � .04 –.23;
Choudhury, Pimentel, & Kendall, 2003). In a prospective study of
90 children ages 10 to 16 exposed to a single-event trauma and
their parents conducted by Meiser-Stedman et al. (2007), children
were found to be significantly more likely than their parents to
report meeting the dissociation and re-experiencing criteria for
Acute Stress Disorder (ASD), as well a diagnosis of ASD. Disso-
ciation was particularly underreported by parents and was ex-
cluded from a study diagnosis of “early PTSD” (PTSD without the
duration criterion); however, parents were still significantly less
likely to report that their child met the criteria for the early PTSD
diagnosis than the children themselves (Meiser-Stedman et al.,
2007). A study by Kassam-Adams and colleagues of 219 injured
children found similar results, with parent– child agreement low
for the overall diagnosis of child ASD (� � 0.22) and for the
specific symptoms in the dissociation, re-experiencing, avoidance,
arousal/anxiety and impairment clusters, with � values ranging
from .02 to .43. Parent and child ratings of child ASD severity
were moderately correlated (r � .35; Kassam-Adams, Garcia-
Espana, Miller, & Winston, 2006).

A study of 83 children hospitalized for traumatic injury found
parents underrated their child’s level of PTSD symptoms when
compared with the child’s report gathered within 24 hours of
hospital admission; this under report approached statistical signif-
icance at the 1-month assessment (Schreier et al., 2005). The early
discrepancy was attributable to how children and parents reported
re-experiencing and avoidance/numbing symptoms, with children
reporting higher symptoms in these categories. These discrepan-
cies decreased at the 1-month and 18-month assessment points.
There was good agreement on report of hyperarousal symptoms at
all time points (Schreier et al., 2005). These data highlight the
divergence of reporting within the peritraumatic period where
preventative interventions might be the most successful. However,
in a study of similar population, Ladakakos (2000) found that
parents significantly underrate their child’s level of PTSD symp-
toms at all data points.

Some studies have shown superior parent– child agreement for
the PTSD diagnosis compared with the ASD diagnosis. Meiser-
Stedman et al. (2007) examined parent– child agreement for these
disorders in a prospective study of assault and motor vehicle
accident (MVA) child survivors, assessed at 2 to 4 weeks and 6
months’ post-trauma. They found an improvement in parent– child
agreement between the initial assessment, with poor parent– child
agreement for the diagnosis of ASD (� � �0.04), but fair agree-
ment (� � �0.21) for PTSD diagnosis at 6-month follow-up. In
addition, parent reports of child ASD symptoms failed to correlate
with later child PTSD (Meiser-Stedman et al., 2007). These find-
ings have significant implications for the assessment of children in
the peritraumatic period and how a child’s need for acute inter-
vention is determined.

Agreement by Age and Gender

It is unclear what role child age plays in the degree of parent–
child concordance about child PTSD symptoms. Discrepancies
have been shown to be more pronounced between younger chil-
dren and their parents (Dyb, Holen, Braenne, Indredavik, & Aars-
eth, 2003). However, Shemesh et al. (2005) found that the gap
between adolescents’ (n � 47) reports of their PTSD symptoms
and that of their parents was greater than that of children under age
12 (n � 29), although the authors note that this finding may be
compromised by the small numbers of children in each group.
Choudhury et al. (2003) found agreement about the presence of an
anxiety disorder diagnosis was poor for both age groups (� � 0.16
for children �11 and � � 0.05 for children �10). They found
younger children to have higher rates of agreement for the pres-
ence of general anxiety disorder, social anxiety disorder, and
specific phobia while showing greater agreement between parents
and older children for the principal diagnosis of specific phobia.
Studies to date have not looked at age relationship to parent– child
agreement in reporting of PTSD diagnosis.

Inconsistent findings in studies examining the relation between
child age and informant discrepancies may be attributable to
inconsistencies in sample characteristics including categorization
of child age and the research methodology (De Los Reyes &
Kazdin, 2005). They may also be influenced by parents’ own
symptoms as parent report of child symptoms has been shown to
be significantly correlated with parents’ own symptoms following
a traumatic event (Kassam-Adams et al., 2006). There is evidence





t i

















f i













r t



l u


f t




r a










that this relationship is particularly true for parents of young
children (Laor, Wolmer, & Cohen, 2001).

It is equally unclear how child gender relates to level of agree-
ment between parent and child reports. The meta-analysis con-
ducted by Achenbach et al. found the results inconclusive in this
area (Achenbach et al., 1987). Choudhury et al. found poor agree-
ment for presence of Anxiety in the diagnostic profile and for the
principal diagnosis for both girls (� � 0.31 and � � 0.35, respec-
tively) and boys (� � �0.01 and � � 0.26, respectively). Review-
ing mixed findings from numerous investigators, De Los Reyes
and Kazdin (2005) surmise that in the aggregate, child gender may
not be related to informant discrepancies, but in specific popula-
tions, child gender effects may be present. No studies were found
that examined gender differences in parent– child agreement for
Depression or PTSD symptoms following a traumatic event.

While there are consistent findings in the literature on the poor
concordance between caregiver and child report of trauma expo-
sure, PTSD symptoms and diagnosis, this study adds to the body
of literature on parent– child concordance in several areas. First,
this study aims to further assess discrepancies between parent and
child reports of specific types of traumas experienced by the child.
Second, this study assesses agreement about the impact of these
traumas, exploring how the trauma affected the child both at the
time of the incident and at the time of the baseline assessment.
Third, differences in concordance between parent and child are
explored by gender and age for type of traumas, impact of the
traumas, and posttraumatic symptoms.



Seventy-six youth aged 7 to 17 years who were exposed to a PTE
and endorsed at least one new symptom of posttraumatic stress
disorder when screened by telephone using the Posttraumatic Check-
list (Amaya-Jackson, McCarthy, Newman, & Cherney, 1995) were
recruited into the Child and Family Traumatic Stress Intervention
study (CFTSI) at the Trauma Section of the Yale Child Study Center.
Children were referred by police, the hospital sexual abuse program,
or a pediatric emergency department due to exposure to a PTE to
participate in a randomized trial of a 4-session secondary prevention
model for children exposed to a PTE.


For purposes of these analyses, data from all baseline interviews
completed from December 2006 through July 2008, for the interven-
tion study were utilized. Following written informed consent proce-
dures, youth participants and their adult caregivers were interviewed
using a standard set of measures. Baseline interviews were completed
separately with each child and a caregiver, administered by a trained
research assistant within 30 days of the PTE.


For the purposes of the present study, six measures included in
the baseline interview were analyzed: the Traumatic History Ques-
tionnaire Parent Report and Child Report Version (Berkowitz &
Stover, 2005), the UCLA Posttraumatic Stress Disorder Reaction

Index (PTSD-RI) Parent Report and Child Self-Report versions
(Pynoos, Rodriguez, Steinberg, Stuber, & Frederick, 1998), and
the Mood and Feelings Questionnaire Parent Report and Child
Self-Report versions (Angold & Costello, 1987).

Traumatic History Questionnaire (THQ). This question-
naire contains 13 identical items presented as child or parent report
developed at Trauma Section at the Yale Child Study Center to
assess the child’s history of previous traumatic events and the
intensity of their reactions. The items were derived from the
Traumatic Events Screening Inventory Child and Parent Report
(TESI; Ford, 2002; Ghosh-Ippen et al., 2002; Ribbe, 1996). For
each item endorsed, participants are asked to indicate the level of
impact the trauma had on the child, both at the time of the event
and at present. Respondents select a number from 0 to 4, using a
scale in which 0 � “not at all” and 4 � “extremely.” A total
trauma history past impact score and total trauma impact current
scores are derived by tallying all items endorsed.

University of California–Los Angeles (UCLA) Posttrau-
matic Stress Disorder Reaction Index (PTSD-RI). This is one
of the most widely used instruments for the assessment of post-
traumatic symptomatology related to subjective distress. It as-
sesses diagnostic criteria B (re-experiencing), criteria C (avoid-
ance), and criteria D (hyperarousal) symptom clusters to diagnose
PTSD (Pynoos, Rodriguez, & Steinberg, 2000). This measure is
among the most extensively studied assessments of childhood
PTSD and has strong convergent validity with regard to Diagnos-
tic and Statistical Manual of Mental Disorders–Fourth Edition
(DSM–IV) diagnosis. It allows for calculation of the severity of
symptoms on each cluster (B, C, and D) and a total severity score.
Numerous studies have found consistently higher Reaction Index
scores among traumatized samples compared with control subjects. It
has good convergent validity, 0.70 in comparison to the PTSD Mod-
ule of the Schedule of Affective Disorders of School Aged Children
and .82 with the Child and Adolescent Version of Clinician Admin-
istered PTSD Scale. Cronbach’s alphas fall in the range of .90 for
internal consistency across versions and with test–retest reliability at
.84. (Steinberg, Brymer, Decker, & Pynoos, 2004).

Short Mood and Feelings Questionnaire (SMFQ). This
questionnaire is a 13- item version of the Mood and Feelings
Questionnaire (Angold & Costello, 1987) designed to detect de-
pression in children and adolescents (Angold et al., 1995). Items
are rated by participants as 0 � “not true,” 1 � “sometimes,” or
2 � “true” yielding a total depression score. The SMFQ is highly
correlated with the longer version of the MFQ and correlates
moderately high with the Child Depression Inventory and Diag-
nostic Interview Schedule for Children. Internal consistency is
good for both the child report (� � .85) and parent report (� � .87;
Angold et al., 1995). The MFQ’s use in diagnosing depression has
been validated in several studies (Messer et al., 1995; Wood, Kroll,
Moore, & Harrington, 1995).

Data Analysis

First, Cohen’s kappa (Cohen, 1960) was used to assess agreement
between parent and child reports of specific traumas experienced by
the child over his or her lifetime on the THQ. Kappa (�) is a
correlational statistic that examines agreement while correcting for
chance. � values of greater than 0.75 are considered to indicate high
agreement, 0.40 to 0.75 represent moderate agreement, and less than





t i

















f i













r t



l u


f t




r a










0.40 indicate poor agreement (Mannuzza et al., 1989). Next, the group
was split by gender and age (school age � 7–12 and adolescents �
13–17) and Kappa statistics computed to determine whether parent–
child agreement differed for trauma types based on either of these
characteristics. Next, prevalence-adjusted bias-adjusted kappas
(PABAK; Byrt, Bishop, & Carlin, 1993) were used to assess agree-
ment between parent and child reports of the three PTSD symptoms
criteria on the PTSD-RI. PABAK is a measure of agreement ranging
from �1 to �1 that depends solely on the observed proportion of
agreement between raters. It adjusts Kappa (Cohen, 1960) for imbal-
ances caused by differences in the prevalence and bias (Byrt et al.,
1993). It addresses the problem of high interrater agreement and low
kappa scores known as the “kappa paradox” (Cicchetti & Fein-
stein,1990), which was observed in our PTSD-RI data. PABAK was
used rather than the standard kappa statistic. Several other recent
studies have used PABAK in their reliability studies (Cibere et al.,
2008; Girianelli et al., 2007). Last, Pearson correlations were calcu-
lated to assess the strength of the relationship between parent and
child report of: (1) the impact of previous traumas on the child both
at the time of the trauma and at the present time; (2) the severity of
PTSD symptoms for Criteria B (Re-Experiencing Cluster), C (Avoid-
ance Cluster), D (Hyperarousal Cluster) and Total scores; and (3)
severity of depression symptoms reported on the SMFQ. Correlations
of .30 or less were considered to reveal little relationship between the
variables (Hinkle, Wiersma, & Jurs, 1988).


Demographics and Descriptive Statistics

The total sample of 76 included 44 girls and 32 boys. The
majority of caregivers were mothers (89%),with only 11 fathers
(11%) participating. Youth ranged in age from 7 to 17 years (M �
12.05, SD � 2.87). The ethnic makeup of youth in the sample was

31.6% Caucasian, 36.8% African American, 19.7% Hispanic,
11.8% multi-ethnic or other. Youth were referred for the following
traumatic events: 21.1% sexual abuse; 19.7% assault; 23.7% motor
vehicle accident; 21.1% witnessing violence; 5.3% threatening;
5.3% injury; and 2.6% animal bite.

Concordance Between Parent and Child Report for
Specific Types of Previous Traumas

Cohen’s kappa was used to assess agreement between parent
and child reports of the child’s trauma history. Kappa statistics for
parent– child agreement with respect to specific traumas are shown
in Table 1. Overall, agreement between parent and child report was
poor to moderate, with � ranging from 0.12 to 0.58. The strongest
agreements, which fell in the moderate range and were statistically
significant, were for whether the child had been a victim or witness
of sexual activities (� � 0.58, p � .001), experienced the death of
someone close (� � 0.41, p � .001), or had a family member
arrested or in jail (� � 0.40, p � .001). Although in the poor range,
agreement for dog/animal bites, suicide of someone close, having
been a victim or witness of a mugging or the victim or witness of
physical violence were statistically significant (Table 2). Negligi-
ble and nonstatistically significant agreement was found for seri-
ous injuries, separation from significant others, and physical as-
sault (Table 2).

When compared by gender, agreement between parent and child
report remained poor to moderate across trauma types. Males and
their parents had statistically significant agreement across more
trauma types than girls. Boys and their parents had better agree-
ment for victim/witness of mugging (� � .52, p � .01), suicide of
someone close (� � .52, p � .01), and family member arrested and
jailed (� � .43, p � .05) than girls who had nonsignificant
agreement in these categories (Table 1). Girls and their parents had
significant agreement for animal bites (� � .42, p � .01) and

Table 1
Kappas for Parent and Youth Agreement for Traumas Reported


Males (n � 32) Females (n � 44)


Total (n � 76)

� Child (%) Parent (%) � Child (%) Parent (%) � Child (%) Parent (%)

Serious accident .19 43.8 18.8 .07 70.5 29.5 5.7� .13 59.2 25.0
Severe illness or injury .23� 6.3 34.4 .19� 2.3 18.2 .76 .22�� 3.9 25.0
Death of someone close .51�� 65.6 46.9 .31� 72.7 65.9 .43 .41��� 69.7 57.9
Separation from

significant others .19 18.8 43.8 .10 52.3 38.6 9.7�� .12 38.2 40.8
Suicide of someone

close .52�� 9.4 12.5 .07 20.5 6.8 1.7 .23� 15.8 9.2
Physical assault or

threatening .29 31.3 21.9 .07 27.3 20.5 .14 .17 28.9 21.1
Victim or witness of

mugging .52�� 12.5 9.4 �.04 11.4 2.3 .02 .25� 11.8 5.3
Attacked by a dog or

other animal .24 31.3 15.6 .42�� 29.5 13.6 .03 .34��� 30.3 14.5
Witnessing physical

violence .18 68.8 71.9 .29� 70.5 47.7 .03 .24� 69.7 57.9
Family member arrested

or in jail .43� 53.1 62.5 .39 56.8 34.1 .10 .40��� 55.3 46.1
Victim or witness of

sexual activities .35� 9.4 6.3 .64��� 18.2 20.5 1.1 .58��� 14.5 14.5

� p � .05. �� p � .01. ��� p � .001.





t i

















f i













r t



l u


f t




r a










witnessing physical violence (� � .29, p � .05), while boys did
not. Girls and boys and their parents both had statistically signif-
icant agreement for severe illness or injury, and victim of sexual
assault (Table 1). Despite statistical significance, agreement across
all trauma types was in the moderate range at best and very
low/poor at worst.

Overall, girls reported significantly higher rates of serious ac-
cidents, F(1, 75) � 5.7 and separation from someone close, F(1,
75) � 9.7 than boys (Table 1), but agreement in these categories
between parent and youth report was comparable across genders
and in the very low range.

When evaluated by age group (school age vs. adolescents),
agreement was variable. Agreement between school-age children
and their parents was in the moderate range and was statistically
significant for family member arrested/jailed (� � .47, p � .001)
and having been a victim or witness of mugging (� � .39, p � .01)
but agreement on these variables was not statistically significant
for adolescents and their parents. Adolescents had statistically
significant agreement with parents for severe illness/injury (� �
.33, p � .05) while school-aged children did not (Table 2). Both
groups had statistically significant agreement with parents for
death of someone close, animal bites, and sexual abuse. Adoles-
cents self-reported significantly higher rates of witnessing physical
violence, F(1, 75) � 7.2 and family member jailed, F(1, 75) � 6.9
than school-age children. No other significant age differences were
found for self-reported trauma exposure (Table 2).

Relationship Between Parent and Child Report of the
Impact of Traumas

Correlations of parent and youth report were not significant for
the total impact of earlier potential traumas on youth at the time of
the incident (r � .21), and as reported at the baseline interview

(r � .21). When assessed by gender, there were significant positive
correlations for male youth and their parents for the impact of the
events at the time of the incident (r � .55, p � .000) and at the
time of the baseline report (r � .43, p � .02). These correlations
fell in the moderate range. However, correlations were nonsignif-
icant for females (Table 3). Girls reported significantly higher
impact of prior traumas at the time of the event than boys based on
their own reports, F(1, 75) � 5.6 (Table 4). No significant gender
differences were found based on parent report (see Table 4). When
analyzed by age, the association between parent and child report
was not significant for school-aged children or adolescents. In fact
the correlation between adolescent and parent report was near zero
(Table 4).

Concordance Between Parent and Child Report of
PTSD and Depression Symptoms

PABAK was used to assess agreement between parent and child
reports on the three symptom criteria for PTSD. There was high
agreement for the …

error: Content is protected !!