Response to Trauma in Haitian Youth at Risk

SUMMARY

In order to characterize undesirable behavior (drug use, fighting, criminal activity) among Haitian youth at risk and determine the relationship between traumatic experience and that kind of behavior, investigators recruited 291 Haitian youths via networks of informal social relations in two zones of Miami/Dade County strongly idenitified with Haitian ethnicity. Each recruit responded to an interview schedule eliciting sociodemographic information and self-reported activities, including involvement in youth-dominated groups. They also reported traumatic experience. Clinicians administered CAPS to a subset of those respondents who self-reported traumatic experience. Staff ethnographers selected respondents for in-depth interviews and family studies to provide contextual depth for findings of the interview schedule and the CAPS assessments. Although traumatic experience may still play a role in mental health outcomes among children, childhood victimization among Haitian children does not appear to be related to the drug use and undesirable behaviors associated with unsupervised youth, including formation of gangs.

INTRODUCTION AND BACKGROUND

Haitians in Miami have had the misfortune of encountering some of the most daunting barriers ever presented to an immigrant group. Not only have some of them made the trip to South Florida in the flimsiest of craft through dangerous seas, but many have sojourned for months or years under hostile conditions in the Bahamas before finally making their way to Miami. Once on shore in Dade County, the new arrivals have faced a community that has routinely ghettoized populations of color during most of its existence. Urbanized areas have well-defined zones inhabited by people of African descent whether they are North American, Bahamian, Jamaican, or other Caribbean. Although all housing is officially open by law, the population of the County clearly has compartmentalized phenotypes by geographic areas (Dunn, 1997; Peacock, Morrow, and Gladwin, 1997; Portes and Stepick, 1993). In this kind of racialized setting, the continued marginalization of a population descended from enslaved Africans hardly surprises.

Factors that have exacerbated this tendency to marginalize Haitian immigrants in Miami/Dade County include their supposed linkage with disease and superstition, and recently, association with juvenile crime. Just as the population of Haitian immigrants achieved a recognizable size and cohesion, about 1982, they registered on the Centers for Disease Control's monitors for the newly emergent disease complex, AIDS (Nachman and Dreyfus, 1986). Haitian immigrants' tuberculosis also brought them unwanted attention from North American health and immigration authorities (Nachman, 1993). Some authors (e.g., Moore and LeBaron, 1986) hinted darkly that Haiti had been a major conduit for HIV into the United States. Later analyses of the pandemic have shown the reverse to be true (Farmer, 1998). Nevertheless, the early indication by the CDC that having Haitian background constituted a risk factor for AIDS did gratuitous and irreparable damage to the Haitian community's image, especially in South Florida. This blaming of Haitians as unclean, disease-ridden interlopers has led to ongoing prejudice against them, and it has moved them, as a community, to distrust overtures by researchers, especially if the researchers want to study AIDS or tuberculosis (Wingerd and Page, 1997).

Conditions of poverty, unemployment, and underemployment have persisted in South Florida's Haitian population despite the notable willingness of these newcomers to spend additional energy acquiring schooling that had not been available in Haiti. Although gathering large-scale demographic data on the Haitian population in South Florida presents difficulties (see Stepick and Stepick, 1990), it has become clear that a large proportion of that population still lives in conditions of overcrowding and poverty (Stepick and Stepick, 1992). The location of these Haitian Americans in the mix of ethnicities living in Dade County is contiguous with neighborhoods inhabited by other groups that had been marginalized earlier, including African Americans and African descended Caribbean people. This location is hostile territory, especially for Haitian children trying to attend public schools in these neighborhoods. Furthermore, the lifeways of the “host” cultural context in which these children attempt to learn how to become “Americans” often communicate the futility of education and the attractiveness of “outlaw” or “gangsta” life, perhaps as a reflection of the “host's” own marginality. Formation of youth gangs has occurred among almost every culturally distinctive immigrant group that has arrived in the United States since 1860 (Goldstein, 1990), but under the conditions described above, the specific adaptations of inner city Haitian youth in the U.S. demand close examination.

The present study of Haitian Youth in Miami/Dade County attempts to define the responses of these youth to a hostile social and economic environment. Although the literature on gangs provides numerous templates for characterizing these responses (cf. Arnold, 1965; Cloward and Ohlin, 1955; Cohen, 1960; Elliott, Ageton, and Canter, 1979; Fagan, 1989; Furfey, 1926; Goldstein, 1991; Thrasher, 1927), we shall attempt here to use self-reported and observed behaviors to define the adaptations of Haitian youth, holding in abeyance our comparisons of these adaptations with those in the rest of the literature.

Our preliminary work in preparing to study the emergent phenomenon of reported drug use and gang activity among Haitian youth first led us to examine the reports of these activities for veracity. Community contacts at the Center for Haitian Studies reported that some of the at-risk youth participating in their street diversion projects claimed to belong to gangs. Newspaper reports, beginning in 1996, described criminal activity among Haitian youth, a group that previously had not demonstrated much inclination toward delinquency of any kind. Conferences with gang specialists in the Metro-Dade County police force and the forces of local municipalities in Miami and North Miami Beach identified approximately 20 “gang entities” that had some of the characteristics of youth gangs seen in other parts of Miami/Dade County. These characteristics included having a recognizable name for the group, involvement in fighting, and criminal behavior involving drugs. Rather than accepting these descriptions as starting points for our investigation, we collected them for later comparison with our own experience with the intention of developing our own taxonomy of peer-dominated groups among Haitian youth. This taxonomy is still under development.

Previous work in the Haitian community revealed traumatic experience as a frequently reported feature of young Haitians' lives. In the course of other studies (Wingerd and Page, 1997), we found that some young Haitians had experienced notable trauma in a number of different contexts, including witness to political killings and torture, death of fellow travelers on the journey to the United States, mistreatment in holding camps and temporary quarters, and abuse at the hands of step-parents.

Because we realized that response to traumatic experience can reduce a person's ability to concentrate (Yehuda et al., 1995), which in turn would interfere with their performance in school, we theorized that Post-Traumatic Stress Disorder (PTSD) might help to explain heavy involvement in peer-dominated and drug-involved social groups (Duke et al., 2003). Our pilot assessments of a group of seven at-risk Haitian youths using the Clinician Administered Post-Traumatic Stress (CAPS) interview had the intention of detecting the presence of symptoms or diagnosability in this population (Blake et al., 1990). We found that all but one of the seven youths assessed by a child psychiatrist either had symptoms or were diagnosable with PTSD. We therefore proposed to broaden our study of PTSD among Haitian youth as a possible explanation for difficulties in school and delinquent behavior.

Brief Description of Gangs

In the course of intervening among Haitian youth for the prevention of HIV infection, a local community-based organization, the Center for Haitian Studies, began to receive reports from the participants in the intervention that Haitian young people had engaged in activities such as street-level drug trafficking, organized fights with other youths, and wearing of “colors.” Some Haitian youths identify themselves as members of groups that use the word “zo” (Creole for “bone”) in their names to express that they are “Haitian to the bone.” These group names often take the forms: “Zo (location), or “Zo (address),” or “(attribute) Zo,” or “(noun) Zo.” In behavior and concept, the zo groups appeared to be gangs. News coverage of gangs in the community showed that what are now called Haitian gangs had not attracted any public attention before about 1994 (cf. Herald Staff, 1988, 1994, 1996, 1997).

Theory of Drug Involvement and Gang Development

When Haitians arrive in Miami, a place with a reputation of not being hospitable to people who are black and poor, they lack the proficiency in English and other educational characteristics to be eligible for any but the lowest-paying jobs. Too quickly, Haitian immigrants find themselves in the least desirable jobs, barely able to meet their family's basic needs with what they earn. In their encounters with the “native” population, they hear and see evidence that they are not held in high regard by their neighbors. This prejudice is linked to their skin color, but it extends to accusations of bringing disease. To add a final degree of difficulty to the process of adjustment, there are few extended family members available to moderate the stress and burden of immigrant status in daily life, no aunt Marie to take the children when one has to work late, no grandmother to spoil them, no community to monitor their behavior.

To varying degrees, the racial prejudice, lack of education, accusations of bringing disease, and lack of extended family resources affect all Haitian immigrants in South Florida, yet most have succeeded in making some sort of life without resorting to crime or delinquent activity. According to our theory, those who become involved in crime and delinquent behavior may have something more in their backgrounds, the experience of trauma.

The clinical construct most strongly identified with traumatic life experience is Post Traumatic Stress Disorder (Kessler et al., 1995) which holds that traumatic life events are processed on an unconscious level (for the “structural violence” variant of this theory, see Baer, Singer, and Susser, 2003: 213-216). When they have been too severe and/or persistent in the life of the individual, the memories of traumatic events can interfere with the functioning of adaptive mechanisms on cognitive, emotional, and behavioral levels (Brenner et al., 1995; Breslau et al., 1992; Yehuda et al., 1995).

Post-Traumatic Stress Disorder (PTSD) can be an extremely debilitating condition that occurs after exposure to a terrifying event or ordeal in which grave physical harm occurred or was threatened. Traumatic events that can trigger PTSD include violent personal assaults such as rape or mugging, natural or human-caused disasters, accident, or military combat. Survivors of traumas, people who witnessed traumatic events, and families of victims are among those who develop PTSD (Romero-Daza and Singer, 1997; Romero-Daza, Weeks, and Singer, 2003).

Many people with PTSD repeatedly reexperience the ordeal in the form of flashback episodes, memories, nightmares, or frightening thoughts, especially when they are exposed to events or objects reminiscent of the trauma. Anniversaries of events can trigger symptoms. People with PTSD also experience emotional numbness and sleep disturbances, depression, anxiety, and irritability or outbursts of anger. Feelings of intense guilt are also common. Most people with PTSD try to avoid any reminders or thoughts of the ordeal. PTSD is diagnosed when symptoms last more than one month.

About 3.6% of U.S. adults (5.2 million people) have PTSD during the course of a given year (NIMH, 2004). PTSD can develop at any age, including in childhood. Symptoms typically begin within three months of a traumatic event, although occasionally they do not begin until years later. Once PTSD occurs, the severity and duration of illness vary. Some people recover within six months, while others suffer much longer. Co-occurring depression, alcohol or other drug abuse, or another anxiety disorder are not uncommon. The likelihood of treatment success is increased when these conditions are appropriately diagnosed and treated. People who have been abused as children or who have had other previous traumatic experiences are more likely to develop the disorder.

Nevertheless, not all people who have traumatic experiences develop symptoms of PTSD. Cultural background and social support (cf. Eisenman et al., 2003), history of head trauma (Keenan et al., 2003), and hippocampal volume (Gilbertson et al., 2003) may affect whether or not people exposed to trauma develop symptoms. Experience with trauma may be more likely in some cultural contexts than in others. For example, people living in Israel or some parts of the developing world have high rates of exposure to violent acts (Eisenman et al., 2003). How people respond to traumatic experience may vary from one cultural context to another. Israelis exposed to terrorism demonstrated rates of PTSD symptoms similar to those found in Guatemalans 20 years after traumatic experience (Bleich, Gelpoff, and Solomon, 2003; Sabin et al., 2003). Detection of traumatic experience and diagnosis of PTSD may present added difficulties in culturally distinct populations (Eisenman et al., 2003).

We theorized that young people who have seen terrible acts of violence or gruesome events, or who live in the thrall of abuse in their own homes would exhibit some symptoms linked with PTSD that would predict poor performance in school, anti-social behavior, and violence. Intrusive thoughts about trauma would disrupt attention in school, either directly, or indirectly through sleep disturbance. Attempts to use drugs to prevent or moderate these symptoms could ensue. Startle reactions and defensive behavior would attract attention and ridicule of fellow students. This array of difficulties, coupled with feeling embattled in surroundings populated by people of different cultural backgrounds might result in the youth's seeking social contexts in which participants have similar life experiences and are similarly angry and frustrated about them. Individuals with these characteristics could easily be expected to form peer-controlled groups that engage in delinquent behavior, or, as they are often called, gangs. Recent research on moderation of the effects of trauma on children (cf. Stein et al., 2003) indicates that if the preceding ethologic theory were correct, cognitive behavioral intervention might have the effect of preventing this kind of behavior. The study described here attempted to focus attention on traumatic experience among Haitian youth as a strategy for understanding the emerging formation of peer-controlled groups that engaged in drug-selling and other delinquent activity.

METHODS

Participants

We spent three years identifying and recruiting 291 Haitian youths at risk through direct, street-based contact with networks of informal social relations. Each of the new recruits, after a consent procedure involving parents of those under 18 years of age, responded to an interview schedule consisting of questions on sociodemographic characteristics, immigration history, self-reported involvement in use of illegal drugs, involvement in peer-dominated social groups, and traumatic life experiences. The latter set of questions inquired about the participants' exposure to parental abuse, forcible sex, witnessing death or serious injury by accidental or intentional violence, serious illness, or imprisonment. If the respondents answered in the affirmative to any of these questions, they became eligible for a later administration of the CAPS.

Because the interview schedule's sensor for traumatic experience tended to include rather than exclude, we could reasonably expect a significant portion of the included respondents not to show symptoms for PTSD. The 228 people identified as having experience with trauma included 29% who only had one variety of trauma to report, but this criterion for identifying traumatic experience did not reflect duration, intensity, or severity of that experience. A person exposed to constant beatings and threat of physical harm over two years had different reasons for showing signs of Post-Traumatic Stress Disorder than a person who had witnessed a single violent incident, yet both would show the same number of traumatic experiences in our crude sensor. The CAPS provided an opportunity to characterize traumatic experience in terms of variety, frequency, or intensity. Although it would have beets ideal to administer the CAPS to all individuals who self-reported traumatic experience in their responses to the interview schedule, we were able to do so only with a subset of these individuals.

Several problems in re-recruitment of this subpopulation presented barriers to their continued participation, including change of address before the field team could recontact them, difficulties in scheduling the administration of the CAPS, and refusal to participate in the second stage of the study. A total of 47 of the 228 participants who had a history of traumatic experience took part in the CAPS.

Procedures and Assessment of Trauma

The Clinical Assessment of PTSD Severity (CAPS) is a structured interview that has demonstrated reliability for detecting symptoms of PTSD and measuring its severity (Blake et al., 1990). The Structured Clinical Interview for DSM-IV (SCID) (First et al., 1996) is the standard method for reliably diagnosing PTSD and ruling out other Axis I disorders. The investigators used it to rule out anxiety disorders other than PTSD, mood disorders, psychosis, and drug-use disorders. The Bellevue Adolescent Interview schedule (BAIS) (Lewis et al., 1988) assesses child abuse in both physical and sexual victimization. It distinguishes between physical and sexual abuse and helps to compare the impact of these two types of trauma. A psychiatrist, with the help of one other clinician, administered these instruments to a total of 47 study participants. The clinicians administering the CAPS received training in its administration at the same time from the same instructor, Dr. Daniela David of the University of Miami's Department of Psychiatry and Behavioral Sciences. This training included establishment of high reliability among raters through repeated practice administrations.

Analyses

Responses to the structured interview were entered into an SPSS file for analysis, producing frequency distributions of sociodemographic characteristics of the study population. The item-by-item responses to the CAPS as well as summary scores also were entered into an SPSS file and that was merged with the structured interview file for the purpose of comparing CAPS respondents with non-respondents. Summary CAPS scores were hypothesized to be independent variables predicting self-reported gang involvement, as well as violent and delinquent behavior.

RESULTS

The team of field workers recruited 291 respondents from whom they gathered sociodemographic and self-reported data on risk and delinquent behavior to contextualize the clinical assessments performed by the clinicians. Table 1 shows that the participants included 35% females, and the participants' ages ranged from 12 to 25, with 70% 19 or under. Two-thirds of the participants in the structured interview were teenage boys born in the U.S., and 91.7% of the participants, when asked about their primary cultural identification, identified themselves as Haitians or Haitian-Americans. The remaining 8% had claimed some form of Haitian identity during the recruitment process, but when responding to the formal question they identified African American, Black, Bahamian, or “other.” Preliminary work had identified three terms to identify groupings of Haitian youths not supervised by adults: gang, clique, and group. More than one-third of the participants had taken part in activities of one or more of these agglomerations, and 15% were currently active in a self-identified gang at the time of the interview. Most of them were exposed to gangs, cliques or groups and their activities early in their teenage years, between 12 and 15 years old, specifically during middle school.

TABLE 1.

Age and Sex of Participants

				Count
		SEX		Total	%
		MALE	FEMALE
Age at Interview	12	1	1	2	0.7
	13-14	9	10	19	6.5
	15-17	42	58	100	34.4
	18-22	113	32	145	49.8
	23-up	22	3	25	8.6
Total		187	104	291	100.0

Youths recruited from Homestead/Florida City were significantly more likely to report traumatic experience than those recruited in Little Haiti (see Table 2). Accidents and natural disasters were by far the most frequently reported traumatic experiences for the total group reporting trauma (Table 3). Among the individuals assessed for PTSD (see Table 4), assault by strangers was the most frequently reported source of trauma, but it was not strongly associated with PTSD symptoms or other psychological problems.

TABLE 2.

Trauma by Neighborhood

				Count
		NEIGHBORHOOD
		LITTLE HAITI	HOMESTEAD	Total	%
TRAUMA	NO	52	11	63	21.6
	YES	149	79	228	78.4
Total		201	90	291	100.0

Chi-square = 6.827 p = .009

TABLE 3.

Self-Reports of Trauma

Types of Trauma	N*	%
War Zone	6/228	2.6
Sex with Older Person	52/228	22.8
Imprisonment	45/228	19.7
Serious Accident	115/228	50.4
Natural Disaster	171/228	75.0
Life Threatening Illness	37/228	16.2
Sexual Assault by Family Member	8/228	3.5
Assault by Family Member	39/228	17.1
Sexual Assault by Stranger	25/228	11.0
Assault by Stranger	59/228	25.8
Forced into Sex	27/228	11.8
Forced into Sex by You	8/228	3.5

^*Respondents reported more than one variety of trauma, so numerators add up to more than 228.

TABLE 4.

CAPS and SCID Results

Trauma	N*	%
War Zone	4/47	8.5
Sex with Older Person	14/47	29.8
Imprisonment	31/47	66.0
Torture	6/47	12.8
Serious Accident	16/47	33.3
Natural Disaster	27/47	55.6
Life Threatening Illness	11/47	23.4
Sexual Assault by Family Member	6/47	12.8
Assault by Family Member	15/47	31.9
Sexual Assault by Stranger	4/47	8.5
Assault by Stranger	32/47	68.9

^*Respondents reported more than one variety of trauma, so numerators add up to more than 47.

Assault by family members and assault by police were two varieties of trauma reported by the respondents to the CAPS (see Table 5). Respondents who reported these kinds of assaults attributed great distress to them. Of the CAPS respondents, 71% received a total CAPS score of one or higher (range of CAPS–frequency 1-55; combined frequency and intensity 2-109; intensity 1-57). Of these individuals, 13 were diagnosable and one more was symptomatic but not diagnosable for PTSD. Symptoms reported by these individuals (both diagnosed and symptomatic) included memory lapses, outbursts of anger, and emotional distress.

TABLE 5.

Traumas Causing the Most Distress

Triggering Traumas	N*	%
War Zone	1/47	2.1
Imprisonment	4/47	8.5
Torture	1/47	2.1
Serious Accident	4/47	8.5
Natural Disaster	1/47	2.1
Life Threatening Illness	1/47	2.1
Sexual Assault by Family Member	3/47	12.8
Assault by Family Member	9/47	19.1
Sexual Assault by Stranger	1/47	2.1
Assault by Stranger	7/47	14.9
Other Traumatic	11/47	23.4
Did not find any of these experiences distressing	4/47	8.5

The relationships among traumatic experience, traumatic disorder, and delinquent behavior did not emerge from the data as theorized. Self-reported gang affiliation and experience with trauma were not related (Chi-square = 1.811, P = .404), nor was self-reported drug use related to traumatic experience (Chi-square = .135, P = .713; see Table 6). Similarly, experience with trauma was not related to specific delinquent activities, such as carjackinig, burglary, drug dealing, or robbery, using dichotomous versions of the summary traumatic experience variable and the delinquent behavior variables. In the 47 cases where the participants' traumatic experiences were more thoroughly characterized by the CAPS score and diagnostic criteria, the relationships between the effects of trauma and specific aspects of delinquent behavior also proved not to be statistically significant (see Table 7).

TABLE 6.

Relationships (Chi-square) Between Traumatic Experience and Delinquent Behavior

Self-reported Behavior	Yes	No	Chi-square	Significance
Former affiliation with gang	95	192*	1.595	P = .450
Current affiliation with gang	45	243*	1.811	P = .404
Current affiliation with clique	33	217**	2.820	P = .420
Current affiliation with group	8	242**	2.563	P = .464
Does current gang/clique fight?	56	20***	.332	P = .255
Have you fired a gun in a fight?	32	70***	1.249	P = .264
Gang/clique dealt drugs in last 12 mo.	58	52***	3.006	P = .222
Gang/clique carjack in last 12 mo.	47	64***	2.019	P = .364
Gang/clique burglary in last 12 mo.	38	73***	1.517	P = .468
Gang/clique robbery in last 12 mo.	41	70	1.597	P = .450
Gang/clique used mj in last 12 mo.	80	33***	4.625	P = .201
Gang/clique used Coc. in last t2 mo.	11	102***	2.644	P = .450
Gang/clique used Heroin in last 12 mo.	2	111***	2.488	P = .477
Gang/clique used Crack in last 12 mo.	7	106***	2.164	P = .539
Respondent ever dealt drugs	62	225	4.174	P = .124
Respondent ever car-jacked	26	260	6.844	P = .033a
Respondent ever burglarized	28	258	2.865	P = .239
Respondent ever robbed	31	255	3.120	P = .210
Respondent ever used illegal drugs	113	174	.135	P = .713
Respondent ever used Marijuana	110	174	.850	P = .654
Respondent ever used Cocaine	18	269	2.649	P = .266
Respondent ever used Crack	2	285	.643	P = .725

^aRespodents without traumatic experience were significantly more likely to have personally participated.

^*Total does not equal 291 because of refusals to answer.

^**Total does not equal 291 because not all youths recognize all terms for youth groups.

^***Analysis incladed only respondents with current or historic aiffiliation with gangs, cliques, or groups.

TABLE 7.

Relationships (ANOVA) Between CAPS Score and Delinquent Behavior

Self-reported Behavior	Yes	No	F	Significance
Former affiliation with gang	18	28*	.266	P = .609
Current affiliation with gang	16	30*	.828	P = .424
Current affiliation with clique	6	26**	.703	P = .501
Current affiliation with group	2	30**	1.228	P = .303
Does current gang/clique fight?	15	5***	.206	P = .815
Have you fired a gun in a fight?	11	15***	.533	P = .473
Gang/clique dealt drugs in last 12 mo.	19	9***	.387	P = .682
Gang/clique carjack in last 12 mo.	15	13***	.575	P = .567
Gang/clique burglary in last 12 mo.	13	15***	1.659	P = .202
Gang/clique robbery in last 12 mo.	16	12***	1.085	P = .347
Gang/clique used mj in last 12 mo.	24	6***	.270	P = .765
Gang/clique used Coc. in last 12 mo.	5	25***	.353	P = .704
Gang/clique used Heroin in last 12 mo.	1	29***	.624	P = .541
Gang/clique used Crack in last 12 mo.	2	28***	.604	P = .521
Respondent ever dealt drugs	18	29	.748	P = .392
Respondent ever car-jacked	9	38	.294	P = .590
Respondent ever burglarized	9	38	.011	P = .918
Respondent ever robbed	10	37	1.525	P = .223
Respondent ever used illegal drugs	26	20	1.219	P = .275
Respondent ever used Marijuana	26	20	1.219	P = .275
Respondent ever used Cocaine	6	40	1.010	P = .373
Respondent ever used Heroin	1	45	1.267	P = .292
Respondent ever used Crack	2	44	1.058	P = .356

^*Total does not equal 47 because of refusals to answer.

^**Total does not equal 47 because not all youths recognize all terms for youth groups.

^***Analysis included only respondents with current or historic affiliation with gangs, cliques, or groups.

Slightly fewer than one-third of the CAPS interviewees, all of whom had traumatic experience, met criteria for PTSD, and one additional individual was assessed to have some symptoms of that disorder (see Table 8). More than one-quarter of the CAPS respondents (13) had zero scores on the CAPS, demonstrating no apparent effects of their self-reported traumatic experiences that included seeing a brother drown, having a friend who was shot to death, being hit by a car, and being flooded out by Hurricane Andrew. The trauma of individuals with zero scores on the CAPS involved natural disasters, attacks by police or other strangers, or sex with a person more than five years older than the respondent. Nevertheless, some individuals with scores on the CAPS had single traumatic experiences of short duration.

TABLE 8.

CAPS Results

	Yes	%	No	%	Symptoms/Sub-threshold	%	Total
Diagnosis of PTSD	13	27.6	33	70.2	1	2.1	47
Self-Report of Memory Lapses	6	12.8	35	74.5	6	12.8	47
Self-Report of Angry Outbursts	12	25.5	31	66.0	4	8.5	47
Self-Report of Emotional Distress	11	23.4	26	55.3	10	21.2	47

In an attempt to probe the intensity of delinquent behavior among CAPS respondents, the investigators constructed a composite score of that behavior by combining self-reports of seven different varieties of delinquent behavior, including fighting, carjacking, auto theft, robbery, drug trafficking, burglary, and fencing of stolen goods. This composite score was also statistically unrelated to CAPS score (R =.275, R-square = 0.076; F = 2.045, Sig. = 0.165), or diagnosis of PTSD (F = 1.359, Sig. = 0.255).

We pursued the alternative hypothesis that a specific aspect of the symptoms attributed to PTSD might predict violent behavior among youths with traumatic experience. A separate part of the CAPS involves irritability and sudden violence, and we used this variable to attempt to predict the array of undesirable behaviors presented earlier. Table 9 presents the individual behaviors studied in this analysis, which produced no significant relationships. We also analyzed the composite variable of combined self-reported behaviors, and the one-way ANOVA yielded an F of .726. Regardless of how we constructed bivariate analyses, among the Haitian youth recruited into this study, we could find no relationship between delinquent behavior and PTSD and its symptoms.

TABLE 9.

Relationship Between Outbursts of Anger and Violent Behaviors

	Yes	No	Total	Chi-square	P
Gang membership	15	28	43	1.601	.659
Car jacking	8	35	43	0.839	.840
Burglary	8	35	43	3.675	.299
Auto theft	14	29	43	4.114	.249
Robbery	10	33	43	2.198	.532
Fired gun in fight	11	31	42	0.342	.559

How these youth came to circumstances in which they were exposed to abuse by household members and brutal treatment by police appears to depend primarily on how they and their families have fared after immigration, rather than what happened before or during that process. The following clinical notes taken during the administration of the Clinician-administered post-traumatic stress (CAPS) illustrate the participants' perceptions of their traumatic experiences.

Case #1

A 16-year-old boy witnessed his 1-year-old sister raped by a family friend two years ago. The friend was never caught. He felt used by the friend. He carries feelings of guilt.

Case #2

The boy says that being stabbed was an act of injustice because police did not do anything about it even though he told them where to find the person at school. His attacker was never arrested. He also said, if it were he [who had been accused] he would have been arrested; he still wants revenge.

Case #3

At age 1, this female teenager was forced into sex with the next door neighbor while the aunt was working. The neighbor was a friend and trusted by the aunt to watch over her. Also, her uncle once walked into the bathroom while she was taking a shower and talked to her about sex. She became terrified of the uncle.

Case #4

He saw his little brother pushed into the lake and drowned in front of him. He was pushed into the water, too, but was saved after swallowing a lot of water. He feels lucky to be alive. He has guilty feelings and anger about this incident.

Case #5

This 20-year-old male witnessed a “lady friend” getting shot in front of him. The police never showed up, even after they called several times. Referring to the police, he said, “It's all about the system. The system is set up a way to mess everything up. It's all Babylon. This society is about only the strong who can survive. It's not about black and white anymore.”

Case #6

This boy was hit in the head by his father and sustained a long gash. The father is in jail now. The boy feels responsible. He saw his father hit in the head by someone else a year ago.

Case #7

While resisting arrest, this boy had sustained a broken leg at age 9. He also sustained a broken arm and a skull fracture by his mother at age 11 and 12 respectively. He was involved in shoplifting and car theft at an early age.

These clinician's assessments take the violent incidents in the study participants' backgrounds out of context, pointing out their key features and alluding to an aftermath. Responses of the young people in these seven cases vary, from guilt in cases 6 and 1, to cynicism in case 5, to desire for revenge in case 2, to ongoing terror in case 3. All responses hold the potential for future harm to the victim. The guilty feelings expressed by the boy in case 6, on the most immediate level may cause him to feel that he deserves further physical abuse, if he remains in the living arrangement in which his father injured him. In both cases 6 and 1, the youths' guilt may lead to feelings of reduced self-worth, which can in turn lead to self-destructive behavior and even suicide. The cynicism shown in case 5 is apparently channeled into a broader condemnation of “Babylon” in Rastafarian style. The boy who wants revenge may resort to violence, and the young woman who is afraid of men and sex may have major problems in adapting to adult life. These residual feelings, however, do not necessarily translate into measurable disorder, and even measurable disorder does not predict delinquent behavior.

DISCUSSION

The administration of the CAPS to people from a cultural background as distinctive as that of Haitians would seem to present problems of language and cultural interpretation, but in fact, the young Haitians encountered in street environments primarily spoke English. None were monolingual Creole speakers. Furthermore, the pretest revealed no difficulties in administering the CAPS to young, English-speaking Haitians. In the absence of full translation, back-translation, and item-by-item checks for cultural appropriateness, we cannot assume that the administration of the CAPS was completely without misunderstandings of a cultural nature. Nevertheless, the data presented here suggest that no systematic problems with administrations occurred during this study. Responses varied significantly, and the summary scores ranged from 0 to 109 with scores distributed throughout the range, suggesting wide variation in frequency and initensity of response to traumatic experience.

The lack of any demonstrable relationship between CAPS scores or PTSD diagnosis and involvement in delinquent behavior or unsupervised groupings of Haitian youth suggest that the process of entering and sustaining this kind of activity has a much subtler relationship with traumatic experience than we originally hypothesized. At minimum, we can conclude that traumatic experience as measured by CAPS is neither a necessary nor sufficient cause of joining gang-like groups or engaging in delinquent activity. Just because the CAPS and SCID instruments do not detect the effects of trauma, however, does not mean that trauma has not had some other kind of impact on those who have experienced it. Recent studies by DeBelles (2002), Heim and Nemeroff (2001), and Yehuda (2001) suggest that, especially in children, traumatic experience causes the nervous system to assume a state of constant arousal that is not susceptible to measurement by means of accepted methods. This subtle, sub-clinical state may be related to misuse of drugs (DeBelles, 2002) or anxiety disorders (Heim and Nemeroff, 2001). We cannot assume that the Haitian youths who have had a traumatic experience now live with a state of constant arousal, but such a state could help to explain the receptivity of some youths to the overtures of their delinquent peers. In the absence of a measurement for detecting this subtle condition, we can only speculate on its effects based on correlations with misuse of drugs (cf. DeBelles, 2002) and anxiety (cf. Heim and Nemeroff 2001).

Findings of very recent studies in populations heavily exposed to trauma but variable in PTSD and its symptoms (Bleich, Gelpoff, and Solomon, 2003; Sabin et al., 2003; Stein et al., 2003) suggest that people from culturally distinct groups may have ways of handling traumatic experience that affect the rates at which PTSD is detectable among them. Our results would place Haitian youth between the Latin Americans and the Israelis in this gradient of response to traumatic experience.

On the other hand, the two-thirds of study participants who took part in the CAPS assessment did not meet criteria for PTSD. Furthermore, the clinicians' qualitative assessments of their conduct in the interviews found most of them to have normal social skills and no outward signs of psychopathology. Most responded to uncomfortable questions with grace and calm, and they presented themselves as having adapted to their difficult situations. This information gives rise to another hypothesis about the emergence of gang-like and delinquent behavior among these young people-that they represent attempts by human beings who are not yet mature to adapt to difficult circumstances of socioeconomic status, truncated extended family, race/ethnic stigma, and traumatized history, among other serious problems. The present study has collected data on these environmental factors, and will address them in subsequent articles.

CONCLUSIONS

Although traumatic experience may still play a role in mental health outcomes among children, childhood victimization among Haitian children does not lead inevitably to PTSD, and it does not appear to be related to the undesirable behaviors associated with unsupervised youth, including formation of gangs. Youngsters who demonstrate high resilience to the effect of serious trauma require further study to differentiate among reaction to trauma that is not measurable by existing instrumentation, intrapsychic resilience, and culturally patterned resilience. A prospective study to assess responses of Haitian youth to stressors found in their local immigrant communities would help to identify appropriate interventions that address these stressors and moderate responses to them. Response to trauma may have more importance in Haitian youth than CAPS-detectable PTSD, necessitating detection of low cortisol, and high CRF and norepinephrinie levels in response to stress and trauma. The present study engaged in none of these assessments, but they could be important in unraveling the relationship between trauma and undesirable behavior among Haitian youth.