ARCHIVED: Chapter 3: Coping with Becoming a Teen: When You Have Been Maltreated as a Child

OBJECTIVE 1

Using parallel substance use items drawn from the 2005 version of OSDUS and the MAP, what are the overall rates of substance use in the MAP sample and how do rates of substance use among MAP youth compare with Ontario youth who do not report some level of lifetime involvement with Ontario child welfare?

Findings from MAP and OSDUS Youth

The first set of analyses explored the overall rates of substance use in the MAP sample and examined differences between rates of substance use among youth participants in the MAP study compared with youth participants in the OSDUS study who reported no lifetime involvement with child welfare.

Table 1 depicts rates of self-reported substance use among youth in the MAP and OSDUS samples, showing percentage differences in reported use across the two samples. As maltreatment among OSDUS youth was not assessed, the OSDUS subsample used may include youth who have experienced maltreatment (but were not involved with Ontario child welfare). To better understand the reported proportions, the relative risk ratios were calculated. The highlighted relative risks are those considered interpretable as detailed in the analysis plan. Risk values indicating a substantial heightened risk (i.e., where the confidence interval had values above 1.0) were found for cannabis and other drugs.

Discussion of Findings

MAP youth are significantly less likely to use alcohol in the past year and, to a lesser extent, to have ever used alcohol. This finding may relate to access, among other things, as most MAP youth were in alternate care environments. Substances other than alcohol may be more likely to be acquired outside of the residential environment, at school, or through friends or previously established connections.

Alternatively, youth who have been involved with child welfare may potentially avoid alcohol due to early experiences with parental alcohol use in which the negative consequences of drinking were particularly salient (i.e., parental intoxication co-occurred with parental abuse or neglect, see Wekerle & Wall, 2002). Perhaps, other drug use was either less frequent among parents or less salient (i.e., parents hid use of drugs from child's view, but not alcohol).

Finally, alcohol involvement may have been avoided due to the involvement of child welfare caseworkers. Caseworkers may be particularly alert to alcohol use and alcohol problems, given greater common knowledge. Without specific substance abuse training, workers may be less attuned to the effects of other drugs, making it easier for youth to hide cannabis and other substance use. Such speculations can be considered in future research on alcohol use and child welfare youth. However, it is advisable not to over- interpret the finding of a lesser risk of alcohol use, given that different results may emerge with a larger or different sample.

MAP youth were more likely than OSDUS youth to report lifetime and past-year cannabis use, lifetime other drug use, and frequent other drug use in the past year. Thus, youth involved with child welfare may deem drug use as more acceptable, more accessible, or better suited functionally to their coping needs.

The implications of these findings are that child welfare staff and caseworkers may not be as sensitive to the signs and symptoms associated with youth's use of these other drugs. Schools offering social work curricula provide a generalist degree, rather than specific child welfare training, and do not comprehensively cover mental health and addiction issues. When training in the child welfare setting, workers similarly do not receive strong support for addiction-related knowledge and a continuing education workshop is insufficient to provide required psychopharmacology, assessment, and interviewing skills. Knowledge dissemination and public health programming is primarily focused on alcohol use and the harms associated with binge drinking. Receiving far less attention in terms of prevention programming and popular media campaigns are cannabis and, to a greater extent, other drug use.

Although the public health focus on alcohol use and binge drinking responds to the needs of the general youth population where rates of past year alcohol use are much higher than rates of past-year cannabis use (Adlaf & Paglia-Boak, 2007; Johnston et al., 2007), the present findings suggest different needs for youth within the child welfare system. In particular, to identify youth based on need and type of intervention (e.g., Drug Abuse Screening Test [DAST] for adolescents, Martino, Grilo, & Fehon, 2000), child welfare caseworkers could receive training to recognize signs and symptoms of drug use, as well as assessment tools, such as surveys and youth self-report questionnaires, for spotting drug use and problems.

While not shown in Table 1, analyses indicated no significant difference between MAP youth and OSDUS youth on self-reported past-year driving after drinking (i.e., driving after consuming two or more drinks of alcohol), or driving after marijuana use (i.e., driving within one hour of using cannabis). This latter finding is not emphasized, however, as there are no data comparing whether involvement with child welfare influences whether a youth obtained a driver's licence.

OBJECTIVE 2

Among youth in the MAP sample, what is the relationship between specific types of childhood maltreatment experienced and types of substance use? Is there an association among multiple forms of maltreatment and particular substances, or particular types of maltreatment and multiple substance types used?

Findings from the MAP

The second set of analyses examined the relationship between the type of childhood maltreatment reported by MAP youth and self-reported substance use. Both specific (i.e., specific maltreatment type, specific substance type) and general (i.e., multiple forms of maltreatment and specific substance type, multiple forms of maltreatment and multiple substance types) relationships were explored.

As shown in Table 2, rates of substance use were based on self-reports of severe levels (e.g physical and sexual abuse) based on items from the Childhood Experiences of Violence Questionnaire (CEVQ). These analyses were limited to severe abuse experiences to approximate the child welfare threshold for defining and/or obtaining evidence related to maltreatment; such an approach has proved useful in studies with community samples (Wekerle & Wolfe, 1998) and child welfare samples (Wekerle et al., 2001). As such, physically abusive behaviours do not refer to spanking or slapping, even though these could be maltreatment events. We selected such behaviours as hitting, punching, biting, kicking, choking, scalding, or attacking in some way. For sexual abuse, flashing or exposure were not included. We selected such behaviours as fondling, being forced to touch in a sexual way, or having sex forced upon the child. For witnessing domestic violence, while witnessing verbal abuse may well be damaging, we selected only witnessing physical assault among adults in the home (i.e., physical assault involving parents, step-parents, or guardians; or a parent, stepparent, or guardian and another adult in the home). This approach also serves to maintain a definitional consistency with the MAP caseworker data, where caseworkers were required to dichotomously indicate (yes/no) whether there was substantiated maltreatment or substantial risk of maltreatment, as per Ontario law.

Table 2 also shows that a history of severe physical abuse and witnessing domestic violence is associated with a greater relative risk of substance use across a range of substances. MAP youth who reported severe physical abuse or witnessing domestic violence on the CEVQ are more likely to report lifetime and past year alcohol use, cannabis use, and other drug use, compared with youth with other maltreatment forms. For example, compared with youth without histories of severe physical abuse, those who reported severe physical abuse were at a 68% higher risk to report cannabis use in the past 12 months and 41% higher risk to report lifetime cannabis use. Findings also showed a greater relative risk of lifetime alcohol and other drug use among youth who had witnessed domestic violence. While not recorded in the table, the results also show that the number of types of severe maltreatment reported was associated with an 50-90% increased likelihood of lifetime and past-year alcohol use, cannabis use, and other drug use.

Among the four interpretive or perception-focused Childhood Trauma Questionnaire (CTQ) items (i.e., "I believe that I was...."), perceiving oneself as having been emotionally abused ("I believe that I was emotionally abused") was associated with greater relative risk of alcohol and cannabis use. Specifically, youth who believed that they were emotionally abused were at a 37% greater risk of reporting lifetime alcohol use, a 32% greater risk of reporting lifetime cannabis use, and a 116% greater risk of reporting problematic drinking. Also, the greater the number of perception-focused maltreatment items youth reported, the greater the likelihood (20% to 50%) of ever having used alcohol and cannabis.

In summary, youth who reported a history of severe physical abuse and witnessing domestic violence showed greater relative risk of reporting in the alcohol and cannabis categories. Youth with cumulative maltreatment scores were more consistently associated with alcohol and drug use, compared with scores pertaining to a single maltreatment type. Those who self-reported emotional abuse showed greater relative risk of lifetime alcohol and cannabis use, as well as greater risk of problematic drinking. In addition, the likelihood of ever reporting lifetime and past-year alcohol use, as well as past-year cannabis use, was higher among those who reported perceiving themselves as experiencing multiple forms of maltreatment, compared with those who reported single or fewer types of experiences.

MAP caseworker information on substantiated forms of maltreatment was obtained from a subsample of MAP youth (n = 50) at the time of intake. Table 3 shows substance use data among youth whose caseworkers reported a history of maltreatment compared with those from youth whose caseworkers did not report maltreatment history. These results, based on a smaller intake sample, should be considered preliminary but potentially suggest of patterns reflecting relationships between official detection of maltreatment type and type of substance use self-reported by youth. Although the sample size is small, this represents a good starting point for examining relationships between official detection of maltreatment and substance use. While the relative risk of workers' assessment on alcohol and substance use is mostly in the predicted direction, the result is not significant at conventional levels due to the small sample size.

 

Discussion of Findings

The present results highlight physical and emotional abuse, as well as multiple forms of maltreatment, as contributing factors to youth substance use among child welfare teens, when youth self-reports are taken into account. When examining documented abuse (i.e., reported by caseworker), trends emerge for sexual abuse, which can be examined from an exploratory approach, considering potential clinical significance. Thus, it is suggested that youth whose caseworkers reported a history of sexual abuse were more likely to use alcohol and cannabis in the past year. This is consistent with previous findings from the literature indicating a significant association between childhood sexual abuse and alcohol and drug use (e.g., Dube et al., 2005; Moran et al., 2004). That childhood sexual abuse was associated with substance use with caseworker reports, but not youth self-reports, may be due to a reporting bias among youth. That is, youth may be less likely to self-report sexual abuse experiences and more likely to report physical or emotional abuse experiences, given the greater shame and guilt components suggested with sexual abuse (e.g., Wekerle et al., 2006). In addition, caseworker reports may be biased toward the most severe forms of childhood sexual abuse. Although the analyses of self-reported data involved severe sexual abuse, the most severe cases may be those detected by caseworkers and, within a small sample, it may be these individuals whose histories of childhood sexual abuse are most likely to predict adolescent substance use. To clarify these relationships, future research should pursue a comparison of caseworker reports of substantiated maltreatment alongside youth self- reports.

Although analyses did aggregate across caseworker reports of maltreatment types (0 = no maltreatment to 4 = positive on all forms of maltreatment), this cumulative score was not significantly associated with substance use in this small subsample. Similarly, since witnessing domestic violence and perceptions of abuse were not assessed by caseworker reports, no comparisons can be made between self-reports and caseworker reports for these types of maltreatment experiences. Again, interpretive caution is advised when considering non-significant results.

Further examination of the link between childhood maltreatment and substance use is clearly warranted. Specifically, understanding the mechanisms involved in the relationship will provide important information regarding prevention and intervention targets. The MAP is currently linking to participating agency databases to abstract the entered maltreatment codes per MAP youth. However, analyses involving MAP intake data will not enable comparisons between MAP youth and youth participants from the OSDUS due to differences in assessing substance use. When substance use information is collected at intake, questions are based on items compiled from both the Monitoring the Future Study (Johnston et al., 2007) and the Youth Risk Behavior Surveillance Study (Centers for Disease Control and Prevention, 2006). The MAP intake substance use items form a continuous measure of days or times of use during a specified period (last year, last month), as well as querying the age of onset of use.

To continue to look at the potential association between maltreatment and substance use among a sample of child welfare involved youth, we return to the MAP Year-1 dataset with substance use questions directly comparable to the OSDUS.

OBJECTIVE 3

How do rates of substance use differ for male and female youths in the MAP versus the OSDUS data?

Gender Differences in Substance Use: MAP vs. OSDUS

The third set of analyses examined whether there were any gender differences in substance use within the MAP and the OSDUS samples. Results for these analyses are shown in Table 4 (gender differences between MAP and OSDUS youth) and Table 5 (gender differences within the MAP and OSDUS samples). Results are considered preliminary considering the drop in sample size and power to detect significant differences when the data are split on gender.

Table 4 shows that, overall, there were few differences between MAP and OSDUS male youth. There were, though, several significant differences between groups for female youth. In contrast to female youth in the OSDUS sample, MAP females had a greater relative risk of lifetime cannabis and lifetime other drug use. In particular, past-year frequent other drug use appears to be over three times the risk for MAP females compared with OSDUS females.

As illustrated in Table 5, comparisons within the MAP and the OSDUS samples examine how males are faring compared with females (i.e., the relative risk of males). As can be seen for child welfare youth, few gender comparisons emerge. The only significant difference is for frequent cannabis consumption (six or more times/past year), where MAP males showed 70% greater likelihood than MAP females to use cannabis frequently in the past year.

In contrast to child welfare youth, there are several male-female differences among Ontario high school youth, all in the direction of greater male use. Ontario non-child welfare involved males reported greater substance use than females on: frequent alcohol drinking; lifetime, past 12 months, and frequent cannabis use; lifetime and past year other drug use; greater hazardous drinking (i.e., more negative consequences); and greater problem drug use. This pattern of gender differences within samples raises the issue of child welfare females being a high risk group, since the gender pattern is opposite to the normative youth data.

To put into context the findings reported in Tables 4 and 5, Table 6 illustrates correlations between substance use and age for MAP youth participants only. For males, increased age was associated with more frequent alcohol use in the past 30 days, more frequent binge drinking in the past 12 months, and more frequent cannabis use in the past 12 months. For females, the majority of findings were in the opposite direction. That is, younger age was associated with more frequent cigarette smoking in the past 12 months and 30 days, and more frequent cannabis use in the past 12 months. Again, this points to MAP females being a relatively higher risk group.

Discussion of Findings

Male youth in the OSDUS reported increased involvement in alcohol, cannabis, and other drug use. These findings are consistent with other national epidemiologic surveys (e.g., www.monitoringthefuture.com ; Johnston et al., 2007) in that male youth, particularly those in the 15-19 year age range, have higher levels of involvement in alcohol (e.g., increased binge drinking), cannabis, and other drug use. Findings from the most recent OSDUHS (Adlaf & Paglia-Boak, 2007) indicate few gender differences in substance use, but this may reflect inclusion of child welfare-involved youth in that sample (whereas child welfare-involved youth were excluded from the present OSDUS analyses), or fewer gender differences at the lower grades.

 

 

That female youth in the MAP sample were more likely to report lifetime cannabis and lifetime and frequent other drug use suggests that child welfare adolescent females may be particularly likely to access drugs. The implication is that females involved with child welfare may be more involved in situations that place them at risk of interpersonal violence (e.g., involvement in drug trafficking resulting in greater exposure to weapons, threats and assaults; entering dangerous settings to procure drugs; having romantic partners that provide and/or use substances; being more likely to associate with substance using peers). These findings are consistent with Wolfe, Scott, Wekerle and Pittman (2001), who found that females with histories of maltreatment were more likely to be victims of violence than males, and reported a greater range of negative outcomes associated with their histories of maltreatment. Similarly, in their sample of male and female adolescent in-patients, Becker and Grilo (2006) found that a history of childhood abuse was associated with alcohol abuse for both males and females, but childhood abuse history was associated with drug abuse only for female adolescents.

The present findings also suggest some gender differences in the relationship between age and substance use. These findings are somewhat different from those documented in the 2007 OSDUHS (Adlaf & Paglia-Boak, 2007) in which results indicated few gender differences in the frequency of alcohol, cigarette, and cannabis use. There were, however, several significant differences in drug and alcohol use associated with grade in school, with overall rates increasing with increasing grade (Adlaf & Paglia-Boak, 2007). In the 2005 OSDUS analyses, we examined age differences by gender, which may account for some of the difference from the OSDUHS. We found that for male youth, frequent past-30-day alcohol use and frequent past-12- month cannabis use increased with increasing age. For female youth, however, frequent past-30-day and past- 12-month cigarette smoking and frequent past 12-month cannabis use decreased with increasing age.

These findings may be due to a more restricted age range for the one-year MAP sample. The overall OSDUS results are based on a sample of youth across a greater age range (grades 7- 12). In addition, grade differences in use in the OSDUS sample are largest between grades 8 and 9, and MAP youth enter the study at an average age of 15.67 years, which approximates the age of grade 9 youth. Thus, the inconsistent findings may be due to youth entering the study past the age at which the greatest increases in substance use occurs.

Given that the majority of the significant relationships between substance use and age were weak in the present analyses, further research is needed to elucidate the specific relationships between age and substance use among male and female youth. Taken together, these findings suggest that the impact of adverse childhood experiences on substance use is more of a factor for females than males, highlighting the need for further research on the mechanisms underlying these relationships, as well as gender- specific screening and intervention strategies. To date, however, research concerning gender-specific adolescent interventions has been scarce.

One exception is a recent randomized controlled trial with adolescent girls of Seeking Safety (Najavits, 2002), a structured treatment intervention for women with substance use and post-traumatic stress disorder. Findings from this pilot study found favourable outcomes in terms of substance use disorder symptoms, substance-related consequences, and trauma symptoms associated with sexual distress and sexual concerns (Najavits, Gallop, & Weiss, 2006). Further research is needed to examine gender-specific intervention strategies and the most effective methods for working with both male and female child welfare involved adolescents.

OBJECTIVE 4

Given the significant link between childhood maltreatment and symptoms of PTSD, and the negative long-term consequences associated with PTSD symptoms, is there a significant gender-moderated mediation of the relationship between childhood maltreatment and substance use by PTSD symptomatology considering sub-clinical rather than disorder levels?

Childhood Maltreatment and Symptoms of PTSD: MAP Overall Rates and Gender Differences

No study to date has considered mediation in a child welfare sample with respect to teen substance use. The purpose of the final set of analyses was to examine symptoms of post-traumatic stress disorder (PTSD) among MAP youth, and to determine whether there was evidence to support gender-moderated mediation in the relationship between PTSD symptoms and childhood maltreatment.

The data that follows is based on the MAP youth at intake (N = 388). PTSD symptomatology was assessed using the Trauma Symptom Checklist for Children (TSCC; Briere, 1996). In the present report, severity of trauma symptoms is based on the number of scales on the TSCC that exceeded the clinical cut-off score. This method for determining trauma symptom severity includes youth who self-reported clinical levels in some areas of trauma symptoms (even if not all areas were reported). This reflects a symptom cluster approach, as opposed to a total score cut-off approach, and is consistent with a developmental traumatology perspective in which the overall PTSD picture may be sub-clinical, but chronically impairing in some domain. Although the measure used in the current study cannot be used as a substitute for clinician diagnosis of PTSD (i.e., based on diagnostic criteria outlined by the DSM-IV), symptoms reported on the TSCC may be used as a proxy for severity of trauma symptoms. The OSDUS does not query trauma symptoms; thus, the present consideration of mediators explores potential causal mechanisms among the child welfare sample (MAP) only. These analyses may point to useful constructs to explore within community samples who report some lifetime child welfare involvement and a history of maltreatment.

Given that issues with MAP youth do not predominantly lie in the area of alcohol, but consistently in the cannabis area, the focus was on understanding the prediction of youth-reported cannabis use. Table 7 shows the results of the multiple regression analyses predicting past-year frequency of cannabis use. While there are differences in rates of non-cannabis use, seemingly due primarily to the contributions of female youth, the cannabis multiple regression results are presented with the proviso that a similar pattern is observed when other drug use is examined. Both cannabis and other drug categories appear to consistently emerge as substances where significant differences are found, particularly among females. On the other hand, no statistically significant relationship was observed between childhood maltreatment and substance use among male youth.

To capture this gender-specific effect, analyses presented here are based on MAP youth at intake, which represent a much larger sample size than the 1- year assessment analyses presented earlier in this report (i.e., MAP females at intake, n = 200). In the hierarchical multiple regression analyses we also control for youth age and child welfare status, entering these control variables before entering maltreatment variables in predicting the frequency of past-year cannabis use. To test mediation, PTSD is entered after maltreatment variables to be able to assess for a drop in significant prediction of maltreatment for substance use. As a means of comparison, male findings are also presented in Table 7.

As noted, preliminary analyses examining the factor structure of the Childhood Trauma Questionnaire (CTQ) for the MAP intake sample indicated a different factor structure for females compared with males (Appendix F). For females, four factors emerged, where physical abuse and emotional abuse items loaded on a single factor, and physical and emotional neglect items loaded together. It is unclear as to why this may occur. It may be that for females physical maltreatment is often accompanied by emotional maltreatment, or that females perceive the physical maltreatment to also be emotionally maltreating, or that females are more likely than males to also perceive the emotional maltreatment that issues from the perpetrator along with the physical maltreatment.

Following is the factor structure of maltreatment for MAP females:

• Factor 1: Physical Abuse & Emotional Abuse items
• Factor 2: Sexual Abuse
• Factor 3: Emotional Neglect & Physical Neglect (medical; no one to protect)
• Factor 4: Physical Neglect (not eat; parent drunk/high; wore dirty clothes)

The sexual abuse factor contained the same items for males as for females. For males, the discrete maltreatment categories may be more relevant. The factor structure for maltreatment for MAP males is:

• Factor 1: Physical Abuse
• Factor 2: Sexual Abuse
• Factor 3: Emotional Neglect and Physical Neglect
• Factor 4: Physical Neglect
• Factor 5: Emotional Abuse

Table 7 represents the regression results for both females and the males in the MAP sample, based on their intake data. The table shows two numbers separated by a slash, with the left number being the female result and the right number being the male result. In the model, step 1 shows that age is significant, indicating that younger MAP youth age is predicting frequency of past-year cannabis use for females. Interestingly, age holds as a significant factor across the regression, pointing to the concern around early entry into cannabis use. Step 2 shows that child welfare variables (e.g., status, Crown ward vs. not Crown ward) are not significant.

In step 3, the emotional-physical abuse factor (females) significantly predicts frequent past-year cannabis use, such that the more such maltreatment is experienced, the greater the number of times of past-year cannabis use. In the model, step 3 also shows that sexual abuse is significant for females only.

Step 4 shows that the TSCC information on the number of subscales exceeding the clinical elevation was entered, there by testing the mediation separately for females and for males. We see the diminishing significance of the sexual abuse for females, indicating that PTSD mediates the sexual abuse-cannabis use relationship. However, for males, sexual abuse becomes a significant predictor of frequency of cannabis use in the past 12 months, but only when PTSD symptoms are taken into account. This suggests a direct relationship for males: male sexual abuse histories directly predict cannabis use, when cannabis use is measured more along a continuum rather than a dichotomy.

Although the models in Table 7 predict a moderate amount of variance in the frequency of past-year cannabis use (i.e., 13% of variance accounted for in the female regression model; 12% of variance accounted for in the male regression model), the addition of background and psychosocial variables would likely result in an increase in model R2. For example, Best et al. (2005) found that age of cannabis use initiation, spending time with mother, and spending time with drug-using friends accounted for 27% of the variance in current cannabis use. Inclusion of personality, genetic predisposition (i.e., family history of substance use) and peer network variables would likely enhance prediction of substance use in the current sample (Feldstein & Miller, 2006).

Prior adolescent research (Wekerle et al., 2001; Wolfe et al., 2001) shows more consistent PTSD mediation results for females. In a similar vein, consistent results were found for females in the multiple regression analyses. For females, there is mediation by PTSD symptoms of the physical-emotional factor in predicting frequency of past-year cannabis use. This is the first significant mediation model to be reported for child welfare involved females in predicting substance variables. These findings show that child welfare females emerge as a high-risk group, where earlier age of entry and greater cannabis use is of concern, and that PTSD symptomatology may be an explanatory variable. While the male sexual abuse-substance use link remains to be further considered, these findings taken together consistently identify young women within child welfare at high risk for problematic substance use.

Historically, there has been a "double standard" suggested, whereby female substance use and abuse has been under-valued as a salient issue for prevention, casework management and monitoring, and treatment (Wekerle & Wall, 2002). The importance of a gendered response to female substance use and gendered sensitivity or gender-specific treatment is articulated in the recent edited compilation on addiction in women (see Greaves & Poole, 2007). The high-risk status of young women is underscored in studies of homeless youth (Erickson, King et al., 2007; Leslie, 2007), which point to child welfare involvement as prevalent among homeless youth struggling with substance, mental health, repeat pregnancy, and housing issues.

Discussion of Findings

The experience of maltreatment may be a single episode, or it may be a chronic pattern of interactions within the family. As noted above, research has documented the differing clinical features of a single- event versus chronic maltreatment (De Bellis, 2002b). Although not conducted with a random sampling of child welfare youth, prior research indicates that PTSD symptoms may function as a mediator between maltreatment and adolescent dating violence (Wekerle et al., 2001). The current report results suggest that trauma symptoms may provide the mechanism for linking additional negative outcomes with childhood maltreatment experiences, thereby highlighting PTSD symptoms as a potential broad target for reducing a range of adolescent problems. The present findings are consistent with the literature on PTSD symptoms and the self-medication hypothesis (e.g., Stewart & Israeli, 2002): substance use is used as a coping mechanism, albeit a maladaptive one, to reduce aversive affective and cognitive symptoms of trauma associated with a history of childhood maltreatment. Again, these findings have important assessment and intervention implications for child welfare involved youth.

Female youth should be carefully assessed for substance use and trauma symptomatology. Although general substance use reduction strategies are warranted among all child welfare involved youth, when a specific symptom profile is detected, tailored intervention approaches may be more effective for ensuring adequate care. Providing female youth with adaptive methods for coping with trauma symptoms may be an important entry into reducing negative outcomes, including substance use.