Homeless Youths’ Caretakers: The Mediating Role of Childhood Abuse on Street Victimization and Housing Instability

Abstract

Homeless youths who live on the streets are particularly vulnerable to victimization and continued homelessness. Identifying factors associated with housing stability and victimization while homeless can offer useful guidance for those who serve these youths. The current study examined the relationship between multiple caretakers and the unique effect of childhood abuse (physical, sexual, neglect) on past-year housing and victimization experiences. Seventy-nine substance-using, street-living youths ages 14 to 24 years completed the survey. Findings confirmed a relationship between multiple caretaker transitions and childhood sexual abuse and neglect, but not physical abuse. Sexual abuse was further associated with higher street victimization and reduced housing stability. In addition, sexual abuse mediated the relationship between multiple caretakers and past-year victimization and housing instability. These findings suggest that sexually abused homeless youths are at particular risk for future victimization and housing instability compared with other youths, and specialized intervention for these youths is indicated.

The childhood of homeless youths is often characterized by conflict, chaos, abuse, and neglect (Rabinovitz, Desai, Schneir, & Clark, 2010; Tyler, Hoyt, & Whitbeck, 2000). It is not unusual for homeless youths to have multiple transitions in living arrangements and to be raised by different caretakers including birth parents, extended family members, and foster parents in their early life course (Tyler, 2006). Overall, experiences of family instability and childhood abuse are common among homeless youths, and as these youths grow older, they often continue to experience problems such as housing instability and street victimization (Robertson & Toro, 1999). Although all homeless youths experience challenges, it is unclear whether those with more troublesome childhoods are more vulnerable and thus require targeted interventions. Our study examined the influence of early experiences of instability (number of caretakers) and childhood abuse on later experiences of housing instability and street victimization. An understanding of how early childhood experiences interact to predict future housing instability and victimization among homeless youths can provide insight into the diversity among these youths. That is, an understanding of who is most at risk for harm while on the streets, and who is likely to show the least housing stability, can guide the development of appropriate assessment tools and direct intervention or prevention efforts.

Previous research has noted that homeless youths often experience multiple transitions in caretakers and often have histories of foster care placements (Hyde, 2005; Tyler, 2006). Tyler (2006) noted that homeless youths report multiple living arrangements prior to becoming homeless, ranging from biological parents to group homes and detention facilities. There appears to be a link between placement in foster care and later homelessness, with an estimated 31 percent to 46 percent of foster care youths experiencing homelessness by age 26 (Dworsky, Napolitano, & Courtney, 2013). However, less is known regarding how multiple transitions of caretakers, including foster care and other placements, affect future homeless experiences. Some research with foster care youths indicates that multiple transitions negatively affect youths (Newton, Litrownik, & Landsverk, 2000; Oosterman, Schuengel, Slot, Bullens, & Doreleijers, 2007; J. P. Ryan & Testa, 2005). For example, multiple foster care placements are associated with higher rates of internalizing and externalizing behavior problems (Newton et al., 2000) and an increased likelihood of experiencing homelessness as a young adult (Dworsky et al., 2013). Frequent caretaker transitions can set individuals on a lifelong path of instability because multiple transitions increase the difficulty for individuals to develop support systems (Tyler, 2006). In addition, having cycled through multiple caretakers, these youths often have exhausted all of the resources these caretakers were able or willing to provide. Consequently, during times of crisis, youths with histories of multiple caretaker transitions may have a limited safety net and experience increased vulnerability to not only homelessness, but also other risks, such as victimization.

EXPERIENCES OF CHILDHOOD ABUSE

Childhood abuse is commonly reported among homeless youths, with between 21 percent and 60 percent of these youths reporting sexual abuse, between 16 percent and 40 percent reporting physical abuse, and 54 percent reporting neglect (Molnar, Shade, Kral, Booth, & Watters, 1998; Tyler & Cauce, 2002; U.S. Department of Health and Human Services [HHS], 2007). Experiences of childhood abuse or maltreatment are often a catalyst to caregiver transitions or becoming homeless (Hyde, 2005; Tyler, 2006). For example, abuse and neglect are a primary reason that youths are placed in foster care, and many homeless youths report that their experience of family abuse or neglect motivated their desire to leave home (Hyde, 2005). These experiences can have a long-term impact on youths’ future life trajectories (K. D. Ryan, Kilmer, Cauce, Watanabe, & Hoyt, 2000; Thrane, Hoyt, Whitbeck, & Yoder, 2006); homeless youths with abuse histories tend to experience more mental health problems and display more risk behaviors than those without a history of childhood abuse (K. D. Ryan et al., 2000; Thrane et al., 2006).

In addition, a connection between experiences of childhood abuse and later street victimization has been noted (Thrane et al., 2006; Tyler et al., 2000). Studies report that 62 percent of youths living on the streets experience physical or sexual victimization, 30 percent report intimate partner violence, and one-third report other victimization experiences such as being robbed (Gaetz, 2004; Slesnick, Erdem, Collins, Patton, & Buettner, 2010). Our study seeks to expand previous research by exploring how experiences of childhood abuse mediate the negative relationship between caretaker transitions and housing instability and victimization among homeless youths. Previous research indicates that different forms of childhood abuse have different outcomes. For example, youths who report prior neglect and sexual abuse are more likely to also report street victimization (Thrane et al., 2006; Tyler, 2006). However, the relationship between physical abuse and victimization is less clear, with some finding a relationship between physical abuse and street victimization (K. D. Ryan et al., 2000) and some finding no relationship (Thrane et al., 2006). Thus, we will further explore the separate influences of sexual abuse, physical abuse, and neglect on victimization as well as housing instability.

CURRENT STUDY

Overall, the early lives of homeless youths are often chaotic and involve several caretaker transitions and experiences of abuse (Robertson & Toro, 1999; Tyler, 2006). These early experiences of instability and childhood abuse can be destructive to youths throughout their life course (K. D. Ryan et al., 2000; Thrane et al., 2006; Tyler, 2006). This study examines how childhood experiences of instability, assessed as the number of caretaker transitions, and child maltreatment are related to later housing instability and victimization experiences among a sample of high-risk homeless youths. Based on previous literature, the three hypotheses are as follows: (1) A higher frequency of caretaker transitions will be associated with lower rates of housing stability and with higher rates of victimization and childhood abuse experiences. (2) A higher frequency of childhood abuse experiences will be directly associated with lower levels of housing stability and higher rates of victimization. The separate effects of each form of abuse (physical, sexual, and neglect) on housing instability and victimization will be explored. (3) Childhood abuse experiences will mediate the direct effect of multiple caretaker transitions on housing instability and victimization.

METHOD

Participants

We used baseline data from a clinical trial designed to evaluate the effectiveness of outreach services for an especially vulnerable group of homeless youths who reported recent substance use, three months of continuous homelessness, and service disconnection. Seventy-nine homeless youths (M age = 20.84 years, SD = 2.13) were recruited through outreach efforts in a large midwestern city. Youths were eligible to participate if they (a) were between the ages of 14 and 24 years; (b) reported no service use through drop-in centers, shelters, or substance or mental health treatment programs in the last three months; (c) reported using alcohol or drugs at least six times in the last 30 days; (d) met the criteria of homelessness, defined by McKinney–Vento Homeless Education Assistance Improvements Act of 2001 as lacking a fixed, regular, and adequate nighttime residence, or living in a welfare hotel, or place without regular sleeping accommodations, or living in a shared residence with other people due to the loss of one’s housing or economic hardship; and (e) had been homeless for the past three months. Table 1 presents a summary of the demographic characteristics of the current sample in regard to youths’ gender, race or ethnicity, abuse history, experiences with multiple caretakers, and homelessness.

Table 1:

Demographic Characteristics of the Total Sample (N = 79)

Variable	n (%)	M (SD)
Gender
Female	37 (46.8)
Male	42 (53.2)
Race/ethnicity
White, not of Hispanic origin	45 (57.0)
Other	34 (43.0)
Abuse history
Sexual abuse	33 (41.8)
Physical abuse	36 (45.6)
Neglect	34 (43.0)
Raised by different caretakers before 18 years old
Both birth parents	36 (45.6)
Birth mother only	51 (64.6)
Birth mother plus partner	22 (27.8)
Birth father only	15 (19.0)
Birth father plus partner	5 (6.3)
Other relatives (grandparents, aunt, uncle, and so on)	33 (41.8)
Adoptive parents	5 (6.3)
Foster parents	20 (25.3)
The number of nights participants stayed in different settings over the past 12 months
In stable housing (paying rent)		41.32 (92.09)
With family members in their home		89.10 (115.74)
With friends in their home		94.42 (107.59)
With romantic partner in his/her home		18.59 (51.23)
In a shelter or mission		6.97 (30.82)
In abandoned building		11.09 (46.74)
In jail		8.38 (22.77)
Someplace indoors (for example, a bus or a train station)		4.38 (26.57)
Someplace outdoors (for example, on the street or in a park)		71.94 (105.39)
In a residential treatment program		6.89 (28.51)
Anywhere else		7.91 (32.47)

Procedure

A research assistant engaged and screened youths during outreach at soup kitchens, parks, libraries, and other locations that homeless youths are known to frequent. Eligible and interested youths provided informed consent and assent; once the form was signed, the baseline assessment began. This assessment was conducted in the youth’s environment at any location acceptable and convenient to the youth, while also negotiating for privacy and confidentiality. Participants received a $40 gift card at completion of the assessment. All research procedures were approved by the Ohio State University’s institutional review board.

Measures

Multiple Caretakers.

The participants were asked how many years they were raised by different caretakers, including both birth parents, birth mother only, birth mother plus partner, birth father only, birth father plus partner, other relatives, adoptive parents, and foster parents before 18 years old. Answers were dichotomized into endorsement of number of years versus no endorsement. The dichotomous responses were summed, with a higher value representing a higher frequency of change in caretakers (range = 1–5).

Childhood Abuse History.

Following a method used by Bonomi and colleagues (Bonomi, Cannon, Anderson, Rivara, & Thompson, 2008), which successfully captures childhood abuse history, youths were asked about their experiences of abuse prior to age 18. The perpetrators of abuse could include family members, friends, strangers, or relatives. Sexual abuse was measured by two questions: (1) Have you ever been kissed in a sexual way or touched in a way that made you uncomfortable? (2) Has anyone had oral, anal, or vaginal intercourse with you or inserted a finger or object in your anus or vagina? Physical abuse was assessed by asking participants whether anyone had ever hit, punched, kicked, shaken, tortured, or otherwise physically hurt them. Neglect was assessed by asking whether participants had been left alone without sufficient parental care or lacked parental affection. If youths responded “yes” to any of the questions, they were asked how frequently the abuse happened. The frequency for each form of abuse was measured on a six-point Likert scale (0 = none, 1 = once, 2 = twice, 3 = three to five times, 4 = six to 10 times, 5 = 11 to 20 times, 6 = more than 20 times) and was used for the data analysis. A higher value represented higher rates of abuse.

Housing Instability.

Participants were asked how many nights they spent in their own room or apartment (stable housing, paying rent) over the last 12 months.

Victimization Experiences.

Participants were asked about their experiences of intimate partner violence (IPV) and street victimization in the last 12 months. IPV was assessed with five questions from the Behavioral Risk Factor Surveillance Survey (BRFSS). BRFSS has been widely used to measure the prevalence of IPV in the United States (Centers for Disease Control and Prevention, 1994; Saltzman, Johnson, Gilbert, & Goodwin, 2003). Sexual IPV was measured by the questions, “Has an intimate partner forced you to participate in a sex act (oral, vaginal, or anal penetration) against your will?” and “Has an intimate partner threatened, coerced, or physically forced you into any sexual contact that did not include penetration or intercourse?” Physical IPV was measured by the questions, “Has an intimate partner hit, slapped, shoved, choked, kicked, shaken, or otherwise physically hurt you?” and “Has an intimate partner frightened you for your safety or that of your family or friends because of anger or threats?” The question “Has an intimate partner put you down, or called you names repeatedly, or controlled your behavior?” was used to measure verbal abuse. The responses to sexual, physical, and verbal abuse were dichotomized (1 = yes, 0 = no). The measure for street victimization included five questions that asked participants to report their sexual and physical victimization and experiences of being robbed or burglarized. Responses were dichotomized (1 = yes, 0 = no). The responses for both IPV and street victimization were then summed, with higher values indicating greater prevalence of victimization experiences (range = 0–6).

Analytic Strategy

Multivariate path analysis was conducted using Mplus software (version 7.3) (Muthén & Muthén, 2014) to examine the relationship between multiple caretakers, childhood abuse experiences, housing instability, and victimization experiences (see Figure 1). Path analysis can provide insight into which variables are proximally related to other variables. Age and gender were added to the model as control variables. In addition, the literature has typically highlighted the negative effects of foster care experiences on homeless youth outcomes. To capture the unique impact of foster care, we examined the relationship between foster care experiences and housing instability and victimization experiences by adding foster care experience as a covariate to the model. Foster care experience was dichotomized to reflect whether participant had been raised by foster parents (1 = yes, 0 = no).

Figure 1:

Proposed Path Model of Multiple Caretakers, Childhood Abuse, Housing Instability, and Victimization Experiences

Notes: The values shown are standardized path coefficients. Gender, age, and foster care are not included in the model for visual simplicity. Significant paths are represented by solid lines. Nonsignificant paths are represented by dashed lines.

*p < .05. **p < .01. ***p < .001.

RESULTS

Means and standard deviations of the variables and bivariate correlations between the variables are presented in Table 2. Housing instability was positively skewed and was square-root transformed. After transformation, the skewness of all the variables fell between −1.96 and 1.96. Childhood abuse experiences (sexual, physical, and neglect) were significantly correlated with each other. Having multiple caretakers was significantly correlated with sexual abuse and neglect experiences but not physical abuse. Furthermore, sexual abuse was significantly correlated with lack of housing stability. The correlation between sexual abuse and victimization experiences approached significance (r = .21, p =.06).

Table 2:

Means, Standard Deviations, and Zero-Order Correlations among Research Variables (N = 79)

Variable	M	SD	1	2	3	4	5	6
1. Multiple caretakers	2.37a	0.98	—
2. Sexual abuse	1.21b	1.89	.25*	—
3. Physical abuse	2.33b	2.80	.18	.43**	—
4. Neglect	2.35b	2.87	.24*	.38**	.47**	—
5. Victimization experiences	1.14c	1.40	.02	.21	.18	.15	—
6. Number of nights stably housed	41.32	92.09	−.04	−.23*	−.18	−.17	−.05	—

^aThe average number of caretaker transitions youths experienced before age 18.

^bThe average level of frequency for each form of abuse (sexual, physical, and neglect) youths experienced before age 18.

^cThe average level of prevalence for victimization experiences reported by youths in the past 12 months.

^*p < .05.

^**p < .01.

Results of the multivariate path analysis showed a good fit of the model to the data: χ²(3) = 3.28, p = .35; comparative fit index = .99; Tucker–Lewis Index = .95; root mean square error of approximation = .03; standardized root mean square residual = .04. Results of path analyses (see Figure 1) suggested that having multiple caretakers was associated with child abuse experiences, which, in turn, was related to housing instability and victimization experiences. Physical abuse was not included in the final model given that it did not relate to either change in caretakers, housing instability, or victimization experiences.

Our first hypothesis was partially supported. As expected, the higher frequency of change in caretakers was associated with higher frequency of sexual abuse (β = .28, p < .01) and neglect experiences (β = .23, p < .05). However, we did not find a direct linkage between multiple caretakers, housing instability, and victimization experiences. Moreover, the higher frequency of sexual abuse experiences predicted a greater prevalence of victimization experiences (β = .23, p < .05) and reduced housing stability (β = −.32, p < .01), supporting our second hypothesis. Compared with sexual abuse experiences, the frequency of neglect was not related to either housing instability or victimization experiences.

The nonsignificant direct effect of multiple caretaker transitions on housing instability and victimization may be attributed to childhood abuse experiences being a more proximal factor influencing housing instability and victimization experiences. A bootstrap procedure was carried out to determine whether childhood abuse experiences mediated the effect of multiple caretakers on housing instability and victimization experiences. One thousand bootstrap samples were used. The bias-corrected interval method was applied to estimate the indirect effect. Results showed a 95 percent confidence interval (CI) of the indirect effect of multiple caretakers on housing instability excluded zero. Therefore, sexual abuse significantly mediated the link between multiple caretakers and housing instability (β = .09, 95% CI [−0.169, −0.003]). Thus, our third hypothesis was supported. The bootstrap results also showed that sexual abuse did not mediate the association between multiple caretakers and victimization. However, findings showed that having multiple caretakers was associated with sexual abuse, and sexual abuse, in turn, was associated with victimization. According to Frazier, Tix, and Barron (2004), given possible measurement error of sexual abuse and the fact that the path connecting multiple caretakers and sexual abuse (β = .28) was larger than the path between sexual abuse and victimization (β = .23), the mediating effect of sexual abuse on the association between multiple caretakers and victimization could be underestimated.

Age was significantly associated with housing stability, with older adolescents exhibiting higher levels of housing stability (β = .32, p < .01). Gender was not significantly associated with either housing instability or victimization experiences. Foster care did not predict either housing instability or victimization.

DISCUSSION

To our knowledge this is the first study to examine the role of multiple caretaker transitions and childhood abuse on housing instability and victimization among a high-risk sample of street-living and substance-using homeless youths. Most studies do not examine childhood abuse experiences separately, which can mask the differential impact of physical and sexual abuse and neglect experiences. This study therefore assessed the separate impact of these three types of childhood abuse on outcomes. Furthermore, little is known about the impact of changes in caretakers among homeless youths and how changes in caretakers influence homeless youths’ future housing instability and street victimization experiences. The goal of our study was to increase understanding of the relationship between early caretaking experiences and housing instability and victimization experiences among a high-risk sample of homeless youths. Of interest was the mediating role of childhood physical and sexual abuse and neglect.

The first study hypothesis, positing that multiple caretaker transitions would be associated with childhood abuse, housing instability, and victimization, was partially supported. Multiple caretaker transitions were associated with sexual abuse and neglect. This study cannot determine at what point in the caretaking history that sexual abuse and neglect occurred. Possibly, sexual abuse and neglect resulted in multiple changes in caretakers, or multiple changes in caretakers increased the risk for sexual abuse and neglect to occur. No relationship was found between multiple caretaker transitions and physical abuse. Possibly, sexual abuse increases with multiple caretakers because vulnerable single mothers are pursued by predators, or have more contact with strangers in general who prey on their children, whereas childhood physical abuse is less linked to such predators. Future research identifying the individuals who perpetrated abuse on the youths could shed light on this finding.

In addition, having multiple caretakers was not directly linked to street victimization or housing instability, suggesting that multiple changes in caretakers does not directly affect these later outcomes. However, in support of the second hypothesis, childhood sexual abuse was associated with later housing instability and victimization. Furthermore, childhood sexual abuse mediated the relationship between multiple caretakers and housing instability, supporting our third hypothesis. In other words, change in caretakers appears to exert its influence on future experiences among homeless youths through the experience of sexual abuse. These findings suggest that sexual abuse has significantly more deleterious effects on homeless youths’ homeless experiences than physical abuse or neglect. This adds to the mix of findings, with some indicating that childhood sexual abuse is associated with more negative homeless experiences than physical abuse (K. D. Ryan et al., 2000; Tyler et al., 2000) and others showing no differences (Thrane et al., 2006). Identifying between-group differences associated with different childhood abuse experiences, rather than comparing those with a history of abuse to those without, provides more detailed information on the effects of childhood victimization (Edwards, Holden, Felitti, & Anda, 2003) and offers better clinical implications for intervention development. In sum, multiple caretaking experiences are not in and of themselves predictive of future problems for homeless youths, but if associated with sexual abuse, the likelihood of future problems associated with homelessness increases.

Neither a history of foster care involvement nor youth’s gender predicted housing instability or victimization. A significant number of studies report that foster care is associated with high rates of homelessness (HHS, 2007). Future research may confirm that among already homeless youths, history of foster care involvement may be less salient for understanding housing instability than the factors that led to youths being placed in foster care (for example, sexual abuse). Finally, older youths reported higher levels of housing stability than younger youths, suggesting that younger homeless youths are more at risk for housing instability. Older youths may have developed more social capital and resources over time and are perhaps more motivated to exit street life than younger youths.

Limitations

Several limitations should be noted. This study relied on participants’ retrospective reports of their different caretaking situations and childhood abuse experiences, increasing potential recall bias. The study used a cross-sectional design, which does not allow conclusions regarding the temporal ordering of events, such as between change in caretakers and childhood abuse experiences. The sample size was small, and a larger sample would be able to offer a more well-specified model. Despite these limitations, this mediating model with a small sample suggests robust findings.

Conclusion and Future Directions

This study offers several important implications. Having multiple caretakers as a child exerts a negative influence on housing stability through the experience of childhood sexual abuse. Having had multiple caretakers itself was not associated with the more distal outcomes of street victimization and social stability, though it was associated with childhood neglect and sexual abuse. Future studies will need to examine whether the increased risk associated with a history of sexual abuse requires interventions that intervene differently with homeless youths reporting sexual abuse experiences. For example, although all homeless youths need assistance in exiting street life, this study suggests that sexually abused street youths may have more barriers associated with maintaining stable housing compared with those who experienced physical abuse or neglect. This study did not identify those barriers, but they could be powerful targets for interventions, and additional study is necessary. For example, it could be especially important for sexually abused youths to receive an intervention component that helps them evaluate risk situations while living on the streets and to develop safety plans. In summary, among youths in the midst of their homeless experiences, this study showed that youths present with diverse histories of childhood abuse, which affect later homelessness and victimization differently. Evaluation of these histories, combined with targeted intervention approaches, may be key for improving housing stability and keeping youths safe from future victimization.