Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Feb 1.
Published in final edited form as: Child Abuse Negl. 2021 Dec 17;124:105439. doi: 10.1016/j.chiabu.2021.105439

Modeling changes in adolescent health risk behaviors approaching and just after the time of emancipation from foster care

Keith J Martin a,b, Katie Nause b, Mary V Greiner b,c, Sarah J Beal b,c
PMCID: PMC8799516  NIHMSID: NIHMS1765716  PMID: 34923298

Abstract

Background

Emancipated foster youth frequently engage in behaviors that contribute to poor health. Whether health risk behaviors increase following emancipation or are established while in foster care remains unclear.

Objective

This secondary data analysis examined substance use and attitudes toward sexual risk behaviors to understand continuity in risk behaviors among foster youth before emancipation and following emancipation.

Participants and Setting

Youth ages 16–20 (N = 151) who had been in foster care for at least 12 months and were expected to emancipate were recruited. The urban county where the study was conducted allowed youth to remain in foster care until 21 years of age.

Methods

Participants completed surveys assessing substance use and attitudes toward sexual risk behaviors at baseline, 6- and 12-months. Multilevel models estimated trajectories of health behaviors and attitudes, with emancipation timing as the primary predictor. Individual and child welfare characteristics were included as covariates.

Results

Substance use did not change with emancipation (Bs = 0.01, p = 0.81) and positive attitudes about risky sexual behavior significantly decreased as youth approached emancipation (Bs = 1.67, p < .01). Placement instability and adversity were not associated with either outcome (ps >.08). Females reported more positive attitudes about higher-risk sexual behavior than males (B = 3.09, p < 0.01) and less substance use (B = −1.15, p = 0.03).

Conclusions

Substance use and attitudes about sexual risk behaviors are established before emancipation; interventions prior to emancipation are necessary to improve health outcomes.

Keywords: adolescent, foster care, emancipation, health risk behavior, sexual risk behavior, substance use

Every year, more than 20,000 youth emancipate from or “age out” of foster care (i.e., in the custody of a county or state child protection system and placed in out-of-home care (United States Department of Health and Human Services Administration for Children and Families, 2020)). Emancipation occurs because of a combination of age, case plan goal, and/or legal status (Beal et al., 2020). This exit from foster care can occur at any point between a young person’s 18th and 21st birthdays.

Previous studies have demonstrated that, compared to a nationally representative sample of all young adults, young adults who have emancipated from foster care report more psychosocial challenges than their peers, including higher rates of mental health conditions, incarceration, and homelessness (Courtney & Dworsky, 2006). These poor health outcomes may stem from health risk behaviors such as substance use and sexual behaviors, which lead to persistent health problems, e.g., human immunodeficiency virus (HIV) infection, substance use disorder (SUD). Youth who are approaching or have recently emancipated from foster care may represent a unique subset of youth at risk for more frequent engagement in health-compromising behaviors (Courtney et al., 2010; Courtney et al., 2007). Importantly, much of what we know about the health of young people with child welfare involvement comes from studies of youth at highest risk (e.g., those in treatment programs, therapeutic foster care (Vaughn et al., 2007)) or retrospective reporting after youth leave foster care (Courtney & Dworsky, 2006; Courtney et al., 2007); as a result, the timing, onset, and evolution of health risk behaviors while youth are in foster care and how that changes as they leave foster care is unclear. In other words, it is unclear whether emancipation itself, which is often associated with a reduction in access to services and less adult oversight (Courtney et al., 2007), is an important contributor to the onset of health risk behaviors.

This study focuses on two specific health behaviors: sexual risk and substance use. Sexual risk behaviors are commonly reported by emancipated young adults, with higher rates of unintended pregnancy and diagnosis of HIV than the general population (Courtney et al., 2007). Furthermore, almost half of emancipated young adults have used illicit substances in their lifetimes, and more than a third meet criteria for SUD (Vaughn et al., 2007). Understanding whether emancipation itself alters trajectories of risk is important because such knowledge could illuminate opportunities for interventions to improve health outcomes. If emancipation increases risk, delaying emancipation with extended foster care may be a helpful intervention. If risk occurs prior to emancipation, other supports such as targeted substance use and sexual health screening and interventions may also be necessary.

Adverse experiences (e.g., (Garrido et al., 2018)), placement in non-familial settings prior to emancipation (e.g., independent living, congregate care; e.g., (Freundlich et al., 2007; Palmer et al., 2020)), and earlier timing of emancipation (i.e., at age 18 vs. 21; e.g., (Miller et al., 2017)) are associated with more health-compromising behaviors. The original adverse childhood experience (ACE) study demonstrated that abuse and neglect during childhood increases a wide range of adult health risk behaviors, including sexual risk behaviors and substance use (Felitti et al., 1998). Recent study of ACEs has extended past general population samples of adults to examine unique stressors of diverse populations of children, including youth with juvenile delinquency and court-involvement (Logan-Greene et al., 2016; Wolff et al., 2018). National data indicates that exposure to ACEs is higher among youth in foster care compared to the general population, those living in poverty, and in non-traditional families (e.g., single-parent, grandparent, non-relatives; (Turney & Wildeman, 2017)). It would therefore be expected that more exposure to adversity would increase health risk behaviors among youth in or emancipating from foster care. In addition to maltreatment and adversity, many youth in foster care experience placement instability, family dysfunction, and educational deficits while in custody that may increase vulnerability to health risk behaviors into adulthood (Courtney et al., 2011). Emancipated youth, who make up 11% of the population of children exiting foster care annually, are of particular concern because they are often left to navigate adulthood alone, lacking family or social support to mitigate health compromising behaviors (University of Hawai’i Center on the Family, 2012). While health risk behaviors are higher for adolescents in or emancipated from foster care compared to the general population (Courtney et al., 2010; James et al., 2009), it is unclear whether these behaviors manifest before or after emancipation. Therefore, there is a need for more research focused on the emergence of health risk behaviors prior to age 18, given that the persistence of health risk behaviors into adulthood is greatest for youth who begin engaging in these behaviors before age 12 (Stevens et al., 2011). Research and legislation extending foster care to age 21 suggests that youth need more services and supports from child welfare in order to, among other things, avoid negative health outcomes, including substance use and sexual risk behavior. However, if those behaviors are actually initiated prior to emancipation, then extending foster care to age 21 or later may not, on its own, result in reduced risk for poor outcomes, because risk behaviors are already occurring before emancipation.

Therefore, this study sought to examine two indicators of health risk behaviors (substance use and attitudes toward sexual risk behaviors) of youth in foster care as they approached and experienced emancipation, using time to emancipation (number of days between study visits and date of emancipation) as a predictor and health risk behaviors (past 30-day substance use and attitudes about sexual risk behavior) as outcomes. Here, we employed methods commonly used in cohort sequential designs to align time with the metric of change relevant to our hypotheses, time to emancipation, rather than time in study (Ram & Grimm, 2007). Covariates were chosen given known or hypothesized relationships with increased substance use and more positive attitudes about sexual risk behavior (e.g., longer lengths of stay in foster care, placement instability, maltreatment, and adversity; (Landsverk et al., 2002; Resnick et al., 1997; Turney & Wildeman, 2017)). Age in years was included as a time-varying (level 1) predictor in our multilevel model (MLM) in order to examine time to emancipation separate from youth aging. Two exploratory, hypothesis-generating possibilities were tested: (1) health risk behaviors among youth in foster care exhibit high prevalence with no change over time to emancipation (before vs. after) – suggesting that interventions before emancipation are necessary - or (2) health risk behaviors among youth in foster care increase over time to emancipation (after vs. before), indicating that delaying emancipation could lead to health benefits.

Methods

Participants, Procedures, Setting

This is a secondary analysis of data from the ICare2CHECK study, which investigated whether healthcare education materials promoted engagement with healthcare services in preparation for emancipation (Beal et al., 2020). ICare2CHECK recruited adolescents who were in foster care for at least 12 months and were expected to emancipate due to a combination of age, case plan goal, and/or legal status. All adolescents were in the custody of a single county child welfare agency that had extended foster care to age 21 years. Youth in this county emancipate from foster care at 18 (67%), 19 (20%), 20 (10%), or on their 21st birthday (3%) (Economics Center on behalf of Hamilton County Job and Family Services and the Higher Education Mentoring Initiative, 2017). Adolescents were eligible for the parent study if they were aged 16–20 years, were without a diagnosis of cognitive disability, were English speaking, and were residing within a 1-hour driving distance of the academic medical center where the study was based. Adolescents aged <16 years, those with a cognitive disability, youth who had been in foster care for fewer than 12 months, and youth placed in inpatient treatment settings, nursing homes, juvenile detention, or who were reunited with family or adopted were considered ineligible to participate. All youth who participated in ICare2CHECK provided assent (if aged <18 years) or consent (if aged ≥18 years) to participate in the study, which was approved by institutional review boards at the academic medical center and the child welfare agency. Of the 365 youth screened for eligibility to participate in ICare2CHECK, 204 were deemed eligible, 154 provided initial verbal consent (if aged ≥18 years) or assent (if aged 16 or 17 years) to participate, and 151 were enrolled. Three of the 154 verbally consented participants were lost to enrollment because of changes in placement (n = 2) or incarceration (n = 1). Study staff scheduled in-person visits with youth in their homes or public locations (e.g., community centers, libraries, and restaurants). During visits, participants provided written informed assent or consent and completed surveys. For youth under the age of 18, the county child welfare agency provided consent, and caseworkers, guardians ad litem, court appointed special advocates, and caregivers were notified of the study and provided the opportunity to opt the adolescent out of the study if there were concerns about youth participating. Enrolled participants completed in-person survey measures assessing health risk behaviors at baseline, 6 months after baseline, and 12 months after baseline. Surveys were administered on computers with headphones provided by study staff, with audio assistance for survey questions to ensure literacy was not a barrier to participation while also supporting confidentiality. Of those enrolled, 61% completed all study visits, 21% completed two study visits, and 17% completed only the baseline survey.

Measures

Attitudes toward sexual risk behavior were measured using 12 items adapted from the sexual attitudes and activities questionnaire (SAAQ; Noll et al., 2003). The SAAQ is a self-reported measure of attitudes of voluntary sexual and sexual risk behaviors. Responses are on a 5-point Likert scale from “strongly agree” to “strongly disagree.” Items were recoded and averaged to a single score where higher scores corresponded to attitudes that increase sexual risk behavior. The SAAQ was developed for and initially validated with maltreated adolescent females (Noll et al., 2003) where test-retest reliability and validity of attitudes associated with behavior was established. Findings were replicated in separate studies and samples (Noll & Shenk, 2013) and with samples of women who had not experienced maltreatment (Bouchard et al., 2009). As with Bouchard et al., this study used the short form of the SAAQ, and focused specifically on attitudes about sexual risk behaviors, which are established predictors of behavior. Across samples, α ranged between .79 and .90. In the current study, α = .73. Of the 12 items adapted from the SAAQ, 4 items reflected attitudes toward pregnancy (e.g., “It wouldn’t be all that bad if you got pregnant at this time in your life,” 1 item reflected attitude toward HIV/AIDS (e.g., “If you got the AIDS virus, you would suffer a great deal,” 1 item reflected attitude toward contraception (e.g., “It would be a big hassle to do the thing necessary to completely protect yourself from getting a sexually transmitted disease,” and 6 items reflected attitudes towards sexual intercourse (e.g., “If you had sexual intercourse, your friends would respect you more.”) Parallel items were available for males and females to prevent the number of items responded to from impacting the SAAQ score, e.g. for males “Getting someone pregnant at this time in your life is one of the worst things that could happen to you.” while for females “Getting pregnant at this time in your life is one of the worst things that could happen to you.”

Substance use questions were adapted from the Youth Risk Behavior Surveillance Survey (YRBSS) (Centers for Disease Control and Prevention, 2021). Participants reported the number of days over the past 30-days that they engaged in the following: smoking cigarettes, using chewing tobacco, drinking alcohol, binge drinking, getting drunk, using marijuana, using illegal drugs, and using illegal drugs with a needle. Participant responses were coded to create a dichotomous variable indicating ≥1 day of any substance use during the past 30-days (1) or 0 days of substance use during the past 30-days (0). YRBSS substance use items are reliable (kappas ≥ 41–61%; (Brener et al., 2002)). In this study, among adolescents who participated in all three timepoints, 38% reported consistent patterns of substance use/non-use (i.e., either always yes or always no) across all three timepoints, and 62% reported inconsistently (i.e., sometimes yes and sometimes no) across all three timepoints.

The Childhood Trust Events Survey (CTES; (Boat et al., 1996)) screened for exposure to childhood adversity (Hamel, 2016). Respondents indicate whether (1) or not (0) each of 26 specific events had occurred. Three subscales have been validated with a Rasch measurement model demonstrating that all factor loadings were constrained to be equivalent within each of 3 subscales (RMSEA = 0.05, Tucker-Lewis Index = 0.90, comparative fit index = 0.90; (Beal et al., 2019)). The current study used subscales for unexpected tragedy (9 items; e.g., natural disaster, parental divorce) and family instability (6 items; e.g., parental substance abuse or mental health concerns; (Beal et al., 2019)).

Time to emancipation was a count of the number of days between study visits and emancipation date extracted from child welfare administrative data. Time was set as positive for observation periods before emancipation and negative after emancipation. There were three assessment timepoints (baseline, 6 months, and 12 months) in the study, and the number of days between each assessment timepoint and a youths’ date of emancipation varied across individuals. The study design/personnel had no control or influence as to when emancipation would occur. The date of emancipation was reported to study personnel by children’s services. The point at which emancipation occurred is coded as “0” in the continuous variable “time to emancipation.”

Time in custody was calculated as the number of months between the date of entry into foster care (from child welfare administrative data) and the date of the baseline study visit. Total number of placements for each participant at the time of enrollment was reported in child welfare administrative data, as was primary reason for removal with categorical responses (yes/no) for neglect, abuse (physical, sexual, or emotional), and child behavior problems.

Demographic characteristics included age (in years), race/ethnicity (non-Hispanic White = 0, Black and Multiracial =1) and gender (Male=0, Female=1). There were no Hispanic/Latino, Native American, or Asian participants in the dataset.

Statistical Analyses

Demographic and bivariate descriptive statistics were estimated to understand the data and inform model development. To predict substance use and attitudes about sexual risk behaviors longitudinally, multilevel modeling (MLM; (Raudenbush & Bryk, 2002)) was used, with linear models estimated for attitudes about sexual risk behaviors and logistic models estimated for substance use. Baseline models that included only time to emancipation centered at 0 days (representing emancipation date with post-emancipation values as negative) were estimated to describe change in substance use and attitudes about sexual risk behaviors over time. Additional models included predictors that were time-varying (level 1) and time invariant (level 2). Level 1 predictors included age in years and grand-mean centered at 18, and time in county custody (grand-mean centered at 46 months). Level 2 predictors, which did not vary across observation periods, included minority status, gender, lifetime number of placements (grand-mean centered at 6 placements), and self-reported exposures to unexpected tragedy (grand-mean centered at 3 exposures) and family instability (grand-mean centered at 3 exposures). Analyses were conducted using restricted maximum likelihood estimation with an estimated degrees of freedom procedure (Kenward & Roger, 1997) to arrive at valid parameter estimates under the assumption of ignorable missing data (Hofer et al., 2007) in SAS, version 9.4 (SAS Institute, Cary, NC). To account for the non-normative distribution of substance use over time, a Gaussian response distribution was specified. Two-tailed tests of significance were used (p < .05).

Results

Descriptive statistics are provided in Table 1. Approximately half of youth were female, and two-thirds were minority race (Black and Multi-racial). Participants had an average of 6 placements, 3 potentially traumatic experiences, and 3 indicators of family instability. Most (78%) participants emancipated during the study. Across time in study, substance use did not significantly change (χ2 [5] = 0.46, p = .80), but attitudes toward sexual risk behaviors did significantly decrease (F = 8.88, p < .001). Substance use was not related to the number of visits completed, χ2 (5) = 1.96, p = 0.37. Those with less positive attitudes about sexual risk behaviors completed more study visits (F = 6.93, p < .01).

Table 1.

Demographic characteristics of study participants at three observation windows

Mean
(SD)/%		 Baseline
N = 151		 6-Month Follow-up
N = 109		 12-Month Follow-up
N = 107
Variable Mean
(SD)/%		 Mean
(SD)/%		 Mean
(SD)/%
Female gender 53.6 -- -- --
Non-White race# 65.6 -- -- --
Lifetime number of placements 6.11 (4.61) -- -- --
Potentially traumatic experiences 3.19 (1.92) -- -- --
Family instability 3.08 (1.67) -- -- --
Age (years) -- 18.1 (1.36) 18.7 (1.35) 19.1 (1.35)
Average time to emancipation* (months) -- 14.3 (10.3) 9.34 (10.1) 3.62 (9.47)
Percent emancipated -- 3.3 12.8 29.0
Substance use -- 46.4 42.2 43.9
Sexual risk behaviors -- 2.60 (.64) 2.34 (.63) 2.29 (.66)
Time in custody (months) -- 45.6 (38.3) 51.1 (38.0) 56.4 (37.2)
Open in a new tab
*

Dates are positive for study date before emancipation date, and negative for study date after emancipation date.

#

Black and Multi-racial

Descriptive Trajectories

Baseline models for substance use and attitudes about sexual risk behaviors by time to emancipation are provided in Table 2. The baseline model for substance use indicated no significant change (slope) over time to emancipation, while the slope for attitudes toward sexual risk behavior suggested a significant decline as youth approached and experienced emancipation.

Table 2.

Baseline models for substance use and attitudes toward sexual risk behaviors*

Substance Use Sexual Risk
Variable B SE P B SE P
Intercept 0.50 0.04 <.0001 32.37 0.58 <.0001
Fixed Slope 0.01 0.04 0.81 1.67 0.62 <.01
Open in a new tab
*

Dates are positive for study date before emancipation date, and negative for study date after emancipation date.

Multivariable Models

Multivariable models predicting attitudes about sexual risk behavior over time to emancipation from foster care are provided in Table 3. Of note, once other variables were included in the model, the effect of time to emancipation (slope) was no longer significant. There were significant effects of age and history of or exposure to family instability, such that younger participants and those with less family instability had more positive attitudes about sexual risk behaviors. Female participants endorsed more positive attitudes about sexual risk behaviors than male participants. As participants spent more time in custody, they exhibited more positive attitudes about sexual risk behaviors. Time to emancipation, potentially traumatic experiences, and race were not significantly associated with attitudes about sexual risk behavior.

Table 3.

Linear regression model estimating attitudes toward sexual risk behaviors by time to emancipation

Variable B SE P rp$
Intercept 32.40 1.22 <0.001 --
Time to emancipation (slope) 0.78 0.57 0.17 0.09
Lifetime number of placements −0.27 0.15 0.08 −0.14
Potential traumatic experiences 0.35 0.32 0.27 −0.05
Family instability −0.89 0.38 0.02 −0.15
Substance use −0.66 1.00 0.51 0.001
Age −1.23 0.43 0.005 --
Female gender 3.09 1.07 0.004 --
White race vs non-White# −0.86 1.70 0.61 --
Time in custody (days) 0.04 0.02 0.02 --
Model fit
Residual covariance 60.1 5.27 <0.0001 --
AIC 1871.1 -- -- --
BIC 1874.6 -- -- --
Open in a new tab
#

Black and Multi-racial

$

rp indicates the partial correlation of each variable with sexual risk behaviors, with covariates accounted for.

Models predicting past 30-day substance use over time to emancipation from foster care are represented in Table 4 and indicated that past-month substance use did not change over time to emancipation (slope). Results indicated no variation in likelihood of substance use with age. A significant gender effect was detected, where young men had an increased risk of past 30-day substance use compared to young women. Other demographic characteristics were nonsignificant. Higher exposure to family instability decreased odds of substance use; potentially traumatic experiences were not significantly associated with substance use. A post-hoc analysis of the effect of family instability on substance use revealed a nonsignificant interaction with age (B = −0.19, SE = 0.10, p = 0.07, OR = 0.83, 95% confidence interval [CI] = 0.67 – 1.01), such that high family instability was associated with high risk for substance use at age 16 that declined as youth got older, while low family instability was associated with low risk for substance use at age 16, which increased as youth got older (Supplemental Figure 1).

Table 4.

Logistic regression model estimating substance use by time to emancipation

Variable B SE P OR +/− 95% CI
Intercept 0.78 0.96 0.42 -- --
Time to emancipation (slope) 0.05 0.26 0.85 1.05 0.62–1.77
Lifetime number of placements 0.02 0.07 0.81 1.02 0.88–1.17
Potentially traumatic experiences −0.02 0.15 0.91 0.98 0.73–1.33
Family instability −0.59 0.19 0.003 0.55 0.38–0.81
Sexual risk behaviors −0.02 0.02 0.44 0.98 0.93–1.03
Age 0.09 0.21 0.68 1.09 0.72–1.64
Female gender −1.15 0.53 0.03 3.16 1.11–8.99
White race vs non-White# 0.10 0.81 0.91 0.91 0.19–4.46
Time in custody (days) −0.0006 .009 0.95 1.00 0.98–1.02
Model fit
Residual covariance 2.59 1.18 0.01 -- --
AIC 347.41 -- -- -- --
BIC 380.24 -- -- -- --
Open in a new tab
#

Black and Multi-racial

Discussion

The purpose of this secondary study was to examine the unique impact of emancipation from foster care on health risk behaviors in young people ages 16–21 in long-term community-based foster care. This study found that during any 6-month period, half of young people in foster care self-reported attitudes and/or behaviors that would suggest increased health risks (i.e., more past-month substance use, more positive attitudes about sexual risk behaviors) compared to the general population of US youth, consistent with previous work (Courtney et al., 2010; James et al., 2009). In other words, health risks of youth in foster care were established before emancipation and emancipation did not change health risk trajectories of youth in foster care. To our knowledge, this study is the first to compare indicators of health risk in youth in foster care beginning at age 16 and across time through emancipation. Results indicate that targeted substance use and sexual health screening and intervention before youth emancipate may be more beneficial than offering services targeted at emancipated youth, and extending foster care alone may not reduce health risk. The mechanisms typically associated with increased substance use and more positive attitudes about sexual risk behavior (e.g., longer lengths of stay in foster care, placement instability, maltreatment, and adversity; (Landsverk et al., 2002; Resnick et al., 1997; Turney & Wildeman, 2017)), were generally not associated with past-month substance use or attitudes about sexual risk behavior. Gender differences were observed, with female participants reporting more positive attitudes about sexual risk behaviors than males, and male participants reporting more past-month substance use than females.

Some states have extended foster care to age 21 years to ease the challenges associated with emancipation (University of Hawai’i Center on the Family, 2012). The effects of delayed emancipation in reducing health risk behaviors among youth are still being evaluated (Smith et al., 2015), with preliminary evidence indicating benefits; a prospective study found that 19 year-old women who stay in foster care past age 18 years report fewer pregnancies than those who age out at 18 years (Courtney & Dworsky, 2006). It is worth noting that extending foster care to age 21 can only be beneficial for reducing health risk behaviors if the services provided by extended foster care are associated with decreased health risk behaviors. Although this research question was not tested directly, this study’s findings provide some initial indication that extended foster care will not, by itself, decrease health risk behaviors. The initiation of health risk behaviors prior to emancipation, as seen in our study, suggests that supports such as targeted substance use and sexual health screening and interventions may also be necessary.

Our findings paint a complex picture of age, time in custody, and time to emancipation. While younger participants and those who spent more time in custody exhibited more positive attitudes about sexual risk behavior, there was no significant association between time to emancipation and sexual risk behavior. There were also no significant associations between the likelihood of substance use and the individual covariates of age, time in custody, and time to emancipation. Largely, our findings suggest that health risk behaviors begin at younger ages, with results of multivariate models implying that youth in foster care who engage in high risk behaviors do so at similar (substance use) or slightly higher (attitudes toward sexual risk behavior) frequencies at younger and older ages. Further, these results suggest already heightened risk by age 16. Mechanisms not evaluated in this study, including beliefs about harm in using substances, perceived ease of obtaining substances while still attending school, and attitudes about peer substance use could explain the increase in substance use among youth in foster care before emancipation which persists following emancipation (Siegel et al., 2016). Preventive services at younger ages and while youth are in foster care are likely necessary to reduce substance use and sexual risks following emancipation. Interventions to reduce substance use and sexual risk behavior among youth in foster care have been developed (Trejos-Castillo & Trevino-Schafer, 2018); however, levels of efficacy vary, and programs are not widely implemented. Studies evaluating STI prevention among youth in foster care have been uncontrolled (Becker & Barth, 2000) or have failed to produce long-term reduction in high-risk behaviors for HIV infection (Slonim-Nevo et al., 1996), while substance use intervention studies have been limited by small sample sizes and low participation rates (Haggerty et al., 2016; Kim et al., 2017). Delivery of targeted substance use and sexual health screening interventions through evidence-based screening, brief intervention, and referral to treatment (SBIRT) at mandated clinic visits for foster youth (Greiner & Beal, 2018), primary care visits, and in other settings may also be effective and offers an opportunity to leverage motivational interviewing to support healthy choices (Mitchell et al., 2013), including for those who are not endorsing health risks. Given that half of youth did not report indicators of health risk behaviors, SBIRT may offer an opportunity to reinforce positive choices about health behaviors.

Our study found that the mechanisms typically associated with increased substance use and more positive attitudes about sexual risk behavior (e.g., longer lengths of stay in foster care, placement instability, maltreatment, and adversity; (Landsverk et al., 2002; Resnick et al., 1997; Turney & Wildeman, 2017)), were generally not associated with past-month substance use or attitudes about sexual risk behavior. Differences in these findings may reflect variation in our study approach and sample. Specifically, we assessed more recent substance use and attitudes about sexual risk behavior, which vary over time more than commonly used measures of lifetime substance use or sexual risk behaviors, and this was a community sample of youth rather than youth recruited from one clinic or program. Local policies regarding lengths of stay, placement, and extended care may also explain differences from other studies, which have been cross-sectional and with nationally representative samples of youth (Resnick et al., 1997; Turney & Wildeman, 2017).

Findings indicated that while risk did not vary by time before vs. after emancipation, types of adverse experiences and age may be important. Potentially traumatic experiences (e.g. natural disasters, parental divorce) were not significantly associated with attitudes toward sexual risk behaviors or substance use, reflecting that events that could happen to anyone may not have a large impact on youth in foster care compared to a general community sample of youth (Beal et al., 2019). In contrast, history of high family instability was associated with substance use and attitudes about sexual risk behavior, but not in the direction expected. Nonsignificant interactions with age indicate the potential that family instability impacts youth differently at differing ages. These findings may help to understand previous findings that increased family instability predicts increased cigarette and marijuana use among adolescents in foster care (Beal et al., 2019). Additionally, our results emphasize the importance of not treating different types of adversity as equivalent – doing so may contribute to investigators missing the nuance of differing types of adversity and their effects on various sub-populations of youth. Other studies, for example, found that increased severity of sexual abuse is associated with increases in self-reported sexual risk behaviors among youth in foster care (Elze et al., 2001). Interactions between different types of adversity may also be an important consideration (Afifi et al., 2020). Specific types of trauma (e.g. separation from a caregiver) also differ in context for youth in foster care compared to other populations, and may therefore have different implications for exposure to trauma. Additionally, our nonsignificant findings of family instability decreasing odds of substance use and its association with risk for substance use at age 16 that declined with age may benefit from replication of our analysis in a larger, more heterogeneous sample of youth.

A prior study with this same cohort found that health risk behaviors were more likely to be identified in the electronic health records (EHRs) of young women compared to young men, with sexual risk behaviors particularly higher for women (Beal et al., 2018). In the current study, young women self-reported more positive attitudes toward sexual risk behaviors and less substance use than young men. These results may reflect gender differences in how patients access healthcare and how practices screen and document in the EHR. Further, our results indicate more positive attitudes toward sexual risk behavior in young women around the time of emancipation, consistent with other studies that have demonstrated different motivations for engaging in sexual risk behavior (Cooper et al., 1998; Cuffee et al., 2007; Hendrick et al., 1985; Ramiro-Sánchez et al., 2018; Werner-Wilson, 1998), likely related to socialization (Bumpus et al., 2001; Henry et al., 2007), for males and females. Importantly, gender differences in sexual risk behaviors (Courtney et al., 2011; Murphy et al., 1998; Twenge et al., 2015) and substance use (Kuhn, 2015) have been identified as well. Our findings lend corroboration to others (e.g., (Murphy et al., 1998)), which suggests that gender-specific interventions targeting health behaviors for youth preparing to emancipate from foster care may be beneficial.

These findings must be considered in the context of several limitations. First, adolescent self-reported attitudes about sexual risk behaviors were used, and adolescents self-reported their substance use in the past 30 days. While youth were assured that data would be kept confidential and were provided with a laptop and private space to complete surveys, it is possible that some reports were inaccurate. Second, no other data sources were used, e.g. from the EHR, to verify the data used in this study. Third, this study did not examine mechanisms known to be protective of health risks (e.g., caregiver relationships, academic engagement, future planning) which would explain the absence of health risks in 50% of the sample. Fourth, the timing of study enrollment varied from time to emancipation, which does not occur strictly on a child’s 18th birthday in the community where data were collected. This provided variability in the timing of emancipation as compared to age, as well as in how long data were collected before and following emancipation. While this was statistically managed, a study designed to align study visits more closely to emancipation would be ideal. Fifth, it would have been ideal to examine trajectories of health risk behaviors with more frequent observations within young people before and after emancipation, over longer periods of time, and with a large enough sample to explore differences in trajectories for youth who emancipated closer to age 18 compared to those who emancipated closer to age 21. Our analysis was limited to the available data that was collected at the three study visits of baseline, 6-months, and 12-months, with individual variation in the temporal proximity of those observations to an individual’s date of emancipation. Sixth, our analysis of substance use data was limited to those data collected at study time points and did not include potentially important alternative tobacco and nicotine delivery products such as e-cigarettes, vaping, or waterpipes. Seventh, while many factors can contribute to youth in our county emancipating on their 18th birthday vs. sometime between their 18th and 21st birthdays (e.g., an individual youth choosing to emancipate regardless of county recommendations, eligibility of an individual youth for after-care and post-emancipation services), our analysis was limited to the information available to us about when a young person emancipated, and did not include information about why emancipation was or was not delayed until a young person’s 21st birthday. Eighth, our analysis was limited to a small (N=151) population of young people from a single county that resided within a 1-hour driving distance from our academic medical center, and was fairly homogenous in terms of geographic variation – the experiences these young people had may or may not generalize to other urban communities, and likely do not reflect experiences in rural communities. Finally, we were unable to examine mechanisms driving gender differences in attitudes about sexual risk behaviors, which may have been due to how questions were worded. The examination of mechanisms contributing to attitudes about sexual risk behavior by gender is an opportunity for future work.

Despite these limitations, this study makes an important contribution to our understanding of health risks for young people emancipating from foster care. Specifically, in these longitudinal data collected from a sample that is representative of young people ages 16 to 20 years and emancipating from foster care, no variation in indicators of health risk behaviors was observed leading up to or following emancipation. This indicates that, at least for substance use and attitudes toward sexual risk behavior, interventions targeted much earlier than the time of emancipation, and likely before the age of 16, are necessary in order to alter poor health outcomes documented for young adults who have emancipated (Courtney & Dworsky, 2006; Courtney et al., 2010; Courtney et al., 2007). It is possible that to be effective, these preventative interventions must be delivered at multiple points during adolescence, with sensitivity to the unique needs and experiences of youth in foster care, and with attention to the effects of both gender and adversity. Healthcare and child welfare systems should work collaboratively to develop and implement such programs, especially given the instability young people experience during this critical time in development.

Supplementary Material

1

Funding Source:

This work was undertaken with support from the T32 HP10027 Ruth L. Kirschstein National Research Service Award (NRSA) in Primary Care Research and with support from the CareSource Foundation. Dr. Beal’s time was further supported by K01-DA041620.

Abbreviations:

ACE

adverse childhood experience

CTES

Childhood Trust Events Survey

CI

confidence interval

EHR

electronic health record

HIV

human immunodeficiency virus

MLM

multilevel modeling

SBIRT

screening, brief intervention, and referral to treatment

SAAQ

sexual attitudes and activities questionnaire

SD

standard deviation

SUD

substance use disorder

YRBSS

Youth Risk Behavior Surveillance Survey

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Financial Disclosure: The authors have no financial relationships relevant to this article to disclose.

Conflict of Interest: The authors have no conflicts of interest to disclose.

References

  1. Afifi TO, Taillieu T, Salmon S, Davila IG, Stewart-Tufescu A, Fortier J, Struck S, Asmundson GJG, Sareen J, & MacMillan HL (2020, Aug). Adverse childhood experiences (ACEs), peer victimization, and substance use among adolescents. Child Abuse and Neglect, 106, 104504. 10.1016/j.chiabu.2020.104504 [DOI] [PubMed] [Google Scholar]
  2. Beal SJ, Nause K, Crosby I, & Greiner MV (2018). Understanding Health Risks for Adolescents in Protective Custody. J Appl Res Child, 9(1). [PMC free article] [PubMed] [Google Scholar]
  3. Beal SJ, Nause K, Lutz N, & Greiner MV (2020, Jun). The Impact of Health Care Education on Utilization Among Adolescents Preparing for Emancipation From Foster Care. J Adolesc Health, 66(6), 740–746. 10.1016/j.jadohealth.2019.12.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Beal SJ, Wingrove T, Mara CA, Lutz N, Noll JG, & Greiner MV (2019, Jun). Childhood Adversity and Associated Psychosocial Function in Adolescents with Complex Trauma. Child Youth Care Forum, 48(3), 305–322. 10.1007/s10566-018-9479-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Becker MG, & Barth RP (2000, May-Jun). Power through choices: The development of a sexuality education curriculum for youths in out-of-home care. Child Welfare, 79(3), 269–282. [PubMed] [Google Scholar]
  6. Boat B, Baker D, & Abrahamson S (1996). The Childhood Trust Events Survey. Department of Psychiatry. University of Cincinnati. Cincinnati, OH. [Google Scholar]
  7. Bouchard S, Godbout N, & Sabourin S (2009). Sexual attitudes and activities in women with borderline personality disorder involved in romantic relationships. Journal of Sex and Marital Therapy, 35(2), 106–121. 10.1080/00926230802712301 [DOI] [PubMed] [Google Scholar]
  8. Brener ND, Kann L, McManus T, Kinchen SA, Sundberg EC, & Ross JG (2002, Oct). Reliability of the 1999 youth risk behavior survey questionnaire. J Adolesc Health, 31(4), 336–342. 10.1016/s1054-139x(02)00339-7 [DOI] [PubMed] [Google Scholar]
  9. Bumpus MF, Crouter AC, & McHale SM (2001). Parental autonomy granting during adolescence: Exploring gender differences in context. Developmental Psychology, 37(2), 163–173. 10.1037/0012-1649.37.2.163 [DOI] [PubMed] [Google Scholar]
  10. Centers for Disease Control and Prevention. (2021). State and Local Youth Risk Behavior Survey. https://www.cdc.gov/healthyyouth/data/yrbs/pdf/2021/2021-YRBS-Standard-HS-Questionnaire.pdf
  11. Cooper ML, Shapiro CM, & Powers AM (1998, Dec). Motivations for sex and risky sexual behavior among adolescents and young adults: a functional perspective. Journal of Personality and Social Psychology, 75(6), 1528–1558. 10.1037//0022-3514.75.6.1528 [DOI] [PubMed] [Google Scholar]
  12. Courtney M, Dworsky A, Brown A, Cary C, Love K, & Vorhies V (2011). Midwest evaluation of the adult functioning of former foster youth: Outcomes at age 26. Chapin Hall at the University of Chicago. Retrieved April 2, 2021 from https://www.chapinhall.org/wp-content/uploads/Midwest-Eval-Outcomes-at-Age-26.pdf [Google Scholar]
  13. Courtney ME, & Dworsky A (2006). Early outcomes for young adults transitioning from out-of-home care in the USA. Child & Family Social Work, 11(3), 209–219. [Google Scholar]
  14. Courtney ME, Dworsky A, Lee JS, & Raap M (2010). Midwest Evaluation of the Adult Functioning of Former Foster Youth: Outcomes at Ages 23 and 24. Chapin Hall at the University of Chicago. Retrieved April 2, 2021 from https://www.chapinhall.org/wp-content/uploads/Midwest-Eval-Outcomes-at-Age-23-and-24.pdf [Google Scholar]
  15. Courtney ME, Dworsky AL, Cusick GR, Havlicek J, Perez A, & Keller TE (2007). Midwest evaluation of the adult functioning of former foster youth: Outcomes at age 21. Chapin Hall at the University of Chicago. Retrieved April 2, 2021 from https://www.chapinhall.org/wp-content/uploads/Midwest-Eval-Outcomes-at-Age-21.pdf [Google Scholar]
  16. Cuffee JJ, Hallfors DD, & Waller MW (2007, Jul). Racial and gender differences in adolescent sexual attitudes and longitudinal associations with coital debut. J Adolesc Health, 41(1), 19–26. 10.1016/j.jadohealth.2007.02.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Economics Center on behalf of Hamilton County Job and Family Services and the Higher Education Mentoring Initiative. (2017). Economic Analysis of Hamilton County’s Emancipated Foster Youth. https://fosteractionohio.files.wordpress.com/2019/09/2017-hemi-research-study.pdf
  18. Elze DE, Auslander W, McMillen C, Edmond T, & Thompson R (2001, Aug). Untangling the impact of sexual abuse on HIV risk behaviors among youths in foster care. AIDS Education and Prevention, 13(4), 377–389. 10.1521/aeap.13.4.377.21427 [DOI] [PubMed] [Google Scholar]
  19. Felitti VJ, Anda RF, Nordenberg D, Williamson DF, Spitz AM, Edwards V, Koss MP, & Marks JS (1998). Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults. The Adverse Childhood Experiences (ACE) Study. American Journal of Preventive Medicine, 14(4), 245–258. [DOI] [PubMed] [Google Scholar]
  20. Freundlich M, Avery RJ, & Padgett D (2007, Feb). Care or scare: the safety of youth in congregate care in New York City. Child Abuse and Neglect, 31(2), 173–186. 10.1016/j.chiabu.2006.09.002 [DOI] [PubMed] [Google Scholar]
  21. Garrido EF, Weiler LM, & Taussig HN (2018, May). Adverse Childhood Experiences and Health-Risk Behaviors in Vulnerable Early Adolescents. J Early Adolesc, 38(5), 661–680. 10.1177/0272431616687671 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Greiner MV, & Beal SJ (2018, Jul). Developing a Health Care System for Children in Foster Care. Health Promot Pract, 19(4), 621–628. 10.1177/1524839917730045 [DOI] [PubMed] [Google Scholar]
  23. Haggerty KP, Barkan SE, Skinner M, Ben Packard W, & Cole JJ (2016, Winter). Feasibility of Connecting, a Substance-Abuse Prevention Program for Foster Teens and their Caregivers. J Soc Social Work Res, 7(4), 639–659. 10.1086/686986 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Hamel KG (2016). Resilience in Latin American Immigrant Families in Toronto: Defining the Roles of Parenting and Culture in Promoting Healthy Child Development. Retrieved April 2, 2021 from https://yorkspace.library.yorku.ca/xmlui/bitstream/handle/10315/33390/Hamel_Kayla_G_2016_Masters.pdf?sequence=2&isAllowed=y
  25. Hendrick S, Hendrick C, Slapion-Foote MJ, & Foote FH (1985). Gender differences in sexual attitudes. Journal of Personality and Social Psychology, 48(6), 1630–1642. 10.1037/0022-3514.48.6.1630 [DOI] [Google Scholar]
  26. Henry DB, Schoeny ME, Deptula DP, & Slavick JT (2007). Peer Selection and Socialization Effects on Adolescent Intercourse Without a Condom and Attitudes About the Costs of Sex. Child Development, 78(3), 825–838. 10.1111/j.1467-8624.2007.01035.x [DOI] [PubMed] [Google Scholar]
  27. Hofer S, Hoffman L, Little T, Bovaird J, & Card N (2007). Modeling contextual effects in longitudinal studies. Routledge. [Google Scholar]
  28. James S, Montgomery SB, Leslie LK, & Zhang J (2009, Sep). Sexual Risk Behaviors Among Youth in the Child Welfare System. Child Youth Serv Rev, 31(9), 990–1000. 10.1016/j.childyouth.2009.04.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kenward MG, & Roger JH (1997, Sep). Small sample inference for fixed effects from restricted maximum likelihood. Biometrics, 53(3), 983–997. [PubMed] [Google Scholar]
  30. Kim HK, Buchanan R, & Price JM (2017, Jul). Pathways to Preventing Substance Use Among Youth in Foster Care. Prev Sci, 18(5), 567–576. 10.1007/s11121-017-0800-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kuhn C (2015, Sep). Emergence of sex differences in the development of substance use and abuse during adolescence. Pharmacology and Therapeutics, 153, 55–78. 10.1016/j.pharmthera.2015.06.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Landsverk J, Garland A, & Leslie L (2002). APSAC handbook on child maltreatment. Sage. [Google Scholar]
  33. Logan-Greene P, Kim BKE, & Nurius PS (2016, Fall). Childhood adversity among court-involved youth: Heterogeneous needs for prevention and treatment. J Juv Justice, 5(2), 68–84. [PMC free article] [PubMed] [Google Scholar]
  34. Miller EA, Paschall KW, & Azar ST (2017). Latent classes of older foster youth: Prospective associations with outcomes and exits from the foster care system during the transition to adulthood. Children and Youth Services Review, 79, 495–505. 10.1016/j.childyouth.2017.06.047 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Mitchell SG, Gryczynski J, O’Grady KE, & Schwartz RP (2013, May-Jun). SBIRT for adolescent drug and alcohol use: Current status and future directions. Journal of Substance Abuse Treatment, 44(5), 463–472. 10.1016/j.jsat.2012.11.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Murphy DA, Rotheram-Borus MJ, & Reid HM (1998, Apr). Adolescent gender differences in HIV-related sexual risk acts, social-cognitive factors and behavioral skills. Journal of Adolescence, 21(2), 197–208. 10.1006/jado.1997.0141 [DOI] [PubMed] [Google Scholar]
  37. Noll JG, & Shenk CE (2013, Apr). Teen birth rates in sexually abused and neglected females. Pediatrics, 131(4), e1181–1187. 10.1542/peds.2012-3072 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Noll JG, Trickett PK, & Putnam FW (2003, Jun). A prospective investigation of the impact of childhood sexual abuse on the development of sexuality. J Consult Clin Psychol, 71(3), 575–586. 10.1037/0022-006x.71.3.575 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Palmer L, Ahn E, Traube D, Prindle J, & Putnam-Hornstein E (2020). Correlates of entry into congregate care among a cohort of California foster youth. Children and Youth Services Review, 110, 104772. [Google Scholar]
  40. Ram N, & Grimm K (2007). Using simple and complex growth models to articulate developmental change: Matching theory to method. International Journal of Behavioral Development, 31(4), 303–316. 10.1177/0165025407077751 [DOI] [Google Scholar]
  41. Ramiro-Sánchez T, Ramiro MT, Bermúdez MP, & Buela-Casal G (2018). Sexism and sexual risk behavior in adolescents: Gender differences. International Journal of Clinical and Health Psychology, 18(3), 245–253. 10.1016/j.ijchp.2018.04.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Raudenbush SW, & Bryk AS (2002). Hierarchical linear models: Applications and data analysis methods (Vol. 1). Sage. [Google Scholar]
  43. Resnick MD, Bearman PS, Blum RW, Bauman KE, Harris KM, Jones J, Tabor J, Beuhring T, Sieving RE, Shew M, Ireland M, Bearinger LH, & Udry JR (1997). Protecting Adolescents From Harm: Findings From the National Longitudinal Study on Adolescent Health. JAMA, 278(10), 823–832. 10.1001/jama.1997.03550100049038 [DOI] [PubMed] [Google Scholar]
  44. Siegel A, Benbenishty R, & Astor RA (2016). A comparison of adolescents in foster care and their peers in high school: A study of substance use behaviors and attitudes. Journal of Child & Adolescent Substance Abuse, 25(6), 530–538. [Google Scholar]
  45. Slonim-Nevo V, Auslander WF, Ozawa MN, & Jung KG (1996, Summer). The long-term impact of AIDS-preventive interventions for delinquent and abused adolescents. Adolescence, 31(122), 409–421. [PubMed] [Google Scholar]
  46. Smith T, Clark JF, & Nigg CR (2015, Jan). Insights in public health: Building support for an evidence-based teen pregnancy and sexually transmitted infection prevention program adapted for foster youth. Hawaii J Med Public Health, 74(1), 27–32. [PMC free article] [PubMed] [Google Scholar]
  47. Stevens SB, Brice CS, Ale CM, & Morris TL (2011). Examining depression, anxiety, and foster care placement as predictors of substance use and sexual activity in adolescents. Journal of Social Service Research, 37(5), 539–554. [Google Scholar]
  48. Trejos-Castillo E, & Trevino-Schafer N (2018). Handbook of Foster Youth. Routledge. [Google Scholar]
  49. Turney K, & Wildeman C (2017, Feb). Adverse childhood experiences among children placed in and adopted from foster care: Evidence from a nationally representative survey. Child Abuse and Neglect, 64, 117–129. 10.1016/j.chiabu.2016.12.009 [DOI] [PubMed] [Google Scholar]
  50. Twenge JM, Sherman RA, & Wells BE (2015, Nov). Changes in American adults’ sexual behavior and attitudes, 1972–2012. Archives of Sexual Behavior, 44(8), 2273–2285. 10.1007/s10508-015-0540-2 [DOI] [PubMed] [Google Scholar]
  51. United States Department of Health and Human Services Administration for Children and Families. (2020). The AFCARS Report (AFCARS Report #27, Issue. https://www.acf.hhs.gov/sites/default/files/documents/cb/afcarsreport27.pdf
  52. University of Hawai’i Center on the Family. (2012). Hawai’i KIDS COUNT Issue Brief. Improving Outcomes for Youth Transitioning Out of Foster Care. https://uhfamily.hawaii.edu/sites/uhfamily.hawaii.edu/files/publications/ImpOutcomesYouthFosterCare.pdf
  53. Vaughn MG, Ollie MT, McMillen JC, Scott L Jr., & Munson M (2007). Substance use and abuse among older youth in foster care. Addictive Behaviors, 32(9), 1929–1935. 10.1016/j.addbeh.2006.12.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Werner-Wilson RJ (1998, Fall). Gender differences in adolescent sexual attitudes: the influence of individual and family factors. Adolescence, 33(131), 519–531. [PubMed] [Google Scholar]
  55. Wolff KT, Cuevas C, Intravia J, Baglivio MT, & Epps N (2018, Nov). The effects of neighborhood context on exposure to adverse childhood experiences (ACE) among adolescents involved in the juvenile justice system: Latent classes and contextual effects. Journal of Youth and Adolescence, 47(11), 2279–2300. 10.1007/s10964-018-0887-5 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS1765716-supplement-1.docx (28.9KB, docx)

RESOURCES