Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Oct 1.
Published in final edited form as: J Drug Issues. 2022 Nov 12;53(4):621–636. doi: 10.1177/00220426221139423

Examining Gender Differences in the Relationship Between School Bonding and Opioid Misuse Among Justice-Involved Adolescents

Enya B Vroom 1,2, Micah E Johnson 2, Zahra Akbari 2, Zachary Frederick 1, Skye C Bristol 2
PMCID: PMC10691556  NIHMSID: NIHMS1865058  PMID: 38046931

Abstract

Justice-involved adolescents (JIAs) have an increased risk for opioid use disorder and overdose related to opioid misuse (OM). Consequences of untreated OM include recidivism and poor educational outcomes, which can be harsher for female JIA. Therefore, identifying relevant factors and settings that reduce the risk for OM is critical. Schools are a central institution in adolescent development. Drawing on social control theory, JIA with higher levels of school bonding was hypothesized to attenuate risk for OM. Cross-sectional data on 79,960 JIA from the Florida Department of Juvenile Justice were examined. Multivariate and stratified logistic regression analyses were employed. On average, for every one-unit increase in school bonding, JIA had 22%, female JIA had 23%, and male JIA had 22% lower odds of OM. Results suggest school bonding and the school context should be considered in treatment and how this setting may impact OM intervention outcomes among JIA.

Keywords: juvenile justice, school involvement, opioid misuse, social control theory, school bonding

Introduction

Opioid misuse (OM) among adolescents remains a chief public health concern within the United States (U.S.). Justice-involved adolescents (JIA) – youth involved with the juvenile justice system before the age of 18 – show significantly higher levels of lifetime prevalence of OM when compared to the general adolescent population (Wojciechowski, 2019). Recent research has shown that correctional populations have an extraordinarily higher risk of relapse and overdose upon reentry into the community (Green et al., 2018). JIA are an understudied high-risk population and are crucial to understanding the etiology of OM, particularly among correctional populations. In addition, females tend to suffer harsher consequences from both justice involvement and substance misuse, including recidivism, trauma, and limited access to behavioral health services (Herrera & Boxer, 2019). Addressing female JIA substance use treatment needs has been a consistent challenge within correctional settings (Barrett et al., 2015). Therefore, identifying the factors and institutions associated with decreasing the likelihood of OM among JIA is a critical step to inform treatment practices. Social control theory, also known as social bonding theory, submits that an individual’s behavior is shaped by the attachments or bonds they form with meaningful individuals and/or institutions (Chriss, 2007). Schools are critical institutions in youth development as they provide an outlet, especially during adolescence, for further development of self-regulation and social and emotional skills (Osher et al., 2014). Youth bonding in schools with peers and educators may be a vital factor influencing the risk for OM among JIA, especially among females.

Opioid Misuse (OM)

Adolescence is a critical period of development and brain maturation. Research has shown the adolescent brain to be highly sensitive to reward, with elevated sensitivity to sensation-seeking and impulsiveness (Reniers et al., 2016), which may contribute to the early initiation of drug use (Winters & Arria, 2011). Specifically, misusing opioids can increase adolescents’ vulnerability to being opioid-dependent in adulthood (Windisch & Kreek, 2020). JIA are at greater risk of opioid use disorder, overdose, and death related to OM (Binswanger et al., 2013; Johnson & Cottler, 2020). OM and opioid-related overdoses among adolescents have tripled since 2000 (Linton et al., 2021; Riley & Johnson, 2020). Justice involvement, coupled with OM, has the potential to disrupt healthy development, potentially leading to long-term adverse health and life outcomes (e.g., addiction and recidivism) (Rudes et al., 2021). Understanding OM among JIA within the school context can provide insight into the analysis of OM among a population with a high risk of misuse and overdose. Identifying strategies to prevent and reduce OM among this unique population is essential.

School Bonding

Travis Hirschi’s social control theory, also known as social bonding theory, submits that individuals who have strong attachments to conventional society are less likely to engage in deviant behaviors when compared to persons who have weak bonds (Hirschi, 1969). Attachment refers to developing significant relationships with key individuals and institutions. Developing strong prosocial attachments are theorized to deter deviant behavior due to an individual’s desire to maintain and appease their prosocial bonds (Cassino & Rogers, 2016). Schools are one of the most important social institutions for adolescents, having widespread and lasting impacts, including the development of identity, social, and cognitive learning skills (Osher et al., 2014; Verhoeven et al., 2019). School bonding refers to adolescents’ connections with their schools, agents of the school, and other elements of their academic life (Maddox & Prinz, 2003). Drawing on social bonding theory, adolescents who are more connected to school are theorized to be less likely to engage in substance misuse (e.g., OM) because they are socialized to meet the schools’ expectations and are willing to endorse the established norms in the school context. This connectivity provides access to prosocial relationships with peers, teachers, and other positive adults, which promote prosocial development (Cassino & Rogers, 2016). In addition, other elements of social bonding theory are also important for school bonding. The additional elements include: (1) commitment to conventional goals (i.e., the importance of obtaining an education); (2) involvement in pursuing socially acceptable activities (i.e., attending school or completing homework); and (3) the level of belief in shared norms/values (i.e., youth belief in the value of obtaining an education or respect for authority) (Gilmore et al., 2005; Hirschi, 1969). Cogent evidence underscores the importance of considering the intricacies of social bonds, especially school bonds, because of the potential impact they have on delinquent behavior (i.e., OM) and service outcomes (Lee et al., 2019; Nash et al., 2019).

A study examining factors that influence the development of school bonding found that adolescent students misusing substances had lower rates of school bonding. The study also found that increases in substance misuse and lower academic grades were associated with a more substantial decline of school bonding over time (Oelsner et al., 2011). Yang and Anyon (2016) examined the relationship between race and risk behaviors (e.g., poor grades, fighting, and substance use) among adolescents and the mediating role of school bonding. Among the different risk behaviors, results showed school bonding as the strongest mediator for substance use for Black, Latinx, and multiracial adolescent students. The strength of this result may be related to the actual relationship between substance use and school bonding, given that students who use substances also tend to be truant more frequently, resulting in fewer opportunities to bolster their attachment to school and school staff (Yang & Anyon, 2016). Several other empirical studies have also found that youth who are more bonded to their schools tend to engage in lower levels of substance use overall (Henry et al., 2012; Li et al., 2011; Snyder et al., 2015).

Gender

Hirschi originally suggested that social controls and the effects of bonding were gender-neutral (Hirschi, 1969). However, more recent literature suggests differences between genders and how gender affects social bonds. Dever et al. (2012) conducted a study examining the influence of protective (i.e., factors that are influential based on those who are at-risk) and promotive factors (i.e., factors associated with positive outcomes among all people regardless of risk) on substance misuse among adolescents. Results showed that school bonding emerged as having both protective and promotive effects on substance use among females in the 10th grade, especially females who were considered high risk, when compared to males. This suggests the effect of school bonding on substance use may be moderated by gender (Dever et al., 2012). Similarly, Oelsner et al. (2011) found that gender impacted the trajectory of the development of school bonding, with males having lower values of school bonding than females, suggesting a more ‘unfavorable trajectory’ of school attachment among males.

Current Study

Although extant literature has shown that school bonding can be a protective factor for substance misuse among adolescents, there is limited existing research that specifically targets JIA and OM. Therefore, the purpose of this study was to examine the association between school bonding and OM among JIA and how this relationship varied by gender. Drawing on empirical evidence and social bonding theory, this study hypothesized that JIA who reported higher levels of school bonding would be less likely to meet criteria for past 30-day (P30D) OM. It was also hypothesized that female JIA with higher levels of school bonding would have a lower likelihood of OM when compared to male JIA. The study examined statewide data provided by the Florida Department of Juvenile Justice (FLDJJ) to test these hypotheses. The study adjusted for sociodemographic covariates shown to be correlated with OM, including: race, gender, school type, age at first offense, and household income.

Methods

Data Collection

The FLDJJ is a state-level agency that oversees and manages juvenile detention centers and youth arrests in Florida. From 2004 to 2015, FLDJJ conducted comprehensive data collection on all youth arrested within Florida. Youth formally processed into the FLDJJ were administered the Positive Achievement Change Tool (PACT) assessment. This analytical tool has been validated across several samples of FLDJJ data and published in a wide array of peer-reviewed journals (Baglivio & Jackowski, 2013). The PACT assessment is administered by trained personnel using structured interviews. The PACT assessed the risk to re-offend and criminogenic need domains, including: criminal history, school, free time, employment, relationships, family, substance use, mental health, antisocial attitudes, aggression, and social skills. PACT interviews were conducted at FLDJJ intake centers. The total FLDJJ dataset consisted of 79,960 JIA. For the current study, 2,197 cases were omitted due to insufficient data related to OM or school bonding, resulting in a total sample size of 77,763 JIA.

Sample

The sample of 77,763 represent JIA who were enrolled in the FLDJJ system, completed the PACT assessment, turned 18 years old by 2015, and had sufficient OM and school bonding data. Of this sample, 38% were non-Latinx White (n=29,453), 46% were non-Latinx Black (n=35,789), 15.5% were Latinx (n=12,142), and .5% were another race (n=379). Of the total sample, 21.9% were female (n=17,010), and the mode age of JIA between 2004–2015 was 13–14.

Measures

Opioid Misuse.

The FLDJJ criteria for OM was a positive urine analysis, youth self-disclosure of illicit opioid use or non-medical use of prescription opioids, or other evidence of unlawful opioid consumption such as police or medical records within the past 30-day. Urine analysis took place during intake and was facilitated by a credentialed health professional (e.g., physician or nurse). OM included both illicit opioids (e.g., heroin) and prescription opioids (e.g., Oxycodone, Fentanyl, and Hydrocodone). Responses were coded as a dichotomous measure with response items, (0) no, did not meet FLDJJ criteria for past 30-day (P30D) OM, or (1) yes, met criteria for P30D OM indicated by urine analysis, self-disclosure, or evidence.

School Bonding Index.

The School Bonding Index (SBI) was developed to examine the association between school bonding and OM among JIA. Eight binary variables from the PACT assessment were combined to create the SBI. These variables included: (1) history of school expulsions/suspensions, which was coded as (1) if JIA did not have a history of expulsions/suspensions, and (0) if otherwise; (2) problems with school conduct, which was coded as (1) if JIA had received recognition for good behavior or had no problems with school conduct, and (0) if otherwise; (3) JIA belief in their school providing an encouraging environment, which was coded as (1) if JIA believed their school offered an encouraging environment, and (0) if otherwise; (4) JIA involvement in school activities, which was coded as (1) if JIA were involved in one or more school activities, and (0) if they were not involved with any school activities; (5) JIA school attendance, which was coded as (1) if JIA had good attendance with few absences or had no unexcused absences, and (0) if they had some partial-day or full-day unexcused absences or were considered habitual truants; (6) JIA likelihood of staying in school, which was coded as (1) if JIA indicated it was very likely for them to stay in school and graduate, and (0) if they were uncertain if they were going to stay in school or it was not possible for them to stay in school; (7) JIA belief in the value of getting an education, which was coded as (1) if the JIA believed there was value in getting an education, and (0) if either JIA somewhat thought there was value or believed there was no value in getting an education; (8) number of school staff (e.g., teachers, coach) JIA felt comfortable talking with, which was coded as (1) if JIA had at least one person they were comfortable communicating with, and (0) if otherwise. Higher SBI scores indicated stronger school bonding among JIA. Cronbach’s alpha was calculated to test the internal consistency of the SBI. Cronbach’s alpha for the SBI was .72, indicating an acceptable level of reliability (Tavakol & Dennick, 2011).

Control Variables.

This study adjusts for control variables that are known to be correlated with P30D OM: race, gender, school type, age at first offense, and household income. Race was measured using a four-item categorical variable (0= White, 1= Black, 2= Latinx, 3= other). Gender is a social norm that incorporates the sex categorizations of males and females (West & Zimmerman, 1987). Sexual minorities largely acknowledge gender as being more inclusive and less offensive (National Institutes of Health (n.d.)). Therefore, gender is used in place of sex to refer to males and females. Gender was measured by a self-reported binary variable (0= male, 1= female). School type was measured using a nine-item categorical variable (0= private/public, 1= vocational, 2= alternative, 3= GED or graduated, 4= home school, 5= college, 6= other, 7= expelled, 8= dropped out). Age at first offense was measured via a five-item categorical variable (0= 12 and under, 1= 13 to 14, 2= 15, 3= 16, 4= over 16). Household income was operationalized using a four-item ordinal variable reporting the combined annual income of JIA and their families (0= under $15,000, 1= from $15,000 to $34,999, 2= from $35,000 to $49,999, 3= $50,000 and over).

Analytical Procedures

The Complete Case Analysis approach to missing data was appropriate given that there was minimal missing data (<3% [2,197]) and the sample size was large (n= 77,763). Demographic data were examined using descriptive statistics. A chi-squared test of independence was performed to assess whether there was a significant association between the categorical variables and P30D OM. Multivariate logistic regression was used to calculate adjusted odds ratios (AORs) and 95% confidence intervals for P30D OM. Additionally, two stratified regression models were conducted separately for male and female JIA. The Hosmer–Lemeshow test confirmed adequate model fit. Data analysis was performed using STATA v. 17 (StataCorp, 2017).

Results

Of the total sample with data on OM (77,763), approximately 2.66% (2,069) JIA met the criteria for P30D OM (i.e., P30D OM group). Table 1 shows the descriptive statistics of the main and control variables for JIA in both P30D OM and no-P30D OM groups (i.e., JIA that did not meet the criteria for P30D OM). Overall, JIA in the P30D OM group had lower SBI scores when compared to the no-P30D OM group. The average SBI score was 1.96 for JIA in the P30D OM group, while the average SBI score was 2.85 for JIA in the no-P30D OM group.

Table 1.

Descriptive Statistics by Past 30-Day (P30D) Opioid Misuse (OM) Status.

Variable
 Observations
(n = 77,763)
 No-P30D OM
(n = 75,694)
 P30D OM (n= 2,069)
n % n % Or mean* n % Or mean*
Main Variables
 Opioids 77,763 75,694 2,069
 School Bonding Index 77,763 75,694 2.85* 2,069 1.96*
Control Variables
 Gender
  Female 17,010 21.9 16,351 21.6 659 31.8
  Male 60,753 78.1 59,343 78.4 1,410 68.2
 Race/Ethnicity
  White 29,453 38.0 27,789 37.0 1,664 80.4
  Black 35,789 46.0 35,618 47.0 171 8.2
  Latinx 12,142 15.5 11,927 15.5 215 10.5
  Others 379 0.5 360 0.5 19 0.9
 School Type
  Public/Private 30,750 40.0 30,311 40.0 439 21.2
  Vocational 747 1.0 731 1.0 16 0.8
  Alternative 12,652 16.3 12,319 16.3 333 16.1
  GED or Graduated 7,570 10.0 6,990 9.3 292 14.3
  Home School 1,586 2.2 1,514 2.0 72 3.5
  College 603 0.7 581 0.8 22 1.2
  Other 9,496 12.3 9,173 12.2 323 15.7
  Expelled 8,060 10.4 7,728 10.3 332 16.1
  Dropped Out 5,994 7.8 5,765 7.7 229 11.1
 Age at First Offense
  12 and Under 19,187 24.7 18,793 24.8 394 19.0
  13–14 28,501 36.6 27,696 36.6 805 38.9
  15 13,990 18.0 13,559 17.9 431 20.8
  16 10,396 13.4 10,101 13.4 295 14.3
  Over 16 5,689 7.3 5,545 7.3 144 7.0
 Household Income
  Under $15,000 20,239 26.0 19,784 26.1 455 22.0
  $15,000–34,999 40,797 52.5 39,807 52.6 990 47.9
  $35,000–49,999 11,435 14.7 11,057 14.6 378 18.3
  $50,000 and Over 5,292 6.8 5,046 6.7 246 12.0
Open in a new tab

Note:

*

Indicates variable mean.

There also were differences in the control variables in the P30D OM and no-P30D groups. Of the JIA in the P30D OM group, 31.8% were female, which is approximately 10% higher than the percentage of females in the no-P30D OM group (21.6%). Among JIA in the P30D OM group, 68.2% were male, 80.4% were White, 8.2% were Black, and 10.5% were Latinx. White JIA had the largest proportion of the total sample in the P30D OM group (80.4%), while Black JIA had the largest percentage in the no-P30D OM group (47%). JIA in the P30D OM group were also more likely to be expelled or dropped out of school, 16.1% and 11.1%, respectively, compared to JIA in the no-P30D OM group (10.3% and 7.7%, respectively). Considering age at first offense, JIA who were 12 or younger accounted for 19% of the P30D OM group, compared to 24.8% in the no-P30D OM group. In addition, JIA who were age 15 at their first offense accounted for 20.8% of the P30D OM group. This percentage was higher than their proportion in the no-P30D OM (17.9%). JIA with lower household income (under $15,000 and $15,000 to $34,999) accounted for less of the P30D OM group (22% and 47.9%, respectively) when compared to the no-P30D OM group (26.1% and 52.6%, respectively). JIA with higher household income ($35,000–49,000 and $50,000 and over) accounted for more of the P30D OM group (18.3% and 12%, respectively) than JIA in the no-P30D OM group with lower household income (14.6% and 6.7%, respectively).

Table 2 shows the multivariate logistic regression results that were conducted to examine how school bonding was associated with the likelihood of P30D OM within the entire sample. For every one-unit increase in the SBI, JIA were 22% (OR: .78, 95% CI, .76–.80) less likely to meet the criteria for P30D OM. Overall, female JIA had 72% higher odds of meeting the criteria for P30D OM when compared to male JIA. Black and Latinx JIA were less likely, 92% and 68%, respectively, to meet the criteria for P30D OM compared to White JIA. Results showed that any school type other than private/public was associated with higher odds of P30D OM. More specifically, JIA who were expelled were 2.36 times more likely to meet the criteria for P30D OM compared to JIA enrolled in private/public schools. Also, JIA in alternative schools had 68% higher odds of meeting the criteria for P30D OM compared to JIA in private/public schools. On average, JIA who were 13 or older at the time of their first offense were more likely to meet the criteria for P30D OM than JIA who were 12 or younger at the age of their first offense. JIA with a household income of $50,000 and over were 51% more likely to meet the criteria for P30D OM compared to JIA with household income under $15,000.

Table 2.

Association Between School Bonding and Past 30-Day Opioid Misuse.

Variables Odds Ratios CI
School Bonding Index 0.78*** (0.76–0.80)
Gender (ref=Male)
 Female 1.72*** (1.58–1.87)
Race (ref= White)
 Black 0.08*** (0.06–0.11)
 Latinx 0.32*** (0.23–0.44)
 Other 0.98 (0.56–1.70)
School Type (ref= Public/Private)
 Vocational 1.56* (0.98–2.50)
 Alternative 1.68*** (1.43–1.97)
 GED or Graduated 2.25** (1.93–2.61)
 Home School 2.00*** (1.51–2.66)
 College 2.20** (1.32–3.67)
 Other 2.24*** (1.92–2.62)
 Expelled 2.36*** (2.05–2.72)
 Dropped Out 1.24*** (1.06–1.45)
Age at First Offense (ref= 12 and Under)
 13–14 1.25*** (1.11–1.40)
 15 1.36*** (1.18–1.56)
 16 1.28*** (1.08–1.53)
 Over 16 1.28** (1.00–1.66)
Household Income (ref= Under $15,000)
 $15,000 -34,999 1.01 (0.88–1.16)
 $35,000 -49,999 1.15 (0.95–1.40)
 $50,000 and Over 1.51*** (1.25–1.83)
 Constant 0.05*** (0.04–0.06)
 Observations 77,458 77,458
Open in a new tab

Note: Robust CI in parentheses.

***

p<.01

**

p<.05

*

p<.1

Table 3 shows the results of two stratified logistic regression models that were conducted to evaluate how school bonding was associated with the odds of P30D OM, separately for male and female JIA. The first and third columns indicate a slight difference between male and female JIA in how school bonding is correlated with the odds of P30D OM. For every one-unit increase in the SBI, Female JIA were 23% less likely, and male JIA were 22% less likely to meet the criteria for P30D OM. The association between school bonding and P30D OM was almost identical between male and female JIA. For Black female JIA, the odds of meeting the criteria for P30D OM were 94% less than White female JIA. In comparison, Black male JIA were 91% less likely to meet the criteria for P30D OM compared to White male JIA. Latinx female and male JIA were less likely, 73% and 66%, respectively, to meet the criteria for P30D OM compared to White female and male JIA.

Table 3.

Association Between School Bonding and Past 30-Day Opioid Misuse by Gender.

Variables Female
 Male
Odds Ratios CI Odds Ratios CI
School Bonding Index 0.77*** (0.74–0.80) 0.78*** (0.76–0.81)
Race (ref= White)
 Black 0.06*** (0.04–0.10) 0.09*** (0.07–0.12)
 Latinx 0.27*** (0.16–0.45) 0.34*** (0.25–0.45)
 Other 1.56 (0.68–3.55) 0.74 (0.36–1.52)
School Type (ref= Public/Private)
 Vocational 2.48*** (1.27–4.82) 1.28 (0.68–2.41)
 Alternative 1.65*** (1.25–2.19) 1.69*** (1.39–2.06)
 GED or Graduated 2.69*** (2.14–3.38) 2.05*** (1.67–2.51)
 Home School 1.73*** (1.13–2.65) 2.15*** (1.53–3.03)
 College 2.89*** (1.25–6.72) 1.83* (0.97–3.43)
 Other 2.34*** (1.85–2.96) 2.20*** (1.80–2.68)
 Expelled 2.13*** (1.52–2.99) 2.42*** (2.00–2.91)
 Dropped Out 1.33* (0.98–1.80) 1.18 (0.94–1.49)
Age at First Offense (ref= 12 and Under)
 13–14 1.11 (0.89–1.39) 1.29*** (1.11–1.50)
 15 1.12 (0.86–1.44) 1.45*** (1.22–1.73)
 16 1.11 (0.86–1.43) 1.34*** (1.07–1.67)
 Over 16 1.06 (0.72–1.57) 1.37** (1.05–1.80)
Household Income (ref= Under $15,000)
 $15,000 -34,999 0.98 (0.77–1.24) 1.03 (0.90–1.20)
 $35,000 -$49,999 1.00 (0.70–1.43) 1.23** (1.01–1.50)
 $50,000 and Over 1.30* (0.95–1.76) 1.63*** (1.34–1.98)
 Constant 0.10*** (0.07–.13) 0.04*** (0.03–0.06)
 Observations 16,940 60,518
Open in a new tab

Note: Robust CI in parentheses.

***

p<.01

**

p<.05

*

p<.1

In addition, the results of the stratified models showed differences between male and female JIA in how school type was associated with the odds of P30D OM. Female JIA that were expelled were 2.13 times more likely to meet the criteria for P30D OM compared to female JIA who were enrolled in public/private schools. In contrast, male JIA that were expelled were 2.42 times more likely to meet the criteria for P30D OM than male JIA who were enrolled in public/private schools. Female JIA enrolled in vocational schools were 2.48 times more likely to meet the criteria for P30D OM compared to female JIA enrolled in public/private schools. Female and male JIA enrolled in alternative schools were more likely, 65% and 69%, respectively, to meet the criteria for P30D OM than female and male JIA enrolled in public/private schools. Overall, male JIA 13 or older at the age of their first offense were 29%–45% more likely to meet the criteria for P30D OM when compared to male JIA who were 12 or younger at their age of first offense. Lastly, male JIA with household income from $34,999 to $49,999 and $50,000 and over were more likely, 23% and 63%, respectively, to meet the criteria for P30D OM compared to male JIA with household income under $15,000.

Discussion

Gender differences in accessing care have been well-established (Abram et al., 2008; Barrett et al., 2015). However, there is still a dearth of research related to the factors associated with disparities in treatment services within the justice system – despite being the primary referral source for adolescents in the U.S. (Chassin, 2008). The current study addresses a gap in the literature by making innovatory use of the FLDJJ dataset with the creation of the SBI to examine the relationship between school bonding and OM among JIA. The study’s first hypothesis was supported with results indicating that JIA with increases in school bonding scores were less likely to report OM. These results align with past literature reporting school bonding to be a negative predictor of substance use (Dever et al., 2012; Henry et al., 2012; Oelsner et al., 2011; Snyder et al., 2015; Yang & Anyon, 2016).

Research has found that school bonding tends to decrease among youth in middle school, which may be related to schools not being adequately equipped to meet the developmental needs of students (Oelsner et al., 2011). In turn, reductions in school bonding, coupled with increased sensation seeking, may explain increased substance use during adolescence. The compelling evidence of the current study suggests that interventions targeting general adolescent risk behaviors and JIA risk behaviors should consider the school context and how this critical setting can impact behavior and intervention outcomes (Yang & Anyon, 2016). For example, clinicians treating this unique population should be aware of how institutions, such as schools, can influence intervention outcomes. In addition, schools may be an essential setting for JIA and their caregivers to not only access OM treatment but also education materials related to proper use of opioid medications, safe storage and disposal of opioid medications, and the risks of misuse and diversion (McCabe et al., 2020).

The study’s second hypothesis was supported by results indicating female JIA with increased levels of school bonding scores were less likely to misuse opioids when compared to male JIA. However, after stratifying the regression models by gender, the difference between females and males was slight. This small difference could indicate that school bonding may be important for both genders. This slight difference contrasts with past research that has shown female JIA continue to have an elevated risk of substance misuse despite males being overrepresented in the substance misuse literature (Herrera & Boxer, 2019) and that gender may moderate the effect school bonding has on substance misuse overall (Dever et al., 2012). The results also showed Black and Latinx JIA were less likely to report misusing opioids when compared to White JIA, especially among Black and Latinx female JIA. Again, these results contrast with previous literature. In their nationally representative sample of Black adolescents, Yockey et al., (2021) found that Black female adolescents were more likely to report non-medical prescription opioid use when compared to males. Female non-medical use of opioids may be used as a coping mechanism to relieve emotional or physical stressors (Piquero & Sealock, 2004). Different biopsychosocial disadvantages females experience can cause stressors such as gender inequality and trauma (e.g., sexual assault) (Docherty et al., 2016; Fasula et al., 2018; Moss, 2002). Given the dearth of research related to gender differences in the substance misuse literature (McHugh et al., 2018), and the contrasting results of the research that does exist, these results highlight the importance of the need to further investigate gender differences among this unique population to more accurately assess how this may impact OM and treatment.

Results of the study also showed that the type of school and/or enrollment status of JIA was associated with the likelihood of OM. More specifically, JIA who had been expelled from school were significantly more likely to misuse opioids when compared to JIA who were enrolled in public or private schools. School bonding has been shown to delay the onset of substance misuse and reduce the average level of substance misuse (Dever et al., 2012). If JIA are not present in schools, they will not have the ability to bond with an institution known to be a protective factor related to substance misuse. Research has shown that exclusionary discipline, a form of punishment that excludes a student from their normal classroom environment (i.e., detentions, suspensions, and expulsions), can be associated with school dropout and involvement with the juvenile justice system (Butler-Barnes & Inniss-Thompson, 2020).

In addition, school disengagement, known as a lack of interest, intrinsic motivation, and eagerness for school, has been linked to delinquent behavior, dropout, and substance misuse (Skinner et al., 2008). Typically, the process of school disengagement occurs before dropping out and can result in exclusionary discipline. Therefore, screening for and preventing school disengagement may be critical in not only preventing expulsion and dropout but also key in preventing and treating OM. Implementing interventions earlier in the process of school disengagement may potentially reduce OM both proximally and distally (Henry et al., 2012). These interventions may be especially important for females, as research has shown that females are more likely to rely on interpersonal relationships (i.e., teachers) that are built and exist within the school setting when compared to males (Burns et al., 2019). These sources of bonding within the school context are known protective factors for disengagement. Therefore, school bonding may be a logical intervention focus point for female JIA. It may serve as a practical approach to capitalize on a potential protective factor that could reduce the likelihood of problem behaviors (i.e., OM) among this specific group (Oelsner et al., 2011). Given the results of the current study, it is also necessary for future research to further investigate the effect school bonding has on male JIA and determine how school bonding may be a protective factor for OM among male JIA.

Both male and female JIA attending alternative schools were over 60% more likely to misuse opioids. This finding may be important for two reasons. First, adolescents attending alternative schools may be at increased risk for substance use and risk behaviors due to experiencing higher levels of social risk factors (e.g., adverse childhood experiences). Second, adolescents attending alternative schools are less likely to be monitored by local, state, or national school-based surveys (e.g., Youth Risk Behavior Survey) that provide prevalence estimates for substance misuse that directly inform policy and program development (Johnson et al., 2019). The current study’s results indicate a school setting where additional OM programming and resources may be needed and where the concept of school bonding should be considered within prevention and treatment.

Further, female JIA attending vocational schools were significantly more likely to misuse opioids. Research has shown individuals attending vocational schools report high rates of substance use, which may be related to relieving work-related stress (Tomczyk et al., 2016). Currently, there is limited literature focused on vocational schools in the U.S. More research is needed to examine how the association between school bonding and gender affects OM outside of mainstream school settings (i.e., public and private) within the U.S. The results of such research may be used to further inform policy change and resource allocation for OM prevention and treatment as well as encourage collaboration between the juvenile justice and education systems.

Lastly, age at first offense was associated with higher odds of misusing opioids exclusively for male JIA. This finding may be particularly important for when to implement an intervention as well as implementation dosage, adaptation, and program reach (Durlak & DuPre, 2008). For example, it may be necessary to screen and intervene at-risk males during middle adolescence (i.e., 13–15 years of age) and provide booster lessons and/or additional treatment as they progress from middle to late adolescence.

Limitations

There are important limitations of the current study that must be discussed. First, this study used cross-sectional data. This type of data limits the ability to infer causal relationships from the associations presented between school bonding and OM long-term from the study’s results. Future research would benefit from conducting longitudinal studies that would allow for the examination of associations over time and temporal sequencing. Secondly, the study did not have access to data measuring JIA’s ability to access opioids and did not differentiate between types of opioids used. Therefore, the study could not examine how these factors may influence OM among JIA. Future studies should collect and analyze data focused on accessibility as well as differentiate between opioid types to examine how these variables may impact use. Thirdly, this study used data from JIA exclusively located within Florida. Therefore, study findings may not be generalizable for JIA nationwide without first considering state-level characteristics and policy. Lastly, the SBI was limited to only including variables that were captured by the PACT assessment, which may have resulted in missing information related to school bonding. Despite this limitation, the SBI included variables that considered all four elements of the social bonding theory (i.e., attachment, commitment, involvement, and beliefs) and had an acceptable level of reliability in terms of the internal consistency of the measure.

Conclusions

This study adds to the existing literature by improving the understanding of the relationship between school bonding and OM and how that may differ by gender among JIA. The study results indicated that higher school bonding scores were associated with a decreased likelihood of OM. However, results of the stratified regression models suggest that school bonding may be important for both male and female JIA. More research is needed to determine how school bonding has the potential to be a protective factor for OM among male JIA. Future research would also benefit from a closer examination of the association of school bonding and its effect on OM stratified by school type. Examining school type may be critical as different school settings differ in funding, access to resources, teacher workload, student expectations, and policies in disciplinary actions. JIA, especially female JIA, continue to be a group that is underrepresented in substance misuse research. The study’s findings highlight the need to utilize schools as an intervention setting for not only OM treatment but also for the prevention of OM through initiatives aimed at enhancing school protective factors (e.g., engagement) and educating youth and their families about safe medication storage and risks of misuse.

Acknowledgments

The data in this study were developed by and obtained from the Florida Department of Juvenile Justice (FLDJJ) in Tallahassee, Florida. Our team would like to especially acknowledge the dedicated professionals at FLDJJ for managing the data and collaborating with investigators.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research reported in this publication was supported by the National Institutes of Health under award numbers T32DA035167 (Dr. Linda B. Cottler), 1K01DA052679 (Dr. Micah E. Johnson, PI), R25DA050735 (Dr. Micah E. Johnson, PI), R25DA035163 (Dr. Micah E. Johnson, USF Site PI), and U01DA051039 (Dr. Micah E. Johnson, USF Site PI). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Florida Department of Juvenile Justice.

Biographies

Enya B. Vroom, PhD, MS, is an NIH NIDA T32 Postdoctoral Fellow in the Department of Epidemiology at the University of Florida. Her research focuses on substance misuse among justice-involved adolescents, with an emphasis on disparities and the implementation of treatment services.

Micah E. Johnson, PhD, MA, is an Assistant Professor in the Department of Mental Health Law and Policy at the University of South Florida. His research focuses on the epidemiology of substance use among pediatric health disparity populations. He also works to improve quality/equity in treating substance use disorder.

Zahra Akbari is a Research Assistant in the Department of Mental Health Law and Policy at the University of South Florida. She is also a PhD candidate in Economics Department. Her research focuses on behavioral and mental health, with emphasis on substance use disorder and racial/gender disparities among justice-involved adolescents.

Zachary Frederick is a Research Trainee in the Study of Teen Opioid Misuse and Prevention lab at the University of Florida, founded by Dr. Micah E. Johnson. His research focuses on the intersection between criminology and opioid misuse epidemiology among pediatric populations.

Skye C. Bristol is a Research Assistant in the Department of Mental Health Law and Policy in the College of Behavioral and Community Sciences at the University of South Florida. Her research focuses on substance misuse, specifically alcohol use disorder, and racial disparities among the Florida Department of Juvenile Justice.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethical Approval

All procedures were in accordance with the ethical standards of the institutional review boards and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. For this type of study, formal consent was not required.

References

  1. Abram KM, Paskar LD, Washburn JJ, & Teplin LA (2008). Perceived barriers to mental health services among youths in detention. Journal of the American Academy of Child and Adolescent Psychiatry, 47(3), 301–308. 10.1097/CHI.0b013e318160b3bbb [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baglivio MT, & Jackowski K (2013). Examining the validity of a juvenile offending risk assessment instrument across gender and race/ethnicity. Youth Violence and Juvenile Justice, 11(1), 26–43. 10.1177/1541204012440107 [DOI] [Google Scholar]
  3. Barrett DE, Ju S, Katsiyannis A, & Zhang D (2015). Females in the juvenile justice system: Influences on delinquency and recidivism. Journal of Child and Family Studies, 24(2), 427–433. 10.1007/s10826-013-9853-8 [DOI] [Google Scholar]
  4. Binswanger IA, Blatchford PJ, Mueller SR, & Stern MF (2013). Mortality after prison release: Opioid overdose and other causes of death, risk factors, and time trends from 1999 to 2009. Annals of Internal Medicine, 159(9), 592–600. 10.7326/0003-4819-159-9-201311050-00005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Burns EC, Bostwick KCP, Collie RJ, & Martin AJ (2019). Understanding girls’ disengagement: Identifying patterns and the role of teacher and peer support using latent growth modeling. Journal of Youth and Adolescence, 48(5), 979–995. 10.1007/s10964-019-00986-4 [DOI] [PubMed] [Google Scholar]
  6. Butler-Barnes ST, & Inniss-Thompson MN (2020). ‘My teacher doesn’t like me’: Perceptions of teacher discrimination and school discipline among African-American and Caribbean Black adolescent girls. Education Sciences, 10(2), 1–14. 10.3390/educsci10020044 [DOI] [Google Scholar]
  7. Cassino PP, & Rogers WS (2016). Hirschi’s social bonding theory nearly 45 years later: A comparison of a traditional, contemporary and hybrid model. Theory in Action, 9(2), 21–44. 10.3798/tia.1937-0237.16009 [DOI] [Google Scholar]
  8. Chassin L (2008). Juvenile justice and substance use The Future of Children, 18(2), 165–183. 10.1353/foc.0.0017 [DOI] [PubMed] [Google Scholar]
  9. Chriss JJ (2007). The functions of the social bond. The Sociological Quarterly, 48(4), 689–712. 10.1111/j.1533-8525.2007.00097.x [DOI] [Google Scholar]
  10. Dever BV, Schulenberg JE, Dworkin JB, O’Malley PM, Kloska DD, & Bachman JG (2012). Predicting risk-taking with and without substance use: The effects of parental monitoring, school bonding, and sports participation. Prevention Science, 13(6), 605–615. 10.1007/s11121-012-0288-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Docherty M, Boxer P, Veysey BM, & Ostermann M (2016). Gender differences in prevalence of internalizing and externalizing symptoms in a justice-referred sample of youth. Journal of Juvenile Justice, 5(2), 104–120. https://www.ojp.gov/pdffiles/251065.pdf#page=109 [Google Scholar]
  12. Durlak JA, & DuPre EP (2008). Implementation matters: A review of research on the influence of implementation on program outcomes and the factors affecting implementation. American Journal of Community Psychology, 41(3–4), 327–350. 10.1007/s10464-008-9165-0 [DOI] [PubMed] [Google Scholar]
  13. Fasula AM, Gray SC, Vereen RN, Carry M, Sales JM, Abad N, & Gelaude DJ (2018). Multiple psychosocial health problems and sexual risk among african American females in juvenile detention: A cross-sectional study. Children and Youth Services Review, 88, 74–80. 10.1016/j.childyouth.2018.02.041 [DOI] [Google Scholar]
  14. Gilmore AS, Rodriguez N, & Webb VJ (2005). Substance abuse and drug courts: The role of social bonds in juvenile drug courts. Youth Violence and Juvenile Justice, 3(4), 287–315. 10.1177/1541204005278803 [DOI] [Google Scholar]
  15. Green TC, Clarke J, Brinkley-Rubinstein L, Marshall BDL, Alexander-Scott N, Boss R, & Rich JD (2018). Postincarceration fatal overdoses after implementing medications for addiction treatment in a statewide correctional system. JAMA Psychiatry, 75(4), 405–407. 10.1001/jamapsychiatry.2017.4614 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Henry KL, Knight KE, & Thornberry TP (2012). School disengagement as a predictor of dropout, delinquency, and problem substance use during adolescence and early adulthood. Journal of Youth and Adolescence, 41(2), 156–166. 10.1007/s10964-011-9665-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Herrera CM, & Boxer P (2019). The role of gender in risk for substance use among justice-involved youth. Children and Youth Services Review, 100, 485–493. 10.1016/j.childyouth.2019.03.023 [DOI] [Google Scholar]
  18. Hirschi T (1969). Causes of delinquency. University of California Press. [Google Scholar]
  19. Johnson KE, Sales A, Rew L, Haussler Garing J, & Crosnoe R (2019). Using polytomous latent class analysis to compare patterns of substance use and co-occurring health-risk behaviors between students in alternative and mainstream high schools. Journal of Adolescence, 75, 151–162. 10.1016/j.adolescence.2019.07.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Johnson ME, & Cottler LB (2020). Optimism and opioid misuse among justice-involved children. Addictive Behaviors, 103(1), 106226. 10.1016/j.addbeh.2019.106226 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Lee MT, Post SG, Wylie AB, Leibowitz GS, Gelboin-Burkhart C, Holiprosad D, Johnson BR, Turk AB, Zacharoff KL, & Zakirkhodjaev N (2019). Transposing the adverse social dynamics of adolescent substance use disorders into more effective treatment and clinician resilience. Alcoholism Treatment Quarterly, 37(4), 513–531. 10.1080/07347324.2018.1559004 [DOI] [Google Scholar]
  22. Li Y, Zhang W, Liu J, Arbeit MR, Schwartz SJ, Bowers EP, & Lerner RM (2011). The role of school engagement in preventing adolescent delinquency and substance use: A survival analysis. Journal of Adolescence, 34(6), 1181–1192. 10.1016/j.adolescence.2011.07.003 [DOI] [PubMed] [Google Scholar]
  23. Linton SL, Winiker A, Tormohlen KN, Schneider KE, McLain G, Sherman SG, & Johnson RM (2021). "People don’t just start shooting heroin on their 18th birthday": A qualitative study of community stakeholders’ perspectives on adolescent opioid use and opportunities for intervention in baltimore, Maryland. Prevention Science, 22(5), 621–632. 10.1007/s11121-021-01226-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Maddox SJ, & Prinz RJ (2003). School bonding in children and adolescents: Conceptualization, assessment, and associated variables. Clinical Child and Family Psychology Review, 6(1), 31–49. 10.1023/a:1022214022478 [DOI] [PubMed] [Google Scholar]
  25. McCabe SE, Schulenberg J, McCabe VV, & Veliz PT (2020). Medical use and misuse of prescription opioids in US 12th-grade youth: School-level correlates. Pediatrics, 146(4), Article e20200387. 10.1542/peds.2020-0387 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. McHugh RK, Votaw VR, Sugarman DE, & Greenfield SF (2018). Sex and gender differences in substance use disorders. Clinical Psychology Review, 66, 12–23. 10.1016/j.cpr.2017.10.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Moss NE (2002). Gender equity and socioeconomic inequality: A framework for the patterning of women’s health. Social Science & Medicine, 54(5), 649–661. 10.1016/S0277-9536(01)00115-0 [DOI] [PubMed] [Google Scholar]
  28. Nash A, Collier C, Engebretson J, & Cron S (2019). Testing the feasibility of measuring recovery in adolescent participants of an alternative peer group: Lessons learned and next steps. Journal of Adolescent Research, 34(6), 655–682. 10.1177/0743558418822332 [DOI] [Google Scholar]
  29. National Institutes of Health.. Sexual and gender minority. In National Institutes of Health Office of Equity, Diversity, and Inclusion. https://www.edi.nih.gov/people/sep/lgbti/safezone/terminology.
  30. Oelsner J, Lippold MA, & Greenberg MT (2011). Factors influencing the development of school bonding among middle school students. Journal of Early Adolescence, 31(3), 463–487. 10.1177/0272431610366244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Osher D, Kendziora K, Spier E, & Garibaldi ML (2014). School influences on child and youth development. In Sloboda Z, & Petras H (Eds.), Defining prevention science (pp. 151–169). Springer. [Google Scholar]
  32. Piquero NL, & Sealock MD (2004). Gender and general strain theory: A preliminary test of broidy and agnew’s gender/GST hypotheses. Justice Quarterly, 21(1), 125–158. 10.1080/07418820400095761 [DOI] [Google Scholar]
  33. Reniers RLEP, Murphy L, Lin A, Bartolomé SP, & Wood SJ (2016). Risk perception and risk-taking behaviour during adolescence: The influence of personality and gender. Plos One, 11(4), Article e0153842. 10.1371/journal.pone.0153842 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Riley TO, & Johnson ME (2020). Parental perceptions linked to opioid misuse among justice-involved children. Children and Youth Services Review, 116(1), 105203. 10.1016/j.childyouth.2020.105203 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Rudes DS, Viglione J, Sheidow AJ, McCart MR, Chapman JE, & Taxman FS (2021). Juvenile probation officers’ perceptions on youth substance use varies from task-shifting to family-based contingency management. Journal of Substance Abuse Treatment, 120(1), 108144. 10.1016/j.jsat.2020.108144 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Skinner E, Furrer C, Marchand G, & Kindermann T (2008). Engagement and disaffection in the classroom: Part of a larger motivational dynamic? Journal of Educational Psychology, 100(4), 765–781. 10.1037/a0012840 [DOI] [Google Scholar]
  37. Snyder SM, Gwaltney AY, & Landeck E (2015). What social bonds have the greatest influence on patterns of substance use among child-welfare-involved youth? Journal of Psychoactive Drugs, 47(4), 308–316. 10.1080/02791072.2015.1075091 [DOI] [PubMed] [Google Scholar]
  38. Tavakol M, & Dennick R (2011). Making sense of Cronbach’s alpha. International Journal of Medical Education, 2(1), 53–55. 10.5116/ijme.4dfb.8dfd [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tomczyk S, Pedersen A, Hanewinkel R, Isensee B, & Morgenstern M (2016). Polysubstance use patterns and trajectories in vocational students – a latent transition analysis. Addictive Behaviors, 58(1), 136–141. 10.1016/j.addbeh.2016.02.027 [DOI] [PubMed] [Google Scholar]
  40. Verhoeven M, Poorthuis AMG, & Volman M (2019). The role of school in adolescents’ identity development. A literature review. Educational Psychology Review, 31(1), 35–63. 10.1007/s10648-018-9457-3 [DOI] [Google Scholar]
  41. West C, & Zimmerman DH (1987). Doing gender. Gender and Society, 1(2), 125–151. http://www.jstor.org/stable/189945 [Google Scholar]
  42. Windisch KA, & Kreek MJ (2020). Review of addiction risk potential associated with adolescent opioid use. Pharmacology Biochemistry and Behavior, 198(1), 173022. 10.1016/j.pbb.2020.173022 [DOI] [PubMed] [Google Scholar]
  43. Winters KC, & Arria A (2011). Adolescent brain development and drugs. The Prevention Researcher, 18(2), 21–24. https://pubmed.ncbi.nlm.nih.gov/22822298 [PMC free article] [PubMed] [Google Scholar]
  44. Wojciechowski TW (2019). Developmental trajectories of opioid use among juvenile offenders: An epidemiological examination of group characteristics and criminological risk factors. Substance Use & Misuse, 54(7), 1203–1213. 10.1080/10826084.2019.1573837 [DOI] [PubMed] [Google Scholar]
  45. Yang J, & Anyon Y (2016). Race and risk behaviors: The mediating role of school bonding. Children and Youth Services Review, 69(1), 39–48. 10.1016/j.childyouth.2016.07.019 [DOI] [Google Scholar]
  46. Yockey A, King K, & Vidourek R (2021). “Here’s the ‘kicker’: School factors associated with opioid misuse among US African American youth. Journal of Ethnicity in Substance Abuse. Epub ahead of print. 10.1080/15332640.2021.1877231 [DOI] [PubMed] [Google Scholar]

RESOURCES