Brief Report: Parents Matter in HIV-Risk Among Probation Youth

Abstract

We investigated the relationship between parenting practices (i.e., parental monitoring, parent permissiveness, and parent-teen communication), and probation youth’s HIV-related sexual risk behavior (i.e., ever having sex, condom use, alcohol and marijuana use before sex). Participants were 61 male and female juvenile offenders, ages 13–17, on probation and awaiting sentencing. Results indicated different relationships between parenting and HIV-related sexual risk behavior for probation boys and girls. Parental monitoring, parenting permissiveness, and parent-teen communication were collectively related to whether girls’ ever had sex and with boys’ use of alcohol and marijuana use before last sex. Findings underscore the important role of parenting on probation teens’ HIV risk behaviors.

According to recent estimates, people under 25 years old account for 15% to 30% of HIV cases in the United States (Morris et al, 2006). Juvenile offenders are among those at greatest risk because of high rates of risky sexual behavior (Teplin, Mericle, McClelland, & Abram, 2003), substance use (McClelland, Teplin, & Abram, 2004), mental illness (Fazel, Doll, & Langstrom, 2008), and sexually transmitted infections (STI) (Morris, Baker, & Huscroft, 1992) that increase their vulnerability to HIV/AIDS.

Parent-teen communication about sexual risk (e.g., being able to talk openly about sex-related topics) and instrumental parenting styles, such as high parental monitoring (e.g., parents knowing youth’s whereabouts) and low permissiveness (e.g., having firm rules and expectations for youth to follow), are consistently related to adolescent sexual risk taking (e.g., delayed sexual debut, condom use, fewer partners) (Crosby, DiClemente, Wingood, Lang, & Harrington, 2003; Hadley et al., 2009), but few studies focus on juvenile offenders. Although, poor parenting practices are often implicated in adjudicated teens’ problems, research suggests parenting heterogeneity, with instrumental parenting associated with fewer risk behaviors (Robertson, Baird-Thomas, & Stein, 2008; Steinberg, Blatt-Eisengart, & Cauffman, 2006). Adjudicated youth with permissive parents have been found to engage in more risk behaviors than those with authoritarian parents (Steinberg et al., 2006). Similarly, Robertson and colleagues (2008) found parental monitoring to be negatively associated with sexual risk behavior among incarcerated youth. These findings suggest parenting may provide important protective effects from sexual risk behaviors for adjudicated youth. However, no published studies have examined the role of parenting on sexual risk taking among probation youth specifically. Unlike incarcerated youth, most probation youth live at home, affording parents an opportunity to influence offending youth’s sexual risk.

The present study tested the associations between parenting behaviors (parental monitoring, parental permissiveness, parent-adolescent communication) and four HIV-related sexual risk behaviors identified among samples of adjudicated youth (ever having sex, sex without a condom, drinking alcohol before sex, and smoking marijuana before sex) (Teplin, et al., 2003). Based on previous research, we expected more parental monitoring, more parent-child communication, and less parental permissiveness to be related to less sexual-risk behavior. Evidence of gender differences in parental monitoring and parent-child communication (Jacobson & Crockett, 2000; Svensson, 2003) led us to evaluate the hypotheses separately for boys and girls.

Methods

Overview

This study reports baseline data from a developmental pilot project to design and test an HIV/AIDS, mental health, and substance use prevention program for teens on probation (i.e., PHAT Life). Participants were recruited from Illinois Cook County’s Evening Reporting Centers (ERC), an alternative to incarceration following arrest. Research staff presented the project to youths as a group at the ERC. All teens were given an informational flyer with options to indicate interest in the study. Interested teens provided parental contact information to obtain consent. Written consent and assent were obtained for all participants. Baseline assessments were approximately 2.5 hours long, and most participants completed them over multiple sessions (M = 1.39, SD = .49, R =1–2). Youth were compensated $50 for completing the assessments. All study procedures were approved by the University of Illinois at Chicago’s Institutional Review Board, with special attention to vulnerable populations.

Participants

Participants were 61 juvenile offenders, ages 13–17 (M = 15.23, SD = 1.09), on probation and awaiting sentencing (typically 21 days). Eighty two percent (N = 63) of eligible youths (N = 78), agreed to participate, but two failed to complete the baseline assessment. Demographic characteristics are shown in Table 1.

Table 1.

Characteristics of the Sample and Rates of HIV-Risk Behaviors

	Full Sample (N = 61)		Female (N = 28)		Male (N = 33)
	Mean (SD)	%	Mean (SD)	%	Mean (SD)	%
Age	15.68 (1.06)		15.65 (1.08)		15.24 (1.05)
Gender (female)		46
Ethnicity
African American		93.4		96.4		91.2
Hispanic/Latino		4.9		3.6		5.9
Non-Hispanic White		1.6		0.0		2.9
HIV-risk behaviors
Ever had sex a		65.6		57.1		72.7
HIV-risk behaviors of youth who have had sex (Girls, N = 16; Boys, N = 24)
Condom use during last sex		65.0		68.8		62.5
Sex after drinking alcohol		17.5		18.8		16.7
Sex after smoking marijuana		32.5		25.0		37.5

NOTE: Means and standard deviations (SD) are reported for continuous variables, and percentages (%) are reported for categorical variables

^aVariables with statistically significant gender differences.

Measures

Risky sexual behavior

Teens self-reported sexual and drug behaviors on the AIDS-Risk Behavior Assessment (ARBA; Donenberg, Emerson, Bryant, Wilson, & Weber-Shifrin, 2001), a structured interview designed for urban teens. The ARBA was administered via audio-computer assisted interview (A-CASI). Analyses tested youth reports of 1) ever having vaginal/anal sex (yes/no), 2) condom use during last sex (yes/no), 3) alcohol use at last sex (yes/no), and 4) marijuana use at last sex (yes/no).

Parental monitoring and permissiveness

The Parenting Style Questionnaire (PSQ) (Oregon Social Learning Center, 1990) assessed parental monitoring (e.g., “How often do you check in with your parents/caretakers or an adult after school/work before going out?”) and permissiveness (e.g., “Your parents/caretakers let you go any place you please without asking.”) on a 5-point scale. The PSQ was administered via A-CASI. Each item is rated on a Likert scale from 1 (never/almost never) to 5 (always/almost always). Parental monitoring and parental permissiveness were each assessed with 4 items, with a range of 4-20 for each subscale. Higher scores represent more monitoring and more permissiveness. Internal consistency was r =.84 for monitoring and r = .63 for permissiveness.

Parent-adolescent communication about risk

Teens reported on the degree of openness and receptivity of parents during discussions about sex using a 9-item adapted version of the Miller et al. (1998) scale (e.g., “I can ask my mother/father/guardian the questions I really want to know about sexual topics.”) administered via A-CASI. Each item is rated on a Likert scale from 1 (strongly disagree) to 4 (strongly agree), with a range of 9-36. A higher score indicated more open and receptive communication. Internal consistency was r =.55.

Results

Data Analysis

Overview

We conducted four multivariate logistic regression analyses to examine the associations among parenting practices (parental monitoring, parental permissiveness, and parent-teen communication) and HIV-related sexual risk behaviors (ever having sex, condom use at last sex, drinking alcohol before last sex, and smoking marijuana before last sex) among youth on probation. Due to studies suggesting age differences in adolescent sexual risk behavior (Fergus, Zimmerman, & Caldwell, 2007), participants’ age range of 13 – 17 years, and the study’s small sample size, we controlled for age in all analyses. Analyses were conducted separately for boys and girls. In each analysis, age was entered into the first block, and the three parenting variables were entered into the second block. Overall models were considered significant at p ≤ .05.

Predictors of HIV-related Sexual Risk Behaviors Among Girls on Probation

The overall models testing condom use at last sex (χ² = 4.68, df = 4, p = .32), alcohol use before last sex (χ² = 7.40, df = 4, p = .11), and marijuana use before last sex (χ² = 4.16, df = 4, p = .38) failed to reach statistical significance. However, the overall model for ever having sex for girls was statistically significant, χ² = 9.58, df = 4, p = .04. In this model, age was not statistically significant, but the three parenting variables predicted ever having sex, χ² = 7.89, df = 3, p = .04. Examination of the individual parenting factors revealed no single parenting variable accounted for unique variance in girls’ ever having sex.

Predictors of HIV-related Sexual Risk Behaviors Among Boys on Probation

The overall models testing ever having sex and condom use at last sex among boys were not statistically significant (χ² = 4.29, df = 4, p = .36 for ever having sex, and χ² = 3.13, df = 4, p = .53, for condom use at last sex). However, the models examining alcohol and marijuana use before last sex were, χ² = 11.62, df = 4, p = .02 and χ² = 9.40, df = 4, p = .05 for alcohol and marijuana use before sex, respectively. For both models, age was not statistically significant but the three parenting variables were, χ² = 9.30, df = 3, p = .02 and χ² = 9.32, df = 3, p = .02 for the models predicting alcohol use before sex and marijuana use before sex, respectively. No single parenting variable accounted for unique variance in probation boys’ use of alcohol prior to sex. However, the relationship between parental permissiveness and boys’ use of marijuana before sex was marginally significant (p = .05); more parental permissiveness was associated with greater likelihood that boys reported having sex after smoking marijuana (see Table 2).

Table 2.

Odds Ratios (OR) and 95% Confidence Intervals (CI) for Predicting Sexual Risk, Using Parenting Variables, Controlling for Age

Predictors	Model 1 Ever Have Sex			Model 2 Condom Use			Model 3 Alcohol Use			Model 4 Marijuana Use
	β	OR	CI	β	OR	CI	β	OR	CI	β	OR	CI
Females (n = 28)
Age	0.82	2.27	[.80, 6.46]	0.25	1.28	[.26, 6.37]	−.78	0.45	[.26, 07.95]	−.53	0.58	[.07, 4.57]
Parenting
Monitoring	−.28	0.75	[.53, 1.06]	−.33	0.71	[.46, 1.10]	−.37	0.68	[.30, 01.55]	−.35	1.43	[.75, 2.70]
Permissiveness	−.07	0.93	[.72, 1.19]	−.12	1.13	[.78, 1.64]	1.44	4.22	[.38, 46.95]	−.10	0.89	[.59, 1.35]
Communication	−.20	0.81	[.62, 1.07]	−.13	0.87	[.60, 1.26]	0.19	1.21	[.82, 01.80]	−.43	0.64	[.34, 1.22]
Males (n = 33)
Age	0.22	1.25	[.56, 2.78]	0.57	1.77	[.64, 4.89]	−4.07	0.01	[.00, 09.52]	−.10	0.90	[.24, 3.30]
Parenting
Monitoring	0.14	1.15	[.94, 1.40]	−.12	0.88	[.70, 1.09]	−0.36	0.69	[.34, 01.40]	−.11	0.89	[.69, 1.14]
Permissiveness	−.45	0.95	[.72, 1.26]	−.01	1.01	[.72, 1.43]	0.80	2.24	[.80, 06.20]	−.41	1.51*	[.99, 2.29]
Communication	−.16	0.84	[.64, 1.10]	−.09	0.91	[.69, 1.19]	−1.14	3.13	[.64, 15.36]	−.25	1.29	[.89, 1.86]

^*p = .05

Discussion

This study examined the relationship between parenting practices and probation youth’s HIV-related sexual risk behavior. Results indicate that parenting plays an important role in probation teens’ sexual risk behavior. Parenting was related to whether girls’ had ever had sex and boys’ substance use before sex. Parenting was not related to condom use at last sex for either girls or boys.

The different relationship between parenting and sexual risk behaviors for boys and girls, suggests distinct pathways based on gender. Our finding of an association between parenting and sexual experience for only girls is consistent with previous research revealing gender differences in parental predictors of youth becoming sexually active (Browning, Levanthal, & Brooks-Gunn, 2005). On the other hand, results indicating a relationship between parenting and substance use among boys are contrary to previous findings, which found this relationship to exist among girls only (Choquest, Hassler, Morin, Falissard, & Chau, 2008; Yeh, Chiang, & Huang, 2006). However, our findings are consonant with a recent study showing a relationship between parent-adolescent communication and substance use among boys only (Luk, Farhat, Iannotti, Simons-Morton, 2010). The inconsistency in findings may be due to studies measuring different aspects of parent-teen communication (e.g., communication ease, comfort, content, frequency, etc.). With the exception of parental permissiveness related to boys’ marijuana use, no single parenting factor accounted for girls or boys’ risky sexual or drug use behavior. Rather, the combination of parenting practices best explained teens’ risk. Given probation youths’ multiple co-morbid problems, it is not surprising that less risk taking was related to a collection of parenting behaviors. It is possible that multiple protective parenting behaviors (monitoring, low permissiveness, and parent-teen communication), rather than any single one, are necessary to reduce risk in this highly vulnerable multi-problem population.

Contrary to our expectations, parenting was not related to probation youth’s use of condoms at last sex. Although research has documented the influence of parent-adolescent communication on youth’s sexual risk behaviors, recent studies suggest this relationship to be based on direct parent-adolescent communication about condom use, rather than general warmth or openness with sexual communication (Hadley, et al., 2009). Drawing from these findings, our own results indicating no association may be due to our measure of communication, which focused on the type of communication (e.g., warmth and openness), rather than the content of communication (e.g., condom use).

Results should be interpreted in the context of study limitations. The small sample size along with the low internal consistency of some measures may have reduced accuracy in detecting effects, yet results revealed important patterns that can inform new research questions. Caution must also be taken in inferring gender differences based on comparisons across analyses, as interpretations may be misleading in the absence of a formal significance test. However, the results are consistent with research indicating gender differences in adolescent sexual behavior and how youth are parented. This study relied on adolescent report of parenting practices, which may not reflect actual parent behavior. Also, youth’s self-reported HIV-risk behaviors may be subject to biases, but research indicates general correspondence between biological tests and self-reported sexual and drug behaviors (DiClemente, Sales, Danner, & Crosby, 2011). Finally, study findings are based on a sample of youth on probation who are predominantly African American and may not generalize to other populations.

This study suggests that parenting matters for probation youths, a population typically viewed as disconnected and disengaged from their families. Results contribute to the growing body of evidence that probation teens’ parents display a range of parenting practices, and when taken together, they can protect girls and boys from high-risk sexual behavior. Clinicians working with probation youth and families may help reduce youth’s HIV risk by helping parents to implement instrumental parenting practices, and by strengthening parent-youth communication regarding sexual risk reduction strategies. Findings argue in favor of developing family-based HIV prevention programs for probation youth (Tolou-Shams, Stewart, Fasciano, & Brown, 2009). Results suggest such programs be tailored by gender, paying special attention to particular sexual risk behaviors for boys (e.g., substance use during sex) and girls (e.g., sexual initiation). Future research is needed to clarify the process by which parenting strategies differentially protect boys and girls on probation.