The primary risk for acquiring HIV/AIDS during adolescence comes from high-risk sexual behavior and drug use. Typical behaviors initiated during adolescence increase risk, such as sexual and drug experimentation, but little is known about the causal mechanisms associated with these behaviors among young people. Youths with mental health problems are at even greater risk of exposure because they engage in the same behaviors as their school-age peers but at higher rates (
Brown et al., 1997a;
Donenberg et al., 2001). Specifically, when compared with youths attending public schools, psychiatrically hospitalized teens are twice as likely to be sexually active, twice as likely not to use condoms, and more than twice as likely to use intravenous drugs (
DiClemente and Ponton, 1993). Teens in inpatient psychiatric care also engage in more self-cutting and sharing cutting utensils than nonpsychiatrically ill youths (
DiClemente et al., 1991), and adolescents with psychiatric disorders are more likely to have a history of STDs than youths without psychiatric disorders (
Baker and Mossman, 1991). Teens in outpatient mental health care also report high rates of HIV/AIDS risk behavior, including sexual intercourse (54.7%) and having been pregnant (8.3%). Among sexually active teens in outpatient psychiatric care, 22% report early sexual initiation (≤14 years old), 12.8% report having had an STD, 48.9% report using drugs/alcohol while having sex, and 55.3% report having had sex without using a condom (
Donenberg et al., 2001,
2003). Early sexual debut increases risk of infection because of added opportunities for sexual encounters and multiple partners. Finally, as many as 30% of youths with psychiatric disorders report comorbid substance abuse (
Arrufo et al., 1994), and alcohol and drugs interfere with safe decision making and effective management of affective arousal.
Understanding the processes associated with HIV transmission among teens with mental health problems requires a broad contextual framework that includes individual and social factors. A social-personal framework of HIV risk behavior among teens in psychiatric care includes maturational constructs implicated in adolescent risk taking (e.g., pubertal development, perceptions of immortality, identity exploration, limitations in abstract thinking) and key psychosocial and contextual risk factors (). The model underscores the interplay of four factors described below: personal attributes, family context, peer and partner relationships, and environmental circumstances (
Donenberg and Pao, 2003).
HIV Risk and Personal Attributes
Personal attributes implicated in adolescent risk behavior include cognitions about HIV/AIDS (knowledge, attitudes and beliefs, impaired decision making), affect dysregulation, mental health problems (internalizing and externalizing), history of sexual abuse, and personality traits (sensation seeking, achievement orientation, value on health).
Cognitions Adolescents generally report accurate information about HIV/AIDS (e.g., cause of infection, modes of transmission and prevention), but increasing knowledge has not altered sexual risk taking (
Morrison-Beedy et al., 2003). Positive attitudes and beliefs about HIV prevention, however, are related to health-promoting behavior. Perceptions of personal vulnerability, motivation to prevent transmission, and behavioral intentions to practice safer sex are among some of the cognitive factors that predict safe sex behavior among youths (
Fisher and Fisher, 1995;
Katz et al., 1995;
Lawrence, 1993). Greater self-efficacy to use condoms is associated with delayed sexual debut (
Santelli et al., 2004).
Many teens in psychiatric care have important cognitive deficits that place them at elevated risk of exposure to HIV. Dysfunctional thinking about relationships and problems in accurate judgment interfere with their ability to assess risk and lead to poor decision making, diminished problem solving, and poor reality testing (
Brown et al., 1997a). Although perceptions of self-efficacy to practice prevention are linked to less risky behavior, many youths with psychiatric problems lack the self-confidence and interpersonal and social skills (e.g., assertiveness, effective communication) necessary to negotiate safe sex practices (
Brown et al., 1997b;
Carey et al., 1997). Thus, sexual behavior among teens in psychiatric care is more likely to be impulsive and spontaneous and not the result of calm decision making (
Donenberg et al., 2005).
Affect Dysregulation Affect regulation skills facilitate successful coping with distress and may help reduce risk behaviors (
DiClemente et al., 2001). Poor affect regulation limits teens’ ability to weigh the pros and cons of using risky sex to obtain short-term relief, thereby increasing their risk of HIV infection (
Cooper et al., 2003). Moreover, deficits in emotion regulation are associated with mental health problems, such as conduct disorder and depression (
Catanzaro, 2000;
Eisenberg et al., 2001). Many youths in psychiatric care experience high levels of negative emotional arousal and lack effective emotion regulation skills. Instead, they respond impulsively to negative arousal to decrease distress. Individuals who lack a cognitive link between the negative stimuli and their impulsive action are unable to weigh the risks/benefits of sex or to plan for protective behavior. Alternatively, these teens adopt short-term self-soothing behaviors (e.g., risky sex, sharing cutting utensils) over self-regulation (
Tice et al., 2001). Distress and anxiety in sexual situations may be particularly stressful for these youths because of relationship maintenance concerns (e.g., “What if my partner rejects me?”) or previous trauma. Emotion regulation can be influenced through learning and modeling, especially in a therapeutic context (
Ciccheti et al., 1995).
Mental Health Problems Two categories of mental health problems, externalizing (e.g., aggression, delinquency) and internalizing (e.g., depression, anxiety), are differentially linked to HIV risk behavior (
Donenberg et al., 2001;
Tubman et al., 2003). Youths with externalizing problems engage in a broad array of risky behaviors, including frequent sexual activity, early sexual debut, low rates of condom use, high numbers of sexual partners, and high rates of prostitution, drug use, needle sharing, exchanging sex for drugs, and drug/alcohol use before and during sex (
Koopman et al., 1994;
Rotheram-Borus and Koopman, 1991;
Stiffman and Cunningham, 1991). Additional risks come from inadequate sexual communication skills and susceptibility to peer norms that encourage deviant sexual behavior. Internalizing problems are also related to HIV risk, including low perceived self-efficacy, decreased assertiveness, and minimal ability to negotiate safe sex with a partner (
Brooks-Gunn and Paikoff, 1997;
Brown et al., 1997a). Depression and low self-esteem are linked to sexually permissive attitudes, having sexually active friends, low contraceptive use, high risk of pregnancy, and nonvirgin status (
Dolcini and Adller, 1994;
Rotheram-Borus et al., 1995;
Whitbeck et al., 1993). Risk may not be uniform across mental health problems, however. Hospitalized teens with conduct disorder report higher rates of risk behavior than do teens with an affective disorder (
DiClemente et al., 1989), and
Donenberg et al. (2001) found that externalizing but not internalizing problems were linked to greater sexual risk taking and substance use.
Sexual Abuse Childhood sexual abuse is consistently associated with elevated rates of HIV risk behavior (
Brown et al., 1997b,
2000). Sexually abused youths report earlier sexual debut, more frequent sexual activity, less consistent condom use, lower self-efficacy for condom use, increased concern with conforming to peer sexual norms, anxiety about partner rejection, and more lifetime sexual partners than nonabused peers (
Lodico and DiClemente, 1994). Abused youths are more likely to have gotten someone pregnant or to have been pregnant and to have engaged in coercive sex (
Brown et al., 1997b). In role-plays, abused teens lacked appropriate assertiveness by communicating their views less consistently, clearly, and directly than nonabused teens. Deficits in sexual communication may reflect abused teens’ fear of partner rejection, feelings of powerlessness, and general passivity in relationships (
Brown et al., 2000).
Personality Traits Three personality traits are particularly relevant for adolescent risk behavior, value on health, sensation seeking, and achievement motivation. Greater health concerns are associated with decreased risk behavior (
Costa et al., 1996), and youths’ reported value on health predicts delayed sexual intercourse with new partners (
Rosengard et al., 2004). Sensation seeking or the willingness to take physical and emotional risks to obtain novel and complex experiences is associated with inconsistent condom use, behavior problems among youths (
Brown et al., 1992;
Zuckerman, 1994), and more sexual partners and unknown sexual partners among adults (
Fisher and Misovich, 1990). High sensation seekers are more likely than low sensation seekers to have had sex, had unwanted sex, and used alcohol or marijuana (
Donohew et al., 2000). Negative attitudes about school, poor academic performance, and low achievement motivation are related to risky sexual behavior (
Kirby, 2002b). High achievement motivation, positive school involvement, and strong academic performance are linked to delayed sexual debut, increased contraceptive use, and decreased rates of pregnancy and childbearing (
Brooks-Gunn et al., 1993b;
Resnick et al., 1997).
HIV Risk and Family Context
A large body of research documents a central role of parents and families in adolescents’ sexual values, attitudes, and behavior (
Crosby and Miller, 2002;
Perrino et al., 2000). The family is the primary source of sexual socialization for children (
Fisher and Feldman, 1998;
Perrino et al., 2000). Four areas of family functioning are consistently related to youths’ sexual risk taking and attitudes: affective characteristics (warmth, support, hostility), instrumental characteristics (monitoring, supervision, control), parent-adolescent communication, and parental attitudes and behavior.
Affective Characteristics Family affective characteristics can mitigate high-risk behavior among teens. Family availability, support, connectedness, and cohesion; high relationship satisfaction between mothers and adolescents; and low parental hostile control and family conflict are all related to reduced sexual experience, less risky sexual behavior, increased use of birth control, and delayed sexual debut (
Borawski et al., 2003;
Donenberg et al., 2003;
McBride et al., 2003;
Miller et al., 2001).
Instrumental Characteristics Parental monitoring, supervision, strictness/permissiveness, and involvement are also linked to teens’ sexual experience and sexual risk taking. Parental monitoring lessens youths’ opportunities for sexual activity (
Paikoff, 1995), predicts less risky sex (
Metzler et al., 1994), and reduces other high-risk behaviors that often co-occur with risky sex, such as drug and alcohol use and delinquency. In one study, however, negotiated unsupervised time (knowing with whom and where the adolescent would be) was related to both safer sex behavior
and increased risky sexual activity (
Borawski et al., 2003). These data suggest that negotiation may be less important than actual supervision in reducing risk taking. Among teens in psychiatric care, greater parental monitoring and supervision and less parental permissiveness are related to reduced sexual risk taking, but links are stronger for girls than for boys (
Donenberg et al., 2002).
Parent–Teen Communication Parent–teen communication about sex influences adolescent sexual behavior (
DiIorio et al., 2003;
Jaccard et al., 2000). Studies are mixed, however, and suggest that more frequent communication is related to both more and less risk taking (
Miller et al., 1998;
Whitaker and Miller, 2000). There is evidence that the quality of communication matters more than frequency (
Wilson and Donenberg, 2004). For example, positive parent–teen communication (open, receptive, comfortable) about sexual topics is related to less sexual experience and reduced risky sexual behavior among adolescents (
Hutchinson et al., 2003;
Miller et al., 1998). There may be a unique relationship between sexual risk taking and parent–teen communication for teens in psychiatric care.
Wilson and Donenberg (2004) found higher rates of risky sexual behavior among teens in which the parent–teen communication was characterized as more mutual or in which parents tended to act more like peers than authority figures. Moreover, those teens whose parents disagreed with them (i.e., behavior that was intended to correct or change the adolescents’ actions or opinions) and were more directive (i.e., parents behaved in a demanding or dominant manner) reported less sexual risk taking. Appropriate intergenerational boundaries (
Minuchin, 1974), absent in many families in psychiatric care, may be critical to preventing adolescent risk taking.
Parental Attitudes and Behavior Parents influence youths’ sexual health and development through modeling and example (
Jaccard and Dittus, 2000;
Wickrama et al., 1999). Mothers’ sexual behaviors are associated with sexual risk taking by their daughters (
Kotchick et al., 1999), and, compared with women who were older at first childbirth, teen mothers are more likely to have daughters who are sexually active as adolescents (
Hardy et al., 1998). Daughters of teen mothers often become teen mothers themselves (
Serbin et al., 1998). Perceptions of maternal approval of birth control are related to increased likelihood of sexual initiation and birth control use (
Jaccard and Dittus, 2000), and perceptions of parental disapproval of teen sex are related to delayed sexual debut, fewer partners, and decreased sexual activity and teen pregnancy (
Meschke et al., 2002;
Miller et al., 1999).
HIV Risk and Peer and Partner Relationships
Peers and romantic partners become increasingly important during adolescence. Three areas of peer and partner relationships are especially salient for adolescent risk behavior: relationship concerns, peer influence, and partner communication.
Relationship Concerns Romantic relationships often produce intense fears of rejection and abandonment (
Welsh et al., 2003). Responsible sexual behavior, such as abstinence or condom use, is a potential source of conflict with and even rejection by romantic partners (
Eyre et al., 1998). Thus, safer sex may be a low priority for teens who believe that maintaining relationships is more important than preventing HIV. Evidence of this is particularly strong for girls and women (
Eyre et al., 1998). Desire for intimacy, love, and affiliation has been linked to sexual behavior among boys and girls (
Ott et al., 2004;
Sanderson and Cantor, 1995) and a primary reason for first sexual intercourse (
Rodgers, 1996).
Eyre et al. (1998) found that adolescent girls believed asking a partner to use a condom would endanger the trust between partners. It is less clear whether the same concerns exist for boys. Pressure to maintain relationships to meet intimacy needs may be particularly powerful for youths in psychiatric care because they tend to have strained, conflictual, and unsupportive relationships with family, peers, and partners (
Brown et al., 1997a;
Seefeldt et al., 2003). These teens may respond less assertively to partner pressure to avoid disconnection (
Donenberg et al., 2001;
Welsh et al., 2003).
Peer Influence The desire for conformity peaks during adolescence, and peers are an important source of sexual information. Adolescent sexual behavior often reflects perceptions of normative peer behavior, including decisions to use birth control and condoms, early sexual initiation, and oral sex (
Kinsman et al., 1998;
Prinstein et al., 2001). In one study, positive peer norms about abstinence were the single best predictor of delayed sexual initiation among middle school students (
Santelli et al., 2004). Among youths in psychiatric care,
Wilson et al. (unpublished, 2004) found that more negative peer influence (i.e., peer support and approval of high-risk behavior, including drinking alcohol, smoking marijuana, smoking cigarettes, and having sex) was associated with less likelihood of using a condom and increased likelihood of having sex while using drugs/alcohol. Negative peer influence has been strongly associated with risky sex among youths in outpatient psychiatric care and mediated links between psychopathology and risk (
Donenberg et al., 2001). Moreover, peer influence is highly correlated with alcohol and drug use (
Donenberg et al., 2001), and substance use is associated with increases in risky sexual behavior.
Partner Communication Assertive communication with sexual partners is essential to practice safe sex, but teens often have difficulty communicating assertively and negotiating safe sex practices (
Crosby et al., 2003;
Hutchinson and Cooney, 1998;
Whitaker et al., 1999). Communication between teenage sexual partners is related to less sexual risk taking, more HIV prevention self-efficacy, fewer partners, and more consistent condom use (
Hutchinson and Cooney, 1998;
Tschann and Adler, 1997;
Whitaker et al., 1999). Teens who express a desire to use condoms to their partner are more likely to use them, and youths who discuss their sexual history have fewer partners. Adolescents who feel confident about initiating safer sex discussions with a partner are also more likely to use a condom and/or resist pressures to engage in high-risk sex in real sexual encounters (
Lawrence, 1993). Unfortunately, those teens in psychiatric care who are more vulnerable to rejection will avoid these discussions for fear of losing the relationship, thereby placing themselves at even greater risk of exposure. Moreover, mental health problems reduce effective communication and self-assertion with peers and partners (
Brown et al., 1997a).
HIV Risk and Environmental Circumstances
Environmental factors shape adolescent sexual behavior. Neighborhood disadvantage (e.g., poverty, violence), ethnic composition, exposure to community violence, neighborhood disorganization (e.g., physical deterioration, drug trafficking), and stressful life events (e.g., divorce) are associated with early sexual initiation, multiple sexual partners, premarital childbearing, increased sexual activity, and greater permissiveness (
Brooks-Gunn et al., 1993a;
Ramirez-Valles et al., 2002;
Sucoff and Upchurch, 1998). Some studies suggest, however, that environmental context explains only a small fraction of the problem (
Blum et al., 2000;
Santelli et al., 2000). Nevertheless, the presence of supportive family or other adult relationships (e.g., teacher, counselor, aunt), commitment to education, and consistent involvement in church activities (
Hawkins et al., 1992;
Wills et al., 1992) can mitigate the impact of environmental stress on sexual risk taking.
Taken together, the research supports a broad social-personal framework of HIV/AIDS risk among youths with mental health problems. This model offers several directions for HIV prevention targets to reduce transmission behaviors and underscores the need for programs to address a multitude of factors that affect risk.