摘要：We systematically searched 9 biomedical and social science databases (1980–2012) for primary and secondary interventions that prevented or reduced 2 or more adolescent health risk behaviors (tobacco use, alcohol use, illicit drug use, risky sexual behavior, aggressive acts). We identified 44 randomized controlled trials of universal or selective interventions and were effective for multiple health risk behaviors. Most were school based, conducted in the United States, and effective for multiple forms of substance use. Effects were small, in line with findings for other universal prevention programs. In some studies, effects for more than 1 health risk behavior only emerged at long-term follow-up. Integrated prevention programs are feasible and effective and may be more efficient than discrete prevention strategies. Adolescence is associated with an increased prevalence of health risk behaviors, including substance use, sexual risk, and aggressive behavior. 1 The vast majority of substance use is initiated in adolescence. 2,3 In the United Kingdom, adolescence is associated with higher rates of sexually transmitted disease and abortion relative to other age groups. 4,5 The majority of young people will experience bullying or aggression during adolescence. 6 In addition, adolescent mortality has increased relative to other age groups, largely because of accidents and unintentional injuries. 7 Early initiation of health risk behaviors is associated with negative outcomes throughout adolescence and adulthood, such as addiction and substance abuse; poor sexual, mental, and physical health; and lower occupational and educational attainment. 8,9 The social and economic costs associated with adolescent risk behaviors have made them a key focus of public health policy initiatives internationally. 10 A growing body of research suggests that health risk behaviors often do not occur in isolation. Smoking, drinking, illicit drug use, sexual risk, and aggressive behaviors are all mutually predictive. 11 For drug use and some forms of sexual risk, co-occurrence with other risk behaviors is essentially normative. Previous research suggests that co-occurrence of risk behaviors is driven by shared risk factors such as peer influences or sensation seeking or by state-specific traits such as the direct effects of substance use or aggression on other risk behaviors. Common risk factors can be found in many domains, including social, psychological, family, school, and neighborhood. 12–14 Evidence also suggests gateway effects, whereby participation in a given health risk behavior leads to increased risk for others, partially attributable to exposure effects and decreases in perceived danger of such behaviors. 15 For example, adolescent smoking and drinking have been linked with subsequent illicit drug use. 16 This typical co-occurrence is often not reflected in the organization of policies and interventions to reduce adolescent risk behavior. National policy regarding adolescent health risk behavior is often organized in nonoverlapping risk-specific policies. 10 Some intervention developers recognize that single-risk interventions for adolescents may trigger effects on other risk behaviors, particularly on multiple forms of substance use. For several reasons, targeting multiple health risk behaviors (MHRBs) simultaneously may be more effective and efficient than targeting a single risk behavior. Limited funding for prevention interventions requires that interventions reduce health risks efficiently, highlighting the importance of synchronized prevention efforts. Time constraints, for example in schools, also make coordinated intervention for multiple risks attractive. Furthermore, it is unclear how discrete interventions might interact in cases where they are not coordinated both theoretically and practically, raising the possibility that uncoordinated interventions could be ineffective or cause harm. 17 Beyond these logistic concerns, research regarding the mechanisms for MHRBs suggests that integrated interventions may be essential for the effective prevention of risk behaviors. If common risk factors explain co-occurrence of risk behaviors, then targeting those risk factors should prove effective for MHRBs. Gateway theories offer further support for integrated intervention strategies; if a given risk behavior increases risk for another, effective prevention strategies for the latter must also focus on the former. For example, sexual intercourse accompanied by alcohol or illicit drug use is linked to a lower likelihood of condom use, 18 so targeting substance misuse may be a feasible approach to reducing unsafe sex. Although the development of integrated interventions for MHRBs requires an understanding of their mechanisms, including common risk factors and gateway effects, the existing literature regarding effective interventions is also a key source of evidence for the development of interventions. The majority of evaluations report on interventions that target 1 risk behavior. However, identifying interventions that have reduced MHRBs can help inform the development of future interventions by indicating which combinations of risk behaviors can be targeted in coordinated approaches, what contexts and approaches are most successful, and what are the other attributes of coordinated interventions, such as duration and participant age. Limited data exist on effective intervention programs to prevent MHRBs. To date, we are aware of only 1 published review that assessed the effectiveness of interventions on MHRBs in young people. 13 That review focused exclusively on studies reporting concurrently on substance use and sexual risk outcomes. We expanded on this work by reviewing additional combinations of outcomes. We undertook a systematic review designed to identify randomized controlled trials that reported significant universal or selective intervention effects for at least 2 health risk behaviors among adolescents.
