Positive Youth Development Interventions Impacting the Sexual Health of Young Minority Adolescents: A Systematic Review

Participants

The original intention of the literature search was to identify studies that focused solely on minority adolescents 10 to 14 years old. However, the search did not uncover any studies that met the specified age criterion. Therefore, the criterion was adjusted to incorporate studies that included adolescents 10 to 14 years old, which resulted in a variety of age ranges. One study targeted adolescents that were at the elementary, middle, and high school levels (CPPRG, 2014). Seven studies (nine articles) focused solely on the middle school aged adolescents (Dilorio et al., 2007; Dilorio et al., 2006; Dilorio et al., 2002; Flay et al., 2004; O’Donnell et al., 1999; O’Donnell et al., 2002; Philliber et al., 2002; Prado et al., 2007; Trenholm et al., 2007). Four studies (five articles) focused on middle and high school aged adolescents (Clemons et al., 2011; East et al., 2003; Sieving et al., 2011; Sieving et al., 2013; Tello et al., 2010).

The majority of participants in all 12 studies were minority youth. Four studies (five articles) were targeted to African Americans (Clemons et al., 2011; Dilorio et al., 2007; Dilorio et al., 2006; Dilorio et al., 2002; Flay et al., 2004). Two studies were intended for Latino youth (Prado et al., 2007; Tello et al., 2010). The remaining 6 studies comprised a diverse group of youth but the majority of participants were minority (CPPRG, 2014; East et al., 2003; O’Donnell et al., 1999; O’Donnell et al., 2002; Philliber et al., 2002; Sieving et al., 2011; Sieving et al., 2013; Trenholm et al., 2007). Most studies also characterized nearly all or all of their participants as high risk due to family socioeconomic status (SES). Two studies also targeted recruitment to adolescents with an additional risk factor, either because of past high rates of aggressive-disruptive behavior (CPPRG, 2014) or a sibling was pregnant or got someone pregnant (East et al., 2003). Thus, these studies were implemented with high-risk adolescents who might be less likely than youth in high-resource families and communities to make and sustain positive sexual health behavior changes.

Study design

The 12 studies utilized either a quasi-experimental (Clemons et al., 2011; East et al., 2003; Tello et al., 2010) or experimental (CPPRG, 2014; Dilorio et al., 2007; Dilorio et al., 2006; Dilorio et al., 2002; Flay et al., 2004; O’Donnell et al., 1999; O’Donnell et al., 2002; Philliber et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013; Trenholm et al., 2007) study design. Three studies utilized a pre and posttest design (Clemons et al., 2011; East et al., 2003; Tello et al., 2010), and nine studies evaluated outcomes at multiple time periods (CPPRG, 2014; Dilorio et al., 2007; Dilorio et al., 2006; Dilorio et al., 2002; Flay et al., 2004; O’Donnell et al., 1999; O’Donnell et al., 2002; Philliber et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013; Trenholm et al., 2007). The length of the intervention varied by study, ranging from 7 weeks (Dilorio et al., 2007) to 12 years (CPPRG, 2014).

Intervention settings

The interventions were all delivered in either school-based or community settings. Five of the interventions were predominantly school based (Clemons et al., 2011; CPPRG, 2014; Flay et al., 2004; O’Donnell et al., 1999; O’Donnell et al., 2002; Trenholm et al., 2007). Seven interventions were community based (Dilorio et al., 2007; Dilorio et al., 2006; Dilorio et al., 2002; East et al., 2003; Philliber et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013; Tello et al., 2010).

Intervention components

The 12 identified studies targeting young minority adolescents’ sexual health using PYD concepts varied widely in approach. The PYD organizing framework developed by Catalano and colleagues (2004) served as the organizing framework for this review. The 15 PYD constructs serve as a guideline for developing PYD program components and was used here to organize PYD constructs operationalized across studies. Table 3 provides a summary of the program components along with the authors’ attempt to categorize intervention components by the 15 PYD objectives identified by Catalano et al. (2004). Looking across interventions, the PYD constructs utilized most frequently were as follows: promotes bonding (11 of 12 studies), promotes cognitive competence (11 of 12 studies), promotes social competence (11 of 12 studies), and promotes behavioral competence (11 of 12 studies). In addition, provides opportunities for prosocial involvement was utilized by 10 of the 12 studies, fosters belief in future was utilized by 9 of 12 studies, and two constructs (promotes emotional competence and fosters self-efficacy) were utilized in 8 of the 12 studies.

OPEN IN VIEWER

Table 3.

Study Components and Positive Youth Development Constructs.

Study	Promotes bonding	Fosters resilience	Promotes social competence	Promotes emotional competence	Promotes cognitive competence	Promotes behavioral competence	Promotes moral competence	Fosters self-determination	Fosters spirituality	Fosters self-efficacy	Fosters clear and positive identity	Fosters belief in the future	Provides recognition for positive behavior	Provides opportunities for prosocial involvement	Fosters prosocial norms
Conduct Problems Prevention Research Group (2014)	X	X	X	X	X	X						X
Clemons, Wetta-Hall, Jacobson, Chesser, and Moss (2011)	X		X	X		X	X	X		X		X		X
Dilorio et al. (2002) and Dilorio et al. (2006)	X		X		X	X				X	X	X		X
Dilorio, McCarty, Resnicow, Lehr, and Denzmore (2007)	X		X		X	X				X	X	X		X
East, Kiernan, and Chavez (2003)	X	X	X	X	X	X						X		X
Flay, Graumlich, Segawa, Burns, and Holliday (2004)	X	X	X	X	X	X	X	X		X	X	X		X
O’Donnell et al. (1999) and O’Donnell et al. (2002)	X	X	X	X	X	X				X		X		X
Philliber, Kaye, Herrling, and West (2002)	X		X	X	X	X						X		X
Prado et al. (2007)	X				X	X								X
Sieving et al. (2011) and Sieving et al. (2013)	X		X	X	X	X				X				X	X
Tello, Cervantes, Cordova, and Santos (2010)	X		X	X	X				X	X	X
Trenholm et al. (2007)			X		X	X	X			X		X		X

Note. These components were assigned to PYD constructs by the authors of the systematic review based on the study descriptions in the peer-reviewed manuscript. This is for illustrative purposes only.

Cultural tailoring

Tailored interventions are often recommended in health promotion as a way to improve program components to make them more culturally relevant to special populations, with the goal of increasing recruitment, retention, and health-promoting behavior change (Ewing, Wray, Mead, & Adams, 2012; Harrington & Noar, 2012; Kong, Singh, & Krishnan-Sarin, 2012). Four studies had tailored PYD curricula that emphasized the cultural norms and values of African American (Clemons et al., 2011; Flay et al., 2004) and Hispanic adolescent populations (Prado et al., 2007; Tello et al., 2010).

Recruitment

Recruitment efforts varied based on the intervention setting. School-based interventions recruited students through school orientations, referrals from instructors, other school personnel, and/or parents, and self-report screenings (Clemons et al., 2011; CPPRG, 2014; Flay et al., 2004; Trenholm et al., 2007). One school-based intervention did not recruit participants directly; instead, they randomized the intervention by school and allowed parents to consent (O’Donnell et al., 1999; O’Donnell et al., 2002). The clinic or community-based settings recruited participants primarily through screenings, clinic visits, and community organization membership lists (Dilorio et al., 2007; Dilorio et al., 2006; Dilorio et al., 2002; East et al., 2003; Philliber et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013; Tello et al., 2010).

Sample size

Baseline sample sizes varied from 253 to 1,538 participants. Nine of the 12 studies had sample sizes of more than 500 total participants (Clemons et al., 2011; CPPRG, 2014; Dilorio et al., 2007; Dilorio et al., 2006; Dilorio et al., 2002; East et al., 2003; Flay et al., 2004; O’Donnell et al., 1999; Tello et al., 2010; Trenholm et al., 2007). Two interventions were implemented with paired participants, with either a mother-child pair (Dilorio et al., 2002) or a father-son pair (Dilorio et al., 2007); the total number of pairs for each of the paired interventions was more than 250.

Retention

Retention rates differed greatly, with rates ranging from 29% to 93.3% One study with nonsignificant results had a retention rate of 29% (CPPRG, 2014) and the other study with nonsignificant results had a retention rate of 65% (Trenholm et al., 2007). Alternatively, most interventions with significant findings had attrition rates below 20% (East et al., 2003; Philliber et al., 2002; O’Donnell et al., 1999; O’Donnell et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013), one of which retained 93.3% of the participants at the final assessment (Sieving et al., 2013). One study with significant findings only retained about half of the original participants for the duration of the intervention (Flay et al., 2004). Two of the studies with significant findings that did not explicitly state their retention rates noted that either all participants completed posttest assessments or reported that the participant numbers increased at the posttest assessment (Clemons et al., 2011; Tello et al., 2010). The other two studies with significant findings that did not explicitly state retention rates noted the percent of participants that completed each follow-up assessment or provided approximate retention rates based on the maximum number of withdrawals per site (Dilorio et al., 2007; Dilorio et al., 2006).

Interventions with high retention rates had some common factors. First, two of the five studies with the highest retention rates offered clinical services (Philliber et al., 2002; Sieving et al., 2011; Sieving et al., 2013). Also, there was a focus on reducing attrition noted in three of the five studies, which included efforts such as contacting students several times throughout the year and offering additional services (East et al., 2003; Philliber et al., 2002; Sieving et al., 2011; Sieving et al., 2013). Furthermore, program components were implemented primarily by trained facilitators who developed trusting relationships with the participants and fostered a supportive environment (O’Donnell et al., 1999; O’Donnell et al., 2002; Philliber et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013). Lastly, four of the five studies with the highest retention rates offered monetary incentives to their program participants (East et al., 2003; Philliber et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013).

Dose of program components

The dosage of the intervention components varied widely and was not always related to intervention length or significant program effects. One intervention with significant findings was implemented for 3 years, 5 days a week, 3 hours a session (Philliber et al., 2002). Another was implemented monthly for 9 months, and averaged about 2 hours of programming a month (East et al., 2003). A third was implemented for 18 months (Sieving et al., 2011; Sieving et al., 2013). One study with nonsignificant findings was implemented for 35 sessions in 1st and 2nd grade, had monthly boosters in 3rd through 5th grade, and featured six child-only sessions and four family sessions from 6th through 10th grade (CPPRG, 2014); the other study with nonsignificant findings was implemented for 1 school year, every day of the school year (Trenholm et al., 2007).

Among the studies that documented a higher dose, most studies included an educational component, mostly curricula, that were implemented by trained facilitators over multiple time periods, with the shortest curriculum lasting six sessions (Clemons et al., 2011; CPPRG, 2014; Dilorio et al., 2007; Dilorio et al., 2006; Flay et al., 2004; O’Donnell et al., 1999; O’Donnell et al., 2002; Philliber et al., 2002; Prado et al., 2007; Tello et al., 2010; Trenholm et al., 2007).

Sexual health outcomes

Sexual health outcomes assessed in minority adolescents included initiation of sexual intercourse, ever had sexual intercourse, perceived likelihood of contraceptive use, and actual contraceptive use (see Table 4). Table 4 depicts a breakdown of the sexual health outcomes measured. Seven of the 12 studies did limit their focus to younger adolescents (Dilorio et al., 2007; Dilorio et al., 2006; Flay et al., 2004; O’Donnell et al., 1999; O’Donnell et al., 2002; Philliber et al., 2002; Prado et al., 2007; Trenholm et al., 2007). These studies focusing on younger youth found significant increases in abstinence (Dilorio et al., 2007), refusing unwanted sex (Philliber et al., 2002), contraceptive use (Dilorio et al., 2007; Dilorio et al., 2006; Flay et al., 2004; Philliber et al., 2002; Prado et al., 2007), sexuality-related knowledge (Dilorio et al., 2006; Philliber et al., 2002), communication skills (Dilorio et al., 2007), and decreases in intimate behaviors (Dilorio et al., 2007), recent sex (Flay et al., 2004; O’Donnell et al., 1999; O’Donnell et al., 2002), pregnancy and birth (Philliber et al., 2002), and STDs (Prado et al., 2007). One of the studies that limited its focus to younger adolescents did not find any significant findings in any of the measured categories (Trenholm et al., 2007).

OPEN IN VIEWER

Table 4.

Summary of Outcomes Evaluated.

	Timing of outcome measurement	Initiation of sex/ever had sex/recent sexual behavior	Birth control/protection	Pregnancy	Sexually transmitted disease	Attitudes	Sexuality-related knowledge	Communication skills
Conduct Problems Prevention Research Group (2014)	Every summer between 7th and 11th grades	ns		ns	ns
Clemons, Wetta-Hall, Jacobson, Chesser, and Moss (2011)	Posttest immediately following program (end of the academic year)					sig ↑ intent to delay until marriage
Dilorio et al. (2002) and Dilorio et al. (2006)	Posttest at 4 (post-intervention), 12, and 24 months after baseline	ns	sig ↑				sig ↑
Dilorio, McCarty, Resnicow, Lehr, and Denzmore (2007)	Posttest at 3-, 6-, and 12-month after baseline	At 3 months; nsAt 6 months: Sons sig↑ abstinence, sig↓ intimate behaviors	At 12 months: Sons sig↓ unprotected sex			At 12 months: Sons sig↑ intent to delay until marriage		At 3 months: Fathers sig↑ sons; nsAt 6 months: ns At 12 months; fathers sig ↑
East, Kiernan, and Chavez (2003)	Posttest immediately following program (9 months after enrollment)	Females: sig ↓ sexMales: ns	Females: nsMales: sig ↑	Females: sig ↓Males: ns	ns
Flay, Graumlich, Segawa, Burns, and Holliday (2004)	At the end of each academic year, 5th to 8th grade	Females: nsMales: sig↓ recent sex	Females: nsMales: sig ↑
O’Donnell et al. (1999) and O’Donnell et al. (2002)	6 months after baseline and 2 years post-intervention	At 6 months:CYS: sig↓ recent sex, sig↓ sex behavior indexCurriculum only: nsAt 2 years post:CYS: sig↓ initiation Curriculum only: sig↑ recent sexBoys: sig↑ initiation of and recent sex
Philliber, Kaye, Herrling, and West (2002)	Every summer for 3 years (last assessment 3-years after enrollment)	Females: sig↑ say no to pressured sexMales: ns	Females: sig↑Males: ns	Females: sig ↓Males: ns			sig ↑
Prado et al. (2007)	6, 12 (post-intervention), 24, and 36 months post-baseline		sig↑		sig ↓
Sieving et al. (2011) and Sieving et al. (2013)	12 and 24 months after enrollment (last assessment 6 months after end of program)	At 12 months: nsAt 24 months: ns	At 12 months: sig↑At 24 months: sig↑			At 12: nsAt 24: sig↓ importance of having sex for materialssig ↑ self-efficacy to refuse sex		At 12: nsAt 24: sig ↑
Tello, Cervantes, Cordova, and Santos (2010)	Posttest immediately following curriculum	sig↑ ever had sexsig↓ sex in the last 30 dayssig↓ ever had sex for money	ns			sig ↑ attitudes towards abstinence	sig ↑
Trenholm et al. (2007)	Final assessment 4-6 years after baseline	ns	ns	ns	ns	ns

Note. ns= not significant; sig↑= significant increase; sig↓= significant decrease; CYS = Community Youth Service.

Outcomes by age

Studies that used a wide age range did not report outcomes by age. While all participants may have benefitted equally from exposure to the intervention, it was not possible to confirm this from reported results. Of the seven studies that focused on adolescents younger than 15, the most common significant outcomes were increased contraceptive use (Dilorio et al., 2007; Dilorio et al., 2006; Flay et al., 2004; Philliber et al., 2002; Prado et al., 2007), increased sexuality-related knowledge (Dilorio et al., 2006; Philliber et al., 2002), and decreased recent sexual intercourse (Flay et al., 2007; O’Donnell et al., 1999; O’Donnell et al., 2002).

Outcomes by gender

Nine studies included both girls and boys, while three studies were gender specific. Common significant outcomes after exposure to the PYD interventions for the girls were that they were significantly more likely to remain abstinent (East et al., 2003; Philliber et al., 2002), use a contraceptive method (Philliber et al., 2002; Sieving et al., 2011; Sieving et al., 2013), and less likely to have had sex (East et al., 2003), or become pregnant (East et al., 2003; Philliber et al., 2002). Females also had greater self-efficacy to refuse unwanted sex (Sieving et al., 2013) and reductions in perceived importance of having sex for material reasons (Sieving et al., 2013).

Common significant outcomes from the PYD interventions for the boys were that they were significantly less likely to ever have sex (Dilorio et al., 2007), to have recent sex (Flay et al., 2004; Tello et al., 2010), to have sex for money (Tello et al., 2010), and to use drugs before sex (Tello et al., 2010). Boys were significantly more likely to increase condom use (Dilorio et al., 2007; East et al., 2003; Flay et al., 2004), greater sexuality-related knowledge (Tello et al., 2010), intent to delay having sex (Dilorio et al., 2007), and attitudes toward abstinence (Tello et al., 2010). Three iatrogenic effects emerged for boys in two studies: Boys were more likely to have had sex after one study (Tello et al., 2010) and more likely to have initiated sex and had recent sex after another study (O’Donnell et al., 2002).

Outcomes by cultural tailoring

Four studies (Clemons et al., 2011; Flay et al., 2004; Prado et al., 2007; Tello et al., 2010) reported tailoring their interventions to make them more culturally relevant to minority adolescents. Clemons and colleagues (2011) tailored the Pursue Your Dreams intervention by utilizing a development model that incorporates key African American principles, such as the importance of faith, self-determination, unity, and working together. The Aban Aya project was developed with a focus on the African American community and based the project on two Ghanaian principles: “aban” which represents social protection and “aya” which represents self-determination (Flay et al., 2004). The Familias + PATH intervention was tailored to the Hispanic culture by focusing on strengthening the family unit and placing parents as leaders in the sexuality education of their children, which is consistent with the importance of the family unit in the Hispanic culture (Prado et al., 2007). Tello and colleagues (2010) tailored the Joven Noble curriculum by focusing on key factors that are important to the Hispanic culture (respect, family, interpersonal relationships, and working together) and dispelling misconceptions associated with certain cultural underpinnings, such as the “meaning of machismo.” Across studies that utilized a tailored approach, program outcomes were increased commitment to abstain from sex until marriage (Clemons et al., 2011), increased attitudes toward abstinence (Tello et al., 2010), contraceptive use (Flay et al., 2004; Prado et al., 2007), condom use (Flay et al., 2004; Prado et al., 2007), sexuality-related knowledge (Tello et al., 2010), decreased recent sex (Flay et al., 2004; Tello et al., 2010), use of alcohol before sex (Tello et al., 2010), and sex for money (Tello et al., 2010).

Outcomes by studies promoting abstinence and promoting safer sex practices

Most studies focused on safer sex practices, including abstinence and contraceptive use. Across the studies that promoted safer sex practices, common program outcomes were increased contraceptive use (Dilorio et al., 2007; Dilorio et al., 2006; East et al., 2003; Flay et al., 2004; Philliber et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013), increased refusal of unwanted sex or self-efficacy to refuse (Philliber et al., 2002; Sieving et al., 2013), increased rates of youth remaining abstinent (Dilorio et al., 2007; East et al., 2003), increased sexuality-related knowledge (Dilorio et al., 2006; Philliber et al., 2002), and decreased pregnancies and births (East et al., 2003; Philliber et al., 2002).

Two studies focused solely on promoting abstinence (Clemons et al., 2011; Trenholm et al., 2007). One study resulted in increased commitment to abstaining from sex until marriage (Clemons et al., 2011), and the other did not find any significant outcomes (Trenholm et al., 2007).

Two studies had abstinence or safer sex practices as secondary objectives (CPPRG, 2014; Tello et al., 2010). One study focused on aggressive behavior, which resulted in nonsignificant findings for initiation of sex, pregnancy, and STDs (CPPRG, 2014). Alternatively, the Joven Noble PYD Program focused primarily on character development through cultural values, positive identity, and trust with an emphasis on reducing pregnancies. The program resulted in decreases in number of youth who had sex in the last 30 days, and deceases in those who had sex for money. In addition, the program resulted in increases in sexuality-related knowledge (Tello et al., 2010).

Outcomes based on time of assessment

Studies assessed outcomes immediately following the intervention (Clemons et al., 2011; East et al., 2003; Tello et al., 2010), several years following the intervention (Trenholm et al., 2007), at multiple points during the intervention (CPPRG, 2014; Flay et al., 2004; Philliber et al., 2002), or at multiple points including follow-up assessments after the post-intervention assessment (Dilorio et al., 2007; Dilorio et al., 2006; O’Donnell et al., 1999; O’Donnell et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013). Of the studies that assessed outcomes immediately following the intervention, there was increased intention to remain abstinent (Clemons et al., 2011; East et al., 2003), increased contraceptive use (East et al., 2003), increased sexuality-related knowledge (Tello et al., 2010), decreased sex for money (Tello et al., 2010), and decreases in adolescents who have ever had sex (East et al., 2003; Tello et al., 2010).

The study that assessed outcomes solely several years following the intervention found no significant results (Trenholm et al., 2007). Of the studies that assessed outcomes at multiple points during the intervention, there was increased condom use (Flay et al., 2004; Philliber et al., 2002), increased sexuality-related knowledge (Philliber et al., 2002), increased ability to resist peer pressure to have sex (Philliber et al., 2002), decreases in pregnancies and births (Philliber et al., 2002), decreased ever had sex (Philliber et al., 2002), and one study resulted in no significant findings (CPPRG, 2014).

Of the studies that assessed outcomes at multiple points and followed up after the post-intervention assessments, there was increased contraceptive use (Dilorio et al., 2007; Dilorio et al., 2006; Prado et al., 2007; Sieving et al., 2013), increased self-efficacy to refuse unwanted sex (Sieving et al., 2013), increased abstinence (Dilorio et al., 2007), decrease in intimate behaviors/sexual behavior index (Dilorio et al., 2007; O’Donnell et al., 1999), decreased initiation of sex (O’Donnell et al., 2002), increased knowledge and decreased STDs (Prado et al., 2007), and increased communication skills, decreased importance of having sex, and decreased recent sex (Sieving et al., 2013).

Outcomes by study components

Several of the study researchers stated the importance of multicomponent PYD interventions to influence several behaviors, yet only one study looked at the outcomes by the type of component (East et al., 2003). There were several significant associations: (a) fostering belief in the future (school and job-related services) was associated with increased contraceptive use for females, (b) promoting more time devoted to sexuality education was associated with reduced likelihood of early parenting for all participants, (c) fostering resilience and promoting social and behavioral competence (psychosocial services—stress management, peer pressure, and decision-making skills) was associated with increased contraceptive use with most recent partner, and (d) providing opportunity for prosocial involvement (community and recreational services) was associated with increased male’s intentions to remain abstinent. There was an iatrogenic effect in that more time in sexuality education was related to a decrease in the male’s likelihood of remaining abstinent. Services offered in individual settings were associated with reduced likelihood of early parenting for all participants, while group setting services were associated with more certainty of contraceptive use for all participants, more intentions to remain abstinent, and delayed initiation of sex among males.

The most common characteristics of all of the studies that produced significant results across the varied sexual-health-related outcomes were that the program promoted bonding (e.g., focusing on parent-child relationships and relationships with community members), social competence (e.g., opportunities to develop interpersonal skills), cognitive competence (e.g., opportunities to develop reasonable ways to overcome stressors and solve problems), behavioral competence (e.g., practice expressing oneself through role-plays), and provided opportunities for prosocial involvement (e.g., community service involvement).

Although all studies did not assess the effectiveness of the program components as East and colleagues (2003), there were some common themes among the study components of the studies with significant findings. The six studies that found significant results in the sexual behavior category (decreases in initiation of sex, ever had sex, recent sex, intimate behaviors, saying no to unwanted sex or pressured sex, and sex for money), all studies included program components and activities that promoted bonding, social, cognitive, and behavioral competence; five of the six studies included components that promoted emotional competence and provided opportunities for prosocial involvement (Dilorio et al., 2007; East et al., 2003; Flay et al., 2004; O’Donnell et al., 1999; O’Donnell et al., 2002; Philliber et al., 2002; Tello et al., 2010). Among the seven studies that found significant results in the birth control/protection category (decreases in unprotected sex, and increases in protected sex, condom use, and other birth control use), all studies included components that promoted bonding, cognitive and behavioral competence, and provided opportunities for prosocial involvement (Dilorio et al., 2007; Dilorio et al., 2006; East et al., 2003; Flay et al., 2004; Philliber et al., 2002; Prado et al., 2007; Sieving et al., 2011; Sieving et al., 2013). Among the two studies that found significant results in reducing pregnancy, both included components that promoted bonding, social, emotional, cognitive, and behavioral competence, fostered belief in the future, and provided opportunities for prosocial involvement (East et al., 2003; Philliber et al., 2002). Among the four studies that found significant results in the attitudes category (increased self-efficacy to refuse unwanted sex and intent to delay sex until marriage, decreased importance of having sex for material items), all studies included components that promoted bonding and fostered self-efficacy; three included components that promoted social, emotional, behavioral, and cognitive competence and provided opportunities for prosocial involvement (Clemons et al., 2011; Dilorio et al., 2007; Sieving et al., 2013; Tello et al., 2010). Two studies improved communication skills; both studies included components that promoted bonding, social, cognitive, and behavioral competence; fostered self-efficacy; and provided opportunities for prosocial involvement (Dilorio et al., 2007; Sieving et al., 2013).

Focus on young adolescents

The adolescents that have had sex by age 13 are at a higher risk of pregnancy and STDs. Early initiation of sex has been linked to early initiation of tobacco and substance use relationships with peers who engage in risky behaviors, depression, and school disconnectedness and disruption (Albert et al., 2003). Therefore, a focus on developmental outcomes to reduce the likelihood of initiating sexual activity could have a positive, long-term impact on the adolescents’ lives.

Researchers should consider developing programs for age-specific populations. Obtaining a large-enough sample size to assess age-specific data is important. An intervention designed to be implemented across a large developmental age range could potentially be beneficial for one subset and not for the other subsets of the age range. Four of the 12 interventions implemented the interventions to students across the middle and high school age ranges, but did not assess intervention effects by age ranges. Researchers that implement programming across developmental age ranges should report effect sizes and conduct power analysis to determine whether age-specific outcomes are appropriate.

De-link ethnic minority status from socioeconomic status

It was difficult to determine whether intervention effects for participants were due to SES or other challenges specific to cultural and social experiences youth have as a result of their ethnic minority status. The exploration of these differences is necessary to determine the particular cultural influences each contributes to early sexual initiation and sexual risk behaviors.

Focus on retention

Most of the studies with significant findings retained the majority of the program participants. Retention is particularly important for long-term developmental interventions due to the longitudinal nature of the effects of PYD interventions. For instance, the Fast Track Program attempted to retain students for 12 years, implementing intervention components in elementary, middle, and high school, but was only able to retain 29% of the original program participants, and experienced no significant program effects at the posttest analysis (CPPRG, 2014). It was difficult to assess whether the attrition of program participants or the actual program itself resulted in the nonsignificant findings. Although some PYD interventions found significant findings in short-term follow-ups, it is important to retain program participants and assess program effectiveness at several points post-intervention to gain a true understanding of program effectiveness. Attrition is particularly a concern for mobile, high-risk populations; so retention efforts should be a part of the planning process. PYD objectives are not likely to change sexual health outcomes after implementation of a few lessons. Developmental changes cannot be addressed in one sitting; therefore, it is important to focus on retaining the majority of participants throughout the entire program by including retention methods in the planning of the program, such as maintaining contact with the students by phone or home visits, using activities they can enjoy, vary the activities to keep them entertained, and provide incentives. Free clinical services may be an attractive incentive, especially for students of lower SES who may not have access to quality medical care. In addition, it is important to consider conducting attrition analyses to assess the impact of those that remained in the program compared with those that dropped out. Additional analysis should also be employed if participants are able to join the intervention, similar to the analysis done by the Reach for Health intervention researchers (O’Donnell et al., 1999; O’Donnell et al., 2002). The students who do not participate in the entire program may have responses that vary from the other participants, which should be captured and assessed for the program evaluation.

Focus on evaluation

Evaluation is the key to furthering PYD programming. Evaluating results of PYD interventions several years after initiation to determine whether short-term positive results withstand time would provide much-needed information about the long-term effects of PYD programming. Understanding the relationship between timing and dose for specific intervention components is key to the effectiveness of PYD interventions. A higher dose of intervention components could improve the likelihood of producing significant results. It remains unclear whether implementing interventions at multiple periods in the adolescents’ life span is most effective in reducing risky behaviors (e.g., CPPRG, 2014), or if it may be more important to concentrate intervention implementation at early adolescence (especially middle school aged) and evaluate the outcomes at multiple time periods post-intervention. More research is needed to determine the most beneficial approach. However, it is clear that PYD interventions should be implemented during early adolescence and incorporate long-term evaluations in order to assess long-term intervention effects into adulthood. In addition, developing a compendium of PYD measures and instruments and having a venue in which to provide this information to fellow researchers would strengthen consistent evaluations of PYD interventions.