eFigure 1. Study Flow Diagram
eFigure 2. Forest Plot for Delayed Sexual Activity Outcome
eFigure 3. Forest Plot for Condom Use Outcome
eFigure 4. Forest Plot for Parent-Child Sexual Communication Outcome
Customize your JAMA Network experience by selecting one or more topics from the list below.
Widman L, Evans R, Javidi H, Choukas-Bradley S. Assessment of Parent-Based Interventions for Adolescent Sexual Health: A Systematic Review and Meta-analysis. JAMA Pediatr. 2019;173(9):866–877. doi:10.1001/jamapediatrics.2019.2324
Are parent-based sexual health interventions associated with improved adolescent sexual health outcomes?
This meta-analysis synthesized the results of 31 randomized clinical trials comprising 12 464 adolescent participants. Across studies there was a significant association of parent-based interventions with improved condom use and parent-child sexual communication compared with control conditions, but there was no significant mean association of these interventions with delaying adolescents’ sexual activity.
Overall, parent-based interventions may improve several aspects of adolescents’ sexual health and decision-making.
Parent-based sexual health interventions have received considerable attention as one factor that can increase safer sexual behavior among youth; however, to our knowledge, the evidence linking parent-based interventions to youth sexual behaviors has not been empirically synthesized.
To examine the association of parent-based sexual health interventions with 3 primary youth outcomes—delayed sexual activity, condom use, and parent-child sexual communication—as well as several secondary outcomes. We also explored potential moderators of intervention effectiveness.
A systematic search was conducted of studies published through March 2018 using MEDLINE, PsycINFO, Communication Source, and CINAHL databases and relevant review articles.
Studies were included if they: (1) sampled adolescents (mean age, ≤18 years), (2) included parents in a key intervention component, (3) evaluated program effects with experimental/quasi-experimental designs, (4) included an adolescent-reported behavioral outcome, (5) consisted of a US-based sample, and (6) were published in English.
Data Extraction and Synthesis
Standardized mean difference (d) and 95% confidence intervals were computed from studies and meta-analyzed using random-effects models. A secondary analysis evaluated potential moderating variables.
Main Outcomes and Measures
The primary outcomes were delayed sexual activity, condom use, and sexual communication.
Independent findings from 31 articles reporting on 12 464 adolescents (mean age = 12.3 years) were synthesized. Across studies, there was a significant association of parent-based interventions with improved condom use (d = 0.32; 95% CI, 0.13-0.51; P = .001) and parent-child sexual communication (d = 0.27; 95% CI, 0.19-0.35; P = .001). No significant differences between parent-based interventions and control programs were found for delaying sexual activity (d = −0.06; 95% CI, −0.14 to 0.02; P = .16). The associations for condom use were heterogeneous. Moderation analyses revealed larger associations for interventions that focused on younger, compared with older, adolescents; targeted black or Hispanic youth compared with mixed race/ethnicity samples; targeted parents and teens equally compared with emphasizing parents only; and included a program dose of 10 hours or more compared with a lower dose.
Conclusions and Relevance
Parent-based sexual health programs can promote safer sex behavior and cognitions in adolescents, although the findings in this analysis were generally modest. Moderation analyses indicated several areas where future programs could place additional attention to improve potential effectiveness.
Adolescents disproportionately experience the negative consequences of sexual risk behavior. Young people aged 15 to 24 years constitute half of the 20 million annual cases of sexually transmitted infections (STIs) in the United States even though they make up only 25% of the sexually active population.1 As many as 1 in 4 sexually active girls has an STI2; recent data suggest that 17% of new HIV infections occur among young boys and men younger than 24 years.3 Youth in the United States are also at heightened risk for unintended pregnancy.4,5
To address these risks, for nearly 3 decades researchers and public health professionals have included parents in sexual health education programs with the aim of improving adolescent sexual decision-making. Targeting parents with sexual health programming makes sense for theoretical and practical reasons. According to the ecological systems theory,6 children develop within a series of nested systems, including the family system, that can directly and indirectly affect decision-making and health behavior. Parents shape the attitudes, norms, values, and sexual decision-making capacity of their children and thus act as particularly important socializing agents.7-10 Parents are also uniquely positioned to monitor adolescents’ behavior; higher parental monitoring has been linked to a higher likelihood of contraceptive use and delayed sexual activity.9,11,12 Additionally, parent-based interventions may allow parents to tailor the content and delivery of information based on their child’s developmental level and interpersonal context.13,14
While there are many reasons to expect that parent-based sexual health interventions may affect adolescents’ behaviors, the empirical evidence is somewhat mixed on the success of these programs.15,16 Several studies have shown that parent-based interventions increase (or are associated with an increase in) the likelihood that adolescents will delay sexual activity and use condoms or other contraceptives more consistently when sexually active.17-19 However, other studies of parent-based interventions have not shown significant findings with regard to adolescent sexual decision-making or behavior.20-22
To our knowledge, the literature currently lacks a meta-analytic review that synthesizes the findings of studies of parent-based interventions on adolescent sexual behavior. A few systematic reviews and meta-analyses have examined the association of cross-cutting parent-based interventions that target multiple adolescent risk outcomes (eg, substance use, sexual risk, and violence)23 and others that target specific components of parent-based interventions, such as those focused exclusively on improving parent-child sexual communication.14,24,25 However, we are not aware of a comprehensive meta-analysis that exclusively focuses on the overall association of parent-based interventions with adolescent safer sex behaviors, including delayed sexual activity and condom use. Such a review could assess the pooled findings of these interventions and identify specific participant characteristics and/or program components associated with program success.
The purpose of this study is to systematically review the literature on parent-based sexual health interventions for US adolescents and meta-analyze their overall association with 3 key behavioral outcomes: delayed sexual activity, condom use, and parent-child sexual communication. These outcomes were selected because of their relevance to preventing unintended pregnancy, HIV, and other STIs.1 In addition, we examined 3 secondary outcomes identified as important components of safer sexual decision-making within health behavior theories,26,27 including intentions to delay sex, sexual health knowledge, and safer sex self-efficacy.
A secondary goal was to identify specific components that are likely to make parent-based interventions successful at decreasing adolescent sexual risk behavior. We considered several potential moderating variables, including sample (eg, sex, age, and race/ethnicity), intervention (eg, design and dose), and methodological characteristics (eg, length of follow-up).
We conducted a comprehensive search of MEDLINE, PsycINFO, Communication Source, and CINAHL databases to extract relevant studies published through March 2018. We used the following combination of key words: adolescent* or teen* or youth; sexual health or safe* sex or sex* education or sexually transmitted disease or sexually transmitted infection or STD or STI or HIV or AIDS or pregnancy or reproductive health or condom* or contracept* or protected sex or unprotected sex or abstinence; intervention or program or education or prevention or promotion or trial; and parent* or caregiver or family or mom or mother or dad or father. Additional studies were located by examining prior reviews and meta-analyses14-16,24,25,28-32 and examining the reference lists of all included articles. This search produced an initial 5268 articles.
Studies were included if they met the following criteria: (1) focused on adolescents (ie, mean sample age, ≤18 years); (2) included parents in a key intervention component (school-based programs with a minor parent component, such as a worksheet, were excluded33-35); (3) evaluated program effects with an experimental or quasi-experimental design; (4) included at least 1 of 3 behavioral outcome measures reported by youth: delayed sexual activity/abstinence, condom use (effect sizes for unprotected sex were recoded so that the direction of the effect always indicated greater protection), or parent-child sexual communication; (5) included a US-based sample; (6) were published in English; and (7) provided sufficient statistics to calculate effect sizes. For studies with multiple follow-up points, we used the longest-term follow-up with adequate data to calculate effect sizes as the most conservative estimate of treatment effects. For studies with more than 1 intervention group, we selected the parent-based intervention that was most comprehensive. When studies included more than 1 indicator for an outcome (eg, multiple indicators of condom use), we used a random number generator to select 1 outcome. Finally, studies that included only parent-reported outcomes were excluded.36 These selection criteria resulted in a final sample of 31 articles (eFigure 1 in the Supplement). From these articles, we calculated 20 independent effect sizes for delayed sexual activity, 16 for condom use, 14 for sexual communication, 6 for intentions to delay sex, 5 for sexual health knowledge, and 4 for safer sex self-efficacy.
Two of the authors (R.E. and H.J.) independently coded the primary studies. The following data were abstracted: (1) demographic and sample characteristics, (2) intervention characteristics (eg, dose intended, defined as the amount of time parents were intended to be involved in the intervention, and dose received, defined as the percentage of parents who completed the full dose intended), and (3) methodological characteristics (eg, length of follow-up). The mean percentage agreement across all coding categories was 92%. Discrepancies between coders were resolved through discussion with the first author (L.W.).
The standardized mean difference (d) was used as the indicator of effect size. Effect size d can be interpreted as small (0.20), medium (0.50), or large (0.80).37 When d’s were reported in an article, they were directly extracted. If d’s were not reported, other statistics that could be converted to d’s (eg, summary statistics and odds ratios) were calculated using Comprehensive Meta-Analysis, version 2.0 (Biostat)38 and the Practical Meta-Analysis Effect Size Calculator.39 When no statistics in the study could be converted to a d, the study authors were contacted and appropriate data were requested. To ensure the consistency and interpretability of effect sizes for all studies, higher values always indicated that the parent-based intervention group performed better than the control group.
We used random-effects meta-analytic procedures for the primary analyses across all independent effect sizes; this procedure allowed for the possibility of differing variances across studies.40 The Q statistic and I2 were used to examine whether significant heterogeneity existed among effect sizes. Effect sizes for hypothesized moderators were calculated along with their 95% confidence intervals, and those effect sizes were compared using the Qb statistic. For these analyses, mixed-effects models were used to allow for the possibility of differing variances across subgroups. These models use random-effects assumptions while stratifying the effect sizes by fixed factors, such as age and study dose.40 Analyses were conducted using Comprehensive Meta-Analysis, version 2.0, and statistical significance was set at P < .05.38
Table 117-21,41-66 provides a summary of the studies included in this meta-analysis. A total of 12 464 adolescents (weighted mean age, 12.3 years) were enrolled across 31 parent-based interventions. Most programs targeted boys and girls (29); there was 1 program for only girls18 and 1 for only boys.19 The racial/ethnic makeup of participants across studies was diverse, with some focused exclusively or primarily (more than 85%) on black (7), Hispanic (7), and white youth (2), and about half were of mixed race/ethnicity (15).
Regarding the parents in these studies, descriptive characteristics were missing from many studies. From what we could assess, most programs focused exclusively or primarily (more than 85%) on mother/maternal figures (13) or mixed-sex parent samples (7), with only 1 study focused on father/paternal figures and 10 studies that did not report parent sex. The mean parent age among the 11 studies that reported age was 39.8 years.
The intervention dose varied widely across studies, from fewer than 3 hours of parent training (6) to 20 hours or longer (6). For most studies, the dose intended did not match the final dose received: only 10 studies actually delivered the full intervention dose to all parents. In addition, most of the programs were delivered to parents in person (28), with only 2 programs delivered online.
Individual study effect sizes for delayed sexual activity ranged from a d of −0.75 to 0.36, with an overall weighted mean effect size across studies of a d of −0.06 (95% CI, −0.14 to 0.02; P = .16) that was not significant. This indicates that, overall, parent-based interventions were not associated with a delay in adolescents’ sexual activity (Table 2; eFigure 2 in the Supplement). Further, there was no significant heterogeneity among studies regarding the delayed sexual activity outcome (Q = 26.59; df = 19; P = .12; I2 = 28.54); thus, we did not examine moderators of this outcome.
Whereas individual effect sizes of studies ranged from a d of −0.18 (95% CI, −0.49 to 0.13) to 1.28 (95% CI, −0.09 to 2.65), the weighted mean effect size for condom use was a d of 0.32 (95% CI, 0.13-0.51; P = .001), indicating that parent-based interventions have a small but significant association with youth condom use behavior (Table 2; eFigure 3 in the Supplement). To examine whether publication bias may have inflated the effect size, a fail-safe number was calculated. This fail-safe number was 139, suggesting that 139 nonsignificant studies would need to exist to reduce the effect size to a trivial level.
Statistical testing indicated heterogeneity among studies regarding the condom use outcome (Q = 55.38; df = 15; P < .001; I2 = 72.91); thus, we examined several potential moderating variables. As shown in Table 3, adolescent age, sex, and race/ethnicity moderated the association of interventions with condom use as well as intervention design and the extent of parental involvement. Intervention associations were stronger for samples with adolescents younger than 14 years (d = 0.64; P < .001) compared with older samples (d = 0.15; P = .10). Intervention associations were stronger for samples with black teens (d = 0.29; P = .07) and for samples with Hispanic teens (d = 0.54; P < .001) compared with interventions with mixed race/ethnicity samples (d = 0.09; P = .51). Intervention associations were stronger for programs that involved parents and teens equally (d = 0.58; P < .001) than for programs that were parent-only (d = 0.09; P = .74), mostly parent (d = 0.07; P = .50), or mostly teen (d = 0.12; P = .39). Lastly, intervention associations were stronger for programs including 10 to 19 hours (d = 0.41; P = .02) or 20 hours or more of parent training (d = 0.44; P < .001) compared with shorter programs (ie, 3-9 hours of parent training; d = −0.03; P = .73). No significant differences were found by parent sex, intervention dose completion, or follow-up duration. All 16 studies that included condom use as a primary outcome were delivered in-person; thus, we could not examine the associations of in-person vs online programs with condom use. Also, because there was only 1 intervention specifically for adolescent boys and none specifically for adolescent girls that examined condom use as an outcome, the sex moderation finding should be interpreted with caution.
Individual study effect sizes for parent-child sexual communication ranged from a d of −0.03 (95% CI, −0.27 to 0.21) to 1.03 (95% CI, 0.38-1.67), with an overall weighted mean effect size of a d of 0.27 (95% CI, 0.19-0.35; P = .001). This indicates that parent-based interventions were significantly associated with increased parent-child communication compared with control programs (Table 2; eFigure 4 in the Supplement). The fail-safe number was 257, suggesting that 257 nonsignificant studies would need to exist to reduce the effect size to a trivial level. There was no statistically significant heterogeneity among the studies in the communication outcome (Q = 21.02; df = 13; P = .07; I2 = 38.13), so we did not consider moderators for this outcome.
The results for all secondary outcomes are shown in Table 2. Parent-based interventions were associated with all 3 secondary outcomes—intentions to delay sexual activity (d = 0.24; 95% CI, 0.10-0.38; P = .001), sexual health knowledge (d = 0.40; 95% CI, 0.06-0.75; P = .02), and safer sex self-efficacy (d = 0.26; 95% CI, 0.16-0.36; P < .001)—with small to medium-sized effects.
This meta-analysis synthesized nearly 3 decades of research on the development and evaluation of parent-adolescent sexual health interventions. These results highlight the potential of these programs to improve adolescents’ safer sex behaviors. Pooling data from 31 studies with 12 464 adolescent participants, we found a significant positive association of parent-based sexual health interventions with condom use and parent-child communication about sexuality. There was no significant association with delayed sexual activity. Parent-based interventions were also associated with increases in 3 secondary outcomes: intentions to delay sexual activity, sexual health knowledge, and safer sex self-efficacy. These findings suggest that parent-based interventions have an overall protective association with several aspects of adolescents’ sexual health and are not associated with sexual activity at earlier ages.
There was significant variability in the findings regarding parent-based interventions and adolescents’ condom use. Thus, we examined several potential moderator variables. These revealed that adolescents’ age and race/ethnicity, as well as the intervention design and extent of parental involvement, were all important factors associated with condom use. Regarding age, intervention associations were stronger among studies with younger compared with older adolescents. Early adolescence is a key period for developing sexual attitudes and norms67,68 that go on to shape adolescents’ sexual behavior.69 Researchers and sex educators should consider these developmental factors, as an earlier delivery of parent-based interventions could increase their potential for success.
Regarding race/ethnicity, intervention associations were stronger for samples with predominantly black or Hispanic individuals compared with those of mixed race/ethnicity. These findings align with other work that demonstrates the power of tailoring interventions to increase relevance, adherence, and cultural appropriateness.70-72 There are marked health disparities in the rates of HIV/STIs and unintended pregnancy among minority youth; thus, interventions that target these groups are high-priority.73,74 It is encouraging that parent-based interventions show success when specifically focusing on black and Hispanic families. Future studies could examine whether it is the cultural tailoring or some other aspect of programs that make them most effective at improving condom use (for example, by changing communication patterns, parental monitoring, or other aspects of family functioning and youth development).25,75
Findings from moderation analyses also demonstrate the potential importance of the intervention dosage and level of parent and teen involvement. Specifically, interventions with a dose of 10 hours or more were associated with stronger improvements in adolescent condom use than shorter programs. Associations were also stronger when programs included parents and teens equally as opposed to targeting parents independently or holding mostly parent-only sessions.25 Increased time and contact between parents and adolescents may be necessary to produce the kind of family-level change needed to improve adolescent sexual decision-making.
While we have highlighted several of the key significant findings, we did not find a significant association across studies between parent-based interventions and delaying sexual activity among youth. There are powerful, normative increases in sexual activity over the course of adolescence76,77 and factors more proximal than parents may influence sexual initiation. For example, biological factors (eg, pubertal hormones) are strongly associated with adolescents’ sexual initiation,78 and peer-related factors (eg, peer norms and popularity) also have strong associations, especially among girls.79 This meta-analysis demonstrates that interventions can help parents communicate sexual knowledge and shape some of adolescents’ sexual choices, such as using condoms, once they become sexually active, but abstinence may be more challenging to address. However, while these programs were not associated with delayed sexual activity, they also were not associated with an earlier initiation of sexual activity. This should be reassuring for parents who are concerned that talking about sex with their children may somehow result in their children initiating sex. This meta-analysis shows that across the dozens of interventions for parents, youth were no more or less likely to initiate sex at the conclusion of the interventions.
In conducting this meta-analysis, several issues associated with intervention design and execution of the included studies became apparent, providing important directions for future research. First, there was substantial variation in the way that the outcome measures were defined across studies. For example, for abstinence and condom use, some investigators focused exclusively on acts of vaginal sex,17,57 whereas others included acts of vaginal, anal, and/or oral sex18,42,45,56 or referred only to “sex” or “intercourse.”19,52 For sexual communication, some studies focused on communication frequency,47,53,63 whereas others assessed communication quality.44,50 The timeline for measuring behaviors also differed. For example, some studies focused on whether youth had ever initiated sexual activity, used condoms, or communicated about sex,17,19,57 whereas others focused on a specified timeframe for these behaviors.42,44,52,56 It is clear that the sexual health intervention literature lacks a criterion standard for measuring sexual risk and protective behavior and associated attitudes.80-82 The measurement variation across studies could be obscuring the ability to detect the most precise estimates of intervention efficacy.
Second, most parent-based interventions in this meta-analysis were delivered in person. Only 2 studies (6.5%) used newer technology-based strategies,47,66 and neither of these programs measured condom use as an outcome. Recent systematic reviews and meta-analyses of eHealth and mobile health interventions for youth show that online programs are associated with increased condom use, delayed sexual activity, and an increased uptake of HIV/STI testing.82-84 Programs delivered to parents and youth online could benefit from the known advantages of technology-based programs, including an increased reach, customization, and fidelity to treatment delivery.85-87 There are several of these new programs in development.88-90
A third methodological issue involves the parents included in each study. It was notable how little information was provided about parents in some study descriptions. Demographic information on parent sex was missing in 10 of the 31 studies. Of the studies that reported sex, participation by mothers substantially outpaced participation by fathers. Only 1 study focused specifically on fathers19 and none of the studies with mixed-sex samples included more fathers than mothers. Although parent sex was not a significant moderator among the programs that included this information, it is possible that these findings may have changed had there been more studies on fathers. Fathers can play a key role in contributing to adolescents’ sex education and influencing the sexual behavior of their children,91 although they often feel unprepared to have conversations about sex.92 Investigating sex-specific interventions remains an important avenue for future research, given that fathers may need additional training in sexual communication skills14 and that parent and adolescent sex can affect communication about sexual health.28,32,93
A fourth issue associated with intervention delivery and execution is that the full dose intended did not match the final dose received in many studies, which may have reduced the overall effects of these programs. For example, some studies included extensive parent components, but few families actually completed all activities.21,33,41 In fact, only in 10 studies did all parents complete the full intervention. Engagement and retention, especially of the most at-risk families and youth, remains an ongoing challenge for health interventionists and health care professionals.94
An additional concern is that none of the interventions we identified specifically addressed the needs of sexual minority youth and most studies did not report the sexual identity of participants. Sexual minority youth experience disproportionate rates of sexual risk95 and unique psychosocial challenges associated with sexual development.79 Few sexual minority youth receive comprehensive, inclusive sexual education in schools that addresses these unique concerns.96,97 Thus, parents of sexual minority adolescents could play a critical role in providing sexual health information.98 Parent-based interventions focused specifically on sexual minority youth remain a critically important area for future research.
A final issue worth considering is that we focused this meta-analysis exclusively on randomized clinical trials because they provide the most rigorous test of the efficacy of parent-based interventions. However, randomized clinical trials also likely provide a conservative test; a pooled analysis of quasirandomized and pre-and-post studies may produce stronger associations than an analysis limited only to randomized clinical trials. As the field moves more toward implementation science and additional studies are added to the literature that move beyond efficacy and into effectiveness in real-world settings, a new review should be conducted.
This meta-analysis addresses an important gap in knowledge about parent-based sexual health interventions and adolescent sexual health. Results pooled from 31 studies and 12 464 adolescents indicate a small but significant positive association of parent-based interventions with improved parent-child sexual communication as well as adolescents’ reports of condom use, safer sex intentions, sexual health knowledge, and sexual self-efficacy. However, there was no association between the intervention and delayed sexual activity. Taken together, these results indicate that parent-based programs can promote safer sex behavior; however, the findings were generally modest, and moderation analyses indicate a few areas where future programs could place additional attention to boost program effectiveness (eg, focus on younger adolescents, increase cultural tailoring, and ensure a sufficient dose). The findings of this meta-analysis highlight areas for future research, including increasing the focus on online programs, conducting additional father-based interventions, attending to the dose received as well as the dose intended, and focusing attention on sexual minority youth. Such changes may increase the efficacy of sexual health programs and demonstrate their potential for the long-term health of youth.
Accepted for Publication: April 19, 2019.
Corresponding Author: Laura Widman, PhD, Department of Psychology, North Carolina State University, 640 Poe Hall, #7650, Raleigh, NC 27695 (firstname.lastname@example.org).
Published Online: July 29, 2019. doi:10.1001/jamapediatrics.2019.2324
Author Contributions: Dr Widman had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Widman, Evans, Choukas-Bradley.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Widman, Evans, Choukas-Bradley.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Widman, Evans, Javidi.
Obtained funding: Widman.
Administrative, technical, or material support: Evans.
Conflict of Interest Disclosures: Dr Widman and Mses Evans and Javidi reported grants from the Office of Population Affairs for the US Department of Health and Human Services during the conduct of the study. No other disclosures were reported.
Funding/Support: This research was funded by the Office of Population Affairs of the US Department of Health and Human Services under contract HHSP23320095626WC task order HHSP23337007T. Support was also provided by the North Carolina State University Center for Family and Community Engagement.
Role of the Funder/Sponsor: This funding provided support for the design and conduct of the study, and the collection, management, and analysis of the data but did not have a role in the preparation, review, or approval of the manuscript and decision to submit the manuscript for publication.
Disclaimer: The findings and conclusions in this study are those of the authors and do not necessarily represent the views of the Office of Population Affairs or the US Department of Health and Human Services.
Create a personal account or sign in to: