Introduction

By the time American adolescents graduate from high school, more than 60% have had sexual intercourse (Abma et al. 2010). Although recent trends in adolescents’ sexual health are generally positive (e.g., later age at first sex, greater contraceptive use), sexual activity among adolescents continues to raise important health concerns. For example, more than 15% of teens have had sexual intercourse by the age of 15 (Finer and Philbin 2013), and somewhere between 30 and 40% of teens use condoms inconsistently or never during sexual activity (National Center for Health Statistics 2010), placing them at substantial risk for sexually transmitted infections and unintended pregnancies.

Parent–adolescent sexual communication is increasingly regarded by scholars as a means for reducing sexual risk behavior among adolescents. A recent meta-analysis showed that, overall, adolescents whose parents talk to them about sexuality generally exhibit safer sexual behaviors (e.g., more consistent condom use; Widman et al. 2015). Despite these overall trends, research and theory is unclear regarding the specific pathways (i.e., the mechanisms of change) that explain how parent–adolescent sexual communication can influence adolescents’ sexual behaviors. Studies report predominantly on direct relations between parent–adolescent sexual communication and adolescents’ sexual intentions and/or behaviors, leaving its explanatory pathways largely unexplored. An understanding of these pathways is critical for both research and practice that aims to promote the sexual health of adolescents.

Two goals guide the current study. The first goal is to establish a conceptual model that explains the processes by which parent–adolescent sexual communication influences adolescents’ sexual intentions and behaviors. To this end, a theoretical extension of the Integrative Model of Behavioral Change (Fishbein and Ajzen 2010) is proposed. The second goal is to systematically review empirical evidence for the conceptual model, identifying emergent themes, gaps, and persisting methodological barriers. The organization of this review maps onto these goals. First, the conceptual model is proposed. Then, the procedures and results of the systematic review are described in detail. It ends with a discussion of findings and their implications for research and practice related to parent–adolescent sexual communication.

Conceptual Model of How Parent–Adolescent Sexual Communication is Associated with Adolescents’ Sexual Intentions and Behaviors

The conceptual model describes the processes by which parent–adolescent sexual communication might influence adolescents’ sexual intentions and/or behaviors. This model consists of two main components, each borrowed from an existing field of theory and research. The first component describes parent–adolescent sexual communication as influencing adolescents’ sexual cognitions, which in turn influence their sexual intentions and behaviors. The second component describes the characteristics of the communication itself that make it more or less persuasive over adolescents’ sexual cognitions.

Sexual Intentions and Behaviors as a Function of Sexual Cognitions

The first component of the conceptual model draws upon The Integrative Model of Behavioral Change (Fishbein and Ajzen 2010) to explain an adolescent’s sexual intentions and behaviors as a function of his/her sexual cognitions. Derived from the Theory of Planned Behavior (TPB; Ajzen 1991) and the Theory of Reasoned Action (Ajzen and Fishbein 1980), the Integrative Model’s central premise is that cognitions regarding the self-performance of a specific behavior are the determinants of that behavior. According to this model, three fundamental social cognitions influence health-related intentions and behaviors. Attitudes refer to an individual’s evaluation, positive or negative, of performing a particular behavior for the self. Perceived norms consist of both descriptive norms—the perceived prevalence and normativity of the behavior—and injunctive norms—the perceived approval or disapproval from others regarding performance of the behavior. Self-efficacy refers to one’s perceived ability to perform or refrain from a behavior (sometimes measured as perceived behavioral control). These three cognitions work to influence one’s intention to enact the behavior, which in turn influences his/her performance of the behavior.

Reviews and meta-analyses verify the predictive power of attitudes, perceived norms, and self-efficacy on sexual intentions and behaviors. Albarracín et al. (2001) conducted a meta-analyses of nearly 100 separate data sets and found that condom use intentions were predicted by attitudes toward condoms (r = .58), perceived norms of condom use (r = .39), and perceived behavioral control over the use of condoms (r = .45). Sheeran and Taylor’s (1999) meta-analysis also showed medium to strong average correlations among condom use attitudes, norms, behavioral control, and actual condom intentions and condom use. These sexual cognitions are significantly associated with sexual behaviors among adolescent populations as well (van de Bongardt et al. 2015; Buhi and Goodson 2007).

Because these cognitions have proven successful in predicting sexual intentions and behaviors, they find straightforward theoretical application to the present question of how parent–adolescent sexual communication can influence adolescents’ sexual intentions and behaviors (see Fig. 1 for conceptual model). Specifically, parent–adolescent sexual communication may influence change in adolescents’ attitudes toward engaging in sexual behavior, their perceived norms regarding sexual behaviors, and/or their self-efficacy toward sexual and safe sexual behaviors. In turn, these social cognitions can influence adolescents’ sexual intentions, which then influence adolescents’ sexual behaviors. The review that follows gathers evidence for attitudes, perceived norms, and self-efficacy as pathways through which parent–adolescent sexual communication influences adolescents’ sexual intentions and/or behaviors. Note that, in addition to behaviors, intentions are considered an outcome of interest in the review, their mediating influence to behaviors being assumed given their strong predictive relation therewith.

Fig. 1
figure 1

Conceptual model of pathways that explain associations between parent–adolescent sexual communication and adolescents’ sexual intentions/behaviors. Note This is the full conceptual model proposed in the study. However, the review focused only on mediated paths from parent–adolescent sexual communication to adolescents’ intentions and/or behaviors via their sexual cognitions (attitudes, perceived norms, and efficacy), and direct paths from parent–adolescent sexual communication to adolescents’ sexual cognitions

Full size image

Finally, it should be noted that the Integrative Model also considers the importance of external variables (i.e., those outside the individual’s control) that can modify the processes previously described. These include environmental constraints, knowledge, and tendencies/habit. For example, a couple’s intention to use a condom during their next sexual experience may not translate to behavior if they lack access to condoms and the knowledge of how condoms work. These external variables are acknowledged theoretically as an important determinant of adolescents’ sexual behaviors, though they are not examined in the review. Instead, the review focuses specifically on the aforementioned mediated pathways through social cognitions (attitudes, perceived norms, and self-efficacy) to sexual intentions and/or behaviors.

Communication Characteristics that Modify the Effectiveness of Parent–Adolescent Sexual Communication

Literature on persuasion and attitudinal change suggests that the features of and surrounding a given communication (e.g., its source, its recipient, its message, its context) determine its effectiveness (Berlo 1960). Drawing upon this idea, Jaccard et al. (2002) called for greater scholarly attention to the various features of parents’ sexual communication that determine its effectiveness in changing adolescents’ sexual cognitions, and thereby their sexual intentions and behaviors. For example, parents may be the source of sexual communication, but parents vary by gender, sexual values, expertise on sexual topics, parenting styles, and trustworthiness, to name only a few of these characteristics. Similarly, adolescents may be the recipients of sexual communication, but they also vary in their gender, development, personality traits, and peer and familial contexts. Many such features are likely to produce meaningful variability in how effective parents’ sexual communication actually is. The review cannot address these in their fullness, but uses extant literature to provide a starting point.

Of growing emphasis in the literature on parent–adolescent sexual communication are the characteristics of the sexual communication itself. Initially, researchers focused primarily on the overall extent of this communication (i.e., the depth, breadth and frequency of sexual conversations). Studies now also consider its content (i.e., the actual messages conveyed, ranging from reproductive physiology to safe sexual behaviors) and its quality (i.e., the interpersonal dynamics of the conversation, often conceptualized as the level of comfort and openness; Lefkowitz 2002). These characteristics may produce meaningful variability in the effectiveness of parent–adolescent sexual communication to influence adolescents’ sexual cognitions, and thereby their intentions and behaviors. For example, abstinence messages might successfully convey parental disapproval of sexual behavior (a perceived norm), but will likely fail in teaching safe sex communication skills with a partner (self-efficacy). Further, the level of respect, comfort, and openness (i.e., quality) during these conversations may influence how receptive adolescents are to parents’ sexual messages (Rogers et al. 2015). As of yet, no studies have synthesized these particular characteristics and their associations with adolescents’ sexual outcomes in a systematic manner. Thus, the review also gives specific attention to how the overall extent, content, and quality of parent–adolescent sexual communication differentially relate to adolescents’ sexual cognitions (i.e., attitudes, perceived norms, and self-efficacy).

The Current Study

Figure 1 joins the various characteristics of parents’ sexual communication and the Integrative Model of Behavioral Change (Fishbein and Ajzen 2010) to explain how parent–adolescent sexual communication may influence adolescents’ sexual intentions and behaviors. Specifically, when parents engage their adolescent children in discussions about sex, the characteristics of that communication (e.g., the extent, the content, and the quality) work uniquely and/or interactively to influence adolescents’ attitudes, perceived norms, and self-efficacy regarding the behavior under discussion. These cognitions, in turn, influence adolescents’ intentions to perform the target behavior, which, when accompanied by a lack of environmental constrains, can translate into adolescents’ enactment of the target behavior.

The current review will systematically locate and summarize empirical evidence for the proposed conceptual model. Primarily, the review will summarize evidence for the influence of parent–adolescent sexual communication on sexual intentions and/or behaviors via attitudes, perceived norms, and self-efficacy. To do so, it will document direct relations between parent–adolescent sexual communication and adolescents’ sexual attitudes, perceived norms, and/or self-efficacy; and indirect relations in which parent–adolescent sexual communication is associated with adolescents’ sexual intentions and/or behaviors via these sexual cognitions. Direct relations between parent–adolescent sexual communication and sexual intentions or behavior are beyond the scope of the review (see Widman et al. 2015, for a more thorough treatment). Then, it will summarize how the communication characteristics of parent–adolescent sexual communication (e.g., its extent, content, and quality) differentially relate to adolescents’ sexual cognitions. The methodological rigor of the studies is also considered.

Method

Selection Criteria

Articles had to meet five criteria to be included in the review. First, studies had to model parent–adolescent sexual communication as an independent variable predicting at least one of the three focal social-cognitive mechanisms. Thus, studies were included if they modeled indirect or mediated pathways between parent–adolescent sexual communication and sexual intentions and/or behaviors via sexual attitudes, perceived norms, and/or self-efficacy. Also included were studies that modeled relations between parent–adolescent sexual communication and one of the social cognitive mechanisms (representing the first part of a mediated process). Studies that examined direct associations between parent–adolescent sexual communication and intentions/behaviors were excluded, as such were beyond the scope of the review.

Second, and related to the first criterion, studies had to operationalize sexual cognitions (attitudes, perceived norms, self-efficacy) in a manner consistent with the Integrative Model’s conceptualization of these constructs. Under this model, a sexual cognition is specific to the performance of the behavior, meaning that more generalized attitudes about sexual topics do not necessarily apply. For example, attitudes about broader social issues, such as gendered norms of sexual expression, are not directly in line with this conceptualization, whereas an attitude reflecting a judgment on whether or not to have sex with a partner is. Careful attention was given such that the sexual cognitions measured were specific to the performance of a behavior.

Third, studies had to report on participants between ages 10–24. This introduces emerging adult samples into the reviewed studies. This decision was made because many studies on parent–adolescent sexual communication rely on emerging adult samples, which are easier to access for the study of sensitive topics such as this. As the key objective of this review was to gather evidence for a new conceptual model, these studies were deemed potentially valuable for the current purposes.

Fourth, studies had to report data from independent samples. No studies among the final selection of studies reporting on the same constructs did so using statistically dependent samples. Fifth, articles had to be published in peer-reviewed journals. Studies that did not meet this criterion, such as dissertations and theses, were omitted. No exclusion criteria were specified for the year of the publication or the country in which the study was conducted.

Retrieval of Studies

Studies were retrieved in May and June of 2016. Retrieval began by searching three online databases, PsycInfo, PubMed (MedLine) and Web of Science, with the following search string: ab(parent* or mother* or father* or family) and ab(adolescen* or youth or teen*) and ab(communicat* or convers* or discuss* or talk*) and ab(sex or sexual* or condom* or contracept* or “birth control”). This search strategy yielded 2325 unique articles published in English and in refereed journals. Titles and abstracts were read and articles that were not substantively relevant, or which clearly did not meet the selection criteria were discarded. This process yielded 352 substantively relevant studies. To determine which of these met all selection criteria, abstracts and, if necessary, full texts were read. This process yielded 36 total studies. Then, these studies were reviewed for references to studies that were not located by the online database searches (i.e., the ancestry method). This method yielded an additional 2 studies. In total, 38 studies were located and comprise the present review (see Fig. 2). Given that the author alone conducted the search, this retrieval process was conducted twice in full to ensure its accuracy.

Fig. 2
figure 2

Flow diagram for selection of studies

Full size image

Data Abstraction

Results of statistical tests representing the focal mediated pathways were indexed. When both multivariate and univariate results were reported, multivariate results were given preference under the assumption that they control for the influence of potentially confounding variables (e.g., family climate variables). These results were then considered on the basis of statistical significance and direction of effects to provide evidence for the conceptual model.

Then, characteristics of parents’ sexual communication were indexed, with particular attention to the overall extent, the content, and the quality of parent–adolescent sexual communication. Studies were indexed as assessing overall extent when the frequency or extent of parents’ communication of sexual topics (e.g., reproductive physiology, condom use) was aggregated for an overall score. Some studies assessed parents’ communication on a specific topic (e.g., safe sex practices, sexual values); these studies were indexed as assessing content. Finally, studies that examined the degree of openness, comfort, and respect during sexual conversations were indexed as assessing quality.

Finally, key methodological characteristics were indexed and reported in the results to indicate the overall methodological quality of the reviewed studies. These characteristics were sample size, racial and/or ethnic breakdown of sample, sampling strategy (e.g., random vs. non-random; national vs. non-national), study design (e.g., cross sectional vs. longitudinal; survey, observational, etc.), analytic strategies (e.g., SEM, logistic regression, ANOVA), and statistical adjustment for family climate variables (e.g., general family communication). Because only the author reviewed the literature, the indexing process was completed twice to help ensure accuracy of the data abstraction.

Results

Summary of the Studies

The 38 studies were published between 1986 and 2015 across 30 peer-reviewed journals. Fifteen of these journals were social or behavioral science publications, thirteen were medical or health related journals, one addressed communications topics, and one was sex-specific. Twenty-eight studies (74%) represented adolescents (ages 10–18), whereas 10 studies (26%) examined emerging adults (ages 18–24). Most studies included male and female participants (k = 28; 74%), with some female-only samples (k = 9; 24%) and one male-only sample (2%). Predominant ethnicities were Caucasian (k = 15; 39%) African-American (k = 14; 37%), Latina/o (k = 4; 11%), mixed (k = 2; 5%) and not reported (k = 3; 8%). Most studies took place in the U.S. (k = 36, 95%).

Methodological Characteristics

Among the studies, sample sizes ranged from 55 to 1587 with a median sample size of 297 (see Table 1). Not quite half of the studies (k = 18; 47%) had sample sizes greater than 300. Thirty-four studies (89%) used non-random sampling strategies and four (11%) used randomized procedures. The vast majority of studies used data from localized samples (e.g., community, school); only one reported on state-wide data. Thirty-two studies (84%) used cross-sectional designs versus longitudinal designs (k = 6; 16%). Seven studies (18%) used an established measure for parent–adolescent sexual communication (i.e., psychometrically tested and/or empirically validated), meaning that 31 studies (82%) used self-developed assessments. Only one study (3%) used an observational design with the remaining 37 (97%) using self-report questionnaires. Family climate variables were statistically controlled for in 10 of the studies (26%). These variables included parental monitoring, parent–child relationship quality/closeness/satisfaction, and general family communication. The statistical procedures used as the primary test of hypothesized relations included multiple regression (k = 18; 47%: OLS, logistic), path analysis, usually SEM (k = 9; 24%), correlation (k = 5; 13%: Pearson, Spearman, Canonical), Analysis of Variance (ANOVA; k = 3; 8%), and simple group difference tests (k = 4; 11%: t test, Chi square).

Table 1 Summary of the 38 reviewed studies, methodological information, and major findings
Full size table

Direct Relations Between Parent–Adolescent Sexual Communication and Sexual Attitudes, Perceived Norms, and Self-Efficacy

Attitudes

In total, 24 studies examined associations between parent–adolescent sexual communication and sexual attitudes. These attitudes included adolescents’ overall evaluation of engaging in sexual behavior (k = 17) from positive (typically labeled permissive) to negative (typically labeled conservative); sexual expectancies (adolescents’ perceived health and emotional consequences of engaging in sexual behavior; k = 4); and condom-related attitudes (adolescents’ evaluations, positive or negative, toward using condoms; k = 6) (see Table 2 for summary of findings). Some studies assessed multiple types of attitudes, which is why the sum of these is greater than 24.

Table 2 Summary table of pathways through which parent–adolescent sexual communication (PASC) influences adolescent sexual behavior
Full size table
Conservative/Permissive Sexual Attitudes

Nine studies showed that parent–adolescent sexual communication significantly predicted less permissive/more conservative attitudes toward engaging in sex (Cederbaum et al. 2013; DiIorio et al. 1999; Harris et al. 2013; Holman and Kellas 2015; Hutchinson and Montgomery 2007; Kowal and Blinn-Pike 2004; Miller et al. 1998; Moore and Davidson 1999; Treboux and Busch-Rossnagel 1995). Seven studies reported mixed findings (i.e., an empirical relation was only found for a meaningful subset of the sample and/or a particular variable of interest), which are detailed as follows. Nikken and de Graaf (2013) found parent–adolescent sexual communication to longitudinally predict less permissive sexual attitudes for girls, but not for boys. Werner-Wilson (1998) reported that parent–adolescent sexual communication was significantly related to less conservative sexual attitudes for girls, but not for boys. Somers and Ali (2011) found that greater extent of fathers’, but not mothers’ parent–adolescent sexual communication was correlated with more permissive sexual attitudes. Bynum (2007) found that parent–adolescent sexual communication was associated with emerging adult daughters’ more permissive sexual attitudes for African American, but not White, emerging adults. Somers and Paulson (2000) reported that less extent of parent–adolescent sexual communication was related to less permissive attitudes only among female adolescents ages 14–16. Romo et al. (2002) found that receiving messages of personal disclosure from mothers about her own romantic/sexual experiences (versus receiving cautions or advice) predicted daughters’ more conservative sexual attitudes. Lefkowitz and Espinosa-Hernandez (2007) found that receiving more abstinence messages, but not messages regarding dating, sexual feelings, and the dangers of having sex, was significantly related to young adults’ conservative sexual attitudes. Finally, in three studies, parent–adolescent sexual communication was unrelated to conservative and/or permissive sexual attitudes (Beadnell et al. 2007; Hutchinson et al. 2003; Fisher 1986).

Expectancies (Perceived Consequences)

Three studies reported that parent–adolescent sexual communication was related to adolescents’ greater awareness of the health consequences (e.g., could contract an STD) and emotional consequences (e.g., could feel guilty) of having sex (Bleakley et al. 2009), particularly when parents spoke specifically about these topics (Dittus et al. 1999; Guilamo-Ramos et al. (2007). One study found no association between parent–adolescent sexual communication and adolescents’ sexual expectancies (Ragsdale et al. 2013).

Condom Related Attitudes

In three studies, parent–adolescent sexual communication was significantly associated with adolescents’ more favorable views of using condoms (Cederbaum et al. 2013; Malcolm et al. 2013; Romer et al. 1999). One study reported mixed findings: higher quality parent–adolescent sexual communication and receiving messages of condoms as preventative predicted more favorable attitudes toward using condoms in young adults; the reception of other messages (e.g., abstinence, dangers of sex) were unrelated (Lefkowitz and Espinosa-Hernandez 2007). One study reported a non-significant longitudinal association with condom related attitudes (Hutchinson et al. 2003).

Perceived Norms

Perceived norms were examined in 13 studies. Of these, studies examined peer descriptive norms (k = 4), peer injunctive norms (k = 6), and parent injunctive norms (k = 5). Some studies assessed multiple norms.

Peer Descriptive and Injunctive Norms

One study reported that parent–adolescent sexual communication significantly predicted adolescents’ greater perceived prevalence of sexual intercourse and birth control use among their peers (Leland and Barth 1993). Three studies reported non-significant associations between parent–adolescent sexual communication and adolescents’ perceptions of their peers’ sexual activity (Bleakley et al. 2009; Whitaker and Miller 2000) and condom use (Romer et al. 1999; Whitaker and Miller 2000).

Six studies examined peer injunctive norms. One study found parent–adolescent sexual communication to significantly predict adolescents’ greater perception of eliciting social disapproval by having sex (Dittus et al. 1999). Two studies reported mixed findings: receiving messages about the social consequences of having sex predicted adolescents’ greater expectancies of social disapproval of getting pregnant, but not of having sex (Guilamo-Ramos et al. 2007). Cederbaum et al. (2013) reported greater extent of parent–adolescent sexual communication was associated with a greater perceived approvability of condom use, but not delayed intercourse, among peers. Two studies reported non-significant results between parent–adolescent sexual communication and adolescents’ perceived social consequences and benefits of having sex (Ragsdale et al. 2013) and adolescents’ evaluative judgments (positive or negative) of their peers who are having sex and their peers who are abstaining (Wills et al. 2003). Finally, one study reported that greater extent of parent–adolescent sexual communication was associated with greater perceived approvability of sex among peers (Beadnell et al. 2007).

Parent Injunctive Norms

Four studies showed parent–adolescent sexual communication was significantly associated with parental injunctive norms, including greater perceived parental disapproval of sex (Dittus et al. 1999; Jaccard et al. 1998; Guilamo-Ramos et al. 2007) and greater perceived parental approval of using condoms (Malcolm et al. 2013). One study reported a non-significant longitudinal association between parent–adolescent sexual communication and perceived maternal approval of sexual intercourse and condom use (Hutchinson et al. 2003).

Self-Efficacy

Self-efficacy was examined in 18 studies. Specifically, studies examined condom-use and safe sex communication efficacy (perceived ability to acquire, use and/or talk to a sexual partner about condoms and other safe sex practices; k = 16), and efficacy to resist unwanted sexual encounters (k = 5). Some studies assessed both types of efficacy.

Condom-use and Safe Sex Communication Efficacy

Twelve studies reported a significant relation between parent–adolescent sexual communication and adolescents’ greater perceived ability to acquire and use condoms or contraceptives (Harris et al. 2013; Hutchinson et al. 2003; Hutchinson and Cooney 1998; Hutchinson 2010; Kowal and Blinn-Pike 2004; Malcolm et al. 2013; Mastro and Zimmer-Gembeck 2015), successfully negotiate the use of condoms with a resistant or new partner (DiClemente et al. 2001; Hutchinson and Montgomery 2007; Kowal and Blinn-Pike 2004), and to discuss health-related topics such as STDs and AIDS with a current or hypothetical partner (DiIorio et al. 2000; Sales et al. 2007; Shoop and Davidson 1994). Three studies reported mixed findings: Fletcher et al. (2015) found that receiving messages endorsing sex as normative predicted emerging adult women’s greater condom-use efficacy, though messages endorsing sex as strictly relational or marital did not; Levin et al. (2012) reported that the reception of sex-positive messages positively predicted condom-use and safe-sex communication efficacy in emerging adult men, but not women, and that the reception of abstinence messages predicted a decrease in the same for both emerging adult men and women; Lefkowitz and Espinosa-Hernandez (2007) reported that higher quality parent–adolescent sexual communication predicted fewer perceived barriers to using condoms, but receiving messages pertaining to dating, fertility issues, dangers of sex, and abstinence, were unrelated. One study found no empirical relation between parent–adolescent sexual communication and adolescents’ perceived ability to use condoms (Cederbaum et al. 2013).

Efficacy to Avoid or Refuse Sexual Encounters

One study reported a significant relation between parent–adolescent sexual communication and adolescents’ greater efficacy to avoid or refuse unwanted sexual encounters (Kowal and Blinn-Pike 2004); however, four studies found this relation to be non-significant (Beadnell et al. 2007; Cederbaum et al. 2013; Hutchinson et al. 2003; Somers and Ali 2011).

Indirect Relations to Intentions/Behaviors via Attitudes, Norms, and Efficacy

Under the current framework, intentions and behaviors are modeled as being indirectly influenced via sexual attitudes, perceived norms, and self-efficacy. Generally speaking, attitudes and self-efficacy were more consistently implicated as the pathways of influence between parent–adolescent sexual communication and sexual intentions or behaviors. Little evidence emerged for perceived norms as an intervening mechanism.

Attitudes

Seven studies modeled sexual attitudes as intervening between parent–adolescent sexual communication and sexual intentions/behaviors. Three of these showed significant indirect effects, which are described as follows. Cederbaum et al. (2013) showed parent–adolescent sexual communication to significantly predict less permissive sexual attitudes, which in turn significantly predicted greater abstinence intentions. Malcolm et al. (2013) reported that higher quality sexual communication predicted more positive condom-use attitudes, which in turn predicted greater condom-use intentions and behaviors. Miller et al. (1998) showed that showed higher quality parent–adolescent sexual communication longitudinally predicted less permissive sexual attitudes, which then predicted later abstinence intentions, which in turn predicted abstinence. Two studies showed mixed findings: Treboux and Busch-Rossnagel (1995) showed that the relation between parent–adolescent sexual communication and greater intentions to delay sexual intercourse was statistically mediated by more conservative sexual attitudes, but only for the younger portion of their sample (ages 14–15); Romo et al. (2002) found that receiving messages of maternal disclosure, but not other messages (e.g., beliefs and values) was related to lower levels of later sexual behavior, a relation that was statistically mediated by more conservative sexual attitudes. Two studies reported that sexual attitudes did not statistically mediate between extent of parent–adolescent sexual communication and adolescents’ sexual intentions (Beadnell et al. 2007) and sexual risk behavior (Hutchinson et al. 2003).

Perceived Norms

Perceived norms were modeled as intervening mechanisms in five studies. One study showed greater extent of parent–adolescent sexual communication to be associated with higher perceived norms of condom use, which in turn was associated with greater condom use intentions (Cederbaum et al. 2013). Three studies failed to show perceived norms statistically mediating the relation between parent–adolescent sexual communication and sexual intentions/behaviors: in two of these, maternal and peer norms were predictive of later safe sex behaviors, but were not actually associated with greater extent of parent–adolescent sexual communication (Hutchinson et al. 2003; Wills et al. 2003); Malcolm et al. (2013) found that higher quality parent–adolescent sexual communication predicted greater perceived maternal approval of condom use, but that maternal approval of condom use did not predict later intentions and behaviors. Finally, Beadnell et al. (2007) found that greater extent of parent–adolescent sexual communication was related to increased sexual intentions and behaviors, which was statistically mediated by more favorable peer norms toward sex.

Self-Efficacy

In four studies, self-efficacy was modeled as an intervening mechanism. In three studies, path models and/or mediation analyses showed that parent–adolescent sexual communication was related to greater efficacy to use condoms and to communicate with a sexual partner about safe sex practices, which were then associated with greater safe sex behavior (e.g., fewer sexual partners, fewer instances of unprotected sex) (DiIorio et al. 2000; Hutchinson et al. 2003; Malcolm et al. 2013). One study showed that parent–adolescent sexual communication was not indirectly related to sexual intentions via self-efficacy to refuse unwanted sex (Beadnell et al. 2007).

Communication Characteristics: Extent, Messages, and Quality

The conceptual model also suggests that the communication characteristics of parent–adolescent sexual communication may determine how effective it is in predicting adolescents’ sexual attitudes, perceived norms, and self-efficacy. Of interest to the review was the overall extent, the content, and the quality of parent–adolescent sexual communication. A handful of studies assessed other topics, including parents as main sex educators (k = 2; 5%), parents as first sex educators, (k = 1; 2.5%), whether parent–adolescent sexual communication ever took place growing up (k = 1; 2.5%), and not described (k = 1; 2.5%).

Overall Extent of Parent–Adolescent Sexual Communication

Several of the studies assessed the overall amount of parents’ sexual communication (k = 13, 33%). This involved assessing parent–adolescent sexual communication on a wide array of topics, but aggregating these for an overall score indicating how much sexual communication has taken place. The overall extent of parent–adolescent sexual communication predicted less permissive sexual attitudes in two studies (Nikken and de Graaf 2013; Treboux and Busch-Rossnagel 1995), more permissive attitudes in three studies (Bynum 2007; Somers and Ali 2011; Werner-Wilson 1998), and displayed no empirical relation to sexual attitudes in two studies (Fisher 1986; Ragsdale et al. 2013). Regarding perceived norms, it predicted greater perceived peer sexual activity (i.e., peer descriptive norms) in one study (Leland and Barth 1993), more perceived social consequences of sex in one study (i.e., injunctive norms; Dittus et al. 1999), greater parental disapproval of sex in one study (Jaccard et al. 1998), and was unrelated to peer norms in one study (Ragsdale et al. 2013). Finally, overall extent of parent–adolescent sexual communication was significantly related to greater safe-sex efficacy in three studies (DiIorio et al. 2000; Mastro and Zimmer-Gembeck 2015; Shoop and Davidson 1994) and unassociated with efficacy to refuse unwanted sex in one study (Somers and Ali 2011).

Messages Communicated

Nineteen (50%) studies examined specific sexual messages communicated by parents. For the most part, these included communication about the health risks of sex and safe sex practices (e.g., discussing STDs, birth control, condom use; k = 14, 36%). A handful of other studies examined communication about sex-related values (k = 3, 8%) and the emotional/social consequences of sexual behavior, (k = 2, 5%).

Receiving messages about sexual risk and safe sex practices predicted less permissive sexual attitudes in six studies (Cederbaum et al. 2013; DiIorio et al. 1999; Harris et al. 2013; Holman and Kellas 2015; Hutchinson and Montgomery 2007; Kowal and Blinn-Pike 2004), more favorable condom-related attitudes in three studies (Cederbaum et al. 2013; Lefkowitz and Espinosa-Hernandez 2007; Romer et al. 1999), greater awareness of the health consequences of sex in two studies (Dittus et al. 1999; Guilamo-Ramos et al. 2007), and was unrelated to attitudes in two studies (Beadnell et al. 2007; Hutchinson et al. 2003). Regarding perceived norms, sexual-risk messages were linked to more perceived peer sexual activity in one study (Beadnell et al. 2007), though three studies found non-significant relations with perceived norms (Hutchinson et al. 2003; Romer et al. 1999; Wills et al. 2003). Regarding self-efficacy, eight studies linked sexual risk messages to greater condom-use and sexual communication efficacy (DiClemente et al. 2001; Harris et al. 2013; Hutchinson and Cooney 1998; Hutchinson et al. 2003; Hutchinson and Montgomery 2007; Hutchinson 2010; Kowal and Blinn-Pike 2004; Sales et al. 2007) whereas one study found this link to be non-significant (Cederbaum et al. 2013); one study linked sexual risk messages to greater efficacy to refuse unwanted sex (Kowal and Blinn-Pike 2004), although three studies found this relation to be non-significant (Beadnell et al. 2007; Cederbaum et al. 2013; Hutchinson et al. 2003).

Messages about sexual values were assessed in three studies and included abstinence and sex-positive values. Communication of abstinence values predicted more conservative sexual attitudes in one study (Lefkowitz and Espinosa-Hernandez 2007), but also failed to predict condom-related attitudes (Lefkowitz and Espinosa-Hernandez 2007) and safe-sex efficacy in others (Fletcher et al. 2015; Lefkowitz and Espinosa-Hernandez 2007; Levin et al. 2012). Sex-positive messages predicted greater safe-sex efficacy in one study (Levin et al. 2012), though it was unrelated to safe-sex efficacy in another (Fletcher et al. 2015). Finally, two studies found that messages about the emotional and social consequences of sex were related to adolescents’ greater expectancies of these same consequences (Dittus et al. 1999; Guilamo-Ramos et al. 2007).

Quality of Parent–Adolescent Sexual Communication

Fewer studies (k = 4, 10%) examined the quality of parent–adolescent sexual communication, or the degree of comfort, openness, and respect during sexual conversations. Among these, higher quality parent–adolescent sexual communication predicted more conservative sexual attitudes (Miller et al. 1998), more favorable condom attitudes (Lefkowitz and Espinosa-Hernandez 2007; Malcolm et al. 2013), greater perceived parental approval of condom use (Malcolm et al. 2013), and greater safe-sex efficacy (Mastro and Zimmer-Gembeck 2015). One study reported a non-significant relation between quality of parent–adolescent sexual communication and sexual attitudes (Lefkowitz and Espinosa-Hernandez 2007).

Discussion

Parent–adolescent sexual communication can help adolescents make healthier sexual decisions (Widman et al. 2015). However, sexual conversations are challenging for many families (Malacane and Beckmeyer 2016), and pediatricians and other helping professionals may seek to help parents more effectively tailor their sexual conversations with adolescents. The current review advances theory on parent–adolescent sexual communication by identifying pathways that explain how parent–adolescent sexual communication is associated with adolescents’ sexual intentions and behaviors, as well as examining the characteristics of the communication that may contribute to its effectiveness.

Parent–Adolescence Sexual Communication and Pathways of Association with Sexual Intentions and Behaviors

The conceptual model guiding this study posits that parent–adolescent sexual communication predicts change in adolescents’ sexual cognitions (i.e., attitudes, perceived norms, and self-efficacy), which in turn can influence their sexual intentions and/or behaviors. The reviewed studies provided support for several aspects of this model, particularly for pathways through attitudes and self-efficacy. Parent–adolescent sexual communication was consistently and empirically related to less permissive attitudes toward engaging in sexual intercourse, greater awareness of the various consequences of having sex, and more positive attitudes toward using condoms. Parent–adolescent sexual communication was also consistently linked to a greater perceived ability to buy and use condoms and to talk with sexual partners about healthy sexual practices, although studies did not necessarily report greater efficacy to avoid or refuse unwanted sexual behavior. A handful of studies then looked at how these cognitions, in turn, predicted sexual intentions or behaviors. These generally provided evidence that adolescents’ sexual attitudes intervened between parent–adolescent sexual communication and adolescents’ sexual intentions and behaviors (e.g., delayed intercourse). Similarly, several studies supported self-efficacy as a mechanism of change, such that parent–adolescent sexual communication predicted greater safe-sex efficacy, which in turn predicted greater safe-sex intentions and behaviors (e.g., more consistent condom use).

Findings regarding perceived norms were mixed at best. Although parent–adolescent sexual communication was generally associated with greater perceived parental disapproval of having sex (parent injunctive norms), it was often unassociated with adolescents’ perceptions of peers’ sexuality (peer descriptive and injunctive norms). Not surprisingly, perceived norms often failed as an intervening mechanism to explain the relation between parent–adolescent sexual communication and sexual intentions/behaviors, as the majority of interested studies found these indirect paths to be non-significant. Of course, null findings must be interpreted with caution, and so concluding that parent–adolescent sexual communication is unimportant in this regard would be premature. Adolescents speak about sexual topics with their friends much more frequently than they do with their parents (Lefkowitz and Espinosa-Hernandez 2007), and so peer norms may be more strongly conveyed by peers themselves. Some research shows that parent–adolescent sexual communication may actually moderate associations between peer norms and sexual intentions and/or behaviors. In a recent longitudinal study, descriptive and injunctive peer norms predicted Dutch adolescents’ greater intentions to have sexual intercourse (van de Bongardt et al. 2014). When controlling for these peer effects, parent–adolescent sexual communication had no relation to sexual intentions. However, for adolescents who reported more frequent sexual communication with their parents, both descriptive and injunctive peer norms were unassociated with their sexual intentions. This notion of parent–adolescent sexual communication as a buffer against peer sexual norms has been suggested in other studies (Whitaker and Miller 2000), and may represent a better pathway (than a mediated one) for understanding the role of parent–adolescent sexual communication in regards to adolescents’ perceived peer norms around sexuality.

These findings represent the first systematic synthesis of the specific mechanisms of change through which parent–adolescent sexual communication is associated with adolescents’ sexual intentions and behaviors, evidencing several theoretically coherent patterns. It should be noted, of course, that studies on a particular pathway or process were sometimes too few to draw definitive conclusions. As such, these findings should be considered preliminary, and more work is needed to further understand and verify these pathways of influence. Nevertheless, these findings can help orient researchers to the pathways by which parent–adolescent sexual communication can contribute to adolescents’ sexual outcomes.

Communication Characteristics

This review was also the first to systematically summarize evidence, among a set of studies, for how the specific characteristics of parent–adolescent sexual communication (i.e., extent, content, quality) may influence its effectiveness in predicting adolescents’ sexual cognitions. The most commonly assessed characteristic among the reviewed studies was content, and in particular messages about sexual risk and safe sex practices, which were examined in over one-third of the studies. With striking consistency, these safe-sex messages predicted more conservative/less permissive sexual attitudes and greater self-efficacy to enact safe sex practices. This contrasts somewhat with studies that assessed the overall extent of parent–adolescent sexual communication. For example, the overall extent of parent–adolescent sexual communication predicted less permissive attitudes, more permissive attitudes, and failed to predict attitudes with fairly equal frequency. Such inconsistencies are likely due, in part, to methodological limitations: measures of overall extent confound the specific topics discussed among families, lowering its overall predictive power (Lefkowitz 2002). Another explanation for these inconsistencies is that parents’ general treatment of sexual topics with their adolescents may often fail to equip their adolescents with safe-sex competencies; safe-sex topics should be addressed specifically.

Other aspects of parent–adolescent sexual communication were examined, and although these studies were too few to draw definitive conclusions, they exhibit notable patterns that warrant future research. For example, abstinence messages, though predictive of less permissive sexual attitudes, were unrelated to other safe-sex competencies such as condom-related attitudes and efficacy. Research on broader sex education programs shows that adolescents receiving abstinence-only sex education may actually be more at risk for health-related consequences of sex (e.g., unintended pregnancy) than those who receive comprehensive sex-education (Kohler et al. 2008). Initially, a similar pattern appears to characterize parents’ sexual communication; such findings should be followed up in relation to these social cognitive pathways specifically. Finally, studies in the review showed that higher quality parent–adolescent sexual communication (i.e., that was open, comfortable, and respectful) was typically associated with greater safe-sex self-efficacy and less permissive sexual attitudes. This dovetails with other research showing that lower quality sexual conversations (e.g., domineering, dismissive) predicts less sexual knowledge (Lefkowitz et al. 1998) and more sexual behavior in adolescents (Rogers et al. 2015). The implication is that the actual interpersonal dynamics of sexuality conversations might influence the degree to which adolescents internalize parents’ messages. Continued research representing these less studied characteristics is a fruitful avenue for future studies, which may help show how they produce meaningful variability in the effectiveness of parent–adolescent sexual communication.

In a similar vein, researchers should also bear in mind that these communication characteristics are not the only factors surrounding parent–adolescent sexual communication that can influence its effectiveness. Situational factors also matter a great deal. Two such factors that could produce variation in the reviewed pathways, but were not represented well enough among the reviewed studies for inclusion in the synthesis, include gender (of the parent and the adolescent) and the timing of parent–adolescent sexual communication relative to the child’s sexual development. Greater evidence now indicates that parent–adolescent sexual communication is more predictive of sexual behaviors when mothers (versus fathers) are the source (DiIorio et al. 1999) and when daughters (versus sons) are the recipient (Widman et al. 2015). This likely represents different gender-based socialization around dating and sexuality (Perilloux et al. 2008). Parent–adolescent sexual communication may also be more effective among younger adolescents who are less sexually developed and who do not face the same social pressures around sexuality as do older adolescents (Beckett et al. 2010). Research should remain open to these (and other) situational features of parents’ sexual conversations.

Limitations of the Reviewed Studies

The reviewed studies had a number of limitations that should be addressed in future work. A majority of the reviewed studies reported on cross-sectional data. As such, the directionality of effects cannot be determined. Operating under the present conceptual model, causal interpretations from parent–adolescent sexual communication to cognitions and from cognitions to intentions and/or behaviors can become taken for granted. The reader is reminded that the results of many of these studies can potentially be interpreted as proceeding in the opposite direction given the lack of longitudinal evidence and untested reciprocal relations. A reliance on cross-sectional data is also concerning for interpretations of indirect effects and statistical mediation. Such effects are best supported by temporally-ordered data, given that a mediated process assumes the passage of time as a sequence of events unfolds (Maxwell and Cole 2007), but this information is not provided by cross sectional data. Although this does not altogether invalidate interpretations of mediation, it is a stipulation that must be considered when making such interpretations. Longitudinal data on parent–adolescent sexual communication can be particularly valuable in addressing these limitations.

Measurement issues were also of concern. Only seven of the reviewed studies (18%) used empirically validated measures, meaning that most measures of parent–adolescent sexual communication were idiosyncratic assessments developed by the studies’ authors. This could create challenges in drawing conclusions across a body of literature, and so the continued development and use of empirically validated measures is a critical next step for research on parent–adolescent sexual communication. Future studies should also consider the value of observational methods, which can be particularly useful for understanding the dynamics of families’ sexual conversations from a more objective lens (e.g., Rogers et al. 2015).

Finally, most studies did not account for other socio-contextual variables in their analyses. Parent–adolescent sexual communication is one influence among many, such as the general family climate, friends, and romantic partners, that can exert influence on adolescents’ sexual cognitions and behaviors (Buhi and Goodson 2007; Huebner and Howell 2003). This omission is particularly concerning in regards to family climate variables, as parent–adolescent sexual communication is associated with more positive parenting in general (DiIorio et al. 2004). Thus, associations between parent–adolescent sexual communication and adolescent sexual outcomes could potentially be an artifact of the influence of more general family functioning. It should be noted that out of 10 studies that did account for one or more of these family climate variables, most found parent–adolescent sexual communication to still uniquely predict adolescents’ sexual outcomes (n = 9). Still, future work would do well to more carefully account for these and other socio-contextual factors.

Limitations of the Present Review

The present review is also limited in a few ways. First, this review is not a representation of all studies on parent–adolescent sexual communication, but those that mapped on to the conceptual model (e.g., included an association with sexual cognitions). Readers should not conclude that these represent all the literature available. Relatedly, being guided by a specific framework, the review was sensitive to the theoretically delineated pathways of influence, but insensitive to other possible pathways, such as moderated relations (see van de Bongardt et al. 2014). Finally, the review used ten studies on emerging adult samples. Though they gave useful insight, emerging adulthood and adolescence are distinct periods, particularly in relation to sexual development. Research is needed to understand these indirect pathways differentially for adolescents and emerging adults.

Conclusion

This review makes a timely, theoretical contribution to the study of parent–adolescent sexual communication. Most scholarly attention is given to direct relations between parent–adolescent sexual communication and adolescents’ sexual intentions and/or behaviors, and so little is known regarding the explanatory processes that account for these associations. The present review addresses this need by sensitizing researchers to the social-cognitive mechanisms through which parent–adolescent sexual communication might influence adolescents’ sexual behavior, as well as the characteristics of that communication that can make it more or less effective over these social-cognitive mechanisms. The greater consideration of these explanatory processes in future studies, in addition to making key methodological improvements, can help advance scholarly research on parent–adolescent sexual communication and ultimately aid in the development of family-based prevention programs that aim to promote adolescents’ sexual health.