Sexual Risk Behavior of Children and Adolescents With a History of Sexual Violence
A Meta-Analytic Review
Abstract
Abstract:Theoretical Background: Experiencing sexual violence (SV) at a young age can impede sexual development and lead to sexual risk behavior. Objective: This meta-analysis aims to provide a systematic overview regarding the development of sexual risk behavior in sexually violated children and adolescents. Method: A systematic literature search was carried out (PubMed, ERIC, Cochrane, and PubPsych). Meta-analyses were conducted, investigating the link between child SV and sexual behaviors (teenage pregnancy, safer sex, substance use during intercourse, transactional sex, and number of sex partners). Results: A total of 19 studies were identified. Survivors of SV are more likely to have higher numbers of sex partners, be involved in teenage pregnancy, and participate in transactional sex. Discussion: Early identification of sexually violated youth is important in order to intervene at an early stage and reduce the effects of sexual violence preventively. PROSPERO registration number CRD42020145139.
Zusammenfassung:Theoretischer Hintergrund: Das Sexualverhalten von Kindern und Jugendlichen ist bisher kaum untersucht worden, obwohl die Pubertät als eine hochsensible Phase bekannt ist, in der eine zweite Welle der zerebralen Umstrukturierung stattfindet. Die hohe Plastizität des adoleszenten Gehirns ermöglicht es, dass Umwelteinflüsse besonders starke Auswirkungen auf kortikale Schaltkreise haben. Dies ermöglicht zwar die intellektuelle und emotionale Entwicklung, bedingt aber auch potenziell schädliche Einflüsse. Das Erleben sexueller Gewalt in jungen Jahren kann den normalen Entwicklungsverlauf stören, die sexuelle Entwicklung behindern und zu sexuell unangemessenem Verhalten führen. Zielsetzung: Ziel dieser Meta-Analyse ist es, einen systematischen Überblick über den aktuellen Forschungsstand zur Entwicklung von sexuellem Risikoverhalten bei Kindern und Jugendlichen mit sexualisierten Gewalterfahrungen zu geben, da der Zeitraum der Kindheit und Jugend als besonders anfällig für abnorme Erfahrungen beschrieben wird und sexualisierte Gewalt ein weit verbreitetes Phänomen ist. Dementsprechend lautet die Forschungsfrage: Inwieweit ist sexualisierte Gewalt bei Kindern und Jugendlichen mit deren sexueller Entwicklung verbunden? Methode: Es wurde eine systematische Literaturrecherche in den Datenbanken PubMed, ERIC, Cochrane und PubPsych durchgeführt. Es wurden Meta-Analysen durchgeführt, in denen der Zusammenhang zwischen sexualisierter Gewalt gegen Kinder und Jugendliche und sexuellem Verhalten in Bezug auf Teenagerschwangerschaften, Safer Sex, Substanzkonsum beim Geschlechtsverkehr, transaktionalem Sex und Anzahl der Sexualpartner untersucht wurde. Studien wurde eingeschlossen, wenn die Teilnehmer jünger als 21 Jahre alt waren, aktuell oder in der Vergangenheit sexualisierte Gewalt erfahren haben, die Ergebnisse sexuelle Risikofaktoren gemäß UNAIDS umfassten, eine Kontrollgruppe ohne eine Vorgeschichte sexualisierter Gewalt eingeschlossen war und es einen zeitlichen Vorlauf des Vorfalls sexualisierter Gewalt gab. Ergebnisse: Es konnten 19 Studien zum sexuellen Risikoverhalten nach dem Erleben sexualisierter Gewalt identifiziert werden. Meta-Analysen ergaben signifikante Ergebnisse für Teenager-Schwangerschaften, Anzahl der Sexualpartner und Transaktionssex. Diskussion und Schlussfolgerung: Bei Betroffenen sexualisierter Gewalt ist die Wahrscheinlichkeit höher, dass sie eine größere Anzahl von Sexualpartnern haben, im Teenageralter schwanger werden und an Transaktionssex beteiligt sind. Ein tieferes Verständnis dafür, wie sich sexualisierte Gewalt auf sexuelles Risikoverhalten auswirken kann, erhöht die Wahrscheinlichkeit, dass die tatsächlichen Ursachen sexueller Folgeerscheinungen wirksam angegangen werden können. Daher ist es wichtig, Jugendliche mit sexualisierten Gewalterfahrungen frühzeitig zu identifizieren, um frühzeitig einzugreifen und die Auswirkungen sexualisierter Gewalt präventiv zu verringern. Diese Übersichtsarbeit wurde von PROSPERO registriert und genehmigt (Registrierungsnummer CRD42020145139).
Sexual violence, that is, the “unwanted sexual activity, with perpetrators using force, making threats or taking advantage of victims not able to give consent” (Kazdin, 2000), is a reality for millions of people worldwide.
Reporting prevalence rates is a major challenge, as sexual violence is mainly underreported, with an elevated number of non-reporting in the case of sexual violence against men and boys (United Nations High Commissioner for Refugees, 2012). Sexual violence against children is hardly reported at the time that the violence occurs, and in many cases is never disclosed with most data stemming from retrospectively asking adults about their history of sexual violence experiences (Andrews et al., 2002). Globally, about 1 out of 3 women has experienced sexual violence, with widely varying prevalence estimates (García-Moreno et al., 2013). Every 73 s, an American is sexually assaulted (Department of Justice, 2019) and every 9 min, that victim is a child (Department of Justice, 2019). Available data from studies conducted in different parts of the world suggest that between 7 % and 36 % of girls, and between 3 % and 29 % of boys, have experienced some form of sexual violence (World Health Organization, 2003). Even though estimated prevalence rates are high and the consequences of sexual violence against children are significant, many countries lack a reliable reporting system and criminal prosecution for sexual violence against children and adolescents (World Health Organization, 2003).
Consequences for Sexual Behavior
Since the period of childhood and adolescence is described as being highly vulnerable to abnormal experiences, it is of great importance to detect child sexual violence as early as possible. The developmental phase of children and adolescents is marked by social, behavioral, psychological, biological, and sexual changes (DeBellis et al., 1994). Experiencing sexual violence at a young age can disturb the normal development course (DeBellis et al., 1994). Biologically, chronic sexual violence can stimulate stress hormone levels and predispose adolescents to a premature onset of puberty (DeBellis et al., 1994). In adolescents, both hyper- and hypo-secretion have been linked to increased sexual behaviors, early sexual behaviors, and coercive sexual experiences among adolescents (Hasson et al., 2022; Miller et al., 2001; Vicary et al., 1995). Sexual violence against children and adolescents can impede sexual development and lead to sexually inappropriate and hyperaroused behavior (Finkelhor & Browne, 1986).
Paolucci and colleagues (2001) conducted a meta-analysis on the effects of childhood sexual violence and found that across 37 studies (N = 25,367), there was a minimum 14 % increase in the risk of engaging in early sexual behavior for those children. Another meta-analysis of the relationship between childhood sexual abuse and HIV risk behavior among women found that across 46 studies there was an increased risk of unprotected sexual activity; having multiple partners; engaging in sexual intercourse in exchange for money, drugs, or shelter; and sexual revictimization in adulthood (Arriola et al., 2005).
Teenage Pregnancy
Regarding teenage pregnancy, a birth cohort study of 1,265 New Zealanders showed that, of the 533 women with complete data, those who became pregnant by age 20 were significantly more likely to have retrospectively reported experiencing sexual violence in childhood (Woodward et al., 2001). Arriola and colleagues (2005) found that a history of childhood sexual violence is associated with greater likelihood to engage in sexual risk behaviors in adolescence and in adulthood, which increases the risk for unwanted pregnancy. Besides, Maharaj and colleagues (2007) reported similar findings for young women in South Africa, stating that youths who experienced coerced first sex were significantly more likely to have ever been pregnant and to report their pregnancy as unwanted. In a community-based case–control study (Ochen et al., 2019), 480 female participants between 14 and 19 years of age were studied. The results reveal that sexual violence experiences were significantly associated with teenage pregnancy. These results concur with some studies that have postulated that sexual violence places girls in particular at higher risk of being involved in teenage pregnancies (Harner, 2016; Hillis et al., 2001; Pallitto & Murillo, 2008). Finally, in a meta-analysis of 21 studies, those who experienced sexual violence were more than twice more likely to experience an adolescent pregnancy than those who had not experienced sexual violence, even after controlling for study heterogeneity (e. g., prospective vs. retrospective designs; Noll et al., 2009).
Sex Partners
In a prospective study following 251 adolescents (169 with documented maltreatment) longitudinally, maltreatment was related to subsequent risky sexual behaviors (i. e., ≥ 4 sexual partners and lack of condom use during the last sexual encounter) for both males and females in adolescence (Negriff et al., 2015). Pengpid and Peltzer’s (2020) results concur with these findings, stating that physical and/or sexual violence was associated with sexual risk behavior including having had multiple sexual partners.
Safer Sex
In a prospective study, maltreatment was related to subsequent risky sexual behaviors including lack of condom use during last sexual encounter for both males and females in adolescence (Negriff et al., 2015). A study of South African female youths aged 15 – 24 years showed that young women who experienced forced sex with their partner in the past 12 months were less likely to use condoms consistently compared to women who had not experienced sexual violence (Pettifor et al., 2004). In sexually experienced male adolescents, child sexual violence was associated with a lower likelihood that a condom was used during the last sexual activity (Raj et al., 2000; Shrier et al., 1998).
Substance Use During Intercourse
Negriff and colleagues (2015) explicitly examined differing types of maltreatment within the same predictive models. They examined how differing types impacted risky sexual behaviors at mean age of 18 years. Findings indicated that sexually abused females were at higher risk of having had sex under the influence of alcohol or drugs. Pengpid and Peltzer (2020) found that physical and/or sexual violence was associated with alcohol use in the context of sex. Moreover, Raj and colleagues (2000) and Shrier and colleagues (1998) both reported a greater likelihood of substance use at the last sexual encounter in male adolescents with sexual violence experiences compared to adolescents without these experiences.
Transactional Sex
Many studies have found that female adults who experienced child sexual violence are more likely to be involved in sex trading (Ahrens et al., 2012; Senn et al., 2011; Stroebel et al., 2012; Wilson & Widom, 2008), depending on participant age (London et al., 2017; Scheidell et al., 2017) and on type of abuse (Stroebel et al., 2013). For male adults, some studies have shown an association between sexual violence experiences and involvement in the sex trade (Scheidell et al., 2017), depending on participant age (London et al., 2017), but others have demonstrated no association (Ahrens et al., 2012; Wilson & Widom, 2008). Research has generally shown a positive connection between child sexual violence and involvement in transactional sex for mixed samples (Ahrens et al., 2012; Meade et al., 2012; Senn et al., 2007). Nevertheless, all these studies collect their data retrospectively. Studies in which children and adolescents are directly interviewed are strongly underrepresented.
The Dilemma
Overall, the importance of examining various facets of sexual violence has gained prominence (Noll, 2008). Sexual violence against children and adolescents has led to numerous studies over the past decades (Modelli et al., 2012). This research has alerted the general public and the scientific community to possible short- and long-term consequences, made significant contributions to the importance of targeting this topic, and has provided essential information on detecting, treating, and preventing sexual violence. Nevertheless, it is difficult to generalize sexual developmental results in sexual violence research. On the one hand, there are many biases affecting retrospective reporting of sexual violence, most of which are thought to give rise to underestimation of the phenomenon as compared to prospective reports. There exists the potential bias of retrospective recall (Andrews et al., 2004). The size of discrepancy between the prospective and retrospective estimates is a matter of concern, as there is good reason to believe that the retrospective estimates are biased downward by post facto rationalization (Andrews et al., 2004). But there are also biases of retrospective reports that can give rise to overestimates, for example, memories, which are, nevertheless, less common than previously thought (Brewin & Andrews, 2016).
On the other hand, sexual behaviors of children have been rarely investigated, even though puberty is known to be a highly vulnerable phase in which a second wave of cerebral restructuring occurs (Konrad et al., 2013). The high plasticity of the adolescent brain enables environmental influences to induce particularly strong effects on cortical circuitry (Konrad et al., 2013). While this makes intellectual and emotional development possible, it also conditions potentially harmful influences (Konrad, Firk & Uhlhaas, 2013).
Objectives
This systematic review and meta-analysis will thus focus on the sexual development of children and adolescents who have experienced sexual violence, in order to shed light on sexual changes that may impact sexual behaviors negatively during the course of their lives. The main aim of this meta-analysis is to initially determine an overall effect of sexual violence while excluding other types of adversities. The results of this meta-analysis call for future work to establish the extent to which these effects are attributable to actual sexual violence per se or to other forms of adversity and maltreatment that are known to typically co-occur with sexual violence. In order to identify relevant studies, we employed the Participants, Interventions, Comparisons, Outcomes, and Study type (PICOS) scheme (Moher et al., 2009). We restrict our study to children and adolescents (P) up to the age of 21 years to avoid retrospective recall (Slaymaker, 2004). Studies were eligible if children and adolescents had been or were currently sexually violated (I) and a control group without a history of sexual violence was included (C). Outcome measures were based on indicators of sexual risk behavior as recommended by UNAIDS. These recommendations target the types of sexual partner(s), safer sex, age at first sexual encounter, number of partners, commercial sex, and age mixing between partners (Slaymaker, 2004) (O). Single case and case–control studies, books, and dissertations were excluded (S).
Method
This review is registered and approved by PROSPERO (https://www.crd.york.ac.uk/prospero/). For more detailed information, use registration number CRD42020145139. The Method section is further structured according to the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) group (Moher et al., 2009).
Eligibility Criteria
The literature search included studies of human participants with no limitation on the year of publication. An article was included if it met the following criteria: (1) participants were younger than 21 years, (2) had currently or in the past been sexually violated (see definition of sexual violence), (3) outcomes comprised sexual risk factors according to UNAIDS (as described above), (4) a control group without a history of sexual violence was included, and (5) there was a temporal precedence of the incident of sexual violence. Single case and case–control studies, books, and dissertations were excluded. Studies had to be published in English, German, or Dutch. Studies that merged the included and excluded (part of the) age ranges were omitted. There were no further restrictions.
Information Sources
PubMed, ERIC, Cochrane, and PubPsych electronic databases from inception to September 2019 were searched with an update in 2021. These were chosen because of the psychological focus that encompasses psychosocial effects of sexual violence in childhood. Additionally, references from current reviews and meta-analyses were included.
Search Strategy
The search term combined concepts of the population (i. e., child or adolescent), the incident (e. g., sexual violence, assault, molestation) and several outcomes concerning the development of sexual behaviors (e. g., sexuality, intimacy, drive, sexual risk behavior, pregnancy, sexual distortions etc.). The search term was kept broad with a focus on sensitivity instead of specificity, initially resulting in a large number of articles, which were then reduced based on our inclusion criteria. For more detailed information, see Table 1.
Selection Process
The study selection process was based on the PRISMA flow chart (Moher et al., 2009). After feeding the electronic databases with the Boolean search term, records were screened independently, based on title and abstract, by the two reviewers and a third team member. Records were included in the full-text screening if they met the previously described eligibility criteria. Final inclusions were determined by consensus.
Data Collection Process
Descriptive characteristics (e. g., sample size, mean age, age range, and gender) were extracted. Different dimensions of sexual violence were included, as the categorization of sexual violence depicted in the studies differs from vague distinction to narrowly formulated definitions. As outcomes, various sexual risk factors were coded, including teenage pregnancy, transactional sex, multiple sex partners, safer sex, or substance use during intercourse.
The sampling method (random sample from the general population vs. specific population), the assessment (survey, interview), and the method of data collection (self-report, file analysis, interview with caretaker) were extracted as methodological moderators.
Depending on the information available for each study, outcomes of the total sample or separately for males and females were reported. Items were coded by two researchers independently, ensuring an interrater reliability.
Data Items
Information extracted from the selected papers included publication year, university/research institution, sample size, sample demographics (mean age, age range, percentage female), assessment, method of data collection, type of sexual violence, main outcome variable, comparison groups, and outcomes (variable, statistical measure, significance level).
Risk of Bias in Individual Studies
After full-text evaluation, the risk of bias and the quality of the selected studies were assessed by two reviewers independently, based on the risk-of-bias instrument for nonrandomized studies of exposures (ROB-NRSE) dealing with effects of environmental exposures on health outcomes (Morgan et al., 2019). Key domains of the risk-of-bias assessment were confounding, participant selection, classification of exposures, departures from intended exposures, missing data, outcomes, selection of reported results, and overall bias. The risk of bias was assessed separately for each study and classified as low, moderate, serious, or critical risk of bias, or alternatively as no information. In the case of disagreements, final classifications were determined by consensus. For a detailed overview of the quality assessment, see Table 2.
Data Synthesis and Analysis
Statistical analyses were based on the recommendations of Borenstein (Borenstein et al., 2009). The analyses were conducted by using the software Meta-Mar v2.7.0 (http://www.meta-mar.com/meta). A standardized mean difference effect size was calculated for each continuous outcome, whereas a logged odds ratio was calculated for each dichotomous outcome in each study. Logged odds ratios were converted back to odds ratios for ease of interpretation. The homogeneity of the effect size distribution for each mean effect size was calculated using the Q statistic, which follows the chi-square distribution. Compared to fixed-effects models, random-effects models provide a more conservative estimate and are often applied in meta-analysis when distributions are heterogeneous (Lipsey & Wilson, 2001). The random-effects model takes into account sources of variation within and between studies (Borenstein et al., 2007), while the fixed-effects model often augments effect size precision (Hunter & Schmidt, 2000). Consequently, using the random-effects model was deemed appropriate. Moreover, a fail-safe population size was computed to assess the file drawer problem inherent in meta-analyses in order to detect the number of studies needed to bring the calculated significance level of the pooled effect near the critical significance level.
Results
Study Selection
Figure 1 illustrates the results of the study screening process. The electronic database search in 2019 provided a total of 16,753 papers. Of these, 16,352 citations did not meet the inclusion criteria and were discarded. The full texts of 401 studies were examined in depth. After literature review articles, case studies, and other article formats not meeting the search criteria were removed, 19 publications remained. Articles were re-examined for relevance and design qualifications for final study selection. In 2021 the database search was updated, providing an additional 749 articles. Nine articles underwent full-text screening, but none was included in the final sample. As such, findings from 19 studies published between 1995 and 2019 constituted the final study sample. Revictimization studies were excluded if the unrelated variable did not refer to the initial sexual violence experienced but explicitly to revictimization. If the independent variable related to sexual violence experiences and there was subsequent revictimization, the studies were included.
Study Characteristics
Nearly 70 % of the studies came from the United States. Approximately 20 % of the studies were conducted in Sweden, one study came from Brazil, and one study was carried out in Uganda. Study sample sizes ranged from 162 to 29,187 with a mean sample size of 3,562 and a total sample size of 67,682 children and adolescents. Age ranged from 10 to 19 years with a mean age range of 12.7 – 18.0. The mean percentage of female participants was 77.4 %. For a detailed overview of the study characteristics, see Table 3.
Overall Effects
Effect sizes for each study are presented in Table 4. Analysis of mean effect sizes (displayed in Table 5) revealed significant group differences for three out of five analyses. Continuous and dichotomous measures of sexual risk factors yielded significant positive group differences, such that victims of sexual violence have significantly more sex partners, are more likely to be involved in pregnancy, and sell sex more often than comparison groups. Supported by statistically significant Q values and relatively high τ2 and I2 values, the assumption that the true effect was identical across studies could not be made. Examination of the Q statistics revealed that the distributions for all five risk outcomes were heterogeneous, suggesting that there was a substantial between-study variation. Therefore, the random-effects model was appropriate for the present analyses.
Pregnancy
The number of studies included was 10. Mean effect sizes of this dichotomous outcome demonstrated significant group differences. Odds ratios revealed that individuals with a history of sexual violence are 3.2 times, 95 % CI [1.92, 5.39], more likely to become pregnant or be involved in pregnancy before age 21 than individuals from comparison groups. Besides, a fail-safe population (Rosenberg, 2005) was computed to address the file drawer problem, and results showed that 544.6 null studies would be necessary to change the α value to a nonsignificant value.
Number of Sex Partners
Using the random-effects model, the adjusted pooled effect size Hedges g (Hedges & Olkin, 1985) was −0.34, 95 % CI [−0.63, −0.05], indicating a significant, albeit small effect size. Children and adolescents with a history of sexual violence are more likely to encounter a higher number of sex partners compared to individuals without such a history. The calculation of the fail-safe population (Rosenberg, 2005) revealed a number of 537.1 studies to reduce this effect to nonsignificance. When interpreting the results, the small underlying number of studies (n = 5) should be kept in mind.
Safer Sex
The adjusted pooled effect size Hedges g (Hedges & Olkin, 1985) was 0.21, 95 % CI [−0.26, 0.68], indicating no significant effect. Therefore, individuals with a history of sexual violence seem to engage in safer sex as often as comparison groups without a history of sexual violence. Four studies were included in the analyses.
Substance Use During Intercourse
The number of studies included was 4. The mean effect sizes indicate no significant group differences, OR = 1.0, 95 % CI [0.66, 1.51]. Odds ratios revealed that children and adolescents with a history of sexual violence used alcohol and drugs during intercourse as often as those without a history of sexual violence.
Transactional Sex
Using the random-effects model, the mean effect sizes indicate significant group differences. Odds ratios disclose that individuals with a history of sexual violence are 8.1 times, 95 % CI [3.79, 17.26], more likely to engage in transactional sex than individuals from comparison groups without a history of sexual violence. The calculation of the fail-safe population (Rosenberg, 2005) revealed that 56.34 studies would be necessary in order to bring the calculated significance level of the pooled effect near the critical significance level. Here, the small underlying number of studies of 3 should also be taken into account.
Discussion
Findings of the present meta-analysis support the conclusions of many studies by demonstrating that sexual violence was systematically related to higher rates of subsequent sexual risk factors, including involvement in teenage pregnancy, multiple sex partners, or engaging in teenage transactional sex. The only exceptions to these significant results were safer sex and substance use during intercourse, although the nonsignificant results for those analyses most likely stemmed from a lack of power.
These results are in line with Lalor and McElvaney’s (2010) review of numerous studies that examined the relationship between sexual violence and later engagement in sexual risk behavior in adulthood. Nevertheless, it cannot be said that sexual violence causes later sexual risk behavior. Whether deviant sexual behavior occurs after experiencing sexual violence or not seems to depend on a number of risk factors, such as family dysfunction or childhood emotional abuse (Lee et al., 2002). A large number of studies have shown an accumulation of risky sexual behaviors – such as frequently changing sexual partners, early onset of sexual activity, and unprotected sexual intercourse – in adult women after sexual violence experiences during childhood (Senn et al., 2008). However, other findings (e. g., Büttner et al., 2014) pointed to pronounced avoidance behavior with regard to sexuality in adult patients with posttraumatic stress disorder. The majority of findings supports an association between sexual violence and sexual risk behavior; however, the underlying mechanisms explaining this connection need further investigation.
Nevertheless, the present analysis shows an association between sexual violence and the development of sexual risk behaviors. A possible explanation for the association between sexual violence experiences and sexual risk behavior in childhood and adolescence is provided by Finkelhor and Browne’s (1985) traumagenic dynamics model. This eclectic but comprehensive model posits that the experience of sexual violence can be dissected in terms of four factors that cause trauma, also called “traumagenic dynamics.” These are traumatic sexualization, betrayal, powerlessness, and stigmatization. These dynamics distort children’s cognitive and emotional conception of the world, and elicit traumatization by vacillating children’s self-concept, world view, and affective capacities. For example, the dynamic of powerlessness detrimentally changes children’s sense of their ability to control their lives. Furthermore, the dynamic of stigmatization deteriorates children’s sense of their own worth and value. Trying to deal with the world through these distortions may precipitate the behavioral problems that are usually noticed in survivors of sexual violence (Finkelhor & Browne, 1985).
Consequently, the dynamic of traumatic sexualization might foster the development of dysfunctional sexual behavior scripts derived from rewarding negative sexual patterns during experiences of sexual violence (Senn et al., 2012). This may lead to the victim’s motivation to have a larger number of sexual partners, engage in risky sexual behaviors, or to have intercourse in exchange for material rewards (Senn et al., 2012). Another aspect of the model, the traumagenic dynamic of betrayal, is evoked when the child feels deceived by the perpetrator or by the reactions of attachment figures after finding out about the abuse, and may be related to the difficulty of trusting others. This, in turn, may lead to the rejection of stable, long-term relationships and favoring multiple and sporadic partners, since the possibility of re-experiencing betrayal is reduced in sporadic relationships (Senn et al., 2008).
In addition to the traumagenic dynamics model (Finkelhor & Browne, 1985), another explanation for an increased risky sexual behavior may be altered attachment patterns due to experiences of sexual violence (Kwako et al., 2010). Several general theories of childhood trauma, including those proposed by Briere (2002) and Cicchetti and Toth (2005), include mechanisms by which childhood maltreatment, including sexual violence, could lead to adult sexual risk behavior. These theories suggest that childhood maltreatment leads to insecure attachment, which could affect adult sexual risk behavior (Briere 2002; Cicchetti & Toth, 2005). Attachment disorders were found disproportionately often in victims of sexual violence, which can lead to insecure or even fearful and avoidant bonds in dyadic relationships (Kwako et al., 2010). Such a fearful or avoidant bond can result in a lack of communication about sexual needs and boundaries (Kwako et al., 2010) and, accordingly, could possibly encourage risky sexual behavior, such as refusing the use of condoms when no appropriate communication about the use of condoms takes place. Moreover, an insecure attachment style can manifest in dysfunctional romantic attachments, but may also result in sexual behavior serving to satisfy attachment needs instead of personal needs (Noll et al., 2019). Regarding teenage pregnancy and parenting, insecure attachment styles could amplify the need of having a child in order to compensate for damaged relationships with loved ones, as being a parent creates an unbreakable bond (Noll et al., 2019).
Another model, the differential effects model, holds that specific types of child maltreatment are linked to specific outcomes (Davis & Petretic-Jackson, 2000; Higgins & McCabe, 2001). In line with this model, literature reviews have found that children who experienced sexual violence consistently exhibit more sexualized behavior as children (Kendall-Tackett et al., 1993) and more sexual risk behavior as adults (Senn, et al. 2008), suggesting that sexual behavior may be a specific outcome related to sexual violence in childhood.
Limitations and Implications
These results should be interpreted with the limitations of the study in mind. The random-effects model was used to adjust the present heterogeneity of the studies. Nevertheless, construing the summarized effect sizes should take note of heterogeneity in order to draw overall conclusions. In addition, the overall risk of bias across studies was alarmingly high, which underlines the key problem of heterogeneity of the present studies in this field. However, we did not exclude such studies so as not to limit the number of included studies further. Thus, the risk of bias should be kept in mind when interpreting the individual study results. A second limitation involves the correlational nature of the data. Obviously, such data limit causal conclusions. Given the temporal sequence of sexual violence and risky sexual behavior, this limitation may be less concerning. Consequently, the limits may be less concerning in this context. Nevertheless, unexplored variables that may be related to sexual violence and risky sexual behavior (e. g., additional adverse experiences; Dong et al., 2003) should be taken into account in future investigations.
Concerning the drawing up of the search term, a further limitation is the absence of some significant keywords, which turned out to be relevant midst the screening process. The Boolean search term could be broadened by adding words such as “rape,” “condom,” “commercial sexual exploitation,” or “sexual peer victimization.”
Additionally, the mechanisms through which sexual violence causes sexual risk-taking that were discussed in this article are on the one hand hypothesized, and on the other hand they lack empirical evidence. Understanding the mechanisms underlying the relationship between sexual violence and sexual risk behavior remains a critical gap in both psychology and public health fields (Davis et al., 2018). Sexual violence and sexual risk behaviors may coexist due to certain attitudinal and trait characteristics (Davis et al., 2018). For example, impulsivity is associated with both (Davis et al., 2016; Zawacki et al., 2003), suggesting that certain constructs may underlie both sexual violence and sexual risk behavior. Future research should identify such underlying mechanisms and develop interventions that contain these shared mechanisms.
Furthermore, the restricted number of studies that met the inclusion criteria in the present meta-analysis was somewhat startling considering the large body of literature on sexual violence and subsequent risk factors. While qualitative reviews are able to make broad assessments of long-term sexual risk factors during adulthood (e. g., Senn et al., 2007), the present analysis was more constrained due to the restricted age range of childhood and adolescence. Thorlund and colleagues (2011) have warned that if meta-analyses are performed too early, before enough studies are available, there is a danger that incorrect conclusions may be drawn – in one direction or the other. Ideally, extremely underpowered meta-analyses should not be performed. Nevertheless, it is often difficult for researchers to know how many eligible studies provide data for any given outcome until most of the reviewing work has been performed, and at that stage withholding the results might lead to reporting bias. As Jackson and Turner (2017) state, in some settings, a lack of power is caused by high observed heterogeneity rather than by the number of studies, and this is even more difficult to predict in advance.
The relatively small number of studies that were eligible for the present analysis gives prominence to a potential weakness in the current literature on sexual violence and sexual risk factors during childhood and adolescence, and connotes an urgent need for more empirical studies. These studies should have a longitudinal design in order to identify developments and changes over time. In addition, they should establish a clear definition of sexual violence, consider and record other forms of violence in context, and operationalize and standardize sexual risk variables.
Another limitation is the absence of the indices for interrater agreements for all stages of the study assessment and data extraction processes, including the RoB assessments.
Conclusion
Overall, the available published evidence indicates that there is an association between sexual violence and sexual risk behavior and that it may have an effect on the whole process of development of children and adolescents, as it alters the children’s self-concept, world view, and emotional capacities permanently (Finkelhor & Browne, 1985). As such, it is of high importance to protect children and adolescents against sexual violence. Furthermore, another focus should be on the prevention of children and adolescents becoming victims of sexual violence in the first place.
Recommendations
With regard to the challenge of treating children and adolescents with experiences of sexual violence, valuable insights could be gained: Sexual violence is associated with the impairment of many areas of sexuality. A deeper understanding of how sexual violence can affect sexual problems increases the likelihood of effectively addressing the real causes of sexual complaints (Hall, 2008). Accordingly, it is important to explore and address the development of sexuality in children and adolescents and to create open communication. Hall (2008, p. 547) reports that those affected by sexual violence often feel like “damaged goods,” especially with regard to their sexuality. In particular, adolescents are often uncertain about what to expect and what is expected of them in terms of sexuality, when they have their first sexual contacts (Kavemann, 2016). These uncertainties can be reduced by encouraging them to identify their own sexual wishes and needs and, more importantly, their limits. In order to develop an adequate understanding of sexuality in children and adolescents with experiences of sexual violence, we need information about sexuality and guidance on reflection and open communication (Kavemann, 2016).
Literatur
(2012). Association between childhood sexual abuse and transactional sex in youth aging out of foster care. Child Abuse & Neglect, 36 (1), 75 – 80. https://doi.org/10.1016/j.chiabu.2011.07. 009
(2002). Comparative risk assessment: Child sexual abuse. Final report WHO Collaborating Centre on Evidence and Health Policy in Mental Health. St. Vincent’s Hospital, Sydney, Australia.
(2004).
Child sexual abuse . In M. EzzatiA. D. LopezA. RodgersC. J. L. Murray (Eds.), Comparative quantification of health risks: Global and regional burden of disease attributable to selected major risk factors, Vol. 2 (pp. 1851 – 1940). World Health Organization.(2001). Health risk behaviors and associated risk and protective factors among Brazilian adolescents in Santos, Brazil. Journal of Adolescent Health, 28 (4), 295 – 302. https://doi.org/10.1016/s1054-139x(00)00197-x
(2005). A meta-analysis of the relationship of child sexual abuse to HIV risk behavior among women. Child Abuse and Neglect, 29, 725 – 746. https://doi.org/10.1016/j.chiabu.2004.10.014
(2009). Introduction to meta-analysis. John Wiley & Sons.
(2007). Meta-analysis: Fixed effect vs. random effects (pp. 1 – 30). Meta-analysis.com. https://www.meta-analysis.com/downloads/Meta-analysis%20fixed%20effect%20vs%20random%20effects%20072607.pdf
(2016). Creating memories for false autobiographical events in childhood: A systematic review. Applied Cognitive Psychology, 31 (2) https://doi.org/10.1002/acp.3220
(2002).
Treating adult survivors of severe childhood abuse and neglect: Further development of an integrative model . In J. E. B. MyersL. BerlinerJ. BriereC. T. HendrixT. ReidC. Jenny (Eds.), The APSAC handbook on child maltreatment (pp. 175 – 202). Newbury Park, CA: SAGE.(2000). Impact of sexual abuse on the HIV-risk-related behavior of adolescents in intensive psychiatric treatment. American Journal of Psychiatry, 157 (9), 1413 – 1415. https://doi.org/10.1176/appi.ajp.157.9.1413
(2014). Trauma und sexuelle Störungen [Trauma and sexual disorders]. Psychotherapeut, 59 (5), 385 – 391. https://doi.org/10.1007/s00278-014-1068-y
(1996). Gender-specific outcomes for sexually abused adolescents. Child Abuse & Neglect, 20 (12), 1219 – 1231. https://doi.org/10.1016/s0145-2134(96)00117-2
(2005). Child maltreatment. Annual Review Clinical Psychology, 1, 409 – 438. https://doi.org/10.1146/annurev.clinpsy.1.102803.144029
(2018). The intersection of men’s sexual violence perpetration and sexual risk behavior: A literature review. Aggression and Violent Behavior, 40, 83 – 90. https://psycnet.apa.org/doi/10.1016/j.avb.2018.04.001
(2000). The impact of child sexual abuse on adult interpersonal functioning: A review and synthesis of the empirical literature. Aggression and Violent Behavior, 5 (3), 291 – 328. https://psycnet.apa.org/doi/10.1016/S1359-1789(99)00010-5
(2016). Distal and proximal influences on men’s intentions to resist condoms: Alcohol, sexual aggression history, impulsivity, and social-cognitive factors. AIDS and Behavior, 20 (1), S147 – S157. https://doi.org/10.1007/s10461-015-1132-9
(1994). Urinary catecholamine excretion in sexually abused girls. Journal of the American Academy of Child & Adolescent Psychiatry, 33, 320 – 327. https://psycnet.apa.org/doi/10.1097/00004583-199403000-00004
(2019). National crime victimization survey, 2018. https://bjs.ojp.gov/content/pub/pdf/cv18.pdf
(2003). The relationship of exposure to childhood sexual abuse to other forms of abuse, neglect, and household dysfunction during childhood. Child Abuse & Neglect, 27 (6), 625 – 639. https://doi.org/10.1016/s0145-2134(03)00105-4
(2001). Untangling the impact of sexual abuse on HIV risk behaviors among youths in foster care. AIDS Education and Prevention, 13 (4), 377 – 389. https://psycnet.apa.org/doi/10.1521/aeap.13.4.377.21427
(1985). The traumatic impact of child sexual abuse: A conceptualization. American Journal of Orthopsychiatry, 55 (4), 530 – 541. https://doi.org/10.1111/j.1939-0025.1985.tb02703.x
(2013). Global and regional estimates of violence against women: Prevalence and health effects of intimate partner violence and non-partner sexual violence. World Health Organization.
(2009). Risk factors for pregnancy among adolescent girls in Ecuador’s Amazon basin: A case-control study. Revista Panamericana de Salud Pública, 26, 221 – 228. https://doi.org/10.1590/S1020-49892009000900006
(2008). Attitudes toward using condoms and condom use: Difference between sexually abused and nonabused African American female adolescents. Behavioral Medicine, 34 (2), 45 – 54. https://doi.org/10.3200/bmed.34.2.45-54
(2008). Childhood sexual abuse and adult sexual problems: A new view of assessment and treatment. Feminism & Psychology, 18 (4), 546 – 556. https://psycnet.apa.org/doi/10.1177/0959353508095536
(2016). Childhood sexual abuse, teenage pregnancy, and partnering with adult men: Exploring the relationship. Journal of Psychosocial Nursing and Mental Health Services, 43 (8), 20 – 28. https://doi.org/10.3928/02793695-20050801-09
(2022). Community violence exposure and stress reactivity in African American and non-Latino White adolescents with overweight/obesity. Journal of Interpersonal Violence, 37(23 – 24), NP22784–NP22810 https://doi.org/10.1177/08862605211073091
(1985). Statistical methods for meta-analysis. Academic Press.
(2001). The development of the comprehensive child maltreatment scale. Journal of Family Studies, 7 (1), 7 – 28. https://psycnet.apa.org/doi/10.5172/jfs.7.1.7
(2001). Adverse childhood experiences and sexual risk behaviors in women: A retrospective cohort study. Family Planning Perspectives, 33 (5), 206 – 211.
(2010). Sexually abused children. Characterization of these girls when adolescents. International Journal of Adolescent Medicine and Health, 22 (2), 291 – 300. https://doi.org/10.1515/ijamh.2010.22.2.291
(2010). Sexual abuse and sexual risk behavior: Beyond the impact of psychiatric problems. Journal of Pediatric Psychology, 35 (5), 473 – 483. https://doi.org/10.1093/jpepsy/jsp111
(2000). Fixed effects vs. random effects meta‐analysis models: Implications for cumulative research knowledge. International Journal of Selection and Assessment, 8 (4), 275 – 292. https://doi.org/10.1111/1468-2389.00156
(2017). Power analysis for random‐effects meta‐analysis. Research Synthesis Methods, 8 (3), 290 – 302. https://doi.org/10.1002/jrsm.1240
(2016). Sexualpädagogik oder Gewaltprävention? – Sexualität vor dem Hintergrund sexueller Gewalterlebnisse [Sex education or violence prevention? – sexuality against the background of experiences of sexual violence]. Forum Gemeindepsychologie, 21 (1), 1 – 11.
(2000). Encyclopedia of psychology (Vol. 2, eds. by American Psychological Association). American Psychological Association.
(1993). Impact of sexual abuse on children: A review and synthesis of recent empirical studies. Psychological Bulletin, 113 (1), 164 https://doi.org/10.1037/0033-2909.113.1.164
(2013). Brain development during adolescence: Neuroscientific insights into this developmental period. Deutsches Ärzteblatt International, 110 (25), 425 https://doi.org/10.3238/arztebl.2013.0425
(2010). Childhood sexual abuse and attachment: An intergenerational perspective. Clinical Child Psychology and Psychiatry, 15 (3), 407 – 422. https://doi.org/10.1177/1359104510367590
(2010). Child sexual abuse, links to later sexual exploitation/high-risk sexual behavior, and prevention/treatment programs. Trauma, Violence, & Abuse, 11 (4), 159 – 177. https://doi.org/10.1177/1524838010378299
(2002). Developmental risk factors for sexual offending. Child abuse & Neglect, 26 (1), 73 – 92. https://doi.org/10.1016/s0145-2134(01)00304-0
(2001). Practical meta-analysis. SAGE publications.
(2017). Adverse experiences in childhood and sexually transmitted infection risk from adolescence into adulthood. Sexually Transmitted Diseases, 44 (9), 524 – 532. https://doi.org/10.1097/OLQ.0000000000000640
(1997). Sexual abuse history and number of sex partners among female adolescents. Family Planning Perspectives, 29 (5), 204 – 211.
(2007). Coerced first sexual intercourse and selected reproductive health outcomes among young women in KwaZuluNatal. South Africa. Journal of Biosocial Science, 39 (2), 231 – 244. https://doi.org/10.1017/s0021932006001325
(2012). Methamphetamine use is associated with childhood sexual abuse and HIV sexual risk behaviors among patrons of alcohol-serving venues in Cape Town, South Africa. Drug and Alcohol Dependence, 126(1-2), 232 – 239. https://doi.org/10.1016/j.drugalcdep.2012.05.024
(2001). Family relationships and adolescent pregnancy risk: A research synthesis. Developmental Review, 21, 1 – 38. https://psycnet.apa.org/doi/10.1006/drev.2000.0513
(2012). Child sexual abuse. Forensic Science International, 217(1-3), 1 – 4. https://doi.org/10.1016/j.forsciint.2011.08.006
(2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Annals of Internal Medicine, 151 (4), 264 – 269.
(2019). A risk of bias instrument for non-randomized studies of exposures: A users’ guide to its application in the context of GRADE. Environment international, 122, 168 – 184. https://doi.org/10.1016/j.envint.2018.11.004
(1995). Adverse factors associated with forced sex among southern adolescent girls. Pediatrics, 96 (5), 944 – 946.
(2015). Child maltreatment and sexual risk behavior: Maltreatment types and gender differences. Journal of Developmental and Behavioral Pediatrics, 36 (9), 708 https://doi.org/10.1097/dbp.0000000000000204
(2008). Sexual abuse of children—Unique in its effects on development? Child Abuse & Neglect, 32 (6), 603 – 605. https://doi.org/10.1016/j.chiabu.2007.09.008
(2019). Is sexual abuse a unique predictor of sexual risk behaviors, pregnancy, and motherhood in adolescence? Journal of Research on Adolescence, 29 (4), 967 – 983. https://doi.org/10.1111/jora.12436
(2013). Teen birth rates in sexually abused and neglected females. Pediatrics, 131 (4), e1181 – e1187. https://doi.org/10.1542/peds.2012-3072
(2009). Childhood sexual abuse and adolescent pregnancy: A meta-analytic update. Journal of Pediatric Psychology, 34 (4), 366 – 378. https://doi.org/10.1093/jpepsy/jsn098
(2019). Predictors of teenage pregnancy among girls aged 13 – 19 years in Uganda: A community based case-control study. BMC Pregnancy and Childbirth, 19 (1), 1 – 14. https://doi.org/10.1186/s12884-019-2347-y
(2008). Childhood abuse as a risk factor for adolescent pregnancy in El Salvador. Journal of Adolescent Health, 42 (6), 580 – 586. https://doi.org/10.1016/j.jadohealth.2007.11.148
(2001). A meta-analysis of the published research on the effects of child sexual abuse. Journal of Psychology, 135, 17 – 36. https://doi.org/10.1080/00223980109603677
(2020). Associations of physical partner violence and sexual violence victimization on health risk behaviours and mental health among university students from 25 countries. BMC Public Health, 20 (1), 1 – 10. https://doi.org/10.1186/s12889-020-09064-y
(2004). Sexual power and HIV risk, South Africa. Emerging Infectious Diseases, 10 (11), 1996 – 2004. https://doi.org/10.3201/eid1011.040252
(2000). The relationship between sexual abuse and sexual risk among high school students: Findings from the 1997 Massachusetts Youth Risk Behavior Survey. Maternal and Child Health Journal, 4, 125 – 134. https://doi.org/10.1023/a:1009526422148
(2005). The file‐drawer problem revisited: A general weighted method for calculating fail‐safe numbers in meta‐analysis. Evolution, 59 (2), 464 – 468.
(1996). Sexual abuse history and associated multiple risk behavior in adolescent runaways. American Journal of Orthopsychiatry, 66 (3), 390 – 400. https://doi.org/10.1037/h0080189
(2004). Teenage pregnancy and associated risk behaviors among sexually abused adolescents. Perspectives on Sexual and Reproductive Health, 36 (3), 98 – 105. https://doi.org/10.1363/psrh.36.98.04
(2017). Child sexual abuse and HIV-related substance use and sexual risk across the life course among males and females. Journal of Child Sexual Abuse, 26 (5), 519 – 534. https://doi.org/10.1080/10538712.2017.1319004
(2011). Self-defining as sexually abused and adult sexual risk behavior: Results from a cross-sectional survey of women attending an STD clinic. Child Abuse & Neglect, 35 (5), 353 – 362. https://doi.org/10.1016/j.chiabu.2011.01.013
(2012). Mediators of the relation between childhood sexual abuse and women’s sexual risk behavior: A comparison of two theoretical frameworks. Archives of Sexual Behavior, 41 (6), 1363 – 1377. https://doi.org/10.1007/s10508-011-9897-z
(2008). Childhood and adolescent sexual abuse and subsequent sexual risk behavior: Evidence from controlled studies, methodological critique, and suggestions for research. Clinical Psychology Review, 28 (5), 711 – 735. https://doi.org/10.1016/j.cpr.2007.10.002
(2007). Characteristics of sexual abuse in childhood and adolescence influence sexual risk behavior in adulthood. Archives of Sexual Behavior, 36 (5), 637 – 645. https://doi.org/10.1007/s10508-006-9109-4
(1998). Gender differences in risk behaviors associated with forced or pressured sex. Archives of Pediatrics & Adolescent Medicine, 152 (1), 57 – 63. https://doi.org/10.1001/archpedi.152.1.57
(2004). A critique of international indicators of sexual risk behaviour. Sexually Transmitted Infections, 80 (2), 13 – 21. https://doi.org/10.1136/sti.2004.011635
(1997). Adolescent pregnancy and sexual risk-taking among sexually abused girls. Family Planning Perspectives, 200 – 227.
(2012). Father–daughter incest: Data from an anonymous computerized survey. Journal of Child Sexual Abuse, 21 (2), 176 – 199. https://doi.org/10.1080/10538712.2012.654007+D153
(2013). Sister–sister incest: Data from an anonymous computerized survey. Journal of Child Sexual Abuse, 22 (6), 695 – 719. https://doi.org/10.1080/10538712.2013.811140
(2007). Selling sex in a population-based study of high school seniors in Sweden: Demographic and psychosocial correlates. Archives of Sexual Behavior, 36 (1), 21 – 32. https://doi.org/10.1007/s10508-006-9083-x
(2013). Adolescents selling sex: Exposure to abuse, mental health, self-harm behaviour and the need for help and support. A study of a Swedish national sample. Nordic Journal of Psychiatry, 67 (2), 81 – 88. https://doi.org/10.3109/08039488.2012.679968
(2011). The number of patients and events required to limit the risk of overestimation of intervention eGects in meta-analysis. A simulation study. PLOS One, 6 (10),
e25491 . https://doi.org/10.1371/journal.pone.0025491(2012). Working with men and boy survivors of sexual and gender-based violence in forced displacement. https://www.refworld.org/pdfid/5006aa262.pdf
(1995). Risk factors associated with date rape and sexual assault of adolescent girls. Journal of Adolescence, 18, 289 – 306. https://doi.org/10.1006/jado.1995.1020
(2008). An examination of risky sexual behavior and HIV in victims of child abuse and neglect: A 30-year follow-up. Health Psychology, 27 (2), 149 – 158. https://doi.org/10.1037/0278-6133.27.2.149
(2001). Risk factors and life processes associated with teenage pregnancy: Results of a prospective study from birth to 20 years. Journal of Marriage and Family, 63 (4), 1170 – 1184. https://doi.org/10.1111/j.1741-3737.2001.01170.x
(2003). Guidelines for medico-legal care of victims of sexual violence. https://apps.who.int/iris/bitstream/handle/10665/42788/924154628X.pdf;jsessionid=67428848C6E9991339A18D70BC229FF5?sequence=1
(2003). A profile of adolescent females with a history of sexual assault in Texas: Familial environment, risk behaviors, and health status. Journal of Pediatric and Adolescent Gynecology, 16 (4), 207 – 216. https://doi.org/10.1016/s1083-3188(03)00123-2
(2003). Perpetrators of alcohol-involved sexual assaults: How do they differ from other sexual assault perpetrators and nonperpetrators? Aggressive Behavior, 29 (4), 366 – 380. https://doi.org/10.1002/ab.10076
