Gender and Race Informed Pathways From Childhood Sexual Abuse to Sexually Transmitted Infections: A Moderated Mediation Analysis Using Nationally Representative Data

Abstract

Purpose:

Although data suggest child sexual abuse is linked with increased risk of contracting asexually transmitted infection (STI), the mechanisms through which these experiences are connected remain understudied. Moreover, there is a need to explore how race/ethnicity and gender influence these processes.

Methods:

The present study examined the mediational pathways from child sexual abuse to risky sexual behavior to STIs and further evaluated the role of depressive symptomatology and nonmedical prescription drug use on the relationship between child sexual abuse and risky sexual behavior. In addition, race and gender were examined as moderators to account for potential different effects of these mechanisms on females and males and on different racial and ethnic groups. A nationally representative sample of 4,181 youth from the Add Health dataset was used.

Results:

Results from a moderated mediation model indicated risky sexual behavior partially mediated the pathway from child sexual abuse to STI contraction and depressive symptomatology and nonmedical prescription drug use partially mediated pathway from child sexual abuse to risky sexual behavior. Race and gender moderated the relationship between risky sexual behavior and STI contraction.

Conclusions:

Findings underscore the need for STI prevention efforts among adolescents to focus on risk factors beyond risky sexual behaviors, such as childhood sexual abuse and mental health screening that includes depressive symptomatology and nonmedical prescription drug use. In addition, findings emphasize the need to further examine the different effects on different racial/ethnic and gender subgroups, particularly black women.

Among the U.S. population aged 15–24 years, approximately 10 million new cases of sexually transmitted infections (STIs) occur annually [1,2]. There are racial/ethnic disparities in rates of STIs that affect individuals identified as black/African-American and Latinx/Hispanic; this disparity is greater among females [3]. STIs among the U.S. adolescent and young adult population are typically contracted through risky sexual behaviors (RSBs) such as engaging in sexual intercourse without using a condom or engaging in sexual activity while under the influence of alcohol or other drugs [4]. Although adolescents who engage in RSBs do so for various and multiple reasons, research has increasingly focused on the link between RSB to child sexual abuse (CSA). Studies have shown that CSA exposure is associated with early sexual debut [5,6] and other RSBs including having multiple sexual partners [7], engaging in multiple-person sex [6,8], and exchanging sex for money or drugs [5,6,9]. Although the adolescent health literature supports the links between CSA and RSB and between RSB and STI, few studies have examined the relationship between CSA to STI through RSB (i.e., explicitly modeling the CSA-RSB-STI pathway).

Research on CSA suggests that mental health problems such as depressive symptomatology are common among survivors, including adolescent and young adult survivors [10–13]. Studies have also demonstrated adolescents with mental health problems have higher rates of STIs compared with those without reported or diagnosed mental health issues [5,7,14]. Mental health therefore potentially mediates the relationship between CSA and STI. Specifically, for the current pathway of interest (CSA-RSB-STI), mental health likely mediates the relationship between CSA and RSB. A review of the literature on CSA by Lalor and McElvaney [15] reported various models that attempt to explain the relationship between CSA and RSB, proposing the inclusion of mental health symptom presentation as a mediator of this association.

Depressive symptomatology is a well-known mental health consequence of CSA [10,13]. What is less well understood is the mediational role of depressive symptomatology in the association between CSA and RSB. It has been hypothesized that those with depressive symptoms might seek to avoid rejection; thus, their ability to negotiate safer sex might be compromised, thereby increasing their risk of engagement in RSBs [16–18]. Another pathway through which CSA likely increases engagement in RSBs and contracting STIs is through drug abuse, as prior research has observed that abuse survivors often turn to drugs and substances as a coping mechanism [19,20]. One substance that has recently emerged as problematic is nonmedical prescription drug use, which is associated with increased RSBs [21–23]. For example, a study of nationally representative sample of youth found that nonmedical prescription drug use was positively associated with a range of RSB, even when controlling for the use of other substances [24]. Similarly, in a local sample of youth (aged 14–20 years), nonmedical prescription drug use increased the likelihood of inconsistent condom use and engaging in sexual activity while under the influence of alcohol and other illicit substances [25]. As individuals exposed to CSA are more likely to use prescription medication in a nonmedical capacity as a coping mechanism [19,20], they might be at increased risk for engaging in RSB, and consequently, for contracting STIs.

There are major racial/ethnic differences in rates of RSBs, particularly when comparing black and Latinx/Hispanic Americans to white/European Americans [26]. Thus, scholars have suggested including race and ethnicity as potential moderators in studies about RSBs [26]. With respect to gender, a recent meta-analysis indicated that females with substantiated CSA were significantly more likely to engage in RSB compared with their male counterparts [27]. Similarly, in a longitudinal study with a nationally representative sample, CSA was positively associated with STI self-report among females, whereas this association was not statistically significant among males [14]. Therefore, gender might also play a moderating role in the associations between CSA and STIs.

In this study, we use the National Longitudinal Study of Adolescent to Adult Health (Add Health) to conduct a moderated mediation analysis of the pathways from CSA to STIs. Figure 1 illustrates our proposed model, which stipulates that [1] RSB mediates the relationship between CSA and STI (i.e., CSA-RSB-STI); [2a] mental health and [2b] nonmedical prescription drug use mediate the relationship between CSA and RSB; and that [3] race/ethnicity and gender moderate the pathways.

Figure 1.

Conceptual moderated mediation model of the pathway from child sexual abuse to subsequently contracted sexually transmitted infections.

Method

Data for this study come from Add Health, a nationally representative sample of youth followed over 15 years, with the first wave of data collected from youth in grades 7–12 during the 1994–1995 school year, and successive data collections occurred in 1996 (Wave II), 2001–2002 (Wave III), and 2008–2009 (Wave IV). Add Health collects data pertaining to social, economic, psychological, and physical well-being, along with demographic and other information. This study’s analytical sample included participants with complete data on the main independent (CSA; missing n = 25) and dependent (STI; missing n = 728) variables, Latinx (missing n = 14), sexual orientation (missing n = 43), child physical abuse (missing n = 148), income (missing n = 847) and Wave IV weights (missing n = 19), resulting in 4,181 cases. The present study was deemed exempt from the authors’ university’s institutional review board.

Measures

Demographic information and covariates.

The present study controlled for demographic characteristics and other variables shown in previous research to have an association with STI such as sexual identity/orientation [28] and physical abuse [26]. Demographic characteristics such as gender, age, socioeconomic status, race/ethnicity, and sexual orientation were included in the present study. Gender/sex was measured dichotomously from Wave I using the follow-up question to “What is your sex?”: “Interviewer, please confirm that [the participant]’s sex is (male) female. (Ask if necessary.)” Responses were coded as female or male². Age at Wave III was used and was included as a continuous variable. Annual household income at Wave I was used as a continuous variable indicating socioeconomic status. Race was coded from Wave I data as non-Hispanic black (black) or non-Hispanic white (white). Ethnicity was coded separately from race as Latinx/Hispanic American (Latinx) or non-Latinx/Hispanic American. Intersectional identity categories were not coded for racial/ethnic grouping due to limited power³. Finally, sexual identity was coded dichotomously using participant responses at Wave III and/or Wave IV, with participants identifying themselves as mostly heterosexual, bisexual, mostly homosexual, or 100% homosexual coded as gay, lesbian, or bisexual, and those identifying as 100% heterosexual coded as straight/heterosexual¹. Childhood physical abuse was measured using two items assessing how often participants had been slapped, hit, or kicked by a parent or adult caregiver before sixth grade (Wave III) and before age 18 years (Wave IV). Items were recoded as “0” if no physical abuse and “1” if participant had experienced physical abuse at least one time.

Childhood sexual abuse.

A dichotomous variable was created for childhood sexual abuse or sexual assault based on four items using retrospective report in Waves III and IV. The items for CSA assessed how often participants had been touched by a parent or other adult caregiver in a sexual way [1] prior sixth grade (Wave III) and [2] before age 18 years (Wave IV) and whether they have ever been forced in a [3] physical or [4] nonphysical way to have any type of sexual activity against their will before age 18 years (both items assessed in Wave IV). Because there were two few instances of cases with multiple CSA reports and responses to waves III and IV questions could have been about the same instance of CSA, a dichotomous variable (ever vs. never) was created, indicating if participants reported at least one instance of sexual abuse or assault in one of the four items.

Depressive symptomatology.

Similar to prior research using Add Health, symptoms of depression were measured using nine items from the Center for Epidemiologic Studies Depression Scale [29]. The Center for Epidemiologic Studies Depression Scale is administered regarding symptoms over the past week; the score from Wave III was used in the present study, as we used this measure as a mediator between CSA and STIs assessed in the adulthood (Wave IV), range 0–27.

Nonmedical prescription drug use.

Nonmedical prescription drug use was measured in Wave III using a question asking whether (yes/no) the participant has taken any of the following drugs without a doctor’s permission (since Wave I): sedatives, tranquilizers, stimulants, pain killers. A continuous variable was created; range 0–4.

Risky sexual behavior.

A continuous measure of RSB assessed in Wave III was developed using five different items indicating RSB. These included engaging in sexual activity with multiple partners over the past year, having condomless sexual intercourse, whether they had paid someone to engage in sexual activity, whether they had engaged in prostitution, and whether they had engaged in sexual activity while under the influence of alcohol or other drugs or had sex with someone who had used intravenous drugs. All items were coded dichotomously (yes/no), and RSB was calculated based on the sum of all items, range 0–5.

Sexually transmitted infections.

STIs were measured as a continuous variable using the total number of reported sexually transmitted diseases at Wave IV. To calculate this value, participants were queried about 13 STIs separately (e.g., chlamydia, genital warts, and syphilis) and asked to respond to each question, range 0–6.

Data analysis

All analyses were completed using SPSS Version 24. Descriptive characteristics are presented by gender and race/ethnicity in Table 1. All possible pairwise correlations and variance inflation factor collinearity statistics were examined, indicating that there was no multicollinearity (Table 2). Data were further examined for normality and outliers. Socioeconomic status/family income was leptokurtic and negatively skewed due to two outlier cases; no other problems were observed. The income variable was logarithmically transformed, but the variable remained kurtotic and skewed. Cases identified as outliers were then removed, and the variable still was kurtotic and skewed. Therefore, models were run with the outliers included and with them omitted; model results remained consistent. Missing data² were imputed using multiple imputation [30]. Wave IV weighting was used during all analyses.

Table 1.

Means (M) and standard deviations (SDs) of selected demographic and health variables across racial/ethnic groups within gender

	Females, n = 2,329			Males, n = 1,852			Significant main effects
	AA, M (SD), n = 564	LA, M (SD), n = 211	EA, M (SD), n = 1,581	AA, M (SD), n = 400	LA, M (SD), n = 194	EA, M (SD), n = 1,286	Identifier
Prescrip Drugs	.27 (.61)	.38 (.78)	.46 (.99)	.27 (.68)	.38 (.90)	.68 (1.21)	AA**
Depression	8.17 (3.27)	8.24 (3.23)	8.19 (2.86)	7.66 (2.67)	7.76 (3.33)	7.53 (2.39)	Gender*
Risky sexual beh	.87 (.82)	.78 (.67)	.82 (.67)	1.00 (.98)	.85 (.80)	.82 (.76)	AA**
STIs	1.02 (1.29)	.40 (.79)	.40 (.78)	.46 (.96)	.27 (.74)	.17 (.65)	AA***; Gender***
Income ($10K)	37.30 (51.92)	36.47 (54.15)	54.24 (63.17)	40.51 (48.47)	36.74 (30.02)	51.01 (57.69)	AA***; LA**
Wave I age	15.82 (1.77)	15.83 (1.73)	15.86 (1.77)	16.03 (1.80)	16.08 (1.81)	16.04(1.74)	Gender**

AA = African-American/black race; EA = European-American/white race; LA = Latinx/Hispanic-American; Income = value/1,000; Prescrip drugs = number of nonmedical prescription drugs used; Risky sexual beh = number of risky sexual behaviors; STIs = number of sexually transmitted infections reported.

^***p < .001;

^**p < .010;

^*p < .050.

Table 2.

Pearson correlations by gender and race

	1	2	3	4	5	6
Bivariate correlations by gender
1. Prescription drugs	–	.146***	.205***	.048*	.026	−.010
2. Depression symptoms	.093***	–	.137***	.131***	−.009	−.037
3. Risky sexual behavior	.173***	.092***	–	.295***	−.041*	.086***
4. Count of STIs	.046*	.064*	.194***	–	.004	−.002
5. Income	.066**	−.020	.008	−.019	–	.006
6. Age	−.014	−.011	.051*	−.001	.028	–
Bivariate correlations by race
1. Prescription drugs	–	.110***	.168***	.058*	−.013	−.037
2. Depression symptoms	.114***	–	.119***	.166***	−.062*	.016
3. Risky sexual behavior	.219***	.114***	–	.272***	−.035	.125***
4. Count of STIs	.065***	.112***	.212***	–	.018	.007
5. Income	.046***	.008	−.007	.042*	–	.024
6. Age	−.012	−.034	.049**	.013	.005	–

For gender correlations, female values above dashed lines; male values below dashed lines. For race correlations, African-American values above dashed lines; European-American values below dashed lines.

STIs = sexually transmitted infections.

^*p < .05;

^**p < .01;

^***p < .001.

To examine whether and how CSA and potential moderators influenced rates of STI through the proposed pathways over time, we used moderated mediation analysis. Use of this special case of structural equation modeling is supported for such analyses examining underlying mechanisms of health behaviors [31]. This model examined the relationship from CSA exposure to STIs included RSB as a mediator between CSA and STI; depressive symptomatology and nonmedical prescription drug use as mediators between CSA and RSB; and race/ethnicity and gender as moderators of the relationship between RSB and STI². Age, Latinx/Hispanic ethnicity, history of physical abuse, and sexual orientation were included in the model as control variables. The full model included the following mediated pathway from CSA to STI: CSA to depressive symptomatology and nonmedical prescription drug use, RSB, and STI. This model was run using a bootstrapping procedure with 10,000 samples using the SPSS version of the PROCESS macro [32]. Furthermore, the model was customized to allow for further examination of moderated pathways by race and gender.

Results

Descriptive characteristics of the study sample are presented in Tables 1 and 2. Results from a multivariate analysis of variance with race and ethnicity (black, white, and Latinx) and gender as predictors of included study variables indicated black race had a main effect on multiple outcomes, including less nonmedical prescription drug use (p = .007), more RSB (p = .001), and more STIs (p < .001) compared with Latinx and white participants, and lower household income than whites (p < .001). The main effect of Latinx ethnicity was also noted for household income compared with participants identified as black and white (p =.001). There was also a main effect of gender on select outcomes such that females had more depressive symptomatology (p =.001), more STIs (p < .001), and younger age (p =.001), and black and Latinx female participants had lower RSB (p < .005). Race by gender and Latinx ethnicity by gender interactions were not observed. With respect to childhood physical and sexual abuse, a series of Three-Way Chi-Square Tests for Independence indicated Latinx females were significantly more likely to report both forms of abuse compared with black and white participants (p < .001), and white females reported significantly higher rates of abuses than did black females (p < .001; Table 3).

Table 3.

Child sexual abuse and child physical abuse within gender by race/ethnicity

	Female child sexual abuse		Male child sexual abuse		χ2	p	OR (CI)
	n (%)	OR (CI)	n (%)	OR (CI)
Total sample	461 (19.8)	.71 (.67–.75)	162 (8.7)	1.83 (1.59–2.10)	99.28	<.001	.39 (.32–.47)
AA (non-LA)	104 (18.4)	.87 (.76–.99)	54 (13.5)	1.26(1.00–1.58)	4.16	.041	1.34 (1.03–1.75)
LA	44 (20.9)	.73 (.59–.89)	21 (10.8)	1.58 (1.09–2.27)	7.54	.006	.46 (.26–.81)
EA (non-LA)	308 (20.6)	.66 (.62–.70)	84 (7.0)	2.26 (1.86–2.74)	98.08	<.001	.29 (.23–.38)
	Female child physical abuse		Male child physical abuse		χ2	p	OR
	n (%)	OR (CI)	n (%)	OR (CI)
Total sample	817 (35.1)	1.12 (1.05–1.18)	755 (40.8)	.88 (.82–.94)	14.22	<.001	1.27 (1.12–1.44)
AA (non-LA)	190 (33.7)	1.13 (1.01–1.27)	162 (40.5)	.85 (.73–.98)	4.69	.030	1.34 (1.03–1.75)
LA	92 (43.6)	1.07 (.89–1.30)	92 (47.4)	.92 (.75–1.13)	.60	.440	1.17 (.79–1.73)
EA (non-LA)	509 (34.0)	1.11 (1.04–1.20)	473 (39.6)	.88 (.81–.96)	8.88	.003	1.27 (1.09–1.49)

AA = African-American; CI = confidence interval; EA = European-American; LA = Latinx-American; OR = odds ratio.

Among male participants, those identifying as black reported significantly more STIs and risky sex (p < .001) and less nonmedical prescription drug use (p < .001), compared with both Latinx and white participants. Males who identified as Latinx had the highest levels of depression symptoms (p < .001) and lowest family income levels (p < .001). Childhood sexual abuse rates were highest among males who identified as black, and childhood physical abuse rates were highest among males who identified as Latinx (p < .001). Rates of CSA were consistently higher among female participants compared with male participants (within racial/ethnic grouping), whereas rates of childhood physical abuse were higher among male participants within racial/ethnic group.

Tested model of the relationship between CSA and STI

A moderated mediation analysis with multiple mediators was run with STI as the outcome variable (F(11, 4158) = 67.90, p < .001, R² = .15; Figure 2). After entering the covariates into the full model (i.e., age, history of childhood physical abuse, family income level, Latinx ethnicity, sexual orientation), CSA positively and directly predicted STI (B = .04, p < .010). Furthermore, depression symptoms, nonmedical prescription drug use, and RSB partially mediated the relationship between CSA and STI. The effect of mediation was significant for all three of the tested mediation pathways and was significant for both males and females and black and non-black participants.

Figure 2.

Moderated mediation model of the pathway from child sexual abuse to subsequently contracted sexually transmitted infections. **p < .01; ***p < .001.

Symptoms of depression significantly mediated the relationship between CSA and STI through RSB for female black: confidence interval (.001, .003), female non-black participants (.001, .004), and for male non-black participants (.001, .002) but was not significant for male black participants (.00, .01). Indices of partial moderated mediation were significant for gender (−.002, −.001) and Black race (.001, .004).

Nonmedical prescription drug use significantly mediated the pathway from CSA to RSB to STI. This effect was observed for all groups except for male non-black participants: female black participants (.001, .004), female non-black participants (.001, .006), and for male black participants (.001, .003). Indices of partial moderated mediation were significant for gender (−.004, −.001) and black race (.001, .002).

There was also evidence that gender (−.03, −.01) and black race (.001, .02) significantly moderated the pathway from CSA to STI through RSB. Specifically, this mediating pathway was significant for female black participants (.01, .03), female non-black participants (.02, .05), and male non-black participants (.01, .02) but not for male black participants (.00, .01). Overall, these moderation effects indicated that identifying as a black female was associated with a more robust influence of RSB on STI than other identifiers, and white females also had a strong pathway from RSB to STI (p < .001). Individuals identified as male had less robust pathways from RSB to STI; the strength of this association was significant for black males (p < .001) but was only marginally significant for white males (p = .072). Accordingly, identifying as female is associated with a stronger influence of RSB on STI.

Discussion

Using population-based longitudinal data, our study examined a moderated mediation model of the pathway from CSA to depression symptoms and nonmedical prescription drug use to RSBs to STIs. In addition to confirming results of prior research indicating that RSB mediates the relationship between CSA and STIs, results from the present study indicated that the relationship between CSA and RSB was mediated by levels of depression symptoms and nonmedical prescription drug use, and that gender and race moderated the association between CSA to STI through RSBs. Furthermore, this study expands understanding of the mechanisms through which disparities in STI rates may occur.

The results regarding depression symptomatology as a mediator of the relationship between CSA and RSB add to the scarce but growing body of evidence of the mediating role of mental health problems between CSA and later health consequences. Another study, which used the Adverse Childhood Experiences data, found depression symptomatology mediated the relationship between childhood adversities and ischemic heart disease [33]. In a study using the Ontario Health Survey, history of mental health problems strongly mediated the relationship between childhood abuse and poorer adult physical health [34]. Altogether, symptoms of depression appear to be an important mental health factor in studies of adult survivors of CSA and other traumatic childhood experiences.

How drug abuse, and particularly nonmedical prescription drug use, influences the relationship between CSA and RSB is not well understood, despite evidence of a relationship between RSB and nonmedical prescription drug use [21,23]. In the present study, we observed a mediational pathway from CSA to RSB through nonmedical prescription drug use. This effect validates the need to further explore how nonmedical prescription drug use can uniquely influence the relationship between CSA and RSB [24]. These results indicate that interventions to reduce RSB and STIs should screen for history of CSA and should include assessments of both symptoms of depression and nonmedical prescription drug use, and results of this study validate the need to further explore how nonmedical prescription drug use can uniquely influence the relationship between CSA and RSB.

Examination of moderating effects of gender and race indicated that the pathway between CSA and STI (through RSB) is different between females and males, between black and non-black participants and between groups of different gender and racial/ethnic backgrounds. Results regarding moderation by gender suggest females with a history of CSA, particularly black females, are at a higher risk for STI via engagement in RSBs than their male and white counterparts. The racial/ethnic group analyses represent one of the few studies that examine the differential effects of racial/ethnic background on the relationship of interest, and to our knowledge, this is the first known study to present such an intersectional analysis of race and gender in the pathway between CSA and STI risk among adolescents. Altogether, findings from the current indicate that disparities in STI rates may occur as a result of other observed disparities, including rates of RSB, but emphasize that gender meaningfully can influence the relationship between race/ethnicity and STIs.

Limitations and future directions

As the present study used an existing dataset, it has some limitations. There were many cases for which data were missing. However, given the longitudinal nature of the dataset, it is impressive that so many data points were available for most participants. Apart from depression symptomatology [29], no standardized measures used in the present study were available from the Add Health survey. In particular, standardized measures were unavailable for substance use or abuse and RSB. It is recommended that future studies replicate the tested model using standardized measures of these constructs.

Analyses for this study used self-reported data, including data regarding gender/sex, CSA experiences, and STI contraction. Gender/sex identification was not clearly defined for participants (especially at Wave I); therefore, it is unknown which construct participants reported regarding their biological sex/gender. However, given the inclusion of only two genders and lack of different gender categories (e.g., agender, gender unknown, nonbinary), it appears that all items in the Add Health survey are measuring sex/gender dichotomously. It is worth mentioning that although gender was a noteworthy variable in the present study, gender is not synonymous with sex. Furthermore, as these two separate constructs have different associations with sexual risk [35,36], particularly for the transgender community [37], future research on gender and sex and sexual risk and STI contraction is strongly encouraged to ensure which construct is being measured. Future research should account for more nuanced definitions of sex versus gender. The question used to assess for CSA does not clearly describe the types of “sexual activity”; therefore, the study was unable to assess for differential effects of penetrative versus nonpenetrative sexual activities. In addition, the STI measure may be a proxy of STI help-seeking, as the questions used in the present study were related to official medical diagnoses. Indeed, some individuals are less likely to seek medical care; STI rates should be interpreted as an approximation that is potentially low. As participants may not be aware of their previous diagnoses, they may not have gotten tested for STIs or may feel uncomfortable sharing such personal information. Moreover, there are differences by STIs for some social groups—syphilis and gonorrhea cases are significantly more prevalent among individuals identified as low socioeconomic studies, black, and LGBTQ [38]. Accordingly, future research should use sexual network differences or social geographical differences between population segments in their analysis. For medical or physical conditions such as STIs, it would be preferable to obtain such data using more valid methodology. However, although STI and other biomarkers were collected at Wave III, the longitudinal design of our study required us to look at STIs from Wave IV data collection for which biomarker data were not available.

Finally, similar to many other studies examining CSA and RSBs, the present study was limited demographically, particularly regarding racial/ethnic identification and sexual orientation. There were too few persons of racial/ethnic backgrounds such as Asian-American, Native American/American-Indian and too few individuals who identified as gender/sexual minorities to use these designations in the study. Future studies are strongly encouraged to include, and possibly oversample, these and other understudied populations.

Although studies have focused on the impact of CSA on sexual health, there is a lack of understanding of the mechanisms that link CSA to negative sexual health outcomes such as STIs. This is one of the few studies that examined the pathway from CSA to STI through RSB and further delineated the pathway from CSA to depressive symptomatology and nonmedical prescription drug use to RSB. The longitudinal dataset from Add Health was used, allowing for more comprehensive assessment of the relationships and mediational pathways from CSA to STIs across individuals’ transition to adulthood and young adulthood. Such work advances understanding of the impact of CSA on sexual health over the life course. In addition, using this nationally representative sample, the present study overcame some limitations regarding the generalizability of findings of prior research that have used mostly clinical samples [27] and community cohorts [7]. Furthermore, through inclusion of race/ethnicity and gender as potential moderators of the pathway from CSA to STI contraction, understanding of the complex interactions within these pathways is enhanced, thereby better informing future intervention efforts that aim to mitigate the iatrogenic effects of CSA.

Altogether, findings from the moderated mediation analysis conducted in the present study indicate that intervention to reduce RSBs and STIs should screen for history of CSA and include assessments of both symptoms of depression and nonmedical prescription drug use. Furthermore, it is strongly encouraged that research extending the scope of the present study account for the different effects among groups of individuals identified as different races/ethnicities and genders and continue to expand examination of intersectional identification categories in future work. This is particularly important in areas of research with distinct disparities, such as STI and substance use fields. The model tested in the present study elucidates contributing factors to the STI rate disparity and provides a framework for follow-up work.

Finally, this longitudinal dataset enabled an evaluation of the relationships and mediational pathways from CSA to STIs across individuals’ transition to adulthood and young adulthood, advancing the understanding of CSA’s impact on sexual health over the life course. Through a more comprehensive examination of race/ethnicity and gender as potential mediators between CSA and STIs, the complex interactions within these pathways are more clearly elucidated, thereby better informing future intervention efforts that aim to mitigate the negative effects of CSA.

IMPLICATIONS AND CONTRIBUTION.

Observed pathways from child sexual abuse to sexually transmitted infections (STIs) across adolescents’ transition to adulthood provide important insights into the mechanisms of how child sexual abuse impacts individuals’ sexual health over the life course. Findings underscore the need for screening and intervention targeting mental health sequelae among abuse survivors to reduce STIs.