Age of occurrence and severity of childhood sexual abuse: impacts on health outcomes in men who have sex with men and transgender women

Abstract

Childhood sexual abuse (CSA) is associated with a wide range of health outcomes and is more prevalent among men who have sex with men (MSM) compared to men who have sex with women exclusively and among transgender women (TW) compared to cisgender peers; however, there is a lack of consensus regarding an explanatory theoretical model. This analysis examines these models across health outcomes using baseline data from a longitudinal study of MSM and TW in the Chicago (n=1,035) collected 2015 to 2019. Severity of CSA was measured for two age ranges (prior to 13 and 13–17). Logistic regression and negative binomial regressions were estimated. Teenage experiences of CSA were associated with alcohol problems, cannabis problems, suicide ideation/attempt, depression, condomless anal sex partners, rectal STI, and HIV. Pre-teenage CSA was associated with alcohol use and depressive symptoms. Significant interactions across age of experience of CSA were found for alcohol problems, cannabis problems, and condomless anal sex partners. Consistent with previous literature, this analysis found CSA to be a significant influence on various health outcomes. No single explanatory framework emerged; however, adolescent exposures may be more closely linked to health outcomes and risk behaviors than pre-teenage or cumulative exposures.

Introduction:

Childhood sexual abuse (CSA) causes detrimental health effects for survivors, including a wide range of negative physical, behavioral, and mental health outcomes (Maniglio, 2009). Furthermore, CSA is more prevalent among men who have sex with men (MSM) compared to men who have sex exclusively with women and transgender women (TW) compared to cisgender peers, with CSA prevalence estimates as high as 50% among MSM (an average of 27% across all studies and 22% among probabilistic samples in meta analysis) (Friedman et al., 2011; Lloyd & Operario, 2012). While similar meta-analyses do not exist for TW samples several studies have estimates above 20% with one study estimating as high as 61% of TW reporting CSA, though it should be noted that these were not probabilistic samples (Martinez et al., 2016; Parsons, Antebi-Gruszka, Millar, Cain, & Gurung, 2018; Sherman et al., 2019; White Hughto, Pachankis, Willie, & Reisner, 2017). This elevated prevalence of CSA among MSM and TW may contribute to known health disparities and related risk behaviors, including cannabis use, alcohol use, adverse mental health outcomes, sexual risk behavior, and HIV infection (Boroughs et al., 2015; Giano, Hubach, Currin, & Wheeler, 2019; Lloyd & Operario, 2012; Reback & Fletcher, 2014; Schneeberger, Dietl, Muenzenmaier, Huber, & Lang, 2014; Williams et al., 2015; Wu, 2018). Although the relationship between CSA and negative health outcomes is well established in the broader literature, there is a lack of consensus regarding the explanatory theoretical model for this relationship (Ben-Shlomo & Kuh, 2002; Kuh, Ben-Shlomo, Lynch, Hallqvist, & Power, 2003). Moreover, less is known about the impact of certain theoretically relevant constructs, such as timing and severity of CSA exposure in MSM and TW (Lloyd & Operario, 2012; Relf, 2001). This analysis applies a life course perspective to examine the effect of timing and severity of CSA on mental health, substance use, and sexual risk among a sample of MSM and TW.

CSA is by definition an exposure that occurs in a developmental period before adulthood, and therefore our understanding of the long-term effects of CSA is situated in a life course perspective. Life course theories situate adverse health exposures (such as CSA) along developmental stages, suggesting that the timing of exposure may influence the degree to which the exposure contributes to negative health outcomes (Ben-Shlomo & Kuh, 2002; Kuh et al., 2003). In these analyses, we explore three possible life course perspectives that may clarify the relationship between CSA and long term health consequences (Dunn et al., 2018). The “sensitive period” model suggests that an adverse exposure, such as CSA, may have a stronger effect if it occurs in specific periods of development during which an individual is particularly susceptible (Kuh et al., 2003). For example, periods of accelerated brain development, such as in childhood, puberty, or in early adolescence, may leave an individual particularly vulnerable to deleterious effects of early stressful exposures (Andersen & Teicher, 2008; Pechtel, Lyons-Ruth, Anderson, & Teicher, 2014). Exposure to stress during these developmental periods is associated with enduring consequences in both brain structure and function, which may predispose an individual for the development of depression and substance abuse disorders (Andersen & Teicher, 2008; Teicher, Tomoda, & Andersen, 2006). By contrast, the “recency model” posits that the effect of a stressor is strongest more immediately following exposure, regardless of the developmental period (Dunn et al., 2018; Shanahan, Copeland, Costello, & Angold, 2011). Recency models of CSA have been explored in relation to mental health in prior research, and studies using this perspective have found that the impact of CSA on depressive symptoms is strongest in the immediate months that follow a potentially traumatic event (Classen, Palesh, & Aggarwal, 2005; Dunn et al., 2018). A third model is known as “cumulative risk,” (or “accumulation” models) which suggests that the experience of multiple exposures over time, or the accumulation of different types of CSA exposures, may have a stronger cumulative effect than a single isolated exposure (Dunn et al., 2018; Evans, Li, & Whipple, 2013). Cumulative risk models have been supported in general samples; for example, individuals who experience 5 or more forms of abuse across childhood and early adolescence are more likely to experience mental health symptoms and alcohol problems compared to individuals with fewer experiences of abuse (Fisher et al., 2015; Merians, Baker, Frazier, & Lust, 2019). Lastly, the severity of the CSA incident may be relevant to the development of negative health outcomes in any of these three life course models. Measurement of severity of CSA varies across studies; generally speaking, the research literature has considered experiences of forced sexual touching to be lower in severity, whereas coerced or forced penetrative CSA has been considered higher in severity (Boroughs et al., 2015; Lloyd & Operario, 2012; Relf, 2001). Severity of CSA has been linked to increased sexual risk behavior in heterosexual men, including earlier sexual initiation, higher numbers of sexual partners, and a higher likelihood of drinking during sex (Schraufnagel, Davis, George, & Norris, 2010).

While these life course models have not been explicitly applied to CSA in MSM samples, CSA has previously been linked to cannabis use and alcohol use in MSM samples (Giano et al., 2019; Marshall et al., 2015; Martinez et al., 2016). MSM and TW who report CSA are more likely to report heavier drinking patterns, such that in a longitudinal study, MSM who reported CSA had nearly twice the odds of an increasingly heavy drinking trajectory (Marshall et al., 2015; Martinez et al., 2016). Similar findings occurred in another sample that included both MSM and TW participants; in this study, those who reported CSA had roughly 2.5 times the odds of heavy drinking (Martinez et al., 2016). Further, the effects of CSA are not just limited to alcohol use, as another study documented 2.5 times the odds of substance use disorders among MSM who experienced CSA perpetrated by a family member (Boroughs et al., 2015). Though cannabis use is common in urban MSM and TW populations little research has examined the influence of CSA on this specific substance (Swann, Bettin, Clifford, Newcomb, & Mustanski, 2017). One study found that MSM who reported adverse childhood experiences such as CSA are more likely to use cannabis (Giano et al., 2019; Swann et al., 2017). Despite the evidence for the relationship between CSA and substance use, researchers have not elucidated differences in theoretical explanations for this relationship in MSM and TW, including whether the timing or severity of CSA impacts its effect on substance use outcomes among these populations.

Similar to substance use outcomes, research has begun to establish a relationship between CSA and a range of adverse mental health outcomes in MSM and TW. For example, MSM who report CSA are more likely to report high levels of perceived stress, clinically significant depressive symptoms, and symptoms of PTSD (Boroughs et al., 2015; Lloyd & Operario, 2012; Mattera et al., 2018). Although literature addressing the relationship between CSA and mental health in TW samples is sparse, both CSA and other forms of victimization have been found to predict depressive symptomology in TW (Schneeberger et al., 2014). Further, another study observed that transgender people who reported childhood maltreatment were 3.4 times as likely to have psychiatric disorders in their lifetime (Bandini et al., 2011). Similar patterns have been observed in suicide-related outcomes. A relationship between CSA and suicide attempts is well-established in general populations, with a meta-analysis showing 1.9 times the odds of suicide attempts among people who experience CSA (Ng, Yong, Ho, Lim, & Yeo, 2018). There is less research about CSA and suicide among MSM and TW, but some studies have suggested higher likelihood of suicide attempts and suicide ideation among young MSM who report CSA (Arreola, Neilands, Pollack, Paul, & Catania, 2008; Bandini et al., 2011; Ratner et al., 2003; White Hughto et al., 2017). Much like substance use outcomes, limited literature has begun to compare the relationship between CSA and mental health outcomes across theoretically relevant constructs in these populations; however, limited evidence supports a sensitive period model such that MSM who reported their first experience of CSA in adolescence had 0.41 times the odds of major depressive disorder than those who experience first CSA before the age of 13 (Boroughs et al., 2015). Additionally, evidence suggests that the severity of CSA is relevant to mental health outcomes. One study found that MSM who reported CSA with penetration had 3.17 times the odds of PTSD than those who did not report penetration. Further MSM who reported CSA with physical injury had 4.04 times the odds of PTSD compared to CSA without physical injury, and MSM who reported CSA with intense fear had 5.16 times the odds of PTSD compared to those who did not report experiencing intense fear (Boroughs et al., 2015). More evidence is needed to examine the relative influences of relevant theoretical constructs in relation to mental health outcomes in MSM and TW.

In addition to substance use and mental health outcomes, CSA has been examined as a predictor of HIV risk behaviors and acquisition in MSM and TW. CSA predicts sexual risk behavior, such as condomless anal sex, number of recent sexual partners, number of casual sex partners, and likelihood of engaging in sexwork (Boroughs et al., 2015; Lloyd & Operario, 2012; Williams et al., 2015; Wu, 2018). For example, MSM who reported CSA had 1.85 the odds of condomless anal sex compared to MSM who did not report CSA (Abajobir, Kisely, Maravilla, Williams, & Najman, 2017; Lloyd & Operario, 2012). Moreover, TW who reported CSA had 1.72 times the risk of HIV transmission risk behaviors (Parsons et al., 2018). In addition to behavioral outcomes, CSA also predicts HIV and sexually transmitted infection status among MSM and TW (Boroughs et al., 2015; Lloyd & Operario, 2012). One meta-analysis shows that MSM who report CSA have 1.54 times the odds of being HIV positive (Lloyd & Operario, 2012; Wu, 2018). Though the literature is limited, a similar pattern was found in a sample of TW sexworkers where participants who reported CSA had 3.39 times the odds of testing positive for HIV (Sherman et al., 2019). Evidence testing theoretical models of CSA in relation to sexual risk in MSM and TW is limited; however, there is some evidence that severity of CSA may influence sexual risk. In one study, MSM who reported CSA with penetration had 2.72 times the odds of being at high sexual risk for HIV (Boroughs et al., 2015). Not only should further research examine these frameworks in relation to sexual risk, but research is needed to identify the extent to which these models are consistent across a range of health outcomes.

In summary, the literature has established that CSA is prevalent among MSM and TW, and that CSA predicts many health outcomes in these populations; however, there are a number of unaddressed gaps in the literature (Boroughs et al., 2015; Lloyd & Operario, 2012; Marshall et al., 2015; Mattera et al., 2018; Ratner et al., 2003). Researchers have suggested that sensitive period, recency, and cumulative risk models are relevant to relationships between adverse childhood experiences and various outcomes in general populations, but these models have not been extensively examined in relation to CSA among MSM and TW populations. While there is some evidence indicating that these theoretical models are relevant to CSA in these populations, little has been done to systematically test these theoretical explanations of the relationship between CSA and health outcomes among a single sample of MSM and TW.

Current Study:

This analysis examines the effect of history of CSA on sexual health, substance use, and mental health outcomes among a sample of young MSM and TW. More specifically, we examine three possible theoretical models (sensitive period, cumulative risk, and recency), as well as the theoretically relevant construct of CSA severity, to observe how the relationship between CSA and health outcomes may be explained by existing life course perspectives. We hypothesize that CSA will be correlated with increased alcohol consumption, increased cannabis use, higher levels of depressive symptoms, increased prevalence of suicidal ideation, suicidal attempts, condomless anal sex, and HIV status. We also hypothesize that increased severity and exposure across developmental periods will be correlated with negative health outcomes. We do not make specific hypotheses about the relative influence of the three models, given the limited existing research in this area.

Methods

Participants and procedures

This analysis uses baseline data from an ongoing cohort study of MSM and TW living in the Chicago metropolitan area called RADAR (n=1,035) (Mustanski et al., 2019). TW were included in the analysis though the sample is small (n=88). This is because of two primary reasons relating to research ethics: 1) to account for their contributions to the study; and 2) to acknowledge that TW are an underrepresented population in this research. In analysis, the authors control for gender in each adjusted estimate to account for anticipated differences by gender. The cohort includes a sample of young MSM and TW assigned male at birth (ages 16–29 at baseline). At enrollment, eligibility criteria included: male-assigned at birth, male-identified, between the ages of 16 and 29, and identify as gay/bisexual/queer or report engaging in oral or anal sex with a man in the past year. Recruitment for this study used an incentivized snowball sampling approach meaning that an initial set of participants are recruited directly who can then refer up to 5 peers. The recruitment for the initial set of participants included venue-based (e.g., community organizations) and social media advertisements (e.g., Facebook). Data were collected between February 2015 and August 2019 using computer assisted self-interview (CASI) software as well as the collection of biological samples for sexually transmitted infection testing. The primary goal of RADAR is to examine multi-level influences on HIV and substance use. See Table 1 for a full characterization of the sample.

Table 1.

Univariates describing sample of MSM and TW (n=1,129)

Variable	N (%)	Mean (SD)
Age	21.4 (3.0)
Race
White	286 (25.3%)
Black	382 (33.8%)
Hispanic/Latinx	338 (29.9%)
Other Race	123 (10.9%)
Gender
Male	1041 (92.2%)
Other Gender	88 (7.8%)
Sexual Orientation
Gay	781 (69.2%)
Bisexual	244 (21.6%)
Other Sexual Orientation	104 (9.2%)
Childhood Sexual Abuse
Touching (age <13)	125 (11.1%)
Penetration (age <13)	116 (10.3%)
Touching (age 13–17)	56 (4.9%)
Penetration (age 13–17)	87 (7.7%)
Childhood Sexual Abuse Combinations
None	819 (73.7)
Touching (age <13) only	83 (8.4)
Penetration (age <13) only	60 (5.4)
Touching (age 13–17) only	30 (2.7)
Penetration (age 13–17) only	27 (2.4)
Touching (age <13)*Touching (age 13–17)	14 (1.3)
Touching (age <13)*Penetration (age 13–17)	13 (1.2)
Penetration (age <13)*Touching (age 13–17)	10 (0.9)
Penetration (age <13)*Penetration (age 13–17)	46 (4.14)
Alcohol Problems (AUDIT Score)		6.0 (5.5)
Moderate or no problematic drinking (15 or less)	1044 (93.9)
High levels of problematic drinking (16 or more)	68 (6.1)
Cannabis Problems (CUDIT Score)		6.2 (6.3)
Moderate or no problematic cannabis use (12 or less)	891 (80.1)
High levels of problematic cannabis use (13 or more)	221 (19.9)
Stimulant Use (6 mo)
Yes	253 (22.8)
No	859 (77.3)
Suicide Ideation (6 mo)
Yes	141 (12.5%)
No	988 (87.5%)
Suicide Attempt (6 months)
Yes	47 (4.2%)
No	1082 (95.8%).
Depression (PROMIS Score)		15.7 (7.5)
Any Rectal STI (Gonorrhea or Chlamydia)
Positive	179 (16.1%)
Negative	933 (83.9%)
PrEP Use
Yes	70 (6.2%)
No	1059 (93.8%)
Condomless Sex Partners (6 mo)		1.7 (1.3)
HIV Status
Positive	183 (16.2%)
Negative	846 (83.8%)

Measures

Demographics included race, gender, and sexual orientation. Race/ethnicity was reduced to four categories for these analyses: White, Black, Latinx or “Other Race” (i.e., Asian, multi-racial, and other). All participants were assigned male at birth. Gender identity was reduced to two categories: cisgender man and gender minority (i.e., transgender women, non-binary individuals, and other gender identity). Sexual orientation was reduced to three categories: gay, bisexual, and other sexual orientation (i.e., queer, questioning, heterosexual, other).

Childhood Sexual Abuse

The measure of CSA included 6 items that were administered to assess the prevalence of self-reported unwanted childhood sexual experiences/abuse (Leserman, 2005). Three questions pertained to experiences prior to the age of 13 (labeled “pre-teenage” in the results), and three questions pertained to experiences between the ages of 13 and 17 (labeled “teenage” in results). This item set was framed as unwanted experiences and each question noted that the experiences were unwanted. Each dichotomous item addressed a different type of CSA experience that varied in intensity, including being touched, being forced to touch another person, and penetration (e.g. “Before your 13^th birthday, did an adult or someone at least five years older than you ever touch the sex organs of your body when you did not want this? By touch we mean with their hands, mouth or objects on your penis, vagina, pubic area or anus.”) (Leserman, 2005). Two categorical variables were constructed from the six CSA items. The first variable reflects experiences of CSA that had occurred before the age of 13, and the second reflected experiences between the ages of 13 and 17. The categories for each variable were “None,” “Touching only,” or “Penetration.” If a participant endorsed both touching and penetrative sexual abuse for an age range, the value was coded as “Penetration.” For each of these variables, the reference category was “None” (i.e. no sexual abuse). Interaction terms were also constructed from these two items. Interactions between these two variables are interpreted such that the effect is compared to having had no sexual abuse at either age range (i.e. the reference categories for both variables).

Cannabis Problems

Cannabis problems were measured using the Cannabis Use Disorder Identification Test (CUDIT-R) which is the sum of an 8-item screening instrument used to identify problematic cannabis use (Adamson et al., 2010; Adamson & Sellman, 2003). Questions addressed several dimensions of cannabis use including frequency, and symptoms (e.g. “How often in the past 6 months have you had a problem with your memory or concentration after using marijuana?”). Response options for the first 7 items are on a 5-item Likert scales, ranging from “Never” to “Daily or almost daily.” The final item is the question “Have you ever thought about cutting down, or stopping, your use of marijuana?” which includes responses options: “never,” “yes, but not in the past 6 months,” or “Yes, during the past 6 months.” The scale ranges from 0 to 32 and had an alpha of 0.78 in this sample. Previous research has defined a score of 13 or above as indicative of likely having a cannabis use disorder (Adamson et al., 2010). CUDIT-R scores are hereafter referred to as cannabis problems.

Alcohol Problems

The Alcohol Use Disorders Identification Test (AUDIT) is the sum of a 10-item instrument from the World Health Organization that measures problematic drinking (World Health Organziation, 2001; Saunders, Aasland, Babor, De la Fuente, & Grant, 1993). Questions addressed several dimensions of alcohol use including quantity, frequency, and symptoms of alcohol dependence, and alcohol-related problems (e.g. “Have you or someone else been injured because of your drinking?”). Response options are on a 5-item Likert scales, ranging from “Never” to “Daily or almost daily.” Scores ranging from 8–15 indicate moderate problematic drinking and 16+ indicates more severe problematic drinking. The scale ranges from 0 to 40 and had an alpha of 0.83 in this sample. AUDIT scores are here after referred to as alcohol problems.

Stimulant use

Stimulant use (specifically methamphetamine use or cocaine) was measured at first HIV-positive visit using self-report for the past 6 months. Participants were asked “In the past 6 months, have you used any of the following non-prescription drugs? - Cocaine or crack (also called Coke, Snow, Blow, Rock, or Freebase)” and “In the past 6 months, have you used any of the following non-prescription drugs? - Methamphetamines (also called Meth, Crystal Meth, Tina, parTy, or Crank).” Response options were “yes” or “no.” Participants were then classified as having any non-prescription drug use if they said yes to one or both substance.

Depression

Depression was measured using the Patient-Reported Outcomes Measurement Information System (PROMIS), which is calculated from an 8 item instrument (Choi, Schalet, Cook, & Cella, 2014; Kaat, Newcomb, Ryan, & Mustanski, 2017; Pilkonis et al., 2011). Item responses are on a 5-point Likert Scale, with response options ranging from “Never” to “Always.” The items included a range of feelings and symptoms (e.g. “In the past 7 days I felt hopeless”). Scores are calculated by multiplying the summed score of items by 8 and dividing by the number of items completed. The scale ranges from 8 to 40 and had an alpha of 0.95 in this sample.

Suicide

Suicidal ideation and attempts in the past 6 months were measured using 2 items from the World Health Organization’s Composite International Diagnostic Interview (World Health Organization, 1993). These items were dichotomous (yes/no) the first asking “In the past 6 months did you ever seriously consider attempting suicide?”, and the second asking “In past 6 months did you actually attempt suicide?.”

Sexual Risk

Sexual risk was assessed using the HIV-Risk Assessment for Sexual Partnerships (H-RASP) which is a computerized self-administered interview to assess sexual behavior (Swann, Newcomb, & Mustanski, 2018). Participants reported sexual behaviors and relationship characteristics for up to 4 partnerships during the preceding 6 months. Those who had more than 4 partnerships reported sexual behaviors for additional partnerships in aggregate. Using these variables, we derived a total number of condomless anal sex partners for the preceding 6 months.

Pre-exposure Prophylaxis Use

Pre-exposure Prophylaxis (PrEP) use in the preceding 6 months was assessed using a single self-report dichotomous item: “In the past 6 months have you taken any pre-exposure prophylaxis (PrEP) medication such as Truvada to reduce your risk of HIV transmission?” This was only included in the model for sexual risk, because of the possible correlation between PrEP use and condomless sex.

Sexually Transmitted Infection Status

Rectal swab samples were collected at the time of visit to assess for gonorrhea and chlamydia. Results of rectal sexually transmitted infections (STI) were combined to represent any positive STI result and negative for all rectal STIs. Alere™ Determine™ 4^th Generation HIV-1/HIV-2 Ab/Ag Combo rapid test was used to test for the presence of HIV-1 antibodies, HIV-2 antibodies and free HIV-1 p24 antigen. Laboratory confirmation followed the Centers for Disease Control and Prevention (CDC) guidelines for HIV testing (Centers for Disease Control and Prevention and Association of Public Health Laboratories, 2016). The resulting categories are “HIV positive” or “HIV negative.”

Statistical Analysis

All statistical analyses were run in R version 3.6.0 (Team, 2019). Logistic regression was used to estimate models for the dichotomous outcomes: suicidal ideation, suicide attempts, and HIV status. Negative binomial regressions were used to estimate models for non-normal outcomes: alcohol problems, cannabis problems, number of condomless anal sex partners, and depression. All models included both CSA variables and demographics. The model estimating condomless sex partners also controlled for PrEP use. Results for logistic models are reported in odds ratios (ORs) with 95% confidence intervals. Results from negative binomial regressions are reported in incidence rate ratios (IRRs). Interaction effects between the 2 CSA variables were assessed for each model (severity of CSA before the age of 13 and severity of CSA between 13 and 17) with reference categories being no history of CSA. In the case that interaction effect models were statistically significant, we report both the base model and the model including interaction effects.

In order to test the 3 proposed theoretical models we included 2 categorical items for CSA representing 2 different time periods (pre-teenage, and teenage), and tested interaction terms between these two time periods. If the pre-teenage CSA item is consistently significant across models, this may be indicative of an early sensitive period model. In contrast, if teenage experience of CSA is consistently significant across models, this may indicative of a recency model or a later sensitive period model. Finally, consistently statistically significant interaction terms between the two CSA items will be indicative of a cumulative model in that the interaction terms represent experiences of CSA across both time periods (e.g. the interaction of pre-teenage forced touching, and teenage forced touching). The categorical response options for CSA items account for severity (forced touching being less severe and penetrative CSA being more severe). If the more severe (penetrative CSA) response is consistently significant across outcomes this is indicative of severity being a relevant construct. Assessment of theoretical models will be made considering all of these indicators together across models.

Results

Of the 1135 people were enrolled in RADAR at the time of the study all participants who had adequate baseline data were included (n=1129) except for 7 people who were excluded due to missing data. Additionally, 17 participants were missing rectal STI results due to at least one of the following reasons 1) refusal, 2) specimen collection error, or 3) indeterminate results. For this reason the models which included rectal STI have a smaller sample than the other models. Table 1 shows CSA and demographic characteristics of the sample. The vast majority identified as male (92.2%), with 7.8% identifying as female or non-binary. The largest racial/ethnic group in this sample was Black (33.8%), followed by Hispanic/Latinx (29.9%), White (25.3%), and other racial groups combined (10.9%).¹ In bivariate analyses, pre-teenage CSA and teenage experiences of CSA were strongly associated X²=218.8 (df=4.3, p<0.001).

Assessment of Main Effects

Alcohol Problems

The results of regressions are presented in Table 2. Prior to considering interaction terms, pre-teenage forced touching (IRR= 1.3, 95% CI 1.1–1.6, p<0.01) and teenage touching (IRR= 1.3, 95% CI 0.1–1.7, p<0.05) were both associated with alcohol problems. Race also had a significant effect, with Black (IRR=0.5, 95% CI 0.4–0.6, p<0.001), Latinx (IRR=0.8, 95% CI 0.7–0.9, p<0.001), and “Other Race” participants (IRR=0.7, 95% CI: 0.6–0.8, p<0.001) having lower IRRs compared to white participants for alcohol problems. Age also had a positive significant effect (IRR=1.1, 95% CI 1.0–1.1, p<0.001).

Table 2.

Negative binomial regressions examining alcohol problems, cannabis problems, suicide, depression, condomless anal sex partners, rectal STI, and HIV status (n=1,129)

	Alcohol Problems AUDIT (RAW)			Cannabis Problems CUDIT (RAW)			Stimulant Use			Suicide Ideation			Suicide Attempt			Depressive Symptoms PROMIS (RAW)			CAS Partners			Rectal STI1			HIV Status
	IRR (95% CI)			IRR (95% CI)			OR (95% CI)			OR (95% CI)			OR (95% CI)			IRR (95% CI)			IRR (95% CI)			OR (95% CI)			OR (95% CI)
	1	2	3	1	2	3	1	2	3	1	2	3	1	2	32	1	2	3	1	2	3	1	2	3	1	2	3
Intercept	2.8*** (1.8–4.2)	2.8*** (1.9–4.2)	2.9*** (1.9–4.3)	3.4*** (1.9–6.2)	3.5*** (1.9–6.3)	3.4*** (1.9–6.1)	0.0 *** (0.0–0.0)	0.0*** (0–0.0)	0.0*** (0–0.0)	1.1 (0.3–4.5)	1.2 (0.3–5.2)	1.3 (0.3–5.6)	0.1 (0.0–1.5)	0.2 (0.0–2.4)		18.0*** (15.8–22.0)	18.7*** (15.3–22.7)	18.8*** (15.4–22.9)	0.7 (0.4–1.1)	0.7 (0.4–1.2)	0.7 (0.4–1.2)	0.06*** (0.0–0.2)	0.1*** (0.0–0.2)	0.1*** (0.0–0.2)	0.0*** (0.0–0.0)	0.0*** (0.0–0.0)	0.0*** (0.0–0.0)
Age	1.1*** (1.0–1.1)	1.1*** (1.0–1.1)	1.1*** (1.0–1.1)	1.0 (1.0–1.1)	1.0 (1.0–1.1)	1.0 (1.0–1.1)	1.2*** (1.1–1.3)	1.2*** (1.1–1.3)	1.2*** (1.1–1.3)	0.9* (0.9–1.0)	0.9** (0.8–1.0)	0.9** (0.8–1.0)	0.9 (0.8–1.0)	0.9* (0.8–1.0)		1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0 (1.0–1.0)	1.0* (1.0–1.1)	1.0* (1.0–1.1)	1.0* (1.0–1.0)	1.0 (1.0–1.1)	1.0 (1.0–1.1)	1.0 (1.0–1.1)	1.4*** (1.3–1.5)	1.6*** (1.3–1.5)	1.4*** (1.3–1.5)
Race (ref=white)
Black	0.5*** (0.5–0.6)	0.5*** (0.4–0.6)	0.5*** (0.4–0.6)	1.3* (1.0–1.6)	1.2 (1.0–1.5)	1.2 (1.0–1.5)	0.3 *** (0.2–0.5)	0.3*** (0.2–0.4)	0.3*** (0.2–0.4)	0.6* (0.4–1.0)	0.5* (0.3–0.9)	0.5** (0.3–0.8)	4.0** (1.6–12.3)	3.3* (1.3–10.2)		0.9*** (0.8–0.9)	0.8*** (0.8–0.9)	0.8*** (0.8–0.9)	0.9 (0.7–1.1)	0.8 (0.7–1.0)	0.8* (0.7–1.0)	4.7*** (2.9–7.9)	4.3*** (2.6–7.4)	4.5*** (2.7–7.6)	29.9*** (12.8–87.7)	28.1*** (12.0–82.9)	30.3*** (12.9–89.7)
Hispanic/Latinx	0.8** (0.7–0.9)	0.8*** (0.7–0.9)	0.8*** (0.7–0.9)	1.2 (0.9–1.4)	1.1 (0.9–1.4)	1.1 (0.9–1.4)	0.7* (0.5–1.0)	0.6* (0.4–0.9)	0.6* (0.4–0.9)	0.7 (0.4–1.1)	0.6 (0.4–1.0)	0.6 (0.4–1.0)	2.0 (0.7–6.5)	1.7 (0.6–5.6)		1.1 (1.0–1.1)	0.9 (0.9–1.0)	0.9 (0.9–1.0)	1.1 (0.9–1.3)	1.1 (0.9–1.3)	1.1 (0.9–1.3)	2.1* (1.2–3.7)	2.0* (1.2–3.5)	2.0* (1.2–3.5)	12.0*** (5.9–35.9)	11.3*** (4.6–34.1)	11.3*** (4.6–34.2)
Other Race	0.7* (0.6–0.8)	0.7*** (0.6–0.8)	0.7*** (0.6–0.8)	1.1 (0.8–1.5)	1.1 (0.8–1.5)	1.1 (0.8–1.5)	0.5** (0.3–0.8)	0.4** (0.2–0.7)	0.4** (0.2–0.7)	0.6 (0.3–1.1)	0.5 (0.3–1.0)	0.5 (0.3–1.0)	3.5* (1.1–12.2)	2.8. (0.8–10.)		0.9* (0.8–1.0)	0.9** (0.8–1.0)	0.9** (0.8–1.0)	1.1 (0.8–1.4)	1.1 (0.8–1.4)	1.0 (0.8–1.3)	1.7 (0.8–3.4)	1.6 (0.8–3.3)	1.6 (0.8–3.4)	9.3*** (3.3–30.1)	8.8*** (3.1–29.0)	8.8*** (3.1–29.2)
Gender (ref=cis male)
Other Gender	1.2 (0.9–1.5)	1.1 (0.9–1.4)	1.1 (0.8–1.3)	1.0 (0.7–1.4)	1.0 (0.7–1.3)	1.0 (0.7–1.4)	1.4 (0.8–2.4)	1.2 (0.7–2.2)	1.2 (0.7–2.2)	3.2*** (1.7–5.8)	2.8** (1.5–5.2)	3.0*** (1.6–5.5)	4.1** (1.6–9.5)	3.2* (1.2–8.0)		1.1 (1.0–1.2)	1.1 (1.0–1.2)	1.1 (1.0–1.2)	1.1 (0.8–1.5)	1.0 (0.8–1.4)	1.0 (0.8–1.4)	1.4 (0.7–2.7)	1.2 (0.6–2.4)	1.2 (0.6–2.3)	1.7 (0.9–3.5)	1.6 (0.8–3.3)	1.8 (0.8–3.8)
Sexual Orientation (ref=gay)
Bisexual	1.0 (0.9–1.1)	1.0 (0.9–1.1)	1 (0.9–1.14)	1.1 (1.0–1.4)	1.1 (0.9–1.4)	1.1 (0.9–1.4)	1.5 (1.1–2.2)	1.5* (1.1–2.2)	1.5* (1.1–2.2)	1.3 (0.7–2.3)	1.3 (0.9–2.0)	1.3 (0.8–2.0)	1.2 (0.6–0.6–2.5)	1.2 (0.6–2.5)		1.1 (1.0–1.1)	1.1 (0.8–1.1)	1.1 (1.0–1.1)	0.7*** (0.6–0.9)	0.7*** (0.6–0.9)	0.7** (0.6–0.9)	0.7 (0.4–1.0)	0.7 (0.4–1.0)	0.7 (0.4–1.0)	0.3*** (0.2–0.6)	0.3*** (0.2–0.5)	0.3*** (0.2–0.6)
Other	0.8 (0.7–1.1)	0.9 (0.7–1.1)	0.9 (0.7–1.1)	1.0 (0.8–1.4)	1.1(0.8–1.5)	1.1 (0.8–1.5)	0.8 (0.4–1.4)	0.8 (0.4–1.4)	0.7 (0.4–1.3)	1.2 (0.7–2.3)	1.2 (0.6–2.3)	1.3 (0.6–2.3)	1.4 (0.5–3.6)	1.4 (0.5–3.9)		1.1 (1.0–1.2)	1.1 (1.0–1.2)	1.1. (1.0–1.2)	0.7* (0.6–1.0)	0.8* (0.6–1.0)	0.8* (0.6–1.0)	0.5 (0.2–1.0)	0.5 (0.2–1.0)	0.5 (0.2–1.1)	0.4* (0.2–0.8)	0.4* (0.2–0.8)	0.3** (0.1–0.7)
PrEP Use 6 mo. (ref=none)																			2.1*** (1.6–2.7)	2.1*** (1.6–2.7)	2.1*** (1.6–2.7)	1.1 (0.5–2.1)	1.0 (0.5–2.0)	1.1 (0.5–2.1)
CSA (pre-teenage)
Touching		1.3** (1.1–1.6)	1.3* (1.0–1.5)		1.2 (1.0–1.6)	1.2 (0.9–1.6)		1.3 (0.8–2.1)	1.3 (0.7–2.2)		1.3 (0.7–2.3)	1.4 (0.7–2.7)		0.9 (0.3–2.5)			1.1 (1.0–1.2)	1.1 (1.0–1.2)		1.2 (0.9–1.5)	1.2 (1.0–1.6)		1 (0.6–1.7)	0.8 (0.4–1.5)		0.9 (0.5–1.6)	1.1 (0.5–2.0)
Penetration		1.1 (0.9–1.3)	1.0 (0.8–1.3)		1.1 (0.9–1.5)	1.4. (1.0–2.0)		1.2 (0.7–2.0)	1.7. (0.9–3.2)		1.5 (0.8–2.8)	1.8 (0.8–3.7)		2.2. (0.9–5.0)			1.1* (1.0–1.2)	1.2** (1.1–1.3)		1.1 (0.8–1.4)	1.1 (0.8–1.5)		1.5 (0.8–2.5)	1.6 (0.8–3.0)		1.2 (0.7–2.2)	1.1 (0.5–2.3)
CSA (teenage)
Touching		1.3* (1.0–1.7)	1.2 (0.9–1.7)		1.3 (0.9–1.8)	1.5 (1.0–2.5)		2.3** (1.2–4.2)	2.3 (1.0–5.1)		2.5** (1.2–5.0)	4.7*** (1.9–10.6)		2.6 (0.8–7.1)			1.2** (1.1–1.4)	1.3*** (1.1–1.6)		1.5* (1.1–2.1)	2.0*** (1.4–3)		1.2 (0.5–2.4)	0.5 (0.1–1.6)		2.7** (1.3–5.5)	1.4 (0.5–3.9)
Penetration		1.2 (1.0–1.5)	1.0 (0.7–1.5)		1.4* (1.0–1.9)	1.7* (1.1–2.9)		2.6*** (1.5–4.5)	4.4*** (1.9–9.7)		1.8 (0.9–3.5)	1.3 (0.4–3.6)		2.7* (1.1–7.0)			1.1* (1.0–1.3)	1.1 (1.0–1.3)		1.4* (1.1–1.9)	1.2 (0.8–1.9)		2.0* (1.1–3.5)	2.5 (0.9–6.0)		1.7 (0.9–3.3)	5.5** (1.7–16.8)
Interactions
Touching (<13)*Touching (13–17)			0.9 (0.5–1.6)			0.8 (0.4–2.0)			1.1 (0.2–4.7)			0.3 (0.1–1.7)						0.9 (0.7–1.2)			0.3** (0.1–0.7)			4.59 (0.65–33.14)			2.4 (0.3–15.6)
Penetration (<13)*Touching (13–17)			1.46 (0.8–2.9)			0.4* (0.1–1.1)			0.8 (0.1–4.0)			0.1 (0.0–1.0)						0.7 (0.5–1.0)			0.6 (0.3–1.5)			4.7 (0.71–35)			5.7. (0.9–40.03)
Touching (<13)*Penetration (13–17)			1.9* (1.0–3.5)			0.9 (0.4–2.2)			0.7 (0.2–3.3)			1.5 (0.2–10.1)						1.2 (0.9–1.6)			1.5 (0.7–3.3)			1.6 (0.32–8.21)			0.1* (0.0–0.7)
Penetration (<13)*Penetration (13–17)			1.2 (0.7–2.0)			0.6 (0.3–1.1)			0.3 (0.1–1.0)			1.5 (0.4–7.5)						0.9 (0.7–1.2)			1.2 (0.6–2.3)			0.58 (0.16–2.18)			0.3 (0.1–1.4)
R-squared	0.04	0.05	0.05	0.01	0.04	0.04	0.07	0.10	0.10	0.04	0.05	0.06	0.02	0.03	0.03	0.03	0.05	0.07	0.03	0.04	0.04	0.05	0.07	0.07	0.25	0.26	0.27

^*p<0.05

^**p<0.01

^***p<0.001

¹Rectal STI results were missing for 17 participants resulting in a smaller sample for this model.

²Interaction model did not converge due to small numbers

Cannabis Problems

Only teenage forced pentration was associated with cannabis problem score (IRR=1.4, 95% CI 1.0–1.9, p<0.05).

Stimulant Use

Teenage forced touching (OR=2.3, 95% CI 1.2–4.2, p<0.01) and teenage forced penetration (OR= 2.6, 95% CI 1.5–4.5, p<0.001) were associated with higher odds of using stimulants. Bisexual participants had higher odds of using stimulants compared to gay participants (OR= 1.5, 95% CI 1.1–2.2, p<0.05) and age was positively associated with stimulant use (OR=1.2, 95% CI 1.1–1.3, p<0.001). Black (OR=0.3, 95% CI 0.2–0.4, p<0.001), Hispanic/Latinx (OR=0.6, 95% CI 0.4–0.9, p<0.05), and “Other Race” (OR=0.4, 95% CI 0.2–0.7, p<0.01) participants were less likely to use stimulants than white participants.

Suicidal Ideation and Attempt

Logistic regressions were run for suicidal ideation and suicide attempts. Teenage forced touching (OR= 2.5, 95% CI 1.2–5.0, p<0.001) was associated with a more than two times higher odds of suicidal ideation in the past 6 months. Gender was also significantly associated with suicidal ideation, such that non-cisgender individuals had higher odds of reporting suicidal ideation compared to cisgender males (OR=2.8, 95% CI 1.5–5.2, p<0.01). Black participants had lower odds of suicide ideation compared to white participants (OR= 0.5, 95% CI 0.3–0.9, p<0.05). A similar pattern was observed for suicide attempts: teenage forced touching was associated with significantly higher odds of reporting a suicide attempt in the past 6 months (OR=2.5, CI 95% 1.2–5.0, p<0.01). However, Black participants had higher odds of suicide attempts compared to white participants (OR=3.3, 95% CI 1.3–10.2, p<0.05).

Depressive symptoms

A negative binomial regression was utilized for examining the association between CSA and depressive symptoms to account for non-normality in the PROMIS Depression raw score. Pre-teenage forced penetration (IRR=1.1, CI 95% 1.0–1.2, p<0.05), teenage forced touching (IRR=1.2, 95% CI 1.1–1.4, p<0.01) and teenage forced penetration (IRR= 1.1, 95% CI 1.0–1.3, p<0.05) were each associated with higher depressive symptom score. Further, Black participants had a significantly lower depressive symptom score when compared to white participants (IRR=0.8, 95% CI 0.8–0.9, p<0.001).

Sexual Risk

Negative binomial regression was used to assess the association between CSA and total number of condomless anal sex partners in the past 6 months. Teenage forced touching (IRR= 1.2, 95% CI 1.1–1.4, p<0.01), and teenage forced penetration (IRR=1.1, 95% CI 1.0–1.3, p<0.05) were each associated with a higher rate of condomless anal sex partners. Bisexual participants (IRR=0.7, 95% CI 0.6–0.9, p<0.001) and participants with other sexual orientations (IRR=0.7, 95% CI 0.6–1.0, p<0.05) had lower rates of condomless anal sex partners compared to gay participants. PrEP use was also included in the model to account for possible correlations between PrEP use and condom use. PrEP use in the last 6 months was associated with the number of condomless sex partners (IRR=2.1, 95% CI 1.6–2.7, p<0.001).

Logistic regression was used to assess the association between CSA and testing positive for any rectal STI. Participants who reported teenage forced penetration had twice the odds of testing positive for any rectal STI compared to people who did not experience CSA (OR= 2.0 95% CI 1.1–3.5, p<0.05). Black participants had higher odds (OR= 4.3, 95% CI 2.6–7.6, p<0.001) and Hispanic/Latinx participants had higher odds (OR= 2.0, 95% CI 1.2–3.5, p<0.05) of testing positive for a rectal STI than white participants.

Lastly, logistic regression models were used to assess the association between CSA and HIV status. Teenage forced touching was associated with a higher likelihood of being HIV-positive (OR=2.7, 95% CI 1.3–5.5, p<0.01). Black participants had higher odds of being HIV positive (OR=28.1 95% CI 12.0–83.0, p<0.001), Latinx participants had higher odds of being HIV positive (OR=11.3, 95% CI 4.6–34.1, p<0.01), and “Other Race” participants had higher odds of being HIV positive (OR = 8.8, 95% CI 3.1–29.2, p<0.001) compared to white participants. Bisexual participants had lower odds (OR= 0.3, 95% CI 0.2–0.6, p<0.001) and other sexual orientation participants had lower odds of testing HIV positive compared to gay participants (OR=0.3, 95% CIU 0.1-.07, p<0.01).

Assessment of interactions

Alcohol Problems

An interaction between pre-teenage forced touching and teenage forced penetration was associated with alcohol problems (IRR= 1.9, 95% CI 1.0–3.5, p<0.05). As depicted in Table 2 along with the significant main effect of pre-teenage touching, those who experienced pre-teenage touching and teenage penetration had a significantly increased IRR compared to participants who reported no CSA. The main effects of pre-teenage touching (IRR= 1.3, 95% CI 1.0–1.5, p<0.05) remained significant. Other significant effects included Black race (IRR=0.5, 95% CI 0.4–0.6, p<0.001), Latinx (IRR=0.8, 95% CI 0.7–0.9, p<0.001) and “Other Race” (IRR=0.7, 0.6–0.8, p<0.001) compared to white participants.

Cannabis Problems

An interaction between pre-teenage forced penetration and teenage forced touching (IRR= 0.4, 95% CI 0.1–1.1, p<0.05) was associated with cannabis problems, with the main effect teenage penetration (IRR=1.7, 95% CI 1.1–2.9, p<0.05) remaining significant. As seen in Table 2, that those participants who reported both pre-teenage penetration and teenage touching were less likely to use cannabis compared to those who did not experience CSA. The main effect of teenage touching was not significant in the model including the interaction term.

Condomless Anal Sex Partners

An interaction between pre-teenage touching and teenage touching was associated with the number of condomless anal sex partners in the last 6 months (IRR=0.3, 95% CI 0.1–0.7, p<0.01). As seen in Table 2, this meant that those who reported both pre-teenage and teenage forced touching had fewer condomless anal sex partners compared to those who didn’t report CSA. The main effect of teenage forced touching remained significant (IRR=2.0, 95% CI 1.4–4.0, p<0.001). PrEP use in the last 6 months (IRR=2.1, 95% CI 1.6–2.7, p<0.001) remained significant. The main effect of bisexual identity (IRR=0.7, 95% CI 0.6–0.9, p<0.01), and other sexual orientation (IRR=0.8, 95% CI 0.6–1.0, p<0.05) remained significant compared to gay participants.

HIV Status

An interaction term between pre-teenage touching and teenage penetration was associated with HIV status (OR=0.1, 95% CI 0.0–0.7, p<0.05). As seen in Table 2 this meant that those who reported both pre-teenage forced touching and teenage forced penetration had lower odds of being HIV positive than those who didn’t report CSA. In this model the only significant main effect of CSA was forced teenage penetration (OR=5.5, 95% CI 1.7–16.8, p<0.01). Identifying as Black (OR=30.3, 95% CI 12.9–89.7, p<0.001), Latinx (OR=11.3, 95% CI 4.6–34.2, p<0.001), or “Other Race” (OR=8.8, 95% CI 3.1–29.2, p<0.001) remained significant compared to white participants. Being bisexual (OR=0.3 95% CI 0.2–0.6, p<0.001) or other sexual orientation (OR=0.3, 95% CI 1.1–1.7, p<0.01) remained significant compared to gay sexual identity. Age remained significant (OR=1.4, 95% CI 1.3–1.5, p<0.001).

Discussion

This analysis examined the impact of timing and severity of CSA on a range of health outcomes in a sample of MSM and TW. Consistent with previous literature, this analysis found CSA to be a salient influence on various substance use, mental health, and sexual risk outcomes; however, a singular explanatory framework did not emerge (Boroughs et al., 2015; Dunn et al., 2019; Lloyd & Operario, 2012; Marshall et al., 2015; Mattera et al., 2018; Ratner et al., 2003). While there was some evidence for each theoretical model examined in these analyses, teenage experiences of CSA emerged as the most consistently predictive set of CSA experiences across the included health outcomes. Though these findings are not definitive, the pattern may align more closely with two possible theoretical models: 1) a sensitive period model where teenage years represent a sensitive period where CSA may have greater impact, or 2) the recency framework, suggesting that among MSM and TW, more recent exposures may be more closely linked to health outcomes and engagement in risk behaviors than pre-teenage or cumulative exposures (Dunn et al., 2018; Shanahan et al., 2011). The results of this analysis are informative to CSA research, prevention, and health interventions.

As hypothesized, we found significant main effects of CSA across all outcomes. These findings are consistent with previous research that found an association of CSA with a range of negative outcomes among MSM (Arreola et al., 2008; Lloyd & Operario, 2012); however, the current analysis differentiated between age and severity of experience when assessing the association of CSA experience with various negative health outcomes. More specifically, we found significant positive main effects of teenage experiences of CSA across all outcomes and significant positive main effects of pre-teenage experiences of CSA only for alcohol use and depressive symptoms. Although it is not conclusive, this may align with either a recency model, or a sensitive period model where teenage years represent a sensitive period (Kuh et al., 2003). Teenage experience of CSA may be associated with these outcomes for a number of reasons. Broader literature has identified a range of potential pathways that may link teenage experiences of CSA to health outcomes including increased impulsivity (Braquehais, Oquendo, Baca-García, & Sher, 2010; Corstorphine, Waller, Lawson, & Ganis, 2007), increased emotion dysregulation (Braquehais et al., 2010), early maladaptive schema (Wright, Crawford, & Del Castillo, 2009), disruption of sexual identity formation (Relf, 2001), and maladaptive relationship behaviors (Colman & Widom, 2004; Taylor, Goshe, Marquez, Safren, & O’Cleirigh, 2018). While the current analysis is unable to test these pathways they warrant further inquiry. Future research must seek to understand how these mechanisms may function in MSM and TW populations as well as if these mechanisms function differently or similarly across these various health outcomes.

In the current analysis, the assessment of interaction terms yielded results that were contrary to our hypothesized direction of association in three cases (cannabis problems, number of casual anal sex partners, and HIV status) in that interactions between preteenage and teenage experiences of CSA were negatively associated with these outcomes. In regard to condomless anal sex partners this is inconsistent with literature that suggests that CSA is associated with frequent condomless sex and more condomless sex partners, as well as a tendency engage in relationships characterized by instability, disruption, and nonmonogamy, which all may represent factors increasing STI and HIV risk among this population (Lloyd & Operario, 2012). Previous literature addressing heterosexual, cisgender males found that more severe experiences of CSA led to earlier sexual debut, riskier sex behaviors, and higher numbers of partners which may relate to associated disruptions in impulse control or coping through substances (Schraufnagel et al., 2010). It is not possible to draw definitive conclusions from the current analysis in regard to mechanism, but it may be that MSM and TW who experience early less-severe CSA may experience a form of sexual inhibition due to these experiences resulting in taking more precautions or being more hesitant to have multiple partners or multiple partners without condoms. Future analyses should examine this pattern and potential explanations for this pattern.

Similarly, pre-teenage experience of forced penetration combined with teenage forced touching was associated with lower cannabis problem scores as compared to individuals who didn’t experience CSA. This finding contradicts both the cumulative risk model and previous research finding that CSA is associated with cannabis use, early cannabis use, and frequent cannabis use (Hayatbakhsh et al., 2009). It appears that a more severe pre-teenage experience of CSA may inoculate an individual in some way from possible cumulative effects. It is possible that this is impacted by other more nuanced dynamics, for example in previous studies where individuals who received a supportive response when disclosing their CSA experiences, or were able to successfully seek help that ended their experiences of CSA were less likely to experience negative health outcomes such as substance dependence later on (Bulik, Prescott, & Kendler, 2001). If for example, more severe experiences of CSA are more likely to be disclosed and addressed this may have a dampened or reversed effect relative to the expected cumulative effect. Future research should examine additional factors that may be associated with this cumulative pattern in relation to cannabis use.

Differences in the impact of preteenage and teenage experiences of CSA across a range of outcomes in MSM and TW warrants further research with a more granular lens. This may include details further characterizing violent experiences (e.g. use of force, periodicity, duration, etc.), the perpetrators of the violence (e.g. relationship to perpetrator, etc.), and the survivor’s perceptions of the violent experiences (e.g. attribution, disclosure, affect, etc.) in order to further examine the relevance of theoretical frameworks that center adolescent experiences (Diamond, 2017; Steel, Sanna, Hammond, Whipple, & Cross, 2004). For example, people who experience CSA are more likely to exhibit internalizing behaviors if they perceive themselves to have participated in their abuse or to be responsible for their abuse, which may be more common in adolescence (Steel et al., 2004). Adolescents are more likely to be sexually involved with the perpetrator than younger children and subsequently more likely to perceive themselves as participating in the abuse (Diamond, 2017). These potential contextual differences between CSA experiences that occur in younger children and in adolescents warrant further inquiry.

Further research, particularly prospective longitudinal research, is needed to tease out this theoretical difference as well as the persistence of these effects across developmental periods (Abajobir et al., 2017; Lloyd & Operario, 2012; Ng et al., 2018). In particular, it is important to further examine possible mechanisms through which age of CSA exposure impacts these outcomes such as through externalizing and internalizing behaviors as well as to differentiate possible psychological and neuro-biological mechanisms that may contribute to this difference across pre-teenage and teenage experience of abuse.

While more detailed measurement of these contextual factors exists in the broader literature, fewer studies have endeavored to measure these dimensions among MSM and TW. For example, in a review of literature addressing CSA’s impact on HIV in MSM in 12 studies only 3 studies differentiated CSA by age of experience, 2 differentiated by frequency of experience, and 1 differentiated by affect relating to the CSA experience (Lloyd & Operario, 2012). More nuanced measurement of CSA is needed in studies of MSM and TW in order to examine to what extent these contextual factors and abuse characteristics account for differences in effect across developmental periods. Additionally, research is needed to examine unique contextual factors that may surround MSM and TW experiences of CSA. For example, successful help-seeking and social support are documented protective factors for long-term health outcomes (Steel et al., 2004); however, MSM and TW may be less likely to have social support and more likely to experience stigma related to their sexual and gender identities.

Lastly, these outcomes may be to some extent mutually-reinforcing. For example, literature addressing heterosexual cisgender men found CSA to be associated with alcohol use before and during sex, which may exacerbate sexual risk behaviors such as increased number of partners, and condomless sex (Schraufnagel et al., 2010). Future analyses may consider examining how CSA may relate to clusters of negative health outcomes, or how negative health outcomes that result from CSA may also be mutually reinforcing.

This analysis highlights the importance of considering nuances of population and experience in CSA prevention and intervention. Prevention and intervention pose considerable challenges given the sensitivity of the topic, the inherent power dynamics that impact children and young adults, and the limited resources allotted to child protection (Scott, Lonne, & Higgins, 2016). Current approaches mostly address CSA through tertiary intervention, addressing the effects of CSA long after the exposure (Scott et al., 2016). Previous research addressing interventions to diminish the risk of PTSD for example, are underdeveloped and often do not take into account the heterogeneity of survivors including gender and sexual identity as well as trauma experiences such as age of experience (Qi, Gevonden, & Shalev, 2016). Some evidence suggests that a needs-based survivor-centered approach that is responsive to specific experiences and needs of a survivor of CSA may be more effective than generic clinical interventions (Qi et al., 2016). We also know from review of the research that even among adults who seek mental health services that CSA screening isn’t commonly applied (Read, Harper, Tucker, & Kennedy, 2018). A review of the research suggests that 0–22% of mental health service users report being asked about CSA with men being less likely to be asked than women (Read et al., 2018; Scott et al., 2016). Real change in the approach to addressing CSA may require a measured, multifaceted, and nuanced public health approach with consideration of potential unintentional consequences of well-intentioned interventions (Felitti et al., 2019; Qi et al., 2016; Scott et al., 2016).

The salience of CSA across a range of health outcomes in this analysis suggests that existing interventions addressing outcomes, such as substance abuse and sexual risk behaviors, should also consider incorporating trauma-informed practices to maximize impact and mitigate the influence of potentially traumatic experiences on MSM and TW populations. More comprehensive approaches should enhance coordination between adolescent healthcare, social services, schools, and public health (Felitti et al., 2019). Interventions to identify and address known maladaptive coping strategies among adolescents who are impacted by CSA should be prioritized, as well as interventions that are tailored to the specific needs of MSM and TW such as the recently tested cognitive behavioral therapy model (titled CBT-TSH intervention) designed to reduce HIV risk behavior and address CSA in MSM (O’Cleirigh et al., 2019; Taylor et al., 2018).

Like intervention, primary prevention of CSA requires consideration of the complexities of the context in which CSA occurs. Fewer resources are currently allocated to preventive strategies for CSA as compared to tertiary response (Kyte, Trocme, & Chamberland, 2013; Scott et al., 2016). Even so, prevention may occur at multiple levels of the social ecology targeting the child, parent/guardian, or the environment (Scott et al., 2016). The OAK Foundation and the Centers for Disease Control and Prevention have released guidelines based on reviews of the literature, which suggest addressing CSA prevention through broader environments (laws, norms, etc.), parent/caregiver support, economic strengthening, response/support services, and individual interventions addressing knowledge and life skills (Health & Human Services, 2007; Ligiero, Hart, Fulu, Thomas, & Radford, 2019). Some strategies have promising evidence or demonstration of effectiveness such as community mobilization programs, parent/caregiver educational programs, support services such as Trauma-focused Cognitive Behavioral Therapy, and some educational programs for children and adolescents (Ligiero et al., 2019; Scott et al., 2016). these programs: (1) have limited evidence, (2) are primarily designed and tested for cisgender girls, and (3) don’t address unique experiences or needs of MSM and TW. To increase the likelihood of accessing these resources, they should be informed by MSM and TW experiences and culturally appropriate. In addition, prevention strategies may include voluntary educational resources for parents of adolescents, especially parents of MSM or TW adolescents. These may be made accessible in places that parents of MSM or TW adolescents seek support such as LGBT-affirming primary care providers, parent support groups, and websites. Moreover, social and physical environments can be modified to reduce stigma for seeking help, increase social support, increase the likelihood of detecting/interrupting CSA, and reduce situations where teenagers may be at risk for CSA (Scott et al., 2016). Given the high prevalence of CSA among MSM and TW adolescents, it may be beneficial to tailor environmental prevention strategies to LGBT community (physical, social, and online). However, as stated before development and testing of interventions should be measured and consider the potential for unintentional negative impacts (Ligiero et al., 2019).

Limitations

This analysis has considerable strengths, including a large racially diverse sample of MSM and TW. There are some notable limitations. While it’s a strength that the participants were sampled at ages closer to exposure to facilitate better recall and the evaluation of recency, it’s not possible to discern between a recency and an teenage sensitive period model. This analysis is cross-sectional, which limits causal inference, though by definition all CSA experiences occurred prior to the observation of the dependent variables allows for some assessment of temporality. All participants had oral or anal sex with a man a year prior to enrollment, which may pose selection effects on these analyses. The subset of participants who were assigned male at birth and identify as genders other than male is too small to make specific inferences, further research with larger non-binary and/or transgender samples should be conducted. Given the measures used in this study we are unable to disentangle age and severity of abuse from other important contextual factors such as relationship to the abuser, or attribution-style. In particular, future research needs to disentangle the use of force from the severity of abuse act in experiences of MSM and TW. Moreover, it should be examined if these contextual factors are more likely to occur in young MSM or TW. An additional consideration is that given the current data we are unable to discern if CSA experiences before and after 13 were perpetrated by the same person among participants who have had experiences across these developmental periods.

Conclusion

While no single theoretical model emerged in regard to the timing of CSA, teenage experiences of CSA were consistently predictive across all health outcomes indicating more support for the recency model or a sensitive period in adolescence. Further research should examine specific mechanisms through which teenage CSA may impact health among MSM and TW populations such as internalization. Prevention strategies should seek to target MSM and TW. Public health interventions across substance abuse, mental health, and sexual risk for MSM and TW should consider incorporating an understanding of potentially traumatizing experiences to maximize impact.