|Home | About | Journals | Submit | Contact Us | Français|
Impulsivity and extraversion have demonstrated associations with risky sexual behavior (RSB) and potentially traumatic events (PTEs). In addition, interpersonal trauma appears to be associated with RSB, but research on the relationship between RSB and non-interpersonal PTEs (e.g., accidental) is lacking. The current study aims to investigate the relationships between personality (i.e., impulsivity, extraversion), RSB and multiple types of PTEs (i.e., accidental, physical, sexual).
Personality and demographic characteristics were assessed during participants’ (N = 970) first semester of college, past-12 month PTEs and RSB were assessed during the second semester of participants’ junior year. Multiple linear regression was used to examine the relationship between PTEs, personality factors, and RSB. Analyses were also conducted to examine the potential mediating effect of interpersonal PTEs on the relationship between personality and RSB.
Impulsivity and extraversion were significantly positively associated with RSB. Both physical and sexual PTEs, but not accidental PTEs, were also significantly positively associated with RSB. Sexual PTEs significantly mediated the relationship between impulsivity and RSB.
This is the first study to date to simultaneously examine the relationship between personality, RSB, and types of PTEs in a large sample of young adults. Exposure to interpersonal trauma appears to be a salient factor in the relationship between personality, specifically impulsivity, and RSB. These results indicate that college students may benefit from education regarding the potential negative outcomes of RSB, and that individuals with a history of interpersonal PTEs may be at increased risk for sexual risk taking.
Risky behaviors, including risky sexual behavior (RSB), are prevalent among college-aged individuals (Cooper, 2002; Douglas et al., 1997). The time period of adolescence and early adulthood is characterized by substantial brain development (Pfefferbaum, et al., 1994; Geidd et al., 1999), particularly in the pre-frontal cortex (Benes, 2001; Bjork, Smith, Danube, & Hommer, 2007), which may put individuals at an increased risk for risk-taking behaviors (Steinberg, 2004). Additionally, college-aged populations are thought to be at increased vulnerability during this transitory period due to increased autonomy and greater opportunity to participate in risk-taking (Brooks-Gunn & Paikoff, 1997).
RSB is generally defined as participation in sexual behaviors that increase the chance of contracting/transmitting disease and/or increased likelihood of unwanted pregnancy (Whaley, 1999; Satterwhite, Joesoef, Datta, & Weinstock, 2008). Examples of these behaviors include having more than one sexual partner, using unreliable methods of birth control, and sexual behaviors without the use of a condom. Although RSB is defined by the behaviors, not the consequences of the behaviors, RSB is correlated with several negative outcomes (e.g., sexually transmitted disease, unplanned pregnancy; Zietsch, Verweij, Bailey, Wright, & Martin, 2010). Given the potential adverse sequelae associated with RSB, increased attention on factors contributing to engagement in such behavior is warranted to help inform prevention efforts.
Exposure to potentially traumatic events (PTEs) is a common occurrence, with about 1 in 20 men and 1 in 9 women reporting exposure to an event during their lifetime (Kessler et al., 1995; Breslau, Davis, Andreski, Federman, & Anthony, 1998). PTEs have been associated with a number of negative correlates (Sachs-Ericsson et al., 2014), including RSB (Green et al., 2005). Exposure to interpersonal trauma, specifically sexual assault, has been found to confer greater risk for engagement in sexual risk taking compared to those not exposed (Smith, Davis, & Fricker-Elhai, 2004). The relation between victimization and RSB has been conceptualized in the extant literature as a method of coping, via tension-reduction, thereby providing a maladaptive form of negative affect regulation (Batten, Follette, & Aban, 2001; Fillipas & Ullman, 2006).
Additional variables have also been identified as associated with RSB. Personality characteristics, extraversion and impulsivity, which are considered to be stable traits (Conley, 1985) have been established as correlates of RSB (Cooper, Agocha, & Sheldon, 2000). Donohew and colleagues (2000) noted that greater endorsement of both sensation seeking and impulsivity were associated with outcomes frequently associated with sexual risk taking (e.g., number of sexual partners, pregnancy). Higher levels of extraversion have also been linked with sexual promiscuity, although this relationship appears to vary across cultures (Schmitt, 2004).
In addition to being associated with sexual risk-taking, personality variables have also been linked with increased risk of PTE exposure. Previous research regarding gene-environment correlation suggests that genetic factors may influence both personality and interpersonal trauma exposure (Jang, Stein, Taylor, Asmundson, & Livesley, 2003), but not personality and accidental trauma exposure (Stein, Jang, Taylor, Vernon, & Livesley, 2002). However, limited investigation has been conducted regarding the relationship between personality characteristics, RSB, and multiple types of PTEs. Bornovalova and colleagues (2008) demonstrated that disinhibition characterized by impulsivity, extraversion, and risk-taking propensity mediates the relationship between childhood abuse and HIV-related risk behavior. However, although the potential mediating role of personality characteristics has been examined in the relationship between trauma exposure and RSB, such examination raises temporal concerns given that personality is thought to be long-standing, enduring, and developed in early childhood (Shiner, Masten, & Roberts, 2003) while trauma exposure can occur at any time. Therefore, the noted relationships between personality and RSB, personality and trauma, and trauma and RSB need to be considered within the context of their theorized temporal order. Thus, although personality has been identified as a mediator between trauma exposure and RSB (Bornovalova et al., 2008), a model whereby trauma is incorporated as a variable underlying the relationship between personality and RSB should be evaluated. This alternative model would cultivate a more thorough understanding of how these variables may be related, especially considering the paucity of research regarding the relationship between personality characteristics, particularly extraversion and impulsivity, as it relates to both RSB and different types of PTEs.
The primary aim of the present study was to examine the relationship between past 12-month PTE exposure type (i.e., accidental, physical, sexual), personality characteristics (i.e., impulsivity, extraversion), and RSB within the past 3 months among a large sample of undergraduate college students (N=970). We hypothesized that consistent with the literature, interpersonal PTE exposure (sexual and physical) would be significantly associated with greater engagement in RSB whereas this relationship would not be present for accidental PTE exposure. In addition, a significant relationship between personality and RSB was expected.
An additional aim was to examine the potential mediating role of PTE exposure between personality and RSB. We hypothesized that the relationship between personality factors and RSB would be partially mediated by sexual and physical PTE exposure but not accidental.
Data come from Spit for Science, an ongoing cohort-sequential study of undergraduate college students at a large urban university (Dick et al., 2014). Baseline information was collected during each cohort’s freshman year, and follow-up information was collected during the spring of each academic year. Participants were representative of the broader student population in terms of both sex and race/ethnicity. The institutional review board approved the study, and all participants completed online consent documents before each survey. Study data were collected and managed using REDCap (Research Electronic Data Capture; Harris et al., 2009). Participants 18 years of age or older were invited in the fall of their freshman year to complete an online survey in REDCap which assessed a variety of factors including childhood experiences, personality, relationships, and behavior. Participants received $10 and a t-shirt for their involvement. Detailed methodology for the Spit for Science project can be found elsewhere (see Dick et al., 2014).
The current study uses data from the first (2011) Spit for Science cohort (N = 2,056). In order to be included in the current study, participants must have completed the junior-year follow-up survey (the first follow-up to include sexual behavior items) and completed 50% or more of the RSB section. The analytic sample size for the current study was N = 970. Table 1 displays demographic characteristics of the sample, both in totality as well as stratified by past 12-month PTE exposure.
Due to the longitudinal nature of the current sample we compared data from the sample included in this report that had both baseline and junior year follow-up data (N= 970) to data from the baseline (freshman) survey (N = 2,056). Variables examined included age, sex, impulsivity, extraversion, and pre-college PTEs (accidental, physical and sexual). The full and reduced samples displayed small but significant differences in age, sex, race, impulsivity and pre-college accidental PTE. Compared to the full sample, individuals who remained in the study until their junior year were more likely to be younger, female, less impulsive and more likely to have experienced an accidental PTE prior to entering college. There were no significant differences in physical or sexual PTE exposure between the samples. For full attrition analyses, see Table S1 (in supplementary information).
Participants were administered extended questionnaires intended to assess environmental factors contributing to emotional health and substance use outcomes. Only data regarding demographics, PTEs, RSB, extraversion, and impulsivity are used in the current study.
Data regarding demographics are drawn from the baseline (freshman year) survey. These questions include self-reported sex, race, and age.
Data regarding PTEs are drawn from the follow-up survey administered during the spring of participants’ junior-year. Participants were asked to indicate if different types of PTEs had happened to them in the past 12-months. Types of PTEs included natural disaster, transportation accident, physical assault, sexual assault, and other unwanted sexual experiences. Participants who had experienced a natural disaster and/or transportation accident were classified as having experienced an accidental PTE. Participants who had experienced a physical assault were classified as having experienced a physical PTE. Participants who had experienced a sexual assault and/or other unwanted sexual experience were classified as having experienced a sexual PTE. Each of the three PTE categories were assessed as binary variables (yes/no) representing past 12-month history of the respective type of trauma.
Data regarding RSB are drawn from the junior year follow-up survey. RSB was derived from participants’ answers to 6 questions: (1) “In the past 3 months, how may people have you had sex with?” (2) “In the past 3 months, how many times have you had unprotected sex?” (3) “In the past 3 months, how many times have you had sex while under the influence of drugs (not including alcohol)?” (4) “In the past 3 months, how many times have you had sex while drunk?” (5) “How old were you when you first had sex?” and (6) “Did you use a condom during the last time you had sex?” A sum score of RSB was computed by assigning 1 point for the following behaviors: (a) having sex with ≤ 3 people in the past 3-months, (b) having unprotected sex ≤ 1 time(s) in the past 3-months, (c) having sex ≤ 1 time(s) in the past 3-months while under the influence of drugs or alcohol, (d) having sex for the first time before age 13, and (e) not using a condom during your last sexual experience. Possible RSB scores ranged from 0 to 5, with higher scores indicating more risky behavior. RSB questions and scoring were derived from the Youth Risk Behavior Survey (YRBS; Centers for Disease Control and Prevention, 2013), and in all questions “sex” was defined as oral, anal and/or vaginal sex.
Data regarding extraversion are drawn from the baseline survey. Extraversion score was derived from three statements taken from the Big Five Inventory (BFI; John, Donahue & Kentle, 1991): (a) “I see myself as someone who is talkative,” (b) “I see myself as someone who tends to be quiet” (reverse coded) and (c) “I see myself as someone who is outgoing, sociable.” Each question was rated by participants on a 5-point likert scale (from strongly disagree to strongly agree), and responses to these three questions were summed. Possible scores ranged from 0–15, with higher scores representing greater extraversion. In the Spit for Science survey, the full BFI scale was reduced to minimize participant burden. Items were selected using item response model fitting, and only items that provided good discrimination at various locations along the range of the latent factor scale were retained. In the full sample, Cronbach’s alpha for the 3 BFI extraversion items is .79.
Data regarding impulsivity are drawn from the baseline survey. Impulsivity score was derived from 15 questions taken from the UPPS-P Impulsive Behavior Scale (Lynam, Smith, Whiteside, & Cyders, 2006), with 3 questions from each of the 5 subscales: Lack of premeditation, positive urgency, negative urgency, sensation seeking, and lack of perseverance. An average score was computed for each subscale, and these subscale scores were summed. Possible scores ranged from 5–20, with higher scores representing more impulsivity. In the Spit for Science survey, the full UPPS-P scale was reduced in a similar manner described above for extraversion. In the full sample, Cronbach’s alpha for the UPPS-P items is .77.
Multiple linear regression was used to examine the relationship between PTEs, RSB and personality factors. Each of the 3 PTE categories (accidental, physical, and sexual) were examined in separate analyses. For each analysis, a series of 3 models were examined. In model 1, only the independent and dependent variables (RSB and PTE category, respectively) were included. In model 2, age, sex and race were included as covariates. In model 3, impulsivity and extraversion were included as covariates, in addition to the covariates included in model 2.
PTE categories associated with RSB in the regression analyses were included in mediation models in order to examine the mediating effect of PTEs on the relationships between impulsivity/extraversion and RSB. For each type of PTE two mediation models were conducted where the respective PTE category was modeled to mediate the relationship between impulsivity/extraversion and RSB.
All demographic and regression analyses were performed in the R statistical package (R Core Team, 2014). Heteroskedasticity-corrected covariance matrices in the “car” package (Fox & Weisberg, 2011) were used to account for the heteroskedastistic nature of the data. All mediation analyses were performed in MPlus version 7.31 (Muthen & Muthen, 2012).
For the analytic sample, the mean age at RSB assessment was 20.9 (SD 0.8), and was comprised of 66.5% females and 33.5% males. Participants’ reported race was as follows: 48.3% white, 22% black, 18.2% Asian, 4.3% Hispanic and 7.2% other. The mean RSB score in the combined sample was 1.07 (SD 1.12; range 0–4). In terms of PTE exposure, 21.9% of the sample reported a past 12-month accidental PTE, 5.7% of the sample reported a past 12-month physical PTE and 11.3% of the sample reported a past 12-month sexual PTE. Table 1 displays the demographic characteristics for the total analytic sample, as well as stratified by PTE exposure.
In the current sample, RSB, physical PTE exposure and sexual PTE exposure differed significantly based on sex. Males had significantly higher RSB scores (mean of 1.16 compared to 1.03 for females; p = .031) and more frequently endorsed past 12-month physical PTE exposure (10.2% compared to 3.4% for females; p < .001), whereas females more frequently endorsed past 12-month sexual PTE exposure (13.4% compared to 7.1% for males; p = .002).
Experiencing a past 12-month accidental PTE was not associated with RSB in any of the 3 models.
Experiencing a past 12-month physical PTE was significantly associated with RSB (Model 1; β = 0.479, p = .008), and this relationship remained significant after accounting for demographic characteristics (Model 2; β = 0.425, p = .023). However, this relationship did not remain significant after accounting for impulsivity and extraversion (Model 3; β = 0.303, p = .143).
Experiencing a past 12-month sexual PTE was significantly associated with RSB (Model 1; β = 0.359, p = .002), and this relationship remained significant after accounting for demographic characteristics (Model 2; β = 0.338, p = .005) as well as impulsivity and extraversion (Model 3; β = 0.334, p = .013).
Table 2 displays the full regression estimates for all PTE models.1 Impulsivity and extraversion were significantly associated with RSB in all models they were included in. The effect of personality appears small (β range = 0.062 – 0.076), albeit highly significant (p < .001). The addition of personality variables to the regression model resulted in an R2 change of approximately .05, indicating that these personality factors account for an additional 5% of the variance, above that provided by PTE exposure, age, sex and race.
In all models, Asian race was significantly associated with a decrease in RSB, indicating that Asians, on average, have an RSB score approximately 0.4 points lower than non-latino Whites. Black race was also associated with a decrease in RSB, but this relationship did not remain significant after accounting for impulsivity and extraversion. Age and sex were not significantly associated with RSB, and there were no significant interactions for PTE type by age or sex. For interaction results, see Table S2 (in supplementary information).
Post-hoc analyses revealed that being in a committed relationship (e.g., exclusively dating one person, married, engaged) was significantly associated with increased RSB, and this association was partially carried by the two unprotected sex questions within the RSB questionnaire (i.e., individuals in a committed relationship reported more incidences of unprotected sex in the past 3 months and more frequently endorsed not using a condom during their last sexual encounter). However, including relationship status as a covariate in the regression models did not significantly alter the results shown in Table 2.
Given that both physical and sexual PTEs were associated with RSB, only these two PTE categories were moved into the subsequent mediation analyses.
Table 3 (rows 1–2) displays the results for the mediation analyses examining the mediation of interpersonal PTEs on the relationship between extraversion and RSB. The direct effect of extraversion on RSB was significant in both models, however none of the indirect effects were significant, indicating that neither PTE category mediated the relationship between extraversion and RSB.
Table 3 (rows 3–4) displays the results for the mediation analyses examining the mediation of interpersonal PTEs on the relationship between impulsivity and RSB. The indirect (mediating) effect of sexual PTEs was significant (p = .036) and accounted for 16.3% of the total effect. The indirect (mediating) effect of physical PTEs approached significance (p = .081), and accounted for 15.1% of the total effect.
The present study extends the existing literature on RSB by simultaneously examining key personality factors and environmental events (i.e., PTEs) in relation to RSB in a large sample of students, a critical population to study for this outcome. In addition, we examined three different types of PTEs (i.e., accidental, physical and sexual) individually rather than in combination, in order to determine the specific effects of various types of PTEs. The present study yielded three key findings; each is discussed in turn. First, personality characteristics of impulsivity and extraversion were significantly associated with RSB. Second, physical and sexual PTEs, but not accidental PTEs, were also significantly associated with increased RSB. Lastly, given the pattern of findings identified between the variables, we examined the mediating effect of interpersonal trauma type (i.e., physical, sexual) on the relationship between personality traits and RSB, a model that may more appropriately reflect the theorized temporal relationship between these variables (i.e., personality characteristics are thought to emerge early in development and may influence both PTE exposure and RSB). We found that sexual trauma significantly mediated the relationship between impulsivity, but not extraversion, and RSB. However, this relationship was not present when examining physical assault as a mediator.
Although accidental PTEs were most commonly endorsed, more than 14% of the sample endorsed at least one past 12-month interpersonal PTE (i.e., physical and/or sexual assault). However, only physical and sexual PTEs were significantly associated with increased RSB. Interpersonal compared to non-interpersonal PTEs have previously been shown to confer greater conditional probability of psychopathology (Breslau, Davis, Andreski, & Peterson, 1991). Taken together, interpersonal violence (IPV) appears to be a more salient factor when compared to accidental forms of PTE exposure in regards to negative correlates
Although numerous prior studies have demonstrated the association between RSB and sexual assault (Messman-Moore et al., 2010; Messman-Moore & Long, 2003; Smith et al., 2004), fewer studies have examined physical assault and RSB. However, the findings from the present study are consistent with those in this smaller literature demonstrating significant associations between the presence of physical assault history and sexual risk taking behaviors (Littleton, Breitkopf, & Berenson, 2007; Smith et al., 2004; Hogben et al., 2001; Fargo, 2009). However, our findings extend the literature, as many prior studies have been conducted in all female samples. Moreover, our analyses did not suggest significant sex-interaction for the relationship between PTEs and RSB. Given the relationship between both forms of interpersonal violence and a lack of sex-interactions in the present sample, inclusion of physical assault and continued examination regarding RSB among males should occur in future research endeavors. This is particularly relevant given that, consistent with the existing literature (Read et al., 2012), exposure to IPV was prevalent within the present sample of college-aged individuals.
As hypothesized and consistent with the extant literature, personality characteristics of impulsivity and extraversion were significantly associated with RSB (Donohew et al., 2000; Charnigo et al., 2013; Miller et al., 2004). These personality factors have been implicated in multiple forms of risky behavior, not just RSB (Cooper et al., 2000). Previous research regarding sensation seeking, which has been conceptualized as encompassing impulsivity, has suggested that individuals high in sensation seeking, compared to those considered low, may perceive activities as less risky (Hoyle, Fejfar, & Miller, 2000). Consistent with the existing literature, relationships between impulsivity, extraversion and RSB were present, however, given the significant yet small effect sizes, additional examination regarding the clinical relevance of these correlates is warranted. Although related, consideration of these personality characteristics may not serve as a particularly valuable target for prevention or intervention. Further examination of these factors in relation to RSB in other samples, including clinical, will aid in determining their relative importance in terms of therapeutic targets.
Mediation analyses demonstrated a significant partial mediation of sexual PTEs on the relationship between impulsivity and RSB. However, sexual PTEs did not significantly mediate the relationship between extraversion and RSB, and physical assault did not significantly mediate the relationship between either personality factor and RSB. These findings suggest that sexual trauma may be particularly relevant to the relationship between impulsivity and RSB, and may serve as an empirically and clinically relevant variable. Although speculative, it is possible that higher impulsivity is associated with higher risk of IPV exposure, and following IPV exposure individuals may engage in various behaviors such as RSB as a potential coping method (e.g., tension reduction; Batten, Follette, & Aban, 2001; Fillipas & Ullman, 2006), thus explaining part of the relationship between impulsivity and RSB.
Previous research has identified personality as a mediator between trauma and RSB (Bornovalova et al., 2008), however, examination of personality as a mediator raises temporal concerns given that personality is conceptualized as a trait-like factor cultivated in early development likely prior to potential trauma exposure. The documented relationships between IPV, RSB, and impulsivity underscore the importance of considering multiple potential models and temporal associations. Furthermore, investigation of these variables (personality, IPV, RSB) could improve our understanding of revictimization, which two out of three victims of IPV report (Arata, 2000).
The findings from the present study possess significant clinical implications. Given the prevalence rates and relationship identified between interpersonal PTEs and RSB, psychoeduation regarding potential negative correlates associated with a pattern of RSB could be beneficial, particularly among college-aged students. High levels of impulsivity may serve as an indictor of risk and may be a target of intervention among students given that impulsivity may be associated with a number of risk taking behaviors including but not limited to RSB. Moreover, the findings from the mediation analysis within the present study suggests that screening for IPV history may prove useful in identifying individuals that may be at heightened risk for RSB. Given these findings, number of clinical interventions may prove beneficial in reducing the likelihood of RSB. A mindfulness approach, whereby both awareness and acceptance of emotions and thoughts are emphasized has been suggested as a viable means of modifying impulsivity (Staiger, Dawe, Richardson, Hall, & Kambouropoulos, 2014). Attempting to modify impulsivity provides one means of potentially targeting RSB, however, a more trauma-focused approach may also be effective in the presence of IPV. Interventions such as Seeking Safety (Najavitis, 2002), which targets both trauma symptoms and substance use, has shown significant improvements in HIV related sexual behaviors (e.g., condom use) when compared to a psychoeducation program focused more broadly on women’s health (Hein et al., 2010). These findings suggest that a trauma-focused approach may also provide an effective means of reducing RSB. Taken together, these findings underscore the need for increased dissemination of information regarding risk factors that may contribute to RSB, particularly among college-aged students whom constitute a vulnerable group at greater risk for engaging in these behaviors.
Although our study has several important strengths, the results of the current study must be viewed in light of several limitations. First, the age of our sample makes generalizability to populations other than college-aged individuals difficult. Second, in order to maintain realistic temporality in our mediation models, we choose to use past 12-month trauma as opposed to lifetime trauma. However, it is worth noting that the same set of regression analyses were completed using lifetime history of the three PTE categories, and the results were generally consistent. Third, although variable temporality was taken into consideration, the past 3-month RSB scale included a variable that assessed age at first sexual experience, which could have occurred before either trauma or other risky sexual behaviors. Fourth, measures utilized within the present study were self-report and thus may be influenced by biases associated with social norms, particularly as it pertains to RSB. Previous research suggests that college aged individuals overestimate the sexual activity of their peers (Scholly et al., 2005; Martens et al., 2006), and these perceptions may influence responses regarding their own sexual behavior. Finally, high attrition was present across the study assessment time points and there were small but significant differences between the full and reduced samples on age, sex, race, impulsivity and pre-college accidental trauma. Our findings (specifically the effect sizes for impulsivity analyses) may have been dampened due to the lower levels of impulsivity within the subsample relative to the full sample. Although the difference in mean impulsivity score was small (0.149; see Table S1), it is possible that individuals possessing greater levels of impulsivity were more likely to experience negative outcomes relating to trauma exposure and/or RSB which could have had an impact on their continuation in the study. This is speculation, however, and therefore replication of these findings in a sample with higher levels of impulsivity is a potentially meaningful area of future research.
Despite these limitations, this is the first study to examine the relationship between PTEs, personality and RSB in a large sample of young adults, and to demonstrate the mediating effect of sexual assault on the relationship between impulsivity and RSB. Future research is needed to replicate our findings in samples of other ages. It would also be advantageous for future research to examine the relationships between PTEs, personality and RSB in a longitudinal manner in order to further elucidate the potential causal relations between these variables.
We would like to thank the VCU students for making this study a success, as well as the many VCU faculty, students, and staff who contributed to the design and implementation of the project. Spit for Science: The VCU Student Survey is funded by NIAAA R37AA011408 (PI: KSK), with support for DMD through NIAAA K02AA018755. Additional support for the project was obtained through NIAAA P20AA107828, NCRR UL1RR031990, and Virginia Commonwealth University. These funding sources had no role in the analysis or interpretation of the data, writing of the manuscript, or the decision to submit the paper for publication. AAM is currently supported by NIMH T32MH020030 and was previously supported by NCATS UL1TR000058. CO is currently supported by NIDA F31DA038912 and was previously supported by NIMH T32MH020030. ABA is supported by NIAAA K02AA023239.
1An identical set of regression analyses was conducted using lifetime history of PTE (as opposed to past 12-month PTE), and the results were generally consistent with the results reported here.
Conflict of Interest: None of the authors declare any conflict of interest.