Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Sep 1.
Published in final edited form as: Psychol Addict Behav. 2019 Jul 25;33(6):529–539. doi: 10.1037/adb0000489

Sexual Victimization Histories and Substance Use Trajectories During the Transition Out of College

Jessica A Blayney 1, Matthew Scalco 2, Sharon Radomski 3, Craig Colder 4, Jennifer P Read 5
PMCID: PMC6742576  NIHMSID: NIHMS1037089  PMID: 31343200

Abstract

The college years are a period of peak vulnerability for sexual victimization (SV) and substance misuse. During college, students with SV histories report riskier substance use patterns, yet little is known about the influence of SV on substance use behaviors as students begin to transition away from the college environment. This was the purpose of the present study. College seniors (N = 480; 61% female) reported on their alcohol and drug use behaviors across five time points spanning one calendar year. For many, this year included the transition out of college. Latent growth curve analysis was used to determine whether trajectories for alcohol and drug use as well as alcohol and drug consequences differed based on SV histories (no SV, pre-college SV, college SV, pre-college + college SV). Results revealed that, at the start of senior year, young adults with SV histories reported greater substance use and consequences relative to non-victimized peers. Over the year, SV histories were associated with steeper declines in substance use and consequences. Despite the declines, those who were revictimized across developmental time periods (pre-college + college SV) consistently reported higher alcohol use and consequences across the transitional year, although this did not replicate for other drugs. In sum, though alcohol and other drug involvement among those with SV histories decreased over time, pre-college + college SV histories continued to be a persistent risk factor for heavier alcohol use behaviors.

Keywords: sexual victimization, alcohol use, drug use, college, developmental transitions

Introduction

Frequent and even heavy substance use in college is common (American College Health Association, 2014; Johnston, O’Malley, Bachman, Schulenberg, & Miech, 2015), and has been linked to a number of negative outcomes including overdose, mental and physical health problems, and risk for substance use disorders (White & Hingson, 2013). Those whose substance misuse persists after they leave college are at risk for progression toward longer-term substance use problems (Jackson, Sher, Gotham, & Wood, 2001; Sher & Gotham, 1999; White, Labouvie, & Papadaratsakis, 2005). Delineation of the factors that influence hazardous substance use patterns - not just during college, but also beyond college - can aid in the earlier identification of and intervention with those at risk. A growing literature suggests that sexual victimization may be one important risk factor.

Young adulthood represents a period of peak vulnerability for sexual victimization (SV), and the college environment appears to confer unique risk (Abbey, 2002). Approximately 20–25% of women and 5–10% of men experience SV during college (Blayney, Read, & Colder, 2016; Krebs, Lindquist, Warner, Fisher, & Martin, 2009; Mouilso, Fischer, & Calhoun, 2012; Snipes, Green, Benotsch, & Perrin, 2014; Tewksbury & Mustaine, 2001). A sizeable proportion of young adults also enter into college with SV histories from childhood or adolescence (~33%; Aosved, Long, & Voller, 2011; Gidycz et al., 2007; Griffin & Read, 2012), which places these students at higher risk for revictimization during the college years (Humphrey & White, 2000). The high rates of SV, including revictimization, on college campuses continues to be a major concern. Understanding the longer-term consequences of SV, including substance misuse, is of critical importance.

Numerous studies have linked SV and alcohol misuse. In cross-sectional work, college students with SV histories report greater alcohol use and alcohol-related consequences relative to their non-victimized peers (e.g., Bedard-Gilligan, Kaysen, Desai, & Lee, 2011; Larimer, Lydum, Anderson, & Turner, 1999; Turchik, 2012). This link is especially strong for those who experienced revictimization (Messman-Moore, Long, & Siegfried, 2000; Ullman & Najdowski, 2009). Prospective examinations of SV and alcohol outcomes are far more scarce. Of those conducted, the majority of studies have focused on alcohol as a risk factor for SV, with data showing that drinking dramatically increases prospective risk for victimization (Kaysen, Neighbors, Martell, Fossos, & Larimer, 2006; Mouilso et al., 2012; Testa & Livingston, 2009). Yet, a handful of studies also have examined how SV histories prospectively increase risk for alcohol use (Griffin, Read, & Wardell, 2013; Testa, Hoffman, & Livingston, 2010; Testa & Livingston, 2009). To date, much of this research is limited to the transition into college (e.g., Mouilso et al., 2012; Parks, Romosz, Bradizza, & Hsieh, 2008; Testa et al., 2010), a time of intensified risk for substance misuse. To our knowledge, the role of SV in alcohol or drug use behaviors during other important developmental transitions, such as the transition out of college, have yet to be examined. Young adults often show a pattern of gradual decline in hazardous drinking patterns (e.g., “maturing out”) as they transition out of college, but not all do (Gotham, Sher, & Wood, 1997). Given the elevated alcohol use patterns among those with SV histories during college, it is unclear to what extent this vulnerable group might also show normative declines during this next important developmental juncture. This is a significant gap in the literature. Further, very little research has examined SV in association with drugs other than alcohol – another notable gap in the knowledge base. With the present study, we sought to address these gaps.

The Present Study

The objective of the present study was to examine the influence of SV histories on substance trajectories as young adults are completing their college careers and preparing for independent adulthood. An important contribution of this study was the investigation of trajectories for both alcohol and other drug use as well as alcohol and other drug consequences. Because revictimization is common (Blayney et al., 2016; Humphrey & White, 2000; Messman-Moore et al., 2000), we also sought to examine the specific impact of repeated victimization as well as the timing of victimization (e.g., pre-college SV, college SV, or both [pre-college + college SV]). Consistent with research focused on the transition out of college (e.g., Gotham et al., 1997; Sher & Gotham, 1999; White et al., 2005), we hypothesized that (1) the overall sample would show reductions in alcohol and other drug use and consequences over the one-year period, but that (2a) those with SV histories would report higher levels of alcohol and other drug use and consequences over time and (2b) those who experienced revictimization across developmental time periods (pre-college + college SV) would be at highest risk. Given the literature on the significance of developmental transitions, we were also interested in whether associations between SV and substance outcomes differed for those graduating college versus those who remained in the college environment. Accordingly, we tested this in exploratory analyses.

Methods

Participants and Procedures

The present study was drawn from a larger longitudinal study on trauma and substance use in college. Using a staggered cohort design, three cohorts of college freshmen were invited to a screening survey at college matriculation (i.e., upon entry into college). A total of 3,391 students completed this initial screening. Those invited to participate in the baseline survey included both trauma-exposed (n = 649) and a random sample of non-trauma-exposed (n = 585) students. A total of 997 students completed the baseline. This sample was then assessed four times a year for up to five additional years after college entry (see Figure 1 for study timeline). For each completed survey, participantswere compensated with gift cards. All procedures were approved by the University’s Institutional Review Board.

Figure 1.

Figure 1

Open in a new tab

Study timeline

Of the three cohorts included in the larger study, two (n = 665) were followed into the 4th and 5th post-matriculation years. In the larger study, these two cohorts were recruited one year apart. In order to examine SV and the transition out of college, only those students who reported entering their fourth year of college as seniors and who provided data on college SV (n = 500; 62% female) were included in the present study. Time points during the fourth college year included September, December, February, and April (T1–T4). Participants were assessed again in September of the following year (T5). Of our sample, 20 participants (65% female) with no prior history of college SV (during the first three post-matriculation years of college) reported victimization during senior year. In order to establish temporal precedence between SV and substance outcomes, these 20 participants were dropped from analyses. Thus, the final N was 480 (61% female). The sample was 73% White (n = 349), 16% Asian (n = 76), 5% African American (n = 25), 3% Hispanic/Latino (n = 15), and 3% multi-racial or other (n = 15). At the start of senior year, the average age was 22.09 years old (SD = 0.35) and 96% (n = 458) identified as heterosexual.

Measures

Figure 1 depicts the time points at which study constructs were assessed. Assessments occurred during the school year, which ran from late August to early May.

Sexual victimization history

Pre-college sexual victimization.

Sexual victimization that occurred prior to college was examined at the baseline of the larger study (at college matriculation). Using a modified Sexual Experiences Survey (R-SES; Testa, VanZile-Tamsen, Livingston, & Koss, 2004), participants reported on unwanted sexual experiences that occurred in their lifetimes. Sexual victimization included unwanted sexual contact, attempted rape, and completed rape (response options: 0 = no, 1 = yes). Based on endorsement of SV prior to college, responses were coded as 0 = no pre-college SV and 1 = pre-college SV.

College sexual victimization.

Sexual victimization after college matriculation was assessed at the end of the 4th (cohort 3) and 5th (cohort 1) post-matriculation years using a modified R-SES. Participants reported on unwanted sexual experiences that occurred since the start of college (response options: 0 = 0 times to 4 = 4+ times). In addition, participants specified the year(s) in college that victimization(s) occurred. College SV history was based on any endorsement of SV from college matriculation through the end of the third post-matriculation year (e.g., during the first three years of college). To further characterize frequency, college SV was coded as 0 = no college SV, 1 = college SV during 1 college year only, and 2 = college SV during 1+ college years.

Substance use

Alcohol use.

Alcohol use (T1–T5) was examined using a grid based on the Daily Drinking Questionnaire (Collins, Parks, & Marlatt, 1985). Participants were asked to report the typical number of drinks they consumed on each day in the past month (i.e., typical Monday, Tuesday, etc). From this, drinking days were dichotomized (0 = no, 1 = yes) and then summed to represent typical number of drinking days in the past month. A Quantity × Frequency index was calculated for each time point to represent past month alcohol use.

Other drug use.

At each time point (T1–T5), participants also reported on their use of drugs other than alcohol. This was assessed with items that inquired how frequently each of seven drugs (marijuana, cocaine, stimulants, inhalants, sedatives, hallucinogens, and opioids) were used in the past month. Response options ranged from 0 = never in the past month to 6 = every day. For analyses, we combined responses across the seven drugs and took the mean to represent past month drug use.1

Substance-related consequences

Alcohol consequences.

Alcohol consequences (T1–T5; α = .91 to .93) were assessed using the Young Adult Alcohol Consequences Questionnaire (YAACQ; Read, Kahler, Strong, & Colder, 2006). Participants were presented 48 items and asked which consequences they experienced in the past month (response options: 0 = no, 1 = yes). A sum score was calculated for each time point to indicate the total number of alcohol consequences experienced.

Other drug consequences.

Other drug consequences (T1–T5; α = .86 to .88) were assessed using a modified Brief YAACQ (B-YAACQ; Kahler, Strong, & Read, 2005). The BYAACQ includes 24 items from the full YAACQ measure and was used to examine the consequences associated with drug use in the past month (response options: 0 = no, 1 = yes). A sum score was calculated for each time point to indicate the total number of drug consequences experienced.

Data management and analyses

Variables were assessed for normality and outliers, with identified outliers recoded to the next highest, non-outlying value (Tabachnick & Fidell, 2013). Latent growth curve models (LGCM) were conducted to examine change in substance use and consequences over five time points. Full Information Maximum Likelihood was used, which allowed us to statistically handle missing data. Missing data was minimal and ranged from 0.5%−2.0% across time points. Analyses were run in Mplus version 7.2 (Muthen & Muthen, 1998–2012). Given that substance outcomes were still skewed and kurtotic, the maximum likelihood robust estimator was used and nested models were compared with the Satorra-Bentler scaled chi-square difference test (Satorra & Bentler, 2001).

First, unconditional LGCM’s were run to establish the form of growth and determine if there was individual variability in growth for alcohol and drug outcomes (see Figure 2 for conceptual model). Fixed time intervals were coded as 0, .2, .4, .6, and 1. As such, the intercept reflects baseline levels of substance use or consequences at T1 (start of senior year). Second, multiple group structural equation models were run to determine whether parameters in the model differed by graduation status. Graduation occurred around T4 and so we tested the multiple group models in two ways: (1) coding the intercept at T1 (start of senior year) and (2) coding the intercept at T5 (after expected graduation). Third, conditional LGCM’s were run in which pre-college SV, college SV, the interaction between pre-college SV and college SV were added as predictors of the latent intercept and growth factors for each model. All models also included gender and prior substance use or consequences. Independent variables were mean centered to test interactions (Aiken & West, 1991). Cut-offs for ‘good’ model fit cannot be generalized across all models and so ranges were used (RMSEA: poor ≥ .08, acceptable = .05–.07, excellent ≤ .05; CFI/TLI: poor ≤ .90, acceptable = .90–.94, excellent ≥ .95; SRMR: poor ≥ .09, acceptable = .06–.09, excellent ≤ .06; Hu & Bentler, 1999; Marsh, Hau, & Wen, 2004).

Figure 2.

Figure 2

Open in a new tab

Conceptual model Note. All models controlled for prior substance use or consequences.

Results

Descriptive statistics

Sample characteristics, means, and standard deviations for study variables are presented in Table 1. Approximately 7% of men (n = 32) and 29% of women (n = 138) reported SV before entering college. Across the sample, 6% of men (n = 31) and 26% of women (n = 127) experienced SV during the first three years of college. A sizable proportion of those victimized in college (8% men, n = 12; 39% women, n = 54) reported SV across multiple college years. Of those with pre-college SV histories, 58% were revictimized during college (8% men, n = 14; 50% women, n = 84).

Table 1.

Sample characteristics and substance use outcomes

n (%)
Total sample
(n = 480)		 College women
(n = 294)		 College men
(n = 184)
Pre-college SV
No 308 (64%) 156 (32%) 152 (32%)
Yes 170 (36%) 138 (29%) 32 (7%)
College SV
No 320 (67%) 167 (35%) 153 (32%)
Yes 158 (33%) 127 (26%) 31 (6%)
Pre-college + college SV
No 380 (80%) 210 (44%) 170 (35%)
Yes 98 (20%) 84 (18%) 14 (3%)
M (SD)
T1 T2 T3 T4 T5
Alcohol use (quantity) 7.91 (8.74) 6.24 (7.46) 6.84 (8.10) 6.88 (7.82) 6.52 (7.21)
Alcohol use (frequency) 1.85 (1.55) 1.68 (1.63) 1.67 (1.45) 1.73 (1.48) 1.87 (1.66)
Alcohol consequences 3.99 (5.90) 3.45 (5.51) 3.52 (5.46) 3.56 (5.53) 2.90 (4.38)
Other drug use 0.09 (0.21) 0.09 (0.22) 0.07 (0.19) 0.07 (0.19) 0.07 (0.18)
Other drug consequences 0.47 (1.49) 0.41 (1.39) 0.33 (1.14) 0.25 (0.97) 0.23 (0.91)
Open in a new tab

Note. N’s ranged from 462–480 due to missing data. SV = sexual victimization. T1 = September. T2 = December. T3 = February. T4 = April. T5 = September.

At the start of senior year (T1), the sample reported drinking 7.91 drinks (SD = 8.74) in a typical week and experiencing approximately four consequences (M = 3.99, SD = 5.90) in the past month. Approximately 20% of the sample (n = 96) used other drugs, and for those who used, the average number of drug consequences was 2.59 (SD = 2.65). One year later (T5), 66% (n = 318) of the sample had graduated college. Graduation rates did not differ among those with and without a SV history, χ2 (1) = 0.65, p = .42. At T5, roughly 27% (n = 130) of the sample was working full-time, 43% (n = 206) working part-time, and 30% (n = 144) were unemployed. Participants reported 6.52 drinks (SD = 7.21) in a typical week and 2.90 consequences (SD = 4.38) in the past month. Approximately 15% (n = 73) used other drugs, and those who did experienced 1.71 consequences (SD = 1.89) in the past month.

Unconditional LGCM’s – Substance use behaviors across the sample

Consistent with the transition out of college literature, our first hypothesis was that the overall sample would show a general decline in alcohol and other drug use and consequences over the one-year assessment. Models (see Figure 2) were specified such that only the variance for the intercept was estimated and chi-square difference tests were used to determine if adding linear, quadratic, or cubic effects of time improved model fit.

For alcohol use, adding a linear and quadratic effect resulted in a negative variance estimate (Heywood case) for the slope factor, suggesting that the model was over-parameterized. Accordingly, fixed effects were explored. Sequentially adding a fixed linear, quadratic, and cubic effect improved model fit (X2 [1, N = 480] = 10.40, p < .001; X2 [1, N = 480] = 5.30, p = .03; and X2 [1, N = 480] = 27.83, p < .001). Modification indices suggested the addition of a covariance between T3 and T4 measurement occasions, which included part of winter break and spring break. This estimation made conceptual as well as empirical sense and so we added this covariance. The final model provided excellent fit to the data: x2 [9] = 19.82, p = .02, RMSEA= .05 (90% C.I. = .02 – .08); CFI/TLI = .98/.98; SRMR = .06. A large amount of variance was explained in each repeated measure of alcohol use by the growth factors (R2 range: .68 – .76). The mean of the intercept factor suggested that T1 alcohol Quantity × Frequency was just over 11 (M = 11.14, p < .001). Means of the growth factors revealed a negative linear slope (M = −14.93, p < .001), a positive quadratic slope (M = 34.55, p < .001), and a negative cubic slope (M = −21.44, p < .001), suggesting a sharp decline in alcohol use between T1 and T2 (T1 M = 11.18; T2 M = 9.27), a slight increase at T4 (M = 9.97) and then a slight decrease at T5 (M = 9.32).

For alcohol consequences, chi-square tests indicated that adding a linear and quadratic effect improved model fit (X2 [3, N = 480] = 38.04, p < .001 and X2 [4, N = 480] = 24.74, p< .001). After adding the quadratic effect, the linear slope variance became non-significant and was set to 0. Similar to the alcohol use model, modification indices suggested adding a covariance between T3 and T4. A nested test suggested that this addition improved model fit (X2 [1, N = 480] = 8.50, p < .001). The final model provided excellent fit to the data: x2 [8] = 12.33, p = .14, RMSEA = .03 [90% C.I. = .00 – .07]; CFI/TLI = .99/.99; SRMR = .03. A large amountof variance was explained by the growth factors (R2 range: .65 – .79). The quadratic factor correlated negatively with the intercept (r = −.58) suggesting that those who started with higher T1 alcohol consequences had larger decreases in consequences across the year. At T1, the mean number of alcohol consequences was just under 4 (M = 3.84, p < .001). Means of the growth factors indicated a non-significant negative linear (M = −.72, p = .27) and quadratic effect (M = −.16, p = .80), indicating that alcohol consequences remained relatively stable between T1–T5, although there was variability in the trend.

For other drug use, adding a linear effect improved model fit, while adding a quadratic effect did not (X2 [3, N = 480] = 21.12, p < .001 and X2 [2, N = 480] = 0.81, p = .99). The final model provided an excellent fit to the data: x2 [10] = 17.80, p = .06, RMSEA = .04 [90% C.I. = .00 – .07]; CFI/TLI = .98/.98; SRMR = .05 and a large amount of variance was explained by the growth factors (R2 range: .72 – .85). While there was significant variability in the intercept factor, the slope factor variance fell just short of conventional criteria for significance (p = .06). The linear slope had a significant negative correlation with the intercept (r = −.57) indicating that those who started with higher T1 drug use had larger decreases in use over time. At T1, the level of drug use was just over 0, suggesting that drug use in the overall sample was relatively infrequent (M = 0.09, p < .001) and declined over time (M = −0.02, p < .001).

For other drug consequences, adding a linear effect improved model fit, while adding a quadratic effect did not (X2 [3, N = 480] = 62.79, p < .001 and X2 [2, N = 480] = 7.56, p = .10, respectively). The final model provided an excellent fit to the data: x2 (10) = 14.45, p = .15, RMSEA = .03 [90% C.I. = .00 – .06]; CFI/TLI = .98/.98; SRMR = .07. A moderate to large amount of variance was explained by the growth factors (R2 range: .44 – .79). There was significant variability in the intercept and linear slope factor (ps < .01) and the linear slope had a significant negative correlation with the intercept (r = −.93). This would suggest that those who started with more T1 drug consequences had larger decreases in consequences over time. At T1, the mean number of drug consequences experienced was less than 1 (M = 0.44, p < .001) and declined over time (M = −0.24, p < .001).

To test whether these models differed by graduation status, chi-square difference tests compared models where parameters were constrained to be the same across graduation status and then freely estimated. Nested tests suggested that trajectories of alcohol use (X2 [11, N = 480] = 10.97, p = .45), alcohol consequences (X2 [11, N = 480] = 5.63, p = .90), drug use (X2 [10, N = 480] = 10.47, p = .40), and drug consequences (X2 [10, N = 480] = 11.78, p = .30) did not differ by graduation status. The intercept was then recoded to T5 for each model to test for mean differences after expected graduation. Mean levels of the latent intercept for alcohol consequences (X2 [1, N = 480] = 1.63, p = .20) and drug consequences (X2 [1, N = 480] = 1.38, p = .24) were not statistically different, but alcohol use (X2 [1, N = 480] = 4.87, p = .03) and drug use were (X2 [1, N = 480] = 6.55, p = .01). Surprisingly, both alcohol and drug use were higher in those who graduated (Alcohol M = 9.94; Drug M = 0.08) relative to those who did not (Alcohol M = 8.12; Drug M = 0.04).

Summary for Unconditional Models

Alcohol and drug use and consequences were assessed across the fourth year of college, and for 2/3 of the sample, the transition out of college. Results revealed declines in alcohol use, drug use, and drug consequences over time. Alcohol consequences remained relatively stable. These trajectories did not differ by graduation status. Follow-up analyses (four months after expected graduation) indicated that college graduates engaged in higher alcohol and drug use relative to non-graduating peers, but groups did not differ in alcohol or drug consequences.

Conditional LGCM’s – SV histories and substance use behaviors

Next, we hypothesized that (2a) those with SV histories would report higher levels of alcohol and other drug use and consequences over time and (2b) those with pre-college + college SV histories would be at the highest substance use risk. To test this, gender, pre-college SV, college SV, the interaction between pre-college SV and college SV, and prior substance use or consequences were added to the final unconditional LCGM’s to assess the effect of SV histories above and beyond gender and prior substance use patterns. Table 2 contains the unstandardized and standardized beta coefficients for all four models.

Table 2.

Latent growth curve model standardized and unstandardized coefficients across models

Predictor Alcohol Use Alcohol Consequences Other Drug Use Other Drug Consequences
Intercept Slope Intercept Slope Intercept Slope Intercept Slope
B β B β B β B β B β B β B β B β
Pre-college SV 0.27 .02 --- --- 0.69 .07 −3.40 −.19 0.02 .06 0.01 .03 0.14 .05 −0.03 −.02
College SV 0.58 .06 --- --- 0.66 .10 −0.53 −.04 0.02 .09 −0.02 −.18 0.12 .07 −0.12 −.10
Pre-college + College SV 1.36 .06 --- --- 1.49 .12 −0.24 −.01 0.02 .05 −0.01 −.02 0.36 .10 −0.37 −.15
Gender 0.80 .05 --- --- 0.41 .04 −1.04 −.06 0.01 .04 0.01 .04 0.01 .01 0.15 .08
Prior substance use 0.64 .80 --- --- 0.55 .76 −0.10 −.27 — --- --- ---
Prior substance consequences — --- --- --- 0.60 .70 −0.14 −.09 — --- --- --- 0.81 .75 −0.44 −.56
Open in a new tab

Note. SV = sexual victimization. Intercept results reflect differences found at T1 (September, start of senior year). Significant results (p < .05) are presented in bold. Marginally significant results (p = .051–.089) are in italics. The alcohol use model was a fixed effects model and so no slope effects were assessed.

For alcohol use (x2 [29] = 49.28, p = .01, RMSEA = .04 [90% C.I. = .02 – .06]; CFI/TLI= .99/.98; SRMR = .04), college SV was a marginally significant predictor (p = .07). Pre-college SV (p = .60) and gender (p = .12) were not associated with T1 alcohol use, but as expected, prior alcohol use was. The first order effects were qualified by a significant interaction (p = .04). The interaction (see Figure 3, Panel A) suggested that those with pre-college SV + college SV during 1+ college years had the highest levels of alcohol use across time points, followed by pre-college SV + college SV during 1 college year. Alcohol use for the remaining groups were clustered tightly together. In order to determine whether alcohol use remained significantly higher for those with SV histories over the transitional year, we examined models where the intercept was recoded to T4 (April - near expected graduation) and T5 (September - after expected graduation). In both models, the interaction remained significant (ps < .05), suggesting that those with pre-college + college SV histories continued to drink at higher levels during this time. College SV continued to marginally predict higher alcohol use at T4 and T5 (p = .07).

Figure 3.

Figure 3

Open in a new tab

Substance trajectories by sexual assault histories over time Note. All models controlled for prior substance use or consequences.

In the alcohol consequences model (x2 [16] = 25.63, p = .06, RMSEA = .04 [90% C.I.= .00 – .06]; CFI/TLI = .99/.98; SRMR = .03), college SV and prior alcohol consequences predicted T1 alcohol consequences, although neither were associated with growth over time (ps = .42–.57). Gender and pre-college SV were not associated with T1 alcohol consequences (ps= .12–.32). However, pre-college SV predicted change in consequences over time (quadratic factor) in that those with pre-college SV showed steeper declines in consequences across the time points. The first order effects were qualified by a significant interaction (see Figure 3, Panel B). The interaction suggested a similar rank ordering of SV histories that was observed for alcohol use, but with a different longitudinal pattern. Although the combination of (a) pre-college SV + college SV during 1+ college years and (b) pre-college SV + college SV showedthe highest levels of consequences across the year, consequences decreased faster over time for these groups. In order to determine whether alcohol consequences remained significantly higher for those with SV histories, we recoded the intercept to T4 and T5. The pre-college + college SV interaction became marginally significant at T4 (p = .07) and non-significant at T5 (p = .27). In addition, pre-college SV (p < .001) and college SV (p = .04) predicted higher consequences at T4, but both became non-significant at T5 (ps = .35–.40).

For other drug use (x2 [25] = 45.97, p = .01, RMSEA = .04 [90% C.I. = .02 – .06]; CFI/TLI = .97/.96; SRMR = .04), pre-college SV, college SV, and prior drug use predicted T1 levels of drug use. College SV and baseline drug use predicted linear change in drug use. Gender was not associated with T1 drug use (p = .27) or change in drug use over time (p = .61). For this model, the Pre-college × College SV interaction was not significant (p = .20). We then recoded the intercepts to T4 and T5 and found that pre-college SV (ps < .01) continued to predict drug use over the year, while college SV did not (ps = .27–.92).

Within the other drug consequences model (x2 [25] = 37.24, p = .05, RMSEA = .03 [90% C.I. = .00 – .05]; CFI/TLI = .97/.96; SRMR = .05), prior drug consequences were associated with T1 drug consequences as well as changes in drug consequences over time. Neither gender, pre-college SV, or college SV were associated with T1 drug consequences (ps = .09-.95) or changes in consequences over time (ps = .18–.80). The first order effects were qualified by a significant interaction (p = .03), but marginally significant changes in drug consequences over time (p = .07). The model implied trajectories (see Figure 3, Panel C) of drug consequences over time by SV histories suggested the same rank ordering as alcohol consequences. We recoded the intercepts to T4 and T5 and found that the interaction did not predict drug consequences at T4 or T5 (ps = .12–.94). Moreover, pre-college SV (p = .02) predicted drug consequences at T4, but not at T5 (p= .11).

Finally, to test whether the effects differed by graduation status, graduation status as well as two-way and three-way interactions were entered into each model. For all models, neither graduation status nor the interactions with graduation status predicted the latent intercept or slope.

Summary for Conditional Models

In these models, alcohol and drug trajectories were conditioned on SV histories, controlling for gender and prior substance use behaviors. Findings revealed that, at the start of senior year, those who experienced SV before and during college engaged in heavier drinking. Elevations in alcohol consequences were also found at the start of senior year for college SV and pre-college + college SV histories. In both alcohol use and consequences models, prior use and consequences predicted elevations in those same behaviors at the start of senior year. Pre-college SV was associated with greater declines in alcohol consequences over time. For drug use, pre-college SV, college SV and greater prior drug use had higher drug use at the start of senior year, although those with college SV and prior drug use showed steeper declines over the year. Finally, greater drug consequences at the start of senior year was associated with pre-college + collegeSV histories and those with greater prior drug consequences.

Discussion

To our knowledge, this study is the first to examine the role that SV may play in substance trajectories as students prepare to leave the college environment. An important innovation of this study was our examination of the unique effects of the timing of SV (i.e., before entering college, during college, or both) in substance trajectories as well as our inclusion of not only alcohol outcomes, but other drug outcomes as well. Findings suggest that the substance use risk associated with SV decreases over time, yet pre-college + college SV histories confer vulnerability for heavier alcohol-related outcomes during the transitional year. Notably, these patterns were observed even after controlling for the strong effects of prior substance use behaviors. We elaborate on these findings and their implications below.

Sexual Victimization and Revictimization

Consistent with the literature (Aosved et al., 2011; Gidycz et al., 2007; Humphrey & White, 2000), 36% of our sample entered into college with a SV history. Strikingly, of those who entered with pre-college SV, more than half were revictimized during their time in college. In this study, we were also able to shed some light on the extent to which men experience SV before and during college (6% before college, 7% during college). This is important, as prior research has tended to focus on women as victims and men as perpetrators, which has greatly limited our understanding of college men’s victimization. Taken together, the rates of SV reported in this study are consistent with prior work in this area (e.g., Krebs et al., 2009; Humphrey & White, 2000; Mouilso et al., 2012), and highlight the continued need for researchto help reduce risk across the college years, especially for individuals who experienced SV before coming to college.

Substance Trajectories – Overall Sample

As hypothesized, rates of alcohol and drug use across the sample were higher at the start of senior year and gradually declined over the one-year assessment. Drug consequences also declined over time. For nearly 2/3 of our sample, this coincided with the transition out of college. Rates of graduation in this sample (66%) were higher than at least some prior work, which found that just under half of college students graduated by the end of their fourth post-matriculation year (Patrick, Schulenberg, & O’Malley, 2016). More commonly, it may take students at public universities five to six years to graduate with their four-year degree (National Center on Education Statistics, 2017). The higher rates of graduation in this sample are likely due to the targeted efforts of this university to facilitate a “Finish in 4” track for undergraduates. Diminution in substance use during this time, often referred to as “maturing out,” is thought to be normative as young adults move increasingly toward independent adulthood (i.e., graduate college, get married, start a family; Sher & Gotham, 1999; O’Malley, 2004; White et al., 2005). Though we observed a general pattern of gradual decreases regardless of graduation status, at the final time point, college graduates reported higher levels of alcohol and drug use relative to non-graduating peers. This may be a function of when this last time point fell. The final time point was four months after graduation (September), and thus corresponds to the summer following college departure. This may be a time when many young adults are still celebrating the end of college, and have not yet moved on to new roles and responsibilities (i.e., full time job, parenting).

In contrast, alcohol consequences did not show a pattern of “maturing out” by the end of the year. Instead, alcohol consequences remained relatively stable over time. This stability appears to be due to the persistence of less severe consequences (i.e., saying something embarrassing, hangover), ones that occur even at lower levels of consumption. A growing literature suggests that the process of “maturing out” of substance use and consequences in young adulthood may be more complex and less abrupt than once was thought. Instead, maturing out may unfold gradually over time in response to changing circumstances (Egerton & Read, in press; Littlefield et al., 2009). Our data are consistent with this, and suggest that a more fine-grained examination of the many factors that influence changes during this time is warranted.

Substance Trajectories as a Function of Sexual Victimization Histories

In line with prior work identifying SV and alcohol disparities early in college (e.g., Bedard-Gilligan et al., 2011; Larimer et al., 1999; Turchik, 2012), and in partial support of our second hypothesis, we found that students with SV histories (college SV, pre-college + college SV) began senior year with greater alcohol use and/or consequences relative to non-victimized peers. However, one year later, many of these differences were no longer statistically significant. To our surprise, the trajectories of those with SV histories revealed a pattern of “maturing out” of alcohol use over time that was similar to those without SV, although the quantity of use behaviors was still comparatively higher at each time point. For the victimized groups, beginning senior year with higher problematic use likely provided more opportunity for diminution over time (particularly for those with pre-college SV histories). Alternatively, research from the early college years has found that women with SV histories tend to be heavier drinkers, even before victimization occurred (e.g., Mouilso et al., 2012), which might also explain the quantity differences, despite the overall pattern of decline over time. The early college findings of Mouilso et al. (2012), considered along with the present findings pertaining to the later college years, could suggest that there is something unique about the college environment that facilitates problematic alcohol use for students at greater SV risk (including those with SV histories). Within the college literature, considerable research has shown that peer influences and social environments play a large role in college drinking (e.g., Borsari & Carey, 2001; Neighbors, Lee, Lewis, Fossos, & Larimer, 2007; Weitzman, Nelson, & Wechsler, 2003). While in the college environment, it may be that students with SV histories (and those at greater SV risk) self-select into peer groups or social contexts where heavier drinking patterns are accepted and even promoted. Heavy drinking in these social contexts can also increase SV risk by placing young adults in close proximity to potential perpetrators. Continued research in this area can help better determine how and why the college context contributes to SV risk. Fortunately though, at this stage of college, alcohol risk appears to gradually decline over the transitional year for those with SV histories. These declines may help prevent progression toward longer-term substance use problems in later adulthood.

We found a similar pattern for drug use and consequences, such that students with SV histories (pre-college SV, college SV, pre-college + college SV) reported higher drug use and/or consequences at the start of senior year. For drug use, individuals with pre-college SV continued to use drugs at a higher frequency one year later, although they did not experience greater drug consequences at the end of the year. When examining change over time, college SV was associated with steeper declines in drug use. Drug use as a consequence of SV has received relatively little research attention in the literature, however some studies have linked SV during early development to an earlier onset of drug use and more hazardous drug use patterns (Nelson et al., 2006; Wilsnack, Vogeltanz, Klassen, & Harris, 1997). Yet, at this developmental juncture, risk for problematic drug use for those with SV histories appears to be in decline.

Sexual Victimization Before and During College

In our final hypothesis, we expected that those with combined pre-college + college SV histories would be at greatest substance use risk over time. This hypothesis was partially supported for alcohol, but not other drugs. First, to our surprise, those with the repeated experience of SV – occurring both before and during college – showed a pattern of normative decline in alcohol outcomes over time that was similar to other groups. However, though declining, levels of alcohol use nonetheless were higher for those with this pattern of pre- and during college victimization throughout the transitional year. This group also showed significantly greater alcohol consequences up to the point of expected graduation, but not after. This is consistent with prior work showing that revictimization is associated with worse outcomes compared to those with and without SV histories, including greater alcohol use (e.g., Messman-Moore et al., 2000; Ullman & Najdowski, 2009). Yet, these are the first data to characterize this risk as students prepare to transition out of college. Our findings point to continued alcohol-associated risk for those with pre-college + college SV, even in the context of normative and gradual reductions in alcohol use behaviors over an important transitional year.

Limitations and Future Directions

This study had limitations. First among these was that SV histories were assessed via retrospective, self-report of experiences that occurred prior to and during college. Self-report has the potential to be influenced by memory biases or decay due to the passage of time. Further, the number of men with SV histories in our sample was relatively small, and thus we were not able to model trajectories conditioned on gender and SV history. Also, in this study, we focused only on alcohol as a consequence of SV. Given that studies have found prospective support for alcohol as a risk factor as well as a consequence of SV (Testa & Livingston, 2009), the exploration of both directions will be an interesting continued direction for future research. Our measurement of drugs other than alcohol could also represent a limitation. That is, to reduce participant burden and to simplify our prospective models, we did not examine use of or consequences from specific drug types (e.g., marijuana, cocaine, etc.). As such, we cannot draw conclusions about associations between SV and specific types of drugs. This too may be an interesting next step for future investigations. Further, our combined scoring of other drug use likely underestimated typical use during this transition. Also, in this study, we examined substance differences according to SV status. Some research suggests that it is psychological sequelae related to the trauma (e.g., psychological distress, PTSD), rather than trauma exposure per se that results in greater substance use risk (Epstein, Saunders, Kilpatrick, & Resnick, 1998; Read et al., 2012; Stewart & Conrod, 2003). SV has certainly been found to be associated with high levels of psychological distress and substance abuse/dependence (for review, see Dworkin, Menon, Bystrynski, & Allen, 2017). In future studies, the examination of SV-related distress may build on the findings presented here, and could help to explain the elevated substance use risk among this group (especially for pre-college + college SV). Lastly, we were only able to follow participants for four months after expected graduation. Future studies can extend our work with examinations of SV-related substance risk that reach further out into adulthood.

Clinical Implications and Conclusions

In this study, we examined the influence of SV histories on substance trajectories as young adults prepare to leave the college environment. We found that SV is a risk factor for substance involvement during this period, but that this risk decreases with time, as the college years draw to an end. Despite these normative reductions, those with pre-college + college SV histories continue to report higher levels of alcohol use behaviors across the transitional year. This has the potential to confer vulnerability for a range of negative, alcohol-related outcomes for these individuals as they progress into adulthood. Thus, the time leading up to college departure may represent a critical period for targeted intervention. Such interventions, developed to reduce hazardous drinking during times of transition and offered prior to college departure for those at greatest risk, could serve to ameliorate or prevent negative outcomes, ultimately ensuring a stronger transition into independent adulthood for this vulnerable group.

Acknowledgments

Data collection and manuscript preparation was supported by the National Institute on Drug Abuse, Grant R01 DA018993 awarded to J. P. Read.

Footnotes

1

We computed a composite drug use variable in three ways (mean, sum [total frequency of past month drug use], and maximum [highest past month frequency]). Correlations among these three composite scores were all above .93. We also examined whether using the three different composites produced different patterns of change over time. All three ways of computing this variable suggested a modest decrease in substance use over time from T1 to T5. Comparison of the T1 and T5 values based on a mean and a sum were both statistically significant (p < .05) and the difference based on max was marginal (p < .06). However, effect sizes suggested small degree of decline for each variable (Cohen’s ds = .07 to .13). Given that the method of computing a substance use composite variable did not appreciably affect the pattern of our findings, we opted for the mean in our substantive analyses, to be consistent with the alcohol use variable.

Contributor Information

Jessica A. Blayney, State University of New York - University at Buffalo

Matthew Scalco, State University of New York - University at Buffalo.

Sharon Radomski, State University of New York - University at Buffalo.

Craig Colder, State University of New York - University at Buffalo Jennifer.

Jennifer P. Read, State University of New York - University at Buffalo

References

  1. Abbey A (2002). Alcohol-related sexual assault: A common problem among college students. Journal of Studies on Alcohol, Supplement 14, 118–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Aiken LS, & West SG (1991). Multiple regression: Testing and interpreting interactions. Thousand Oaks, CA US: Sage Publications, Inc. [Google Scholar]
  3. American College Health Association. (2014). National College Health Assessment II: Undergraduate Students Reference Group Executive Summary Spring 2014. Hanover, MD US: American College Health Association. [Google Scholar]
  4. Aosved AC, Long PJ, & Voller EK (2011). Sexual re-victimization and adjustment in college men. Psychology of Men & Masculinity, 12, 285–296. doi: 10.1037/a0020828 [DOI] [Google Scholar]
  5. Bedard-Gilligan M, Kaysen D, Desai S, & Lee CM (2011). Alcohol-involved assault: Associations with posttrauma alcohol use, consequences, and expectancies. Addictive Behaviors, 36, 1076–1082. doi: 10.1016/j.addbeh.2011.07.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Blayney JA, Colder C, & Read JP (2016). The role of alcohol in college sexual victimization and post-assault adaptation. Psychological Trauma: Theory, Research, Practice, & Policy, 8, 421–430. doi: 10.1037/tra0000100 [DOI] [PubMed] [Google Scholar]
  7. Borsari B, & Carey KB (2001). Peer influences on college drinking: A review of the research. Journal of Substance Abuse, 13, 391–424. doi: 10.1016/S0899-3289(01)00098-0 [DOI] [PubMed] [Google Scholar]
  8. Collins RL, Parks GA, & Marlatt GA (1985). Social determinants of alcohol consumption: The effects of social interaction and model status on the self-administration of alcohol. Journal of Consulting and Clinical Psychology, 53, 189–200. doi: 10.1037/0022-006X.53.2.189 [DOI] [PubMed] [Google Scholar]
  9. Dworkin ER, Menon SV, Bystrynski J, & Allen NE (2017). Sexual assault victimization and psychopathology: A review and meta-analysis. Clinical Psychology Review, 5665–81. doi: 10.1016/j.cpr.2017.06.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Egerton G, & Read JP (In press). Relationship role transitions and problem alcohol use in emerging adulthood. Emerging Adulthood. [Google Scholar]
  11. Epstein JN, Saunders BE, Kilpatrick DG, & Resnick HS (1998). PTSD as a mediator between childhood rape and alcohol use in adult women. Child Abuse & Neglect, 22, 223–234. doi: 10.1016/S0145-2134(97)00133-6 [DOI] [PubMed] [Google Scholar]
  12. Gidycz CA, Loh C, Lobo T, Rich C, Lynn SJ, & Pashdag J (2007). Reciprocal relationships among alcohol use, risk perception, and sexual victimization: A prospective analysis. Journal of American College Health, 56, 5–14. doi: 10.3200/JACH.56.1.5-14 [DOI] [PubMed] [Google Scholar]
  13. Gotham HJ, Sher KJ, & Wood PK (1997). Predicting stability and change in frequency of intoxication from the college years to beyond: Individual-difference and role transition variables. Journal of Abnormal Psychology, 106, 619–629. doi: 10.1037/0021-843X.106.4.619 [DOI] [PubMed] [Google Scholar]
  14. Griffin MJ, & Read JP (2012). Prospective effects of method of coercion in sexual victimization across the first college year. Journal of Interpersonal Violence, 27, 2503–2524. doi: 10.1177/0886260511433518 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Griffin MJ, Wardell JD, & Read JP (2013). Recent sexual victimization and drinking behavior in newly matriculated college students: A latent growth analysis. Psychology of Addictive Behaviors, 27, 966–973. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Hu L, & Bentler PM (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. doi: 10.1080/10705519909540118 [DOI] [Google Scholar]
  17. Humphrey JA, & White JW (2000). Women’s vulnerability to sexual assault from adolescence to young adulthood. Journal of Adolescent Health, 27, 419–424. doi: 10.1016/S1054-139X(00)00168-3 [DOI] [PubMed] [Google Scholar]
  18. Jackson KM, Sher KJ, Gotham HJ, & Wood PK (2001). Transitioning into and out of large-effect drinking in young adulthood. Journal of Abnormal Psychology, 110, 378–391. [DOI] [PubMed] [Google Scholar]
  19. Johnston LD, O’Malley PM, Bachman JG, Schulenberg JE & Miech RA (2015). Monitoring the Future national survey results on drug use, 1975–2014: Volume 2, College students and adults ages 19–55. Ann Arbor, MI US: Institute for Social Research, The University of Michigan. [Google Scholar]
  20. Kahler CW, Strong DR, & Read JP (2005). Toward efficient and comprehensive measurement of the alcohol problems continuum in college students: The Brief Young Adult Alcohol Consequences Questionnaire. Alcoholism: Clinical and Experimental Research, 29, 1180–1189. [DOI] [PubMed] [Google Scholar]
  21. Kaysen D, Neighbors C, Martell J, Fossos N, & Larimer ME (2006). Incapacitated rape and alcohol use: A prospective analysis. Addictive Behaviors, 31, 1820–1832. doi: 10.1016/j.addbeh.2005.12.025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Krebs CP, Lindquist CH, Warner TD, Fisher BS, & Martin SL (2009). College women’s experiences with physically forced, alcohol- or other drug-enabled, and drug-facilitated sexual assault before and since entering college. Journal of American College Health, 57, 639–647. doi: 10.3200/JACH.57.6.639-649 [DOI] [PubMed] [Google Scholar]
  23. Larimer ME, Lydum AR, Anderson BK, & Turner AP (1999). Male and female recipients of unwanted sexual contact in a college student sample: Prevalence rates, alcohol use, and depression symptoms. Sex Roles, 40, 295–308. doi: 10.1023/A:1018807223378 [DOI] [Google Scholar]
  24. Littlefield AK, Sher KJ, & Wood PK (2009). Is “maturing out of alcohol involvement related to personality change? Journal of Abnormal Psychology, 118, 360–374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Marsh HW, Hau K, & Wen Z (2004). In search of golden rules: Comment on hypothesis-testing approaches to setting cutoff values for fit indexes and dangers in overgeneralizing Hu and Bentler’s (1999) findings. Structural Equation Modeling, 11, 320–341. doi: 10.1207/s15328007sem1103_2 [DOI] [Google Scholar]
  26. Messman-Moore TL, Long PJ, & Siegfried NJ (2000). The re-victimization of child sexual abuse survivors: An examination of the adjustment of college women with child sexual abuse, adult sexual assault, and adult physical abuse. Child Maltreatment, 5, 18–27. doi: 10.1177/1077559500005001003 [DOI] [PubMed] [Google Scholar]
  27. Mouilso ER, Fischer S, & Calhoun KS (2012). A prospective study of sexual assault and alcohol use among first-year college women. Violence & Victims, 27, 78–94. doi: 10.1891/0886-6708.27.1.78 [DOI] [PubMed] [Google Scholar]
  28. Mutheén LK, & Mutheén BO (1998–2012). Mplus User’s Guide (7th Ed.). Los Angeles, CA US: Mutheén & Mutheén. [Google Scholar]
  29. Neighbors C, Lee CM, Lewis MA, Fossos N, & Larimer ME (2007). Are social norms the best predictor of outcomes among heavy-drinking college students?. Journal of Studies on Alcohol And Drugs, 68, 556–565. doi: 10.15288/jsad.2007.68.556 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Nelson EC, Heath AC, Lynskey MT, Bucholz KK, Madden PA, Statham DJ, & Martin NG (2006). Childhood sexual abuse and risks for licit and illicit drug-related outcomes: A twin study. Psychological Medicine, 36, 1473–1483. [DOI] [PubMed] [Google Scholar]
  31. O’Malley PM (2004). Maturing out of problematic alcohol use. Alcohol Research & Health, 28, 202–204. [Google Scholar]
  32. Patrick ME, Schulenberg JE, & O’Malley PM (2016). High school substance use as a predictor of college attendance, completion, and dropout: A national multicohort longitudinal study. Youth & Society, 48, 425–447. doi: 10.1177/0044118X13508961 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Parks KA, Romosz AM, Bradizza CM, & Hsieh Y (2008). A dangerous transition: Women’s drinking and related victimization from high school to the first year at college. Journal of Studies on Alcohol and Drugs, 69, 65–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Read JP, Colder CR, Merrill JE, Ouimette P, White J, & Swartout A (2012). Trauma and posttraumatic stress symptoms predict alcohol and other drug consequence trajectories in the first year of college. Journal of Consulting and Clinical Psychology, 80, 426–439. doi: 10.1037/a0028210 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Read JP, Kahler CW, Strong DR, & Colder CR (2006). Development and preliminary validation of the Young Adult Alcohol Consequences Questionnaire. Journal of Studies on Alcohol, 67, 169–177. [DOI] [PubMed] [Google Scholar]
  36. Satorra A, & Bentler PM (2001). A scaled difference chi-square test statistic for moment structure analysis. Psychometrika, 66, 507–514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Snipes DJ, Green BA, Benotsch EG, & Perrin PB (2014). The non-medical use of prescription drugs and lifetime experiences of sexual victimization among college men. Journal of Interpersonal Violence, 29, 2482–2496. doi: 10.1177/0886260513520229 [DOI] [PubMed] [Google Scholar]
  38. Sher KJ, & Gotham HJ (1999). Pathological alcohol involvement: A developmental disorder of young adulthood. Development and Psychopathology, 11, 933–956. doi: 10.1017/S0954579499002394 [DOI] [PubMed] [Google Scholar]
  39. Stewart SH, & Conrod PJ (2003). Psychosocial models of functional associations between posttraumatic stress disorder and substance use disorder In Ouimette P, Brown PJ, Ouimette P, Brown PJ (Eds.), Trauma and substance abuse: Causes, consequences, and treatment of comorbid disorders (pp. 29–55). Washington, DC US: American Psychological Association. doi: 10.1037/10460-002 [DOI] [Google Scholar]
  40. Tabachnick BG, & Fidell LS (2013). Using Multivariate Statistics (6th ed.). Boston, MA US: Allyn & Bacon. [Google Scholar]
  41. Tewksbury R, & Mustaine EE (2001). Lifestyle factors associated with the sexual assault of men: A routine activity theory analysis. Journal of Men’s Studies, 9, 153–182. doi: 10.3149/jms.0902.153 [DOI] [Google Scholar]
  42. Testa M, Hoffman JH, & Livingston JA (2010). Alcohol and sexual risk behaviors as mediators of the sexual victimization–revictimization relationship. Journal of Consulting and Clinical Psychology, 78, 249–259. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Testa M, & Livingston JA (2009). Alcohol consumption and women’s vulnerability to sexual victimization: Can reducing women’s drinking prevent rape?. Substance Use & Misuse, 44, 1349–1376. doi: 10.1080/10826080902961468 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Testa M, VanZile-Tamsen C, Livingston JA, & Koss MP (2004). Assessing women’s experiences of sexual aggression using the sexual experiences survey: Evidence for validity and implications for research. Psychology of Women Quarterly, 28, 256–265. doi: 10.1111/j.1471-6402.2004.00143.x [DOI] [Google Scholar]
  45. Turchik JA (2012). Sexual victimization among male college students: Assault severity, sexual functioning, and health risk behaviors. Psychology of Men & Masculinity, 13, 243–255. doi: 10.1037/a0024605 [DOI] [Google Scholar]
  46. Ullman SE, & Najdowski CJ (2009). Re-victimization as a moderator of psychosocial risk factors for problem drinking in female sexual assault survivors. Journal of Studies on Alcohol and Drugs, 70, 41–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. White A, & Hingson R (2013). The burden of alcohol use: Excessive alcohol consumption and related consequences among college students. Alcohol Research: Current Reviews, 35, 201–218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. White HR, Labouvie EW, & Papadaratsakis V (2005). Changes in substance use during the transition to adulthood: A comparison of college students and their noncollege age peers. Journal of Drug Issues, 35, 281–306. [Google Scholar]
  49. Weitzman ER, Nelson TF, & Wechsler H (2003). Taking up binge drinking in college: The influences of person, social group, and environment. Journal of Adolescent Health, 32, 26–35. doi: 10.1016/S1054-139X(02)00457-3 [DOI] [PubMed] [Google Scholar]
  50. Wilsnack SC, Vogeltanz ND, Klassen AD, & Harris TR (1997). Childhood sexual abuse and women’s substance abuse:Nnational survey findings. Journal of Studies on Alcohol, 58, 264–271. [DOI] [PubMed] [Google Scholar]

RESOURCES