Traditionally aged college students (18–24 years old) are a segment of the age group with the highest burden of new human immunodeficiency virus (HIV) diagnoses in the United States.1 Despite a low reported prevalence of HIV among national reference groups of college students (0.2%),2 students are embedded in an environment that is conducive to engaging in sexual risk factors (SRFs) conducive to sexually transmitted disease/HIV transmission including sex with casual partners,3 having a higher number of partners in the past year, and failing to always use preventive strategies during intercourse.2,4 It is important to note that despite the Centers for Disease Control and Prevention recommendations for HIV testing, only 28.9% of college students have ever been tested for HIV,2 and HIV testing intention was higher for those who have been previously tested and those who have higher perceived susceptibility.5,6 Disengagement in HIV testing contributes to an estimated 50% of adolescents and young adults living with HIV experiencing a 2.4-year delay from time of infection to time of diagnosis.7 Identifying characteristics that can increase at-risk college student intent to get tested for HIV can help to increase testing utilization, which can simultaneously increase reach of risk reduction education to people at risk while increasing linkage to care among persons living with HIV (PLWH). Accordingly, the purpose of this study was to examine if the indicators of sexual risk behaviors were related to perceived susceptibility to HIV infection and HIV testing intention among sexually active college students who have not been tested.
Sample and Procedure
This study used existing data collected using a Web-based survey in January 2016 from a sample of college students attending a public university in Florida where HIV testing services are widespread both on the campus and in the surrounding community (see Supplemental File 1, http://links.lww.com/OLQ/A373). Upon receiving ethics approval, the university provided a simple random sample of 10,000 students older than 18 years who attend class on-campus. Before taking the survey, participants read and agreed to a brief informed consent. See James and Ryan8 for detailed methods descriptions. We used a subset of the sample who were sexually active in the past 12 months and who had never been tested for HIV (n = 1173; 50.1% of the total sample of 2343 students who responded to the study). Analytic sample characteristics are presented in Table 1.
Sexual Risk Factors
Three SRFs were measured: (1) if the participant ever had sex without knowing their partner's HIV status; (2) having 3 or greater oral, anal, or vaginal intercourse partners in the past 12 months; and (3) having sexual relationships in the past 12 months mostly comprised of commercial/transactional (i.e., having sex for drugs, money, or other items), casual (i.e., someone the respondent knew and had sex with occasionally), or 1-time partners, opposed to regular partners (i.e., an ongoing relationship with a partner, spouse, boyfriend, or girlfriend).
Human immunodeficiency virus–related knowledge was measured using the 18-item version of the HIV Knowledge Questionnaire9 with 5 additional items from the original questionnaire.10 A percent correct knowledge score was calculated for each respondent by dividing the number of correct answers by the total number of items. Reliability of the knowledge scale in the total sample was acceptable (ρ(KR-20) = 0.79).8
Perceived Susceptibility and Testing Intention
Perceived susceptibility was measured using a single 5-point Likert-type item: “What are the chances that you might catch HIV?” with the following corresponding response options: (1) none, (2) small, (3) moderate, and (4) large. Similarly, participants self-reported the likelihood of getting tested in the next 12 months (i.e., “How likely is it that you will be tested for HIV within the next 12 months?” with the following response options:  very unlikely,  unlikely,  unsure, [4} likely,  very likely).
Demographic items assessed gender, self-reported race and ethnicity, and self-identified sexual orientation, which were dichotomized and controlled for in the main analysis.
A path analysis model was estimated using Mplus 8.0.11 Sexual risk factors were modeled to predict HIV testing intention directly and indirectly through perceived susceptibility, and HIV-related knowledge was hypothesized to predict perceived susceptibility. Covariances between exogenous variables that were significantly correlated were freely estimated, whereas nonsignificant covariances were fixed to 0. We used a robust maximum likelihood (MLR) estimator because of the nonnormality of outcome variables. Estimated mediated effects were obtained by the products of path coefficients involved in the mediational pathways, and results from the MLR estimator were compared with a maximum likelihood (ML) estimator with 5000 bootstrapped samples and reported with 95% confidence intervals (CI)12 (little difference was found between the MLR and ML bootstrapped standard errors). Missing data were adjusted using full information maximum likelihood procedures implemented in Mplus. Overall model fit was assessed using multiple fit indices (ie, χ2, root mean square error approximation, comparative fit index, Tucker Lewis index, and standardized room mean square residual).13,14
The hypothesized model (Fig. 1) fit the data very well (MLR:χ2(df = 15) = 16.932, P = 0.323; root mean square error approximation = 0.010 [90% CI, 0.000–0.030]; comparative fit index = 0.994; Tucker Lewis index = 0.994; standardized room mean square residual = 0.018). All of the SRFs were related to higher perceived susceptibility and to higher intent to test in the next 12 months (see Table 2). Mediation analyses suggested perceived susceptibility partially mediated the relation between SRFs and HIV testing intent. Specifically, sex without knowing a partner's HIV status was related to higher perceived susceptibility, which was associated with higher test intention (estimated mediated effect, 0.099; SE, 0.020; 95% CI, 0.066–0.133; P < 0.001). Having more sexual partners (estimated mediated effect, 0.0768; SE, 0.021; 95% CI, 0.043–0.113; P < 0.001) and nonregular types of partner (estimated mediated effect, 0.085; SE, 0.021; 95% CI, 0.051–0.120; P < 0.001) were also related to test intention through perceived susceptibility. Knowledge was not a significant predictor of perceived susceptibility.
Half of the individuals who completed the survey were sexually active and had never been tested for HIV. Respondents in our analytic sample reported varying levels of SRFs, with most students having less than 3 (77%) sexual partners and maintaining “regular” sexual relationships (e.g., with a spouse, boyfriend, girlfriend, etc.; 74%) during the 12 months preceding the survey. Almost half of respondents (42%) indicated that they had previously had sex without knowing their partner's HIV status. Our data supported the hypothesis that participants with higher risk sexual behavior have increased perceived susceptibility and increased intent to get tested for HIV in the next 12 months. Health behavior theory, such as the Health Belief Model, suggests that perceived susceptibility depends on the individual's level of knowledge related to the condition15; interestingly, this hypothesized association was not supported. It might be due to high level of HIV knowledge in this sample (M = 76%), as indicated in prior research.16,17
These results have implications for clinical practice and health education messaging regarding HIV testing. In health education and promotion, mass media campaigns have been used worldwide for HIV prevention since the 1980s. The majority of these campaigns have focused on increasing condom use and providing skills-based training for HIV prevention and have been effective for increasing knowledge and behavior related to preventing HIV, especially for the populations with low knowledge.18 Thus, we might expect to see smaller effects in a knowledgeable population. Future campaigns in populations with high HIV knowledge among the untested population should consider tailoring messages to the SRFs of that population. This message framing is also applicable to clinical practice. When young adult patients present to the clinic, providers should have a discussion regarding HIV-related risk behaviors, including SRFs, when offering opt-out HIV testing; this framing removes the burden of asking for a test from the patient, standardizes care for patients regardless of their level of HIV knowledge, and may increase patient perceived susceptibility and intention to get tested for HIV.
These findings should be interpreted with respect to the limitations of using a cross-sectional survey. Taboo topics, such as HIV, could lead to social desirability bias and underreporting of risk behaviors.19 Further, we operationalized perceived susceptibility and testing intent using one item each. Future analyses should expand the domains of SRFs (e.g., measuring serodiscordant partnerships, sex with virally undetectable PLWH, PrEP use, sex with injection drug users, and use of strategic positioning) and use psychometrically rigorous scales to estimate constructs from health behavior theory, including perceived susceptibility and testing intention. The inclusion of social network constructs, such as knowing a PLWH, could also help better explain HIV testing intention. Due to the small number of respondents in our analytic sample reporting knowing a PLWH (6%), we were unable to include this variable in our model. Due to the cross-sectional nature of our data, we cannot presume causality between any of the variables. However, prior empirical evidence and theory provide a framework for the relations described by our model. A final limitation is that participants were recruited at a single, predominantly white public institution and, thus, the findings may not be generalizable to all college students. Future studies should explore the relation between HIV testing intention and SRFs among students from multiple institutions while accounting for the availability of HIV testing services on that campus.
1. Centers for Disease Control and Prevention. Sexually transmitted disease surveillance 2017. Atlanta, GA: US Department of Health Human Services, 2018.
2. American College Health Association. American College Health Association National College Health Assessment—Spring 2017 Reference Group Data Report. Hanover, MD, 2017.
3. Uecker JE, Regnerus MD. Bare market: Campus sex ratios, romantic relationships, and sexual behavior. Sociol Q 2010; 51:408–435.
4. LaBrie JW, Hummer JF, Ghaidarov TM, et al. Hooking up in the college context: The event-level effects of alcohol use and partner familiarity on hookup behaviors and contentment. J Sex Res 2014; 51:62–73.
5. Caldeira KM, Singer BJ, O'Grady KE, et al. HIV testing in recent college students: Prevalence and correlates. AIDS Educ Prev 2012; 24:363–376.
6. Anwuri GC, Dunn MS, Schulze F. Determinants of voluntary HIV/AIDS counseling and testing among community college students in the United States. Int J MCH AIDS 2017; 6:109–120.
7. Dailey AF, Hoots BE, Hall HI, et al. Vital signs: Human immunodeficiency virus testing and diagnosis delays—United States. MMWR Morb Mortal Wkly Rep 2017; 66:1300–1306.
8. James TG, Ryan SJ. HIV knowledge mediates the relationship between HIV testing history and stigma in college students. J Am Coll Health 2018; 66:561–569.
9. Carey MP, Schroder KE. Development and psychometric evaluation of the brief HIV knowledge questionnaire. AIDS Educ Prev 2002; 14:172–182.
10. Carey MP, Morrison-Beedy D, Johnson BT. The HIV-knowledge questionnaire: Development and evaluation of a reliable, valid, and practical self-administered questionnaire. AIDS Behav 1997; 1:61–74.
11. Muthén LK, Muthén BO. Mplus User's Guide. 6th ed. Los Angeles: Muthén & Muthén, 2011.
12. MacKinnon D. Introduction to Statistical Mediation Analysis. Routledge, 2012.
13. Hu L, Bentler PM. Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Struct Equ Model Multidiscip J 1999; 6:1–55.
14. Schumacker RE, Lomax RG. A Beginner's Guide to Structural Equation Modeling
. 4 edition. New York, NY: Routledge, 2015.
15. Glanz K, Rimer BK, Viswanath K. Health Behavior: Theory, Research, and Practice. San Francisco, CA: John Wiley & Sons, 2015.
16. Ndugwa Kabwama S, Berg-Beckhoff G. The association between HIV/AIDS-related knowledge and perception of risk for infection: A systematic review. Perspect Public Health 2015; 135:299–308.
17. Catania JA, Kegeles SM, Coates TJ. Towards an understanding of risk behavior: An AIDS risk reduction model (ARRM). Health Educ Q 1990; 17:53–72.
18. LaCroix JM, Snyder LB, Huedo-Medina TB, et al. Effectiveness of mass media interventions for HIV prevention, 1986-2013: A meta-analysis. J Acquir Immune Defic Syndr 2014; 66(Suppl 3):S329–S340.
19. Krumpal I. Determinants of social desirability bias in sensitive surveys: A literature review. Qual Quant 2013; 47:2025–2047.