Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Aug 1.
Published in final edited form as: AIDS Educ Prev. 2012 Aug;24(4):363–376. doi: 10.1521/aeap.2012.24.4.363

HIV testing in recent college students: Prevalence and correlates

Kimberly M Caldeira 1, Barbara J Singer 2, Kevin E O’Grady 3, Kathryn B Vincent 1, Amelia M Arria 1,4
PMCID: PMC3408631  NIHMSID: NIHMS373815  PMID: 22827905

Abstract

Prevalence and correlates of HIV testing were examined in a sample of 957 unmarried recent college students in the U.S. Participants were asked about HIV testing, past-six-months sexual activities, lifetime treatment for sexually transmitted infections (STI), past-year health service utilization, and DSM-IV criteria for alcohol and other drug (AOD) dependence during the 2008–2009 academic year. Two in five (41.9%wt) were ever tested for HIV. Holding constant demographics, HIV testing was positively related to AOD dependence, frequency of unprotected sex, number of sex partners, having a physical exam by a medical professional, number of visits to a health provider for physical health problems, and lifetime STI treatment. Women were more likely than men to be tested for HIV despite similar levels of risky sex. Results demonstrate the feasibility of achieving high HIV testing rates in a college population.

Keywords: College students, HIV testing, Risky sex, Sexually transmitted infections

Introduction

Human Immunodeficiency Virus (HIV) remains a major public health challenge worldwide and a persistent risk to young people. One-third of all new HIV infections occur among people under age 30 (Prejean, Song, An, & Hall, 2008), and the number of new infections annually has changed little since the late 1990s (Hall et al., 2008). Of major concern is that, of the 1.1 million individuals living with HIV in the U.S., one in five do not know they are infected (Campsmith, Rhodes, Hall, & Green, 2008). HIV testing can increase the proportion of HIV-infected individuals who are aware of their serostatus, and thus remains an important preventive measure in reducing HIV transmission. To that end, in 2006 the Centers for Disease Control and Prevention (CDC) revised its HIV testing guidelines to recommend universal opt-out testing for all adults visiting a health care setting, while still targeting high-risk individuals for more frequent repeat testing (Branson et al., 2006). The new approach reflects the changing face of HIV infection, which increasingly includes individuals traditionally considered lower-risk, such as youth under age 20 (Ruiz et al., 2000), and has the advantage of promoting testing even in individuals who are unlikely to seek HIV testing due to their low perceived risk, thus enabling them to access treatment at an earlier stage of disease and take steps to avoid transmitting the virus to others. Despite these efforts, the CDC recently estimated that more than half (55%) of U.S. adults ages 18 to 64 had never been tested for HIV, and in young adults ages 18 to 24, 66% remain untested, despite having similar HIV diagnosis rates as other adults (Johnson et al., 2010).

College students’ HIV testing behaviors are important to understand, given their high rates of risky sexual behaviors (American College Health Association, 2010)and epidemiologic evidence of increasing heterosexual transmission of HIV (Centers for Disease Control and Prevention, 2010). Among students surveyed at 57 post-secondary institutions, 24% were tested for HIV at least once in their lifetime (American College Health Association, 2010), yet HIV testing rates in various college samples have ranged between 10% and 58% (Bontempi, Mugno, Bulmer, Danvers, & Vancour, 2009; Buhi, Marhefka, & Hoban, 2010; Crosby, Miller, Staten, & Noland, 2005; Marelich & Clark, 2004; Scholly, Katz, Gascoigne, & Holck, 2005; Thomas et al., 2008; Trieu, Modeste, Marshak, Males, & Bratton, 2010).

The behavioral correlates of HIV testing in college students are similar to the general population, and generally reflect the previously recommended approach of targeting high-risk individuals. Individuals engaging in high-risk sexual behaviors, such as having sex without condoms (Bontempi et al., 2009; Mattson, 2002), using condoms infrequently (Bontempi et al., 2009), and having a greater number of sexual partners (Crosby et al., 2005; Thomas et al., 2008)are more likely to get tested for HIV, and these associations have been replicated in racially diverse samples (Thomas et al., 2008; Trieu et al., 2010). Demographically, HIV testing is higher for women (Crosby et al., 2005), members of racial and ethnic minorities (Buhi et al., 2010; Crosby et al., 2005), older students (Crosby et al., 2005; Thomas et al., 2008), and those with more years of schooling (Thomas et al., 2008).

Three potentially important correlates of HIV testing in college students remain understudied. First, heavy involvement with alcohol and other drugs (AOD)—and consequent risky sexual behaviors—are common in college students, yet their relationship with HIV testing has not been studied. Annually, 11% of full-time college students engage in unprotected sex as a result of drinking (American College Health Association, 2010), and in 2009, 43.5% engaged in binge drinking and 22.7% used illicit drugs (Substance Abuse and Mental Health Services Administration, 2010). Second, under current universal testing guidelines, one would expect health service utilization to predict increased likelihood of HIV testing. One study of students at historically Black colleges and universities (HBCU) supported this association (Thomas et al., 2008), but it remains to be seen whether or not it persists in other college populations. Third, it is expected that college students being treated for sexually transmitted infections (STIs) would have high rates of HIV testing, because the presence of a STI both increases the potential for contracting or transmitting HIV (Eng & Butler, 1997)and signals a pattern of HIV-risk behaviors. Thomas et al. (2008) found that students with an STI in the past year were more than twice as likely to have been tested for HIV. However, in most survey research, HIV and other STI testing are assessed as one variable, despite evidence that college students perceive HIV testing as more serious than STI testing (Barth, Cook, Downs, Switzer, & Fischhoff, 2002).

This study documents the prevalence of HIV testing in a sample of young adults recently in college and examines the demographic and behavioral correlates of HIV testing. Three measures of health risk behaviors (AOD dependence, number of sex partners, frequency of unprotected sex) and three measures of health service utilization (past-year physical exam, number of provider visits for health problems, and history of STI treatment) are hypothesized to be associated with greater likelihood of HIV testing.

Methods

Data were collected as part of the College Life Study, an ongoing longitudinal study of 1,253 young adults originally ascertained as incoming first-time first-year college students, ages 17 to 19, in 2004 at a large public university in an urban region of the mid-Atlantic U.S. (Arria et al.,2008; Vincent et al., in press). After screening 82% of all incoming first-time, first-year students ages 17 to 19 at summer orientation (N=3,401), a representative sample was selected for the longitudinal study. Individuals who used illicit drugs or nonmedically used a prescription drug at least once in high school were purposively over sampled to ensure adequate statistical power for analyses on drug use (Arria et al., 2008). Participants completed a two-hour baseline assessment (n=1,253; 87% response rate), consisting of a personal interview and self-administered questionnaires sometime during their first year of college, and similar follow-up assessments annually thereafter, regardless of continued college attendance. Eighty-one percent (n=1,019) completed the Year 5 assessment (2008–2009), from which the present study emanates. Participants provided written informed consent and received cash payments for each assessment. University IRB approval and a federal Certificate of Confidentiality were obtained.

Sample

Of the 1,019 Year 5 participants, 1,003 completed all self-administered items on HIV testing and sexual activity. Thirty individuals responding “don’t know” to the item on HIV testing were excluded, plus another 16 married individuals whose sexual activities were unlikely to be comparable to the unmarried majority. The final analytic sample was 957 individuals (ages 21 to 24, 54.2% female, 71.3% White; Table 1).

TABLE 1.

Sample characteristics, by lifetime history of HIV testing, among 957 unmarried recent college students

Never been tested for HIV (n=502) Tested for HIV (n=455) Total (N=957)
n % or M (SD) n % or M (SD) n % or M (SD)
PERSONAL CHARACTERISTICS
Gender***
 Female 228 45.4% 291 64.0% 519 54.2%
 Male 274 54.6% 164 36.0% 438 45.8%
Race/ethnicity**
 White 368 73.3% 314 69.0% 682 71.3%
 Black 36 7.2% 55 12.1% 91 9.5%
 Asian 58 11.6% 31 6.8% 89 9.3%
 Other 22 4.4% 30 6.6% 52 5.4%
 Hispanic 18 3.6% 25 5.5% 43 4.5%
Age
 21 21 4.2% 30 6.6% 51 5.3%
 22 376 74.9% 335 73.6% 711 74.3%
 23 103 20.5% 89 19.6% 192 20.1%
 24 2 0.4% 1 0.2% 3 0.3%
Sexual orientation***
 Non-heterosexual 13 2.6% 34 7.5% 47 4.9%
 Heterosexual 489 97.4% 421 92.5% 910 95.1%
College enrollment status
 Not in school 296 59.0% 283 62.2% 579 60.5%
 Part-time undergraduate 21 4.2% 27 5.9% 48 5.0%
 Full-time undergraduate 73 14.5% 70 15.4% 143 14.9%
 Graduate school 112 22.3% 75 16.5% 187 19.5%
Neighborhood incomea 499 74.8 (34.3) 451 71.0 (32.4) 950 73.0 (33.4)
Religiosity
 Not Important 155 31.1% 154 34.1% 309 32.5%
 Slightly Important 116 23.2% 103 22.8% 219 23.1%
 Moderately Important 143 28.7% 115 25.5% 258 27.2%
 Extremely Important 85 17.0% 79 17.5% 164 17.3%
HEALTH RISK BEHAVIORS
Had at least 1 same-sex partner, lifetime*** 13 2.6% 45 9.9% 58 6.1%
 Number of same-sex partners, lifetime 13 4.8 (5.4) 45 7.4 (16.5) 58 6.8 (14.7)
 Age at first same-sex encounter 13 17.9 (3.1) 45 17.8 (3.1) 58 17.8 (3.1)
Had at least 1 same-sex partner, past 6 months*** 8 1.6% 27 5.9% 35 3.7%
 Number of same-sex partners, past 6 months 8 2.4 (2.0) 27 2.9 (3.3) 35 2.8 (3.0)
 Number of unprotected same-sex encounters, past 6 months 8 7.9 (17.1) 27 17.5 (31.7) 35 15.3 (29.1)
Had at least 1 opposite-sex partner, lifetime*** 402 80.1% 427 93.8% 829 86.6%
 Number of opposite-sex partners, lifetime*** 402 5.1 (5.5) 427 9.4 (8.8) 829 7.3 (7.7)
 Age at first opposite-sex encounter*** 402 17.7 (1.9) 426 16.9 (1.7) 828 17.3 (1.8)
Had at least 1 opposite-sex partner, past 6 months*** 350 69.7% 386 84.8% 736 76.9%
 Number of opposite-sex partners, past 6 months*** 350 1.4 (1.0) 386 1.8 (1.4) 736 1.6 (1.2)
 Number of unprotected opposite-sex encounters, past 6 monthsb,*** 341 22.2 (31.0) 377 32.2 (37.1) 718 27.4 (34.7)
Dependent on alcohol and/or marijuana** 32 6.4% 56 12.3% 88 9.2%
HEALTH SERVICE UTILIZATION
Physical exam by medical professional, past year*** 276 55.0% 331 72.7% 607 63.4%
Number of visits for physical health problems, past year** 502 1.7 (3.0) 455 2.7 (7.5) 957 2.2 (5.6)
Treated for STI, lifetime*** 31 6.2% 93 20.6% 124 13.1%
Intent to get HIV test within next year (%Yes/Probably Yes)*** 78 15.5% 245 53.8% 323 33.8%
Open in a new tab

Note. Statistically significant differences between tested and untested groups denoted as

***

p<.001,

**

p<.01,

*

p<.05

a

The mean adjusted gross income for each participant’s home ZIP code during their last year in high school, measured in ten thousands.

b

Excludes 16 outliers with >180 opposite-sex encounters in the past 6 months.

Measures

HIV testing

Self-administered questions asked if participants had ever been tested for HIV, and whether or not they intended to be tested in the next year. Participants were not asked their HIV status. One individual voluntarily disclosed being HIV-positive in response to interview questions about health conditions.

AOD dependence

Dependence on alcohol and/or marijuana during the past year was assessed via standard methods (Substance Abuse and Mental Health Services Administration, 2003), following DSM-IV definitions (American Psychiatric Association, 1994). Responses were later consolidated into one dichotomous variable indicating AOD dependent and non-dependent.

Number of sex partners

Sexual behavior items were adapted from standard surveys (Centers for Disease Control and Prevention, 1997)and expanded to include same-sex activities and capture counts rather than ranges. Self-administered questions asked about both opposite-sex (i.e., “vaginal sex between a male and a female”) and same-sex activities (i.e., “having oral or anal sex with a partner of the same sex as yourself”), regardless of sexual orientation. Separate questions captured the number of same-sex and opposite-sex partners during the participant’s lifetime and past six months.

Frequency of unprotected sex

Individuals with at least one same-sex partner in the past six months were asked how many times they had sex with any same-sex partner in the past six months, and in how many of those same-sex encounters a condom or other barrier was used. Frequency of unprotected same-sex encounters was later computed as the difference between protected and total encounters. Frequency of unprotected opposite-sex encounters was computed in a similar fashion. Sixteen individuals reporting greater than 180 opposite-sex encounters in the past six months were excluded from analyses as outliers.

Health service utilization

Interview questions assessed past-year physical examination by a medical professional (yes vs. no) and past-year number of visits to a medical health professional for physical health problems. Lifetime treatment for any STI (yes vs. no) was self-administered.

Personal characteristics

Sexual orientation, age at first sexual encounter, and religiosity (How important is religion in your life? (Sher & Rutledge, 2007)) were self-reported. Gender was coded as observed. As a proxy for neighborhood income, the mean adjusted gross income for each participant’s home ZIP code (i.e., parents’ home) during the last year the participant was in high school was abstracted from publicly available data (Melissa DATA, 2003). Current school enrollment status was asked by interviewers. Race and ethnicity were self-reported with multiple response options permitted. A five-level race/ethnicity variable was later created: Hispanic (regardless of race), White, Black, Asian, and “Other,” which encompassed 52 non-Hispanic individuals endorsing either multiple (n=26) or no available race/ethnicity options (n=26). Age at interview was computed from date of birth.

Statistical Analyses

Because individuals who used illicit drugs in high school were purposively over sampled (see above), it was necessary to statistically weight our prevalence estimates for HIV testing(denoted by %wt) so that they would be generalizable to the original target population. To that end, sampling weights were computed within each race-sex-drug use cell as the number of individuals in the sampling frame divided by the number of individuals in the sample. All remaining analyses were conducted using unweighted data to avoid artificially inflating statistical power. First, characteristics of individuals who were and were not tested for HIV were compared, and comparisons were evaluated for statistical significance using Chi-square and one-way ANOVA, including the six hypothesized predictors of HIV testing (AOD dependence, unprotected sex, number of sex partners, past-year physical exam, number of provider visits in the past year, and lifetime STI treatment). Intercorrelations of gender and race/ethnicity with each of the six hypothesized predictors were examined. Next, a series of six multiple logistic regression analyses were conducted testing the associations between each of the six hypothesized predictors and HIV testing, holding constant gender, race/ethnicity, neighborhood income, and sexual orientation. (Due to the potential for multicollinearity amongst the six hypothesized predictors, they were never entered simultaneously in a combined model.) For any significant bivariate intercorrelations between the six hypothesized predictors and gender or race/ethnicity (as identified in the preceding step), the corresponding first-order interaction was also entered into the logistic regression model. Small cell sizes necessitated creating a combined Hispanic/Other category for use in the logistic regression modeling. Due to the small number of individuals engaging in same-sex activities, data on opposite-and same-sex activities were summed to compute overall variables on number of sex partners and unprotected sexual encounters.

Results

Sample Characteristics

In Year 5, one in five participants were still in college, either full-time (14.9%) or part-time (5.0%; Table 1). Another 19.5% were in graduate school. The vast majority were heterosexual (95.1%). Most (76.9%) were sexually active with opposite-sex partners in the past six months. Few had sex with same-sex partners (6.1% lifetime, 3.7% past six months). Fewer men than women had lifetime same-sex experience (3.9% vs. 7.9%, p<.01), but their same-sex involvement in the past six months was similar (3.2% vs. 4.0%; Table 2). Notably, 110 individuals (11.5%) reported never having sex, several of whom had been tested for HIV (n=15, 13.6%; not shown in table).

TABLE 2.

Intercorrelations of gender and race/ethnicity with health risk behavior and service utilization variables

Gender
Race/Ethnicity
Women (n=519) Men (n=438) p White (n=682) Black (n=91) Asian (n=89) Hispanic (n=43) Other (n=52) p
Had at least 1 same-sex partner, past 6 months,a % 4.0% 3.2% .485 2.9% 7.7% 2.2% 9.3% 3.8% .047
Number of same-sex partners, past 6 months, Mean (SD) 1.7 (1.2) 4.4 (4.1) .006 2.8 (3.3) 3.3 (3.4) 1.0 (0.0) 16.5 (16.1) 1.5 (.7) .914
Number of unprotected same-sex encounters, past 6 months, Mean (SD) 19.0 (33.4) 9.8 (21.0) .368 23.3 (36.0) 3.3 (5.3) 1.0 (1.4) 10.5 (16.5) 1.5 (.7) .462
Had at least 1 opposite-sex partner, past 6 months,b % 79.2% 74.2% .068 79.9% 65.9% 60.7% 72.1% 88.5% <.001
Number of opposite-sex partners, past 6 months, Mean (SD) 1.6 (1.1) 1.7 (1.3) .314 1.6 (1.2) 1.8 (1.2) 1.4 (.7) 1.6 (1.0) 1.6 (1.6) .497
Number of unprotected opposite-sex encounters, past 6 months, Mean (SD) 27.4 (33.1) 27.4 (36.8) .998 29.0 (35.6) 17.4 (25.1) 23.5 (35.3) 30.2 (35.5) 25.5 (32.7) .130
Dependent on alcohol and/or marijuana, % 7.3% 11.4% .029 9.2% 6.6% 4.5% 14.0% 17.3% .079
Physical exam by medical professional, past year, % 70.5% 55.0% <.001 61.6% 69.2% 62.9% 69.8% 73.1% .271
Number of visits for physical health problems, past year,c Mean (SD) 2.6 (6.9) 1.6 (3.4) .007 2.0 (3.9) 2.2 (4.9) 1.6 (2.8) 4.6 (19.2) 2.8 (5.0) .034
Treated for STI, life time,d % 17.7% 7.6% <.001 12.0% 15.7% 5.6% 27.9% 23.1% <.001
Open in a new tab

Note. Count variables are restricted to individuals who had that type of partner in the past six months.

a

No pairwise comparisons by race/ethnicity were significant at p<.05.

b

Whites and Others were significantly different from both Blacks and Asians (all p<.05).

c

Whites and Asians were significantly different from Hispanics (both p<.05).

d

Whites were significantly different from Hispanics, and Asians were significantly different from Hispanics and Others (both p<.05).

Prevalence of HIV Testing

Nearly half the sample (n=455, 47.5%) had been tested for HIV at least once in their life. After statistically adjusting for the sampling design, the corresponding weighted estimate was 41.9%wt. HIV testing was more prevalent in women (49.8%wt) than men (32.9%wt), and non-heterosexuals (77.7%wt) than heterosexuals (39.8%wt). HIV testing was similarly high for Blacks (59.7%wt), Hispanics (59.1%wt), and Others (52.3%wt), and lower for Whites (40.3%wt) and Asians (28.3%wt).

Correlates of HIV Testing

HIV testing was significantly related to gender, race/ethnicity, sexual orientation, lifetime and recent same-sex and opposite-sex activity, AOD dependence, having a physical exam in the past year, number of past-year provider visits for physical health problems, lifetime STI treatment, and plans to get tested for HIV in the next year (Table 1). For opposite-sex activities, HIV testing was significantly associated with earlier sexual debut, greater number of sex partners, and more unprotected sexual encounters. For same-sex activities, power was insufficient to detect significant differences. HIV testing was unrelated to age, neighborhood income, religiosity, and college enrollment status.

Gender Differences

Among those who engaged in same-sex activities during the past six months, men had more same-sex partners than women (4.4 vs. 1.7, p=.006; Table 2). No other gender differences were observed on the sexual behavior measures we tested. Women consistently differed from men on health service utilization variables, with more provider visits (2.6 vs. 1.6, p=.007), greater likelihood of lifetime STI treatment (17.7% vs. 7.6%, p<.001), and greater likelihood of a past-year physical exam (70.5% vs. 55.0%, p<.001). Men were more likely than women to exhibit AOD dependence (11.4% vs. 7.3%, p=.029).

Race/Ethnicity Differences

Few significant race/ethnicity differences were found for sexual behaviors in the past six months. Blacks (65.9%) and Asians (60.7%) were less likely than either Whites (79.9%) or Others (88.5%, all ps<.05) to have had an opposite-sex partner, but among those who did, they had similar numbers of partners and unprotected encounters. The only race/ethnicity differences in health service utilization were that Hispanics had significantly more provider visits (4.6) than Whites (2.0) or Asians (1.6; both p<.05) and were more likely to have had STI treatment (27.9% vs. 12.0% and 5.6%, respectively; both p<.05).

Logistic Regression Predicting HIV Testing

Men were half as likely as women to have been tested for HIV (AOR=.48, 95%CI=.37–.62, p<.001; Table 3), controlling for race/ethnicity, sexual orientation, and neighborhood income. Both Blacks (AOR=2.46, 95%CI=1.31–4.63, p=.005) and Hispanics/Others (AOR=2.30, 95%CI=1.25–4.25, p=.008) were more than twice as likely as Asians to have been tested for HIV, and heterosexuals were one-third as likely as non-heterosexuals (AOR=.34, 95%CI=.17–.66, p=.001). All other variables tested were significantly associated with greater likelihood of HIV testing, even holding constant sexual orientation, gender, neighborhood income, and race/ethnicity). AOD dependent individuals were 2.33 times as likely as non-dependent individuals to have been tested for HIV (95%CI=1.45–3.75, p=.001), and individuals with STI treatment history were 3.09 times as likely (95%CI=1.99–4.81, p<.001). The statistical effects of sexual orientation, gender, neighborhood income, and race/ethnicity were essentially unchanged regardless of inclusion of each behavioral variable, with the sole exception that when frequency of unprotected sex was held constant, the Black vs. White contrast became significant (AOR=1.82, 95%CI=1.13–2.93, p=.015; not shown in table). Neighborhood income was the only variable tested that was not related to HIV testing (not shown in table).

TABLE 3.

Results of logistic regression testing the association of HIV testing with six hypothesized behavioral predictor variables, holding constant personal characteristics

Bivariate
Multivariate
OR (95% CI) p AOR (95% CI)a pa
PERSONAL CHARACTERISTICS
 Gender=Male .47 (.36–.61) <.001 .48 (.37–.62) <.001
 Race/ethnicity (Reference=Asian)
  Black 2.86 (1.56–5.24) .001 2.46 (1.31–4.63) .005
  Hispanic/Other 2.57 (1.42–4.67) .002 2.30 (1.25–4.25) .008
  White 1.60 (1.01–2.53) .047 1.59 (.987–2.55) .057
 Race/ethnicity (Reference=White)
  Hispanic/Other 1.61 (1.04–2.49) .031 1.45 (.93–2.27) .103
  Asian .63 (.40–.99) .047 .63 (.39–1.01) .057
  Black 1.79 (1.15–2.80) .011 1.55 (.97–2.48) .065
 Heterosexual .33 (.17–.63) .001 .34 (.17–.66) .001
BEHAVIORAL VARIABLES
 Dependent on alcohol and/or marijuana 2.06 (1.31–3.25) .002 2.33 (1.45–3.75) .001
 Number of unprotected sexual encounters, past 6 months (in units of 10) 1.14 (1.09–1.19) <.001 1.15 (1.10–1.20) <.001
 Number of sex partners, past 6 months 1.47 (1.31–1.66) <.001 1.47 (1.30–1.67) <.001
 Physical exam by medical professional, past year 2.19 (1.67–2.87) <.001 1.97 (1.48–2.61) <.001
 Number of visits for physical health problems, past year 1.06 (1.02–1.10) .005 1.04 (1.01–1.08) .027
 Treated for STI, lifetime 3.92 (2.55–6.02) <.001 3.09 (1.99–4.81) <.001
Open in a new tab
a

Multivariate results for personal characteristics are adjusted for neighborhood income (not shown) and all other personal characteristics shown. Multivariate results for each behavioral variable are adjusted for neighborhood income (not shown) and all personal characteristics shown. Neighborhood income was not related to HIV testing in either bivariate or multivariate analyses.

Two significant first-order interactions were found. First, gender moderated the statistical effect of AOD dependence (p=.005) on HIV testing, such that dependence was more strongly associated with HIV testing in women than men. The estimated probability of HIV testing in dependent and non-dependent women was .94 and .66, respectively (p<.001), compared with .56 and .50 in men (p=.386). Secondly, gender interacted with STI treatment (p<.001) such that the estimated probability of HIV testing in men with and without a history of STI treatment was .91 and .43, respectively (p<.001), whereas in women the difference was less pronounced (.77 and .63, p=.006). Race/ethnicity did not interact significantly with any of the six behavioral variables tested.

Discussion

In this study of young adults recently enrolled in college, 41.9%wt had been tested for HIV at some point in their lifetime, considerably more than previous estimates from national data on either college students (American College Health Association, 2010)or young adults (Johnson et al., 2010), but less than in one large study of HBCU (Thomas et al., 2008). These in consistencies could be attributable to methodological differences or institutional differences in access to campus-based health services. As in previous studies, participants engaging in higher-risk sexual behaviors—including earlier sexual debut, more sexual partners, and more unprotected sex—were more likely to be tested for HIV. Results suggest modest success in adopting traditional HIV-testing strategies targeting high-risk individuals in this sample.

The finding that contact with health care providers—for either physical examination or STI treatment—was associated with greater likelihood of HIV testing adds to the small body of evidence that individuals who seek health services, especially campus health services, are more likely to be tested (Thomas et al., 2008). Another important contribution is the finding that HIV testing was positively associated with AOD dependence, which is not surprising given prior evidence that substance use and high-risk sexual behaviors are highly correlated (Brown & Vanable, 2007; Wechsler, Dowdall, Davenport, & Castillo, 1995). To the extent that HIV testing services were in higher demand among AOD-dependent individuals than their non-dependent counterparts—whether due to greater perceived susceptibility to contracting HIV, greater acceptability of HIV testing, or other reasons—findings suggest that AOD treatment programs might be an important setting for providing routine HIV testing, especially for women. Prior studies indicate that the availability of HIV testing (Chriqui, Terry-McElrath, McBride, & Eidson, 2008; Knudsen & Oser, 2009)and testing rates (Pollack & D’Aunno, 2010)are relatively low within these settings, yet one study of alcohol treatment clients documented high levels of sexual risk-taking and HIV infection, even in the absence of injection drug use (Woods et al., 2000). Further study is needed to understand the specific settings in which college students seek HIV testing.

Race/ethnicity differences were largely consistent with prior studies. By some measures, Blacks appeared to be at lower risk for HIV than Whites, such as by being less likely than Whites to have had an opposite-sex partner and having fewer unprotected sexual encounters. On the other hand, Blacks were more likely to have had a same-sex partner and to have a history of STI treatment or HIV testing. This apparently conflicting pattern of differences echoes prior findings from the most recent national study of college students, wherein Blacks had higher rates of HIV testing and STI, yet also had higher rates of protective behaviors like condom use (Buhi et al., 2010). Nationally, as in our sample, both Blacks and Hispanics are more likely than Whites to be tested for HIV (Duran, Beltrami, Stein, Voetsch, & Branson, 2008), whereas Asians remain less likely to be tested (Zaidi et al., 2005). Further research is needed to understand the factors underlying this association, such as possible race-related differences in attitudes and beliefs about HIV testing.

Results must be interpreted in light of certain limitations. All data were subject to self-report bias, and despite excellent follow-up rates, we cannot rule out the possibility of attrition bias. Participants were recruited from one university, so generalizability to other populations is limited. Although the overall sample was large, cell sizes were insufficient to permit comparisons amongst some minority groups, such as non-heterosexuals. Sexual risk behavior is a notoriously complex and difficult construct to operationalize, and while we are confident in the validity of our measures, we must acknowledge their limitations. Because we had no information about sex partners’ HIV status or testing behaviors, we could not account for differences in the relative HIV transmission risk associated with each partner or encounter, and therefore counted all sex partners and unprotected sexual encounters equally. We did not ask about all possible types of opposite-sex encounters (i.e., oral, anal) and cannot say how much we might have underestimated the number of opposite-sex partners or encounters in which vaginal sex did not occur.

Even with these limitations, this study has important strengths. Our method of measuring sexual behavior in self-reported numbers of encounters, both protected and unprotected, provides more precision and flexibility than many other standard methods that capture condom use frequency in subjectively defined ranges (e.g., “sometimes,” “always”) and permits meaningful comparisons between individuals having sex at very different frequencies. Other strengths include the large overall sample size and ability to examine the possible moderating effects of gender on the relationships between HIV testing and health risk behaviors and service utilization. Lastly, this study examines AOD dependence within the college population, a previously understudied correlate of HIV testing.

Conclusions

The finding that nearly half of our college-educated sample had been tested for HIV is encouraging and demonstrates the feasibility of achieving high rates of HIV testing in a college population. This finding reflects positively on the efforts of campus-based health services. The sample had broad access to HIV testing during the years they were enrolled: their campus health center was providing confidential testing via both rapid and traditional testing methods for a small fee on a walk-in basis at the lab during normal business hours, and interaction with a health care provider was not required. There is some evidence that access to testing on a college campus may increase testing rates. In one study of individuals seeking voluntary testing at a college health center, 73.1% said they would not have sought a HIV test had it not been offered on campus (Anastasi, Sawyer, & Pinciaro, 1999).

Despite this apparent success, the feasibility of achieving universal HIV testing on a college campus remains to be seen. The present sample had high levels of health service utilization, and although we did not ask participants if they ever opted-out of HIV testing when it was offered to them, it seems unlikely that this could account for a substantial proportion of the untested individuals. Moreover, the health center’s standard practice was to conduct risk-based rather than universal HIV testing. The American College Health Association’s (ACHA) most recent guidance on HIV prevention strategies (Hoban, Ottenritter, Gascoigne, & Kerr, 2003)was published before the CDC recommendation and makes no mention of universal testing. Yet HIV infection in young people remains a grave public health concern, with one in 1,000 young adults already infected according to recent estimates (Campsmith et al., 2008; Morris et al., 2006). Revision of ACHA guidance on HIV testing and enhanced provider outreach might be needed to encourage college health center providers to implement universal testing. Further study of college students’ attitudes about HIV testing is also needed to inform the design of testing campaigns (Boshamer & Bruce, 1999).

In this study women were more likely than men to be tested for HIV, even controlling for their higher health service utilization, which extends prior evidence from college-student and young-adult populations (Buhi et al., 2010; Thomas et al., 2008; Ward, Barnes, Freeman, & Schiller, 2010). Considering that men account for 73% of new HIV infections nationally (Prejean et al., 2008), this finding underscores the need for continued outreach to increase HIV testing in young men. Present findings point to the importance of STI treatment settings in promoting HIV testing for men, especially considering that men tend to have less contact with health care providers in general relative to women, and therefore fewer opportunities for HIV testing. It is also likely that men might benefit from alternative outreach strategies that provide HIV testing outside medical settings.

A considerable proportion of our sample (17.7% of women, 7.6% of men) had undergone STI treatment, and while most of them (75%) were tested for HIV, it is surprising that so many were not. The American College of Physicians recognizes patients presenting for STI testing and treatment as a high-risk population for targeted HIV testing (Qaseem, Snow, Shekelle, Hopkins, & Owens, 2009), so any college student who does not undergo HIV testing while in care for STI treatment represents an important missed opportunity. Extrapolating to the entire U.S. population of 18.7 million undergraduate students in the U.S. (United States Department of Education, 2007), this figure translates to more than 600,000 untested individuals at high risk for HIV. Further research may be needed to understand the barriers to HIV testing in STI treatment situations.

Acknowledgments

This work was supported by the National Institute on Drug Abuse at the National Institutes of Health (R01-DA14845, Dr. Arria PI). Special thanks are extended to Rebecca Baron, Laura Garnier-Dykstra, Jessie Tsai, Emily Winick, the interviewing team, and the participants.

References

  1. American College Health Association. American College Health Association-National College Health Assessment II: Reference Group Data Report Fall 2009. Baltimore, MD: American College Health Association; 2010. [Google Scholar]
  2. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders: DSM-IV. 4. Washington, DC: American Psychiatric Press; 1994. [Google Scholar]
  3. Anastasi M-C, Sawyer RG, Pinciaro PJ. A descriptive analysis of students seeking HIV antibody testing at a university health service. Journal of American College Health. 1999;48(1):13–19. doi: 10.1080/07448489909595667. [DOI] [PubMed] [Google Scholar]
  4. Arria AM, Caldeira KM, O’Grady KE, Vincent KB, Fitzelle DB, Johnson EP, Wish ED. Drug exposure opportunities and use patterns among college students: Results of a longitudinal prospective cohort study. Substance Abuse. 2008;29(4):19–38. doi: 10.1080/08897070802418451. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Barth KR, Cook RL, Downs JS, Switzer GE, Fischhoff B. Social stigma and negative consequences: Factors that influence college students’ decisions to seek testing for sexually transmitted infections. Journal of American College Health. 2002;50(4):153–159. doi: 10.1080/07448480209596021. [DOI] [PubMed] [Google Scholar]
  6. Bontempi JB, Mugno R, Bulmer SM, Danvers K, Vancour ML. Exploring gender differences in the relationship between HIV/STD testing and condom use among undergraduate college students. American Journal of Health Education. 2009;40(2):97–105. [Google Scholar]
  7. Boshamer CB, Bruce KE. A scale to measure attitudes about HIV-antibody testing: Development and psychometric validation. AIDS Education and Prevention. 1999;11(5):400–413. [PubMed] [Google Scholar]
  8. Branson BM, Handsfield HH, Lampe MA, Janssen RS, Taylor AW, Lyss SB, Clark JE. Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. Morbidity and Mortality Weekly Report. 2006;55(RR14):1–17. [PubMed] [Google Scholar]
  9. Brown JL, Vanable PA. Alcohol use, partner type, and risky sexual behavior among college students: Findings from an event-level study. Addictive Behaviors. 2007;32(12):2940–2952. doi: 10.1016/j.addbeh.2007.06.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Buhi ER, Marhefka SL, Hoban MT. Thestate of the union: sexual health disparities in a national sample of US college students. Journal of American College Health. 2010;58(4):337–346. doi: 10.1080/07448480903501780. [DOI] [PubMed] [Google Scholar]
  11. Campsmith ML, Rhodes P, Hall HI, Green T. HIV prevalence estimates—United States, 2006. Morbidity and Mortality Weekly Report. 2008;57(39):1073–1076. [PubMed] [Google Scholar]
  12. Centers for Disease Control and Prevention. CDC surveillance summaries, youth risk behavior surveillance: National College Health Risk Behavior Survey, United States, 1995. Morbidity and Mortality Weekly Report. 1997;46(SS-6):1–64. [PubMed] [Google Scholar]
  13. Centers for Disease Control and Prevention. Diagnoses of HIV infection and AIDS in the United States and dependent areas, 2008. Atlanta, GA: National Center for HIV/AIDS Viral Hepatitis STD and TB Prevention; 2010. [Google Scholar]
  14. Chriqui JF, Terry-McElrath Y, McBride DC, Eidson SS. State policies matter: The case of outpatient drug treatment program practices. Journal of Substance Abuse Treatment. 2008;35(1):13–21. doi: 10.1016/j.jsat.2007.08.012. [DOI] [PubMed] [Google Scholar]
  15. Crosby RA, Miller KH, Staten RR, Noland M. Prevalence and correlates of HIV testing among college students: An exploratory study. Sexual Health. 2005;2(1):19–22. doi: 10.1071/SH04047. [DOI] [PubMed] [Google Scholar]
  16. Duran D, Beltrami J, Stein R, Voetsch AC, Branson BM. Persons tested for HIV-United States, 2006. Morbidity and Mortality Weekly Report. 2008;57(31):845–849. [PubMed] [Google Scholar]
  17. Eng TR, Butler WT. The hidden epidemic: confronting sexually transmitted diseases. Washington, DC: Institute of Medicine, National Academy Press; 1997. [PubMed] [Google Scholar]
  18. Hall HI, Song R, Rhodes P, Prejean J, An Q, Lee LM, Karon J, Brookmeyer R, Kaplan EH, McKenna MT, Janssen RS. Estimation of HIV incidence in the United States. Journal of the American Medical Association. 2008;300(5):520–529. doi: 10.1001/jama.300.5.520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Hoban MT, Ottenritter NW, Gascoigne JL, Kerr DL. Campus HIV prevention strategies: Planning for success. Washington, DC: American Association of Community Colleges, American College Health Association; 2003. [Google Scholar]
  20. Johnson AS, Heitgerd J, Koenig LJ, VanHandel M, Branson BM, Connelly E, Hall HI, Valleroy LA. Vital Signs: HIV testing and diagnosis among adults-United States, 2001–2009. Morbidity And Mortality Weekly Report. 2010;59(47):1550–1555. [PubMed] [Google Scholar]
  21. Knudsen HK, Oser CB. Availability of HIV-related health services in adolescent substance abuse treatment programs. AIDS Care. 2009;21(10):1238–1246. doi: 10.1080/09540120902803182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Marelich WD, Clark T. Human immunodeficiency virus (HIV) testing and false disclosures in heterosexual college students. Journal of American College Health. 2004;53(3):109–115. doi: 10.3200/JACH.53.3.109&#x02013;116. [DOI] [PubMed] [Google Scholar]
  23. Mattson M. Impact of HIV test counseling on college students’ sexual beliefs and behaviors. American Journal of Health Behavior. 2002;26(2):121–136. doi: 10.5993/ajhb.26.2.5. [DOI] [PubMed] [Google Scholar]
  24. Melissa DATA. Income tax statistics lookup. 2003 Retrieved May 28, 2008, from http://www.melissadata.com/lookups/taxzip.asp.
  25. Morris M, Handcock MS, Miller WC, Cohen MS, Ford CA, Schmitz JL, Hobbs MM, Harris KM, Udry JR. Prevalence of HIV infection among young adults in the United States: Results from the Add Health Study. American Journal of Public Health. 2006;96(6):1091–1097. doi: 10.2105/AJPH.2004.054759. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Pollack HA, D’Aunno T. HIV testing and counseling in the nation’s outpatient substance abuse treatment system, 1995–2005. Journal of Substance Abuse Treatment. 2010;38(4):307–316. doi: 10.1016/j.jsat.2009.12.004. [DOI] [PubMed] [Google Scholar]
  27. Prejean J, Song R, An Q, Hall HI. Subpopulation estimates from the HIV incidence surveillance system—United States, 2006. Morbidity and Mortality Weekly Report. 2008;57(36):985–989. [PubMed] [Google Scholar]
  28. Qaseem A, Snow V, Shekelle P, Hopkins R, Jr, Owens DK. Screening for HIV in health care settings: A guidance statement from the American College of Physicians and HIV Medicine Association. Annals of Internal Medicine. 2009;150(2):125–131. doi: 10.7326/0003-4819-150-2-200901200-00300. [DOI] [PubMed] [Google Scholar]
  29. Ruiz MS, Gable AR, Kaplan EH, Stoto MA, Fineberg HV, Trussell J. No time to lose: Getting more from HIV prevention. Washington, DC: Institute of Medicine, National Academy Press; 2000. [PubMed] [Google Scholar]
  30. Scholly K, Katz AR, Gascoigne J, Holck PS. Using social norms theory to explain perceptions and sexual health behaviors of undergraduate college students: An exploratory study. Journal of American College Health. 2005;53(4):159–166. doi: 10.3200/JACH.53.4.159&#x02013;166. [DOI] [PubMed] [Google Scholar]
  31. Sher KJ, Rutledge PC. Heavy drinking across the transition to college: Predicting first-semester heavy drinking from precollege variables. Addictive Behaviors. 2007;32(4):819–835. doi: 10.1016/j.addbeh.2006.06.024. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Substance Abuse and Mental Health Services Administration. 2002 National Survey on Drug Use and Health Questionnaire. Rockville, MD: Office of Applied Studies; 2003. [Google Scholar]
  33. Substance Abuse and Mental Health Services Administration. Summary of National Findings. I. Rockville, MD: Office of Applied Studies; 2010. Results from the 2009 National Survey on Drug Use and Health. [Google Scholar]
  34. Thomas PE, Voetsch AC, Song B, Calloway D, Goode C, Mundey L, Nobles J, Sly K, Smith MR, Williams B, Shiloh M, Patterson K, Ward S, Sullivan PS, Heffelfinger JD. HIV risk behaviors and testing history in historically black college and university settings. Public Health Reports. 2008;123(Supplement 3):115–125. doi: 10.1177/00333549081230S314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Trieu SL, Modeste NN, Marshak HH, Males MA, Bratton SI. Partner communication and HIV testing among US Chinese college students. American Journal of Health Behavior. 2010;34(3):362–373. doi: 10.5993/ajhb.34.3.11. [DOI] [PubMed] [Google Scholar]
  36. United States Department of Education. Enrollment, staff, and degrees conferred in postsecondary institutions participating intitle IV programs, by type and control of institution, sex of student, type of staff, and type of degree: Fall 2007 and 2007–08. Washington, DC: National Center for Education Statistics; 2007. [Google Scholar]
  37. Vincent KB, Kasperski SJ, Caldeira KM, Garnier-Dykstra LM, Pinchevsky GM, O’Grady KE, Arria AM. Maintaining superior follow-up rates in a longitudinal study: Experiences from the College Life Study. International Journal of Multiple Research Approaches. doi: 10.5172/mra.2012.6.1.56. (in press) [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Ward BW, Barnes PM, Freeman G, Schiller JS. Early release of selected estimates based on data from the January–June 2010 National Health Interview Survey. Atlanta, GA: National Center for Health Statistics; 2010. [Google Scholar]
  39. Wechsler H, Dowdall GW, Davenport A, Castillo S. Correlates of college student binge drinking. American Journal of Public Health. 1995;85(7):921–926. doi: 10.2105/AJPH.85.7.921. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Woods WJ, Lindan CP, Hudes ES, Boscarino JA, Clark WW, Avins AL. HIV infection and risk behaviors in two cross-sectional surveys of heterosexuals in alcoholism treatment. Journal of Studies on Alcohol. 2000;2(61):262–266. doi: 10.15288/jsa.2000.61.262. [DOI] [PubMed] [Google Scholar]
  41. Zaidi IF, Crepaz N, Song R, Wan CK, Lin LS, Hu DJ, Sy FS. Epidemiology of HIV/AIDS among asians and pacific islanders in the United States. AIDS Education and Prevention. 2005;17(5):405–417. doi: 10.1521/aeap.2005.17.5.405. [DOI] [PubMed] [Google Scholar]

RESOURCES