Older Women and HIV Testing: Examining the Relationship Between HIV Testing History, Age, and Lifetime HIV Risk Behaviors

HUMAN IMMUNODEFICIENCY VIRUS (HIV) testing rates among Americans over age 50 are low. Although 44% of US adults have been tested, only 10% to 15% of those over age 45 have been tested.1 Older adults are frequently tested late in their disease course.2–5 As a consequence, they are more likely to present with opportunistic infections, progress to acquired immune deficiency syndrome, and die within a year of diagnosis.2,6–9 Few studies have examined HIV testing and HIV risk behaviors among older persons,4,10–13 particularly among older women from high-prevalence communities.10,13 In this study, we examined HIV testing history among older women in a high-prevalence community. Our objectives were to: (a) determine the prevalence of past HIV testing, (b) identify predictors of HIV testing, and (c) examine the distribution of HIV risk factors among women based on age and HIV testing history. We conducted a cross-sectional questionnaire in a general internal medicine clinic at a large hospital in Atlanta, GA between June 2001 and July 2002. Eligible women were English speakers over age 50 presenting for routine medical visits. Excluded women were incarcerated, acutely ill requiring immediate medical attention, or not fully communicative because of mental retardation, dementia, or intoxication. Of 564 eligible women, 514 (91%) participated. This analysis includes the 488 women who answered the question about their HIV testing history. Our recruitment and consent process is described in detail elsewhere.14–16 The 68-item questionnaire was administered in a face-to-face interview. The Institutional Review Board at Emory University School of Medicine approved the protocol. The main outcome measure was self-reported past HIV testing. We included independent factors that influence HIV testing behavior4,17–24 and HIV risk25,26 including sociodemographics, HIV knowledge, perceived HIV risk, actual HIV risk, whether currently sexually active, and recall that a provider ever recommended HIV testing. Demographic variables included age, race/ethnicity, marital status, education, employment, and sexual orientation. We modified a previously validated item assessing perceived HIV risk27 and 2 previously validated scales measuring HIV knowledge27 and lifetime risk of exposure to HIV.28 A detailed description of each modified scale is provided elsewhere.14 The 9-item HIV knowledge scale was scored from 0 to 9 with higher scores indicating greater knowledge. We dichotomized scores using a median split. The actual HIV risk scale classified women as low, moderate, or high risk for lifetime exposure to HIV28 based on medical history, personal, or partner risk behaviors. Univariate analysis was performed to obtain descriptive statistics. For the bivariate analysis, we used Pearson χ2 (or the Fisher exact test) for categorical variables and the Student t test or Mann–Whitney U test for continuous variables. We performed stratified analyses based on HIV testing history and age groups. A multivariable logistic regression model identified independent predictors of past HIV testing. The model contained sociodemographics, perceived and lifetime HIV risk, sexual activity, and recall that a provider recommended HIV testing. We report the adjusted proportion of women reporting each characteristic rather than odds ratios as the odds ratios might overstate the magnitude of the associations.29 Two-tailed P values were used throughout with statistical significance set at P <0.05. Statistical analysis was performed using STATA version 8.1 (StataCorp, College Station, TX). Women ranged in age from 50 to 84 years (median 61). The majority were black, heterosexual, and not sexually active. Most had low HIV knowledge scores, low HIV risk perception, and moderate- or high-risk factors. Previous HIV testing was reported by 169 (35%) women of whom 156 (92%) knew the results of their test, 2 (1.3%) reported being HIV-positive. Compared to never-tested women, previously tested women were more likely to be younger and recall that a provider ever suggested HIV testing (Table 1. Thirty-nine percent of women aged 50 to 54 had been previously tested compared to 26% of those aged 55 to 59, 21% of those aged 60 to 64 and 14% of women age ≥65 years (P <0.001). Previously tested women also had higher HIV knowledge and were more likely to be sexually active, have moderate- or high-risk factors, and perceive themselves at high risk for HIV.TABLE 1: Characteristics of Sample and Women Based on History of HIV TestingOnly 2 variables predicted a history of HIV testing: younger age and recall that a provider recommended testing. As age increased, the likelihood of previous HIV testing decreased significantly. Fifty-four percent of women aged 50 to 54 had been previously tested compared with 49% of women aged 55 to 59, 37% of women aged 60 to 64, and 16% of women over age 65 (P = 0.001). The strongest predictor of past HIV testing was recall that a provider ever recommended HIV testing: 91% of women who recalled a provider recommending HIV testing had been tested HIV compared with 16% of women who did not recall a provider recommending HIV testing (P <0.001). Women aged 50 to 54 were more likely to have almost every risk factor compared with older women (Table 2). This difference was statistically significant for 3 risk factors: partner with a history of intravenous drug use, prostitution, and number of sexual partners in the past 3 decades. Similarly, previously tested women were more likely to report all of the HIV risk factors compared with never-tested women (data not shown) with this difference statistically significant for 6 risk factors: partner with a history of IVDU (19% vs. 4%, P <0.001), HIV+ partner (1% vs. 0%, P = 0.05), history of exchanging sex for money (9% vs. 3%, P = 0.003), having 2 to 5 (34% vs. 28%, P = 0.02) or >6 partners (3% vs. 1%, P = 0.02) since 1978, and exposure to blood products (8% vs. 3%, P = 0.02).TABLE 2: Distribution of HIV Risk Factors, Stratified by AgeOur findings suggest that the upper age limit for offering HIV testing in the Centers for Disease Control’s (CDC) new HIV testing guidelines may not be warranted in high-prevalence communities. The new guidelines recommend that providers offer HIV testing to adults aged 13 to 64 with the upper limit chosen because persons over age 65 comprise less than 2% of new HIV infections.30 Thus, the guidelines were designed to target the age group most likely to be infected. Our sample had a high HIV prevalence (>1%) with almost two thirds of women over age 50 and nearly half of those over 65 reporting moderate- or high-risk factors for lifetime exposure to HIV. However, less than a third of the women had ever had HIV testing and testing rates decreased with age. A major concern is that the CDC’s upper age limit might inadvertently prevent providers from offering testing to older women who are at risk. This is a real concern given previous studies demonstrating that providers frequently fail to perform sexual risk assessments, provide risk-reduction counseling, or recommend HIV testing for older patients.4,13,31,32 Studies performed in a variety of settings, including the current study, found HIV testing was associated with provider recommendations for testing.18,22,33–35 At a minimum, providers should routinely obtain sexual histories from older women assessing women’s current and past sexual partners’ risk behaviors and should offer HIV testing, if appropriate. Providers should also counsel older women about safe sex behaviors, such as condom use, given previous studies demonstrating that older women are less likely to use condoms.36–39 There are several considerations when interpreting our results. Participants were predominantly nonwhite, from a Southeastern community with a high HIV prevalence and cannot necessarily be generalized to populations with different sociodemographics or a lower HIV prevalence. The cross-sectional design does not allow us to assess the directionality or validity of the observed associations, such as the association between recall that a provider recommended HIV testing and HIV testing history. Selective recall bias may have overinflated the magnitude of this association because previously tested women were more likely to recall a provider recommended testing. Our measures of lifetime HIV risk and HIV prevalence were self-reported and thus subject to recall and social desirability bias. Some HIV-positive women may have chosen not to disclose their HIV status. Similarly, some women may have chosen not to report their own or their partner’s HIV risk behaviors. Despite these potential limitations, we were able to determine risk status for more than 80%, almost two thirds of who reported moderate- or high-risk factors. We did not determine when, why, or even if previous HIV testing had occurred. Understanding why previous testing had occurred would have provided valuable additional information for understanding the HIV testing behaviors of older women in this population, but this data were not collected. Previous studies indicate HIV testing in older persons is often performed for persistent, unexplained symptoms. We did ask women who were interested in HIV testing to provide a reason for their interest. Most were concerned about their health, safety, or a previous or current partner’s risk status.14 Finally, we did not assess the number of women who received a HIV test on the day of the questionnaire. The results of those tests would have provided a better idea of the HIV prevalence than our self-reported measure. This study demonstrates that lifetime prevalence of HIV risk factors can be substantial among older women in high-risk communities. Our findings suggest that the upper age limit for CDC’s new opt-out policy may perpetuate low testing rates among older, at-risk women. The results reinforce the fact that providers should obtain social and sexual histories from older patients assessing current and past HIV risk behaviors and recommend HIV testing as appropriate. HIV prevention programs targeting older persons have been shown to be effective at increasing HIV knowledge, empowering older adults to accurately assess their HIV risk status and that of their partners, and increasing HIV testing rates.30 These programs have not been widely adopted. Given that the US population is aging rapidly, it is important that we replicate effective HIV education prevention and testing programs targeting older adults.