The advent of combination antiretroviral therapy has resulted in great advances in treatment of HIV infection, with a subsequent and dramatic drop in morbidity and mortality from HIV infection.1,2 A complex set of factors, including socioeconomic status, health status, side effects, and drug use, all influence an individual's ability to comply with an antiretroviral drug regimen. Previous studies point to lower adherence in racial minorities.3-5 Few studies, however, have investigated specific factors associated with lower rates of adherence in blacks and Hispanics in the United States. Although blacks and Hispanics are often treated as separate uniform categories, each represents a wide diversity of cultures and genetics that overlap in some areas.
Current usage of racial categories masks rich differences within groups. Ethnicity is frequently used interchangeably with race, although there are subtle differences that make them distinct concepts. Race is often socially imposed and dependent on one's biology. Ethnicity, however, is more closely related to culture and with which group an individual identifies. This distinction is important to note as race and ethnicity may affect health behaviors in different ways.
Hispanics and blacks have been disproportionately affected by the HIV epidemic. Although only 13% of the US population is black, they account for about half (49%) of new infections.6 Similarly, although only 15% of the US population is Hispanic, this group comprises 17% of all new HIV infections.7 Thus, it is very important to address the specific needs of these populations. We investigate how adherence differs by race and what factors are associated with adherence in each racial group. We also investigate how ethnicity affects adherence.
Population and Data Collection
The Multicenter AIDS Cohort Study (MACS) is an ongoing prospective study of HIV infection among homosexual and bisexual men conducted in 4 study sites: Baltimore, Chicago, Pittsburgh, and Los Angeles.8-10 The men are followed every 6 months with interviews, physical examinations, quality of life assessments, neuropsychological testing, and blood sample collections. An interviewer-administered questionnaire to assess self-reported adherence to antiretroviral medication was incorporated into the protocol in October 1998.3 The institutional review board of each center approved the study protocols, and informed consent was obtained from each participant.
Participants in the original cohort were enrolled in 1984-1985 and 1987-1990. Recruitment to include more minorities (specifically blacks and Hispanics) in the MACS began in October 1, 2001 and ended in April 8, 2003. Data for this study were taken from the newly recruited participants and the participants already enrolled in the MACS. Over 90% of the new cohort was recruited after April 1, 2002. Thus, analysis was limited to subjects who (1) came in for an assessment between April 2002 and October 2006 and (2) had been administered the adherence questionnaire within that period. Only participants who had started antiretroviral therapy were given the adherence questionnaire. Subjects who were first given the adherence questionnaire after October 2005 were excluded due to insufficient follow-up time.
To collect information on race, participants were asked, “Which of the following best describes your racial background?” Participants were able to choose from one of the following options: white non-Hispanic, white Hispanic, black Hispanic, black non-Hispanic, American Indian or Alaskan Native, Asian or Pacific Islander, other Hispanic, or other. For analysis, these categories were collapsed into 4 categories: (1) white, (2) Hispanic, (3) black, and (4) other.
To collect information on ethnicity, all participants were asked, “Please tell us which of the following best describes your family's national origins or ethnic background.” Ethnicity data were collected in 19 different categories: Western European, Eastern European Jewish, Eastern European non-Jewish, Mediterranean, Middle Eastern, Central or South American, Central or South Asian, Southeastern Asian, Chinese, Japanese, Pacific Islander, Haitian, Jamaican, Cuban, other Caribbean, East African, Zairian, other African, and other. For analyses, these categories were collapsed into 7 categories: (1) European, (2) Middle Eastern, (3) Central/South American, (4) Asian/Pacific Islander, (5) Caribbean, (6) African, and (7) other.
Self-reported adherence was dichotomized as 100% adherence and less than 100% adherence. Inclusion in the <100% adherence group for a visit resulted when at least one of the following criteria was met: (1) the participant reported taking less medication than prescribed in the past 4 days, (2) the participant reported taking all prescribed doses in the past 4 days, but not in a typical pattern, (3) the participant reported taking fewer pills per dose than were prescribed, (4) the participant reported ever skipping any medications, or (5) the participant reported ever not following his medication schedule.
From the information collected on clinical symptoms, prescription drug use, socioeconomic status, behavioral characteristics, recreational drug use, and biological markers, a subset of variables with a theoretical relationship with adherence were chosen for analysis. Variables with missing data, no correlation to adherence, low variability, or that duplicated other variables were excluded from the analysis. Viral load was chosen for analysis over CD4 because impact of adherence on change in CD4 count tends to be delayed.11
Multilevel logistic regression using generalized estimating equations was used to analyze the longitudinal clustered data. Similar models were then separately built for whites, Hispanics, and blacks. Rates of adherence within Hispanic and black subpopulations were next examined to determine whether ethnic and cultural differences within race affect adherence.
A total of 1102 men participated in this study, comprising 6203 visits for analysis. There were 644 white men (58.4%), 151 Hispanic men (13.7%), 289 black men (26.2%), and 18 men (1.6%) of other racial backgrounds. Racial breakdown was different for the cohort recruited before April 2001 and that recruited after April 2001. The older cohort had 509 white men (82.8%), 36 Hispanic men (5.9%), 63 black men (10.2%), and 7 (1.1%) men of other racial backgrounds. The newer cohort had 135 white men (27.7%), 115 Hispanic men (23.6%), 226 black men (46.4%), and 11 (2.3%) men of other racial backgrounds.
The most obvious differences between racial groups were in socioeconomic status. The proportion of men who reported individual gross income less than $10,000 a year at a visit was 10.2% for whites, 31.3% for Hispanics, and 43.5% for blacks. The proportion of men reporting financial difficulty meeting expenses at a visit was 14.0% for whites, 37.6% for Hispanics, and 48.1% for blacks. Substantial differences were also found in illicit drug use. Hash or marijuana use was similar in whites and blacks (36.5% and 34.8%, respectively), but lower in Hispanics (25.0%). Whites had a higher proportion of popper use (33.8%) than Hispanics and blacks (19.1% and 14.3%, respectively). Blacks had a higher proportion of crack use (21.8%) compared with whites and Hispanics (3.9% and 7.7%, respectively).
Whites had the highest percentage of visits with self-reported 100% adherence (44.2%). Hispanics had 31.8% adherent visits, blacks had 28.1% adherent visits, and all others had 21.9% adherent visits.
All variables included in the multivariate analysis were significantly associated with nonadherence in the univariate analysis. After controlling for center, age, financial difficulties, popper use, crack use, and viral load, whites had the least reported nonadherence. Hispanics had 2.16 (CI, 1.47 to 3.18) and blacks had 1.37 (CI, 1.05 to 1.79) times more nonadherence compared with whites. Viral load was significantly associated with adherence.
Different factors were associated with nonadherence in the multivariate analysis within each separate racial group. For whites, younger age, joint pain, and viral load were all significantly associated with nonadherence. For Hispanics, younger age and viral load were associated with nonadherence. Determining variables associated with nonadherence in blacks was complex. Surprisingly, age and viral load were not associated with adherence. A new skin rash, doubling the cost of personal expenditure for prescription medication, financial difficulty meeting expenses, and crack use were all significantly associated with nonadherence. Conversely, numbness in feet/legs and crystal/methamphetamine use were significantly associated with lower rates of nonadherence.
Among Hispanics, participants primarily reported European, Central/South American and Caribbean ethnicity. Those reporting European descent had the highest rates of adherence (44.6%). Those of Central/South American and Caribbean descent had much lower rates of adherence (27.5% and 22.2%, respectively). Among blacks, participants primarily reported European, Caribbean, and African ethnicity. There may have been some confusion as to how to classify ethnicity, as approximately one-third of participants had no response or listed “Other.” Highest rates of adherence were again reported in those of European descent (37.5%). Lower rates were reported in those of African descent (28.2%), but the lowest rates were reported in those of Caribbean descent (13.2%) (Table 1).
Overall, rates of adherence were relatively low. This is due in part to our strict definition of 100% adherence, which was similar to the definition used by Kleeberger et al,3,4 but additionally required participants to report never skipping medications on the adherence questionnaire and to report following their medication schedule closely all the time. A lack of standardization of adherence makes it difficult to make direct comparisons with other studies. However, in a review of studies employing self-reported measures of adherence, the most common cutoff for optimal adherence was 100%, appearing in 44% of the studies.12 Even so, data collection methods and definitions vary and can greatly affect presentation of results.
The higher rates of nonadherence in black and Hispanics observed in our study have been found by others.3,5,13,14 However, few studies have investigated differences within these broad racial categories. We found that Hispanics of Central or South American descent had much lower rates of adherence than those of European descent. Also, both Hispanics and blacks from the Caribbean had very low rates of adherence. At the same time, self-reported racial and ethnic data is difficult to interpret as individuals often have complex family histories, which do not neatly fit into questionnaire response categories.
There were not enough participants or follow-up time in this study to do an appropriate analysis of factors affecting each ethnic subpopulation. Furthermore, more specific information on acculturation was not collected. Prior work has shown that an individual's degree of acculturation can strongly affect his/her health-related behaviors. This is true for both Hispanic and black populations.15,16 Although a question on native language was included in the MACS questionnaire, less than 12% of participants reported their native language. Regardless, proxy measures alone such as language and numbers of years spent in the United States may not be enough to determine degree of acculturation, especially in the black population.
One major strength of this study is that data were taken from a large prospective study, and over 4 years of follow-up data were used for the analysis. Multiple data points were recorded for a large number of participants, including Hispanic and black men. At the same time, because many of the participants have a long history of participation with the MACS, these individuals are more likely to be highly motivated and to have higher adherence to antiretroviral therapy. In addition, the MACS cohort does not include women and children. Thus, our findings might not be generalizable to all HIV-infected persons.
Another weakness of this study is that only self-reported data were used. Individuals are more likely to overreport adherence; thus, reported rate of adherence represents the maximum true rate. However, because the data collection was not done by their primary physicians, participants were less likely to overreport for social desirability. The significant association of viral load with adherence supports the reliability of self-report.
We conclude that adherence is multifactorial and varies significantly by race and ethnicity. We found that blacks and Hispanics are more likely to be nonadherent and that individuals from Central and South America and the Caribbean are especially at risk for nonadherence. Future adherence research and interventions should focus not solely on racial groups but also the ethnic differences within groups.
Data in this article were collected by the MACS with centers (principal investigators) at The Johns Hopkins University Bloomberg School of Public Health (Joseph B. Margolick and Lisa Jacobson), Howard Brown Health Center and Northwestern University Medical School (John Phair), University of California, Los Angeles (Roger Detels), and University of Pittsburgh (Charles Rinaldo).
1. Detels R, Munoz A, McFarlane G, et al. Effectiveness of potent antiretroviral therapy on time to AIDS and death in men with known HIV infection duration. JAMA
2. Hammer SM, Saag MS, Schechter M, et al. Treatment for adult HIV infection. 2006 recommendations of the International AIDS Society-USA Panel. JAMA
3. Kleeberger CA, Phair JP, Strathdee SA, et al. Determinants of heterogeneous adherence to HIV antiretroviral therapies in the Multicenter AIDS Cohort Study. J Acquir Immune Defic Syndr
4. Kleeberger CA, Buechner J, Palella F, et al. Changes in adherence to highly active antiretroviral therapy medications in the Multicenter AIDS Cohort Study. AIDS
5. Sullivan PS, Campsmith ML, Nakamura GV, et al. Patient and regimen characteristics associated with self-reported nonadherence to antiretroviral therapy. PLoS ONE
6. CDC. HIV/AIDS Surveillance Report, 2005
. Atlanta, GA: US Department of Health and Human Services; 2007.
7. Hall HI, Song R, Rhodes P, et al. Estimation of HIV incidence in the United States. JAMA
8. Chmiel J, Detels R, Kaslow RA, et al. Factors associated with prevalent human immunodeficiency virus (HIV) infection in the Multicenter AIDS Cohort Study. Am J Epidemiol
9. Detels R, Phair JP, Saah AJ, et al. Recent scientific contributions to understanding HIV/AIDS from the Multicenter AIDS Cohort Study. J Epidemiol
10. Kaslow RA, Ostrow DG, Detels R, et al. The Multicenter AIDS Cohort Study: rationale, organization, and selected characteristics of the participants. Am J Epidemiol
11. Press N, Tyndall MW, Wood E, et al. Virologic and immunologic response, clinical progression, and highly active antiretroviral therapy adherence. J Acquir Immune Defic Syndr
. 2002;31(Suppl 3):S112-S117.
12. Simoni JM, Kurth AE, Pearson CR, et al. Self-report measures of antiretroviral therapy adherence: a review with recommendations for HIV research and clinical management. AIDS Behav
13. Lazo M, Gange SJ, Wilson TE, et al. Patterns and predictors of changes in adherence to highly active antiretroviral therapy: longitudinal study of men and women. Clin Infect Dis
14. Singh N, Squier C, Skvik C, et al. Determinants of compliance with antiretroviral therapy in patient with human immunodeficiency virus: prospective assessment with implications for enhancing compliance. AIDS Care
15. Landrine H, Klonoff EA. Culture change and ethnic-minority health behavior: an operant theory of acculturation. J Behav Med
16. Lara M, Gamboa C, Kahramanian MI, et al. Acculturation and Latino health in the United States: a review of the literature and its sociopolitical context. Annu Rev Public Health