| |
Characteristics of HIV Infection Among Hispanics, United States 2003-2006
[Epidemiology and Social Science]
|
| |
| |
JAIDS Journal of Acquired Immune Deficiency Syndromes:Volume 49(1)1 September 2008pp 94-101
Espinoza, Lorena DDS, MPH; Hall, H Irene PhD; Selik, Richard M MD; Hu, Xiaohong MS
From the Division of HIV/AIDS Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention, Centers for Disease Control and Prevention, Atlanta, GA.
"In summary, risk factors, diagnosis trends, HIV-to-AIDS intervals, and survival of Hispanics with a diagnosis of HIV infection varied by place of birth. These findings suggest that conceptualizing Hispanics as a homogeneous group may be inappropriate. HIV prevention efforts may not be equally effective for US-born and foreign-born Hispanics; thus, educational efforts should address the important cultural and behavioral differences among Hispanic subgroups. Moreover, the Hispanic population is expected to triple between 2000 and 2050, so it will be necessary to continue to address the impact of HIV on this population....
....The transmission categories for HIV infection in US-born Hispanics were distributed as follows: male-to-male sexual contact (50%), followed by high-risk heterosexual contact (26%), IDU (18%), and male-to-male sexual contact and IDU (4%). Among persons born in Puerto Rico, larger proportions had been infected through high-risk heterosexual contact (40%) and IDU (33%) than through male-to-male sexual contact (23%)......By transmission category, there was a nonsignificant decrease in the number of diagnoses among Hispanic IDU (males and females) but a significant increase among Hispanic males exposed through male-to-male sexual contact (5.3%; 95% CI = 2.3 to 8.4)...... Significant changes occurred among Hispanic males aged 20-29 years: the EAPCs (estimated annual percent change) increased (5.5%; 95% CI = 1.8 to 9.4).....In 2005, a total of 7561 cases of HIV infection were diagnosed among Hispanic adults and adolescents in the 33 states and dependent areas; of these, 42% progressed to AIDS in less than 12 months."
Abstract
Background: Hispanic subgroups of varied national origin differ culturally; overall, Hispanics in the United States are disproportionately affected by HIV infection.
Methods: We analyzed cases of HIV infection that were diagnosed among Hispanics in 33 states and US-dependent areas during 2003-2006 and reported to the Centers for Disease Control and Prevention through June 2007. We used Poisson regression to calculate the estimated annual percent change in the number and rate of HIV diagnoses and used logistic regression to analyze the association between birthplace and a short (<12 months) HIV-to-AIDS interval.
Results: HIV infection was diagnosed among 30,415 Hispanics. Of 24,313 with reported birthplace, 61% were born outside the continental United States. The annual number of diagnoses increased among Mexican-born males [estimated annual percent change = 8.8%; 95% confidence interval (CI) = 3.5 to 14.5] and Central American-born males (18.6%; 95% CI = 9.4 to 28.6) and females (24.6%; 95% CI = 8.8 to 42.7) but decreased among US-born Hispanic females (-8.2%; 95% CI = -13.3 to -2.8). A short HIV-to-AIDS interval was more common among Mexican-born Hispanics than among US-born Hispanics.
Discussion: Diagnosis trends and HIV-to-AIDS intervals varied by place of birth. To decrease the incidence of HIV infection among Hispanics, prevention programs need to address cultural differences.
In the United States, Hispanics are disproportionately affected by HIV infection. Although Hispanics accounted for 15% of the US population in the 2006 census estimates,1 19% of US residents with AIDS that year were Hispanic.2 Although the highest HIV diagnosis rates in 2006 (in 33 states) were those for non-Hispanic blacks, the second highest rates were for Hispanics: 51 HIV diagnoses per 100,000 Hispanic men (3 times the rate for non-Hispanic white men) and 15 HIV diagnoses per 100,000 Hispanic women (5 times the rate for non-Hispanic white women).2
US Hispanic subgroups of varied national origin differ from one another culturally. Likewise, HIV knowledge,3,4 behavioral risk factors,5 perceptions of risk,4,6 and use of prevention services7 differ between foreign-born and US-born Hispanic populations. For example, Hispanics born in Puerto Rico are more likely to contract HIV through injection drug use (IDU) than are other foreign-born or US-born Hispanics.2,7 Using birthplace data can help identify subpopulations that differ in HIV testing,6,8 use of health care services,9 or survival. The importance of these differences is underscored by the large proportion of US Hispanics who are foreign born. According to the 2000 US Census, 35 million Hispanics were residing in the United States and 14 million (40%) of them were foreign born.10 Foreign-born Hispanics are expected to make up 25% of the increase in the Hispanic population from 2000 to 2010.11 The largest Hispanic subgroup was Mexican (64%), followed by Puerto Rican (9%) and Cuban (3.5%).1 Few studies have reported the epidemiology of HIV infection by using birthplace data.12,13
We examined the characteristics of Hispanics with a diagnosis of HIV infection by analyzing, by place of birth, (1) recent trends in HIV diagnosis, (2) the association between place of birth and a short interval (<12 months) between diagnoses of HIV infection and AIDS, and (3) survival after diagnosis of AIDS.
DISCUSSION
Over half of Hispanics with reported birthplace and diagnosed with HIV infection in the 33 states and dependent areas were born outside the continental United States. Although we cannot say definitively where these Hispanics became infected, some evidence suggests that most were infected in areas of the United States where the prevalence of HIV infection was higher.13 Hispanics have a history of migrating to the United States for work in agricultural industries. Migration patterns may contribute to the risk for HIV infection, perhaps because change in residence can result in homelessness, loneliness, isolation, low ratios of women to men, and financial instability. In turn, these factors can result in new sex partners, drug use, and inadequate access to health care services.20
Our finding that a larger proportion of HIV-infected Hispanics born in Puerto Rico, compared with US-born and other foreign-born Hispanics, had been infected through IDU supports the finding that a larger proportion of Puerto Ricans contracted infection through IDU.2,21 Even within the IDU category, differences in behavior associated with drug use could contribute to higher risk for HIV infection. Thus, the finding that a larger proportion of Puerto Ricans, compared with US-born Hispanics who inject drugs, are infected could be due to not only a higher prevalence of IDU but also higher levels of risky behaviors related to IDU (eg, sharing syringes and other drug paraphernalia).22-24 It may also reflect the fact that services to reduce HIV risk behaviors (eg, needle exchange programs and methadone treatment) are used more often in the United States than in Puerto Rico.25
The lack of changes in the rate of diagnosis of HIV infection among Hispanics may be due to no significant increases or decreases in incidence of HIV infection or could also be due to a reduction in HIV testing. The increase in the number of diagnoses of HIV infection among Hispanic MSM may suggest a resurgent epidemic among MSM-a finding in other studies of MSM.26,27 Improved treatments for HIV disease may have led to changes in belief patterns regarding disease severity28 and consequently to increases in the prevalence of high-risk sexual behavior among MSM.29,30 Our finding of decreases in HIV diagnosis among injection drug users is consistent with other reports of reduced HIV infection incidence among injection drug users.31,32
A short HIV-to-AIDS interval apparently indicates a delay in testing until late in the course of HIV disease, when symptoms are likely to have developed. It also may reflect inadequate care and treatment. A short interval increased with age, which may be explained by the fact that HIV disease progression tends to occur more rapidly among older persons. Another possible explanation is that older persons are assumed to be not at risk and therefore are not the focus of testing programs. A short interval has been shown to be more common among Hispanics than persons of other races/ethnicities.33,34 The proportion of persons with a short HIV-to-AIDS interval varied by place of birth. A short interval was more common among Hispanics born in Mexico than among US-born Hispanics and may reflect differing testing behaviors among Hispanic subgroups. Persons born in Puerto Rico were more likely than those born in Mexico to report having been tested or planning to be tested.8,35 Migrant and seasonal workers who are undocumented are less likely than documented workers to have been tested for HIV infection.36 Historically, Hispanics, compared with non-Hispanic whites, have had less access to general medical treatment and prevention services.37 The barriers of language, lack of knowledge, and lack of health insurance have been obstacles to care and may be more prevalent among foreign-born Hispanics.38,39
We found that rates of survival after the diagnosis of AIDS differed among Hispanic subgroups. Few studies have examined survival differences among Hispanic subgroups,40 although studies have reported that the survival of Hispanics is similar to that of non-Hispanic whites.41-44 This finding is surprising because Hispanics are more likely to be uninsured and those who are uninsured are more likely to delay or not receive medical care because of cost.45,46 An explanation of this apparent paradox may be differential underreporting on death certificates (eg, foreign-born Hispanics with HIV infection may return to their country of birth).47 Subgroup differences in such return migration may explain the apparently lower survival rate among persons born in Cuba compared with those born in Mexico. However, other research has found shorter survival time among HIV-infected Hispanics compared with HIV-infected non-Hispanic whites.48 This difference may be attributed to the inconsistencies between census information about Hispanic origin and the information recorded on death certificates.49 Another possible contribution to apparent differences in survival is the completeness of ascertainment of death data by the HIV/AIDS surveillance programs of state health departments-data on which Centers for Disease Control and Prevention depends.
Our data are subject to at least 5 limitations. First, HIV infection data from some states are not included in the national surveillance system. Although our data are from the largest set of population-based data on persons with HIV infection, the 33 states and dependent areas used in this analysis may not be nationally representative, as they reported only 65% of all AIDS cases diagnosed among adults and adolescents in the United States during 2003-2006 and 68% of AIDS cases among Hispanics. The exclusion of data from some states with high AIDS morbidity and a large Hispanic population (eg, California) may result in an underrepresentation of cases among Hispanics. In addition, Hispanic immigrants, who may not access health care services because of cost or fear of deportation, may be underreported. Second, the assumptions on which we based the reclassification of the transmission category of cases reported without risk factors may no longer be valid: these assumptions are being reevaluated. However, limited studies have indicated an underestimation of the number of cases attributed to high-risk heterosexual contact.50,51 Third, detailed information about the HIV-related behavioral risk factors of sex partners of HIV-infected persons is limited, reducing the usefulness of the surveillance data in evaluating the effect of sexual behavior on HIV transmission. The correlation between high-risk behavior and acculturation among Hispanics cannot be assessed by our surveillance data. Other studies have suggested that a higher level of acculturation is associated with an increase in HIV knowledge,3,52,53 a higher prevalence of behavioral risk factors for HIV infection,4,54 and a greater ability to obtain health care services.55 In addition, ancestry of US-born Hispanics is not collected, thus differences among subgroups cannot be evaluated. Fourth, the misclassification of Hispanics as members of other races/ethnicities may have resulted in an underestimation or overestimation of the number of Hispanics and Hispanic subgroups. However, in a validation study of race/ethnicity and transmission mode in the HIV/AIDS Reporting System, self-reported race/ethnicity agreed well.56 Last, in the HIV/AIDS surveillance system, birthplace information is commonly missing (true of 20% of the records in this analysis).
In summary, risk factors, diagnosis trends, HIV-to-AIDS intervals, and survival of Hispanics with a diagnosis of HIV infection varied by place of birth. These findings suggest that conceptualizing Hispanics as a homogeneous group may be inappropriate. HIV prevention efforts may not be equally effective for US-born and foreign-born Hispanics; thus, educational efforts should address the important cultural and behavioral differences among Hispanic subgroups. Moreover, the Hispanic population is expected to triple between 2000 and 2050, so it will be necessary to continue to address the impact of HIV on this population.
METHODS
We analyzed cases of HIV infection that were diagnosed among Hispanics during 2003-2006 and reported to the Centers for Disease Control and Prevention through June 2007 by state and local health departments. Data have been available from 33 states (Alabama, Alaska, Arizona, Arkansas, Colorado, Florida, Idaho, Indiana, Iowa, Kansas, Louisiana, Michigan, Minnesota, Mississippi, Missouri, Nebraska, Nevada, New Jersey, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, South Carolina, South Dakota, Tennessee, Texas, Utah, Virginia, West Virginia, Wisconsin, and Wyoming) and 5 US-dependent areas (American Samoa, Guam, Northern Mariana Islands, Puerto Rico, and US Virgin Islands) that had been reporting HIV diagnoses since at least 2003-long enough for data collection to stabilize and for adjustment of the data to monitor trends. The 33 states represent 57% of Hispanics in the United States with the exclusion of the states (California and Illinois) with a high population of Hispanics.
The categorization of race/ethnicity is generally based on information in medical records, but that information is recorded by health care providers who use unspecified criteria. We classified the place of birth of Hispanics as one of the following: the United States (limited to the 50 states and the District of Columbia), Puerto Rico (an US-dependent area), Mexico, Cuba, Central America, South America, other countries, and unknown. We used the US Census definition of Hispanic origin as those who indicated that their origin was Mexican, Puerto Rican, Cuban, Central American, South American, or some other Hispanic origin.1 We defined US-born Hispanics as those whose place of birth was in the continental United States (excludes US-dependent areas) or unknown. Hispanics can be of any race.
Demographic Characteristics of Hispanics With HIV Infection
Using data from the 33 states and dependent areas, we examined sex, age group, transmission category, and year of diagnosis for HIV-infected Hispanics by place of birth. The HIV transmission categories were (1) male-to-male sexual contact (regardless of whether the men also had sex with women), (2) IDU, (3) male-to-male sexual contact and IDU, (4) high-risk heterosexual contact (with a sex partner known to have, or to be at high risk for, HIV infection, eg, a man who has sex with men or an injection drug user), (5) all other specified HIV risk factors (eg, receipt of HIV-contaminated blood transfusion, blood product, or tissue), and (6) no HIV risk factor identified.
We adjusted the number of diagnoses for expected reporting delay (time between diagnosis and report).14,15 We also adjusted the distribution of diagnoses by transmission category for missing risk factor information based on historical patterns of cases originally reported with no risk factor that were later found to have risk factors and were reclassified into the corresponding transmission categories.16,17
Trends in HIV Infection Diagnoses
We analyzed trends in the diagnosis of HIV infection among Hispanics whose diagnosis was made during 2003-2006 and who resided in the 33 states and dependent areas. We calculated the annual rates of HIV diagnosis per 100,000 population for adults and adolescents (age 13 years) by sex and age group in the 33 states. We used Poisson regression to calculate the estimated annual percent change (EAPC) in the annual number of diagnoses, by sex, age group, place of birth, and transmission category, from 2003 through 2006.18 The significance of a trend was determined by whether the 95% confidence interval (CI) for the EAPC included 0.
Determinants of a Short HIV-to-AIDS Interval
We defined a short HIV-to-AIDS interval as the diagnosis of AIDS less than 12 months after the diagnosis of HIV infection. We conducted logistic regression analysis to examine associations between a short HIV-to-AIDS interval and characteristics (sex, age, place of birth, and transmission category) of Hispanic adults and adolescents whose diagnosis of HIV infection was made during 2005 in the 33 states and dependent areas. Differences between groups in the proportion of cases with a short interval were considered significant if the 95% CI for the adjusted odds ratio did not include 1. Cases in persons whose month of diagnosis of HIV infection was unknown (n = 11) were excluded from this analysis. Cases in children (n = 29) were also excluded because the number was too small for the derivation of statistically stable estimates.
Survival Time After AIDS Diagnosis
We analyzed survival time after diagnosis of AIDS among Hispanic adults and adolescents whose diagnosis of AIDS was made during 1996-2003 in the 50 states, the District of Columbia, and 5 US-dependent areas and who were reported to Centers for Disease Control and Prevention through June 2007. We used data on deaths that occurred during 1996-2006 and were reported by June 2007. We used the Kaplan-Meier method to estimate the proportions of Hispanics who survived more than 12 months and the proportions who survived more than 36 months after diagnosis.19 We adjusted the data for sex, age group, place of birth, transmission category, and CD4 count at the time of diagnosis. We did not adjust for reporting delays or for unknown risk factors.
RESULTS
Demographic Characteristics of Hispanics With HIV Infection
An estimated 30,415 cases of HIV infection were diagnosed for Hispanics during 2003-2006 in the 33 states and dependent areas, accounting for 21% of all cases (including cases in non-Hispanics) during this period in these states and dependent areas. Of 24,313 Hispanics with reported birthplace, over half (61%) were born outside the continental United States (Table 1). The transmission categories for HIV infection in US-born Hispanics were distributed as follows: male-to-male sexual contact (50%), followed by high-risk heterosexual contact (26%), IDU (18%), and male-to-male sexual contact and IDU (4%). Among persons born in Puerto Rico, larger proportions had been infected through high-risk heterosexual contact (40%) and IDU (33%) than through male-to-male sexual contact (23%).
Trends in HIV Infection Diagnoses
The EAPC in the annual number of diagnoses of HIV infection during 2003-2006 in the 33 states and dependent areas remained stable among Hispanic females (-1.4%; 95% CI = -4.6 to 2.0) and among Hispanic males (1.4%; 95% CI = -0.5 to 3.2) (Table 2). Significant changes occurred among Hispanic males aged 20-29 years: the EAPCs increased (5.5%; 95% CI = 1.8 to 9.4). The annual number of diagnoses increased significantly among Hispanic males born in Mexico (8.8%; 95% CI = 3.5 to 14.5) and Hispanic males (18.6%; 95% CI = 9.4 to 28.6) and Hispanic females (24.6%; 95% CI = 8.8 to 42.7) born in Central America. In contrast, the annual number of HIV diagnoses decreased significantly among US-born Hispanic females (-8.2%; 95% CI = -13.3 to -2.8). By transmission category, there was a nonsignificant decrease in the number of diagnoses among Hispanic IDU (males and females) but a significant increase among Hispanic males exposed through male-to-male sexual contact (5.3%; 95% CI = 2.3 to 8.4).
During 2003-2006 in the 33 states, the annual rate of HIV infection diagnosis per 100,000 in Hispanics decreased significantly from 37.0 in 2003 to 33.7 in 2006. The annual EAPC decreased overall (-3.2%; 95% CI = -5.6 to -0.7) and was significant for males (-2.8%; 95% CI = -5.3 to -0.1) and females (-5.8%; 95% CI = -9.7 to -1.7). The average annual rates decreased significantly among Hispanic males aged 30-59 years and Hispanic females aged 30-49 years (data not shown).
Determinants of a Short HIV-to-AIDS Interval
In 2005, a total of 7561 cases of HIV infection were diagnosed among Hispanic adults and adolescents in the 33 states and dependent areas; of these, 42% progressed to AIDS in less than 12 months. After adjustment for covariates, a short interval was significantly more common among Hispanic males than among Hispanic females and increased with age. Likewise, a short interval was more common among Hispanics born in Mexico and Central America than among Hispanics born in the United States (Table 3). A short interval was more common among Hispanics in the IDU and high-risk heterosexual contact categories than among Hispanic men who have sex with men (MSM).
Survival Time After AIDS Diagnosis
During 1996-2003, a total of 69,484 cases of AIDS were diagnosed among Hispanic adults and adolescents in the 50 states, the District of Columbia, and dependent areas-20% of all AIDS cases diagnosed during this period. Of the Hispanic males and females who survived more than 36 months, a significantly larger proportion were males (Table 4). The proportion who survived more than 36 months decreased with increasing age at the time of diagnosis. After 36 months, the smallest proportion of surviving Hispanics had been born in Puerto Rico, followed by those born in Cuba. The largest proportion surviving were those born in South America and those whose place of birth was unknown.
|
|
| |
| |
|
|
|
