The Geography of HIV in Dallas County, Texas, 1999-2003

Abstract: 

HIV/AIDS has become a growing concern for major metropolitan areas in the United States such as Dallas, Texas. This research seeks to understand the spatial distribution of HIV in Dallas County zip codes and the factors associated with areas of high concentration. Zip code level reports of HIV/AIDS from the Texas Department of Health from 1999–2003 are used as the dependent variable. Education, ethnicity, and income data from the 2000 Census are used as explanatory variables with Spearman’s rank correlation analysis. The results suggest that race/ethnicity, level of education, and income are significant determinants of the HIV rate in Dallas County. Neighborhoods with a higher percentage of African-American or Hispanic residents, a high percentage of the population with educational attainment of ninth grade or less, or a lower median household income tend to have higher HIV rates. Because HIV/AIDS is not uniformly distributed through zip codes in Dallas County, establishing a clear perception of risk areas for HIV in a large metropolitan area will ensure effective and efficient HIV prevention interventions.

Table of Contents: 

    Introduction

    Beginning in the early 1980s, HIV/AIDS began to grow into a global pandemic. In 2003, about 4,802 Texans were diagnosed with HIV, and 3,689 were diagnosed with AIDS (Texas Department of Health, 2003a). In the United States, most of the HIV/AIDS cases are found in large metropolitan areas. Dallas County has the second highest HIV rate (37.8) and AIDS rate (29.2) per 100,000 people in Texas (Texas Department of Health, 2003b). HIV is a growing concern for major metropolitan areas such as Dallas, but the characteristics of the disease vary from one area to another. To be effective, interventions must address the unique characteristics of HIV/AIDS in particular locales.

    This research seeks to understand the spatial distribution of HIV in Dallas County zip codes and the factors associated with areas of high prevalence. Analysis of the spatial pattern of HIV will provide insights for planning intervention and control programs by targeting communities where individuals have a greater risk of contracting HIV. Another goal of the study is to examine whether the characteristics of HIV nationwide are similar to the characteristics of HIV in Dallas County.

    Literature Review

    Human Immunodeficiency Virus (HIV) attacks and kills white blood cells known as CD4 + T-lymphocytes, leaving the body’s immune system defenseless against infections and illnesses. In time, HIV progresses to an advanced stage called Acquired Immune Deficiency Syndrome (AIDS). AIDS usually takes about 2–10 years to develop and is diagnosed once a patient has a CD4 + T-lymphocyte count of less than 200 (University of California at San Francisco, 2004). HIV is not a genetic phenomenon. It is transferred through contact of blood or bodily fluids directly from one person to another. Common modes of transmission of HIV include:

    1. Male-male sex
    2. I.V. drug use
    3. Heterosexual sex
    4. Pediatric (from mother to infant), and
    5. Blood transfusions and other.

    Studies show HIV rates vary among ethnic groups. According to the U.S. Centers for Disease Control and Prevention (2005), the HIV rate (per 100,000) in 2004 was 79.6 for African Americans, 29.5 for Hispanics, and 9.0 for Whites. Similar HIV rates were reported in Texas according to the Texas State Department of Health Services (2003b). The rate of reported HIV infections per 100,000 in 2003 for African Americans was 79.2, more than five times higher than both the HIV rate of Whites at 14.9 and the rate of Hispanics at 14.1.

    Initially, when the HIV/AIDS pandemic was emerging, many studies found a direct positive relationship between level of education and HIV/AIDS rates. Ironically, this means that as a person’s education increased, so did their chances of contracting HIV/AIDS. Since then, studies have shown that the HIV/AIDS pandemic is changing. An inverse relationship now exists between education and HIV/AIDS prevalence ( Columbia University, n.d.). Those who are illiterate and uneducated are at a greater risk of acquiring HIV/AIDS.

    A powerful indicator of the quality of life in a particular neighborhood is its income. Disadvantaged neighborhoods are commonly associated with low income, low education, high crime, and higher drug use. Galea, Ahern, and Vlahov (2003) found that neighborhood income played an important role in the amount of drug use in each neighborhood. Low-income neighborhoods are sites of increased drug use, and a drug user living in a disadvantaged neighborhood commonly uses more drugs than a drug user who lives in a more advantaged neighborhood. Stresses that come along with the disadvantaged neighborhood might account for increased high-risk behavior and increased drug use. Disadvantaged neighborhoods usually have less support and help networks to help those in need (Galea et al., 2003).

    Methodology

    The Texas State Department of Health Services provided zip-code-level cumulative counts of reported cases of HIV/AIDS from 1999 to 2003 to Dr. Joseph Oppong. Reported HIV/AIDS per 100,000 people in the population is used as the dependent variable. Education, ethnicity, and income data from the 2000 Census are used as explanatory variables with Spearman’s rank correlation analysis. Zip codes of residences at diagnosis were reported in the data for both HIV and AIDS; however, this study selected only HIV zip codes/data. Of the 4,897 total HIV cases in Dallas County, 252 cases were not used due to incorrectly entered zip codes, leaving 4,325 cases analyzed. The HIV prevalence rate is defined by the number of HIV cases per zip code region divided by the population and multiplied by 100,000. Ethnicity data were defined as the percentage of each of the major race/ethnicity groups in each zip code, i.e., percent White, percent Black, and percent Hispanic. Education was defined as the percentage of the population over age 25 with education of ninth grade or less. Neighborhood income was conceptualized by using median household income. SPSS 12.0 was used for statistical analysis and ArcGIS was used for mapping.

    Proposed Hypotheses

    Three main hypotheses are examined in this study:

    Hypothesis 1: Race is a predictor of HIV prevalence. Neighborhoods/zip codes with a higher percentage of African Americans will have higher HIV rates. In contrast, neighborhoods with a high percentage of Whites will have a lower HIV prevalence rate.

    Hypothesis 2: Income, defined by median household income, will have a negative relationship with a neighborhood’s HIV prevalence rate.

    Hypothesis 3: Education is crucial to the prevention of HIV spread. Education prior to exposure to HIV is very important. Therefore, a direct relationship will exist between the percentage of the population in a zip code above the age of 25 with an education of ninth grade or less and HIV prevalence.

    Results

    HIV in Dallas County

    Before exploring the spatial distribution of HIV in Dallas County, it is important to understand the descriptive statistics of HIV for the United States and Dallas County. Nationally, HIV disproportionately affects males, African Americans, and men who have sex with men (Centers for Disease Control and Prevention [CDC], 2005, 2006). The data in Table 1, Table 2, and Table 3 and Figure 1 suggest that HIV in Dallas County disproportionately affects African-American males, patients 33–37 years old, and men who have sex with men.

    The Geography of HIV in Dallas County

    The spatial distribution of HIV in Dallas County is shown in Figure 2. The highest HIV prevalence rates are found clustered in central Dallas County, with a ring of moderately high rates radiating from the center, particularly in the northwest and southeastern portion of the county.

    Race/Ethnicity and HIV Prevalence Rate

    It was hypothesized that zip codes in Dallas County with a greater percentage of African Americans would have higher HIV rates. Table 4 shows a strong positive relationship between the HIV rate and the percentage of African Americans living in the neighborhood. This means that, as the percentage of African Americans increases, a neighborhood’s HIV rate will also increase. Comparing Figure 2 and Figure 3 helps to illustrate the relationship between the percentage of African Americans and HIV rate, which is statistically significant at less than 0.01. A fairly strong and significant positive relationship also exists between the percentage of Hispanics in a neighborhood and the neighborhood’s HIV rate. On the contrary, a significant negative relationship exists for neighborhoods with a high percentage of White residents and HIV rates. This positive relationship suggests that as the percentage of White inhabitants increases, the HIV rate for that zip code decreases.

    Income and HIV Prevalence

    As hypothesized, neighborhood income has a strong negative relationship with the HIV prevalence rate. Therefore, as a neighborhood’s median household income decreases, the neighborhood’s HIV prevalence rate increases. Statistically, this relationship was significant at the 0.01 level.

    Education

    Education less than ninth grade was hypothesized to have a positive relationship with a neighborhood’s HIV rate. Statistically, this relationship was significant for Dallas County neighborhoods. Many of the neighborhoods with high HIV rates also have a high percentage of the population of adults over age 25 with an education of ninth grade or less. Refer to Figure 4.

    Conclusion

    HIV in Dallas County appears to be strongly influenced by race, income, and education. The correlations presented in this study not only point to general trends in the spatial distribution of HIV in Dallas County, but also show how nationwide risk markers affect the local variation of HIV in Dallas County. Understanding the local variation of HIV allows public officials and local authorities to produce Dallas-County-specific HIV intervention programs, which will efficiently target communities where inhabitants are at a greater risk of HIV. One limitation of this research, as noted earlier, is that it did not include AIDS. Studying AIDS can show similarities and/or differences between the distribution of HIV and AIDS. Further study may include an analysis of the spatial distribution of HIV in other Texas metropolitans, allowing for the comparison of local variations of HIV throughout Texas and an understanding of how nationwide risk markers affect a demographically different metropolitan area such as San Antonio, which has a higher Hispanic population. Obtaining zip code + 4-level HIV data would provide a more detailed study into the spatial distribution of HIV in an urban area.

    References

    • Centers for Disease Control and Prevention. (2006). A glance at the HIV/AIDS epidemic. Retrieved March 27, 2006, from http://www.cdc.gov/hiv/resources/factsheets/PDF/At-A-Glance.pdf
    • Centers for Disease Control and Prevention. (2005). HIV/AIDS Surveillance Report, 2004. Vol. 16. Atlanta: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, 14. Retrieved March 27, 2006, from http://www.cdc.gov/hiv/topics/surveillance/resources/reports/2004report/
    • Columbia University. (n.d.). The “education vaccine” against HIV. Retrieved November 22, 2005.
    • Galea, S., Ahern, J., & Vlahov, D. (2003). Contextual determinants of drug use risk behavior: A theoretical framework. Journal of Urban Health: Bulletin of the New York Academy of Medicine, 80.
    • Texas Department of Health. (2003a). Texas HIV/STD cases and rates (cases per 100,000) by year of report. Retrieved November 22, 2005.
    • Texas Department of Health. (2003b). Summary of HIV/AIDS in Texas. Retrieved November 22, 2005.
    • niversity of California at San Francisco. (2004, September). HIV INSITE: What are HIV and AIDS? Retrieved November 22, 2005, from http://hivinsite.ucsf.edu/hiv?page=basics-00-01.

    Table 1: HIV Cases by Sex

      Frequency Percent Valid Percent Comulative Percent
    Valid Male 3697 78.1 78.1 78.1
    Female 1038 21.9 21.9 100.0
    Total 4735 100.0 100.0  

     

    Table 2: HIV Cases by Ethnicity

      Frequency Percent Valid Percent Cumulative Percent
    Valid White 1962 41.4 41.4 41.4
    Black 1945 41.1 41.1 41.1
    Hispanic 754 15.9 15.9 98.4
    Other 74 1.6 1.6 100.0
    Total 4735 100.0 100.0  

     

    Table 3: Age at HIV Diagnosis

      N Minimum Maximum Mean Std. Deviation
    Age At HIV Dx 4735 0 99 35.59 9.873
    Valid N (listwise) 4735        

     

    Table 4: HIV Rate, Income, Education, and Race/Ethnicity Correlations

    Correlations

      HIV_RATE INCOME EDUCATION %White %Black %Hispanic
    Spearman's rho HIV_RATE Correlation Coefficient
    Sig. (2-tailed)
    N
    1.000
    .
    96
    -.533**
    .000
    96
    .453**
    .000
    96
    -.515**
    .000
    96
    .503**
    .000
    96
    .395**
    .000
    96
    INCOME Correlation Coefficient
    Sig. (2-tailed)
    N

    -.533**
    .000
    96

    1.000
    .
    96
    -.693**
    .000
    96
    .869**
    .000
    96
    -.527**
    .000
    96
    -.563**
    .000
    96
    EDUCATION Correlation Coefficient
    Sig. (2-tailed)
    N
    .453**
    .000
    96
    -.693**
    .000
    96
    1.000
    .
    96
    -.632**
    .000
    96
    .355**
    .000
    96
    .902**
    .000
    96
    %White Correlation Coefficient
    Sig. (2-tailed)
    N
    -.515
    .000
    96
    .869**
    .000
    96
    -.632**
    .000
    96
    1.000
    .
    96
    -.634**
    .000
    96
    -.547**
    .000
    96
    %Black Correlation Coefficient
    Sig. (2-tailed)
    N
    .503**
    .000
    96

    -.527**
    .000
    96

    .355**
    .000
    96
    -.634**
    .000
    96
    1.000
    .
    96
    .247*
    .015
    96
    %Hispanic Correlation Coefficient
    Sig. (2-tailed)
    N
    .395**
    .000
    96
    -.563**
    .000
    96
    .902**
    .000
    96
    -.547**
    .000
    96

    .247**
    .015
    96

    1.000
    .
    96

    **. Correlation is significant at the .01 level (2-tailed).
    *. Correlation is significant at the .05 level (2-tailed).

    Figure 1: Mode of Transmission of HIV

    Figure 2: HIV Prevalence Rate for All Races, Dallas County, Texas, 1999-2003.

    Figure 3: Percent of African Americans Living in Dallas County by Zip Code, 2000

    Figure 4: Population with Education of Ninth Grade or Less in Dallas County, 2000.