Association between area- level poverty and HIV diagnoses, and differences by sex, New York City Ellen Wiewel, HIV Epidemiology & Field Services Program, New York City Department of Health & Mental Hygiene APHA abstract #296498, Session , “Structural and Institutional Issues in HIV/AIDS,” Tuesday, November 18, 2014
Presenter Disclosures The following personal financial relationships with commercial interests relevant to this presentation existed during the past 12 months: Ellen Wiewel No relationships to disclose
Place & social conditions matter Most investigations of determinants of HIV infection focus on individual traits (risk, race) Social conditions (e.g., neighborhood poverty) contribute to health disparities of many kinds, including HIV infection
Gap in research on HIV and neighborhood poverty Link between HIV and neighborhood poverty has not been explored by ZIP code in a US city and could inform local policy Are differences by neighborhood poverty level explained by differences in age and racial/ethnic composition?
Objectives Assess the association of neighborhood poverty with HIV diagnosis rates, after controlling for other neighborhood factors such as age and racial/ethnic composition Determine whether neighborhood poverty has different effects on HIV rates by sex
Methods (1) Ecological analysis Outcome: ZIP code-level HIV diagnosis rates among New York City residents 13+ years old, by sex
Methods (2) Sex-stratified negative binomial regression models measured effect of neighborhood-level poverty on diagnosis rates Covariates: neighborhood-level education, racial/ethnic composition, age distribution, and percent men who have sex with men
Data Sources Data source NYC HIV surveillance registry NYC Community Health Survey American Community Survey US Census 2010 Population All NYC HIV cases Representative sample of NYC adults Representative sample of US population Every US resident Measures HIV diagnosesPercent MSM Poverty Education Sex Race/Ethnicity Age
Text Explanation of Data Sources Table Table of four data sources and the measures gleaned from each Data sources include NYC HIV surveillance registry, NYC Community Health Survey, American Community Survey, and US Census
Neighborhood Poverty Definition Percent in a ZIP with incomes below the federal poverty threshold Categories: 0-<10% (“low-poverty”) 10-<20% (“medium-poverty”) 20-<30% (“high-poverty”) % (“very-high-poverty ZIPs”)
Results (1) 180 residential NYC ZIP codes included 6,184 persons newly diagnosed with HIV in NYC in : 4,754 males and 1,430 females Overall annual HIV diagnosis rate: 44.8 per 100,000
Results (2) ZIP-level diagnosis rates in NYC ranged from 0.0 to per 100,000 (median: 34.8 per 100,000) Median diagnosis rates by sex: – 56.2 per 100,000 males – 10.4 per 100,000 females
Figure 1b. Male HIV diagnosis ratesFigure 1c. Female HIV diagnosis rates Mapping Poverty and HIV in NYC
Text Explanation of Maps Three NYC ZIP code maps showing poverty rates and male and female HIV diagnosis rates High- and very-high-poverty ZIP codes appear to have higher male and female HIV diagnosis rates
Box Plots of Annualized HIV Diagnosis Rates per 100,000 Persons, by Neighborhood Poverty, NYC Males Females
Text Explanation of Box Plots Two box plots of male and female HIV diagnosis rates by NYC ZIP code Males and females have higher HIV diagnosis rates as neighborhood poverty increases
Regression of Neighborhood Poverty and HIV Diagnosis Rates Among Males and Females in 180 NYC ZIP Codes,
Text Explanation of Regression Table Table of crude and adjusted rate ratios (RRs) from Poisson regression of ZIP code-level HIV diagnosis rates among males and females by neighborhood poverty level Crude and adjusted RRs increase with increasing poverty, and increase more among females than males
Summary of Findings (1) HIV diagnosis rates among males and females increased with increasing neighborhood poverty At all neighborhood poverty levels, higher diagnosis rates among males than females
Summary of Findings (2) Living in very-high- vs. low-poverty neighborhoods was associated with an increase in HIV diagnosis rates of 63% for males and 114% for females Relative difference in rates between neighborhoods with very high vs. low poverty particularly large for women
Limitations ZIP code is larger area than is optimal for detection of disparities (Census tract and Census block) Diagnosis rates rather than incidence rates (only diagnoses are reportable) Explained neither how poverty influences HIV nor why relative impact of poverty is larger on females
Strengths First ecological analysis of HIV diagnosis rates and poverty at the ZIP code level and in a US city One of few analyses to assess sex differences in the poverty-HIV relationship
Conclusions Area-based poverty was associated with HIV diagnoses in NYC, supporting interventions making structural and social changes Other structural urban HIV prevention should focus on higher-poverty areas Prevention strategies addressing poverty are important for males and (especially) females
Next Steps NYC health department already focuses HIV prevention, care, and treatment on higher-poverty areas Support other local place-based approaches to HIV prevention Future research: poverty’s influence on care continuum, e.g., linkage to care and viral suppression
Acknowledgments Coauthors Angelica Bocour, Laura Kersanske, Sara Bodach, Qiang Xia, and Sarah Braunstein (all with the New York City Department of Health & Mental Hygiene at the time of this analysis) NYC Department of Health & CUNY SPH Dean Dr. Ayman El-Mohandes for travel stipends
Contact Ellen Wiewel
Appendix. Distribution of model outcomes, covariates, and poverty level among New York City ZIP codes (N=180), for persons 13+ years old