Measures of prison release, inequality and poverty were independently associated with higher HIV prevalence in a study of 12 U.S. cities presented at IDWeek 2016. Every one-unit increase in the prison-release rate translated into a 4% higher HIV prevalence.
In the United States, HIV infection and incarceration disproportionately involve blacks. Researchers at Boston's Brigham and Women's Hospital who conducted this study noted that HIV prevalence stands about five times higher among incarcerated people than those not in prison. Prison release can affect neighborhood-level HIV prevalence directly if released HIV-positive people are not linked to care and treatment, and indirectly when it increases inequality and promotes resumption of HIV risk behaviors.
The Boston team undertook this study to determine how prison release and associated economic factors affect HIV prevalence in 12 U.S. cities: Tampa (Florida), Columbia (South Carolina), Orlando (Florida), Detroit (Michigan), Jacksonville (Florida), Houston (Texas), Baton Rouge (Louisiana), Atlanta (Georgia), Miami (Florida), Philadelphia (Pennsylvania), New Orleans (Louisiana) and New York City. The researchers determined HIV prevalence per 100,000 people by reviewing 2012 ZIP code-level HIV surveillance data in 1,002 ZIP codes in the study cities. Besides 2008 prison-release rate, other variables in the analysis included education, race/ethnicity, distance of each ZIP code from the city center, poverty and inequality measured by the GINI index, which gauges social inequality on a fractional scale from 0 (perfect equality) to 1 (greatest inequality).
HIV prevalence in 2012 ranged from a low of 0.42 per 100,000 people in Tampa to a high of 1.5 per 100,000 in both New Orleans and New York City. All these rates were higher than the overall U.S. prevalence of 0.39 per 100,000. Among men, HIV prevalence ranged from a low 0.64 (Tampa) to a high of 2.30 per 100,000 (New York City) in the 12 U.S. cities, and among women from a low of 0.26 (Tampa) to a high of 0.82 (New Orleans).
The prison release rate averaged 4.4 per 1000 adults in the 12 U.S. cities, while the GINI index averaged 0.4. More than one in six people living in these ZIP codes (17.7%) lived in poverty. Racial/ethnic proportions were 61.6% white, 24.7% black and 9.8% Hispanic.
Variables that correlated with higher HIV prevalence were higher prison release rate (Spearman correlation 0.26 overall, 0.23 in men, 0.34 in women), higher GINI index (0.49 overall, 0.53 in men, 0.32 in women), higher poverty (0.49 overall, 0.43 in men, 0.62 in women), less education (-0.29 overall, -0.20 in men, -0.45 in women), black race (0.41 overall, 0.53 in men, 0.53 in women), white race (-0.47 overall, -0.40 in men, -0.58 in women) and every additional kilometer from the city center (-0.49 overall, -0.50 in men, -0.37 in women) (P < .01 for all).
Logistic regression analysis identified five independent predictors of HIV prevalence in the 12 target cities. Every 1-unit increase in GINI index was associated with 15% higher overall HIV prevalence. Every 1-unit increase in prison-release rate meant a 4% higher overall HIV prevalence. Every 1-unit increase in distance from city center translated into a 3% lower overall HIV prevalence. Every 1-unit increase in percent living in poverty was linked to a 2% higher overall HIV prevalence. And every 1-unit increase in percent of black population meant a 1% higher HIV prevalence.
The researchers noted that the correlation they identified between inequality and HIV prevalence confirms earlier findings. Their study adds prison-release rate as an important population-level factor associated with neighborhood HIV prevalence. They underlined the need for better interventions to improve prison and jail discharge planning for people with HIV.