Table 3.

Bivariate and multivariate results of a multilevel regression of HIV viral suppression on individual- and United Hospital Fund- (UHF-) level covariates in a sample of heterosexuals (N = 3,159) in New York City diagnosed with HIV between 2009–2013

Variable	Bivariates		Multivariable model
	Relative Risk	P-value	Adjusted Relative Risk	P-value
UHF level variables
% residents who were food distressed (ref = high)
Low <1.00	1.25	0.04	1.18	0.09
Medium 1.00–<5.00	1.18	0.13	1.12	0.23
% residents who were non-Hispanic Black (ref = high)
Low <5.00	1.21	0.01	1.13	0.09
Medium 5.00-<30.00	1.04	0.44	1.02	0.72
% residents who were 21–54 years old	1.00	0.68
Alcohol outlet density	1.08	0.32
Residential vacancy	1.00	0.11	1.00	0.16
Stops per resident (ref = high)
Low <0.04	1.06	0.49
Medium 0.04-<0.20	1.10	0.19
Stops without arrest per resident (ref = high)
Low <0.06	1.12	0.12	1.03	0.55
Medium 0.06-<0.22	1.06	0.38	1.06	0.29
UHF stabilityⱡ	1.01	0.30
UHF affluence₸	1.01	0.47
Neighborhood disadvantage†	0.99	0.26
Poor access to healthcare	0.94	0.87
Individual-level variables
Male (ref = Female)	0.89	0.01	0.87	0.001
Race/ethnicity (ref = Black)
Hispanic	1.14	0.0002	1.06	0.07
White	1.05	0.47	0.98	0.78
Foreign-born	1.30	<.0001	1.23	<.0001
History of homelessness	0.86	0.16
Age at HIV diagnosis (ref = 40–49)
13–19	0.85	0.08	0.87	0.15
20–29	0.99	0.86	0.96	0.40
30–39	1.05	0.30	1.04	0.38
50-59	1.08	0.10	1.06	0.19
60+	1.12	0.05	1.05	0.35
Year of HIV diagnosis (ref = 2009)
2010	1.15	0.01	1.14	0.01
2011	1.22	<.0001	1.21	0.0001
2012	1.31	<.0001	1.31	<.0001
2013	1.43	<.0001	1.36	<.0001

^ⱡHigher values mean more stability.

^₸Higher values mean more affluence.

^†Higher values mean more disadvantage.