. 2020 Aug 12;15(8):e0237269. doi: 10.1371/journal.pone.0237269

Table 2. A multivariate analysis of county level interactions involving Latino disparity.

Category	Variable^δ	Model 1: N = 485^a					Model 2: N = 675^b
Category	Variable^δ	Beta	Std^† Beta	SE	p-value	Role of Factor	Beta	Std^† Beta	SE	p-value
HIV Characteristics	County Prevalence Rate^β	—	—	—	—		—	—	—	—
	Male Prevalence Rate^β	—	—	—	—		—	—	—	—
	Female Prevalence Rate^β	-0.07	-0.06	0.03	0.044^*	Effect Modifier	-0.02	-0.02	0.03	0.574
	% MSM	—	—	—	—		—	—	—	—
	% IDU	0.03	0.21	0	< 0.001^*	Effect Modifier	—	—	—	—
	% Heterosexual	—	—	—	—		—	—	—	—
	New Diagnosis Rate^β	0.03	0.03	0.03	0.312	Homogeneous	—	—	—	—
	IDU Prevalence Rate^β	—	—	—	—		—	—	—	—
	MSM Prevalence Rate^β	—	—	—	—		—	—	—	—
Socioeconomic	Median Household Income^β	—	—	—	—		—	—	—	—
	GINI Index	-0.01	-0.02	0.01	0.629		-0.02	-0.06	0.01	0.115
	% High School Education	-0.01	-0.04	0.01	0.223		-0.02	-0.08	0.01	0.031^*
	% Unemployed	—	—	—	—		—	—	—	—
	Income Inequality Ratio^β	0.5	0.07	0.28	0.078	Homogeneous	0.81	0.11	0.28	0.004^*
	Non-White/White Segregation Index	—	—	—	—		—	—	—	—
	Black/White Segregation Index	—	—	—	—		—	—	—	—
	% Non-Latino White Living Below Poverty	—	—	—	—		—	—	—	—
	% Living Below Poverty	—	—	—	—		—	—	—	—
Community Environment	% Single Parent Household	—	—	—	—		—	—	—	—
	Social Association Rate	-0.2	-0.07	0.07	0.007^*	Effect Modifier	-0.14	-0.05	0.07	0.037^*
	Violent Crime Rate	-0.07	-0.04	0.05	0.161		-0.1	-0.06	0.04	0.033^*
	% Severe Housing Problems	-0.03	-0.13	0.01	< 0.001^*	Effect Modifier	-0.03	-0.12	0.01	< 0.001^*
	% Rural	0	-0.06	0	0.046^*	Effect Modifier	0	-0.02	0	0.625
Health Behaviors	% Excessive Drinking	—	—	—	—		—	—	—	—
	Chlamydia Rate^β	—	—	—	—		—	—	—	—
	Drug Overdose Mortality Rate^β	-0.26	-0.12	0.05	< 0.001^*	Effect Modifier	-0.24	-0.11	0.05	< 0.001^*
	Gonorrhea Rate^β	0	0	0.03	0.921		-0.03	-0.02	0.03	0.389
Access to Health Care	% Uninsured	0	0	0.01	0.932		0	-0.01	0.01	0.85
	Primary Care Provider Rate^β	—	—	—	—		—	—	—	—
	Mental Health Provider Rate^β	—	—	—	—		—	—	—	—
	Preventable Hospitalization Rate^β	-0.13	-0.04	0.11	0.235		-0.07	-0.02	0.1	0.527
	Healthcare Costs^β	-0.32	-0.04	0.23	0.162	Homogeneous	-0.38	-0.05	0.23	0.096
Latino Characteristics	Latino Population % Change 2000–2014	-0.05	-0.05	0.02	0.035^*	Effect Modifier	-0.07	-0.07	0.02	0.001^*
	% Mexican	0	-0.07	0	0.052	Homogeneous	0	-0.04	0	0.277
	% Puerto Rican	0.01	0.09	0	0.006^*	Effect Modifier	0.01	0.12	0	< 0.001^*
	% Latino	—	—	—	—		—	—	—	—
	% Not English Proficient	0.04	0.13	0.01	0.003^*	Effect Modifier	0.03	0.1	0.01	0.015^*
	% Latino Living Below Poverty	0.02	0.12	0	< 0.001^*	Effect Modifier	0.01	0.09	0	< 0.001^*
Latino/Non-Latino White Ratios	Poverty Ratio (% Latino / % Non-Latino White) ^β	—	—	—	—		—	—	—	—
	Population Ratio (% Latino / % Non-Latino White) ^β	-0.35	-0.4	0.04	< 0.001^*	Effect Modifier	-0.33	-0.38	0.03	< 0.001^*
	Income Ratio (% Latino / % Non-Latino White) ^β	—	—	—	—		—	—	—	—

^δ Each row represents a factor in a multivariable model, only the interaction terms from the models are displayed. A dashed line indicates this variable was non-significant in the individual level analyses of disparity and therefore not included in this multivariate analysis, or they were redundant and excluded from the final model.

^†Std = standardized.

^a Model 1 includes all variables found significant in the individual level analyses, excluding redundant variables (% MSM, MSM Prevalence Rate while retaining % IDU; % Latino while retaining % Mexican and % Puerto Rican).

^b Model 2 additionally excludes two HIV characteristics with large amounts of missing/suppressed data: % IDU, and New Diagnosis Rate.

^β Denotes log transformation of variable.

* Significant at p<0.05.