Table 2. Obesity Odds Ratios by Individual and Contextual Variables from the Multivariate Regression Model That Adjusted for Individual and County-Level Variables

Variables	Odds Ratio	95% Confidence Interval
PCP supply in county of residence
Quintile 1 (most access)	1.000	Index
Quintile 2	1.081^*	1.031–1.133
Quintile 3	1.053^{^†}	1.004–1.105
Quintile 4	1.088^*	1.039–1.139
Quintile 5 (least access)	1.187^*	1.136–1.241
Individual variables
Demographic
Age	1.120^*
Education
Some high school	1.080^*	1.023–1.141
High school graduate	1.00	Index
Some college	0.966	0.931–1.002
College graduate	0.668^*	0.642–0.694
Parental status
No children	1.00	Index
Parent	1.063^*	1.023–1.104
Race/ethnicity
Non-Hispanic white	1.00	Index
Black	1.611^*	1.527–1.701
Hispanic	1.124^*	1.061–1.191
Other race	0.681^*	0.631–0.735
Marital status
Married	1.00	Index
Never married	1.048	0.999–1.100
Divorced	0.993	0.956–1.030
Insured**	1.076	1.025–1.130
Income ($/year)
<15,000	1.119^*	1.025–1.130
15,000–20,000	1.022^{^†}	1.025–1.222
20,000–25,000	1.002	0.924–1.086
25,000–35,000	0.890^*	0.821–0.966
35,000–50,000	0.827^*	0.766–0.894
50,000–75,000	0.842^*	0.766–0.894
>75,000	0.659^*	0.682–0.713
Smoking status
Never smoker	1.00	Index
Smoker	0.768^*	0.735–0.802
Former smoker	1.175^*	1.136–1.215
County-level variables
Black race (%)	1.042	0.916–1.185
Rural-urban continuum codes
1 (Most urban)	1.000	Index
2	1.133^*	1.089–1.178
3	1.171^*	1.117–1.227
4	1.222^*	1.156–1.292
5	1.179^*	1.098–1.265
6	1.187^*	1.122–1.256
7	1.190^*	1.110–1.274
8	1.081	0.950–1.231
9 (Most rural)	1.245^*	1.070–1.448
County poverty
Low	1.000	Index
Low/mid	1.038	0.994–1.084
Mid/high	1.049	1.006–1.094
High	1.095	1.043–1.150

Data are based on N = 392,535 individuals. Obesity is defined as body mass index ≥30 kg/m².
↵* P ≤ .05.
↵† P ≤ .01.
Data Sources: 2012 Behavioral Risk Factor Surveillance System, 2012 AMA Masterfile, 2010 Census estimates of county population, 2003 rural urban continuum codes, and 2012–2013 Area Resource Files.
PCP, primary care physician.