Focusing on fast food restaurants alone underestimates the relationship between neighborhood deprivation and exposure to fast food in a large rural area

Table 3 Association between 1-mile and 3-mile coverage of fast-food opportunities and area deprivation, using multivariate linear regression model

Model 1	Access (number of fast-food opportunities within 1 network mile)
	Fast-food restaurant	Fast-food opportunities	Variety healthier entrée options	Variety healthier side dish options
Deprivation	b (SE)	b (SE)	b (SE)	b (SE)
High	0.24 (0.59)	2.39 (0.90)^†	1.02 (0.43)*	0.83 (0.18)^‡
Medium	0.24 (0.50)	0.67 (0.76)	0.23 (0.36)	0.16 (0.15)
R²	0.611	0.708	0.662	0.509
P	<0.001	<0.001	<0.001	<0.001
Model 2	Access (number of fast-food opportunities within 3 network miles)
	Fast-food restaurants	Fast-food opportunities	Variety healthier entrée options	Variety healthier side dish options
Deprivation	b (SE)	b (SE)	b (SE)	b (SE)
High	2.75 (1.31)*	5.91 (2.42)*	2.92 (0.98)^†	1.37 (0.35)^‡
Medium	0.45 (1.11)	1.11 (2.05)	0.49(0.83)	0.11 (0.29)
R²	0.713	0.716	0.722	0.658
P	<0.001	<0.001	<0.001	<0.001

NOTE: In both models, the four equations were simultaneously estimated, controlling for population density. In model 1 (fast-food opportunities within 1 network mile), deprivation entered as categorical variable; low deprivation is referent group. Model 2 estimates the relationships between deprivation and the number of fast-food opportunities within 3 network miles. In both models, population density entered as continuous variable. Results reported as multivariate-adjusted b (SE). Statistically significant variables are indicated as: *< 0.05 ^†< 0.01 ^‡< 0.001

ISSN: 1475-2891