Assessing the association between food environment and dietary inflammation by community type: a cross-sectional REGARDS study

Table 3 Model-based associations of the food environment with dietary inflammation score (DIS)

	Main effects model		Interaction model
	β (SE)	p-value	β (SE)	p-value
Supermarkets, tailored^a	0.33 (0.22)	0.13	0.34 (0.22)	0.13
Fast-food restaurants, tailored^a	0.59 (0.12)	<0 .001	0.42 (0.23)	0.06
Community type
Higher density urban	− 0.26 (0.07)	< 0.001	0.35 (-0.71)	0.94
Lower density urban	− 0.09 (0.05)	0.06	0.29 (− 0.44)	0.67
Suburban/small town	− 0.04 (0.05)	0.45	0.42 (0.35)	0.01
Rural (reference)
Fast-food restaurants X Community type^b
Higher density urban			− 0.03 (0.35)	0.94
Lower density urban			0.12 (0.29)	0.67
Suburban/small town			1.18 (0.42)	0.01
Rural (reference)

Bold denotes statistically significant at α < 0.05 level. Supermarkets and fast-food restaurants were modeled together. We controlled for individual-level covariates, NSEE, and total food outlets in all models. Higher DIS scores indicate more proinflammatory diets (theoretical range: − 14.9–12.8)
^aWe tailored buffer sizes to each community type using 2-, 3-, 10-, and 16 km (1-, 2-, 6-, and 10-mile) buffers for higher density urban, lower density urban, suburban/small town, and rural areas, respectively. Buffer sizes are represented in kilometers rounded to the nearest whole number
^bP-value for overall test of interaction = 0.02

ISSN: 1476-072X