Table 5.

Logistic regression estimates of the short- and long-term effects of built environment on first onset of fair/poor self-rated health (N = 15,372)

1. Short-term exposure	Conventional model		MSM
	Odds ratio	SE	Odds ratio	SE
Single-point-in-time
Built environment category (ref = Average)
High density	1.050	.131	.805	.125
Low density	1.115	.092	.993	.103
Service dense	0.977	.139	.868	.150
Commercially declined	1.319	.241	1.531	.388
2. Cumulative exposure	Conventional model		MSM
	Odds ratio	SE	Odds ratio	SE
Running average
Built environment category (ref = Average)
High density	.985	.113	.882	.132
Low density	1.224*	.110	1.187	.153
Service dense	1.314	.204	.952	.189
Commercially declined	1.493*	.288	1.851*	.517

Note: Combined estimates from 10 multiple imputation data sets. All models include all baseline and time-varying individual and zip code covariates, although their effects are not reported.

*p < .05.