Table 2.

Relative risks of hospitalization for psychotic disorders per interquartile increases in levels of PM_2.5 (μg/m³), NO₂ (ppb) in total study population and stratified by sex, age range, area level poverty, percent of blacks and Hispanics

	PM2.5		NO2
Group	RRa	95% CI	RR	95% CI
Total	1.11	1.09, 1.13	1.27	1.24, 1.31
Sex
Male	1.11	1.08, 1.13	1.33	1.28, 1.37
Female	1.14	1.12, 1.16	1.24	1.20, 1.27
P for differenceb	0.062		0.001
Age range
Younger age <30 years	1.00	0.97, 1.02	1.28	1.24, 1.32
Older age >=30 years	1.17	1.14, 1.19	1.30	1.26, 1.34
P for difference	<0.001		0.438
Poor vs. less poor areasc
Poor	1.29	1.24, 1.34	1.42	1.33, 1.51
Less poor	1.00	0.98, 1.02	1.19	1.16, 1.22
P for difference	<0.001		0.001
Percent of Blacksd
High	1.08	1.05, 1.11	1.07	1.02, 1.13
Low	1.08	1.05, 1.11	1.17	1.14, 1.21
P for difference	0.387		<0.001
Percent of Hispanicse
High	1.18	1.14, 1.22	1.18	1.11, 1.25
Low	1.02	0.99, 1.04	1.21	1.18, 1.23
P for difference	<0.001		<0.001

^aRelative risks (RR) of hospitalization for psychotic disorders per interquartile (IQR) increase in PM_2.5 (μg/m³) and NO₂ (ppb);

^bP value for effect heterogeneity (difference) z test between the comparison sub-groups;

^cPoor areas defined as above 67% percentile of community poverty level, less poor areas defined as less than 67% percentile of community poverty level;

^dHigh indicates high percent of blacks in the area vs. low indicates (relatively) low percent of blacks in the area using 67% as the cutoff;

^eHigh indicates high percent of Hispanics in the area vs. low indicates (relatively) low percent of Hispanics in the area using 67% as the cutoff;

^*Analyses were conducted for areas with population on the subgroup based on multi-exposure model including air pollutants, temperature, precipitation, relative humidity, daylight intensity, admission state, admission year, population density, area-level long-term socioeconomic, healthcare access and behavioral factors as described in Covariates section.