Table 3.

Percent increase in risk of death (95% CI) for a 10-μg/m³ increase in NO₂: single-pollutant models and models adjusted for PM₁₀ and O₃ (pooled results, 10 cities), EpiAir Study, Italy, 2001–2005.

Model	Percent increase in risk	95% CI	HETa
All natural mortality
NO2 (lag 0–5), single-pollutant model		2.09		0.96 to 3.24		0.001
Model with PM10 (lag 0–5)		1.95		0.50 to 3.43		0.003
April–September
NO2 (lag 0–5), single-pollutant model		4.46		3.14 to 5.80		0.109
Model with O3 (lag 0–5)		4.55		3.32 to 5.79		0.175
Cardiac mortality
NO2 (lag 0–5), single-pollutant model		2.63		1.53 to 3.75		0.658
Model with PM10 (lag 0–5)		2.58		1.05 to 4.13		0.265
April–September
NO2 (lag 0–5), single-pollutant model		4.77		2.92 to 6.65		0.871
Model with O3 (lag 0–5)		4.69		2.74 to 6.67		0.925
Cerebrovascular mortality
NO2 (lag 0–5), single-pollutant model		2.35		–0.13 to 4.89		0.266
Model with PM10 (lag 0–5)		2.55		–0.71 to 5.92		0.247
April–September
NO2 (lag 0–5), single-pollutant model		7.87		4.78 to 11.05		0.628
Model with O3 (lag 0–5)		7.26		3.51 to 11.14		0.420
Respiratory mortality
NO2 (lag 1–5), single-pollutant model		3.48		0.75 to 6.29		0.000
Model with PM10 (lag 0–5)		3.39		0.77 to 6.08		0.512
April–September
NO2 (lag 1–5), single-pollutant model		9.63		4.08 to 15.47		0.016
Model with O3 (lag 0–5)		10.07		3.69 to 16.83		0.004
ap-Value from the heterogeneity test (the null hypothesis being homogeneity of the city-specific results).