Table 2: Sensitivity analyses for the association of nonaccidental mortality with every 5°C change in cold and hot temperatures across 27 selected census divisions in Ontario, 1996-2010
ModelPooled mean percent change in daily nonaccidental mortality (95% CI)*
Cold season
(n = 188 415)
Warm season
(n = 164 403)
Modelled different temperature metrics
Daily mean air temperature3.0(1.8 to 4.2)2.5(1.3 to 3.8)
Daily maximum air temperature3.0(1.9 to 4.2)2.2(1.0 to 3.4)
Daily minimum air temperature2.4(1.3 to 3.5)1.8(0.6 to 2.9)
Daily mean humidex- -1.6(0.8 to 2.3)
Daily maximum humidex--1.5(0.7 to 2.3)
Daily mean windchill1.8(0.7 to 3.0)--
Daily maximum windchill2.3(0.8 to 3.8)--
Modelled different lag structure for temperature
Lag 00.8(-0.1 to 1.7)2.5(1.3 to 3.8)
Lag 11.3(0.4 to 2.2)0.1(-1.1 to 1.3)
Lag 21.2(0.1 to 2.2)-1.0(-2.2 to 0.1)
Lag 31.5(0.6 to 2.5)-0.5(-1.4 to 0.5)
Lag 42.1(1.3 to 3.0)-0.2(-1.2 to 0.8)
Lag 51.6(0.8 to 2.4)-0.5(-1.5 to 0.5)
Lag 60.9(0.0 to 1.9)-0.6(-1.9 to 0.6)
Lag 0-11.4(0.3 to 2.5)1.7(0.3 to 3.1)
Lag 0-32.1(0.9 to 3.3)0.2(-1.2 to 1.7)
Lag 0-63.0(1.8 to 4.1)-0.1(-1.5 to 1.3)
Adjusted for snow fall†
No3.6(2.0 to 5.2)--
Yes3.6(1.9 to 5.3)--
Adjusted for fine particulate matter‡
No2.7(1.7 to 3.8)2.5(1.3 to 3.8)
Yes2.8(1.6 to 3.9)2.6(1.2 to 4.1)

*Cold season: daily mean temperature at lag 0-6; warm season: daily mean temperature at lag 0.

†Restricted to census divisions with available data on snow fall.

‡Restricted to days with available data of fine particulate matter (lag 0-2).