Mean rate ratios and 95% CIs
| Number of deaths | Ozone effect on mortality | p Value | Heat effect | p Value | |
| Analyses using temperature variables constructed from daily mean temperature | |||||
| All deaths | 592320 | 1.003 (1.001 to 1.005) | 0.01 | 1.071 (1.050 to 1.093) | <0.01 |
| CVD | 233219 | 1.003 (1.000 to 1.006) | 0.08 | 1.055 (1.025 to 1.087) | <0.01 |
| Respiratory | 82757 | 1.003 (0.996 to 1.011) | >0.20 | 1.139 (1.079 to 1.202) | <0.01 |
| Other causes | 276344 | 1.002 (0.999 to 1.005) | 0.18 | 1.083 (1.061 to 1.106) | <0.01 |
| Ages 0–64 | 117221 | 1.002 (0.997 to 1.006) | >0.20 | 1.019 (0.978 to 1.062) | >0.20 |
| Ages 65–74 | 125635 | 1.004 (0.997 to 1.011)H | >0.20 | 1.067 (1.031 to 1.103) | <0.01 |
| Ages 75–84 | 191520 | 1.001 (0.998 to 1.005) | >0.20 | 1.092 (1.063 to 1.121) | <0.01 |
| Ages 85+ | 153786 | 1.003 (0.999 to 1.007) | 0.10 | 1.124 (1.093 to 1.155) | <0.01 |
| Sensitivity analyses (all deaths only) | |||||
| A: using maximum temperature | 592320 | 1.000 (0.998 to 1.002) | >0.20 | 1.087 (1.062 to 1.113) | <0.01 |
| B: six-knot temperature spline | 592320 | 1.003 (1.001 to 1.005) | 0.01 | 1.074 (1.054 to1.095) | <0.01 |
| C: plus season (4 d.f./year) | 592320 | 1.003 (1.001 to 1.005) | <0.01 | 1.087 (1.065 to 1.109) | <0.01 |
| D: 97.5th percentile for hot-day threshold | 592320 | (same model as Model 1) | |||
| E: ‘4+-day heatwave’ | 583917 | (same model as Model 1) | |||
| F: ‘5+-day heatwave’ | 579746 | (same model as Model 1) | |||
| G: lag 0–2 of ozone | 588111 | 1.002 (1.000 to 1.004) | 0.02 | 1.073 (1.053 to 1.095) | <0.01 |
| H: lag 0 for ozone and temperature | 594477 | 1.001 (0.999 to 1.002) | >0.20 | 1.087 (1.067 to 1.107) | <0.01 |
| I1: without humidity | 535922 | 1.003 (1.000 to 1.005) | 0.03 | 1.074 (1.052 to 1.096) | <0.01 |
| I2: with humidity | 535922 | 1.004 (1.000 to 1.008)H | 0.04 | 1.073 (1.053 to 1.094) | <0.01 |
| J: excluding three conurbations | 554457 | 1.003 (1.001 to 1.005) | 0.01 | 1.075 (1.052 to 1.099) | <0.01 |
| K: O3 hockey-stick model (threshold at 65 μg/m3) | 592320 | 1.005 (1.002 to 1.008) | <0.01 | 1.072 (1.052 to 1.093) | <0.01 |
| Analyses using temperature variables constructed from daily maximum temperature | |||||
| CVD | 233219 | 1.001 (0.997 to 1.004) | >0.20 | 1.072 (1.039 to 1.106) | <0.01 |
| Respiratory | 82757 | 1.002 (0.995 to 1.009) | >0.20 | 1.153 (1.104 to 1.205) | <0.01 |
| Other | 276344 | 0.999 (0.996 to 1.002) | >0.20 | 1.098 (1.068 to 1.129) | <0.01 |
| Ages 0–64 | 117221 | 1.000 (0.995 to 1.006) | >0.20 | 1.018 (0.961 to 1.078)H | >0.20 |
| Ages 65–74 | 125635 | 1.002 (0.994 to 1.010)H | >0.20 | 1.070 (1.022 to 1.121) | <0.01 |
| Ages 75–84 | 191520 | 0.998 (0.994 to 1.002) | >0.20 | 1.110 (1.079 to 1.142) | <0.01 |
| Ages 85+ | 153786 | 1.000 (0.995 to 1.004) | >0.20 | 1.148 (1.112 to 1.185) | <0.01 |
| Sensitivity analyses (all deaths only) | |||||
| B: six-knot temperature spline | 592320 | 1.000 (0.998 to 1.003) | >0.20 | 1.087 (1.062 to 1.112) | <0.01 |
| C: plus season (4 d.f./year) | 592320 | 1.000 (0.997 to 1.002) | >0.20 | 1.105 (1.075 to 1.135)H | <0.01 |
| G lag 0–2 of ozone | 588111 | 1.000 (0.998 to 1.002) | >0.20 | 1.089 (1.066 to 1.113) | <0.01 |
| H: lag 0 for ozone and temperature | 594477 | 0.999 (0.997 to 1.001) | >0.20 | 1.094 (1.078 to 1.111) | <0.01 |
| I1: without humidity | 535922 | 1.001 (0.997 to 1.004) | >0.20 | 1.084 (1.053 to 1.115)H | <0.01 |
| I2: with humidity | 535922 | 1.002 (0.998 to 1.007)H | >0.20 | 1.083 (1.051 to 1.116)H | <0.01 |
| J: excluding three conurbations | 554457 | 1.000 (0.998 to 1.002) | >0.20 | 1.093 (1.066 to 1.121) | <0.01 |
| K: O3 hockey-stick model (threshold at 130 μg/m3) | 592320 | 1.011 (0.999 to 1.025) | 0.08 | 1.085 (1.061 to 1.109)H | <0.01 |
Left hand side: rate ratios from Model 1 per increase of 10 μg/m3 of 0–1-day average of 8-h max ozone. Right hand side: rate ratios from Model 1 for effect of heat (mortality rate for 2-day mean temperature at its 97.5th percentile, relative to that at the 75th percentile). All rate ratios are adjusted for PM10, residuals from previous 3 days, day of week, bank holiday and month. Those for ozone are also adjusted for spline of mean temperature. Those for heat are adjusted for ozone. Note: deaths in age groups do not add up to the total number of deaths because of records with missing ages. The superscript ‘H’ indicates evidence of heterogeneity (p≤0.10) between conurbation-specific results.