Estimated attributable fractions (%), attributable numbers and 95% CIs for cumulative cold and heat effects on daily hospitalisations for coronary heart disease (CHD) and stroke over a lag of 21 days in Ontario, Canada, 1996–2013
| Temperature range | CHD | Stroke | ||
| Attributable fractions (%) (95% CI) | Attributable number (95% CI) | Attributable fractions (%) (95% CI) | Attributable number (95% CI) | |
| Total cold* | 2.49 (1.03 to 3.67) | 34 456 (14 244 to 50758) | 1.71 (0.77 to 2.64) | 6051 (2744 to 9361) |
| Extreme cold† | 0.23 (0.14 to 0.29) | 3119 (1969 to 4072) | 0.29 (0.21 to 0.36) | 1038 (752 to 1277) |
| Moderate cold‡ | 2.27 (0.88 to 3.47) | 31 351 (12 152 to 47 993) | 1.42 (0.38 to 2.33) | 5019 (1336 to 8261) |
| Total heat* | 1.20 (0.22 to 2.14) | 16 628 (3040 to 29 559) | 1.82 (0.06 to 3.21) | 6469 (223 to 11 378) |
| Extreme heat† | 0.16 (0.09 to 0.22) | 2176 (1216 to 3093) | 0.19 (0.11 to 0.26) | 680 (401 to 921) |
| Moderate heat‡ | 1.05 (0.07 to 1.87) | 14 472 (929 to 25 823) | 1.63 (−0.06 to 3.10) | 5795 (−204 to 10 986) |
*Total cold and heat were defined as temperatures lower and higher than the optimum temperature.
†Extreme cold was defined as temperatures lower than the 2.5th percentile, while extreme heat was defined as temperatures higher than the 97.5th percentile.
‡Moderate cold was defined as temperatures in the range between the optimal temperature and the 2.5th percentile, while moderate heat was defined as temperatures in the range between the optimum temperature and the 97.5th percentile.