Table 2.

Association between daily temperature and homicide, with odds ratios reported per 1 °C increase in temperature. Numbers in parentheses are 95% confidence intervals

	Deaths included		Odds ratio
	Definitea	Totalb	Definitea	Totalb
Max
Lag 0c	63,343	223,733	1.015 (1.013, 1.018)	1.012 (1.011, 1.013)
Lag 1d	63,335	223,735	1.012 (1.009, 1.014)	1.008 (1.007, 1.010)
Lag 0-1e	63,131	223,035	1.018 (1.015, 1.021)	1.014 (1.012, 1.015)
Meanf
Lag 0c	62,601	220,854	1.021 (1.017, 1.024)	1.017 (1.015, 1.019)
Lag 1d	62,661	220,999	1.016 (1.012, 1.019)	1.011 (1.009, 1.013)
Lag 0-1e	62,200	219,234	1.023 (1.019, 1.027)	1.018 (1.015, 1.020)
Min
Lag 0c	62,850	221,724	1.010 (1.007, 1.013)	1.009 (1.007, 1.011)
Lag 1d	62,890	221,874	1.007 (1.004, 1.010)	1.005 (1.004, 1.007)
Lag 0-1e	62,502	220,320	1.011 (1.008, 1.015)	1.010 (1.008, 1.012)

^aDefinite homicides, ICD-10: X85-Y09

^bIncludes definite homicides and probable homicides (ICD-10: W25-W26, W32-W34, W50, Y22-Y24, Y28-Y29)

^cSame day’s temperature

^dPrevious day’s temperature

^eMoving average of same and previous day’s temperature

^fHere the mean is defined as the average between the maximum and the minimum