Heat acclimatization and sunshine cause false indications of mortality due to ozone

We used graphic analysis and generalized additive modeling to assess whether mortality that is often attributed to ozone and sometimes to < 10-microm particulate matter (PM10) and sulfur dioxide in hot weather results from confounding by neglected weather factors. When mean daily air temperatures exceeded 18 degrees C, mortality at age > or = 65 years, in Greater London 1991-2002 rose and at each temperature was generally higher in early summer than later after exposure to hot weather. Ozone, and to lesser degrees PM10 and sulfur dioxide, followed similar patterns. At each temperature, high levels of ozone and PM10 tended to be associated with sunshine, and high PM10 and sulfur dioxide with low wind, both of which will increase heat stress. With allowance for these confounding factors, generalized additive modeling showed no significant mortality due to ozone, to PM10, or to sulfur dioxide. We conclude that the pollutants played little part in excess mortality associated with hot weather.