The widely held belief that European levels of air pollution might seriously affect human health has been hard to verify. Most studies investigate the immediate effects on mortality, comparing day to day variations in atmospheric pollutants, temperature, and humidity with variations in deaths. High and low temperatures increase mortality while the effects of cold at least are prolonged1 and temperature is associated with pollution. Most investigators attempt to adjust for this confounding, although methods vary. Because both relations are complex, simple adjustments for confounding may be inadequate and some or all the apparent effects of pollutants may be indirect associations caused by the effects of temperature. A recent paper by Keatinge and Donaldson now questions whether this is true for some recent studies such as those from the “air pollution and health—a European approach” (APHEA) project, citing sulphur dioxide as an example.2

Several reports suggest that atmospheric sulphur dioxide is associated with short term mortality, although the variability in results needs explanation. The APHEA project was designed to investigate this by, among other things, using several years of daily measurements from 15 European cities and standardised analyses.3 Katsouyanni et al concluded from it that rising sulphur dioxide levels increased mortality in western but not eastern Europe, although their results also varied.4 The protocol for temperature adjustment for …