Commentary: Cancer in the air

the work of aircrews entails occupational exposure to cosmic radiation mostly in the range of 2–4 mSv/year. In 1991, the International Commission on Radiological Protection recommended an occupational exposure limit of 20 mSv/year.1 The actual exposure level of aircrews is thus normally well below the recommended limit. This limit is based, however, on disease occurrence in populations exposed to other types of ionizing radiation, e.g. following the Hiroshima and Nagasaki bombing. A detailed monitoring of the long-term health consequences of aircrew work is therefore well justified. The study by Haldorsen et al.2 reported in this issue of the International Journal of Epidemiology forms part of the monitoring activity. Epidemiological studies of mortality and cancer incidence in aircrews have recently been reviewed.1,3 The studies fall into three categories. Firstly, proportional mortality, PMR, studies of military and commercial aircrews. Aircrew work requires health certification, and the group is subject to a strong healthy worker selection. The PMR studies may therefore be used only as a crude method of hypothesis generation. Secondly, cohort studies have been conducted of military pilots. Military pilots are, however, unlikely on average to receive the same cosmic radiation dose as commercial aircrews as they do not fly long hours in high altitudes. Thirdly, cohort studies have been conducted of commercial pilots and cabin attendants. These studies are clearly the most informative concerning long-term health effects of exposure to cosmic radiation. When it comes to assessment of the potential cancer risk, it is furthermore advantageous when these cohorts are followed for cancer incidence, as this implies more cases and more accurate diagnosis than the follow-up for cancer mortality. At present, five cohort studies of cancer incidence in commercial air pilots have been published, two from Canada,4,5 one from Denmark,6 one from Iceland,7 and one from Norway.8 Two cohort studies of cabin attendants have been published from Finland9 and from Iceland.10 Data from one small cohort of retired cabin attendants from the US have been reported in letters,11,12 and in a review.3 Selected results from a small Danish cohort of cabin attendants have been published in a letter.13 Haldorsen et al. in this issue of the journal report on a cohort study of Norwegian cabin attendants. Unfortunately, 830 INTERNATIONAL JOURNAL OF EPIDEMIOLOGY