Effect of incomplete exposure assessment on epidemiologic dose-response analyses

Occupational sunlight exposure and mortality from non-Hodgkin lymphoma among electric utility workers

This case-control study examined occupational sunlight exposure and death from non-Hodgkin lymphoma (NHL) and NHL subtypes among 188 cases and 1880 controls selected from a cohort of 138,905 male electric utility workers. Exposure was classified according to work history linked to indices of cumulative sunlight exposure. Odds ratios and 95% confidence intervals were derived from conditional logistic regression models and were conditioned on the matching factors birth year and ethnicity. Mortality from NHL and intermediate/high-grade lymphomas was not related to cumulative sunlight exposure, with odds ratios around the null. For low-grade lymphomas, a dose-response gradient was observed for exposure in the past 12 to 21 years, but this result seemed to be sensitive to cut points for categorization of cumulative exposure. These data do not provide evidence for an association between occupational sunlight exposure and mortality from NHL or NHL subtypes.

High-voltage overhead power lines in epidemiology: patterns of time variations in current load and magnetic fields

In epidemiological studies of electromagnetic fields and health effects, exposure classification is crucial. There is no generally accepted biophysical interaction mechanism, but many studies are based on the hypothesis of a causal relationship with the strength of magnetic field. Some definition of the magnitude of exposure must be used, e.g., mean magnetic flux density, the integral of magnetic flux and time, or a peak value. Magnetic fields around a particular power line depend on the current load. The aim of the present study was to follow variations in line current load in the power supply system of the largest Norwegian city on a yearly, monthly, daily and diurnal basis. Fairly large variations in load were found, but increases in consumption were not necessarily reflected in current load on high voltage lines. The correlation between outdoor temperature and current load varied widely, depending on the type of power station feeding the line in question. The registered time variations are large enough to interfere with epidemiological classification of residences and testing of epidemiological hypotheses.