Assessing historical exposures of children to power-frequency magnetic fields

One risk factor for human cancer currently being studied is residential exposure to power-frequency magnetic fields. A key problem in such research is how best to use contemporaneous measurements to assess past magnetic-field exposures. The main goal of the research presented in this paper was to examine the effectiveness of various surrogate measures in predicting historical exposures and to determine if residential power consumptions and the loads served by neighborhood electric networks could be used to improve the accuracy of such predictions. Residential magnetic-field data were collected during 24-h periods in the spring of 1990 and, again, in the winter of 1990-1991 for 35 children living in Western Massachusetts and Northern California. Measurements included spot magnetic fields in rooms occupied by subjects for an average of one or more hours per day, 24-h recordings at locations selected to emphasize ground-current and power-line fields, personal exposures, wire codes, residential power consumptions, and loads served by neighborhood electric networks. The geometric means of time-weighted-averaged (TWA) room spot magnetic fields measured during earlier and later visits to each home were 0.052 microT and 0.060 microT, respectively. Geometric-mean personal exposures for these visits were 0.084 microT and 0.111 microT and were significantly larger. Wertheimer-Leeper wire codes were associated with exposure. These codes, TWA spot fields, and the 24-h averages of the magnetic-field recordings taken to emphasize power-line contributions were about equally effective in explaining between-home variability in personal exposures measured eight months in the past or future. In contrast, personal exposure measurements were ineffective surrogates for past or future exposure. The study yielded little evidence suggesting that residential power consumption or neighborhood electric power flow are helpful in explaining temporal changes in personal exposure.