Analyses of magnetic-field peak-exposure summary measures

Two previous epidemiologic studies reported an association between the maximum magnetic field exposure logged during a 24-h measurement period and risk of miscarriage. A hypothesis was put forth which argued that the observed association may be the result of behavioral differences between women with healthy pregnancies (less physically active) and women with miscarriage. We analyzed four existing data sets with power-frequency magnetic-field personal exposure (PE) measurements to investigate the characteristics of peak-exposure measures. We found that the value of the measured maximum magnetic-field exposure varied inversely with the sampling interval between magnetic-field measurements and that maximum values demonstrated less stability over time in repeated measurements, compared to time-weighted average and 95th and 99th -percentile values. We also found that the number of activity categories entered by study subjects could be used to estimate the proportion of subjects with exposure above various threshold values. Exposure metrics based on maximum values exceeding thresholds tend to classify active people into higher exposure categories. These findings are consistent with the hypothesis suggesting that the association between maximum magnetic fields and miscarriage are possibly the result of behavioral differences between women with healthy pregnancies and women who experience miscarriages. Thus, generalization from a given study to more global exposure characterization should be made with particular caution and with due consideration to sampling interval and other characteristics of the measurement protocol potentially influencing the measured maximum. Future epidemiologic studies of peak magnetic field exposure and spontaneous abortion should carefully evaluate the potential confounding effect of the women's activity level during pregnancy.