The ultimate inferences from wire codes pertain not to characterizing the home's magnetic fields but rather characterizing the home occupant's magnetic-field exposure. A few studies mentioned above have obtained data pertinent to addressing the question of the extent to which wire codes provide valid estimates of individual exposure.

Wire codes reflect only power-line fields, which, as noted above, are only part of the residential exposure. Spot and long-term-average measurements are generally made near the centers of rooms away from obvious sources of magnetic fields, and personal-exposure measurements are made while the subjects move around the house. Because the subjects occasionally use or are close to appliances, personal-exposure measurements are generally higher than the spot or long-term-average magnetic-field measurements. Based on the measurements reported above, it appears that the average grounding system contributes about 0.02 µT (0.2 mG) to residential personal exposure (regardless of wire code); appliances contribute between 0.02 and 0.03 µT, and power lines contribute the remainder (0.02, 0.04, 0.06, and 0.10 µT for VLCC, OLCC, OHCC, and VHCC wire codes, respectively).

It should not be surprising, therefore, that wire codes account for only a relatively small fraction of the variance in residential background fields and residential personal exposure (approximately 18%). Most of the variance between people in personal exposure is a result of variation in exposure to residential sources (appliances and grounding system); in a recent study of children, the correlation coefficient between log-transformed 24-hr residential background fields and total personal exposure was 0.97 (Kaune et al. 1994).

Although the literature on wire codes, residential measurements, and personal exposure to magnetic fields is still quite incomplete, some tentative conclusions can be drawn. The replicated association of wire codes with residential magnetic fields extends to an association between wire codes and personal time-integrated exposure. As shown in the last section, use of wire codes to classify personal exposure into broad magnetic field categories (e.g., greater than 0.2 µT and less than 0.2 µT) appears to be somewhat more accurate than estimation of actual levels of exposure, for which the proportion of magnetic fields accounted for is