Use of spot measurements for assessing residential ELF magnetic field exposure: A validity study

Abstract

The purpose of this study was to evaluate residential short term "spot" measurements as surrogates for long term personal magnetic field (MF) exposure. In an epidemiological study on birth weight and pregnancy delay, MF exposure was assessed by taking five spot measurements in each room. For a subsample of 30 subjects 24 h personal MF measurements were made, and the following exposure metrics were calculated: 24 h arithmetic mean, 24 h median, percentage of time above 0.15 microT, and percentage of time above 0.29 microT. The 24 h exposure metrics were used as gold standards, when evaluating the validity of various summary measures calculated from spot measurements for assessing personal exposure. Based on Spearman correlation coefficient (r), specificity and sensitivity, the average of the spot measurements of a residence resulted in least exposure measurement error (misclassification). Also the above bed spot value correlated better with the 24 h metrics than any room average. Spot measurements performed about equally well in predicting different types of exposure metrics.

Indoor transformer stations as predictors of residential ELF magnetic field exposure

“…In the reference apartments on other floors, only one (3.3%) reading equal to or higher than 1 mT (1.30 mT) was found. The whole apartment mean, which has been shown to be the best spot-measurement-based metric for predicting personal 24-h exposure [Eskelinen et al, 2002], was at least 0.2 mT in all apartments above transformer stations. This value was reached in 10 first floor reference apartments (36%) and in only 3 reference apartments (10%) on other floors.…”

Use of short-term measurements for assessing temporal variability of residential ELF magnetic field exposure

“…Rankin et al (2002) reported a Spearman correlation of 0.74 between point-in-time MF measurements and personal exposures, but the personal exposure measurement was limited to exposure in the home. Studies on the correlation of spot measurements with total 24 h personal exposure have shown correlations from 0.37 to 0.77 for timeweighted average exposure (Armstrong et al, 2001;Eskelinen et al, 2002) and from 0.41 to 0.74 for threshold-type exposure metrics (Eskelinen et al, 2002). It is of interest that, in the present data, the correlations with total 24 h exposure are not generally lower than those with the 12 h spent at home.…”

“…However, the 20-min measurements were not better than spot measurements for predicting long-term average exposure. For example, the residence-specific mean of spot measurements produced Spearman correlations of 0.77 for arithmetic mean, and 0.62 for median of the 24-h measurement (Eskelinen et al, 2002), which are higher than the correlations shown in Table 1. Also, the sensitivities and specificities (Table 1) were lower for the 20-min measurements than those found for the spot measurements (Eskelinen et al, 2002).…”

“…It has been shown that residential point-intime ''spot'' measurements can be used to estimate (in addition to time-average MF) time spent above an MF threshold (Juutilainen et al, 1996;Eskelinen et al, 2002).…”

Comments on “Maternal exposure to extremely low frequency magnetic fields: Association with time to pregnancy and foetal growth”

“…for example, the known spatial variability of EMF exposure combined with movement of the person for whom exposure is estimated can have non-trivial consequences (Savitz, 1995;Miller and Green, 2010), and for this population has been shown to have low to high correlation with 24-hr personal exposure, depending on the location of sampling (Eskilen et al 2002). Similarly, expert assessment of occupational EMF exposure is notoriously difficult (Kheifets et al 2009;Savitz, 1995;Miller and Green, 2010), while the inference of personal EMF exposure from appliances based on self-reported information is also known to be subject to considerable error (Savitz, 1995).…”