Probabilistic approach to ranking sources of uncertainty in ELF magnetic field exposure limits

Dosimetric Uncertainties: Magnetic Field Coupling to Peripheral Nerve

The International Commission on Non-ionizing Radiation Protection (ICNIRP) and the Institute for Electrical and Electronic Engineers (IEEE) have established magnetic field exposure limits for the general public between 400 Hz (ICNIRP)/759 Hz (IEEE) and 100 kHz to protect against adverse effects associated with peripheral nerve stimulation (PNS). Despite apparent common purpose and similarly stated principles, the two sets of limits diverge between 3.35-100 kHz by a factor of about 7.7 with respect to PNS. To address the basis for this difference and the more general issue of dosimetric uncertainty, this paper combines experimental data of PNS thresholds derived from human subjects exposed to magnetic fields together with published estimates of induced in situ electric field PNS thresholds to evaluate dosimetric relationships of external magnetic fields to induced fields at the threshold of PNS and the uncertainties inherent to such relationships. The analyses indicate that the logarithmic range of magnetic field thresholds constrains the bounds of uncertainty of in situ electric field PNS thresholds and coupling coefficients related to the peripheral nerve (the coupling coefficients define the dosimetric relationship of external field to induced electric field). The general public magnetic field exposure limit adopted by ICNIRP uses a coupling coefficient that falls above the bounds of dosimetric uncertainty, while IEEE's is within the bounds of uncertainty toward the lower end of the distribution. The analyses illustrate that dosimetric estimates can be derived without reliance on computational dosimetry and the associated values of tissue conductivity. With the limits now in place, investigative efforts would be required if a field measurement were to exceed ICNIRP's magnetic field limit (the reference level), even when there is a virtual certainty that the dose limit (the basic restriction) has not been exceeded. The constraints on the range of coupling coefficients described in this paper could facilitate a re-evaluation of ICNIRP and IEEE dose and exposure limits and possibly lead toward harmonization.

Accounting for human variability and sensitivity in setting standards for electromagnetic fields

Biological sensitivity and variability are key issues for risk assessment and standard setting. Variability encompasses general inter-individual variations in population responses, while sensitivity relates to unusual or extreme responses based on genetic, congenital, medical, or environmental conditions. For risk assessment and standard setting, these factors affect estimates of thresholds for effects and dose-response relationships and inform efforts to protect the more sensitive members of the population, not just the typical or average person. While issues of variability and sensitivity can be addressed by experimental and clinical studies of electromagnetic fields, investigators have paid little attention to these important issues. This paper provides examples that illustrate how default assumptions regarding variability can be incorporated into estimates of 60-Hz magnetic field exposures with no risk of cardiac stimulation and how population thresholds and variability of peripheral nerve stimulation responses at 60-Hz can be estimated from studies of pulsed gradient magnetic fields in magnetic resonance imaging studies. In the setting of standards for radiofrequency exposures, the International Commission for Non-Ionizing Radiation Protection uses inter-individual differences in thermal sensitivity as one of the considerations in the development of "safety factors." However, neither the range of sensitivity nor the sufficiency or excess of the 10-fold and the additional 5-fold safety factors have been assessed quantitatively. Data on the range of responses between median and sensitive individuals regarding heat stress and cognitive function should be evaluated to inform a reassessment of these safety factors and to identify data gaps.

Comparisons of peak SAR levels in concentric sphere head models of children and adults for irradiation by a dipole at 900 MHz

The aim of this study is to examine the scale and significance of differences in peak specific energy absorption rate (SAR) in the brains of children and adults exposed to radiofrequency emissions from mobile phones. Estimates were obtained by method of multipole analysis of a three layered (scalp/cranium/brain) spherical head exposed to a nearby 0.4 lambda dipole at 900 MHz. A literature review of head parameters that influence SAR induction revealed strong indirect evidence based on total body water content that there are no substantive age-related changes in tissue conductivity after the first year of life. However, it was also found that the thickness of the ear, scalp and cranium do decrease on average with decreasing age, though individual variability within any age group is very high. The model analyses revealed that compared to an average adult, the peak brain 10 g averaged SAR in mean 4, 8, 12 and 16 year olds (yo) is increased by a factor of 1.31, 1.23, 1.15 and 1.07, respectively. However, contrary to the expectations of a recent prominent expert review, the UK Stewart Report, the relatively small scale of these increases does not warrant any special precautionary measures for child mobile phone users since: (a) SAR testing protocols as contained in the CENELEC (2001) standard provide an additional safety margin which ensures that allowable localized SAR limits are not exceeded in the brain; (b) the maximum worst case brain temperature rise (approximately 0.13 to 0.14 degrees C for an average 4 yo) in child users of mobile phones is well within safe levels and normal physiological parameters; and (c) the range of age average increases in children is less than the expected range of variation seen within the adult population.

Dealing with uncertainty in formulating occupational and public exposure limits

Scientists are forced to address many sources of uncertainty in the setting of exposure limits. This paper discusses the rationale for some of the approaches that have been used for many years to address uncertainties encountered in the setting of occupational and public exposure limits to protect against non-cancer effects of chemicals. These approaches are quite different from the methodology that has been adopted for the setting of exposure limits to power-frequency electric and magnetic fields to avoid adverse effects of induced electric fields and currents. The limitations of the non-standard electric and magnetic fields guideline methodology are described. To address these limitations, a probabilistic approach to developing exposure limits for electric and magnetic fields has been proposed. As a demonstration, this approach has been applied to the estimation of magnetic field exposures that would not pose significant risks of cardiac stimulation from electric fields induced in the torso. The results of this probabilistic analysis demonstrate the importance of explicitly characterizing potential sources of uncertainty and variability. Furthermore, the use of fixed, single point estimates of dosimetric parameters by electric and magnetic fields guideline-setting organizations provides more than adequate protection to individuals from the risk of cardiac stimulation. The mean magnetic field level that would increase the risk of cardiac stimulation is shown by this approach to be about three times higher than is predicted by dosimetric formulas used by the ACGIH and ICNIRP. If input data for parameters affecting the risk of adverse effects with lower thresholds, i.e., nerve stimulation, are developed and validated, this method could be used in the setting of electric and magnetic fields standards based more strongly on quantitative assessments of experimental data.

Health effects relevant to the setting of EMF exposure limits

To date, electric and magnetic exposure limits for frequencies below 100 kHz have been based on vaguely defined neurobiological responses to electric fields induced in tissues in vivo by magnetic fields and on perceptual responses to external electric fields. Advances in tissue dosimetry, risk assessment methods, and biological research on stimulation thresholds and mechanisms are providing new bases for exposure limits. This paper reviews the historical basis for current electric and magnetic exposure limits in preparation for the development of the "next generation" of electric and magnetic occupational and public exposure guidelines. This is followed by an overview of reported neurobiological effects of electric and magnetic stimulation that should be considered in new exposure guidelines. For magnetic fields, there is stronger evidence for setting exposure limits to protect against adverse effects of nerve stimulation than for protecting against visual magnetophosphenes. Magnetophosphenes are not adverse, and the evidence that these perceptual responses of the eye are a precursor or surrogate for other adverse neurologic responses is weak. Rather than relying just on theoretical models to set exposure limits, data from human subjects exposed to pulsed magnetic fields should be used to estimate nerve stimulation thresholds. Such data can provide a solid basis for setting magnetic field exposure limits if uncertainties in the data and inter-individual variability are addressed. Research on sensory perception, spontaneous and evoked potentials, and epidemiologic studies of neuropsychiatric conditions in electric and magnetic exposed populations does not suggest a need for lower exposure limits. However, a report that a 60-mT magnetic field (below the threshold for peripheral nerve stimulation) produces prolonged alterations of brain excitability and "indisposure" of subjects should be investigated in future research.

A transmission utility's experience of applying EMF exposure standards

Exposure standards for power-frequency electric and magnetic fields are often structured in terms of basic restrictions and investigation levels. For uniform exposures, investigation levels help the user operate within the standard without concern about exceeding the basic restriction. For non-uniform fields, however, numerical calculations of induced currents in the body may be necessary to determine compliance with the basic restriction. Utilities have strong incentives to apply exposure standards so that they do not impose unnecessary and unjustified constraints on their operations. They are therefore likely to be prepared to invest considerable effort in assessing compliance. However, the standards are based on round numbers, the values of tissue conductivity are not well characterized, and the calculations needed to apply the standards are still evolving. This suggests that, scientifically, this level of effort devoted to distinguishing compliant and non-compliant exposure scenarios is not justifiable.

Assessment of guidelines for limiting exposures to emf using methods of probabilistic risk analysis

Allowable limits of human exposure to radiofrequency fields commonly include a "factor of safety," typically between 10 to 50, which is somewhat arbitrary. The broad objective in our work is to assess radiofrequency exposure limits, hazard thresholds, and safety factors using methods of probabilistic risk analysis. We focus our analysis on the variables affecting peak radiofrequency specific energy absorption rate (SAR) values in the brain from digital mobile telephones operating at approximately 900 MHz. As SAR is defined as a product of positive random variables, it is not unreasonable to assume that SAR has a lognormal distribution. Our analysis of component SAR variables such as conductivity and permittivity of grey brain matter and radiated field strengths using experimental and numerical modeling data strongly supports our hypothesis that SAR values are distributed lognormally. It then follows that the probability that the SAR exceeds a certain threshold can be derived directly and is shown to be very low for handset SARs relative to presently allowable standard limits.