We strongly agree with the authors that, although there is no compelling evidence to suggest that nonionizing electromagnetic fields represent a public health hazard, there is sufficient evidence of magnetic- and electric field-induced biologic effects to continue scientific investigation of this issue.
Just as it would be inadequate to investigate the action of a few chemicals and decide that there is no harm possible from any chemical in the environment, a more informed strategy is needed in studies of electric and magnetic fields. Salvatore et al rightly suggest that nonionizing electromagnetic fields are characterized by a number of parameters. In fact, because fields are uniquely characterized only when the value of each parameter (frequency, magnitude, orientation, duration, and so on) is defined, it is generally not realistic to draw direct analogies between field exposure and single chemical exposures. For example, "high exposure" to fields refers only to the measured field magnitude, and not necessarily to its biologic effectiveness.
Although epidemiologic studies have, in some cases, shown a correlation between nonionizing electromagnetic field exposure and the development of certain cancers, the studies have not yet clearly identified specific details of nonionizing electromagnetic fields that might be causative. Laboratory studies, conducted under well-defined conditions, may be useful for determining whether the exposure categories used in epidemiologic studies are adequate, or how they can be improved.
Essentially, all studies of nonionizing electromagnetic fields focus on magnetic field effects because magnetic fields are rarely shielded (attenuated) in the inhabited environment (home, work, or transportation). Furthermore, laboratory studies have shown that magnetic fields can have direct action on biologic processes (eg, Blackman et al ), and epidemiologic data appear to be more strongly correlated with estimates of magnetic fields than with electric field estimates .
The Need to Define "Dose"
Until now, most laboratory animal studies using particular cancer models with various nonionizing electromagnetic field exposures showed negative results. However, at least two models (a skin tumor model [3,4], and a breast tumor model [5-7]have recently shown positive effects under specific, precise nonionizing electromagnetic field treatment conditions. These studies suggest the need for refining what is meant by "dose" under nonionizing electromagnetic field exposures.
Three particularly promising areas of laboratory research for providing critical focus to research are:
1. Investigations of subtle distinctions in field exposure conditions that are necessary to elicit biologic changes,
2. Determination of whether and how certain magnetic field exposures may augment the action of chemicals known to influence cancer development, and
3. Studies examining how certain magnetic field exposures may induce changes in the production, availability, and action of critical oncostatic hormones, such as melatonin(Drug information on melatonin).
Characterizing Field Exposure Conditions
Recent laboratory research suggests that nonionizing electromagnetic field interactions with biologic systems may be resonance-based to some extent. Several resonance interaction models, with different degrees of predictivity and validation, exist. For example, the ion cyclotron resonance (ICR) model  identifies combinations of alternating-current (AC) frequency and direct-current (DC) magnetic flux density required to establish resonance conditions for biologically active ions. Lednev  augmented the ICR model by including the predicted influence of the AC magnetic field flux density (Bac) on the resonance process.
More recently, the ion parametric resonance (IPR) model  identified a Bac-based response form different from the Lednev model, and considers the combined influence of multiple ion resonances at any given exposure condition. This IPR model has growing experimental support [11,12]. Other well-established magnetic resonance models, such as nuclear magnetic resonance and electron spin resonance, also appear to be useful in identifying the precise conditions of nonionizing electromagnetic field exposure to create biologic changes . In each model, fundamental magnetic field conditions must be closely controlled, including AC frequency(ies), AC and DC flux densities, and the relative orientation between the AC and DC magnetic field vectors.
Nonionizing electromagnetic fields may also interact through induced electric current in conductive biologic materials. It has been suggested, for example, that externally generated electric transients may lead to more biologically significant induced currents within the body.
Other studies examined the time signature of exposure conditions. Historically, epidemiologic studies assumed that time-weighted average (TWA) magnetic flux density was a sufficient measure of exposure. The most compelling of these studies  showed that averaging exposure intensity over 1-year periods produced a higher correlation with disease incidence than spot measurements taken for only a 15-minute-period. However, several recent studies (for example, Morgan and Nair ), challenge the use of the TWA metric, and have searched for more appropriate measures.
Synergy With Chemicals
The biologic effects of nonionizing electromagnetic field exposures may be influenced by the presence of certain chemicals within the body. These include environmental chemicals, specifically, tumor promoters, and indigenous signaling molecules, such as hormones. Tumor promoters are believed to influence biologic systems only when their concentrations are above a certain threshold value. As noted by Salvatore et al, it is possible that magnetic fields can change (reduce) that apparent threshold.
Hormones normally carry information between tissue systems to provide timely stimuli, causing specific tissue/cell metabolic responses. For example, the pineal hormone, melatonin, dramatically influences the body's circadian clock. Melatonin is the most potent free-radical scavenger in the body (with potential implications for protection against cancer-initiating events)  and also acts as a modulator of gap junction intercellular communication , an information channel that both coordinates cellular functions and allows neighboring cells to control the growth of "initiated" cells. Recent studies suggest that nonionizing electromagnetic field exposures may control the amount of circulating melatonin and influence its specific functions, making the biologic system more susceptible to disease.
Because nonionizing electromagnetic fields cannot be characterized solely by their magnitude, they are more complicated than earlier studies have assumed, and hence, the single chemical analogy is inappropriate. Present research suggests that (1) specific combinations of nonionizing electromagnetic fields may be more effective at creating biologic effects than others, and (2) synergistic effects between chemicals and nonionizing electromagnetic fields are possible within a biologic system. It is too early to claim categorically that nonionizing electromagnetic fields do or do not have carcinogenic properties. More directed research is clearly required to elucidate the biologic significance of ubiquitous nonionizing electromagnetic fields exposure(s).