To address widespread concerns about cancer risks from residential magnetic field exposures, investigators from NCI and from the Childrens Cancer Group have conducted a case-control study of childhood acute lymphoblastic leukemia. Measurement data were collected in current and former residences for 638 cases and 620 controls under age 15, and interview data on prenatal and postnatal exposures to electrical appliances for 640 cases and 640 matched control children. Measurements of magnetic fields in the childs bedroom, the family room, the kitchen and the area immediately outside the front door were used to assess exposure, as were exposures based on the distance and configuration of power lines (e.g., wire codes) near the subjects home. The exposure of the mother during pregnancy was similarly evaluated. Papers were published in 1997 on the methodology for exposure assessment (Kleinerman et al. Epidemiology 1997;8:575-83)and the finding of little evidence that risk of acute lymphoblastic leukemia was increased among children residing in homes with high measured time- weighted average magnetic fields or high wire code levels (Linet et al. N Engl J Med 1997;337:1-7). In 1998 the investigators reported excellent reproducibility between the two wire coders and similar correlation between wire codes levels and magnetic field measurements as found in earlier U.S. studies (Tarone et al. Occup Environ Med 1998;55:333-9). Childhood acute lymphoblastic leukemia was found to be increased among children whose mothers used an electric blanket or mattress pad during pregnancy, and among children exposed postnatally to electric blankets and mattress pads, hair dryers, arcade video machines, and television-connected video games (Hatch et al. Epidemiology 1998;9:234-45). No dose-response was found for years of use or frequency of use for most of the appliances. In an analysis to evaluate the impact of measuring only one home, then imputing or not imputing magnetic field estimates for the second home among children residing in two homes, a marked attenuation in risk estimates was observed for children residing in homes with high exposure when measurements considered were restricted to one home, particularly when the only measurements utilized were those taken in the childs residence at diagnosis and imputed values were used for homes lived in previously (Baris et al. Occup Environ Med 1999;56:562-66). These findings argue against attempting to estimate lifetime magnetic field exposure by using imputed values to fill in gaps from unmeasured homes resided in previously. Papers submitted for publication during 1999-Aug 9 and currently in press include: evaluation of the relationship of childhood acute lymphoblastic leukemia with 1)avariety of exposure metrics other than time-weighted average levels for measured residential magnetic field levels (including threshold values, peak values, rate of change, and other metrics); 2) distance of power lines in relation to residences and characteristics of power lines; and 3) possible effects of selection bias and potential confounders of residential magnetic field measurements (the latter including sociodemographic factors, residential characteristics, and other prenatal and postnatal exposures potentially associated with childhood acute lymphoblastic leukemia). Data collection was completed late in 1998 and analysis is currently underway for a case-control study of adult brain cancer (glioma, meningioma and acoustic neuroma) to investigate cellular telephones and other forms of wireless communication devices, and the role of occupational and other possible risk factors. In collaboration with investigators in Boston, Phoenix and Pittsburgh, we enrolled approximately 800 cases (including approximately 500 gliomas, 200 meningiomas, and 100 acoustic neuromas) and 800 controls (the latter including patients hospitalized with a variety of nonmalignant conditions.) It is anticipated that papers will be submitted in 2000 on risk of brain tumors according to type of brain tumor in relation to: sociodemographic characteristics; use of cellular telephones and other wireless communication devices; and use of electrical appliances. Analysis will continue on risk of brain tumors according to type of brain tumor in relation to: occupational exposures to solvents; dietary use of artificial sweeteners; brain tumors and other types of cancer in first-degree relatives. Extraction of DNA from blood samples collected from participating subjects will continue as will analyses of the DNA for the family of GST polymorphisms. Characterization of individual subjects DNA repair mechanisms will be initiated in collaboration with investigators at MD Anderson Comprehensive Cancer Center. In response to concerns about cancer risk from non-ionizing radiation of microwave and radio frequencies, an earlier study of cancer mortality among Korean War Naval Service veterans exposed to microwave radiation has been reactivated. Two cohorts of 20,000 men each, who served as shipboard radar operators and maintenance workers, respectively, during their Navy service were followed for cancer mortality through 1993. Results on the relationship between cancer risk and non-ionizing radiation dose are expected to be submitted late in 1999 or early in 2000.
