Acute lymphoblastic leukemia (ALL) accounts for 70 percent to 80 percent of all childhood leukemias and one-third of all childhood cancers in the United States. However, only a small proportion of cases have an identifiable cause. Beginning in 1979, some studies have suggested that magnetic fields (EMFs) may increase the risk for ALL while others have found no evidence for risk. EMFs are produced by power lines, electrical wiring and household electrical appliances.

An earlier report published in July 1997 in the New England Journal of Medicine showed little evidence that high-current power lines or high levels of magnetic fields measured in the home (produced mainly by outside power lines) were associated with an increased risk of ALL in children.

The current study, published in the May issue of Epidemiology, investigated whether use of household electrical appliances by the mother during pregnancy and by the child might be associated with an increased risk of ALL. Compared to background exposures from power lines, the contribution of appliances to a person's total exposure to EMF is thought to be small because most appliances are used for short periods of time and EMF exposures are elevated only when close to the appliance.

Two small studies previously examined the use of electrical appliances in relation to childhood leukemia, but could not examine details on patterns of use during pregnancy and childhood.

Both the 1997 report and the current study are a collaborative effort between National Cancer Institute (NCI) scientists and the Children's Cancer Group (CCG) to study residential magnetic field exposures and childhood ALL. The same cases and controls were used in both studies.

CCG, a multi-center network of pediatric oncologists and other researchers from 38 institutions and affiliated hospitals in the United States, is conducting an ongoing study to look at a wide range of possible causes (including EMF) of childhood ALL.

2. How was this study conducted?

The NCI and CCG collaborated on the study. Elizabeth Hatch, Ph.D., and Sholom Wacholder, Ph.D., of NCI's Division of Cancer Epidemiology and Genetics, analyzed the data on the appliances and Martha Linet, M.D., of NCI's Radiation Epidemiology Branch in the Division of Cancer Epidemiology and Genetics directed the study. The researchers compared the reported use of household electrical appliances of 640 children 14 years or younger diagnosed with ALL between l989 and l993 to the appliance use of 640 matched controls. Over half of the cases (59.4 percent) were less than 5 years old when they were diagnosed with ALL. The study participants lived in Illinois, Indiana, Iowa, Michigan, Minnesota, New Jersey, Ohio, Pennsylvania, and Wisconsin. The data are based on the mothers' responses to a detailed questionnaire administered in their homes. No measurements were taken of magnetic fields associated with the actual appliances used.

The mothers were asked to describe how often during their pregnancy and in which trimester specific appliances were used. The interview also included questions about the child's use of the appliances before the diagnosis of ALL, for cases and during a comparable time period for controls.

3. Which electric appliances were included in the study?

Electric appliances included in the study were electric blankets, mattress pads, heating pads, water beds, stereo or other sound systems, television and video games connected to a television, video machines located in arcades, computers, microwave ovens, sewing machines, hair dryers, curling irons, ceiling fans, humidifiers, night lights and electric clocks.

These appliances were chosen for several reasons. The level of magnetic fields from an appliance to which a person is actually exposed depends mainly on three factors: 1) distance from the appliance, 2) length of time that an appliance is used, and 3) strength of the magnetic field from the appliance. The level of magnetic fields from appliances falls off very rapidly with distance, so the likelihood of any exposure from appliances at a distance of more than three to four feet away is very small. Therefore, appliances that are generally used for a long time and close to the body, such as electric blankets, may contribute very much to a person's total magnetic field exposure. Other appliances, such as microwave ovens and hair dryers, may have high magnetic fields very close to the appliance. However, they are generally used for short periods of time and are unlikely to contribute to an individual's overall cumulative exposure. Finally, the researchers selected some appliances that have only a small likelihood of any exposure to magnetic fields (ceiling fans, night lights, and stereo systems) to assess the possibility that mothers of cases might report their exposure to appliances more thoroughly and accurately than mothers of controls.

4. What were the results?

Maternal exposure:
Overall, there was little evidence for any association between appliances used during pregnancy and the risk of childhood ALL. No significant association was found with the mothers time spent watching TV or distance from TV during pregnancy. Childhood ALL was also not associated with use of water beds, hair dryers, curling irons, electric clocks, microwave ovens, ceiling fans, sound systems with headsets, electric stoves, electric heat or air conditioning during pregnancy. Mothers of ALL cases were less likely to report having used a sewing machine during pregnancy. Sewing machines have been found to have rather high magnetic field exposures.

Mothers of ALL cases were slightly more likely to report having ever used electric blankets or mattress pads, heating pads, and humidifiers during pregnancy. However, these associations did not strengthen with increasing use of any of these appliances. For example, the relationship between childhood ALL and maternal electric blanket use was the same whether mothers reported that they used electric blankets less than once a week or more than six hours per day for more than two months during their pregnancy.

Mothers' patterns of use of electrical appliances during their pregnancy were the same in each trimester, with the exception of electric blankets and heating pads. For heating pads, an association was found for use by the mother only in the first trimester of pregnancy, whereas for electric blankets, an association was found only for use by the mother in the second and third trimesters.

Children's exposure:
For each appliance, the researchers examined whether an increased risk of ALL was associated with any use (versus no use) of the appliance since birth. In addition, researchers examined both the number of years and the frequency of use of the appliances during the year prior to diagnosis. The number of years that children had watched TV was not assessed, since virtually all children watched TV from a very early age.

The researchers found no increased use among cases for night lights, stereo systems without headsets, water beds, humidifiers, electric clocks, personal computers, microwave ovens, electric stoves, air conditioning, or electric heat.

Four appliances (electric blankets, hair dryers, video machines in arcades, and video games connected to TVs) were associated with ALL when any reported exposure (regardless of the amount of exposure) during the child's lifetime was examined. But only two appliances (video machines in arcades and stereo systems with head sets) showed an increased risk with the number of years of use. There were no trends of increasing risk with longer reported use for electric blankets; the association among children who were reported to use the blankets for less than one year was actually stronger than for those who were reported to use the blankets for three or more years.

When frequency of use during the year prior to diagnosis was examined, there was little indication of increasing risk with increasing reported use of the appliances. The researchers did find that mothers of ALL cases tended to report that their children watched more television than mothers of control children. However, the results were the same regardless of the reported distance between the television and the child's usual position while watching. If the association between television watching and ALL was due to an effect of magnetic fields, the association would be expected to be stronger for children who sit closer to the television.

The researchers also estimated the child's total exposure to electric blankets by summing the estimates of the number of hours that electric blankets were used during the pregnancy and child's life. There was no trend of increasing risk of ALL with increasing number of hours that an electric blanket was used.

5. What are the limitations of the study?

• No measurement was taken of magnetic fields associated with the actual appliances used.

• No detailed lifetime histories of use were collected because it was very difficult for mothers to recollect lifetime patterns of use of the appliances by their children.

• Except for television watching, no information was collected on the subject's distance from the appliance, which is a critical determinant of the strength of magnetic fields.

• With the exception of TV watching and playing video games connected to TV, data for each appliance were analyzed individually, instead of combining appliances to obtain an overall estimate of exposure to magnetic fields. The researchers did not analyze appliances jointly because not enough is known about magnetic fields from appliances to combine them in a meaningful way.

• There were some limitations to the accuracy of the data: 1) Children with ALL may change their behaviors following diagnosis and treatment (for example, they may increase sedentary activities, such as television watching), and it may be difficult for mothers to remember accurately their children's habits prior to the diagnosis of leukemia; 2) A mother whose child has suffered a life-threatening disease like ALL may be searching for a cause that led to the illness, and may remember more details than a control mother. Since electric blankets have been most frequently cited by the media as having high magnetic fields, they may be most prone to this kind of reporting bias; 3) Mothers were asked in two separate interviews about their use of electric blankets during pregnancy and their answers were often not consistent, suggesting that the data on electric blankets, as well as other appliances, may not be accurate; 4) Although the controls and cases were well matched by age, sex, and race, the control mothers were interviewed, on average, later than case mothers and thus had to remember farther back in time.

6. How was this study different from previous studies?

Two other studies have evaluated appliance use in relation to the risk of childhood leukemia. A 1990 study of 70 childhood leukemia cases in the Denver area suggested a link to prenatal electric blanket use. A second study in 1991 of 232 childhood leukemia cases in Los Angeles found associations with children's use of hair dryers and black and white televisions, but no information was provided on length of use or frequency of use of these appliances. Both studies were based on rather small numbers of cases and had limited ability to estimate exposure to appliances accurately.

The current study is the largest to date, covering a nine-state geographic area. It had the same eligibility requirements for cases and controls, and the participants were interviewed closer in time to diagnosis than previous studies.

7. What conclusions can be drawn from the results?

Although the results from this study are preliminary, the authors tentatively conclude that magnetic fields from electrical appliances are unlikely to increase the risk of childhood ALL. Further detailed studies with measurements would be needed to determine whether the findings, if replicated, could represent any effects of magnetic field exposure.

This conclusion is based on several factors: A major factor is the failure to see a consistent pattern of increasing risk with increasing exposures for most of the appliances. In addition, the results for television were similar regardless of the child's distance from the TV, suggesting that other factors associated with TV use (not magnetic fields) may have been responsible for this result. Preliminary findings from a small study that measured actual magnetic field levels at varying distances from TVs found that magnetic fields are nearly the same as background levels at distances of more than three or four feet from television sets. As the proportion of U.S. households with a television increased from 9 percent to 95 percent between 1950 and 1970, whereas there was no corresponding increase in childhood leukemia rates during this time period, suggesting that television watching is unlikely to be a causal factor. Children's exposure to other electrical appliances has also increased dramatically over the last 50 years. Finally, the companion study with the same cases and controls, failed to find consistent evidence based on measurements and wire coding (an indirect measure of magnetic fields based on the thickness, configuration, and distance from the home of nearby power lines) that high magnetic fields levels in the home (produced mainly by outside power lines) increase the risk for ALL.

Background on EMFs and childhood cancers:

8. What are EMFs? How could EMFs cause cancer?

Power lines, electrical wiring and appliances all produce electric and magnetic fields (EMFs) which are fields of electrical and magnetic force. The electric and magnetic fields produced by power lines and appliances have relatively long wavelengths compared to most other forms of electromagnetic fields. For example, radiation from power lines has wavelengths of more than 3,100 miles compared to microwaves with wavelengths of several inches, and X-rays and gamma rays which are about 4 billionth of an inch or smaller.

No form of electromagnetic energy at energy levels below those of X-rays and ultraviolet radiation has been shown to cause cancer. Only shorter wavelength electromagnetic fields are able to disrupt molecular bonds, potentially causing mutations and damage to the DNA. Therefore, EMFs from power lines or appliances are not capable of causing cancer by this means.

There are several theories about the mechanism by which EMFs may act on biological tissue. EMFs can induce weak electric currents in the bodies of people and animals. But the amount of the current, even directly beneath a large transmission line, is very small and seems too weak to penetrate cell membranes and cause damage inside cells. Some scientists argue that it is impossible for EMFs to have any important biological effects. Others, however, contend that EMFs may amplify normal electric currents in tissues and cells, or that EMFs may transfer energy to cell membranes and alter their function in some manner.

Presently, there is no consistent experimental data that prove that EMFs have any detrimental effects on living organisms. In spite of this, because of the heavy use of electrical power and new electric technologies in the U.S., public interest in EMF is high.

9. What is known about which household electrical appliances produce the highest sources of magnetic field exposures?

The level of exposure to the magnetic field of an appliance depends on the strength of the field it produces, how often and for how long the appliance is used, and the distance between the user and the appliance. The magnetic field strength from an appliance drops off very rapidly with distance. That means that the magnetic fields a few inches from an appliance are many times greater than at two feet away.

In general, magnetic fields generated by large household appliances tend to be lower than those from portable devices. This is probably because the motors in these appliances are better shielded. In addition, most motors on large appliances are placed at the bottom and rear of the cabinets.

In contrast, some smaller appliances have higher magnetic fields than larger ones. For example, the field for microwave ovens at distance of one foot is ten times higher than refrigerators. Also, can openers, mixers, blenders, vacuum cleaners, hair dryers, electric shavers, hand-held drills and saws generate higher than average magnetic fields. In general, the higher levels of magnetic fields from these appliances are due to the light-weight motors that have little magnetic shielding, or as in the case of microwave ovens, have large power transformers.

Electric blankets and mattress pads, although not commonly used, may contribute to an individual's overall cumulative exposure to magnetic fields because they are used close to the body for long time periods. Estimates of the intensity of magnetic field exposure of these products vary, but blankets manufactured after 1989 were redesigned to have much lower magnetic fields (by a factor of 10) than blankets made before this date.

10. How common is ALL?

Every year about 30 out of every 1 million children under age 15 in the United States are diagnosed with ALL and about five die from ALL. About 1600 children are expected to be diagnosed with ALL this year. The disease is most commonly diagnosed in white children under 5 years old. It is twice as common in white children as in black children and is slightly more common in boys than in girls.

ALL is much more treatable now than in the past. Most children with ALL can now be cured, and about 80 percent of patients under age 15 survive at least five years after diagnosis, compared with about 1 percent in the 1950s.

11. What are the known risk factors for ALL? What others have been proposed?

Only a few risk factors are known, although many have been proposed and studied. Children with Down syndrome have a greatly increased risk of ALL, reported to be 10 to 40 times the risk of other children. Other, more rare chromosomal and genetic abnormalities may also increase risk for ALL. Children whose mothers had diagnostic X-rays during pregnancy are about one-and-a-half times more likely to have ALL compared with children whose mothers had no X-rays. Risk factors that have been proposed, but not proven, include certain birth characteristics such as high birth weight; medical conditions or drugs affecting delivery; mothers prior reproductive problems such as repeated miscarriages; pesticides and other chemicals; certain viruses; and natural background ionizing radiation.

12. Are other studies of magnetic fields and childhood leukemia under way? Are studies underway of other risk factors for childhood leukemia?

Population-based studies of residential magnetic fields and childhood leukemia are under-way in Canada and the United Kingdom. The results are expected within one to two years.

The NCI/CCG magnetic field study is part of a larger CCG study of more than 1,900 children diagnosed with ALL between 1989 and 1993, and 1,900 controls. This ongoing study is designed to evaluate the risk of ALL associated with a wide range of factors, including maternal diseases and medication use during pregnancy, childhood infectious and other diseases, parental occupational exposures, prenatal and postnatal environmental exposures, parental smoking and alcohol use, lifestyle, and genetic factors. Results are expected in about two years.

13. Where can I find more out about EMFs?

The National Institute of Environmental Health Sciences maintains an EMF home page at the website address: http://www.niehs.nih.gov/emfrapid/home.htm. Both the booklet "Questions and Answers About EMF" and their newest publication, "EMF in the Workplace - Q&A." are available on this website.

The two studies conducted by NCI/CCG are:

• The current study is titled "Association Between Childhood Acute Lymphoblastic Leukemia and Use of Electrical Appliances During Pregnancy and Childhood." The authors are Elizabeth E. Hatch, Martha S. Linet, Ruth A. Kleinerman, Robert E. Tarone, Richard K. Severson, Charleen T. Hartsock, Carol Haines, William T. Kaune, Dana Friedman, Leslie L. Robison, and Sholom Wacholder. Epidemiology, May 1998.

• The previous study is titled "Residential Magnetic Field Exposures and Childhood Acute Lymphoblastic Leukemia." The authors are Martha S. Linet, Elizabeth E. Hatch, Ruth A. Kleinerman, Leslie L. Robison, Willian T. Kaune, Dana R. Friedman, Richard K. Severson, Carol M. Haines, Charlene T. Hartsock, Shelly Niwa, Sholom Wacholder, and Robert E. Tarone. New England Journal of Medicine, July 3, 1997

For more information about cancer visit NCI's Website for patients, public and the mass media at http://rex.nci.nih.gov or NCI's main website at http://www.nci.nih.gov.