Breast cancer risk and protracted low-to-moderate dose occupational radiation exposure in the US Radiologic Technologists Cohort, 1983-2008.

BACKGROUND: Although high-dose ionising radiation is associated with increased breast cancer risks, the association with protracted low-dose-rate exposures remains unclear. The US Radiologic Technologist study provides an opportunity to examine the association between low-to-moderate dose radiation and breast cancer incidence and mortality. METHODS: One thousand nine hundred and twenty-two self-reported first primary cancers were diagnosed during 1983-2005 among 66 915 female technologists, and 586 breast cancer deaths occurred during 1983-2008 among 83 538 female cohort members. Occupational breast dose estimates were based on work histories, historical data, and, after the mid-1970s, individual film badge measurements. Excess relative risks were estimated using Poisson regression with birth cohort stratification and adjustment for menopause, reproductive history, and other risk factors. RESULTS: Higher doses were associated with increased breast cancer incidence, with an excess relative risk at 100 mGy of 0.07 (95% confidence interval (CI): -0.005 to 0.19). Associations were strongest for technologists born before 1930 (excess relative risk at 100 mGy=0.16; 95% CI: 0.03-0.39) with similar patterns for mortality among technologists born before 1930. CONCLUSIONS: Occupational radiation to the breast was positively associated with breast cancer risk. The risk was more pronounced for women born before 1930 who began working before 1950 when mean annual doses (37 mGy) were considerably higher than in later years (1.3 mGy). However, because of the uncertainties and possible systematic errors in the occupational dose estimates before 1960, these findings should be treated with caution.

Central adiposity, obesity during early adulthood, and pancreatic cancer mortality in a pooled analysis of cohort studies.

BACKGROUND: Body mass index (BMI), a measure of obesity typically assessed in middle age or later, is known to be positively associated with pancreatic cancer. However, little evidence exists regarding the influence of central adiposity, a high BMI during early adulthood, and weight gain after early adulthood on pancreatic cancer risk. DESIGN: We conducted a pooled analysis of individual-level data from 20 prospective cohort studies in the National Cancer Institute BMI and Mortality Cohort Consortium to examine the association of pancreatic cancer mortality with measures of central adiposity (e.g. waist circumference; n = 647 478; 1947 pancreatic cancer deaths), BMI during early adulthood (ages 18-21 years) and BMI change between early adulthood and cohort enrollment, mostly in middle age or later (n = 1 096 492; 3223 pancreatic cancer deaths). Multivariable hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using Cox proportional hazards regression models. RESULTS: Higher waist-to-hip ratio (HR = 1.09, 95% CI 1.02-1.17 per 0.1 increment) and waist circumference (HR = 1.07, 95% CI 1.00-1.14 per 10 cm) were associated with increased risk of pancreatic cancer mortality, even when adjusted for BMI at baseline. BMI during early adulthood was associated with increased pancreatic cancer mortality (HR = 1.18, 95% CI 1.11-1.25 per 5 kg/m(2)), with increased risk observed in both overweight and obese individuals (compared with BMI of 21.0 to <23 kg/m(2), HR = 1.36, 95% CI 1.20-1.55 for BMI 25.0 < 27.5 kg/m(2), HR = 1.48, 95% CI 1.20-1.84 for BMI 27.5 to <30 kg/m(2), HR = 1.43, 95% CI 1.11-1.85 for BMI ≥30 kg/m(2)). BMI gain after early adulthood, adjusted for early adult BMI, was less strongly associated with pancreatic cancer mortality (HR = 1.05, 95% CI 1.01-1.10 per 5 kg/m(2)). CONCLUSIONS: Our results support an association between pancreatic cancer mortality and central obesity, independent of BMI, and also suggest that being overweight or obese during early adulthood may be important in influencing pancreatic cancer mortality risk later in life.

Extremely low-frequency magnetic fields and survival from childhood acute lymphoblastic leukemia: an international follow-up study.

A previous US study reported poorer survival in children with acute lymphoblastic leukemia (ALL) exposed to extremely low-frequency magnetic fields (ELF-MF) above 0.3 µT, but based on small numbers. Data from 3073 cases of childhood ALL were pooled from prospective studies conducted in Canada, Denmark, Germany, Japan, UK and US to determine death or relapse up to 10 years from diagnosis. Adjusting for known prognostic factors, we calculated hazard ratios (HRs) and 95% confidence intervals (CI) for overall survival and event-free survival for ELF-MF exposure categories and by 0.1 µT increases. The HRs by 0.1 µT increases were 1.00 (CI, 0.93-1.07) for event-free survival analysis and 1.04 (CI, 0.97-1.11) for overall survival. ALL cases exposed to >0.3 µT did not have a poorer event-free survival (HR=0.76; CI, 0.44-1.33) or overall survival (HR=0.96; CI, 0.49-1.89). HRs varied little by subtype of ALL. In conclusion, ELF-MF exposure has no impact on the survival probability or risk of relapse in children with ALL.

Risk of treatment-related esophageal cancer among breast cancer survivors.

BACKGROUND: Radiotherapy for breast cancer may expose the esophagus to ionizing radiation, but no study has evaluated esophageal cancer risk after breast cancer associated with radiation dose or systemic therapy use. DESIGN: Nested case-control study of esophageal cancer among 289 748 ≥5-year survivors of female breast cancer from five population-based cancer registries (252 cases, 488 individually matched controls), with individualized radiation dosimetry and information abstracted from medical records. RESULTS: The largest contributors to esophageal radiation exposure were supraclavicular and internal mammary chain treatments. Esophageal cancer risk increased with increasing radiation dose to the esophageal tumor location (P(trend )< 0.001), with doses of ≥35 Gy associated with an odds ratio (OR) of 8.3 [95% confidence interval (CI) 2.7-28]. Patients with hormonal therapy ≤5 years preceding esophageal cancer diagnosis had lower risk (OR = 0.4, 95% CI 0.2-0.8). Based on few cases, alkylating agent chemotherapy did not appear to affect risk. Our data were consistent with a multiplicative effect of radiation and other esophageal cancer risk factors (e.g. smoking). CONCLUSIONS: Esophageal cancer is a radiation dose-related complication of radiotherapy for breast cancer, but absolute risk is low. At higher esophageal doses, the risk warrants consideration in radiotherapy risk assessment and long-term follow-up.

Leukemia-related chromosomal loss detected in hematopoietic progenitor cells of benzene-exposed workers.

Benzene exposure causes acute myeloid leukemia and hematotoxicity, shown as suppression of mature blood and myeloid progenitor cell numbers. As the leukemia-related aneuploidies monosomy 7 and trisomy 8 previously had been detected in the mature peripheral blood cells of exposed workers, we hypothesized that benzene could cause leukemia through the induction of these aneuploidies in hematopoietic stem and progenitor cells. We measured loss and gain of chromosomes 7 and 8 by fluorescence in situ hybridization in interphase colony-forming unit-granulocyte-macrophage (CFU-GM) cells cultured from otherwise healthy benzene-exposed (n=28) and unexposed (n=14) workers. CFU-GM monosomy 7 and 8 levels (but not trisomy) were significantly increased in subjects exposed to benzene overall, compared with levels in the control subjects (P=0.0055 and P=0.0034, respectively). Levels of monosomy 7 and 8 were significantly increased in subjects exposed to <10 p.p.m. (20%, P=0.0419 and 28%, P=0.0056, respectively) and ≥ 10 p.p.m. (48%, P=0.0045 and 32%, 0.0354) benzene, compared with controls, and significant exposure-response trends were detected (P(trend)=0.0033 and 0.0057). These data show that monosomies 7 and 8 are produced in a dose-dependent manner in the blood progenitor cells of workers exposed to benzene, and may be mechanistically relevant biomarkers of early effect for benzene and other leukemogens.

Mobile phone use and glioma risk: comparison of epidemiological study results with incidence trends in the United States.

OBJECTIVE: In view of mobile phone exposure being classified as a possible human carcinogen by the International Agency for Research on Cancer (IARC), we determined the compatibility of two recent reports of glioma risk (forming the basis of the IARC's classification) with observed incidence trends in the United States. DESIGN: Comparison of observed rates with projected rates of glioma incidence for 1997-2008. We estimated projected rates by combining relative risks reported in the 2010 Interphone study and a 2011 Swedish study by Hardell and colleagues with rates adjusted for age, registry, and sex; data for mobile phone use; and various latency periods. SETTING: US population based data for glioma incidence in 1992-2008, from 12 registries in the Surveillance, Epidemiology, and End Results (SEER) programme (Atlanta, Detroit, Los Angeles, San Francisco, San Jose-Monterey, Seattle, rural Georgia, Connecticut, Hawaii, Iowa, New Mexico, and Utah). PARTICIPANTS: Data for 24,813 non-Hispanic white people diagnosed with glioma at age 18 years or older. RESULTS: Age specific incidence rates of glioma remained generally constant in 1992-2008 (-0.02% change per year, 95% confidence interval -0.28% to 0.25%), a period coinciding with a substantial increase in mobile phone use from close to 0% to almost 100% of the US population. If phone use was associated with glioma risk, we expected glioma incidence rates to be higher than those observed, even with a latency period of 10 years and low relative risks (1.5). Based on relative risks of glioma by tumour latency and cumulative hours of phone use in the Swedish study, predicted rates should have been at least 40% higher than observed rates in 2008. However, predicted glioma rates based on the small proportion of highly exposed people in the Interphone study could be consistent with the observed data. Results remained valid if we used either non-regular users or low users of mobile phones as the baseline category, and if we constrained relative risks to be more than 1. CONCLUSIONS: Raised risks of glioma with mobile phone use, as reported by one (Swedish) study forming the basis of the IARC's re-evaluation of mobile phone exposure, are not consistent with observed incidence trends in US population data, although the US data could be consistent with the modest excess risks in the Interphone study.

Childhood cancer in the offspring born in 1921-1984 to US radiologic technologists.

We examined the risk of childhood cancer (<20 years) among 105 950 offspring born in 1921-1984 to US radiologic technologist (USRT) cohort members. Parental occupational in utero and preconception ionising radiation (IR) testis or ovary doses were estimated from work history data, badge dose data, and literature doses (the latter doses before 1960). Female and male RTs reported a total of 111 and 34 haematopoietic malignancies and 115 and 34 solid tumours, respectively, in their offspring. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated using Cox proportional hazards regression. Leukaemia (n=63) and solid tumours (n=115) in offspring were not associated with maternal in utero or preconception radiation exposure. Risks for lymphoma (n=44) in those with estimated doses of <0.2, 0.2-1.0, and >1.0 mGy vs no exposure were non-significantly elevated with HRs of 2.3, 1.8, and 2.7. Paternal preconception exposure to estimated cumulative doses above the 95th percentile (82 mGy, n=6 cases) was associated with a non-significant risk of childhood cancer of 1.8 (95% CI 0.7-4.6). In conclusion, we found no convincing evidence of an increased risk of childhood cancer in the offspring of RTs in association with parental occupational radiation exposure.

Antibiotic use and risk of non-Hodgkin's lymphoma: a population-based case-control study.

Antibiotic use in 759 non-Hodgkin's lymphoma (NHL) patients and 589 controls was compared. Neither total antibiotic use (odds ratio=0.7, 95% confidence interval=0.5-1.2), nor antibiotic use by site, was associated with total NHL, or NHL subtypes. There were no trends with frequency or age at first use (P trend=0.23 and 0.26, respectively).

Single-nucleotide polymorphisms in selected cytokine genes and risk of adult glioma.

A role of immunological factors in glioma etiology is suggested by reports of an inverse relationship with history of allergy or autoimmune disease. To test whether single-nucleotide polymorphisms (SNPs) in cytokine genes were related to risk of adult glioma, we genotyped 11 SNPs in seven cytokine genes within a hospital-based study conducted by the National Cancer Institute and an independent, population-based study by the National Institute for Occupational Safety and Health (overall 756 cases and 1190 controls with blood samples). The IL4 (rs2243248, -1098T>G) and IL6 (rs1800795, -174G>C) polymorphisms were significantly associated with risk of glioma in the pooled analysis (P trend = 0.006 and 0.04, respectively), although these became attenuated after controlling for the false discovery rate (P trend = 0.07 and 0.22, respectively). Our results underscore the importance of pooled analyses in genetic association studies and suggest that SNPs in cytokine genes may influence susceptibility to glioma.

Body mass index and all-cause mortality in a nationwide US cohort.

OBJECTIVE: To investigate whether the nature of the relationship between body mass index (BMI (kg/m2)) and all-cause mortality is direct, J- or U-shaped, and whether this relationship changes as people age. DESIGN: Prospective nationwide cohort study of US radiologic technologists (USRT). SUBJECTS: Sixty-four thousand seven hundred and thirty-three female and 19 011 male certified radiation technologists. METHODS: We prospectively followed participants from the USRT study who completed a mail survey in 1983-1989 through 2000. During an average of 14.7 years of follow-up or 1.23 million person-years, 2278 women and 1495 men died. Using Cox's proportional-hazards regression analyses, we analyzed the relationship between BMI and all-cause mortality by gender and by age group (<55 years; > or = 55 years). We also examined risk in never-smokers after the first 5 years of follow-up to limit bias owing to the confounding effects of smoking and illness-related weight loss on BMI and mortality. RESULTS: Risks were generally J-shaped for both genders and age groups. When we excluded smokers and the first 5 year of follow-up, risks were substantially reduced in those with low BMIs. In never-smoking women under the age of 55 years (excluding the initial 5-year follow-up period), risk rose as BMI increased above 21.0 kg/m2, whereas in older women, risk increased beginning at a higher BMI (> or = 25.0 kg/m2). Among younger men who never smoked (excluding the initial 5-year follow-up period), risk began to rise above a BMI of 23.0 kg/m2, whereas in older men, risk did not begin to increase until exceeding a BMI of 30.0 kg/m2. CONCLUSIONS: In younger/middle-aged, but not older, women and men, mortality risks appear directly related to BMI. The more complicated relationship between BMI and mortality in older subjects suggests the importance of assessing whether other markers of body composition better explain mortality risk in older adults.

Incidence of haematopoietic malignancies in US radiologic technologists.

BACKGROUND: There are limited data on risks of haematopoietic malignancies associated with protracted low-to-moderate dose radiation. AIMS: To contribute the first incidence risk estimates for haematopoietic malignancies in relation to work history, procedures, practices, and protective measures in a large population of mostly female medical radiation workers. METHODS: The investigators followed up 71,894 (77.9% female) US radiologic technologists, first certified during 1926-80, from completion of a baseline questionnaire (1983-89) to return of a second questionnaire (1994-98), diagnosis of a first cancer, death, or 31 August 1998 (731,306 person-years), whichever occurred first. Cox proportional hazards regression was used to compute risks. RESULTS: Relative risks (RR) for leukaemias other than chronic lymphocytic leukaemia (non-CLL, 41 cases) were increased among technologists working five or more years before 1950 (RR = 6.6, 95% CI 1.0 to 41.9, based on seven cases) or holding patients 50 or more times for x ray examination (RR = 2.6, 95% CI 1.3 to 5.4). Risks of non-CLL leukaemias were not significantly related to the number of years subjects worked in more recent periods, the year or age first worked, the total years worked, specific procedures or equipment used, or personal radiotherapy. Working as a radiologic technologist was not significantly linked with risk of multiple myeloma (28 cases), non-Hodgkin's lymphoma (118 cases), Hodgkin's lymphoma (31 cases), or chronic lymphocytic leukaemia (23 cases). CONCLUSION: Similar to results for single acute dose and fractionated high dose radiation exposures, there was increased risk for non-CLL leukaemias decades after initial protracted radiation exposure that likely cumulated to low-to-moderate doses.

Season of birth and risk of brain tumors in adults.

BACKGROUND: Recent studies demonstrated an excess of winter births in children with brain tumors and in adults with various neurologic or psychiatric diseases relative to the general population. OBJECTIVE: To investigate a possible association between month of birth and risk of brain tumors in adults using data from a large, hospital-based case-control study. METHODS: Cases were patients with incident glioma (n = 489) or meningioma (n = 197) diagnosed at hospitals in Boston, MA, Phoenix, AZ, and Pittsburgh, PA. Controls (n = 799) were patients hospitalized for a variety of nonmalignant conditions and frequency matched to cases by hospital, age, sex, race/ethnicity, and distance of residence from hospital. Odds ratios (ORs) were calculated using multivariate unconditional logistic regression allowing for cyclic variation in risk with month of birth. RESULTS: A relationship between month of birth and risk of adult glioma and meningioma was found, best described by a 12-month periodic function with peaks in February and January and troughs in August and July. The association between month of birth and risk of glioma differed significantly by handedness, with left-handed and ambidextrous subjects born during late fall through early spring being at particularly high risk of adult glioma as compared with those born at other times of the year. CONCLUSION: These findings suggest the importance of seasonally varying exposures during the pre- or postnatal period in the development of brain tumors in adults.

Occupation and the risk of adult glioma in the United States.

OBJECTIVE: Previous studies have observed increased glioma incidence associated with employment in the petroleum and electrical industries, and in farming. Several other occupations have also been associated with increased risk, but with inconsistent results. We evaluated associations between occupational title and glioma incidence in adults. METHODS: Cases were 489 patients with glioma diagnosed from 1994 to 1998 at three United States hospitals. Controls were 799 patients admitted to the same hospitals for non-malignant conditions. An experienced industrial hygienist grouped occupations that were expected to have similar tasks and exposures. The risk of adult glioma was evaluated for those subjects who ever worked in an occupational group for at least six months, those who worked longer than five years in the occupation, and those with more than ten years latency since starting work in the occupation. RESULTS: Several occupational groups were associated with increased glioma incidence for having ever worked in the occupation, including butchers and meat cutters (odds ratio [OR] = 2.4; 95% confidence limits [CL]: 1.0, 6.0), computer programmers and analysts (OR = 2.0; 95% CL: 1.0, 3.8), electricians (OR = 1.8; 95% CL: 0.8, 4.1), general farmers and farmworkers (OR = 2.5; 95% CL: 1.4, 4.7), inspectors, checkers, examiners, graders, and testers (OR = 1.5; 95% CL: 0.8, 2.7), investigators, examiners, adjustors, and appraisers (OR = 1.7; 95% CL: 0.8, 3.7), physicians and physician assistants (OR = 2.4; 95% CL: 0.8, 7.2), and store managers (OR = 1.6; 95% CL: 0.8, 3.1), whereas occupation as a childcare worker was associated with decreased glioma incidence (OR = 0.4; 95% CL: 0.2, 0.9). These associations generally persisted when the subjects worked longer than five years in the occupation, and for those with more than ten years latency since starting to work in the occupation. CONCLUSIONS: This is our first analysis of occupation and will guide future exposure-specific assessments.

Magnetic fields produced by hand held hair dryers, stereo headsets, home sewing machines, and electric clocks.

A recent epidemiologic study reported associations between leukemia risk in children and their personal use of television (TV) sets, hair dryers, and stereo headsets, and the prenatal use by their mothers of sewing machines. To provide exposure data to aid in the interpretation of these findings, extremely and very low frequency (ELF and VLF) magnetic fields produced by a sample of each type of appliance were characterized in a field study of volunteers conducted in Washington DC and its Maryland suburbs. Questionnaire data regarding children's or mothers' patterns of usage of each type of appliance were also collected. ELF magnetic fields measured 10 cm from the nozzles of hair dryers were elevated over the ambient by a mean factor of 17 when these devices were in use. Fields near headsets being used to listen to music were not distinguishable from ambient levels except at frequencies below and well above 60 Hz and, even then, field levels were < 0.01 microT. Home sewing machines produced ELF magnetic fields that were elevated by a factor of 2.8 over ambient levels at the front surfaces of the lower abdomens of mothers. Estimated mean daily times of usage of hair dryers, stereo headsets, and sewing machines were 2.6, 19, and 17 minutes, respectively. These data and previously published data on TV sets, do not provide a consistent picture of increased (or decreased) leukemia risk in relation to increasing peak or time weighted average (TWA) ELF magnetic field exposure. The data could, however, conceivably be compatible with some more complex biophysical model with unknown properties. Overall, the results of this study provide little evidence supporting the hypothesis that peak or TWA ELF magnetic fields produced by appliances are causally related to the risk of childhood leukemia in children.

Static magnetic field measurements in residences in relation to resonance hypotheses of interactions between power-frequency magnetic fields and humans.

Bowman et al. used epidemiologic data to test a model in which subjects were classified as being "in-resonance" or "not-in-resonance" for 60-Hz magnetic-field exposures depending on single static magnetic-field measurements at the centers of their bedrooms. A second paper by Swanson concluded that a single static magnetic-field measurement is insufficient to meaningfully characterize a residential environment. The main objective of this study was to investigate exposure-related questions raised by these two papers in two U.S. data sets, one containing single spot measurements of static magnetic fields at two locations in homes located in eight states, and the other repeated spot measurements (seven times during the course of one year) of the static magnetic fields at the centers of bedrooms and family rooms and on the surfaces of beds in 51 single-family homes in two metropolitan areas. Using Bowman's criterion, bedrooms were first classified as being in-resonance or not-in-resonance based on the average of repeated measurements of the static magnetic field measured on the bed where the presumed important exposure actually occurred. Bedrooms were then classified a second time using single spot measurements taken at the centers of bedrooms, centers of family rooms, or on the surfaces of beds, as would be done in the typical epidemiologic study. The kappa statistics characterizing the degree of concordance between the first (on-bed averages) and second (spot measurements) methods of assessing resonance status were 0.44, 0.33, and 0.67, respectively. This level of misclassification could significantly affect the results of studies involving the determination of resonance status.

Household solvent exposures and childhood acute lymphoblastic leukemia.

OBJECTIVES: This study explored the risk of childhood acute lymphoblastic leukemia (ALL) associated with participation by household members in hobbies or other home projects involving organic solvents. METHODS: Participants in this case-control study were 640 subjects with ALL and 640 matched controls. RESULTS: Childhood ALL was associated with frequent (> 4 times/month) exposure to model building (odds ratio [OR] = 1.9; 95% confidence interval [95% CI] = 0.7, 5.8) and artwork using solvents (OR = 4.1; 95% CI = 1.1, 15.1). We also found elevated risk (OR = 1.7; 95% CI = 1.1, 2.7) among children whose mothers lived in homes painted extensively (> 4 rooms) in the year before the children's birth. CONCLUSIONS: In this exploratory study, substantial participation by household members in some common household activities that involve organic solvents was associated with elevated risks of childhood ALL.

Cellular-telephone use and brain tumors.

BACKGROUND: Concern has arisen that the use of hand-held cellular telephones might cause brain tumors. If such a risk does exist, the matter would be of considerable public health importance, given the rapid increase worldwide in the use of these devices. METHODS: We examined the use of cellular telephones in a case-control study of intracranial tumors of the nervous system conducted between 1994 and 1998. We enrolled 782 patients through hospitals in Phoenix, Arizona; Boston; and Pittsburgh; 489 had histologically confirmed glioma, 197 had meningioma, and 96 had acoustic neuroma. The 799 controls were patients admitted to the same hospitals as the patients with brain tumors for a variety of nonmalignant conditions. RESULTS: As compared with never, or very rarely, having used a cellular telephone, the relative risks associated with a cumulative use of a cellular telephone for more than 100 hours were 0.9 for glioma (95 percent confidence interval, 0.5 to 1.6), 0.7 for meningioma (95 percent confidence interval, 0.3 to 1.7), 1.4 for acoustic neuroma (95 percent confidence interval, 0.6 to 3.5), and 1.0 for all types of tumors combined (95 percent confidence interval, 0.6 to 1.5). There was no evidence that the risks were higher among persons who used cellular telephones for 60 or more minutes per day or regularly for five or more years. Tumors did not occur disproportionately often on the side of head on which the telephone was typically used. CONCLUSIONS: These data do not support the hypothesis that the recent use of hand-held cellular telephones causes brain tumors, but they are not sufficient to evaluate the risks among long-term, heavy users and for potentially long induction periods.

Benzene and lymphohematopoietic malignancies in China.

While this study is larger than previous investigations and includes workers with a wide range of exposures to benzene, the estimates of risk, as measured by statistical confidence intervals, are still fairly broad, and would benefit from the larger numbers that could be provided by continued follow-up of this population. Nevertheless, the study confirms earlier findings of increased risk for ANLL and aplastic anemia among benzene-exposed workers, provides the first substantial evidence that MDS is linked to benzene exposure, and provides evidence that benzene increases risk for ANLL/MDS at lower levels of exposure than had previously been demonstrated. Currently we are evaluating the potential for extending the follow-up of workers included in this study. A new study would include expanded data collection for cases of hematopoietic malignancy and related disorders and for an appropriate control series.

Do confounding or selection factors of residential wiring codes and magnetic fields distort findings of electromagnetic fields studies?

In contrast with several previous studies, our recent large case-control study found little association between childhood acute lymphoblastic leukemia (ALL) and electric-power-line wire codes. Here we examine internal evidence from our study to assess the possibility that selection bias and/or confounding may have affected the findings. We compared the relation between childhood ALL and wire codes and direct measurements of magnetic fields in subjects who participated in all phases of the study with the relation in all subjects, including those who declined to allow access inside the home. We found that the odds ratio for ALL among those living in homes with very high current configurations increased by 23% when 107 "partial participants" were excluded. We found similar, but slightly smaller, increases in the odds ratios when we performed the same comparisons using direct measurements of magnetic fields, excluding subjects who allowed only a measurement outside the front door. "Partial participants" tended to be characterized by lower socioeconomic status than subjects who participated fully, suggesting possible selection bias. We also examined the relation between a large number of potential confounding variables and both proxy and direct measurements of magnetic fields. Univariate adjustment for individual variables changed the odds ratio for ALL by less than 8%, while simultaneous adjustment for several factors reduced the estimate by a maximum of 15%. We conclude that while confounding alone is unlikely to be an important source of bias in our own and previous studies of magnetic fields, selection bias may be more of a concern, particularly in light of the generally low response rates among controls in case-control studies.