Epidemiology

The second session of the workshop focused on epidemiologic studies. The following sections are organized into cancer and noncancer effects, and effects in children.

Epidemiologic studies are of primary importance in health risk assessment, because they are able to provide direct information on the health of people exposed to an agent. While epidemiologic investigations are difficult to conduct and easy to criticize, they offer unique advantages over experimental models. In particular, epidemiologic studies permit for associations between an environmental exposure, under actual rather than artificial conditions, and a health outcome to be evaluated in human populations. Moreover, because epidemiologic research is conducted in natural settings, the joint influence of multiple factors on disease occurrence can be studied, while taking into account an individual’s susceptibility. Equally important, extrapolation from high doses to low doses is not necessary because epidemiologic studies examine a wide range of exposures at relevant doses. However, epidemiologic studies may be affected by bias,1 confounding,2 and exposure misclassification.3 A large amount of exposure misclassification is likely to be present for radiofrequency (RF) exposure. Furthermore,

1

Bias is any trend in the collection, analysis, interpretation, publication, or review of data that can lead to conclusions that are systematically different from the truth. Types of bias include recall and selection.

2

Confounding occurs when an observed association between an exposure and disease is distorted by other risk factors for the disease that are also associated with exposure.

3

Misclassification is inaccuracies in how subjects are categorized by exposure or disease status.



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 19
Epidemiology The second session of the workshop focused on epidemiologic studies. The following sections are organized into cancer and noncancer effects, and effects in children. Epidemiologic studies are of primary importance in health risk assess- ment, because they are able to provide direct information on the health of people exposed to an agent. While epidemiologic investigations are difficult to conduct and easy to criticize, they offer unique advantages over experi- mental models. In particular, epidemiologic studies permit for associations between an environmental exposure, under actual rather than artificial conditions, and a health outcome to be evaluated in human populations. Moreover, because epidemiologic research is conducted in natural settings, the joint influence of multiple factors on disease occurrence can be studied, while taking into account an individual’s susceptibility. Equally important, extrapolation from high doses to low doses is not necessary because epi- demiologic studies examine a wide range of exposures at relevant doses. However, epidemiologic studies may be affected by bias,1 confounding,2 and exposure misclassification.3 A large amount of exposure misclassifica- tion is likely to be present for radiofrequency (RF) exposure. Furthermore, 1 Bias is any trend in the collection, analysis, interpretation, publication, or review of data that can lead to conclusions that are systematically different from the truth. Types of bias include recall and selection. 2 Confounding occurs when an observed association between an exposure and disease is distorted by other risk factors for the disease that are also associated with exposure. 3 Misclassification is inaccuracies in how subjects are categorized by exposure or disease status. 

OCR for page 19
0 IDENTIFICATION OF RESEARCH NEEDS large studies are needed to have enough statistical power to detect effects when exposures and outcomes are rare. Thus, while the observational nature of the epidemiologic studies makes them difficult to conduct and interpret, they remain indispensable by providing direct information on the health of people exposed to an agent. CANCER EFFECTS Recent concerns about possible health effects of exposure to RF radia- tion from cellular phones have centered on brain cancer (Krewski 2007), and a series of epidemiologic studies looking at the association between cellular phone use and malignant and benign brain tumors, as well as salivary gland tumors, have been published since 1999 (Johansen et al. 2001; Lonn 2004; Shoemaker et al. 2005; Schuz et al. 2006; Hardell et al. 2006; Lahkola 2007). The majority of these have been case-control stud- ies.4 Limitations of these studies have included inclusion of only a small number of heavy users, inability to account for a sufficiently long latency5 period, differential participation by cases and controls according to his- tory of cellular phone use, inclusion of both prevalent and incident cases,6 and relatively crude exposure assessment. All of these case-control studies relied on self-reported histories of cellular phone use, which can be prone to both random and systematic errors.7 The latter could occur if cases or their proxies tended to over-report cellular phone use relative to controls or if tumor laterality8 influenced reported laterality of phone use. It also is possible that cognitive impairment would cause brain cancer cases to incor- rectly report phone use. Analyses conducted as part of the INTERPHONE study, a multi-center international case-control study of brain and parotid gland tumors, point to selection bias9 due to likely under-enrollment of nonexposed controls, 4 A case-control study is a type of observational study in which subjects are selected on the basis of whether they have or do not have a disease under investigation. Characteristics such as previous exposure are then compared between cases and controls. 5 Latency is the period of subclinical disease following exposure that ends with the onset of disease. 6 Prevalence is the number or proportion of cases or events or conditions present in a given population. Incident cases are those who are newly diagnosed with the disease of interest. 7 Random errors vary in a nonreproducible way around a limiting mean. These errors can be treated statistically by use of the laws of probability. Systematic errors are reproducible and tend to bias a result in one direction. Their causes can be assigned, at least in principle, and they can have constant and variable components. 8 Tumor laterality is the preference in location of tumor in one portion of the body over other locations in the body. 9 Selection bias occurs when the association that is observed among those who participate in a study is different from what would have been observed because individuals who were eligible to participate refused or were not selected.

OCR for page 19
 EPIDEMIOLOGY which would tend to bias estimates of the odds ratio downward (Vrijheid et al. 2006). Random errors in reported use of cellular phones also could obscure a small increase in risk. In the simulation study based on different assumptions about random and systematic errors using preliminary data of a study on short-term recall of mobile phone use among healthy volun- teers, sensitivity analyses indicated that random errors in recall were likely to outweigh plausible systematic errors in recall that can lead to a large under-estimation, if there is a true risk, in the risk of brain cancer associ- ated with use of wireless communication devices (Vrijheid et al. 2006). The pending results of the INTERPHONE study, by far the largest case-control study of head and neck tumors to date, are likely to have a major influence on the direction and scope of future research concerning the use of cellular phones and cancer. An international collaborative study with prospective follow-up of mobile phone use is being launched in Europe. A key advantage of cohort studies10 is the ability to evaluate multiple outcomes, providing the op- portunity for a more comprehensive assessment of possible health risks. In addition, more accurate exposure data may be obtained because recall of exposure information would be independent of the outcome of interest and the self-reported mobile phone use could be complemented more read- ily with information from mobile service providers. Furthermore, selection bias should not be an issue as it can be in a case-control study. Sufficiently long exposure and follow-up would allow for the detection of effects that occur with a latency of several years. Additionally, it might be possible to collect data to evaluate dose-response relationships in a prospective cohort study. Mortality is probably inadequate as the sole measure of effect and incidence11 could also be included. Occupational studies12 have been performed over a longer time period, but the exposure frequencies may not always be relevant for an assessment of effects at frequencies of interest from wireless communication devices, and we are only beginning to measure and learn about RF exposures in various occupations. The most extensive literature addresses brain tumors and leukemia, but also of note are cancers of the breast (Tynes et al. 1996; Morgan et al. 2000), testis (Hayes et al. 1990), lung (Armstrong et al. 1994; Armstrong Groves et al. 2002), and uveal melanoma13 (Stang et al. 2001). These stud- ), ies have several methodological weaknesses related to the fact that none 10 A cohort study is a type of epidemiologic study where a group of individuals are followed over time to assess the occurrence of a given disease or condition. Enrollment into the study is based on exposure characteristics or membership in a group. 11 Incidence is a measure of disease occurrence that quantifies the number of new cases of a disease that develop in a population of individuals at risk during a specified time period. 12 Occupational studies are studies of workers. 13 Uveal melanoma is cancer of the eye.

OCR for page 19
 IDENTIFICATION OF RESEARCH NEEDS of the studies measured RF exposures. The included occupations did not necessarily represent truly highly exposed occupations. Another weakness is that exposure classification has often been based on a single job title. In addition, limited control of confounding, if any, has been made. Evaluation of effects and adjustment for exposure measurement error and selection bias in case-control studies is needed. By better characterizing exposure measurement error and evaluating the magnitude of selection bias, such information may be used to adjust for exposure misclassification in existing and future studies. Several ecologic studies14 (Hocking et al. 1996; McKenzie et al. 1998) have examined cancer risk, including risk of childhood leukemia, among populations living in proximity to radio and television broadcast towers. Often driven by a previously identified cluster, these analyses are based simply on distance from the source and often include an extremely small number of cases. Such studies have been mostly uninformative. More rigorous investigations might be feasible with a development of new instru- ments capable of capturing personal RF exposures. Furthermore, in addition to the rapid growth in the number of base stations, there has been a growth of other sources of RF radiation from cordless phones and wireless systems. The last general survey of RF levels in U.S. cities was conducted in the 1970s (Tell and Mantiply 1980), and ), an updated survey of RF intensities would be useful background for future epidemiologic studies. Measurements of the differences in the exposures generated by the use of cell phones and other wireless systems will be of value in determining if there are any health effects resulting from exposures to low levels and/or intermittent sources of RF radiation. NONCANCER HEALTH EFFECTS Few studies have been conducted on health effects other than cancer risk of RF electromagnetic fields from wireless communication devices (Auvinen 2007). The existing studies have been small and have meth- odological limitations such as lack of rigorous exposure assessment, in- adequate control of confounding, and cross-sectional design. Exposure settings have included mobile phone use, residence close to base stations, and occupational exposures (broadcasting). The health endpoints evaluated include reproductive effects, cardiovascular health (Morgan et al. 2000; Groves et al. 2002), and nonspecific symptoms such as headache, sleep disturbances, fatigue, and depressed mood (Chia et al. 2000; Sandstrom et al. 2001). Several reports of increased prevalence of symptoms among 14 Ecologic studies contain only information on population averages and lack joint distribu- tions of individual-level variables.

OCR for page 19
 EPIDEMIOLOGY subjects exposed to communication devices have been published, but the quality of the research is generally low. At present, the knowledge base of health effects other than cancer is grossly inadequate. Due to lack of knowledge on possible mechanisms and the wide range of potential effects, assessment of research needs cannot be very specific. Little indication to guide future efforts is available from other fields of re- search related to wireless communications devices (in vivo studies, animal studies, or laboratory studies of volunteers). Thus, multi-endpoint studies are needed, and in this case a cohort approach of both children and adults is most suitable. In addition, there is the possibility of obtaining more accurate exposure data, which is important considering how rapidly wireless tech- nology is changing in both exposure characteristics (e.g., frequencies and modulation) and usage pattern (e.g., phone use vs. text messaging or web surfing). The following outcomes are of particular interest: cancer, cardio- vascular, neurological, neurodegenerative, and cerebrovascular diseases; and reproductive, behavioural, cognitive, neurophysiologic, and neuro- psychological effects, including headaches, sleep disturbances, tinnitus,15 psychiatric disorders, and well-being (health-related quality of life). Children With the rapid advances in technologies and communications utilizing RF fields, children are increasingly exposed to RF energy at earlier ages (starting at age 6 or before). Environmental exposures could be particularly harmful to children because of their vulnerability during periods of develop- ment before and after birth. Although it is unknown whether children are more susceptible to RF exposures, they may be at increased risk because of developing organ and tissue systems, particularly of the nervous system (Kheifets et al. 2005). In addition, they may have a greater specific absorp- tion rate (SAR) and higher absorption of RF energy at frequencies utilized in wireless communications devices and greater RF penetration relative to head size. Finally, the current generation of children will experience a longer period of exposure to RF fields from mobile phone use than adults since they started using mobile phones at an early age and are likely to continue to use them. Data from the INTERPHONE study show that both the prevalence of regular mobile phone users and daily use were highest in the younger age groups (Cardis et al. 2007). Moreover, several recent trends (such as increased popularity, reduced price, and advertising to children) have led to increased mobile phone use among children. A steep increase in mobile phone ownership among children has been reported in 15 Tinnitus is ringing of the ears.

OCR for page 19
 IDENTIFICATION OF RESEARCH NEEDS several surveys. In spite of this, there is virtually no relevant epidemiology at present that examines health effects in children. Children are potentially exposed from conception through maternal wireless device use and then postnatally when they themselves become users of mobile phones. Both short-term and long-term health consequences of mobile phone use can be efficiently studied by adding information on the mobile phone use to the existing birth cohorts using a life course approach. Cognitive and behavioral effects are of particular importance and can be evaluated in the near term. Owing to widespread use of mobile phones among children and adoles- cents and the possibility of relatively high exposures to the brain, investiga- tion of the potential effects of RF fields in the development of childhood brain tumors is warranted. Brain cancer is an important endpoint to study but is rare in children and so a cohort study is unlikely to be feasible. There is at present a lack of information concerning the health effects associated with living in close proximity to base stations. Epidemiologic studies of mobile phone base stations present unique challenges that need to be addressed to make such studies rigorous. One particular difficulty in a study of populations near transmitters such as TV, radio, and base stations, with both residential exposure and cell phone use, is the development of accurate indexes of exposure that are closely correlated to the exposures that a person receives. Assuming a nonspecific effect from RF exposure implies that in an epidemiologic study all major RF sources should be evalu- ated (i.e., from both base stations and TV and radio towers as exposure from these sources are whole body and of similar magnitude; additionally, exposure from one’s own cell phone use should be considered). Further investigation into improved measurements is a critical step in better captur- ing exposure data from these sources and in determining the feasibility of epidemiologic studies of children living in the vicinity of these sources. The committee’s evaluation of presentations and discussions at the workshop has resulted in the identification of the following research needs and gaps. Adults Research Needs 1. Prospective Cohort Studies. A prospective cohort study will allow for the evaluation of diverse outcomes, but a very large sample size and extended follow-up is required for rare outcomes or those that occur only with very long latencies. 2. Occupational Cohorts with Medium to High Exposure. None of the occupational studies to date have been based on an adequate exposure

OCR for page 19
 EPIDEMIOLOGY assessment. Much work is needed to identify occupations with potentially high RF exposures and to characterize them. Among the particular find- ings in need of follow-up are uveal cancer, leukemia, lung, breast, testicular cancers, and sperm quality. If feasible, epidemiologic studies could focus on the health effects of potential RF exposures at high levels that might occur in a few workplace tasks (e.g., antenna construction and maintenance). In addition, several groups of firefighters have expressed concerns about their exposure from base stations located in firehouses and possible exposure to antennas on the fire trucks. A feasibility study might be useful in determin- ing whether these groups have a sufficient exposure range and if an infor- mative study is feasible. Research Gaps Judged to Be of Lower Priority 1. Case-control Studies of Rare Diseases. For rare diseases, such as adult leukemia, non-Hodgkin lymphoma and neurodegenerative diseases, additional case-control studies may be justified as a cohort study will likely not have sufficient statistical power. Such case-control studies would better characterize exposure and try to adjust for exposure measurement error and selection bias as needed. 2. Observational Studies on Subjective Outcomes.16 Subjective out- comes from long-term exposures can be addressed in appropriately de- signed, observational studies (e.g., prospective cohort study) that include provisions for avoiding or minimizing reporting and selection and misclas- sification biases. Children Research Needs 1. Prospective Cohort Studies of Pregnancy and Childhood. Children are potentially exposed from conception through maternal wireless device use and then postnatally when they themselves become users of mobile phones. Both short-term and long-term health consequences of mobile phone use can be efficiently studied by adding information on the mobile phone use to the existing birth cohorts using a life course approach. Cogni- tive and behavioral effects are of particular importance and can be evalu- ated in the near term. 16 Subjective outcomes are outcomes or symptoms that are difficult to quantify objectively (e.g., pain, headaches, sleep disturbances).

OCR for page 19
 IDENTIFICATION OF RESEARCH NEEDS 2. Case-control Study of Children Mobile Phone Users and Brain Cancer. Owing to widespread use of mobile phones among children and adolescents and the possibility of relatively high exposures to the brain, investigation of the potential effects of RF fields in the development of childhood brain tumors is warranted. Brain cancer is an important end- point to study but is rare in children and so a cohort study is unlikely to be feasible. Research Gaps Research Ongoing 1. Case-control studies on childhood cancer with improved exposure assessment taking into account all major fixed point sources of RF expo- sure (base stations, AM, FM, TV antennas, and other sources). There is at present a lack of information concerning the health effects associated with living in close proximity to base stations. Epidemiologic studies of mobile . phone base stations present unique challenges that need to be addressed to make such studies rigorous.