NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

National Research Council (US) Panel on Impact of Video Viewing on Vision of Workers; National Research Council (US) Committee on Vision. Video Displays, Work, and Vision. Washington (DC): National Academies Press (US); 1983.

Cover of Video Displays, Work, and Vision

Video Displays, Work, and Vision.

Show details

3Radiation Emissions and Their Effects

Types and Levels of Radiation Emitted by VDTs

Most electronic products, including video display terminals, emit electromagnetic radiation. The types and levels of radiation emitted vary with the device. VDTs are designed to emit visible radiation (light), and all electronic products that increase in temperature, including VDTs, emit infrared (thermal) radiation. Electronic products in which certain types of high-voltage components are used can produce X radiation (i.e., X rays). For example, VDTs that use a cathode-ray tube (CRT) display produce internal X radiation; the tube face, however, is designed to filter out this radiation so that it does not leave the tube. CRTs also have sweep oscillator circuits, which emit radio frequency radiation. The radiation emitted in any part of the spectrum may be either broadband (e.g., most infrared radiation emitted by VDTs is broadband) or limited to discrete wavelengths or frequencies, as is most common for radio frequency radiation emitted by CRT circuits.

The radiation emissions of a wide variety of different types of VDTs have been measured, and most types of VDTs in current use have been adequately surveyed for all types of potentially hazardous electromagnetic radiation. In many cases, measurements were undertaken because of concerns expressed by VDT users about potential radiation hazards. Field surveys and laboratory studies have been conducted by government agencies in this country and abroad (Moss et al., 1977; Cox, 1980; Terrana et al., 1980; Bureau of Radiological Health, 1981; National Institute for Occupational Safety and Health, 1981), private organizations (Weiss and Petersen, 1979), and independent groups (Wolbarsht et al., 1980). All of the studies have reached the same conclusion: emissions of all types of electromagnetic radiation—X rays, ultraviolet (UV), visible (light), infrared (IR), and radio frequency (RF) radiation, including microwaves—are well below accepted occupational and environmental health and safety standard limits. The principal studies reported in the literature are summarized in Table 3.1.

TABLE 3.1. Maximum Values Reported From Several Radiation Measurement Studies.


Maximum Values Reported From Several Radiation Measurement Studies.

Some differences appear in the levels of radiation reported in different studies (see Table 3.1); the largest differences occur between field surveys and laboratory studies. In the field studies reviewed, VDTs and measuring instruments were not shielded from ambient radiation; thus, the readings obtained represent the sum of the VDT emissions and all other sources of radiation present. Furthermore, in some of the field surveys, instruments of limited sensitivity were used and the emissions were very weak, so that the actual emission levels could not be determined; in these cases it could be determined only that the emission levels were less than some value (usually the radiation exposure standard). Under controlled laboratory conditions it is possible to use shielding, special time-averaging techniques, and more sensitive instruments to measure the actual emission levels or to set a much lower limiting value. Thus, the laboratory studies cited in Table 3.1 provide the best measures of actual emissions from VDTs.

Studies of Emission Levels

The 1977 NIOSH Study

The first detailed measurements of radiation emitted by VDTs were made by the National Institute for Occupational Safety and Health (NIOSH) of the U.S. Department of Health, Education, and Welfare (Moss et al., 1977). The measurements were undertaken in response to a request from the Newspaper Guild and the New York Times for an evaluation of possible radiation hazards to which employees working with VDTs might be exposed.

The emissions from three units (Harris 1500A, Incoterm SPD 10/20, and Telco 40) in use at the newspaper facility were measured in the UV, visible, IR, and RF bands. (The Harris and Telco units had been used by two employees who developed cataracts.) Luminance measurements were also made on the Telco and Harris units and on an IBM 3277 unit. Measurements of RF emissions and luminance values were made on an additional 20 units of several models (Harris 1500A, Harris 2200, IBM 3277, and Telco 40). Measurements of X-ray emissions were not made during this survey because previous measurements made by NIOSH and other groups on the same and similar VDT models, some of which were located at the newspaper facility, found no X radiation above background levels.

Emissions of visible, IR, and RF radiation could not be detected, even when the most sensitive detector scales on the measurement instruments were used. Measured luminance values were less than one percent of the standard of the American Conference of Governmental Industrial Hygienists (ACGIH). No UV radiation below a wavelength of 300 nm was detected. At wavelengths between 300 and 400 nm (UV-A), the maximum measured emission was 1/500,000 of the occupational exposure standards presently recommended by the ACGIH.

Because the values obtained for the emissions of all types of radiation measured were considerably below currently recommended occupational exposure standards, NIOSH concluded that "the VDTs surveyed do not appear capable of producing levels of radiation presenting an occupational ocular radiation hazard" (Moss et al., 1977:14).

The Bell Telephone Laboratories Study

The next widely circulated, published radiation survey was conducted by Weiss and Petersen (1979) for Bell Telephone Laboratories. The 33 VDTs surveyed represented 13 different models used at Bell Telephone Laboratories. Measurements were made across the entire spectrum, although not all units were measured in each band. This study also reviewed the problems associated with accurate measurement of weak emissions.

The only radiation emissions detected were from 1.5 kHz to 1.42 GHz in the RF spectrum, and from 350 nm in the UV to 600 nm in the red part of the visible spectrum. The emissions of RF radiation were 1/100 of the most stringent safety standard in the world, that of Czechoslovakia; emissions of UV were 1/10,000 or less of occupational exposure limits.

Measurements of X radiation emissions were reported for 18 CRT units of 11 different models. Only one unit yielded emissions above background levels: the emissions were caused by a faulty high-voltage power supply, and after the unit was repaired, no further X radiation was detected. Each of the other units surveyed yielded levels that were less than 0.5 mR/h, measured 5 cm from any accessible surface.

A Collaborative Study

In a study reported in 1980 a group of independent investigators from Duke University, the U.S. Army Environmental Hygiene Agency, the University of Washington, and the IBM Corporation performed careful, detailed laboratory measurements of emissions from an IBM Model 3277 VDT (Wolbarsht et al., 1980). RF radiation (including microwave) from 10 kHz to 10 GHz, optical radiation from 200 nm (UV) to 10 µm (far infrared), and X radiation from 5 keV to over 40 keV were measured. Because measurements were performed in a screened laboratory, it was possible to measure very weak emissions that cannot be measured in field surveys.

The emission values for all bands at the tube face and other locations were reported to be from 1/10 to 1/100 of existing safety standards, even under artificially induced overvoltage fault conditions designed to maximize emissions. In many instances, it was difficult to measure emissions because they were below or near background environmental levels. Measurements of RF emissions from black-and-white and color television sets were also made for comparison purposes; the television sets emitted more RF radiation at frequencies of less than 1 MHz and slightly less RF radiation at higher frequencies.1 This study concluded that there was no radiation hazard associated with emissions from VDTs.

Two European Studies

Terrana and coworkers (1980) measured emissions of RF and X radiation on 13 VDTs of 5 different models used in a newspaper facility and in a commercial firm in Italy. One of the VDT units was also examined under controlled laboratory conditions. Measurements of RF emissions were made in contact with all surfaces, at keyboard level, and at operator position, and all measurements were made with a completely illuminated screen. X radiation was measured by scanning over the screen and all other surfaces, with the screen completely filled with characters. Measurements of X-radiation emissions made by the Enrico Fermi Nuclear Research Centre on an additional 72 units of 9 models are also reported by Terrana and coworkers. The emission levels are shown in Table 3.1. This survey concluded: "The values obtained for X-ray and radio frequency radiation were far lower even than the most restrictive permissible exposure levels established by any agency or government" (Terrana et al., 1980:13).

A radiation survey commissioned by the Health and Safety Executive of the British government and conducted by the National Radiological Protection Board was reported by Cox (1980). Measurements across the entire spectrum were made on more than 200 different types of VDTs in 60 companies. Emission levels were measured at maximum screen luminance with full screen illumination. X-ray emissions were 1/50 of the British emission standard for household electronic products, UV emissions were 1/100 of permissible limits, visible and near infrared emissions were 1/25 and 1/2,000 of applicable limits, and RF intensities were 1/10 of the appropriate standard limit values in most cases. Based on the measured levels of emission and standards for exposure, the study concluded that the "radiation normally emitted from a VDU [a VDT] does not pose a hazard to operators either in the long or short term" (Cox, 1980:31).

The 1981 NIOSH Study

In 1981 NIOSH performed a second set of radiation measurements on VDTs in work settings in three San Francisco-Oakland area newspaper facilities. Of the 530 VDTs in use, 136 units, produced by 6 manufacturers, were surveyed. Based on the results of those measurements (see Table 3.1) and on its previous investigations, NIOSH concluded that "the VDT does not present a radiation hazard to the employees working at or near a terminal" (National Institute for Occupational Safety and Health, 1981:68). The report discussed the danger of obtaining erroneously high RF readings with some RF survey instruments as a result of inductive coupling with the flyback transformer. Early press accounts of such erroneous readings had led to unwarranted concern about a potential hazard from RF emissions. In this study, RF emissions were not detectable.

The Bureau of Radiological Health Study

Because of an increasing number of inquiries from private citizens, union representatives, and government agencies about the possibility of harmful levels of radiation emitted by VDTs, the Bureau of Radiological Health (1981) of the U.S. Department of Health and Human Services recently undertook a comprehensive survey and evaluation of electromagnetic radiation and acoustic emissions from VDTs (see Table 3.1). This survey measured emissions of ionizing (X rays) and nonionizing (UV, visible, IR, and RF) radiation from 34 units representing a cross-section of models of all known manufacturers of VDTs used in the United States. Data on measurements of ionizing radiation emissions from 91 units tested between 1975 and 1980 were included and reanalyzed to provide a more comprehensive analysis for both monochromatic and color units. All units were tested under conditions designed to maximize emissions. Potential sources of emissions were carefully analyzed, and measurements were made under controlled laboratory conditions.

The emission levels of all bands of ionizing and nonionizing radiation were well below current state, federal, and international standards and guidelines for exposure for all 34 of the units tested in this study. Under worst-case conditions (including artificially induced component failures and misadjusted service and user controls), 8 of the 91 units tested for ionizing radiation between 1975 and 1980 exceeded the 0.5 mR/h standard for X-radiation emission for television receivers. Under normal conditions, however, no X radiation was detected from any of the 91 units. The eight sets that exceeded the X-radiation standard represented three models that were either subsequently recalled by the manufacturer for modification to comply with the federal performance standard for television receivers, or were not permitted entry into the U.S. market.2

Radiation Safety Standards

There are occupational and environmental safety standards for most types of radiation. For some spectral regions, there is widespread acceptance and confidence in national and international standards (see the last column in Table 3.1); for other spectral regions, only a few guidelines exist. For those forms of radiation for which few guidelines exist, there is generally little demand for standards, either because few people are exposed to such radiation or because there is no general concern that such radiation is hazardous at present levels of occupational or environmental exposure. For example, exposure standards have not been promulgated for RF radiation at wavelengths greater than 10-30 m (i.e., frequencies less than 10-30 mHz) because body absorption of radiation at very long wavelengths is far below that occurring at microwave wavelengths; see Guy and Chou (1982) and Schaefer and coworkers (1982) for discussions of absorption characteristics and safety considerations for very long wavelength radiation.

Some public concerns about VDTs may stem from lack of knowledge or misinterpretation of existing studies of radiation emission from VDTs. For example, a booklet prepared for workers by the Ontario Public Service Employees Union (DeMatteo et al., 1981) expresses concern that field surveys reported X-radiation values apparently close to the 0.5 mR/h standard. As noted above, however, field studies do not accurately measure the actual emission levels; laboratory studies indicate that VDT emissions of X radiation are usually orders of magnitude below 0.5 mR/h. The booklet also expressed concern that the X-radiation standard did not consider risks of biological damage other than "fatal radiation-induced cancer." In fact, radiation standards are generally based on all known serious biological effects, both acute and long term, and take into consideration the cumulative exposure of workers to various human-made and natural sources of radiation. The appropriateness of the 0.5 mR/h standard has also been questioned on the grounds that it is designed for television viewers sitting across the room from their sets, and thus the standard would not suffice for operators working close to VDTs on a daily basis. The standard, however, specifically limits emissions 5 cm from the screen, considerably closer than the usual 50 cm-70 cm working distance fom VDTs. The Bureau of Radiological Health (1981) stated that this standard is appropriate for occupational use of VDTs.

For some bands of radiation (e.g., in the UV-B portion of the UV spectrum and the microwave portion of the RF spectrum) it has been suggested that chronic exposure to levels below threshold for acute effects might produce a cumulative effect and over many years lead to the development of cataracts. To our knowledge there have been no studies of chronic exposure to VDTs; however, we again emphasize that the levels of radiation emitted by VDTs are much lower than background levels.

VDT Emissions and Ambient Radiation

The previous sections compared VDT emission levels with occupational and environmental standards. For perspective, VDT radiation emissions should also be compared with exposure to ambient radiation. The sources of ambient, or natural background, radiation are cosmic radiation, terrestrial radiation, and naturally occurring radionuclides deposited in the body primarily from inhalation and ingestion in air, food, and water.

A person is exposed to greater radiation levels in all parts of the spectrum from ambient sources than from a VDT. According to the National Research Council (1980:3), "The major sources of the ionizing radiation to which the general population is exposed continue to be natural background (with a whole-body dose of about 100 mrem/yr) and medical applications of radiation (which contribute similar doses to various tissues of the body)." This level exceeds by far the exposure likely from a typical VDT.

Studies indicate that the level of UV radiation emitted by VDTs is far less than the UV radiation emitted by ordinary fluorescent light that falls on a VDT screen and is reflected. Outdoor UV radiation levels are more than 10,000 times greater than the level of ultraviolet radiation emitted by VDTs (Wolbarsht et al., 1980). The ambient visible and IR environment outdoors is 100 to 1,000 times greater than reported emissions of visible and infrared radiation from VDTs (Wolbarsht et al., 1980). Radio frequency radiation emissions from VDTs, averaged over the 1 kHz to 300 GHz frequency range, have been reported to be of the order of the ambient levels produced in metropolitan areas by radio, television, and communications transmitters (Wolbarsht et al., 1980).

Biological Effects of Radiation

A complete review of the adverse health effects of electromagnetic radiation is clearly beyond the scope of this report; however, this section briefly summarizes the biological effects generally accepted by the scientific community as constituting health hazards, and the final section discusses in some detail the issue of cataracts. In general, the eye is one of the organs most sensitive to injury by most types of radiation emitted by VDTs. The specific effects depend on which structures in the eye absorb significant amounts of radiation; the effects vary with the frequency and wavelength of the incident radiation.

Ionizing Radiation

Few, if any, health hazards have been studied more extensively than those related to exposure to ionizing radiation, including X rays and other types of ionizing radiation. Several reviews of current knowledge concerning the biological effects of ionizing radiation have recently been published, including a report of the National Academy of Sciences (National Research Council, 1980).

VDT workers have expressed concern that exposure to radiation emitted by VDTs might lead to formation of cataracts (the nature and causes of cataracts are discussed below). Available data indicate that the threshold doses for induction of cataracts in humans by X radiation are from 200 to 500 rad for a single exposure and around 1,000 rad for exposure spread over a period of several months (National Research Council, 1980).3 Dose-response data on human exposure for longer periods have not been reported. In comparison, a VDT worker exposed to 0.01 mR/h would absorb less than 1 rad in 40 years of work at VDTs.4

VDT operators have also expressed concern that adverse effects on pregnancy may result from exposure to radiation emitted by VDTs. There have been reports of clusters of spontaneous abortions, miscarriages, and birth defects among VDT operators in four offices in the United States and Canada (Microwave News, 1981). To our knowledge only two of these clusters have been formally studied, one by the Centers for Disease Control (1981) and the other by the U.S. Army Environmental Hygiene Agency (1981). In both cases, VDT work was judged unlikely to be a causal factor, but the reports have not been publicly disseminated.

The level of ionizing radiation generally believed to significantly increase the risk of birth defects in humans is more than 1 rad for acute exposure. For comparison, one may consider that, on the average, people in the United States receive a total dose from natural and human-made sources of about 60 mrad (0.06 rad) during intrauterine life, about 0.2 mrad/d (National Research Council, 1980.) This low dose rate is generally thought not to be a factor in the normal incidence of birth defects. A worker exposed to 0.01 mR/h from a VDT would absorb roughly 14 mrad in 9 months. The report, Biological Effects of Ionizing Radiation (National Research Council, 1980:586), states that "it is impossible on the basis of human studies alone to determine with certainty a dose below which teratologic effects in man are not induced by exposure at sensitive stages in development. Such thresholds do, however, probably exist, and they may be higher for protracted or fractionated radiation than for acute single exposures."

Nonionizing Radiation

Several comprehensive reviews of the biological effects of nonionizing radiation have been published in recent years (Acton and Carson, 1967; International Radiation Protection Association, 1977; National Council on Radiation Protection, 1977; United Nations Scientific Committee on the Effects of Atomic Radiation, 1977; Parrish et al., 1978; World Health Organization, 1979; National Research Council, 1980, 1981; Sliney and Wolbarsht, 1980; Williams and Baker, 1980; National Council on Radiation Protection and Measurements, 1981). These studies cover UV, visible, IR, and RF radiation.

Ultraviolet Radiation

The health hazards of ultraviolet radiation have been reviewed in an environmental health criteria document of the World Health Organization (1979). This report and others (Parrish et al., 1978; Sliney and Wolbarsht, 1980) note that the UV irradiance necessary to elicit acute effects is strongly dependent on wavelength; radiation of less than 320 nm is far more hazardous than near-UV radiation (320 nm to 400 nm). Occupational exposure levels are less than those required to cause erythema sunburn or photokeratitis. Levels as low as 4 mJ/cm 2 (at 270 nm) have been shown to produce photokeratitis, although more than 10 J/cm2 is required at wavelengths greater than 330 nm (Parrish et al., 1978).

Visible and Infrared Radiation

Ocular effects from visible and near-IR radiation are largely limited to the retina, where the radiation is focused. Visible radiation levels required to cause retinal injury are lowest at 440 nm (blue light), the wavelength at which 30 J/cm2 in one day can cause a retinal lesion (Sliney and Wolbarsht, 1980; Williams and Baker, 1980). This level is more than 100 times greater than the visible radiation emitted by VDTs or television sets (Sliney and Wolbarsht, 1980). Exposure to infrared radiation in the 700 nm-1400 nm region at levels approaching 1,000 J/cm2 (277.8 mW-h/cm2) has caused IR cataracts; however, these levels are typically found only in foundries and in heavy industries. The IR radiation emitted by VDTs is no more than 1 mW/cm2 above ambient levels. Exposure-limit guidelines for chronic exposure are typically 10 mW/cm2 to the eye or for the entire body (see, e.g., Sliney and Wolbarsht, 1980).

Radio Frequency Radiation

The level of radiation required to cause microwave cataracts experimentally in animals is generally about 100 mW/cm2. A previous report of the National Academy of Sciences concluded, "Existing evidence does not suggest that microwave fields of less than 10 mW/cm2 can induce cataracts" (National Research Council, 1981:3). However, the report did state: "Although there is currently no evidence that long-term human exposure to field intensities around 10 mW/cm2 can induce cataracts, that possibility cannot be unequivocally excluded on the basis of existing knowledge" (National Research Council, 1981:4).

Teratogenic effects caused by microwave radiation have been reported in experimental animals at levels exceeding 10 mW/cm2 (Czerski et al., 1974), which is much greater than the measured VDT emission levels in the entire RF range. A World Health Organization (1981) criteria document on microwave and RF fields concluded that human occupational exposure to RF and microwave fields between 0.1 and 1.0 mW/cm2 incorporated a sufficient safety factor so as not to lead to adverse effects.

Skin Rashes

Several incidents of skin rash (especially face rashes) in VDT operators have been reported in Norway, Sweden, and England (W. C. Olsen, 1981; Nilsen, 1982). Attempts to correlate these generally isolated findings to VDT use have been met with strong criticism since VDTs emit neither chemical contaminants nor any form of radiation that would explain the rashes. Unfortunately, studies have not compared the incidence of skin rash in VDT workers with that in non-VDT workers. Nilsen (1982) argued that the static electric potential that could build up in dry air between an operator's body and a CRT screen would attract airborne particulate contaminants to the skin.

It is possible that under certain conditions airborne contaminants (e.g., fiberglass particles) could be precipitated out of the air onto the skin by electrostatic charge induced in the operator's body, but this has not been demonstrated to occur at VDT workstations. If this effect were to occur, it would probably be most likely to appear as contact dermatitis when an operator touched a CRT screen, where the electrostatic potential difference would be comparatively high. This effect would probably occur only for operators who are very sensitive to whatever airborne contaminants might be in the office. W. C. Olsen (1981) suggested that synthetic fiber carpets not given antistatic treatment (and thus a source of electrostatic charge) might contribute to the occasionally reported incidents of face rashes.

Reported incidents of skin rash have been rare, probably because three conditions would be necessary for such an effect to exist (if it exists at all): dry air to permit charge buildup; presence of airborne contaminants, which would precipitate out on a charged surface; and skin sensitivity to the particulates. The combination of static charge, very low humidity, and the presence of airborne contaminants would presumably be uncommon in most properly designed offices. There is also the possibility of psychosocial stress as a causal factor in skin rashes, as reported by House and coworkers (1979) in chemical industry employees.

VDT Use and Cataracts

It has recently been claimed that VDT use causes cataracts (Zaret, 1980a, 1980b). Although no scientific medical study of the issue is yet completed, available data do not support the claim.

Prevalence and Causes of Cataracts

The normal eye contains a crystalline lens that helps focus light on the retina behind it. When the lens contains opacities it is said to be cataractous. Lens opacities involving the central axis of the lens that are sufficient to degrade the optical image and to reduce visual acuity are considered abnormal.

Small, inconsequential opacities in the lens are extremely common; many are probably congenital. As many as 25 percent of the general population may have congenital or developmental lens opacities that do not affect vision (Duke-Elder, 1969). Other minor opacities, often termed precataract, do not in themselves affect vision but are thought to be precursors of senile cataract. These opacities increase in frequency with increasing age: they have been found in 43 percent of people aged 50-64 and 61 percent of those aged 75-85 (Ederer et al., 1981a).

Visually disabling cataracts are far less common. In a sample of nondiabetics in the United States,5 the prevalence of cataracts that ''entirely account for a visual acuity deficit of 6/9 (20/30) or more" has been found to be 1.3 percent in people aged 50-64; aphakia, the result of surgical cataract extraction, is found in an additional 1.4 percent (Ederer et al., 1981a). For people aged 75-85, cataracts were found in 22.6 percent and aphakia in an additional 9.0 percent (Ederer et al., 1981a). Of all people with central lens opacities (cataracts), however, less than 7 percent have vision deficits of more than 6/60 (20/200) (Milne and Williamson, 1972).

The causes of most cataracts are not known. Some cataracts are present at birth or develop in early childhood and are usually related to intrauterine infection (e.g., congenital rubella) or inborn errors of metabolism (e.g., galactosemia). Cataracts whose onset is later in life are related to trauma, metabolic and degenerative disorders (e.g., juvenile diabetes, hypoparathyroidism, myotonic dystrophy, high myopia, retrolental fibroplasia), or exposure to cataractogenic agents (e.g., high levels of ionizing or infrared radiation, chronic steroid use, ocular surgery). However, the known causes of cataract account for only a tiny proportion of all cataract cases; the vast majority of visually disabling cataracts are associated with aging (Duke-Elder, 1969). It is thought, but not proven, that age-related changes in glucose metabolism (Caird, 1973; van Heyningen, 1975), chronic exposure to high levels of ultraviolet radiation (or other components) of ambient sunlight (Taylor, 1980), and dietary factors (Chatterjee et al., 1982) may be important in inducing lenticular opacities. The wide variation in cataract rates that are reported from different parts of the world certainly suggests that environmental agents may play a role in their pathogenesis (Sommer, 1977; Zigman et al., 1979; Taylor, 1980).

The Evidence Regarding VDT Use and Cataracts

The Claims

A single investigator has presented two brief reports on a total of 10 patients said to have developed cataracts from the use of VDTs (Zaret, 1980a, 1980b). Neither report has been published in a refereed scientific journal. The 10 cases seem to have been referred to the investigator specifically because they were VDT (or radarscope) operators with lens opacities.

The information given in the two reports is not sufficient adequately to evaluate the clinical conditions of the 10 patients. However, it appears that in 6 of the patients the "cataracts" (termed "incipient" by the investigator) were actually inconsequential opacities that did not appreciably reduce visual acuity. Of the four patients with significant lens opacities:

  • One had been a radar technician for 15 years and had required cataract surgery on his right eye long before he had ever used a VDT.
  • One, with unilateral nuclear sclerosis6 at age 53, had had extensive X radiation to her face early in life.
  • One had already had one eye removed for retrolental fibroplasia (RLF); the remaining eye, with RLF, high myopia, and previous ocular surgery, had a nuclear sclerotic and posterior capsular cataract.
  • One, a 54-year-old with a unilateral "mature" lens, had used a radarscope for 24 years.

In summary, of the 10 anecdotal cases, only 4 had significant lenticular opacities, and each of them had known preexisting disease or exposure to cataractogenic agents.

Zaret has also claimed to have seen a total of 500 people with cataracts due either to exposure to microwaves or to VDT use, but discloses neither the proportion he attributes to VDTs nor any additional details by which his claim can be assessed (Zaret, 1981). He claims that the reported opacities had the typical appearance of radiation cataracts, but the descriptions given do not support the claim. It should be noted that the appearance of radiation cataracts (typically posterior capsular and cortical cataracts7 ) is not pathognomonic.8 Posterior capsular and cortical cataracts may occur idiopathically without any recognized environmental exposure.

Response to the Claims

The weight of available evidence indicates that an association between VDT use and development of cataracts is highly unlikely:

The 10 cases reported by Zaret (1980a, 1980b) span a wide age-range in which idiopathic cataracts (and insignificant opacities) have been commonly recognized since long before the invention of VDTs. Since many people have lens opacities and many people use VDTs, it is hardly surprising that some VDT users have lens opacities. As VDT use increases, so, inevitably, will the number of people with cataracts who use them.

With regard to Zaret's claim to have seen 500 VDT- or microwave-caused cataracts, even if some of the people had significant lens opacities and had been chronic VDT users, it would not prove that their cataracts were either associated with or caused by VDTs. Such a claim would require demonstration that those cataracts are of a peculiar, unusual nature or that cataracts occur more frequently in VDT users than in nonusers. One can always collect a subgroup of people with cataracts who also have in common some occupation, avocation, or other attribute.

Short-term experimental studies of animals indicate that the types and levels of radiation emitted by VDTs are highly unlikely to produce cataracts. Studies of accidental or occupational exposure of humans are consistent with this conclusion (see "Biological Effects of Radiation" above). Studies of long-term (years) exposure to radiation of the wavelengths and levels emitted by VDTs have not been done, and few studies of chronic exposure exist for radiation of any wavelength. In two studies of long-term exposures to UV or microwave radiation many orders of magnitude higher than those produced by VDTs, some lens changes were reported with UV radiation in mice (Zigman and Vaughan, 1974), but microwave irradiation of rabbits had no lenticular effect (Ferri and Hagan, 1976). Based on present knowledge and as discussed above, it appears that the radiation emitted by VDTs would be insufficient to cause ocular damage even with prolonged exposure.

Methods of Studying Whether There Is a Relationship Between VDT Use and Cataracts

The question is not whether VDT users can also have lens opacities or cataracts, but whether VDT use increases the risk of developing a cataract. (If it did, one would still have to determine whether VDT use actually caused the cataract.) In principle, experimental studies on animals or epidemiological studies or both could be used. In an experimental study, monkeys (whose lenses have radiosensitivity similar to that of humans) might be chronically exposed (for years) to VDTs and examined periodically (with slit lamp and ophthalmoscope and by histologic examination). We believe the value of such a study is dubious, however, given that the effects of radiation have been studied extensively (see above), and levels of radiation many orders of magnitude higher than those emitted by VDTs have been found necessary to produce cataracts. For similar reasons we doubt that extensive epidemiological studies are warranted.

If epidemiological studies are undertaken (two recent pilot studies are discussed below), they must be appropriately designed to be of value. In general, three types of epidemiologic studies can be done to assess the absolute or relative risk of cataracts associated with VDT use (Lilienfeld, 1971; Sommer, 1980): concurrent longitudinal studies, nonconcurrent longitudinal studies, and case-control studies.9

Concurrent Longitudinal Study

Two groups of individuals, one randomly assigned to use VDTs and the other not, are followed with periodic examinations for the development of lenticular opacities. In such a study, care must be taken in ensuring that the two groups are as much alike as possible (age, sex, work and home environments, general and ocular health, etc.) to limit the effects of confounding variables. This is the most definitive type of study available, since one knows the baseline status of both groups and can therefore exercise tight control over selection before they are placed at "risk" of developing differential cataract rates. This is the only kind of study that provides true incidence data (the rate at which cataracts develop per unit of population per unit of time, e.g., per 100,000 per year). Its major limitation is the need for long-term follow-up, which delays results and increases costs.

Nonconcurrent Longitudinal ("Cohort") Study

In this form of study, rather than selecting two groups and following them over time, a group of VDT users and a carefully matched group of non-VDT users are examined once and the prevalence of cataracts in the two compared. As the investigator does not know the baseline cataract prevalence preceding VDT exposure and has less control over the selection and matching of the groups, the results are more susceptible to a variety of biases (Hill, 1971; Lilienfeld, 1971; Sommer, 1980). The obvious advantage to this type of study is that one need not follow subjects for years, which reduces costs and provides earlier results.

Both concurrent and nonconcurrent longitudinal studies pose problems: It is generally impractical to randomly assign workers to use or not use VDTs, and one cannot eliminate the possibility of confounding variables (bias) in any matched sample. Although there is no reason to suspect this would be the case, individuals already with, or at higher risk of developing cataracts, might be more likely to seek out (or alternatively, avoid) VDT use. For example, diabetics might be selected for more sedentary (e.g., VDT) work. Because diabetics are already at increased risk of developing cataracts, a study of this sort might demonstrate that VDT users had a higher rate of cataract development than nonusers, not because users were exposed to VDTs, but because they contained a larger proportion of diabetics.

More generally, a practical, standardized, reproducible method for demonstrating and quantifying the presence of lens opacities is yet to be developed, complicating the diagnosis and comparison of "cataract" rates in population groups. Existing techniques for comparing cataract rates are far from ideal, and have either compared the presence of any opacities, opacities of specified types, or the presence of opacities combined with a reduction in visual acuity of specified amount with or without exclusion of patients with other forms of ocular pathology (Kahn et al., 1977; Ederer et al., 1981a, 1981b; Sommer, 1981).

In addition, choice of an appropriate sample size is complicated by the fact that neither the "background" prevalence or incidence of the various types of lenticular opacities is known nor the magnitude of the risk that VDT use might add to it. Because the background rate of cataracts is expected to be low, the need to prove the absence of significant risk requires use of enormous sample sizes. For example, if one accepts the estimate that 4 percent of non-VDT users have the particular asymptomatic lens opacities of the type claimed for VDT users and decides that a 25 percent increase in risk is "significant" (i.e., an additional 1 percent for a total prevalence of 5 percent), a cross-sectional study requires studying at least 12,000 VDT users and 12,000 (preferably 24,000) well-matched nonusers to disprove a clinically meaningful association with an alpha error (two-tailed) of 0.05 and a beta error (one-tailed) of 0.20. With smaller sample sizes, an alpha or beta error, or both, is increased, and only greater risks of cataract development can be detected.

Only if VDT use produced a specific, unusual form of cataract (such as changes associated with acute exposure to high levels of ionizing radiation) in a large proportion of people might differences in prevalence (and incidence) be detectable in samples of smaller sizes. And to date, there has been nothing unusual in the type of cataract described among VDT users.

Case-Control Study

A third type of epidemiologic study is available that is more economical in terms of sample size. Rather than comparing the rate of lens opacities among VDT users and nonusers, a casecontrol study compares the rate of VDT use among people with and without lens opacities. Unfortunately, such an approach has severe limitations: it would be difficult to devise a technique for identifying asymptomatic opacities since people do not usually seek medical attention for asymptomatic conditions; and so such a study would, of necessity, be limited to cases of symptomatic cataracts. In addition, selection of appropriate controls is particularly difficult within this study design.

Well-designed longitudinal (prospective and nonprospective) studies could be carried out, but would require careful attention to selection of the study population and large populations; consequently, they would also be very costly. Ideally, several investigations would be performed: by using different groups and multiple controls and looking for dose-related changes in risk, the problems of confounding variables and hidden bias could be largely dealt with.

Two Ongoing Studies

One pilot epidemiological study has recently been completed and another study is presently under way. Both suffer from the potentially biased volunteer nature of the study population and from small sample sizes.

The NIOSH Baltimore Sun Study

The Baltimore Sun VDT study was carried out by NIOSH (Appendix B includes a detailed review of the preliminary results of this study). The study sought to examine the relationship between refractive errors, demographic variables, ergonomic factors, and somatic complaints, as well as the relationship between cataracts and VDT use. Sample size requirements for the cataract analyses were calculated to be 511, assuming a "background" prevalence of 5 percent and a doubling of this rate (to 10 percent) by VDT use. It was recognized, however, that this expected background prevalence and the expected impact of VDT use were both high. The major limitations of the study related to the sample under investigation.10 Of 588 workers listed on the Newspaper Guild roster, only 283 (48 percent) participated in the full examination. This very high degree of self-selection introduces a large degree of potential bias. Of the participants, 71 percent were current users of VDTs, leaving a very small number of controls. In addition, there was no information on previous VDT use by these controls, an important omission if one wished to compare effect of VDT use (presumably a dose-related phenomenon) on cataractogenesis. Because of these limitations, particularly the small and self-selected nature of the study population, the investigators concluded that the study did not permit an assessment of a relationship between VDT use and development of cataracts. Nonetheless, the study does provide some extremely interesting and valuable information on the relationship between somatic complaints and ergonomic factors for which the sample sizes are adequate (see Appendix B).

The Mt. Sinai Study

A second study is being carried out by the Mt. Sinai School of Medicine for the Newspaper Guild. This study has a wider scope and is covering nonophthalmic as well as ophthalmic conditions. The sample size is larger than in the Baltimore Sun study (4,000 workers); it is being done by mailed questionnaires, but consideration is being given to conducting an eye examination on a subsample of the total study population. Until study procedures are final and the composition of the examined group known, it is impossible to predict the degree to which this study will answer the question of cataract risk from VDT use.

Conclusions About Radiation Hazards

The variety and depth of the studies reviewed here represent a comprehensive study of the issues and lead to our conclusion: Present knowledge indicates that the levels of radiation emitted by VDTs are highly unlikely to be hazardous to health.

Perhaps of equal importance, the emission levels from VDTs are far below those emitted by many common electronic products or those present from natural sources in the environment (Smith and Brown, 1971; United Nations Scientific Committee on the Effects of Atomic Radiation, 1977; World Health Organization, 1979; National Research Council, 1980; Sliney and Wolbarsht, 1980). Even if a large number of VDTs were arranged near each other in an office, the summed levels of radiation would be less than ambient levels of radiation from other sources. Although there are differences around the world in standards for occupational and public exposure, by none of the standards would VDT emissions be considered cause for concern (Acton and Carson, 1967; American Conference of Governmental Industrial Hygienists, 1981; National Council on Radiation Protection and Measurements, 1981).

Results from the Mt. Sinai study may suggest whether larger, more definitive studies of cataracts among VDT users are likely to prove productive. Until that time, however, because of the lack of evidence suggesting a real association between VDT use and visually disabling cataract, and the extraordinary size, complexity, and cost of the definitive study needed to disprove the possibility of such an association, it would seem unreasonable and unjustifiable to embark on such studies.



In 1967 national attention was drawn to the emission of X radiation from color television receivers. Some color televisions and some VDTs in use at that time used high-voltage shunt regulators that emitted higher than acceptable levels of X radiation (see Bureau of Radiological Health, 1981:6); the shunt regulators were redesigned to reduce radiation leakage. Solid-state circuitry has now eliminated the use of shunt regulator tubes in color televisions and VDTs, and only the cathode-ray tube (CRT) remains as a potential source of X radiation. As noted above, the face of the CRT is shielded to prevent unacceptable levels of X radiation from passing outward.


Personal communication, William A. Herman, Associate Director, Division of Electronic Products, Bureau of Radiological Health, July 22, 1982.


The rad is a unit of absorbed dose corresponding to a deposition of 100 ergs/gram of tissue. The roentgen, R, is a unit of exposure dose of X or gamma radiation that would produce ions with 1 electrostatic unit of charge per 0.001293 g of air. For practical purposes 1 R produces about 0.876 rad in tissue.


Assuming 40 hours work per week in close proximity to VDTs, 52 weeks per year.


Diabetes mellitus is frequently associated with the development of senile cataract.


Hardening of the central portion of the lens. In advanced cases, nuclear sclerosis can lead to nuclear cataract.


Donald Pitts, College of Optometry, University of Houston, remarks at the Symposium on Video Display Terminals and Vision of Workers, Washington, D.C., August 20-21, 1981.


David G. Cogan, Chief of Neuroophthalmology, National Eye Institute, personal communication, August 6, 1982.


The terminology used here follows the convention used in epidemiology; for further discussion see Sommer (1980).


Though primarily targeted at local Newspaper Guild members, a small sample of non-guild workers was also included because of a lower-than-expected rate of participation by the guild members.

Copyright © 1983 by the National Academy of Sciences.
Bookshelf ID: NBK216487


  • PubReader
  • Print View
  • Cite this Page
  • PDF version of this title (4.3M)

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...