Environmental Neurotoxicology (1992)

Neurologic disease and dysfunction after exposure to toxic chemicals in the environment could be prevented at least in part through premanufacturing and premarket screening (Chapter 4) of all newly synthesized chemicals, coupled with retroactive screening of chemical substances already in commerce. Such screening is intended as primary prevention; it is meant to prevent neurologic disease and dysfunction of environmental origin by identifying new neurotoxicants before they enter the environment and before human exposure has occurred and then handling them in such a way that human exposure will be as close to zero as possible.

Current American programs for detecting neurotoxicants, including industrial chemicals, through premarket testing have serious deficiencies (NRC, 1984). A major problem is that many potentially neurotoxic compounds that came into use before passage of the Toxic Substances Control Act (TSCA) in 1976 remain untested and are still not required to be tested. Current neurotoxicity testing procedures developed under the TSCA mandate have not been standardized for new industrial chemicals (other than pesticides) and are not sensitive enough to detect the effects of chronic low-dose exposures. Neurotoxins in natural products are most likely to escape testing and control, in part because of confusion as to which federal agencies are responsible for this control. Primary prevention is far from perfect; it has sometimes failed in the past, and it cannot be expected to be totally successful in the future.

The early identification of disease or dysfunction and acting on the identification constitute secondary prevention. Its goals are to prevent progression of disease in identified cases, to treat early enough to cure if that is possible, and to prevent the occurrence of additional cases by providing a basis for primary prevention. The methods of postmarketing secondary prevention of environmental neurotoxic disease include environmental monitoring and surveillance of the human population to detect neurotoxic disorders as early as possible, which are combined with efforts to keep potentially toxic exposures to a minimum. Epidemiologic studies and clinical screening are complementary facets of maintaining alertness so that neurotoxic illness will not go unnoticed. Secondary prevention constitutes a necessary and critical backup to primary prevention.

Secondary prevention is not a substitute for premarket testing of chemicals for potential neurotoxicity; rather, the data from a secondary prevention program complement and supplement the data derived from premarket laboratory testing. Although toxicologic information derived from future in vitro and in vivo investigations might be expected to form an ever-increasing proportion of the neurotoxicity data base, responsibility to conduct epidemiologic surveillance, to monitor toxic exposures, and to study incidents of human exposure to neurotoxicants when they occur will remain.

Surveillance and monitoring play oversight or exploratory roles. Fundamentally, they should be viewed as tools for identifying possible problems-hypothesis-generating exercises. In contrast, classical epidemiologic studies of populations exposed to neurotoxicants (e.g., Needleman, 1986), clinical trials of pharmaceuticals, and occasionally laboratory experiments on humans (e.g., Dick and Johnson, 1986; Johnson, 1987) are hypothesis-testing and will not be considered here in detail.

Effective surveillance systems are characterized by systematic and rapid collection of data, efficient analysis and evaluation of data, and prompt communication of the results obtained. In discussing general principles of health surveillance, Foege et al. (1976) noted that dissemination of a program's information often enhances cooperation in contributing more information to the system.

Surveillance of occupational and environmental health can be divided broadly into systems for monitoring of exposure (to detect hazards) and monitoring of disease (to detect health responses). Hazard surveillance systems monitor the use, storage, and environmental release of chemical toxicants and tabulate the numbers of people exposed to them. Their purpose might be to detect instances of poor industrial hygiene, inappropriate or excessive use of pharmaceuticals, or improper waste disposal or to generate exposure data to complement epidemiologic and clinical data on toxic end points. Health surveillance systems record the occurrence of designated diseases. They might have the general population as their target or focus on specific high-risk groups (Baker, 1989; Landrigan, 1989). Both types of surveillance are necessary, and both are practiced by public-health agencies in the United States and elsewhere; the data generated by the two types of systems are complementary. Surveillance systems can be categorized according to their purpose: to monitor the handling of known neurotoxicants, to follow the effect of control measures, or to detect previously unrecognized problems.

Recognizing a previously unrecognized problem is most difficult, because no system can be expected to find reliably what is not specifically sought or to engender confidence that something not specifically looked for is really not present. Because the measurement of environmental concentrations of chemicals requires discrete decisions about what substances will be sampled and where and when, it is clear what information has and has not been sought. Medical monitoring can be somewhat more open-ended, and that raises the possibility of uncertainties about the definitiveness of negative findings; but the storage and retrieval of observations necessitate specification of format and limitations on the type of information accepted. In addition, having a data base truly useful for quantitative estimation of occurrence rates demands that the means by which the data were ascertained be clearly defined, so that it is known what the sample represents.

A perennial problem in epidemiologic surveillance is that information on the baseline incidence and prevalence of neurologic diseases is sparse. Thus, it is difficult to detect anything but large outbreaks or clus-

ters of neurologic disease. The development of a substantial data set would require significant resources. Much knowledge about neurotoxicants and neurotoxic disease has, however, come from observation of humans—from case reports to large-scale epidemiologic studies. Most reports of outbreaks of neurologic diseases have come from medical clinics, state health departments, industrial medical groups, or consultative programs, such as the Health Hazard Evaluation Program of the National Institute for Occupational Safety and Health (NIOSH). Outbreaks have all too often served as late-stage sentinel events identifying the neurotoxicity of chemical substances. Many times, apparent excesses or clusters of disease have, on close examination, proved to be false leads.

In Chapter 1 (Table 1-1), the great diversity of forms that neurotoxicity can take was demonstrated by the extensive list of neurologic signs and symptoms attributed to exposure to various chemicals. Table 5-1 shows that a temporal pattern of exposure and response can characterize particular chemicals. A complete neurologic examination and occupational history (Table 5-2) are essential for identifying specific neurologic syndromes (Johnson, 1987). The examination summarized in Table 5-2 is typically supplemented by tests of CNS function, which can be as simple as asking the subject to name the current president of the United States or asking the subject at the end of the examination for three words (e.g., hat, dog, and tree) that were presented at the beginning of the examination. More sophisticated neuropsychologic tests can be used when there is reason to suspect CNS problems.

Testing human subjects for neurotoxic injury combines clinicians' efforts to identify the cause of diseases or disorders and neuropsychologists' or experimental psychologists' use of techniques to assess performance. Field investigators might adopt a unique battery of tests for each chemical to be assessed or select from among the 5-10 standardized batteries now in use. If latitude is possible in selecting tests, selection is based on signs reported in the exposed group and the known neurotoxic effects of the chemical under study (or of structurally related chemicals). That general approach has led to the use of hundreds of tests in various worksite studies (Johnson and Anger, 1983; Anger, 1990) and has typified NIOSH research (Anger, 1985). Several human neurotoxicity test batteries have been developed and are being validated. Otto and Eckerman (1985), Anger (1990), and Johnson et al. (1990) have described the major batteries, including several implemented on computers.

Finland's Institute of Occupational Health pioneered neurotoxicity testing at the worksite in the 1950s. Investigators at the institute developed a test battery well adapted to monitoring the main concerns in Finnish industry (primarily solvents). Their tests (Table 5-3), which have been optimized by factor analysis over the years, appraise various psychologic domains (Hänninen and Lindström, 1979). The battery is now used in neurotoxicity evaluations of several worker groups in Finland, including prospective studies on new workers.

An apical test battery (Table 5-3) based on information-processing theory (Marteniuk, 1976) has been developed and tested in the field in Australia (Williamson et al., 1982; Williamson, 1990). For example, battery and lead-smelter workers with blood lead concentrations of 1.2–3.9/µmol/L (25–81 µg/dL) demonstrated poorer performance on all tests than a set of matched controls (Williamson and Teo, 1986).

At a 1983 meeting sponsored by the World Health Organization (WHO), an international team of experts on testing for

neurotoxic effects in humans recommended the Neurobehavioral Core Test Battery (NCTB). The NCTB (Table 5-3) can be used to identify a broad range of neurotoxic effects. The primary goal was to generate uniform, more consistent data from a broad spectrum of occupations and neurotoxic exposure conditions. Tests included in the set had been used successfully in worksite studies (they had identified group differences associated with chemical exposures) and were believed to reflect a wide range of functional areas in humans. A supplemental series of tests was specified for further characterization (Johnson, 1987).

The Neurobehavioral Evaluation System (NES), available on IBM PC and portable Compaq computers, combines several neurobehavioral tests that have been used successfully in clinical settings or in field studies (Table 5-3). Tests are occasionally added to the battery, which is following an evolutionary course dictated by current interest in the field. The NES includes variants of five of the seven WHO NCTB tests, and the developers of the NES recommend that the user select tasks to be used according to specific exposure situations (Baker et al., 1985). That is, the battery provides a menu of tests without an explicit decision strategy for selecting among them for screening or other purposes. The NES has been used in more laboratory and field studies than any other battery in the last 5 years (Letz, 1991).

Anger (1989) matched the components of the WHO NCTB, Finnish, and NES batteries to the neurobehavioral effects that he had found to be reported most commonly after chemical exposures (Table 1-1, Chapter 1). He concluded that the NCTB would be the most comprehensive in detecting those effects. The Finnish battery lacks the ability to detect affective changes, and the NES does not have an established test of motor function. Although cognitive and, to a smaller degree, motor effects should be recognized fairly well by all three batteries, they all have limited potential for detecting sensory deficits. All three batteries include tests to assess the more subtle CNS deficits produced by lower exposures, which are the subject of increasing concern.

The tests noted above have been developed primarily for screening, but other tests have been used to characterize the effects found in the screen. They are not fundamentally different, but add information about the nature of the problem. They include laboratory-based behavioral tests, clinical neuropsychologic examinations, and imaging evaluations. Some neurologic and behavioral tests also have been adapted to the monitoring for early identification of disease or indicators of neurotoxicity. Screening and monitoring tests have been developed to identify peripheral problems (such as toxic distal axonopathies) resulting from exposure to n-hexane, methyl n-butyl-ketone, acrylamide, and related neurotoxicants. The neurologic examination often identifies neurotoxicity at an early, but clinically observable, stage. Specific monitoring or screening devices (e.g., the Optacon and the Vibratron II for assessment of vibratory sensitivity and NTE II to detect losses in temperature sensitivity) have successfully identified neuropathy at the preclinical stage (Arezzo and Schaumburg, 1980, 1989). A more specific test has been used to assess vibratory sensitivity in monkeys, and it is now used in the workplace to identify peripheral neuropathy (Arezzo et al., 1983).

Tests of auditory threshold have been used to detect hearing loss after exposures to noise and to drugs that affect the auditory system (e.g., quinine, arsenic, and glycoside antibiotics) (Stebbins et al., 1987). Similarly, visual function has been found to be impaired in animals and people exposed to methylmercury (Evans et al., 1975). Each of those tests has been used to identify early sensory loss so that exposures can be stopped before the development of irreversible effects.

Although there is debate about some screening tests for field applications, the

tests can identify some forms of peripheral neuropathy when they are still reversible. Some companies test routinely to monitor for neuropathy alleging the sensory axons—a kind of secondary prevention system. Of course, such strategies cannot be applied to the testing of newly developed chemicals, to chemicals with effects that are not clearly defined, or to situations in which it is not known what chemical, if any, could cause any health effect.

It can be difficult to interpret the results of test batteries. On the whole, they do not yet have established norms based on extensive population testing, but they do provide objective measures suited to test-retest (before-and-after) assessments. Baseline (pre-exposure) performance data have not been routinely collected from workers before exposure, so unexposed people must be used for comparison. These performance effects are now usually evaluated not by self-matching or against established standards, but by comparison to a referent or control group.

The immense diversity of behavior makes it difficult for a simple test battery to screen adequately in a comprehensive manner; a battery cannot detect an effect for which it lacks a test. Although it is not likely that an entirely new type of health effect will be generated by a new chemical or class of chemical, it is likely that tests for additional end points will be needed as more chemicals are tested. As understanding of the various mechanisms of neurotoxicity increases, confidence that a comprehensive battery can be assembled should grow.

The established batteries of human tests are clearly limited in their screening capacity. No comprehensive rationale for test selection has been posited. The Australians Williamson et al. (1982) developed their screening battery around cognitive information theory, but this theory is now dated and is insufficiently related to medical approaches to neurologic evaluation or to neuropsychologic approaches. Eckerman et al. (1985) proposed a battery in which the tests are selected to assess eight cognitive functions identified by factor analysis; this battery has not been used in worksite evaluations. The field needs to develop a comprehensive rationale for the development of a neurotoxicity screening battery. Ideally, it would be based on the functions of the nervous system and the range of potential neurotoxic effects, and that is not yet possible.

In the absence of a comprehensive rationale for the selection of tests to compose a neurotoxicity screening battery, a good rationale would be to select tests that are sensitive to the range of neurotoxic effects that have been identified. In a comprehensive review of the literature, Anger and Johnson (1985) identified more than 120 neurotoxic effects. The range of neurotoxic effects might be far too broad to test comprehensively. Anger (1986) reviewed the list of 120 effects and identified the 35 reported as occurring after exposure to 25 or more chemicals (Table 1-1). The result formed a basis for selecting the most important functions to assess, given current knowledge. None of the batteries assesses sensory effects thoroughly, and the computer-based NES does not use a proven test of motor performance. However, the NES and some other batteries have the advantage of being administered by computer, which reduces the costs of administering the tests substantially. Some of the most common neurotoxic effects—particularly some forms of peripheral neuropathy and affective symptoms, weakness, ataxia, and many sensory effects—would not be detected by any of the batteries. (Of course, CNS changes can be correlated with some of these effects, in which case these batteries would be performing their function of detecting health effects. As noted earlier, a screening battery is developed for detection, not for characterization.)

The WHO NCTB contributors first suggested human neurotoxicity testing based on

the core-test idea, which uses a standard set of tests in all studies, supplemented by additional tests when appropriate (this is a variant of the tier-testing idea) and is consistent with the idea of evaluating the most common health effects. The NES can be used to test most of the same functions tested by the WHO NCTB, and new tests are added periodically. That will allow adaptation to new problems as they arise. The tests of the NES are largely cognitive and so will approach the complex functions that are not identifiable without sophisticated tests, although tests of motor and sensory function are certainly required for truly comprehensive assessment. Both the NES and the NCTB have been used in different countries and languages; that fosters the development of an international data base that will allow exchange of test results.

Although those test batteries appear to be best suited to assessments of neurotoxic exposures in humans, their usefulness needs constant examination in light of new types of neurotoxicity, possible changes in the most frequently reported health effects (as chemical use shifts), and new cognitive or neurobiologic theories of nervous system function. That is, NES and NCTB batteries are the best choice today, but this judgment is based on identified health effects, rather than on understanding of the nervous system and how it functions. Fundamental progress will be based on advances in basic and theoretical neurobiology; data from continued testing of exposed population can be expected to lead only to minor improvements in test selection. An intermediate goal is the collection of data with a small number of reliable and valid test batteries, to provide a solid base of relatable data on various chemicals. Such a collection could lead to the development of specific criteria for prediction of effects (most desirably on the basis of chemical structure) or the development of a better screening battery.

Several sources of information provide data on potential for occupational exposure to neurotoxic substances in the United States. However, none of them is comprehensive, represents a well-defined worker population or industrial setting, or corresponds to any existing set of information on adverse health effects (NRC, 1987a).

The Integrated Management Information System (IMIS) of the Occupational Safety and Health Administration (OSHA) contains information obtained as a result of inspections for accidents or complaints, followups on violations, and general scheduled inspections. The only exposure data systematically entered into the IMIS are data obtained through analysis of environmental samples taken at worksites where citations were issued. Because IMIS samples are not random samples, generalizable information on the distribution of neurotoxic occupational exposures does not exist.

Industry-generated data constitute a second source of information on neurotoxic hazards. For 11 toxic substances (including lead) for which OSHA has set standards, industry is required to maintain records of exposure. However, because the records are not part of a central data system, such as the system maintained by OSHA, they are most useful for monitoring specific industries and employers and are far less useful for surveillance and discovery.

NIOSH conducts health hazard evaluations (HHEs) at the request of employers, employees, or their representatives. At times, both exposure monitoring and biologic sampling are done. The records provide insight into particular occupations and industries, but, like the OSHA data, they are not readily generalizable. NIOSH has conducted two national surveys of chemical exposure in nonagricultural businesses: the National Occupational Hazard Survey (NOHS) from 1972 to 1974 and the National

Occupational Exposure Survey (NOES) from 1981 to 1983. Each entailed walk-through inspections of approximately 5,000 plants, and potential exposures were tabulated by trained observers. However, no industrial-hygiene samples were collected. As a result, data on actual exposures are not available. The NOHS-NOES data have been used to identify general industrial groups in which neurotoxic exposures may be anticipated, and they have served as a valuable resource for epidemiologists and public-health officials (ASPH, 1988). Information on exposure of workers and consumers to pesticides is available from the Environmental Protection Agency (EPA).

Evidence of exposures to toxic substances in the general environment might be detected by the National Human Adipose Tissue Survey (NHATS), periodic sample surveys done by the National Center for Health Statistics (the National Health and Nutrition Examination Surveys, or NHANES), and toxic inventories maintained by states and cities under right-to-know legislation. Those data bases provide information on a range of chemicals—some that are known to be neurotoxic (e.g., lead, surveyed under NHANES II [NCHS, 1984]), some that accumulate in fat (such as pesticides, surveyed in the NHATS), and some that are stored and released in substantial quantities in industrial and other sites. The National Research Council has published a review of the NHATS program and made several recommendations for its modification and improvement (NRC, 1991b).

A particularly valuable new source of information is the Toxic Release Inventory (TRI), compiled annually by EPA since 1988 to monitor more than 300 chemicals released by industry into the environment. The inventory is mandated under the Superfund Amendments and Reauthorization Act (SARA). Seventeen of the 25 toxic substances with the highest volumes of release into the environment have neurotoxic potential (OTA, 1990).

Efforts are also needed to improve the availability and use of exposure-surveillance data. IMIS, the OSHA inspection data base, would be more useful if it covered all states in a more uniform fashion and if it included a larger, more representative sample of exposure data. Other types of exposure-reporting systems (such as those under federal and state right-to-know laws) should be evaluated for their utility in surveillance. For the preponderance of industrial settings, adequate exposure data are not available. Such information should be collected according to explicit sampling plans with up-to-date measurement techniques and stored in accessible data bases. It could then be used to target efforts to control known dangerous exposures or matched to health-effects files in an effort to detect new problems, as discussed below. Exposure surveillance is a continuous effort both to monitor changes in exposure and to monitor new industries or new uses of well-known chemicals (Fowler and Silbergeld, 1989).

When the objective is to assess broad trends in population exposure to neurotoxicants and to gather information on groups potentially at risk, selection of an appropriate surveillance strategy will depend on the expected degree of toxicity of the chemicals under study, the size of the exposed population, and the intensity of exposures. In essence, hazard or exposure surveillance consists of obtaining information on the distribution and use of chemicals, on the distribution of exposures, and on the size and distribution of the exposed population.

A disease-surveillance system directed toward detecting problems caused by unrecognized neurotoxicants must be much more open and versatile than a system intended only to tabulate the incidence and prevalence of and mortality from a known

disease. Neurotoxic disorders are particularly difficult to monitor (NIOSH, 1986; Friedlander and Hearne, 1980), and present surveillance systems are not very effective. Neurotoxic disorders are easy to overlook or misdiagnose, because their signs and symptoms often develop slowly and subtly. And some can be reversible or self-limiting. The patterns of effects produced by many neurotoxic chemicals are similar, consisting of common, nonspecific complaints, often mimicking diseases with other etiologies. The association between an observed health effect and the culpable environmental agent is rarely obvious. Clinical tests used in diagnosis are usually too imprecise or insensitive to identify the full spectrum of adverse health effects. A latent period between exposure and overt response will further complicate determination of causation.

Several existing data systems have been suggested as potentially useful for surveillance for neurotoxic disease that might be attributable to environmental exposures (Gable, 1990). They include vital records, health surveys, specifically designed surveillance systems, and the U.S. Census.

Since 1971, the Bureau of Labor Statistics (BLS) has conducted the Annual Survey of Illness and Injury with a well-constructed (and well-defined) probability sample; the most serious problem with this data base for the detection of neurotoxic illness is that the occurrence of illnesses, especially those involving latent periods, is vastly under-reported. BLS also maintains a Supplementary Data System in which data from the states' workers' compensation programs are compiled; the data gathered differ from state to state, and the lack of total reference populations precludes the calculation of rates.

The National Center for Health Statistics (NCHS) compiles data (from death certificates) on every death in the country through the National Death Index. The basic data on death certificates are fairly uniform across states, and demographic cause-of-death information and some occupational data are almost always available. Occupational data are of uneven quality, which usually depends on whether anyone uses them for research purposes. The categories ''housewife'' and "retired" are often reported in the place of out-of-home employment or one-time major employment. Data from the Bureau of the Census are appropriate as denominators in calculating rates. NCHS makes the encoded mortality data available to researchers on computer tapes. The National Death Index is valuable for determining the vital status of subjects in epidemiologic followup studies; for subjects who have died, researchers are directed to the appropriate state for detailed information. NCHS also collects data from the states on every birth and fetal death, including birthweight, Apgar scores, congenital anomalies, and complications of pregnancy or birth. The birth records are not as uniform across states as are death certificates, and only a few states record information on parental occupation. NCHS has conducted the National Health Interview Survey, a large (120,000 persons), stratified probability sample of the civilian, noninstitutionalized population, every year since 1957. In addition to standard demographic information, both occupational status and data on illnesses, injuries, disabilities, and use of medical services are gathered. The fourth NHANES, a probability sample of the civilian, noninstitutionalized population also conducted by NCHS, is currently being carried out on a full-scale basis (the first was carried out in 1970). Physical examinations, laboratory tests, and responses to a questionnaire yield a detailed medical picture, but occupational information was not gathered in the previous surveys. In the current survey, job histories are being requested, and neurotoxicity is among five work-related conditions about which information is sought.

Other data bases are of potential value for the detection of neurotoxic disease. Regional poison-control centers, insurance

companies, the Social Security Administration, and the Department of Veterans Affairs might have information useful for a specific surveillance project, but most fall short by not having both health and occupational data (data on actual exposures are almost never found), by being drawn from an inadequately defined population, by having restricted availability, or by being limited to a period that is not of interest.

The National Research Council Panel on Occupational and Health Statistics (NRC, 1987a), after reviewing the various sources of information on adverse health events, concluded that the information available on occupational injuries is poor and that the information on occupational illnesses is considerably worse. Those data sources were not designed to determine the causes of disease, nor do they emphasize occupational or environmental exposures. Thus, the contribution of such exposures to the etiology of a given disease is not usually recognized (ASPH, 1988). The end points selected for surveillance are usually restricted to overt clinical disease or death.

Once cases are reported, officials can collect further information and provide appropriate followup and suggestions for prevention in the workplace. The reporting systems must be interactive and provide a means for followup of reported cases to ensure continued physician involvement in the surveillance program. State and local health departments, in collaboration with occupational clinic groups, probably provide the best basis for such reporting systems.

When an excess or cluster of a particular symptom associated with neurotoxicity (National Conference on Clustering of Health Events [1990]) or a new syndrome with involvement of the nervous system is detected, determination of the likely causes is the next stage of the epidemiologic process. It requires careful comparison between observed physical and temporal patterns of exposures to possible toxicants and the occurrence of the cases.

The NIOSH Health Hazard Evaluation Program and similar activities of academic or government groups provide mechanisms for rapid evaluation of reported outbreaks and can therefore confirm initial reports of a possible neurotoxic problem. California has pioneered the development of a reporting system for outbreaks of pesticide poisoning in which collaborating health and agricultural departments can intervene. However, similar systems for the reporting of such diseases by physicians are not well established elsewhere in the United States (ASPH, 1988).

The committee has considered several ideas for improving disease surveillance systems:

• Training of medical professionals. Many medical professionals who are in direct contact with persons with neurotoxic disorders of environmental origin lack the knowledge necessary to recognize these ailments or to identify their etiology. The ability to diagnose the disorders and to associate them with workplace or community exposures must be improved (U.S. House of Representatives, 1986). Physicians need to be made more aware that environmental exposures can produce toxic effects and should be trained to obtain an account of exposures received occupationally, avocationally, etc., as well as medically, when taking a history (Goldman and Peters, 1981). Better initial training in occupational and environmental medicine and the dissemination of more current information on neurotoxic illnesses would benefit both health practitioners and their patients (NRC, 1987a). The development of improved diagnostic tests and criteria would facilitate the recognition of an occurrence of neurotoxicity when it appears.

• Standardized disease definitions. Uniform clinical definitions of the disorders in question are needed to provide a common reporting basis for physicians. Such a nomenclature would also make the assembled

data more amenable to analysis and interpretation. NIOSH has a project to develop definitions for a series of occupational disorders.

• Sentinel health events. A sentinel health event (SHE) is a preventable disease, disability, or untimely death whose occurrence serves as a warning that the quality of preventive or therapeutic medical care might need to be improved (Rutstein et al., 1976). The idea has been refined for application to occupational situations; a SHE(O) is a disease, disability, or untimely death that is related to an occupation and whose occurrence can provide the impetus for epidemiologic or industrial-hygiene studies or serve as a warning signal that material substitution, engineering control, personal protection, or medical care might be required (Rutstein et al., 1983). Wagener and Buffler (1989) showed how NCHS's Compressed Mortality File could be screened with SHE(O) codes to derive mortality rates for as small an area as a county, which would be a step toward identifying geographic areas with increased incidences of sentinel diseases. The industrial or occupational information routinely entered on death certificates is encoded on higher-level, more easily retrievable records in only a few states, however, and that diminishes the usefulness of the SHE(O) idea and of other occupational-surveillance programs that use mortality data. Recently, the SHE(O) principle has been adopted as the core of demonstration reporting projects between NIOSH and 10 state health departments (Baker, 1989). Lead poisoning is one of the eight target occupational-health conditions focused on by this project, the Sentinel Event Notification System for Occupational Risk (SENSOR). This trial surveillance project was described in conjunction with the reporting of a case of adult lead poisoning (blood lead was 170 µg/dl, and reporting is required for concentrations above 25 µg/dl). The index case led to finding secondary cases in the subject's workplace and family (Johnson et al., 1989). The investigators at NIOSH and Harvard Medical School (Rutstein et al., 1983), who developed the original list of SHE(O)s, intended that new associations between toxic environmental exposures and disease conditions be added as they were recognized. The extent to which neurotoxic health effects are indistinct entities will interfere with their conversion into SHEs or SHE(O)s. Neurotoxic conditions that might be added to the list include occupational neuropathy due to Lucel 7 (Horan et al., 1985), paralysis of the urinary bladder after exposure to the NIAX catalyst dimethylaminopropionitrile (Gad et al., 1979; Pestronk et al., 1979), and the Kepone (chlordecone) syndrome (Guzelian, 1982).

• Biologic markers. The complexity and inaccessibility of many parts of the nervous system make it difficult to use biologic markers to study neurotoxicology. But events in physiologic systems that are substantially controlled by neuronal processes, such as some aspects of endocrine and immune function, can be monitored. And objective, computer-based systems for assessment of the function of the central and peripheral nervous systems have been developed to yield biologic markers of neurologic function in exposed populations (Letz and Baker, 1986); the tests are rapid, inexpensive, and noninvasive. Some indicators of exposure to neurotoxicants are measurable in easily sampled media, such as blood, urine, and hair. Surveillance for exposure to neurotoxicants is possible with such easily sampled surrogates, as in the measurement of peripheral esterase in workers exposed to organophosphates to monitor CNS effects indirectly (see also Chapter 3).

• Disease and exposure registries. Groups of people with known or suspected large exposures to neurotoxicants, such as occupationally exposed populations or residents near hazardous-waste sites, can be included in programs of targeted medical surveillance or followup, e.g., by the Agency for Toxic Substances and Disease Registry (ATSDR).

ATSDR is charged under the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA, or Superfund) with maintaining national registries of people exposed to hazardous substances. ATSDR's current registries are of people exposed to trichloroethylene, dioxin, or ß-naphthylamine. The latter two are not regarded as a neurotoxicants, but the program will be expanded and could provide some unanticipated information. Available information from the NOHS (1972–1974) or the more recent NOES (1981–1983), OSHA compliance monitoring, health-hazard evaluations, and similar programs can be reviewed to identify occupational and industrial groups with the potential for substantial exposure. Another group of persons for potential followup consists of patients (probably of HMOs) who are on long-term pharmaceutical treatment regimens with potentially neurotoxic drugs. Depending on the nature of the chemical, targeted medical surveillance might consist of periodic determination of blood concentrations, measurement of cholinesterase, nerve-conduction testing, or other testing of neurologic or psychiatric function.

Prospective surveillance of populations exposed to neurotoxicants in defined settings, such as the workplace, is a promising approach to the early and sensitive detection of functional impairment of the nervous system. The methodologic advantage of prospective surveillance over one-time or cross-sectional surveillance is that each subject can serve as his or her own control. Initial information on a person can be taken to represent baseline or pre-exposure measures of function, and later data can be compared with the baseline data to detect possible toxic injury. In this study design, effects of modifiers, such as drug or alcohol abuse, on the nervous system would be largely canceled out. Thus, compared with cross-sectional evaluations, prospective surveillance should be able to detect neurologic damage at an early stage and at low levels of exposure to such neurotoxic agents as lead in children and solvents in adults. Such surveillance is ideally targeted to the identification of functional impairment while it is still potentially reversible; detection of subclinical impairment can, in such instances, trigger a person's removal from exposure or steps to reduce overall exposures so others will not be harmed.

In another epidemiologic approach to the detection of environmental causes of neurologic disease, time trends or geographic patterns of occurrence of illness are measured. For example, in studies of multiple sclerosis, a clear north-south gradient in incidence is noted, with higher rates at more northerly latitudes. Moreover, people who spend their early years in northern latitudes and then move south seem to carry much of their earlier risk of multiple sclerosis into later life. Those observations suggest some undefined environmental factors in the etiology of this chronic neurologic disease. Likewise, strong time trends have been seen in the incidence of Parkinson's disease and motor neuron disease. The explanation is not known, but again an environmental factor is suggested (Lilienfeld et al., 1989).

Case-control studies of groups with chronic neurologic disease provide another approach to the detection of environmental causation. People with and without a chronic neurologic disease are interviewed to determine whether there are systematic and statistically significant differences in exposure between the two groups (Hertzman et al., 1990). Such studies help to detect diseases of long latency caused by exposure many years in the past.

• Imaging techniques. New methods are becoming available for studying the nervous system noninvasively. Computed axial tomography (CAT) allows visualization of the structure of internal organs; CAT scanning has been used, for instance, to reveal structural abnormalities in the brains of schizo-

phrenics; (Zec and Weinberger, 1986). The recently developed positron- emission tomography (PET) and magnetic resonance imaging (MRI) allow visualization of physiologic and biochemical processes as they occur in the brain (Battistin and Gerstenbrand, 1986). PET scanning has been used in diagnosis of neurotoxicity in people exposed to MPTP (Calne et al., 1985). At present, however, PET and MRI are not routine screening tools.