Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Neurobehavioral Toxicity of Selected Environmental Chemicals: Clinical and Subclinical Aspects Gerhard Winneke The present chapter discusses established as well as controversial associations between human exposure to environmental chemicals and nervous system dysfunction and, in particular, examines if and to what extent psychological theory and methods may contribute to an early detection and evaluation of chemically induced neurotoxic- ity. Because animal models used in neurotoxicity testing for purposes of screening and of clarifying mechanisms of action, are covered in other parts of this volume, this chapter is restricted largely to findings in human populations which describe the outcome of exposure in terms of neurological or psychological dysfunction. Many chemicals are known or suspected to affect nervous system functioning. Anger and Johnson (1985) point out that over 850 work- place chemicals can be classified as neurotoxic, and for 65 of them, the exposed population is estimated to exceed one million (Anger, 1986~; it should be noted, however, that the nervous system may not be the primary target for some of these chemicals. This review must be selective; therefore, only a few chemicals will be dealt with for which human environmental exposure is likely to occur, and for which neurotoxic effects in humans have been reported by using signs and symptoms of neurological or psychological dete- rioration as the endpoint. This is true for some metals (e.g., alumi- num, lead, and mercury), as well as for some organic compounds such as specific solvents and solvent mixtures as well as polychlori- nated biphenyls (PCBs). Secondary neurotoxicity due to carbon monoxide- 226
TOXICITY OF SELECTED ENVIRONMENTAL CHEMICALS 227 induced hypoxia and delayed neurotoxicity of organophosphorus (OF) compounds is not covered in this chapter. CLINICAL VERSUS SUBCLINICAL NEUROTOXICITY Nervous system (NS) diseases may roughly be grouped into the categories of focal or nonfocal syndromes (Schaumberg and Spencer, 1987). This distinction is based on pathology. Typical examples or focal syndromes are the neurodegenerative diseases associated with the process of aging, namely, Parkinson's (PD), Alzheimer's (AD), and motoneuron (MD) diseases (Caine et al., 1986). In PD the pri- mary areas of cell death are the dopaminergic neurons in the zone compacta of the substantia nigra. Typical of AD is the loss of cholinergic neurons in the medial basal forebrain, and the primary focus of MD pathology is the loss of upper (higher brain areas) and lower (spinal cord brain stem) motor neurons. Such circumscribed pathology is typically absent in clinical syn- dromes induced by many neurotoxic chemicals, which therefore may be called nonfocal syndromes (Schaumberg and Spencer, 1987). Clinical manifestations of toxicant-induced damage to the central nervous system (CNS) usually take the form of toxic encrephalopathies, whereas in- sult to the peripheral nervous system (PNS) may give rise to different types of neuropathy or polyneuropathy. Such full-blown clinical syndromes are rare in occupational and environmental exposure to- day, due to increased hygienic awareness, improved preventive countermeasures, and improved early diagnosis. It is recognized however that subtle, insidious, and rather nonspe- cific alterations of NS functioning, often classified as subclinical effects indicative of asymptomatic neurotoxic disease, do occur, may even be widespread under certain circumstances (Schaumberg and Spencer, 1987), and thus constitute a real hazard at the workplace and in the environment particularly with respect to sensitive subgroups of the population. The solvent literature is a rich source of signs and symptoms often carrying the label "subclinical" effects. These may be grouped into the broad categories of subjective complaints, neurophysiological changes, and psychological test results. If the size of the observed effect is sufficiently large, or if manifestations from different outcome catego- ries occur together and thus support each other, clinical syndromes well known from the solvent literature can be diagnosed. Examples, particularly from Scandinavian researchers, are the "psycho-organic syndrome," the "neurasthenic syndrome," or especially in Denmark, "presenile dementia," if the symptomatology is indicative of CNS - `-r-
228 GERHARD WINNEKE involvement or, in case of predominant ENS involvement, neuropathy or polyneuropathy. If, however, one is dealing with rather isolated effects instead, which can only be extracted from background noise by elaborate sta- tistical procedures, the label subclinical is attached to them. What, now, are the main features of subclinical effects? First, and above all, they are usually weak, with individual values typically within the normal range of fluctuation. This creates diagnostic problems and usually means that statistical group comparisions rather than single cases form the source of information. Second, they are essentially nonspecific which, together with their small size, creates problems of interpretation in terms of causality. If, for example, age, alcohol consumption, or deprived social status is also associated with the exposure to neurotoxic agents, it is by no means a trivial task and frequently an impossible one to disentangle those cause-effect strings in which occupational or environmental neurotoxicology is interested. Third, although the observed effects must resemble aspects of the pathological condition (e.g., general slowing of mental speed) to jus- tify their classification as subclinical, they never fulfill all the diag- nostic attributes of the clinical condition. Fourth, for this very reason, it is often difficult to identify their implications for health/adjustment. Just two examples: How relevant is an average slowing of nerve conduction velocity by 4 m/s if this is still within the range of nor- mal variability? How important is an average drop of IQ by 4 points, if it is still well above the expected mean value of 100? These are the types of questions being asked not only in the scientific community but also by health administrators and, above all, by the affected indi- viduals who serve as subjects (Ss) in such studies. After these more general remarks, specific examples illustrate how neurotoxic effects of some environmental chemicals are being charac- terized in terms of neurobehavioral deterioration and the extent to which psychological test methods can be used to detect early, subclinical signs of NS involvement before irreversible damage occurs. Although clinical conditions are described to some extent, the majority of find- ings deal with subclinical neurobehavioral effects in asymptomatic subjects. NEUROBEHAVIORAL EFFECTS OF SELECTED INDUSTRIAL CHEMICALS Chemicals of more widespread environmental impact, for which neurological and psychological effects have been described, are con- sidered here. These include metals namely, aluminum, lead, and
TOXICITY OF SELECTED ENVIRONMENTAL CHEMICALS 229 mercury and organic chemicals namely, polychlorinated biphenyls and solvents. The emphasis here is on general population exposure, although experience from workplace exposure is taken into account whenever it helps to clarify the issue at hand. Potent biological neu- rotoxins from edible plants and animals, which have given rise to endemic neurotoxicity in developing countries (e.g., Spencer et al., 1986, 1987), are beyond the scope of this chapter. Effects in the Developing Nervous System For some chemicals, namely, lead and mercury, children have been shown to be more vulnerable than adults in terms of neurotoxicity. These chemicals are considered first; in addition, PCBs are discussed under this heading because recent evidence suggests that young chil- dren may also be at particular risk for environmental PCB exposure. I=ead Lead has been used by man since antiquity, and its detrimental health effects have been well known for centuries. It is probably the best studied neurotoxic compound, and comprehensive reviews cov- ering chemical, environmental, and biological aspects in great de- tail e.g., the recent Environmental Protection Agency (EPA, 1986) report should be consulted for more in-depth information. Chemi- cally speaking, this metal occurs in inorganic form, namely in the form of lead salts of widely different water solubility, as well as in organic form. Although the organometallic compounds have been found to be highly neurotoxic in acute occupational exposure (Grandjean, 1984), chronic low-level exposure to inorganic lead constitutes a more important Dublic health issue (Lansdown and Yule. 19861. , , v' 1 1 ~ ' ~ Inorganic lead enters the body by way of inhalation and ingestion; absorption is better in infants than in adults. Blood lead concentra- tion (PbB) is an representative marker of current lead exposure, whereas tooth lead concentrations have been used as markers of past expo- sure (Needleman et al., 1979~. Both placental transfer and blood- brain transfer of lead occur, so that prenatal exposure and CNS in- volvement are possible. Lead is considered a nonessential metal. Its toxicity may be explained largely by interference with different en- zyme systems: lead inactivates these enzymes by binding to sulfydryl (SH) groups or by competitive interaction with other essential metal ions. Therefore, almost all organs or organ systems can be considered potential targets for lead: depending on duration and degree of exposure, a wide range of biological effects has been documented, the more
230 GERHARD WINNEKE critical of which are those on heme biosynthesis, erythropoiesis, and the nervous system. For a number of reasons related both to exposure and to CNS vulnerability, children between the ages of 9 months and 6 years are particularly at risk according to the Centers for Disease Control (CDC, 1985~. Both the peripheral and the central nervous systems may be in- volved in lead neurotoxicity, although PNS effects seem to be more prominent in occupational lead exposure of adults, whereas CNS in- volvement is more characteristic of childhood lead exposure. Acute symptomatic lead poisoning has been linked to the swallowing of lead-based paint and is often associated with encephalopathy at PbB exceeding 100 ,ug/dL (Chisolm, 1971~. Lead encephalopathy is clini- cally characterized by some or all of the following symptoms (CDC, 1985~: coma, seizures, ataxia, apathy, incoordination, vomiting, clouded consciousness, and loss of previously acquired skills. Children surviving lead encephalopathy typically present with neurological and psycho- logical sequelae, including focal EEG abnormality, cramps, intelligence deficit, hyperactivity, distractibility, and reduced impulse control (Byers and Lord, 1943; Perlstein and Attala, 1966; Smith et al., 1963~. Such clinical findings have led to the hypothesis that long-term low-level childhood exposure to lead might be associated with subclinical neurobehavioral deficit in asymptomatic children as well, which due to its subtlety may often go undetected. Since the early 1970s this hypothesis has been tested in about 30 cross-sectional studies using different psychological tests as well as behavior ratings to assess the degree of CNS involvement, and PbB or tooth lead levels as markers of current or past exposure. The variety of psychological functions and tests covered in these studies in different combinations may be given roughly as follows: psycho- metric intelligence in most studies was assessed by means of the Wechsler Scales (WISC, WPPSI), although other tests were used in some studies (e.g., the McCarthy Scales of children's ability, the Stanford- Binet, or the British Ability Scales). The following have been added to IQ measures in several of these studies: perceptual motor integra- tion using the Bender Gestalt Test, the Benton Test, or the Frostig Scales; gross or fine motor coordination using the Purdue Pegboard; finger-wrist tapping and the Osertesky Motor Scales; reaction and attentional performance using delayed and serial choice reaction times; vigilance performance; behavior ratings by means of the Connor, Wherry- Weiss-Peters, and Rutter Scales, and measures of educational attain- ment. Inconsistency of outcome, differences in study design, and con- founder structure interfere with any simple, straightforward conclu- sion as to whether neurobehavioral deficit in asymptomatic children
TOXICITY OF SELECTED ENVIRONMENTAL CHEMICALS 231 is truly associated with, or even caused by, low-level childhood lead exposure. A comprehensive review (EPA, 1986, p. 145) arrived at the following careful conclusions: "As for CNS-effects, none of the avail- able studies on the subject, individually, can be said to prove conclu- sively that significant cognitive (IQ) or behavioral effects occur in children at blood lead levels <30 ,ug/dL." However, the most recent neurobehavioral studies of CNS cognitive (IQ) effects collectively demonstrate associations between neuropsychological deficits and low- level lead exposures in young children resulting in blood lead levels ranging to below 30 ,ug/dL. Some more recent cross-sectional stud- ies (Fulton et al., 1987; Hatzakis et al., 1987; Norby-Hansen et al., 1989), not covered in the above mentioned review, generally agree with this conclusion, in that small but significant cognitive and attentional deficit was observed at low blood or tooth lead levels. In addition to these cross-sectional approaches, typically lacking precise exposure histories, first results from several prospective stud- ies are now beginning to be published, four of which have recently been reviewed (Davis and Svendsgaard, 1987~. In all of these studies, repeated blood lead sampling starting at birth (cord blood) was done to describe the early exposure history, and repeated outcome assess- ment was done at regular intervals using the Bayley Scales of Infant Development as the instrument. In their review, Davis and Svendsgaard (1987, p. 299) conclude: "There can now be little doubt that exposure to lead, even at blood-levels as low as 10-15 g/dL, and possibly lower, is linked with undesirable developmental outcomes in human fetuses and children. These effects include impaired neurobehavioral development, reduced gestational age, lowered birth weight, and other possible effects on early development and growth." In view of some divergent findings between studies, in the present author's opinion, this must be qualified as a bold statement, although it is certainly true that some of these results do raise concern about persistent neurobehavioral effects of low-level lead exposure at early stages of brain maturation. Such concern is supported by animal studies showing long-lasting neurobehavioral deficit in different spe- cies after perinatal lead exposure associated with blood levels below 30 ,ug/dL (Winneke, 1986~. It should be added, however, that no convincing mechanism has yet been proposed to account for such deficit. Mercury Mercury also belongs to those metals known to, and used by, man since ancient times; for centuries it has been employed primarily for therapeutic purposes. For more detailed information on chemical,
232 GERHARD WINNEKE environmental, and biomedical aspects, comprehensive reviews should be consulted (Berlin, 1986; Clarkson et al., 1984~. There are different physical and chemical forms of mercury namely, metallic, mercurous, and mercuric mercury. Vapors of metallic mer- cury are primarily an occupational problem and are not considered here; instead, the emphasis is on those stable organometallic compounds known as methylmercury (MeHg). It is through these compounds that increased environmental exposure may occur in segments of the general population because inorganic mercury released into the envi- ronment from a variety of sources is methylated by microorganisms present in bodies of fresh and ocean waters and thus enters the aquatic food chain. The highest MeHg concentrations have been measured in large predatory fish, such as shark or tuna. Consequently, popula- tions with high fish consumption must be considered at risk from MeHg exposure. Absorption of MeHg in the gastrointestinal tract is almost com- plete, and its distribution in the body is rather uniform. There is no placental barrier for MeHg, which is found in all fetal tissue; MeHg concentrations in fetal blood are typically higher than those of the mother. Methylmercury also enters the hair as soon as it is formed; thus, MeHg concentration in hair is an excellent noninvasive marker of exposure. Knowledge about the neurological and psychological sequelae of high MeHg exposure was gained primarily in two catastrophic inci- dents of mass poisoning, namely, the Minamata Bay tragedy in Japan in 1950 (Harada, 1966) and an outbreak in Iraq 20 years later (Bakir et al., 1973~. Population exposure in the Minamata incident was through contaminated fish from Minamata Bay, which had been polluted for years by metallic mercury from industrial sources; this was then methylated by marine microorganisms and thus introduced into the food chain. Thousands of people were exposed, and hundreds of cases of MeHg poisoning have been documented. Methylmercury ex- posure in Iraq was through ingestion of seed grain treated with an MeHg fungicide; the grain had been ground into flour to make bread. About 7,000 people were hospitalized with signs and symptoms of poisoning, and more than 400 of them died. The clinical picture of MeHg poisoning is characterized by sen- sory, motor, and cognitive deficit. The earliest effects are nonspecific symptoms (e.g., complaints of paresthesia, general malaise, and blurred vision). Later on, signs of neurotoxicity appear, such as constricted visual field ("tunnel vision"), deafness, dysarthria, and ataxia. Men- tal disturbances and alterations of the chemical senses may occur as well. Clarkson et al. (1984) mention three important features of MeHg
TOXICITY OF SELECTED ENVIRONMENTAL CHEMICALS 233 effects, namely, their irreversibility, their selective neurotoxic charac- ter with predominant CNS involvement, and the long latency between cessation of exposure and onset of symptoms, which may extend from a few weeks to several months or even years. Whereas irreversibility of CNS effects is most likely due to loss of neurons, the reason for the long latency period is not known. Clarkson et al. (1984, p. 302) speculate that latency periods of several years "may be partially explained by psychogenic overlay which modifies the symptoms or subclinical lesions which may be revealed by the aging factor." Another important feature of MeHg neurotoxicity is the particular vulnerability of the developing CNS, which has been observed in human cases as well as in animal models. In both the Minamata and the Iraq outbreaks, pregnant women with only minor symptoms of MeHg poisoning occasionally gave birth to children with severe CNS damage. The clinical picture was dose-dependent. At high maternal MeHg blood levels, microcephaly, hyperreflexia, and severe motor and mental impairment were prominent. For lower degrees of expo- sure, subtle deficits were difficult to diagnose shortly after birth but became increasingly pronounced later on. Psychomotor impairment and persistent abnormal reflexes were found at hair levels exceeding 50 mg/kg. The mildest cases presented with signs of the minimal brain dysfunction syndrome, characterized by hyperactivity and attention deficit (Amin-Zaki et al., 1974~. The likelihood of mental retardation increased with increasing maternal MeHg hair levels. In the Minamata case, follow-up studies revealed strong associations between cord blood MeHg levels and mental retardation in 20-year-old victims of prena- tal exposure (Harada et al., 1977~. Thus far, only two studies deal with subclinical signs of MeHg exposure at environmentally elevated levels. In one such study (McGill Group, 1980) in Cree Indians exposed to MeHg from fish, some asso- ciations were found between tone and reflexes of Cree boys and MeHg hair levels of their mothers during pregnancy. Such effects occurred at much lower MeHg levels than those previously found to be associ- ated with neurotoxicity. Because these effects were mild and somewhat isolated, however, doubts as to their substantive nature have been raised (Clarkson et al., 1984~. A more recent, ongoing study in New Zealand (Kjellstrom et al., 1986) used a pair-matching approach to study developmental retardation due to low-level in utero exposure to MeHg from fish at hair levels exceeding 6 mg/kg. Developmental status was assessed by means of the Denver Developmental Screen- ing Test. From a basic cohort of 11,000 mother-child pairs, 31 with elevated MeHg levels between 6 and 20 mg/kg were compared with
234 GERHARD WINNEKE pair-matched controls. Significant dose-related developmental delay was found at age 4. Results from subsequent psychometric testing at age 6 have not yet been published. PoluchZorinated BiphenyZs Polychlorinated biphenyls typically are mixtures of several com- pounds differing in terms of number and position of chlorine substituents. In most industrialized countries, PCBs are no longer used in "open" systems but continue to be used in "closed" systems such as hydrau- lic pumps, transformers, or heat exchangers. The PCB compounds are biologically persistent and, therefore, ac- cumulate in the food chain. Marine mammals are a particular target: average PCB concentrations of 160 mg/kg of fat have been measured in marine mammals in the North Sea. In human fat tissue, average PCB concentrations are between 1 and 2 mg/kg of fat. There is an age-related increase of PCBs in fat tissue. Toxicological effects of PCB exposure in man were first observed in the context of mass poisoning in Japan in 1968 (Kuratsune, 1972~. The first and most obvious signs were skin affections resembling chloracne in about 1,000 persons. The cause of poisoning was found to be PCB-contaminated rice oil. Besides skin affections, the clinical picture of the disease, which soon became known as yusho (oil disease), was characterized among others by pigmentation of fingernails, alopecia, porphyria, and decreased concentrations of immunoglobulin M (IgM). An increased number of stillbirths and of small-for-age babies was observed. Apart from such outbreaks of acute poisoning, low-level chronic PCB exposure occurs through foodstuffs. Human breast milk con- tains elevated PCB concentrations: average values of 1-2.5 mg of PCB per kilogram of fat were measured in West Germany (DFG, 1984~. The PCBs are known to cross the placenta. Neonatal effects of transplacental PCB exposure have been studied recently by using psychological and neurological criteria (Jacobson et al., 1985; Rogan et al., 1986~. In one of these studies (Rogan et al., 1986) the Brazelton Neonatal Assessment Scale was used to assess reflexive and motor behavior and to track the state of about 900 neonates regularly from the first three weeks after birth to 24 months of age. The PCB level was measured in cord and maternal serum at term. Multiple regres- sion analysis revealed significant PCB associations only for hypoto- nicity and hyporeflexia; birth weight and head circumference were not related to PCB serum levels. The second study examined the effects of transplacental and neo-
TOXICITY OF SELECTED ENVIRONMENTAL CHEMICALS 235 natal PCB from fish in 242 children born to women who ate fish from Lake Michigan and in children form non-fish-eating mothers (rein et al., 1984; Jacobson et al., 1985~. The PCB level in cord serum predicted lower birth weight and smaller head circumference (rein et al., 1984~. In addition, cognitive performance in the Visual Recognition Memory Test (Fagan and McGrath, 1981) at 7 months of age exhibited signifi- cant association with cord serum PCB levels but not with breast milk PCB levels after control for confounding (Jacobson et al., 1985~. Thus, there is some evidence for the ability of prenatal PCB exposure to affect cognitive and neuromuscular development in the neonate, al- though unfortunately, possible simultaneous MeHg exposure was not taken into account. EFFECTS IN THE ADULT NERVOUS SYSTEM The effects of aluminum and solvents are considered here, the neurotoxicity of which has been studied only in adults so far. Aluminum Aluminum is abundant in the earth's crust. Its toxicity for humans has been rated low in the past because it was considered to be almost nonabsorbable from the gastrointestinal tract. It is now clear, how- ever, that both inhaled and ingested aluminum is absorbable to some extent. For a detailed review of chemical, environmental, toxicological, and biomedical aspects, the reader is referred to Elinder and Sjogren (1986~. Aluminum exists in organic and inorganic form. The inor- ganic aluminum salts have different water solubilities. The neurotoxicity of aluminum was first detected in animal stud- ies: Epileptic cramps and neurofibrillary degeneration were observed after direct brain injection or parenteral application of different alu- minum salts (De Boni et al., 1976; Sorensen, 1974~. Information about aluminum-induced neurotoxicity in humans was gained when dialy- sis patients developed a progressive cementing illness, which often proved fatal if untreated by chelation. Clinical signs of this type of brain damage include speech and motor disturbances, memory defi- cit, personality changes, dementia, and seizure disorders. Although there was some debate as to the etiological contribution of aluminum to this disease, which has become known as dialysis dementia, it is now accepted that the use of aluminum-containing phosphate-binding gels or of water with high aluminum content was the cause of this cementing illness. Reduced renal function may also lead to a significant accumulation of aluminum in the body, associated with dialysis dementia
236 GERHARD WINNEKE in patients who had never undergone dialysis treatment, but who had taken large doses of aluminum hydroxide. A possible role of aluminum in the pathogenesis of Alzheimer's disease is being discussed (Crapper and De Boni, 1980~. This hypoth- esis rests on partial similarity of dialysis dementia to presenile and senile features of Alzheimer's disease, in both clinical and pathological terms. It has been shown, for example, that in Alzheimer's patients, aluminum selectively accumulates in the nucleus of the brain cells that form the neurofibrillary tangles, typical of the Alzheimer condi- tion; neurofibillary tangles have also been observed subsequent to injection of aluminum salts in cats and rabbits. In some studies, elevated aluminum levels in the gray matter of Alzheimer's patients with normal kidney function were found as well. There are, however, several contradictory findings, so that the evidence supporting an association between environmental aluminum exposure and Alzheimer's disease must still be considered circumstantial (Elinder and Sjogren, 1986~. Organic Solvents Organic solvents represent a large, chemically heterogeneous group of chemicals which are liquids between 0 and 250°C. Traditionally they are used for the extraction, solution, or suspension of water- insoluble materials namely, fats, lipids, resins, and polymers. Sol- vents may be grouped into aliphatic hydrocarbons (e.g., hexane), aro- matic hydrocarbons (e.g., toluene), halogenated hydrocarbons (e.g., trichloroethylene), alcohols, ketones (e.g., methyl ethyl ketone), esters (e.g., butyl acetate), and mixtures (e.g., white spirit). _ Due to the wide variety of applications, occupational exposure and, to a lesser degree, general population exposure are frequent. The nervous system is the primary target for inhaled solvents be- cause of their lipophilic characteristics. Whereas narcotic action is the predominant biological effect in the CNS, functional and struc- tural effects ranging from neurophysiological changes to severe poly- neuropathies have been reported to occur in the PNS. Comprehen- sive reviews covering relevant aspects of chemistry, exposure, and biomedical effects are available [Riihimaki and Ulfvarson, 1986; World Health Organization (WHO), 1985~. The PNS neurotoxicity of hexacarbons, carbon disulfide, and acrylamide is well established both clinically and experimentally. The range of effects covers subclinical neurophysiological alterations and full-blown polyneuropathies with slow recovery, depending upon the degree and duration of exposure, as well as potentiation by "innocent by- stander" chemicals (Schaumberg and Spencer, 1987~. Much of the CNS neurotoxicity of organic solvents is explainable
TOXICITY OF SELECTED ENVIRONMENTAL CHEMICALS 237 in terms of narcotic action. Short-term human exposure under ex- perimental conditions has resulted in prenarcotic reversible effects such as psychomotor slowing or vigilance decrement at low levels of exposure (Dick and Johnson, 1986~. As yet it is not clear whether repeated prenarcotic exposure over years may eventually give rise to irreversible brain damage. It has, however, been shown that for some compounds such as trichloroethylene, styrene, and carbon disulfide, as well as solvent mixtures, chronic low-level exposure is associated with perceptual and motor retardation which, from the very design of the different studies, could not be explained as an acute reversible effect. In summarizing several such studies the conclusion was drawn (Gamberale, 1986, p. 217) "...that the measurement of behavioral per- formance has been demonstrated to possess more general applicabil- ity in human studies than other methods." Case control studies from Scandinavian countries, using records from disability pensions, generally support psychological findings from cross-sectional studies in workers chronically exposed to or- ganic solvents, in that a higher prevalence of neuropsychiatric disor- ders or toxic encephalopathies was found after long-term occupational exposure to solvents or solvent mixtures (Hogstedt and Axelson, 1986~. Syndromes in such cases have become known as psycho-organic syndromes in Finland and Sweden or, primarily in Denmark and Norway, as presenile dementia. Despite a number of methodological drawbacks the conclusion has been drawn (Hernberg, 1984) that sol- vent-induced toxic encephalopathies do exist, although they may present with considerable problems of differential diagnosis in individual cases. CONCLUSION The preceding examples have been selected to illustrate how psy- chological tests have been used to detect subtle psychological deficit resulting from neurotoxic insult of environmental/occupational chemicals. In some instances, functional deficit has been reported to occur at exposure levels that generally do not induce alterations of neurologi- cal or neurophysiological functions. Despite this apparent sensitivity, such findings have been met with considerable skepticism, in the scientific world as well as among administrators. This appears to be due mainly to the following shortcomings: (1) lack of consistency, (2) lack of theory, and (3) lack of significance for health or adjustment. Lack of Consistency It is true that neurobehavioral findings reported from one labora- tory often do not prove replicable in other laboratories. This is very
238 GERHARD WINNEKE obvious in neuropsychological lead research and interferes with efforts to establish dose-response modeling. Apart from analytical differences and differences in study design or confounder structure, part of this inconsistency is due to lack of standardized testing. This refers both to instrumental and to procedural aspects of psychological testing. Two examples are given for the instrumental aspect to illustrate this point: (1) The Bender Gestalt Test, which has been used in several studies, is a well-established clinical tool for the early detection of brain damage, with well-standardized stimulus material and stan- dardized instructions, but different scoring systems (e.g., Koppitz, GET system). These different systems have in fact been shown to produce different exposure-related effects for the same lead levels (Trillingsgaard et al., 1985~. (2) Reaction time (RT) is another case in point. The paradigm is apparently simple but actually covers a wide range of cognitive demands (e.g., simple RT, choice RT, delayed RT, serial choice RT). The Shakov-derived delayed RT paradigm of Needleman has produced different lead-related outcomes depending on similarity to the original procedure (Hunter et al., 1985; Winneke et al., 1985~. Conclusion There is a need for more rigorous standardization of test procedures, and computerized testing could be an important step forward. Whereas for adults in occupational exposure settings such developments, based on the WHO core battery, have proved promis- ing, very little has been done at the lower end of the age continuum. In addition to standardization efforts, strategies of quality assurance for psychological data must be developed to be able to compare out- comes from different studies. In this context the advantages of para- metric variation (Weiss, 1978), e.g., increase of task difficulty, should also be studied and exploited in a more systematic manner to increase the sensitivity and the comparability of psychological outcome mea- sures and, possibly, to clarify their validity. One such example is the interaction of task difficulty and lead-induced neurobehavioral deficit in serial choice reaction performance (Winneke et al., 1989~: It was shown that lead-induced deficit occurred for high but not for low signal rates. The type of deficit resembles clinical observations in children presenting with attention deficit disorders. Lack of Developmental Perspective It is true that the selection of psychological tests in many neuro- toxicity studies has been guided primarily by availability and convenience, rather than by considerations based on experience from developmen-
TOXICITY OF SELECTED ENVIRONMENTAL CHEMICALS 239 tat psychology. Just two examples illustrate the point: Psychometric intelligence has been a preferred outcome measure in many studies. Even in prospective studies of early developmental lead exposure, the Bayley Scales have been used, which are known to possess poor predictive validity. Instead, the work of Pagan and coworkers on visual recognition-memory (Fagan and McGrath, 1981) offers more promising features, in terms of across-age continuity, because its pre- dictive validity for later cognitive development has been shown to be higher than that of the Bayley Scales. Another important aspect of developmental continuity is extrapolation of neurobehavioral effects across species. Cognitive performance is a case in point. Much of the abundant animal literature on agent-induced learning and memory deficits is difficult to extrapolate to the human level, because the preferred cognitive models developed and used in clinical or basic contexts are almost incompatible with the typical models used in behavioral pharmacology and toxicology. Conclusion There is a need for the development and use of more specific measures with known functional significance and greater va- lidity across different age groups. In addition, closer collaboration between those engaged in animal and human research in the field of neurobehavioral toxicology is necessary to develop cognitive para- digms that allow for a more direct extrapolation across species. Uncertain Health Significance Subclinical effects are necessarily difficult to evaluate in terms of their implication for health and adjustment. One such example is the ongoing discussion about the relevance of an average exposure-related IQ drop in later academic achievement. For psychology to contribute more successfully to the solution of environmental health problems it is necessary to systematically exploit biological or disease states as frames of reference for the interpretation of results from psychologi- cal toxicity studies. Examples for studies in infants and children are small-for-age or low-birthweight babies, perinatal hypoxia, infectious or traumatic brain injuries, minimal brain dysfunction, dyslexia, and epilepsy. Examples for studies in adults are the effects of normal aging and of age-related neurological disorders (e.g., Alzheimer's or Parkinson's diseases and other cementing illnesses), particularly in their early degenerative stages. Conclusion Systematic validation of functional tests that use bio- logical processes or neurological disease entities as frames of reference
240 GERHARD WINNEKE is needed for the interpretation of Neurobehavioral findings in terms of health adjustment. REFERENCES Amin-Zaki, L. S. Elhassani, M. A. Mjeed, T. W. Clarkson, R. A. Doherty, and M. R. Greenwood. 1974. Intrauterine methylmercury poisoning in Iraq. Pediatrics 54:587- 595. Anger, W. K. 1985. Neurobehavioral tests used in NIOSH-supported worksite stud- ies, 1973-1983. Neurobehavioral Toxicology 7:357-368. Anger, W. K. 1986. Workplace exposures. In Neurobehavioral Toxicology, A. Annau, ed. Baltimore: The Johns Hopkins Press. Anger, W. K., and B. L. Johnson. 1985. Chemicals affecting behavior. Pp. 51-148 in Neurotoxicity of Industrial and Commercial Chemicals, J. L. O'Donoghue, ed. Boca Raton, Fla.: CRC Press. Bakir, F., F. Damlougi, L. Amin-Zaki, M. Murtadha, A. Khadlidi, N. Y. Al-Rawi, S. Tikriti, H. I. Dhahir, T. W. Clarkson, J. Smith, and R. A. Doherty. 1973. Methyl- mercury poisoning in Iraq. Science 181:230-241. Berlin, M. 1986. Mercury. Chap. 16 in Handbook on the Toxicology of Metals, second edition, L. Fridberg, G. F. Nordberg, and V. Vouk, eds. Amsterdam: Elsevier. Byers, R. K., and E. E. Lord. 1943. Late effects of lead poisoning on mental develop- ment. American Journal of Diseases of Children 66:471-494. Caine, D. B., A. Eisen, E. McGeer, and P. Spencer. 1986. Alzheimer's disease, Parkinson's disease, and motoneuron disease: Abiotropic interaction between ageing and envi- ronment? Lancet 2:1067-1070. Centers for Disease Control. 1985. Preventing Lead Poisoning in Young Children. Document 99-2230. U.S. Department of Health and Human Services, Atlanta, Geor- g~a. Chisolm, Jr., J. J. 1971. Lead poisoning. Scientific American 224:15-23. Clarkson, T. W., R. Hamada, and L. Amin-Zaki. 1983. Mercury. Pp. 32~389 in Changing Metal Cycles and Human Health, J. O. Nriagu, ed. Berlin: Springer. Crapper, D. R., and U. De Boni. 1980. Aluminium. Pp. 326-335 in Experimental and Clinical Neurotoxicity, P. S. Spencer and H. H. Schaumburg, eds. Baltimore: Wil- liams and Wilkins. Davis, J. M., and D. J. Svendsgaard. 1987. 329:297-300. Lead and child development. Nature De Boni, U., A. Otvos, J. W. Scott, and D. R. Crapper. 1976. Neurofibrillary degenera- tion induced by systemic aluminium. Acta Neuropathologica 35:285-288. DFG DeutscheForschungsgemeinschaft. 1984. Rueckstaende und Verunreinigungen in Frauenmilch. Mitteilung VII der Kommission zur Pruefung von Rueckstaenden in Lebensmitteln. Weinheim: Verlag Chemie. Dick, R. B., and B. L. Johnson. 1986. Human experimental studies. Pp. 348-387 in Neurobehavioral Toxicology, Z. Annau, ed. Baltimore: Johns Hopkins University Press. Elinder, C. G., and B. Sjogren. 1986. Aluminium. Chap. 1 in Handbook on the Toxicology of Metals, second edition, L. Friberg, G. F. Nordberg, and V. Vouk, eds. Amsterdam: Elsevier. Fagan, J. F., and S. K. McGrath. 1981. Infant recognition memory and later intelli- gence. Intelligence 5:121-130. Fein, G. G., J. L. Jacobson, S. W. Jacobson, P. M. Schwartz, and J. K. Dowler. 1984. Prenatal exposure to polychlorinated biphenyls: Effect on birth size and gestational age. Journal of Pediatrics 105:315-320.
TOXICITY OF SELECTED ENVIRONMENTAL CHEMICALS 241 Fulton, M., G. M. Raab, G. Thomson, D. Laxen, R. Hunter, and W. Hepburn. 1987. Influence of blood lead on the ability and the attainment of children in Edinburgh. Lancet 1:1221-1226. Gamberale, F. 1986. Application of psychometric techniques in the assessment of solvent toxicity. Pp. 203-224 in Safety and Health Aspects of Organic Solvents, V. Riihimaki and U. Ulfvarson, eds. New York: Liss. Grandjean, P., ed. 1984. Biological Effects of Organolead Compounds. Boca Raton, Fla.: CRC Press. Harada, Y. 1966. Study group on Minamata disease. In Minamata Disease, M. Matsuma, ed. Humamato: Humamato University Press. Harada, M., T. Fujino, and K. Kabashima. 1977. A study of methylmercury concentra- tion in the umbilical cords of the inhabitants born in the Minamata area. Brain and Development 9:79-84. Hatzakis, A., A. Kokkevi, C. Maravelias, K. Katsouyanni, F. Salaminos, A. Kalandidi, A. Koutselinis, C. Stefanis, and D. Trichopoulos. 1987. Psychometric intelligence deficits in lead-exposed children. In Lead Exposure and Child Development, M. A. Smith, L. D. Grant, A. I. Sors, eds. Boston: Kluwer Academic Publishers. Hernberg, S. 1984. Die subklinische Wirkung von Losungsmitteln and Loseungs- mittelgemischen auf das Nervensystem. Pp. 47-56 in Verhandlungen der Deutschen Gesellschaft fur Arbeitsmedizin, H. Konietzko and F. Schuckmann, eds. Stuttgart: Gentner Verlag. Hogstedt, L., and O. Axelson. 1986. Longterm health effects of industrial solvents A critical review of the epidemiological research. Medicina del Lavorno 77:11-22. Hunter, J., M. A. Urbanowicz, W. Yule, and R. Lansdown. 1985. Automated testing for reaction time and its association with lead in children. International Archives of Occupational and Environmental Health 57:27-34. Jacobson, S. W., G. G. Fein, J. L. Jacobson, P. M. Schwartz, and J. K. Dowler. 1985. The effect of intrauterine PCB exposure on visual recognition memory. Child Develop- ment 56:853-860. Kjellstrom, T., P. Kenenedy, S. Wallis, and C. Mantell. 1986. Physical and mental devel- opment of children with prenatal exposure to mercury from fish. Stage 1: Prelimi- nary tests at age 4. National Swedish Environmental Protection Board, Report 3080. Kuratsune, M. 1972. Epidemiologic study on yusho, a poisoning caused by ingestion of rice oil contaminated with a commerical brand of polychlorinated biphenyls. Environmental Health and Perspectives 1:119-128. Landsdown, R., and W. Yule, eds. 1986. The Lead Debate: The Environment, Toxicol- ogy and Child Health. London: Croom Helm. Needleman, H. L., C. Gunnoe, A. Leviton, R. Reed, H. Peresie, C. Maher, and P. Barrett. 1979. Deficits in psychologic and classroom performance of children with elevated dentine lead levels. New England Journal of Medicine 300:689-695. Norby-Hansen, O., A. Trillingsgaard, I. Beese, T. Lyngbye, and P. Grandjean. 1989. Neuropsychological profile of children in relation to dentine lead level and socio- economic group. In Lead Exposure and Child Development, M. A. Smith, L. D. Grant, A. I. Sors, eds. Boston: Kluwer Academic Publishers. Perlstein, M. A., and R. Attala. 1966. Neurologic sequelae of plumbism in children. Clinical Pediatrics 5:292. Riihimaki, V., and U. Ulfvarson, eds. 1986. Safety and Health Aspects of Organic Solvents. New York: Liss. Rogan, W. J., B. C. Gladen, J. D. McKinney, N. Carreras, P. Hardy, J. Thullen, J. Tingelstad, and M. Tully. 1986. Neonatal effects of transplacental exposure to PCBs and DDE. Journal of Pediatrics 109:335-341. Schaumburg, H. H., and P. S. Spencer. 1987. Recognizing neurotoxic disease. Neurol- ogy 37:276-278.
242 GERHARD WINNEKE Smith, H. D., R. L. Baehner, T. Carney, and W. J. Majors. 1963. The sequelae of pica with and without lead poisoning. A comparison of the sequelae of five or more years later. I. Clinical and laboratory investigations. American Journal of Diseases of Children 105:609-616. Spencer, P. S., P. B. Nunn, J. Hugon, A. Ludolph, and D. N. Roy. 1986. Motoneuron disease on Guam: Possible role of a food neurotoxin. Lancet 1:965. Spencer, P. S., J. Hugon, A. Ludoph, P. B. Nunn, D.N. Ross, and H. H. Schaumburg. 1987. Discovery and partial characterization of primate motor-system toxins. Pp. 221-231 in Selective Neuronal Death, G. Bock and M. O'Connor, eds. Ciba Founda- tion Symposium 126. Trillingsgaard, A., O. Norby-Hansen, and I. Beese. 1985. The Bender Gestalt Test as a neurobehavioral measure of preclinical visual-motor integration deficits in children with low-level lead exposure. Pp. 189-198 in Neurobehavioral Methods in Occupa- tional and Environmental Health. Copenhagen: World Health Organization. U.S. Environmental Protection Agency. 1986. Air Quality Criteria for Lead, Vol. I-IV. EPA Report No. 60/8-83/028. Springfield, Va.: National Technical Information System. Weiss, B. 1978. The behavioral toxicology of metals. Federation Proceedings 37:22- 27. Winneke, G. 1986. Animal Studies. Pp. 217-234 in The Environment, Toxicology and Child Health, R. Lansdown and W. Yule, eds. London: Croom Helm. Winneke, G., U. Beginn, T. Ewert, C. Havestadt, U. Kramer, C. Krause, H. L. Thron, and H. M. Wagner. 1985. Comparing the effects of perinatal and later childhood lead-exposure on neuropsychological outcome. Environmental Research 38:155- 167. Winneke, G., A. Brockhaus, W. Collet, and U. Kramer. 1989. Modulation of lead- induced performance deficit in children by varying signal rate in a serial choice reaction-task. Neurotoxicology and Teratology 11(6). World Health Organization. 1985. Chronic Effects of Organic Solvents on the Central Nervous System and Diagnostic Criteria. Copenhagen.
