SECTION 8
DISCUSSION
A key purpose of the TCE Subregistry is to determine if there is an excess of adverse health conditions for registrants when compared with a national sample. To date, this objective has been pursued by comparing TCE Subregistry data about health conditions with National Health Interview Survey (NHIS) data; comparing TCE cancer outcomes with cancer incidence data from the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) program; and comparing TCE Subregistry mortality rates with National Center for Health Statistics (NCHS) published rates. Health, demographic, occupational, and environmental information was collected on 4,280 TCE-exposed persons (4,041 living, 239 deceased) who met the eligibility criterion for National Exposure Registry participation: resided for more than 30 consecutive days during the period of exposure at a site address. The source of contamination for the wells at site addresseslocated in 13 sites in Indiana, Illinois, and Michiganvaried from site to site, as did the levels and numbers of contaminants besides TCE.
The response rate of eligible persons who were located was 98% or higher at each site (one site had 100% participation). Such a high participation rate minimized bias in the data that might have been associated with nonresponse.
TCE Subregistry and NHIS data were both self-reported; however, the responses of the TCE Subregistry members might, in part, have been influenced by the participants' knowledge of and concern about TCE exposure, which must be taken into consideration. The subregistry and NHIS questions about health conditions shared important similarities, but differed in two ways: (1) restrictions on the source of diagnosis, and (2) the wording of some of the questions about health conditions.
Concerning the source of diagnosis, the TCE Subregistry questions specified that the source of diagnosis must be a "physician or other medical provider." This restriction was added to each subregistry question about health conditions in an effort to minimize overreporting by registrants whose health awareness might have been heightened by activities occurring at their sites (for example, ongoing litigation). The inclusion of this qualifier had the potential to reduce the rates of reporting for the subregistry when compared with the NHIS rates, all other factors being equal; however, reporting of medical diagnoses was not confirmed. In keeping with this thinking, the reporting was significantly increased for the NHIS population for response rates for some health conditions (that is, arthritis, asthma, and respiratory allergies). Of particular interest is that these conditions are more commonly self-diagnosed and would predictably be reported at a higher rate by the NHIS participants. In other words, the conditions with increased reporting in NHIS are those conditions that are most limited or affected by the TCE Subregistry restriction of health care provider confirmation of the condition. The results were consistent with that prediction. The increased reporting for asthma was not age or sex dependent; the increased reporting by NHIS respondents for arthritis and respiratory allergy problems also was not age dependent, but differed between males and females. There was an exception to the predicted increased reporting rates by NHIS participants for commonly self-diagnosed problems--the reporting of skin problems; in this case, there was an excess reporting by TCE Subregistry members. The excess reporting of skin conditions occurred primarily in the lower age groups; reporting became equal or reversed in some of the higher age groups.
The comparability of the wording of TCE Subregistry and NHIS health conditions was addressed in Section 5. Nine of the health condition questions matched exactly or very closely and eight others were considered similar. The implications of health condition wording comparability are addressed further in the following discussion.
An additional problem, one of file comparability (NHIS is a composite of time frames and other retrictions that could not be duplicated using NER data), precluded making a comparison of reporting rates for heart disease. ATSDR is exploring alternative national norms for reporting rates for heart disease to use for this comparison. When this is completed, the results of the comparison will be published.
The mortality analyses covered the period from 1970 through 1990. Before 1985, the mortality rates of the TCE Subregistry data were below those reported in the national data. After 1985, the rates were reversed; the TCE Subregistry rates of mortality were higher than the national rates.
Site-by-site comparisons in terms of health outcomes (for example, Southeast Rockford versus Gemeinhardt) were not provided; nevertheless, it is important to compare the similarities of registrants from site to site in terms of demographic, occupational, and environmental data. This comparison serves as an indicator or check for any aberrant site that might have skewed the data. As was discussed in Section 7, the sites were comparable in terms of sex ratios and occupational exposure. There were slight differences in the average age and education attainment; however, these differences were small and would not be expected to have any marked influence on the rate of reporting of health outcomes.
ENVIRONMENTAL DATA
As is discussed at length in Section 3, the environmental data available were collected and evaluated, but had limitations that merit reiteration. The samples were not taken for the purpose of quantifying exposures over time. Rather, they were taken for the purpose of verifying contamination. Regardless, the data serve the Registry's purpose in that they establish that exposure to TCE in water used for household purposes occurred at site addresses (that is, individual residences). Most of the environmental data represent one-time samples of well water at site addresses. These samples might or might not have been taken when TCE and other chemical exposures were at a maximum level.
In Section 6, the analyses focused on whether there was an association between the levels of exposureboth for TCE and other chemicalsand the rate of reported health outcomes. The results of these analyses, after controlling for the effects of sex, age, cigarette smoking history, and occupation are shown in Table 8-1. Although there was not a consistent increase of health symptom reporting with increased levels of exposure and length of exposure, using the specified groups, some associations were found. Given the limitations of the environmental data, associations between health conditions and exposure levels must be interpreted with caution. Future studies by the Agency for Toxic Substances and Disease Registry (ATSDR) and others should pursue methods for the investigation of dose-response relationships and the selection of grouping variables for modeling time-dependent covariates such as cumulative exposures (31).
Table 8-1.--Summary of environmental groups analyses.
| Outcome | Environmental Group (Variables in Model) | |||||||||
| Maximum TCE | Cumulative TCE | Cumulative Chemical | Length of Exposure | Max TCE Quartiles | Cum Chem Quartiles | |||||
| (Age/Sex) | (All) | (Age/Sex) | (All) | (Age/Sex) | (All) | (Age/Sex) | (All) | |||
| Hearing
impairment |
X | |||||||||
| Respiratory
allergies |
X | |||||||||
| Stroke | X | |||||||||
X--denotes statistical significance.
HEALTH OUTCOMES
The comparisons of TCE Subregistry and NHIS data on reported health conditions revealed several statistically significant differences. Section 5 provides the results, setting the
Type I error (the level) at .01 and the large sample size served to control the likelihood of
false negative or false positive results. These findings and the results of the mortality analyses are summarized in Tables 8-2 and 8-3 and are discussed in the following sections with respect to the results of the environmental group comparisons and the relevant literature.
Skin Rashes, Eczema, or Other Skin Allergies
The analyses revealed an excess of "skin rashes, eczema, or other skin allergies" among registrants compared with the NHIS sample (Table 8-2). The differences noted were consistent for males and females, and consistent across age groups. In addition to real differences between the two samples in the percentages of people experiencing skin rashes, this finding might also reflect differences in the NHIS and TCE Subregistry wording of this health condition. Table 5-2 indicates that the NHIS and TCE Subregistry health conditions for "skin rash" did not match exactly but were similar. The NHIS question was restricted to psoriasis, dermatitis, or dry (itching) skin; the TCE Subregistry health condition was phrased more generallyskin rashes, eczema, or other skin allergies.
Effects seen following dermal exposure to TCE are usually the consequence of direct skin contact with concentrated solutions, but occupational exposure also involves vapor contact. Persons working with TCE for intermediate periods sometimes develop skin burns or rashes and dermatitis (32). It has been reported that some people can be particularly sensitive to TCE and develop allergies when exposed to high levels in the air or on their skin during occupational exposures of intermediate duration (33,34).
Table 8-2.Summary results of NHIS comparison.
Disease Category |
Age Groups (years) | ||||||||||||||||||
| 0-9 | 10-17 | 18-24 | 25-34 | 35-44 | 45-54 | 55-64 | 65 | All | All | ||||||||||
| M | F | M | F | M | F | M | F | M | F | M | F | M | F | M | F | M | F | ||
| Speech impairment | X | X | |||||||||||||||||
| Hearing impairment | X | X | R | R | R | R | R | R | R | R | R | R | |||||||
| Stroke | X | X | X | X | X | X | |||||||||||||
| Liver problems | X | ||||||||||||||||||
| Anemia and other blood disorders | X | X | X | X | X | X | X | ||||||||||||
| Diabetes | X | X | |||||||||||||||||
| Kidney disease | X | ||||||||||||||||||
| Urinary tract
disorders |
X | X | X | X | X | X | X | X | |||||||||||
| Skin rashes | X | ||||||||||||||||||
| Asthma, emphysema | R | ||||||||||||||||||
| Arthritis | R | R | |||||||||||||||||
| Respiratory allergies | R | R | |||||||||||||||||
X = Statistically significant differences, TCE subregistry rate higher.
R = Statistically significant differences, NHIS rate higher.
= Insufficient data.
Although neither the etiology nor pathogenesis of progressive systemic sclerosis (scleroderma) has been established, this rare skin condition has been associated with a wide variety of seemingly unrelated compounds, including exposure to organic solvents (35-38). It should be noted that most literature associating scleroderma with exposure to TCE has been based on the reporting of single cases. Flindt-Hansen and Isager (39) reported three cases of scleroderma that developed after occupational exposure to TCE and trichloroethane (TCA). Lockey et al. (40) reported a case of progressive scleroderma that developed in a previously healthy 47-year-old woman who was occupationally exposed to TCE. Renal and skin biopsies were consistent with progressive systemic sclerosis. Saihan (41) reported scleroderma, pigmentation, and Raynaud's phenomenon in a middle-aged man following prolonged exposure to TCE. Finally, Yanez Diaz et al. (42) reported progressive systemic sclerosis as being TCE induced.
Significantly increased symptoms of lupus erythematosus were reported in a resident study population in Tucson, Arizona, reported to have been exposed to TCE and other chemicals
Table 8-3.--Summary of mortality analyses, 1985-1989 (analysis restricted to registrants 15 years of age or older at time of death).
| Age Group (years) | Cause of Death | SMR* |
| 75 - 79 | Malignant neoplasm of respiratory system | 6.62 |
| 75 - 79 | Other diseases of circulatory system | 4.64 |
| All ages | Diseases of the heart | 1.59 |
| All ages | Accidents | 3.91 |
| 25 - 29 | All causes | 6.28 |
| 60 - 64 | All causes | 2.67 |
| 75 - 79 | All causes | 2.54 |
| All ages | All causes | 1.84 |
*Significant two-sided Poisson - p<.01.
and metals at levels ranging from 6 to >500 parts per billion (ppb) for up to 25 years. Females in the exposed group reported more symptoms than the control females (43).
Some of the participants in the Woburn, Massachusetts, study reported skin lesions (maculopapular rashes) that occurred approximately twice yearly and lasted from 2 to 4 weeks (44). These lesions ceased from 1 to 2 years after cessation of exposure to TCE. Serious limitations of all of the Woburn studies have been extensively reviewed (45-49).
One study (50) in which TCE was applied to the skin of guinea pigs reported the presence of karyopyknosis, karyolysis, spongiosis, and pseudoeosinophilic infiltration in biopsies taken at different times of exposure. Another animal study reported that guinea pigs exhibited considerable erythema, edema, and increased epidermal thickness following an uncovered, dermal exposure to TCE (51).
In general, based on the literature, the finding of excess skin rashes, eczema, or other skin allergies in a population exposed to TCE and other volatile organic compounds is not unexpected. Although some studies suggest additional skin problems following TCE exposures, these were not specifically identified in the TCE Subregistry population. It might be of value for researchers to further investigate the cases of skin conditions in order to identify the specific disorders included in the subregistry.
The analyses revealed an excess of speech impairment among registrants aged 0 through 9 years compared with the NHIS sample of the same age group (no sex difference) (risk ratio = 2.45, 6 expected and 17 observed). The difference in the NHIS and TCE Subregistry wording of this health condition (the subregistry question was stated as speech impairment, the NHIS question as stammering and stuttering, other speech impairment) might have introduced differences in the reported outcomes.
Another consideration when exploring this finding is the fact that the TCE Subregistry population had a lower percentage of children 4 years of age or less than the NHIS population (the percentages were equal or close for those 5 through 9 years of age). Therefore, the TCE Subregistry group 9 years of age or younger would have a larger proportion of children older than 4 years of age than the comparable group for the NHIS population. This is important because many of the health conditions, including speech impairment, would be diagnosed when the children entered school. In looking at the rate of reporting in the two age groups, however, there was a 2.7-fold increase in reports of speech impairment for the 4 years of age and younger group and a 2.5-fold increase for the 5 through 9 years of age group. The reporting was consistent in the two age groups and combining them did not create results that were not applicable to each of the age groups.
Following the observation that TCE produces selective analgesia of the face in the area of the trigeminal nerve distribution (52), TCE was used as a treatment for trigeminal neuralgia. (Note: Spencer and Schaumburg (53) have postulated that dichloroacetylene, a breakdown product of TCE, is probably responsible for trigeminal nerve toxicity.) TCE has also been used in the past as an inhalant analgesic-anesthetic for dental procedures because of its recognized analgesic properties (54). TCE is no longer employed as a general anesthetic or an analgesic, however, due to residual neuropathy characterized by nerve damage (55-60). Neuropathy of this type is characterized by facial numbness (indicating damage to cranial nerve VII), and jaw weakness and facial discomfort (indicating damage to cranial nerve V, which is the trigeminal nerve) that can persist for several months (61,62). In addition, Feldman et al. (63) reported a patient with symptoms, including extensive anesthesia of the face, oral cavity, and tongue mucosa, similar to those reported by Plessner in 1915.
Trigeminal nerve impairment is frequently seen in chronic trichloroethylene intoxication (64-66). Facial hypesthesia, when present, is typically predominant in the mandibular and maxillary nerve areas and might or might not be associated with absent reflexes. Chronic exposure in the workplace has also been associated with damage to cranial nerves in several cases (67-73) and has more recently been evaluated by means of the masseter and blink reflex (74). Reported neuropathies can persist for several months (61,63). Persons who have died from overexposure have shown degeneration of cranial nuclei in the brain stem (61). Complaints such as vertigo, fatigue, headache, short-term memory loss, fewer word associations, and increased misunderstanding occurred with higher frequencies in workers exposed for up to 15 years to a mean of 167 parts per million (ppm) TCE than in workers with lower exposures. Workers with discontinued exposure, however, were not evaluated (75).
Trigeminal nerve impairment has also been reported in populations exposed to TCE in drinking water. Residual damage to facial and trigeminal nerves, measured by a decreased blink reflex (indicating damage to cranial nerves V and VII), was reported 6 years after exposure in 28 persons in Woburn, Massachusetts, who were reported to have been exposed to TCE in their drinking water (65); however, dose-response relationships could not be determined.
In an animal study, Kossler (76) reported that TCE depressed the force development of both twitch and tetanic tension of isolated muscles prepared from frogs and rats in a dose-dependent manner. The results suggested that the interaction of TCE with membrane sites is responsible for Ca2+ release for contractile processes.
It should be noted that the greatest number of individuals in the TCE Subregistry sample exhibiting speech impairments were males and females in the 0 through 9 years of age group. Although several factors might have played a role in the observed increase in speech impairment in this particular group, there were three TCE-related factors that should be considered. First, children of this age could have been exposed to TCE at most of the subregistry sites while in utero, as well as following birth. Second, the actual dose received by the target tissues for children is greater than for adults, given the same exposures. Third, given that maturation of the central nervous system (CNS) continues after birth, it is possible that it is during this period that TCE might elicit an effect. Finally, TCE exposure has been linked to short-term memory loss, fewer word associations, and increased misunderstandings, which would be detrimental to the development of speech because sounds might not be remembered or might be misunderstood. It would be informative to investigate the TCE Subregistry population in order to determine, if possible, what actual physical manifestations, if any, contributed to the reported increase in speech impairments. The small number of reported positive responses, while significantly statistically increased, might limit the scope of further research.
A significant increase in reported hearing impairment by TCE registrants was found for the 0 through 9 years of age group (risk ratio = 2.13, 8 expected and 16 observed) compared with the NHIS group. In the analysis described in Section 5, age was a significant factor, sex was not. The ratio of the number of registrants reporting a hearing impairment to the NHIS sample number reporting that they had a hearing impairment decreased with increasing age (the risk ratio became less than 1 from 18 years of age to less than 0.30 at later ages). Only 4 out of 33 registrants in the 0 through 9 years of age group who reported having a speech or hearing impairment reported having both types of impairment. The TCE Subregistry and NHIS version of this health condition were classified as being similar (hearing impairment versus deafness in one or both ears or any other trouble hearing with one or both ears) (see Table 5-2). It should be noted that most school systems in the United States systematically evaluate hearing in children, with subsequent referrals to physicians when abnormalities are detected. Therefore, parents in both populations would be equally likely to have knowledge of hearing deficiencies in their children.
As for speech impairment, the concern that a larger percentage of individuals older than 4 years of age included in the 0 through 9 years of age group for the TCE Subregistry population than the comparable 0 through 9 years of age group in the NHIS population would affect the data was operative. In the case of hearing impairment, there was a considerable difference in the two age groups--a 5.2-fold increase over NHIS rates in hearing impairments for the younger than 5 years of age group was reported, while a 1.8-fold increase was reported for those in the 5 through 9 years of age group. Combining the age groups, therefore, did affect the comparison of the combined groups.
In 1971, Tomasini and Sartorelli reported symmetrical bilateral eighth cranial nerve deafness, with subsequent recovery, in a patient who had been overexposed to TCE during operation of a dry-cleaning unit (77). Other examples of deafness following occupational exposure to either TCE (78) or organic solvents containing TCE and noise (79,80) have been reported.
Flanagan et al. (81) reported tinnitus in patients following acute poisoning with organic solvents from consumer products. Morrow et al. (82) determined that the latencies of the N250 and P300 components of the auditory event-related potential were significantly delayed in patients exposed to organic solvents (including TCE) when compared with those in normal controls. (Note: See Morrow et al. (82) for references to the history of the association between organic brain dysfunction and increased latency of event-related potential waveforms, especially the later components such as P300.) The delay in N250 and P300 provides evidence that solvent exposure slows CNS mechanisms that evaluate and/or process relevant auditory stimuli. In addition, complaints such as fewer word associations and increased misunderstanding have been reported in workers exposed for longer than 15 years to TCE (75).
Rebert et al. (83) reported hearing loss in male Long-Evans and Fischer-344 rats following exposure to TCE. The hearing loss reported was predominantly high frequency. Another study found an antagonistic interaction between TCE and noise at varying levels and durations of exposure to solvents, including TCE (84).
As mentioned in the discussion of speech impairments, maturation of the nervous system is still occurring in children younger than 10 years of age; therefore, exposure that might have occurred in utero or during early childhood, when the dose received would be proportionally greater than in later years, might affect this development. In addition, perceived hearing losses might be, in part, due to "fewer word associations and increased misunderstandings." In light of these possibilities, determination of exact medical diagnoses would be needed. In addition, further evaluation of the cranial nerves of this subset of the TCE Subregistry population by researchers would be informative.
When the registrants were grouped by length of exposure to TCE, a statistically significant association between length of exposure and reported hearing impairments was found. However, when the population was limited to those over 18 years of age, the results were no longer statistically significant; this suggests that the differences found were in the subpopulation less than 19 years of age.
It should be noted that hearing impairments could be due to ear infections, which were not investigated but might have occurred in excess in this population. This possibility, as well as the potential skewing of the outcome by the decreased number of registrants less than 4 years of age, should be addressed in future studies. Also, as with the hearing impairment studies, the small number of reported positive cases, while statistically significantly increased, might limit the scope of further research.
The wording of the NHIS and TCE Subregistry health condition for diabetes matched. Diabetes was reported at statistically significant higher rates for females in age groups 18 through 24 years and 45 through 54 years only (there was no excess of diabetes for males). Diabetes is typically underdiagnosed (85), therefore the actual incidence of diabetes in the TCE Subregistry population and the NHIS population could be greater than reported. Regional variations in the incidence of diabetes have also been reported (Table 8-4); however, the number of cases of diabetes for the TCE Subregistry population exceeded the rate for the Midwest (where all of the sites are located).
Few studies are available that explore a possible relationship between chemical exposure and the development of diabetes. In a study of workers in the corn wet-milling industry who were occupationally exposed to grain dusts, pesticides and fumigants, acids, solvents, sulphur dioxide, and other chemicals used in the manufacture of starch, oil, syrup, and dextrins, Thomas et al. (86) reported an elevated frequency of deaths due to diabetes and a threefold excess of pancreatic cancer deaths among blacks. In addition, diabetes has been related to exposure to various drugs and a few chemicals, including carbon disulfide, nicotinic acid, edetic acid (EDTA), ethanol, nickel chloride, and Vacor (a rodenticide) (87).
It has been reported that the administration of glucose or insulin increases the amount and speed of excretion of metabolites of TCE (88). It should be noted that one of the major metabolites of TCE, trichloroacetic acid, is a suspected teratogen (89); therefore, an insulin-dependent population might be more susceptible to TCE exposure.
In light of the fact that female registrants in the 18 through 24 years of age group (with TCE exposures occurring during childhood and teen years) reported an excess of diabetes, it would be important for researchers to confirm the diagnoses of diabetes in this subgroup. In addition, screening for diabetes in the TCE subpopulation would provide important additional information. Finally, a more in-depth occupational history is needed in order to identify those persons known to have been exposed to chemicals thought to be associated with diabetes.
The wording of the NHIS and TCE Subregistry health condition for the effects of a stroke was a close match. Stroke was reported in statistical excess in the total population for all groups 35 years of age and older; however, the excess in the group 55 through 64 years of age was not statistically significant. (Note: For 0 through 24 years of age, the data were too sparse for analyses.) Of the nine registrants in the 35 through 44 years of age group who reported the effects of a stroke, six also reported having hypertension; of the eight registrants who reported stroke effects in the age group 45 through 54 years, five also reported hypertension. For the 65 years and older age group, 15 of the 43 registrants who reported the effects of a stroke also reported having hypertension.
One study was located that reported a borderline significance for relative risk of stroke in a population that had been exposed to chlorinated domestic drinking water for 12 years (90). (Note: The by-products of chlorination can include TCE, although it is considered unusual.) It is important to note, however, that hypertensionwhich occurs approximately twice as
Table 8-4.--Number of cases of diabetes per 1,000 persons, by geographic region and place of residence: United States, 1989 to 1990 (85).
| Location | Number of Cases
Per 1,000 Persons |
| Geographic Region | |
| Northeast | 25.2 |
| Midwest | 25.9 |
| South | 29.8 |
| West | 23.9 |
| Place of Residence | |
| All MSA* | 23.9 |
| Central city MSA* | 28.3 |
| Not central city MSA* | 21.1 |
| Not MSA* | 36.2 |
*Metropolitan Statistical Area
frequently in persons with diabetes (91)contributes to stroke, the incidence of which is also twice as high among diabetics as among nondiabetics (92,93). In the TCE Subregistry population, of the 32 individuals reporting stroke, 11 were diabetic (6 males and 5 females, all over 55 years of age). No animal studies specific to the effects of a stroke following exposure to TCE were located.
In addition to the NHIS comparison results, the analyses of the environmental subgroups indicated a very strong association of reported stroke effects with TCE levels. For the mortality data analyses, there was a significantly higher standardized mortality rate (SMR) for the category other diseases of the respiratory system (which included stroke) for the 75 through 79 years of age group for the years 1985 through 1989.
Again, it should be noted that smoking also contributes to the incidence of stroke. It is unknown if smoking enhances the effects of TCE. Given the limited amount of information available on TCE and stroke, this area needs further investigation.
Urinary Tract Disorders, Including Prostate Trouble
The wording of the NHIS and TCE Subregistry health condition for urinary tract disorders, including prostate trouble, is not an exact match (see Table 5-2 and Appendix B). It should be noted that the NHIS question was specific for bladder disorders, while the TCE questionnaire included the broader category of all urinary tract disorders, which might account for a portion of the reported cases in the TCE population. The excess positive reporting for urinary tract disorders was significant for all females, with the exception of the 45 through 54 years of age group (the data were too sparse in the 10 through 17 years of age group for analysis). For males, as well as females, there were an excess number of cases reported in the 18 through 24 and 25 through 34 years of age groups (the data were too sparse for the 0 through 17 years of age group for analysis). It is of interest to note that, for females, the risk ratios decreased from 12.79 (1 expected and 11 observed) in the 0 through 9 years of age group to 2.62 in the 65 years and older age group. Also, the risk ratios for the males and females were approximately equal for the 18 through 24 (risk ratio = 6.11 for males, risk ratio = 6.38 for females) and 25 through 34 years (risk ratio = 6.38 for males, risk ratio = 5.56 for females) of age groups for which both sexes had an excess number of urinary tract disorders or prostate problems.
Few studies are available that investigate the potential impact of TCE exposure on the urinary tract. In a study of workers exposed to solvents in the paint industry, Lundberg (94) reported that three cleaners had died from infectious urinary tract disease, in comparison to 0.2 deaths expected from all genitourinary diseases. Although the population was small and very heavily exposed, the author believed this association might have indicated a modifying effect on the prognosis of urinary tract morbidity from solvents. One study is available that suggests an association between cumulative exposure to solvent-contaminated well water and increased urinary tract infection in children (31); however, exposure was to a number of solvents and no clinical tests were run on the urine.
Urinary tract infection has been noted to occur more frequently in diabetics than in nondiabetics. Of the 145 individuals reporting urinary tract disorders in the TCE Subregistry population, 6 were diabetic (all over the age of 45 years). The prevalence of infection was higher in female diabetics (18% to 41%) than in male diabetics (1% to 18%) (95,96). In the TCE population, 7% of the diabetic females also reported urinary tract disorders, as did 2% of the diabetic males. Bladder paralysis and urinary retention, which contribute to infections, have also been reported with diabetes (97,98).
Given the sparsity of information pertaining to the relationship between TCE exposure and urinary problems, additional research is needed. Also, researchers should explore making the questions more comparable between comparison groups for future studies.
The wording of the NHIS and TCE Subregistry question for kidney disease was a close match. The comparable question in the NHIS questionnaire was kidney stones, infections, other kidney trouble. Kidney problems were reported statistically more often among the TCE subregistry for females in the 55 through 64 years of age group (risk ratio = 4.57, 2 expected and 11 observed).
David et al. (99) studied one worker in a degreasing operation who developed acute renal failure due to acute allergic interstitial nephritis with secondary tubular necrosis; however, this reaction could not be definitely attributed to TCE. Slight renal effects, indicated by changes in urinary proteins and enzymes, have been reported in workers exposed to TCE in mixtures with other substances (100,101).
In a study of workers in the jewelry industry, Dubrow and Gute (102) reported elevated proportionate mortality ratios (PMRs) for nonmalignant kidney disease in males. Exposure to known renal toxins (heavy metals and solvents) used in the jewelry industry was thought to account for the excess numbers of deaths from kidney disease. The authors cautioned that, because of the lack of information about specific occupational exposures of the decedents, their study should be viewed as an exploratory investigation requiring further follow-up.
According to Kluwe et al. (103), although the data on humans are not definitive, causal associations appear to exist between occupational exposure to some hydrocarbon solvents and chronic kidney disease. They reviewed studies that they believe indicated a potential for organic chemicals, especially halogenated hydrocarbons and aromatic amines, to produce chronic kidney injury in humans and other mammalian species.
One study reported hepatorenal failure as the cause of death following oral exposure to TCE, but a dose was not determined (105); however, another case report stated that urine and blood analyses revealed no hepatic or renal injury in a man who drank several tablespoons of TCE (105).
Renal failure is one of the major chronic complications of diabetes (106,107). Compared with the rest of the population, people who have diabetes are at a 17-fold increased risk for end-stage renal disease (97) related to hypertension. Of the TCE Subregistry diabetics, 11% also reported hypertension and less than 1% reported kidney disease. Given this information, it will be important to observe the rates for kidney disease in future followups of the TCE Subregistry. Finally, proteinuria, a clinical marker for nephropathy, develops in from 18% to 30% of persons who have had Type I or Type II diabetes for at least 15 years (97).
Slight renal effects have been found in animals following their exposure to TCE. In animal studies, renal enlargement has been associated with acute or intermediate duration inhalation or oral exposure. Kidney enlargement seemed to be less pronounced and occurred less consistently than liver enlargement. Enlargement has been associated with altered biochemical indices of renal function, but not abnormal histology in some of the intermediate-duration animal studies (108).
Chronic experiments did reveal toxicity. Administration of high doses of TCE by gavage for 78 weeks to Osborne-Mendel rats and B6C3F1 mice resulted in treatment-related chronic nephropathy, characterized by degenerative changes in the tubular epithelium (109). In two carcinogenicity studies of rats and/or mice, nonneoplastic renal effects included toxic nephrosis (characterized as cytomegaly) at doses of 500 and 1,000 milligrams per kilogram (110,111), and cytomegaly of the renal tubular cells coupled with toxic nephropathy (112). Both studies recorded treatment-related neoplastic renal lesions. In animal cancer studies, however, deaths occurred by tumors or general demise (body weight loss, respiratory infection, renal failure, and CNS depression) due to a very high level of exposure.
In conclusion, the literature concerning TCE exposure and subsequent kidney disease is mixed. This result is also reflected in the data from the TCE Subregistry. It would be instructive for future researchers to identify the specific conditions in the subgroup reporting excess kidney disease.
Anemia or Other Blood Disorders
The wording of the NHIS and TCE Subregistry health condition for anemia was a close match (see Table 5-2). Anemia was reported statistically more often among TCE Subregistry members for selected age groups for both males and females: 0 through 9 years of age, males (risk ratio = 3.54, 1 expected and 5 observed); 18 through 24 years of age group, females (risk ratio = 2.18, 9 expected and 19 observed); 35 through 44 years of age group, males and females; 45 through 55 years of age group, females; and 65 years of age and older group, males. Looking in detail at the 0 through 9 years of age group increase for male registrants the results were mixed. The younger than 4 years of age group had very similar rates (1.6% for the TCE Subregistry population compared with 1.8% for the NHIS population); however, the 5 through 9 years of age registrant group rates increased to 2.9% and then declined, while the NHIS group rates declined to 0.2% for the same age group and continued to decline.
Reports in the literature of effects of TCE exposure on the blood and/or blood-forming organs of humans are also mixed. Most hematologic findings reported have been the result of occupational exposures. Mild macrocytic anemia has been observed in some workers (113,114), in addition to intolerance to alcohol, abnormal liver function tests, and an increase in total blood lipids. Microangiopathic hemolytic anemia was also reported in one case of progressive systemic sclerosis attributed to dermal TCE exposure (40). Among corn wet-milling workers exposed to solvents and other chemicals, mortality from lymphatic and hematopoietic malignancies was elevated in whites (86).
A decrease in the OKT11 (all) T-cell fraction and a decrease in the OKT4 helper cells were reported in solvent-exposed workers. This same group also showed an increase in the anti-Leu 7 positive cells (mostly natural killer cells), an important increase in anti-Leu 12 labeled T-cells (that is, human B-lymphocytes), and no differences in the OKT8 suppressor cells. It should be noted that similar changes in lymphocyte subpopulations are found in states of immunodeficiency and immunogenetic forms of aplastic anemia, a disease whose etiologic relationship might be due to long-term exposure to organic solvents (115). Immunologic abnormalities were also found in adults in Woburn, Massachusetts, who were family members of children with leukemia. The abnormalities reported included persistent lymphocytosis, increased numbers of T-lymphocytes, and a depressed helper:suppressor T-cell ratio. It should be noted, however, that there was a possible bias in reported risk factors (44). In contrast, Stewart et al. (116) reported that humans acutely exposed to TCE exhibited no adverse hematologic effects on blood cell counts, sedimentation rates, serum lipid levels, serum proteins, or serum enzymes.
In animal studies, TCE was reported to have direct action on the bone marrow of rabbits, causing myelotoxic anemia (117). Mice exposed to a mixture of organic solvents in their drinking water developed suppressed marrow granulocyte macrophage progenitors (118). This study showed that even 10 weeks after cessation of the chemical mixture treatment when all hematologic parameters were normal, a residual effect of the chemical mixture was still demonstrated as lower progenitor cell numbers following irradiation.
Mice, following intermediate oral exposure to TCE, showed minor changes in hematology (including a 5% lower hematocrit), but no changes in white cell counts, platelet counts, or prothrombin time (119). The changes were not considered to be toxicologically significant since they were not dose related and were within normal values for mice. Various hematologic effects have been reported in other animal studies (120,121), but since hemoglobin concentration in erythrocytes did not change the changes were not considered to be adverse.
In view of the reported literature, it is possible that chronic TCE exposure can result in blood disorders. It must be noted, however, that the reported excesses in rates did not show a pattern with respect to age or sex. In addition, medical confirmations of these diagnoses have not been conducted. Further study is needed before definitive conclusions can be reached.
Cancer
The TCE Subregistry question asked for information on all cancers, as did the NHIS question. The differences in the reporting rates for cancer were not statistically significant for the morbidity analyses (or for the mortality data except for malignant neoplasms of the respiratory system for males and all persons in the 75 through 79 years of age category for the years 1985 through 1989). The comparison with the SEER data showed that the relative risk ratios for females increased from 1973 through 1988. The male and combined ratios increased less consistently over time. The relative risk ratio was 0.99 for males, 1.89 for females, and 1.50 for both sexes in 1988. More definitive information is needed to interpret these findings. It is, however, important to review the current controversy surrounding the carcinogenic potential of TCE.
Studies of the potential carcinogenic properties of TCE have been varied. Two studies that reviewed mortality statistics for the period 1969 through 1979 for a group of people in Woburn, Massachusetts, who drank water from a domestic water supply reported to be contaminated with industrial solvents (including TCE, tetrachloroethylene, and 1,2-trans-dichloroethylene) concluded that there was a statistically significant elevated rate of childhood leukemia (122,123); however, etiologic factors were not identified.
A fifth study (31) found a potential association between ingestion of drinking water contaminated with solvents in Woburn and increased risk of childhood leukemia, particularly acute lymphocytic leukemia. Other researchers (45-49) reevaluated the data and did not concur because of the numerous problems associated with this study, including residents' exposure to a mixture of chemical contaminants and the use of inadequate statistical methods. Another study of the Woburn population also found an unusually high incidence of leukemia in children exposed in utero (44). However, this study was also found to be flawed (108).
In a nonconcurrent prospective study of persons who used chlorinated surface water as a domestic water source for 12 years versus persons who used unchlorinated well water as a domestic water source, Wilkins and Comstock (90) reported that incidence rates for cancer of the bladder among men and for cancer of the liver among women were nearly twofold for persons exposed to the chlorinated water. A comparative mortality study also suggested an association of chlorinated water with cancers of the liver and urinary tract. A significantly elevated risk was observed for cancer of the breast.
An ecologic study of persons exposed to volatile organic compounds (VOCs), particularly TCE, in drinking water supplies (124) found the standardized incidence ratio of leukemia was elevated for females in towns in the highest of three exposure categories. No association was observed in males in any of the exposure categories. The increase in risk among females with increasing levels of contamination appeared to be distributed evenly among all age groups. The observed association appeared to suggest that drinking water contaminated with VOCs might increase the incidence of leukemia among exposed females, but caution was advised in the interpretation of the results due to uncertainties inherent in ecologic studies.
Mallin (125) reported a high incidence of bladder cancer in a section of Winnebago County in northwest Illinois. One of four drinking water wells had been closed in that area due to contamination with TCE, tetrachloroethylene, and other solvents. The reported excess in the number of bladder cancers was primarily confined to one town in which standardized incidence ratios were significantly elevated in males (1.7) and females (2.6). Investigation of this cluster is ongoing.
Previous studies have also detected associations between chlorinated water and genitourinary and gastrointestinal cancers, but most of the studies have been ecologic in design (126,127). One case control study, however, detected a twofold relative risk for bladder cancer in individuals with 40 or more years of exposure to chlorinated surface water, after adjusting for known risk factors (128).
Results from occupational studies have also been mixed. Following a report of a cancer cluster, Zoloth et al. (129) initiated a proportionate analysis of cause of death in 1,401 commercial pressmen who were occupationally exposed to solvents. They found a statistically significant elevated risk of all cancers, colorectal cancer, and cancers of the lymphatic and hematopoietic system, with non-Hodgkin's lymphomas responsible for much of the excess risk. Pressmen who had been employed for 20 or more years had a significantly elevated risk for cancers of the liver and pancreas; however, no excess risk of bladder or lung cancer or leukemias was seen.
Elevated frequencies of deaths from leukemia have been reported among white maintenance workers in the corn wet-milling industry who were exposed to multiple chemicals, including solvents (86). The same study reported a threefold excess number of pancreatic cancer deaths among black workers.
An update of the Axelson et al. (130) epidemiologic study of occupational exposures, which evaluated an expanded cohort of 1,424 men (levels of TCE not specified), found significant excess numbers of bladder cancers and lymphomas, but a lower than expected incidence of total cancer mortality, possibly due to the healthy worker effect (131).
Spiegelman and Wegman (132) reported suggestive evidence for colon cancer risk in males with potentially high exposure to solvents, abrasives, and fuel oil, and those with jobs with high stress; suggestive evidence emerged for females with potentially high exposure to dyes, solvents, and grinding wheel dust. Fredriksson et al. (133) found a slightly increased risk for colon cancer following exposure to TCE in general. In a study of highway workers exposed to solvents, herbicides, asphalt and welding fumes, diesel and auto exhausts, asbestos, abrasive dusts, hazardous materials spills, and moving motor vehicles, white retirees with 5 or more years of service experienced an excess number of deaths due to cancers of the colon, skin, and brain, and to lymphosarcomas and reticulosarcomas (134).
In the followup and confirmation of a previous study of occupational exposures to solvents, Hernberg et al. (135) reported the odds ratio as greater than 3 for liver cancer in solvent-exposed women, irrespective of reference group, and no increase for men. Among jewelry workers, elevated PMRs were reported for liver cancer in males and for stomach cancer in females. The elevated PMR for liver cancer might have been due to exposure to solvents (TCE, tetrachloroethylene, and carbon tetrachloride) that cause liver cancer in animals. However, because of the lack of information about specific occupational exposures of the decedents, this should be viewed as an exploratory investigation requiring further follow-up (102). It is important to note that chronic exposure to TCE results in hepatocellular cancer in mice, but not in rats (136).
Garland et al. (137) found a statistically significant elevated age-adjusted rate of testicular cancer in naval aviation support equipment technicians, enginemen, and construction mechanics when compared with the U.S. population and the total Navy population. These occupations involve maintenance of internal combustion engines and exposure to the attendant lubricants, solvents, paints, and exhausts.
Various other cancers have been reported following exposure to TCE. In a study of workers exposed to TCE in the paint industry, Lundberg (94) reported three workers had died from multiple myeloma (0.6 expected) and related the excess number of deaths to solvent exposure. Maizlish et al. (134) found that white male highway workers exhibited a statistically significant excess number of cancers of the digestive organs, lymphopoietic cancers, and benign neoplasms.
Kluwe et al. (103) found preliminary evidence that suggested an association between organic chemical exposure and cancers of the urinary tract in humans. Many literature reports indicating that short-chain halogenated hydrocarbons appear to have a propensity for causing a low incidence of renal tubular carcinoma in exposed rodents tend to support these findings. Comparative analyses of the National Toxicology Program/National Cancer Institute studies did not indicate consistent sex or species differences in the nonneoplastic chronic toxic response, but chemically induced urinary tract cancers occurred more commonly in rats than in mice and chemically induced cancers of the kidney occurred more commonly in males than in females. Viewed collectively these data could indicate a potential for organic chemicals, especially halogenated hydrocarbons and aromatic amines, to produce chronic kidney injury in humans and other mammalian species. Harrington et al. (138) found no relation between occupational exposures to solvents and renal cancer. However, the statistical power of this observation has been questioned.
Several epidemiologic studies (130,131,139,140) of occupational exposures to TCE did not find significant increases in the overall incidence of cancer, but were limited by relatively small numbers of subjects and lack of lengthy follow-up periods. Thus, such studies would not be expected to detect uncommon cancers or weak carcinogens (108). One study (141) investigated whether working with solvents, particularly TCE, posed any excess risk of mor-tality. Significant deficits occurred for mortality from all causes and all malignant neoplasms in solvent-exposed workers; however, mortality was increased for multiple myeloma and non-Hodgkin's lymphoma in white women. The inconsistent mortality patterns by sex, multiple and overlapping exposures, and small numbers made it difficult to ascribe these excesses to any particular substance.
O'Leary et al. (140) studied the relationship of parental occupational exposures to childhood malignancies and concluded that the preponderance of evidence supported the hypothesis that occupational exposures of parents to chemicals increased the risk of malignancies in their children. The parental occupational exposures implicated in childhood malignancy risk were to chemicals such as paints, petroleum products, solvents (especially chlorinated hydrocarbons), pesticides, and metals. The available data, however, did not allow the identification of specific etiologic agents. Odom et al. (142) reported that the occurrence of acute monoblastic leukemia in young children appeared to be associated with in utero exposure to marijuana and parental exposure to pesticides and solvents.
No studies have been reported regarding cancer in humans after dermal exposure to TCE. In three animal studies, no significant tumor incidences were observed and doses were well below the maximum tolerated level following dermal exposure to TCE (108).
Animal studies have linked TCE exposure to various types of cancers in rats and mice. Statistically significant increases in cancers include mouse hepatocellular carcinomas, forestomach tumors in Swiss mice, renal tubular cell adenocarcinomas in male rats but not female, increased incidence of leukemia in male rats but not female, renal tubular cell adenomas in male rats but not female, interstitial cell tumors of the testis in rats, and mouse hepatocellular carcinomas and hepatocellular adenomas (108). Crebelli and Carere (143) reported that exposure to high doses of TCE, administered to rodents during long-term carcinogenicity studies, resulted in the induction of liver and lung tumors in the mice, and tumors of the kidney and testes in rats. Among TCE metabolites, trichloroacetic acid was reported to be carcinogenic to the livers of mice. The incidence data for lung tumors in female Swiss mice, together with other tumor incidences, were used by the U.S. Environmental Protection Agency (EPA) (144) to derive a carcinogenicity potency estimate; however, this cancer rating was recently withdrawn (145).
One problem with some animal studies is the use of TCE containing small amounts of epoxide stabilizers to preserve the TCE from rapid degradation. Since these epoxides form free radicals, they themselves might be carcinogens, and might have contributed to the carcinogenic potential of industrial TCE. Thus, the significance of these studies could not be determined. Another serious problem with many of the studies is the poor survival rate (111,112,146).
One study that did use epoxide-free TCE showed liver tumors in mice, some indication of renal tumors in male rats, and no evidence of carcinogenicity in female rats (111). Acute oral exposure to TCE or its metabolites preferentially induces peroxisome in the liver of mice, which might be related to the carcinogenic response in the liver of mice (147).
Sprague-Dawley rats and B6C3F1 mice administered TCE intraperitoneally displayed minor decreases in splenocyte viability, inhibition of LPS-stimulated mitogenesis in rat cells, and marked inhibition of rat and mouse hepatic natural killer cells and rat natural cytotoxic cell activities in all groups of effector cells (148). These results indicate that TCE is able to inhibit the activity of lymphocytotoxic cells that are involved in the immune surveillance of cancerous cells, suggesting the possibility that compromised immune function might play a role in the carcinogenic responses in experimental animals exposed to TCE.
The carcinogenic status of TCE is currently being reviewed by EPA. The TCE Subregistry sample does not report an excess of cancers, except for older males (mortality data). Given the type of cancer (malignant neoplasms of the respiratory system) and the incidence of smoking in the TCE Subregistry sample, it would be inappropriate to relate this excess solely to TCE exposure. The increases in female rates from the TCE Subregistry data compared with SEER rates should be explored by researchers. Future studies using the TCE Subregistry sample should attempt to address specific types of cancer with respect to comparison data in order to answer questions posed by findings in the literature.
SUMMARY
This section has reviewed the scientific literature for TCE exposure and health effects. The literature suggests that there might be associations between TCE exposure and some of the health outcomes that are reported in excess by the Subregistry data. The reported literature has many limitations: case reports of human poisonings and occupational studies usually involve exposure levels much higher than those reported in environmental exposures; high dose animal studies may not be relevant to humans; and human health studies often lack sufficient exposure characterization, controls for important confounding factors, and sample sizes large enough to investigate low-dose effects. These and other limitations must be considered by the reader. This section does not support a cause and effect relationship between TCE exposure and human health outcomes. However, interesting areas for further investigation are mentioned.
Agency for Toxic Substances and Disease Registry, 1825 Century Blvd, Atlanta, GA 30345
Contact CDC: 800-232-4636 / TTY: 888-232-6348
