ATSDR -The goals and objectives of the TCA Subregistry

ATSDR ›
NER ›
TCA ›
The goals and objectives of the TCA Subregistry

The goals and objectives of the TCA Subregistry

SECTION 1 INTRODUCTION

This is a report on the baseline activities and findings from the analyses of data collected from members of the 1,1,1-Trichloroethane (TCA) Subregistry of the Volatile Organic Compounds (VOCs) Registry. The VOCs Registry is part of the National Exposure Registry (NER), which were created and are being maintained by the Agency for Toxic Substances and Disease Registry (ATSDR).

In 1988, the policies and procedures proposed for the NER were reviewed extensively by several committees composed of independent scientists, state representatives, representatives of other federal agencies, and other interested people. The revised policies and procedures were published in the NER Policies and Procedures Manual (1). The VOCs Registry was one of the first registries to be established as part of the NER program. The NER currently contain four substance-specific registries (VOCs, Dioxins, Heavy Metals, and Radioactive Substances) (see Figure 1-1). The VOCs Registry currently contains three chemical-specific subregistries (Trichloroethylene, Benzene, and TCA).

The goals and objectives of the TCA Subregistry reflect those of the NER; specifically, the TCA Subregistry will be used to facilitate epidemiologic or health studies and surveillance and will provide information that can be used to assess the effects of exposure to TCA on a general population. In addition, the TCA Subregistry will enable federal, state, and local officials to provide exposed persons with timely, relevant information about TCA exposure, potential adverse effects related to that exposure, preventive measures, and therapeutic advances that were not understood when the TCA Subregistry was established. The methodology used to accomplish these goals follows the multisite concept.

The Policies and Procedures Manual (1) describes in detail all policies, procedures, and operational details pertinent to establishing the TCA and other subregistries of the NER. Specific topics from the policies and procedures document are reiterated in this report, where necessary, for clarity.

The objective of this report is to present the results of the statistical analyses comparing the reporting rates of registrants for specific health outcomes with national norms. The report is the third of a set of ongoing reports and publications that will summarize the latest VOCs Registry findings for specific subregistries and suggest specific hypotheses for future research. The research will focus on these and potentially other residential populations that have experienced similar exposures to VOCs. This report highlights some health outcomes and predictive variables that should be considered for analysis during future epidemiologic or health studies related to TCA exposure.

Section 2 of the report reviews the rationale for the selection of TCA as a primary contaminant for the NER and provides a discussion of the information available at the time of its selection on the toxicity of TCA, based on the related epidemiologic and toxicologic studies, as well as information on data gaps that subregistry-generated information could fill. In addition, Section 2 presents general information on the number of National Priorities List sites where TCA was found in 1995, as well as a detailed discussion of the site meeting the criteria for inclusion in the TCA

Figure 1-1.-Design of the National Exposure Registry.

Subregistry. Section 2 also provides descriptive and summary information on the environmental data and period of exposure.

This report provides an overview of the characteristics and health status of registrants who took part in this baseline effort. Section 3 includes a summary of the TCA Subregistry registrants data. Section 3 also includes a comparison of the TCA Subregistry data with national survey data files for smoking habits, demographic characteristics, and reported rates of adverse health outcomes; Section 4 includes the same information for the TCA Subregistry data collected for the first followup. Section 5 summarizes the findings of the report and discusses the findings in relationship to the published literature. Section 6 states the conclusions of the analysis of the TCA Subregistry baseline data and outlines future activities related to the TCA Subregistry.

SECTION 2 BACKGROUND OF THE 1,1,1-TRICHLOROETHANE SUBREGISTRY

SELECTION OF 1,1,1-TRICHLOROETHANE AS A PRIMARY CONTAMINANT

In keeping with the National Exposure Registry (NER) procedures for substance selection (1), the factors that led to the selection of 1,1,1-trichloroethane (TCA) included the prioritization of TCA on the Hazardous Substance Priority List (2); the ubiquitousness of TCA in the environment; the published evidence of TCA toxicity in worker populations and in toxicologic studies; and the paucity of information on low-level, long-term exposures to TCA. Each of these factors suggested that the NER could contribute significantly to the detection of an excess in adverse human effects, should they exist, following long-term, low-level exposure to TCA in the environment.

Environmental Information

TCA is a synthetic compound that is released to the environment as a result of anthropogenic activity (3). TCA is used as a solvent for adhesives (including food packaging adhesives), metal degreasing, and textile processing and is used in the manufacturing of pesticides, cutting fluids, aerosols, lubricants, cutting oil formulations, drain cleaners, shoe polishes, spot cleaners, printing inks, and stain repellents. TCA is emitted during use of items prevalent in the average home, such as liquid detergent, wallpaper glue, carpets, spray and solid insecticides, carpet glue, and chlorine bleach scouring powder.

TCA-contaminated groundwater was found at 59% (n = 380) of the National Priorities List (NPL) sites with TCA (Figure 2-1). Of these sites, 42% (n = 161) used private well systems, 19% (n = 72) had municipal systems, and 39% (n = 147) used both private and municipal systems to provide residents with drinking water (4).

The available literature indicates that TCA is toxic to humans and animals. A brief synopsis of the literature available in 1992, based on the Toxicological Profile for 1,1,1-Trichloroethane (3) follows. A more detailed discussion of the updated literature is given in Section 4.

Clinical symptoms associated with high-level exposure to TCA that have been reported in humans include hypotension, diarrhea and vomiting, central nervous system (CNS) depression, and dermal and ocular irritation (3). Mild hepatic effects can also occur in humans. Deaths have been attributed to cardiac arrhythmia and respiratory failure secondary to CNS depression. Effects reported in humans that also occur in animals include hypotension, cardiac arrhythmia, CNS depression, mild hepatic effects, and dermal irritation. Effects that have been observed in animals but not investigated in humans include mild developmental effects.

Immunological effects of TCA have not been reported in humans (Figure 2-2), and have not been studied extensively in animals (Figure 2-3). Acute inhalation exposures had no effect on survival from a bacterial pathogen challenge in mice. Histological evaluations of immune system tissues from rats and mice (including the lymph nodes, thymus, and spleen) have not revealed any lesions attributable to TCA exposure. However, more extensive studies of immune function would be required to adequately evaluate the immunotoxic potential of TCA in humans.

Figure 2-1.-Number and types of TCA sites in 1995.

Developmental effects of TCA in humans have not been reported. Mild embryotoxic effects were reported in rats and rabbits exposed to high concentrations of TCA by inhalation. Effects included decreased fetal weights, increased minor soft tissue and skeletal abnormalities, and delayed ossification. For one of these studies, developmental defects could have been associated with the significant maternal toxicity observed. Neither an inhalation study using a lower concentration nor a drinking water study found any developmental effects.

Reproductive effects of TCA in humans have not been reported. Histological evaluation of reproductive organs and tissues from male and female rats and mice revealed no lesions attributable to TCA exposure. More sensitive tests are required before a full evaluation of the potential for reproductive effects in humans can be made.

Although TCA was mutagenic in a few assays with Salmonella, induced chromosomal aberrations in a Chinese hamster ovary cell assay, and was positive in mammalian cell transformation assays, the existing genotoxicity data are largely negative. In addition, there is a possibility that positive results were produced by the stabilizers and not by TCA itself. Therefore, a firm conclusion regarding the genotoxic potential of TCA in humans is not possible.

Evidence for or against an association between exposure to TCA and cancer in humans has not been reported. Among animals, no effects were found in an inhalation study at exposure levels up to 1,500 parts per million (ppm). The results of an oral study indicated the TCA might have produced an increase in the occurrence of immunoblastic lymphosarcoma in rats. However, the biological and statistical significance of the results of the study were questionable because of the limitations of the study design.

There is also limited information on the role of TCA metabolites in the toxicity of the parent compound. Reactive metabolites are important in the carcinogenicity of other chloroethanes. Binding to deoxyribonucleic acid (DNA), which is correlated with carcinogenicity in chlorinated

Figure 2-2.-Existing information on human health effects of 1,1,1-trichloroethane (3).

Figure 2-3.-Existing information on health effects of 1,1,1-trichloroethane in animals (3).

ethanes, was weak in both in vivo and in vitro tests. Even weak binding, however, indicates an ability to interact with DNA. Cell biotransformation tests were positive for TCA. The results of these assays might have been confounded by the presence of stabilizing agents, however.

In summary, the TCA Subregistry was created in response to the pervasiveness of TCA at the nation's waste sites; the relative lack of information on human health outcomes associated with long-term, low-level exposures to TCA in drinking water supplies; and the knowledge of adverse health outcomes among people experiencing different types of TCA exposure (for example, short-term exposures at high concentrations). Subsequent knowledge concerning TCA exposure and human health, available since the selection of TCA and the establishment of the TCA Subregistry of the VOCs Registry and discussed in Section 4, further supports the establishment and maintenance of a TCA Subregistry.

SELECTION AND DESCRIPTION OF THE TCA SUBREGISTRY SITE AND POPULATION

The site selection process used to develop subregistries for the NER is described in detail in the Policies and Procedures Manual (1).

During the site selection process, 380 potential sites were identified as having TCA as a contaminant of groundwater. After reviewing the sites with respect to specific criteria, the site described in this section was selected as most appropriate for inclusion in the TCA Subregistry. The criteria for selection included documentation of exposure levels and duration of exposure, identification and estimated size of the exposed population, identification of susceptible sub-populations, and identification of the number and levels of secondary contaminants. Sites where state and local officials supported development of the subregistry were given priority. Although not a criterion for selection, preference was given to sites with available hydrogeologic information in order to better characterize registrants' exposure.

One site was selected for inclusion in the TCA Subregistry of the VOCs Registry. The name of the site-as provided for in the Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (5), amended by the Superfund Amendments and Reauthorization Act of 1986 (6)-is the Vestal Water Supply Well #1-1 site in Vestal, New York. It should be noted that this site used a public water system; therefore, for the purpose of this report, exposures for all registrants at the site were considered to be the same. Following is a summary of the site included in the TCA Subregistry of the VOCs Registry. This summary includes a brief description of the source of contamination, the period of documented contamination, and the contaminants detected during sampling.

Description of the TCA Subregistry Site

Vestal Water Supply Well #1-1 is located in Water District #1 of the Town of Vestal, Broome County, New York (Figure 2-4). The well was the primary of three wells (Figure 2-5) that supplied drinking water to approximately 9,000 residents in the western part of the town, as well as to residents in Water Districts #2, #8, and #9.

Figure 2-4.-Location of Vestal, New York.

Figure 2-5.-Vestal, New York, water districts.

In August 1978, the U.S. Environmental Protection Agency (EPA) conducted a community water supply survey of 66 systems in the state of New York. The Vestal system was included in this survey. TCA was detected in the distribution system at 1,600 parts per billion (ppb) (Table 2-1). On January 31, 1980, Well #1-2 was sampled for halogenated organic compounds; trichloroethylene and 1,1,1-trichloroethane were discovered. Because of the high level of halogenated compounds, in June 1980 the town began pumping Well #1-1 to waste in the nearby Susquehanna River. EPA determined that this action captured the contaminant plume and kept the contaminants from reaching the other two supply wells in the area. The site was placed on the National Priorities List in December 1982.

Table 2-1.—Contaminants for Well #1-1, Vestal, New York.

Contaminant	Maximum Level (ppb)*
1,1,1-Trichloroethane	1,600.0
Trichloroethylene	470.0
Benzene	27.0†
Tetrachloroethylene	6.1
Chloroform	35.0
Dibromochloroethane	7.0
1,1,2-Trichlorofluoromethane	4.0
Freon	15.0
Vinyl chloride	2.0
1,1-Dichloroethylene	8.0
1,1-Dichloroethane	81.0
1,2-Dichloroethylene	82.0
1,2-Dichloroethane	1.8
Trichloroethane	86.0
Chloroethane	7.0

*ppb - parts per billion.

†The first reading for benzene was 43 ppb. However,

it was noted that benzene coeluted with trichloroethylene;

therefore, this reading was not included.

Businesses in the area east of Well #1-1 were implicated in the contamination. These businesses were gasoline service stations, a dry cleaner, automobile sales and services, automobile body shops, and small manufacturing plants. Although the exact businesses have changed in the last 35 years, the nature of the commercial community has not. Commercial development in the area began between 1944-1955, with the current level achieved between 1968-1977.

Given the information available on groundwater flow and industrial growth in the area, a conservative date for the start of contamination in Well #1-1 is 1969. The duration of exposure, and thus the time period of interest for the Subregistry, would be from January 1, 1969, until May 31, 1980.

On-Site Activities

In the spring of 1991, data collection for the TCA Subregistry began with face-to-face interviews with potential Subregistry members from the Vestal Well #1-1 site. People who were then residing or had previously resided at the addresses served by the TCA-contaminated well were considered to be "potential" registrants.

Preparations for on-site data collection included a mailing to potential registrants. The mailing contained a question-and-answer brochure with information about ATSDR and the NER. At that time, benzene was also of concern; therefore, the mailout also contained a chapter from the ATSDR Toxicological Profile for Benzene, which summarized all that was known about the association between benzene exposure and adverse health outcomes (7). This information packet was sent to residents before a public meeting about the NER and the data collection effort. Soon after the public meeting, interviewers began contacting residents at the designated household addresses in search of eligible Subregistry members.

For each Subregistry included in the NER, the selection process uses definitions of eligibility and exposure that incorporate three key components: (1) valid information that indicates the presence of the contaminant(s) of interest in one or more of the media of interest; (2) evidence, for a given individual, of an appropriate route(s) of exposure; and (3) evidence of indicated transmission from the contaminated source to the potential registrant during the period of exposure as verified by that individual. For example, in the case of the TCA Subregistry, the well water had to have been tested and validated for the presence of TCA. Also, the well water had to have been the sole source of water for drinking, bathing, or cooking for all individuals at the site residential addresses. Finally, a registrant would had to have reported using the TCA-contaminated well water for drinking, cooking, or bathing during the exposure period.

Each eligible person or a proxy for that person was administered the TCA Subregistry baseline questionnaire, which included a set of questions about health conditions that the registrant currently had or had ever had and that had been either confirmed or treated by a health practitioner. Each time the respondent reported the presence of one of these health conditions, a set of follow-up questions was asked about the date of first treatment by a physician, current treatment, prescribed medication, and hospitalization related to the condition.

Information on the identity of past residents at an exposure address was solicited from current residents and other knowledgeable persons (such as neighbors). Any names provided were traced by a professional tracing service. Individuals located were contacted by telephone to determine if they were eligible to participate; if so, their participation was solicited and, if they agreed, the core questionnaire was administered at that time.

For the Vestal site, information was obtained for 3,665 persons (3,204 living and 461 deceased). The response rate (the percentage of those eligible who were contacted and agreed to participate) was 97%.

SECTION 3 COMPARISON OF THE TRICHLOROETHANE (TCA)

SUBREGISTRY DATA AND NATIONAL DATA

REGISTRANT DESCRIPTIVE DATA

Tables 3-1 and 3-2 contain information about the characteristics of the Trichloroethane (TCA) Subregistry members. The total TCA Subregistry has 3,665 members; 3,204 were alive at time of baseline data collection, 461 were deceased. Only those living at the time of the baseline interview are included in the analyses reported in this document. The response rate (the percentage of those eligible who were contacted and agreed to participate) was 97%. For the deceased registrants, the Agency for Toxic Substances and Disease Registry (ATSDR) is currently obtaining death certificates; when this effort is completed, pertinent information will be abstracted, a mortality file constructed, comparisons using the appropriate national norms made, and the results published.

Of the total living population (n = 3,204), 3,166 persons (98.8%) answered white to the race question; 38 persons (1.2%) answered other than white. The percentage of nonwhite people was significantly less than that for the comparison group-the 1990 National Health Interview Survey (NHIS) population. The analyses are, therefore, restricted (for both the Subregistry and NHIS files) to those answering white to race (n = 3,166). No further analyses were performed on the nonwhite groups because of the small numbers and the potential for violating the confidentiality of the respondents.

The exposure period in Vestal ended in 1980; therefore, all eligible people in the Vestal community were older than 10 years of age and, consequently, all registrants are 10 years of age or older. Table 3-1 indicates that, at baseline, 52% of the living registrants were female. Approximately 45% of the registrants were between the ages of 25 and 44 at the time of the baseline interview; 5% were between the ages of 10 and 17.

Among registrants who were 19 years of age or older, 94% had a high school diploma. Slightly more females than males had not completed high school (6% versus 5%). A higher percentage of males (40%) than females (29%) had completed college or some post-college education. About three-fourths (72%) of the male registrants who were 19 years of age or older were currently employed either full- or part-time, compared with 60% of the females.

Table 3-2 provides data on registrants' use of tobacco products. Of the registrants 18 years of age or older, 19% reported being current cigarette smokers (that is, they had smoked at least 100 cigarettes during their lifetime and smoked at time of data collection). Slightly more males than females (21% versus 18%) were current cigarette smokers. Fewer females than males (42% versus 52%) reported having ever smoked cigarettes (that is, were current smokers or had smoked at least 100 cigarettes in their lifetime). Overall, a small number of the registrants reported ever using other tobacco products: pipes (8%), cigars (7%), snuff (2%), and chewing tobacco (3%). Smoking rates for registrants are compared with national data later in this section.

Table 3-1.—Descriptive data for living registrants (white only).

Variable	Males		Females		Total
	N	(%)	N	(%)	N	(%)
Number	1,519	(48)	1,647	(52)	3,166
Age (years) £17 18-24 25-44 45-64 65-74 ³75 R*	76 142 707 343 170 79 2	(5) (9) (46) (23) (11) (5) (<1)	71 128 709 401 198 133 7	(4) (8) (43) (24) (12) (8) (1)	147 270 1,416 744 368 212 9	(5) (8) (45) (23) (12) (7) (<1)
Education (³19 years of age) Not high school graduate High school graduate Some college College graduate or more D† R*	73 429 356 568 - 1	(5) (30) (25) (40) - (<1)	94 587 426 444 2 2	(6) (38) (27) (29) (<1) (<1)	167 1,016 782 1,012 2 3	(6) (34) (26) (34) (<1) (<1)
Occupational status (³19 years of age) Currently employed Previously employed Never employed Missing	1,032 391 4 -	(72) (27) (<1) -	925 594 35 1	(60) (38) (2) (<1)	1,957 985 39 1	(66) (33) (1) (<1)
Place of residence (at time of interview) On site Off site D† R*	608 907 3 1	(40) (60) (<1) (<1)	738 907 2 -	(45) (55) (<1) -	1,346 1,814 5 1	(42) (57) (<1) (<1)
Type of interview Subject Proxy	1,408 111	(93) (7)	1,525 122	(93) (7)	2,933 233	(93) (7)

*R-Refused to answer question.

†D-Response was don't know (usually proxy reporting).

Table 3-2.—Tobacco use for registrants 18 years of age or older (white only).

Variable	Males		Females		Total
	N	(%)	N	(%)	N	(%)
Cigarettes Current smoker Ex-smoker Never smoked R*	297 447 696 1	(21) (31) (48) (<1)	285 379 905 -	(18) (24) (58) -	582 826 1,601 1	(19) (27) (53) (<1)
Pipe Current smoker Ex-smoker Never smoked Other D† R	21 223 1,194 - 2 1	(2) (16) (83) - (<1) (<1)	- - 1,567 2 - -	- - (100) (<1) - -	21 225 2,761 2 2 1	(<1) (8) (91) (<1) (<1) (<1)
Cigars Current smoker Ex-smoker Never smoked Other D R	41 176 1,221 - 2 1	(3) (12) (85) - (<1) (<1)	- - 1,568 1 - -	- - (100) (<1) - -	41 176 2,789 1 3 1	(1) (6) (93) (<1) (<1) (<1)
Snuff Current user Ex-user Never used Other D R	15 37 1,387 - 1 1	(1) (3) (96) - (<1) (<1)	- - 1,568 1 - -	- - (100) (<1) - -	15 37 2,955 1 2 1	(<1) (1) (98) (<1) (<1) (<1)
Chewing tobacco Current user Ex-user Never used Other D R	26 61 1,352 - 1 1	(2) (4) (94) - (<1) (<1)	- - 1,568 1 - -	- - (100) (<1) - -	26 62 2,920 1 1 1	(<1) (2) (97) (<1) (<1) (<1)

*R-Refused to answer question .

†D-Response was don't know (usually proxy reporting).

RATIONALE FOR THE COMPARISON OF THE TCA SUBREGISTRY DATA WITH NATIONAL DATA

This section includes comparisons of TCA Subregistry data with data from national surveys. These comparisons are consistent with NER objectives and goals, as stated in the Policies and Procedures Manual (1), which are to provide a preliminary assessment of the extent to which TCA Subregistry members might have an excess of adverse health conditions and to generate-rather than test-hypotheses about TCA exposure and health outcomes.

In addition to a comparison of the TCA Subregistry health data with national health data norms, this section includes comparisons of registrant demographic and smoking data with national data. The comparisons of demographic characteristics and smoking rates indicate the extent to which TCA Subregistry members are similar to the general population. These comparisons are important because both demographic characteristics and smoking are known to be correlated with or are probable causes of many adverse health conditions.

TCA Subregistry data were compared with data obtained from the 1990 NHIS (8). Subsets of the NHIS data were used in the comparisons with demographic, smoking, and health data components of the TCA Subregistry. The NHIS data were selected for comparison with the TCA Subregistry data for the same general reasons that the NHIS was selected for comparison with the NER. The NHIS is an appropriate comparison population because it is a subset of the residential, noninstitutionalized U.S. population, the population of interest for comparisons of the health status of the NER members. As of 1985, a stratified, multistage cluster sample design was used in the NHIS to obtain a representative sample of the target population; this information was used to create representative national norms. The NHIS, similar to the NER, consists of self-reported data that were obtained using face-to-face interviews.

Also, because of the similarity of the data collection instrument and methods used by the NHIS and the NER, the NHIS data were appropriate for the calculation of selected prevalence and incidence statistics and could be used for exploratory comparison with NER data for health outcomes. The weighting factors (9) provided by the National Center for Health Statistics (NCHS) were applied when using the data. The 1990 NHIS file used for selected comparisons in this report included approximately 155,000 respondents.

As was discussed in Section 2 of this report, members of the TCA Subregistry sample are located primarily in the Northeast United States (New York); others are located throughout other regions of the country. The influence of region on reported disease outcome rates for the TCA Subregistry, a concern when comparing the Subregistry reporting rates with the national rates reflected by the NHIS numbers, was explored. ATSDR's review of the NHIS regional rates for selected outcomes found no definitive evidence indicating that the overall health status of persons located in the Northeast region differed significantly from that of the general U.S. population. Rates for selected reported chronic conditions for the geographic regions Northeast, Midwest, South, and West (all races) are listed in an NCHS publication (8). Very few of the reported health rates were highest for the Northeast region and none were related to conditions reported in excess by the registrants. Therefore, differences between the TCA Subregistry file and the NHIS file were not expected to be and did not appear to be the result of regional differences.

METHODS OF DESCRIPTIVE VARIABLE COMPARISONS

Demographic Characteristics

The NHIS and TCA Subregistry samples were compared in terms of four demographic characteristics-sex, age, race, and education level-as well as cigarette smoking rates. Each of these variables is a potential correlate of health status.

Sex

The distribution of the male-female ratio was assessed on an age-specific basis. The proportion of males and females in each age category was based on the NHIS data and compared with the corresponding proportions in the TCA Subregistry. Each age-specific proportion in the TCA Subregistry was compared with the corresponding proportion in the NHIS by testing that the binomial proportion was equal to a specified theoretical value. No statistically significant differences were found between the two files for this variable.

Age

The descriptive comparisons of age used a 10-category measure. The regression analyses in this section involved a regrouping of age categories. An eight-category measure of age (combining the lower two and upper two groups) was used because of the sparsity of positive reports in some of the age strata. For the TCA Subregistry file, there were no registrants less than 10 years of age and the number of categories was reduced to seven.

It should be noted that because the health outcome analyses involved summarizing age- and sex-specific comparisons rather than analyzing age-adjusted summaries, whether the age distribution of the NHIS file matched the age distribution of the TCA Subregistry file was not directly relevant unless distribution differed within the age groups.

Race

Race is an established correlate of socioeconomic status (10) and health status (11). National data indicate that nonwhites have lower rates for cigarette smoking (12). For these reasons, race is a potential control variable for the comparisons of health status and smoking rates. However, as was discussed at the beginning of this section, there are too few nonwhites in the TCA Subregistry to use race as a variable; all analyses were restricted to registrants responding white to the race question.

Education Level

For education level (the highest level attained as reported by a respondent), the descriptive analyses included comparisons in which education level was measured as a seven-category ordinal variable (that is, 0 through 5 years, 6 through 8 years, 9 through 11 years, 12 years or the equivalent of a high school diploma, 13 through 15 years or some college, 16 years or the equivalent of a college degree, and 17 or more years).

For the regression analyses, education attainment was collapsed into four groups: 0 through 11 years, 12 years, 13 through 15 years, and 16 or more years. Because education is difficult to interpret as a surrogate for socioeconomic status for school-aged children, analyses involving adjustments for education were restricted to adults (19 years of age or older).

Cigarette Smoking

Rates for current and past smoking behaviors were compared across sex, age, and education attainment categories. A current smoker ("current rate") was defined as anyone who reported being a smoker at the time of the interview, and who had smoked at least 100 cigarettes in his or her lifetime. Past smoking behavior ("ever" rates) was assessed by calculating the rates for people who had ever smoked at least 100 cigarettes during their lifetime. People who had ever smoked included both current and ex-smokers. For the NHIS, only one adult per household was asked the smoking-related questions; for the NER, all adults were asked. This lack of comparability precluded using smoking as a factor in the model in the statistical analysis.

METHODS OF COMPARING HEALTH OUTCOMES

Question Comparability

TCA Subregistry and NHIS data were compared for health conditions reported by respondents. Such comparisons were preceded by an assessment of the comparability of NHIS and TCA Subregistry health condition questions. The questions about health conditions in these two surveys differed in three respects: restrictions on the source of diagnosis; the time frame of occurrence or treatment; and, in some cases, the wording of the health condition. A discussion of each potential source of variation in health condition questions follows. The NHIS health-related questions are presented in Appendix A; the TCA Subregistry health-related questions are in Appendix B.

Source of Diagnosis

TCA Subregistry questions about health conditions specified that the source of diagnosis had to have been a "physician or other medical provider". This qualification was intended to minimize self-diagnoses or the biased reporting of health problems by registrants who might have had a greater awareness of health because of their known exposure and publicity related to the exposure. The NHIS questions did not include any type of qualification concerning the source of diagnosis. Therefore, if all other factors were equal or similar, an increased reporting by NHIS respondents when compared with the registrants might be expected. The increases would be expected to be greater for health conditions often self-diagnosed (for example, arthritis, hearing impairment, and some respiratory problems).

Time Frame

As was the case for all components of the NER, the TCA Subregistry baseline questions about health conditions asked about diagnoses of or treatment for conditions from the point of birth through the date of the interview ("Has a physician or other medical provider ever told you/SUBJECT that you/he/she/ had or treated you/SUBJECT for CONDITION?"). Only one time frame was addressed: ever had (registrant's lifetime). Respondents who reported "yes" to this question were also asked whether the registrant was ever treated for the condition, when the registrant was first treated for the condition, and whether the registrant was currently being treated for the condition.

The NHIS questionnaire included questions that focused on three time frames-ever had the condition, had the condition within the last 12 months, or currently had the condition. With the exception of heart disease, only one time frame was used to create a response rate for any given health condition. The NHIS data for heart disease rate reflects a composite of responses to heart-related questions that were asked in both the "ever had" and "12-month" time frames. The NHIS questionnaire asked whether respondents had ever had the heart conditions rheumatic, congenital, or coronary heart disease; angina pectoris; myocardial infarction; or any other heart attack. The NHIS questionnaire further asked whether in the last 12 months respondents had had a damaged heart valve; tachycardia or rapid heart beat; heart murmur; or other heart trouble. In addition, for some heart-related questions, a positive response was discarded if the respondent did not answer positively to one or more other selected questionnaire items (13). A comparable response rate could not be created for the heart-related questions in the NER file; therefore, a comparison to national norms could not be made for the heart condition variable.

For the other health conditions, the time frames were standardized to make the NHIS and TCA Subregistry rates directly comparable. Table 3-3 provides a comparison of NHIS and TCA Subregistry questions in terms of the time frame for each health condition. One NHIS health condition question, the effects of a stroke, was asked and the rate calculated in the context "have you ever had". The questions and time frames for the Subregistry and NHIS matched on this condition.

Eleven of the NHIS questions were asked in the time frame "within the last 12 months". For hypertension, the NHIS 12-month response rate was calculated using the "ever had" positive responses; however, the positive response was retained in the file only if the respondent also answered positively to one or more of nine other selected questionnaire items (13). This additional restriction might have reduced the NHIS response rate for this condition. For these 12 health conditions (see Table 3-3), the TCA Subregistry time frames for comparison were adjusted. In the TCA Subregistry, a health condition was defined as occurring "within the last 12 months" if (1) the reported date of first treatment was within the 12 months preceding the interview or (2) the subject was receiving treatment at the time of the baseline interview. This adjustment could have resulted, however, in an underestimation of these 12 conditions for TCA Subregistry data for the following reason. A year or more before the baseline interview, a registrant might have been told that he or she had (or was treated for) one of these 12 health conditions, but was not being treated at the time of the interview. If so, such a registrant would not have been included in the rates for these 12 health conditions.

Three health conditions in the NHIS questionnaire were asked about in terms of the time frame "do you now have". These conditions were speech impairment, hearing impairment, and mental

Table 3-3.—Comparison of time frames for health condition questions.

(TCA Subregistry Conversion from "ever had")	NHIS* Time Frame for Condition
(TCA Subregistry Conversion from "ever had")	"ever had"	"in the last 12 months"	"now have"
"Ever had" (same)	Stroke
"In the last 12 months" ("Ever had" and "currently have" or date of first treatment within last 12 months)		Cancer, rash, anemia, kidney disease, urinary tract disorders, ulcer, liver problems, asthma, respiratory problems and allergies, diabetes, arthritis, hypertension
"Now have" ("Ever had" and "currently have")			Speech impairment, hearing impairment, mental retardation

*National Health Interview Survey

retardation. The time frame for the comparable TCA Subregistry health conditions was adjusted by counting only registrants who reported that they were "currently receiving treatment" for one of these three conditions. Again, if all other factors were equal or the same, an increased reporting by the NHIS respondents when compared with the TCA registrants would have been expected.

Health Conditions

TCA Subregistry and NHIS questions were also compared in terms of the phrasing of health conditions. As Table 3-4 indicates, some health conditions matched exactly, others did not. An ATSDR panel of scientists and physicians determined matches for the TCA Subregistry health conditions and specific NHIS conditions (ICD-9 codes [14], or NHIS condition recodes [8]).

The nine health conditions in Class A of Table 3-4 either matched exactly or the TCA Subregistry version was inclusive of the NHIS version. That is, the NHIS wording of the health condition and the NHIS classification of the condition in the recodes were the same as or paralleled closely the corresponding TCA Subregistry item. As Table 3-4 indicates, nine health conditions were in Class A. Class B included seven health conditions that did not match as closely, but were considered to be sufficiently similar for the purposes of the NHIS and TCA Subregistry comparisons.

In addition to the heart disease outcome, for six other health conditions on the TCA Subregistry questionnaire there were no parallel items in the NHIS questionnaire. These conditions pertained to symptoms including "frequent periods of fatigue or tiredness"; "frequent periods of nausea"; "seizures, tremors, spells, or epilepsy"; "weakness or paralysis or numbness in the arms or legs"; "frequent periods of anxiety, nervousness, or depression"; and "frequent or severe headaches."

Unlike some of the national health surveys (15), environmental studies commonly ask about symptoms as well as health outcomes. Data on these symptoms, while not directly comparable with

Table 3-4.—Comparison of the Trichloroethane (TCA) Subregistry and National Health Interview Survey (NHIS) health questions.

Q#*	Wording in TCA Survey	National Health Interview Survey Definition	NHIS Chronic Recodes†	ICD-9§
Class A¶
6	High blood pressure (hypertension)	Essential hypertension Hypertensive heart disease Hypertensive renal disease Hypertensive renal and heart disease	C508	401-05
8	Kidney disease	Kidney stones Kidney infections Other kidney trouble	C409-11	592 590 581-3 593
10	The effects of stroke	Cerebrovascular disease	C509	430-38
14	Liver problems	Liver disease, including cirrhosis	C302	571-2 573.0, .3-.9
15	Asthma, emphysema, or chronic bronchitis	Same	C601-2 609	490-1 492 493
16	Other respiratory allergies or problems such as hay fever	Hay fever Allergic rhinitis without asthma	C603	477
17	Diabetes	Same	C403	250
22	Hearing impairment	Deaf - both ears Other hearing impairment	C203-4	X05 X06-9
25	Mental retardation	Same	C208	X19
Class B**
3	Cancer	Some cancers queried directly; other ascertained indirectly		140-208
5	Skin rashes, eczema, or other skin allergies	Psoriasis Dermatitis Dry (itching) skin	C112-4	696 690-94 698.9
7	Anemia or other blood disorders	Anemia of any kind	C404	280-85
9	Urinary tract disorders, including prostate trouble	Disorders of the bladder (other than bladder infections) Diseases of prostate	C413-14	594.1 596 600-602 (except 601.4)

Table 3-4.—Continued.

Q#*	Wording in TCA Survey	National Health Interview Survey Definition	NHIS Chronic Recodes†	ICD-9§
Class B**
13	Ulcers, gallbladder trouble, or stomach or intestinal problems	Gallbladder stones Gastric, duodenal, or peptic ulcer Abdominal hernia Gastritis and duodenitis Disease of esophagus Other functional disorders of stomach or digestive system (not indigestion) Enteritis and colitis Spastic colon Diverticula of intestines Other stomach and intestinal disorders (not constipation)	C301 C303 C303-8 C310-3 C315	574 530-7 550-3 555 556 558 560-562 564.1 569 787
18	Arthritis, rheumatism, or other joint disorders	Arthritis Rheumatism Gout Sciatica (and lumbago) Intervertebral disc disorders Bone spur and tendinitis Disorders of bone or cartilage Bursitis	C101-7 C109	711.0, .9 712.8-.9 714-16 720.0 721 729.0 724,.2-.3 722, 726 727.0, .2-.9 730.0-.3, .9 731.0, .2 732-3
20	Speech impairment	Stammering and stuttering Other speech impairment	C205-6	X10 X11
No Match
19	Rheumatic fever, heart disease, or other heart problems	Rheumatic fever Ischemic heart disease Heart rhythm disorders Congenital heart disease Other select heart diseases	C501-7	390 392-9 410-4 427.0-.6, .8-.9 785.0-.2 745-6 415-7 420.9 421.0, .9 422.9 423-4 425.0-.5, .9 426, 428 429.0-.6, .8-.9

*Question in TCA Subregistry questionnaire.

†Chronic Recodes, NHIS, Public Use Data Tape Documentation (9).

§ICD-9 is the International Classification of Diseases, 9th Revision, World Health Organization (WHO) (14).

¶Class A indicates questions match exactly or closely.

**Class B indicates questions are similar.

NHIS data, are important in assessing the impact of the environment on health and will be useful for comparisons with past and current epidemiologic environmental studies, as well as for future longitudinal studies. Symptom response rates were compared in intrafile exposure groups.

The TCA Subregistry questionnaire was used to record information on all types of cancer via an open-ended question. The NHIS questionnaire, however, was used to obtain direct information on only some types of cancers, such as skin, stomach, intestinal, colon, rectal, lung, breast, and prostate cancer; information on other cancers was obtained indirectly from the respondent through information on hospital stays, doctor visits, and restricted activity. The NHIS question is worded, "In the last 12 months, did anyone in the family have ... cancer?" The time frame restriction and the possible restriction on the types of cancers reported makes this comparison with the TCA registrant data questionable and the interpretation tenuous.

Statistical Analysis of Health Data

The statistical analyses performed treated the NHIS population as a standard population and applied the age- and sex-specific prevalence or period prevalence rates obtained from the NHIS data to the corresponding age- and sex-specific denominators in the TCA Subregistry. The observed age- and sex-specific numerators for the TCA Subregistry were compared with the expected numerators based on the NHIS rates.

This one-sample approach ignored sampling variability in the NHIS data because of the large size of the NHIS database relative to the TCA Subregistry data file. Treating the NHIS versus TCA Subregistry comparison as a two-sample problem might have resulted in a dramatic underestimation of the variability associated with the TCA Subregistry data if any pooled variance estimates were used.

For the NHIS data set, methods established by the NCHS for estimating rates were used (9). That is, the complex sampling design of the NHIS was accounted for when computing health condition rates. Numerators for a given sex-age stratum were computed by summing the cross product of the "condition weight" and the "basic final weight" for people who were asked the appropriate "condition list" and responded positively for the condition of interest. Because the presence of cancer could be ascertained from several sources within the NHIS questionnaire, stratum-specific numerators were computed by summing the "basic final weight" for all people with cancer. Denominators were obtained from the NHIS "person" records by summing the "basic final weight" for all people in a particular stratum.

All health outcomes were analyzed in the following manner. Taking the NHIS as the standard population, weighted age- and sex-specific prevalence or period prevalence rates were constructed. These "standard" rates were applied to the corresponding TCA Subregistry denominators to obtain expected counts in each age and sex stratum. Standardized relative risks-the ratios of observed-to-expected age- and sex-specific counts-were modeled using Poisson regression in the Generalized Linear Interactive Modeling program (13). The Poisson regression approach is described in Breslow and Day (16). In the Poisson regression analysis, the null model was specified as log(observed) = log(expected) + grand mean. By adding terms for the age and sex effects to the null model and computing likelihood ratio statistics (by subtracting deviances of nested models), the homogeneity of the stratum-specific ratios was assessed; that is, it was determined whether the overall relative risk or the age-, sex-, or age, sex-specific ratios must be presented. Confidence intervals for the observed-to-expected ratios (collapsed at the appropriate level) were generated using exact methods; the observed counts were assumed to follow the Poisson distribution.

To take education into account, each analysis was repeated for adults (aged 19 years or older) using age-, sex-, and education-specific rates. Problems arose with the inclusion of all factors in the model; numerous cell sizes were very small or zero. In some cases, the model did not converge; in others, the zero cells made the results suspect and interpretation tenuous. Therefore, for most outcomes, the inconclusive models were not considered valid and the results are not reported here.

Statistical Analysis of Cancer Data

For the cancer health outcome, cancer incidence data from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) (17) program and NHIS data were used to generate expected numbers of events for comparison with the TCA Subregistry observed numbers. The SEER data can be used to generate rates that are strictly a measure of incidence (new cases in a particular year) and include all reported cancers. To make the TCA Subregistry database comparable and to generate similar rates for comparison, the year of first treatment was used as the date of onset or the year of "new case."

Both the TCA Subregistry and the SEER program are complete enumerations for data on an entire population. In the first case, the TCA Subregistry is the population of citizens with documented environmental exposure to TCA; in the latter case, SEER is the population of all (invasive) cancer cases in the 10 geographic areas constituting the SEER study area. Neither involves probability-based sampling. Thus, comparisons that were made are descriptive rather than inferential in nature and lend themselves directly to deterministic decision making. The sparsity of the data for specific cancers in each age category, particularly the younger age groups, precludes statistical comparisons for specific cancers or even for specific age groups. The rates for total cancers by sex and year reported were calculated. No statistical comparisons were made; the observed and expected values are provided.

RESULTS OF DESCRIPTIVE COMPARISONS

This section provides a discussion of the comparability of descriptive data for the TCA Subregistry file and the NHIS file. The results of this section were used to plan the subsequent analyses of the health outcome data. That is, the results were used to determine what variables were appropriate to include as covariates in modeling the health outcome comparisons.

Demographics

Race

In the NHIS sample, 18.0% of the participants were nonwhite. In the TCA Subregistry sample, 1.2% (n = 38) of the registrants were nonwhite. Given this difference in the proportion of nonwhites and the diversity of races reported among those reporting race as nonwhite in the TCA Subregistry, all nonwhite participants from the NHIS and TCA Subregistry data were excluded from the analysis reported in this and subsequent sections of the report. The statistical analyses were performed on the 3,166 living, TCA-exposed, white registrants. The small number and diversity of the nonwhite subpopulation (and also the potential for violating confidentiality) precluded conducting any analyses of this subpopulation.

Age

Table 3-5 shows the percentage of people in each age group. As expected, the percentages in some age groups were affected by the timing of data collection relative to the exposure period. For example, the percentage of people in the 17 years of age or younger group was less for the TCA Subregistry than for the NHIS sample; there were no registrants in the group 0 through 9 years of age. This result is consistent with ATSDR's knowledge about the demographics of the TCA site. The exposure period ended more than 10 years before data collection and precluded the eligibility of young children (that is, children born after the exposure period ended). Table 3-6 provides a comparison of the age-specific sex distribution of males in the TCA Subregistry and NHIS samples (the proportion of females can be calculated as one minus the proportion of males), and the corresponding p values for the differences between the TCA Subregistry and NHIS proportions. Table 3-6 indicates that the proportion of males (and females) did not differ statistically significantly between the TCA Subregistry and NHIS samples for any of the age groups examined. Nine of the registrants refused to answer this question (2 males and 7 females); therefore, the sample size is decreased by 9 when age is a factor in the statistical analyses.

Table 3-5.—Comparison of National Health Interview Survey (NHIS) and Trichloroethane (TCA)

Subregistry population age groups (white only).

Age Group (Years)	NHIS	TCA Subregistry
Age Group (Years)	% of Total	Number	% of Total
All	100.0	3157*	100.0
£4	7.2	0	0.0
5 - 9	7.1	0	0.0
10 - 17	10.4	147	4.7
18 - 24	9.8	270	8.6
25 - 34	17.2	645	20.4
35 - 44	15.5	771	24.4
45 - 54	10.6	409	12.9
55 - 64	9.0	335	10.6
65 - 74	8.0	368	11.7
³75	5.2	212	6.7

*9 refused to answer this question.

Table 3-6.—Comparison by age group of percentage of white males for the Trichloroethane (TCA)

Subregistry and National Health Interview Survey (NHIS) populations.

Age (Years)	TCA Subregistry	NHIS*	p value†
0-4	---	0.52	---
5-9	---	0.52	---
10-17	0.52	0.51	p = 0.92
18-24	0.53	0.49	p = 0.24
25-34	0.50	0.50	p=0.94
35-44	0.50	0.50	p=0.84
45-54	0.47	0.49	p=0.51
55-64	0.44	0.48	p=0.26
65-74	0.46	0.45	p=0.58
³75	0.37	0.37	p=0.94

*Cell values are proportions of total males in age group.

†The p values are two-sided.

Education Level

The summary statistics for education attainment are provided in Table 3-7. For both males and females, the overall difference in the age-specific education level of people aged 19 years or older was statistically significant at the p £0.05 level. These overall differences were attributable to a smaller proportion of TCA Subregistry members in the 0 through 11 years of education category (especially among younger adults), and a greater proportion who had at least a college degree compared with the NHIS sample. These findings indicate that the TCA Subregistry participants had attained a higher level of education than the NHIS population. This difference could modify the comparison of health outcome rates in that the TCA Subregistry population might be expected to have fewer outcomes related to lower socioeconomic factors. Therefore, not including the factor in the statistical model would not have increased the number of positive findings.

Cigarette Smoking

Table 3-8 provides comparative information on current and past cigarette use for people 18 years of age or older. Smoking rates are indicated for TCA Subregistry members, all NHIS respondents, and the NHIS Northeast regional subpopulation. As can be seen, the smoking rates-

Table 3-7.—Education level attained for Trichloroethane (TCA) Subregistry and National Health

Interview Survey (NHIS) populations (white, 19 years of age or older).

Age (Years)	NHIS Education Level (years)				TCA Subregistry Education Level (years)
Age (Years)	0-11	12	13-15	16+	0-11	12	13-15	16+
Males and Females
All	19.5	38.9	20.8	20.8	5.6	34.1	26.3	34.0
19 - 25	15.1	40.9	32.3	11.8	3.4	29.4	45.4	21.8
26 - 45	11.8	39.0	22.6	26.7	1.3	27.0	28.7	43.0
46 - 64	21.1	40.6	17.1	21.1	4.4	41.3	20.4	33.8
³65	39.8	34.7	12.7	12.7	18.9	45.8	17.3	18.0
Males
All	19.5	36.6	20.3	23.7	5.1	30.1	25.0	39.8
19 - 25	16.3	42.2	30.8	10.7	3.9	28.6	40.9	26.6
26 - 45	12.3	37.4	21.8	28.5	1.4	24.9	26.8	46.9
46 - 64	21.6	35.2	16.8	26.5	3.2	33.7	19.7	43.5
³65	40.2	31.3	12.1	16.5	18.9	41.4	16.5	23.3
Females
All	19.5	41.0	21.2	18.2	6.1	37.8	27.5	28.6
19 - 25	13.8	39.6	33.7	12.9	2.9	30.2	50.4	16.5
26 - 45	11.3	40.5	23.3	24.9	1.1	29.1	30.6	39.2
46 - 64	20.7	45.7	17.5	16.1	5.5	47.9	21.1	25.5
³65	39.5	37.2	13.2	10.1	18.9	49.1	18.0	14.0

Note: Education data were missing for five registrants.

Table 3-8.—Reported rates of cigarette smoking for the Trichloroethane (TCA) Subregistry and

National Health Interview Survey (NHIS) populations (white only).

Age Group	TCA Subregistry Rates (%)			NHIS Rates (%)
	TCA Subregistry Rates (%)			All		Northeast Region
	N	Current*	Ever†	Current	Ever	Current	Ever
³18 years All Males Females	3,009 1,440 1,569	19.3 20.6 18.2	46.8 51.7 42.3	25.3 27.6 23.3	51.7 59.8 44.2	23.6 25.9 21.8	52.0 58.8 45.9
18-25 years All Males Females	321 168 153	17.1 16.7 17.6	25.2 23.2 27.5	26.8 27.8 25.8	38.2 38.6 37.9	27.6 27.6 27.6	39.6 37.8 41.2
26-34 years All Males Females	594 296 298	24.1 24.7 23.5	41.1 38.2 44.0	29.8 31.0 28.6	48.8 51.4 46.3	28.7 30.2 27.3	50.6 52.6 48.7
³35 years All Males Females	2,094 976 1,118	18.3 20.1 16.8	51.7 60.7 43.9	23.5 26.3 21.0	55.7 67.9 45.0	21.3 24.0 19.0	55.0 65.6 46.1

*Reported currently smoking.

†Reported having smoked more than 100 cigarettes in lifetime.

both current and ever rates-were lower for both the male and female registrants when compared with the NHIS national rates. The registrants' rates were also lower than the NHIS Northeast regional rate.

Table 3-9 shows the rates of reported current cigarette smoking by education attainment levels. As with the NHIS rates, TCA Subregistry rates (current smokers) are correlated with highest level of education attainment. It should be noted that within each age-education level, with few exceptions, the NHIS rates were either equal to or exceeded the TCA Subregistry rates. The exception was the 18 through 34 years of age group with a high school diploma-the registrant rate was slightly higher than the NHIS population rate. Since the smoking rates were lower, with minor deviations, for the TCA Subregistry population, the omission of smoking as a covariate had the potential to moderate any excess reporting of health outcomes by registrants.

Table 3-9.—Current smoking rates (percentage) by education attainment for the Trichloroethane

(TCA) Subregistry and National Health Interview Survey (NHIS) populations (white only).

Age Group	Education Level Attained
	No High School Diploma		High School Graduate		Some College		College Graduate*
	TCA Subregistry	NHIS	TCA Subregistry	NHIS	TCA Subregistry	NHIS	TCA Subregistry	NHIS
³18 years All Males Females	16.2 17.4 15.2	31.8 36.1 28.1	26.4 29.5 24.1	29.7 33.1 27.0	22.0 25.1 19.5	23.0 25.6 20.7	10.9 11.6 9.9	13.5 14.3 12.4
18-25 years All Males Females	17.6 15.8 20.0	42.2 44.1 40.2	31.1 31.1 31.1	30.4 30.1 30.6	12.8 12.7 12.9	18.7 19.7 17.8	6.3 7.3 4.3	7.8 7.9 7.8
26-34 years All Males Females	37.5 20.0 66.7	48.9 49.1 48.7	40.7 42.0 39.5	38.3 41.1 35.6	28.2 30.9 25.8	25.0 26.1 24.0	9.4 10.1 8.7	11.0 11.0 11.0
³35 years All Males Females	14.8 17.7 12.6	26.7 31.6 22.7	22.7 26.0 20.5	26.5 30.6 23.6	22.4 26.5 19.0	23.8 27.4 20.4	11.8 12.6 10.8	15.0 16.0 13.6

*May be additional years post-bachelor degree.

RESULTS OF HEALTH OUTCOME COMPARISONS

A summary of the reporting rates for health conditions for the TCA Subregistry file and the NHIS file is provided in Table 3-10 (percentages by age groups are shown in Appendix C for the TCA Subregistry). At the time of the interview, if a registrant reported having been told that he or she had cancer or had been treated for cancer by a health care provider, additional questions were

Table 3-10.—Comparison of the Trichloroethane (TCA) Subregistry and National Health Interview

Survey (NHIS) participants reporting health condition (white only).

Condition	TCA Subregistry		NHIS
Condition	Male (%)*	Female (%)	Male (%)	Female (%)
Hypertension†	11.0	12.8	10.2	11.9
Stroke§	2.0	2.1	1.3	1.1
Diabetes†	3.0	3.3	2.2	2.6
Kidney disease†	0.7	1.1	1.1	1.6
Urinary tract disorders†	3.8	4.1	1.6	1.3
Skin rashes†	6.3	7.8	5.4	8.2
Anemia and other blood disorders†	0.7	5.3	0.4	2.3
Asthma, emphysema†	4.2	6.1	8.5	10.5
Respiratory allergies†	6.2	10.6	9.0	10.3
Stomach problems, ulcers†	4.1	6.4	6.3	8.9
Liver problems†	0.2	0.6	0.3	0.3
Arthritis†	5.1	10.6	14.0	19.5
Mental retardation¶	0.1	0.2	0.7	0.5
Speech impairment¶	0.1	0.2	1.1	0.7
Hearing impairment¶	2.8	2.6	12.6	8.3

*Percentage of subpopulation (white only) reporting positively.

†Indicates time frame is "last 12 months."

§Indicates time frame is "ever had."

¶Indicates time frame is "now have."

asked about the type(s) of cancer. Although multiple types of cancers might have been reported, only one primary type of cancer was assigned to each registrant. The summary table entries and data analysis are based on the reported primary cancers; the number of secondary cancers is also included in the tables. A summary of the TCA Subregistry population's reporting rates for specific cancers (total population and by sex) is shown in Table 3-11 for the "ever had" time frame and Table 3-12 for the "within the last 12 months" time frame. The "within the last 12 months" time frame

Table 3-11.—Summary of the Trichloroethane (TCA) Subregistry reporting rates for cancer (primary

site), "ever had" time frame (white only).

Cancer >	Sex						Total
	Male			Female			Total
	N*		%	N*		%	N*		%
None	1,438		94.7	1,535		93.2	2,973		93.9
Lip, oral, pharynx	1		<0.1	0		0.0	1		<0.1
Digestive system	9		0.6	13	(1)	0.8	22	(1)	0.7
Respiratory system	4	(1)	0.3	3	(1)	0.2	7	(2)	0.2
Skin	33	(4)	2.2	33	(2)	2.0	66	(6)	2.1
Breast	0		0.0	19	(3)	1.2	19	(3)	0.6
Genital organs	18	(3)	1.2	34	(2)	2.1	52	(5)	1.6
Urinary organs	3		0.2	2		0.1	5		0.2
Lymphatic tissues	3		0.2	2		0.1	5		0.2
Leukemia	2	(1)	0.1	1		< 0.1	3	(1)	0.1
Other	8	(1)	0.5	5	(1)	0.3	13	(2)	0.4
Total cancers	81	(10)	5.3	112	(10)	6.8	193	(20)	6.1

*The numbers in parentheses are number of reported secondary sites.

(Table 3-12) is comparable to the NHIS time frame; the 12-month rates were used in the statistical comparisons of the data files.

Controlling for Age and Sex

Table 3-13 provides a summary of the results of the NHIS and TCA Subregistry file comparison using Poisson regression analysis. For each health outcome, the table indicates the likelihood ratio statistics with the associated degrees of freedom and p values for the effects of age (categorized into eight levels) and sex, based on a model containing age and sex. The residual deviance and the associated degrees of freedom are also given as a global lack-of-fit measure for this model, which specifies multiplicative effects of the age (i) and sex (j) ratios Oij/Eij. For each outcome, the age- and sex-specific numerators Oij were obtained from the TCA Subregistry data,

Table 3-12.—Summary of the Trichloroethane (TCA) Subregistry reporting rates for cancer (primary site), "within last 12 months" time frame (white only).

Cancer	Sex					Total
	Male			Female		Total
	N*		%	N*	%	N*		%
None	1,483		97.6	1,608	97.6	3,091		97.6
Lip, oral, pharynx	0		0.0	0	0.0	0		0.0
Digestive system	3		0.2	3 (1)	0.2	6	(1)	0.2
Respiratory system	3		0.2	3 (1)	0.2	6	(1)	0.2
Skin	15	(1)	1.0	12	0.7	27	(1)	0.9
Breast	0		0.0	7 (1)	0.4	7	(1)	0.2
Genital organs	6		0.4	10	0.6	16		0.5
Urinary organs	2		0.1	1	<0.1	3		0.1
Lymphatic tissues	1		<0.1	1	<0.1	2		0.1
Leukemia	2	(1)	0.1	0	0.0	2	(1)	0.1
Other	4		0.3	2	0.1	6		0.2
Total cancers	36	(2)	2.4	39 (3)	2.4	75	(5)	2.4

*The numbers in parentheses are the numbers of reported secondary sites.

while the expected numerators Eij were based on the suitably person-weighted age- and sex-specific ratios from the NHIS data. For the purpose of detecting structure in these age- and sex-specific ratios, a significance level of 0.05 was adopted.

As is shown in Table 3-13, the model was adequate and neither age nor sex was a statistically significant predictor in the models for the health outcomes speech impairment; kidney disease; skin rashes; mental retardation; cancers; and asthma, emphysema, or chronic bronchitis (p > 0.10). For the health outcome respiratory allergies or problems such as hay fever, the effects of sex were statistically significant; the effects of age were not. For the outcomes hypertension, hearing impairment, and stomach problems, statistically significant variations in the ratios were seen as a function of age but not of sex. In addition to age effects, statistically significant sex effects were seen for urinary tract disorders and arthritis, rheumatism, or other joint disorders. The multiplicative model for anemia exhibited statistically significant lack of fit, suggesting that the effects of age

Table 3-13.—Summary of Poisson regression modeling.

Condition	Age/Sex*		Sex/Age*		Residual Deviance (p value)	df
Condition	LR Stat† (p value)	df§	LR Stat (p value)	df	Residual Deviance (p value)	df
Skin rashes	6.53 (p = 0.37)	6	2.07 (p = 0.15)	1	9.66 (p = 0.14)	6
Arthritis	13.95 (p = 0.03)	6	9.43 (p £ 0.01)	1	9.68 (p = 0.14)	6
Mental retardation¶	5.62 (p = 0.47)	6	1.83 (p = 0.18)	1	1.95 (p = 0.92)	6
Speech impairment	3.82 (p = 0.70)	6	1.56 (p =0.21)	1	4.91 (p = 0.56)	6
Hearing impairment	19.61 (p £ 0.01)	6	1.99 (p = 0.16)	1	5.79 (p = 0.45)	6
Liver problems¶	---	---	---	--	---	-
Stomach problems, ulcers**	13.41 (p = 0.04)	6	0.89 (p = 0.34)	1	3.48 (p = 0.75)	6
Anemia and blood disorders**	5.93 (p = 0.43)	6	0.01 (p = 0.93)	1	39.14 (p£ 0.01)	6
Diabetes**	---	---	---	--	---	---
Kidney disease	4.03 (p = 0.67)	6	0.12 (p = 0.73)	1	7.98 (p = 0.24)	6
Urinary tract disorders**	29.32 (p £ 0.01)	6	4.45 (p = 0.03)	1	6.26 (p = 0.39)	6
Hypertension**	27.66 (p £ 0.01)	6	0.37 (p = 0.54)	1	5.85 (p = 0.44)	6
Stroke**	---	---	---	--	---	---
Respiratory allergies	6.92 (p = 0.33)	6	10.08 (p £ 0.01)	1	8.57 (p =0.20)	6
Cancer	10.77 (p = 0.10)	6	0.33 (p = 0.56)	1	3.52 (p = 0.74)	6
Asthma, emphysema	7.56 (p = 0.27)	6	0.19 (p = 0.66)	1	4.34 (p = 0.63)	6

*Indicates order of inclusion in the model.

†LR Stat - Likelihood Ratio Statistic.

§df - degrees of freedom.

¶Indicates no model constructed.

**Indicates some infinite estimates obtained.

depended on sex for these outcomes. Because of the sparseness of the data, an adequate model could not be constructed for the liver outcome; models also could not be constructed for the outcomes diabetes and stroke.

Table 3-14 summarizes the observed and expected numbers and their ratios for each of the health conditions considered separately. The results for age- and sex-specific analyses are provided in Appendix D. For the descriptive analysis, a ratio of 0/0 was considered to be undefined. For models in which the sex effect was not statistically significant, summary estimates based on the

Table 3-14.—Summary of observed and expected health outcomes using multivariate models.

Condition	Observed	Expected	Risk Ratio	99% CI*
Arthritis, rheumatism, and other joint disorders	253	693.41	0.37	0.31, 0.43
Speech impairment	4	21.62	0.19	0.03, 0.58
Hearing impairment	86	412.18	0.21	0.16, 0.27
Kidney disease	29	50.38	0.58	0.34, 0.91
Urinary tract disorders	125	58.29	2.15	1.68, 2.69
Hypertension	378	458.56	0.82	0.72, 0.94
Respiratory allergies	269	341.82	0.79	0.67, 0.92
Asthma, emphysema	164	296.57	0.55	0.45, 0.67
Anemia and other blood disorders	99	48.04	2.06	1.57, 2.66
Diabetes	101	99.57	1.01	0.77, 1.31
Liver	13	10.76	1.21	0.52, 2.37
Mental retardation	5	14.71	0.34	0.07, 0.96
Skin rashes	224	242.83	0.92	0.77, 1.09
Stroke	66	50.91	1.30	0.92, 1.77
Stomach problems, ulcers	168	302.94	0.56	0.45, 0.68
Cancer	75	59.79	1.25	0.91, 1.68

*CI - 99% confidence interval for risk ratio.

age-specific ratios for the combined sexes were given. For outcomes in which, at most, one effect was statistically significant, summary estimated ratios based on the simplest model that did not exhibit lack of fit were given together with 99% Wald confidence intervals (CIs). Reliable estimates of the standard errors of estimated risk ratios of zero could not be obtained, so no confidence intervals are presented whenever this occurred. For outcomes for which a simple model was not adequate, 99% CIs are presented for the individual age- and sex-specific ratios. For the purpose of identifying statistically significant excesses, a significance level of 0.01 was used. For some outcomes the models were mathematically adequate; however, because the data were very sparse, interpretation of the results should be carried out with caution.

A summary of the appropriate model for each health condition and the results using that model are summarized in Table 3-15. Interpretation of the results must be carried out judiciously; other factors-such as age at time of exposure, time since exposure, and sample size within age groups-must be taken into consideration. A discussion of the statistical results for each health outcome follows. The time frame "ever had" was used for reporting rates for the health outcomes hypertension, stroke, and heart condition; the time frame "now have" for the health outcomes mental retardation, speech impairment, and hearing impairment; and the time frame "had in the last 12 months" for the remainder of the health outcomes.

For the results of the statistical analyses presented in the following text, the age- and sex-specific risk ratios are defined as the number reported by the TCA Subregistry registrants (observed, O) divided by the number expected based on the number reported by the NHIS participants (expected, E). When the expected value (the number reported by the NHIS participants) was zero, the risk ratio (O/E) was not estimated. The following discussion is based on the statistical results summaries found in Tables 3-15 and 3-16 and Appendix D (details of statistical analyses).

Anemia or Other Blood Disorders

The results for the TCA Subregistry and NHIS comparisons of reporting rates for anemia and other blood disorders are detailed in Appendix D-1. For the outcome anemia, the time frame of reporting was "within last 12 months." No summary estimates are presented because there was not a summary model; the estimated effect of sex depended on the age group. Excess reporting by registrants was found for most age- and sex-groups; that is, the O/E ratios were greater than 1. However, only excesses seen in female age groups 25 through 34 years (24 observed versus 11.8 expected; O/E = 2.04, 99% CI =1.12,3.37) and 65 years of age or older (2.5 observed versus 9.6 expected; O/E = 2.59, 99% = 1.45, 4.25) were statistically significantly increased (p £.01).

Arthritis, Rheumatism, and Other Joint Disorders

The age-, sex-specific estimates (sex and age were statistically significant factors) for the outcomes related to arthritis, rheumatism, and other joint disorders are shown in Appendix D-2. The reporting time frame for this outcome was "within the last 12 months". The sex-specific risk ratios were greater than 1 for both males and females in the age group 10 through 17 years of age. The increase in TCA Subregistry reporting was statistically significantly increased for the males (4 observed versus 0.5 expected; O/E = 8.01, 99% CI = 1.35, 25.21). The sex-specific risk ratios were less than 1 for both males and females in the age group 18 years of age or older; the O/E ratio was less than 0.50 and statistically significantly less than 1 for all age groups over 35 years of age. This indicated that the reporting rates were generally lower in the TCA Subregistry data than expected based on the NHIS reporting rates-a phenomenon expected given the health care provider confirmation restriction placed on registrant responders.

Table 3-15.—Summary of results of statistical analyses comparing the Trichloroethane (TCA)

Subregistry and National Health Interview Survey (NHIS) rates.

1. Grand Mean: No Structure with Age or Sex.

Structure	Condition
Structure	Speech Impairment	Asthma, Emphysema	Kidney Disease	Mental Retardation
Overall Summary	R	R	R	R

2. Sex Effect Only

Sex	Condition
Sex	Respiratory Allergies
Males	R
Females
R	R

3. Age Effect Only

Age (Years)	Condition
Age (Years)	Hypertension	Stomach Problems, Ulcers	Hearing Impairment
0-9	---	---	---
10-17
18-24			R
25-34	R		R
35-44	R	R	R
45-54		R	R
55-64		R	R
³65		R	R

Age (Years)	Condition
Age (Years)	Urinary Tract Disorders	Anemia	Arthritis, Rheumatism, and Other Joint Disorders	Stroke	Liver	Diabetes
0 - 9	---	---	---	---	---	---
10-17	---		X
18-24		---			---
25-34			R
35-44	X		R
45-54		---	R
55-64			R
³65			R

(b) Females

Age (Years)	Condition
Age (Years)	Urinary Tract Disorders	Anemia	Arthritis	Stroke	Liver	Diabetes
0 - 9	---	---	---	---	---	---
10-17	X					---
18-24				---
25-34	X	X		---
35-44			R
45-54			R
55-64			R
³ 65	X	X	R

X = Statistically significant differences, Subregistry rate higher.

R = Statistically significant differences, NHIS rate higher.

--- = No cases reported.

Table 3-16.—Summary results of statistical analyses comparing the Trichloroethane (TCA)

Subregistry reporting rates to National Health Interview Survey (NHIS) reporting rates.

Condition	Age Groups (Years)
	0-9		10-17		18-24		25-34		35-44		45-54		55-64		³ 65		All		All
	M	F	M	F	M	F	M	F	M	F	M	F	M	F	M	F	M	F	All
Anemia and other blood disorders	-	-			-			X			-					X
Arthritis, rheumatism, and other joint disorders	-	-	X				R		R	R	R	R	R	R	R	R
Asthma, emphysema	-	-																	R
Hearing impairment	-	-			R	R	R	R	R	R	R	R	R	R	R	R
Hypertension	-	-					R	R	R	R
Kidney disease	-	-																	R
Respiratory allergies	-	-															R
Speech impairment	-	-																	R
Ulcers or other stomach problems	-	-							R	R	R	R	R	R	R	R
Urinary tract disorders	-	-	-	X				X	X							X
Mental retardation	-	-																	R

X = Statistically significant differences, Subregistry rate higher.

R = Statistically significant differences, NHIS rate higher.

─ = Insufficient data.

Asthma, Emphysema, or Chronic Bronchitis

The results for the outcomes related to asthma, emphysema, or chronic bronchitis showed that the effects of age and sex were not statistically significant factors (see Appendix D-3 for details). The reporting time frame was "within the last 12 months". The estimated overall risk ratio of 0.55, (164 observed, 296.6 expected, 99% CI = 0.45, 0.67) was statistically significantly less than 1, indicating a statistically significantly decreased reporting in this group relative to the NHIS population. None of the age-,sex-specific O/E ratios exceeded 1. This result was consistent with what was expected given the impact of the health care provider restriction on the registrants; these conditions are often self-diagnosed.

Diabetes

For diabetes, summary model estimates are not presented because the estimated effect of age depended upon the sex group. Age- and sex-specific risk ratio estimates are presented in Appendix D-4. The reporting time frame was "within the last 12 months". With few exceptions, the age-,sex-specific risk ratios were less than 1.

Hearing Impairment

For hearing impairment, there was statistically significant variability in the age-specific O/E for males and females (see Appendix D-5 for details). There was consistent decreased reporting for all age groups of the TCA Subregistry population when compared with the NHIS population. The risk ratios were statistically less than 1 for all age groups except the 10 through 17 years of age; risk ratios were less than 0.50 for all age-,sex-groups. The reporting time frame was "now have" for this outcome.

High Blood Pressure (Hypertension)

The estimated age- and sex-specific (age was a statistically significant factor) risk ratios for hypertension are given in Appendix D-6. The reporting time frame was "within the last 12 months". The age-specific risk ratios were statistically significantly (p £.01) reduced for two TCA Subregistry age groups; that is, the risk ratios were statistically less than 1. The two age groups were 25 through 34 years (O/E = 0.40, 10 observed versus 24.9 expected, 99% CI = 0.15, 0.86) and 35 through 44 years (O/E = 0.52, 33 observed, 63.6 expected, 99% CI = 0.32, 0.80).

Kidney Disease

Neither age nor sex was a statistically significant factor when introduced in the model. The age- and sex-specific risk ratios for kidney disease are given in Appendix D-7. The time frame for reporting was "within the last 12 months". An overall decreased reporting by the TCA Subregistry population was found but was only marginally statistically significant (O/E = 0.58, 29 observed versus 50.4 expected, 99% CI = 0.34, 0.91).

Liver Disease

There was no summary model; the data are exceedingly sparse for this outcome (13 observed, 10.8 expected). The age- and sex-specific risk ratios for liver disease are given in Appendix D-8. The time frame for reporting was "the last 12 months." Both deficits and excesses were seen; there was no overall pattern. The number expected for the total population (n = 10.8) and the number reported for the total population (n = 13) were small and precluded establishing an adequate model.

Mental Retardation

No summary estimates are given for mental retardation (see Appendix D-9 for details). Five registrants reported being currently treated for mental retardation (the time frame for reporting was "now being treated"); 14.7 were expected based on NHIS rates. The five responses occurred across age and sex groups with no overall pattern. There was an overall statistically significant decrease in reporting of mental retardation for the TCA registrants. The overall O/E ratio was 0.34 (5 observed versus 14.7 expected, 99% CI = 0.073, 0.962).

Skin Rashes, Eczema, or Other Skin Allergies

The summary model was adequate; neither age nor sex was a significant factor when introduced in the model. The sex and age group rates for the TCA Subregistry registrants and those expected based on the NHIS population are shown in Appendix D-10. The time frame for reporting was "the last 12 months". The overall risk ratio (224 observed versus 242.8 expected, O/E = 0.92; 99% CI = 0.77, 1.10) was not statistically significantly different from 1.

Other Respiratory Allergies or Problems, Such as Hay Fever

The model was adequate; sex was a statistically significant factor. Sex- and age-,sex-specific ratios for respiratory allergies are shown in Appendix D-11. The reporting time frame was "the last 12 months". The sex-specific estimated risk ratio for respiratory allergies was statistically significantly decreased for males only (O/E = 0.62, 94 observed versus 152.5 expected, 99% CI = 0.46, 0.8).

Speech Impairment

Four registrants reported being told they currently had or receiving treatment for a speech impairment. On the basis of this small sample size, there was an overall statistically significant decrease in reporting by the TCA Subregistry population when compared with the NHIS population. There was not statistically significant variability in the age- or sex-specific rates. The overall O/E ratio was 0.18 (4 observed, 21.6 expected, 99% CI = 0.03, 0.58). (See Appendix D-12 for details.) Speech impairment was another of the outcomes for which the restriction on health care provider confirmation was expected to reduce the number of registrant positive reports, all other factors being equal, when compared to NHIS reports.

Effects of Stroke

The age-specific estimates for stroke are shown in Appendix D-13. The time frame for reporting for this outcome was "ever had". Overall estimates are not given. There was not an adequate summary model. No statistically significant differences were found for age-sex groups; parameters could not be estimated for several groups because of sparse data (the denominator was zero).

Ulcers, Gallbladder Trouble, and Stomach or Intestinal Problems

The results for outcomes related to ulcers, gallbladder trouble, and stomach or intestinal problems appear in Appendix D-14. An overall summary model was obtained; age was a statistically significant factor when introduced in the model. The time frame for reporting for this outcome was "in the last 12 months". Overall, the TCA Subregistry population reporting rates were lower than the NHIS reporting rates (302.9 expected versus 168 observed). The decrease in reporting by the Subregistry population was statistically significant for the following age groups: 45 through 54 years of age (22 observed versus 38.9 expected, O/E = 0.56, 99% CI = 0.30, 0.95); 55 through 64 years of age (19 observed versus 46.9 expected, O/E = 0.40, 99% CI = 0.20, 0.71); and 65 years of age and older (48 observed versus 103.5 expected, O/E = 0.46, 99% CI = 0.31, 0.67). This outcome is also one for which self-diagnosis, rather than health care provider confirmation, could affect the reporting rates.

Urinary Tract Disorders, Including Prostate Trouble

The age- and sex-specific risk ratios for urinary tract disorders are given in Appendix D-15. Risk ratios and standard errors could not be calculated for the youngest male group; there were no reported cases in the TCA Subregistry or NHIS files. A statistically significant increase in reporting was found for the following TCA Subregistry female age groups: 10 through 17 years (O/E = 12.52, 5 observed versus 0.4 expected, 99% CI = 2.70, 35.43); 25 through 34 years (O/E = 5.82, 12 observed versus 2.1 expected, 99% CI = 2.40, 11.72); and 65 years of age or older (O/E = 2.45, 29 observed versus 11.83 expected, 99% CI = 1.44, 3.89). The following TCA Subregistry male group reported statistically significantly higher numbers of urinary tract disorders, including prostate trouble: 35 through 44 years of age (O/E = 3.78, 17 observed versus 4.5 expected, 99% CI = 1.83, 6.84).

Cancers

A summary of the reported primary and secondary cancers (time frame "within the last 12 months") is shown in Table 3-12. (Table 3-11 gives a summary of the reported primary and secondary cancers for the "ever had" time frame.) Appendix D-16 shows the results for all primary cancer outcomes when the TCA Subregistry reported cancers were compared with those reported by the NHIS. The model was adequate; age was a borderline statistically significant factor (p = 0.10). It should be noted that the age grouping was changed; all ages below 14 years were combined because of the sparsity of data for these age groups; because of the larger number of reports, the age group 65 years or older was split into 65 through 74 years and 74 years of age or older groups. The overall risk ratio of 1.25 (75 observed versus 59.8 expected, 99% CI = 0.91, 1.68) was not statistically significantly different from 1.

Comparison with SEER Data

Tables 3-17 through 3-19 show the calculated cancer rates for the TCA Subregistry population and the SEER rates (whites only) by sex and for all registrants for the years 1973 through 1989 (the years for which SEER data were available), along with the crude relative risks. An adjustment of the rates for survival until 1989 was made for the SEER data (17) (this adjustment is built in for the TCA Subregistry data); therefore, there was no known bias when the SEER rates were compared with the TCA Subregistry rates. As can be seen, for males and females, the rates in the TCA Subregistry were smaller for some years and greater for other years than those recorded by SEER; there was no consistent pattern.

SUMMARY

All of the results are summarized in Tables 3-15 and 3-16. These tables present a summary of the statistically significant (p £0.01) risk ratios observed in the Poisson regression analysis. Results were grouped according to the structure found in the corresponding Poisson regression model in Table 3-15; the results for the total Subregistry are shown in Table 3-16.

Table 3-17.—Comparison of the Trichloroethane (TCA) Subregistry and Surveillance, Epidemiology, and End Results (SEER) Program cancer rates (white males).

Year of Diagnosis	Count*	TCA Subregistry Rate†	SEER Rate§	Relative Risk¶
1973	0	0.0	40.9	0.00
1974	0	0.0	55.3	0.00
1975	1	65.8	62.2	1.06
1976	1	65.8	69.1	0.95
1977	3	197.5	77.1	2.56
1978	0	0.0	82.3	0.00
1979	1	65.8	92.0	0.72
1980	3	197.5	101.9	1.94
1981	4	263.3	115.5	2.28
1982	5	329.2	124.9	2.64
1983	5	329.2	146.2	2.25
1984	2	131.7	161.3	0.82
1985	1	65.8	189.5	0.35
1986	3	197.5	217.1	0.91
1987	5	329.2	267.4	1.23
1988	4	263.3	320.5	0.82
1989	8	526.7	434.6	1.21

*Number of male registrants reporting any cancers.

†Rate per 100,000 per year (not adjusted for age).

§SEER rates (per 100,000) not adjusted for age; adjusted for

survival until 1989.

¶Calculated crude relative risks using the crude TCA Subregistry

divided by the crude SEER rate.

Table 3-18.—Comparison of the Trichloroethane (TCA) Subregistry and Surveillance, Epidemiology,

and End Results (SEER) Program cancer rates (white females).

Year of Diagnosis	Count*	TCA Subregistry Rate†	SEER Rate§	Relative Risk¶
1973	3	182.1	105.4	1.73
1974	2	121.4	142.2	0.85
1975	2	121.4	162.8	0.75
1976	3	182.1	168.1	1.08
1977	1	60.7	167.5	0.36
1978	2	121.4	171.7	0.71
1979	0	0.0	176.5	0.00
1980	8	485.7	185.3	2.62
1981	3	182.1	200.6	0.91
1982	4	242.9	212.0	1.15
1983	7	425.0	230.5	1.84
1984	2	121.4	257.3	0.47
1985	6	364.3	288.3	1.26
1986	7	425.0	314.0	1.35
1987	6	364.3	357.8	1.02
1988	6	364.3	403.3	0.90
1989	8	485.7	481.7	1.01

*Number of female registrants reporting any cancers.

†Rate per 100,000 per year (not adjusted for age).

§SEER rates (per 100,000) not adjusted for age; adjusted for

survival until 1989.

¶Calculated crude relative risks using the crude TCA Subregistry

rate divided by the crude SEER rate.

Table 3-19.—Comparison of the Trichloroethane (TCA) Subregistry and Surveillance, Epidemiology,

and End Results Program (SEER) cancer rates (all white registrants).

Year of Diagnosis	Count*	TCA Subregistry Rate†	SEER Rate§	Relative Risk¶
1973	3	94.8	73.8	1.28
1974	2	63.2	98.6	0.64
1975	3	94.8	113.4	0.84
1976	4	126.3	119.5	1.06
1977	4	126.3	123.2	1.03
1978	2	63.2	127.9	0.49
1979	1	31.6	135.1	0.23
1980	11	347.4	144.5	2.40
1981	7	221.1	158.9	1.39
1982	9	284.3	169.3	1.68
1983	12	379.0	189.2	2.00
1984	4	126.3	210.2	0.60
1985	7	221.1	239.8	0.92
1986	10	315.9	266.4	1.19
1987	11	347.4	313.4	1.11
1988	10	315.9	362.7	0.87
1989	16	505.4	362.7	1.39

*Number of registrants reporting any cancers.

†Rate per 100,000 (not adjusted for age).

§SEER rates (per 100,000) not adjusted for age; adjusted for

survival until 1989.

¶Calculated relative risks using the crude TCA Subregistry rate

divided by the crude SEER rate.

SECTION 4 COMPARISON OF THE TRICHLOROETHANE (TCA) SUBREGISTRY FOLLOWUP 1 DATA AND NATIONAL DATA

The Trichloroethane (TCA) Subregistry Followup 1 data collection occurred one year after the Baseline data collection. The same questionnaire and procedures used at Baseline were used; the only difference was that telephone interviews rather than face-to-face interviews were used to collect the data. A concerted effort was made to locate and interview each of the 3,200 remaining registrants (4 had withdrawn from the National Exposure Registry [NER]) who were alive at the time of the Baseline data collection).

REGISTRANT DESCRIPTIVE DATA

The TCA Subregistry total at Baseline was 3,665 members; 3,204 were alive at time of Baseline data collection, 461 were deceased. Four of the registrants were removed from the Subregistry at their request, leaving 3,200 living registrants. At Followup 1, there were 3,469 members; 2,963 were alive, 510 deceased (49 additional deaths since Baseline). Only those living at the time of the Baseline and Followup 1 interviews are included in the statistical analyses reported in this document.

A summary of the registrants lost to followup is shown in Table 4.1. Complete information was obtained for 3,012 (94.1%) of the 3,200 eligible for a Followup 1 interview; 3% were lost to refusal, 2.3% to unable to contact, 0.3% to unable to schedule interview, and the remainder to multiple causes (for example, no proxy or litigation). For the deceased registrants, the Agency for Toxic Substances and Disease Registry (ATSDR) is currently obtaining death certificates; when this effort is completed, pertinent information will be abstracted, a mortality file constructed, comparisons using the appropriate national norms made, and the mortality analysis results published.

Of the total living population who completed an interview (n = 2,963), 2,929 persons (98.8%) answered white to the race question. The percentage of people reporting their race other than white was statistically significantly less than that for the comparison group-the 1991 National Health Interview Survey (NHIS) population. The analyses are, therefore, restricted (for both the Subregistry and NHIS files) to those answering white to the race question (n = 2,929). No further analyses were performed on the nonwhite groups because of the small numbers and the potential for violating the confidentiality of the respondents.

Tables 4-2 and 4-3 contain information about the characteristics of the TCA Subregistry Followup 1 members. The exposure period in Vestal, New York, ended in 1980; therefore, all eligible people in the Vestal community were older than 10 years of age at Baseline; consequently, all registrants were 11 years of age or older at Followup 1. For all the Followup 1 data age groups, one year has been added to age group limits to make the grouping consistent with those at Baseline. Table 4-2 indicates that at Followup 1, as at Baseline, 52% of the living registrants were female. Approximately 45% of the registrants were from 25 through 44 years of age at the time of the Followup 1 interview (consistent with Baseline); 4% were from 10 through 17 years of age (a decrease from 5% at Baseline).

Table 4-1.—Summary of Trichloroethane Subregistry registrants response at Followup 1.

Recontact Outcome	Number	Percent of Baseline (living only)
Completed interview, subject living	2,963	92.6
Completed interview, subject deceased	49	1.5
Refusal, some data	3	0.1
Refusal, no data	93	2.9
Unable to contact	74	2.3
Unavailable during interview period	9	0.3
No proxy available	2	0.1
In litigation	1	<0.1
Other	6	0.2
Total	3,200	100.0

The educational summary data did not vary from Baseline. Among registrants who were 20 years of age or older, 94% had a high school diploma. Slightly more females than males had not completed high school (6% versus 5%). A higher percentage of males (40%) than females (29%) had completed college or some post-college education.

Table 4-3 provides data on registrants' use of tobacco products. Of the registrants 19 years of age or older, 17% reported being current cigarette smokers (that is, they had smoked at least 100 cigarettes and smoked at time of data collection) compared with 19% at Baseline. Slightly more males than females (17% versus 16%) were current cigarette smokers. Fewer females than males (40% versus 49%) reported having ever smoked cigarettes (that is, were current smokers or had smoked at least 100 cigarettes in their lifetime). Overall, a small number of the registrants reported ever using other tobacco products: pipes (7%), cigars (7%), snuff (2%), and chewing tobacco (3%). Smoking rates for registrants at Followup 1 were comparable with the Baseline rates.

RATIONALE FOR THE COMPARISON OF THE

TCA SUBREGISTRY FOLLOWUP 1 DATA WITH NATIONAL DATA

This section includes comparisons of TCA Subregistry Followup 1 data with data from national surveys. These comparisons are consistent with NER objectives and goals as stated in the Policies and Procedures Manual (1) as discussed in the previous section. This section includes comparisons of Followup 1 registrant demographic and smoking data with national data. As discussed in Section 3, these comparisons are important because both demographic characteristics

Table 4-2.—Descriptive data for Trichloroethane Subregistry Followup 1 living registrants (white only).

Variable	Total Population N(%)
Variable	Total		Males		Females
Number	2,929		1,397		1,532
Age (years) 01-18 19-25 26-45 46-65 66-75 ³76	114 249 1,320 740 335 171	(4) (9) (45) (25) (11) (6)	57 131 653 342 152 62	(4) (9) (47) (24) (11) (4)	57 118 667 398 183 109	(4) (8) (44) (26) (12) (7)
Education (³20 years of age) Not high school graduate High school graduate Some college College graduate or more D* R†	123 867 802 987 1 1	(4) (31) (29) (36) (<1) (<1)	52 356 372 543 - -	(4) (27) (28) (41) - -	71 511 430 444 1 1	(5) (35) (30) (30) (<1) (<1)
Place of residence (at time of interview) On site Off site D*	1,195 1,726 8	(41) (59) (<1)	527 866 4	(38) (62) (<1)	668 860 4	(44) (56) (<1)
Type of interview Subject Proxy	2,724 205	(93) (7)	1,295 102	(93) (7)	1,429 103	(93) (7)

*D - Response was don't know (usually proxy reporting).

†R - Refused to answer question.

and smoking are known to be correlated with or are probable causes of many adverse health conditions.

TCA Subregistry Followup 1 data were compared with data obtained from the 1991 NHIS (18). Subsets of the NHIS data were used in the comparisons with demographic, smoking, and health data components of the TCA Subregistry Followup 1. The rationale for selecting the NHIS data for

Table 4-3.—Tobacco use for Trichloroethane Subregistry Followup 1 registrants 19 years of age or

older (white only).

Variable	Males N (%)		Females N (%)		Total N (%)
Cigarettes Current smoker Ex-smoker Never smoked	236 427 677	(17) (32) (51)	239 356 880	(16) (24) (60)	475 783 1,557	(17) (28) (55)
Pipe Current smoker Ex-smoker Never smoked D* R†	15 177 1,146 1 1	(1) (13) (86) (<1) (<1)	- - 1,475 - -	- - (100) - -	15 177 2,621 1 1	(<1) (6) (93) (<1) (<1)
Cigars Current smoker Ex-smoker Never smoked D* R†	44 132 1,162 1 1	(3) (10) (87) (<1) (<1)	1 - 1,474 - -	(<1) - (100) - -	45 132 2,636 1 1	(2) (5) (93) (<1) (<1)
Snuff Current user Ex-user Never used D* R†	19 22 1,298 1 1	(1) (2) (97) (<1) (<1)	- - 1,475 - -	- - (100) - -	19 22 2,733 1 1	(<1) (<1) (99) (<1) (<1)
Chewing tobacco Current user Ex-user Never used R†	20 48 1,271 1	(1) (4) (95) (<1)	- - 1,475 -	- - (100) -	20 48 2,746 1	(<1) (2) (98) (<1)

*D-Response was don't know (usually proxy reporting).

†R-Refused to answer question.

comparison with the TCA Subregistry data, for the National Exposure Registry in general, is commented on in Section 3.

As was discussed in Section 2 of this report, the TCA Subregistry sample is located primarily in the northeastern United States (New York); other registrants are located throughout other regions. The influence of the regionality on reported disease outcome rates for the TCA Subregistry, a concern when comparing the Subregistry reporting rates with the national rates reflected by the NHIS numbers, was explored. ATSDR's review of the NHIS regional rates for selected outcomes found no definitive evidence indicating that the overall health status of those located in the northeastern region differed statistically significantly from that of the general U.S. population. Rates for selected reported chronic conditions for the geographic regions Northeast, Midwest, South, and West (included all races) are listed in an NCHS publication (18). Analogous to the 1990 survey data used for the Baseline comparisons (see Section 3), very few of the reported health rates were highest for the northeastern region and none were related to conditions reported in excess by the registrants. Therefore, differences between the TCA Subregistry file and the NHIS file were not expected to be and did not appear to be the result of regional differences.

METHODS FOR STATISTICAL ANALYSES

TCA Subregistry Followup 1 and NHIS participant reporting rates for demographic and health variables were compared using the same statistical methods described in detail in Section 3. The strengths and limitations of those comparisons are also relevant to the statistical comparisons carried out for this section.

RESULTS OF DESCRIPTIVE COMPARISONS

The results of analyses exploring the comparability of the descriptive data for the TCA Subregistry Followup 1 file and the 1990 NHIS file and a discussion of these findings are presented.

Demographics

Race/p>

As discussed earlier in this report, because of the sparsity of registrants reporting race as other than white, those so reporting (17%) were not included for either the Subregistry or the NHIS file in the statistical comparisons of the reporting rates for health outcomes.

Age

Table 4-4 shows the percentage of registrants and NHIS participants in each age group. As discussed previously, the percentages in some age groups were affected by the timing of data collection relative to the exposure period. The information gives assurance that the ages of populations examined were not skewed. Table 4-5 provides a comparison of the age-specific sex distribution of males in the TCA Subregistry Followup 1 and NHIS samples and the corresponding p values; Table 4-5 also indicates that the proportion of males was not statistically significantly different within the age groups represented in the TCA Followup 1 file.

Table 4-4.—Comparison of National Health Interview Survey (NHIS) - Trichloroethane (TCA)

Subregistry Followup 1 population age groups (white only).

Age Group (Years)	NHIS*	TCA Followup 1
Age Group (Years)	% of Total	Number	% of Total
All	100.0	2,929	100.0
1 - 5	7.2	0	0.0
6 - 10	7.1	0	0.0
11 - 18	10.5	114	3.9
19 - 25	9.5	249	8.5
26 - 35	17.2	580	19.8
36 - 45	16.0	740	25.3
46 - 55	10.9	412	14.1
56 - 65	8.9	328	11.2
66 - 75	7.8	335	11.4
³76	4.9	171	5.8

*NHIS - National Health Interview Survey.

Education Level

The summary statistics for education attainment are provided in Table 4-6. As at Baseline, the percentage of persons with more than a high school education was greater for the TCA registrants; the percentage with less than a high school education was lower than that found for NHIS participants. These findings indicate that the TCA Subregistry Followup 1 participants had attained a higher level of education than the comparison population and, therefore, suggest that there is not a positive bias in registrant reporting rates related to socioeconomic level.

Cigarette Smoking

Table 4-7 provides comparative information on current and past cigarette use for people 19 years of age or older. The smoking rates for both males and females were lower when compared with the NHIS national and the NHIS Northeast region rates. Table 4-8 shows smoking rates by education level. With few exceptions, reduced smoking rates by registrants were the norm.

Table 4-5.—Comparison of percentage of males by age group for the Trichloroethane (TCA)

Subregistry Followup 1 and National Health Interview Survey populations (NHIS) (white only).

Age	Followup 1	NHIS*	p value†
1 - 5	---	0.51	---
6 - 10	---	0.51	---
11 - 18	0.50	0.51	0.84
19 - 25	0.53	0.49	0.30
26 - 35	0.49	0.50	0.63
36 - 45	0.50	0.50	0.91
46 - 55	0.46	0.49	0.20
56 - 65	0.46	0.48	0.66
66 - 75	0.45	0.45	0.87
³ 76	0.36	0.37	0.91

*Cell values are proportions of total males in age group.

†The p values are two-sided.

RESULTS OF HEALTH OUTCOME COMPARISONS

A summary of the reporting rates for health conditions for the TCA Subregistry file and the NHIS file is provided in Table 4-9. As at Baseline, at the time of the interview, if a registrant reported having been told that he or she had cancer or had been treated for cancer by a health care provider, additional questions were asked about the type(s) of cancer. Although multiple types of cancers might have been reported, only one primary type of cancer was listed for each registrant. The summary table entries and data analysis are based on the reported primary cancers; the number of secondary cancers is also included in the tables. A summary of the TCA Subregistry population's reporting rates for specific cancers (total population and by sex) is shown in Table 4-10 for the "ever had" time frame (which combines Baseline and Followup 1 information) and Table 4-11 for the "within the last 12 months" time frame (Followup 1 information only). The "within last 12 months" time frame (Table 4-11) is comparable to the NHIS time frame and was used in the statistical comparisons of the data files.

Table 4-6.—Education level attained for the Trichloroethane (TCA) Subregistry Followup 1 and

National Health Interview Survey (NHIS) populations (white, 20 years of age or older).

Age (Years)	NHIS Education Level (average years)				Followup 1 Education Level (average years)
Age (Years)	0-11	12	13-15	16+	0-11	12	13-15	16+
Males and Females
All	18.8	38.0	21.4	21.9	4.4	31.2	28.9	35.5
20 - 26	13.3	37.8	32.4	16.5	2.3	17.8	48.8	31.0
27 - 46	11.4	37.4	23.8	27.4	1.1	24.0	31.3	43.6
47 - 65	20.8	40.8	17.3	21.2	3.9	38.8	22.9	34.4
³66	39.1	35.3	12.7	12.9	15.1	47.0	20.0	17.9
Males
All	18.6	35.6	20.6	25.2	3.9	26.9	28.1	41.0
20 - 26	13.7	39.0	31.7	15.6	2.2	17.3	46.0	34.5
27 - 46	12.1	36.3	22.0	29.6	1.2	21.7	29.5	47.6
47 - 65	21.0	35.5	16.8	26.8	2.9	31.6	22.6	42.9
³66	39.4	30.5	12.5	17.6	15.0	42.5	20.1	22.4
Females
All	18.9	40.1	22.1	18.9	4.9	35.1	29.5	30.5
20 - 26	12.9	36.7	33.0	17.3	2.5	18.5	52.1	26.9
27 - 46	10.7	38.5	25.4	25.3	1.0	26.3	33.0	39.7
47 - 65	20.6	45.7	17.8	15.9	4.7	45.0	23.2	27.1
³66	38.8	38.8	12.9	9.5	15.2	50.3	20.0	14.5

Table 4-7.—Reported rates of cigarette smoking for the Trichloroethane (TCA) Subregistry Followup 1 and National Health Interview Survey (NHIS) populations (white only).

Age Group	N	Followup 1 Rates (%)		NHIS Rates (%)
		Followup 1 Rates (%)		All		Northeast Region
		Current*	Ever†	Current	Ever	Current	Ever
³19 years All Males Females	2,815 1,340 1,475	16.9 17.6 16.2	44.7 49.5 40.3	25.2 26.9 23.6	51.4 59.2 44.4	23.7 24.7 22.7	52.4 58.7 46.7
19-26 years All Males Females	292 156 136	19.2 23.1 14.7	27.7 30.1 25.0	26.2 26.2 26.1	36.9 36.9 36.9	26.4 26.9 25.9	38.1 36.4 39.9
27-35 years All Males Females	537 257 280	20.3 19.5 21.1	39.9 35.4 43.9	30.3 32.2 28.5	48.7 51.0 46.4	30.8 30.3 31.3	51.4 50.9 51.9
³36 years All Males Females	1,986 927 1,059	15.6 16.2 15.1	48.5 56.6 41.4	23.3 25.2 21.6	56.3 68.4 45.9	20.8 22.2 19.7	56.4 67.0 47.2

*Reported currently smoking.

†Reported having smoked more than 100 cigarettes in lifetime.

Table 4-8.—Current smoking rates (percentage) by education attainment for the Trichloroethane

(TCA) Subregistry Followup 1 (F1) and National Health Interview Survey populations (NHIS)

(white only).

Age Group	Education Level Attained
	No High School Diploma		High School Graduate		Some College		College Graduate*
	F1	NHIS	F1	NHIS	F1	NHIS	F1	NHIS
³19 years All Males Females	12.7† 11.3 13.7	32.1 36.7 28.2	21.6 22.2 21.2	29.6 32.3 27.3	21.0 24.3 18.1	22.9 24.1 22.0	9.8 10.5 9.0	13.4 14.2 12.5
19-26 years All Males Females	55.6 50.0 60.0	46.6 49.7 43.3	36.8 42.1 30.0	32.5 33.7 31.4	15.6 21.2 10.1	17.1 14.3 19.7	6.3 8.3 3.1	9.1 7.8 10.3
27-35 years All Males Females	16.7 0.0 33.3	52.8 56.2 49.5	33.3 31.7 34.6	39.6 42.2 37.0	25.9 28.2 24.2	24.5 26.3 23.0	8.7 8.9 8.3	11.0 11.4 10.5
³36 years All Males Females	9.0 8.7 9.2	26.3 30.7 22.9	17.7 17.3 18.0	25.7 28.3 23.9	20.8 24.1 18.1	24.7 27.4 22.4	10.7 11.3 9.9	15.1 16.0 13.8

*May include additional years post-bachelor degree.

†Percentage of group reporting currently smoking.

Table 4-9.—Comparison of the Trichloroethane (TCA) Subregistry Followup 1 and National Health

Interview Survey (NHIS) participant reporting rates for health conditions (white only).

Condition	Followup 1 (%)*		NHIS (%)*
Condition	Male	Female	Male	Female
Hypertension†	8.9	10.1	11.1	11.5
Stroke§	0.6	0.5	1.4	1.2
Diabetes†	2.6	2.5	2.8	3.1
Kidney disease†	0.7	0.8	1.1	1.7
Urinary tract disorders†	4.3	6.1	1.5	1.3
Skin rashes†	5.7	8.0	6.3	8.6
Anemia and other blood disorders†	0.8	3.3	0.4	2.1
Asthma, emphysema†	5.0	6.7	9.0	11.5
Respiratory allergies†	5.3	8.0	10.6	11.5
Stomach problems, ulcers†	3.1	7.1	7.3	9.9
Liver problems†	0.5	0.5	0.4	0.3
Arthritis†	6.2	9.5	15.2	21.0
Mental retardation¶	0.1	0.1	0.7	0.5
Speech impairment¶	0.2	0.1	1.4	0.8
Hearing impairment¶	0.5	0.8	12.9	8.7

*Percentage of subpopulation (white only) reporting positively.

†Indicates time frame was "last 12 months."

§Indicates time frame was "ever had."

¶Indicates time frame was "now have."

Table 4-10.—Summary of the Trichloroethane (TCA) Subregistry Followup 1 reporting rates for

cancer (primary site), "ever had" time frame (white only).

Cancer	Sex						All
	Male			Female			All
	N		%	N		%	N		%
None	1,365		97.7	1,494		97.5	2,859		97.6
Lip, oral, pharynx	0		0.0	0		0.0	0		0.0
Digestive system	2		0.1	4		0.3	6		0.2
Respiratory system	1		0.1	3		0.2	4		0.1
Skin	14	(1)*	1.0	12		0.8	26	(1)	0.9
Breast	0		0.0	10		0.7	10		0.3
Genital organs	11	(2)	0.8	4		0.3	15	(2)	0.5
Urinary organs	3	(1)	0.2	0		0.0	3	(1)	0.1
Lymphatic tissues	0		0.0	1		0.1	1		<0.1
Leukemia	0		0.0	1		0.1	1		<0.1
Other	1		0.1	3	(1)	0.2	4	(1)	0.1
Total cancers	32	(4)	2.3	38	(1)	2.5	70	(5)	2.4

* - The numbers in parentheses are the numbers of reported secondary sites.

Table 4-11.—Summary of the Trichloroethane (TCA) Subregistry Followup 1 reporting rates for

cancer (primary site), "within last 12 months" time frame (white only).

Cancer	Sex					All
	Male			Female		All
	N		%	N	%	N		%
None (People)	1,375		98.4	1,504	98.2	2,879		98.3
Lip, oral, pharynx	0		0.0	0	0.0	0		0.0
Digestive system	0		0.0	3	0.2	3		0.1
Respiratory system	1		0.1	3	0.2	4		0.1
Skin	11	(1)*	0.8	6	0.4	17	(1)	0.6
Breast	0		0.0	9	0.6	9		0.3
Genital organs	9	(1)	0.6	3	0.2	12	(1)	0.4
Urinary organs	1		0.1	0	0.0	1		<0.1
Lymphatic tissues	0		0.0	0	0.0	0		0.0
Leukemia	0		0.0	1	0.1	1		<0.1
Other	0		0.0	3 (1)	0.2	3	(1)	0.1
Total cancers	22	(2)	1.6	28 (1)	1.8	50	(3)	1.7

* - The numbers in parentheses are number of reported secondary sites.

Controlling for Age and Sex

Table 4-12 provides a summary of the results of the NHIS and TCA Subregistry file comparison using Poisson regression analysis. For each health outcome, the table indicates the likelihood ratio statistics with the associated degrees of freedom and p values for the effects of age (categorized into eight levels) and sex, based on a model containing age and sex. The residual deviance and the associated degrees of freedom are also given as a global lack-of-fit measure for this model, which specifies multiplicative effects of the age (i) and sex (j) ratios Oij/Eij. For each outcome, the age- and sex-specific numerators Oij were obtained from the TCA Subregistry Followup 1 data, while the expected denominators Eij were based on the suitably person-weighted age- and sex-specific ratios from the NHIS data. For the purpose of detecting structure in these age- and sex-specific ratios, a significance level of 0.05 was adopted.

Table 4-12.—Summary of Poisson regression modeling.

Condition	Age/Sex*		Sex/Age*		Residual Deviance (p value)	df
Condition	LR Stat† (p value)	df§	LR Stat (p value)	df	Residual Deviance (p value)	df
Skin rashes	4.61 (p = 0.60)	6	0.01 (p = 0.91)	1	7.79 (p = 0.25)	6
Arthritis	7.28 (p = 0.30)	6	1.49 (p = 0.22)	1	7.60 (p = 0.27)	6
Mental retardation¶	---	---	---	---	---	---
Speech impairment	---	---	---	---	---	---
Hearing impairment	2.96 (p = 0.81)	6	3.80 (p = 0.05)	1	3.70 (p = 0.72)	6
Liver problems¶	---	---	---	--	---	---
Stomach problems, ulcers**	12.84 (p = 0.05)	6	10.16 (p £ 0.01)	1	8.92 (p = 0.18)	6
Anemia and blood disorders**	17.57 (p £ 0.01)	6	1.18 (p = 0.28)	1	36.26 (p £ 0.01)	6
Diabetes**	10.11 (p = 0.12)	6	0.52 (p = 0.47)	1	8.49 (p = 0.20)	6
Kidney disease	6.61 (p = 0.36)	6	0.18 (p = 0.67)	1	13.86 (p = 0.03)	6
Urinary tract disorders**	14.05 (p = 0.03)	6	8.83 (p £ 0.01)	1	2.24 (p = 0.90)	6
Hypertension**	9.90 (p = 0.13)	6	0.34 (p = 0.56)	1	4.90 (p = 0.56)	6
Stroke¶	---		---		---
Respiratory allergies	8.87 (p = 0.18)	6	6.14 (p = 0.01)	1	3.94 (p =0.68)	6
Asthma, emphysema	11.19 (p = 0.08)	6	0.35 (p = 0.55)	1	6.16 (p = 0.41)	6
Cancer	9.48 (p = 0.15)	6	0.06 (p = 0.81)	1	6.92 (p = 0.33)	6

*Indicates order of inclusion in the model.

†LR Stat - Likelihood Ratio Statistic.

§df - degrees of freedom.

¶Indicates no model constructed.

**Indicates some infinite estimates obtained.

As is shown in Table 4-12, the model was adequate and neither age nor sex was a statistically significant predictor in the models for the health outcomes skin rashes; cancers; asthma, emphysema, or chronic bronchitis; arthritis, rheumatism or other joint disorders; diabetes; and hypertension (p > 0.05). For the health outcomes hearing impairment and respiratory allergies or problems such as hay fever, the effects of sex were statistically significant; the effects of age were not. In addition to age effects, statistically significant sex effects were present for urinary tract disorders and stomach problems. The multiplicative model for anemia and kidney disease exhibited statistically significant lack of fit, suggesting that the effects of age depended on sex for these outcomes. Because of the sparseness of the data, an adequate model could not be constructed for the liver, mental retardation, stroke, and speech impairment outcomes.

Table 4-13 summarizes the observed (O) and expected (E) numbers and their ratios (O/E) for each of the health conditions considered separately. The results for age- and sex-specific analyses are provided in Appendix E. For the descriptive analysis, a ratio of 0/0 was considered to be undefined. For models in which the sex effect was not statistically significant, summary estimates based on the age-specific ratios for the combined sexes were given. For outcomes in which, at most, one effect was statistically significant, summary estimated ratios based on the simplest model that did not exhibit lack of fit were given together with 99% Wald confidence intervals (CIs). Reliable estimates of the standard errors of estimated risk ratios of zero could not be obtained, so no CIs are presented whenever this occurred. For outcomes for which a simple model was not adequate, 99% CIs are presented for the individual age- and sex-specific ratios. For the purpose of identifying statistically significant excesses, a significance level of 0.01 was used. For some outcomes the model was mathematically adequate; however, because the data were very sparse, results should be interpreted cautiously.

A summary of the appropriate model for each health condition and the results using that model are summarized in Table 4-14. Interpretation of the results must be carried out judiciously; other factors-such as age at exposure, time since exposure, and sample size within age groups-must be taken into consideration. A discussion of the statistical results for each health outcome follows. The time frame "ever had" was used for reporting rates for the health outcomes hypertension, stroke, and heart condition; the time frame "now have" for the health outcomes mental retardation, speech impairment, and hearing impairment; and the time frame "had in the last 12 months" for the remainder of the health outcomes.

For the results of the statistical analyses presented here, the age- and sex-specific risk ratios are defined as the number reported by the TCA Subregistry registrants divided by the number expected based on the number reported by the NHIS participants. When the expected value (the number reported by the NHIS participants) was zero, the risk ratio was not estimated. The following discussion is based on the statistical results summaries found in Tables 4-12 through 4-15, and Appendix E (details of statistical analyses).

Anemia or Other Blood Disorders

The results for the TCA Subregistry Followup 1 and NHIS comparisons of reporting rates for anemia and other blood disorders are detailed in Appendix E-1. For the outcome anemia, the time frame of reporting was "had in the last 12 months." As at Baseline, no summary estimates are presented because there was not a summary model; the estimated effect of sex depended upon the age group. Excess reporting by registrants (that is, the O/E ratios were greater than 1) was found for all age- and sex-groups older than 46 years of age. However, only the excess reporting seen in the female age group 26 through 35 years (17 observed versus 7.9 expected; O/E = 2.15, 99% CI = 1.04,3.89) was statistically significantly increased (p £ .01).

Table 4-13.—Summary of observed and expected health outcomes using multivariate models.

Condition	Observed	Expected	Risk Ratio	99% CI*
Arthritis	232	699.36	0.33	0.28, 0.39
Speech impairment	4	24.59	0.16	0.03, 0.51
Hearing impairment	19	402.02	0.05	0.02, 0.08
Kidney disease	22	49.96	0.44	0.24, 0.75
Urinary tract disorders	153	50.46	3.03	2.44, 3.72
Hypertension	280	436.52	0.64	0.55, 0.75
Respiratory allergies	197	348.67	0.57	0.47, 0.68
Asthma, emphysema	173	297.63	0.58	0.47, 0.71
Anemia and other blood disorders	62	42.72	1.45	1.02, 2.00
Diabetes	75	112.94	0.66	0.48, 0.89
Liver	15	13.06	1.15	0.53, 2.16
Mental retardation	4	9.26	0.43	0.07, 1.36
Skin rashes	202	236.58	0.85	0.71, 1.02
Stroke	15	50.95	0.29	0.14, 0.55
Stomach problems, ulcers	152	316.92	0.48	0.39, 0.59
Cancer	50	65.38	0.77	0.52,1.09

*CI - Confidence interval for risk ratio.

Arthritis, Rheumatism, or Other Joint Disorders

The age-,sex-specific estimates (sex and age were not statistically significant factors in the model) for the outcomes related to arthritis, rheumatism, or other joint disorders are shown in Appendix E-2. The reporting time frame for these outcomes was "had in the last 12 months." The sex-specific risk ratios were greater than 1 for females in the group 11 through 18 years of age. The sex-specific risk ratios were less 0.50 and statistically significantly less than 1 for all age groups (232 observed versus 699.4 expected; O/E = 0.33; 99% CI = 0.28,0.39). As discussed in Section 3,

Table 4-14.—Summary of results of statistical analyses comparing the Trichloroethane (TCA)

Subregistry Followup 1 and National Health Interview Survey (NHIS) rates.

1. Grand Mean, No Structure with Age or Sex.

Structure	Condition
Structure	Arthritis	Asthma, Emphysema	Diabetes	Hypertension	Hearing Impairment	Stroke
Overall Summary	R*	R	R	R	R	R

2. Sex Effect Only

Sex	Condition
Sex	Respiratory Allergies
Males	R
Females	R

3. Age and Sex Effect

Age (Years)	Condition
	Anemia		Kidney Disease		Speech Impairment		UrinaryTract Disorders		Stomach Problems, Ulcers
	Males	Females	Males	Females	Males	Females	Males	Females	Males	Females
11-18							---†
19-25								X¶
26-35		X		R			X	X
36-45							X	X	R
46-55						---		X	R	R
56-65								X	R	R
³66						R	X	X	R	R

*R=Statistically significant differences, NHIS rate higher. †--=Insufficient data. ¶X=Statistically significant differences, Subregistry rate higher.

Table 4-15.—Summary results of statistical analyses comparing the Trichloroethane (TCA) Subregistry reporting rates with National Health Interview Survey (NHIS) reporting rates.

Condition	Age Groups (Years)
	11-18		19-25		26-35		36-45		46-55		56-65		³66		All		All
	M	F	M	F	M	F	M	F	M	F	M	F	M	F	M	F	All
Anemia or other blood disorders			-		-	X
Arthritis																	R
Asthma, emphysema																	R
Diabetes																	R
Hearing impairment																	R
Hypertension																	R
Kidney disease						R
Respiratory allergies															R	R
Speech impairment										-				R
Ulcers or other stomach problems							R		R	R	R	R	R	R
Stroke																	R
Urinary tract disorders	-			X	X	X	X	X		X		X	X	X

X = Statistically significant differences, Subregistry rate higher.

R = Statistically significant differences, NHIS rate higher.

─ = Insufficient data.

this indicated that the reporting rates were generally lower in the TCA Subregistry data than expected based on the NHIS reporting rates-a phenomenon expected for this outcome given the health care provider confirmation restriction placed on registrant responders.

Asthma, Emphysema, or Chronic Bronchitis

The results for the outcomes related to asthma, emphysema, or chronic bronchitis showed that the effects of age and sex were not statistically significant factors (see Appendix E-3 for details). The reporting time frame was "within the last 12 months". The estimated overall risk ratio of 0.58 (173 observed, 297.6 expected, 99% CI = 0.47, 0.70) was statistically significantly less than 1, indicating a statistically significantly decreased reporting in this group relative to the NHIS population. Only the male, 11 through 18 years, age-,sex-specific O/E ratio exceeded 1. This result was consistent with what was expected given the impact of the health care provider restriction on the registrants; these health conditions are often self-diagnosed. In addition, the results are consistent with Baseline results.

Diabetes

For diabetes, age- and sex-specific risk ratio estimates are presented in Appendix E-4. The reporting time frame was "had in the last 12 months." With two exceptions (females in the group 11 through 18 years of age and females and males in the groups 19 through 25 years of age) the age-,sex-specific risk ratios were less than 1. The model was adequate, there were no statistically significant sex or age effects, and the O/E ratio was statistically significantly less than 1 (75 observed versus 112.9 expected; O/E = 0.66; 99% CI = 0.48, 0.89). No statistically significant differences were found at Baseline, although decreased reporting was also found at Baseline.

Hearing Impairment

For hearing impairment, there was consistent decreased reporting for all age groups of the TCA Subregistry Followup 1 population when compared with the NHIS population. The risk ratios were statistically less than 1 for all age groups (19 observed, 402.0 expected; O/E = 0.05; 99% CI = 0.02, 0.08); risk ratios were less than 0.20 for all age-, sex-groups. The reporting time frame was "now have" for this outcome. (See Appendix E-5 for further details.) These results are consistent with Baseline results and with the reporting rates expected given the health care provider confirmation restriction on the registrants.

High Blood Pressure (Hypertension)

The estimated age- and sex-specific ratios for hypertension are given in Appendix E-6. The reporting time frame is "within the last 12 months." The model was adequate, and age and sex were not statistically significant factors. The grand mean was statistically significantly (p £ .01) reduced for TCA Subregistry Followup 1 age groups (280 observed, 436.5 expected; O/E = 0.40; 99% CI = 0.55, 0.74). The reduced reporting by registrants is the trend also noted at Baseline.

Kidney Disease

An adequate summary model could not be constructed for this outcome. The age- and sex-specific risk ratios for kidney disease are given in Appendix E-7. The time frame for reporting was "had in the last 12 months." An overall decreased reporting by the TCA Subregistry Followup 1 population was found; none of the differences were statistically significant. The results were consistent with the Baseline findings.

Liver Disease

There was no summary model; the data were exceedingly sparse for this outcome (15 observed, 13.6 expected). The age- and sex-specific risk ratios for liver disease are given in Appendix E-8. The time frame for reporting was "the last 12 months." Both deficits and excesses were seen; there was no overall pattern.

Mental Retardation

No summary estimates are given for mental retardation (see Appendix E-9 for details). Four registrants reported being currently treated for mental retardation (the time frame for reporting was "now being treated"); 9.3 were expected based on NHIS rates. Although there was no overall pattern, 2 of the four responses occurred in the 26 through 35 years of age group.

Skin Rashes, Eczema, or Other Skin Allergies

As at Baseline, the summary model was adequate and neither age nor sex was a significant factor when introduced in the model. The differences between the TCA Subregistry sex and age group rates and those expected based on the NHIS populations are shown in Appendix E-10. The time frame for reporting was "the last 12 months." The overall risk ratio (202 observed versus 236.6 expected, O/E = 0.85; 99% CI = 0.71, 1.02) was not statistically significantly different from 1. This is an outcome where the restriction of health care provider confirmation was expected to (and did) result in statistically significant decreased reporting by the registrants.

Other Respiratory Allergies or Problems, Such as Hay Fever

As at Baseline, the model was adequate; sex was a statistically significant factor. Sex- and age-,sex-specific ratios for respiratory allergies are shown in Appendix E-11. The reporting time frame was "the last 12 months." The sex-specific estimated risk ratio for respiratory allergies was statistically significantly decreased for males (74 observed versus 159.9 expected; O/E = 0.42; 99% CI = 0.34, 0.62) and females (123 observed versus 188.8 expected; O/E = 0.65; 99% CI = 0.51, 0.82).

Speech Impairment

A total of 4 registrants reported currently being told they had or receiving treatment for a speech impairment; 24.6 were expected based on NHIS reporting rates. Based on this small sample size, an adequate model could not be constructed (see Appendix E-12 for details.) This was another of the outcomes for which the restriction on health care provider confirmation was expected to reduce the number of registrant positive reports, all other factors being equal, when compared with NHIS participants.

Effects of Stroke

The age-specific estimates for stroke are shown in Appendix E-13. The time frame for reporting for this outcome was "ever had"; the results of Baseline and Followup 1 were combined to calculate the reporting rates. There was not an adequate summary model. No statistically significant differences were found for age-,sex-groups; parameters could not be estimated for several groups because of sparse data (the denominator was zero).

Ulcers, Gallbladder Trouble, and Stomach or Intestinal Problems

The results for outcomes related to ulcers, gallbladder trouble, and stomach or intestinal problems appear in Appendix E-14. An overall summary model was obtained; age and sex were statistically significant factors when introduced in the model; sex was not statistically significant in the Baseline model. The time frame for reporting for this outcome was "in the last 12 months." Overall, the TCA Subregistry population reporting rates were lower than the NHIS reporting rates (317.0 expected versus 152 observed). The decrease in reporting by the Subregistry population was statistically significant for the group 46 through 55 years of age. This outcome is also one for which self-diagnosis rather than health care provider confirmation could potentially affect the reporting rates.

Urinary Tract Disorders, Including Prostate Trouble

The age- and sex-specific risk ratios for urinary tract disorders are given in Appendix E. As at Baseline, risk ratios and standard errors could not be calculated for the youngest male group; there were no reported cases in the TCA Subregistry Followup 1 or NHIS files. A statistically significant increase in reporting was found for TCA Subregistry females aged 19 years or older, and for males 26 through 45 years of age and 66 years of age and older (see Appendix E-15 for details). A total of 153 cases were reported; 50.4 were expected based on NHIS rates.

Cancers

A summary of the reported primary and secondary cancers (time frame "had in the last 12 months") is shown in Table 4-11. (Table 4-10 gives a summary of the reported primary and secondary cancers for the "ever had" time frame.) Appendix E-16 shows the results for all primary cancer outcomes when the TCA Subregistry reported rates (12-month time frame) were compared with the NHIS rates. The model was adequate. It should be noted that age grouping was changed because of the sparsity of data for the younger age groups and because of the larger number of reports for older age groups; the age group 65 years or older was split into 65 through 74 years of age and 74 years of age or older. The overall risk ratio of 0.76 (50 observed versus 65.4 expected, 99% CI = 0.51, 1.09) was not statistically significantly different from 1.

Summary

All of the results are summarized in Tables 4-14 and 4-15. These tables present a summary of the statistically significant (p £ 0.01) risk ratios observed in the Poisson regression analysis. Results were grouped according to the structure found in the corresponding Poisson regression model in Table 4-14; the results for the total Subregistry are shown in Table 4-15.

SECTION 5 DISCUSSION

A key purpose of the Trichloroethane (TCA) Subregistry is to create a database that can be used to determine if there is an excess of adverse health conditions for registrants with known exposure when compared with a national sample. To date, this objective has been pursued by comparing TCA Subregistry data about reported health conditions with National Health Interview Survey (NHIS) data and comparing TCA Subregistry reporting of cancer outcomes with cancer incidence data from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program and rates derived from state data. Health, demographic, occupational, and environmental information was collected on 3,665 TCA-exposed persons at Baseline (3,204 living, 461 deceased) and 3,469 TCA-exposed persons at Followup (2,963 living and 510 deceased). The analysis of the mortality data is not included in this report.

The registrants met the eligibility criterion of documented exposure. They resided for more than 30 consecutive days during the period of exposure at a site address known to have a contaminated water source, in this case contamination with TCA. It should be noted that the environmental samples were taken to verify contamination and not for the purpose of quantifying exposures over time. These samples might or might not have been taken when TCA and other chemical exposures were at a maximum level.

Several sources of potential bias in the TCA Subregistry population reporting rates were identified. One potential source of bias was nonresponse. However, the response rate of eligible persons who were located was 97% at Baseline and 94.1% at Followup; such a high participation rate minimized or eliminated bias in the data that might have been associated with nonresponse. Another potential source of reporting bias was knowledge of exposure. TCA Subregistry and NHIS data are similar in that both were self-reported; however, the responses of the registrants might have been influenced, in part, by their knowledge of and concern about TCA exposure.

The Subregistry and NHIS health condition questions, while sharing important similarities, differed in the restrictions on the source of diagnoses or treatment leading to a positive response to a health question. Also, there was a difference in the wording for some of the health questions. For those reasons, response rates might have been altered or biased. Concerning the source of diagnosis and treatment, the TCA Subregistry questions specified that the source of diagnosis or treatment had to be a "physician or other medical provider;" reports of medical diagnoses were not confirmed from medical records. This restriction was added to each Subregistry question about health conditions to minimize increased biased reporting 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 similar or equal, and to modulate any increased reporting as a result of bias. Some conditions are more commonly self-diagnosed and these would predictably be reported at a higher rate by the NHIS participants; the results were consistent with that prediction. There was a deficit in reporting by the registrants for the conditions asthma and emphysema, ulcer or other stomach problems, speech impairment, arthritis, hearing impairment, and respiratory allergies-conditions often self-diagnosed-and kidney disease, mental retardation, and hypertension.

The comparability of the wording of TCA Subregistry and NHIS health conditions was addressed in Section 3. Nine of the health condition questions matched exactly or very closely and eight others were considered similar. However, when the questions did not match exactly, again all other factors being the same, it was likely the dissimilarity would have resulted in decreased reporting by the TCA Subregistry population in all categories save one-urinary tract disorders. The implications of health condition comparability are addressed further in the following discussion.

HEALTH OUTCOMES

As discussed in Section 3, the comparisons of TCA Subregistry and NHIS data on reported health conditions revealed several statistically significant differences. It should be noted that setting the Type I error (the α level) at £0.01 and the large sample size controlled the likelihood of false-negative or false-positive results. The results of the morbidity analyses are summarized in Table 5-1 and are discussed in the following text with respect to the relevant literature. As discussed previously, there are some differences in the NHIS and the National Exposure Registry (NER) questionnaires-time frame of reporting; wording of questions; and the requirement that positive reporting of health conditions by the NER respondents had to be confirmed by a health care provider. There is even a greater expectation of increased reporting by NHIS respondents for the group of health outcomes that are typically self-diagnosed; that is, the NHIS has no confirmation that a health care provider told the participant that he or she had the condition or treated the participant for the condition, as is required by the NER.

Metabolism

Following exposure, TCA is readily absorbed into the body through the lungs and gastrointestinal tract; it is absorbed less efficiently through intact skin (19-24). The compound is then rapidly redistributed from the blood into the tissues (25). TCA has a low rate of metabolism; approximately 90% of the absorbed dose is excreted unchanged through the lungs (19, 26-30). A small percentage of the solvent is converted to trichloroethanol, trichloroacetic acid, and trichloroethanol-glucuronide, and excreted in urine (19). In addition, some TCA is transformed to carbon dioxide and eliminated in exhaled air (31). It is eliminated rapidly at first, but elimination from fatty tissue results in a half life of 53 hours (19,22,32).

Anemia or Other Blood Disorders

Anemia and other blood disorders were reported in excess in females who were 25 through 34 years of age at Baseline and in this same group at Followup; females 65 years of age or older at Baseline also reported statistically significantly more cases of this outcome. It should be noted that the wording of the TCA Subregistry and NHIS questions for nemia was a close match (see Table 3-4).

Hematological parameters have been reported to remain unchanged in humans acutely exposed to high or moderate concentrations of TCA (33-38) and in workers chronically exposed to low to moderate levels of TCA (39). These variables included white blood cell count, red blood cell

Table 5-1.—Summary results of statistical analysescomparing the Trichloroethane (TCA)

Subregistry Baseline and Followup 1 reporting rates with National Health Interview Survey

(NHIS) reporting rates.

Condition	Age Groups At Baseline* (Years)
	10-17		18-24		25-34		35-44		45-54		55-64		³65		All		All
	M	F	M	F	M	F	M	F	M	F	M	F	M	F	M	F	All
Anemia and other blood disorders
Baseline
Followup 1			-		-	X
Arthritis
Baseline	X				R¶		R	R	R	R	R	R	R	R
Followup 1																	R
Asthma, emphysema
Baseline																	R
Followup 1																	R
Diabetes
Baseline
Followup 1																	R
Hearing impairment
Baseline			R	R	R	R	R	R	R	R	R	R	R	R
Followup 1																	R
Hypertension
Baseline					R	R	R	R
Followup 1																	R
Kidney disease
Baseline																	R
Followup 1						R

Table 5-1.—Continued.

Condition	Age Groups (years)
	10-17		18-24		25-34		35-44		45-54		55-64		³65		All		All
	M	F	M	F	M	F	M	F	M	F	M	F	M	F	M	F	All
Mental retardation
Baseline																	R
Followup 1
Respiratory allergies
Baseline															R
Followup 1															R	R
Speech impairment
Baseline																	R
Followup 1										-				R
Stroke
Baseline
Followup 1																	R
Ulcers or other stomach problems
Baseline							R	R	R	R	R	R	R	R
Followup 1							R		R	R	R	R	R	R
Urinary tract disorders
Baseline	X				X	X								X
Followup 1	-			X	X	X	X	X		X		X	X	X

*Age groups at Followup 1 are 1 year older.

†-=Insufficient data.

§X=Statistically significant differences, TCA Subregistry rate higher.

¶R=Statistically significant differences, NHIS rate higher.

count, hemoglobin, and hematocrit. Stewart and coworkers (34,40) have reported the only hemolytic disease in humans exposed to high levels of TCA, as suggested by increases in urinary urobilinogen; however, since there was no clear association between TCA exposure and elevated urinary urobilinogen levels, any possible associations were discounted.

Most animal studies located were also negative with respect to finding hematological effects following exposure to TCA. No changes in hematological parameters were found following acute, intermediate, and chronic exposure to moderate to high TCA concentrations in several species of animals (28,33,37,41-47). In addition, chronic oral exposure of rats to 1,500 milligrams TCA per kilogram body weight per day (mg/kg/day) and mice to 5,615 mg/kg/day did not affect the incidence of non-neoplastic lesions in the bone marrow (48).

One study was located in which mature crossbred dogs acutely exposed to TCA showed a statistically significant decrease (p < 0.05) in neutrophils and an increase in lymphocytes (49). All types of leukocytes showed a dose-dependent decrease, based upon leukocyte differential counts. A similar response was seen in mice and guinea pigs when TCA was injected into the peritoneum (49). Finally, Maltoni et al. (50) reported an increase in the incidence of leukemia, particularly immunoblastic lymphosarcomas, in rats; however, the authors stated that the design and size of the experiment did not allow them to draw definite conclusions.

It should be noted that the reported excesses in rates were limited to females in two age groups at Baseline and to one female age group at Followup. Individuals in the latter group would have been 15 years of age or younger at exposure; however, the significance of this observation could not be determined given the limited literature. Medical confirmation of these diagnoses would provide additional information needed for more definitive conclusions to be drawn.

Arthritis

Arthritis was reported in excess in male TCA Subregistry members who were 10 through 17 years of age at the time of the Baseline interview; however, it was reported statistically significantly less for males 25 years of age or older and for females 35 years of age or older at Baseline and for all registrants at Followup. No studies could be identified regarding musculoskeletal effects following TCA exposure. Given the immunological aspects of rheumatoid arthritis, however, studies addressing the immunological effects of TCA exposure were also reviewed.

In humans, only one case of an immune system effect, dermal sensitization, was reported following exposure to TCA (51). Few studies pertaining to immunological aspects of TCA exposure in animals were found. Aranyi and coworkers (52) exposed mice to 359 parts per million (ppm) TCA for 3 hours by means of inhalation, then challenged their immune systems with Streptococcus zooepidemicus; test mice were no more susceptible to bacterial infection than were control mice. A similar study was conducted with mice exposed to radiolabeled 35S-Klebsiella pneumoniae; again, no difference was found between test and control mice. The same results were found in both tests when mice were exposed under similar conditions for 5 days (51).

Although extensive immune function tests were not conducted, several animal studies have not revealed any lesions attributable to TCA exposure upon histological evaluation of immune system tissues (including lymph nodes, thymus, and spleen) from rodents (33,37,42,46,53,54). In addition, no exposure-related nonneoplastic effects were found upon histopathological examination of the spleen and thymus after chronic exposure to £1,750 ppm in rats and 1,500 ppm in mice (41,55) and 1,500 mg/kg/day to rats and 5,615 mg/kg/day to mice (48). Chronic oral exposure of rats to 1,500 mg/kg/day and mice to 5,615 mg/kg/day did not affect the incidence of nonneoplastic lesions in the muscles or bones (48).

The decreases seen in the reporting rates for arthritis were not totally unexpected; this is one of the conditions likely to be self-diagnosed and, therefore, increased reporting by the NHIS population would be expected. Of concern is the excess reporting for males 10 through 17 years of age at Baseline. These individuals could have been exposed in utero and into their childhood years. Given the limited information concerning arthritis and any relationship to TCA exposure, however, no conclusions can be drawn at this time.

Asthma, Emphysema, or Chronic Bronchitis and Respiratory Allergies

Asthma, emphysema, and chronic bronchitis and respiratory allergies were statistically significantly reduced in the entire TCA Subregistry population at both Baseline and Followup when compared with the NHIS population. The effects of acute exposures of humans to TCA have been well documented. Some of these effects, such as generalized central nervous system (CNS) depression, cardiac arrhythmias, and aspiration of gastric contents, can produce respiratory depression or respiratory arrest, leading to death (20,24,33,56-62). In one case of solvent abuse, D'Costa (56) found edema present in the lungs of the victim; Kelly and Ruffing (63) reported acute onset eosinophilic pneumonia following abuse of TCA. It should be noted, however, that these outcomes are rare following chemical exposure (64). Finally, a case of hypoxia with atelectasis was reported following TCA exposure (65); therefore, since hypoxia can exacerbate arrhythmias (66), acute lethal exposure levels might be lower in individuals exposed during physical exertion (33,67-69). Acute exposure to lower concentrations of TCA has not been shown to affect respiratory rate or volume in humans (36,37,70), nor has TCA been shown to produce irritation of respiratory mucous membranes in humans.

Respiratory failure has been reported for animals acutely exposed via inhalation to high concentrations of TCA (53), although tissue lesions of the respiratory system have not been found in most species following exposure (28,37,41,45-47,53,71-73). It should be noted, however, that according to the National Institute for Occupational Safety and Health (NIOSH) (26), repeated exposure of animals to concentrations from 1,000 to 10,000 ppm TCA resulted in lung changes in some species. Finally, chronic oral exposure of rats to 1,500 mg/kg/day and mice to 5,615 mg/kg/day did not affect the incidence of nonneoplastic lesions in the lungs, trachea, or nasal passages (33,48).

Two studies have suggested that rapid and shallow breathing following exposure to higher TCA concentrations might be related to the increase in the activity of the lung stretch receptor (53,74). Furthermore, Kobayashi et al. (74) suggested that a part of rapid and shallow breathing in TCA-anesthetized dogs was not based on anesthetic effects but enhancement of the Hering Breuer reflex, which tends to limit both inhalation and exhalation.

Asthma, emphysema, and chronic bronchitis are conditions for which a decreased reporting by registrants would have been expected; the results confirm this. It should be noted that, although the literature reports respiratory effects following TCA exposure, the levels were extremely high-that is, several orders of magnitude beyond that to which TCA Subregistry participants might have been exposed. Given the results and the literature, it appears that respiratory effects are not related to TCA exposure at the levels and durations of exposure reported for the TCA Subregistry population.

Diabetes

Diabetes was reduced in the TCA Subregistry population overall at Followup when compared with the NHIS population. The wording of the NHIS and TCA Subregistry health condition for diabetes matched. No studies were located that discussed the occurrence of diabetes in either humans or animals following exposure to TCA.

Hearing Impairment, Speech Impairment, and Mental Retardation

Hearing impairment was statistically significantly reduced in TCA Subregistry participants 18 years of age or older at Baseline and for the TCA population overall at Followup. Mental retardation and speech impairment were both statistically significantly reduced for the TCA Subregistry population overall at Baseline; speech impairment was also statistically significantly reduced for females 66 years of age or older at Followup. The wording of the questions for mental retardation on the TCA Subregistry and NHIS questionnaires is an exact match, for hearing impairment the wording matches closely, and for speech impairment the wording is similar. Because these health conditions all relate to nervous system effects, they will all be considered together with respect to the current literature.

As mentioned previously, CNS depression has been reported as a primary effect in humans after exposure to high levels of TCA (24,75-80). Dizziness, lightheadedness, and mild motor impairment are caused by exposure to moderate concentrations (34,35,37). According to ATSDR, (22) at high concentrations TCA causes euphoric narcosis followed by stupor, coma, and seizures. With chronic exposure, agitation and lethargy are apparent during sedentary periods. Decreased memory and sleep disturbances have been reported for chronic, low-level exposures. As mentioned previously, CNS depression following exposure might result in death due to respiratory arrest.

Mackay et al. (81) conducted an exposure chamber study in which volunteers were exposed to 0, 175, and 350 ppm TCA for 3.5 hours. These exposures resulted in a pattern of psychomotor performance deficits; however, syntactic reasoning was not affected and distractibility improved, suggesting that impairment by TCA might be task specific (33). Gamberale and Hultgren (82) have shown similar decrements in psychomotor performance on a number of tasks including simple and choice reaction time. TCA exposure has also been discovered in cases of solvent abuse. D'Costa (56) reported a case of a death due to inhalation of Tipp-Ex®, which contains TCA. Postmortem examination showed a grossly edematous brain, also reported in other cases (58), and a marked tonsillar herniation and uncal grooving.

In an occupational study, Kelafant and coworkers (32) examined 28 workers with long-term repetitive high exposures to TCA and found that, as a group, they had statistically significant deficits in memory, intermediate memory, rhythm, and speed. In addition, deficits in vestibular, somatosensory, and ocular components of balance were reported. According to NIOSH (26), exposure to concentrations in excess of 1,000 ppm for 15 minutes, or 2,000 ppm for 5 minutes, can be expected to produce a disturbance of equilibrium in the majority of adults.

House et al. (75) described a case of peripheral sensory neuropathy in a woman who had daily exposure to TCA used as a degreasing agent. The woman also experienced reduced amplitudes of sural sensory responses bilaterally and normal nerve conduction velocities and sensory latencies consistent with a toxic axonopathy. This case was similar to two described by Liss (83) in which distal sensory peripheral neuropathy was experienced by two women who also used TCA as a degreaser. In both cases, there were abnormalities of nerve conduction of the upper extremities, and sural nerve biopsies showed chronic neuropathy (axonopathy and myelonopathy). Howse et al. (84) also reported peripheral sensory effects in workers in the same plant as the two workers described by Liss (83). It should be noted that the cases reported by Liss (83) and House et al. (75) had facial sensory changes, in addition to peripheral neuropathy.

TCA is considered to be one of the least toxic solvents with respect to neurotoxicity (85-86) and is often reported to have no CNS effects (35,81,87-89). Indeed, there have been no reports of irreversible neurological impairment in humans (33). For example, Stewart et al. (35) were unable to show any impairment in a series of psychomotor tasks following several days exposure to 500 ppm of TCA. Similarly, Salvini et al. (88) failed to show any behavioral effects after two periods of exposure to 450 ppm for 4.5 hours with a 1.5-hour break in between. Maroni et al. (90) found no exposure-related neurological effects in workers exposed to up to 990 ppm TCA for an average of 6.7 years. Kramer et al. (39) compared exposed and non-exposed individuals in adjacent textile plants and could find no statistically significant differences in symptoms, physical examination, clinical chemistry, or electrocardiogram results.

CNS depression is also the predominant effect of TCA in animals. Signs include ataxia, unconsciousness, and death at increasing concentrations (71,91-94). Widely reported behavioral changes in animals include impaired performance of neurobehavioral tests and increased motor activity (43,95-103). TCA has been found in several animal models to reduce response rates in operant behavior of different reinforcing schedules (80,95,96,98). Neurophysiological changes such as intoxication and incoordination in rats and mice (53,91,93,94); ataxia, followed by unconsciousness and death due to respiratory failure in mice (103); a progression from ataxia to lethargy, loss of motor function, and prostration in rats, mice, dogs, and monkeys (37,53,71,92,94); and deficit in forelimb grip performance (86,97) have also been reported from animal studies.

Animal studies indicate TCA exposure can result in biochemical alterations in the CNS. For example, lasting physical changes to the brain, as indicated by increased glial fibrillary acid (GFA) protein and decreased deoxyribonucleic acid content, have been reported in the brains of gerbils after intermediate exposure to low levels of TCA (104,105). This GFA protein is considered to be a marker for astrogliosis, which is associated with brain injury. Another astroglial protein, S-100, has also been used to track chemically induced changes in the astroglia; brain weights of animals exposed to 1,000 ppm were found to have statistically significantly less S-100 protein in the frontal cerebral cortex and less total protein in the occipital cerebral cortex (31). Although this evidence of neurotoxicity has been questioned (106), the investigators maintain that TCA is a neurotoxicant (75,107).

Other animal studies have also provided evidence of possible neurotoxicity following exposure to TCA. For example, following inhalation exposure to TCA there was decreased cyclic guanosine monophosphate content in the brain stem and cerebral cortex in mice, altered levels of cyclic nucleotides, and changes in brain metabolism in rats and mice (108-110); decreased glucose consumption and blood flow have also been documented in the brains of rats (93). Conversely, Savolainen and coworkers (111) found no effect on the levels of protein, glutathione, acid proteinase, or ribonucleic acid (RNA) in the brains of rats acutely exposed to low concentrations of TCA. Continuous exposure to 1,200 ppm but not lower doses of TCA for 30 days altered the fatty acid composition of ethanolamine phosphoglyceride isolated from the cerebral cortex in male Sprague-Dawley rats (29,33,112).

With respect to mental retardation, in utero exposures must also be considered. Recently, concern has been expressed about the potential for TCA to have an adverse effect on pregnancy outcomes in humans. This concern arose primarily from an epidemiologic study conducted by the California Department of Health Services in which statistically significant excesses of spontaneous abortions and birth defects were reported for persons living in an area that water company personnel believed received TCA-contaminated water (113-115). Wrensch et al. (116) reanalyzed the information, however, and found the data did not support the hypothesis that adverse pregnancy outcomes were associated with exposure to TCA in drinking water.

The results of several animal studies have shown that TCA is not a reproductive toxicant or teratogen even at doses of 30 ppm in drinking water (25,117-120). In a study of rats exposed to high levels of TCA via inhalation, York et al. (121) observed some embryotoxic and fetotoxic effects; however, these effects were attributed by the authors of the study to maternal toxicity.

All of these conditions-hearing impairment, speech impairment, and mental retardation-are sometimes self-diagnosed; therefore, the TCA Subregistry qualifier "has a physician told you you have or treated you for (condition)" would tend to cause NHIS rates, which were created without the qualifier, to be higher. This was indeed the case. The literature reports nervous system effects from TCA exposure; however, these effects were not specific to the conditions of concern and were usually reported following exposure to extremely high levels of TCA. Additional information, such as trend analysis from further followups, will be of interest.

Hypertension

Reports of hypertension were statistically significantly reduced in the TCA Subregistry population 25 through 44 years of age at Baseline and in the population overall at Followup when compared with the NHIS population. The wording of this question was a very close match with the wording of the NHIS question.

TCA has been demonstrated to lower blood pressure in humans following high (10,000 to 26,000 ppm) exposures (70); however, the effects were not permanent and subsided shortly after exposure ceased. Daily exposure to low levels for up to 6 years did not affect blood pressure, heart rate, or electrocardiogram results in humans (37,39,82). In humans, cardiac dysrhythmias have also been reported following exposure to high levels of TCA (such as occurs with solvent abusers); tachycardia has been noted in exposures to low doses, and bradycardia has been observed in exposures to high doses (22).

TCA has been shown to produce a dose-dependent drop in blood pressure, stroke output, and myocardial contractility in dogs (73,122); the mechanism by which this occurs appears to involve cardiac depression and peripheral vasodilation (73). Animal studies have also revealed that fatal arrhythmias might result from sensitization of the heart to epinephrine (123-125).

Honma (126) studied the effect of TCA exposure on the plasma contents of lipoproteins in rats (male Fischer 344). He found that levels of very low density lipoproteins (VLDL) and low density lipoproteins (LDL) were increased by exposure to TCA at 100 to 300 mg/kg; however, the high density lipoprotein (HDL) levels rose at 100 mg/kg but fell at 1,000 mg/kg. Honma believed the decreases in HDL at high doses of TCA resulted from the inhibition of HDL synthesis.

Given the results of the TCA Subregistry and the information in the literature, it is possible that the reduction in reporting of hypertension among TCA Subregistry members might be due to the hypotensive effect of TCA. This could not be demonstrated with the TCA Subregistry information per se, but should be considered as a hypothesis generated by the Subregistry for further testing by an appropriately constructed epidemiological study.

Kidney Disease

The wording of the NHIS and TCA Subregistry question for kidney disease was a close match. The comparable question in the NHIS questionnaire queried respondents about kidney stones, infections, and other kidney trouble. Kidney problems were reported statistically significantly fewer times in the entire TCA Subregistry population at Baseline and in females 26 through 35 years of age at Followup.

Kidney dysfunction has been reported in humans following acute exposures to TCA (22). Transient renal damage, as evidenced by microscopic hematuria and proteinuria and sometimes accompanied by an increase in blood urea nitrogen (BUN) or serum creatinine, has also been reported (22,56,127).

Other studies have found no evidence of nephrotoxicity, although the end points examined (such as BUN) are adequate only for detecting serious decrements in function. Acute exposure to TCA had no effect on BUN or uric acid levels in humans exposed to high or moderate concentrations (35-36,38,40). In a study of workers exposed to up to 250 ppm TCA for 1 to 3 years there was no evidence of any impairment in kidney function, as determined by BUN, uric acid, or other serum indicators of nephrotoxicity (39).

In animals, exposure of varying durations to TCA did not produce kidney lesions (28,37,41,42,45-47,53,54,71,72,128). It should be noted, however, that relative kidney weight was increased slightly at 12,000 ppm in the one study in which it was measured (53). No effects on kidney weights or histology were found in rats given a single gavage dose of 4,000 mg/kg/day, repeated doses of 10,000 mg/kg/day, or intermediate-duration exposure to 5,000 mg/kg/day; however, there was a slight transient increase in BUN in the rats repeatedly given 10,000 mg/kg/day (129). The hematology values, urinalysis values, organ weights, and relative organ weights (g/100 g body weight) for male and female rats and mice at the various time intervals throughout the study were unaffected by treatment (41).

Given the negative reports found in the literature and the reduced rates of kidney disease reported for the TCA Subregistry population, it is unlikely that exposure to TCA would result in kidney disease.

Ulcers or Other Stomach Problems

The wording for the TCA Subregistry and NHIS questions for ulcers or other stomach problems were categorized as Class B; that is, the questions were similar but not closely or exactly the same. Ulcers or other stomach problems were found at a statistically significantly higher rate in the NHIS population than in the TCA Subregistry population for males and females 35 years of age and older at Baseline and for males 36 years of age or older and females 46 years of age or older at Followup.

Nausea, vomiting, and diarrhea reportedly occur in humans, but not animals, following acute oral or inhalation exposure to high TCA levels (40,59,130,131). Other gastrointestinal end points, however, have not been examined in humans (33). In a cross-sectional study comparing workers exposed to up to 350 ppm TCA with an unexposed population, Kramer et al. (39) reported that a greater proportion of exposed than unexposed workers (28 versus 21) experienced "abnormalities of the gastrointestinal system"; however, the authors felt this finding related more to liver disease. Finally, in a death caused by solvent abuse, postmortem examination showed edema present in the gut (56).

Gastrointestinal effects have not been reported in animals exposed by any route (28,33,37,41,42,45-47,53,54,105,112,128,132-134). In addition, chronic oral exposure of rats to 1,500 mg/kg/day and mice to 5,615 mg/kg/day did not affect the incidence of nonneoplastic lesions in the stomach, intestines, and pancreas (48). It should be noted that reductions in body weight gain have been reported in several animal studies following exposure to TCA (41,46,48,50,53,128, 129,135).

Again, the reduced reporting in the TCA Subregistry population and the preponderance of negative reports in the literature for ulcers or related problems following TCA exposure make it unlikely that this outcome is related to such exposure.

Urinary Tract Disorders, Including Prostate Trouble

Urinary tract disorders were reported statistically significantly more often in the TCA Subregistry population, specifically in males 10 through 17 years and 25 through 34 years of age and females 25 through 34 years of age and 65 years of age or older at Baseline. At Followup, females 19 years of age or older and males 26 through 45 years of age and 66 years of age or older reported statistically significantly higher rates of urinary tract disorders when compared with the NHIS population. The wording of the TCA Subregistry and NHIS questions for the health condition urinary tract disorders, including prostate trouble, is not an exact match (see Table 3-4 and Appendices A and B).

No studies were found that investigated the potential effect of TCA exposure on the urinary tract of humans. Since urinary tract problems can be caused by infections, the reader is referred to the information in the previous section on arthritis that discusses what is known about the effect of TCA on the immune system. As mentioned previously, studies of the effect of TCA on the human immune system have not been identified; only one case of dermal hypersensitivity was reported for humans (51).

The information on the effect of TCA on the urinary tract in animals is also limited. Quast and coworkers (41) reported that male and female rats exposed to TCA were normal with respect to urinalysis results.

It should be noted that the NHIS question was specific for bladder disorders, while the TCA questionnaire included the broader category of all urinary tract disorders, which might account for a portion of the reported cases in the TCA Subregistry population. Given the sparsity of information pertaining to the relationship between TCA exposure and urinary problems, additional research is needed.

SUMMARY

This section has reviewed the most current scientific literature for TCA exposure and health effects. The literature suggests that there might be associations between TCA exposure and some of the health outcomes that have been reported as excess in the Subregistry data. However, the reported literature has many limitations-case reports of occupational studies usually involve only males; exposure levels in occupational exposures are much higher than those reported in environmental exposures; and high-dose animal studies might not be relevant to humans. In addition, human health studies often lack sufficient exposure characterization, lack controls for important confounding factors, and do not include 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 TCA exposure and human health outcomes among the TCA Subregistry population.

SECTION 6 CONCLUSIONS AND RECOMMENDATIONS

CONCLUSIONS

The Trichloroethane (TCA) Subregistry is a database containing self-reported health information for 3,665 persons (3,204 living and 461 deceased) at Baseline and 3,469 (2,963 living and 510 deceased) at the first Followup who had been exposed to TCA through the drinking water used in their homes. Using the available environmental data, probable start dates for exposure were established; the end dates were known. The registrants of Vestal, New York, had been potentially exposed to a maximum level of 1,600 parts per billion (ppb) for up to 12 years.

As discussed in Section 2, the environmental data available were collected and evaluated; however, their limitations merit reiteration. Historically, the samples were not taken for the purpose of quantifying individuals' exposures over time. Rather, they were taken for the purpose of verifying the level of contamination at a past point in time. Regardless, the data serve the National Exposure Registries' (NER) purpose in that they establish that exposure to TCA occurred, in this case at site addresses (that is, individual residences) through contaminated water. Most of the environmental data represent multiple samples of the source well, taken for the purpose of quantifying the level of contamination. These samples might or might not have been taken when TCA and other chemical exposures were at the maximum level ever experienced.

Dose calculations will be considered in the future for the TCA Subregistry population. Recently, several computer models (136-138) have become available with which exposure patterns can be simulated with a minimum amount of information. Perhaps these models could be used in conjunction with statistical models to develop a best estimate of actual environmental levels for all exposure periods and sites. If this is feasible, individual doses might be reconstructed for all routes of exposure for further use in dose-response evaluations.

Cause-and-effect relationships between TCA exposure and reported health outcomes cannot be established using the TCA Subregistry data; however, potential associations can be identified. There are biases and limitations of the Subregistry data files that should be considered in assessing the validity of the suggested associations. A bias in reporting might have occurred because of the registrants' awareness of their exposure and possible resulting actions-for example, seeking medical care more often. To moderate this possible reporting bias, the qualifier that a health care provider must have told the registrant about or treated the registrant for the reported health outcome was required (the qualifier was not required of the comparison population). The wording for the health questions (potentially two) of the Agency for Toxic Substances and Disease Registry (ATSDR) and the National Health Interview Survey (NHIS) data collection instruments was different. Finally, although the significant level for the statistical analyses was low (α = 0.01) and the sample size was large, given the large number of comparisons made, there was a potential for false positives.

The TCA Subregistry population reported positively more often for some health outcomes, less often for others, than the NHIS population. Statistically significantly increased reporting was found for TCA Subregistry members in most cases for select age and sex groups; for the health conditions anemia and other blood disorders (Baseline and Followup); arthritis (Baseline); and urinary tract disorders (Baseline and Followup).

Statistically significantly decreased reporting by the TCA Subregistry members was found for the conditions diabetes, speech impairment, hearing impairment, hypertension, stomach problems, asthma and emphysema, respiratory allergies, kidney disease, and arthritis. This confirmed expectations, given that some of the conditions are often self-diagnosed and would therefore be most affected by the limitations of health care provider confirmation of their existence.

Several of the findings were of particular note, including the excess reporting of arthritis in children. Confirmation of these cases of arthritis in the TCA Subregistry population is an important first step for future research. In addition, the follow-up data for the TCA Subregistry population should prove to be useful for clarifying the Baseline and first Followup information.

Also of note were the excess numbers of anemia or other blood disorders. The findings of the Subregistry suggest that additional followup of registrants and verification of these conditions would be worthwhile. More information from registrants on potentially confounding factors, such as diet, specific health risk behaviors (for example, smoking and alcohol consumption), and other lifestyle factors will help identify the contributing factors.

RECOMMENDATIONS

Information obtained from this database can and will be used to suggest possible TCA exposure-health outcome associations and to determine appropriate future activities and research. Considerations for additional research using this database include modification of existing data collection procedures and methods; exploration of recognized sources of bias and reduction or elimination of these biases; acquisition of additional definitive information on and confirmation of selected outcomes that appear to be in excess; and substance-specific research with specific hypotheses clearly identified.

It is uncertain whether wording differences between TCA Subregistry and NHIS questions or recall bias might have generated different response rates. To better understand the excess reporting of some of the health conditions, ATSDR will conduct an evaluation of how these wording differences could have created differences in responses and will modify the questions appropriately for future NER activities. Future research will be conducted to aid in assessing the potential impact of exposure awareness on both recall of health conditions and health care utilization by registrants. It should be noted that even if a determination is made that members of the exposed population sought health care more often than members of the comparison population, it will not be possible to separate whether the increase in number of visits to health care providers was the result of an awareness of exposure or an actual increase in health problems.

Improvement of individual exposure estimates for the TCA Subregistry population and other environmentally exposed populations is a top priority for ATSDR. Better exposure characterization of registrants will allow improved assessment of the relationship between exposure and health outcomes.

Preliminary findings from the completed analyses leave many questions unanswered; it is hoped that additional information and further investigations will help answer these questions. Routine longitudinal followup of the TCA Subregistry population will continue. The additional information will be used to assess trends in reporting and to assist in answering or clarifying some of the issues and questions previously discussed.

It should be noted that all of the health conditions reported in excess by the TCA Subregistry population are preventable or treatable, if not curable. Early detection frequently plays a role in whether a health condition can be arrested or reversed by existing medical technologies. Given the results of the Baseline analyses for this population, it was imperative that this information be shared with the TCA Subregistry population in a responsible manner.

A summary version of the TCA Subregistry Baseline and Followup 1 Technical Report was composed in nontechnical language for release to the registrants. A Subregistry Technical Assistance Panel (STAP), composed of representative membership from the states involved, other federal agencies, and other knowledgeable persons, reviewed the TCA Subregistry registrant report and gave ATSDR suggestions and recommendations about future activities that would be appropriate for followup to release of the information.

Following review of the registrant report by the STAP, community representatives and interested county, state, and federal agency representatives met to review the report and to assist ATSDR with planning site-specific activities and with development of a one-page fact sheet summarizing the report. Immediately following those meetings, information packets containing the registrant report, fact sheet, and cover letter were mailed to all TCA Subregistry members. Similar information was mailed to the media shortly after the registrants received their information packets.

A public meeting was held in Vestal, New York. The public meeting was conducted to review the purpose of the TCA Subregistry and the information provided to the registrants. Registrants were encouraged to ask questions of ATSDR, either in person or by telephone.

Registrants expressing specific health concerns were encouraged to see their personal physicians. A concerted effort was made by ATSDR to share the findings of the Baseline and first Followup TCA Subregistry analyses with health care providers in the area. In addition, specific information about TCA and training about health effects associated with TCA exposure will be made available to them.

Although no definitive conclusions can be drawn from the information contained in this report, it is hoped that researchers will use the information to better assess the potential for adverse health outcomes following exposure to substances in the environment. As has been pointed out, additional studies of the TCA Subregistry data are warranted and encouraged by ATSDR.

REFERENCES

Agency for Toxic Substances and Disease Registry. National Exposure Registry policies and procedures manual (revised). Atlanta: US Department of Health and Human Services, Public Health Service, 1994.
Agency for Toxic Substances and Disease Registry. Hazardous Substances Priority List, Toxicological Profiles, Second List. Federal Register 1988;53:41280.
Agency for Toxic Substances and Disease Registry. Toxicological profile for 1,1,1-trichloroethane. Atlanta: US Department of Health and Human Services, Public Health Service, 1990 Dec. Report No.: TP/91/180463/AS.
Agency for Toxic Substances and Disease Registry. ATSDR HazDat Version 1.5. Atlanta: US Department of Health and Human Services, Public Health Service; 1992 Apr.
Comprehensive Environmental Response, Compensation, and Liability Act of 1980, Pub. L. 95-510 (1980 Dec 11).
The Superfund Amendments and Reauthorization Act of 1986, Pub. L. 99-499.
Agency for Toxic Substances and Disease Registry. Toxicological profile for benzene. Atlanta: US Department of Health and Human Services, Public Health Service, 1989 May. Report No.: TP-88-03.
National Center for Health Statistics. Current estimates from the National Health Interview Survey, United States, 1989. Vital and health statistics; series 10, no. 164. Washington, DC: US Department of Health and Human Services, Public Health Service, 1990. Publication No.: (PHS)90-1504.
National Center for Health Statistics. Public Use Data Tape Documentation, Part I, National Health Interview Survey, 1989 (Machine readable data file documentation). Springfield (VA): US Department of Commerce, National Technical Information Service, 1990.
Sweet JA, Bumpass LL. American families and households. New York: Russell Sage, 1987.
Aday LA, Andersen R. The national profile of access to medical care: where do we stand? Am J Public Health 1985;74:12.
Alcohol, Drug Abuse, and Mental Health Administration. National household survey on drug abuse 1990 population estimates. Washington, DC: US Department of Health and Human Services, Public Health Service, 1991. DHHS Publication No.: (ADM)91-1732.
Baker RJ, Nelder JA. Generalized linear interactive modeling. Release 3.77. Oxford: Royal Statistical Society, 1978.
World Health Organization. Internal classification of diseases: manual of the international statistical classification of diseases, injuries, and causes of death. Geneva: World Health Organization, 1977.
Lipscomb JA, Satin KP, Neutra RR. Reported symptom prevalence rates from comparison populations in community-based environmental studies. Arch Environ Health 1992;47:4.
Breslow NE, Day NE. Statistical methods in cancer research, vol. 2. The design and analysis of cohort studies. Lyon: International Agency for Research on Cancer, 1987.
National Cancer Institute. Surveillance, epidemiology and end results program (SEER) data. Bethesda (MD): National Cancer Institute, 1989.
National Center for Health Statistics. Current estimates from the National Health Interview Survey, United States, 1991. Vital and health statistics; series 10, no. 184. Washington, DC: US Department of Health and Human Services. Public Health Service, 1990. Publication No.: (PHS) 93-1512.
Aitio A, Jarvisalo J, Riihimaki V, Hernberg S. Biologic monitoring. In: Zenz C, editor. Occupational medicine: principles and practice. Chicago: Year Book, 1988:190.
Stewart RD. The toxicology of methyl chloroform. J Occup Med 1963;5:259-62.
Wilcosky TC, Simonsen NR. Solvent exposure and cardiovascular disease. Am J Ind Med 1991;19:569-86.
Agency for Toxic Substances and Disease Registry. Case studies in environmental medicine: 1,1,1-trichloroethane toxicity. Atlanta: US Department of Health and Human Services, Public Health Service, 1993.
Holmberg B, Jakobson I, Sigvardsson K. A study on the distribution of methylchloroform and n-octane in the mouse during and after inhalation. Scand J Work Environ Health 1977;3:43-52.
Stewart RD. The toxicology of 1,1,1-trichloroethane. Ann Occup Hyg 1968;11:71-9.
George JD, Price CJ, Marr MC, Sadler BM, Schwetz BA, Birnbaum LS, et al. Developmental toxicity of 1,1,1-trichloroethane in CD rats. Fundam Appl Toxicol 1989;13:641-51.
National Institute for Occupational Safety and Health (NIOSH). Occupational health guideline for methyl chloroform. Cincinnatti (OH): US Department of Health and Human Services, Public Health Service, September 1978.
Hake CL, Waggoner TB, Robertson DN, Rowe VK. The metabolism of 1,1,1-trichloroethane by the rat. Arch Environ Health 1960;1:101-5.
Eben A, Kimmerle G. Metabolism, excretion and toxicology of methylchloroform in acute and subacute exposed rats. Arch Toxicol 1974;31:233-42.
Kyrklund T, Haglid KG. Exposure of rats to high concentrations of 1,1,1-trichloroethane and its effects on brain lipid and fatty acid composition.Pharmacol Toxicol 1990;67:384-6.
Hoffmann P, Breitenstein M, Toraason M. Calcium transients in isolated cardiac myocytes are altered by 1,1,1-trichloroethane. J Mol Cell Cardiol 1992;24:619-29.
Bogen KT, Hall LC. Pharmacokinetics for regulatory risk analysis: the case of 1,1,1-trichloroethane (methyl chloroform). Regul Toxicol Pharmacol 1989;10:26-50.
Kelafant GA, Berg RA, Schleenbaker R. Toxic encephalopathy due to 1,1,1-trichloroethane exposure. Am J Ind Med 1994;25:439-46.
Agency for Toxic Substances and Disease Registry. Toxicological profile for 1,1,1-trichloroethane, update. Draft for public comment. Atlanta: US Department of Health and Human Services, Public Health Service, October 1993.
Stewart RD, Gay HH, Erley DS, Hake CL, Schaffer AW. Human exposure to 1,1,1-trichloroethane vapor: relationship of expired air and blood concentrations to exposure and toxicity. Am Ind Hyg Assoc J 1961;22:252-62.
Stewart RD, Gay HH, Schaffer AW, Erley DS, Rowe VK. Experimental human exposure to methyl chloroform. Arch Environ Health 1969;19:467-72.
Stewart RD, Hake CL, Wu A. 1,1,1-Trichloroethane: development of a biologic standard for the industrial worker by breath analysis. Cincinatti (OH): National Institute for Occupational Safety and Health, US Department of Health and Human Services, Public Health Service, 1975. NTIS# PB82-151879.
Torkelson TR, Oyen F, McCollister DD, Rowe VK. Toxicity of 1,1,1-trichloroethane as determined on laboratory animals and human subjects. Am Ind Hyg Assoc J 1958;19:353-62.
Wright MF, Strobl DJ. 1,1,1-Trichloroethane cardiac toxicity: report of a case. J Am Osteopath Assoc 1984;84:285-8.
Kramer CG, Ott MG, Fulkerson JE, Hicks N, Imbus HR. Health of workers exposed to 1,1,1-trichloroethane: a matched-pair study. Arch Environ Health 1978;33:331-42.
Stewart RD. Methyl chloroform intoxication: diagnosis and treatment. JAMA 1971;215:1789-92.
Quast JF, Calhoun LL, Frauson LE. 1,1,1-Trichloroethane formulation: a chronic inhalation toxicity and oncogenicity study in Fischer 344 rats and B6C3F1 mice. Fundam Appl Toxicol 1988;11:611-25.
Calhoun LL, Quast FJ, Schumann AM. Chloroethene VG. Preliminary studies to establish exposure concentrations for a chronic inhalation study with rats and mice. Midland (MI): Toxicology Research Laboratory, Health and Environmental Sciences, Dow Chemical Company, 1981.
Horiguchi S, Horiguchi K. An experiment of 1,1,1-trichloroethane vapor exposure to mice. Japanese Journal of Industrial Health 1971;13:226-7.
Koizumi A, Kumai M, Ikeda M. Dose-dependent induction and suppression of liver mixed-function oxidase system in chlorinated hydrocarbon solvent metabolism. J Appl Toxicol 1983;3:208-17.
MacEwen JC, Vernot EH. The biological effect of continuous inhalation exposure of 1,1,1-trichloroethane (methyl chloroform) on animals. Toxic Hazards Research Unit Annual Technical Report. Wright-Patterson Air Force Base (OH): Harry G. Armstrong Aerospace Medical Research Laboratory, 1974;81-90.
Prendergast JA, Jones RA, Jenkins LJ, Siegel J. Effects on experimental animals of long-term inhalation of trichloroethylene, carbon tetrachloride, 1,1,1-trichloroethane, dichlorofluoromethane, and 1,1-dichloroethylene.Toxicol Appl Pharmacol 1967;10:270-89.
Truffert L, Girard-Wallon C, Emmerich E, Neauport C, Ripault J. Early experimental demonstration of the hepatotoxicity of some chlorinated solvents by the study of the synthesis of hepatic DNA. Archives des Maladies Professionnelles de Medecine du Travail et de Securite Sociale 1977;38:261-3.
National Cancer Institute. Bioassay of 1,1,1-trichloroethane for possible carcinogenicity. National Cancer Institute Carcinogenesis Technical Report Series 3, 1977. NTIS PB265082.
Hobara T, Kobayashi H, Higashihara E, Kawamoto T, Sakai T. Acute effects of 1,1,1-trichloroethane, trichloroethylene, and toluene on the hematologic parameters in dogs. Arch Environ Toxicol 1984;13:589-93.
Maltoni C, Cotti G, Patella V. Results of long-term carcinogenicity bioassays on Sprague-Dawley rats of methyl chloroform, administered by ingestion. Acta Oncol 1986;7:101-17.
Ingber A. Occupational allergic contact dermatitid from methyl chloroform (1,1,1-trichloroethane). Contact Dermatitis 1991;25:193.
Aranyi C, O'Shea WJ, Graham JA, Miller FJ. The effects of inhalation of organic chemical air contaminants on murine lung host defenses. Fundam Appl Toxicol 1986;6:713-20.
Adams EM, Spencer HC, Rowe VK, Irish DD. Vapor toxicity of 1,1,1-trichloroethane (methylchloroform) determined by experiments on laboratory animals. Archives of Industrial Hygiene and Occupational Medicine 1950;1:225-36.
Kjellstrand P, Bjerkemo M, Adler-Maihofer M. Effects of solvent exposure on testosterone levels and butyrylcholinesterase activity in mice. Acta Pharmacol Toxicol (Copenh) 1985;57:242-9.
Quast JF, Rampy LW, Balmer MG, Leong BJL, Gehring PJ. Toxicologic carcinogenic evaluation of a 1,1,1-trichloroethane formulation by chronic inhalation in rats. Midland (MI): Toxicology Research Laboratory, Health and Environmental Research, Dow Chemical Company, 1978.
D'Costa DF. Fatal cerebral oedema following trichloroethane abuse. J R Soc Med 1990;83:533-4.
Guberan E, Fryc O, Robert M. Sudden death from ventricular fibrillation after voluntary inhalation of chloroethene in a mechanics apprentice. Schweiz Med Wochenschr 1976;106:119-21.
Hall FB, Hine CH. Trichloroethane intoxication: a report of two cases. J Forensic Sci 1966;11:404-13.
Jones RD, Winter DP. Two case reports of deaths on industrial premises attributed to 1,1,1-trichloroethane. Arch Environ Health 1983;38:59-61.
MacDougall IC, Isles C, Oliver JS, Clark JC, Spilg WGS. Fatal outcome following inhalation of Tipp-Ex®. Scott Med J 1987;32:55.
Stahl CJ, Fatteh AV, Dominguez AM. Trichloroethane poisoning: observations on the pathology and toxicology in six fatal cases. J Forensic Sci 1969;14:393-7.
Travers H. Death from 1,1,1-trichloroethane abuse: case report. Mil Med 1974;139:889-90.
Kelly KJ, Ruffing R. Acute eosinophilic pneumonia following intentional inhalation of Scotchguard. Ann Allergy 1993;71:358-61.
Mahatma M, Haponik EF, Nelson S, Lopez A, Summer WR. Phenytoin-induced acute respiratory failure with pulmonary eosinophilia. Am J Med 1989;87:93-4.
Woo OF, Healey KM, Sheppard D, Tong TG. Chest pain and hypoxemia from inhalation of a trichloroethane aerosol product. J Toxicol Clin Toxicol 1983;20:333-41.
Reinhardt CF, Azar A, Maxfield ME, Smith PE Jr, Mullin LS. Cardiac arrhythmias and aerosol "sniffing." Arch Environ Health 1971;22:265-79.
King GS, Smialek JE, Troutman WG. Sudden death in adolescents resulting from inhalation of typewriter correction fluid. JAMA 1985;253:1604-6.
Ranson DL, Berry PJ. 1986. Death associated with the abuse of typewriter correction fluid. Med Sci Law 1986;26:308-10.
Troutman WG. Additional deaths associated with the intentional inhalation of typewriter correction fluid. Vet Hum Toxicol 1988;30:130-2.
Dornette WHL, Jones JP. Clinical experiences with 1,1,1-trichloroethane: a preliminary report of 50 anesthetic administrations. Anesth Analg 1960;39:249-52.
Bonnet P, Francin JM, Gradiski D, Raoult G, Zissu D. Determination of the median lethal concentration of principle chlorinated aliphatic hydrocarbons in the rat. Archives des Maladies Professionnelles de Medecine du Travail et de Securite Sociale 1981;41:317-21.
Cornish HH, Adefuin J. Ethanol potentiation of halogenated aliphatic solvent toxicity. Am Ind Hyg Assoc J 1966;27:57-61.
Herd PA, Lipsky M, Martin HF. Cardiovascular effects of 1,1,1-trichloroethane. Arch Environ Health 1974;28:227-33.
Kobayashi H, Hobara T, Satoh T, Sakai T. Respiratory disorders following 1,1,1-trichloroethane inhalation: a role of reflex mechanism arising from lungs. Arch Environ Health 1986;41:149-54.
House RA, Liss GM, Wills MC. Peripheral sensory neuropathy associated with 1,1,1-trichloroethane. Arch Environ Health 1994;49:196-9.
US Environmental Protection Agency. Health assessment document for 1,1,1-trichloroethane (methyl chloroform): final report. Washington, DC: US Environmental Protection Agency, 1984. EPA-600/8-82-00JF.
Health and Safety Executive Toxicity Review. 1,1,1-Trichloroethane. London: Her Majesty's Stationery Office, 1984.
Kleinfeld M, Feiner B. Health hazards associated with work in confined spaces. J Occup Med 1966;8:358-64.
Torkelson TR, Rowe VK. Halogenated alpihatic hydrocarbons containing chlorine, bromine, and iodine. In: Clayton GD, Clayton E, editors. Patty's industrial hygiene and toxicology, third edition. New York: Wiley, 1981:3502-10.
You L, Muralidhara S, Dallas CE. Comparisons between operant response and 1,1,1-trichloroethane toxicokinetics in mouse blood and brain. Toxicology 1994;93:151-63.
Mackay CJ, Campbell L, Samuel AM, Alderman KJ, Idzikowski C, Wilson HK, et al. Behavioral changes during exposure to 1,1,1-trichloroethane: time-course and relationship to blood solvent levels. Am J Ind Med 1987;11:223-40.
Gamberale F, Hultengren M. Methylchloroform exposure. II. Psychological functions. Scand J Work Environ Health 1973;10:82-92.
Liss G. Peripheral neuropathy in two workers exposed to 1,1,1-trichloroethane. JAMA 1988;260(15):2217.
Howse DC, Shanks GL, Nag S. Peripheral neuropathy following prolonged exposure to methyl chloroform. Neurology 1989;39(Suppl 1):242.
Hartman DE. Neuropsychological toxicology. New York: Pergammon Press, 1988;153-4.
Mattsson JL, Albee RR, Lomax LG, Beekman MJ, Spencer PJ. Neurotoxicologic examination of rats exposed to 1,1,1-trichloroethane vapor for 13 weeks. Neurotoxicol Teratol 1993;15:313-26.
Rowe VK, Wujkowski T, Wolf MA, Sadek SE, Stewart RD. Toxicity of a solvent mixture of 1,1,1-trichloroethane and trichloroethylene as determined by experiments on laboratory animals and humans. Am Ind Hyg Assoc J 1963;24:541-54.
Salvini M, Binaschi S, Riva M. Evaluation of the psychophysiological functions in humans exposed to the threshold limit value of 1,1,1-trichloroethane. Br J Ind Med 1971;28:256-92.
Savolainen H, Riihimaki V, Laine A, Kekoni J. Short term exposure of human subjects to m-xylene and 1,1,1-trichloroethane. Int Arch Occup Environ Health 1981;49:89-98.
Maroni M, Bulgheroni C, Cassitto MG, Merluzzi F, Gilioli R, Foa V. A clinical and neurophysiological and behavioral study of female workers exposed to 1,1,1-trichloroethane. Scand J Work Environ Health 1977;3:16-22.
Clark DG, Tinston DJ. Acute inhalation toxicity of some halogenated and nonhalogenated hydrocarbons. Human Toxicology 1982;1:239-47.
Gehring PJ. Hepatotoxic potentcy of various chlorinated hydrocarbon vapours relative to their narcotic and lethal potencies in mice. Toxicol Appl Pharmacol 1968;13:287-98.
Hougaard K, Ingvar M, Wieloch T, Siesjo BK. Cerebral metabolic and circulatory effects of 1,1,1-trichloroethane, a neurotoxic industrial solvent. I. Effects on local cerebral glucose consumption and blood flow during acute exposure. Neurochemical Pathology 1984;2:39-53.
Lazarew NW. 1929. Concerning the strength of the narcotic effects of the vapors of the chlorine derivatives of the methanes, ethanes, and ethylenes. Archiv fuer Experimentelle Pathologie und Pharmakologie 1929;141:19-24.
Balster RL, Moser VC, Woolverton WL. Concurrent measurement of solvent vapor concentrations and effects on operant behavior using a dynamic exposure system. Journal of Pharmacological Methods 1982;8:299-309.
DeCeaurriz J, Desiles JP, Bonnet P, Marignac B, Muller J, Guenier JP. Concentration-dependent behavioral changes in mice following short-term inhalation exposure to various industrial solvents. Toxicol Appl Pharmacol 1983;67:383-9.
Dow Chemical Company. Initial 8e submission. Neurotoxicological study in rats exposed to 1,1,1-trichloroethane vapor for 13 weeks (final report) with cover letter dated 6/4/92. Washington, DC: US Environmental Protection Agency Public Files, 1991. Fiche #OTS0540067.
Geller I, Mendez V, Hartmann RJ, Gause E. Effects of 1,1,1-trichloroethane on a match-to-sample discrimination task in the baboon. J Toxicol Environ Health 1982;9:783-95.
Kjellstrand P, Holmquist B, Jonsson I, Romare S, Mansson L. Effects of organic solvents on motor activity in mice. Toxicology 1985;35:35-46.
Moser VC, Balster RL. Acute motor and lethal effects of inhaled toluene, 1,1,1-trichloroethane, halothane, and ethanol in mice: effects of exposure duration. Toxicol Appl Pharmacol 1985;77:285-91.
Moser VC, Balster RL. The effects of inhaled toluene, halothane, 1,1,1-trichloroethane, and ethanol on fixed-interval responding in mice. Neurobehavioral Toxicology and Teratology 1986;8:525-32.
Mullin LS, Krivanek ND. Comparison of unconditioned reflex and conditioned avoidance tests in rats exposed by inhalation to carbon monoxide, 1,1,1-trichloroethane, toluene, or ethanol. Neurotoxicology 1982;3:126-37.
Woolverton WL, Balster RL. Behavioral and lethal effects of combinations of oral ethanol and inhaled 1,1,1-trichloroethane in mice. Toxicol Appl Pharmacol 1981;59:1-7.
Karlsson J-E, Rosengren LE, Kjellstrand P, Haglid KG. Effects of low-dose inhalation of three chlorinated aliphatic organic solvents on deoxyribonucleic acid in gerbil brain. Scand J Work Environ Health 1987;13:453-8.
Rosengren LE, Aurell A, Kjellstrand P, Haglid KG. Astrogliosis in the cerebral cortex of gerbils after long-term exposure to 1,1,1-trichloroethane. Scand J Work Environ Health 1985;11:447-55.
Verschuuren HG, Wilmer JW. Neurotoxicity of 1,1,1-trichloroethane questioned [letter]. Scand J Work Environ Health 1990;16:144-5.
Haglid KG, Rosengren LE, Karlsson J-E. Neurotoxicity of 1,1,1-trichloroethane questioned: author's reply [letter]. Scand J Work Environ Health 1990;16:145-6.
Nilsson KB. Effects of 1,1,1-trichloroethane on the c-AMP metabolism in mouse brain. Acta Pharmacol et Toxicol 1986;59:362-9.
Nilsson KB. Effects of 1,1,1-trichloroethane on the c-GMP metabolism in mouse brain. Acta Pharmacologica et Toxicologica 1986;58:318-26.
Folbergrova J, Hougaard K, Westerberg E, Siesjo BK. Cerebral metabolic and circulatory effects of 1,1,1-trichloroethane, a neurotoxic industrial solvent. 2. Tissue concentrations of labile phosphates, glycolytic metabolites, citric acid cycle intermediates, amino acids, and cyclic nucleotides. Neurochem Pathol 1984;2:55-68.
Savolainen H, Pfaffli P, Tengen M, Vainio H. Trichloroethylene and 1,1,1-trichloroethane: effects on brain and liver after five days intermittent inhalation. Arch Toxicol 1977;38:229-37.
Kyrklund T, Kjellstrand P, Haglid KG. Effects of exposure to Freon 11, 1,1,1-trichloroethane or perchloroethylene on the lipid and fatty-acid composition of rat cerebral cortex. Scand J Work Environ Health 1988;14:91-4.
Epidemiological Studies Section, California Department of Health Services. Pregnancy outcomes in Santa Clara County 1980-1982: reports of two epidemiological studies. Sacramento (CA): California State Publications Section, 1985. Report No.: 7540-958-1301-5.
Deane M, Swan SH, Harris JA, Epstein DM, Neutra RR. Adverse pregnancy outcomes in relation to water contamination in Santa Clara County, California. Am J Epidemiol 1989;129:894-904.
Swan SH, Shaw G, Harris JA, Neutra RR. Congenital cardiac anomalies in relation to water contamination, Santa Clara County, California, 1981-1983. Am J Epidemiol 1989;129:885-93.
Wrensch M, Swan S, Lipscomb J, Epstein D, Fenster L, Claxton K, et al. Pregnancy outcomes in women potentially exposed to solvent-contaminated drinking water in San Jose, California. Am J Epidemiol 1990;131:283-300.
George JD, Price CJ, Marr MC, Morrissey RE, Schwetz BA. Developmental toxicity evaluation of 1,1,1-trichloroethane (CAS no. 71-55-6) administered to CD rats. Part I. Postnatal evaluation, final study report. Rockville (MD): National Toxicology Program, National Center for Toxicological Research, August 1987. NTIS No.: PB88131321/AS.
George JD, Price CJ, Marr MC, Morrissey RE, Schwetz BA. Developmental toxicity evaluation of 1,1,1-trichloroethane (CAS no. 71-55-6) administered to CD rats. Part II. Teratology evaluation, final study report. Rockville (MD): National Toxicology Program, National Center for Toxicological Research, September 1987. NTIS No.: PB88134101.
Lane RW, Riddle BL, Borzelleca JF. Effects of 1,2-dichloroethane and 1,1,1-trichloroethane in drinking water on reporductive and development in mice. Toxicol Appl Pharmacol 1982;63:409-21.
Schwetz BA, Leong BKJ, Gehring PK. The effect of maternally inhaled trichloroethylene, perchloroethylene, methyl chloroform, and methylene chloride on embryonal and fetal development in mice and rats. Toxicol Appl Pharmacol 1975;32:84-96.
York RG, Sory BM, Hastings L. Evaluation of teratogenicity and neurotoxicity with maternal inhalation exposure to methyl chloroform. J Toxicol Environ Health 1982;9:251-66.
Kobayshi H, Hobara T, Kawamoto T, Sakai T. Effect of 1,1,1-trichloroethane inhalation on heart rate and its mechanism: a role of autonomic nervous system. Arch Environ Health 1987;42:140-3.
Carlson GP. Effect of alterations in drug metabolism on epinephrine-induced cardiac arrhythmias in rabbits exposed to methylchloroform. Toxicol Lett 1981;9:307-13.
Clark DG, Tinston DJ. Correlation of the cardiac sensitizing potential of halogenated hydrocarbons with their physicochemical properties. Br J Pharmacol 1973;49:355-7.
Reinhardt CF, Mullin LS, Maxfield ME. Epinephrine-induced cardiac arrhythmia potential of some common industrial solvents. J Occup Med 1973;15:953-5.
Honma T. Effects of trichloroethylene, 1,1,1-trichloroethane and carbon tetrachloride on plasma lipoproteins of rats. Industrial Health 1990;28:159-74.
Halevy J, Pitlik S, Rosenfeld J. 1,1,1-Trichloroethane intoxication: a case report with transient liver and renal damage. Clinical Toxicology 1980;16:467-72.
Viola A, Sigon M, Pittoni G, Sarto F, Pennelli N. Serum enzyme activities and histological changes after percutaneous application of methylchloroform. Med Lav 1981;72:410-5.
Bruckner JV. Findings of toxicological studies of 1,1,1-trichloroethane. Progress report. Washington, DC: US Environmental Protection Agency, 1983.
McCarthy TB, Jones RD. Industrial gassing poisonings due to trichloroethylene, pentachloroethylene and 1,1,1-trichloroethane, 1960-1980. Br J Ind Med 1983;40:450-5.
Stewart RD, Andrews JT. Acute toxication with methylchloroform. JAMA 1966;195:904-6.
Toftgaard R, Nilsen OG, Gustafsson JA. Changes in rat liver microsomal cytochrome P-450 and enzymatic activities after the inhalation of n-hexane, xylene, methyl ethyl ketone and methylchloroform for four weeks. Scand J Work Environ Health 1981;7:31-7.
Kallman MJ, Kaempf GL. Efficacy of choice testing to predict chronic ingestion of drinking solutions adulterated with chemicals. Pharmacol Biochem Behav 1984;20:195-200.
Kinkead ER, Leahy HF. Evaluation of the acute toxicity of selected groundwater contaminants. Wright-Patterson Air Force Base (OH): Harry G. Armstrong Aerospace Medical Research Laboratory, 1987. AAMRL-TR-87-021.
Wahlberg JE, Boman A. Comparative percutaneous toxicty of ten industrial solvents in the guinea pig. Scand J Work Environ Health 1979;5:345-51.
Aral MM. Ground water modeling in multilayer aquifers: steady flow (SLAM). In: Chelsea (MI): Lewis Publishers, 1989;114.
Tang Y, Aral MM. Contaminant transport in layered aquifer media (CLAM). Atlanta: Georgia Institute of Technology, 1992. Report No.: CE512.
GisPlus, geographic information system, version 2.1. Newton (MA): Caliper Corporation, 1992.

EXECUTIVE SUMMARY

This is a report on the baseline activities and resultsof the statistical analyses of the baseline and first follow-up data forthe 1,1,1-Trichloroethane (TCA) Subregistry of the Volatile Organic Compounds(VOCs) Registry of the National Exposure Registry. The National ExposureRegistry (NER) was created in response to the congressional mandate inthe 1980 Comprehensive Environmental Response, Compensation, and LiabilityAct and reiterated in the Superfund Amendments and Reauthorization Actof 1986. This mandate directed the Agency for Toxic Substances and DiseaseRegistry (ATSDR) to create registries of people exposed to hazardous substancesin the environment. The VOCs Registry is one of four existing substance-specificregistries in the NER. The TCA Subregistry is one of three existing subregistriesin the VOCs Registry.

The purpose of the NER is to assess the long-term healthconsequences to the general population of long-term exposure to low levelsof environmental contaminants. One of the goals for meeting this purposeis to establish a database that will furnish the information needed togenerate appropriate and valid hypotheses for future activities-such asepidemiologic studies. The VOC Registry is not a definitive study. Cause-and-effectrelationships cannot be established using only Registry-based information.However, the results of the data analyses will identify reported healthoutcomes that should be considered for future activities.

As with the other registries in the NER, the VOCs Registryis a database containing information on people who have been exposed tospecific substances, in this case specific VOCs. The purpose of the VOCsRegistry is to assess the long-term health consequences, if any, of long-termpast exposures to low levels of VOCs in drinking water. It should be notedthat the VOCs Registry itself is not a definitive study; cause and effectrelationships cannot be established using only registry-based information.However, the registry will furnish the information needed to generateappropriate and valid hypotheses for future activities, such as epidemiologicstudies.

The data collected for each member of the TCA Subregistryinclude environmental data, demographic information, smoking and occupationalhistories, and self-reported responses to 25 general health status questions.The data files for each subregistry are established at the time baselinedata are collected. Follow-up surveys are conducted at the end of thefirst year, then at 2-year intervals thereafter to update the data files.

The TCA Subregistry baseline data contains informationon 3,665 persons (3,204 living and 461 deceased at the time of data collection)who had documented exposure to TCA in their drinking water and were exposedfor at least 30 days; the follow-up files contain information on 3,469persons (2,963 living and 510 deceased). The individuals included in theTCA Subregistry resided at a site in Vestal, New York. The participationrate for individuals identified as eligible was 97%.

Reported health outcome rates were calculated for membersof the TCA Subregistry and compared with morbidity data from the 1990National Health Interview Survey (NHIS) and with the National Cancer Institute'sSurveillance, Epidemiology, and End Results Program data.

When interpreting the statistical results and planningfuture activities based on these results, the limitations of the subregistrydata files must be kept in mind. For instance, a bias in the rate of reportingcould have existed because people were aware of their TCA exposure, hadbeen advised of the potential effects on their health, and might havesought medical care more often than the general population. To moderatethis potential bias, the subregistry data were collected with the restrictionthat a health care provider had to have told the person that he or shehad the condition or had to have treated him or her for the condition.Also, some of the questions in the two data collection instruments wereworded differently, making direct comparisons of the reported rates moredifficult to interpret. There is the possibility that, given the largenumber of comparisons used in the analyses, there might be some falsepositive findings. These limitations and restrictions are discussed inthis report.

The morbidity data analysesindicated an increased reporting of several health outcomes by TCA Subregistryregistrants. Statistically significant increases (p£0.01significance level) were reported at Baseline for anemia and other blooddisorders for females 25 through 34 years of age and 65 years of age orolder; arthritis in males 10 through 17 years of age; and urinary tractdisorders for females 10 through 17 years, 25 through 34 years, and 65years of age or older, and males 35 through 44 years of age. At followup,statistically significant increases were seen for anemia in females 26 through35 years of age and for urinary tract disorders for females 19 years ofage and older and males 26 through 25 years and 66 years of age and older.Statistically significant deficits were reported for the TCA Subregistrypopulation (for specific age and sex groups) at either baseline or followupor both for the following health conditions: arthritis, asthma and emphysema,diabetes, hearing impairments, hypertension, kidney disease, respiratoryallergies, speech impairments, ulcers or other stomach problems, and mentalretardation. It should be noted that the rates of these conditions mighthave been affected by the limitation imposed on registrants' reporting,that is, that a health care provider had told the registrants they hadthe condition or had treated them for the condition. This limitation wasnot part of the NHIS.

The findings in this report cannot be used to identifya causal relationship between the health outcomes and TCA exposure. Inaddition, there are some methodological differences in data collectionthat might have biased the reporting rates, resulting in false positivefindings. The findings of this report do, however, reinforce the needto continue regular followup of this population. In addition, the findingssuggest several areas needing further study or clarification.

New methods need to be explored to characterize the environmental exposures of registrants. Dose-response relationships need to be further explored by means of strategies currently being investigated by ATSDR.

Additional information is needed about the people in the TCA Subregistry population who reported health outcomes identified as excess. The types of information that might prove useful include further details of health conditions from medical files, information on lifestyle factors (such as alcohol consumption or diet), and information on occupational exposures that might relate to another potentially causative factor. In the future, ATSDR will contact registrants to obtain their permission to access this information. These data, along with data obtained from continued followup of the registrants and confirmation of medical conditions, will aid in further exploration of possible linkages.

The subregistry questionnaire wording will be altered so that it is more closely aligned with the appropriate NHIS questions.

The data file, without personal identifiers, will be made available to the public. Other researchers are encouraged to study these reported results more extensively and to undertake other analyses.

For eligible people who are deceased, death certificates will be obtained and pertinent information abstracted. A mortality file will be constructed and comparisons will be made using the appropriate national norms.

Another of the stated goals of the NER is to keep registrantsinformed of all current information related to their exposures. Beforerelease of the information in this report, a TCA Subregistry TechnicalAssistance Panel-composed of representative membership from the stateinvolved, other federal agencies, and other knowledgeable people-was formedto advise ATSDR on future research needs. Following the panel meetingand immediately before the release of information to the registrants,a meeting with community officials, along with interested county, state,and federal agency representatives, was held to plan site-specific activities-includingpublic meetings. A registrant report, written for the general public andcontaining the findings of this technical report, was prepared and sentto each registrant. The registrant report was then released to the media.The mailing was followed by a public meeting at the site to discuss thereport.

SECTION 2: BACKGROUND OF THE 1,1,1-TRICHLOROETHANE SUBREGISTRY

SELECTION OF 1,1,1-TRICHLOROETHANE AS A PRIMARY CONTAMINANT

The 1,1,1-trichloroethane (TCA) Subregistry was establishedin 1992 using a process that has been previously described (1).In summary, the factors that led to the selection of TCA included theprioritization of TCA on the Hazardous Substance Priority List (2);the ubiquitousness of TCA in the environment; the published evidence ofTCA toxicity in worker populations and in toxicologic studies; and thepaucity of information on low-level, long-term exposures to TCA. Eachof these factors suggested that the TCA Subregistry could contribute significantlyto the detection of an excess in adverse human effects, should they exist,following long-term, low-level exposure to TCA in the environment.

This report provides details on the presence of in theenvironment, as well as a summary of evidence of adverse effects observedin both human and animal studies. The available literature indicates thatTCA can be toxic to humans and animals and was described previously (3).A more detailed discussion of the updated literature is given in Section 6.

SELECTION AND DESCRIPTION OF THE TCA SUBREGISTRY SITE

The site selection process used to develop subregistriesfor the NER is described in detail in the Policies and Procedures Manual (1). Details on the selection of Vestal Water Supply Well #1-1as the site for the TCA Subregistry are provided elsewhere (3)and are only summarized here.

Vestal Water Supply Well #1-1 is located in Water District#1 of the Town of Vestal, Broome County, New York (Figure 2-1). The wellwas the primary of three wells (Figure 2-2) that supplied drinking waterto approximately 9,000 residents in the western part of the town, as wellas to residents in Water Districts #2, #8, and #9. It should be notedthat this site used a public water system; therefore, for the purposeof this report, exposures for all registrants at the site were consideredto be the same. The duration of exposure, and thus the time period ofinterest for the Subregistry, was determined to be January 1, 1969 throughMay 31, 1980, which represents documented beginning dates and confirmedending dates of use of contaminated water (Table 2-1). A complete listof contaminants found in the Well #1-1, Vestal water supply is providedin Table 2-2.

On-Site Activities

Data collection for the TCA Subregistry began in 1992and to date three followups have been completed (Table 2-3). People whowere then residing or had previously resided at the addresses served bythe TCA-contaminated well were considered to be "potential" registrants.Baseline interviews were conducted face-to-face while followup interviewswere conducted via computer-assisted telephone interviews (CATI). Beforeall interviews, registrants were first sent a mailing that contained informationabout ATSDR, the NER, the TCA Subregistry, and the chemical TCA. For

Figure 2-1.—Location of Vestal, New York.

Figure 2-2.—Vestal, New York, water districts.

Table 2-1.—Summary of environmental data.

TCA Subregistry Site	Year Exposure Began	Exposure Period* (Number of Years)	Maximum Level of TCA Reported (ppb)†
Vestal Water Supply Well #1-1	1969	11.5	1,600.0

*Exposure period is based on best available evidence of when contamination occurred and when exposure ceased following switch to an alternative water source.

†ppb = parts per billion.

the Baseline, a public meeting about the Registry was held in the area prior to the start of data collection.

the contaminated source to the potential registrant during the period of exposure as verified by that individual. For example, in the case of the TCA Subregistry, the well water had to have been tested and validated for the presence of TCA. Also, the well water had to have been the sole source of water for drinking, bathing, or cooking for all individuals at the site residential addresses. Finally, a registrant would had to have reported using the TCA-contaminated well water for drinking, cooking, or bathing during the exposure period.

Information on deceased eligible persons was obtained from a knowledgeable proxy (usually the spouse) and a death certificate was requested from the appropriate state office. Information on the cause of death, along with other pertinent information, is being extracted from the death certificates and coded as copies of death certificates are obtained from the states. These procedures were the same for the baseline and for all follow-up interviews. Analysis of the mortality data is not included in this report; a separate report on mortality is in progress.

The procedures used for locating registrants for each followup were as follows: four to five weeks before the start of data collection activities, ATSDR began tracing efforts of those registrants known to have moved since the last interview. These tracing cases were usually identified through registrant mailings that were returned to ATSDR as undeliverable. During data collection, cases requiring tracing were identified through attempts made to the telephone number on record. A case was forwarded to a locating specialist if the registrant or proxy had moved (no new telephone number provided), if the telephone number had been disconnected, or if calling the number resulted in "ring no answer" or "busy" (this would be considered a non-working number). The locating specialist would then attempt to locate the registrant by one of the following methods (in the order listed): (1) calling directory assistance; (2) calling contacts provided during the last interview; (3) credit bureau searches; and (4) state departments of motor vehicles searches.

At each followup, and using CATI technology, each registrant or proxy for the registrant was again administered the NER core questionnaire, which includes a set of questions about health practitioner confirmed or treated health conditions that the registrant currently had or had since the last interview. If a respondent reported a health condition, further questions were asked about the date of first treatment, current treatment, prescribed medication, and hospitalization history.

Table 2-4 summarizes the response information from interviews of the exposed registrants at each of the data collection time points and overall. At Baseline, 95% of the eligible people who were contacted and asked to take part in the TCA Subregistry did participate. Total participation rates were calculated by dividing the number of registrants (living and deceased) who completed interviews by the number of potentially eligible persons who were contacted and asked to participate. By the Followup 3 interviews, 94% of the registrants contacted had agreed to participate. Overall, from Baseline through Followup 3, the TCA Subregistry has retained 81 % (N = 3,139) of the original registrants (including all deceased registrants).

Table 2-2.—Contaminants for Well #1-1, Vestal, New York.

Contaminant	Maximum Level (ppb)*
1,1,1-Trichloroethane	1,600.0
Trichloroethylene	470.0
Benzene	27.0†
Tetrachloroethylene	6.1
Chloroform	35.0
Dibromochloroethane	7.0
1,1,2-Trichlorofluoromethane	4.0
Freon	15.0
Vinyl chloride	2.0
1,1-Dichloroethylene	8.0
1,1-Dichloroethane	81.0
1,2-Dichloroethylene	82.0
1,2-Dichloroethane	1.8
Trichloroethane	86.0
Chloroethane	7.0

*ppb - parts per billion. †The first reading for benzene was 43 ppb. However, it was noted that benzene coeluted with trichloroethylene; therefore, this reading was not included.

Table 2-3—Summary of data collection activities.

Interview Period	Date
Baseline	January 1992-April 1992
Followup 1	October 1993-September 1994
Followup 2	September 1995-January 1996
Followup 3	September 1997-December 1997

Table 2-4.—Summary of registrant response in the TCA Subregistry, Baseline through Followup 3, all races.

Outcome	Baseline	Followup 1	Followup 2	Followup 3
Outcome	Number(%)	Number (%)	Number (%)	Number (%)
Completed interview, registrant living	3,204 (83.0)	2,963 (92.5)	2,769 (93.5)	2,536 (91.6)
Completed interview, registrant deceased	461 (11.9)	49 (1.5)	37 (1.2)	56 (2.0)
Refusal	142 (3.7)	96 (3.0)	82 (2.7)	104 (3.8)
Noninterviewed registrants*	51 (1.3)	96 (3.0)	75 (2.5)	73 (2.7)
Total Eligible	3,858	3,204	2,963	2,769

% = Percent of Total Eligible

* Includes unable to locate or contact, unavailable during interview period, language barrier, in litigation, and mentally or physically incapable, with no available proxy

top