Document 2wdYby13yr3jjd7JN77OkxJp

A Preliminary Study of Potential Human Health Effects in Private Residences Following Chlordane Applications for Termite Control SUSAN MENCONI, M.S. J. MILTON CLARK, Ph.D. Environmental and Occupational Health Sciences PATRICIA LANCENBERC, Ph.D. DANIEL HRYHORCZUK, M.D. Epidemiology and Biometry School of Public Health University of Illinois at Chicago 2121 W. Taylor Street Chicago, Illinois ABSTRACT. A cross-sectionalepidemiological investigation of the health status of 261 people from 85 private households previously treated with chlordane for termite control was conducted to assess potential human health effects. When sinusitis, bronchitis, and migraine responses were compared to measured indoor air levels of chlordane in categories designated as low (< 1C(edm3),medium (1-5 pglm'), and high 0 5 )cglm3),a doseresponse relationshipwas found after controlling individuallyand simultaneously for age, sex, and smoking. Homes studied in this investigation having a proper chlordane application were observed to have chlordane air levels exceeding acceptable exposure limits. CHLORDANE has been the leading chemical agent used for controlling termite infestations and damage to private homes in the United States.',' Indoor air pollution by chlordane was first reported by the United States Air Force in 1970 and again in 1974 and 1978.3 In 1979, the National Academy of Sciences issued a temporary guideline of 5 &m3 for chlordane, but without a guarantee that biological effects in exposed humans would not occur below this 1eveL3In an attempt to determine the veracity of health complaints related to chlordane, a cross-sectional epidemiological inwstiga- tion was conducted in 85 chlordane-treated house- holds with 261 people, where indoor air conceatra- ..tions of chlordane or other cyclodiene termitiddes Were known.` pling. Indoor air concentrations of hlor, and other terntiticides $&de living areas from 107 homes were obtained from a customer database made available by an environmental consulting firm. The termiticide sampling approach employed was similar to that described by Wright and L e i d ~ T. ~he National Institute of Occupational Safety and Health`s (NIOSH) analytical method for chlordane detection and quantification was utilized.6The concentrations of termiticides were reported .as micrograms per cubic meter of air (pg/m3). The detection limit for chlordane, based upon an average sample volume of 0.24 m3, was 0.42 pg/m3. Results of the air analyses were provided to the customer in a written report but without any health interpretations. Questionnaire survey. A health questionnaire was designed to collect data from the study population. A total of 20 questions addressed acute health symptoms experienced immediately after the chlordane termiti- cide treatments were prcwmed and included symp toms often associated with Kute exposure to chlori- n pe~ticides."T~wenty-three health cOndStiOns exwrienced bv individuals adthe past ylear were 'listed in the questionnaire followed the format used by the Niittional Center for Health Statistics in their 1979 National Health Intenkv Survey (NHIS).' As a measure of potential bias, some health conditions believed unrelated to termiticide exposure such as goiter, diabetes, hearing impairments, and ulcers were included. A mail survey approach similar to that described by Dillman was used." Quebtionnaires were mailed to the 107 households in June 1984. A total of 86 questionnaires were returnd, which is considered a high cooperation rate (80%)for a mail survey." Respondent's principal reasons for conducting air testing included news reports on chlordane (52%), perceived health problems ( 19%), and odor problems (10%). Questionnaires were returned from 19 states, with New York and New Jerseyconstituting 54%of the total. Analysis of database. To analyze statistically the reported prevalence of selected chronic disease conditions in the chlordane-treated homes, using the NHlS data as a control, the 261 respondents were divided in- to four age categories: < 17 yr, 17-44 yr, 45-64 yr, and > 65 yr. Observed and expected cases of disease were computed for each selected chronic health condition, based on the age-specific rates from the NHlS study. Sex-Ypecific rates were not computed. Chi-square analysis was performed after the observed and expected cases of disease were summed over the age strata. If the expected cases of disease were less than five for a chronic health condition, then the Poisson approximation to the binomial distribution was used to estimate the probability of such an occurrence. Three exposure groups were selected for evaluation, based upon the NAS guideline for chlordane: nonde- tectable or < 1 pg/m3, from 1-5 pg/m3 (within the NAS guideline), and > 5 pglm' (above the NAS guideline). Because of the structural and toxicological similarity of heptachlor and aldrin to chlordane, measured air levels of these compounds were treated as "chlordane equiv- 'alent" and were included in the exposure groups.' For further analysis of the study population, an internal control was used of those households having meas- ured air levels of termiticides of < 1 pg/m3. Tables of frequency of condition by exposure level were con= stcucted for each chronic health condition. After chisquare analyses were performed for the various disease conditions, the frequency tables were stratified on levels of potential confounders, and Mantel-Haenszel summary chi-square statistics were computed." For those disease conditions found to be statistically significant, a test tor trend (with exposure) was computed to identify an increase in disease response with increasing chlordane air levels." To consider the effects of confounders in the test for trend, a Mantel extention of the Mantel-Haenszel statistic was computed." Finally, to further confirm a dose-response relationship found for disease conditions in the previous tests, a stepwise logistic regression model was constructed. ReSUttS oescripti data. The time pleting the questionnaire and the last term& application at the residence ivaS 1-24 yr. TI time was 3.8 yr i 3.66 (standard deviation [s proximately 50%of households fell in the 1-2 since the last treatment with chlordane at th& idence. Time after chlordane treatment, categorized than or greater than 2 yr, was found to be unreia measurd air levels of chlordane, which supports ous research that demonstrates chlordane decline very slowly with time.5"8"9 No stati tionships between measured air levels of ch other termiticides, either present or absent, fowdhome (full basement, slab or crawl space) was In addition, no relationships were discovered betweqr measured air levels of chlordane and the type ofh& ing or air conditioning system used. aHeadache was the most frequently reported acute symptom (22%) among respondents immediately lowing a chlordane termiticide application. Sore thropt and respiratory infections were also cited freque* (l6%),as were fatigue (1496),sleeping difficulties, M w vision, weakness and fainting, or confusion. For chronic health conditions, sinus trouble was the m a frequently reported symptom (21'10) among re- spondents, followed by bronchitis (13%), migraine (8%), dermatitis (8%), asthma (5%), neuralgia neuritis (5%), anemia (4%),and disease of the 09 and uterus (4%). Statistical analysis of chronic health conditions. The incidence rates of anemia, migraine, neuralgidneuritis, ovarian and uterine disease, bronchitis, sinusitis, &c matitis, and skin neoplasms were found to be signi.i. cantly elevated (p < .OS) in the chlordane-treated homes compared to the NHlS population. The preva- lence of goiter, diabetes, hearing impairment, and ulcers in the study population thought to be unrelated to chlordane were similar to prevalence rates reported in the NHlS survey population, therefore increasingthe Table l.-Resuks of the Test for Trend of Chlordane Ex- posure: Low, Medium, and High, and Self-Reported -. Chronic Health Conditions after Stratifying on Smoking, Age;*, and-fhmberof Chlordane Treatments, Using the .- - Mantel Extension of the Mantd-Haenszd Test* t Number of Health chlordane condition treatments* Smoking Age Sex t Migraine x1 value: 3.26 Sinusitis p value: (.071) x' value: 9.66 p value: (.002) Bronchitis x* value: 7.88 p value: (.005) 1.21 (.271) 7.69 (.004) 6.1 1 (.010) 4.55 (.033) 11.37 (.001) 6.91 (.008) 4.76 (.OM) 10.68 (.001) 7.30 (.007) 350 Archives of EnvironmentalH d Table 2.-Odds Ratios with 95% Confidence Intervals Estimated From a Logistic Model of Health Conditions as a Function of Pesticide Levels Simultaneously Controllingfor Age, Sex, and Smoking Dependent variable Migraine Sinusitis Bronchitis Significant variables > 5 &m3 chlordane Females > 5 pg/m3chlordane Smoking > 5 pg/m3chlordane 17-44 yr age group p value ,0214 ,0145 .a2 .0377 ,0053 ,0297 Odds ratio with 95% confidence interval 3.00 (1.18-7.4s) 3.64 (1.30-10.30) 2.56 (1.31-5.OO) 2.25 (1.05-4.83) 3.06 (1.40-6.73) 2.68 (1.10-6.51) probability of homogeneity between the control and miticides, the observation has been made that the cy- study populations. -___-- -c_lo_diene-t_erm_it_ic_i_des are inhalent and skin irritant^.^*','^ Chronic disease rates in homes having 1-5 &m3 (me- Aplastic and acute refractory megaloblastic anemia and dium exposure) or > 5 pg/m3 (high ex-powref of a ter- effects on the female recmductive system have also miticide were compared to those for the control (lowexposure) group by chi-square analysis. Migraine (p = been associated with chlordane and/or heptachlor exre.10 13.16.17 .031),bronchitis (p = .004), and sinusitis (p = .004) In our study, half the homes judged to have had a were the self-reported health conditions for which a proper termiticide application had detectable air levels statistical relationship to measured air levels of chlor- of chlordane, an average of 2.7 &m3, months to years dane or other termiticide were seen. Using the Mantel- following the last application. Very similar results have Haenszel chi-square test, and controlling for smoking, been reported in other ~tudies.*~'D~e~ri'v~ations of ac- age, and sex, sinusitis, bronchitis, and migraine were ceptable exposure limits to chtordane using animal tox- found to remain statistically significant at probability @) values of .05or less. icological data suggest that indoor air concentrations of chlordane would need to be < 1 pg/m3 (as low as 0.2 The results of the Mantel extension analysis for mi- pg/m3) to provide an adequate margin of safety from graine, sinusitis, and bronchitis (Table 1) indicate that noncarcinogenic effects on the liver.2' the measured pesticide air level was found to be signifi- As the cyclodienes are considered to be animal car- cantly associated with the chronic health conditions of cinogens, long-term exposure to these compounds sinusitis, bronchitis, and migraine. Results of the step warrant ~ o n c e r n . ~ , ~ll's~in'g- ~th~e exposure informa- wise logistic regression (Table 2) revealed statistically tion and termite treatment frequency data from this significant associations between sinusitis, bronchitis, and other studies, potential cancer risks would range and migraine at air levels of chlordane or other termiti- from approximately IO-` to 10-~ (1 in 10,OOO to 1 in >cides 5 &m3. 1,OOO) for a chlordane air concentration of 1-5 pglrn', a Discussion 2-5-yr exposure period, and treatments every 3-7 yr.'9,20Such indoor air cancer risks would be second The study population in this investigation was self-se- only to those derived for indoor radon exposure, esti- lected: individuals concerned about chlordane or mated to be in the range of 10-' to The United health problems paid to have their homes tested. A States Environmental Protection Agency (USEPA) has health survey on randomly selected homes would have provided more definitive conclusions regarding a rela- often considered lifetime cancer risks exceeding 10-6 (one in a million) as ~nacceptabie.A~n~ air level of < tionship between indoor air levels of chlordane and hu- 0.1 d m 3 , which appears unachievable, may need to man health effects. Due to potential, uncontrolled bias be required to have cancer risks within an acceptable in this study, further epidemiological studies and in- range. On August 1 1 , 1987,on the basis of new evalua- door air measurements of chlordane and other termiti- tions regarding the safety of the cyclodiene termiti- cides would be required on larger populations of ran- cides, EPA announced the cancellation of virtually all domly slected residences in order to draw firm conclu- termiticide uses of chlordane, heptachlor, aldrin, and sions. The finding of a dose-response relationship be- dieldrin.26 tween the indoor air concentrations of chlordane and three self-reported chronic health conditions (migraine, sinusitis, and bronchitis), using an internal con- i &; trol and considering potential confounders, suggests that chlordane could have chronic human health impacts. The association found between chlordane and mi- ********** Special thanks go to Dr. Fred Kviz and Dr. Edward Chen of the Epidemiology and Biometry Program, U n i t y of Illinois School of Public Health, who provlded valuabk aput on the design of the chlordane health survey. Submitted for publication August R. 1967; revised; accepted for publication April 18, 1988. Requests for reprints should be sent b:J. Milton Clark, Ph.D., Envi- ronmental and Occupational Health seicnces, sdrooi of Public Health, University of Illinois at ChIL 60612. . 2321 W. Taylor St., -4 Chicago, 7 t I t i I 2. lnternaional Agency for Research on Cancer. 1979.&KC Monographs on the Evaluation of the Carcinogenic Risk of Chenicals to Humans Some Halogenated Hydrocarbons. d.M Lyon, France: International Agency for Research on Cancer. 3. National Academy of Scimces. 1979. Chlordane in Military Housing. Washington, D.C.: Subcommittee on Chlordane in Military Housingof the Committee on Toxicology. 4. United States Environmental Protection Agency (USEPA). 1980. Summary of Reported Pw'ticide Incidents :nvdving CMordane, Report No. 360. Washington, D.C.: USEPA, lieaith Effects Branch, Hazard Evaluation Division, OHice of Pesticides P w gram. 5. Wright, C. C.and Leidy, R. E. 1982.Chlordane and heptachlor in the ambient air of houses treated for termites. Bull Environ Contam Joxicol 28:616-20. 6. United States Department of Health and Human Services. 1980. NlOSH Manual of Analytical Methods. 6 (Method) 5278. Cincinnati, OH: Public Health Service, Centeo for Disease Control, Na- tional Institute for Occupational Safety and Health, Division of Physical Sciences and Engineering. 7. United States Department of Health and Human Services. 1981. Current Estimates from the National Health Interview Survey. C l . 5 , 1970; Series 10 (136).Hyattsville, MD: Public Health Sewice, Office of Health Research, Statistics, and Technology, National Center for Health Statistics. 8. Morgan, D. P. 1982.Recognition and Management of Pesticide Poisonings, 3rd ed. Washington, D.C.: United States Environmental Protection Agency. 9. United States Environmental Protection Agency (USEPA). 1980. Chlordane, Health and Environmental Effects Profile No. 35. Washington, D.C.: USEPA, Office of Solid Waste. 10. Dillman, D. 1978.Mail and Telephone Surveys. New York: John Wiley and Sons, Inc. 1 1 . Schlesselman, 1. 1. 1982.CaseControl Studies. New Yo*. Oxford University Press. 12. Carrettson, L. K.; Guzelian, P. S.; and Elanke, R. V. 1965. Sub acute chlordane poisoning. Clin Toxicol 22:565-71. 13. Olanoff, L. S.; Eristow, W. I.; Colcolough, J.; and Reigart, J. R. 1983. Acute chlordane intoxication. I TOXiCOl Clin Toxicol 2 0 -16. Infante, P.; Epstein, 5.; and Newton, W. 1978. Blood dyscr&i and childhood t u m m and exposure chlor. Scand / Work Environ Health 4 to chlordane 137-50. and hepab. . 17. Welch, R. M.; Levin, W.; Kuntzman, R;Jacobson, M.; andc- ney, A. H. 1971. Effect d halogenated hydrocarbon in- on the metabolism and uterothophic action of estrogens in and mice. Toxicol App/ Phumacol 19234-46. 18. Khasawinah, A. M. 1982. Chlordane. Air Concentrations in Treated Homes: Assessments and Significance. Chicago, W. Velsicol Chemical Cop. 19. Leidy, R. E.; Wright, C. G.;Dupree, Jr., H. E.; and Sheets, T. J. 1985. Subterranean termite control. Chlordane residues in soil surrounding and air within houses. in: Dermal Exposure Relaw to Pesticide Use, R. C. Honeycutt, G.Zweig, and N. N. Re@&, Eds., ACS Sym. Ser. 273, pp. 265-67.Washington, D.C.: Am& can Chemical Society. 20. New York Department of Environmental Conservation. 1983. Drah Environmental lmpact Statement on Amendments to 6 NYCRR Part 326 Relating to the Restriction of the Pesticides Aldrin, Chlordane, Chorpyrib, Dieldrin and Heptachlor. & bany, New York. 21. National Academy of Sciences. 1982. An Assessment of the Health R i s k of Seven Pesticides Used for Termite Control. Wash ington, D.C.: National Academy of Sciences. 22. National Cancer Institute. 1977.Bioassay for Chlordane for Pozri. ble Carcinogenicity. Eethesda, MD: National Institutesof Health, Division of Cancer Cause and Prevention. 23. United States Environmental Protection Agency. 1974. Pesticide products containing heptachlor or chlordane: Intent to cancel registrations. Fed Register 3 9 41298-41Mo. 24. Thomas, D. C.; McNeill, K C.; and Dougherty, C. 1985. Esti- mates of lifetime cancer risks resulting from radon progeny expc- sure. Health Physics 49:825. 25. Thomas, L. 1983.Testimony before the Public Works and Trans portation Subcommittee on investigationsand Oversight. Environ Rep November 18,1983. 26. United States Environmental Protection Agency (USEPA). 1987 (August 11). Announcement of Restrictions in Use of Chlordane. Washington, D.C.: USEPA. 352 Archives of Environmental Health