Document YGE3wnMy1oxZdy9Obp1xJdEvD

Copen- ">.ion from 1, Institute of ay and/or Dr. "sity of Wiirzandards being e Maltoni, md urn Ramazzini i "F. Addarii" e Ercolani 4/2 Bologna, Italy American Journal of Industrial Medicine 13:531-559 (1988) Corporate Influence on Threshold Limit Values Barry I. Castleman, sco, and Grace E. Ziem, mo, orPH Investigations into the historical development of specific Threshold Limit Values (TLVs) for many substances have revealed serious shortcomings in the process followed by the American Conference of Governmental Industrial Hygienists. Unpublished corporate communications were important in developing TLVs for 104 substances; for 15 of these, the TLV documentation was based solely on such information. Efforts to obtain written copies of this unpublished material were mostly unsuccessful. Case studies on the TLV Committee's handling of lead and seven carcinogens illustrate various aspects of corpo rate influence and interaction with the committee. Corporate representatives listed officially as "consultants'- since 1970 were given primary responsibility for developing TLVs on proprietary chemicals of the companies that employed them (Dow, DuPont). It is concluded that an ongoing international effort is needed to develop scientifically based guidelines to replace the TLVs in a climate of openness and without manipulation by vested interests. . Key words: unpublished corporate communications, TLV committee, carcinogen, conflict of interest, industrial experience, OSHA standards _ INTRODUCTION The Threshold Limit Values (TLVs) published by the American Conference of Government Industrial Hygienists (ACGIH) have been widely adopted as workplace exposure standards. The ACGIH values have been very influential over the past 40 years in Belgium, West Germany, Austria, Italy, The Netherlands, Portugal, Den mark, Sweden, Finland, Norway, Spain, Switzerland, the United Kingdom, and Japan [Toyama, 1985; Vigliani et al., 1977]. In the developing countries as well, the TLVs have been relied upon by governmental occupational health authorities [Noweir, 1986], However, it has nonetheless been widely recognized that the TLVs for chemical substances are in most cases poorly supported by scientific evidence. This is clear from even a casual review of the Documentation of the Threshold Limit Values and Biological Exposure Limits (5th Edition, 1986). West Germany adopted the ACGIH values in 1955 and has been influenced by the ACGIH in setting exposure limits ever since. But the German authorities, upon review of the documentary adequacy of their MAKs, concluded that less than 10 percent of the limits were based on "sufficient The authors are independent consultants in occupational and environmental health. Address reprint requests to Barry Castleman, 1722 Linden Ave., Baltimore, MD 21217. Accepted for publication September 14, 1987. 1988 Alan R. Liss, Inc. PLAINTIFF'S EXHIBIT SA-583 532 Castleman and Ziem animal tests and/or field experience" [Henschler, 1984], This finding, based initially on a review of 150 substances, has been more recently corroborated by review of 300 more substances on the German MAK list [Henschler, 1985]. ACGEH's TLVs have been directly criticized by both industry and labor representatives for scientific inadequacy [Henderson, 1975; Samuels, 1981]. This report examines the historic role of industry in the development of the TLVs. Role of Industry in TLV Process The American Conference of Governmental Industrial Hygienists established a Committee on Threshold Limits which issued annual reports starting in 1946. ACGIH was and continues to be a voluntary organization with no formal ties to the U.S. government despite its name. Its members were initially federal, state, and local officials, and within a few years, academics and well-known industry consultants were also included. From the beginning, the TLVs were acknowledged to involve a balancing of health considerations and cost to industry [Report, 1948]. Industry data were invited. In order to understand this interaction, it is necessary to appreciate the dependence of the TLV committee on information from industry, especially prior to the 1970s. In the United States, government toxicologists and industrial hygienists of this era had very limited access to knowledge of dose-effect relationships in industry. There was no federal regulation of general industry workplace hazards until 1971; and state and local agencies were thinly staffed and minimally funded. These agencies had little if any regulatory power and lacked laboratory and other technical resources so vital to the surveillance of hazards in industry. At U.S. universities, faculty occupational health professionals depended upon industry goodwill for research funding, consulting, and field experience and jobs for their students. Government funding for occupational health research was virtually nonexistent. Dr. John Knox, medical officer for Turner and Newall, an asbestos-based multinational corporation headquartered in Britain, recorded his impressions in notes of a 1960 visit to his company's U.S. subsidiary [Knox, I960]: "The legislative framework under which industries operate in the U.S. A. makes it difficult for me here to follow the lines of thought which prompt action over there in the matter of standards of industrial practice. In many industries, the employers seem so far in front of legislation as to have created a special code of practice for themselves." It was well recognized that, to the extent that data existed on exposures to toxic agents and ill health in industry, they had been mostly developed by industry. Industrial concerns in the U.S. were in no way compelled to share what they knew. Under the chairmanship of toxicologist Herbert Stokinger, the TLV committee first tried the approach of prodding industry by issuing a "Notice of Intent" to change some TLVs in 1964. A number of companies responded, supplying data, leading to 9 of 23 new additions that year [Notice of Intent, 1965]. Stokinger wrote to the Manufacturing Chemists' Association (now Chemical Manufacturers Association) [Stokinger, 1964]: "Th has i trial welc By Occupati< TLVs on file and t establishi Ovi at the Me TLVs. Th the Indust Data" [Cc this way Industrial repository the e: industry's Stokinger. responsibi "The to de deve relia Regarding "The long relea legal able The follov the U.S. ( federal sta well have In th rapport wi on at least TLV ic work .i V.K. Rove. S "xsed initially ..ttf iew of 300 Ii7"j TLVs have es for scientific elopment of the ists established a in 1946. ACGIH ties to the U.S. state, and local istry consultants ; a balancing of ita were invited, te dependence of the 1970s. lygienists of this lips in industry, ards until 1971; . These agencies hnical resources depended upon r id jobs for iS virtually i asbestos-based ressions in notes he U.S.A. ch prompt In many as to have posures to toxic cd by industry, hat they knew. TLV committee ntent" to change lata, leading to 9 er wrote to the ers Association) I Corporate Influence on TLVs 533 "This was particularly encouraging in view of the fact that the committee has never had a significant amount of voluntary contributions from (indus trial sources) as long as I can recall (13 years), despite annual exhortations welcoming such information. " By 1966, a committee of the Industrial Medical Association (now American Occupational Medical Association) expressed concern over the growing impact of the TLVs on industry. At the same time, it was acknowledged that industry had data on file and the means to develop more data that could "contribute constructively to the establishment of realistic TLVs" [Golz et al., 1966]. Over the years, Stokinger had had a number of meetings with industry groups at the Mellon Institute/Industrial Hygiene Foundation to discuss proposed changes in TLVs. The TLV committee's 1968 "Notice of Intent" even invited industry data via the Industrial Hygiene Foundation "Repository of Anonymous Occupational Health Data" [Committee, 1968]. However,, little if anything of value was ever obtained in this way [Stokinger, 1986-87]. From the time the idea was first suggested, the Industrial Medical Association had apprehensively observed that documents in a data repository might be subject to subpoena in damage suits [Minutes, 1967]. In 1969, Stokinger described the lack of appropriate industrial hygiene data as the greatest problem facing the TLV Committee. Describing the American chemieal industry's contribution of data on new substances to the TLV committee'as "pathetic", Stokinger, who was employed at the U.S. Public Health Service, addressed industry's responsibility directly [Stokinger, 19691: ' "The TLVs are industry's values. . . industry has the sole responsibility to develop data on its own products; government is not in a position to develop the facilities to handle the problem in total, nor should it, when reliable toxicologic consultants are now available." (Original emphasis) Regarding chronic animal exposure data, Stokinger commented [Stokinger, 1969]: "The data are in short supply because industries either do not develop long-term studies, or if they do, more often than not, do not see fit to release the data in the open literature. Various reasons are given for this: legal protection of their products, lack of staff time to put data in publish able form. Whatever the reason, the data are not forthcoming. " The following year, (1970), the Occupational Safety and Health Act was passed by the U.S. Congress, and virtually the entire 1968 list of TLVs became enforceable federal standards. In future OSHA standards development, the TLV committee could well have been expected to have a considerable influence. In the chemical industry, the Dow Chemical Company had developed some rapport with the TLV committee in the 1960s. Dow had provided unpublished data on at least 5-10 products, commented on the committee's documentation for specific TLVs, and discussed work published by Dow toxicologists and others around the world. In 1970, this relationship deepened, with the enlistment of Dow toxicologist V.K. Rowe as a "liaison member" of the TLV committee and his co-worker Theodore 534 Castteman and Ziem TABLE I. TLV Documentation Assignments Substance (trade name) 2,4.5 -T ethylene glycol ` vinyl chloride methyl bromide propylene glycol methyl ether ("Dowanol PM") methyl chloride 1,2 dibromoethane (ethylene dibromide) -1,2 dichloroethane o-chlorostyrene methylene chloride 1,2.4 trichlorobenzene vinylidine chloride _ dicyclopentadiene clopidol ("Coyden") . tricyclohexyltin hydroxide ("Plictran") chlorpyrifos ("Dursban") picloram ("Tordon") dimetholate 3.5 dinitro-o-tolamide ("Zoalene") dimethyl sulfate tris(2,3-dibromopropyl phosphate) styrene bis-chloroethyl ether 1,2,3 trichlorobenzene chloroform dipropylene glycol methyl ether ("Dowanol DPM") ethanolamine . 2-chloro-6-trichloromethyl pyridine ("N-Serve") crufomate ("Ruelene") chlorodifluoromethane chromates methomyl ("Lannate") ` perfluoroalkanes cyciopentane m-xylene, a,a` -diamine bromacil ("Hyvar X") diuron ("Karmex") dioxane calcium hydroxide cyclopentadiene dibromochloropropane cyanamide azodrin dicrotophos ("Bidrin") m-phthalodinitrile isophthalonitrile dioxin Person assigned Rowe Torkelson Torkelson Torkelson Morgan Morgan and Torkelson Torkelson Morgan Torkelson Morgan Zavon Torkelson Year first assigned 1970 1971 1972 1972 1973 1973 1974 1974 1975 1975 (continued) TABLE I. TLV Doc Substance (trade name) phosgene m-toluene diamine hexamethyl phosphor formamide dimethyl sulfoxide dichloromonofluoror 4,4'-methylene bis ( tetramethyl thiourea hexachlorobutadiene 3-amino, 1.2,4 triaz deodorized kerosene toluene concentrate acrylonitrile Torkelson as alte Morgan joined tl active role in the decade. Torkelsoi mittee on c o The n*..ate bility for review representatives fc which little or no Torkelson ' halogenated hydt "Plictran"). By t know about Du "Moca", hexame pesticides ("Lam substantial econo and other produc full flow of infor The 1970s workers, the env Table I. The derr large number of 1 case would be tf workplace expost ofTLVs, regulat icity by the TLV ations contesting Consumer Produ to reduce the cos More'' 'r. Manufactu of claim that the us  _________ Year first assigned 1970 1971 1972 1972 1973 1973 1974 1974 1975 1975 (continued) Corporate Influence on TLVs 535 TABLE I. TLV Documentation Assignments (Continued) Substance (trade name) Person assigned phosgene m-toluene diamine hexamethyl phosphoramide formamide dimethyl sulfoxide dichloromonofluoromethane 4,4'-methy!ene bis (2-chloroaniline) ("MOCA") tetramethyl thiourea hexachlorobutadiene 3-amino, 1,2,4 triazole ("Amitrol") deodorized kerosene toluene concentrate acrylonitrile Morgan Morgan Zavon Torkelson Year first assigned 1975 1976 1976 1976 Torkelson as alternate industry liaison member. Dupont industrial hygienist James Morgan joined the committee in 1972, and together with Torkelson, he played an active role in the work of the committee for the rest of the 1970s and into the present decade. Torkelson and Morgan became two of the four members of the new subcom mittee on carcinogenic substances established in 1972 [Minutes, 1972], The minutes of the TLV committee in 1972-1976 show that primary responsi bility for reviewing documentation in developing TLVs was borne by corporate representatives for major products of their own companies and new products about which little or nothing had been published. Torkelson was well situated to know about the toxicity of Dow Chemical's halogenated hydrocarbons and pesticides ("Tordon", "Ruelene", "Dursban", and "Plictran"). By the same token, Morgan would appear to have been well placed to know about DuPont's carcinogenic products (dimethyl sulfate, lead chromate, "Moca", hexamethyl phosphoramide), chlorofluorocarbons ("Freon" products), and pesticides ("Lannate", "Hyvar X", and "Karmex"). Dow and DuPont also had substantial economic reasons for wanting to influence the TLV committee on these and other products. But these economic considerations were adverse to the free and full flow of information from the companies. The 1970s would see government legulators charged with the protection of workers, the environment, and consumers very busy with some of the chemicals in Table I. The demonstration of vinyl chloride's carcinogenicity cast a shadow over a large number of halogenated hydrocarbons. A reference point for regulators in every case would be the currently accepted limit for maximum human exposure, namely workplace exposure. And since most of OSHA's limits were from the aging 1968 list of TLVs, regulators looked to the current TLV lists and designations of carcinogen icity by the TLV committee for guidance. The chemical companies and trade associ ations contesting standards at OSHA, the Environmental Protection Agency, and the Consumer Product Safety Commission included Dow and DuPont. High TLVs tended to reduce the costs of regulation to the chemical industry. Moreover, there were liability considerations in addition to regulatory ones. Manufacturers of products involved in damage suits before juries readily resort to the claim that the use of the product was not expected to exceed the TLV and was thus 536 Castleman and Ziem considered "safe." The "TLV defense" offers manufacturers the plausible deniability that any harm sustained was foreseeable. Where a manufacturer has evidence that the exposure involved was in fact below the TLV, this may even be used to support a denial that the product caused health impairment. Duplicity of corporate representatives clearly angered longtime Massachusetts occupational health official and TLV committee member Hervey Elkins, who, writing a letter of retirement to Chairman Stokinger in 1975 [Elkins, 1975], stated: "In looking over the new documentation I was taken aback by that for ethylene glycol; the limit of 100 ppm was found intolerable by sedentary volunteers in a few minutes (or seconds). I believe that {industry represen tative} recommended this figure. In spite of his knowledge he seems to come up with some recommendations for TLVs that are way too high, in my judgment. The same can be said for most of the other industry representatives we have had. In many cases they recommend a TLV much above the action levels used in their own plants. " By the time of Elkins' complaint, Dow Chemical had long been assigning internal corporate exposure limits for toxic substances. Other firms, including Rohm and Haas, had also decided to adopt this practice. Corporate workplace exposure limits have served as a managerial tool both for substances with assigned TLVs and others for which TLVs had not been adopted [Paustenbach and Langner, 1986]. Regulatory and liability concerns appear to have deterred corporate management from publishing these lists and supporting rationales--despite their obvious practical value and potential importance in preventing occupational disease. MATERIALS AND METHODS The 1986 Documentation ofthe Threshold Limit Values and Biological Exposure Indices was reviewed for all chemical substances. Where reference appeared in the text to unpublished communications and internal corporate reports, etc., a determi nation was made as to whether such information had been important in setting the TLV or classifying the substance's carcinogenic status. This was a matter of judge ment based on the full text for each chemical substance listed. Due to the wide variation in type and quantity of information used as a basis for the various TLVs, rigid criteria could not be used; it is presumed that different experts conducting such a review would come up with slightly different lists of TLVs for which unpublished corporate communications would be judged important. The important communications can be generally described as animal data, data from tests on human volunteer subjects, and "industrial experience." Communications coming from corporations and trade associations are in many cases so identified in the Documentation. However, in many other cases only the names of individuals are published in the Documentation. The institutional affiliations of these people at the times they sent information to the Committee on Threshold Limits have been investigated in various ways. The sources checked included: con temporary publications by the same people; past directories of professional associa tions (American Industrial Hygiene Association, American Occupational Medical Association, A' telephone for th Attempts cited in the Doi of specific refei in the Documen pattern of irreti was reanalyzed Committee filej in Cincinnati. 1 who had server the 1970s, cont; Though A Board of Direct RESULTS For a tota reliance on unpi were assigned T (Table Iff. s the above uupu corporations. Of the 89 referenced souri unpublished cor all identified in I There was was placed on i sixth of the nun mentation. Of the 17 committee, nine Jersey Departm unobtainable fro in NIOSH files, longtime commi cals. Stokinger : but came over important unput form for indeper Industrial Expe The TLV included reports content of'of a senten r whether there v. .,-\)ltthe ) support a ssachusetts ho, writing it for ntary esenT1S tO ;h, in ustry much i assigning Jing Rohm i exposure TLVs and ier, 1986]. :ment from .tical value l Exposure ired in the a determisetting the r of judgei the wide ous TLVs, icting such npublished data, data re in many :s only the affiliations Threshold uded: conlal associaal Medical Corporate Influence on TLVs 537 Association, ACGIH); and retired members of the TLV committee contacted by telephone for their recollections. Attempts were made in several ways to obtain copies of unpublished material cited in the Documentation. The New Jersey Department of Health requested copies of specific references on 67 substances in 1985 from ACGIH and companies named in the Documentation for the purpose of developing chemical fact sheets (later, as a pattern of irretrievable unpublished corporate statements emerged, the information was reanalyzed for this paper). An examination was also made of the historic TLV Committee files at the National Institute for Occupational Safety and Health (NIOSH) in Cincinnati. The surviving files kept there by United States government employees who had served on the TLV Committee, covering years from the late 1950s through the 1970s, contained a small number of letters and reports cited in the Documentation. Though ACGIH has copies of TLV committee minutes for the last 10 years, the Board of Directors would not grant access to them [Kelly, 1986-87]. RESULTS For a total of 89 substances, the 1986 TLV Documentation placed important reliance on unpublished corporate communications (Table II). Another 15 substances were assigned TLVs solely on the basis of unpublished corporate studies and reports (Table HI). This investigation was able to locate written copies of far less than half of the above unpublished corporate material from the NIOSH files, ACGIH, and the corporations. Of the 89 substances in the first group above, corporate affiliation of the referenced source person was not published for 25. For the 15 TLVs based solely on unpublished corporate communications, the companies providing information were all identified in the Documentation. There was thus a total of 104 substances for which important or total reliance was placed on unpublished corporate communications. This accounts for over one sixth of the number of (less than 600) chemical substances listed in the 1986 Docu mentation. Of the 17 corporations asked for documentation they had provided to the TLV committee, nine sent old documentation or commented on their work to the New Jersey Department of Health. The unpublished documentation in most cases was unobtainable from the companies (Table IV) and the historic TLV committee records in NIOSH files. There were no files available from ACGIH itself; nor did former longtime committee members (Stokinger, Elkins) have personal files on the chemi cals. Stokinger admits that some of the information was never conveyed in writing but came over the telephone [Stokinger, 1986-87]. In any event, most of these important unpublished corporate communications are now unobtainable in written form for independent scientific examination. Industrial Experience The TLV committee's reliance upon unpublished corporate communications included reports of "industrial experience" on dozens of chemical substances. The content of these reports rendered in the Documentation often appears in just the space of a sentence or two (Table V). The scientific community is left unable to determine whether there was more information originally conveyed; and where there was, no 538 Castleman and Ziem TABLE II. TLVs for Which Unpublished Corporate Data was Important* Substance acrylic acid fonofos acrylonitrile hydroquinone asphalt fumes2 . isooctyl alcohol2 benomyl isophorone benzene2 2-isopropoxyethanol n-butyl acrylate lead chromate sec-butyl alcohol2 manganese and compounds n-butyl glycidyl ether manganese tetroxide caprolactam2 methacrylic acid carbon disulfide2 methomyl catechol 4-methoxyphenol chlorinated camphene (60%) methyl n-butyl ketone chlorinated diphenyl oxide methyl chloride chloracetaldehyde - methyl 2-cyanoacrylate chloracetyl chloride methylene bis-4-cyclohexyl-isocyanate2 chlorodifluoromethane2 methylene bisphenyl isocyanate o-chlorostyrene methylene chloride o-chlorotoiuene 4,4' methylene dianiline chlorpyrifos methyl isocyanate copper2 metribuzin2 cyclopentadiene monocrotophos cyhexatin paraquat2 dibutyl phthalate piperazine dihydrochloride dichlorodifluoromethane propionic acid dichloroethylene quinone2 dichlorofluoromethane resorcinol- 2,2 dichloropropionic acid rosin core solder pyrolysis products2 dichlorotetrafluoroethane silicon tetrahydride dicrotophos silver and compounds2 dicylopentadienyl iron sulfuryl fluoride diethyl phthalate sulprofos2 diglycidyl ether tetraethyl lead2 dimethyl acetamide tetramethyl lead2 dimethylamine . tetrahydrofuran2 dimethylformamide thioglycolic acid dimethyl sulfate 1,2,4 trichlorobenzene2 diphenylamine2 trichlorofluoromethane di-sec-octyl phthalate 1,1,2 trichloro 1,2.2 trifluoroethane endrin2 trimethyl phosphite ethion ' ethylene dichloride2 ethylenimine tungsten compounds2 vinylcyclohexene dioxide xylidine2 n-ethyl morpholine zinc stearate fenamiphos2 Includes substances- assigned carcinogenicity status. Does not include papers presented at scientific conferences. ^Corporate affiliation of correspondent not published in Documentation of TLVs. TABLE III. Docum Substance n-butyl lactate o-sec-butylphenol clopidol dinitolmide divinylbenzene ethyl amyl ketone* 2-hydroxypropy! acrylate isophorone diisocyanate n-isopropyl aniline methyl acetyler" propadiene mixture nitrapyrin phenylphosphine tetrasodium pyrophosphate triphenyl amine m-xylene a.o' diamine Includes an industn. workers exposed to c- way exists to loo statements publis whether the stater Because of industry could pe these communicat The infomu mentation raises t tine" medical ext that "no evidence chloride? V u What was ti ity studies on acn anated :tsa ine ented at scientific Corporate Iijfluence on TLVs 539 TABLE ID. Documentation of TLVs Solely by Unpublished Corporate Communications Substance Animal data, acute Animal data. subacute or chronic Human data ' Source, year n-butyl lactate o-sec-butylphenol clopidol _ dinitolmide divinylbenzene ethyl amyl ketone* 2-hydroxypropyl acrylate isophorone diisocyanate n-isopropyl aniline methyl acetylene- propadiene mixture nitrapyrin phenylphosphine tetrasodium pyrophosphate triphenyl amine m-xylene a.a' diamine X X X X X X X X X X x (2 yrs;teratol) x (2 yrs:teratol) x (30 da.) x (4 wks) x (4 mos.) x(93 days) x(90 day) X X X X X Philips Endoven 1969 British Petroleum 1972 Dow/1977 Dow/1973 Dow/1973 Dow/1977 Shell/1958. 1965 Dow/1977 VeraChemie Dow/1977 Dow/1964 Dow DuPont/ 1970 Dow/1977 Kodak/ 1973 Dupont/ 1973 SherwinWilliams 1978 Includes an industrial hygiene bulletin by Shell Chemical Corporation claiming no systemic effects in workers exposed to concentrations above the TLV recommended by the company. way exists to look up the original source and resolve questions about the basis of statements published in the Documentation, including methodology utilized and whether the statement was based on any study or merely an impression. Because of the weight given to these reports and the great value of studies industry could perform on the workers exposed to these agents, special attention to these communications is warranted. The information provided by companies and published by ACGIH in the Docu mentation raises obvious and fundamental questions. What exactly did Dow's "rou tine" medical examinations and any analysis performed on them show to establish that "no evidence of over-exposure" occurred at the reported concentrations of methyl chloride? What tests were conducted and what analysis was carried out by Dow? What was the scientific content and methodology of the unpublished negative mortal ity studies on acrylonitrile, benzene, dimethyl sulfate, and ethylenimine? What were 540 Castieman and Ziem TABLE rv. Requests of Data from Corporations_____________________ Number of Corporation___________ Chemicals______________________________ Results Dow Hooker Hercules Crown Zellerbach FMC Rohm and Haas American Cyanamid DuPont 33 No information received. 2 The company provided a report for one (chlorotoluene) of the two requested chemicals. Study methods and results were described (animal study). 2 For one chemical (chlorinated camphene), Hercules stated they had sold the operation to Nor-Am and stated any toxicologic information "must now come from Nor-Am." (Hercules did not say they no longer had the information.) Nor-Am stated they no longer produced it, and that much of the correspondence and reports had been discarded. For the other chemical (Rosin core solder pyrolysis products), the smdy was provided with detailed methods and results. However, inflammation and hyperemia in multiple organs for both controls and exposed animals causes one to wonder about (inadvertent) exposure of "controls." 1 The Documentation states, "industrial experience has been good over the years." Crown Zellerbach's correspondence describes 3 yr experience manufacturing the chemical (catechol), with only "a few mild toxic reactions." CZ notes "no physical abnormalities . . . noted by observation" (not stated whether all workers had physical exams) or in "multichannel blood tests" (type, frequency, other methods unspecified). While catechol is an irritant, there is no mention of the use of symptom questionnaires or lung function tests in this 1975 communication. 1 No information received (carbofuran). 2 Significant material sent describing study methods and results for animal studies on both chemicals. Study report noted "Squamous metaplasia" of nasal mucosa, thought secondary to irritation (ethyl acrylate). This effect not noted in Documentation. 2 The company provided information on one (phorate) of the two chemicals requested. This was an inhalation smdy (level unspecified) for 8 hr involving 12 animals observed for 7 days after exposure. The report merely says "there was no evidence that they were affected in any way." There is no mention of whether pathologic studies or biologic monitoring were conducted let alone reporting of such findings. The criteria for no effects were unspecified. 7 Some information (not always complete) was sent for all 7 chemicals. The Documentation states that there were "no complaints of illness" and no abnormal liver function tests in employees exposed at roughly half the TLV for several years (dimethyl formamide). The information provided by the company to NJDOH does not appear to be a reference upon which such a statement could have been based; the original basis for the statement not be located. An epidemiological smdy of 143 workers exposed to dimethyl sulfate showed that few deaths from respiratory cancer occurred among them while employed by DuPont. The work force was not broken down in terms of either time elapsed from onset of exposure or duration of exposure to DMS. No follow-up of ex-employees and retirees was done. (continued) TABLE IV. R Corporation Western Electr Sherwin Willia Mobil Oil Ethyl Corp Koppers Eastman Kodak Union Carbide Shell Oil B.F. Goodrich orotoluenel of the md results were lercules stated J stated any from Nor-Am." he information.) . and that much :n discarded. For >lysis products). .ods and results, multiple organs .ses one to controls." riefice has bee-n s correspondence he chemical actions." CZ ed by ers had physical type, frequency, jl is an irritant, i questionnaires or ation. ethods and results Jv Virt noted :U -.iect not tphorate) of the .alation study inimals observed ely says "there n any way." c studies or lone reporting of were unspecified, as sent for all 7 '. there were "no .ser function tests TLV for several ltion provided by to be a reference seen based: the ated. An posed to dimethyl tratorv cancer . DuPont. The s of either time on of exposure to J retirees was (continued! Corporate influence on TLVs 541 TABLE IV. Requests of Data from Corporations (Continued) Corporation Number of Chemicals Results Western Electric Sherwin Williams Mobil Oil Ethyl Corp Hoppers Eastman Kodak Union Carbide Shell Oil B.F. Goodrich No data were provided to substantiate Zapp's communication (1970) to the ACGIH that methylene bis (4-cyclohexyliso- cyanate) was less toxic on inhalation than TDt. Dupont's subacute study on dogs, found to have no skin irritation or sensitization effects, unlike "results. . . previously reported" (tetrahydrofuran), was given greater credence than the published positive studies by the TLV committee "because of the greater number of animals involved". The DuPont study used 4 dogs. A 90-day study of phenylphosphine contained adequate discussion of methods and results. DuPont's study of m- xylene a.a' diamine found "generally mild" sensitization in all 10 guinea pigs tested. This is mentioned in the Documentation as "evidence of sensitization" without noting.that all animals were affected. The Documentation refers to a subacute study by DuPont in 6 rats as one which "caused no fatalities" (dicyclopentadienyl iron). The Documentation omits data showing that in addition to irritability and weight loss, all 6 rats showed testicular atrophy. It is unclear whether these effects were ever communicated to the TLV committee (report was obtained from the company but not in the TLV files). 1 Western Electric did not provide the correspondence for - isophorone, but it was obtained from the TLV committee files and consisted of five sentences noting two symptom complaints, urinanalysis and "kidney function checks." No methods discription was given, nor was the number of employees noted, nor whether questionnaires were used or if they waited for employees to complain. No medical surveillance data was provided even in summary form. 1 No response to request for information (m-xylene a.a' diamine). 1 Epidemiologic study conducted on employees (for eye effects only) with exposure levels evaluated, study methods described. Unclear whether study was ever published (trimethyl phosphite). 1 No response to request seeking information (tetraethyl lead). I The Documentation states that "a surveyj of 180 men employed in work involving resorcinol revealed that none complained of irritation or discomfort at exposure levels of 10 ppm." The company provided no information about any study but merely sent a safety data sheet on the chemical. No discussion of methods was provided in Koppers' letter to the TLV committee, which was located in the committee's files. 4 Letters on animal studies were located for 2 chemicals (o. chlorotoluene, triphenylamine). Observations on workers could not be located for the other 2 (dibutyl phthalate. di sec-octyl phthalate). 4 No response to request for information. 3 Shell no longer makes the 3 chemicals and states that correspondence concerning them is no longer available. 1 The Documentation states "observations in the rubber industry have revealed no adverse effects from many years of inhalation of zinc stearate dust." The company had discarded the correspondence but stated that there was no "organized study of workers exposed to zinc stearate." 542 Castieman and Ziem TABLE V. Unpublished Industrial Experience Cited in TLV Documentation acrylonitrile asphalt fumes benzene sec-butyl alcohol n-butyl lactate caprolactam carbofuran carbon disulfide catechol chlorinated camphene (toxaphene) chlorodifluoromethane o-chlorotoluene dibutylphthalate _ diethylphthalate di-sec-octy Iphthalate diglycidyl ether dimethylformamide dimethylsulfate diphenylamine endrin ethylene dichloride eth>lenimine hydroquinone Monsanto. 1981, epidemiology negative on carcinogenic effects. Hammond (Humble Oil) 1968--opinion of industrial hygienists that conditions were satisfactory at 10 mg/m3. Ott et al. (Dow), 1975, epidemiology "revealed no excess mortality. " Banks (Shell Chemical Company) 1966--hygienist reports that "many years of industrial experience (at 100 ppm) have resulted in no difficulties." Turner (British Petroleum), 1972, reported that 7 ppm was not found to be objectionable or injurious. Ferguson (Allied Chemical), 1972, reports on 143 workers, "some of whom were exposed up to 17 years to vapor concentration as high as 5-10 ppm without any evidence of damage to health." Tobin (FMC Corporation), undated, given as source: "Workers exposed to concentrations approaching 0.1 mg/m3 per day have not shown any effects." Calhoun (American Viscose), 1968, reports no cases of carbon disulfide poisoning since 1942, when exposures averaged below 2.5 ppm. Crown Zellerbach, 1975, referenced as reporting industrial experience has been good under adequately controlled conditions. Hercules, Inc., 1969. reports that review of records of 137 employees, "some" exposed up to eighteen years, "failed to reveal any adverse effects that could be associated with toxaphene." Reinhardt (DuPont), undated, reports that cardiac arrythmias are not considered a possibility "under currently recommended industrial hygiene practices." Hopton (Hooker Chemical), 1962, reports that no cases or dermatitis or poisoning from this compound had been encountered. Raleigh (Kodak), undated, reports workers exposed to 1-6 ppm of mixed phthalates had no phthalates in their blood and had no peripheral polyneuritis. White (Shell Chemical Company). 1962. recommends a ceiling limit of 0.5 ppm "on the basis of a no-effect level in animal studies and industrial experience." (TLV-TWA: 0.1 ppm) DuPont, undated, reports no complaints of illness and no abnormal liver function tests at about one half the TLV. DuPont, 1972, epidemiology "covering a period of 15 years" and an update in 1976 show no excess of lung cancer in exposed workers. Demehl (Union Carbide), 1967, cites "industrial experience" in recommending a satisfactory operating level. On this basis, the same value was selected as the TLV (10 mg/m3). Jager (Shell): no medical effects seen with 233 workers, comparing them before and after 10 years' exposure to endrin and related pesticides (body weight, blood pressure. WBCs. and SREs). Fassett (Kodak), 1964, "Experience in one plant indicated that concentrations in the range of 25 to 50 ppm were safe for prolonged exposure." Dow report of BASF. 1973, epidemiological study "revealed no evidence" of carcinogenicity in 144 workers "some of whom had 40 years' experience." Fassett (Kodak), undated, reports that clinical and environmental studies of workers "confirm that no systemic effects arise at (the TLV)." (continued) TABLE V. Ur manganese anc methyl n-butyl methyl chlorid> methylene bisf isocyanate 4.4' methylem monocrotophoparaquat propionic acid resorcinol tetraethyl lead tetramethyl U trimethyl phosp tungsten compi vinyl cyclohexe zinc stearate the parameu effect that a t that could be Similar upon by the ` larger numb reliance on u * a' -t'-J-Cffects. lal hygienists that i excess t reports that ml have resulted in ppm was not found ' workers, "some concentration as age to health." rce: "Workers nr' per day have .ses of carbon averaged below industrial itrolled conditions, ds of 137 irs, "failed to ted with arrythmias are not unended industrial i cases or dermatitis Oil'"'ted. -6 ppm of /had no lends a ceiling limit animal studies and s and no abnormal of 15 years" and an n exposed workers, experience" in )n this basis, the orkers. comparing Jrtn and related and SREs). indicated that ere safe for Jv "revealed no some of whom had id environmental effects arise at (the (continued) Corporate Influence on TLVs 543 TABLE V. Unpublished Industrial Experience Cited in TLV Documentation (Continued) manganese and compounds Whitman (Bethlehem Steel), 1976, reports no cases of manganism in workers exposed for years to 1 to 5 mg/m3 of manganese dioxide dust. methyl n-butyl ketone Raleigh (Tennessee Eastman Co.), 1976, reports "no history of muscular weakness, parathesia, loss of coordination or clinical evidence of neuropathy in 37 employees engaged for 3 years in the manufacture of (methyl n-butyl ketone)." methyl chloride Dow. undated, reports that a "routine periodic medical program did not identify evidence of overexposure to methyl chloride" at concentrations averaging 30 ppm. methylene bispheny! Imperial Chemical Industries, 1962, repons no cases of skin isocyanate . irritation during early industrial experience handling this compound. 4,4' methylene dianiline Dow, 1977, reports "no morbidity findings" for exposures ranging from 0.03 to 0.4 ppm over 26 years. monocrotophos Shell Chemical, undated, reports no decrease in field workers' cholinesterase concentrations following exposure. paraquat Gage (Imperial.Chemical Industries), 1968. is cited as reporting that "no serious injury or illness resulted from eight years' agricultural use of paraquat." propionic acid Dow, 1977, reports that, at reported exposure levels, no irritation was noted. Medical reports include mild eye redness and one case of mild cough and asthmatic response. resorcinol Koppers Company, 1974, reports that none of 180 men exposed to 10 ppm complained of irritation or discomfort. tetraethyl lead Linch (DuPont), 1968, reports that exposures averaging about 20% tetramethyl lead over the TLVs produce average urinary lead concentrations "not ' significandy elevated above a high normal" (no values above 0.15 mg/1). Ethyl Corporation, undated, reports that 3/4 of the TLV for tetraethyl lead "is a rough guideline for an allowable (TLV)." trimethyl phosphite Mobil Chemical Co., 1980 reports no ocular changes among 179 workers with exposures reported. "Plant exposure data could be interpreted to indicate that concentrations of I ppm certainly, and very likely 2 to 4 ppm, are without significant adverse effect." TLV raised from 0.5 to 2 ppm in 1982. tungsten compounds Demehl (Union Carbide), 1966, reports that "long industrial - experience" has indicated workers exposed to solely tungsten and its insoluble compounds do not develop pneumoconiosis. vinyl cyclohexene dioxide Demehl (Union Carbide), 1973, referenced as source: "In the U.S.. industrial experience over the past 10 to 20 years has been good." zinc stearate B.F. Goodrich Rubber Co., undated, referenced as source: "Observations in the rubber industry have revealed no adverse effects from many years inhalation of zinc stearate dust." (no concentrations given) the parameters and data underlying Hercules' unpublished communication to the effect that a review of employee medical records "failed to reveal any adverse affects that could be associated with toxaphene?" Similar questions arise over the nature and quality of industrial experience relied upon by the TLV Committee for 32 other chemical substances (see Table V). An even larger number of TLV substances were assigned exposure limits after significant reliance on unpublished corporate communications about animal experiments. 544 Castleman and Ziem In this survey, a TLV Documentation reference was counted as "published" even if it was from a manufacturer's safety data sheet or an unsupported statement published in a text by a corporate health professional. The brevity, age, and obscurity of such documentation raises serious questions of reliability despite the fact of such references being "published." Manufactureres' safety data sheets, while briefly noting chemicals' health ef fects or lack of effects, are not generally useful as primary sources for detailing the scientific basis of health effects statements. Safety data sheets are not written to convey the important data underlying statements like, "no health problems have been attributed to the use of this agent in industry"; or "it is an irritant but not a sensitizer." Yet TLVs are still based on such statements by manufacturers on safety data sheets issued in the 1950s (e.g., ethyl ketone, methylamine, nitromethane). The use of corporate safety data sheets of even recent vintage is inappropriate for documenting TLVs. . Some reports of "no adverse industrial experience" in the 1986 Documentation originally appeared in classical texts but were unsubstantiated by data and are now very old. In the case of morpholine, the text refers to the 1963 edition of Patty's toxicology text as a basis for saying that "no chronic effects have been reported." The primary source cited was a 1948 review on morpholine issued by the American Petroleum Institute. Patty's text was in large part written by industry professionals; and some of the statements appearing in the text, though unexplained there, went on to be cited as the basis for TLVs. Patty himself reported on ethyl acetate concentra tions he had measured during a period of "several months" during which time "no adverse symptoms or illnesses were observed." Patty, who was an industrial hygienist at General Motors, did not explain whether the observations made were those of physicians, himself, or other medically untrained management officials (supervisors, foremen, personnel managers). TLVs for Carcinogens The case studies of six carcinogenic materials will be considered next, in order to examine in some detail the work of the TLV committee in this important area. These summaries illustrate a number of ways in which the committee was informed and influenced by industrial parties. Some of the materials on the first lists of MACs, as they were called in the early years, were known or suspected of being human carcinogens. These agents included asbestos (1946), arsenic (1947), and chromates (1950). Threshold limits for these materials appear not to have been based on their carcinogenic effects, however. Arsenic In the case of arsenic and its compounds, the 1947 value was 100 fig/m3. The following year. Hill and Faning produced strong epidemiological evidence of a lung and skin cancer hazard in a factory making sodium arsenite sheep-dip [Hill and Faning, 1948]. Median room air concentrations of arsenic measured in the chemical plant were 71, 254, 373 and 696 /xg As/m3. Average urinary arsenic concentrations of the workers were in the range of 0.09 to 0.24 mg/liter [Perry et ai.. 1948], For reasons not explained at the time, the threshold limit for arsenic was raised in 1948 from 100 to 500 ftg/m3. In the first published documentation of the TLVs in 1962. the "subsequent experience" of the American Smelting and Refining Company was cited a; medical dir confidential health you v cancer." [S' Pinto' arsenic and cancer than arsenic leve arsenic bur ceded that arsenic !e\Health Adi cases in the The 1963 report sions of the same Asarc 1974], In 1 exposed wi from 500 tc for most su of the Occu The 7 for arsenic environmer of the TLV standard. 7 Stokinger. ment salan OSHA said [Occupatio in 1978. v,; 1978], The ' carcinogen production this proces Asbestos The ` "tentative" [Dreessen than 5 mill plants cleu moderate" exposure.' cWi$ "published" "hd statement jgt. ind obscurity te the fact of such .-micais' health efs for detailing the are not written to roblems have been it not a sensitizer." safety data sheets lane). The use of e for documenting '86 Documentation data and are now edition of'Patty's e been reported." d by the American stry professionals; ned there, went on acetate concentraig which time "no industrial hygienist ade were those of icials (supervisors, ered next, in order lis important area, ittee was informed a called in the early ese agents included >ld limits for these cts, however. as 100 /xg/m3. The evidence of a lung heep-dip [Hill and red in the chemical enic concentrations tal., 1948]. r arsenic was raised tion of the TLVs in Refining Company Corporals Influence on TLVs 545 was cited as supporting 500 fig/m3. The source of this information was the company medical director [Pinto, 1961; Documentation, 1962], In acknowledging Pinto's confidential report, Stokinger replied, "It was surprising to see what a clean bill of health you were able to produce, in view of the many implications of arsenic and lung cancer." [Stokinger, 1961). Pinto's work was published in 1963, showing that both employees exposed to arsenic and employees with "non-arsenic exposure" had a greater incidence of lung cancer than males in the state of Washington. The "exposed" group had urinary arsenic levels of 0.82 mg/liter, and the "unexposed" smelter employees had urinary arsenic burdens averaging 0.13 mg/liter [Pinto and Bennett, 1963). Pinto later con ceded that the latter group in this controversial report was in fact exposed to "low arsenic levels," but denied a suggestion published by the Occupational Safety and Health Administration (OSHA) that there had been under-reporting of lung cancer cases in the 1963 study [Pinto and Nelson, 1976). The National Institute for Occupational Safety and Health evaluated Pinto's 1963 report as showing an increase in lung cancer mortality, contrary to the conclu sions of the authors [Inorganic, 1975]. A 1974 mortality study on the workers at the same Asarco copper smelter confirmed their lung cancer hazard [Milham and Strong, 1974). In 1975, OSHA responded to mounting reports of lung cancer in arsenicexposed workers by proposing a reduction in the workplace standard for arsenic, from 500 to 4 /xg/m3. (The original standards was 500 because the 1968 TLV values for most substances were adopted en masse as enforceable standards with the passage of the Occupational Safety and Health Act of 1970). The Threshold Limits Committee of ACGIH followed by adopting two TLVs for arsenic trioxide in 1977; 50 /xg/m3 at smelters and 250 jtg/m3 in non-smelting environments. This aroused bitter resentment at NIOSH and OSHA, where the actions of the TLV Committee were seen as aiding the industry challenges to the government standard. The government researchers and regulators were especially piqued at Dr. Stokinger, who was then Chairman of the TLV Committee while drawing a govern ment salary at NIOSH. Referring to the actions of the TLV Committee on arsenic, OSHA said: "The detailed basis for arriving at these levels is not clear on the record" [Occupational Exposure, 1978). OSHA's final standard for inorganic arsenic, issued in 1978, was 10 /xg/m3 of air, averaged over an 8-hr period [Occupational Exposure, 1978). The TLV Committee first listed "arsenic trioxide production" as a human carcinogen in Appendix A of the TLV booklet in 1975. In 1980, arsenic trioxide production was reclassified as a suspect human carcinogen; and numerical TLVs for this process and for insoluble arsenic compounds were completely eliminated. Asbestos The TLV adopted by ACGIH in 1946 to 1970 for asbestos was based upon the "tentative" recommendations of a Public Health Service study published in 1938 [Dreessen et al., 1938). The P.H.S. survey showed that workers exposed to more than 5 million particles per cubic foot (MPPCF) of total dust in the air of asbestos plants clearly developed asbestosis. But the P.H.S. survey also found "early to moderate" asbestosis in workers with less than 50 MPPCF--years of cumulative exposure. The P.H.S. findings and those of an earlier medical survey by Pennsylvania 546 Castleman and Ziem labor authorities strongly indicated that workers eventually would develop asbestosis from exposures under 5 MPPCF [Fulton et al., 1935], Lung cancer among asbestos workers was first reported in the mid-1930s, and by 1939, German state insurance carriers were compensating lung cancer in combi nation with even slight asbestosis as an occupational disease [Baader, 1939]. Pathol ogists around the world continued to contribute data and comments on the coincidence of those two diseases through the 1940s. In 1949 the British government published powerful confirmatory statistical evidence: in 235 deaths in which asbestosis had played a role, fully 31 (13.2 percent) also involved cancer of the lung or pleura [Annual Report, 1949]. The old 5 MPPCF threshold was never regarded as safe by leading asbestos industry consultants (Drs. Leroy Gardner, Arthur Vorwald, and Anthony Lanza). A similar lack of faith in this TLV as an index of safety was expressed publicly and privately in the 1940s, 1950s, and 1960s by executives and health professionals of the leading asbestos companies in the United States and the United Kingdom, as well as health authorities in these and other countries [Castleman, 1986]. In 1964, the old TLV for asbestos was repeatedly criticized by government and industry speakers at a widely publicized conference on asbestos held by the New York Academy of Sciences [Ann. N.Y., 1965], By this time, it was evident that nearly half of all asbestos insulation workers, whose average exposure was of the same order of magnitude as the TLV, were dying from occupational cancer and asbestosis. The ACGIH Threshold Limits Committee had included asbestos industry con sultants from its earliest years. Industrial hygienist Manfred Bowditch, who was on the Committee in 1946 and 1947, was then also trying to fulfill contracts the Saranac Laboratory had made with the asbestos industry [Castleman, 1986]. Bowditch's deceased predecessor at Saranac, Leroy Gardner, had performed studies in confidence for asbestos manufacturers, and the manufacturers wanted to publish some of the results (not the animal studies showing asbestos causing lung cancer, however. Dr. Arthur Vorwald, the next director of the Saranac Laboratory, accommodated asbestos industry sponsors with his publication of Gardner's non-cancer related research in 1951 [Castleman, 1986; Vorwald et al., 1951 and Vorwald, 1948], That year, he joined the Threshold Limits Committee, on which he served until 1956. During these years, Vorwald evaluated at least 30 cases of suspected and proven asbestosis and cancer, many of which were the subject of compensation claims, for companies in the United States and Canada. He also conducted a confidential animal inhalation study which appears to have re-confirmed asbestos' carcinogenicity in the early 1950s; however, this was never discussed in Vorwald's publications (Castleman, 1986: Vorwald, 1952], Dr. Paul Gross, at the Industrial Hygiene Foundation (since 1971. Industrial Health Foundation), became a member of the Threshold Limits Committee from 1964-1983. Gross' consulting work on asbestos included case pathology reviews for Johns-Manville in the 1950s and confidential animal research on brake drum dust for Johns-Manville in the 1960s [Castleman, 1986], As a member of a U.S. Public Health Service committee in 1969, Gross secretly provided draft copies of a report to three asbestos companies. Dr. Robert deTreville, President of the Industrial Hygiene Foundation, inviting comment, explained: "(W)e will attempt to see that needed corrections are introduced by Dr. Paul Gross, a member of the Committee'' [deTreville, 1969]. In 1976, Gross resigned from a committee of the National Academy of Sciences, consulted for 1976]. Upon subcommittee ACGIH more stringer was circulate may have an (published) n Commenters or to the "F deTreville at consideratior asbestos: A limit technic: tos. Crc asbestot establis are obt; equippe This idea of: before the pu Over tf TLV for all % microscopy. Finally, in l1-. contrast mic; through forrr the most abu By this time, for 4 yr in tl only (6.2 f-'o asbestos. No since 1980. It is no' follow in de\ vative goverr OS HA and t leading reco: Vinyl Chlori Upon r proposed lo\ Farland. 196 1961. where asbestosis 1930s, and :ancer in combir, 1939]. Patholi the coincidence nment published h asbestosis had i lung or pleura leading asbestos thony Lanza). A sed publicly and >fessionals of the zdom. as well as "in 1964, the old try speaker's at a iemy of Sciences of all asbestos of magnitude as os industry con:ch, who was on acts the Saranac 86]. Bowditch's ies in confidence h' L^-ime of the i-ver. ', accommodated n-cancer related aid, 1948], That rved until 1956. cted and proven ition claims, for nfidential animal nogenicity in the ions [Castleman, 1971, Industrial Committee from logy reviews for ke drum dust for S. Public Health a report to three lustrial Hygiene see that needed Committee" [deationai Academy Corporate Influence on TLVs 547 of Sciences, amid charges of improperly sharing information with a company he consulted for -- the issue was health effects of asbestos in drinking water [Wade, 1976]. Upon joining the Threshold Limits Committee, Gross became chairman of the subcommittee on insoluble respirable dusts [Minutes, 1965]. ACGIH's Threshold Limits Committee briefly considered having a separate, more stringent TLV for the crocidolite variety of asbestos. A 1968 Notice of Intent was circulated, "so that industry-connected individuals principally, but others also, may have an opportunity to help shape the deliberations of the Committee prior to its (published) recommendation of tentative changes in the 1967 Threshold Limits List. " Commenters were asked to write either to Dr. Stokinger at the Public Health Service, or to the "Repository of Anonymous Occupational Health Data" in care of Dr. deTreville at the Industrial Hygiene Foundation (Committee, 1968). "Revisions under consideration. . . proposed for 1968 List (of TLVs)" included the following for asbestos: A limit of 5 MPPCF, based on impinger samples counted by light-field technics (sic), is satisfactory to control exposures to most forms of asbes tos. Crocidolite, however, has been shown to produce, in addition to the asbestotic inflammation, also mesothelioma. Since no safe limit can be established for this form of asbestos at this time, until more definite data are obtained, it is recommended that workers exposed to crocidolite be equipped with air-supplied helmets. This idea of stringently controlling exposure to crocidolite asbestos dust was dropped before the publication of the 1968 book of TLVs. Over the next few years, the ACGIH published notices of intent to lower the TLV for all varieties of asbestos and change the method of analysis to phase contrast microscopy, but the formally adopted value remained 5 MPPCF through 1970. Finally, in 1974, ACGIH listed an adopted TLV of 5 f/cc for asbestos (using phase contrast microscopy), two yr after OSHA had established a standard at that level through formal rulemaking. In 1980, ACGIH lowered its TLV for chrysotile asbestos, the most abundant variety, to 2 f/cc, and set lower limits for crocidolite and amosite. By this time, government standards for chrysotile had been in effect at the 2 f/cc level for 4 yr in the United States and 11 yr in Britain. The TLV for crocidolite asbestos only (0.2 f/cc) is equal to the current (1986) OSHA asbestos standard for all types of asbestos. No notice of intended change for asbestos has been published by ACGIH since 1980. It is noteworthy that, despite the comparatively slow process governments must follow in developing standards under their laws and despite the reluctance of conser vative governments to regulate industry in the 1980s, ACGIH has lagged behind both OSHA and the British government in lowering limits for workplace exposure to the leading recognized cause of occupational cancer. Vinyl Chloride Upon the recommendation of Dr. Robert Scala at Esso, the TLV committee proposed lowering the limit for vinyl chloride gas to 50 ppm from 500 ppm [Mc Farland, 1965], This was largely based on animal tests published by Torkelson in 1961, where effects were noted at 100 ppm and a TLV of 50 ppm was recommended 548 Castleman and Ziem [Torkeison et al., 1961]. Following the circulation of the committee's 1966 Notice of Intent, Chairman Stokinger was invited to the Industrial Hygiene Foundation (IHF) in early 1966 to discuss the proposed changes in the TLV list. There, he met with 50 representatives of companies with membership in IHF in Pittsburgh. Stokinger was told that, "industrial experience suggests that (50 ppm) may be too low" [Report, 1966]. Consequently, the proposed change of vinyl chloride's TLV was "put off, on suggestion of Dr. Torkeison, that the Committee await further accumulating experi ence" [Stokinger, 1966]. The committee lowered the TLV to 200 ppm in 1971, based on unpublished Dow findings of liver dysfunction in workers exposed to 300 ppm (vinyl chloride combined with 5 ppm vinylidine chloride) [Documentation, 1971]. Dow representatives maintain that the company reduced its internal employee expo sure limit to 50 ppm in 1961; but in practice this limit was knowingly exceeded, as Dow first reported the above data in 1968 [Documentation, 1971; Paustenbach and Langner, 1986]. The first U.S. workplace standard for vinyl chloride was 500 ppm, the 1968 TLV. It was revealed in 1974 that vinyl chloride workers had died from angiosarcoma of the liver and that similar tumors had been produced in experimental animals at 50 and 250 ppm. OSHA issued a proposed standard for vinyl chloride, specifying that exposures be below detectability using instrumentation sensitive to 1 ppm. But official U.S. government statements that the safety of the gas had not been demonstrated at any level were publicly denounced by Stokinger as "irrational" and "unfortunate" in a letter to the National Cancer Institute. In an interview with the New York Times, Stokinger went on to say that there was "ample and increasing evidence that there are threshold levels for carcinogens below which there is little risk" [Official, 1974]. OSHA issued a 1 ppm standard for vinyl chloride later in 1974, and the U.S. industry not only met that goal but promptly resumed its growth [PVC, 1976]. In the meantime, the TLV committee had taken on members from industry, including Torkeison of Dow Chemical, a major manufacturer of vinyl chloride. Torkeison had primary responsibility for TLVs for vinyl chloride and a number of other high-volume, halogenated hydrocarbons, starting in 1971 [Minutes and Agenda, 1970-1976]. It was not until 1977 that the committee issued a new TLV for vinyl chloride, 5 ppm, which still stands. The TLV for vinyl chloride was thus set at one tenth the concentration carcino genic to animals for a proven human carcinogen. This conflicts with the current TLV committee claim that safety factors of 100 to 1,000 have "traditionally" been used to determine TLVs for carcinogens [Identification, 1986]. Dimethyl Sulfate The TLV for this vapor, used as a war gas in World War I, was originally set at 1 ppm in 1946. German reports in the late 1960s showed that DMS was carcinogenic in rats and probably also in workers; and the Germans lowered their MAK for this vapor to 0.01 ppm in 1971, as animal studies revealed serious lung damage at 0.5 ppm [Henschler, 1975]. The TLV committee had published its first listing of carcinogens as an appendix to the TLV booklet in 1971, consisting of only nine entries (mostly dye intermediates). In early 1972, the committee's annual Notice of Intended Changes informed readers that this list was being expanded, with separate groupings of human and "experimen tal" carcinoj five chemic; articles and observed to [Stokinger, : A few doctor at B/ that the Ge carcinogen ! DuPont sent conclusion tl potentially e to persons h; The D plants where ployees expc 1970, 97 wa the DMS are DMS worke lowed up. When 1973, Stokin gens" and o Stokinger rel tory cancers was complei Stokinger cit these compa also were no The TI 1976 was Jai Agenda. 197 the limit pre' Benzene The TI in 1947, 35 p a ceiling exp daily 10-min Institute. Cc (NIOSH, 19' British ceiling in fa\ 1966], The adoption unti Hueper 1942 [Huepe among nine s tQ(j6 Notice of ) tion (IHF) , tX met with 50 i. Stokinger was o low" [Report, vas "put off, on mulating experin in 1971, based sed to 300 ppm entation, 1971]. employee expo:ly exceeded, as Paustenbach and ) ppm, the 1968 m angiosarcoma :al animals at 50 . specifying that >pm. But official demonstrated at `unfortunate" in !ew York Times, ce that there are icial, 1974], 4, and the U.S. ? -7;76]. :-t industry, vinyl chloride, nd a number of tes and Agenda, TLV for vinyl Uration carcino:he current TLV y" been used to originally set at as carcinogenic r MAK for this damage at 0.5 ; as an appendix intermediates), iformed readers nd "experimen Corporate Influence on TLVs 549 tal" carcinogens. The listing of DMS in the former category prompted inquiries from five chemical companies. Stokinger replied to them, sending copies of underlined articles and saying: "a sufficient number of human cancers of the lung have been observed to make it highly probable that dimethyl sulfate is a carcinogen for man" [Stokinger, 1972]. A few months later, DuPont provided Stokinger with a copy of a letter from a doctor at BASF, a German manufacturer of dimethyl sulfate. The writer pointed out that the German MAK list denoted dimethyl sulfate as an experimental animal carcinogen but not a human carcinogen [Morgan, J.F., 1972], The next month, DuPont sent Stokinger an epidemiological report "which formed the basis of our conclusion that dimethyl sulfate is not known to have produced human cancers among potentially exposed persons. " Stokinger was asked to limit distribution of the study to persons having a need to see it [Morgan, J.F., 1972]. The DuPont study examined employee lung and lamyx cancer rates at three plants where DMS had been handled. However, "usable data" identifying the em ployees exposed to DMS before 1961 were available for only one plant. During 1932-- 1970, 97 wage roll workers and 46 salaried employees had worked at some time in the DMS area. There were two deaths each from lung and larynx cancer among the DMS workers between 1956-1970, with retirees and ex-employees clearly not fol lowed up. When OSHA issued an Emergency Temporary Standard for carcinogens in 1973, Stokinger argued for a distinction to be made between "known human carcino gens" and others on the OSHA list. Writing as Chairman of the TLV committee, Stokinger relied on the unpublished DuPont report to assert that no excess of respira-' tory cancers had occurred among DMS workers: "Manufacturing exposure control was completely effective, without the requirement for air-pressurized suits. . ." Stokinger cited other unpublished reports from DuPont and Dow to argue that two of these companies' products covered by the OSHA standard (MOCA, ethylenimine) also were not human carcinogens [Stokinger, 1973]. The TLV committed member with responsibility for DMS in.the period 1972-- 1976 was James Morgan of DuPont (sole U.S. manufacturer of DMS) [Minutes and Agenda, 1970-1976]. The committee assigned a TLV of 0.1 ppm in 1977, ten times the limit previously accepted in Germany. Benzene The TLV for benzene was adjusted downward from 100 ppm in 1946, 50 ppm in 1947, 35 ppm in 1948, to 25 ppm in 1957. The TLV committee adopted 25 ppm as a ceiling exposure limit in 1963. An industry consensus "standard" of 10 ppm (with daily 10-min peaks of 50 ppm) was issued in 1969 by the American National Standards Institute. Consequently, 10 ppm was the first benzene limit adopted by OSHA (NIOSH, 1974). British industry and government writers urged Stokinger to abandon the 25 ppm ceiling in favor of a 10 ppm average value as early as 1966 [King, 1970; Stokinger, 1966]. The TLV committee first proposed this change in 1968, but deferred its adoption until 1977. Hueper had assessed benzene as almost certainly a proven cause of leukemia in 1942 [Hueper, 1942]. The German MAK commission had listed benzene in 1971 among nine human carcinogens, "for which zero concentration values are given 550 Castleman and Ziem because the objectionable concentration is not yet known" [Morgan, L., 1972]. Benzene was classified by the TLV committee as a "suspected" human carcinogen in 1975. The 1986 Documentation contains no references less old than 1977 and relies on one report whose findings were reversed in 1977. That year, OSHA issued an emergency temporary standard and proposed a permanent standard of 1 ppm for benzene. An adverse Supreme Court ruling in 1980 based on the record of the benzene standard issued in 1978 prompted OSHA to conduct quantitative cancer risk assessment and again propose a 1 ppm limit in 1985. The 1986 Documentation refers to unpublished work by Ott in 1975 as showing no excess mortality among benzene-exposed workers. However, Ott concluded that this same cohort of Dow Chemical employees demonstrated a significant excess of myelogenous leukemia cases--prompting Dow to announce a new corporate ceiling limit of 10 ppm in 1977 [Benzene, 1977; Ott et al., 1978], Dow epidemiologists have now seen 4 deaths from myelogenous leukemia in this work force, versus 0.9 expected; a fifth worker with leukemia was listed as dying with pneumonia [Bond et al., 1986]. Infante at OSHA notes that average benzene exposure of these workers was 5.5 ppm [Infante, 1987]. Similarly, the 1986 Documentation makes no mention of dose-related chromo somal abnormalities among Dow workers exposed to benzene concentrations below 10 ppm [Infante and White, 1983]. These findings were withheld by Dow during the OSHA benzene hearings in 1977, prompting the researcher involved to quit in 1978 in order to release his results. Because of the company's delay in releasing these findings, the researcher denounced Dow as "unethical" and "immoral" [Picciano, 1979; Scott, 1978]. The TLV committee, which adopted a companion short-term exposure limit of 25 ppm to go with the 10 ppm average for benzene in 1980, is discarding the short term limit in 1987. Exposure at even 10 ppm for eight min is illegal under the OSHA benzene standard published September 11, 1987. The standard requires that exposures average no more than 1 ppm, with 15-min peaks no more than 5 ppm. The committee's position in 1987 thus resembles that of the American Petroleum Institute in its 1978 court challenge to the overturned benzene standard. The past decade of benzene toxicology research has not been incorporated into the TLV Documentation. The research and policy at Dow Chemical (whose senior toxicologist was an active member of the TLV committee), if known to the committee, have been disregarded without mention. Acrylonitrile Following the reports of positive animal studies by inhalation and ingestion, as well as positive epidemiological findings, OSHA regulated acrylonitrile as a carcino gen in 1978. Acrylonitrile was also classed by ACGIH as a human carcinogen in 1978. Following the publication of an inconclusive epidemiological study in Britain and the receipt of epidemiological "communications to the TLV committee" from Monsanto Company in 1981, acrylonitrile was reclassified under "industrial sub stances suspect of carcinogenic potential for man. " The Monsanto conclusions were quoted by the TLV committee; no published study is yet available for scrutiny by the scientific community. Ethylenimii When Dow's Dr. 1 despite this experience 1 report of a ti "Toda; Anilin site of ologic: was 4C Dr. Kilian l experience ' study was si let's append recommends Dow Chemi' Carcinoger The Tl classified as ACGIH has substances. ; 1986], The Society max The A carcinogens; entries as su are 42 comp "justifiably mum, 1984] The T importance, pounds were mental Anir. they set ma> attributed fo Lead Becaus in which the could hardly Inorganic L From 0.15 mg/nr orr'*n, L., 1972]. ' ^^arcinogen in - +' an 1977 and relies OSHA issued an ard of 1 ppm for the record of the titative cancer risk n 1975 as showing Ott concluded that 'nificant excess of ' corporate ceiling idemiologists have force, versus 0.9 leumonia [Bond et of these workers .e-related chromoicentrations below >y Dow during the ed to quit in 1978 in releasing these moral" [Picciano, exposure limit of < \'g the shortI >er the OSHA res that exposures m. nerican Petroleum andard. The past ed into the TLV senior toxicologist mittee, have been and ingestion, as itrile as a carcinolan carcinogen in ll study in Britain committee" from r "industrial sub conclusions were or scrutiny by the Corporate Influence on TLVs 551 Ethylenimine When OSHA proposed to regulate this compound as a carcinogen in 1973, Dow's Dr. D.J. Kilian provided the basis for the TLV committee observation, that despite this chemical's toxic and carcinogenic effects in animal studies, "industrial experience has been good." The entire basis for this was the following second-hand report of a telephone conversation between two major manufacturers: [Kilian, 1973] "Today, I talked by telephone to Dr. Theiss, medical Director of Badische Anilin and Soda-Fabrik in Germany (the only other major manufacturing site of ethylenimine) and he stated that they had just finished an epidemi ological study of 144 of their El workmen. The exposure time on some was 40 years and they found no evidence that El was a human carcinogen." Dr. Kilian also wrote that he and Dr. Theiss planned to combine their companies' experience "in a medical publication in the near future." It does not appear that any study was subsequently published. Ethylenimine was removed from the TLV book let's appendix list of "experimental carcinogens" after 1974, presumably upon the recommendation of the subcommittee on carcinogens, which included Torkelson of Dow Chemical (sole U.S. producer of the material). Carcinogens in General The TLV committee has now stated its intent to "formally" evaluate chemicals classified as carcinogens by other organizations but not ACGEH [Spirtas et al., 1986]. ACGIH has published a table listing the carcinogenic status of more than 300 substances, according to five national and international organizations [Identification, 1986]. The most appropriate comparison is with the list of the German Research Society maximum workplace concentrations (MAK) Commission. The ACGEH classifies 11 materials in the aforementioned table as human carcinogens; the MAK Commission's total is 17. The ACGIH classifies 40 other entries as suspected human carcinogens; the corresponding MAK commission totals are 42 compounds proven carcinogenic in animal experimentation only, and 61 more "justifiably suspected of having carcinogenic potential" [Identification, 1986; Maxi mum, 1984]. The TLV committee avoided listing animal carcinogens of major industrial importance, including trichloroethylene and dioxane. These and other unnamed com pounds were exempted by the "Committee Guidelines for Classification of Experi mental Animal Carcinogens" published in 1976. The guidelines are unique in that they set maximum carcinogenic dosages, above which no "practical importance" is attributed for positive animal experiments. Lead Because of their enormous significance in occupational health and the manner in which their TLVs emerged, the story of inorganic and organic lead compounds could hardly be overlooked in this review. Inorganic Lead From 1946 through 1956, the TLV for lead and its inorganic compounds was 0.15 mg/m3. This followed earlier recommendations of the U.S. Public Health 552 Castleman and Ziem Service and an American Public Health Association committee on lead. Later editions of the Documentation observed that this limit proved "difficult to achieve in many industries." Explaining the 1957 decision to raise the lead TLV to 0.20 mg/m3, the first edition of the Documentation said: "Long industrial experience with the 0.15 mg/m3 limit, however, showed that. . . lead-absorption, as measured by urinalysis, were (sic) not indicative of harmful exposure." No reference for this was given. The 1966 Documentation went on to describe the blood lead concentration of 80 micrograms per 100 ml as "normal", and noted that repeated exposures above 0.20 mg/m3 could cause higher blood lead burdens "indicative of incipient lead poisoning." Pressure for lowering the TLV developed in November, 1968, when an inter national commission on occupational health recommended 0.15 mg/m3. In preparation for discussions with industry, the TLV committee summarized recent developments on lead toxicity and reviewed the "Basis of Present TLV." Under this last heading were three items,'all unpublished corporate communications, from: Bowditch (Lead Industries Association); Dooley (Texaco); and Nelson (Asarco) [Review, 1970]. Neither Stokinger nor Elkins can now recall what information was provided by these individuals over 30 years ago, and no primary written documentation has been found in Stokinger's old files at NIOSH. On May 1, 1970, a meeting was held by TLV committee members Stokinger and Frederick with representatives of the automotive and lead industries, state health officials, and others. Industrial representatives said they used blood lead analyses for health control measures, and urged that air sampling be advised only as an engineering guide. General Motors hygienist Vincent Castrop acknowledged that his company used 0.15 mg/m3 as its guideline [Stokinger, 1970]. The TLV committee then readopted the former value of 0.15 mg/m3, which has remained unchanged since 1973. A short-term (15-min) exposure limit of 0.45 mg/ m3 was also adopted in 1976, later to be discarded in 1986. The current Documenta tion includes an attack on NIOSH for recommending a standard of 0.10 mg/m3 and rejects the OSHA standard of 0.05 mg/m3 promulgated in 1978. Organic Lead Compounds When tetraethyl lead was introduced as a gasoline additive in the 1920s, lead poisoning was a major by-product of the industry. About 80% of the workers at DuPont's New Jersey production facility were believed to have been lead poisoned; and the plant was known to workers as "the House of the Butterflies" because of the hallucinations afflicting employees there. DuPont was accused of suppressing infor mation from the press even in cases where workmen were hospitalized and died from lead poisoning [Rosner and Markowitz, 1985]. Tetraethyl lead (TEL) and tetramethyl lead (TML) were given TLVs of 0.075 mg/m3 in 1963 and 1967, respectively. The main basis for the tetraethyl lead TLV consisted of statements by industry representatives that this limit was observed by Ethyl Corporation without apparent ill effects on the workers [Documentation, 1966]. Publication of the second of these TLVs brought forth a "Confidential" letter of protest in 1967 from Dr. Robert Kehoe, the lead industry's foremost medical expert, its consultant and a defender of the tetraethyl lead industry since the 1920s [Rosner and Markowitz, 1985], Kehoe urged that both TLVs be discarded, "with the least possible fanfare. " His "Dear Herb" letter concluded: [Kehoe, 1967] "I would i conditions ever, and y little doubt authoritativ Kehoe invited St< in Cincinnati catand recalls that aware that Keho industry expert o interests in this n The most dropping the limi sible industries t< Intent", the com alkyls was being At least ons each of these cht organic lead air c mg/m3 for TML significantly abo' this encouraging mentation) was ! chairman is recor The TLV ct 2-3, 1968, and d for TML [Stokin; the same to this found; and giver appears to have b It has been : the exposure limi for many substam result for the lea< OSHA standard finorganic lead cc OSHA adopts the Bias of TLV Cor Dr. Hector member, protesi bias. . . particula members who co pro-industry bias for having made and anti-labor" [E } \ Later editions yiieve in many 20 mg/m3, the first rith the 0.15 mg/m3 by urinalysis, were -as given. The 1966 i of 80 micrograms ; 0.20 mg/m3 could oning." 168, when an inter;/m3. In preparation ecent developments er this last heading m: Bowditch (Lead >) [Review, 1970]. s provided by these tion has been found members Stokinger ustries, state health >d lead analyses for y as an engineering l that his company mg/m3, which has : " of 0.45 mg/ t Documentaot u. LO mg/m3 and in the 1920s, lead of the workers at een lead poisoned; ies" because of the suppressing inforized and died from en TLVs of 0.075 itraethyl lead TLV t was observed by umentation, 1966]. nfidential" letter of ost medical expert, the 1920s [Rosner ;d, "with the least '7] Corporate Influence on TLVs 553 "I would not take the risk of subjecting a group of men to working conditions represented by this atmospheric standard for any reason what ever, and yet this level is being adopted on a worldwide basis, and I have little doubt that it will be applied literally by someone, sometime, as being authoritative. It is not so applied in any part of the industry at present." Kehoe invited Stokinger to be his lunch guest at the Queen City Club, a private club in Cincinnati catering primarily to businessmen [Kehoe, 1967], Stokinger accepted, and recalls that Kehoe "pontificated" without supplying any data. Stokinger was aware that Kehoe had become a wealthy man over decades as the principal U.S. industry expert on lead poisoning. Though Kehoe presumably represented industrial interests in this matter, no firms were specifically named [Stokinger, 1986-87]. The most influential members of the TLV committee rejected the idea of dropping the limits for TEL and TML, and instead cautiously challenged the respon sible industries to produce some dose-response data. In its January, 1968 "Notice of Intent", the committee wrote that a downward revision of the TLVs for both lead alkyls was being considered. No new proposed limits were given [Committee, 1968]. At least one manufacturer of these compounds found that operations involving each of these chemicals exceeded even the then-current TLV of 0.075 mg/m3. But organic lead air concentrations averaging as high as 0.121 mg/m3 for TEL and 0.179 mg/m3 for TML reportedly corresponded to average urinary lead concentrations "not significantly above a high normal" -- meaning, less than 0.15 mg/1. The source of this encouraging news was A.L. Linch, whose employer (never noted in the Docu mentation) was DuPont. The date of this communication to the TLV committee chairman is recorded as April l, 1968. The TLV committee held its semiannual meeting over the next two days, April 2-3, 1968, and decided to raise the TLVs to 0.10 mg/m3 for TEL and 0.15 mg/m3 for TML [Stokinger, 1968]. These limits were formally adopted in 1970, and remain the same to this day. No written communication from Linch to Stokinger has been found; and given the rapid sequence of events here, the cited report from Linch appears to have been a telephone call [Stokinger, 1986-87]. It has been proposed recendy that OSHA try to adopt current TLVs to "update" the exposure limits for hundreds of substances. While this would yield stricter limits for many substances whose OSHA limits are still the 1968 TLVs, the opposite would result for the lead alkyls. This is especially worrisome in view of the fact that the OSHA standard for organic lead compounds is now more permissive than that for the inorganic lead compounds, which are less toxic; this anomaly will be worsened if OSHA adopts the current TLVs for the lead alkyls. Bias of TLV Committee Membership Dr. Hector Blejer, resigning from the committee in 1980 after 10 years as a member, protested what he called "an increasingly stronger pro-industry bias. . . particularly among almost all the Committee consultants and among the members who consult privately for private industry." Blejer went on to blame this pro-industry bias and repeated "unnecessary" disagreements with NIOSH and OSHA for having made the TLV committee and ACGIH appear "and-NIOSH, anti-OSHA, and anti-labor" [Blejer, 1980]. 554 Castleman and Ziem To its discredit, the committee has long turned a blind eye to conflicts of interest, both overt and subtle. Health and safety professionals tend to view policy issues from a spectrum of opinions: from those who would resolve the benefit of doubt in assuring the fullest worker protection to those who are more sensitive to corporate financial priorities where health and safety is in practice regarded as an expenditure to be controlled as much as possible. It is no accident that professionals with the latter point of view are more likely to consult for or be employed by corporations, and those closer to the former viewpoint are more likely to be indepen dent of corporate funding, perhaps working in government or for labor unions, public interest groups, etc. The TLV committee never acknowledged this reality or attempted to achieve a balance between corporate- and union-affiliated health professionals. Only occasional token efforts were made to get a union industrial hygienist on the TLV committee. There, the union person could expect to be marginalized at least as badly as was Dr. Blejer (a NIOSH expert on lead, arsenic, cadmium, and asbestos), by the sheer force of numbers and adversaries with vastly superior technical resources. The TLV committee never offered unions and other strong advocates of worker protection a chance to participate on a fully equal basis. The occasional token offers for partici pation in effect only gave unions the "choice" of participating in an unequally balanced arena and depleting their resources with little chance of being heard -- or of no participation at all. CONCLUSIONS AND RECOMMENDATIONS.. While earlier reviews of the TLVs themselves have been critical, the process of TLV development has not been critically examined in the past. The unavailiabity of unpublished corporate "documentation" precludes scientific scrutiny of the primary basis for nearly one sixth of the "documented" TLVs. At the same time, the TLV committee's uncritical acceptance of industry assertions based on scant, unpublished "data" raises yet greater concern. The documentation of TLVs for their own companies' products by industry members of the TLV committee constitutes a major conflict of interest. This happened on a large scale in the 1970s, with the Dow Chemical representative primarily responsible for TLV development for major Dow products (vinyl chloride, vinylidine chloride, chloroform, methyl chloride, ethylene dichloride, ethylene dibromide, tri chlorobenzene, dioxane, ethanolamine, dipropylene oxide methyl ether, styrene, ethylene glycol.dibromochloropropane, "Tordon", "Ruelene", "Dursban", and "Plictran"); and the DuPont representative doing the same for major DuPont products (dimethyl sulfate, "'MOCA", lead chromate, formamide, dichloromonofluoromethane, "Lannate", "Karmex", and "HyvarX") [Chemical Week, 1975; Minutes and Agenda 1970-1976). The listing of dominant corporate TLV committee members as "consultants" and the issuance of statements to the effect that they did not officially vote on the TLVs were deceptive [Lee, 1987). The concealment of industry influence on the TLVs is a serious matter, quite apart from the exercise of that influence itself. Aside from the participation of industry-employed health professionals, the TLV committee has extended full membership to full-time industry consultants as early as 1951 (Dr. Arthur Vorwald of the Saranac Laboratory). To this day, TLV committee mei industrial const booklet. ACGI1 cases or requii parties. Simila' committee men and research gr The TLV As part of the budget of $30.( meetings [Kelly technical resou: puter searches, and/or their em| calls). Over the TLVs on their c As a result, do minimal review The TLVs cally based lim over a lifetime < to be first worlc such misplaced the ACGIH lim as an obstacle tc Industrial TLVs. Hygienis TLV can be reg more adequately need increased t field inspections "cranking out assuming all is v OSHA is r from the 1968 1 improvement, it adequate. Since chemicals for w should adopt NI> since many chei OSHA should ci not a substitute l With the n workplace expos and elsewhere, i gradually replaci respected organ world, with sufl ?, "'conflicts of iew policy ve^the benefit of note sensitive to j regarded as an hat professionals be employed by [y to be indepenor unions, public pted to achieve a Only occasional TLV committee, badly as was Dr. y the sheer force irees. The TLV -ker protection a iffers for particiin an unequally ag heard -- or of il, the process of unavailiabity of v he primary ' jb, the TLV ant, unpublished ucts by industry t. This happened native primarily oride, vinylidine e dibromide, tri ether, styrene, `Dursban", and DuPont products omonofluorome>75; Minutes and as "consultants" ially vote on the influence on the nee itself, rofessionals, the 7 consultants as o this day, TLV Corporate Influence on TLVs 555 committee members can and do earn a substantial fraction of their incomes as industrial consultants, while publishing only their university affiliations in the TLV booklet. ACGIH has no policy either restricting TLV committee membership in such cases or requiring public disclosure of consulting work for financially interested parties. Similarly, there is no policy restricting the chemicals assigned to TLV committee members because of conflicts of interest through employment, consulting, and research grants [Kelly, 1986-87]. The TLV committee's lack of adequate resources is evident from its finances. As part of the ACGIH, a volunteer organization, the committee now has an annual budget of $30,000, most of which goes for travel and lodging expenses to conduct meetings [Kelly, 1986-87]. The members of the committee must rely on whatever technical resources and support services are available to them as individuals (com puter searches, libraries, research assistants, clerical assistants), and borne by them and/or their employers for their unpaid committee work (e.g., long-distance telephone calls). Over the years, this has meant that committee members have had to work on TLVs on their own time and their own expense, with their own resources, unassisted. As a result, documentation on many chemicals seems to have been prepared with minimal review of the literature. The TLVs have nonetheless been widely represented and accepted as scientifi cally based limits that would protect virtually all workers from health impairment over a lifetime of exposure on the job [Lee, 1987]. The TLVs are assumed by many to be first world, "first class" guidelines for worker protection. The consequences of such misplaced confidence in the TLVs are profound and global. The credibility of, the ACGIH limits as scientifically, independently, and verifiably determined persists as an obstacle to a better standard of worker protection. Industrial hygienists need clear instruction regarding the limited nature of the TLVs. Hygienists too often assume or convey to workers that exposure below the TLV can be regarded as safe. They need training which would enable them to assess more adequately the scientific grounds upon which the TLVs are based. They also need increased training in eliciting and evaluating worker complaints of illness during field inspections. This approach should replace the technician approach of simply "cranking out numbers" with monitoring, comparing them to a table, and then assuming all is well if exposures measured are less than the TLVs. OSHA is now considering adopting current TLVs to replace its exposure limits from the 1968 TLV list (Z table). While for some chemicals this may represent an improvement, it is clear that we cannot assume that the current TLVs are scientific or adequate. Since more rigorous and thorough documentation has been done for the chemicals for which NIOSH recommends specific maximum exposure levels, OSHA should adopt NIOSH levels where these are stricter than those of the ACGIH. Finally, since many chemicals have not been assessed by NIOSH and others need updating, OSHA should consider the adoption of TLVs or NIOSH values as a stopgap measure, not a substitute for ongoing rigorous assessment of chemical exposure values. With the more recent emergence of better trained and equipped groups issuing workplace exposure limits and supporting documentation in North America, Europe, and elsewhere, it now seems appropriate for an international effort to be mounted to gradually replace the TLVs. This can be done under the auspices of an internationally respected organization, with the participation of leading experts from around the world, with sufficient financing. Corporations with their own internal lists of occu- 556 Castieman and Ziem pational exposure limits can contribute to this process by publishing these lists and supporting data without further delay. Openness of the process is essential, as is the exclusion of financially interested parties from having leverage in the deliberations. Policies regarding disclosure of income and conflicts of interest must be accepted by the participants so that the highest level of credibility maintained. Policies regarding making any use of and maintaining public repositories for unpublished documentation will also be needed. Public access to minutes of meetings should be assured and provided for. Yet even a panel of the best technical experts would not overcome all obstacles inherent to the process of setting worker exposure limits. There are implicit assumptions in any process of establishing some "acceptable" level of chemical insult to which humans may be exposed. Many scientists reject "safe" exposure levels for carcinogens and certain reproductive hazards. The concept of "safe" exposure limits for other chemicals is less often questioned however, even though scientists are unable in practice to determine "safe" exposures. They can only determine levels below which their limited measurement tools are unable to detect effects in a finite and often very limited number of workers. Thus the very concept of "safe" exposures to any chemical is inherently unscientific. Indeed, the term "threshold limit" embodies this unproven and probably improvable concept that there is some known level of exposure which does not adversely affect the organism. Discarding the term "threshold limit" is a necessary first step in correcting this false ideology of the past. Rather, the numerical values for exposure limits selected as "acceptable" by one social group (scientists) for another social group (workers) is very much a political as well as a scientific process. The Norwegian Administrative Norms, for example, explicidy acknowledge that the chemical exposure limits reflect economic as well as medical and technical considerations. The Norwegian authorities consider that while writing the documentation for chemicals is ideally a scientific process, the setting of numerical limits is a political process. It is time that we all openly acknowledge the political nature of decisions by unexposed scientists and regulators regarding maximum levels of chemicals to which other humans can knowingly be exposed. The decision process therefore must not only be freed from undue corporate influence; it must also include substantial participation by representatives of exposed persons. REFERENCES Ann NY Acad Sci 132:1-765 (1965). "Annual Report of the Chief Inspector of Factories for the Year 1947" (1949) London, HM Stationery Ofc., pp. 79-81. Baader EW (1939): Asbestosis. Deut Med Woch 69:407-408. Benzene Ceiling Level Lowered to 10 PPM (July 14, 1977). Dow Today No. 41 (Dow Chemical, Pub. Relations Dept.). . Biejer HP (June 5, 1980). Letter to Col. VL Carter (Chairman, TLV committee). Bond GG, McLaren EA, Baldwin CL, Cook RR (1986): An update of mortality among chemical workers exposed to benzene. Br J Ind Med 43:685-691. Castieman BI (1986): "Asbestos: Medical and Legal Aspects" 2nd Ed. Clifton, N.J.: Prentice Hall/Law & Business. "Chemical Week 1976 Buyers' Guide Issue" (Oct. 29, 1975) New York: McGraw-Hill. Committee on Th 1968 Notic< "Criteria for a Rt 69. de Treville R (Sep "Documentation o "Documentation o "Documentation c 278. Dreessen WC, Di Study of As Elkins HB (May 2 Fulton WB et al. ( Golz HH, Culver investigatio Henderson R (197: Ann Occup Henschler D (July- Toxic Worl Medicine D Henschler D (1984 ' 92. Henschler D (1985 Hill RH, Faning El of arsenic. 1 Hueper WC (1942. pp. 594-591 Identification and Limit Value to A-ll(86). Infante PF (1987): Infante PF, White > health effect Inorganic Arsenic: Fed Registei Kehoe R (Oct. 11. files. Kelly WD (1986-F communicac Kilian DJ (Oct. 17. King E (Aug. 11. 1 files. Knox JF (1960): R< Turner and .' Lee JS (1987): Ale MacFarland HN (O "Maximum Concei (1984): Deut Milham S, Strong 7:176-182. Minutes of Meetim NIOSH files Minutes of Meeting pp. Minutes of Meetin Environment Minutes and Agend. 8-9 and Not 16-17 and N> NIOSH files ^ese lists and lancially interested ding disclosure of :ipants so that the ig any use of and ill also be needed, ed for. rcome all obstacles some "acceptable" ny scientists reject zards. The concept ned however, even res. They can only -e .unable tQ detect s the very concept Indeed, the term : concept that there feet the organism, orrecting this false .s "acceptable" by .) is very much a trative Norms, for s reflect economic ' ties consider 4 c. ,,c process, the aat we all openly ists and regulators can knowingly be m undue corporate itatives of exposed ondon, HM Stationery (Dow Chemical, Pub. nong chemical workers <'.J.: Prentice Haii/Law w-Hill. Corporate influence on TLVs 557 Committee on Threshold Limits of the American Conference of Governmental Industrial Hygienists 1968 Notice of Intent, NIOSH files, 3 pp. "Criteria for a Recommended Standard Occupational Exposure to Benzene" (1974): NIOSH, pp. 68 69. de Treville R (Sept. 10, 1969): To IHF Fibrous Dust Study Sponsors. (Memorandum). "Documentation of Threshold Limit Values" (1962): A.C.G.I.H. Cincinnati, pp. 10-11. "Documentation of Threshold Limit values (2d Ed.) (1966): A.C.G.I.H. Cincinnati, pp. 184-185. "Documentation of the Threshold Limit Values" (3rd Ed.) (1971): A.C.G.I.H. Cincinnati, pp. 277 278. Dreessen WC, Dallavalle JM, Edwards TI, Miller JW, Sayer RR, Easom HF, Trice MF (1938): "A Study of Asbestosis in the Asbestos Textile Industry." Pub. Health Bull. No 241. Washington. Elkins HB (May 24, 1975): Letter to H. Stokinger, NIOSH files. Fulton WB et al. (1935): "Asbestosis." Pennsylvania Dept, of Labor and Ind., Harrisburg. Golz HH, Culver BD, Hardy HL, Miller LH, Raleigh RL, Roush G, Tusing TW (1966): Report of an investigation of threshold limit values and their usage. J Occup Med 8:280-283. Henderson R (1975): U.S. experience with occupational safety and health legislation--an industry view. Ann Occup Hyg 18:335-338. Henschler D (July 31, 1975): Senate Commission of the German Research Society on the Testing of Toxic Workplace materials. Letter to H. Stokinger and enclosed Toxicological Occupational Medicine Documentation on dimethyl sulfate. NIOSH files. Henschler D (1984): Exposure limits: history, philosophy, future developments. Ann Occup Hyg 28:79 92. Henschler D (1985): Personal communication to G. Ziem. Hill RH, Faning EL (1948): Studies in the incidence of cancer in a factory handling inorganic compounds of arsenic. Br J Ind Med 5:1-6. Hueper WC (1942): "Occupational Tumors and Allied Diseases" Springfield, Illinois: Charles Thomas, pp. 594-599. Identification and Classification of Carcinogens (1986): Appendix to "Documentation of Threshold Limit Values and Biological Exposure Indexes" (5th Ed.) Cincinnati: A.C.G.I.H., pp. A-3(86) to A-ll(86). Infante PF (1987): Personal communication to B. Castleman. Infante PF, White MC (1983): Benzene: epidemiologic observations of leukemia by cell type and adverse health effects associated with low-level exposure. Environ Health Perspectives 52:75-82. Inorganic Arsenic: Proposed Exposure Standard. (Jan. 21, 1975) Occupational Safety and Health Admin. Fed Register 40:3392-3404. Kehoe R (Oct. 11, 1967): (Kettering Laboratory, Univ. of Cincinnati). Letter to H. Stokinger. NIOSH files. Kelly WD (1986-87): (Executive Secretary, A.C.G.I.H.). Letter of Jan. 21, 1987 and other personal communications to B. Castleman. Kilian DJ (Oct. 17, 1973): (Dow Chemical Co.) Letter to H. Stokinger. NIOSH files. King E (Aug. 11, 1970): (National Occup. Hyg. Services, Ltd., U.K.). Letter to H. Stokinger. NIOSH files. Knox JF (1960): Report of visit by Dr. J.F. Knox to Keasbey & Mattison Company August 4-5, 1960. Turner and Newall document, 2 pp. Lee JS (1987): A letter from the chair of ACGIH. AppI Ind Hyg 2:F6-F7. MacFarland HN (October 25, 1965): (Hazelton Laboratories, Inc.) Letter to H. Stokinger. NIOSH files. "Maximum Concentrations at the Workplace and Biological Tolerance Values for Working Materials" (1984): Deutsche Forschungsgemeinschaft, Verlag Chemie, Weinheim. Milham S, Strong T (1974): Human arsenic exposure in relation to a copper smelter. Environ. Res. 7:176-182. Minutes of Meeting of the Threshold Limits Committee, ACGIH. (March 11-12, 1965): Washington. NIOSH files. Minutes of Meeting of the Threshold Limits Committee, ACGIH. (March 30-31, 1967): NIOSH files, 7 pp. Minutes of Meeting of the Threshold Limits Committee for Chemical Substances in the Workroom Environment, ACGIH. (Nov. 16-17, 1972): NIOSH files. Minutes and Agenda notices of the TLV committee meetings, April 1-2, and Nov. 19-20, 1970: April 8-9 and Nov. 18-19, 1971; April 13 and Nov. 16-17, 1972; May 3^t and Nov. 1-2, 1973; April 16-17 and Nov. 20-21, 1974; May 6-7 and Nov. 25-26, 1975; April 28-29 and Dec. 6-1. 1976. NIOSH files. . v 558 Castleman and Ziem Morgan JF (1972): (DuPont). Letter to H. Stokinger (June 6, 1972), with attached copy of letter from D.H. Zeller (BASF, Ludwjgshaven) to Dr. J. Zapp (DuPont), April 13, 1972. Morgan JF letter to H. Stokinger (June 6, 1972), with report of Pell, S., "An Epidemiological Study of Dimethyl Sulfate and Cancer of the Respiratory System", 4 pp. NIOSH files. Morgan L (1972): (Chief Med. Officer, International Nickel, Ltd.) Letter to E. Mastromatteo (Ontario Dept, of Health), April 17, 1972. NIOSH files. Date confirmed by Dr. Henschler (Jan. 20, 1987). Notice of Intent 1965 (Jan. 7, 1965): NIOSH files. Noweir MH (1986): Occupational health in developing countries with special reference to Egypt. Am J Ind Med 9:125-141. Occupational Exposure to Inorganic Arsenic: Final Standard (1978): Occup. Safety and Health Admin. Fed Register 43:19584-19631. Official Assails National Cancer Institute (Apr. 6, 1974): New York Times. . Ott MG, Townsend JC, Fishbeck WA, Langner RA (1978): Mortality among individuals occupationally exposed to benzene. Arch Env Health 33:3-10. Paustenbach D, Langner R (1986): Corporate occupational exposure limits: the current state of affairs. Am Ind Hyg Assoc J 49:809-818. Perry K, Bowler RG, Buckell HM, Druett HA, Shilling RSF (1948): Studies in the incidence of cancer in a factory handling inorganic compounds of arsenic. II. Clinical and Environmental Investiga tions. Br J Ind Med 5:6-15. Picciano D (1979): Cytogenic study of workers exposed to benzene. Environ Res 19:33-38. Pinto SS: (March 23, 1961): Letter to H.E. Stokinger. NIOSH files. Pinto SS, Bennett BM (1963): Effect of arsenic trioxide exposure on mortality. Arch Environ Health 7:583-591. Pinto SS, Nelson KW (1976): Arsenic toxicology and industrial exposure. Ann Rev Pharm Tox 16:95 100. PVC Rolls Out of Jeopardy, into Jubilation (Sept. 15, 1976): Chem Week. Report of the Committee on Threshold Limits (Mar. 27-30, 1948): "Trans, of the Tenth Annual Meeting of the American Conference on Governmental Industrial.Hygienists" Boston, pp. 29-31. Report of a Discussional on Threshold Limit Values held at Mellon Institute on February 1-2, 1966. Confidential draft, March, 1966. NIOSH files, 2 pp. Review of TLV for Inorganic Lead/May 11, 1970 AIHA Meeting. NIOSH files. Rosner D, Markowitz G (1985): A `Gift of God'? The public health controversy over leaded gasoline during the 19.20's. Am J Pub Health 75:344-352. Samuels SW (1981): The international context of carcinogen regulation: benzidine. "Banbury Report 9: Quantification of Occupational Cancer" New York: Cold Spring Harbor Laboratory, pp. 497 512. Scott R (June 11. 1978): Danger of low-level benzene reported. Washington Post. Spirtas R, Steinberg M. Wands RC, Weisburger EK (1986): Identification and classification of carcino gens. Amer J Pub Health 76:1232-1235. Stokinger HE (March 2, 1961): Letter to SS Pinto (Asarco). "Not to be sent to files", NIOSH files. Stokinger HE (April 1, 1964): Letter to FG Stephenson (Manufacturing Chemists' Association). NIOSH files. Stokinger HE (May 3, 1966): Letter to RJ Sherwood (Esso, U.K.). NIOSH files. Stokinger HE (March 8, 1966): Letter to WA Knapp (Allied Chemical Corp.). NIOSH files. Stokinger HE (1969): Current problems of setting occupational exposure standards. Arch Environ Health 19:277-281. Stokinger HE (Feb. 16. 1972): Letter to Emii E. Christofano (Hercules, Inc.), cc: Standard Oil, ICI America, and Mallinckrodt Chemical representatives. NIOSH files. Stokinger HE (June 11, 1973): Letter to John Stender (OSHA Administrator. U.S. Dept, of Labor). NIOSH files. Stokinger HE (1986-87): Personal communication to B. Castleman. Note: Frequent reference is made to the "Documentation of Threshold Limit Values and Biological Exposure Indices" (5th Ed.) American Conference of Governmental Industrial Hygienists, Cincinnati, 1986. Also cited are the various annual TLV booklets published by the ACGIH. No citations are given for such references either in the text or bibliography. Many of the annual lists of TLV's and historic articles on TLV's have been republished in "Threshold Limit Values--Discussion and Thirty-Five Year Index with Recommendations" Ann. A.C.G.I.H., Vol. 9, Cincinnati, (1984). Stokinger HE (Ms files. Stokinger HE (Ap Torkelson TR, D\ exposure of Toyama T (1985): Vigliani E et al. ( Permissible Milan. Vorwaid AJ. Durl Med 3:1-4? Vorwaid AJ (May Mining As: Institute of Vorwaid AJ (1948 Dec. 6. 194 Corp. by th< Wade N (1976): ,Y ' of letter from * \gan JF letter i of Dimethyl istromatteo (Ontario tier (Jan. 20, 1987). :nce to Egypt. Am J and Health Admin. duals occupationally rent state of affairs. incidence of cancer onmental Investiga- ':33.-38. _ ,rch Environ Health - Pharm Tox 16:95- nth Annual Meeting ,pp. 29-31. :ebruary 1-2, 1966. >ver leaded gasoline iry Report 9: ,, ,bry, pp. 497- .ification of carcino- ", NIOSH files, association). NIOSH SH files. Arch Environ Health :: Standard Oil, ICI S. Dept, of Labor). dues and Biological gienists, Cincinnati, o citations are given ' TLV's and historic ind Thirty-Five Year Corporate Influence on TLVs 559 Stokinger HE (May 19, 1970): Semi-formal meeting for review of the TLV for inorganic lead. NIOSH files. Stokinger HE (April 1968): TLV committee changes in the list for 1968. NIOSH files. . Torkelson TR, Dyen F, Rowe VK (1961): The toxicity of vinyl chloride as determined by repeated exposure of laboratory animals. Amer Ind Hyg Assoc J 22:354-361. Toyama T (1985): Permissible and control limits at places of work in Japan. Amer J Ind Med 8:87-89. Vigliani E et al. (1977): "Methods Used in Western European Countries for Establishing Maximum Permissible Levels of Harmful Agents in the Working Environment." Fondazione Carlo Erba: Milan. Vorwald AJ, Durkan TM, Pratt PC (1951): Experimental studies of asbestosis. Arch Ind Hyg Occup Med 3:1-43. Vorwald AJ (May 7, 1952): First Interim Report/Asbestosis and Pulmonary Cancer. To Quebec Asbestos Mining Assoc, Saranac Laboratory, Saranac Lake, N.Y. Vorwald Archives, Armed Forces Institute of Pathology, Washington, D.C. Vorwald AJ (1948): (Saranac Laboratory). Letter to Dr. A.J. Lanza (Metropolitan Life Insurance Co.), Dec. 6, 1948; and letter from Lanza to Vorwald, Dec. 14, 1948. Obtained from files of Manville Corp. by the U.S. Justice Dept. Wade N (1976): NAS committee on asbestos: discovery of a special relationship. Science 193:661-664.