<< Back to Final Report Metalworking Fluids Standards Advisory Committee


The major health issues related to MWFs include: dermatitis, acute and chronic respiratory disease, skin cancer and other cancers (M9:21). These issues were addressed by the committee primarily during the second, fifth and seventh meetings and discussed at other meetings. This chapter addresses the presentations, literature review and discussion related to these health issues. This chapter is organized in a different way from the other chapters with each major category represented under the individual health issues.

In preparation for discussions about the health issues, committee member Dr. David Wegman provided an overview of epidemiology. He explained the different types of epidemiological studies, and the pros and cons of each. In addition, comments by Dr. Daniel Hoffman also provided background for the committee. More information about basic epidemiology can be found in their handouts (Wegman, 1997; Hoffman, 1998a) and in Monson's Occupational Epidemiology (1990) and Research Methods in Occupational Epidemiology by Checkoway (1989). The NIOSH Criteria for a Recommended Standard - Occupational Exposure to Metalworking Fluids was the starting point for the committee's discussion of health issues.


2.2.1 Speakers and Presentations

Dermatitis was discussed in detail during the second meeting of the committee and was discussed at later meetings. Dr. Robert Adams, Professor Emeritus from Stanford University and practicing clinical dermatologist addressed the group (M2:2-3). Adams, and Dr. Boris Lusniak from NIOSH, explained the skin problems associated with exposure to MWFs (M2-3;13-15). Lusniak described the studies cited in the NIOSH Criteria Document. Stephen Gauthier, a machinist at a large East Coast manufacturer described his own experiences with dermatitis and MWFs (M2:17-18; M8:18-19). Dr. Larry Fine, NIOSH, mentioned dermatitis in his overview of the NIOSH Criteria Document (M2:2). Greg Piacitelli, NIOSH, noted dermatitis cases in his description of the NIOSH Small Business Study (M4:3; M7:4). Tom Beeman, a machinist at a mid to large facility in the Western part of the US provided some limited information about his own skin disorder (M5:3). Dr. William Lucke, Cincinnati Milicron, in his presentation noted formulators' efforts to reduce dermatitis (M5:21). Dr. Ed Stein, OSHA, provided background information on previous OSHA and NIOSH recommendations for dermatoses (M5:28-29). John Burke, Eaton Corporation, noted dermatitis during his discussion of problems in middle size facilities using MWFs (M627). Michelle Lantz, Caterpillar Corporation, cited incidents of dermatitis in her presentation on systems management (M8:10-11). Committee member, David Burch, noted dermatitis in his presentation on machining in small business (M2:4). The OSHA Office of Regulatory Analysis provided an overview of statistical information on occupational dermatitis. Presentation notes of Laura Nakoneczny, Precision Metalforming Association (PMA) were provided by David Burch, committee member. Other committee members and alternates provided their experiences and interpretations throughout the meetings.

In the discussion of rates of adverse health effects, three types of data were presented: anecdotal or case reports, surveys of plant experience and formal cohort or cross-sectional studies. The first type provides only evidence that the problem exists in the setting from which the report comes and may exist in comparable settings. The second type is limited by the quality of the different reporting units (plants) and no effort has been made to determine that each was equally aggressive in identifying and recording adverse health effects. Generally these surveys were based on OSHA 200 logs which may or may not have been complete. These survey results, therefore, should be seen as offering a different type of information than case reports with less quantitative reliability than systematic scientific studies. These survey results are limited by the sources of data. The third type of data, formal studies, is the most reliable, although these types of studies have been carried out only to a limited extent in occupational environments using MWFs (Wegman, tenth meeting)

2.2.2 Background Information

Lusniak cited Bureau of Labor (BLS) statistics indicating that occupational skin diseases represent 12% of all occupational illnesses, are the second highest reported occupational illness category, and are responsible for annual expenditures of $22 million (1984 dollars) (M2:13). From 1973 though 1987, dermatitis was the leading occupational illness in the US (Stein, 1998, NIOSH,1998). Twenty-one percent of all occupational skin disease reported to the BLS result in days away from work, according to Lusniak (M2:13). The rate of skin disorders was 76/100,000 workers in 1993 and health professionals have a target of 55/100,000 by the year 2000 (M2:14).

Lusniak believed that the BLS data underestimates values because this agency bases its numbers on OSHA 200 logs and he has seen individual cases that were not reported (M2:13,14). He stated that dermatitis represents 90-95% of all occupational skin disease (M2:13).

The OSHA Office of Regulatory Analysis used BLS data to provide estimates of skin disease (OSHA Office of Regulatory Analysis, 1998). In 1996 there were 58,100 recordable skin diseases and disorders in private industry. This is an average rate of 6.9 per 10,000 full time equivalent workers (FTEs) (OSHA Office of Regulatory Analysis, 1998).

2.2.3 Experiences and Resources Presentations and Related Discussions

Lusniak and Stein cited deBoer's study indicating a 14% prevalence rate of dermatitis among those exposed to MWFs (M2:13). In his handout, Stein cites studies by Rycoft with a prevalence rate of 30% and Sprince with 27.2% (Stein, 1998a). NIOSH Health Hazard Evaluations (HHE) indicate a prevalence rate from 14-67% (M2:13; Stein, 1998a). Adams disagreed with the HHE numbers and estimated that 1-2% of machinists have dermatitis (M2:3).

Burch noted that in his survey of members of the PMPA, 120 out of 580 member companies sent in 667 OSHA 200 summaries (M2:4). In these 667 summaries, a total of 410 dermatitis cases were listed (M2:4). Burch explained that this was a rate of 1.6 OSHA recordable cases per hundred full time employees (M3:13). This works out to 1 worker in 61 may develop dermatitis related to MWFs (Nakoneczny,1998).

For another small business organization, PMA, 1 worker in 3,339 may develop dermatitis based on OSHA 200 logs (Nakoneczny, 1998). For PMA, in 66,739 employee-years of exposure, there were 20 reports of dermatitis for a 0.001 incidence rate (Nakoneczny, 1998).

Piacitelli cited 4 OSHA 200 logs out of 39 collected during the NIOSH Small Business Study that indicated dermatitis (M4:3). In a later update, Piacitelli cited 2 cases of dermatitis out of 30 OSHA 200 logs evaluated (M7:4).

Lick noted the dermatitis case observed at one of the site visits to a small business (M4:6). Day explained that every MWF plant he visits has cases of dermatitis (M3:13). McGee believed that dermatitis is the leading cause of lost work time in MWF plants (M3:14).

Mirer cited a MIOSHA inspection of a MWF plant with 60 employees and 2 cases of dermatitis, at exposure levels slightly above the NIOSH REL (M6:37). Mirer noted the disparity between another MWF plant whose OSHA 200 logs listed one dermatitis case (M6:37). An HHE report for this second plant indicated that of 8 randomly selected employees, five had visible rashes and all eight complained of skin symptoms (M6:37).

Lusniak explained that the factors contributing to the development of MWF related skin disease include: degree of skin contact, individual susceptibility, personal protective equipment, overall work environment, climate, machine types, control methods and the types of MWFs used (M2:13). Barrier creams, cleansers, work habits, machine type, and workplace climate are additional factors (Adams, 1997). Adams noted the MWF exposure to the machinist's skin during the changing of machine parts and during maintenance (M2:2). Burch explained that different machining operations have different degrees of contact with the fluids (M2:20). Stein cited reuse of MWF soaked clothing or materials as a potential source of problems (Stein, 1998a).

Burke explained his experiences with severe dermatitis in his mid size facilities, citing individual fluid formulations, on-site chemical addition and lack of proper fluid maintenance as the causes (M5:27). Lantz noted dermatitis occurs when tramp oil in the MWF is allowed to rise to approximately 5% of the fluid (M8:10). Howell explained that individual machines provided more opportunity for skin contact with fluids than transfer lines (M8:26).

In 90-95% of occupational dermatitis cases, the primary site is the hands followed by forearms, face and neck, according to Lusniak (M2:13). Gauthier provided an example that his hands were the area for his dermatitis (M2:17). Beeman suffered from rashes on his arms during the first six years of his employment at a plant using MWFs (M5:3). Lantz noted a case of dermatitis in the stomach area which contacted fluid when a worker reached across a machine (T8:274-276). Gauthier explained that skin in contact with shop rags saturated with MWFs often is an area with dermatitis (M8:18).

Lusniak explained that MWFs can cause irritant and contact dermatitis along with folliculitis, oil keratosis, pigmentary changes and oil granulomas (M2:13). He noted that straight MWFS are associated with folliculitis, keratosis and skin cancer (M2:13; M5:28). Folliculitis is associated with infectious agents and these microorganisms are more of a problem for someone with already irritated skin that is not intact (M2:13). Skin cancer was associated with polyaromatic hydrocarbons that are no longer in use, so skin cancer is rarely seen, according to Lusniak (M2:13). He explained that irritant contact dermatitis and allergic contact dermatitis are seen more often with soluble, semi-synthetics and synthetics (M2:13). According to Lusniak, 50 - 80% of MWF related skin diseases are irritant contact dermatitis, while 20-50% are allergic contact dermatitis (M2:13). Lusniak had no doubt that MWFs cause both irritant and allergic dermatitis (M2:14). Stein reiterated many of Lusniak's comments in his talk (M5:28). Gauthier explained the personal pain and embarrassment caused by these disorders (M2:18).

Lusniak explained that irritant contact dermatitis is associated with MWF ingredients such as alkaline emulsifiers and solvents, microtrauma from shaving contaminants and hand-washing with irritating detergents (M2:13). Howell noted that excess alkalinity in MWFs causes defatting of the skin (M2:19). Howell explained that fluid producers take special care to formulate MWFs that will not cause dermatitis on a routine basis (M2:19). Lucke noted that some suppliers conduct dermal toxicity testing of formulations prior to marketing (M5:21).

Allergic contact dermatitis can be due to metals, additives such as biocides, and emulsifiers, according to Lusniak (M2:13). Adams explained that once anyone is sensitized to an allergen it takes contact with a very small amount to elicit a response (M2:2). Adams noted that overuse of biocides can lead to an irritation reaction that eventually becomes an allergic contact dermatitis (M2:2). Biocides known to cause allergic contact dermatitis include: Grotans (BK, HD, HD-2 and K) which are, Biobans (PS-1487, CS-1246, CS 1248, CS-1135), Proxels, Kathons, Forcide 78, 1-H benxothiazole, o-Phenylphenol, p-Chloro-xylenol and Tris Nitro (Adams, 1997). Other sensitizers that can be present include: rosin, mercaptobenzothiazole, nickel, cobalt and chromium (Adams, 1997). Day noted his own allergic contact dermatitis related to MWFs and how the site visit in March, 1998 elicited a hives reaction (M3:13). Adams explained that workers may not realize that the dermatitis they have is related to MWFs until they are tested (M2:2-3). This clarified the lower self reported dermatitis cases than physician reported cases in the Sprince study (M2:2-3). Allergies to some individual MWF components can be determined by patch testing, but according to Adams, the FDA restrictions on allergen test kits hamper the dermatologist's diagnosis and treatment (M2:2). Additional Information from the NIOSH Criteria Document

NIOSH discusses dermatologic conditions at the end of Chapter Five of its Criteria Document (NIOSH, 1998). Lusniak and Stein reviewed many of the issues discussed in the Criteria Document. Some earlier statistics are cited such as the 1991 occupational skin disease incidence rates of 7.7 per 10,000 workers for all industry (NIOSH, 1998). NIOSH cites Coenraads, 1983 indicating a general population prevalence of dermatitis of 4.6% compared to the ranges indicated in the previous section of 14-30% for machining environments (NIOSH, 1998).

The NIOSH Criteria Document provides specific chemical names for some of the components of MWFs that are allergens (NIOSH, 1998). These are found in section 5.4.3, page 141 of the Criteria Document (NIOSH, 1998). The document includes some similar conclusions as Adams about the limits of patch testing (NIOSH, 1998). According to Alomar as cited in the NIOSH Criteria Document, biocides, corrosion inhibitors, coupling agents, and emulsifiers produce the most frequent positives on skin patch testing (NIOSH, 1998).

The document lists control strategies for MWF related dermatitis (NIOSH, 1998). These include: substitution of fluids or additives or constituents; process modification, isolation, and ventilation; work practice and administrative controls to assure fluid maintenance and workplace cleanliness; proper use of personal protective equipment; education and training of employees; minimal contact with fluids; and personal hygiene such as clothing changes and cleaning with non-abrasive soaps (NIOSH, 1998). Other Resources

A handout prepared by Stein provided some statistics on dermatitis and MWFs (Stein, 1998a-d). In 1991, the Department of Labor (DOL) noted that the highest incidence rates for skin disease included fabricated screw machine products with 33.3 cases per 10,000 full time equivalent workers (FTEs), and general industrial machinery with 22.0 per 10,000 FTEs (Stein, 1998a-d).

Information provided by the OSHA Office of Regulatory Analysis used BLS statistics for 1996 to estimate skin disease in SIC codes 33-37 (OSHA Office of Regulatory Analysis, 1998). In these industries, there were 14,300 recordable skin diseases and disorders, accounting for approximately 25% of all recordable skin diseases and disorders in private industry (OSHA Office of Regulatory Analysis, 1998). SIC codes 33-37 had an average rate of 18.3 recordable skin diseases and disorders per 10,000 FTEs, almost three times the average for all of private industry (OSHA Office of Regulatory Analysis, 1998). The transportation equipment industry (SIC 37) had the highest skin disease and disorder rate of any industry group using MWFs, 33.9 cases per 10,000 FTEs (OSHA Office of Regulatory Analysis, 1998). Each of the five industry groups in SIC 33-37 were at least 70% above the rate for all private industry (OSHA Office of Regulatory Analysis, 1998). More information on lost work days and economic costs of dermatitis based on the OSHA Office of Regulatory Analysis work is provided in Chapter Three.

An article by Sluhan (1997) cites the following causes of dermatitis from MWFs: alkalinity, acidity, solvents, metals, the use of straight MWFs, filthy MWFs, misapplication of biocides, handling equipment, concentration problems and misuse of protective creams. Sluhan also warns that the cause could be contamination from an external source or outside activities and not MWFs (Sluhan,1997). He provides examples of solutions that are similar to what has already been reported (Sluhan,1997).

An article by Itschner, 1996 was cited by Teitelbaum at the tenth meeting. Additional references on dermatitis are found in Chapter Eight, Medical Surveillance and in Attachment #6.

The UAW publication How to Prevent Skin Disease outlines causes of dermatitis and how management and workers can prevent these disorders (UAW, 1997). Adams provided a list in his handout of recent references on MWFs and dermatitis (Adams, 1997). Stein provided in his handout, a list of references, a glossary and information on the OSHA Standards Advisory Committee on Cutaneous Hazards and other early efforts by NIOSH and OSHA (Stein, 1998a-d). The two MWF Symposia Proceedings provide additional information and discussion (AAMA, 1996,1998).

2.2.4 Concerns and Limitations Size of Business

As noted earlier, Burch reported that out of 667 OSHA logs for his members, a total of 410 dermatitis cases were listed, for a rate of 1.6 cases per hundred full time employees (M2:4). Comparing PMPA data to the OSHA Office of Regulatory Analysis assessment, and assuming the same method of calculating cases and FTEs was used: the rate for PMPA would be 160 cases per 10,000 FTEs. Using Nakoneczny's data, the rate for PMA would be 2.9 per 10,000 FTEs (Nakoneczny,1998). Nakoneczny's combined dataset of PMA and PMPA produces a rate of 45.6 cases per 10,000 (Nakoneczny, 1998). Assuming the same method was used for calculating cases, these values can be compared with 6.9 cases per 10,000 FTEs for private industry, 18.3 cases per 10,000 for SIC codes 33-37 and 33.9 cases per 10,000 FTEs for SIC code 37 (OSHA Office of Regulatory Analysis, 1998).

Piacitelli cited 2 cases of dermatitis out of 30 OSHA 200 logs evaluated (M7:4). No calculation per FTE was made. Day and McGee explained that dermatitis is seen universally in MWF plants (M3:13,14). Based on MIOSHA inspection reports and NIOSH HHE, Mirer thought there was an underreporting of dermatitis on small business OSHA 200 logs (M6:37). Atopy

Adams stated that 10-15% of the US population is atopic; 3-5% have atopic skin reactions (M2:3). Adams noted that atopic individuals have a greater chance of developing any allergic skin reactions, but not all will react to MWFs (M2:2). Three or more major features are required for a diagnosis of atopic dermatitis: itching, typical morphology and distribution, chronic and relapsing course and personal and family history of atopy (Adams, 1997 citing Hanifen, 1980). Extremes of temperature, presence of irritants and certain microorganisms and emotions aggravate atopic dermatitis (Adams, 1997). According to Lusniak, atopics have a greater risk of irritant contact dermatitis but no increased risk of allergic contact dermatitis (M2:14). Adams questioned the legality of preventing atopics from working with MWFs (M2:2). Newman and McGee warned that even if we exclude susceptible individuals, dermatitis will still occur with MWFs (M2:19). Other Issues

Teitelbaum and Frederick expressed concern about skin absorption (M2:2,14). It was noted by Fine that no urine or other analyses were available for MWFs (M2:2). Lusniak explained that absorption of any workplace chemical will be enhanced if the skin is not intact (M2;14). Howell noted that some MWF components could pass through the skin (M8:26). Teitelbaum noted that after review of the ILMA handout with CAS numbers, he was less concerned about benzene absorption since the benzene content appeared to be de minimus (M8:26; ILMA 1998).

2.2.5 Linkage of Discussions to OSHA Action

Adams and Lusniak emphasized early detection of dermatitis, but Adams thought medical surveillance may not provide enough "return on investment" (M2:3). If an agent is truly removed from an individual's environment, any dermatosis is reversible according to Adams (M2:3). Lusniak cited a study by Pryce, 1989 that showed a poor rate of healing even with removal from the MWF environment, and stressed the importance of prevention (M2:13).

Gauthier noted that his dermatitis improved and went away after reducing exposure (M2:17). Gauthier, Howell, Burke and Kushner explained the importance of good fluid management and Teitelbaum and Shortell agreed (M2:17 & 19; M4:8; M6:27 & 40; M8:4). Shortell noted the presence of dermatitis when fluids were not well managed in plants he worked in during his career as a machinist (M2:19). Mirer agreed based on his experiences (M3:13). Stein showed a downward trend in dermatitis, and Cox believed this trend will continue with the purchase of new machines and better education (M5:28). Ways to reduce exposure are discussed in Chapter Three of this report, Technological Feasibility.

Burch was concerned that we not issue a blanket statement that MWFs cause dermatitis because not everyone working with MWFs experiences dermatitis (M2:19). Newman emphasized that having 100% of workers affected is not required for action (M2:20). At the tenth meeting, Wegman and Teitelbaum agreed with Newman, stating that this is true for any disease and any exposure. Burch agreed with Newman, but questioned what level is required, and what the Americans with Disabilities (ADA) implications of any action would be (M2:20).

Wegman viewed dermatitis as the third most important health effect associated with MWFs, listing it after asthma and HP (M5:31). Mirer cited Gauthier's experience of dermatitis and the potential economic impact of job transfer as an example of material impairment of health (M2:19). Howell agreed that unresolved dermatitis can be a material impairment of health (M2:20). Stein explained that there were no OSHA standards based solely on dermatitis although many include dermatitis as one of the important health effects (M5:28).

2.2.6 Committee Decisions and Rationale

The majority (13) opinion of the committee was that dermatitis is known to be associated with exposure to MWFs (M9:21-22). Members cited their own experiences: working with individuals who had dermatitis, treating employees with dermatitis, and observations at a MWF plant visited by the work groups (M9:21-22). In addition, presentations by Adams, Lusniak and Gauthier (M2:23;13-15; 17-18), the NIOSH Criteria Document and the letters sent by small business to Jeffress were noted as evidence (M9:22; NIOSH, 1998; PMPA, 1999; PMA, 1999). Dermatitis from MWF is a material impairment of health (M2:19; M9:22).

The minority opinion (Burch, Howell) of the committee was the evidence was equivocal (M9:21). These members noted their own experiences and statements that dermatitis is associated with poorly managed fluids (M2:19; M3:13). Manufacturers test and produce fluids that when new, generally do not cause dermatitis (M 2:19).

Two members noted that all MWFs can cause dermatitis (M9:22). Two other members (Day, McGee) explained that all MWF plants they have been in had workers with dermatitis (M9;21-22). Two members (Cox, Burch) stated that although there are dermatitis problems, these problems are controllable (M9:21-22).


2.3.1 Speakers and Presentations

Acute and chronic respiratory effects were discussed in detail during the second and fifth meeting of the committee and were discussed at other meetings. Committee members and alternates provided background information, their own studies, experiences, expertise and interpretations throughout the meetings. Individual researchers presented their work and a worker provided his experience. NIOSH and OSHA provided representatives to explain their analyses of the studies. Industry consultants provided an overview and critique of studies. The NIOSH Criteria Document was used as a starting point for discussion (NIOSH, 1998).

Dr. Lee Newman, committee member, provided an overview of the respiratory system and diseases (M2:5-6). Committee member, David Burch addressed respiratory issues in his presentation on machining in small business (M2:4). Dr. Kenneth Rosenman, Michigan State University, discussed asthma in his presentation on the Michigan SENSOR program (M4:7-8). Dr. Ellen Eisen, University of Massachusetts - Lowell, described her study of asthma and MWFs (M5:1-2). Dr. Kevin Fennelly, National Jewish Research and Medical Center, provided information on the clinical aspects of asthma (M5:2-3). Tom Beeman, a machinist at a mid to large facility in the Western part of the US documented his own experience with asthma and MWFs (M5:3). Dr. Henry Anderson, committee member, explained his study of an HP outbreak in Wisconsin (M5:3-4). Dr. Cecile Rose, National Jewish Research and Medical Center, documented various case studies of HP and addressed medical removal (5-6). Dr. Michael Hodgson, University of Connecticut, described investigatioqs of HP outbreaks in his state (M5:6-7). Dr. Thomas Robins, University of Michigan, described asthma, cross shift pulmonary function tests and other respiratory effects related to MWFs (M5:8-10). Dr. Michelle Schaper discussed her inhalation toxicity tests of MWFs (M5:18-19). Dr. Harold Rossmoore described the importance of the microbiology of MWFs and cited respiratory problems related to these organisms (M5:19-20). Dr. William Lucke, Cincinnati Milicron, in his presentation noted formulators efforts to reduce respiratory effects (M5:21). Dr. Gordon Reeve, Ford Motor Company, presented prevalence data and hospital admissions related to non-malignant respiratory disease at his company (M6:1-6; M7:27-28). John Burke, Eaton Corporation and committee alternate member, Ken Kushner noted respiratory symptoms during their discussions of problems in middle size facilities using MWFs (M6:27-29; M6:33).

Dr. Larry Fine, NIOSH, explained the respiratory disease section of the NIOSH Criteria Document (M2:6-7). William Perry and Dr. Steven Bayard, OSHA, outlined OSHA's progress on risk assessment on non-malignant respiratory disease (M6:19-23, 10th meeting). Dr. Daniel Hoffman, George Washington University, explained his evaluation of the articles summarized in the respiratory effects section of the NIOSH Criteria Document which he did as a consultant for ILMA (M5:10-14).

A panel discussion consisting of Dr. Fennelly and Dr. Rose along with committee members, Dr. David Wegman, Dr. Lee Newman, and Dr. Henry Anderson, fielded committee questions on respiratory disease and MWFs (M 5:7-8). Schaper, Rossmoore and Lucke joined Dr. Daniel Goon, Castrol and committee member, Dr. Daniel Teitelbaum in a panel discussing MWF components (M5:20-26). Dr. Eugene White, NIOSH, provided an update on endotoxins (M9:1).

In the discussion of rates of adverse health effects, three types of data were presented: anecdotal or case reports, surveys of plant experience and formal cohort or cross-sectional studies. The first type provides only evidence that the problem exists in the setting from which the report comes and may exist in comparable settings. The second type is limited by the quality of the different reporting units (plants) and no effort has been made to determine that each was equally aggressive in indentifying and recording adverse health effects. Generally these surveys were based on OSHA 200 logs which may or may not have been complete. These survey results, therefore, should be seen as offering a different type of information than case reports with less quantitative reliability than systematic scientific studies. These survey results are limited by the sources of data. The third type of data, formal studies, is the most reliable, although these types of studies have been carried out only to a limited extent in occupational environments using MWFs (Wegman, tenth meeting).

2.3.2 Background Information

Newman explained that respiratory diseases can be divided into those that affect the conducting airways and those that affect the alveoli, the gas exchange region (M2:5). Asthma is a conducting airway disease, while hypersensitivity pneumonitis (HP) affects the alveoli (M2:5). He noted that much misclassification of respiratory disease occurs (M2:6).

A variety of studies were investigated by the committee. Limited toxicology studies were available (M5:18). Methods for testing toxicity have been developed for water miscible MWFs (ASTM,1993). Epidemiological studies included a range of types as explained by Wegman and Hoffman (M2:7-9; M5:11-14). Studies included exposure measurement or estimation using different exposure metrics. A variety of health endpoints were covered in the epidemiological studies. Symptom reporting using questionnaires was included in some studies (M4:7).

Medical testing was used in some investigations. Pulmonary function changes such as forced expiratory volume (FEV1.0) and forced vital capacity (FVC) were used as indicators of underlying disease or potential disease. Newman explained that forced expiratory volume (FEV1.0) represents the volume of air one can force out in one second (M2:5). He noted that the forced vital capacity (FVC) is the total amount of air that a person can inhale in one breath (M2:5). These tests are part of pulmonary function testing and change with disease (M2:5). Reeve noted the difficulties in obtaining consistent pulmonary function test results (M6:1).The comparison of these pulmonary function tests before and after a work shift is called cross-shift pulmonary function testing.

According to E. White, endotoxins have been associated with respiratory symptoms such as: coughing, wheezing, fever, chills and decreased FEV1.0 (M9;1). Endotoxin can potentiate immunological reactions, exacerbate illness caused by other agents but may not be the instigating cause of a disease (M9:1). Endotoxins are a class of pyrogenic compounds derived from the outer cell membrane of Gram negative bacteria and consist of lipopolysaccharides (E.White, 1999). The lipid A portion of the molecule is responsible for its toxicity (E.White, 1999).

Additional information on general concepts on respiratory disease can be found in Newman's handout (Newman, 1997), and his chapter on Pulmonary Toxicology in the book Clinical Principles of Environmental Health (Newman, 1992).

2.3.3 Experiences and Resources Related to Asthma and Airway Effects Overview

Newman defined asthma as an inflammatory disease of the airways in which an inhaled substance can trigger a narrowing of the bronchial passage (M2:5). Fennelly clinically defined asthma as reversible air flow obstruction with fairly common symptoms usually chest tightness, shortness of breath and may include wheezing (M5:2). Fennelly cited a review of asthma by Chan-Yeung (1995). Rose explained that about 16% of asthma is occupational (M5:8). Occupational asthma has become the most common work-related respiratory disorder, representing 26% of these disorders in the United Kingdom (Newman, 1995). According to Wegman, chronic obstructive pulmonary disease (COPD) refers to significant decrements in lung function and increases in symptoms such as chronic bronchitis, cough and phlegm (M5:31).

Induction of asthma can have a long latent period between initial exposure and first manifestation according to Newman and Fennelly (M2:5; M5:2). Fennelly explained that "without latency type" is called irritant asthma (M5:2). "Latency type" can be divided into "lgE allergic antibody dependent" or "lgE independent", according to Fennelly (M5:2). Ethanolamines, Iow molecular weight compounds may be IgE independent, while microbiological toxins may not follow any distinct pattern stated Fennelly (M5:2). He noted that sensitization may take years of exposure to an agent and may require large amounts of the agent inhaled or on the skin (M5:2). Fennelly noted that once an individual is sensitized, only minute amounts are needed to elicit the allergic reaction (M5:2). He explained that peak exposures may be more important than averages but it depends on how a toxin behaves in the respiratory tract (M5:2).

Teitelbaum noted that an upper airway version of asthma, called reactive upper airway dysfunction syndrome, can be due to large particles (M2:5). Newman explained that sinusitis and asthma are on a continuum that relates upper and lower airway problems (M2:6). Newman noted that asthma and bronchitis are obstructive diseases which make it difficult for the individual to blow air out of the lungs (M2:5). Bronchitis is inflammation of the bronchial tubes and more often has cough associated with it, while asthma is a reactive narrowing of the bronchial tubes according to Newman (M2:6). Burch noted the increase in asthma in the general population and Teitelbaum agreed that this increase is occurring (M2:6). Researchers' Reports of Their Own Studies

Eisen reported on her analysis of asthma in the cross sectional study of auto workers done by Greaves (M5:1). After suspecting her original analysis was affected by the healthy worker effect, Eisen focused on when asthma was diagnosed in these workers (M5:1). One hundred and twelve workers out of 1800 had asthma and of these, 29 cases were post-hire (M5:1). She formed a control group of other workers without asthma who had worked in the plant for the same time as each of the 29 workers, and corrected for date of hire and other variables (M5:1). Eisen specifically examined exposure in the two years prior to asthma onset in the cases, and a comparable two year period for the controls (M5:1). She found an odds ratio exposed/control of 2.1 for synthetic MWFs (M5:1). Based on her study, she thought synthetics were more of an asthma problem than straights which were more of a problem than solubles (M5:2). Additional information can be found in her article (Eisen, 1997).

Robins explained his study comparing two machining departments, "case" and "valve body" in an auto plant using soluble fluids with minimal use of biocides and no Mycobacteria found (M5:8). The mean seniority of machinists was 19 years (M5:8). His study consisted of three rounds of baseline questionnaires, pre and post shift questionnaires and spirometry including FEV1.0, FVC and the ratio between these values (M5:8). Robins study was a case series design according to Wegman (M2:7-8). Area and personal air samples including personal thoracic sampling with a PEM were taken (M5:8). Viable and nonviable bacteria and endotoxin were measured (M5:8). Thoracic particulate means were 0.54 mg/m3 for "case", 0.28 mg/m3 for "valve body" and 0.13 mg/m3 in an assembly area away from MWFs (M5:8). Robins used 1.25 as a "rule of thumb" ratio of total/thoracic particulate (M5:8).

Robins defined the development of post-hire asthma as a 12% or greater drop in cross shift lung function (M5:9). This occurred in 11 out of 83 machinists and 2 out of 44 assemblers (M5:9). Machinists were more likely to have this drop, but it was not significant (M5:9). Six out of 83 machinists had a drop of 19% or more, while none of the assemblers had this great a drop (M5:9). Robins found that machinists were more likely to have a 10% decrement in cross shift lung function than were assemblers (M5:8).

Robins found that twenty five percent of machinists who were obstructed smokers had a greater than 10% decrement in cross shift lung function compared to 3% of non-obstructed non-smokers (M5:8). He compared groups of workers to non-obstructed, non-smoking assemblers and found odds ratios for cross shift lung function changes ranging from 2.75 to 6.22 (M5:9). With obstruction, the ratios increased with exposure, while in the groups without obstruction, the cross shift change did not increase with exposure (M5:9). Bacterial and thoracic particulate fit the health effects well (M5:9). Robins was concerned that repeated cross shift lung function decrements could lead to permanent irreversible changes (M5:9).

Robin's questionnaire data showed a higher proportion of machinists reported symptoms as compared to people in assembly (M5:9). Key symptoms he reported were: phlegm production, dry cough, wheezing, chest tightness and dyspnea, all lower respiratory tract symptoms (M5:9). He explained that there are false positives for answers on questionnaires (M5:9). Despite higher endotoxin levels in the case area, symptoms were different from valve body (M5:10). More details about Robin's study can be found in his presentation handout (Robins, 1998).

Asthma sentinel cases should prompt medical surveillance according to Rosenman (M4:7). His review of symptom questionnaires and medical records of the individuals that report physician identified asthma cases, provided strong enough evidence for him to say there is a cause and effect relationship between exposure to MWFs and asthma (M4:8). Rosenman found for 5400 questionnaires: 1000 reported daily or weekly symptoms of asthma, and only eight of these showed up on OSHA 200 logs (M4:8). Rosenman reported that in Michigan, the 146 MWF work related cases of asthma he studied represented 12.4% of all work related asthma for that time period (M4:7).

Respiratory symptom reporting has a background value in control plants without exposure to known irritants, of about 10%, so values above this number are significant, according to Rosenman (M4:7-8). He explained that his work showed a trend of an increase in the number of symptomatic people with increased exposure (M4:7). Hoffman noted that Rosenman found higher values of asthma in plants using soluble or synthetic fluids (M5:11). Rosenman indicated that there were more symptomatic people in plants with synthetic MWFs, while individuals in plants with straight fluids had less need to see a physician for shortness of breath or sinus problems (M4:7). Rosenman explained that MiOSHA recommends a medical surveillance program for facilities when 20% or more of the workers are symptomatic (M4:7). Such programs catch people in the early treatable phase of the disease and can help remove people from contact with a sensitizing agent, according to Rosenman (M4:7). More information about the characteristics of the workers in Rosenman's study can be found in his handout and articles (Rosenman, 1993;1997;1998).

Burch explained that in his survey of members of the PMPA, 120 out of 580 member companies sent in 667 OSHA 200 summaries (M2:4). In these 667 summaries, a total of 34 respiratory cases were listed and only one of these was related to MWFs (M2:4). This works out to an incidence rate of 0.004, or 1 worker in 25,118 may develop a MWF induced respiratory disease (Nakoneczny, 1998).

Data for PMA showed one reported respiratory condition for an incidence rate of 0.001 (Nakoneczny, 1998). This works out to 1 case per 66,739 workers (N akoneczny, 1998).

Reeve reviewed his assessment of Ford's medical surveillance database which includes visits to the medical department, OSHA logs, workman's compensation and compensated private doctor or emergency room visits (M6:16). He compared the respiratory diagnosis data for 11 MWF-using plants to 6 glass plants without MWF or suspected irritants (M6:2). For 1997, the MWF plants had 20,000 workers representing 45 million hours of work while the glass plants had 4000 workers with 7.7 million hours worked (M6:2). He included the cases with reported symptoms of upper respiratory tract irritation, cough, congestion, throat irritation, tightness in chest, shortness of breath, wheezing, sinus problems, difficulty breathing, allergy, asthma and pneumonia (M6:2). He eliminated non-relevant cases and for 1997, found a rate of 0.66 cases per 200,000 hours worked for the MWF plants and 0.34 cases per 200,000 hours worked for the glass plants (M6:2). There were 148 respiratory cases in the MWF plants and 13 respiratory cases in the glass plants (M6:2). The lost time case rate was 0.05 per 200,000 hours and the severity rate was 0.54 per 200,000 hours for the MWF plants in 1997 and there were 121 lost days of work. There were no lost days of work at the glass plant related to their cases in 1997 (M6:2). Reeve found 7 cases of asthma initiated or made worse by working in MWF plants and none in the glass plants (M6:3). Reeve noted that it is difficult to track asthma treated by private physicians (M6:3). Additional information such as plant by plant data and diagnoses considered can be found in his handouts (Reeve, 1998a).

Reeve provided an additional assessment of the treatment of Ford workers outside of the occupational arena and compared the results of MWF plants to non- MWF plants (M7:27). Reeve assessed the data from the third quarter of 1994 to the second quarter 1997 from five engine plants and 6 transmission/chassis plants (M7:27). He used information about 14,000 workers who represented about 60% of the worker population in these plants (M7:27). The control group of about 5,000 workers were from 3 glass plants, 9 parts depots and 2 parts plants and none of these plants had MWFs (M7:27)

In Reeve's study, for the MWF plants, there were 91 in-patient admissions for a primary diagnosis of non-malignant respiratory disease, and 21 for the control plants (M7:27). These admissions produced age adjusted rates of 6.2 per 1000 workers for the MWF plants compared to 4.1/1000 workers for the control plants (M7:27; Reeve, 1998b). The MWF/control relative rate for various years fluctuated between 1.4 to 1.5, indicating a 50% excess risk of hospital admission for non-malignant respiratory disease in a MWF plant (M7:27). Using one year's data, this translates for a typical Ford plant of 2000 people, to about 3.5 to 4 people put in the hospital each year (M7:27). Reeve compared Ford hospital admissions to other Midwest manufacturing and found that admissions in the MWF plants were 7% higher than average in 1996-1997, and 18% higher in 1995-1996 (M7:28). Lick clarified that Ford uses about 80% solubles, 15% synthetic or semi-synthetic and 5% straight MWFs (M7:28).

Fennelly explained a case study in which a worker sensitized to MWFs had a drop in FEV1.0 from 4 liters when not exposed, to 1.92 liters when exposed to MWFs (M5:3). He noted that it was impossible to discern the causative agent in the MWFs in this case (M5:3). Fennelly noted that this situation could have been prevented by better exposure control or at least secondarily by medical removal (M5:3). Additional details about the clinical treatment Fennelly used can be found in his handout (Fennelly, 1998).

Beeman outlined the temporal nature of his sensitization, his asthma was worse at work (M5:3). Beeman described the physical and emotional trauma he experienced from asthma and his efforts to improve his condition (M5:3). Speakers Analyses of Studies

Fine explained three positive studies, two of which were significant, that related asthma to synthetic MWFs (M2:6). Kennedy's study showed pulmonary function changes in new employees in smaller shops but none had developed asthma (M2:6). NIOSH thinks the evidence linking asthma to synthetics is quite strong, according to Fine (M2:6). The largest study on synthetics had concentrations of 0.36 to 0.91 mg/m3 with a mean of 0.6 mg/m3 (Fine, 1997). He explained that there are five studies (two significant) relating asthma to soluble MWFs (M2:6). Robin's study of solubles found that 11 out of 83 machinists had FEV1.0 decrements of greater than 12% (Fine, 1997). NIOSH believes the overall evidence points to a relation between asthma and exposure to soluble MWFs (M2:6). According to Fine, despite the weaker link between asthma and straight oils, other evidence such as cross shift pulmonary function changes and symptom reporting points to a link (M2:6). Risks are likely lower for straights than for solubles and synthetics (Fine, 1997).

Fine explained that for asthma, the risk is elevated above the REL and maybe below the REL (Fine, 1997). The risk is likely dose-dependent but independent of cigarette use (Fine, 1997). Cases of asthma and MWFs range in severity and not all cases recover after removal from exposure (Fine, 1997).

Fine explained four studies of lung function changes and MWFs (M2:6-7). Two of the three positive studies showed lung function decrements at levels averaging below the REL (M2:7). Greaves calculated a statistically significant relationship between lifetime exposure to straight oils of about a 5 ml drop in function per mg-yr of exposure (M2:7).

Hoffman summarized his work noting for asthma that the surveillance evidence showed a crude association with job title and fluid class but they did not quantify risks or clarify the nature of the association between exposure or fluid category (M5:13). Hoffman explained Gannon and Burge's study that stated a rate of 36 cases of asthma per million for metalworkers compared to 12 cases per million for a non-exposed group (M5:11). Hoffman cited Meredith's study that found a 25 fold increase in asthma in metalworkers compared to clerical workers (M5:11). Hoffman noted that these and Rosenman's study were examples of surveillance studies and that these types of studies may be under-reported or selectively reported (M5:11). Rosenman provided no quantification of risk (Hoffman, 1998a).

According to Hoffman, the European studies found no evidence for association between physician diagnosed asthma and machine fluids (M5:13). Both the Massin and Ameille studies (mean exposures between 2 and 3 mg/m3) found no difference between the total positive test to the methacholine challenge, but did see a similar steep-dose response curve for those who were methacholine positive (M5:13). Methacholine challenge testing is a possible surrogate for airway hyperresponsiveness (Hoffman, 1998a).

Hoffman interpreted the Greaves' study which produced an exposure matrix from measurements taken during the study and past industrial hygiene records (M5:12). Assembly workers had exposures at about 0.12 mg/m3, machinists had 0.45 mg/m3 for straights, 0.55 mg/m3 for soluble and 0.41 mg/m3 for synthetic according to Hoffman (M5:12; Hoffman 1998a). Greaves used a questionnaire and found that odds ratios for asthma and chronic bronchitis did not differ across job category or fluid group although bronchitis values were elevated (M5:12). Grinders had elevated symptoms of chronic bronchitis, cough and phlegm (M5:12).

Kreibel's cross-sectional, nested case control study was explained by Hoffman (M5:12). Kreibel used a 7 hole sampler for aerosol mass concentration and measured microorganisms and endotoxin (M5:12). Assemblers had exposures from non-detectable to 0.28 mg/m3 with a mean of 0.08 mg/m3, while machinists had exposures of 0.24 mg/m3 for straight fluids and 0.22 mg/m3 for soluble (M5:12). Relative risk of 5% or greater decrease in lung function of machinists/non-machinists was 0.4, according to Hoffman (M5:12). Hoffman explained that the relative risk increased significantly from 1 at lowest exposure to 3.2 at the highest exposure (M5:12). Hoffman noted the incidence of FEV1.0 decline found by Kreibel was higher among non-machinists in all concentrations (M5:12). There was no evidence of increased sensitivity to machine fluids for those people who tested positive for atopy in Kreibel's study, according to Hoffman (M5:12).

Hoffman noted other studies such as Sama's who used the same population as Kreibel but focused on sulfur concentrations using X-ray spectroscopy (M5:13; Sama, 1997). Sama found an increase in the relative risk of FEV1.0 decrement with increased sulfur concentrations and thought sulfur was a stronger predictor of this decrement, according to Hoffman (M5:13; Sama, 1997). Sama found a decrease in six-day peak flow with exposure to sulfur (M5:13; Sama, 1997).

Hoffman explained Sprince's study which used a mini-RAM for particulates, and measured airborne microbials, endotoxin and bacteria (M5:13). Total oil measurements were 0.3 mg/m3 for machinist and 0.08 mg/m3 for assemblers (M5:12). Questionnaires indicated elevated relative risks for cough, phlegm and post-shift cough in the exposed but the Monday post-shift pulmonary function tests decrement greater than 5% was not significantly different (M5:12).

Hoffman noted that the lung function studies looked at either point estimates of FEV1.0 at the start of the Monday shift or cross-shift decrements in FEV1.0, FVC, and FEV1.0 as a percentage of FVC as a measure of lung obstruction (M5:13). Hoffman explained that several studies have shown point and cross-shift decrements in FEV1.0, FVC or the FEV11.0/FVC ratio (Hoffman, 1998a). Risks range up to 6.9 in the Kennedy study and 6.2 for persons who were obstructed in Robins' study for exposure to soluble machine working fluids (M5:13). Kennedy's exposures for machinists were 0.16 to 2.03 mg/m3 vs. assembler's exposures from 0.07 to 0.44 mg/m3 (M5:12). Kennedy did not observe a decline at concentrations below 0.2 mg/m3 and found no evidence of asthma (M5:12). According to Hoffman, the exposure levels for studies that demonstrated cross-shift decrements ranged between 0.13 and 0.6 mg/m3, although they were substantially higher in the Kresniak study, up to 99 mg/m3. The prevalence of symptoms were based upon the MRC and ATS survey instruments which have proven to be fairly valid and repeatable instruments, according to Hoffman (M5:13). Ely and Oxhoji did not demonstrate any significant associations or elevations in the prevalence data for these symptoms, even when accounting for smoking status (M5:13). According to Hoffman, the more recent American and French studies did see increased prevalence ratios at levels ranging from 0.16 to 4.6 mg/m3.

According to Hoffman, the American studies are quite mixed (M5:13). Greaves found no association with asthma, although when Eisen corrected for transfer bias she did see an increased risk to exposure to synthetic fluids at a 0.4 mg/m3 thoracic about a 0.6 mg/m3 total particulates concentration (M5:13). Because of the size of the populations they studied, good power, good methods of exposure assessment, the way they measured health effects he gave the GM studies (Greaves, Eisen, Kennedy) more weight as compared with the earlier studies (M5:13). Additional analysis can be found in Hoffman's handout (Hoffman, 1998a).

Perry provided information about how OSHA could approach risk assessment for nonmalignant respiratory disease (M6:19). He noted that OSHA does risk assessment to gain a better understanding of the relationship between exposure and disease and to develop better standards (M6:19). Using Kennedy's 1989 study, Perry's preliminary work showed that at least 16-19% showed at least a 5% decrement in lung function at 0.5 mg/m3 and 2.5-4% had this decrement at 0.1 mg/m3 (M6:20). Based on the Robins 1997 study, Perry showed for 0.5 mg/m3:11.4-35.3% of individuals with obstructive lung problems had cross shift lung function changes (M6:20). Perry noted that Robins attributed some differences to diurnal variation (M6:20). Using Greave's 1997 study on respiratory symptoms and focusing on chronic bronchitis prevalence, 12.1% indicated this problem at 0.5 mg/m3 while 2.7% showed this at 0.1 mg/m3 (M6:20). Looking at the % of workers with FEV1.0 less than 80%, Greaves (1996) was close to Kennedy (1989) with 12% at 0.5 mg/m3 and 2% at 0.1 mg/m3 (M6:20). The dose response was only evident for obstructed individuals in Robins study and Greave's study could be used to determine the percentage of people who have a given percent decrement that OSHA could define as a critical value (M6:20). Perry cited, based on the Kennedy (1989) study, an excess risk estimate of 189/1000 workers at 0.5 mg/m3 and 41/1000 at 0.1 mg/m3 (M6:23). At the tenth meeting of the committee, Bayard provided an update of this preliminary risk assessment and included more studies such as Kennedy's 1999 article. Additional information can be found in Perry's and Bayard's handouts (Perry, 1998; Bayard, 1999).

Information related to asthma and airway effects may overlap with the later section of this chapter on other non-malignant respiratory effects. Check that section for additional information. Additional Information from the NIOSH Criteria Document

Section 5.1.2 of the NIOSH Criteria Document addresses asthma and other airway disorders (NIOSH, 1998). Additional case reports and surveillance program information is provided (NIOSH, 1998). The document explains how repeated exposure to an irritant can evolve into chronic bronchitis (NIOSH, 1998). Tables 5.1 through 5.4 in the Criteria Document summarize respiratory studies (NIOSH, 1998). As noted by Fine in his presentation, the summarized studies provide evidence indicative of an elevated risk of asthma (NIOSH, 1998). Irritation and sensitization appear to be involved (NIOSH, 1998). According to NIOSH, "with the exception of Ely's study, epidemiological studies of respiratory symptoms present generally consistent and (in the case of the more recent studies) compelling epidemiological evidence indicating that occupational exposure to MWF aerosols causes symptoms consistent with airway irritation, chronic bronchitis, and asthma" (NIOSH, 1998). The cross-sectional studies of lung function and three of four cross shift lung function studies generally follow the respiratory symptom data (NIOSH,1998). Other Resources Related to Asthma

Kennedy published a longitudinal study on machinist apprentices in 1999. Questionnaires, spirometry, methacholine challenge and allergy skin tests were used to compare machinist apprentices with apprentices from other trades. The mean "total" particulate exposure for machinists was 0.46 mg/m3. At baseline, before starting their apprenticeship, both groups did not differ. At follow-up after two years apprenticeship, the average change in bronchial responsiveness was double for machinists compared with controls, and machinists were more apt to have developed new bronchial hyperresponsiveness (BHR) with symptoms resembling asthma. In mathematical modeling, duration of exposure to both synthetic and soluble MWFs was the predictor of the increase in BHR.

The UAW submitted a MiOSHA SENSOR study of the Federal Mogul, Greenville, MI plant (UAW, 1999). The investigation was triggered by one physician reported case of asthma (UAW, 1999). One new case of asthma was revealed by survey of 54 people and 5 additional persons had symptoms of asthma or bronchitis (UAW, 1999). This represents an attack rate of 7/54 or 13% (UAW, 1999). None of these cases was reported on the company's OSHA 200 log (UAW, 1999). The highest measured MWF exposure at the plant was 0.33 mg/m3 (UAW, 1999).

The UAW submitted a MiOSHA SENSOR study of the GM Delphi plant (UAW, 1998). Three sentinel asthma cases prompted an investigation using a questionnaire (UAW, 1998). A 10% prevalence rate for reported asthma was found with no reports on the OSHA log (UAW, 1998). Other MiOSHA reports on Carpenter Enterprises and American Axle indicate a similar pattern of underreporting (UAW, 1998; UAW, 1999). A NIOSH HHE report indicated two cases of exacerbation of existing asthma at the Caterpillar Mossville plant (NIOSH, 1998).

Committee member, Dr. David Wegman, provided the committee a summary of many variables associated with non-malignant respiratory disease, including asthma in the MWF studies (Wegman, 1998). The summary consists of tables which define the study, design, population, fluid class, aerosol exposure concentration, health effect, #cases/#exposed, and risk estimate (Wegman, 1998). Hoffman's handout provides additional information on these studies (Hoffman 1998a).

The article, Occupational Asthma, Diagnosis, Management and Prevention by committee member, Dr. Lee Newman provides a detailed review of this disease and how physicians can address it (Newman, 1995).

A report by Cole for Caterpillar provides a critique of the NIOSH Criteria Document's assessment of non-malignant respiratory effects (Cole,1996). Cole disputes the use of surveillance data, case studies and how the studies and NIOSH interpret risk (Cole, 1996).

A general article on asthma is Enarson, D. et al, Asthma, Asthma like Symptoms, Chronic Bronchitis, and the Degree of Bronchial Hyperresponsiveness in Epidemiologic Surveys (1987). This article addresses the use of questionnaires and clinical measurements of bronchial hyperresponsiveness and recommends the development and use of a validated asthma questionnaire.

Chan-Yeung (1995) was cited by Fennelly as a good overview of asthma. Examples of other relevant general asthma articles include: Rijcken, B. et al, Longitudinal Analysis of Hyperresponsiveness and Pulmonary Function Decline (1996); Huovinen, E. et al, Mortality of Adults with Asthma: A Prospective Cohort Study (1997); Lange et al, A Fifteen Year Follow-up Study of Ventilatory Function in Adults with Asthma (1998), and Toren, K et al, Asthma and Asthma-like Symptoms in Adults Assessed by Questionnaires, A Literature Review (1993).

Additional references are cited in Chapter Eight, Medical Surveillance and are also found in Attachment #6.

Other sources of questionnaire design and recommendations on questionnaire use include Appendix C to 1910.134 OSHA Respirator Medical Evaluation Questionnaire; and Rosenman, K. Recommended Medical Screening Protocol for Workers Exposed to Occupational Allergens (Rosenman Handout, 1998). The two MWF Symposia Proceedings provide additional information and discussion (AAMA, 1996,1998).

2.3.4 Hypersensitivity Pneumonitis Overview

Rose defined hypersensitivity pneumonitis (HP) as a granulomatous lung disease resulting from repeated inhalation of, and sensitization to, a wide variety of organic dusts and some Iow molecular weight chemical antigens (M5:5). Rose noted that a complex pathogenesis involves antigen exposure and sensitization that leads to cellular events that cause lung injury and granuloma formation (M5:5). Newman explained that HP affects the alveoli, thickening and stiffening the alveolar lining and making it difficult to exchange gases and move air (M2:5). Anderson noted that HP is a restrictive disease in which the lung volume is smaller and the patient cannot expand the lungs as much as a normal individual (M5:4).

Anderson and Rose cited symptoms of HP including: cough, shortness of breath, fever, crackles, nodules and certain diffusion and pulmonary function changes (M2:5; M5:5). Rose noted precipitins in serum and the presence of a defined white cell profile in the lung lavage are also criteria (M5:5). According to Anderson, HP can resolve quickly in some acute cases or develop into a chronic, disabling and potentially fatal disease (M2:5). Anderson explained that it is very difficult to diagnose HP and define the case criteria (M5:4). Rose stated that if an individual has an acute flu like illness, subtle progressive shortness of breath, chest tightness, coughing, muscle aches, weight loss and decreased appetite, HP should be suspected (M5:5).

Rose explained that a wide range of agents can cause HP including: microbial antigens, animal proteins and reactive chemicals (M5:5). Microbial antigens and reactive chemicals may be the categories of importance with MWFs, according to Rose (M5:5).

Rose noted that HP is seen more often in non-smokers than chronic smokers (M5:5). Rose explained that attack rates can be high when you follow a sentinel HP event to find other cases (M5:5) Experiences and Resources Related to HP and MWFs

Rossmoore indicated that microorganisms may cause HP, but this has not been demonstrated unequivocally (M5:19). Rossmoore stated that in any MWF facility he has investigated that has had an incident of HP, organisms of the genus Mycobacteria were present (M5:19). He noted that he has found this genus in situations where there has been no HP diagnosed (M5:19). According to Rossmoore, gram positive microorganisms such as Streptococcus, Staphylococcus and Mycobacteria can survive aerosolization (M5:19).

The extent of endotoxin's involvement in respiratory illness such as HP related to MWFs is uncertain (E.White, 1999). Symptoms similar to HP have been associated with endotoxin in general (E.White, 1999).

Anderson described an HP outbreak he investigated in Wisconsin (M5:3). Five employees were diagnosed with HP at a plant of 1600 employees, prompting surveillance of employee contacts with the medical department (M5:4). Seventy-one records of specialist referrals were reviewed and of these, 22 mentioned a definite HP diagnosis and 12 mentioned HP/chronic bronchitis as a diagnosis (M5:4). Of these cases, 20 fit the case criteria for HP but only 18 individuals were available for study (M5:4). Twenty four of 32 cases of bronchitis were diagnosed as occupational (M5:4). Forty percent of the HP cases were in non-MWF parts of the plant, but Mirer pointed out that many of these individuals worked close to MWF areas (M5:4). All exposures to particulate, vapor and metals were Iow, but biocide use doubled to tripled in the months prior to HP diagnosis (M5:4). According to Anderson, he found four predictors of disease were: 1)diagnosed having pneumonia by a physician; 2)out for at least one 3 day sick leave; 3)restrictive spirometry pattern or 4)decreased pulmonary diffusion capacity (M5:4). Anderson explained that pneumonia may be a misdiagnosis of HP (M5:4).

In Anderson's study, fifty-one randomly selected workers agreed to participate: filling out questionnaires, giving a medical history and taking a pulmonary function test (M5:4). Of these workers, 65% had at least one hallmark respiratory symptom for HP, but none of these met the medical criteria for HP (M5:4). Thirty-one percent had abnormal spirometry or diffusing capacity (M5:4). Anderson interpreted these results as exposure to many irritants in their environment (M5:4). Time lines and clinical information can be found in Anderson's presentation handout (Anderson, 1998).

Hodgson outlined an HP study he conducted in Connecticut (M5:6). He compared the MWF plant with HP index cases to two control plants (M5:6). All oil mist measurements were below 0.5 mg/m3 and endotoxin levels were not significantly different between the case and control plants (M5:6). Some Mycobacteria were detected (M5:6). He cited many difficulties in diagnosing HP (M5:6). He urged use of a very sensitive and specific questionnaire, and that only 9 out of 13 confirmed cases responded to questions indicative of HP (M5:6). Hodgson believed that HP was endemic in MWF plants (M5:6).

Reeve reviewed data related to HP from the Ford database collected from 10 MWF plants from 1994-1996 with 19,000 workers involved (M6:2). He checked for cases with symptoms of shortness of breath, persistent coughing or difficulty breathing, repeat visits to medical, and symptoms that did not improve with antibiotics (M6:2). After eliminating some cases that were non-relevant based on the patient's narrative, 17 surveillance cases were included (M6:2). These surveillance cases were not confirmed as HP by pathology exam, and Reeve thought that some of these may be asthma, not HP (M6:2,5). Based on these surveillance cases, this represented a rate of 3 cases of potential HP per 10,000 workers (M6:2). Adding the 15 proven cases from the Connersville Ford Plant HP outbreak, produced a rate of 5 per 10,000 workers or 0.5 per 200,000 hours worked (M6:2). If only the confirmed Connersville cases are used, the rate is 2.3 per 10,000 workers or 0.02 per 200,000 hours worked (M6:2).

Fine mentioned HP and eight clusters associated with water based MWFs and that more are occurring (M2:7). Some of these cases are associated with Mycobacterium chelonae (Fine, 1997). Mirer explained that biocide use often makes HP situations worse (M2:13). Lick noted an example of an HP outbreak where the airborne MWF exposure was less than 0.5 mg/m3 but biocides were misused (M2:15). Additional Information about HP from the NIOSH Criteria Document

In section of the NIOSH Criteria document, HP is discussed (NIOSH, 1998). Another term for HP, allergic alveolitis, is explained (NIOSH, 1998). NIOSH cites Merideth's surveillance report of two cases of HP associated with MWFs in a three year period in the United Kingdom (NIOSH, 1998). Work by Bernstein (1995), Rosenman (1994), Rose (1996) and Kreiss (1997), indicates that HP associated with MWFs can occur in different size facilities (NIOSH, 1998). HP may have been occurring unrecognized for years in MWF facilities or could be the result of recent changes in these work environments (NIOSH, 1998). Prevention of contamination and careful use of biocides are noted as control strategies (NIOSH, 1998). Additional Resources about HP

The UAW submitted articles by Stephens (1996), and a UAW Hazard Alert on HP. A NIOSH HHE report indicated four probable cases of HP at the Caterpillar Mossville plant (NIOSH, 1998). A NIOSH HHE report is available about the Chrysler Kenosha Engine plant that was discussed by Anderson (NIOSH, 1997). The UAW submitted articles by Wickham (1997) and Webber (1997) about an HP cases at the GM Flint engine plant.

A NIOSH HHE report on the Ford Connersville, OH plant indicated that 14 workers had HP (NIOSH, 1998). The average total particulate exposure in the Connersville plant was 0.4 mg/m3 with a range from 0.08 to 1.17 mg/m3 (NIOSH, 1998). Mycobacteria chelonae was the dominant organism found in the sump at the Connersville plant (NIOSH, 1998).

A NIOSH HHE report of the Meritor Automotive with a worker with HP, indicated a range of total particulate from 0.33 to 1.29 mg/m3 (NIOSH, 1998). Mycobacterium chelonae was the most common organism found in some sumps with many different gram negative bacteria found in others (NIOSH, 1998). Additional information on recommendations for remediation are included in the report (NIOSH, 1998).

An article by Freeman (1998) reported a case study of a machinist who was diagnosed with HP. The machinist's HP worsened when he returned to his job where he was exposed to MWFs (Freeman, 1998). Removal from the MWF environment improved his clinical conditions (Freeman, 1998).

The two MWF Symposia Proceedings provide additional information and discussion (AAMA, 1996,1998). Additional references are cited in Chapter Eight, Medical Surveillance and are also found in Attachment #6.
2.3.5 Other Non-Malignant Respiratory Effects Overview

Newman explained the disorder lipid or lipoid pneumonia which is damage to the alveoli following inhalation of oil droplets (M2:5). He noted that hard metal disease is an immunological reaction to metals such as cobalt that may be a component of the machined metal (M2:5). Infection and irritation are other responses of the respiratory tract.

Rossmoore discussed the importance of the microorganisms in the MWFs (M5:19-20). According to Newman and Rossmoore, Legionellosis is an infectious disease that can result from exposure to microorganisms in MWFs (M2:5; M5:19). Shortell was concemed that other infections due to immune system impairment may be associated with MWFs (M9:23). Rossmoore explained that endotoxins from gram negative microorganism may be responsible for the acute respiratory syndrome associated with MWFs (M5:19). Speakers' Reports and Analyses of Studies

Rossmoore emphasized the problem of misuse of biocides that can lead to respiratory irritation and alteration of the balance of microbial species present (M5:20). Rossmoore cited an incident in a plant using biocides where workers had respiratory distress and counts of 105 Mycobacteria/ml of MWF were found in the sump (M5:20). Rossmoore was concerned about the development of biocide resistance microorganisms (M5:20). He noted that biocides are the only additive to MWFs required to have acute toxicity testing done to be registered by EPA (M5:22). He has found biocide levels in plant sumps at two to three times the limited defined by the biocide's EPA registration (M5:22). Rossmoore suggested a separate MSDS for the biocide in MWFs (M5:22).

Mirer noted that in microbial related outbreaks, 20-30% of employees have complaints (M5:25). He urged consideration of the problem of formaldehyde release by some biocides used to counter high microbial numbers (M5:26).

Burke noted the potential exposure of formaldehyde, not from biocides but from propane forklift trucks in plants (M6:28). Burke explained that some of the irritant health effects he has seen were associated with grinder use and a sulfurized MWF (M6:29). Burke urged attention be given to the role vapors may play in irritation and other effects (M6:29).

Schaper explained her irritation inhalation studies (M5:18). According to Schaper, sensory irritation refers to the response of the upper respiratory tract, while pulmonary irritation is the reaction in the lower respiratory tract (M5:18). Schaper worked with seven neat fluids that had not as yet been introduced into the work place and included one synthetic, one semi-synthetic, four solubles, and one straight (M5:18). She tested three in-use fluids that matched three of the unused fluids, for a total of 10 tested fluids (M5:18). Schaper also tested fluid ingredients (M5:18). She explained that the ASTM method she used looks for changes in both the respiratory pattern as well as the breathing rate of the animals when they are exposed to either single chemicals or mixtures (M5:18). Schaper calculates the concentration that depresses the respiratory rate to 50% of normal, the RD50 (M5:18).

Schaper noted that most of the components of the fluids do not have any PELs or TLVs (M5:18). For the individual components in the most irritating category, the RD50s ranged from sodium sulphonate around 100 mg/m3 to tolutriazole at 205 mg/m3 (Schaper, 1998). Some of these showed either sensory or respiratory irritation and some showed both of these (M5:18). Other compounds, such as some ethanolamines and oils had much higher RD50s (Schaper, 1998). Some biocides were very irritating and, more importantly, some of the mice died 24 to 72 hours after exposure to them (M5:18). Schaper explained that for the whole MWFs she studied, all were irritating to the respiratory tract, both upper airway, sensory, as well as pulmonary, so it is no surprise that the components were irritating as well (M5:18). Schaper cited work she published in 1991, stating the most irritating fluids were the synthetic and semi-synthetics and her estimated projected occupational exposure limits range from 1 to 2 mg/m3 (M5:23).

Schaper warned about the use of biocides that are so irritating when airborne (M5:18). Schaper noted that the animal bio-assay can be a rapid, inexpensive tool for evaluating respiratory irritancy, sensory and pulmonary of fluids as well as components, and be a starting point for other information so we can protect workers from sensory and pulmonary irritants (M5:18) This information along with epidemiological data can help determine a safe exposure level and the mouse bioassay can be used by MWF manufacturers to screen out those new components that may be very irritating when airborne (M5:19).

Lucke was skeptical about the predictability of the respiratory irritation test (M5:21). Lucke stated that the respiratory irritation test could not distinguish a used fluid from a fresh fluid in a situation that turned out to be an HP problem (M5:24). He believed that workers manifest symptoms at concentrations lower than would be predicted by the RD50 and did not believe RD50s should be included in MSDSs (M5:21).

According to Lucke, some formulators do conduct acute inhalation testing on their MWFs (M5:21). Howell explained that companies generate toxicity information to comply with premanufacturing notification requirements of the Toxic Substances Act (TSCA) (M5:22). TSCA requires an acute toxicity battery to determine what happens after a single acute exposure to relatively high exposure level (M5:22). Inhalation or an RD50 studies are not usually done, according to Howell (M5:22). He noted that few ILMA members get involved in the TSCA registration process because they are formulators, not makers of chemicals (M5:22). Formulators rely on suppliers for this testing (M5:22).

Aerosolization related to components was discussed as a factor in respiratory effects. Tramp oil in the sumps may increase aerosolization according to Lucke (M5:23). Anti-misting additives can reduce aerosolization according to Lick and toxicity testing on at least one compound, PIB has been done (M5:23). Lucke noted that some of the compounds used as anti-misting agents are extensively used in cosmetics and personal care products, so the toxicity should be known (M5:23).

Schaper stated it is important to have some perspective on what it is we are putting in the workplace before we do, and this is one of the values of toxicology (M5:23). Schaper noted that health complaints in the work place may not always be tied strictly to the product itself but to what it has become (M5:23).

Kushner provided information on a questionnaire NIOSH used on a population of 174 people at a roller plant (6:32-33; Kushner, 1998). The primary reported symptoms were upper respiratory irritation such as throat dryness, nose irritation and watery eyes (Kushner, 1998). Comparing grinding and inspection, they reported that there was not any significant difference in symptoms (Kushner, 1998). Exposures were 1.6 to 2.6 mg/m3 for grinders and 0.3 mg/m3 in inspection (Kushner, 1998). Additional Information on other Non-Malignant Respiratory Effectsfrom the NIOSH Criteria Document

Sections through of the NIOSH Criteria Document discuss the diseases lipid pneumonia, hard metal disease and legionellosis (NIOSH, 1998). The relative rarity of lipid pneumonia may be due to current lower concentrations of MWF aerosols as compared to past exposures (NIOSH, 1998). Hard metal disease can develop quickly and is associated with the grinding of hard metal parts such as cutting tools (NIOSH, 1998). These machined parts contain cobalt and/or tungsten carbide (NIOSH, 1998). Species of the Legionella genus have been isolated from MWF reservoirs and one form was associated with an outbreak of Pontiac fever, a self limited form of legionellosis (NIOSH, 1998). Lipid pneumonia, hard metal disease and legionellosis appear to be relatively unusual in MWF environments, according to NIOSH (1998). Other Resources

Committee member, Dr. David Wegman, provided the committee a summary of many variables associated with non-malignant respiratory disease, in the MWF studies (Wegman, 1998). The summary consists of tables which define the study, design, population, fluid class, aerosol exposure concentration, health effect, #cases/#exposed, and risk estimate (Wegman, 1998). See the section on asthma and airway problems in this chapter. Some items from that section and this one overlap, such as Reeve's work and Burch's. The two MWF Symposia Proceedings provide additional information and discussion (AAMA, 1996,1998). Additional references are cited in Chapter Eight, Medical Surveillance and are also found in Attachment #6.

2.3.6 Concerns and Limitations Size of Business

Burch and Cox were concerned that the studies did not reflect all MWF business segments (M5:31; M6:21; M6:22). Perry explained that this would be a problem if there was evidence that the risks are different in different sectors, instead of with different exposures (M6:21). Cox noted that the operating profile of the machines used in auto plants do not fit the profile of a small company (M6:22). Cox cited differences such as the number of shifts, constant running of high misting machines in big business while a small business may run a dirty machine once per month (M6:22).

Combining the data from PMPA and PMA, results in a risk of 1 worker in 45,928 may develop a MWF induced respiratory illness (Nakoneczny, 1998). Nakoneczny stated that only minimal incidence of respiratory illness occurs (Nakoneczny,1998). Smoking

Howell cited Ameille's, Robin's and Greave's studies as evidence for smoking as an issue (M2:11). Howell was concerned that machinists who smoke may be more at risk (M2:11). Teitelbaum warned against blaming the smoker (M2:11).

Fine believed that smoking had at least an additive effect with MWFs and other respiratory toxins (M2:2). All four lung function studies showed a significant interaction between smoking and exposure, according to Fine (M2:7).

Rosenman noted that smoking is reported in his SENSOR questionnaire and any interpretations he makes are based on categorizing workers as current, ex, or non smokers (M4:7). Fennelly explained that the temporal variability in the asthma case he described could not be related to smoking (M5:3). Reeve noted that he could not link Ford's hospital admission dataset to smoking but wanted to investigate this issue (M7:28).

Although he saw more dramatic effects with obstructed smokers, Robins did not think there was good evidence that smoking was related to cross shift changes in pulmonary function (M5:10). Wegman acknowledged that smoking could play a role (M8:17). He noted a study underway that is reassessing Kennedy's 1989 data correcting for smoking and that Robin's conclusions about obstruction may not be borne out (M8:17). The Health Work group viewed that smoking should be banned in MWF plants (M8:26). Sherman recommended differentiating between smoking and smoking where MWFs are present (M8:27). Healthy Worker Effect

The healthy worker or survivor effect was debated. Eisen explained that in her study, for each of her 29 post hire asthmatics, she selected four individuals of the same age and race who did not have asthma, but were exposed to the same conditions as the asthmatic up to the time of diagnosis (M5:2). She tracked all of these workers and found that more asthmatics had transferred to assembly work by the time Greave's study was conducted (M5:2). She followed these workers until 1994 and found that 66% of the asthmatic machinists had terminated their employment compared to 40% termination in the rest of the population (M5:2). She noted that the 1994 study was not corrected for date of hire (M5:2).

Lick did not believe this phenomenon occurred in the auto industry (M2:9). He noted that machinists have the highest paying jobs and transfers would tend to be to machining, not away (M2:9). He also explained the minimal turnover in the auto industry (M2:9). Mirer stated that people leaving machining departments because of health effects has to be considered despite the higher pay, opportunity for overtime and less strenuous work involved in machining (M6:4). Respiratory Disease Diagnosis

Diagnosis of respiratory disease and the relation of this disease with occupational cause were concerns. Newman explained that asthma can be diagnosed using lung function testing and more involved methacholine challenge tests (M2:5). Bronchial responsiveness test and peak flow meter can be used in the workplace to monitor asthma, according to Fennelly (M5:2). Fennelly stated that challenge tests to determine a specific agent are complex and the usefulness of questionnaires is limited (M5:2).

HP is clinically diagnosed, according to Newman, using lung biopsy supplemented by immunological tests (M2:5). Rose was very concerned that medical schools do not provide adequate preparation of physicians to diagnose HP (M5:5). Newman noted that HP is often confused with sarcoidosis and other granulomatous disease (M6:3). Rose noted the poor predictive value of spirometry, chest X-rays and job designation in detecting HP (M5:5). Rose opined that questionnaires could be helpful in finding sentinel HP cases (M5:5). Rose explained that viable microbial sampling is inadequate since the antigens causing HP may be nonviable (M5:5). Rosenman indicated that questionnaires may pick up people that spirometry doesn't and spirometry may pick up what questionnaires do not (M4:8) Symptoms vs. Disease

Another concern addressed was the relevance of symptoms vs. disease. According to Fine, NIOSH looks at symptoms as surrogates of serious health problems because symptoms can represent conditions that are truly impairing and disabling (M2:6). Burch stated that symptoms are not semantically or clinically equal to having a disease (M5:15).

Robins noted that obstruction may be a predictor symptom that an individual would have a problem in a MWF environment (M5:10). He warned that his study had workers with 19 years seniority and obstruction may not be present, nor a predictor in young workers but should be monitored (M5:10).

Hoffman noted that cross-sectional studies cannot address the issue about the long-term effects of people who exhibit short-term decrements in FEV1.0 or who exhibit prevalent symptoms of cough, of phlegm, over a long period of time (M5:13). Hoffman questioned what does transient FEV1.0 decrement and/or a positive methacholine challenge test mean in the long term (M5:13).

White agreed with Burch and Hoffman questioning the meaning of a 5% decrement in lung function and presence of certain symptoms (M5:31). Lick questioned what amount of lung function decrement was significant (M6:22). Newman noted that a 5% lower mean FEV1.0 in a population shows that a sizable number in that population have much larger drops reflective of aggravation of asthma at work (M6:22). The 19% decline seen in some of Robin's workers is a dramatic decline according to Newman (M6:22). Newman explained that cross shift decrements are reflective of embedded cases of asthma and probably HP in these populations (M6:21).

Wegman explained that Becklake reviewed the association between acute and chronic respiratory changes (M5:13). According to Wegman, Becklake's evidence supports the fact that a short-term change in FEV1.0 is predictive of long-term change in several different occupational settings (M5:13; M6:21). Peak Exposures

Another issue discussed was peak values. Sheehan was concerned that peak exposures and work patterns may be different in different sites (M6:22). Lick opined that toxicity data on peak values would be helpful (M6:22). O'Brien noted that peaks may be related to upper airway disease (M7:15). Wegman explained that there are hypotheses that peak exposures may be associated with the onset of reactive airway dysfunction or acute irritant asthma (M8:26). Mirer noted that most of the evidence is based on clinical impression (M8:26). Other Limitations of Studies

Other limitations of studies were discussed. Howell was concerned about the common ventilation systems in areas defined as exposed and unexposed in the respiratory studies (M2:11). Rosenman explained that carryover could occur in the MiOSHA plants (M4:8).

Greave's study was a cross sectional study according to Wegman, and these studies suffer from survivor effect (M2:8). Howell was concerned about the limitations of the odds ratios in Greaves study (M2:11).

Howell noted misclassifications of fluid types (M2:11; M6:21). Howell expressed his concern about Eisen's study misclassifying the synthetics (M5:14,31). Howell suggested that if one is combining classes to combine semi-synthetics with soluble oils because the chemistry is more similar than semi-synthetics with synthetics (M5:14). Howell noted that the effect seen in Eisen's work might disappear with reclassification of fluids (M5:14). Howell recommended that OSHA use the raw data from studies to avoid such limitations (M6:21).

In Hoffman's analysis for ILMA of the studies cited in the NIOSH Criteria Document, he reviewed about 300 papers (M5:13). Hoffman questioned the use of different exposure assessment techniques and that long-term exposure versus short-term exposure limits need to be considered (M5:13). He also questioned the use of total particulate as a surrogate (M5:13). Hoffman noted that Sama used total sulfur concentrations and Robins looked at endotoxins and bacteria (M5:13). Robins found the bacteria exposures were quite substantial in terms of the relative risks (M5:13). Hoffman criticized Krzesniak's study for the wide exposure range (M5:11). Teitelbaum explained that Krzesniak's study was not peer-reviewed but was a report of a conference (M5:12). Hoffman explained that more work needs to be done on exposure assessment (M5:13).

Hoffman viewed the UAW/GM studies of Kennedy, Greaves and Eisen as having the greatest weight due to the size of the populations, the quality assurance in the exposure assessment and the assessment of health outcomes (M5:12). Hoffman was concerned about participation bias in Kennedy's study (M5:12)

Hoffman warned that relative risk values of 1.3 or 1.5 could be easily confounded (M5:11). Wegman stressed the need to look at not only rates but the confidence interval, explaining that wide confidence intervals indicate Iow statistical power and Iow precision (M2:8). Hoffman evaluated the studies in the context of Hill's criteria for causality and noted that there was a temporality of the relationship between exposure and response, but questioned whether there was constantly strong relative risks, consistent across studies (M5:14). Schaper noted that her work on mice provides the basis of biological plausibility and causality for the effects of MWFs (M5:18).

Burch thought Reeve's numbers overestimated the effect of MWFs based on Reeve's inclusion of any case, even if later it was determined by workman's compensation as non-occupational (M6:4). Newman viewed Reeve's estimates as lower than the real number, noting that Reeve did not include all private physician data and could have misclassified cases (M6:7). Shortell thought Reeve's numbers were an under-reporting because in his experience, many workers avoid going to the medical department (M6:5). Shortell stated workers avoid medical due to their lack of knowledge about the occupational basis for their disease and workers view medical as an unpleasant place (M6:5). Reeve viewed the symptom prevalence rate somewhere between the passive and active numbers (M6:4). Lick acknowledged that the Ford database needs improvement but stated that it did show a rationale for controlling MWFs (M6:5).

2.3.7 Linkage of Discussions to OSHA Action

Speakers and committee members linked studies and concepts to potential action by OSHA. They addressed issues such as an exposure limit, the weight of the evidence, health endpoints, risk assessment, relevant studies, material impairment of health and whether an exposure limit would be effective.

Fine explained that NIOSH based its criteria for the REL primarily on the respiratory effects (M2:1). Fine explained that studies by Kreibel, Robbins and Kennedy indicated adverse effects below the REL (M2:2). Rosenman noted that symptoms are reported at levels below the REL (M4:7). Eisen showed asthma in populations exposed to 0.6 mg/m3 according to Fine (M2:6). NIOSH could not find evidence that a limit for endotoxin or bacterial counts would be protective (M2:2).

Mirer stated that the NIOSH Criteria Document presented a very strong body of evidence and compared to most OSHA rule-making, an overwhelming body of research studies and case reports at levels below the REL (M2:10). He noted that NIOSH did do quantitative assessment of statistically significant studies (M2:10).

Hoffman agreed with Mirer that the general approach of public health authorities is to say that there may still be something going on at the no effect level (M5:14). According to Howell and Mirer this level furnishes no assurance of safety because you cannot account for sub-clinical health effects and the limits of statistical power (M5:14).

Teitelbaum explained that from Ely in 1970, articles have shown an increasing awareness and a greater precision in the diagnostic categories of illness which are found in persons exposed to metalworking fluids (M5:14). Teitelbaum noted that there are papers from the 1930's which demonstrate pulmonary disorders associated with metalworking fluids, and after 60 or 70 years of a growing literature, it is clear that there is respiratory disease associated with the exposure (M5:14). Teitelbaum explained that we have a literature which demonstrates consistency and biological plausibility (M5:14). Teitelbaum stated that there is disease and that it is occurring at the level which is currently achievable, which is one-tenth of the level which is permitted under the existing OSHA standard, so we need to respond to this (M5:14).

Infante asked for guidance about endpoints for material impairment (M5:31). Infante questioned if cumulative dose using current exposures should be used (M5:31). Wegman noted lung function changes and other factors would have to be considered but that morbidity studies are more difficult to assess than mortality ones (M5:31).

White agreed that respiratory endpoints could be used, but questioned which ones, and what relevance they have (M6:22). He questioned if the studies used (Kennedy, Greaves, Robins) by Perry were appropriate for risk assessment (M6:20).

Wegman questioned why Kriebel and Eisen studies were not included in Perry's assessment (M6:21). Perry noted that an outside contractor viewed that these were not amenable for this assessment (M6:21). Wegman noted the important papers to review include: Sprince, Greaves, Eisen, Kriebel and Robins for asthma and COPD (M5:31). Anderson's upcoming paper could be used for HP (M5:31). Wegman viewed asthma and HP as first and second priorities and COPD fourth with dermatitis third (M5:31).

Mirer recommended that MiOSHA data be included in OSHA's assessment since it included 50-70 different locations and Rosenman related asthma and symptoms (M6:21). Lick disagreed, stating that this dataset was not on par with the other studies (M6:21). Wegman thought that Rosenman's work could help interpret the risk assessment (M6:22).

Mirer noted that Ford's passive medical surveillance data presented by Reeve showed a significant increase in respiratory complaints to medical in MWF plants (M6:2). Mirer viewed Reeve's estimate as Iow compared to the actual workers exposed since according to Mirer only 1/3 of the workers in a MWF plant would be in machining areas (M6:4). Mirer highlighted active surveillance studies including Anderson's report of HP at the Chrysler Kenosha plant, the Sprince study at Chrysler's Kokomo plant, the 3 GM studies of Greaves, Kennedy and Eisen, Robins study of one GM plant, and Rosenman's surveillance of multiple small and large facilities (M6:4). Mirer noted that 20-25% had substantial respiratory symptoms and occupational asthma rates were 15-30% higher in machining departments (M6:4). Symptom prevalence rates were 15% in active surveys and 0.5 to 1% in passive reporting, according to Mirer (M6:4). Mirer used Reeve's number of 0.02 cases of HP/200,000 hours worked multiplied by a 30 year working life and calculated a rate of 6 cases of HP per 1000 workers (M6:4). Mirer viewed this as evidence of a material impairment of health (M6:4).

Lick stated that respiratory effects are harder than cancer to disregard (M5:15). O'Brien also questioned if straight oils represent the same respiratory hazard that water solubles do (M5:16).

Howell noted that the diseases, chronic bronchitis and asthma, might represent a material health impairment (M2:11). Howell agreed with Burch that he cannot conclude that exposure to metalworking fluids, new or used, caused non-carcinogenic respiratory disease (M5:16). Howell noted that metal removal fluid exposures have declined significantly over the last 25 years and acknowledged there may be consensus that at least long-term exposure to some in use metal removal fluids can be associated with some non-carcinogenic respiratory disease in some individuals (M5:16).

Howell explained that either the absence of good fluid management and/or the presence of microbial contaminants of one form or another are more likely to lead to acute and chronic non-cancer respiratory disease (M5:16). Howell noted that reducing exposure to less than or equal to 0.5 mg/m3 total particulate alone, absent any other program elements, is unlikely to further reduce the prevalence of non-cancer respiratory disease (M5:16). Howell questioned Perry's data, showing effects at the 0.1 mg/m3 level and believed something else may be causing problems at these levels (M6:23). Howell noted that product stewardship is a proper and necessary step (M5:16). Howell explained for employee health, reduction of work illness and injury, that it is the management of fluids and the education of users at all levels that will yield the greatest payoff in the shortest time (M5:16).

2.3.8.Committee Decisions and Rationale

The majority (13) opinion of the committee was that acute and chronic respiratory effects are known to be associated with exposure to MWFs (M9:2223). Members cited the epidemiological studies, the limited toxicology studies and their own experiences: in plants, in discussions with workers and in clinical practice (M9:22). Presentations by Rose, Fennelly, Eisen, Hodgson, Fennelly's patient, the NIOSH Criteria Document and papers by Kennedy were noted as additional evidence (M9:23). Data from Wegman (Wegman,1998), and Rosenman (1998) were cited (M9:23). One member, White, stated that there was some evidence to support the association of acute and chronic respiratory effects and MWFs (clarification provided in meeting #10). Another member, Cox indicated that there was no evidence in small plants, although there was in large ones (M9:22).

The minority opinion (Howell) of the committee was the evidence was equivocal. Concerns were expressed about the categorization of fluids, and other confounders in the studies. Risk ratios were close to one, making them vulnerable to confounders (M6:22). The relevance of some of the health endpoints was questioned (M5:6).

Two members (O'Brien, Sheehan) explained that the effects were associated with end-use fluids (M9:22). One member (McGee) noted that there was more evidence for acute effects than chronic effects (M9:22). One member (Burch) had no comment (M9:22). One member (Mirer) explained that HP is more associated with in-use water-based fluids and asthma is associated with all MWFs (M9:23). This same member (Mirer) viewed that material impairment of health related to respiratory problems is more associated with water-based fluids (M9:23).


2.4.1 Speakers

The issue of cancer was discussed at the fifth and seventh meetings and at other meetings. Dr. Boris Lusniak, NIOSH, addressed skin cancer in his presentation on dermatitis (M2:13). Committee member, Dr. John Howell mentioned cancer testing methods in his discussion of consensus standards (M4:4). Dr. Geoffrey Calvert, NIOSH, summarized the cancer studies noted in the NIOSH Criteria Document (M5:33). Committee member, Ken Kushner, noted a study he did of cancer rates in his company's plants (M6:33). Committee member, Dr. Daniel Teitelbaum presented an overview on cancer in the December, 1998 meeting (M7:6-8).

In the discussion of rates of adverse health effects, three types of data were presented: anecdotal or case reports, surveys of plant experience and formal cohort or cross-sectional studies. The first type provides only evidence that the problem exists in the setting from which the report comes and may exist in comparable settings. The second type is limited by the quality of the different reporting units (plants) and no effort has been made to determine that each was equally aggressive in identifying and recording adverse health effects. Generally these surveys were based on OSHA 200 logs which may or may not have been complete. These survey results, therefore, should be seen as offering a different type of information than case reports with less quantitative reliability than systematic scientific studies. These survey results are limited by the sources of data. The third type of data, formal studies, is the most reliable, although these types of studies have been carried out only to a limited extent in occupational environments using MWFs.

2.4.2 Background Information

Teitelbaum explained the old definition of cancer is a malignant tumor of potentially unlimited growth that expands locally by invasion and systemically by metastases (M7:6-7). Currently, according to Teitelbaum, cancer is viewed as a malady of genes and most if not all causes of cancer act by damaging genes directly or indirectly (M7:7). Scientists accept that cancer is clonal and is due to the progressive accumulations of mutations, interactions of different genetic alterations and complementation between different mutant genes (M7:7).

According to Teitelbaum, in 1990, the age adjusted cancer death rate in the US was 174 per 100,000 (M7:7). The cancer rate has increased steadily since 1930 but if tobacco related deaths are removed, the rate would not have increased for males and would have decreased for females (M7:7). According to Teitelbaum, 1 in 3 people in the US will get cancer and 1 in 4 will die from it (M7:7). Additional rates for specific types of cancers are in Teitelbaum's handout (Teitelbaum, 1998).

Teitelbaum explained if you have had one cancer, your chance of a second, new cancer is vastly increased due to mechanistic reasons (M7:7). Heredity, lifestyle, diet, viruses, bacteria, worms, radiation and workplace exposures are factors related to cancer (M7:7; Teitelbaum, 1998). Host factors include: age, gender, nutritional status, genetic makeup and presence of some infectious and genetic diseases (M7:7-8). Exposure differences and habits such as smoking can account for a wide variety of special risk groups who are more likely to develop cancer (M7:8).

According to Teitelbaum, a series of rare events in the right sequence and right time can lead to cancer (M7:7). The more carcinogens, the more cancer that is seen (M7:7). Many agents cause the same type of cancer and one carcinogen can be multi-potential, i.e. can cause different types of cancer (M7:8). He noted that simplistic approaches to understanding cancer do not work (M7:8). More details on the mechanisms of carcinogenesis are in his handout (Teitelbaum, 1998).

Workers enter a workplace with pre-existing histories and predispositions to cancer (M7:8). Removing carcinogens from the workplace lowers the likelihood of the interactions which will cause more cancers (M7:8).

Teitelbaum provided the designations of carcinogens of the International Agency for Research on Cancer (IARC) (M7:8). Group or Class 1 cause cancer in humans based on evidence in humans and arsenic is an example (M7:8). Group 2A probably causes cancer in humans while 2B possibly causes cancer in humans (M7:8). The two designation is based on animal evidence with limited human evidence and examples include formaldehyde and methylene chloride (M7:8). Group 3 is not classifiable usually due to inadequate evidence and Group 4 include agents for which a fair amount of research has been conducted and all results are negative (M7:8). He noted that the National Toxicology Program also has its own classifications (M7:8).

2.4.3 Review of Studies Presentations

Calvert reviewed the NIOSH Criteria Document section on cancer and his publication in the March,1998 issue of the American Journal of Industrial Medicine (M5:33). He explained that NIOSH identified six animal studies but that these were viewed as inadequate evidence for carcinogenicity of MWFs (M5:33). These animal studies were inconsistent and had inadequate characterization of the MWF fluid (M5:33).

Calvert explained that NIOSH focused on epidemiological studies in their review (M5:33). NIOSH viewed that the Tolbert and Eisen studies conducted in 3 auto plants and including over 23,000 workers, had the most statistical power (M5:33). Workers were studied from 1941 to 1984 (M5:33). The studies indicated an association between MWFs and laryngeal, rectal and pancreatic cancer (M5:33). There was a dose response for laryngeal cancer and it was significant for exposure to greater than 0.5 mg/m3 of straight fluids (M5:33).

NIOSH used Silverman's study as evidence for an association between bladder cancer and MWFs (M5:33). Silverman's study corrected for smoking and had more power than studies by Howe, Schiffler and Gonzales (M5:33). There is a relation between exposure to MWFs and bladder cancer, but the type of fluid responsible cannot be assessed, according to Calvert (M5:33).

Calvert noted that laryngeal, rectal and pancreatic cancer are primarily associated with exposure to straight fluids, and there is some evidence of association with synthetics (M5:33). He explained that scrotal and skin cancer are associated with PAHS in earlier formulations (M5:33). Lusniak noted that straight fluids were associated with skin cancer as well as other disorders (M2:13). Skin cancer was associated with polyaromatic hydrocarbons that are no longer in use, so according to Lusniak, skin cancer is rarely seen (M2:13).

Substantial evidence was found of an increased risk of cancer of the larynx, rectum, pancreas, skin, scrotum and bladder (Calvert, 1998). Studies could not assess a causative agent and changing formulations make interpretation difficult, stated Calvert (M5:33). The conclusions are all based on exposures prior to 1970 and there are not any studies that determine the risk of current exposures to currently used fluids (M5:33). Since exposures have been reduced, Calvert felt that the risk of cancer has been reduced (M5:33).

Howell noted the availability of an ASTM method for formulators and fluid component suppliers to use to assess components and fluids (M4:4). This standard is E-1687 Test Method for Determining the Carcinogenic Potential of Base Oils Used in MWFs (M4:4).

Kushner compared death from cancer rates in his company's bearing plants which use MWFs, to the company's steel plants, which do not use MWFs (M6:33). The rates of deaths from Gl cancer for machinists was 5.6% of the total deaths (M6:33). Kushner noted that this was a lower percentage than rates found in MWF cancer studies (M6:33). He also compared his company's rates to the statistics of the counties in which the plants were located (M6:33). He did not correct using standard mortality analysis (M6:33). Additional information can be found in his handout (Kushner, 1998). Wegman commented on the need for these data to be analyzed by proper epidemiologic methods (clarification at tenth meeting). Additional Information about Cancer from the NIOSH Criteria Document

Calvert explained the information provided in the NIOSH Criteria Document sections 5.2 and 5.3 and this information is in the previous section of this report (NIOSH, 1998). Tables 5-5 through 5-17 in the Criteria Document summarize the various types of studies on cancer (NIOSH, 1998). Table 5-18 addresses aerosol concentrations at different time periods (NIOSH, 1998). Additional Resources about Cancer

Hoffman provided a draft, unpublished document without references that addressed cancers studies and MWFs (Hoffman, 1998b). He highlighted studies on stomach cancer, pancreatic cancer, lung cancer, and laryngeal cancer (Hoffman, 1998b). He noted that the studies he reviewed had problems with small study populations and many were PMR studies which cannot evaluate causation (Hoffman, 1998b). No attempt was made in these studies to estimate individual exposures and analyze dose-response relationships (Hoffman, 1998b). No individual information was obtained by these studies on personal risk factors (Hoffman, 1998b). He cites other studies and provides limitations in this draft document (Hoffman, 1998b). In his report, Hoffman noted that the UAW/GM studies at three auto plants have been given the most weight due to sample size and resulting statistical power to identify rare cancers (Hoffman, 1998b). In the UAW/GM studies Hoffman explained that individual information on potential confounding factors was not obtained (Hoffman, 1998b).

Overall, Hoffman explained in his report that evidence has been presented for slight increased risks for certain types of cancer as a function of exposures to various classes of MWF (Hoffman, 1998b). It is possible that certain subgroups of workers may exhibit slight increased risks of laryngeal cancer. Due to the margin of the increased risk, Hoffman believed that the risk could be due to these confounders (Hoffman, 1998b). The patterns of exposure indicate that historical exposures play a more prominent role than more recent exposures (Hoffman, 1998b). Plant conditions, MWFs used and exposures prior to 1960 are very different from today (Hoffman, 1998b).

The UAW provided a packet of studies to the committee (UAW, 1997). Included in this packet are studies by Park (1996), Silverstein (1988) and a summary of these studies (UAW, 1997). Another article and analysis was provided by UAW, the article was Sullivan et al (1998). UAW provided a variety of articles noting the award of workman's compensation to the family of a worker exposed to MWFs who died from cancer (UAW, 1998).

Lucke provided an article he wrote for Lubrication Engineering, Health and Safety of MWFs, Fluid Formulations: A View into the Future (1996). This article addresses his analysis of the cancer studies on MWFs as well as other issues (Lucke, 1996).

In Cole's report for Caterpillar, he criticizes the draft version of the NIOSH Criteria Document (Cole, 1996). He was concerned about the use of causal interpretation of some studies, the exposure to different types of fluids confounding different studies and the emphasis on certain types of cancer (Cole, 1996).

The two MWF Symposia Proceedings provide additional information and discussion (AAMA, 1996,1998). Additional references are cited in Chapter Eight, Medical Surveillance and are also found in Attachment #6. The NIOSH Criteria Document has been cited throughout and is a comprehensive source of other articles (NIOSH, 1998).

2.4.4 Concerns and Limitations Size of Business

Throughout all health discussions, Burch and Cox were concerned about the relevance of studies done in large auto plants to small business. Other Issues

The committee was concerned about: lifestyle factors, the strength of the evidence, biological plausibility, formulations and consensus. Burch viewed that lifestyle was a very important factor in carcinogenesis (M5:33). Calvert agreed that it is very difficult to control for lifestyle factors in the cancer studies (M5:33).

Mirer viewed that the NIOSH Criteria Document underestimated the strength of the evidence for cancer (M7:9). He referenced page 96 of the Criteria Document on risk ratios (M7:9). NIOSH used agreement among studies and study size to determine the risk, according to Mirer (M7:9). Mirer commented that neither the National Toxicology Program nor IARC have very clear decision rules on how they aggregate epidemiologic studies (M7:8). According to Mirer, there are clear rules for animal carcinogens (M7:8).

Lick cited the inconsistencies among studies (M9:24). He noted no discernible pattern but the articles indicate something is going on even if there is not a good fit (M7:8). He noted NIOSH's interpretations such as rectal cancer associated with straight fluids (M7:10). The Ford Cleveland plant population showed evidence of rectal cancer in the studies but did not use straight fluids (M7:10). He questioned why lung cancer is not seen in the studies (M7:10).

Anderson noted that as an epidemiologist, he did not have a problem with the lack of agreement, noting it could be due to different populations and different plants (M7:9). Anderson cited some consistency in pancreatic cancer (M7:9). The dearth of animal studies bothered Anderson (M7:9). Anderson noted some limited skin painting studies resulted in skin cancer (M7:10).

Howell noted the difficulty of proving biological plausibility (M5:33). Calvert agreed but indicated that early synthetic and semi-synthetic fluids were associated in Wang's animal study with pancreatic cancer (M5:33). Calvert indicated that nitrosamines in animal studies show some biological plausibility (M5:33).

Anderson noted that the polynuclear aromatic hydrocarbons (PAH) animal studies show biological plausibility (M5:34). Mirer felt the human studies show and component studies show biological plausibility (M5:33-34; M7:9). Teitelbaum warned of the differences in biotransformation between animals and humans and the potential for multi-target carcinogens in the MWF mixture (M5:34).

Howell noted that formulations have changed over the years and ingredients known to cause health problems have been removed by fluid manufacturers (M7:9). Anderson cautioned against thinking all potential problems have been eliminated by formulation changes (M7:10). Mirer agreed that nitrosamines and other carcinogenic compounds like ethanolamine have been reduced, but trace amounts still exist (M6:38). McGee and Anderson were concemed that the fluids from the manufacturer are very different from what they become at the plant after contamination (M7:9,10). Kushner warned that PAHs are still a question and that manufacturing processes can increase or decrease certain aromatic compounds (M6:38). Howell explained that some in process cleaners can contain nitrites and any uses of secondary alkanolamines could result in nitrosamines in the used fluids (M7:9).

There was an overall impression from the committee, and especially Howell and Mirer, that disagreement would continue during any cancer debate (M6:40; M7:9). Mirer persuaded the committee not to debate the carcinogenic potential of current fluids since we cannot really know the cancer effect of a currently used fluid for 20-25 years (M6:38). Howell stated that disagreeing on cancer did not preclude or prevent the committee from moving forward (M7:9)

2.4.5 Linkage of Discussion to OSHA Action

The committee addressed issues such as what class of carcinogen MWFs would be in, acknowledging that these discussions were to help determine the weight of the evidence. The committee did not view its role was to determine an actual class for MWFs. How the cancer issue related to other health issues and potential action were discussed.

Teitelbaum indicated that he thought there was enough information to classify MWFs as IARC type 2B or maybe 2A (M7:8). Teitelbaum suggested reviewing Calvert's paper and that this type of paper would be reviewed by IARC in classifying MWFs (M7:8). In response to Burch's question about which MWFs would be classified, Teitelbaum noted that this is a difficult question to answer (M7:8). Ethanolamines are known carcinogens stated Teitelbaum (M7:8). The old straight oils with PAHs would be classified in Class 1 while nitrosamine containing ones could be in 2A or 2B, according to Teitelbaum (M7:8). Teitelbaum explained that the multiple exposures many workers have had makes knowing what they were exposed to a difficult determination (M7:8). Teitelbaum cautioned against waiting for complete information prior to action (M7:8),

Lick commented that if MWFs were Class 1 or even Class 2 carcinogens, he would see more consistent cancers from plant to plant (M7:8). Cancer was not the defining disease for Lick (M7:8). Lick noted that HP and other nonmalignant respiratory diseases are quite significant to act on, but all actions need a systems approach (M7:10). A PEL, the usual way to address carcinogens, would be inadequate, according to Lick (M7:10).

Teitelbaum agreed with Lick that air exposures are not the only concern (M7:10). He recommended getting the workers hands out of the fluids (M7:10).

Hoffman viewed the complexity of MWFs made it very difficult to assess

potential IARC categories (M7:9). He did not think they could be put into one category (M7:9).

Based on the studies of earlier formulations and the precautionary principle, Anderson would categorize MWFs as carcinogens Class 1 or 2A, until proven otherwise, because there was sufficient evidence in humans (M7:9,11). He thought preventing dermatitis may reduce the risk of skin cancer (M7:10). He did not think a standard should be based on carcinogenicity (M7:10).

McGee stated that we know people exposed to MWFs 30-40 years ago have a higher incidence of cancer (M7:10). Mirer noted that the bulk of recent evidence is for soluble, synthetic and semi-synthetic fluids (M7:8). Mirer and McGee were concerned that too many carcinogens are in dispute too long before action is taken (M7:9). Mirer thought there was enough evidence to support MWFs as a human carcinogen but we can protect people without this conclusion (M7:11).

Wegman noted that cancer is a secondary reason, after respiratory disease and dermatitis, to control MWFs (M7:10). He noted that additional cancer studies are on-going (M7:10).

Howell explained that it was impossible to prove today's formulations will not cause cancer under all of the circumstances of use (M7:11). Howell explained that the significant reduction in exposure coupled with a significant reduction in impurities that have been associated with cancer has reduced and will continue to reduce any future possible carcinogenic risk (M7:9). He cited the role of product stewardship by fluid manufacturers in reviewing new health studies and making recommendations for members to remove components of concern (M7:10).

2.4.6. Committee Decisions and Rationale

The committee addressed skin cancer and cancer at other sites as separate issues. Skin cancer was addressed first. The opinions were separated into evaluating "old formulations" versus "current formulations".

The majority (10) opinion was that skin cancer is known to be associated with exposure to old formulations of MWFs (M9:23). The opinions were mixed for current formulations of MWFs (M9:23). White believed that old formulations were a problem, two members (Lick, Teitelbaum) believed there was no evidence for current formulations, three members (Sheehan, Mirer, Frederick) viewed evidence for current fluids as equivocal and one (Anderson) thought it was reasonably anticipated that there was evidence for current fluids (M9:23). Three members (Wegman, Newman, Day) believed there was known evidence for old and current formulations (M9:23).

Chapter Five of the NIOSH Criteria Document was cited (M9:23). An alternate who is a machinist noted his own experience with squamous cell cancer (M9:23). Issues such as the difficulty of assessing the effects of current exposure due to the latency period and the possible presence of co-carcinogens and promoters were noted (M9:23).

The minority (Burch, Cox, Howell) opinion was that the evidence was equivocal for old formulations (Mg: 23). As noted above, the opinions for current fluids were mixed (M9:23). The members who presented the minority view on the older formulations believed there was no evidence for current formulations (M9:23).

Two members (O'Brien, White) did not think they had adequate information to make a decision (M9:23).

The committee addressed the issue of cancer at other sites. The majority (10) opinion was that old formulations of MWFs are known to cause cancer at various sites (M9:24). Epidemiological studies, MSDSs, and the NIOSH Criteria Document were cited (M9:24).

The minority opinion (Burch, Howell, Lick) was that the information on the older formulations was equivocal (M9:24). The inconsistencies among the epidemiological studies regarding sites were noted for a rationale (M9:24).

Two members (Cox, White) had no opinion (M9:24).

The committee was split on the issue of cancer related to current formulations of MWFs (M9:24). Four members (O'Brien, Lick,Teitelbaum, Frederick) viewed that evidence was equivocal for current formulations (M9:24). Four members (Day, Newman, Sheehan, Anderson) viewed the evidence as reasonably anticipating cancer associated with current fluids (M9:24). Three members (Howell, Cox, Burch) thought there was no evidence that currently formulated MWFs cause cancer (M9:24). Three (Wegman, Mirer, McGee) noted that prudence dictates that we view current formulations as carcinogenic, and one (White) had no opinion (M9:24). Latency periods, and reductions in nitrosamines and PAHs were noted as a rationale and concern (M9:23).