Background: In June 2003, NIOSH received a Health Hazard Evaluation (HHE) request from employees at three cemented tungsten carbide manufacturing facilities in Alabama to investigate the risk of hard metal lung disease, asthma, and bronchitis among workers. The three facilities are owned by Metalworking Products and are located within 30 miles of each other in Huntsville, Gurley, and Grant, Alabama. Prior to the HHE request, two former workers had been diagnosed with hard metal disease and one former worker had been diagnosed with occupational asthma. Exposures of concern included cobalt, tungsten carbide, nickel, and metalworking fluids. From September 2003 through May 2005, NIOSH investigators conducted several medical and environmental surveys to identify if particular exposures, production processes, and work practices were associated with increased occupational lung disease risk in these facilities. NIOSH investigators were particularly interested in analyzing relationships between cobalt exposures and respiratory health outcomes because of cobalt's known ability to cause hard metal lung disease and asthma. Assessment: NIOSH investigators conducted a walkthrough of all three facilities in July 2003, and conducted three medical surveys in September 2003, January/February 2005, and April/May 2005. These surveys included: standardized questionnaires; chest x-rays; lung function tests; measurement of metals levels in blood (cobalt, tungsten, nickel and chromium), urine (cobalt and tungsten), and exhaled breath condensate (cobalt, tungsten, and nickel); measurement of levels of biomarkers of inflammation and oxidative stress in exhaled breath condensate; and blood analyses for antibodies to metals and for variants of the gene HLA-DPbeta1, which has been associated with hard metal lung disease. NIOSH investigators conducted three environmental surveys in May 2004, October/November 2004, and November 2004. Environmental samples collected included: bulk metalworking fluid samples for cobalt, nickel, endotoxin, and microbial analyses; air and wipe samples (work area surfaces and workers' skin) for cobalt, tungsten, nickel and chromium levels; and air samples for analyses of particle shape, chemistry, and size. The different particle sizes included particles small enough to enter the nose, mouth, and throat (inhalable-size); particles small enough to enter the large air passageways or deep regions of the lungs (thoracic-size); and particles small enough to enter the deepest regions of the lung (respirable-size). NIOSH investigators also reviewed company air sampling results for total cobalt obtained from 1985 to 2003 and worker blood cobalt levels obtained from 1998 to 2003. Main Findings: 1. Exposures: Based on NIOSH air samples, the a. powder mixing area had the highest average (mean) personal breathing zone (PBZ) total cobalt concentration (574 micrograms per cubic meter of air [microg/m3]), the highest mean PBZ thoracic cobalt concentration (304 microg/m3), and the highest mean PBZ respirable cobalt concentration (78 microg/m3). Based on combined NIOSH and company air sampling results, mean total cobalt levels were highest in reprocessing (427 microg/m3) and powder mixing (414 microg/m3); air concentrations in powder mixing, screening, reprocessing, and blending exceeded the OSHA PEL of 100 microg/m3 for total cobalt; air concentrations in these areas and also in spray drying and reclamation exceeded the NIOSH REL of 50 microg/m3 for total cobalt. b. Wipe samples detected measurable levels of cobalt on all work surfaces across the three facilities. Results of skin wipe samples identified measurable levels of cobalt on all of the hands and most of the necks of participating workers. The two work areas where workers had accumulated the most cobalt on their necks over approximately four hours of work were metal separation and powder mixing. c. In-use metalworking fluid (MWF) contained low levels of cobalt, and no detectable nickel, culturable fungi, or mycobacteria; culturable bacteria and endotoxin were detected. (Recommended or regulatory limits for levels of endotoxin and microbial contaminants in metalworking fluids have not been established.) Levels of metalworking fluid in air in the grinding work area were below the NIOSH REL. (NIOSH recommends a MWF aerosol exposure limit of 0.4 mg/m3 (thoracic particulate mass) as a 10-hour time-weighted average (TWA). This corresponds to approximately 0.5 mg/m3 for total particulate mass). d. Blood cobalt measurements obtained by the company from 1998 through 2003 on workers in areas with potential high risk for cobalt exposure indicate that most workers had exposures above the NIOSH REL. NIOSH measurements of cobalt in blood and urine in 2005 showed that the average blood and/or urine cobalt levels in several work areas indicated exposures above the NIOSH REL. These areas included reclamation, powder mixing/reprocessing/blending (data from these three areas combined for analyses), milling, spray drying, pressing, and shaping/round cell (data from these two areas combined for analyses). 2. Medical findings: a. Survey participation: For the 2003 medical survey, 249 current workers with possible exposure to cobalt at the three facilities were invited to participate. Of these 249 workers, 171 (69%) participated in the 2003 medical survey; an additional 26 of the 249 workers participated for the first time in the 2005 survey, for a total participation rate of 79% in the original invited group. b. Respiratory symptoms and asthma: Data analyses after the 2003 survey showed that the numbers of workers reporting respiratory symptoms and physician-diagnosed asthma were approximately twice as high as expected when compared to national data. In workers who reported adult-onset (age 16 or older) physician-diagnosed asthma, the estimated post-hire asthma incidence rate was approximately two-and-a-half times higher than the pre-hire rate. Information from the questionnaire and medical test results from both the 2003 and 2005 surveys was used to identify workers with possible ("suspected") occupational asthma. The work areas with the highest estimated rates of suspected occupational asthma were milling/spray drying (data from these two areas combined for analyses), pressing, sintering, and product testing. c. Hard metal lung disease: One former worker had lung biopsy findings consistent with hard metal disease and two current workers had findings on lung function tests indicating possible hard metal disease. One worker was working in pressing, another in sintering, and the third in grinding at the time they developed symptoms or were found to have abnormal lung function. No current workers had chest x-ray findings indicating possible hard metal disease. 3. Relationships between exposures and health outcome data: a. Statistical models showed relationships between some exposure measures and some health outcomes. There was a statistically significant association between estimated cumulative total cobalt air concentration and reporting three or more asthma-like symptoms. The amounts of cobalt in the urine, on the wrist, and in exhaled breath condensate were also associated with reporting three or more asthma symptoms. In statistical models that included cobalt and tungsten levels, lung function on spirometry tests was negatively correlated with current and estimated cumulative respirable cobalt air concentration (i.e., lung function declined as exposures increased). b. Three of the areas with the highest rates of suspected occupational asthma had mean total cobalt exposures that were below the NIOSH REL. The former worker with hard metal lung disease and both current workers with possible hard metal lung disease had worked in areas where air sampling results showed exposures below the NIOSH REL, though urine and blood cobalt levels in one of these areas (pressing) indicated exposures above the REL. 4. Findings from other analyses: a. Genetic analyses: There was no association found between the genetic allele HLA-DPbeta1E69 and post-hire physician-diagnosed asthma. The low number of cases of suspected hard metal lung disease did not permit analyses of an association between this outcome and HLA-DPbeta1E69. Antibody analyses: b. There were no statistically significant differences in the levels of total immunoglobulin E among workers with pre-hire physician-diagnosed asthma, workers with suspected occupational asthma, and other workers. Of 140 workers tested, none had immunoglobulin G specific to cobalt, nickel, or chromium. c. Exhaled breath condensate (EBC) analyses: Statistical models showed a weak correlation between levels of cobalt in EBC and total cobalt air concentrations. EBC cobalt levels were moderately correlated with urine and blood cobalt levels. EBC levels of malondialdehyde (MDA), a biomarker of lung oxidative stress, were weakly correlated with total cobalt air concentrations but were not correlated with EBC cobalt or tungsten. Levels of LTB-4 and IL-8 (biomarkers of inflammation) and MDA did not differ significantly between "current asthma" cases (workers who reported currently-active physician-diagnosed asthma or who had airways hyperresponsiveness on methacholine challenge testing with a PC20 of less than or equal to 4 mg/m) and "healthy" workers (workers who did not have an asthma history, respiratory symptoms, or findings suggestive of asthma on NIOSH lung function tests). Conclusions: Among workers at three cemented tungsten carbide facilities owned by Metalworking Products in Alabama, a former worker had evidence of hard metal disease on lung biopsy and two current workers had findings on lung function tests that suggested possible hard metal disease. Cobalt exposures in two of the three areas where these workers worked were below the NIOSH REL. Because some workers may still be at risk for hard metal disease even when exposures are below recommended and regulatory limits, it is important to regularly monitor the health of all exposed workers to identify those who may be potentially affected. Several work areas in these facilities had cobalt exposures that exceeded the NIOSH REL and sometimes also the OSHA PEL, indicating a need for additional exposure controls in these areas. Our finding of an elevated rate of post-hire asthma compared to pre-hire asthma strongly suggests that exposures in these facilities at the time of the medical survey were putting workers at risk for asthma. Some of our analyses showed associations and correlations between different measures of cobalt exposure and respiratory symptoms and lung function. Studies conducted in other cemented tungsten carbide plants in other countries have utilized controlled cobalt inhalation in a laboratory setting to identify workers who had developed asthma due to cobalt exposure; antibody analyses indicated that some of those workers had developed cobalt asthma due to an allergic mechanism. We did not utilize the controlled cobalt inhalation approach to determine if any workers had asthma due to cobalt exposure, and our antibody analyses did not identify evidence of an allergic response to cobalt in any Metalworking Products workers. Therefore, while it is possible that some workers we identified with suspected occupational asthma may have asthma due to cobalt exposures, it is also possible that, in some Metalworking Products workers, respiratory symptoms or asthma may be due to other workplace exposures.