CARBON DISULFIDE

OSHA comments from the January 19, 1989 Final Rule on Air Contaminants Project extracted from 54FR2332 et. seq. This rule was remanded by the U.S. Circuit Court of Appeals and the limits are not currently in force.

CAS: 75-15-0; Chemical Formula: CS₂

OSHA’s former limits for carbon disulfide were 20 ppm as an 8-hour TWA, a 30-minute STEL of 30 ppm, and a ceiling limit of 100 ppm that was never to be exceeded. OSHA proposed revising these limits to 1 ppm as an 8-hour TWA and 10 ppm as a 15-minute STEL, and NIOSH (Ex. 8-47, Table N1) supported these proposed limits. OSHA has evaluated all of the evidence and testimony presented in the record and has determined that a 4-ppm 8-hour TWA limit, a 12-ppm STEL, and a skin notation are necessary to reduce the risk of cardiovascular disease and reproductive effects among carbon-disulfide-exposed workers, and the Agency is establishing these limits for carbon disulfide in the final rule. The need for a lower limit is based on evidence that exposure to carbon disulfide presents risks of cardiovascular, fetotoxic, and neurological material impairment of health.

OSHA’s decision to promulgate a 4-ppm limit rather than the proposed 1-ppm limit is principally based on the feasibility evidence available to OSHA (see Section VII, Technological Feasibility and Economic Impact Assessment). A skin notation has been added because there is evidence that carbon disulfide can cause systemic toxicity via the dermal route. Carbon disulfide is a clear, colorless, or faintly yellow liquid with a strong, disagreeable odor.

OSHA’s proposal to reduce the limits for carbon disulfide was based on a number of human studies reviewed by the ACGIH (1986/Ex. 1-3) and NIOSH (1977b/ Ex. 1-260) that suggested that exposure to carbon disulfide levels between 10 and 40 ppm was associated with an excess risk of coronary heart disease and of adverse neurological effects. These reports comprise a series of studies carried out on carbon-disulfide-exposed workers in Great Britain (Tiller, Schilling, and Morris l968/Ex. 1-92) and Finland (Seppalainen and Tolonen 1974/Ex. 1-100; Tolonen et al. 1975/Ex. 1-392; Tolonen, Nurminen, and Hernberg 1979/ Ex. 1-158). The British cohort was recently followed up by Sweetnam et al. (1987), and the Finnish workers have been followed up by Nurminen and Hernberg (1985).

The study by Tiller et al. (1968/Ex. 1-92) of British rayon workers was the first to relate exposure to carbon disulfide to the development of coronary heart disease. These authors found that, among men employed for more than 10 years in the rayon industry and followed from 1950 to 1964, those exposed to carbon disulfide had death rates from coronary heart disease more than twice the rate in other rayon workers. Thus, the Tiller et al. (1968/Ex. 1-92) study demonstrated that 10 years or more of exposure to carbon disulfide was associated with a significantly elevated risk of coronary disease.

The United Kingdom’s threshold limit value for carbon disulfide, which had been 20 ppm in the 1960s, was subsequently reduced to 10 ppm in the 1970s. To examine the effect of this reduced limit on occupational risk, Sweetnam et al. (1987) conducted a follow-up study on the cohort first described by Tiller et al. (1968/Ex. 1-92). The health status and cause of death for 2,848 members of this cohort were ascertained up to the end of 1982. Exposure scores representing cumulative exposure to carbon disulfide were developed for each cohort member, based on an analysis of personal and area sampling results, job category, and time spent in each job category. Sweetnam et al. (1987) found the pattern of mortality similar to that found by Tiller, Schilling, and Morris (1968/Ex. 1-92): among spinner operators, who had the highest CS(2) exposures of any job category, 73 deaths from ischemic heart disease (IHD) were identified, compared with 42.5 expected deaths (SMR = 172), a finding that was statistically significant. A statistically significant trend was found between cumulative exposure since first exposure and incidence of IHD mortality, which indicates a dose-related effect. A second (and perhaps most important) finding of this study was that recent (or current) exposure to carbon disulfide, as well as total cumulative exposure, were both risk factors for IHD. The authors established this association by examining the relationship between IHD mortality risk and each worker’s total CS(2) exposure in the two years preceding death or the end of the study. The third result of this study was that workers with current CS(2) exposure also had significantly higher risk than workers who had ceased exposure. The dose-related relationship between increased IHD mortality risk and recent exposure to carbon disulfide suggested to the authors of this study that the effect of carbon disulfide on the cardiovascular system was direct and reversible.

Thus, the Sweetnam et al. (1987) follow-up determined that there is a relationship between the risk of IHD mortality and increased cumulative exposure to CS(2). Among workers who terminated exposure, this risk declined to non-statistically-significant levels after one year of no exposure. However, risk continued to be elevated among workers who continued to be exposed or who had not been exposure-free for a full year. OSHA interprets the findings of this important study to indicate that cumulative CS(2) dose from time of first exposure is a risk factor for IHD, and that this elevated risk continues unless exposure is terminated. That is, OSHA finds that workers who have been exposed to CS(2) in the past continue to be at increased risk as long as they are exposed to CS(2), even when their recent exposure is to lower levels (approximately 10 ppm, the current U.K. TLV).

This finding was confirmed by Nurminen and Hernberg (1985) in their follow-up study of 343 Finnish rayon workers who had been exposed to carbon disulfide for at least five years. Health status data were obtained for these workers for the period 1967 to 1982. In 1972, a preventive program had been instituted that included establishing a 10-ppm exposure limit and removing workers at high risk of coronary disease from continued exposure to carbon disulfide. Median exposure levels (largely from area samples) for the period 1975 to 1980 did not exceed 5 to 6 ppm, and third-quartile exposure levels did not exceed 10 ppm. These levels were about half those reported for the period 1967 to 1975.

Nurminen and Hernberg (1985) reported a 4.7-fold increase in IHD mortality incidence for the period 1967 to 1972, prior to the establishment of the protective measures described above. Five years after these measures were instituted, only 19 percent of the cohort continued to be exposed to carbon disulfide (compared to 53 percent of the cohort exposed in 1972). The relative risk for the first seven years of follow-up (1967 to 1974) was 3.2, compared to a relative risk of 1.0 for the last eight years (1974 to 1982). The excess risk of IHD mortality thus declined steadily throughout the follow-up period; this trend was statistically significant. The authors concluded that “…the cardiotoxic effects of CS(2) are reversible in the sense that the cessation of, or a radical decrease in, exposure reduces the risk of cardiovascular mortality to background levels” (Nurminen and Hernberg 1985, p. 34). Thus, the Nurminen and Hernberg (1985) study shows that reducing exposure levels below 10 ppm (combined, in their case, with a rigorous medical removal program to terminate exposure for employees who had developed signs or symptoms of coronary heart disease) can reduce the significant risk of IHD mortality to baseline levels.

In addition to NIOSH (Exs. 8-47 and 193), the AFL-CIO (Ex. 194) and Dr. James Melius, Director of the Division of Occupational Health and Environmental Epidemiology of the New York Department of Health (Ex. 152), supported OSHA’s proposed 1-ppm PEL for carbon disulfide. However, several rulemaking participants criticized the studies relied on by OSHA, primarily on the grounds that the cohorts in which excess deaths from cardiovascular disease had been seen included workers who, these participants argued, were exposed for many years to levels of carbon disulfide much higher than the 10- to 40-ppm levels generally reported in these studies (Exs. 3-747, 3-1158, 8-19, 8-45, 31, 125, and 174; Tr. pp. 4-74 to 4-107). For example, Dr. Ernest Dixon, a toxicology consultant for the Inter-Industry Committee on Carbon Disulfide, testified as follows on these studies, which were also relied on by NIOSH to determine NIOSH’s recommended standard:

• The NIOSH document presents a recitation of the toxic reviews, neurotoxic effects, and the various cardiovascular studies from chiefly Scandinavia, largely epidemiologic studies which attempted to determine whether or not ischemic or other cardiovascular abnormalities caused an excess of deaths among workers exposed to elevated levels of CS(2). Essentially, all of these were from the viscose manufacturing industry. Air sampling for carbon disulfide in the period prior to a decade ago was cumbersome, costly, and took a long time for chemical analysis. As cited in numerous other reports, the practices of that period were to obtain area rather than personal samples. Work practices examined in the studies were such that the area sample results relied upon are believed to have significantly underestimated [both] the actual exposures and [the fact] that there were substantially higher exposures than have existed in more recent years. Accordingly, many of the workers in such studies had encountered many years of greatly higher exposure, especially for the earliest period of their exposure (Tr. p. 4-77).

In discussing the Tiller, Schilling, and Morris (1968/Ex. 1-92) study, Dr. Dixon emphasized that the coronary mortality risk of viscose production workers was not reported in this study to have been elevated, despite the fact that 17 percent of samples taken in production areas exceeded 20 ppm. However, there was a substantial excess in mortality from cardiovascular disease among spinners, where 50 percent of area samples exceeded 20 ppm (Tr. p. 4-80). In addition, Dr. Dixon pointed out that the populations studied by Vigliani (1954/Ex. 1-103) and by Seppalainen and Tolonen (1974/Ex. 1-100) were likely to have been exposed during high-viscose-production periods at the time of World War II, when exposures were higher than in later periods.

As discussed above, OSHA believes that both cumulative exposure and current exposure are risk factors for IHD among CS(2)-exposed workers; the Agency has also determined that excess risk continues for exposed workers as long as exposure continues. As to Dr. Dixon’s point about area samples, OSHA does not agree that it is possible to infer that earlier area samples underestimate exposures. It is common industrial hygiene practice to measure problem areas in a facility to determine where additional control is needed. In addition, there is no way of determining, without actually taking both personal and area samples, whether the results of personal sampling would in fact be higher or lower than area samples taken in the same facility; whether breathing zone samples are higher or lower than area samples depends on a host of factors, including the positioning of the area sample in relation to the source of emissions, the location of the worker in relation to the same source, and the amount of time the worker spends in the vicinity of the emission source.

The Inter-Industry Committee on Carbon Disulfide submitted to the record a recent epidemiologic study by MacMahon and Monson (1988/Ex. 125). The study cohort consisted of 10,418 men employed between 1957 and 1979 in the four principal U.S. viscose rayon plants. The mortality status of the cohort was ascertained up to mid-1983. Cohort members were placed into general exposure categories according to job title; these categories were highest, intermediate, variable, least, and none. The authors found no significant increase in overall mortality in the 4,448 employees with the highest potential for CS(2) exposure compared with the mortality among 3,311 employees with no CS(2) exposure. However, there was a statistically significant excess of arteriosclerotic heart disease (ASHD) among the most heavily exposed workers (242 deaths versus 195.6 expected). No clear relationship was observed between exposure duration or latency and excess ASHD mortality; however, the data suggested that the risk was higher among employees exposed to CS(2) for 15 or more years and among employees hired prior to 1960.

In addition, MacMahon and Monson (1988/Ex. 125) found a statistically significant increase in the SMR (SMR = 150) for ASHD among members of the cohort who had been exposed to CS(2) the year immediately preceding the date of death or the termination date of the study (Ex. 125, Attachment B, Table 7, p. 702); however, there was no general pattern of increased SMRs among cohort members whose time since last exposure exceeded one year. This finding is consistent with the results of the British studies, which also found an increased risk of heart disease among recently exposed employees but not among employees who had left their jobs.

The Inter-Industry Committee on Carbon Disulfide interpreted the MacMahon and Monson (1988/Ex. 125) study to mean that U.S. workers employed since 1960 were not at risk of ASHD (Tr. 4-96), and NIOSH (Ex. 193, Comments on Carbon Disulfide) noted that the study lacked exposure data. However, OSHA finds the results of the MacMahon and Monson (1988/Ex. 125) study supportive and consistent with those of the British and Finnish studies discussed above. First, all of the studies clearly demonstrate a positive association between exposure to carbon disulfide and increased risk of mortality from heart disease. Second, studies from all three countries link the excess risk to cumulative CS(2) exposure. Third, the studies from all three countries demonstrate that significant risk can be substantially reduced or eliminated by reducing or stopping exposure, even after a considerable CS(2) dose has accumulated; both the U.S. and British studies report a significantly increased risk of death from heart disease among workers who were recently exposed. However, no increased risk was seen among workers whose exposures had ended one year or longer prior to death or the end of the study. Moreover, the Finnish study reported steady declines in heart disease mortality among workers after exposure levels were reduced to below 10 ppm and a rigorous medical screening and removal program was instituted. These findings clearly demonstrate that current or continued exposure to carbon disulfide at the levels presently encountered in these facilities is as important a risk factor for heart disease mortality as cumulative exposure.

In addition to evidence that carbon disulfide is a cardiovascular toxin, there is a substantial body of evidence that exposure to carbon disulfide presents a fetotoxic hazard and that this substance may also be a teratogen. Some of the early (pre-1977) animal data on reproductive effects were reviewed in the NIOSH (1977b/Ex. 1-260) criteria document on carbon disulfide. In its posthearing submission, NIOSH (Ex. 193) mentions two relevant reports. One by Cai and Bao (1981, as cited in Ex. 193) reported increased incidences of menstrual disturbances and of pregnancy toxemia, a potentially lethal condition, in rayon workers. These authors also presented evidence that CS(2) can cross the placental barrier and be secreted into mothers’ milk. The second report cited by NIOSH (Hemminki and Niemi 1982, as cited in Ex. 193) found a significantly elevated incidence of spontaneous abortions among women employed in viscose rayon facilities in Finland; however, data on the specific CS(2) exposure levels were generally lacking.

The Rohm and Haas Company submitted a summary (Ex. 10-5) of the evidence on the reproductive toxicity of carbon disulfide to the OSHA docket; this information shows that carbon disulfide has caused fetal deaths and malformations in CS(2)-exposed laboratory animals. Rohm and Haas cite a series of abstracts by Tabacova and others in which oral administration of CS(2) to female rats during gestation produced both teratogenic and fetotoxic effects. These effects were magnified in the F(2) offspring of the prenatally exposed F(1) generation, which suggests that CS(2) has a multigenerational effect that continues to cause malformations in successive generations.

Jones-Price et al. (1984, NTIS/PB84-192343) found both maternal and fetal toxicity in CD rats exposed orally to 200, 400, or 600 mg/kg/d CS(2) during days 6 through 15 of gestation. No dose-related increases in the incidence of teratogenicity were observed. In another report, Jones-Price et al. (1984, NTIS/PB84-192350) found significant dose-related increases in percent resorptions/litter, percent non-live (dead or resorbed)/litter, and percent of fetuses affected (non-live and malformed)/litter among New Zealand White rabbits exposed orally to 25, 75, or 150 mg/kg/d during days 6 through 19 of gestation. The percentage of malformed fetuses per litter increased with dose and was statistically significant at the highest dose tested.

In an inhalation study, Hardin, Bond, Sikov et al. (1981/Ex. 1-699) exposed rats and rabbits to 20 or 40 ppm CS(2) for 6.5 hours per day during days 1 through 19 (rats) or 1 through 24 (rabbits) of gestation. No embryotoxic or fetotoxic effects were noted, indicating that 40 ppm is a no-effect level for these effects in rats and rabbits. According to the analysis by Rohm and Haas, the lowest-reported-effect level (25 mg/kg/d) documented by Jones-Price et al. (1984) for rabbits corresponds to an equivalent airborne exposure of 58 ppm; this lowest-reported-effect level is in close agreement with the no-effect level reported by Hardin et al. (1981/Ex. 1-699) for the same species.

OSHA believes that this evidence, which shows that consistent fetotoxic and teratogenic effects are associated with exposure to carbon disulfide, warrants considerable concern. OSHA is particularly alarmed at the multigenerational effect of CS(2) exposure that has been demonstrated to occur in rats. This risk of reproductive effects, combined with the previously recognized risk of cardiovascular disease, have convinced OSHA that a substantial reduction in the PEL for carbon disulfide is clearly justified.

Several foreign governments and standards-setting organizations have already established 8-hour TWA exposure limits for carbon disulfide that range from 1 to 10 ppm. For example, NIOSH has recommended a 1-ppm TWA limit for this substance, and Rohm and Haas established an internal limit of 4 ppm as an 8-hour TWA (Ex. 10-5). Several foreign countries, including West Germany, Italy, Japan, Sweden, and Switzerland, currently have 10-ppm limits. The ACGIH has established a 10-ppm TLV for CS(2); however, the ACGIH limit does not consider any of the evidence of CS(2)’s fetotoxic or teratogenic effects.

Based on the evidence in the record and the toxicological literature, OSHA concludes that 4 ppm is a reasonable and prudent level at which to establish a revised 8-hour TWA PEL for carbon disulfide. This limit should provide for a substantial reduction in the significant risk both of cardiovascular disease and adverse reproductive effects associated with CS(2) exposures; clearly, these effects constitute material impairments of health and functional capacity. In addition, because of the seriousness of the effects associated with exposure to carbon disulfide, and in accordance with the policy described in Section VI.C.17 on short-term exposure limits, OSHA finds that a STEL is necessary to ensure that the 8-hour TWA limit is not exceeded during operations characterized by intermittent exposures to elevated levels of CS(2). Rohm and Haas (Ex. 10-5) has established an internal guideline of 12 ppm as a short-term limit to ensure that the 8-hour TWA limit is not exceeded, and NIOSH also recommends a short-term limit to ensure that full-shift exposures are maintained under good control. In the final rule, OSHA is accordingly establishing a 12-ppm STEL to supplement the 4-ppm TWA PEL. * OSHA’s assessment of the feasibility of this limit indicates that, under normal operating conditions, a 4 ppm TWA PEL and a 12 ppm STEL are generally achievable by using engineering and work-practice controls. Evidence in the record demonstrates that engineering controls and work practices are not feasible to achieve compliance and respiratory protection may be required during certain operations in industries that regenerate cellulose from viscose to form commercial products such as rayon staple, rayon yarn, cellophane, sponges and casings. Accordingly, respirators may be worn to achieve compliance with the Air Contaminants Standard when employees are performing the following tasks:

– Maintenance-type tasks (regardless of whether such tasks are performed by “maintenance personnel” or by others), such as tank washing, opening and redressing filters, cleaning process liquor screens, and handling unwashed, unpurified viscose and viscose products;

– Opening of production lines, e.g., to troubleshoot production quality, take tank samples, set thickness of cellophane, change spinerettes, clear jams, spin, thread and align film and fiber strands during extrusion, regeneration, and cutting, and manually puncture casings;

– Handling of fibers and filament bundles that have been removed from process equipment;

– Effecting product-line changes; and

– Loading alkali cellulose, and unloading, washing and dissolving xanthate, viscose and viscose products.