Most Popular Links
- All RELs
- Environmental Tobacco Smoke (ETS)
- East Bay Kids Web
- Hot Spots
- Criteria Air Pollutants
- Diesel Exhaust
- Criteria Air Pollutants
- Diesel Exhaust
- Environmental Tobacco Smoke
- Hot Spots Guidelines: Technical Support Documents
- Exposure Assessment and Stochastic
- Risk Assessment Methodology
- MTBE and Ethanol
- Toxic Air Contaminants
- Wildfire Smoke: A Guide for Public Health Officials
- Air Pollution and Children's Health
- Toxicity Criteria Database
- Outside Air Links
resources for kids and educators
- Disney's Planet Challenge
- Education Links
- Guidelines for Safe Use of Art and Craft Materials
- Green California Schools Summit
- Lawrence Hall of Science
- Sally Ride Science Festivals
Air Toxicology and Epidemiology
and UF Resin Manufacturers Associations, Cont.
Up | Down
Comment: The rationale for the uncertainty factor, intraspecies variability, is expressly addressed by other aspects of the BMD:
Furthermore, the TC01 [1% BMR] is consistent with the definition of the proportion of the population that may not be protected by these levels (as defined previously, these levels are intended to protect "essentially all" individuals and a very small proportion may not be protected). Use of the 95 percent lower confidence limit on concentration takes into account the variability of the test population .... TSD, page 24.
The interrelation of the uncertainty factors with other components of the BD has been repeatedly noted by experts in the field. For instance, some at the ILSI Workshop questioned the use of LCL in light of the adjustments that are typically made elsewhere in the BD calculation:
The use of the uncertainty factor for intraspecies variability elicited a discussion on the significance of using the central estimate versus the lower confidence limit (LCL) of the BMD. Some participants suggested that the LCL accounts for intraspecies variability as well as experimental variability. (ILSI Workshop Report, p. 11)
Response: It is true that the relationship between the 95% LCL and intraspecies variation was mentioned in the proceedings; however, the workshop ultimately decided not to recommend any changes to the intraspecies uncertainty factors. OEHHA reduced the intraspecies uncertainty factor to 3 when using a BC approach on a study in normal human subjects since we believe that some degree of individual variation in the test population was accounted for by use of the 95% LCL.
Comment: The EPA Report noted the extensive work that is being done on uncertainty factors17 and highlighted other items that should influence their determination:
...the calculation of the BMD depends on the BMR as well as the size of the statistical confidence bound employed. These additional considerations may need to be accounted for when selecting uncertainty factors for BMDs.
The EPA also noted that, "Some biological considerations (e.g., relating to the possibility of a threshold for the responses under investigation) could affect the selection of uncertainty factors." This is particularly apt in the formaldehyde setting where it is agreed that there is a threshold for acute effects of formaldehyde.l8
Response: OEHHA continues to believe that a threshold exists for the onset of non-cancer health effects, including those caused by formaldehyde. Both the NOAEL and benchmark dose approaches yield results consistent with a practical threshold.
Comment: Another biological consideration of even greater import, as noted above, is the severity of the toxicological endpoint. This is particularly relevant to formaldehyde. When the debate over toxicity turns on whether a substance merely can be detected or whether it is minimally annoying, the uncertainty factors should be viewed in a different light than when effects such as reproductive toxicity or other endpoints are at issue.
Response: The severity of effect is accounted for by the use of severity effect levels and also by OEHHAs use of uncertainty factors. We agree that minimal irritation should be viewed differently than reproductive effects. For this reason, mild irritation is considered a mild adverse effect and reproductive or developmental toxicity is considered a severe adverse effect. The mere detection of a compound, in absence of toxicological signs or irritating or annoying symptoms, is not considered by OEHHA as a mild adverse effect. In addition, OEHHA has reduced the LOAEL to NOAEL uncertainty factor to 3 for mild irritation.
Comment: Again, one can address the overall uncertainty of the BMD calculation in the UF factor, in the modifying factor, in the use of a different BMR, in the use of the MLE versus the LCL, or in a combination of ways. The factors are interrelated. The Associations submit that the underlying data and overwhelming body of knowledge on formaldehyde19 justify either the elimination of the uncertainty factor or further reduction of the modifying factor to 0.1 or 0.2.
Appendix E of the TSD lists chemicals for which complete data base searches have been completed, those for which searches are in progress, and those which are to commence in 1995. Formaldehyde does not appear on any of the lists. The literature on formaldehyde health effects is extensive.
Response: OEHHA has agreed to change the BC01 to BC05 in accordance with the recent workshop by ILSI and reported in the literature (Regul. Toxicol. Pharmacol. 21:296-306, 1995). The UF methodology for the BC approach has been changed. A factor of 3 is used for the interspecies or intraspecies uncertainty factor, depending on whether the key study was performed in humans or animals. The modifying factor is no longer used.
In regard to footnote #19, a literature search has been completed for formaldehyde. Unfortunately, Appendix E incorrectly indicated that the literature search was not yet completed.
Comment: The Hot Spots program requires the facility to evaluate potential risk to the surrounding population in ways that add still additional safety factors. These features, in effect even if not directly, lower the toxicity exposure levels even more.
For example, the facility must create an exposure assessment to determine the ground level concentration at each gridded receptor and the expected exposure to specific, sensitive receptors. Concepts of "maximum impacted offsite locations," "maximum exposed individual residents," and "maximum exposed individual workers" are woven into this exercise. The probability that the maximum exposed individual would be a person whose sensitivity would only be reflected by the 1% BMR is extremely small.
Response: The RELs are used to evaluate impacts of many receptors, not just the maximum impacted receptors (MEIs) as indicated by the comment. The MEI exposures may also be experienced by a large number of people in a given gridded receptor. By protecting the most sensitive receptor, RELs offer protection to the majority of individuals in an exposure area. In accordance with earlier comments, the BC01 will be changed to BC05.
Comment: Second, added responsibilities to prepare a detailed isopleth map arise if the Hazard Index for any endpoint exceeds 0.5. The net effect of this provision is to reduce by a factor of 2 the computed RELs, in spite of the fact that they already reflect the numerous other conservative assumptions and safety factors described above. Although these various safety factors ---1% BMR, use of 95% LCL, uncertainty factors, no accommodation to severity of toxicological endpoint, maximum exposed individual, and 0.5 HI trigger for mapping -- may be logical and defensible when analyzed individually, they have a multiplicative impact when incorporated into the Hot Spots Program together. The result is an overly-protective and burdensome regulation. OEHHA suggests that it is required by statute to employ a margin of safety. Although there is a question as to whether the provisions of AB 1807 and 2728 apply to the Hot Spots program; in any event, this language should not be used repetitively to make the risk assessment scientifically meaningless.
Response: OEHHA has withdrawn the proposed requirement regarding the HI of 0.5. The other comments have been addressed in detail above. The commentators footnote (#20) is in error; Health and Safety Code Section 39660 (Ch 1161, Sec 4) states specifically to use "...an ample margin of safety which accounts for variable effects..."
Comment: The Associations ask that the data from the Kulle study be recharacterized and that appropriate changes in the modifying factors or elsewhere be made to remove the layering of safety factors.
Response: This concern has been addressed in the calculations above. Recharacterizing the data from the Kulle et al. study results in only a minute change in the REL. The current approach is consistent with OEHHAs definition of adverse effects.
Comment: OEHHA has set a Level II exposure for formaldehyde of 10 ppm for "disability or serious effect[s]" based on the AIHA-ERPG-2. However, in the TSD the following description is given for expected effects: "Moderate eye irritation and lacrimation may be experienced with exposure to formaldehyde above this level." The Associations recommend that this description be modified by removing the word "moderate" and changing "may" to a more forceful statement. Exposure to formaldehyde at these levels would be very noticeable and very uncomfortable to almost all individuals.
Response: The level II for formaldehyde has been revised. It is now based on decrements in FEV1 in exercising asthmatics. The level II was revised to 1.6 ppm
Comment: OEHHA has proposed the following RELs for methanol:
Level I Discomfort 2.1 ppm (2.8 mg/m3)
Level II Disability or Severe Effect 5.1 ppm (6.6 mg/m3)
Level III Lethality 40.0 ppm (52.4 mg/m3)
This substance, perhaps more than any other chemical in the Evaluation, reflects the difficulties that are encountered when assumptions are applied without reference to the practical implications of the end result. The Occupational Safety & Health Associations Permissible Exposure Limits (PEL)21 and American Conference of Governmental Industrial Hygienists ("ACGIH") Threshold Limit Value ("TLV")22 for methanol are 200 ppm (262 mg/m3) on an 8-hour time-weighted average basis. These levels, promulgated by two prestigious organizations charged with the protection of worker health are five times higher than the level that OEHHA has alleged "may result in death due to respiratory or cardiac arrest."
Response: The 3 levels for methanol are addressed individually in the following comments and responses.
Comment: The Level I REL For methanol should be changed. The Level I REL is based on a study of 12 healthy young men exposed to methanol at 250 mg/m3 for 75 minutes in controlled chambers (Cook et al. 1992)23. Note that this is below the OSHA and ACGIH levels in both concentration and time of exposure. A battery of neurobehavioral challenges was evaluated for methanol-induced changes. OEHHA made the assumption that 250 mg/m3 (190 ppm) was the LOAEL; it then applied a 100-fold safety factor (10-fold for LOAEL -> NOAEL; 10-fold for interspecies [sic] differences, and extrapolated from the 75-minute exposure of the study to a 60 minute acute exposure using the following relationship Cn * T = K, where n=2, to derive a REL Level I of 2.8 mg/m3 (2.1 ppm). This is a poor choice of study and endpoint. Furthermore, OEHHA differs with the authors of the study in the interpretation of results. For example, a very small number of subjects (n = 12) was followed. Further, of 20 commonly used tests of sensory, behavioral, and reasoning performance before, during, and after each exposure, no detectable effect on the subjects performance was noted, except in "subjective ratings of fatigue" (Cook et al. 1992) and the timing of peaks of brain wave patterns in response to light flashes and sounds, that varied by latency of response. Marginally significant effects were found for subjective levels of concentration and performance on the Sternberg memory task, which although statistically significant, were not outside the range of values observed with the subjects in sham exposures. Therefore, the findings were not clinically relevant. When considered as a whole, there is a lack of internal consistency of response. For example, reported reductions in concentration and increased subjective feelings of fatigue should also be reflected in performance on all of the neurobehavioral tests. Also, reduction in reaction time on the Sternberg memory test should be mirrored by a slower reaction time in the Symbol Digit substitution task. Similar changes in other measures of brain wave patterns should have been observed in addition to the changes in latency of response to light flashes and sounds. The authors of the Cook study indicate clearly in their conclusions: "Given the plausible role of chance in obtaining positive results in this study, and the uncertain significance of changes in the event-related potentials, it is inappropriate to conclude that these positive effects are attributed to methanol exposure". There are other significant shortcomings in the derivation and explanation of the REL. First, the 10-fold safety factor for estimating a NOAEL based on the LOAEL is inappropriate. Even if one accepts the noted changes as adverse effects, a safety factor of 10 is unnecessarily conservative given the mild (benign) nature of the response. Second, a 10-fold safety factor could be employed to account for inter-individual, or intraspecies extrapolation, but the reference to interspecies is misplaced. The 10-fold safety factor for individuals is a standard default assumption. Third, no consideration was given to mechanism of action, and/or ultimate toxic metabolite.
Response: OEHHA concludes that the Cook et al. (1991) study is adequate for setting the mild adverse effect level REL. It is a well-conducted study of neurobehavioral effects due to methanol. Its faults are the small sample size and failure to test at greater concentrations. The Cook et al. (1991) study showed statistically significant CNS effects in methanol-exposed individuals compared to controls. The comment argues that these effects (fatigue and visual evoked potential) are not biologically significant and are inconsistent with the other neurological tests. However, it is well-known that methanol causes acute CNS depressant effects (Andrews and Snyder, 1991; Kavet and Nauss, 1990). The statistically significant effects reported in Cook et al. (1991) can therefore be interpreted as consistent with the onset of CNS effects. The apparent discrepancies between the positive Sternberg test and the negative Symbol Digit substitution test highlighted by the comment are not pertinent, since the Sternberg test results were not statistically significant. The authors of the study were tentative in concluding definite effects from methanol exposure. We continue to believe that the statistically significant effects and other trends observed in the paper were consistent with the hypothesis that methanol exposure causes CNS depression and that these effects warrant concern. However, because of the uncertain significance of the findings of this study as a whole, OEHHA will consider this study to represent a free-standing NOAEL until further research is available. An uncertainty factor of 10 for intraindividual variability will be applied, as will time extrapolation, yielding a REL of 21 ppm (1.7 mg/m3).
Comment: The Level II REL for disability is similarly inappropriate given the other reference standards. The Level II REL is based on a NOAEL of 1,300 mg/m3 (1,000 ppm) for congenital malformations in mice exposed to methanol for 7 hours/day on gestation days 6-15 (Rogers et al. 1993).24 The LOAEL was 2,600 mg/m3 (1,200 ppm), at which an increase was seen in the number of fetuses per litter with cervical ribs (small ossification sites lateral to the 7th cervical vertebra). The next higher effect level was 6,500 mg/m3 (5,000 ppm), for which increased incidences of exencephaly and cleft palate were reported. For 1% and 5% added risk, the study authors calculated maximum likelihood estimates, and lower 95% confidence bound benchmark doses which were adopted by the state of California (see Table 1). A 30-fold safety factor was applied by the OEHHA (100-fold for inter- and intra-species variability, and an additional modifying factor of 0.3 "since the BC approach accounts for some degree of individual variation". The 7-hour value was extrapolated to a 1-hour REL using the formula Cn * T = K, where n=2, with a resulting Level II REL of 6.6 mg/m3 (5.1 ppm). The OEHHA recommends that this REL be revised when a primate reproductive study is available.
Table 1. Exposures Corresponding to Incremental Risk of Cervical Rib Incidence in Methanol-Treated Mice(mg/m3)
1% added risk
5% added risk
|Rogers et al.||
The Rogers et al. report is the best study available in the peer-reviewed literature for setting a reproductive toxicity-derived safe exposure level. This study was well conducted and well-controlled. However, one limitation is the fact that it was performed in a non-primate species. Point (MLE) and lower-bound (BMD) benchmark dose estimates from this data set appear to be reasonable. OEHHAs application of safety/uncertainty factors is not inconsistent with their traditional conservative assumptions. However, in spite of the consistency of approach with normal default assumptions, one must grapple with the paradoxical nature of the 5.1 ppm resulting value for reproductive or developmental toxicity in light of the 200 ppm OSHA and ACGIH values. There is no indication whatsoever that the OSHA and ACGIH levels are inadequate or inappropriate. In the face of these disparities, the Associations submit that reliance on generic uncertainty factors is misplaced. One possible explanation for the problem arises from the study. There is concern that mice are not an appropriate species for human health risk assessment due to the high-dose sensitivity of rodents that is observed and predicted by pharmacokinetics of methanol disposition. A primate study would be preferred because of the potential interspecies differences in methanol metabolism. Although not yet published in the peer-reviewed literature, such a study exists, and has been cited by the World Health Organization (1994)25 and Kavet and Nauss (1990)26. These two groups report that the Japanese Institute for Applied Energy, with the sponsorship of the New Energy Development Organization (NEDO), conducted an extensive research program in which rodents and cynomolgus monkeys were exposed to 13, 130, and 1,300 mg/m3 (10, 100, and 1,000 ppm, respectively) of methanol for up to 30 months. Summaries of the studies indicate that no reproductive (or other toxic or carcinogenic effects) were evident at levels of 130 mg/m3, and no teratogenic effects were observed even at 1,300 mg/m3.
Response: OEHHA thanks the commentator for bringing the Japanese Institute for Applied Energy/New Energy Development Organization data to our attention. It appears that this is a chronic study (30 months), which is inappropriate for use as the basis for an acute, severe adverse effect value. The information presented by the comment is too general and vague to allow thorough evaluation of these data (e.g., no sample size or dose-response information is presented).
The severe adverse effect level for methanol is lower than occupational standards as described by the comment. However, the Rogers et al. (1993) study in mice is more recent than the most recent versions of the ACGIH-TLV (1992), which only describe congenital malformations in rats at high doses of methanol from Nelson et al. (1985). It is possible that, given the new developmental toxicity information, the ACGIH-TLV committee will revise the TLV.
Even if this was not the case, using workplace standards to set acceptable community standards is inappropriate. The occupational standards developed by the ACGIH are clearly intended for purposes other than community health-based values. Even ACGIH acknowledges the limitation of the TLVs utility in the second paragraph of the Introduction section of the TLV document: "These values are intended for use in the practice of industrial hygiene as guidelines or recommendations in the control of potential health hazards and for no other use, e.g., in the evaluation or control of community air pollution or physical agent nuisances...These values are not intended as fine lines between safe and dangerous exposure concentrations...." The TLVs are designed to protect healthy workers and not the infirm, infants, children, pregnant women, or elderly. Thus, the TLVs and RELs cannot be directly compared.
The severe adverse effect level for methanol has been recalculated using the BD05 and this is reflected in the Technical Support Document. The level II REL thus changes from 6.6 to 35 mg/m3. We reiterate that a severe adverse effect level based on reproductive effects will be revised upon publication of a primate study.
Comment: A Level III (lethality) REL that is five times lower than accepted occupational standards is inappropriate. OEHHAs life threatening effect level REL is based on a 1931 study of rhesus monkeys, rabbits, and rats, in which animals were exposed to 1,300 mg/m3 - 52,400 mg/m3 (1,000 - 40,000 ppm) for 1-18 hours per day for up to 4 days. The point selected by OEHHA for derivation of its Level III REL was the LOAEL of 40,000 ppm for 1 hour, "the concentration which resulted in death in all animals several days following exposure." It is not clear from the description of the study in the TSD which animals succumbed, but it appears that monkeys were most likely the target species. To the LOAEL of 40,000 ppm (52,400 mg/m3), a 1,000-fold uncertainty factor was applied: 10-fold for LOAEL-> NOAEL; and 100-fold for inter-and intra-species adjustments. No time adjustment was applied, as the exposure duration was one hour. The resulting Level III REL was 40 ppm or 52.4 mg/m3. It is unclear why this particular study was selected; there is a large body of literature, some it of quite recent (reviewed in Kavet and Nauss, 1990; WHO, 1994) on both acute and chronic effects of methanol intoxication. One of the conclusions that continues to surface in the literature is that route of exposure is critical factor for methanol toxicity. Rather, the intake dose and resulting blood methanol concentrations are critical to methanols toxicity (Clayton and Clayton, 1982)27. Thus, oral exposures to methanol in appropriate species could be employed in developing more reliable estimates of exposure levels protective against acute lethal toxicity. Whatever the strengths or weaknesses of the methodology used, a comparison of OEHHAs Level III REL for lethality (52.4 mg/m3) to the OSHA PEL or ACGIH TLV of 262 mg/m3 for an 8-hour time-weighted average indicates the inappropriateness and unacceptability of the new level.
Response: If inhalation data are available, these are preferred over data from other routes of exposure for setting inhalation RELs. The RAAC recommended that route-to-route extrapolation not be performed for acute REL development. The "large body of literature" mentioned in the comment contains little, if any, inhalation data, and although the review by Kavet and Nauss (1990) is fairly recent, the information reviewed in their paper is not. However, OEHHA agrees that the life-threatening effect level for methanol appears to overestimate lethal effects. The revised NIOSH 1996 IDLH level of 6,000 ppm (7860 mg/m3) has been proposed as the life threatening level for methanol by OEHHA.
Comment: The Associations respectfully request that OEHHA reexamine the RELs that have been proposed for methanol in light of the evaluations of other governmental and private health agencies. Attached as Exhibit D for the Agencys consideration is a "Determination of Alternative Safe Human Exposure Levels for Discomfort, Disability, and Lethality" prepared by Drs. Thomas Starr and Larisa Rudenko of Environ.
Response: OEHHA appreciates the efforts to improve the technical support document. We have reviewed the attachments provided. As noted in our previous response to this commentator, we have revised all three toxicity criteria, including the level I REL.
Comment: Attached as Exhibit E is a description of some concerns raised by Environ regarding the use of additivity when compiling the hazard index from various hazard quotients. Even if the same endpoint is being evaluated, it is not always appropriate to aggregate these quotients.
Response: There are very little data available on effects of chemical mixtures. When two or more toxicants evaluated in a risk assessment impact the same target organ or system, it is prudent to assume additivity. There are a limited number of examples of both synergism and antagonism in the literature. However, in general, these type of data are lacking. The large number of possible combinations of chemicals to which people are exposed simultaneously both from the facility being evaluated as well as other sources cannot be adequately characterized by the available data. If a risk assessor has data pertaining to the specific mixture being evaluated, that data can be supplied with the risk assessment as supplemental information. However, it is not possible to ascertain the combined effects of exposure to emittents from a hot spots facility and exposure to other airborne, water-borne, food-borne or medicinal chemicals to which an individual is also simultaneously exposed. In addition, the physiological status of an individual including such factors as age, health, pregnancy status, lactational status, and nutritional status will likely impact the response to exposure to multiple chemicals. This is not adequately reflected in the available data on chemical mixtures. Assuming additivity is one way to account for the large amount of uncertainty. It is not particularly health-protective in some instances where synergism has been demonstrated.
References cited by the commentator
2/ Kulle, J.T., L.R. Sauder, J. R. Hebel, D. Green, and M.D. Chatham 1987. formaldehyde dose-response in healthy nonsmokers. J. Air Pollution Control Assoc. 37:919-924, at 920, Col 2. A copy of the Kulle study is attached as Exhibit C. 3 Id. at 921.
3/ Id. at 921.
6/ Kulle at 920.
7/ Dourson, M.; Stara, J. (1983) Regulatory history and experimental support for uncertainty (safety) factors. Reg. Toxicol. Pharmacol. 3: 224-238.
8/ EPA, The Use of the Benchmark Dose Approach in Health Risk Assessment. (1995) EPA/630/R-94-007 (the "EPA Report").
9/ See DeRosa, C.T., Stara, J., and Durkin, P. (1985). Ranking chemicals based on toxicity data. Tox. Ind. Hlth. 1:177-199.
10/ International Life Sciences Institute and Risk Science Institute, Report of the Benchmark Dose Workshop, September 28-30, 1993. ("ILSI Report").
12/ See footnote 10, supra.
13/ ILSI Report, at page 8.
14/ Allen, B.C.; Kavlock, R.J.; Kimmel, C.A.; Faustman, E.M. (1994) Dose-response assessment for developmental toxicity: II. Comparison of generic benchmark dose estimates with NOAELs. Fund. Appl. Toxicol. 23: 487-495.
l7/ Hattis, D.; Lewis, S. (1992) Reducing uncertainty with adjustment factors. The Toxicologist. 12: 1327. (EPA Report at 50).
21/ 29 CFR ¤ 1910.1000, Table Z-1.
22/ American Conference of Governmental Industrial Hygienists, "Threshold Limit Values and Biological Exposure Indices for 1994-1995," p.25.
23/ Cook, M.R.; Bergman, F.J.; Cohen, H.D.; Gerkovich, M.M.; Graham, C.; Harris, R.K.; and Siemann, L.G. 1991. Effects of methanol vapor on human neurobehavioral measures. Health Effect Institute, Research Report No. 42.
24/ J Rogers, J.M.; Mole, M.L.; Chernoff, N.; Barbee, B.D.; Turner, C.I.; Logsdon, T.R.; and Kavlock, R.J. 1993. The developmental toxicity of inhaled methanol in the CD-1 mouse, with quantitative dose-response modeling for estimation of benchmark doses. Teratology. 47: 175-188.
25/ World Health Organization, the United Nations Environmental Program, and the International Labor Organization. 1994. International Program on Chemical Safety: Environmental Health Criteria - Methanol. Draft Report. Geneva, Switzerland.
26/ Kavet, R. and Knauss [sic], K.M. 1990. The toxicology of inhaled methanol vapors. Critical Reviews in Toxicology 21(1):22-50.
27/ Clayton, G.C. and Clayton, F.E., eds. (1982) Pattys Industrial Hygiene and Toxicology: Volume 2C: Toxicology with cumulative index for vol. 2. 3rd Edition. John Wiley & Sons. New York, pp. 4527-4551.
Auton, T.R. 1994. Calculation of benchmark doses from teratology data. Reg. Toxicol. Pharmacol. 19:152-167.
Crump, K.S. 1984. A new method for determining allowable daily intakes. Fundam Appl. Toxicol. 4:854-871.
Malsch, P.A., Proctor, D.M., and Finley, B.L. 1994. Estimation of a chromium inhalation reference concentration using the benchmark dose method: a case study. Reg. Toxicol. Pharmacol. 20:58-82.