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
Comment: (1) The guidelines should allow risk managers to deviate from the OEHHA recommendations relative to the factor of safety below the lowest observed adverse effects level.
(2) Evaluations should not be required where RELs are not provided, but a comparison to the LOAEL should be added.
(3) It is more informative to calculate a facilitys maximum potential hourly impact, and compare that to the maximum measured hourly background, than to add the maximum potential hourly impact to the average annual background. Nevertheless, districts should be allowed to determine whether such comparisons are beneficial.
(4) Proposed factors of safety used for each degree of uncertainty should relate to the severity of the effect. If a factor of safety of ten/degree is appropriate for lethal effects, a factor of safety closer to two/degree would be more appropriate for irritation.
The Office of Environmental Health Hazard Assessment (OEHHA) proposes on pages 21-22 of its January 1995 draft "The Determination of Acute Toxicity Exposure Levels for Airborne Toxicants" using factors of safety of ten for each successive area of uncertainty. Specifically one factor of ten is always imposed, and additional factors of safety (which OEHHA refers to as an "uncertainty factor") of ten is proposed whenever the effects data is based on: LOAEL, animal data, average persons, or deficient studies instead
Because these factors are then multiplied together, the combined factor of safety applied can be as high as 105 or 100,000 regardless of how insignificant the effect might be, or as low as 10 regardless of how serious the effect might be, depending entirely upon the nature of the data, not upon the nature of the effect. Since circumstances which cause more serious effects are generally better documented, this policy tends to result in the least protection against most serious, certain and common effects. For example, with respect to acrolein OEHHA proposes a factor of safety of about 5 relative to "may be lethal" effects, but a factor of safety of 1,000 relative to "may result in "effects. That should be reversed!
Response: There are several inaccuracies in the comments statements. First, OEHHA does not use uncertainty factors for "well designed studies" or data gaps in the acute document. Second, the application of uncertainty factors by OEHHA is to account for specific sources of uncertainty and variability not addressed in the scientific literature: individual variation, interspecies differences, and estimation of a NOAEL from a LOAEL. The maximum combined uncertainty factor is 1000, used when deriving an REL from an animal study without a reported NOAEL. The minimum uncertainty factor is 1, as in the case of sulfuric acid, where the REL is based on a study in a human sensitive population (asthmatics). OEHHA accounts for severity of effects in two ways: (1) by categorizing them into one of three levels: mild adverse effect level, severe adverse effect level , and life-threatening effect level; and (2) by using smaller uncertainty factors in predetermined situations. RELs based on LOAELs for mild sensory irritation have smaller uncertainty factors of 3 instead of the traditional 10. OEHHA has incorporated the best available scientific information into the traditional uncertainty factor method where data have allowed such a departure to occur without endangering community health. By applying consistent uncertainty factors for each level, the risk manager is provided with as much information as needed to understand
The uncertainty factor used by OEHHA for acrolein is 100, not 1000 as stated in the comment.
Comment: In proposing large added factors of safety for each added degree of uncertainty, OEHHA creates an obligation to more carefully evaluate whether there is a full degree of added uncertainty.
Response: The NAS explicitly states: "It is vital to select uncertainty factors that reflect the quality and relevance of the data, differences between test species and humans, and variation within the human population. Typically, in the past the permissible human exposure has been reduced by a factor of 10 for each additional source of variation or uncertainty" (NRC (1993), executive summary, p. 6). OEHHA agrees that only using a factor of 10 is too limiting and may ignore contributions made by chemical-specific data. For this reason, the technical support document uses uncertainty factors in the range from 1-10, depending on available data. For example, for benchmark calculations, the interspecies UF for animal studies and the intraspecies UF for human studies has been changed to 3. Similarly, for sensory irritation, the uncertainty factor for LOAEL to NOAEL extrapolation has been reduced to 3. Dourson and Stara (1983, p. 228) stated that, based on their data "it seems somewhat reasonable to employ a 10-fold uncertainty factor to account for intraspecies variability in lieu of chemical-specific toxicity "
Comment: In the case of nickel, insoluble nickel appears unlikely to cause the acute hazard effects indicated. Sources wishing to distinguish between soluble and insoluble nickel compounds emitted should be able to reduce calculated hazards accordingly.
In the Determination of Acute Toxicity Exposure Levels for Airborne Toxicants, OEHHA indicates that the acute hazard impact for nickel is based upon testing in 1978 by Graham et al. This study did report a potential immunotoxicity suppression effect in mice at inhalation doses of 250 m g/m3 nickel, and no effect at inhalation doses of 110 m g/m3. However, there was some uncertainty as to the significance of the effect. At most the effect appears to be an 11% decrease in cell production attributed to a four times higher nickel chloride dose. Also, the report in question includes data for three nickel compounds: nickel chloride, nickel sulfate and nickel oxides. For the sulfate and oxide, the effects were about the same at 12 mg Ni/g body wt as at 3 mg Ni/g body wt, and about equal to the zero exposure data point for NiCl2. Only NiCl2 showed any consistent effect at increasing dose. Furthermore, a greater effect (lower plaque production) was observed in one of three exposures to dilute hydrochloric acid (HCl at pH=6). That leaves it uncertain whether it is the nickel ion, the chloride ion, the specific compound, or some testing artifact which contributed to this limited effect observed. The actual data was:
Log10 plaques per 106 cells
HCl Test Data (Log10 plaques per 106 cells)
Response: The data shown in the comment are not the data upon which the REL for nickel was based. The comment shows the dose-dependent suppression of antibody-forming cells (or "plaque-forming cells") by intramuscular injection of nickel salts and nickel oxide. These data substantiate the systemic immunotoxic effect observed following inhalation exposure to the same compounds. In addition, the degree of suppression of the antibody response stated in the comment, 11%, is inaccurate. The actual suppression in antibody cell production in the data shown above is up to 55% compared to controls because the suppression data are on a logarithmic scale. In addition, the true degree of suppression in the data upon which the REL is based is closer to 25% at the LOAEL based on historical data, although the number of plaque-forming cells in the controls et al. (1978) reference).
The use of the hydrochloric acid pH=6 data is inappropriate since these data were collected following intramuscular injections in a different experiment. These data are also shown out of context in the comment. The original paper shows that the four pH groups (not shown in the comment) showed no significant differences from concurrent controls. The authors therefore concluded, on page 80 of the paper, that despite the single datapoint of 1.9 (which apparently contained one spurious replicate), there was no significant influence of pH on the intramuscular injections of the metals studied.
Comment: If this study were to be the primary basis for addressing nickel hazards at the 100 to 250 m g/m3 range, as OEHHA suggests, then there ought to be considerable doubt whether the nickel sulfate or oxides more typically emitted by power plants pose the same risk as the nickel chloride.
Response: It is true that nickel oxide was not shown by Graham et al. (1978) to have an immunosuppressive effect when given intramuscularly. Nickel oxide was not tested by inhalation by Graham and colleagues. Since no inhalation immunotoxicity data for nickel oxide or other insoluble nickel compounds were collected, there is insufficient evidence to categorize nickel oxide separately from other nickel compounds. Nickel sulfate showed distinct suppression of the antibody response at all levels tested when animals were exposed to nickel intramuscularly. In addition, mice exposed to nickel chloride or nickel sulfate by inhalation were significantly more susceptible to streptococcal infection than controls (Adkins et al., 1979). This finding further strengthens the hypothesis that inhaled nickel produces biologically significant suppression of pulmonary defenses. Nickel oxide was not tested in the Adkins et al. report.
Comment: Furthermore, by definition, insoluble nickel is less likely to be rapidly absorbed by the body, and any acute hazard exposure effect is more likely to be dependent upon soluble nickel exposures than upon insoluble nickel exposures. If insoluble nickel causes adverse effects, that is more likely to be a chronic effect, where the longer term presence of insoluble nickel compounds in the body might be a compensating factor for the poorer body chemistry access to those compounds. We therefore suggest that OEHHA limit its acute nickel hazard REL to soluble nickel. Such a factor would apply to total nickel where sources did not distinguish in their emission inventories between soluble and insoluble nickel emitted. But sources should be able to collect such data, and reduce calculated acute hazards accordingly.
Response: OEHHA agrees that the theories presented in the comment are plausible. However, since there are no actual data to evaluate immunotoxicity separately for each nickel compound, OEHHA cannot develop separate RELs for speciated nickel compounds at the present time.
Comment: Historically, workers have been exposed to peak and chronic soluble and total nickel exposures far in excess of the proposed REL. For example, the Report of the International Committee on Nickel Carcinogenesis in Man, February 1990, reported on page 24 that the medium nickel concentration from 3044 air samples was 130 m g/m3 in Oak Ridge location, and that about 70% of the Oak Ridge workforce were in two other locations where nickel concentrations exceeded 1,400 to 1,800 m g/m3 in at least 10% of the measurements. On page 25, it was noted that there were areas where average nickel concentrations exceeded 10,000 m g/m3 over working lifetimes. With all of those historically high human exposures, OEHHA has made no recommendation for exposures causing lethal or serious effects, and only cites complaints of irritation and headaches after four weeks of exposure in the 70 to 1100 m g/m3 range, (mean = 440 m g/m3 ), and skin sensitivity at a level not stated.
Response: OEHHA thanks the commentator for the review of workplace exposure levels for nickel compounds at Oak Ridge. The data mentioned by the commentator describe health effects that are more severe than those addressed by the RELs and may be useful for the determination of the severe adverse and life-threatening effects.
Comment: As indicated on the attached spreadsheet, it is possible for relatively small and common oil combustion sources to have total calculated acute hazard indices approaching 20% to 50% of the 1.6 m g/m3 level. At that level, ARB guidelines suggest T-BACT should be required (particulate control for all new diesel engines?). Furthermore, several such distillate oil sources located together, or just one residual oil source with a small stack and close neighbors, could easily cause the calculated acute hazard index to approach or exceed one, at which point most risk managers require notification of impacted neighbors, denial of new or modified source permits, and shutdown of existing sources unable to come into compliance within five to ten years. All this for impacts that remain far below any lowest observed effect level. A review of the assumed factor of safety for nickel may be more appropriate than the control, prohibition or shutdown of all oil combustion sources. If OEHHA does not wish to undertake such review, we suggest that OEHHA at least provide for the possibility that the appropriate risk management agencies might prefer to gather information on what effect an alternative factor of safety assumption for nickel might have on the calculated acute hazard index, before deciding whether to actually require such notification, control, prohibition, or shutdown based solely upon the OEHHA factor of safety recommendation.
Response: The RELs in this document are health-based values derived from the most sensitive endpoints reported in the scientific literature. Considerations of mitigation measures requiring expenditures by industries are risk management decisions, in which OEHHA is not involved.
After reviewing the literature, OEHHA has calculated an acute REL for nickel of 3.3 m g/m3 based on the acute inhalation study conducted by Cirla et al. (1985) in asthmatics. Although the animal immunotoxicity data are highly suggestive of acute human health consequences for nickel exposures, the data collected by Cirla et al. contain considerably less uncertainty and are based on effects in sensitive individuals.
Comment: Factors of safety of 1000 relative to reversible effects of no great significance, such as mild eye irritation for acrolein, are unjustified.
Response: OEHHA considers eye irritation to be an adverse effect. Eye irritation is a mild adverse effect, unless it is of a severe or irreversible nature, in which case it is considered a severe adverse effect. Since a NOAEL was not identified in the human subjects in the key reference, an uncertainty factor of 10 was originally applied to the LOAEL for estimation of a NOAEL. As described elsewhere in the responses, an uncertainty factor of 3 is now being applied when adjusting a LOAEL to a NOAEL for mild irritation. In addition, since individual variation exists in the human population, an uncertainty factor of 10 was applied. Therefore, the total uncertainty factor applied to the acrolein REL is 30. The duration of the study was only for 5 minutes, therefore a 1-hour value was estimated using a time-adjustment. The REL therefore changed from 1.2 x 10-1 to 3.6 x 10-1 µg/m³.
Adkins B, Richards JH, Gardner DE. Enhancement of experimental respiratory infection following nickel inhalation. Environ Res 1979;20:33-42.
Dourson M, Stara JF. Regulatory history and experimental support of uncertainty (safety) factors. Regul Toxicol Pharmacol 1983; 3:224-238.
Graham JA, Miller FJ, Daniels MJ, Payne EA, Gardner DE. Influence of cadmium, nickel, chromium on primary immunity in mice. Environ Res 1978;16:77-87.
National Research Council (NRC). Committee on Toxicology. Guidelines for developing community emergency exposure levels for hazardous substances. Washington (DC): National Academy Press; 1993.
Comment: The guidelines do not seem to allow an assessor to offer an alternative to any of the RELs that have been promulgated by OEHHA. The spirit of SB-1732 would seem to be served better by allowing an interested party to offer an alternative assessment based on newer information than used in OEHHAs development of the REL or using newer methods to interpret old information. While OEHHA has expressed its intention to update the RELs as such new information or methods become available, its incentive to do so in a timely fashion is less than that for an interested party.
Response: OEHHA welcomes submission of new pertinent data or methodologies affecting the generation of RELs. Confusion and an uneven playing field would result if each assessor were to submit risk assessments based on different RELs. In addition, OEHHAs RELs undergo public and peer review. For these reasons, issues regarding the derivation of the RELs should be resolved through the public comment process and through review by the ARBs Scientific Review Panel.
Comment: The basic approach for establishing RELs is to apply uncertainty and modifying factors either to benchmark doses or to NOAELs/LOAELs. Both depend on defining what constitutes an "adverse" effect. The language in the Technical Support Document appears to be inconsistent in this regard. For example, Figure 2 on page 11 describes Level 1 as: The level at or below which no adverse effects are expected. Exposure at or below this level may be perceived by mild irritation of the eyes, nose, or throat, or by unpleasant odors, tastes or sight. Other changes of uncertain physiologic significance may also be observed.
This language, and that in the corresponding listing on page 10, seems to imply that mild irritation, sensory clues, and physiologic changes without clear clinical significance would not be considered adverse for the purposes of defining a NOAEL. Yet in Table 1 (on page 9) mild irritation, unpleasant sensations, and findings of statistical significance but uncertain clinical significance are all shown to occur above Level 1, which is described as "discomfort." In its development of the RELs, OEHHA has apparently used the more restrictive definition of adverse. Because of the mechanical interpretation of state guidelines by some local air districts, I fear that the occurrence of "unpleasant odors" on one day per year at a location that may not be inhabited by anyone at that time may be interpreted as equivalent to a chronic exposure to annual averaged concentrations that might cause a serious neurologic condition or cancer. In other words, using such a restrictive definition of "adverse" for acute toxicity may trivialize
Response: Unfortunately, the draft incorrectly described the Level 1. The definitions for the severity effect levels have been corrected in the revised draft. We have tried to define the acute, adverse, non-cancer health effects such that local Air Districts and other risk managers can calculate hazard indices and prioritize risks accordingly. The toxicological endpoint is clearly stated for each acute REL, and for each hazard quotient calculated in any risk assessment.
Comment: The procedure for adjusting observed dose-response relationships to one hour when data were obtained for a different exposure duration uses default procedures that are "conservative" for durations either longer or shorter than one hour. While this procedure may be appropriate for substances with only one duration represented by the available data, it could be excessively conservative if data for durations bracketing one hour are available, because the procedure for durations longer than one hour could underestimate the NOAELs for durations under one hour and vice versa. It is not clear whether OEHHA would
Response: If multiple dose/duration data are available for the same endpoint and species, we have derived the exponential term for the modified Habers relationship (for example, see the ammonia summary). OEHHA welcomes the submission of analyses of new data sets that may help pinpoint this exponent for specific chemicals.
Comment: The discussion of exposure assessment and risk characterization in the guidelines states that 1-hour maximum concentrations for each substance should be calculated for each substance and the hazard quotients and hazard indexes derived accordingly. There is a danger that a user could derive a hazard index based on maximum hazard quotients that occur at different places or at different times of year because of different locations or characteristics of the sources for different substances. While methods exist for overcoming this difficulty, OEHHA should be sure that the exposure assessment guidelines adequately cover this issue and that appropriate cross-references are made in the acute toxicity guidelines.
Response: This issue will be addressed in the upcoming Risk Assessment Guidelines.
Comment: Both Table 1 of the draft guidelines and Appendix B of the Technical Support Document show the toxicologic endpoint for hydrogen sulfide to be "respiratory irritation." The discussion of hydrogen sulfide in Appendix C of the Technical Support Document makes it clear, however, that the REL (which is also the California ambient air standard) is set at the odor detection level. Statements at the workshop confirmed that the REL for H2S is based on odor. The listing in the Tables is probably an artifact of the old CAPCOA guidelines, which erroneously implied that H2S is a respiratory irritant at levels only slightly above the REL. The listing should be changed because users may erroneously calculate the hazard index for respiratory irritation to include the hazard quotient for hydrogen sulfide.
Response: The comment raises a valid and difficult issue. The 1-hour CAAQS values were uniformly adopted by OEHHA for use as 1-hour Level I RELs. This was the case for H2S. The comment is correct that the CAAQS is based on odor perception. An REL based on bronchial obstruction in asthmatics exposed to 2 ppm H2S for 30 minutes (Jappinen et al., 1990) would be 0.1 ppm. However, because the CAAQS is not based strictly on respiratory irritation or other adverse effects that can be summed in a hazard quotient, OEHHA is changing the REL for H2S based on the bronchial reactivity observed in the Jappinen et al. (1990) study. The new 1-hour REL is 0.1 ppm (140 m g/m3). Thus the REL has changed from 42 µg/m³ to 140 µg/m³.