Commentary
EPA Can No Longer Ignore Risks Potentially Posed By Hours-Long PM Excursions
By Robert A. Michaels, PhD, CEP, RAM TRAC Corp.
Since the 1970s the federal Clean Air Act has provided an arsenal of regulatory tools to protect public health against risks potentially posed by airborne hazards. Risks posed by each “criteria pollutant” are managed via a National Ambient Air Quality Standard (NAAQS), consisting of a suite of time-weighted average concentration limits responsive to the mechanisms by which each NAAQS-regulated substance exerts its adverse effects. PM and ozone, for example, each pose different risks depending upon exposure duration. Responding to these durationdependent risks, EPA imposes regulatory limits on ozone’s annual, 24-hour, and one-hour average airborne concentration. This regulatory risk management strategy also should apply to airborne PM, indeed to any substance whose airborne concentration is regulated by a NAAQS, but the agency has shied away from regulating, or even considering regulating, airborne PM via a one-hour NAAQS limit.
Why has EPA been reticent to consider a one-hour PM NAAQS limit? The short answer is, for reasons which are no longer valid. The first place for EPA officially to consider regulating airborne PM via a one-hour NAAQS limit is in its PM Criteria Document, which is now in its second draft because of a court-mandated reconsideration of the PM NAAQS. The purpose of the PM Criteria Document is not to present widely accepted conclusions about PM. Rather, it is to assess “the latest scientific information on the effects of airborne particulate matter on the public health and welfare for use in the next periodic review of the National Ambient Air Quality Standard (NAAQS) for PM” (11). Thus, EPA’s reticence to consider the issue of PM excursions lasting minutes to hours can be neither explained nor justified by the fact that much of the published research is emerging or recently published, and has not
RISK POLICY REPORT - www.insideEPA.com - September 18, 2001 35
Commentary
penetrated yet into most scientists’ general understanding of PM dynamics.
EPA’s PM Criteria Document not only is deficient in failing to conclude that PM excursions are significant to public health. More egregiously, the Criteria Document fails even to re-evaluate the time frame(s) over which airborne PM exerts toxic effects. This is notable because a rich literature now overwhelmingly confirms the hypothesis, advanced in 1996 (4), that brief exposure to PM, over minutes to a few hours, can and does exert adverse effects upon human health. A small fraction of peer-reviewed publications supporting this hypothesis is cited in the Criteria Document (1, 2, 8, 10), but the majority of evidence, amounting to roughly 30 peer-reviewed publications, is not presented, and even the few citations addressing — all confirming — the hypothesis are dispersed in multiple chapters. As a result, the relevant The Criteria Document fails evidence never is collected in one section or subsection and, therefore, never
. is deployed toward addressing the short-term PM excursion issue. Other even to re-evaluate the time issues, such as time lags between PM exposure and discernible response, are frame(s) over which airborne addressed in issue-oriented subsections, but EPA has refrained from PM exerts toxic effects. pursuing this effective strategy with respect to the short-term PM excursion
issue. This strategy should be adopted in the next draft of the Criteria Document as an EPA official verbally had assured the author it would be in the current draft.
The effect of failing to address the short-term PM excursion issue in a focused manner, as in a Criteria Document subsection, is to preclude consideration of PM excursion implications for the form of the PM NAAQS to be proposed by the EPA Administrator. Specifically, the Administrator will be presented with a Staff Paper to be derived from the Criteria Document, in which PM risks are presented along with potential regulatory risk management options including, most notably, the possibility of limiting the one-hour average airborne PM2.5 concentration. The Administrator cannot exert discretion, however, and cannot make a technically supportable decision regarding this issue, if the Criteria Document fails to address it.
Several lines of evidence, whose detailed explication is beyond the scope of this Commentary, are relevant for consideration in a subsection addressing the PM excursion issue. These include animal studies, in vitro or mechanistic studies, and clinical studies. The clinical studies include at least 16, cited in 4, 6, 7, in which short-term PM exposure adversely affected numerous respiratory function parameters in volunteers, mostly asthmatics, and mostly as a result of exertion typical of daily life, simulated by exercise during the studies. The clinical studies, primarily addressing the issue of identifying which atmospheric PM constituents could cause adverse effects, never were interpreted by EPA in the context of potential effects of brief PM excursions, notwithstanding their relevance to this critical issue.
The possibility that PM excursions might exert significant effects upon public health was first raised on theoretical grounds (4, 5, 6, 7). Specifically, the suggestion was made that observed epidemiological associations of 24hour and annual PM levels with adverse public health parameters might occur simply because PM regulations required measurement of PM over these two averaging times. In theory, according to this hypothesis, PM could exert its toxic effects over multiple time frames, including time frames shorter than these two required regulatory averaging times. Indeed, coincidence of the regulatory time frames with the time frames of toxic mechanisms recognized subsequently would be fortuitous, and clearly unexpected. Yet, adverse clinical effects have been associated, if only weakly, with elevated PM over these reporting times, just as daily mortality in a war zone presumably would be associated with 24-hour ambient airborne concentrations of lead, especially if the measurement apparatus included bullets in air samples analyzed for lead. Does this mean that the time frame over which bullets injure or kill is 24 hours? Clearly, the answer is no.
The document cites numerous PM clinical studies to show that the theoretical possibility raised in citation 4 and addressed in subsequent publications actually had materialized in numerous controlled clinical settings (5, 6, 7). This firmly places PM excursion effects in the same category as other PM phenomena accepted by the scientific and regulatory communities based upon retrospective data re-evaluations, for example, the recent Health Effects Institute (HEI) re-evaluation of the Harvard Six Cities database. With respect to the PM excursion issue, the clinical study results were predicted by theory and confirmed by retrospective data re-evaluation. Together, the theoretical studies (4-7), the in vitro or mechanistic studies, the animal studies, and the clinical studies (cited in 4, 6, 7) were highly suggestive, motivating EPA to assign the PM excursion issue the agency’s highest research priority. What was missing then, but not now, was epidemiological confirmation.
After 1996, epidemiological studies began to appear, all supporting the hypothesis, and together confirming it (1, 2, 8-10). Citing Michaels (6), Delfino, ef al. were the first to investigate PM averaging time empirically based upon the premise that “recent experimental findings in humans indicate adverse respiratory effects from short-term exposures to particulate air pollutants delivered over 1 to several hours” (1, page 751). As recently as June of this
36 RISK POLICY REPORT - www.insideEPA.com - September 18, 2001
Commentary
year, Peters, et al. concluded that “elevated concentrations of fine particles may transiently increase the risk [of myocardial infarctions or heart attacks] for several hours . . . after exposure ” (9; page 2,814). (Though details of the epidemiology studies are beyond the scope of this Commentary, relevant studies are cited for the benefit of interested readers.)
The literature now has all the elements in place to evaluate the PM excursion issue. Such excursions are routinely discerned and quantified via EPA-accepted PM monitoring instruments. The time is ripe for the Criteria Document to evaluate this issue, which is essential for NAAQS review in light of the purpose of the Criteria Document: to assess “the latest scientific information on the effects of airborne particulate matter on the public health and welfare for use in the next periodic review of the National Ambient Air Quality Standard (NAAQS) for PM” (11).
EPA has acknowledged the importance of the PM excursion issue.
In 1996, the scientific community was alerted to the PM excursion issue, and to at least 16 studies documenting PM excursion effects exerted on people in clinical settings. (Readers interested in the technical details are referred to citations 4, 6, and 7). In 1997, a letter in Science warned against uncritical attribution of PM effects to PM averaging times specified in the PM NAAQS (5). This warning was illustrated via the analogy Statements about PM of asking whether 24-hour average casualty figures ina war zone would excursion effects are be attributed to the 24-hour average concentration of lead in the atmosphere, versus brief exposure to lead in bullets. Accordingly, for the past . ; several years, EPA at least orally has acknowledged the potential signifi- incompatible. cance of PM excursions. Indeed, EPA included this issue among its highest research priorities and, further, invited the author to present a seminar on the issue at EPA’s facilities in Research Triangle Park, which took place in May of 1998. However, the logical and promised follow-through — consideration of the issue in the Criteria Document — so far has not occurred. Indeed, the second draft differs from the first in having removed relevant citations that had not been considered in the text, instead of retaining the citations and considering the studies in the text of the current draft of the Criteria Document.
disconnected and mutually
Statements on PM excursion effects are dispersed, ineffective, and inaccurate.
Statements about PM excursion effects are dispersed throughout the Criteria Document in Chapter 6 on epidemiology and Chapter 9, the Integrative Synthesis. However, the statements are disconnected, mutually incompatible, and generally presented only incidentally, that is, out of the context of a consideration of the PM excursion issue. For example, a study of PM excursions is mentioned, though unintelligibly for most readers, in Criteria Document Table 6-1 Short-Term Particulate Matter Effects Studies. Elsewhere in the Criteria Document, section 6.4.3.2 Fine- and Coarse-Particulate Matter Effects on Morbidity acknowledges that:
“Morgan et al. (1998) reported that PM2.5 ... yielded a PM2.5 association with [chronic obstructive pulmonary disease hospital admissions] for 1-hr max PM that was more positive than [for] 24-h average PM2.5” (page 6-236; see this comment document, citation 8).
The Morgan, et al. study (8) is also evaluated in Table 6-16, Acute Particulate Matter Cardiovascular Hospital Admission, but misinterpreted. The table states that:
“the PM effect disappeared when NO2 and O3 were added to the model... These results suggest that PM is not robustly associated with heart disease admissions when NO2 is included, similar to the sensitivity of PM to CO in other studies” (Australia section of Table 6-16).
This is an entirely unscientific evaluation. Any epidemiological association can appear to be invalidated by a search for potentially interacting variables. Even the well-known, robust association between flu virus exposure and flu, for example, can be invalidated by this method, when previous immunization against flu is “added to the model.” The apparent negative effect of NO2 and 03 upon adverse effects associated with PM may have many explanations. Maybe NO2 and/or 03 keep people indoors, or more inactive, thereby reducing their susceptibility to cardiovascular effects during days when PM also is elevated. Maybe NO2 is therapeutic for cardiovascular effects, as in the case of medicinal nitroglycerine. The Criteria Document offers no explanation of why NO2 might interact negatively with PM, and in no way undermines the association of PM with cardiovascular effects, especially with respect to the stronger one-hour PM association. Further, the one-hour PM and NO2 associations may refer to different, non-intersecting, non-interacting one-hour periods each day, except on days when the maximum one-hour PM and NO2 averages happen to coincide. Finally, in June, Peters, ef al. reported elevated incidence of myocardial infarctions within hours after PM excursions (9), confirming real-time PM effects on EKGs of dogs reported by Godleski, et al. (3). These confirmatory studies
RISK POLICY REPORT - www.InsideEPA.com - September 18, 2004 37
Commentary
are not cited and not considered by EPA.
Finally, in Chapter 9, the ‘Integrative Synthesis’ EPA fails to cite or consider most of these peer-reviewed reports. As a consequence, Section 9.6.2.3.2, Respiratory Effects of Ambient Particulate Matter Exposures, erroneously asserts that:
“Delfino et al. (1998) is unique in that they report larger effects for 1- and 8-h maximum PM10 than for the 24h mean” (p. 9-76; see this comment document, citation 2).
In light of the Morgan, et al. study (8), the Peters, et al. study (9), and the Simpson, et al. study (10); the Integrative Synthesis should conclude that the Delfino, et al. study (2) is consistent with other literature, not “unique” in concluding that excursion effects occur, and not “unique” in reporting that such excursions may exhibit stronger associations with health
The missing NAAQS limit on effects than longer PM averages. The Delfino, et al. study should also be the one-hour average rises cited in Chapter 6 on epidemiology if it is to be used in Chapter 9, the in importance above the Integrative Synthesis. Of course, the Peters, ef al. study should also be
: considered now that it is available. current vogue of fine-tuning The following quotation from Checkoway, et al. (2000; 1) provides the 24-hour average. a final indication of the inadequacy of EPA’s consideration of PM excursion effects.
“One important comment by the HEI Health Review Committee attached as part of the [Checkoway] report was that the induction time for the occurrence of primary arrhythmia may be much shorter than the 24-h time window studied and, as such, other hazard periods (i. ¢.; 1-, 2-, or 4-h concentrations) might be of more interest for evaluating PM effects on sudden cardiac arrest” (page 6-132; emphasis added).
EPA should add sections in Chapters 6 and 9 to address the PM excursion issue.
Taken together, EPA’s PM excursion statements are consistent with EPA’s expressed belief that PM excursion effects are potentially significant for human health. Yet, the Criteria Document fails to evaluate this issue, instead dispersing or ignoring the critical mass of information on the subject rather than focusing it on this important issue. Although the first three epidemiology studies reporting one-hour associations of PM with adverse health effects (2, 8, 10) do not explicitly address the potential significance of this time frame in the risk management context, the fourth epidemiology study (9) does, and other articles do (4-7), but EPA ignores them all.
To remedy this deficiency, the Criteria Document should add (sub)sections in Chapters 6 and 9 to address the short-term PM excursion issue. It should cite approximately 30 articles that augment the paucity of data dispersed through the current draft of the Criteria Document. Articles that EPA should cite include the peer-reviewed Science letter breaking the traditional link between the time frame over which PM epidemiology data are presented and the time frame over which PM toxic effects may be exerted (5); and other articles (4, 6, 7) which address the PM excursion issue by evaluating animal studies (such as 3), clinical studies, and mechanistic studies. EPA officials have indicated orally that they would follow through on the recommendations detailed in this comment document. Indeed, some of the articles cited above also were cited in the previous draft of the Criteria Document. Although they were not addressed in the text, EPA had assured the author that they would be addressed in the next draft. Well, the next draft now is the current draft. Instead of considering the cited articles, EPA now has removed them from citation lists, and ignored them in the present draft of the Criteria Document.
Over the past half decade, when the author has considered what is missing from the arsenal of public health defenses against airborne PM risks, the author has come to think about that missing NAAQS limit on the one-hour average airborne PM concentration. This omission rises in importance above the current vogue of fine-tuning the 24-hour average and the annual average PM NAAQS limits. Indeed, these latter NAAQS limits are equivocal, having evolved as the net result of conflicting opinions. Some have claimed a stronger PM effect, and the need for more stringent PM limits. Others, in contrast, have claimed that the effects which have been discerned statistically are hardly significant clinically, and might disappear from reality entirely when potential confounding variables such as weather-related parameters and co-occurring pollutants also are considered. The time has come for a credible review of both short-term excursions and these possible confounders.
LITERATURE CITED
1. Checkoway, H.; et al. A case-crossover analysis of fine particulate matter air pollution and out-of-hospital sudden cardiac arrest. Cambridge, Massachusetts: Health Effects Institute; Research Report 99, (http:// www.healtheffects.org/pubs-recent.htm), 2000;
2. Delfino, R. J.; Zeiger, R. S.; Seltzer, J. M.; and Street, D. H. Symptoms in pediatric asthmatics and air
38 RISK POLICY REPORT - www./nsideEPA.com - September 18, 2001
Commentary
pollution: differences in effects by symptom severity, anti-inflammatory medication use and particulate averaging time. Environmental Health Perspectives, 106(11):751-61, November 1998;
3. Godleski, J. J.; Nearing, B. D.; Killingsworth, C. R.; Murthy, G. G. K.; Skornik, W. A.; and Verrier, R. L. Increased cardiac vulnerability during exposure to inhaled environmental particles. Inhalation Toxicology, in press 1999a;
4. Michaels, R. A.. Airborne particle excursions contributing to daily average particle levels may be managed via a one-hour standard, with possible public health benefits. Aerosol Science and Technology, 25:437-44, November 1996;
5. Michaels, R. A.. Particulate matter policy. Science, 278:1,696 (letter); 5 December 1997;
6. Michaels, R. A.. Permissible daily airborne particle mass levels encompass brief excursions to the ‘London
fog’ range which may contribute to daily mortality and morbidity in communities. Applied Occupational and Environmental Hygiene, 13(6):385-94, June 1998;
7. Michaels, R. A.; and M. T. Kleinman. Incidence and apparent health-significance of brief airborne particle excursions. Aerosol Science and Technology, 32:93-105, February 2000;
8. Morgan, G.; Corbett, S.; Wlodarezyk, J. Air pollution and hospital admissions in Sydney, Australia, 1990 to 1994. American Journal of Public Health, 88(12):1,761-6; December 1998;
9. Peters, A.; et al. Increased particulate air pollution and the triggering of myocardial infarction. Circulation. 103:2,810-5; 12 June 2001;
10. Simpson, R. W.; Williams, G.; Petroeschevsky, A.; Morgan, G.; and Rutherford, S. Associations between outdoor air pollution and daily mortality in Brisbane, Australia. Archives of Environmental Health, 52(6):442-54, November/December 1997;
11. US EPA. Air Quality Criteria for Particulate Matter (External Review Draft). Federal Register (Notices), 66(71):18,929: 12 April 2001.
Publication citation
Michaels, RA. EPA can no longer ignore risks potentially posed by hours-long PM excursions. Commentary. Risk Policy Report, 8(9):35-9, 18 September 2001;