UNITED STATES ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
May 17, 1984
OF
THE ApMINlSTRATO**
Honorable William D. Ruckelshaus
Administrator
U.S. Environmental Protection Agency
401 M Street, S.W.
Washirgton, D.C. 20460
Dear Mr. Ruckelshaus!
Ihe Clean Air Scientific Advisory Committee (CASAC) .has completed its
review of two documents related to the development of revised primary
National Ambient Air Quality Standards (ISt&AOS} for Carbon Monoxide (CO).
Ihe documents were the Revised Evaluation of Health Effects Associated with
Carbon Monoxide jjacposureV An''Addendum jto the 1979 Air Quality Criteria ' ' f
Document for Carbon Monoxide written by the staff of the Office of Research
and Development (QRD), and a staff paper entitled Review of the NftAQS _for • •"
Carbon Monoxide! 1983 Reassessment of Scientific and Technical Information
prepared by the Office of Air Quality Planning and Standards (QAQES). The
Cownittee unanimously concluded that both documents represent a scientifically
balanced and defensible summary of the current basis of our knowledge of
the health effects literature for this pollutant.
As you know, the latest CASAC review of the CO documents took^lace in
an atmosphere of great scientific uncertainty and controversy due to the
fact that a group of scientists conducting a review of the protocols for a major
series of peer reviewed studies, carried^out by Dr. Wilbert Aronow> had shortly
before concluded that adequate standardized procedures for scientific
research were not utilized in those studies, Confronted with this situation,
Agency staff in both QRD and QAQPS moved quickly and resolutely to analyze
the remaining scientific basis for the Clean Air Act requirement to finalize
a revised CO standard. The CASAC concludes that, even without the use of
the Aronow studies to determine a critical effects level from CO exposures,
there remains a sufficient and scientifically adequate basis on which to
finalize the CO standard,
As a result of its review of the information contained in these docu-
ments, the CASAC recommends that you consider choosing the 8-hour and
1-hour carbon monoxide standards to maintain approximately current levels
of protection. A wore extended analysis of the factors that led to this
recommendation is contained in the enclosed report.
-------
lhank you for the cpportunity to present the Conatdttee*s views on this
important public health issue.
Enclosure
cc: Mr. Alvin Aim
Mr, Joseph Cannon
Dr. Bernard Goldstein
Dr. iferry Yosie
Sincerely,
^
,
Morton
Clean Air Scientific
Advisory Ccramittee
-------
CASAC Findings and IteccnTOsndations on the Scientific Basis for
a Revised NAAQS for Carbon Monoxide
Addendum to the CD Air Quality Criteria Document
lt A key issue in the evaluation of public health risks from carbon
monoxide (CO) exposures concerns the relation between CO in air and its
/
displacement of oxygen,in blood hemoglobin. The index for this displacement,
known as carbcxyhemoglobin (COHb), is expressed as a percentage of the
blood hemoglobin. There is a scientific consensus that relatively low
levels of COHb are associated with critical (i.e., health impairing) health
effects. The discussion of the scientific evidence thus centers on what
percentage of CQHb causes a critical effect.
CM October 9, 1979, CASAC submitted a report to the Administrator
concluding that the critical CQHb level occurred within a range of 2.7—3.0%.
The Committee reached this finding following an extensive review of the
scientific literature, including a series of studies performed by Dr. Milbert
Aronow. CAS&C expressed some reservations about one of these studies
{Aronow, 1978 which reported effects at levels 11*811 well below the 2.7-
3.0% range) in view of the fact that some**confounding factors in the study
protocols vere not appropriately accounted for. The Ccnraittee further
recoroended that "given the uncertainties stemming from the methodological
approach, [the Agency]...should utilize the [1978 Ironow] study for matgin
of safety considerations rather than using it for the determination of a
threshold value" (CASAC report, October 9, 1979, p.5). On August 31,
1982 CASAC sent a follow-up report on several issues related to the NAAQS
for carbon rcranoxide. In that report the Ccmtnittee reaffirmed its prior
findings on the critical COHb effects level. It should be noted that
1 s 1982 recownendations were reached after the Committee members
-------
had an opportunity to review an additional (1981) study by Dr. Aronow which
concluded that a 10% reduction in the time to onset of an angina attack
occurred during treadmill exercise with 2% ODHb.
A review of the most recent update of this scientific literature in
f
the August 1983 draft EPA Addendum to the CD Mr Quality Criteria Document
persuades CAS&C that there is no significant reason bo substantively alter
its previous findings. An elaboration of CASAC's current reasoning on
several issues will clarify the Committee's position. These include:
A, lhe role of the Aronow studies
A key question raised about Aronow's work was whether or not the
procedures used insured that the studies were double blind. A. double /
blind protocol is one in which neither the subjects nor the laboratory
technicians conducting the experiments and collecting the data are aware of
key parameters of the study (exposure conditions, timing, etc.) and the
results of the responses by the experimental group and the-control jproup.
It is apparent that such double blind procedures were not applied in Aronow's
work because technicians who were directly involved with the subjects knew **
sane ot the important parameters of the study. The lack of quality assurance
checks represents another issue of concern. In these respects, the results
of Aronow's work do not meet a reasonable standard of scientific quality
for a study of the kinds of responses of interest, and thereforer they should
not be used by the Agency in defining the critical COHb level,
B» lhe role of the Anderson study
The 1973 study by Anderson et alt reported that angina patients exposed
to low CO levels while at rest experienced a statistically significant reduction
in time to onset of exercise induced angina at average COHb levels of 2,9% and
4.5%. Hie study further concluded that there was a significantly lengthened
-------
angina attack during exercise at an average COHb level of 4,5%. The 1983
CD criteria document addendum noted concerns expressed by some parties
about the study due to the small number of subjects studied, apparent
inconsistencies between predicted and observed COHb levels, the possibility
that the protocols were not -truly double blind, and the lack of subsequent
confirmatory findings,
/
CASAC reached several conclusions concerning this study. It was trou-
bled that so few patients were included in the study design and that there was
uncertainty about the exposures to which the patients were subjected. The
Committee agreed that it is important to replicate such a study, but the notion
that a study has no validity until it's been replicated is flawed. Based
upon its current knowledge of how the study was conducted, CASAC presumes
that double blind protocols were, in fact, observed and that discrepancies
between observed and predicted CQHb levels are not as great or as serious
as originally suggested. In summary, while CASAC treats the Anderson et
al, study with caution, it can find no substantive reason at this tiros to
dispute the reported values, and it recotmsnds that the Agency not disregard ~
its findings.
C. Additional studies
CASAC wishes to point out two sets of additional studies which lend
support to concerns about low level CO exposures. In 1974, both Raven et al.
and Drinkwater et al* reported statistically significant decreases (less
than 5%) in exercise time for work capacity in healthy, nonsmoking young
and middle aged men at approximately 2.3 - 2,8% COHb. Also, a 1980 controlled
human exposure study by navies & 3nith observed changes in electrocardiogram
(EKG) measurements in a small number of healthy nonsmoking young men at
2.4% QOHb, Such CO induced changes are a cause for public health concern
and should be factored into the Agency's thinking for setting a standard
with an adequate nergin of safety.
-------
D. Use of th<* Coburn-Fo$ter-Kane (CFK) equation
The CFK wxfel is the most important available tool for analyzing a
number of physiologically important variables (blood volume and endogenous
00 production rate, for example) in order to project a relationship
between ambient 03 exposures and resulting GQHb levels* Wiile this
model, like any model f is subject to the need for additional evaluation
of CQHb in different population groups, it is reasonable to conclude that
the CFK equation accurately predicts CO uptake under differing exposure
conditions.
E. Summary of cardiovascular effects
The Committee unanimously agrees that: 1) the key mechanism of CO toxicity
iTS
is the decreased oxygen carrying capacity resulting from the greater af-
finity of blood hemoglobin for carbon ironoxide than for oxygen; 2) reduction
in time to the onset of an angina attack is a medically significant event
and should be considered an adverse health effect? and 3} following a
review of the peer reviewed scientific literature (not including eRe Aronow
-------
literature leads to the conclusion that there is not a sufficient scientific
basis to establish a connection between a CO exposure level and BIDS*
QftOPS Staff Paper Rgvigwj3_f_the___NAAQS, For Carton Mpnoxiae ^
Based upon the addendum to the revised Air Duality Criteria Document
for Carbon Monoxide, OAOPR developed a staff paper analyzing alternative
ranges of concentration levels for a final promulgated standard. The current suite
of primary standards is set at 9 parts per million (ppra) for the 8-hour averaging
tiine and 35 ppm for the 1-hour average.
CASAC was asked to advise the Agency on several issues associated with ,*,
the proposed ranges. The following discussion responds to the Agency request.
1. CASAC reaffirms the judgment it reached in its October 1979 report
s
that reduction in the time to onset of angina aggravation represents
an adverse health effect.
2. The Committee concurs with the Agency that R-hour and 1-hour
standards'are the appropriate averaging times, but it recommends
" "* -1
that there be additional discussion and more explicit comparison
in the regulatory package concerning the relationship between the
two averaging times, particularly in terms of what attainment of
the 8-hour standard portends for the health protection provided by
the 1-hour standard,
3. The factors identified by OftOPS for margin of safety
consideration are appropriate. Underlying CASAC's view of
the margin of safety, however, is its traditional belief that
-------
where the scientific data, as in this case, are subject to large
uncertainties, it is desirable for the Administrator to consider a
greater margin of safety than the numerical values *of QOHb generated
by the Coburn equation migh't otherwise suggest.
4. The QftOPS staff recewends that the AdMnistrator'retain or
select an 8-hour primary standard in the range of 9 to 12 ppm.
With regard to the 1-hour primary standard, the staff reeomnends
that a selection be made within the range of 25 to 35 ppm. CASAC
concurs that the proposed ranges for both the 8-hour and 1-hour
primary standards are scientifically defensible. Given the uncer-
tainties within the scientific data base and Discussion of margin
of safety issues, the Committee recomnenrfs that you consider
choosing, standard limits that maintain approximately current
levels of protection.
------- |