World Health Organization Regional Office for Europe
' Copenhagen
I
Air Quality Guidelines
for Europe
WHO Reg ional Publications, E
uropean Series No. 23< 2
Air qualityguidelines for Europe
Copenhagen : WHO. Regional Office for Europe, 1987 WHO regional publications.Europeanseries ; No.23 ISBN 92-890-1114-9
Ai^/^^r^ny^sis — Air Pollution/Prcventionand Control
TheWorld Health Organization isa specializedagencyofthe United Nations with primaryresponsibiiity for international health matters andpublic health. Through this Organization, which was created in 1948, the he-ahli professions of some
160countries exchange their knowledge and experience with the aim of making possible the attainmentby allcitizensof the worldby the year 2000 of a level of health that will permit them to lead a socially andeconomically productivelife.
The WHO RegionalOffice for Europe is oneof six regional offices throughout the world, each withits own programme geared to the particular health problems of the countriesitserves.The European Regionhas32 active Member States,0 and is unique in that a large proportion of them are industrialized countrieswithhighly advancedmedical services. The Europeanprogramme thereforediffersfromthose of other regions in concentratingon the problems associated with industrial society. In itsstrategyfor attaining the goalof "health forall by the year 2000" the Regional Office isarranging its activitiesin threemainareas:promotion of lifestylesconducive to health; reductionof preventable conditions; and provision of care that is ade
quate, accessfole and acceptable to all.
TheRegion is also characterized by the large number of languagesspoken by its peoples, andthe resulting difficulties in disseminatinginformation to all whomay need it.The Regional Office publishes in four languages — English, French, German and Russian — and applications for rights of translation intoother languagesare mostwelcome.
a Ataama. Austro Belgium. Bdgana. C^cch(1siovakia, Denmark Finland, F^arce, German Demo
cratic R^ipulbJic. Federal Republic of Germany, Greece, Hungary. Iceland, Ireland, Israel. Italy. Luxem
bourg, Malta. Monaco. Netherlands, Norway. Poland. Portugal. Romania, San Marino, Spain, Sweden, Switzerland, Turkey, USSR. United Kingdom and Yugoslavia.
Air quality guidelines
for Europe
Coverdesign: G. Gudmundsson
World Health Organization Regional Office for Europe
Copenhagen
Air quality guidelines for Europe
WHO Regional Publications, European Series No. 23
ISBN 92 890 1114 9 ISSN 0378-2255
© World Health Organization 1987
Publications of the World Health Organization enjoy copyright protection in ac
cordance with the provisions of Protocol 2 of the Universal Copyright Convention.
For rights of reproduction or translation, in part or in toto, of publications issued by the WHO Regional Office for Europe application should be made to the Regional Office for Europe, Scherfigsvej 8, DK-2100 Copenhagen 0, Denmark. The Regional Office welcomes such applications.
The designations employed and the presentation of the material in this publi
cation do not imply the expression of any opinion whatsoever on the part of the Secretariat of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.
The mention of specific companies or of certain manufacturers’ products does not imply that they are endorsed or recommended by the World Health Organiz
ation in preference to others of a similar nature that are not mentioned. Errors and omissions excepted, the names of proprietary products are distinguished by initial capital letters.
The views expressed in this publication are those of the participants in the meetings and do not necessarily represent the decisions or the stated policy of the World Health Organization.
PRINTED IN DENMARK
CONTENTS
Page
Fore-word ... ... ix
Preface ... ... ... xi
PART I. GENERAL 1. Introduction . . ... 1
Nature of the guidelines ... . . . ... ... 3
Procedures used in establishing the guidelines ... 5
References . ... 6
2. Criteriausedinestablishing guideline values ... 7
Criteria common to carcinogens and noncarcinogens ... 7
Criteria for endpoints other than carcinogenicity ... 8
Criteria for carcinogenic endpoint ... . ... 12
Ecological effects ... 17
References . . . ... 18
3. Summary of theguidellnes... 20
Guideline values based on effects other than cancer ... 21
Guidelines based on carcinogenic effects ... 26
Guidelines based on ecological effects on vegetation ... 30
4. Use of the guidelines inprotecting publichealth ... 31
PART II. ORGANIC SUBSTANCES 5 . Acrylonitrile . . .. ... 33
6. Benzene ... ... ... 45 v
7. Carbon disulfide . . ... ... 59
8. 1,2-Dichloroethane . . . ... 69
9. Dichloromethane ... 81
10 . Formaldehyde ... . 91
11. Polynuclear aromatic hydrocarbons (PAH) ... 105
12 . Styrene .. . . ... 118
13 . Tetrachloroethylene ... 127
14. Toluene . .. . ... . . . .. . . .. . . . .. . . ... 137
15 . Trichloroethylene ... ... 148
16 . Vinyl chloride ... 158
PART III. INORGANIC SUBSTANCES 17 . Arsenic ... • 171
18. Asbestos ... 182
19. Cadmium ... 200
20 . Carbon monoxide . ... . ... ... ... ... 210
21. Chromium ... 221
22 . Hydrogen sulfide... 233
23. Lead . . ... . . ... 242
24. Manganese .. . .. .. . . ... ... ... 262
25. Mercury .. .. .. . ... 272
26. Nickel . . .. .. . . .. . ... 285
27. Nitrogendioxide . . . . ... 297
Vi
28. Ozoneand other photochemical oxidants ... 315
29. Radon . ... 327
30. Sulfur dioxide and particulate matter ... 338
31. Vanadium... ... 361
PART IV. EFFECTS OF INORGANIC SUBSTANCES ON VEGETATION 32. Theeffects of nitrogen on vegetation ... 373
33. Theeffects of ozone and otherphotochemical oxidants on vegetation ... 386
34. The effects of sulfur oxides on vegetation ... .... 394
Annex 1. Tobacco smoking ... 405
Annex 2. Participants at WHO air quality guideline meetings ... 411
vii
Forewcord
Following the successful introduction of WHO guidelines for drinking-water quality, the Government of the Netherlands approached the WHO Regional Office for Europe in 1983 to suggest that air quality guidelines should be developed, using the same general philosophy and approach. We realizedfrom the outset that this would be a difficult task and that, for example, sampling procedures were very much more complicated than in the case of water supplies. As was also the case for the drinking-water programme, we agreed that it would be inappropriate to try to formulate ‘‘standards”, these being for governments and regulating agencies to decide on in the context of prevailing exposure levels and environmental, social, economic and cultural conditions.
The approach has been to develop guideline values that in the opinion of the experts are appropriate for the safeguarding of public health and will guide national and local authorities in their risk management decisions. It is im
portant that the numerical values are taken in the context of the descriptive sections of the guidelines: in many cases only an order of magnitude can be given, based on available data., but this is more useful to public health officials and regulators than no figure at all. In some cases, it has not been found useful to give a guideline value, but instead to provide a risk estimate.
Great efforts have been made during recent years in many countries of the European Region to reduce air pollution, and the intense smogs that were frequently experienced in London and other large cities up until 30years ago no longer occur. Progress has, however, been uneven and in particular the burning of 'low-quality soft coals has caused increasing problems in some parts of Europe, especially when associated with atmospheric inversions.
In addition to the major pollutants such as sulfur dioxide and particulates, there is now increasing interest in the emission of small concentrations of potentially toxic inorganic and organic micropollutants.
The vast majority of Europeans spend over 75 % of their lives indoors and the guidelines appropriately include consideration of indoor air quality, though not of occupational exposure. Although this publication is primarily intended to cover public health considerations, it was, I think, a useful decision to include some ecotoxicological dimensions. Most certainly, those involved in environmental management and public health practices increasingly need to view the effects of pollution on man within the context of the health of the entire biosphere. This is particularly important in the case of air pollution, which transcends national frontiers and can have marked effects at great distances from the source.
It is hoped that the guidelines will have a wide application in environmental decision-making throughout the Region as well as in other parts of the world.
They were developed to a very demanding time scale and all those concerned in their preparation are aware that the publication does not represent any final judgement on the subject. In future editions new parameters will have to be’
addressed and the effect of combined exposures may have to be taken into account.
The farsightedness of the initiative by the Government of the Netherlands and the encouragement received from their representatives during the prep
aration of the guidelines are warmly acknowledged. The work involved great dedication and enthusiasm on the part of the 150 experts who participated in the various meetings and who were often under very great pressure to meet apparently impossible deadlines. It is appropriate that the unstinting efforts of the Editorial Consultation Group that met in Copenhagen in March 1986 should be specially recognized.
That the final product was delivered on. time is a great tribute to all concerned and I would like to thank the secretaries and other members o f the Regional Office staff for their great contribution and to mention in particular Dr Reiner Ttirck, project coordinator, and Dr Dinko Kello, project consultant, without whose combined qualities of scientific excellence, power ofpersuasion, tenacity and organizational ability, the work could not have been completed within an acceptable time frame.
J.I. Waddington Dire tor
Environmental Health Service
Preface
The World HealthOrganization has beenconcerned with airpollution and, in particular, its dangers to human health for 30years. In 1957, a WHO conference was held in Milan dealing with thepublic health aspectsof air pollution in Europe.In the same year a special reporton air pollutionbythe WHO Expert Committeeon Environmental Sanitation was published.
Since then,many activities have been undertaken covering different areas of thesubject, such asmeasurement techniques,the compilation ofair quality datafrom different urban areas in the worldand the effects ofair pollution on health. AManualonurban air qualitymanagementpublished in 1976, a Glossary on air pollution, published in 198(1. and a Manual on industrial aii■ pollutants publishedon behalf of WHOby Elsevier in 1983, have been part ofthe work ofthe Regional Office forEurope in this field.
With regard to air quality criteria andguides, thereport ofa WHO Expert Committee was published in 1972. Furthermore,theEnvironmental Health Criteria Programme, which started under thejoint sponsorship of WHO and UNEP and now comes under the trilateral responsibility of WHO, UNEP and ILO,within the Iramework of the International Programmeon ChemicalSafety, has sofar resulted in some60 environmental healthcriteria documents, asignificant number of which deal with air pollutants.
When the industrialized countries of the European Region began to establishenvironmental policies, it soonbecameapparent that a yardstick was needed for the evaluation of air' quality. Air quality standards or objectives for majorurban air pollutants were accordingly developed in many European countries. In this contevt,it was gradually necogni/ed that international cooperation would be needed in certain areas, one of them being thejoint assessment by scientists from variouscountries of the adverse effects of air pollutants. This perception may be demonstrated bya state ment in the WHO GlobalMedium-term Programme for the Promotion of Environmental Health, 1983:
With respect to environmental hazards assessment and control, much can be accomplished through international cooperation. The collection and assessment of information on human exposures to pollutants and on their effects on health is a very time-consuming and expensive effort, and very few countries, even in the industrialized world, have the necessary resources and expertise to do this alone.
The pooling of resources to conduct these assessments can provide the govern
ments with the necessary information which they need to take action.
It is obvious why the WHO Regional Office for Europe therefore felt fully justified inembarking on its projectof establishing air quality guide linesfor the Region. The recent successful development ofguidelines for drinking-waterqualitywas regarded as afurther incentive.
From the verybeginning of the project, it was clear that some basic principleswould haveto be followed.
• The guidelines should describe the lateststate of scientific knowledge.
They would have to be developed together in a short time period in order to guarantee this objective.
• Theinformation providedwould haveto be condensed, describing only the essential factors leading to the final conclusions. Tocreatea one- volume book on28 pollutants orgroups of pollutants, all encyclopaedic types of information would haveto be ignored.
• The description of scientific findings would have to be of a kind that would be .understandable to a broad and rather heterogeneous reader ship. The flow of the arguments would have to be clear.
• The rationale for theguidelinerecommendationsshould also contain a description of uncertainties in the evaluationprocess due to missing, inadequate orequivocal data. Any illusion should be avoided thatit is possibletocondense a very complex situation in reality down to asimple figure without making assumptions.In particular, questions ofsafety or protection factors, combinedeffectsandhigh-risk groups wouldhave to be discussedin thiscontext.
• Another important goalwouldbe tocreatea basic commonstructure for the description of pollutants and the rationale for guidelines, without deleting too much ofthe“handwriting” of specificauthors and working groups.
• Finally, it was a prerequisite that the draftguidelines would have to undergo several intensive reviews, giving everybodyinvolved thechance to look at the whole document at various stagesof development.
It can easily be understoodthattheachievementof these goalsrequired a high spirit ofcooperation among all the experts involved in this publi cation. It is for the reader to decide how successfully their task has been accomplished.
The term“guidelines” should be understood literally, meaning that the main objective is to provide guidance to those interested in air pollution problems as well as to those directly involved in airqualitymanagement.
The guidelines consider various toxic (carcinogenic and noncarcino- genic) substances, and for afew substances alsotheir ecological effects. No differentiation is made in termsofguidelines foroutdoor or indoor pol
lution. Theguidelinesaddressconcentration and exposure times regardless xii
of whether the exposure is inside or outside buildings. However, typical occupational exposures arenot the mainfocus of attention as theseguide lines relate to the general population.
For those compounds that were not reported to induce carcinogenic effects and on which data regarding carcinogenic effects were lacking or insufficient, a threshold assumption was made and guideline values were proposed. For carcinogenic substances,theguidelines provide an estimate of lifetime cancer risk arising from exposure to those substances.
The guidelines are intended to provide background information and guidance to governmentsin makingrisk management decisions, particularly in setting standards. Itshould be strongly emphasizedthat the guideline values arenot to beregarded asstandards in themselves. It is obvious that there isa variety of ways tocontrol air pollution and to protect our health and environment: it is advisableto use all the toolsavailable. Air quality guidelines alonemay not beveryeffectiveinthefight against pollution, but they shouldprove extremely useful inthe framework ofan overall environ
mental policy.
R. Tiirck
ProjectCoordinator AirQuality Guidelines
xiii
Part I
General
1
Introduction
Human beingsneed a continuoussupply offood, air and water to exist. The requirements for air and water arerelatively constant(10-20 m1 and 1-2litres per dayrespectively). Airandwaterare also used inindustrial processes of energy conversion, in manufacturing,and for the removalof wasteprod
ucts,some of whichmaybeinjurious to human health. In a comprehensive setof guidelines for drinking-water developed by WHO (1),guideline values were recommended for specified contaminants, a consistent process of assessmentbeing used.
The WHO Regional Office forEurope has subsequently developed the present air quality guidelines for the European Region. The task of develop
ing such guidelines is more difficult than thatofdrawing up drinking-water guidelines, since for air, unlike water, there is no centrally supplied and controlled source.The development of consistent rules for assessing 28 chemi
cal air contaminants also poseda challenge.
These air qualityguidelines should be seenas a contribution to thetarget on air pollution contained in ■ WHO's regional strategyfor health for all.
This target states that “by 1995, all people ofthe Region should be effec
tively protected against recognized health risks from air pollution” (2).
Accordingly, “theachievement of this targetwillrequirethe introduction of effective legislative, administrative and technical measuresfor thesurveil lance and control of both outdoor and indoor air pollution, in order to comply with criteria to safeguardhuman health” (2).
Various chemicals are emitted into the air from both natural and man-made (anthropogenic) sources. The quantities may range from hundredsto millions oftonnes annually. Natural air pollution stems from various biotic and abioticsources (e.g. plants, radiological decomposition, forest fires, volcanoesand other geothermal sources, emissions from land and water), leading to a natural background concentration that varies according to local sources or specific weatherconditions. Anthropogenic air pollution has existed at least since people learnedto use fire, but it has increased rapidly since industrialization began. Theincrease in air pollution as a consequence of the expanding use of fossil energysources and the growth inthe manufactureand use of chemicalshas beenaccompaniedby mounting public awareness of and concern about its detrimentaleffectson
1
2 AIR QUALITY GUIDELINES
health and the environment. Moreover, knowledge of thenature,quantity, physicochemical behaviour and effects of air pollutants has greatly in
creased in recent years. Nevertheless, more needs to be known. Certain aspects of the health effects ofair pollutants require further assessment;
these include newer scientific areas such as developmental toxicity. The proposed guideline values will undoubtedly be changed as future studies leadto new information.
The impact of air pollution is broad. In man, thepulmonarydeposition and absorption of inhaled chemicals can have direct consequences for health. However, public health can also be indirectly affected by the de position of airpollutants in plants, animals and the other environmental media, resulting in chemicalsentering the food chain or being. present in drinking-water and thereby constitutingadditional sources ofhuman ex posure. Furthermore, the direct effectsof airpollutants on plants, animals and soil can influence the structure and function of ecosystems, including their self-regulation ability, thereby affecting the quality of life.
Although in recent decadesmajorefforts have beenmadeto reduce air pollution, the situation in the European Region is still not satisfactory.
While air pollution has decreasedand peak concentrations have been re duced in many larger citiesand urbanareas,theoverallpollutionin terms of the amounts of pollutants released intotheatmosphere has oftenbeen only slightlyreduced or has remained unchanged, and concentrations haveeven increased insome areas and for some pollutants (2-5).
Many countries of the European Region encounter rather similar air pollution problems, partlybecause pollution sources are comparable, andin any case air pollutiondoes not respect national frontiers. The subject of the transboundary medium- and long-rangetransportof air pollution has re ceived increasing attention in Europe in recent years. International effortsto combatits consequences are underway, for instance withintheframework of theConvention on Long-range TransboundaryAir Pollution established bytheUnitedNations EconomicCommission for Europe (6).
The task of reducing levels ofexposure to air pollutantsis a complex one.
It begins with an analysis to determine which chemicals are presentin the air, at what levels, and whether these levels of exposure are hazardousto human health and the environment. It must then be decided whether an unacceptable risk is present. When a problem is identified, mitigation strategiesaredevelopedand implementedsoas to prevent excessive risk to publichealth in the most efficient way.
Analyses of air pollution problems are exceedingly complicated. Some are national in scope (e.g. definition of actual levels of exposure of the population, determination of acceptable risk, identification of the most efficient controlstrategies),while othersare ofa more basiccharacterand are applicable in all countries (e.g. analysis of the relationships between chemical exposure levels, dosesand theireffects). The latter form the basis of thepresent guidelines.
The most direct and important source of air pollution affecting the health of many people is tobacco smoke. Even thosewho do not smokemay inhale the smoke producedby others (“passive smoking”).Indoorpollution
INTRODUCTION 3 in general and occupationalexposure in particularalsocontribute substan
tially to overall human exposure: indoor concentrations of nitrogendioxide, carbon monoxide, respirable particulates, formaldehyde and radon are oftenhigherthan outdoor concentrations (7).
Outdoor air pollution can originate from a single pointsource which may affect only a relatively small area. More often, outdoor air pollution is caused by a mixtureof pollutants froma variety of diffuse sources, suchas traffic and heating, and from point sources. Finally, in addition to those emitted by local sources, pollutants transported over medium and long distancescontributefurther to the overall level of air pollution.
The relative contributionof emissionsourcestohum.an exposure to air pollution may vary according to regional and lifestyle factors. Although indoor airpollution will be of higherrelevancethan outdoorpollution as far as certain air pollutants are concerned,this does notdiminish the import ance ofoutdoor pollution. In terms of the amounts of substancesreleased, the latter isfarmoreimportant andmayhavedeleterious effectsonanimals, plantsand materials as well asadverse effects on human health.
Nature of the Guidelines
The primary aim of the air quality guidelines" is to provide a basis for protecting public health from adverse effects ofair pollution and for elim inating, or reducing to a minimum, those contaminants of air that are knownor likely to be hazardous to human healthandwellbeing.
The guidelines are intended to provide background information and guidance to governments in making risk management decisions, particu
larly in settingstandards, but their use is not restricted tothis. They also provide information for allwhodeal with air pollution. The guidelinesmay be usedin planningprocesses and various kinds ofmanagement decisionat community or regional level. When guidelinevalues are indicated,this does not necessarily mean that they must take theform of general countrywide standards, monitored by acomprehensive network of controlstations. In the caseof . someagents, guideline values may beof use mainly for carrying out local controlmeasures aroundpoint sources.
It should be emphasized that when airquality guideline values aregiven, these valuesare not standards in themselves. Before standards are adopted, theguideline values must be consideredinthecontextof prevailing exposure levels and environmental, social,economicand cultural conditions(1). In certaincircumstances theremay be valid reason to pursue policies whichwill result in pollutant concentrations above orbelow the guidelinevalues.
Ambientairpollutants can cause several significant effects which require attention: irritation, odour annoyance, acute and long-term toxic effects (including carcinogenic effects). Air quality guidelineseither indicate levels combined with exposure times at which no adverse effectisexpected con cerning noncarcinogenicendpoints, or they provide an ■estimate of lifetime
a Guidelines in thepresent context arenot ■ restricted to suggested numerical values, but also includeany kindof recommendation or guidance in the relevant field.
4 AIR QUALITY GUIDELINES
cancer risk arising from those substances which are provenhuman carcino gens or carcinogens with atleastlimitedevidenceofhuman carcinogenicity (seep. 12).
The guidelinesrepresent the current best scientific judgement, but there isaneed for periodic revision,since muchremains to be determined regard ingthe toxicity of air pollutants forhumans.
Itisbelieved that inhalation of an airpollutant in concentrations and for exposure times below a guideline value will not have adverse effects on health and, in thecase of odorouscompounds,will not create a nuisanceof indirect health significance (see definition of health, Constitution of the
World Health Organization).Compliance with recommendations regarding guideline valuesdoes not guarantee theabsoluteexclusionof effects at levels below such values. For example, highlysensitive groups especially impaired by concurrentdisease or otherphysiological limitations maybe affected at or near concentrationsreferred to in theguideline values. Healtheffects at or below guideline values canalso result from combined exposureto various chemicals orfrom exposure to the same chemical by multiple routes.
It is important to note that guidelines have been established for single chemicals. Chemicals, in mixture, can have additive, synergistic or antag onisticeffects; however, knowledge ofthese interactions isstillrudimentary.
With a few exceptions, such as the combined effectof sulfur dioxideand particulates, there is insufficient information at presenttoestablish guide linesfor mixtures. An adequate margin of safety shouldexist between the guideline values andconcentrations at which toxic effects will occur.
Risk estimates for carcinogens do not indicate a safe level; they are presented so that thecarcinogenic potencies ofdifferent carcinogens canbe comparedandan assessmentof overall risk made.
Although health effects werethe major consideration in establishingthe guidelines, ecologically based guidelines for preventingadverse effects on terrestrial vegetation were also considered and guideline values were recommended for a few substances. These ecological guidelines for veg
etation havebeenestablished because, inthe long term,only a healthytotal environment can guarantee human health and wellbeing(see p. 17). Eco
logical effects on species other than plants have not been discussed, since theyare outside thescope of thisbook.
The guidelines do not differentiate between indoor and outdoor ex posure (with the exception ofexposure to mercury) because, although the sites influence thetypeand concentrationof chemicals,they donot directly affect the basic exposure-effect relationship. Occupational exposure has been considered inthe evaluation process, but it wasnot a main focusof attention as these guidelinesrelate to the general population.However, it should be noted that occupational exposure may add to the effects of environmentalexposure.
The guidelines do not apply to very high short-term concentrations which mayresultfrom accidentsor naturaldisasters.
The health effectsoftobacco smokinghave ■not beenassessedhere, the carcinogeniceffects of smoking having recently been evaluated by IARC (8 and see Annex 1). Neither have the effects of air pollutants on climate
INTRODUCTION 5 been considered, since too many uncertainties remain to allow anevaluation of thepossibleadversehealthandenvironmental effects. However,possible changes of climatehaveto be investigated veryseriously by theappropriate bodies because their overall consequences, for example the “greenhouse effect”, may gobeyond direct adverse effects on humanhealth orecosystems.
Procedures used in Establishing the Guidelines
The first step in the process of establishing air quality guidelines was the selection of pollutants. Air pollutants of special environmentaland health significance to countries of the European Region were identified and selected on thebasis of the following criteria suggestedby a WHO Scientific Group(9):
(a) severityand frequency of observed orsuspected adverse effects on human health, where irreversible effects are of special concern;
(b) ubiquity and abundance of the agent in man’s environment, with emphasison air pollutants;
(c) environmental transformations or metabolic alterations, as these alterations may lead to the production of chemicals with greater toxic potential;
(d) persistence in the environment,particularlyif the pollutantwould resist environmental degradation andaccumulate in humans, the ■ environ
ment or food chains; and
(e) population exposed (size of exposed population and special groups at risk).
Other factorsaffecting the selection werethetimetable of the projectand the factthat onlythose substances could be considered for which sufficient documentation wasavailable(such as the WHOEnvironmental health cri teriadocuments). Onthebasis of these criteria, the following 28pollutants wereselected for evaluation.
Organic air pollutants Acrylonitrile Benzene Carbon disulfide 1,2-Dichloroethane Dichloromethane Formaldehyde Polynuclear aromatic
hydrocarbons (carcinogenic fraction)
Styrene
Tetrachloroethylene Toluene
Trichloroethylene Vinyl chloride
Inorganic air pollutants Arsenic
Asbestos Cadmium Carbon monoxide Chromium Hydrogen sulfide Lead
Manganese Mercury Nickel
Nitrogen oxides
Ozone/photochemical oxidants Particulate matter
Radon Sulfur oxides Vanadium
6 AIR QUALITY GUIDELINES
Alter a planning meeting in early 1984 that offered suggestions on content, format, workplan and timetables for the air quality guidelines project,aseries of nine meetings involving more than 1 30 expertstookplace toevaluate various air pollutants (see Annex 2).
Before the meeting of each workinggroup,scientific background docu ments were prepared as a basis for discussion and for establishing the guidelines.Aftereach meeting,a texton the individual pollutant or pollutant group was drafted on the basis of the amended background documents, incorporating the working group’s conclusions and recommendations. The drafts were then circulated to all participants of the meetings for their comments and corrections. Aneditorial consultation group of scientistswas thenconvened to review the documents for clarity of presentation,adequacy ofdescription of'the rationale supportingeach guideline and consistency in the application ofcriteria. Certain sections in which inconsistencies were noted were again submitted for review, whereupon the final draft was prepared and submittedfor extramural review; it wassent to the govern
ments of MemberStates of the Region, and to organizations and individuals engagedin air quality research or management. The process concludedwith a review in a final meeting,at which the recommendaiions andconclusions ofall the working groups were submittedforfinal appraisal.
References
1. Guidelines for drinking-water quality. Vol. I.Recommendations.Geneva, World Health Organization, 1984.
2. Targets for healthfor all. Copenhagen, WHO Regional Office for Europe, 1985.
3. United Nations Economic Commission for Europe. Airborne sulphur pol
lution: effects andcontrol. NewYork, United Nations, 1984 (Air Pol lution Series, No. 1).
4. International Register of Potentially Toxic Chemicals. List oj environ mentally dangerous chemical substances and processes of globalsignifi cance: scientific monographs. Geneva, United Nations Environment Programme, 1984(UNEP ReportNo. 2).
5. Pilot compendium of environmentaldata. Paris, Organisation forEcon
omic Co-operation and Development, 1984.
6. United NationsEconomicCommission for Europe*.Convention on Long- range Transboundary Air Pollution. International digest ofhealth legis lation- 30(4): 965(1979).
7. Estimating human exposure to air pollutants. GEMS: GlobalEnviron mental Monitoring System. Geneva,WorldHealth Organization, 1982 (WHO Offset Publication, No. 69).
8. Tobacco smoking.Lyon,InternationalAgency for Research onCancer, 1986 (IARC MonographsontheEvaluation of theCarcinogenicRiskof Chemicals to Humans,Vol. 38).
9. Background and purpose ofthe WHO environmental health criteria programme. In: Mercury. Geneva, World Health Organization, 1976 (Environmental Health Criteria, No. 1).
2
Criteria used in establishing guideline values
Relevant information on thepollutantswas carefully considered duringthe process of establishing guideline values. Ideally, guideline values should representconcentrationsofchemicalcompounds in airthat would not pose any hazard to the human population. However, the realistic assessment of human health hazards necessitates a distinction between absolutesafety and acceptable risk. To aim at achieving absolute safety, one wimld need a detailed knowledge of dose-responserelationshipsinindividuals in relation to all sources of exposure, the types of toxic effect elicited by specific pollutantsor theirmixtures, theexistence or nonexistence of “thresholds” for specified toxiceffects,the significance of interactions andthe variation"
in sensitivityand exposure levels within the human population. However, such comprehensive and conclusive data on environmental contaminants are not always available. Very often the relevant data are scarce and the quantitative relationships uncertain; scientific judgement and consensus therefore play an important role inestablishing acceptable levelsofpopu lation exposure. Value judgements are unavoidable,because terms such as
“adverse” and “sufficient evidence” arenotin themselves totally objective, their meaning'being based ongenerally agreedjudgements.
Although it may be accepted that ■ a ' certain risk can be tolerated or is simply unavoidable, the risk within a population may not be equally dis
tributed. There may be subpopulations which are atconsiderablyhigher risk from the same exposure. Therefore, groups at special risk in the general population must be takenspecifically into account in theriskmanagement process. Even if knowledge about groupswithspecificsensitivity is available, unknown factors may existthat change the risk inanunpredictable manner.
Criteria Common to Carcinogens and Noncarcinogens
Sources,levels and routes of exposure
Available data are provided on the current levels ol human exposure to pollutantsfrom all sources, including the air. Special attentionis given to at
mosphericconcentrations in urban andinnonpollutedruralareas andinthe indoor environment. Where appropriate,concern unions in the workplace are also indicated for comparison with enviionmental levels. To provide
7
8 AIR QUALITY GuiiDEl WES
information on the contribution from air in relation to all othersources, data on uptake by inhalation,ingestion from waterand food, and dermal contact are given where relevant. However, for mostchemicals, data on total human exposure ase lackingto sums extent.
Kinetics and metabolism
Availabledata on the toxicokinetics of distribution in humans and experi
mental animalsare indicated for inter- andintraspecies extrapolation, es
pecially toassessthe magnitude of body-burden from long-term, low-level exposuresand to characterize better themode of toxicaction. Data concern ing thedistribution of anagentinthebodyareimportantindetermining the molecular or tissue dose to target organs. High-to-low-dose andinterspecies extrapolations are more easily carried out using equivalent tissue doses.
Metabolites are mentioned, particularly ifthey are known or believed to exert a greater toxicpotential thanthe original agent. Additional data of interestinclude the rate of excretionand the biological half-life.
Criteria for Endpoints other than Carcinogenicity
For those compoundsreportedlywithoutcarcinogeniceffects (orfor which data on carcinogenicity were lacking or insufficient), the starting-point for the derivation of guideline values wastodefinethe lowest concentration at which effects are observed in humans, animals and plants. In doing so, an attempt was made to define a lowest-observed-adverse-effect level. The question whether thelowest-observed-effectlevel or the no-observed-effect level should be used instead is mainlyamatter of availability of data. Ifa series of data fixes the lowest-observed-effect level and the no-observed effect level, either of those levels might be used. The gap between, the lowest-observed-effectlevel andthe no-observed-effect level isamong the factors included in judgements concerningthe appropriate marginof pro
tection. However, a single, free-standing no-observed-effect level which is not defined in reference to a lowest-observed-effect level or a lowest- observed-adverse-effect level is not conclusive. Opinions on this subject differ, but the workingconsensus was that the levelofconcern in terms of human health ismore relatable to the lowest-observed-adverse-effectlevel;
this levelwas therefore used wheneverpossible. In the caseof irritant and sensory effectson humans, it is desirable where possible to determinethe no-observed-effect level.
On the basis of the evidence concerning adverse effects, judgements about theprotection factors (safety or uncertainty factors)needed to mini
mizehealth riskswere made. Averagingtimeswereincluded, since thetime of exposure is critical in determining toxicity. Criteria applied to each of these key factors are described below.
Criteria forselection of a lowest-observed-adver.se.effect level
The distinctionbetween adverseandnonadverse effects poses considerable difficulties (1). Any observable biological change may be considered an adverse effect under certain circumstances. The definition of an adverse
CRITERIA FOR GUIDELINE VALUES 9 effecthas been given as “any effect resultingin functional impairment and/or pathologicallesionsthatmay affectthe performanceofthewhole organism or which contributes to a reduced ability to respond to an additional challenge”.(2). Even with such a definition, a significant degree of subjec tivity and uncertainty remains. Ambient levels of major air pollutants frequently cause subtle -effects thatare typically detected only by sensitive methods. This makes it exceedingly difficult,if notimpossible, to achieve a broad consensusas to which effects are adverse. To resolvethis difficulty, data should be ranked in three categories.
1. Observations, even of potential health concern, which are single findings that havenot been verified byother groups. Becauseof thelackof verification byother investigators, such data could not readily be used as a basis for guidelinevalues. Theydo, however, indicate the-need for further research and may be considered in evaluating a margin of protection.
2. A lowest-observed-effect level: such a level is represented by data which have been supported by otherscientific information. When theresults are in a direction that might result inpathologicalchange, thereis. a higher degreeof health concern. Scientific judgement based on all availablehealth information is used to determinehow effects. in this category canbe used in determining the . pollutant level that is to be avoided so that excessive risk can be prevented.
3. A substantialchange inthe -directionof pathological effects: these findings have had a major influence on guideline considerations.
Criteria for selection ofprotection factors
In previousevaluationsby WHO, protection factors, usually called safety factors, have-been appliedto derive guidelines from accepted criteriafor adverseeffects on health (3,4). The rationale has been that sucha factor allowsfor avariety ofuncertainties,for example, about possibly undetected effectson particularlysensitivemembers of the population, synergistic effects of multiple exposures, and the adequacy of existing data. Traditionally, the safetyfactor has been used to allow for uncertainties inextrapolationfrom animalsto humansand from a small group ofindividuals to a large popu
lation (1).
Intheseguidelines, the terms “protection factor” and“margin of pro
tection” have been used in preference to “safety factor” or “margin of safety”because the word“safety” may convey to thepublicthe impression of absolute freedomfromrisk; this goes beyond what isintended by scientists whenthey refer to - safety factors.These factors are applied in guidelines for the protection of human health. They are not applied to ecological guide
lines, because they already include a kind of protectionfactor with regard to thevariety of species covered.
Awide range of factorsfor protecting human health is usedinthisbook, based on scientificjudgements concerningthe interplay of various criteria.
Thedecision process for developing protection-factors has been complex, involvingthe transformation of mainly nonquantitative informationintoa single number expressing the judgementof agroup of scientists.
10 AIR QUALITY GUIDELINES
Some of thefactors which are taken intoaccount indecidingthe margin of protection can be grouped under the heading ofscientific uncertainty.
Uncertainty occurs becauseof limitations intheextent orqualityof the data base. One can confidently set a lower margin of protection (i.e. use of a smaller number), when a large number ofhigh quality,mutually supportive scientificexperiments in differentlaboratoriesusing different approaches clearly demonstrate the dose-response, includinga lowest-observed-effect leveland a no-observed-effect level.Inreality, difficulties inherentinstudy ing air pollutants andthe failure to perform much neededand veryspecific research usually preclude this situation.
Where a protectionfactorwasadoptedintheairquality guidelines, the reasoning behind this factor is given in the scientific background infor mation. As previously mentioned, exceeding a guideline value with an incorporated protection factor does not necessarily mean that adverse effectswill result. However, therisk to public health will increase,particu larly in situationswherethe most sensitive population group is exposed to several pollutants simultaneously.Itis thereforenecessary to exercisesome kind of judgement regardingthe size of the protectionfactor.
Groups within a population responddifferently to pollutants(5). Indi viduals with pre-existing lung disease,forinstance, can be at higherrisk fromexposure to airpollutants than healthypeople. Differencesin response can be due to factors other than pre-existing health factors, such asage, sex, levelofexercisetaken, or to unknown factors. Thus, the population must be considered very heterogeneous in respect of response to air pollutants.
Existing information does notallowadequate assessment of the proportion of the population that has enhancedresponse. However, an estimate of even afew per cent ofthe total populationentailsa largenumber of people.
Effects observed in laboratory animals inthe absence of human studies generally require a larger protection factor, because humans maybe more susceptible than laboratory animal species.Negative data from human studies willtend to reduce the magnitude ofthe protection factor.Also of import
ance are thenature and reversibility of thereported effect. Apollutant level producingslight alterations in physiological parametersrequiresasmaller protectionfactor thanapollutantlevelproducing a clearly adverse effect.
Scientific judgement about protectionfactorswill also takeinto account the toxicologyof pollutants, including the type of metabolites formed, variability in metabolism orresponsein humans suggestinghypersusceptible groups, andthe likelihood thatthe compound or its metabolites willaccumulatein the body. It is also important to consider the exposure level used in health studies andto make appropriate conversions to environmental situations.
It is obvious, therefore,that diverse factors must be taken intoaccount in proposing amarginof protection. Theprotection factor cannot be assigned by a simple mathematical formula; it requires experience, wisdom and judgement.
Criteria for selection of averaging times
The development of toxicity is a complex function of the interaction between concentration and time of exposure. A chemical may cause acute,
CRITERIA FOR GUIDELINE VALUES 1 1 minor, reversible effects after brief exposure and irreversible or incapaci
tating effects after prolonged exposure. Our knowledge is usually insuf ficient to delineate theseconcentration-time interrelationships. Therefore, expert judgement must be applied, based on the weight of the evidence available (61. Generally, when short-term exposures leadto adverse effects.
short-term averaging times are recommended. The use of a long-term average under such condition- would, be misleading. since the typical pattern of repeated peak exposures is averaged over time and the risk managerhasdifficultyindeciding upon effectivestrategies. In other cases, exposure-iCvponse knowledge is sufficient to recommend a long-term average. Thisfrequently occursfor chemicals thataccumulate in the body over time, thereby resultinginadverseeffects. In such cases, the integral of exposurecan have more impact than the patternof peak exposure.
It should be noted that these averaging times are based on effects.
Therefore, ifthe guidelines are usedas a basis for regulation,the regulator needs to select the most appropriate and practical standards in relation to the guidelines,without necessarilyusingtheguidelines directly.
Asimilar situation occurs for effects on vegetation. Plants are generally damaged by short-term exposures to high concentration as well as by long-term exposures to low concentration.Therefore,both short- andlong term guidelines to protectplants are proposed.
Criteria for consideration of sensory effects
Some of the substances selected forevaluation have malodorous properties at concentrations far below (hose at which toxic clfeets occur. Although odour annoyance cannot beregarded asan adverse health effectin a strict sense,it _affects the quality of life (7). Therefore,odour threshold levels for such chemicals havebeen indicated where relevant and used as a basis for separateguideline values.
Forpracticalpurposes, the following aspectsand respective levelswere considered in the evaluationof sensoryeffects:
(a) intensity,where the detection thresholdlevelisdefined as the lower limit of the perceived intensity range (by convention the lowest concen
trationthat canbedetected in 50% ofthecasesin which it is present);
(b) quality, where therecognition threshold level isdefined as the lowest concentration at which the sensory effect, e.g. odour, can be recognized correctly in 50% ofthe cases;
(c) acceptability and annoyance, where the nuisance threshold level is defined as the concentration at which not more thana smallproportion of the population (less than 5%) experiences annoyance for a small part of the time (less than 2%); since annoyance will be influenced by a number of psychologicaland socioeconomic factors, a nuisancethresholdlevelcannot be defined on the basis ofconcentration alone.
12 AIR QUALITY GUIDELINES
Criteria for Carcinogenic Endpoint
Cancer risk assessmentis basically a two-stepprocedure, involving a quali
tative assessment of how likely it isthat an agent is a human carcinogen, and a quantitative assessment ofthecancerrate theagent is likely to cause at given levels anddurations of exposure (8).
Qualitative assessment of carcinogenicity
'The decision to consider a substance as a carcinogen is based on the qualitative evaluationof all available information on carcinogenicity, en
suring that the association is unlikely to be due to chance alone. Here the classification criteriaofthe International Agency for Research on Cancer have been applied (9). These classify chemicals for carcinogenicity in the following way.
GroupJ — Proven human carcinogens. Thiscategoryincludes chemi
cals or groups of chemicals for which there is sufficient evidence from epidemiological studies to support a causal association between the ex posure and vincet.
Group 2 — Probable human carcinogens. Thiscategory includes chemi cals and groups ofchemicals for which, at one* extreme, the evidence of human carcinogenicityis almost sufficient, andthose forwhich, atthe other extreme,it isinadequate. Toreflect this range,the category is dividedinto two subgroups according to higher (Group 2A) and lower (Group 2B) degrees of evidence.
Group 2A. This groupis usually used for chemicals forwhich there is at least limited evidence ofcarcinogenicity in humans and sufficient evi dence for carcinogenicity in animals.
Group 2B. Thisgroup is usuallyused for chemicals forwhich there is inadequate evidence of carcinogenicity in humans and sufficient evi dence of carcinogenicity in animals. In somecases the known chemical properties of a compound and the results of short-term tests have allowed its transfer from Group 3 toGroup 2B, or from Group 2B to Group 2A.
Group 3 — Unclassifiedchemicals. This group includes chemicals or groups of chemicals which cannot beclassified as to their carcinogenicity in humans.
Withregard to thisclassification scheme,it is likely that in the future some chemicalsin Group 3can be classified asnoncarcinogenic.
It was concluded that thequalitativeevaluation appliedbyI ARC should serveas the baselineinestablishing the air quality guidelines forcarcino
gens,and that the IARC categorization scheme should be used if no new divergent evidence was available. In this respect, it was decided that all chemicals categorized in Groups 1 and2A, i.e.proven human carcinogens
CRITERIA FOR GUIDELINE VALUES 13 and carcinogens with at least limited evidence ofhuman carcinogenicity, shouldbe treatedas human carcinogens, andguidelines should be formu
lated accordingly,indicating onlyrisk estimates. For chemicals classifiedin Group 2B (inadequate evidence inhumans, sufficientevidence inanimals) it was decided that, untilnew evidence appeared, guidelines wouldpoint out the carcinogeniceffects inlaboratory animals and citerisk estimates (ifsuch estimates couldbereasonablyobtained)inthe health risk evaluation part of thescientific background information.However, these risk estimates,based onanimaldataonly,would not be incorporated inthe guidelinerecommen dations because of various uncertainties in this connection. Guideline values based onnoncarcinogenic endpoints wouldbegiven for these pollutants.
Quantitative assessment of carcinogenic potency
The aim of risk assessment is to apply information available from very specific study situations (mainly occupational studies) to thegeneral popu
lation in order to calculate the possible risk to the latter. Therefore, quantitative risk assessment or, more specifically, dose-response assess ment generallyincludes the extrapolationof risk from relativelyhigh dose levels (characteristic of animal experiments or occupational exposures), where cancer responsescan be measured, to relatively low doselevels,which are ofconcern in environmental protection and where such risks are too small to bemeasured directly, eitherinanimalstudiesorin epidemiological studies (10).
The choice of the extrapolation model depends on the current under standing of the mechanisms of carcinogenesis (11). No single mathematical procedure canbe regardedasfully appropriate for low dose extrapolation.
Methods based on alinear,nonthreshold assumption have been usedat the international level (10,12)andthe national level (various health assessment documents producedby the US Environmental Protection Agency (EPA) and the NationalInstitute of Public Health, Netherlands) more frequently thanmodelswhich assume a safe or virtually safe threshold.
In these guidelinestherisk associated with lifetimeexposure to a certain concentration of a carcinogen in the air has generally been estimated by linear extrapolation and the carcinogenic potency expressed as the in
cremental unit risk estimate. The incremental unit risk estimate foran air pollutant is defined as “the additional lifetime cancer risk occurring in a hypothetical population in whichall individuals areexposedcontinuously from birth throughout their lifetimes to a concentration of 1/rg/m5 ofthe agent in the air they breathe” (13).
Calculationsexpressed in unit risk estimates provide theopportunity to compare the carcinogenic potency of different agents and can helpto set priorities inpollutioncontrolaccording to the existing exposure situation.
Byusingunit risk estimates, any reference to the “acceptability” of risk is avoided. The decision on the acceptability of a risk should be made by national authorities in the frameworkof risk management.
For those substances for which appropriate human studies are available, the method known as the “average relative risk model” (14) has been generally used andis therefore described in more detail below.
14 AIR QUALITY GUIDELINES
For animalcancer bioassaysseveral methods have been used to estimate theincrementalrisks.Two general approaches have beenproposed.Astrictly linearized estimate has been used by US EPA generally (11). Nonlinear relations have been proposed by others where either the concentration
tumour response was found experimentally or where metabolism is of limited capacity.Accordingly, risk estimates based on animal bioassays are considered separately.
Quantitative assessment ofcarcinogenicity based on human data
The quantitative assessmentusing the average relative risk model includes four steps: (a) selection of studies; (b) standardized description ofstudy results in terms ofrelative risk, exposure level andduration of exposure;
(c) extrapolation towards zerodose;and (d) applicationto a general (hypo
thetical)population.
First, a reliable human study must be identified, where theexposureof the study population can beestimatedand theexcess of cancer incidence' is statistically significant. Ifseveral studies exist,the best representative study should be selectedor several risk estimates evaluated.
Whena study is identified, the relative risk(R) as a measure of response must be calculated. It is important to note that the 95%confidence limits around the central value for the relative risk can be wide and should be specificallystated and evaluated. The relative risk is thenintroduced in the following formula (average relative risk model) which combines steps (c) and (d) and allows the unit lifetime risk(UR) (i.e. risk associated with a lifetime exposure to 1/tg/m’) tobe calculated:
PAR - 1) ur = _r____ L
x
where: Po = background lifetimerisk; thisis taken from age/cause-specific death or incidence rates found in national vital statistics tablesusing the life tablemethodology,or it is available from a matched control population
R = relative risk, being the ratio between the observed (O) and expected (E) number of cancer cases in the exposed popu
lation; the relativerisk is sometimes expressed as the stan
dardized mortality ratio SMR=(O/E) X 100
X = lifetime average exposure (standardized lifetime exposurefor the study population on a lifetime continuous exposure basis); in the case of occupational studies, X represents a conversion fromthe occupational 8-hour, 240-day exposure over a specific number of working years and can be cal
culatedas X= 8-hourTWA X 8/24 X 240/365 X(average exposure duration [in years])/(life expectancy [70 years]), whereTWA is the time-weighted average Q/g/m1).
CRITERIA FOR GUlDi-LlNl: VALUE s 15 It shouldbenotedthatthe ■ unitlifetime risk depends on Po (background lifetime risk), which is determined from national age-specific cancer inci denceormortality rates. Since theserates are also determined by exposures other than the one of interest and may vary from country to country, it follows thatthe UR may also vary from one country to another.
Necessary assumptions foraverage relative risk method
Before any attempt is made to 'assess the risk in the general population, numerous assumptions are needed at each phase of the risk assessment process to fill invarious gapsinthe underlying scientific database.There fore, as a first step in any given risk assessment, an attempt should be made to identify the majorassumptions that have to be made, indicating their probable consequences. These assumptions areas follows.
1. The response (measured, as relative risk) is somefunction ofcumulative dose or exposure.
2. There isno threshold dose_ for carcinogens.
Many stagesinthebasic mechanism of carcinogenesis are not yetknown or are onlypartlyunderstood. However, taking available scientific findings into consideration, severalscientific bodies (3,10,12,15.17) haveconcluded thattherei.s no scientific basis forassumingathreshold orno-effect level for chemical carcinogens. This view is based on the factthat most agents that cause cancer also cause irreversible damage to deoxyribonucleic acid (DNA). The assumption appliesfor all nonthreshold models.
3. Thelinear extrapolationof the dose-response curvetowards zerogives anupper-bound conservative estimate of the true risk . function if theunknown (true) dose-response cii/ve has a sigmoidal shape.
The scientificjustification for the use of a linear nonthreshold extra
polation model stems from several sources: the similarity between carcino genesis and mutagenesis as processes which both have DNA as target molecules; the strong evidence ofthe linearity ofdose-response relation
ships for mutagenesis; theevidence for the linearity of the DNA binding of chemical carcinogens in the liverandskin; the evidence for the linearity in thedose-response relationship in the initiation stageof the mouse 2-stage tumorigenesis model;and the rough consistencywith the linearity ofthe dose-response relationships for several epidemiologicalstudies (10). This assumption applies for all linear models.
4. There is constancy of the relative riskin the specific study situation.
In a strict sense, constancy of the relative risk means that the background age/cause-specific rate at any time is increased by a constant factor. The advantage ofthe average relative risk method isthat this needs tobe true only for theaverage.
Advantages of the method
The averagerelative risk method wasselected in preference to many other more sophisticated extrapolation models because it has several advmimgcs.
16 AIR QUALli'Y GUIDEl.lMS
the main one being'that it seems to be appropriate for a fairly large class of different carcinogens, as well asfor different human studies. This is possible becauseaveraging doses, i.e. averaging done over concentration and dur
ation ofexposure,give a reasonable measure ofexposurewhen dose rates are not constantin time. This may be illustrated bythe factthat the use of more sophisticated models (13,14,18.19)results in risk estimates very similar to those obtained by theaverage relative risk method.
Another advantageof the method is that the carcinogenic potency canbe calculated when estimates of the average level andduration of exposure are the only known parameters besides the relative risk. Furthermore, the method has theadvantageof being simple to apply, allowing non-expertsin the field of risk models to calculate a lifetime risk from exposure to the carcinogens.
Limitationsof the method
As pointedout earlier,the average relative risk method is basedon several assumptions which appear to be valid in a wide variety of situations.
However, there are specific situations in which the method cannot be recommended, mainly because the assumptions do not hold true.
The cumulative dose concept, forinstance, is inappropriate when the mechanism of the caraiweri suggests that it cannot produce cancer throughout all stages of the cancer development process. Also, specific toxicokinetic properties, such as ahigher excretionrate of a carcinogen at higher doses or a relatively lower production rate of carcinogenic metab
olitesat lower doses, may diminish the usefulness ofthe method in esti
mating cancerrisk.Furthermore, supralinearity ofthedose-response curve or irregular variationsin the relative risk over timewhich cannotbeelimi nated would reducethe value of themodel. However, evidence concerning these limitationseither does not exist or is still too preliminary to make the averagerelativerisk method inappropriate for carcinogensevaluated here.
A factor of uncertainty, rather than of methodological limitation, is that dataonpastexposure are nearlyalways incomplete (12,17). Although it is generally assumed thatinthemajority of studies the historical dose ratecan be determined within an order of magnitude, there are possibly greater uncertainties,even of more than two orders of magnitude, in some studies.
Intherisk assessment process itisof crucialimportance that this degree of uncertainty beclearly stated. This isoften done simply by citing upper and lowerlimits of riskestimates(12). Duration of exposureand theage- and time-dependence ofcancer caused by a particular substance are less un certain parameters, although the mechanisms of relationship are not sowell understood (8).
Riskestimates from animal cancer bioassays
Animal bioassays of chemicals provide important information onthe human risk of cancer fromexposure to chemicals. These data enhance our con fidence in assessing human cancer risksonthebasisof epidemiological data.
Several chemicals considered in this volume have been studied using animal cancer bioassays.The process is continuing and new information on
CRITERIA FOR GUIDELINE VALUES 17 the potential carcinogenicity of chemicals is rapidly appearing. Con
sequently, the status of chemicals isconstantly being reassessed.
During the preparation ofthis book, dichloromethane was classified by IARC as showingsufficient evidence ofcarcinogenicity inanimals, on the basis of studies in ratsand mice (20). Detailed studies of the kinetics of metabolism of dichloromethane have also recently been completed, indi cating that the capacity of mammals to metabolize dichloromethane is limited. Thus, the extent of metabolismofdichloromethane athighdoses wherecancerbioassays are conducted is less thanatlevelsof environmental exposure. The linearized cancer risk modelsmay therefore represent con siderable overestimates ofthe carcinogenicpotential of dichloromethanein humans at levels likely to occur in the environment. As with dichloro methane, there arealso considerable uncertainties in establishing human risk estimates derived from animal data for formaldehyde and 1,2- dichloroethane. Therefore, the significance of such estimates is still very problematic.
There is little doubt of the importance of animal bioassay data in reaching an informed decision on a chemical. Thecollection anduse of data suchas those on saturation mechanisms, absorption, depositionand meta
bolicpathways, aswell as on interaction with other chemicals,isimportant and shouldbecontinued. Regrettably, these datawere notavailable for the above-mentionedchemicals at the time of this evaluation of guidelines for air pollutants. Theprocess ofevaluating guidelines andtheimpact ofhuman exposure to these chemicals shouldcontinue and be revisedas new infor mation becomesavailable.
Interpretation of risk estimates
The risk estimates presentedin this book should not be regardedas being equivalent to the true cancer risk. Quantitative risk estimates canprovide policy-makerswithrough estimatesof riskwhich mayserve well as abasis for setting priorities, balancing risks and benefits, and establishing the degree of urgency ofpublic health problems among subpopulations in
advertently exposed to carcinogens(11).
Ecological Effects
The importance of taking an integrated view of both health andecological effectsin air quality managementwas recognized from the beginning ofthe project. Ecological effects may have a significant indirect influence on human health and wellbeing. For example, most ofthe major urban air pollutants are known to have adverse effects at low levels on plants, in
cluding food crops. Aconsultation groupwas thereforeconvenedto con sider ecological effects of sulfur oxides,nitrogen oxidesand ozone/photo- chemicaloxidants onterrestrialvegetation. These substancesare important both because of the high anthropogenicamounts produced andbecause of their wide distribution. They deserve specialattention because of significant adverse effects on ecological systems in concentrations far below those knownto be harmfulto humans.
