Hazardous Chemical Information System (HCIS)

Carbon monoxide

No standard should be applied without reference to Guidance on the interpretation of Workplace exposure standards for airborne contaminants.

^{Documentation notice:} National Occupational Health andSafety Commission documentation available for these values.

1. SOURCES OF CARBON MONOXIDE EXPOSURE

Occupational exposure to carbon monoxide (CO) occurs in a wide range of industries. Acolourless odourless gas with no inherent warning properties, it may occur whereverorganic or carbonaceous material is burnt in an inadequate supply of air or oxygen.

Industries and processes where CO is a potential hazard include furnaces, kilns andfoundries, catalytic cracking units in petroleum refineries, fire-fighting, and themanufacture of certain chemicals including formaldehyde and synthetic methanol. Carbonmonoxide is present in the exhaust gas of internal combustion engines, and motor vehiclesare responsible globally for the majority of human-made CO emissions. In the domesticenvironment, heating, including space and combustion heaters, is also an important sourceof CO.

Cigarette smokers absorb CO from the inhalation of cigarette smoke. Carbon monoxide isregarded as the commonest single cause of poisoning in both industry and the home

⁽¹⁾ .

2. METABOLISM

Carbon monoxide is absorbed readily through the lungs, and in the bloodstream itcombines reversibly with haemoglobin (Hb) to form carboxyhaemoglobin (COHb). Carbonmonoxide produces cellular hypoxia by several mechanisms. It binds to Hb about 200-240times more strongly than does oxygen. Carbon monoxide therefore directly reduces theoxygen-carrying capacity of the blood. Carbon monoxide also has the effect of loweringtissue oxygen tensions by shifting the oxyhaemoglobin dissociation curve to the left. Thatis, tissue oxygen tension must fall to lower levels, before oxygen will be released fromoxyhaemoglobin.

Carbon monoxide also binds strongly to myoglobin, in this case 40 times more stronglythan oxygen. This factor may be important in cardiac effects of CO, as during heavy work,cardiac oxygen extraction from myoglobin probably contributes significantly to satisfyingincreased oxygen demands. Enzyme systems, involved in cellular respiration and oxygentransport, for example cytochrome P-450, can be directly inhibited by the presence of CO,and this inhibition may be the major factor in CO toxicity

⁽²⁾.

The actual COHb level after inhalation of CO is dependent mainly on the following

^(3,4,5) :

(a) Concentration of CO in the air.

(b) Duration of exposure.

(c) Ventilation rate (in turn dependent on workload).

(d) Pre-inhalation COHb level.

Carboxyhaemoglobin levels in an exposed person rise with increasing duration andintensity of exposure. The time taken to reach equilibrium between CO in the blood and inambient air depends mainly on initial COHb level, concentration of CO in ambient air andventilation rate. Various equations have been developed which can be used to calculate thetheoretical value of COHb given a particular ambient air CO concentration and exposureduration.

Less is known about excretion of CO than about absorption. Carbon monoxide iseliminated nearly completely by the lungs, with a half-life in healthy people of about 4-5hours. The rate of excretion slows with time, and the lower the initial level, the sloweris the rate of excretion. Elimination probably occurs in two phases, a rapid initialexponential decline lasting 20-30 minutes, followed by a slower linear decline

⁽³⁾ .

3. MEASURING EXPOSURE TO CARBON MONOXIDE - BIOLOGICAL AND ATMOSPHERIC MONITORING

Exposure to carbon monoxide may be assessed either by measuring CO concentration inambient air (environmental monitoring) or by measuring COHb concentration in blood(biological monitoring). The COHb percentage in blood may be measured directly in bloodtaken by venepucture, or calculated indirectly from measurement of CO in expired air.

Accurate field and laboratory techniques exist for both blood and expired airmeasurement which show good correlation with each other

^(6,7).

The techniques for biological monitoring of COHb in blood were reviewed by Commins

⁽⁸⁾ . Techniques for monitoring of ambient air were reviewed byHarrison

⁽⁹⁾ .

In general, measurement of COHb level is an accurate measure of exposure to all sourcesof CO and compares favourably with atmospheric measurement. For this reason, bothbiological and atmospheric measurements, alone or together, appear as measures of exposurein studies of the effects of CO.

Some of the international and national standards also incorporate a biological exposurelimit as well as a standard for environmental air. In evaluating biological measures ofexposure to CO, it is important to remember that tobacco smoking is a source of CO andraises COHb levels.

4. DISTRIBUTION OF CARBON MONOXIDE IN THE ENVIRONMENT AND COHb IN THE GENERALPOPULATION

The Swedish carbon monoxide environmental criteria document

⁽⁵⁾gives a range of CO concentrations in air from various sources; these are summarisedbelow:

Sources of CO concentrations in air

Causes of a raised COHb other than occupational CO exposure include the following:

(a) Endogenous production, which increases in some diseases e.g haemolytic anaemia.

(b) Tobacco smoking.

(c) Exhaust emissions from motor vehicles.

(d) Exposure to the solvent methylene chloride.

Without occupational exposure, the following are given as typical saturation levels ofCOHb:

(i) Endogenous production - 0.4-0.7%, increasing to 4-6% in patients with haemolytic anaemia.

(ii) Tobacco smokers - cigarettes, 1 pack/day, average 5-6%; 2-3 packs/day, average 7-9%; cigars, peak up to 20%.

(iii) Commuters on urban highways (USA) - up to 5%.

(iv) Methylene chloride - 100 ppm for 8 hours, 3 to 5%.

5. HEALTH EFFECTS OF CARBON MONOXIDE

Carbon monoxide exerts its main toxic effect via its ability to interfere with oxygendelivery to the tissues. Hence the organs most vulnerable to the effects of CO are thosewith a high metabolic demand for oxygen.

Of these the heart, the central nervous system (CNS) and the foetus can be regarded asthe critical organs.

Although acute effects of high level CO exposure are important and well-documented,much of the research on CO has been directed towards investigation of effects at lowerlevels from both acute and chronic exposure.

These studies are of primary interest in considering the basis for occupationalstandards. The following review will therefore briefly cover acute high-level effects andwill be directed mainly at lower-level effects as observed in the critical organs - theheart, central nervous system and foetus.

5.1 Acute High-level Exposure to Carbon Monoxide

The acute effects of CO at high levels are generally known and well-documented, and areattributable mainly to depression of the central nervous system. High atmosphericconcentrations leading to COHb levels of 50-60% are likely to lead to unconsciousness andconvulsions. Collapse may occur very quickly, before the victim is aware of impendingdanger. A COHb level of 70-80% is probably incompatible with life

^(1,10).

At COHb levels in the range of 10-20% or lower, symptom complaints of headache, nauseaand dizziness may be expected

⁽¹⁾ . These symptoms are, ofcourse, non-specific and there is great variation in the individual reporting of symptoms,so that no great reliance can be placed on the absence of symptoms as an indication thatCOHb is not elevated.

Indeed, the inability of the human senses to detect CO contributes greatly to the acutehazard, as it is invisible, cannot be smelt and produces only variable and non-specificsymptoms. The following chart shows the principal signs and symptoms in relationship toCOHb levels (reproduced from Ref 1).

Principal signs and symptoms with various concentrations of carboxyhaemoglobin ^* (after International Labour Office

⁽¹⁾)

* There is considerable individual variation in the occurrence ofsymptoms.

5.2 Effects on the Foetus

Effects of CO on the foetus have been reviewed by Rylander and Vesterlund

⁽⁵⁾ , WHO

⁽³⁾ and, more recently,Evanoff and Rosenstock

⁽¹¹⁾ and Zielhius et al

⁽¹²⁾ .

Carbon monoxide crosses the placenta by simple diffusion. Foetal uptake of CO appearsto take place more slowly than maternal uptake. At equilibrium, levels in the foetus areslightly above those in the mother, due to particular characteristics of foetalhaemoglobin. The half-life for excretion is longer in the foetus, and is estimated to be 6to 7 hours, when pure air is breathed by the mother

^(5,13) .

Animal experiments on the foetal outcome of pregnant animals exposed to CO have beenconducted in at least five species. Most experiments have used rather high maternal COHblevels, i.e. ranging from 15 to 60%. At levels of COHb above 15%, which could occur in theworkplace without maternal symptoms, there have been consistent findings of foetal effectsin relation to both survival and development. Developmental effects have been found inrelation to malformations, low birth weight, and both behavioural effects and grosspathology of the central nervous system

⁽⁵⁾ .

At levels below 15% COHb, there have been few animal studies, and these have yieldedmixed results. Several animal models of continuous low-level exposure during pregnancyresulted in maternal COHb levels as low as 9-10% and provided evidence of lower birthweight and increased neonatal mortality

⁽¹¹⁾ . One study isreported on pregnant rats, showing adverse effects at around 5% COHb. After continuousexposure throughout the pregnancy to 30 ppm or 90 ppm (bioequivalent to 4.8% and 8.8%COHb), there was no effect on mortality or increase in malformations, but there was agreat reduction in the number of successful pregnancies (Garvey and Longo, cited in Ref5). The chart below summarises animal studies of effects of CO exposure on pregnancyoutcome

⁽⁵⁾ .

Foetal effects of maternal carbon monoxide exposure: animal experiments (afterRylander & Vesterlund

⁽⁵⁾ )

Rylander and Vesterlund

⁽⁵⁾ concluded that there wasinsufficient evidence from the animal experiments to establish dose-response relationshipsfor foetal effects of CO, but considered that CO levels greater than the WHO standard(equivalent to 2.5-3% COHb) would be required for the occurrence of foetal effects.Zielhius et al

⁽¹²⁾ also concluded that the no-effectthreshold was not known. It was considered that the current threshold limit value (TLV),bioequivalent to 8%, may not be protective against adverse foetal effects, and thatreducing exposure to 25-30 CO mg/m³ for eight hours at work (22-26 ppm,bioequivalent to 3-4% COHb) would minimise the risks to pregnancy and offspring.

There are no epidemiological studies of the direct effect of CO on human pregnancies.Indirect information can be obtained from the observation of pregnancy outcomes in womenwho smoke. It has been found that smoking is associated with adverse outcomes includinglow birth weight, an effect similar to the effect of prolonged hypoxia or CO exposure inanimals. It can therefore be suggested that CO may be responsible for the adverse effects,but as cigarette smoke contains high concentrations of other toxic agents, no definiteconclusion can be drawn as to the causative agent

^(3,5) .

5.3 Effects on the Central Nervous System

Behavioural effects of low levels of CO exposure in humans have been investigated bymany researchers. Reviews of the subject are contained in the NIOSH

⁽¹⁰⁾, WHO

⁽³⁾ and Swedish

⁽⁵⁾environmental criteria documents. The subject was also extensively reviewed by Laties andMerigan

⁽¹⁴⁾ .

Two broad categories of studies have been defined. The first consists of vigilancetests such as time discrimination and the performance of tasks requiring attention andconcentration. The second is motor tests, that is tests concerned with the performance ofphysical tasks.

Experimental subjects have in almost all cases been healthy and young. Results frommany of the experiments have been conflicting. The WHO, NIOSH and Swedish environmentalcriteria documents

^(3,5,10) are in agreement that adversebehavioural effects have not been found consistently at levels below 5% COHb. At levelsbetween 5 and 10% COHb, effects have been found on performance of tasks requiringvigilance, and on reaction time.

Chronic neurological effects have been postulated but are not clearly reported in theliterature except in victims who recover from acute high-level exposure

⁽¹⁰⁾. Kurppa and Rantanen (cited in Ref 7) postulated that chronic nervous system complaintssuch as headache, poor memory and depression might arise from accumulated insults to thecentral nervous system, resulting from repeated exposure to CO at levels which may damagenervous tissue without leading to actual loss of consciousness. Chronic effects relatingto the cardiovascular system are reviewed in the next section.

5.4 Effects on the Cardiovascular System

Consideration of both acute and chronic effects of CO on the cardiovascular system iscrucial: the basis for several standards (i.e. the WHO, NIOSH and US EnvironmentalProtection Agency standards) is protection against adverse cardiovascular effects whichhave been shown to occur in humans at low levels of CO exposure.

5.4.1 Evidence for effects from chronic low-level exposure

Evidence from animal studies suggests that CO may act as a risk factor in thedevelopment of atherosclerosis. These studies were reviewed comprehensively in the Swedishcriteria document

⁽⁵⁾ and, more recently by the USSurgeon-General

⁽¹⁵⁾ , and briefly by Atkins and Baker

⁽¹⁶⁾ .

In vitro and in vivo studies have shown physiological changes which mayprecede or accelerate atherosclerosis: altered lipid metabolism, increase in vesselpermeability, increased cholesterol uptake, and hypoxia. Although some studies areconflicting, since 1970, animal studies in three species have shown an enhanced effect ofCO on atherosclerosis in animals fed a high-cholesterol diet

⁽¹⁶⁾.

Stender et al (cited in Ref 5) concluded that CO exposure is probably a weakstimulus for atherogenesis compared to high cholesterol diets. The US Surgeon-General

⁽¹⁵⁾ considered that animal experiments on acceleratedatherogenesis with carbon monoxide were unsatisfactory, and that a cause and effectphenomenon for carbon monoxide and disease of the arterial wall has not been elucidated.

Epidemiological studies examining both morbidity and mortality of workers chronicallyexposed to CO have been carried out in Britain, the US, Finland and Canada. Studies ofBritish blast furnace workers and Finnish foundry workers with documented elevated COHblevels found no increase in mortality from cardiovascular causes

^(17,18). Redmond et al

⁽¹⁹⁾ failed to find a relationshipbetween death from cardiovascular disease and exposure to hot environments in Americansteelworkers. Although CO was not specifically measured, it was postulated that within thesteel industry, exposure to hot environments and to CO are correlated.

The British blast furnace study

⁽¹⁷⁾ also found noincrease in cardiovascular morbidity, as determined by questionnaire, among steelworkerscurrently working in a CO-exposed environment. The Finnish foundry study found, asdetermined by questionnaire, an excess in angina pectoris amongst CO-exposed workers, butno excess in electrocardiographic (ECG) changes.

Kuller and Radford

⁽⁴⁾ have pointed out seriouslimitations in the design of these studies: workers with heart disease or at high risk maymove out of CO-exposed areas because of symptoms related to doing the type of heavy workrequired, and thus prevalence of heart disease might tend to be underestimated.

Two recent studies have been able to show an increased rate of cardiovascular diseasein workers exposed to CO in the base metals industries.

Silverstein et al

⁽²⁰⁾ studying 278 grey ironfoundry workers in the US, found an increase in mortality from circulatory disease innon-whites but not whites in the whole cohort, and an excess of cardiovascular diseasemortality amongst black males working in foundry classifications. The authors regardedthis as being consistent with the fact that heat and CO exposures were likely to behighest in foundry classifications. Historical data from industrial hygiene surveys showedsurvey medians for carbon monoxide of 20-100 ppm in cupola and pouring areas (that is,higher exposure areas) in the foundry.

Theriault

⁽²¹⁾ found an association between incidence ofheart disease (morbidity and mortality) and exposure to CO, sulphur dioxide and noisewhich occurred in the Soderberg and prebake processes of a primary aluminium smelter inCanada. Each exposure was interrelated, but the strongest association was with CO. It wasfelt that this may represent a true association between CO exposure and ischaemic heartdisease. The level of exposure to CO was estimated using data supplied by engineers fromthe company. Although average exposure was "in the order of 20 ppm", there wereseveral operations where peaks of CO exposure occurred. During these operations, CO levelswere approximately 400 ppm for 15 minutes, and this could happen 5 to 10 times a day.

The design of Theriault's study may represent an improvement over previousepidemiological studies in that incidence data were able to be more completely ascertained(i.e. both first attacks of angina pectoris or myocardial infarction and sudden deaths inpreviously healthy individuals).

The biological plausibility of an effect of CO exposure on cardiovascular health isstrengthened by evidence of increased risk of heart attack in smokers. It is not firmlyestablished whether the causative agent in cigarette smoke is nicotine, tar or CO. Kullerand Radford

⁽⁴⁾ argued that the clinical experimentalevidence (see Section 5.4.2) suggests that CO plays an important, if not the primary, rolein the relationship between cigarette smoking and cardiovascular disease.

The increased risk of a heart attack from smoking may occur through different effects.Some effects of smoking may be direct, reversible, non-cumulative effects on theprecipitation of a heart attack. Others may be cumulative and not readily reversible, suchas the chronic development of atherosclerosis.

Certainly, there is evidence that mortality from coronary artery disease drops fairlyrapidly after cessation of smoking, suggesting that some effects of cigarette smoking arereversible

⁽¹⁵⁾ , and may be related to acute mechanisms.There is also evidence that the mortality continues to decline steadily, and thatatherosclerosis in smokers is correlated with higher COHb levels

⁽²⁾. This latter evidence suggests that chronic disease processes play at least a part incardiovascular disease occurrence from smoking.

Animal and epidemiological studies of long-term low-level exposure to CO thus providesuggestive but not conclusive evidence of an adverse effect on the cardiovascular system.

5.4.2 Evidence for effects from acute low-level exposure

On the other hand, clinical experimental studies provide good evidence for an acuteeffect of low-level exposure on the cardiovascular system. The clinical experimentalstudies were reviewed in the NIOSH

⁽¹⁰⁾ , WHO

⁽³⁾and Swedish

⁽⁵⁾ environmental criteria documents and byKuller and Radford

⁽⁴⁾ and the US Surgeon-General

⁽¹⁵⁾ . This group of studies provides the most definitiveevidence for effects from low-level CO exposure.

Specifically, there is good evidence that CO at low levels has acute effects on themyocardium of those with ischaemic heart disease. This is thought to occur via thefollowing mechanism. The increase in oxygen to the myocardium required duringcardiovascular work is mainly achieved through increased coronary artery blood flow. Inthose with coronary artery disease, increased coronary artery blood flow is prevented, sothe combination of a reduced oxygen supply which occurs with an elevated COHb level, andincreased demand, produces ischaemia

⁽¹⁰⁾ .

Experimental work by Ayres

⁽²²⁾ supports this concept.Patients with coronary artery disease who were acutely exposed to CO levels which produceda rapid rise in COHb to about 9% COHb, showed an increase in myocardial lactateproduction, indicating that myocardial metabolism was occurring via anaerobic pathways. Insubjects without coronary artery disease, at around 9% COHb there was increased coronaryblood flow and oxygen extraction; patients with coronary artery disease did not show thesecompensatory changes.

Charts 4 and 5 summarise clinical experimental studies of effects on the cardiovascularsystem. The results of experiments on those with clinical cardivascular disease (Chart 4)demonstrate clear evidence of adverse effects of CO at low levels.

Several groups of investigators (Aronow et al

⁽²³⁾; Anderson et al

⁽²⁴⁾ ; Aronow and Isbell

⁽²⁵⁾ ; and Knelson 1972, cited in Ref 10), studying patientswith angina pectoris, have shown a reduced time to exercise induced angina and decreasedsystolic blood pressure (BP) and heart rate at angina, at COHb levels between 2.5-3% andabove this level, resulting from exposure to CO levels of 40-100 ppm for 1-4 hours.

One group has shown such effects at 2% COHb (Aronow et al 1979, cited in Ref 4).

Experiments in healthy men (Chart 5) have shown limitation of maximum work capacity ormaximum aerobic capacity as indicators of performance. The limitation is dose-related andappears at a COHb level around 4%. This limitation does not occur at levels of 2.5-4%COHb, but these levels reduce the length of time for which maximum work effort can becarried out (Anderson et al 1971, cited in Ref 10, Ekblom and Huot 1972, cited inRef 3; Drinkwater et al

⁽³⁰⁾ ; Horvath et al

⁽²⁹⁾ ; and Aronow and Cassidy 1975, cited in Ref 5).

Of more concern are two experiments showing electrocardiographic (ECG) changessuggestive of ischaemic changes in healthy men with around 2.4-5% COHb. Anderson et al

⁽²⁴⁾ , in addition to showing adverse changes in exercisecapacity, found that clinically healthy middle-aged men exposed to 100 ppm for four hours,with COHb levels of 5-9%, had accentuated ischaemic ECG changes during exertion. At thislevel, no such changes were found in healthy young men. The investigators interpreted thisto mean that "...low level CO exposure may augment the production of exercise-inducedmyocardial ischaemia in persons with pre-existing subclinical heart disease, contribute tothe development of myocardial dysfunction, and may lead to an increased incidence ofarrhythmias in such persons".

Chart 4: Effects of CO exposure on the cardiovascular system in humans withcardiovascular disease (CVD)

⁽³⁾ , NIOSH

⁽¹⁰⁾, Rylander & Vesterlund

⁽⁵⁾ , Kuller & Radford

⁽⁴⁾ and from other sources.*

Chart 5: Effects of CO exposure on the cardiovascular system in healthy humansubjects

⁽³⁾ , NIOSH

⁽¹⁰⁾, Rylander & Vesterlund

⁽⁵⁾ , Kuller & Radford

⁽⁴⁾ and from other sources.

*

Davies and Smith

⁽²⁸⁾ studied clinically healthy youngmen using an exposure regime of eight days continuous exposure to 0, 15 and 50 ppmresulting in COHb levels in the CO-exposed groups of 2.4 and 7.1%. Resting ECGs showedp-wave changes (suggestive of ischaemia) in six of 15 men at 50 ppm, in three of 15 men at15 ppm and in none of the controls.

Thus there is clear evidence that a person with clinical cardiovascular disease is atrisk of acute myocardial ischaemic effects at COHb levels around 2.5-3%. The risk to thosewith pre-existing sub-clinical heart disease (that is, those without symptoms) is lessclear - at least two studies on healthy human subjects give cause for concern that adversecardiovascular effects may occur at levels of 2.5-5% COHb.

5.5 Other Effects on Persons with Pre-existing Illnesses

Studies similar to those investigating exercise time to symptoms in patients withcardiovascular disease have been conducted by Aronow's group for two other patient groupsfor which there is biological plausibility for an effect of CO - patients with chronicobstructive pulmonary disease and patients with intermittent claudication. At a level of4.1% COHb, there was a decrease in exercise time until marked dyspnoea in subjects withchronic obstructive pulmonary disease

⁽⁵⁾ . At a level of2.8% COHb, there was decrease in mean exercise time until the onset of the pain ofintermittent claudication

⁽³⁾ . The WHO report

⁽³⁾ discussed the susceptibility of individuals with anaemia tocarbon monoxide exposure. Although available information was regarded as inadequate, itcould be assumed a priori that anaemic individuals would be more at risk fromcarbon monoxide exposure than normal persons because the capacity of the oxygen transportsystem is reduced in anaemic individuals.

5.6 Exposure to Methane-derived Halogenated Hydrocarbons

Exposure to dichloromethane (methylene chloride) causes biotransformation into CO.Stewart et al (cited in Ref 3) showed that exposure to methylene chloride at aconcentration of 500-1000 ppm for 1-2 hours resulted in COHb levels of more than 14%. Theelevation of COHb levels continues beyond the time of exposure (possibly for many hours),resulting from metabolic production of CO from lipid stores of methylene chloride

⁽³⁾ .

Methylene chloride is widely used in industry, commonly as a paint or varnish remover,but also as a degreaser, an aerosol propellant, a solvent for paint, in plastics textilemaking, in producing photofilms and in preparing certain oils and waxes

⁽²⁾.

Where such exposure occurs concurrently with CO exposure, there is obviously an addeddanger of overexposure to CO.

5.7 Exposure to Carbon Monoxide at High Altitudes

Data referring to the effects of CO at high altitudes is scarce, but there is evidencethat carbon monoxide produces effects that aggravate the oxygen deficiency present at highaltitudes, and that when high altitude and carbon monoxide exposures are combined, theeffects are apparently additive

⁽³⁾ .

6. BASIS FOR EXISTING OCCUPATIONAL AND ENVIRONMENTAL STANDARDS

6.1 World Health Organisation (WHO)

The WHO environmental health criteria document

⁽³⁾ ,published in 1979, was the result of a review of existing evidence by an internationalgroup of experts (Task Group) meeting in Geneva in 1977.

They recommended a range of COHb concentrations of 2.5 - 3.0% as a standard for theprotection of the general population, including those who have impaired health. Thisbiological standard corresponds to the ambient air levels listed as guidelines in the WHOcriteria document.

Guidelines for exposure conditions to prevent carboxyhaemoglobin levels exceeding2.5-3% in nonsmoking populations (after WHO

⁽³⁾ ).

* Suggested by the Secretariat to the WHO Task Group.

For working populations, the Task Group recommended that exposure limits should ensurethat COHb levels were kept below 5%. The basis of the recommendation was that individualsin working populations are assumed to be healthy, physiologically resilient, and underregular supervision. The biological limit corresponds to ambient air levels listed asguidelines in the WHO document..

The WHO Task Group considered that recommendations for exposure limits should beconfined to the protection of non-smokers. Smokers should be told of the evidence that thehabit might be harmful and then be subject to the levels of protection recommended above.

Foetal effects were reviewed in the criteria document, but the Task Group consideredthat more research was needed in this area.

The Task Group regarded the main areas of concern that had arisen from acute or chroniclow-level exposure to CO in experimental and epidemiological research in animals andhumans to be the following:

(a) Its role in the development of arteriosclerotic vascular diseases.

(b) Its role in the aggravation of symptoms of cardivascular diseases.

(c) It contribution to performance deficits in certain psychomotor tasks.

(d) Its role in limiting the working capacity of "exercising man" [sic].

6.2 US Environmental Protection Agency (EPA)

The EPA air quality standard for CO, promulgated in 1971, is an eight-hour annualaverage maximum of 9 ppm and a maximum 1-hour level of 35 ppm not to be exceeded more thanonce a year. "The ambient standard was based on the effects of elevated COHb levelson the cardiovascular system and behavioural responses, especially vigilance tasks, inwhich the individual was asked to report the occurrence of occasional signals over longperiods"

⁽⁴⁾ .

6.3 US National Institute for Occupational Safety and Health (NIOSH)

In 1972, the NIOSH report

⁽¹⁰⁾ recommended anoccupational standard prescribing exposure to any concentration greater than 35 ppm as aneight-hour time-weighted average (

TWA), or to a ceiling concentration of 200 ppm. It wascalculated, using the Coburn equation, that exposure of a non-smoking sedentary worker atthis level would result in a COHb level approaching 5%.

The aim of the standard was to:

"(1) prevent acute CO poisoning;

(2) protect the employee from deleterious myocardial alterations associated with levels of COHb in excess of 5%; and

(3) provide the employee protection from adverse behavioural manifestations resulting from exposure to low levels of CO".

NIOSH considered that smokers might not be provided the same degree of protection asnon-smokers by the standard. Under high altitude conditions, the standard should beappropriately lowered. In addition, the standard would not protect workers withpre-existing illnesses such as anaemia, coronary heart disease and emphysema, to the samedegree as healthy workers. It was recommended that workers with overt cardiovasculardisease, and other diseases rendering the invididual susceptible to CO, be included in amedical program of pre-placement and periodic examinations. Such a medical program couldalso provide the opportunity for conducting anti-smoking programs for high-risk employees.

The NIOSH report discussed the high prevalence of coronary artery disease (CAD) in theUS, and the difficulty of screening for the presence of CAD, as the standard was designedto protect against "deleterious myocardial alterations". It was considered"necessary to assume that the average worker has asymptomatic CAD; especially whenhis first clinical symptom may be sudden death"

⁽¹⁰⁾ .It was also considered that limiting CO exposure to no more than the standard shouldprotect individuals with asymptomatic CAD from developing clinical symptoms. Effects onthe foetus were not considered.

6.4 Swedish Occupational Standard

The Swedish occupational standard

⁽¹³⁾ is 35 ppm as aneight-hour

TWA, designed to keep COHb levels below 5% in non-smokers. The Swedish CriteriaGroup for Occupational Standards considered that the effect of carbon monoxide on heartfunction, embryos and the central nervous system should be borne in mind in theestablishment of exposure limits.

The Group considered that persons with heart or circulatory disease are particularlysensitive to the effects of carbon monoxide, in particular with physical exertion. Itconsidered that experiments had shown effects on the central nervous system at COHb levelsof 5-10%, and that data from animal experiments implied that foetal effects had not beenshown to occur except at levels higher than exposure to about 35 ppm.

6.5 American Conference of Government Industrial Hygienists (ACGIH)

The ACGIH TLV is set at 25 ppm as an 8-hour

TWA.

The ACGIH BEI is set at 3.5% COHb, not to be exceeded at any time during the workshift. This level is regarded as being bioequivalent to the TLV of 25 ppm, assuming aneight-hour exposure, with the proviso that in smokers, COHb over 3.5% is not a necessaryindicator of occupational overexposure

⁽³¹⁾ .

The ACGIH recommends a 25 ppm TLV "to keep blood COHb levels below 3.5%, toprevent adverse neurobehavioral changes, and to maintain cardiovascular exercisecapacity". The recommendation also provides a greater margin of safety for"susceptible" individuals. These individuals include those with cardiovasculardisease both detected and undetected. For such workers, it is considered that a TLV of 25ppm, bioequivalent to COHb of 3.5% is necessary, and that even this concentration may bedetrimental to the health of some workers who may have far advanced cardiovasculardisease.

Other susceptible groups are considered to be those living at high altitudes (over 5000feet (1524 metres) above sea level), those with respiratory disease and the foetus ofpregnant workers.

These susceptible groups may also be at additional risk under conditions of heavylabour and high temperature.

7 RATIONALE FOR THE PROPOSED STANDARD

7.1 Health Effects Considered in Proposing a Standard

Based on the foregoing review, the major potential effects which should be consideredin the setting of a standard limiting occupational exposure to carbon monoxide are asfollows:

(a) Adverse behavioural effects at low levels of exposure (see Section 5.3).

(b) Adverse pregnancy outcomes at low levels of exposure (see Section 5.2).

(c) Acute effects of high levels of exposure (see Section 5.1).

(d) Adverse effects from acute and chronic low-level exposure in workers with pre-existing coronary artery disease (both clinical and asymptomatic) and possibly healthy people at low levels of exposure (see Section 5.4)

7.2 Behavioural Effects

There is general agreement in the WHO, NIOSH and Swedish criteria documents

^(3,10,13) that behavioural effects have been found onperformance of tasks requiring vigilance, and on reaction time at levels between 5 and 10%COHb, and that behavioural effects have not been observed consistently below 5% COHb (seeSection 5.3).

7.3 Foetal Effects

A level of 5% COHb (or "exposure to about 35 ppm CO") is also considered inthe Swedish occupational standard

⁽¹³⁾ to provide protectionfor the foetus, given present information. A recent review of the subject by Zielhius etal

⁽¹²⁾ suggests that "the present TLV[bioequivalent to COHb of 8-10%] may not fully protect against adverse effects on theoffspring" and that "reducing exposure to 25-30 mg CO/m³ for 8 h atwork [bioequivalent to COHb of 3-4%] will minimize risks to pregnancy and offspring"(see Section 5.2).

7.4 Acute Effects of High Level of Exposure

The proposed standard should provide protection against high-level acute effects, thatis, giving rise to overt symptoms, as it is aimed at protecting against low-level effects(see Section 5.1).

7.5 Adverse Effects on Persons with Coronary Artery Disease

7.5.1 Persons with overt coronary artery disease

A level of 2.5-3% COHb is the lowest level at which clearly adverse health effects havebeen well-documented. These health effects are adverse cardiovascular effects on personswith pre-existing clinically overt coronary artery disease, giving rise to symptoms ofangina pectoris (see Section 5.4.2)

Thus a health-based standard which would protect all persons, including susceptibleindividuals, should be set to ensure exposure no higher than this level i.e. 2.5-3% COHband the corresponding ambient air levels for non-smokers (see Section 6.1).

The WHO environmental standard is based on the explicit assumption that workingpopulations are healthy and under regular supervision. The NIOSH report made therecommendation that working populations have routine medical examinations which woulddetect and exclude susceptible individuals with overt CAD from CO exposure at the level ofthe recommended standard.

Individuals with overt CAD may be at risk at the proposed level of 5% COHb.

7.5.2 Persons with undetected (subclinical) CAD

Although it is well-documented that those with clinical coronary artery disease (CAD)may be at risk at a COHb level of 4-5%, the risk to those with sub-clinical CAD is lessclear. This review has identified one study showing adverse effects in middle-agedclinically healthy men at 5% COHb, and one study showing non-specific effects suggestiveof cardiac ischaemia in healthy young men at a level of 2-4% (see Section 5.4.2). There isinadequate information on which to base a dose-response relationship.

The WHO, NIOSH and Swedish Criteria Groups have set an occupational exposure limitbioequivalent to 5% COHb for non-smokers which was considered to protect healthyindividuals from the adverse CAD effects. The ACGIH considered that COHb of 3.5% might benecessary to protect workers with significant CAD, both detected and undetected.

For the purposes of setting standards, the NIOSH

⁽¹⁰⁾stated in 1972 that it was "necessary to assume that the average worker hasasymptomatic CAD", given the high prevalence of the disease in the US and the factthat "the detection of such persons [with CAD] in the absence of overt clinicalsymptoms is virtually impossible". This assumption should also hold for Australiatoday, for the following reasons.

7.5.2.1 Prevalence of subclinical coronary artery disease

Coronary artery disease has a high incidence in Australia, including populations ofworking age. Despite a recent fall in cardiovascular disease mortality, ischaemic heartdisease remains the largest cause of morbidity and mortality in Australia and in mostother developed countries

⁽³²⁾ .

Table 8 is derived from the National Heart Foundation of Australia (NHFA) Risk FactorPrevalance Survey No. 2 (1983)

⁽³³⁾ and shows the percentageof people who answered yes to the question, "Have you ever been told you have or havehad angina/heart attack?"

Table 8 Angina/heart attack

The above data indicate that, especially in the 45 year and above age groups, theprevalence of heart disease becomes quite significant.

There is a high prevalence of subclinical disease in populations of working age. TheNIOSH report

⁽¹⁰⁾ cites evidence from an autopsy study ofyoung soldiers, average age 22 years, killed during the Korean War, which revealed that77.3 percent had gross pathological evidence of CAD, and also cites an estimate of morethan half a million persons per year in the US with "silent" (clinically notapparent) myocardial infarctions.

There is no reason to suppose that the incidence of subclinical CAD is any lower inAustralia than in the US, and indeed mortality rates from ischaemic heart disease inAustralia are high compared with most other countries

^(32,34,35).

Another insight into the prevalence of subclinical CAD can be seen from the PerthMonica Project, which recorded all major cardiac events such as heart attack and suddendeath in Western Australia

⁽³⁶⁾ . This study found that 25%of these major cardiac events occurred in people "unknown" to the Project, thatis people who were previously asymptomatic or not diagnosed as having CAD.

Given that 30% of all causes of death are due to cardiac disease, it is reasonable todeduce that, especially in the 45 years and above age group, there is a significantproportion of the workforce with subclinical or undiagnosed heart disease.

7.5.2.2 Screening for coronary artery disease

There is a continuing difficulty in screening for CAD, on both technical and ethicalgrounds.

Screening methods chosen to detect asymptomatic CAD might be, in the order less to moresophisticated (and accurate, and expensive) standardised symptom questionnaire, e.g. theWHO questionnaire

⁽³⁷⁾ , resting electrocardiograph (ECG)and exercise ECG. These methods also require, from first to last in order, an increasingdegree of cooperation from the worker.

All screening tests produce a proportion of false negative and false positive results,and it might be expected that this proportion would be greater when tests are used toscreen asymptomatic persons than when used for clinical diagnosis.

False negatives raise the potential for mis-diagnosis, whereas false positives raisethe potential for discrimination in employment and may cause unnecessary anxiety toapparently healthy workers.

For these reasons, it is not possible to develop a feasible and equitable strategywhich will detect persons with subclinical CAD (as opposed to overt CAD) in order toremove them from exposure. Instead, it must be accepted that these persons withsubclinical disease should be protected by an occupational standard.

8. RECOMMENDATIONS FOR AN OCCUPATIONAL EXPOSURE STANDARD

Coronary artery disease is widespread in the Australian community and, despite a recentfall in CAD mortality, remains the largest source of morbidity and mortality in thiscountry. The disease has a number of causes and factors which affect its development;these include genetic, lifestyle and occupational contributions.

In assigning an exposure standard for carbon monoxide, the Working Group acknowledgesthe following:

(a) It is well established that persons with clinical (established) cardiovascular disease may be at risk at 4-5% COHb.

(b) Individuals with subclinical cardiovascular disease and foetuses of exposed pregnant women may be at risk at levels above 5% COHb. There is inconclusive evidence of a risk for these two groups below this level.

(c) There is a high prevalence of subclinical (undetected) cardiovascular disease in Australia, especially in older workers.

(d) There is a significant number of deaths from CAD in individuals who previously had no detectable heart disease.

(e) It would be inappropriate and discriminatory to exclude all older workers from jobs where exposure to carbon monoxide was likely on the basis that as a class of people, they were at higher risk.

(f) It is not possible to develop a medical surveillance strategy which will reliably detect persons with subclinical cardiovascular disease.

Accordingly, the Working Group concludes that the exposure standard for carbon monoxidemust afford protection for those significant numbers of working individuals withsubclinical coronary artery disease.

The Working Group recommends that an eight-hour (

TWA) exposure standard of 30 ppm beassigned to carbon monoxide. This airborne concentration is believed to be bioequivalentto 5% COHb under normal temperatures, workloads, atmospheric pressures, and in the absenceof such solvents as methylene chloride which, in their metabolism, contribute to the bodyburden of COHb.

The Working Group believes that compliance with this exposure standard and 5% COHbshould minimise the risk to those persons with subclinical CAD and to foetuses of exposedpregnant women. It should also protect against adverse behavioural effects arising fromcarbon monoxide exposure.

The exposure standard does not exclude an increased risk for individuals with overt CADor for smokers. A COHb of greater than 5% in smokers is not necessarily an indicator ofoccupational over-exposure.

The Working Group acknowledges that because of the ubiquitous nature of carbon monoxidein the environment, there are a number of non-occupational and lifestyle activities whichcould give rise to transient exposures approaching or exceeding 5% COHb. Nonetheless, inthe occupational context, where exposure may be prolonged, the Working Group recommendsthat 5% COHb be used as a guideline for workplace control.

8.1 Excursion Limits

The affinity of carbon monoxide for haemoglobin results in its rapid uptake by the bodyand slow depletion following the cessation of exposure. The biological half-life of carbonmonoxide is in the order of three to four hours. Accordingly, excursions above theeight-hour

TWA of 30 ppm must be carefully controlled if 5% COHb is not to be exceeded.

The Working Group believes that the toxicokinetics of carbon monoxide are notcompatible with the establishment of a conventional

STEL. To maintain compliance with 5%COHb, a

STEL of 100 ppm would have to be applied. However, using the data derived from theCoburn model, the guidelines set out in Table 9 are recommended for the control of theexcursions above the eight-hour

TWA exposure standard.

Table 9. Guidelines for the control of short-term excursions for Carbon Monoxide

^(38).

(a) Short-term excursions should never exceed 400 ppm.

(b) This duration represents the sum of exposures at this level over an 8-hour workday, and assumes no other exposure to carbon monoxide.

The values given in Table 9 should be considered in conjunction with the 8-hour

TWAexposure standard for carbon monoxide.

The principle determinate in evaluating excursions above the

TWA value is themaintenance of a COHb not exceeding 5%.

9. ACKNOWLEDGEMENTS

This documentation is substantially based upon an original review prepared in 1987 byDr Susan Lewis of the Epidemiology and Surveillance Unit of the National Institute ofOccupational Health and Safety. Dr Lewis' review has been amended to incorporate commentsprovided by the Cardiac Society of Australia and New Zealand, by Dr Brian Dare, SeniorOccupational Physician at the Western Australian Department of Occupational Health Safetyand Welfare, and the recommendations of the Exposure Standards Working Group.

REFERENCES

1. ILO, Encyclopaedia of Occupational Safety and Health. 3rd ed, InternationalLabor Office, Geneva, 1983.

2. Zenz C, "The epidemiology of carbon monoxide in cardiovascular disease inindustrial environments: a review", Prev Med, 8: 279-288, 1979.

3. WHO, "Carbon Monoxide", Environmental Health Criteria 13, WorldHealth Organisation, Geneva, 1979.

4. Kuller LH & Radford EP, "Epidemiological bases for the current ambientcarbon monoxide standards", Environ Health Persp, 52: 131-139, 1983.

5. Rylander R & Vesterlund J, "Carbon monoxide criteria", Scand J WorkEnviron Health, 7, Suppl 1, 1-39, 1981.

6. Peterson JE, "Postexposure relationship of CO in blood and expired air", ArchEnviron Health, 21: 172-173, 1972.

7. Jarvis MJ et al, "Expired air carbon monoxide: a simple breath test oftobacco smoke intake", Brit Med J, 16 August, 484-485, 1980.

8. Commins BT, "Measurement of carbon monoxide in the blood: a review of availablemethods", Ann Occup Hyg, 18: 69-77, 1975.

9. Harrison N, "A review of techniques for the measurement of carbon monoxide inthe atmosphere", Ann Occup Hyg, 18: 37-44, 1975.

10. NIOSH, Criteria for a Recommended Standard - Occupational exposure to CarbonMonoxide, National Institute for Occupational Safety and Health, US Department ofHealth, Education and Welfare, Washington DC, 1972.

11. Evanoff BA & Rosenstock L, "Reproductive hazards in the workplace: a casestudy of women fire fighters", Am J Industr Med, 9: 503-515, 1986.

12. Zielhius RL et al, Health Risks to Female Workers in OccupationalExposure to Chemical Agents, Springer-Verlag, Berlin, 1984.

13. Criteria Group for Occupational Standards, Scientific Basis for SwedishOccupational Standards III. Consensus Report for Carbon Monoxide, National Board ofOccupational Safety and Health, Solna, Sweden, 1982.

14. Laties VG & Merigan WH, "Behavioural effects of carbon monoxide on animalsand man", Ann Rev Pharmacol Toxicol, 19: 357-392, 1979.

15. US Surgeon-General, The Health Consequences of Smoking: Cardiovascular Disease:A Report of the Surgeon-General, Office on Smoking and Health, US Department of Healthand Human Services, Public Health Service, Rockville, Maryland, 1983.

16. Atkins EH & Baker EL, "Exacerbation of coronary artery disease byoccupational carbon monoxide exposure; a report of two fatalities and a review of theliterature", Am J Industr Med, 7: 73-79, 1985.

17. Jones JG & Sinclair A, "Arterial disease amongst blast furnanceworkers", Ann Occup Hyg, 18: 15-20, 1975.

18. Koskela R et al, "A mortality study of foundry workers", ScandJ Work Environ Health, 2, Suppl 1, 73-89, 1976.

19. Redmond CK et al, "Mortality of steel workers employed in hotjobs", J Environ Pathol Toxicol, 2: 75-96, 1979.

20. Silverstein M et al, "Mortality among ferrous foundry workers", AmJ Industr Med, 10: 27-43, 1986.

21. Theriault G, "Risk of heart disease following chronic exposure to CO, sulphurdioxide and noise in primary aluminium smelter workers", Paper presented at theFifth International Symposium on Epidemiology and Occupational Health, Los Angeles,September 11, 1986.

22. Ayres SM et al, "Systemic and myocardial hemodynamic responses torelatively small concentrations of carboxyhemoglobin", Arch Environ Health,18: 699-709, 1969.

23. Aronow WS et al, "Effect of freeway travel on angina pectoris",.Ann Intern Med, 77: 669-676, 1972.

24. Anderson EW et al, "Effect of low-level carbon monoxide exposure anonset and duration of angina pectoris", Ann Intern Med, 79: 46-50, 1973.

25. Aronow WS & Isbell MW, "Carbon monoxide effect on exercise induced anginapectoris", Ann Intern Med, 79: 392-395, 1973

26. Adams KF et al, "Acute elevation of blood carboxyhemoglobin to 6%impairs exercise performance and aggravates symptoms in patients with ischemic heartdisease", J Am Coll Cardiol, 12: 900-9, 1988.

27. Allred E, et al, "Short-term effects of carbon monoxide exposure on theexercise performance of subjects with coronary artery disease", N Engl JMed, 321: 1426-32, 1989.

28. Davies LM & Smith DJ, "Electrocardiographic changes in healthy men duringcontinuous low-level carbon monoxide exposure", Environ Res, 21: 197-206,1980.

29. Horvath SM et al, "Maximal aerobic capacity at different levels ofcarboxyhaemoglobin", J Appl Physiol, 38: 300-303, 1975.

30. Drinkwater BL et al, "Air pollution, exercise and heat stress", ArchEnviron Health, 28: 177-181, 1974.

31. ACGIH, Documentation of Threshold Limit Values and Biological Exposure Limits,6th ed, American Conference of Governmental Industrial Hygienists, Cincinnati, Ohio, 1991.

32. Heller RF, "The rise and fall of cardiovascular disease", Med J Aust,144: 686-687, 1986.

33. Natinal Heart Foundation of Australia (NHFA), Risk Factor Prevalence Study No.2(1983), Canberra, NHFA, 1983.

34. Al-Roomi KA & Leeder SR, "Changing patterns of ischaemic heart disease inelderly Australians", Med J Aust, 145: 278-279 1987 [letter].

35. Leeder SR et al, "Attack and case fatality rates for acute myocardialinfarction in the Hunter region of NSW, Australia, in 1979", Am J Epidemiol,118: 42-57, 1983.

36. Turnstall-Pedoe H, "The World Health Organisation Perth Monica Report:(Monitoring Trends and Determinants in Cardiovascular Disease): a major internationalcollaboration", J Clin Epidemiol, 41. : 105-114, 1988.

37. Rose GA et al, Cardiovascular Survey Methods, 2nd ed, World HealthOrganisation, Geneva, 1982.

38. National Occupational Health and Safety Commission, Exposure Standards forAtmospheric Contaminants in the Occupational Environment: Guidance Note [NOHSC:3008(1995)] and National ExposureStandards [NOHSC:1003(1995)], 3rd Edition, Australian Government Publishing Service,Canberra, May 1995.