For environmental
occupational health safe
and responsible use

Science and publications

World Health Organization
Organisation Mondiale de la Santé


Reports prepared by a WHO Meeting 
(Oxford, United Kingdom, 10-11 April 1989)




Summary of the evidence on the adverse health effects of exposure to asbestos

1.  General 
2.  Animal experimental evidence 
3.  Lung burdens of asbestos fibres 
4.  Epidemiological evidence - general 
5.  Epidemiological evidence of dose-response relations 
6.  Implications for occupational exposure limits 
7.  Conclusions on occupational exposure limits 
8.  Considerations on the measures needed to set occupational exposurelimits




Because of the importance of safety in the use of asbestos, the International Labour Conference adopted a Convention (No. 162) in its 72nd session in 1986. This convention contains 30 articles, covering both the administrative and technical aspects. Article 15.1 stipulates that "The competent authority shall prescribe limits for the exposure of workers to asbestos or other exposure criteria for the evaluation of the working environment".

As the occupational exposure limits for asbestos currently applied in a number of industrialized countries are somewhat different for many reasons, WHO was approached by the International Fibre Safety Group to convene urgently a consultation meeting to discuss this specific question, with a view of providing advice to those countries, particularly in the developing world, which are considering the ratification of the ILO Convention.

This report is the work of the consultation meeting held in Oxford, United Kingdom, on 10-11 April 1989. We are very grateful for the advice provided by the participants in the meeting chaired by Sir Richard Doll. The representation of the International Labour Organization, the International Commission on Occupational Health, the International Federation of Chemical, Energy and General Workers' Union, and the Finnish Institute of Occupational Health in this WHO meeting requires special acknowledgement.

Dr. Thomas K. Ng
Acting Chief
Office of Occupational Health
WHO, Geneva 1989



Opening the meeting on behalf of the Director-General, Dr Hiroshi Nakajima, Dr. Thomas Ng, Acting Chief, WHO Office of Occupational Health, pointed out that many countries, particularly those in the developing world, were looking for guidance in establishing occupational exposure limits for asbestos, in order to ratify the ILO Asbestos Convention (No. 162).

The meeting began by reviewing the two special background papers, prepared by Professor N.J. Gardner and Dr. R. Murray. After proposing certain modifications, the meeting recommended the publication of these two papers as annexes to the report.

In the course of the review, the meeting prepared a summary of the evidence on the adverse health effects of exposure to asbestos. Recognizing the problems and limitations of current scientific evidence and methodology, the meeting concluded that despite more than 30 years of research, the evidence was still insufficient to make a definite statement that there was a level of exposure to asbestos below which there was no risk.

On the other hand, the meeting expressed an opinion that a level of control could be achieved, particularly with respect to chrysotile (*) asbestos, at which the risk of any asbestos-related disease that might occur would be very small. This opinion was based on the weight and direction of the evidence, reflecting the best judgment at present. The recommendations were made chiefly according to this opinion.

(*)  Here and elsewhere in this report, except in sections 2(g) and 3(d), "chrysotile" means the fibre as mined and milled and thereafter used in industrial applications; such fibre often contains very small amounts of amphibole fibres or other contaminants.



1.  General

(a)  Occupational exposure to asbestos by inhalation can cause pulmonary fibrosis (asbestosis), lung cancer, mesothelioma of the pleura and peritoneum, and changes to the pleura (thickening, plaques, effusion).

(b)   Exposure levels which are adequate for the control of lung cancer and mesothelioma will also be adequate for the control of all other asbestos-related disease.


2.  Animal experimental evidence

(a)  All asbestos types can produce pulmonary tumours and mesotheliomas.

(b)  In animal inhalation experiments, pulmonary tumour incidence has been found to rise both with increasing duration and with increasing fibre concentration of exposure. For constant concentration, the response was approximately linearly related to duration of exposure. There was no definite evidence for or against a threshold (that is, non-zero exposure level at which no health effect occurs).

(c)  In animal intrapleural and intraperitoneal installation experiments, mesothelioma incidence was related to dose of fibres. There was no definite evidence for or against a threshold.

(d)  Carcinogenicity depends on fibre length: fibres longer than 10 micrometres (µm) being most carcinogenic. Very short fibres (less than 2 or 3 micrometres (µm) have not produced tumours experimentally and are probably not carcinogenic.

(e)  Carcinogenicity depends on fibre diameter: fibres finer than 0.25 micrometres (µm) being most carcinogenic. Fibres thicker than 1.5 micrometres (µm) have not produced tumours experimentally and are probably not carcinogenic.

(f)  There is a body of opinion that carcinogenicity is also affected by durability and surface properties of the asbestos fibres.

(g)  Chrysotile is less durable in lung tissue than amphibole asbestos (crocidolite, amosite, anthophyllite, actinolite, tremolite).


3.  Lung burdens for asbestos fibres

(a)  Asbestos fibre types can be identified in lung tissue and their concentration measured.

(b)  For those occupationally exposed to asbestos, amphibole asbestos concentrations in lung tissue correlate well with estimated workplace cumulative exposure. Chrysotile asbestos concentrations correlate less well. This poor correlation for chrysotile, together with the fact that there are much lower lung concentrations, exposure for exposure, suggests that much chrysotile is eliminated from the lung.

(c)  Average asbestos fibre levels of all types in lung tissue decline from occupational to neighbourhood to general environmental exposure.

(d)  Lung burdens of amphibole, but not of chrysotile asbestos, are higher in mesothelioma cases than in controls.

(e)  Ferruginous (asbestos) bodies have been used as an indicator of exposure to amphibole asbestos. They are of limited use in determining degree of exposure.


4.  Epidemiological evidence - general

(a)  Lung cancer has been produced by all types of asbestos fibre, as experienced occupationally. In general, the human evidence suggests a lower risk from exposure to chrysotile than to crocidolite or amosite, although it is difficult to substantiate this difference firmly after standardization of exposure levels, type of industry, duration and employment, etc.

(b)  The risk of lung cancer is markedly higher among cigarette smokers, the combined risk being much greater than the sum of the separate risks for exposure to asbestos and for cigarette smoking alone.

(c)  Pleural mesothelioma has been produced by all types of asbestos fibre, as experienced occupationally. In general, the human evidence suggests a much lower risk from exposure to chrysotile than to crocidolite or amosite.

(d)  Peritoneal mesothelioma can be produced by crocidolite or amosite. This tumour has probably not been produced by chrysotile.

(e)  For chrysotile asbestos, the risks of lung cancer from employment in mining and in the manufacture of asbestos cement and friction products have been markedly less than in textile production.

(f)  In asbestos cement manufacture the risks of lung cancer and pleural mesothelioma have been lower if chrysotile only has been used than if amphibole asbestos has also been used.

(g)  Asbestos-related diseases (pulmonary fibrosis, lung cancer and mesothelioma) have occurred in the secondary production industries, in the application of asbestos insulation and during construction using materials containing friable asbestos. (As excessive concentrations of asbestos dust may be released during removal of asbestos or demolition of structures containing friable asbestos, the meeting expressed concern about the potential hazard if adequate control measures are not applied.)


5.  Epidemiological evidence of dose-response relations

(a).  The exposure assessments used in epidemiological studies have been mainly quantitative. The exposure of workers employed three or more decades ago was rarely measured, so the estimation of fibre levels reported are subject to a degree of uncertainty.

(b)  Measurement of response - lung cancer and mesothelioma - is less problematical.

(c)  For lung cancer, the relationship with cumulative exposure has been found to be approximately linear in a number of studies, but with wide variations in slope (*). The slopes appear to be related to fibre type and to industry: lowest in chrysotile mining and in the manufacture of friction materials containing chrysotile only, and highest in textile production (irrespective of fibre type) and when amphibole asbestos is used.

(d)  ...Mesothelioma mortality rates are found to increase in relation to about the third power of duration since first exposure to asbestos, with rates of increase related to exposure concentration, to industry and to fibre type and size. The Mesothelioma rates have been substantially lower than exposure to chrysotile only than from exposure to amphibole asbestos, with or without chrysotile.

(e)  It is not possible to be absolutely firm about these relationships and differences because of the limitations inherent in the exposure data, in particular, but also because of limitations of the epidemiological method.

(*)   In this report, "slope" means the slope of the relationship between death rate (usually measured as the Standard Mortality Ratio (SMR) and the assessments of exposure,usually cumulative exposure. The assumption of linearity implies that the slope is constant: that is, the SMR increases by an equal amount for each unit increase in exposure (section 5.C.3 of the paper by Professor N.J. Gardner).


6.  Implications for occupational exposure limits

(a)  The human evidence has not demonstrated that there is a threshold exposure level for lung cancer or mesothelioma, below which exposure to asbestos dust would be free of hazard to health.

(b)  Although the available data are compatible with the simple statistical models used, including the postulate that risk is proportional to exposure, the models are not based on an understanding of the mechanisms by which fibres cause lung cancer or mesothelioma. With some of the possible mechanisms proposed, it might be held that there is a threshold level of exposure below which no cancer is produced.

(c)  The lifetime risks of lung cancer or mesothelioma predicted from the models at the low levels of occupational exposure to chrysotile currently achieved in some countries are so small that in designing a study to detect their presence or otherwise, many thousands of workers have to be recruited into the cohort.

(d)  There are differences between predicted risks by fibre type and by industrial process for the same fibre concentration and duration of exposure. These are relevant to interpretation of the evidence and for targeting of occupational control measures.

(e)  For lung cancer, there are differences between predicted risks for smokers and non-smokers, the risks for smokers being much higher than for non-smokers. Thus, the overall risk for an occupational group is dependent on smoking habits.


7.  Conclusions on occupational exposure limits

(a)  There is no substantial evidence of a threshold for asbestos exposure below which cancer does not occur. However, there is a body of opinion that holds that there is a threshold and in a number of recently published studies of chrysotile asbestos workers, no excess of asbestos-related disease has been demonstrated.

(b)  On current epidemiological models, the lifetime risks of lung cancer and mesothelioma can be calculated for various combinations of exposure level and duration of exposure. Examples of these calculations are given in Annex 2, in the paper by Professor N.J. Gardner.

(c)  For working populations with a low prevalence of smoking, the associated lifetime risks of asbestos-related lung cancer will be lower. The reverse is also true.

(d)  Among chrysotile asbestos applications, the risks of developing asbestos-related disease in mining, in the manufacture of asbestos-cement products and in the manufacture of friction materials are compatible with each other for the same level of fibre exposure. For reasons that are not understood, the risks are higher in the asbestos textile industry for the same estimated level of fibre exposure.


8.  Considerations on the measures needed to set occupational exposure limits

(a)  The occupational exposure limit to be set by any country can be set only by the appropriate national body and should take into account national priorities. The national body should take into account that the limit is aimed at protecting the health of the workers many decades ahead.

(b)  It needs to be recognized that the setting of an occupational exposure limit is only one part of the programme to protect the health of workers. The programme should also include dust control, monitoring, adherence to good work practice and, if appropriate, the use of effective respiratory protective equipment.

(c)  The implementation of an occupational exposure limit thus set particularly requires that the monitoring of its application be adequate. Appropriate occupational hygiene training is essential.

(d)  It was observed at the meeting that:

  1. gravimetric and fibre counting occupational exposure limits are both used;

  2. the biological effects of asbestos depend on the exposure to fibres of certain sizes;

  3. most evidence on health effects is based on particle and fibre counts;

  4. in fibre counting methods, only fibres longer than 5 micrometres (µm) with an aspect ratio of 3 to 1 are considered for occupational hygiene practice;

  5. gravimetric methods have some advantages, in particular for evaluation of the efficiency of control of the occupational environment; and,

  6. little work has been done to show the relation between gravimetric and fibre counting measurements.



  1. The health information contained in this report should be taken into consideration by appropriate national authorities, together with the relevant technological and economic considerations, in setting and revising occupational exposure limits for asbestos. Attention is drawn to 6.(a) in the above summary that human evidence has not demonstrated that there is a threshold exposure level for lung cancer or mesothelioma, below which exposure to asbestos dust would be free of hazard to health.
  2. For chrysotile asbestos, it is recommended that countries currently having high limits should take urgent steps to lower the occupational exposure limit for an individual worker to 2 fibres/ml (8-hour time weighted average), based on health reasons alone. It is also recommended that countries should move quickly to lower the occupational exposure limit for an individual worker to 1 fibre/ml or below (8-hour time weighted average), if they have not yet already done so.
  3. For crocidolite and amosite asbestos, on the basis of health, it is recommended that their use should be prohibited as soon as possible. Restricted use in the interim period should be exercised with great care to ensure that exposure is less than that permitted for chrysotile.
  4. The health effects of exposure to fibrous amphiboles, other than crocidolite and amosite, should be investigated urgently with the intent of providing information for the recommendation of appropriate occupational exposure limits. For the time being, their use should be restricted and the exposure to these fibres should be less than that permitted for chrysotile.
  5. Study is needed of the extent to which contamination of chrysotile asbestos by traces of amphibole asbestos modifies effects on health.
  6. Those countries which use or intend to introduce gravimetric occupational exposure limits should investigate the association between gravimetric and fibre-count measurements to relate their limits to the fibre-count limits stated in paragraph 2 above.
  7. Because implementation of the recommended exposure level is as important as the setting of the level itself, it is recommended that practical assistance be made available to countries, on request, in education, training, monitoring methods and equipment for safety in the use of asbestos. Current international efforts should be supported and promoted.

Back to index of publications

International Chrysotile Association © All rights reserved | Contact us