A copy of the INSERM report was submitted to the Ministry of Labour, June 21, 1996 and formally released on July 3. Unfortunately, only a French version of the Executive Summary of the INSERM report is currently available (73 pages). The comments made below are restricted only to this document. A more in-depth analysis will be undertaken once a copy of the full report, complete with annexes, is obtained.
At best, the INSERM report represents a superficial treatment of what is obviously a very complex subject. What is most striking is that it appears to have ignored or dismissed out-of-hand the latest risk assessment evidence on chrysotile asbestos and recent developments in the understanding of the mechanism of fibre-related disease. In addition, the report totally ignores the many scientific studies which demonstrate no detectable risk to chrysotile workers at low levels of exposure.
What the document does do is to portray chrysotile as an amphibole. And, although the document does not recommend a total ban on chrysotile products, the risk estimates provided in the report lead the reader to reach no other conclusion.
The INSERM Report misrepresents the risks of asbestos in buildings for both school children and general building occupants
In its report, INSERM presented only one set of risk estimates regarding passive exposure to building occupants and school children - a worst case scenario where it was assumed that occupants were exposed to an average level of 0.025 f/ml, the level at which remedial action is triggered by French regulations. The 0.025 f/ml level was selected for the INSERM risk assessment model despite the fact that INSERM itself acknowledged that over 93% of buildings with friable ACM in France have average fibre concentrations less than 0.005 f/ml. The risk estimates of between 4 and 9 increased deaths per 10,000 (based on concentrations of 0.025 f/ml) leave the reader with the mistaken impression that that asbestos insulation in buildings is a public health risk. The comments of France's Minister of Labour in announcing his country's ban did nothing to correct this misconception, in fact, they fueled it.
A similar, but more comprehensive study undertaken in the U.S. by the Health Effects Institute (HEI), an independent research body, found that average concentrations in the buildings studied (including schools, residences and public and commercial buildings) ranged from 0.00004 to 0.00243 f/cc with a mean concentration for all buildings of 0.00027 f/cc. At these levels of exposure, HEI estimated that the risk of premature cancer death for building occupants ranged from approximately 4 to 40 per million people exposed, a risk up to 220 times lower than that estimated by INSERM. These low levels of risk clearly indicate that asbestos in buildings, provided it is properly maintained, does not present a risk of any significance to building occupants.
It is important to note that both the INSERM and the HEI risk estimates are not based on actual documented cases of disease at low level exposures in buildings, but rather on linear extrapolations from much higher occupational exposures. The validity of the linear model continues to be questioned by researchers who have found no direct evidence that there is an actual risk at the low levels of exposure studied.
The INSERM Report is in contradiction with more extensive risk evaluations of occupational exposures at low levels
While it is not known what risk model it has used, INSERM claims that occupational exposures of 0.1 f/cc will result in 30 additional deaths per 10,000 workers exposed. This stands in contrast to the conclusions of WHO's Oxford Group of Experts Report which estimated a similar level of excess mortality, but under exposure conditions 5 times higher than those from INSERM. The Oxford experts also acknowledged that the risks to non-smokers could be one tenth of those projected. As with most risk estimates of this nature, the Oxford Report numbers are based on a linear dose-response curve whose validity many researchers now question. According to statistician Dr. Douglas Liddell, at low levels of exposure for chrysotile asbestos, the dose-response curve model is likely to be sub-linear, thus rendering risk assessments based on a linear dose-response model to be seriously overstated.
Unrealistic estimates of the number of deaths from mesothelioma in workers exposed only to commercial chrysotile
INSERM has used a theoretical approach to evaluating mesothelioma risks and has completely ignored the relevant epidemiological data. On the basis of the risk factors applied in the INSERM model, it can be estimated that there would be a 10% mesothelioma mortality rate (e.g. 1,000/10,000 workers) among workers with cumulative lifetime exposures of 400 f/ml years (e.g. 40 years at a concentration of 10 f/ml ). However INSERM estimates bear little relationship to the actual experience to date. In the past, Quebec miners and millers have had cumulative lifetime fibre exposures well above this level. However of the 8,000 deaths reported to date in the world's largest cohort study of chrysotile workers, only 28 have been due to mesothelioma (i.e. 0.35%) rather than the 1,000 or so which would have been predicted by the INSERM model.
It must be noted that the number of mesothelioma cases directly attributable to chrysotile exposure are very small in number and almost all have occurred amongst working populations from the mining sector exposed to very high dust levels in the past. The strength of the epidemiological data has led Dr. A. Churg to conclude that "as a practical matter, the data indicate that chrysotile will not produce mesotheliomas in those exposed to any current or recently regulated number of fibres." (see Annex III)
The INSERM model is obviously flawed and reflects INSERM's failure to distinguish between asbestos fibre types when predicting mesothelioma risks.
INSERM does not consider evidence supporting a practical occupational threshold for chrysotile asbestos
In its report, it appears that INSERM has ignored a large body of scientific evidence which suggests a practical occupational threshold for chrysotile asbestos. This evidence was reviewed at an International Workshop on Health Risks Associated with Chrysotile Asbestos organized jointly by the International Programme on Chemical Safety, and the Scientific Committee on Fibres of the International Commission on Occupational Health in 1993 (Ann. Occup. Hyg., 1994; 38(4):397-638). Based on the proceedings of the meeting, it was evident that with the exception of the textile industry, the slopes of the exposure response curves for lung cancer in the various chrysotile industry sectors were shallow, with no detectable risk or an extremely low level risk of lung cancer associated with exposure to chrysotile asbestos at and below lifetime cumulative exposures of 30 f/ml€years. No chrysotile related risk was detected at considerably higher exposures in the mining sector. INSERM's estimate of significant risk at 0.1 f/ml for chrysotile asbestos is at best questionable and definitely out of step with recent scientific evidence. (see Annex III)
INSERM dismisses data on the lower biopersistence of chrysotile
One of the major advances in the understanding of the mechanisms of fibre-related disease over the last decade or so has been the role of biopersistence or the durability of a fibre in lung tissue. Based on studies of synthetic fibres, it has been suggested that fibres of low biopersistence pose a lower lung cancer and mesothelioma risk than those of higher biopersistence. INSERM completely dismisses this notion and totally ignores evidence which indicates that chrysotile is substantially less durable in lung tissue than amphibole fibres. The fact that chrysotile clears very rapidly from the human lung following inhalation is viewed as an important factor in explaining why low level exposures are unlikely to present detectable risks to workers or the general population.
INSERM suggests that it is not feasible to have different exposure standards for chrysotile and amphiboles
The report suggests that having a less stringent standard for chrysotile would suggest that "chrysotile" is not carcinogenic. This is patently false and underscores the simplistic approach INSERM has taken in its study. If amphiboles are no longer used in France, it should be possible to deal with the problem of the past using adequate controls in the handling and removal of friable "asbestos" materials. The concern for the future then becomes chrysotile only products.
INSERM makes no attempt to evaluate the risks of substitutes
The report did not consider risks associated with substitutes. This is important, given the decision of many countries and advisory bodies to classify various natural and man-made fibres as probable or possible carcinogens. The fact that the EPA did not adequately consider the risks of substitute products was cited as one of the main reasons why the U.S. Court of Appeals overturned the EPA ban rule.
July 22, 1996