Asbestos: use, bans and disease burden in Europe

Abstract Objective To analyse national data on asbestos use and related diseases in the European Region of the World Health Organization (WHO). Methods For each of the 53 countries, per capita asbestos use (kg/capita/year) and age-adjusted mortality rates (deaths/million persons/year) due to mesothelioma and asbestosis were calculated using the databases of the United States Geological Survey and WHO, respectively. Countries were further categorized by ban status: early-ban (ban adopted by 2000, n = 17), late-ban (ban adopted 2001–2013, n = 17), and no-ban (n = 19). Findings Between 1920–2012, the highest per capita asbestos use was found in the no-ban group. After 2000, early-ban and late-ban groups reduced their asbestos use levels to less than or equal to 0.1 kg/capita/year, respectively, while the no-ban group maintained a very high use at 2.2 kg/capita/year. Between 1994 and 2010, the European Region registered 106 180 deaths from mesothelioma and asbestosis, accounting for 60% of such deaths worldwide. In the early-ban and late-ban groups, 16/17 and 15/17 countries, respectively, reported mesothelioma data to WHO, while only 6/19 countries in the no-ban group reported such data. The age-adjusted mortality rates for mesothelioma for the early-ban, late-ban and no-ban groups were 9.4, 3.7 and 3.2 deaths/million persons/year, respectively. Asbestosis rates for the groups were 0.8, 0.9 and 1.5 deaths/million persons/year, respectively. Conclusion Within the European Region, the early-ban countries reported most of the current asbestos-related deaths. However, this might shift to the no-ban countries, since the disease burden will likely increase in these countries due the heavy use of asbestos.


Introduction
The World Health Organization (WHO), joined by the International Labour Organization (ILO) 1

and the United Nations
Environment Programme, has called on countries throughout the world to eliminate asbestos-related diseases. 2-4 WHO advises that the best way to eliminate such diseases is to stop using all types of asbestos. 2 Although numerous countries have adopted national asbestos bans, many others continue to use asbestos at various levels. The use has declined 55% from its historical peak of 4.7 million metric tonnes per year in 1980, 5 but more than 2 million metric tonnes per year are still used worldwide. 6,7 WHO estimates that 107 000 global annual deaths are caused by mesothelioma, asbestos-related lung cancer and asbestosis. 8 In 2005, occupational exposure to asbestos was estimated to cause 43 000 mesothelioma deaths 9 and 7000 deaths due to asbestosis 10,11 worldwide. Of those caused by mesothelioma, 7000 were attributed to Europe. 9 However, the current and future burden of asbestos-related diseases in Europe has not been fully addressed, nor have such data been examined in relation to national asbestos bans.
Through the Parma Declaration on Environment and Health, member countries of the WHO Regional Office for Europe agreed on the need to eliminate asbestos-related diseases. 12 WHO and ILO specifically urged each country to formulate a national programme for eliminating asbestos-related diseases and develop a national asbestos profile 3 as milestones for implementing the Parma Declaration by 2015. 12 Asbestos use is a key item of a national asbestos profile. The volume of asbestos produced per person has been used to characterize the asbestos situation in various populations, [13][14][15] and can serve as a surrogate for population-level exposure. In addition, a list of asbestos use per capita across countries and over time 16 has been included in a recent monograph of the International Agency for Research on Cancer. 17 Per capita asbestos use has also been employed to estimate and predict asbestos-related diseases in different populations. 16,[18][19][20] We conducted a descriptive analysis of national data on asbestos use and asbestos-related diseases in Europe accounting for the status of national asbestos bans.

Methods
European countries were defined as the 53 countries in the European Region of WHO. Data on raw asbestos in these countries were obtained from the database of the United States Geological Survey 5,6 and its updated data file (RL Virta, United States Geological Survey, personal communication, March 6, 2013). The definition of use -production plus import minus export -followed that of the United States Geological Survey. 5 Data on asbestos use by country were available in 10-year intervals for 1920-1970, in 5-year intervals for 1970-1995 and annually for 1995-2012. We treated a reported negative value of asbestos use (e.g. reflecting storage from previous years) as zero. For years lacking data, reported data from the closest years were interpolated. We also retrieved information on the national status of asbestos bans 21,22 , ratification of the ILO Asbestos Convention 23 and health system ranking. 24 Data on asbestos-related diseases were extracted from the WHO mortality database; 25 these included the number of deaths recorded as mesothelioma (International Classification Objective To analyse national data on asbestos use and related diseases in the European Region of the World Health Organization (WHO). Methods For each of the 53 countries, per capita asbestos use (kg/capita/year) and age-adjusted mortality rates (deaths/million persons/ year) due to mesothelioma and asbestosis were calculated using the databases of the United States Geological Survey and WHO, respectively. Countries were further categorized by ban status: early-ban (ban adopted by 2000, n = 17), late-ban (ban adopted 2001-2013, n = 17), and no-ban (n = 19). Findings Between 1920-2012, the highest per capita asbestos use was found in the no-ban group. After 2000, early-ban and late-ban groups reduced their asbestos use levels to less than or equal to 0.1 kg/capita/year, respectively, while the no-ban group maintained a very high use at 2.2 kg/capita/year. Between 1994 and 2010, the European Region registered 106 180 deaths from mesothelioma and asbestosis, accounting for 60% of such deaths worldwide. In the early-ban and late-ban groups, 16/17 and 15/17 countries, respectively, reported mesothelioma data to WHO, while only 6/19 countries in the no-ban group reported such data. The age-adjusted mortality rates for mesothelioma for the early-ban, late-ban and no-ban groups were 9.4, 3.7 and 3.2 deaths/million persons/year, respectively. Asbestosis rates for the groups were 0.8, 0.9 and 1.5 deaths/million persons/year, respectively. Conclusion Within the European Region, the early-ban countries reported most of the current asbestos-related deaths. However, this might shift to the no-ban countries, since the disease burden will likely increase in these countries due the heavy use of asbestos.

Research
Disease burden related to asbestos use in Europe Takashi Kameda et al. of Diseases, 10th Revision [ICD-10], C45) 26 or any subcategory and asbestosis (ICD-10, J61) between 1994 and 2010. Asbestos-related lung cancers were precluded from our analysis because of difficulties in attributing causation. Separately, deaths recorded for malignant neoplasm of the pleura (ICD-9, 163), a condition generally synonymous with mesothelioma of the pleura, were counted between 1994 and 2009. To investigate countries that did not report data to WHO, we searched PubMed, governmental websites and other websites that we thought were credible [27][28][29][30] for national frequency data on asbestos use and asbestos-related diseases. To calculate rates, national population data for 1920-2012 were obtained in the following order -depending on data availability and reliability -from WHO, 25 the United States Census Bureau 31 or Lahmeyer. 32 To analyse asbestos use and asbestos-related diseases in each country, we calculated per capita asbestos use (kg/ capita/year) and age-adjusted mortality rates (deaths/million persons/year), respectively. Age-adjusted mortality rates were calculated using a direct ageadjustment method with reference to the WHO world standard population. 33 We analysed all countries individually, together and in groups based on national asbestos ban status, i.e. early-ban (ban adopted by 2000; n = 17 countries), lateban (ban adopted 2001-2013; n = 17 countries), and no-ban (no ban adopted as of 2013; n = 19 countries).
To provide continuity with data from currently existing countries, data on historical asbestos use from countries that had undergone political transitions (e.g. dissolution or unification) or had been combined with other countries or entities by the United States Geological Survey (n = 14) were obtained as follows. First, in the United States Geological Survey database, data for the Soviet Union  represented Kazakhstan and the Russian Federation combined. We apportioned the data between Kazakhstan and the Russian Federation according to the ratio of use recorded by these countries between 1995 and 2012. Second, data for West and East Germany  were combined into Germany. Third, data for Czechoslovakia  were apportioned to the ratio of asbestos use recorded by the Czech Republic and Slovakia between 1995 and 2012.
Similarly, the data for Montenegro and Serbia  were apportioned to the ratios recorded by Bosnia and Herzegovina, Croatia, Montenegro, Serbia, Slovenia and the former Yugoslav Republic of Macedonia between 1995 and 2012. In the United States Geological Survey database, the data for Montenegro and Serbia (1999)(2000)(2001)(2002)(2003)(2004)(2005) were combined; we apportioned the data to Montenegro and Serbia according to the sizes of the respective populations during this period. Finally, the combined data for Belgium and Luxembourg  were similarly apportioned to Belgium and Luxembourg.
To exploit all available data, we assessed asbestos use from 1920-2012 and asbestos-related disease mortality from 1994 to 2010 (the disease category for mesothelioma was included in the WHO mortality database in 1994). The period for asbestos use was divided into 1920-1970, 1971-2000 and 2001-2012. An early cut-off point -1970 -was chosen to allow a sufficient interval for observation of asbestos-related diseases and to be coherent with our previous studies. 16,18,34,35 A later cut-off point -2000 -was used to separate recent asbestos trends in both use and related diseases.
Based on our earlier finding that 1.0 kg/capita/year of asbestos use corresponded to 2.4 and 1.6-fold increases in mesothelioma deaths among men and women, respectively, 18 we considered this level to be high. Asbestos use of 2.0 kg/capita/year was considered very high. For asbestos-related diseases, we considered age-adjusted mortality rate levels for mesothelioma and asbestosis exceeding those of the world average (5.2 and 0.8 deaths/million persons/ year, respectively) to be high.
All data were compiled using Microsoft Excel (Microsoft Corporation, Redmond, United States of America). Age-adjusted mortality rates were calculated using SAS Version 9.1 (SAS Institute, Inc., Cary, USA).

Results
Andorra, Monaco and San Marino did not report any data for any of the indicators during the whole study period. Table 1 shows asbestos use and related disease mortality and the national asbestos ban status for each country. From 1920-1970, six countries recorded very high levels of asbestos use: Belgium, Cyprus, Denmark, Israel, Kazakhstan and Luxembourg. An additional 11 countries recorded high levels. From 1971-2000, the number of countries recording high and very high levels of use increased to 14 and 13, respectively. Between 2001 and 2012, most countries in Europe reduced their use, including those that previously used very high or high levels of asbestos: 43 countries used less than 0.5 kg/capita/year, of which 36 countries used less than 0.1 kg/capita/ year. In contrast, Kazakhstan, Kyrgyzstan and the Russian Federation recorded very high levels and Belarus, Ukraine and Uzbekistan reported high levels.
For three consecutive periods Kazakhstan and the Russian Federation reported very high or high levels of asbestos use, while Kyrgyzstan, Ukraine and Uzbekistan recorded such levels for the two latter periods, and 11 countries did so for the two earlier periods.
Deaths due to mesothelioma were reported by 36 countries during 3-17 years and by one country during 1-2 years, while 16 countries did not report at all. Deaths caused by asbestosis were recorded by 26 countries during 3-17 years, by seven countries during 1-2 years and 20 countries did not provide any reports. Among the 36 countries that recorded mesothelioma mortality for three years or more, 21 countries recorded high age-adjusted mortality rates led by Iceland, followed by Malta and the United Kingdom of Great Britain and Northern Ireland. Among the 26 countries that recorded asbestosis mortality for three years or more, 12 recorded high age-adjusted mortality rates, led by Malta, followed by Slovenia and Finland.

Discussion
This descriptive analysis of data in the WHO mortality database shows that 56% of all mesothelioma deaths and 41% of all asbestosis deaths recorded worldwide occurred in Europe, which accommodates 13% of the world's population. Combining these data with those from other sources showed that Europe accounted for 60% of the reported global deaths from asbestos-related diseases, excluding asbestos-induced lung cancer. During the periods 1920-1970 and 1971-2000, Europe used 48% and 58%, respectively, of all asbestos traded throughout the world. Europe can thus be characterized as the historical global centre of asbestos use and the current global centre of reported asbestosrelated diseases.
The three different ban groups had comparable population sizes but showed wide differences in the absolute numbers of asbestos-related disease deaths. The early-ban group reported the highest burden of asbestos related disease, while the no-ban group recorded the lowest. This could reflect differences in the reporting of asbestos-related diseases, as the majority of early-ban and late-ban countries reported asbestos-related disease data, whereas most of the noban countries did not. The three groups also differed in the quality rankings of their health systems, 24 prompting us to speculate that gaps may exist in the level of medical expertise and resources required to diagnose and report asbestosrelated diseases.
Almost all countries (14/17) that used asbestos at very high or high levels during 1920-1970 also demonstrated high mortality rates from mesothelioma and/or asbestosis. Kazakhstan and the Russian Federation did not report such data to WHO and no other comparable data could be identified. Switzerland did not report data to the WHO, but a substantial mesothelioma burden was found in a scientific article reporting national data. 28 Israel constituted the only identified exception to the relationship between asbestos use and asbestosrelated diseases.
We have shown earlier that the level of asbestos use correlates with the subsequent asbestos-related disease burden. 18 The asbestos-related disease burdens observed in the early-ban and late-ban countries are thus likely to be proportional to their levels of earlier asbestos use. The lower asbestos-related disease mortality currently being recorded by the no-ban group -despite higher levels of earlier asbestos use -is based on sparse data and likely reflects underdiagnosis and underreporting.
Asbestos use can be influenced by national policies. Therefore, we assessed the ratification status of the ILO Asbestos Convention. 23 The ratification rate was highest in the early-ban group and lowest in the no-ban group, suggesting a possible influence. It is also plausible that ratification may not have influenced use, but rather reduced asbestos exposure. A directive of the European Union (EU) mandated that all member states ban asbestos from 2005. 36 The early-ban and late-ban groups reduced their average use to less or equal to 0.1 kg/capita/year, respectively, whereas the no-ban group continued to use an average of 2.2 kg/capita/year. In the early-ban group the use varied between 0 and 0.19 kg/capita/year, which might be variable according to the extent of national exemptions. These are considerable reductions from the previously high levels of use observed between 1920 and 2000 in the earlyban group and between 1971 and 2000 in the late-ban group. In contrast the no-ban group recorded very high and high levels of asbestos use throughout the timeframe studied. The six countries with very high and high levels of use in the present century were all no-ban countries. Hence, although asbestos use was historically widespread and substan-tial across most of Europe, more recent use has been concentrated in the no-ban countries. We therefore speculate that the future burden of asbestos-related diseases will likely shift from the earlyban and late-ban countries towards the no-ban countries.
We previously used similar methods to analyse asbestos use and asbestosrelated diseases in Asia. 35 Although the previous and present findings can be roughly compared, some caution should be exercised. The earlier study adopted the United Nations Statistical Division definition of Asia, whereas the present study adopted the WHO definition of Europe, resulting in an overlap of 11 countries, including one early-ban country (Israel), two late-ban countries (Cyprus and Turkey), and eight no-ban countries (Armenia, Azerbaijan, Georgia, Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan). Also, the final years analysed for asbestos use and asbestos-related diseases in the earlier study were 2007 and 2008, respectively.
A strength of this study is the use of quantitative data from public databases to describe the situations in many countries. There were limitations, however, in the representation and comparability of the analysed data. For example, the data on use of asbestos were extrapolated for several countries that lacked specific data due to political transitions. In addition, data on the use of imported asbestos-containing products were not available. Moreover, as asbestos-related diseases are generally rare and difficult to diagnose, serious bias could have been introduced by countries having limited experience with asbestos-related diseases.