Wood dust in France. Trends in the population of exposed workers between 1982 and 2017 based on a job-exposure matrix assessment

Abstract

Objective

Many occupations and industries use wood as a raw material and wood dust is a well-known carcinogen. This study presents trends in occupational exposure to wood dust for all workers (employees and self-employed workers) in France between 1982 and 2017 and focuses on the exposed workers in 2017.

Methods

Exposures to this carcinogen were assessed using the Matgéné job-exposure matrix. Trends in the prevalence and proportion of exposure over the study period were estimated by linking the matrix with population data from the 1982, 1990, 1999, 2007, and 2017 censuses and are described for selected industry groups.

Results

The number of exposed workers to wood dust has decreased significantly over the last 40 years, from 466,900 potentially exposed workers in 1982 to 305,000 workers in 2017. The proportion of exposed workers has also decreased over time, although not uniformly across industries. Increases in the proportion of exposed workers are observed in certain industries, such as “Sawmilling and logging” (from 61.2% to 73.6% over the period for men) and “Finishing of sale premises” (from 3.3% to 6.2% for women).

Conclusion

This article is the first to describe occupational exposure to wood dust in France for all workers and to follow its evolution over the last 40 years. Occupations and industries still at risk in 2017 are also described with the aim of helping to improve prevention policies.

What’s Important About This Paper?

Wood dust is a well-known carcinogen, but industries in which wood is used have changed over time. This study describes the trend of occupational exposure to wood dust of all workers (employees and self-employed) in France between 1982 and 2017 using the Matgéné job-exposure matrix and census data. The proportion of workers exposed to wood dust has decreased significantly since 1982 from 2.1% (466,900 workers) to 1.2% in 2017 (305,000 workers).

Introduction

Wood is one of the most widely natural materials used in many industries, such as construction, furniture making, musical instruments, etc. As defined in the International Agency for Research on Cancer (IARC) Monographs, “Trees are characterized botanically as gymnosperms (principally conifers, generally referred to as ‘softwoods’), and angiosperms (principally deciduous trees, generally referred to as ‘hardwoods’). Hardwoods tend to be somewhat more dense, and have a higher content of polar extractives than softwoods” (International Agency for Research on Cancer, 1995, International Agency for Research on Cancer, 2012). The majority of operations at work that involve wood, from tree cutting to finishing, result in the release of varying levels of dust that can expose workers to wood dust. Depending on the mechanical actions exerted on the wood, the dust generated can be more or less fine. For example, woodcutting tasks emit coarser dust compared to sanding tasks.

Wood dust can cause various respiratory or skin diseases such as asthma, rhinitis, and eczema as well as cancer. The European Union has classified “work involving exposure to hardwood dust as carcinogenic” (European Council, 29 April 2004) and has set occupational exposure limits to 3 mg/m³ for respirable hardwood dust until 17 January 2023, and 2 mg/m³ after this date (European Council, 16 janvier 2019). The IARC classified wood dust (all species) as carcinogenic for humans (Group 1) for cancers of the nasal cavity and paranasal sinus in 1995 (International Agency for Research on Cancer, 1995) and for the nasopharynx in 2012 (International Agency for Research on Cancer, 2012).

Respirable wood dust is defined in the French labor code as any “solid particle whose aerodynamic diameter is not more than 100 micrometres or whose limiting fall velocity, under normal temperature conditions, is not more than 0.25 metres per second”(Ministry of Labour, 7 March 2008). In 2000, France added work involving exposure to respirable wood dust to the list of carcinogenic substances, preparations, and processes (Ministry of Labour, 18 September 2000) and, since 2004, has defined a binding regulatory exposure limit value of 1 mg/m³ over 8 hours in working atmosphere, regardless of the wood type. In addition, it created compensating tables for occupational diseases caused by wood dust among employees for both industrial workers (Ministry of Labour, 14 February 1967) and agricultural workers (Ministry of Labour, 15 January 1976). Approximately 75 cases of occupational diseases are compensated each year in France related to wood dust for a cost of approximately 20 to 30 million euros (National Health Insurance Fund, 2010 to 2018).

In Finland, 2.3% of employees were considered exposed to wood dust (57,500 employees) in 2020 (Kauppinen et al., 2013) and 1.3% (280,000) in Italy in 2003 (Mirabelli and Kauppinen, 2005). In France, some estimates of the total number of exposed employees were published: approximately, 308,000 (1.3%) employees were exposed to wood dust in France in 2000–2003 (Kauppinen and Vincent, 2006), and 444,200 (1.8%) in 2017 (Matinet et al., 2020). However, the recent estimates published in France mainly concern the population of employees, and it is, therefore, important to be able to describe the exposure for both employees and self-employed workers. Over the last 10 years, the number of self-employed workers has been increasing sharply (Moisan, 2016). A single published study estimates the total number of exposed workers (employees and self-employed) in France at 180,000 workers in 1990 (Kauppinen et al., 2000).

The Matgéné program (Févotte et al., 2011) aims to elaborate job-exposure matrices (JEM) for the French population and to estimate occupational exposure indicators for all the population at work, including employees and self-employed workers. Knowledge of the population at work exposed to wood dust helps meet the objectives of Santé publique France’s occupational risk surveillance, and to identify the industry and occupational groups with the largest number of exposed workers. A matrix has been specifically developed to assess occupational exposure to respirable wood dust in France.

The objective of this study was to estimate the number and proportion of exposed workers to wood dust in France, based on the exposure assessment provided by the Matgéné JEM. The trends were described between 1982 and 2017 according to sex and certain industry groups. A focus on the exposed workers by industry and occupation is also provided for 2017.

Methods

The wood dust JEM was developed as part of the Matgéné program (Févotte et al., 2011). This matrix is a table giving an exposure probability for each job (defined as an occupation in an industry) considered exposed to wood dust (Supplementary data).

Exposure assessment

The JEM, which was developed a priori by a team of three occupational hygienists, assessed exposure to respirable wood dust for all jobs in France between 1970 and 2020, regardless of the wood species used. Two occupational hygienists independently assessed the same jobs, then their results were compared and the assessment differences were discussed. The exposure assessments were reviewed by a third occupational hygienist, and a consensus was reached for each discrepancy.

Jobs were defined according to the two French classifications for industries and occupations, the classification of economic activities (NAF) (The National Institute of Statistics and Economic Studies, 1999, The National Institute of Statistics and Economic Studies, 2003a, The National Institute of Statistics and Economic Studies, 2008) and the classification of professions and socio-professional categories (PCS) (The National Institute of Statistics and Economic Studies, 2003b, The National Institute of Statistics and Economic Studies, 1994). The matrix provided a historical exposure assessment by periods. These periods were defined according to successive regulations and changes in the conditions of exposure to wood dust in different industries (Table 1 and Supplementary data).

Table 1.

Characteristics of the “Wood dust” matrix of the Matgéné program.

Job classifications used	Start date	End date	Exposure probability	Exposure periods
PCS a 1982 × NAF b 1993	1970	2020	Very low: [1–10%] Low: [10–30%] Moderate: [30–70%] High: [70–90%] Very high: [90–100%]	All sectors except carpentry • 1970–1995 • 1996–2005 • 2006–2020 Carpentry • 1970–1984 • 1985–2020
PCS 2003 × NAF 2003
PCS 2003 × NAF 2008

^aFrench occupations classification (PCS)

^bFrench economic activities classification (NAF)

The exposure assessment was averaged for each job, averaging across the variability of all the tasks performed and the exposure situations encountered in the job. The exposure probability, defined as the proportion of exposed workers in the job over the given period, was assessed for each job considered to be exposed (Table 1).

Population

To study the trend of occupational exposure to wood dust in France, population census data produced by the National Institute of Statistics and Economic Studies (INSEE) (The National Institute of Statistics and Economic Studies., 2017) and including information on occupation and industry were used. They cover all jobs of workers in France by age, sex, and department of residence for the years 1982, 1990, 1999, 2007, and 2017 and the worker status (employee or self-employed) for 2007 and 2017. The self-employed workers include those working on their own account (farmers, shopkeepers, craftsmen, and liberal professionals) and employees include those working for a company. Specific codes exist in the PCS classification for separate self-employed workers and employees in the same occupation.

The census method has evolved throughout the study period. Between 1982 and 1999, every individual living in France was asked to answer the census. Since 2004, population censuses have become annual and based on a 5-year sample. As a result, the 2007 and 2017 censuses take into account data from the annual censuses of 2005 to 2009 and 2015 to 2019. This was not considered to be a bias given the adjustments made by INSEE to make the data collected from two different methods comparable (The National Institute of Statistics and Economic Studies, 2014).

Exposure indicator estimates

The census data were linked with the wood dust JEM using the occupation and industry codes and the exposure period corresponding to the census year. The number of workers exposed to wood dust was calculated by multiplying the exposure probability provided by the JEM (mid-point value of the probability range) by the number of workers in the job. The proportion of exposed workers was obtained by dividing the sum of exposed workers by the number of workers in the population. A sensitivity interval (SI) was calculated by taking the lower and upper bounds of each probability range. The trend between 1982 and 2017 (increase or decrease) was based on the absence of overlap between the SIs of these two years. All results referred to workers in metropolitan France (defined as the European territory of France) aged between 20 and 74. The indicators were described by sex and by major industry (NAF) over the study period. Changes in job classifications between 1982 and 2017 make it difficult to follow the whole population by occupation and industry over time. However, a selection of industries that can be easily traced in the different classifications was made, allowing the study of the evolution in each group between 1982 and 2017 (Table 2). The distribution of exposed workers by major occupation (PCS 2-digits) was calculated for the study period and the same information was also estimated for 2017 only at a more precise level by industry (NAF 2-digits) and occupation (PCS 3-digits).

Table 2.

Selection of industry sectors where the jobs inside the classification’s codes haven’t changed over time and classification versions.

Industries	Industry codes used in the 2007 and 2017 population census (NAF 2008)a	Industry codes used in the 1982, 1990 and 1999 population census (NAF 1993)
Sawmilling and logging	02.20Z Logging	02.0B Logging
	16.10A Sawmilling and planing of wood, excluding impregnation	20.1A Sawmilling and planing of wood
Impregnation of wood	16.10B Impregnation of wood	20.1B Impregnation of wood
Manufacture of wood-based panels	16.21Z Manufacture of veneer sheets and wood-based panels	20.2Z Manufacture of veneer sheets; manufacture of plywood, laminboard, particle board, fibre board and other panels
Finishing of sale premises	43.32C Finishing of sale premises	45.4L Layout works of sale premises
Floor and wall covering	43.33Z Floor and wall covering	45.4F Floor and wall covering
Manufacture of paper and paper products	17.11Z Manufacture of pulp	21.1A Manufacture of pulp
Services to buildings and landscape activities	81.30Z Landscape service activities	01.4B Landscape gardening

^a: French adaptation of the European classification NACE Revision 2 and of the international classification ISIC Revision 4. For example, NAF2008—02.20Z Logging corresponds to 02.20 in NACE2008 and 0220 in ISIC2008

Results

The proportion of workers exposed to wood dust was estimated, by sex, between 1982 and 2017 (Figure 1). Approximately, 466,900 workers aged between 20 and 74, were exposed in 1982 (2.1% of workers SI [1.7-2.5]) and 305,000 in 2017 (1.2% [0.9-1.5%]). In 2007, 23% of workers exposed to wood dust were self-employed (76,900/ 335,900), versus 30% (93,000/ 305,000) in 2017.

Fig. 1.

Number and proportion of workers exposed to wood dust between 1982 and 2017 by sex.

Trend in workers exposed to wood dust by sex

The number of exposed workers has decreased over the last few decades for both men and women.

Among men, 426,500 workers were exposed in 1982 (3.2% SI [2.6-3.9]) and 280,600 workers in 2017 (2.1% SI [1.6-2.7]). Although the total number of exposed men workers decreased over the period, differences were observed by industry. Indeed, the number of workers exposed to wood dust has sharply increased in the “Services to buildings and landscape activities” industry (2,578 exposed workers in 1982 vs 22,271 in 2017). This industry has expanded greatly over the period (+ 348% workers) and has had the highest increase in the proportion of exposed workers (+ 93%) (Figure 2). Moreover, the number of exposed workers in the “Finishing of sale premises” industry has increased by more than 2-fold (949 vs 2,486) and the proportion by 70% (20.5% in 1982 vs 37% in 2017). The industries with the highest proportion of exposed workers in 2017 were “Sawmilling and logging” industries (73.6% of exposed workers) and the “Manufacture of wood-based panels” (62%).

Fig. 2.

Number and proportion of exposed men and women by industry and census date.

Among women, 40,400 workers were exposed in 1982 (0.4% SI [0.4 to 0.5]) and 24,400 were exposed in 2017 (0.2% SI [0.1 to 0.3]). Two industries showed an increase in the number and proportion of exposed women workers between 1982 and 2017: “Finishing of sale premises” and “Floor and wall covering” industries (3.3% vs 6.2% and 2.1% vs 4.9%, respectively). Although the “Services to buildings and landscape activities” showed a sharp increase in the number of exposed women workers over the period (145 vs 1,015 workers), the proportion of SIs was overlapping. In 2017, the two industries with the highest proportion of exposed workers were “Manufacture of wood-based panels” (32.4%) and “Sawmilling and logging” industries (28.2%).

Distribution of all exposed workers over the period and in 2017

The changes in the occupation distribution in the exposed workers over the period showed that craftsmen has increased from 16.7% in 1982 to 26.6% in 2017. Conversely, the proportion of exposed wood working employees (PCS 62, 63, 67, and 68) has been decreasing over time, although it represents the majority of the exposed workers (68.8% in 1982 vs 54% in 2017); this trend is observed especially in the unskilled workers (PCS 67 and 68: 29% in 1982 vs 17.6% in 2017) and to a lesser extent in the skilled workers (PCS 62 and 63: 39.8% in 1982 vs 36.4% in 2017) (Figure 3).

Fig. 3.

Trend of the distribution of occupations among exposed workers between 1982 and 2017.

In addition, we provide results for all industries in the 2017 census (and not only for workers from the selected industries groups in Table 2). Presently, many workers were still exposed to wood dust in 2017. Almost 70% of the workers exposed are engaged in four industries: “Specialized construction work” industries (NAF=43) with 124,350 workers (41%), “Manufacture of wood and wood products” (NAF=16) with 36,050 workers (12%), “Wholesale trade” (NAF=46) with 25,990 workers (9%) and “Services to buildings and landscape activities” (NAF=81) with 23,950 workers (8%) (Table 3). The three occupations with the highest number of exposed workers in 2017 were “Craftsmen in construction, public works, parks, and gardens” (PCS=211) with 68,310 workers (22%), “Skilled handicraft workers in construction” (PCS=632) with 52,930 workers (17%) and “Industrial-type skilled workers - Other industries” (PCS=627) with 43 350 workers (14%).

Table 3.

Number and distribution of workers exposed to wood dust by occupation and industry in 2017.

Occupations (PCS code)	N exposed workersb	Distribution of exposed workers (%)	Industries (NAF a code)	N exposed workersb	Distribution of exposed workers (%)
Craftsmen in construction, public works, parks and gardens (PCS = 211)	68 310	22	Specialised construction activities (NAF = 43)	124 350	41
Skilled handicraft workers in construction (PCS = 632)	52 930	17	Manufacture of wood and of products of wood and cork, except furniture; manufacture of articles of straw and plaiting materials (NAF = 16)	36 050	12
Industrial-type skilled workers—Other industries (textiles, tanning, clothing, industrial leather work, woodworking, furniture, paper-cardboard, printing) (PCS = 627)	43 350	14	Wholesale trade, except of motor vehicles and motorcycles (NAF = 46)	25 990	9
Unskilled handicraft workers in construction (PCS = 681)	25 730	8	Services to buildings and landscape activities (NAF = 81)	23 950	8
Industrial-type unskilled workers—Other industries (textiles, tanning, clothing, industrial leather work, woodworking, furniture, paper-cardboard, printing) (PCS = 675)	18 640	6	Manufacture of furniture (NAF = 31)	17 240	6
Other occupations	96 000	31	Other industries	77 380	25
Total	304 960	100		304 960	100

^a: 2-digit NAF code equals 2-digit NACE code (Eurostat, 2008).

^b: for Craftsmen in construction, public works, parks and gardens (PCS = 211): 68 310/ 304 960 = 22% (rounded to ten).

Discussion

Results overview

The proportion of workers exposed to wood dust has decreased sharply since 1982, from 2.1% of all workers (466,900 workers) to 1.2% of all workers in 2017 (305,000 workers). However, this decline is not uniform across all industries. In construction in particular, the proportion of exposed workers in men has remained relatively stable, given the constant use of wood over time. Among men, there has been a significant increase in the proportion of exposed workers over time in the “Services to buildings and landscape activities” (from 13.7% in 1982 to 26.5% in 2017). This could be mainly explained by a structural reorganization of jobs within the sector, because assessment for this sector has not changed over the study period. The proportion of exposed women workers increased over the period in the “Sawmilling and logging” and “Finishing of sale premises”. This increase can be explained by the fact that over the studied period, the proportion of women at work has increased, and women are now more likely to be employed in jobs historically reserved for men, especially jobs that involve wood dust exposure.

Wood dust regulations have changed over the period covered by the study. These developments concern occupational exposure limit values and have mainly had an effect on exposure levels, which were not considered in this study. Exposure limits have had relatively less impact on the proportion of workers exposed within each occupational group, given that wood-using industries have generally continued to involve wood. However, there are some exceptions to this observation, such as the introduction of new materials in specific industries (PVC or aluminum in joinery, various plastics in outdoor furniture, etc.). Since workers specialize in handling a particular material (aluminum or PVC joiners, for example), the introduction of these materials has resulted in a lower number of workers exposed to wood dust. However, this use of innovative materials alone cannot explain the overall decrease in the number of exposed workers. The de-industrialization that France has been ongoing for the past 30 years, resulted in a 36% loss of industrial jobs between 1980 and 2007 (Demmou, 2010). These job losses are linked to the increased mechanization of labor, which leads to the loss of manual jobs that are potentially more exposed to wood dust and other hazards, in favor of managerial or administrative jobs. De-industrialization seems to us to be a more plausible explanation for the decline in the number of workers exposed to wood dust, and this decline is not reversed by the increase in the use of wood in some industries (e.g., wood construction in the building industry). There is also an impact linked to globalization and industry transfers. In France, for example, wood logs are sent to China to be sawn and processed, and the wood returns to France in the form of finished products (Delamarche, 2017). This effectively eliminates some industrial industries in France.

Although the total number of exposed workers decreased between 2007 and 2017 from 335,900 to 305,000, the number of exposed self-employed workers increased from 76,900 to 93,000. The specificity of our study is to provide results on the population of self-employed workers exposed to wood dust who do not have medical monitoring by occupational health physicians and are therefore rarely studied. Our results show an increase in the number of self-employed workers exposed. This work presents occupational exposure to wood dust in France for the entire population at work (employees and self-employed).

Job coding

Over the course of the various censuses, the versions of the occupation and industry classifications used to code jobs have evolved, from PCS1982 and NAP1973 in 1982 and 1990, to PCS1982 and NAF1993 in 1999, and to PCS2003 and NAF2008 in 2007 and 2017 (Ministry of Economy and Finance, 1976). In addition, a specific coding system was used for agricultural industries for the years 1982 to 1999. In order to be able to link the JEMs, the 1982 and 1990 censuses had to be cross-walked from NAP1973 to NAF1993 using distribution tables provided by INSEE and expertise for the agricultural industries for the year 1999. The numbers may therefore differ due to cross-walking and caution should be exercised when interpreting the results for this sector. Ultimately, our estimates are based on a large population of more than 20 million workers.

Exposure assessment

The JEM is dependent on the classifications used to build it and has certain limitations related to the coding of jobs. Indeed, the classification codes sometimes group together heterogeneous occupations or industries with potentially different exposure assessments, making it necessary to define an average exposure for a single code. Some occupational exposures could not be included in the JEM, due to the job classification used, which do not always identify a particular occupational situation that may involve wood dust exposure (e.g. parquet layers, pruners, technical high school teachers). In order to take into account each task variability over a year (seasonal variability, specific tasks depending on the required job…), the JEM provides an averaged exposure assessment.

The mid-point values of the exposure probability provided by the JEM were used to estimate the exposure proportion. For example, the value of 40% was used for a probability range between 30% and 50%. The results presented in this article are therefore subject to a level of uncertainty related to the assessment, which is taken into account by the SIs calculated by taking the lower and upper bounds of each probability range.

However, JEMs are still the most suitable tools for assessing exposure in large populations where individual assessment is not possible. In addition, the Matgéné JEM gives an assessment for all jobs in France and provides historical information since 1970, which has made it possible to follow the trend of wood dust exposure for more than 30 years.

Comparison with other studies on the French population

Different studies in the French population have estimated the number of workers exposed to wood dust. One European study that included the French population in 1990 showed that 180,000 workers were exposed (Kauppinen et al., 2000), compared to our results of 422,000 (333,600 to 524,700) workers exposed in 1990. The difference between these two estimates can be explained by the different methods used; the European study was based on Finnish and American exposure data, readjusted by expertise to estimate the number of French exposed workers, and our results are based on the French population data linked to a JEM based on the French regulation. Another European study that took place in 2002, showed 308,000 French employees exposed (Kauppinen et al., 2006), while our result for 2007 gives 259,000 [195,300 to 333,200]. For France, 60% of the workers exposed to wood dust were exposed above the OEL of 1 mg/m³. The difference observed in the number of exposed workers, is probably due to methodological differences. In the European study, a questionnaire survey was used in each European Union member country and analyzed by the Finnish Institute of Occupational Health to calculate prevalence, which was then followed by a national expert assessment. In contrast, our study used the JEM and census data.

Finally, a French study of the French Directorate for Research, Studies and Statistics (DARES) (Matinet et al., 2020) estimated that 444,200 employees were exposed to wood dust in 2017 (1.8%), nearly twice the values in our study, which showed that 212,200 (156,500 to 277,100; 0,9% [0.7 to 1.2]) employees were exposed in 2017. The difference may be explained by the method used, the Dares study was based on the expertise of occupational physicians on the week prior to the interview and statistical analysis was based on a sample of 26,500 employees compared to our study which is based on average exposure over the period and statistical analysis on the entire population of employees in France (23 million in the census of 2017).

Comparison with international studies

At the international level, several teams have assessed the population of workers exposed to wood dust. A Canadian team (Peters et al., 2015) assessed exposure using the expertise of industrial hygienists based on the CAREX database, linked with the 2006 Canadian Population Census. This study showed that 2% of Canadian workers were exposed to wood dust, while our study concluded that 1.3% [1.0 to 1.7] of French workers were exposed. The wooded areas of the two countries—3.5 million km² of forests in Canada vs. 0.18 million in France in 2016 (Atlasocio) and the importance of the respective wood industries, Canada being the world’s largest exporter of forest products (Government of Canada)—can largely explain this difference in prevalence.

Three European studies carried out using a JEM adapted to the local context found results similar to ours, with 1.3% of workers exposed to wood dust between 2000 and 2003 in Italy (Mirabelli and Kauppinen, 2005), around 1.6% in 2000 to 2003 in Hungary, Greece, Sweden (Kauppinen and Vincent, 2006) and 1.5% in Catalonia (Spain) in 2009 (de Grado et al., 2014). Our results are of the same order of magnitude, with 1.6% [1.2 to 1.9] in 1999, 1.3% [1.0 to 1.7] in 2007, and 1.3% [1.0 to 1.7] in 2011. Our results seem to be confirmed by the figures for forests and wooded areas in Europe (Europe Information Center). The wooded area is slightly higher for Spain, with 0.28 million km², and relatively close for Italy, with its 0.11 million km², compared to the 0.18 million km² of wooded areas in France. The common European regulations in these three countries are also a realistic explanation for these relatively comparable figures.

A Finnish study with a similar methodology (JEM linked with Finnish population censuses) shows a significant decrease in wood dust exposure between 1950 and 2020 (Kauppinen et al., 2013): 3.4% in 1950, 2.7% in 1990, 2.4% in 2008, and 2.3% in 2020 (predicted proportion). Our results are somewhat lower with 2.1% [1.7 to 2.5] in 1982, 1.9% [1.5 to 2.4] in 1990, 1.3% [1.0 to 1.7] in 2007, and 1.2% [0.9 to 1.5] in 2017, but they also indicate a decrease of the exposed population. The differences in the industrial context, production, and use of wood between the two countries may explain the difference in exposure proportion. Although less forested in terms of total area than Spain (0.23 million km²), Finland has the largest proportion of forested land in Europe (76% vs. 55% in Spain, 37% in Italy, and 32% in France).

Conclusions

The JEM takes into account every job retrieved in the French working population and therefore it can be used to assess exposure when no question on wood dust exposure is asked in a study (Geoffroy Perez et al., 2012, Goldberg et al., 2017). The Matgéné JEM is unique because it includes self-employed workers, and can be used to establish exposures among a cohort of self-employed workers (Geoffroy Perez et al., 2012). Future development of the JEM will allow exposure levels to be quantified and to identify the occupations and industries with the highest magnitude of exposure.

As part of a second phase of research, attributable risk fraction for the diseases associated with this exposure (cancers of the nasal cavities, paranasal sinus, and nasopharynx) will be estimated using lifetime exposure prevalence estimates from an occupational calendar sample. A first study was carried out in collaboration with IARC in a published study based on our cross-sectional proportion estimates (Marant Micallef et al., 2021). This study shows that in France between 15 and 31 cases of nasal cavity cancers and between 2 and 47 cases of nasopharyngeal cancers were attributable to occupational exposure to wood dust in men (between 0 and 2 cases of nasal cavity and between 2 and 47 cases of nasopharyngeal cancers in women). In 2018, in France, 241 incident cases were observed for nasopharyngeal cancers and 552 for nasal cavity, sinus, and ear cancers for men, and 85 and 254 for women, respectively (Defossez et al., 2019b, Defossez et al., 2019a). The number of cases due to occupational exposure to wood dust can be compared to the number of cases of occupational diseases reported each year (about 69) (National Health Insurance Fund., décembre 2018).

In conclusion, this study identifies the occupations and industries exposing workers to wood dust in France between 1982 and 2017 and highlights the total number of exposed workers. To our knowledge, this is the only study carried out on the entire French population at work (employees and self-employed) over a large period. These results constitute a baseline regarding the trend in occupational exposure to this carcinogenic hazard and are intended to help improve protocols of prevention policies on targeted occupational groups. The Matgéné program actively takes part in the surveillance of the evolution of this exposure in France, by estimating the number of workers exposed in the past decades, presently, and in the future.

Author contributions

CP supervised the project. LG, SD, and CP have developed the JEM. MH analyzed the data and interpreted the data in collaboration with LG, SD, and CP. LG, SD, MH, and CP contributed to the final version of the manuscript. All authors approved the final version.

Funding

None.

Ethical approval

Not applicable.

Conflict of interest statement. The authors declare that they have no competing interests.

Data availability

The exposure indicators will be accessible on the Géodes website of Santé publique France (https://geodes.santepubliquefrance.fr), and the JEM will be available for consultation on an Internet website (www.exppro.fr); the matrix files can also be made available via DSET-matgene@santepubliquefrance.fr.