Exposure to silica dust in the Danish stone industry

E. Exposure to silica dust in the Danish stone industry. Scand J Work Environ Health 1989;15:147-53. Exposure 10 silica dust among Danish stone workers was as sessed from data collected in 1948- 1980. Arter I <nO, the exposure level was given in milligrams per cubic rueter and an exposure index (concentration of respirable dust divided by the threshold limit value for quartz) was calculated. The median index was 2.1 for the road and building material industry and 0.6 for the stonccuuing industry. Crushing showed a median index of 2.6, compared to 1.0 for drilling, 0.9 for sieving, and 0.5 for CUlling. The median index for the road and building material industry was higher for Bornholm than for other parts of Denmark. The median index of the sronecuuing industry was 1.0 for Copenhagen and 0.5 for other parts of Denmark, and no data were available for Bornholm. Before 1970. the exposure level was given in particles per cubic meter, and few data were available on quartz content. Crushing showed the highest exposure before 1970.

C rys ta lline silica is a well-k nown agent of pneumoconio sis, a nd in recent yea rs in vest igat io ns o n h umans a nd an ima ls ha ve st re ng htened th e hypothesis th a t silica may a lso be a carcino gen . The Int erna t io nal A gen cy fo r Research on Cancer (IA RC) r ecen tl y co ncluded t ha t th er e is sufficient eviden ce fo r th e ca rcino ge nicity of crysta lline silica in expe rimental a ni m al s and limited ev id e nce for its carcino ge nicit y in h u mans ( I) .
In order 10 p ro vid e furt her inform ation on th e po ssible ca rc ino gen icity of silica in humans, we ide ntified a co hort o f st o ne workers in De nm ark. An a ssessme nt of th e exposure to silica d ust was carried o ut from dat a retrieved o n all m ea surem ents fro m t he s to ne industry avai lable in th e archives o f th e Danish National Inst it ute of Occu pat iona l H eal th . The an a lysis of t he se ex pos u re data is present ed in th is pa pe r.
T he s to ne industry in Denm a r k ca n be di vided int o two m a in branc he s: t he sto nec utt ing ind ustry and t he road a nd b u ild ing m a te r ial in d us try . T he work processes in th e sto nec u u ing ind ustr y incl ud e sawi ng. edg ing, s ur fa ce fin ishing by g rin d ing , ha mmeri ng , or poli shi ng , a nd ca r ving . In th e ro a d m a te ri al indust ry stones ar e cru shed by machi nes, an d th e g rav el is sieved throug h m eta l screen s to be so rted b y size. The isla nd Born ho lm , lo cat ed so ut h of Swed en , is t he o nly p la ce in De n m ark w here g ra nite ro ck s a re found , a nd g ra nite is wo n in open cast quarr ies . G ra nite has been th e m ai n m a teria l in th e Dan ish s to ne ind ust ry, bu t sa nd- s to ne has al so been used in th e sto nec utt ing industry, a nd flint ha s bee n used in t he road ma te ria l in du st ry . Flint is 100 010 q ua rt z (2) . Sands to ne co ns ists of g ra ins o f q uart z ce me nted toget her wi th cla y (3) . G ra nit e is a mi xt ure of qu art z , fe ldsp ar , a nd mic a (3) . The qu artz con tent in Bornh olm g ranite varies be t ween 15 a nd 35 010 (4); tabl e I gives t he n umber o f wo rkers in the Da nish sto ne in du st ry in 19 30-1 980 (5-10) .

Material and methods
T he d ata set in cluded all meas urem e nt s m ad e in the Dan ish st o ne ind ustry in D enmark b y t he Danish N atio nal In stitute o f Occupa t ional H ealth from 1948 to 1980 . These m ea surements were m ad e a t th e req ues t o f th e lo ca l labor inspection officers and t h us reflect eith er a reas wh er e a special effort for r eg ulation ha s been m ade or areas w he re co m p lai nt s ha d been m a de about work cond itio ns . The records were retrieved ma n ually fro m the arch ives of th e In s t itu te . No record was availa ble before 1948, as t he Institu te was founded in t ha t yea r. The av a ila b le in fo rm a ti on was the dat e of the in vestigation , the na m e of th e com p an y, the na me o f t he employee whe n a per sonal a ir sa m -  pier was used, the work process or the place of work involved during the sampling, and, in some cases, the type of stone material processed. Details on the sampling and analytical methods were also given.
The data set can be divided into twosubsets according to the date of investigation. Before 1970 most of the air samples were taken in the general environment at a fixed location with an impinger or an electrostatic precipitator or by filter sampling, and with sampling periods generally of less than I h. The results were expressed in number of respirable particles per cubic meter (diameter range 0.5-5 urn), counted by optical microscopy. After 1970 most of the samples were taken in the breathing zone with personal air samplers. The sampler met the requirements of the Johannesburg pneumoconiosis conference (II). The sampling period was generally 6-8 h , and the time-weighted average concentration of respirable dust was expressed in milligrams per cubic meter. The content of crystalline silica in both the general environment and the personal samples was determined by X-ray diffraction. No attempt has been made to convert the count of particles into weight or vice versa, and the two subsets of data have been analyzed separately. Measurements of respirable dust were carried out in nine stonecutting companies and in II road and The results of the personal sampler measurements were compared to the threshold limit value (TL V) for respirable dust as established by the American Conference of Governmental Industrial Hygienists (12): The comparison with the TL V was expressed by a wellestablished index, which is the ratio of the dust concentration (in mg/ru') to the TL V. With no cristoballite or tridymite, the TLV is 0.1 mg/rn ' for exposure to pure quartz and 1.0 mg/m ' for exposure to 8°10 quartz. As is a common practice in industrial hygiene (13), the median exposure level was used in the analysis.

Comparison of exposure levels associated with the use oj granite and jlint
Cristoballite was detected in eight measurements from two road material industries and ranged between 2 and 11 070 of the respirable dust. Tridymite was not present in any of the measurements from the Danish stone industry. Since the quartz content of flint is higher than that of granite, an attempt was first made to assess the exposure levels to crystalline silica associated with each of them. The results are shown in table 3 for the personal sampling measurements. The respirable dust concentrations were 1.3 and 1.0 mg/rn ' for flint and granite, respectively. The median quartz content in the dust was 23 % for flint and 13 % for granite. These values lead to a median exposure index of 4.5 for the work operations in which t1int was invo lved , compared to a n ind ex of 2.0 for work operation s with gra nite.

Comparison of exposure levels in the road material industry and in the stonecutting industry
From th e su bset of measurement s carried o ut in the genera l env iro nment , the median of th e respirab le du st concent ration was 125 pa rti cles/em' in the road material indu str y a nd 205 particles/em ' in th e sto necuttin g industry. The quartz content of th e du st was not sta ted in an y of the measurements from th e sto nec uttin g industry and in onl y half of the me asurem ent s from th e road material industry, with a med ian of 23 0/0 qu art z. Th e s ubset of per sonal sa mpling dat a was more informative. The results are summari zed in tab le 4 a nd figure I . As can be see n from tab le 4, the respirable du st co ncentra tio n was high er in th e road mat eria l industry th an in the sto necut ting ind ustry. The med ia n quart z content of th e dust was 13 070 in th e road ma teria l industry and 9 0J0 in th e stonecutting ind ustry for a median exp osure index of 2.1 in the ro ad mate ria l ind us try and of 0.6 in the sto necutt ing indu str y fo r all of Denm ark . Figure I shows that approximately 75 and 45 0J0 o f the measurem ent s exceeded th e TL V in the road material industr y a nd the sto necutt ing ind us try , resp ecti vely. T he a na lysis by region shows that th e medi an expos ure index wa s the highe st in Bornho lm for th e road mat erial industry (table 4). No data were available in Bornholm for the sto necutting industr y. The median exp osure index wa s twice as high in C o pe nhagen as in other parts of Denmark ( 1.0 vers us 0.5), a lthough few me asu rements were a vailable .
Flint was thc most frcq ue ntly used material in the road mat erial ind us try, and gra nite was the main material in th e sto necutt ing industry. Th e differences between the tw o industries cou ld not be ascribed to type of min eral. A compa riso n of work pr ocesses in which onl y granite was used showed the sa me med ian qu a rtz content of respirab le d ust for the two industri es, 13 and 15 11 /0, respecti vely. Th e median expo sure ind ex was , however, 3.0 in the road material industry and 1.0 in th e sto ncc utting ind ustry.  _. . .._ --------a tn addi ti on to quartz . t wo samples co ntained 2 % and one sam ple each con tain ed 4, 5, and 7 % c ris toba llite. Cristoball ite has been taken into account in the ca lculation o f the expos ure in dex. b Co nce ntrat ion of respi rable dus t divided by the th res hold li mit value (12) for quartz. C No data avai lab le fo r Bornh ol m from the sto nec utting ind us try.
5 Figure 1_ Cumulative frequency distributi on s of the exposure indices (concent rat ion of resp irable dust divided by t he thr eshol d limit value for quartz) for crystalli ne silica calculated from person al air samples from the stonecutt ing (1) and the road and bui lding materi al (2)    Com parison of exposu re levels in diff erent work processes T he work pro cesses in the roa d mat erial ind ustr y were di vided in to drilli ng, crush ing, and sieving. T he number o f measurements from t he st onec utti ng industry was sma ll, and hand cutt ing and surface finishing were the refo re gro uped into a single cat egor y.
Th e results of the measu remen ts ca rr ied out in the genera l environment are pre sen ted in tab le 5 and figur e 2. No da ta were avai lable fo r drilling . Fro m seven measurements only, crushing ap peared to be the du stiest wor k op erat ion . In figure 2 the d istribu tion of the part icle conce ntration is pre sente d by wor k o perat ion , a nd the med ia n du st expo sure levels of the work processes are readil y est ima ted fr om the gra phs. It sho uld be noted tha t the qu ar tz con tent of the respirable fra ction has not been taken into acco unt in that figure , since the percentage of q uartz was not measured for cutt ing and was measur ed in fo ur cases only for cru shing .
The subset of personal air sam pling data is presented in tabl e 6. The respirable du st concen tration associated with crushi ng was approximat ely twice as high as th e con centrat ion associ ated with the other wo rk ope rations, for which the median du st co ncentration was below 1.0 mg /rn' . T he qua rtz content ran ged from 7 to 14 070; the high est va lue was ob served fo r crus hing an d the lowest for cutt ing. Th e median exposur e index was 2.6 for crus hi ng, and 80 0J0 of the measu rements were above the TL V. T he med ian expos ure index was 1.0, 0.9 , and 0.5 fo r drilli ng, sieving, and cutting, respectively.
Finally, the dat a previo usly grouped in the cutting catego ry were a na lyzed . A co mparison was mad e between surface fin ishing (five measurements) and hand cutt ing (13 measurements). Th e media n respirabl e dust concentration was 0. 8 (range 0.4-3 .0) mg/rn ' for surface finishing and 0.6 (ra nge 0.2-5.4) mg/rn ' for hand cutting. Three measurements ou t of five exceeded the T LV for surface fin ishing, whereas the correspo ndin g values for hand cutti ng were fou r out of thirteen .

Discussion
The exposur e to siliceous dust in foundries, mines, a nd pottery manu facturin g has been thoroughly investigated, and the se environmenta l da ta have been reviewed by IA RC ( I) . A stud y by Froines et al (14) also evaluated the expos ure to silica in industries in th e United States (US).
Th e exposure levels measured in fou nd ries have been gene rally high . In ferro us foundries in the US 4 1 0J0 of the obser vatio ns exceeded the TL V fo r quar tz , and th e exposure inde x was 1.3 fo r the pe riod 1979-I982 (14). Another US st udy of more than IO 000 samples ana lyzed by the Na tio nal Instit ute for Occupational Safe ty an d Hea lth (N IOS H) an d collected in fou ndri es in 1972-I982 reported that 23 0J0 of th e samples exceeded twice the TLV (\) . In pottery man ufactu ring , significant expos ur e to silica was also o bserved in the lSI US, and 73 010 of the samples exceeded the TLV (14). In addition 23 (t/o of the samples analyzed in the pottery industry by NIOSH exceeded twice the TLV (1). An analysis has been made of 45 000 respirable dust samples collected by the US Mine Safety and Health Administration in 1977-1981. The geometric mean of the measurements of respirable silica dust was below 0.05 mg/m' for most of the mining industries and work operations (I).
The exposure to silica in these industries thusappears to be on the same order of magnitude as found around 1980 in the Danish stone industry. In Denmark 75 % of the measurements exceeded the TL V in the road material industry, and the corresponding value for the stonecutting industry was 45 010. The median exposure to respirable silica was 0.16 rng/rn' for the road material industry and 0.05 mg/rn' for the stonecutting industry.
Unlike the stone industry, however, the exposure to silica dust generally occurs in foundries, pottery manufacturing, and mines in association with other materials (I, [15][16][17][18][19][20][21]. Studies of cancer incidence or mortality in groups of workers with pure exposure to silica, such as stone workers, are therefore important for assessments of the possible carcinogenic properties of silica dust. Environmental data analyses incorporated into cancer mortality studies of granite industry workers are available from the US and Finland. In Vermont (US), a proportional mortality analysis among granite workers found a ratio for lung cancer of 1.2 [95 % confidence interval (95 % CI) 0.9-1.5] in a comparison with the national figures (22). Exposures were estimated from six environmental surveys covering the period 1924-1977. The estimated average exposure to respirable-free silica was 0.07 mg/rn' for cutters after 1950. Other environmental studies (1) in 54 granite industries in Vermont, carried out in 1970, found the highest exposure levels to respirable silica among cutters (0.06 mg/rn ') and sculptors and carvers (0.09 mg/rn'), and data from Vermont and Georgia from 1973-1974 showed mean exposure levels to respirable silica of 0.10 and 0.06 mg/rn", respectively, for stonecutters. By comparison the exposure to respirable silica in the present study was 0.06 mg/rn ' for the stonecutting industry as a whole and 0.04 mg/m' for cutting operations only.
A proportional mortality analysis among the members of the US Granite Cutters Union showed a slight excess of lung cancer (23) (proportional mortality ratio 1.19,95 % CI 0.97-1.46). From the 1905 death certificates analyzed, about two-thirds came from Vermont, where the exposure levels to silica were documented. For the remaining third of the subjects, the study by Froines et al (14) provides relevant exposure data from the stone industries throughout the US. The percentage of inspections with test samples over the TLV was 40 % in this study, and the median exposure index was 0.8. In the Danish stonecutting in-152 dustry, 45 (t/o of the measurements exceeded the TL V and the median exposure index was 0.6. A Finnish study (24) among granite workers showed an overall standardized mortality ratio (SMR) of 129 for lung cancer (22 observed, 17. I expected). The mortality from lung cancer was excessive for workers with at least 15 years since entry into granite work (SM R 221,95 (t/o CI 137-338). Exposure data were available from a survey made in 1970-1972. The highest concentrations of silica dust were observed for drilling, the hygienic standard (0.2 rng/rn') for quartz being exceeded tenfold on the average and the range being 0.3-4.2 rng/rn'. This value is higher than in the present study, in which the median exposure to quartz dust was 0.09 (range 0.03-1.20, mean 0.43) mg/rn' for drilling. The range of the quartz concentration was also high for block surfacing in Finland (0.2-4.9 mg/rn ' of respirable quartz) compared to the findings in Denmark (0.03-0.57 mg/rn' of respirable quartz). Finally, the range of the respirable quartz concentration for other work processes was 0.02-3.6 mg/rn' in Finland, compared to 0.03-1.7 mg/m for other granite operations in Denmark.
The measurements from the stone industry available in the archives of the Danish National Institute of Occupational Health were relatively few, but they were coherent with the findings from larger samples available from other countries and collected in similar industries. In Denmark, differences in exposure levels appeared between the branches of activity, the materials processed, and the work operations. On the average, the road material and building industry was associated with exposure to crystalline silica that was three times higher than that found in the stonecutting industry, and crushing was associated with higher exposure levels than drilling, sieving, and cutting. These environmental data provide a reference for the interpretation of epidemiologic results.