Document 377RbZbGv4zM6XmJ4vjwax0E3

Occupational Risks for Pleural Mesothelioma in Sweden, 1961-79*1 Hans S. R. Malker,2 Joseph K. McLaughlin,3 4 Blrgltta K. Malker,5 B. J. Stone,3 *Jan A. Weiner,2 Jan L. E. Erickson,5 and William J. Blot3 ABSTRACT--From national population-based registries linking cancer incidence from 1961 to 1979 with 1960 census data on industry and occupation for.all employed individuals in Sweden, a systematic assessment was made of pleural mesothelioma occur rence according to occupational and industrial classifications. There were 318 cases of pleural mesothelioma recorded during the 19-year follow-up period among males employed in 1960, with significant variation by industrial and occupational categorizations. The observed number of pleural mesotheliomas for men employed in the sugar refining, cellulose, wood and.pulp, shipbuilding, and railroad equipment manufacturing industries was more than three times the number expected. Occupations with at least twofold excess of mesotheliomas included the craftsman categories of plumbers, mechanics and repairmen, electricians, painters, tire makers, and stationary equipment operators. Our findings are consistent with available data relating mesothelioma to occupa tional asbestos exposure in other countries, although unexpected associations were found that deserve further epidemiologic study.--JNCI 1985; 74:61-66. Malignant pleural mesothelioma is a rare and highly fatal tumor arising from the mesothelial surface lining the pleural cavities. The pleura is the most common site of origin for mesotheliomas in most countries, including Sweden, where 90% of all mesotheliomas among males and 60% among females were from the pleura during the period 1961-79. Throughout the world this neoplasm has been strongly associated with occupational exposure to asbestos (1). The rarity of mesothelioma, however, has made evaluation of occupational associations difficult, except in large heavily exposed groups (2). The CER in Sweden (5, 4), which links cancer incidence data with employment data, provides a unique opportunity to assess systematically the occurrence of pleural mesotheli oma by industry and occupation for an entire country. In this paper we utilize this linked data resource to develop new etiologic hypotheses and provide further information regarding previously reported occupational findings. METHODS The CER links the cancer incidence data (1961-79) of the National Swedish Cancer Registry with occupational information obtained in the 1960 national census. The CER now contains about 605,000 cancer cases (5). Every Swedish citizen has a personal identification number, making linkage between data sources easy. A national law requires cancer cases to be reported to the Registry (6). Nonreporting for cancers of the trachea, bronchus, lung, and pleura is estimated to be less than 5% (7). For the period under study (1961-79), 99% of all pleural cancers (ICD 7th Revision, 162.2; ICD 8th Revision, 163.0) and 100% of all cancers recorded as malignant pleural mesotheliomas were microscopically verified. Only verified malignant pleural mesotheliomas are used for this analysis. An SIR was used to estimate risks of mesothelioma for various occupational and industrial categories for the 19year period. The SIR is the ratio of the observed number of cases of pleural mesothelioma in a particular industry or occupation to the expected number of cases. The expected number of cases is generated by application of the birth cohort- and sex-specific rates for 1961-79 from the genera] population of Sweden to the birth cohortand sex-specific distributions of the various occupational and industrial groupings. Five-year birth cohorts were used in these calculations. Because of variation in cancer incidence rates by region, the SIR were also adjusted for region, with the use of the 3 metropolitan areas and 24 counties in an indirect standardization (8). Statistical significance was tested under the assumption that the observed number of mesotheliomas followed a Poisson distribution (9). Industrial and occupational classifica tions from the 1960 census (Folk-och Bosiadsrakningen 1960) were used to define industrial and occupational cohorts (10). These codes are revised versions of the standards from the International Labour Office and the United Nations (11,12). SIR were calculated for all broad (1-digit code) industrial and occupational groups, for all manufacturing and craftsman-tradesman groups (2-digit code), and for selected specific (3-digit code) industries and occupations. Since at the 3-digit level there are more than 300 industry codes and 300 occupation codes, only those industries and occupations with 3 or more observed cases were included in the analysis at this level. When numbers permitted, the distribution of pleural mesotheli omas was also determined for occupations within specific industries to enable a more detailed evaluation of risk. Abbreviations used: CER = Cancer-Environment Registry; ICD=fn- ternational Classification of Diseases; ind. = industry codefs); NOC=nol otherwise classified; occ. ^occupational codefs); SIR=standardized incidence ratiofs). . 1 Received June 26, 1984; accepted August 15, 1984. * National Board of Occupational Safely and Health, Solna, S-171 84, Sweden. * Environmental Epidemiology Branch, Division of Cancer Etiology, National Cancer Institute,'National Institutes of Health, Public Health Service, U.S. Department of Health and Human Services, Belhesda, MD 20205. 1 Address reprint requests to Dr. McLaughlin, National Institutes of Health, Landow Building, Room SCI6, Belhesda, MD 20205. ' ''National Board of Health and Welfare, Stockholm, S-106 30, Sweden. 61 JNCI, VOL. 74, NO. I. JANUARV 1985 SC-ELEC-08720 Malker, McLaughlin, Malker, etaj. RESULTS There were 318 cases of malignant pleural mesotheli omas during the 19-year follow-up among males employed in 1960. The SIR for the 10 broad industries and occupations in Sweden are presented in table I. Significantly elevated SIR were found for the manu facturing and construction industries (ind. 3 and 4). For occupation, a significantly increased risk was observed for craftsmen and production-process workers (occ. 7). These broad industrial and occupational categories also had the largest numbers of mesotheliomas. Significantly low SIR were observed for farmers (ind. 0; occ. 4) and sales workers (occ. 3). The manufacturing industries (ind. 2 and 3), the construction industry (ind. 4), and the productionprocess workers (occ. 7 and 8) were examined further at the 2- and 3-digit coding levels (tables 2,3). Employment in the general industries (2-digit code level) of paper and allied products (ind. 26), rubber (ind. 30), transportation equipment (ind. 36), and other construction (ind. 41) was associated with a significantly increased SIR (table 2). Other manufacturing industries also showing elevated risks included textile (ind. 22), furniture (ind. 25), chemical (ind. 31), stone, clay, and glass (ind. 33), and machinery and electronics (ind. 35). An examination of the 3-digit level indicates that increased SIR are seen for both cellulose, wood, and pulp (ind. 261) and paper mill industries (ind. 262) within industry 26; that work in shipbuilding and repair (ind. 360) and manufacturing of railroad equipment (ind. 362) account for the high risk within the transportation equipment category (ind. 36); and that painting (ind. 410) and plumbing (ind. 413) are the specific industries accounting for the excess risk within other construction (ind. 41). Significantly elevated SIR were also observed for other specific industries, including sugar factories and refineries (ind. 206) and other manufacturing of industrial chemicals (ind. 312). Results for general and specific occupations are pre sented in table 3. Significantly elevated SIR were seen for the general category of toolmakers, machinists, plumbers, welders, and related jobs (occ. 75) and for painters and paperhangers (occ. 78), whereas nonsignificantly in creased SIR were seen for some other general occupa tional categories. Within the general occupation of toolmakers, machinists, plumbers, welders, and related jobs (occ. 75), all of the specific occupations had elevated SIR, two of them significant: mechanics and repairmen (occ. 752) and plumbers and pipe fitters (occ. 754). Of the 16 mechanics and repairmen (occ. 752) who developed pleural mesotheliomas, 4 were employed in the manufac turing of railroad equipment (ind. 362) for an occupation within industry SIR of 9.5, 2 were employed in sugar factories and refineries (ind. 206) for an SIR of 32.5, and 2 were employed in shipbuilding and repair (ind. 360) for an SIR of 11.5. Other specific occupations associated with significantly elevated risks were electricians and electrical repairmen (occ. 761), painters and paperhangers (occ. 781), cellulose and pulp workers (occ. 834), tire workers and vulcanizers (occ. 851), and operators of stationary engines and related equipment (occ. 871). Twofold and greater (nonsignificant) risks were observed for blacksmiths (occ. 735), sheetmetal workers (occ. 753), bricklayers, plasterers, and cement finishers (occ. 791), and craftsmen and production workers (NOC) (occ. 858). Two of the 3 cases in the last occupation occurred in the cement and light concrete industry (ind. 334), which gives an SIR of 3.4 for the occupation within industry. There was also a more than twofold excess risk among woodworkers (NOC) (occ. 779); 2 of these cases were within the sawmill industry (ind. 240), yielding an SIR of 2.1. A fourfold excess (based on I case) was found for the Table 1.--SIR for pleural mesothelioma by major divisions of industry and occupation among males Code 0 1 2 3 4 5 6 7 8 9 Major industry Farming, forestry, hunting, and fishing Mining and quarrying Manufacturing I Manufacturing II Construction Electric, gas, water, and sanitary services Trade, finance, insurance, and real estate Transportation and communication Services Nonclassifiable estab lishments Cases 21 0 56 103 63 2 20 20 30 3 SIR" 0.4* -- 1.2 1.5* 1.6* 0.4 0.6 0.8 0.9 3.4 Code 0 1 2 3 4 5 6 7 8 9 Major occupation Professional, technical, and related workers Administrative, executive, and managerial workers Clerical workers Sales workers Farmers, fishermen, hunters, and related workers Miners, quarrymen, and related workers Transport and communica tions workers Craftsmen, productionprocess workers, and laborers Craftsmen, productionprocess workers, and laborers (NOC) Service, sport, and recreation workers Cases 41 12 9 9 21 2 20 138 48 18 "Adjusted for age and region. *P<0.01. JNCI. VOL.. 74. NO. I. JANUARY 1985 SIR" 1.2 1.0 0.7 0.4* 0.4* 1.3 0.9 1.6* 1.0 1.2 Occupational Risks for Pleural Mesothelioma 63 Table 2.---SIR for pleural mesothelioma by general (2-digit) and specific (S-digit) manufacturing industries among males Code 20 21 22 23 24 25 26 27 28 30 31 32 33 34 35 36 General industry Food products Tobacco and beverages Textile Apparel (includes shoes and furs) Lumber and wood products Furniture and fixtures Paper and allied products Printing and publishing Leather ' Rubber Chemical ,, Coal and petroleum refining Stone, clay, and glass Fabricated metal products Machinery and electronics Transportation equipment 37 Miscellaneous manufacturing 39 . Manufacturing (NOC) 40 House building construction 41 Other construction Cases 10 0 9 3 4 4 23 3 0 6 9 0 8 11 29 40 0 0 27 36 SIR" 1.1 -- 1.6 0.5 0.5 1.4 2.9* 0.6 -- 3.0" 2.2 -- 1.3 0.8 1.4 2.3* -- -- 1.3 1.9* Code 206 220 221 Specific industry Sugar factories and refineries Wool Cotton . 250 261 262 300 312 334 345 350 352 360 362 . Furniture manufacturing Cellulose, wood, and pulp Paper mills Rubber manufacturing Other manufacturing of industrial chemicals Manufacturing of cement and light concrete Iron and steel manufacturing Machine manufacturing Other electrical industry Shipbuilding and repair Manufacturing of rail- road equipment 400 House building 410 Painting 413 Plumbing 414 Electrical installation 415 Road and power plant construction Cases 7 3 3 3 14 8 5 6 3 6 19 10 25 11 27 13 12 3 6 SIR" 6.3* 2.7 1.3 1.4 4.7* 2.4 2.8 3.7" 2.0 1.1 1.2 2.0 3.0* 7.2* 1.3 2.9* 4.2* 1.6 0.8 "Adjusted for age and region. *P<.01. "P<05. specific occupation of insulators (occ. 794), an uncom mon occupation (1,677 men) in Sweden. Specific occupations not included in table 3 but with significantly elevated SIR were chemical engineers (occ. 004; 6 cases; SIR=3.2), railroad engineers and firemen (occ. 631; 4 cases; SIR=5.0), and transportation workers (NOC) (occ. 639; 3 cases; SIR=16.4). . The small numbers of cases often precluded examina tion of risks associated with occupations within indus tries. For the sugar refining, cellulose, wood and pulp, paper, shipbuilding, and railroad equipment manufac turing industries, however, elevated SIR were found for multiple occupations, although the SIR tended to be highest for mechanics and repairmen. Although only 41 cases of malignant pleural mesothe lioma were observed among women employed in 1960, there were some notable results at the 3-digit industry and occupation levels. A significantly increased SIR was observed for the apparel industry (ind. 233; 6 cases; SIR = 3.1). Five of these cases occurred among seam stresses (occ. 716), with an SIR of 3.4 for this occupation within the apparel industry. The occupations of nurses and midwives (occ. 040 and 043; 6 cases; SIR=3.3) had significantly more cases of mesothelioma than expected. DISCUSSION This survey of malignant pleural mesothelioma inci dence in Sweden and its relationship to occupation and industry revealed findings generally consistent with those reported from other countries (13). The highest incidence ratios at the broadest industry and occupation levels were for manufacturing and construction trades, where the potential for exposure to asbestos is greatest. Within this wide category, industries involving the manufacture of ships, railroad cars, paper, pulp, and sugar and occupa tions involving plumbing, metal, machine and electrical, work, and painting showed the largest elevations in the SIR. The four industries account for 20% of the pleural mesotheliomas but only 4% of the work force. The threefold excess of mesothelioma among shipyard workers adds Sweden to the long list of countries in Europe and North America where this association has been seen (H, 15). Of all pleural mesotheliomas among JNCI, VOL 74. NO. I, JANUARY 1985 64 Malker, McLaughlin, Malker, et al. persons employed in 1960, 8% occurred in the ship-. . building industry, which accounts for about 1% of the total work force. The almost sevenfold excess among railroad car con struction workers confirms an excess recently reported in the United States and other countries (16) and seems due to asbestos exposure in the fitting and repair of car engines. Similarly, the association with plumbing and pipe fitting is consistent with previous reports (17,18) of mesothelioma following asbestos exposure in these occu pations. The findings for the paper and pulp industry are new, and the number of tumors found among the workers in this industry was surprisingly high. Mesotheliomas have not been reported among pulp workers in the United States, where the industry is a major one; nor have lung cancer risks been found to be elevated (19). The excess in Sweden seems related to differences in the mechanical and chemical processes for converting wood to pulp and the consequent heavier use of asbestos as an insulation material in the machines employed in Sweden. Although more pronounced among workers in the cellulose wood and pulp industry, the increased risk also extends to paper production, where chemical and dust exposure to wood products is considered to be lower but where asbestos has been used in processing equipment. The association with the sugar refinery workers appears most likely to be due to asbestos exposure that occurred during Table 3.--SIR for pleural mesothelioma by general (2-digit) and specific (S-digit) craftsman-tradesman occupation among males Code 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 General occupation Spinners, weavers, knitters. and related jobs Tailors, cutters, furriers, and related jobs Shoe and leather workers Metal making and treating workers Precision instrument makers Toolmakers, machinists. plumbers, welders, and related jobs Cases 2 1 0 7 1 59 Electricians and electronic workers Woodworkers Painters and paperhangers Bricklayers and construction workers (NOC) Graphical workers Potters, kilnmen, and glass workers Food industry workers Chemical and cellulose workers Tobacco workers Craftsmen and production- process workers Laborer (NOC) Stationary engine and equipment operators Packers, handlers, and long- shoremen 12 25 12 19 1 1 1 8 0 7 9 9 12 SIR" 0.7 0.4 -- 1.3 0.7 2.0* 1.8 1.4 2.0f 1.4 0.3 0.6 0.2 1.6 -- 1.8 0.9 1.4 1.0 Code Specific occupation 735 Blacksmiths 750 Toolmakers and machinists 751 Assemblers and machine erectors 752 Mechanics and repairmen 753 Sheetmetal workers 754 Plumbers and pipe fitters 755 Welders and flame cutters 756 Structural metal workers 758 Other mechanical and metal workers 761 Electricians and electrical repairmen 771 Carpenters and joiners 772 Cabinet workers 779 Woodworkers (NOC) 781 Painters and paperhangers 791 Bricklayers, plasterers, and cement finishers 793 Construction workers (NOC) 794 Insulators 834 Cellulose and pulp workers 851 Tire workers and vulcanizers 858 Craftsmen and production workers (NOC) 871 Operators of stationary engines and related equipment 882 Longshoremen and freight handlers 883 Stockroom workers Cases 5 11 3 16 4 13 4 4 4 10 12 8 3 12 6 10 1 5 4 3 7 3 9 SIR" 3.2 1.4 1.5 2.4* 2.0 4.8* 1.6 1.8 1.4 2.3c 1.4 1.4 2.5 2.3f 2.1 1.1 4.3 4.0f 4.4C 2.5 3.9* 1.3 1.0 "Adjusted for age and region. *P<01. fP<05. JNCI, VOL. 74, NO. I. JANUARY 1985 Occupational Risks for Pleural Mesothelioma 65 the annual dismantling of the machinery used in sugar the registration of cancer is thought to be essentially production (20, 21). ' . complete and since microscopic verification was obtained We found a small elevation in risk for males in the for all cancers in the study. It cannot be denied, however, textile industry, consistent with reports of mesothelioma that diagnostic criteria may vary among Swedish pathol among asbestos textile makers (22), but the excess was ogists. In addition, some diagnoses may have been limited to several high-risk occupations such as mechan influenced by a knowledge of exposure to asbestos, a ics and repairmen and operators of stationary equipment, recently observed phenomenon that could result in and it was not found among the occupation "textile differential reporting (28). workers." Finally, the statistical techniques used may have Other findings that can be considered as new leads for influenced the differential risks by industry and occu occupationally induced pleural mesothelioma include pation. Because of lack of information on loss to follow elevated SIR among rubber workers, machinists, metal up, either from departure from Sweden or from death, the workers, electricians, and painters, although elevated expected values for the SIR calculations were not based relative risks for mesotheliomas have been reported for on person-years, but rather they were based on the the rubber products industry, electricians, and painters in assumption that the proportions of individuals within a recent case-control study in which data from the each sex-region-birth cohort subgroup for each industry Connecticut Tumor Registry were used (17). In the and occupation remained constant over the 19-year rubber industry in Sweden large amounts of talc, which follow-up. Thus, if total mortality was elevated (or may have contained asbestos (23, 24), have been used. depressed) for a particular industry or occupation, the The results of this analysis linking census and cancer expected number of mesotheliomas, which does not take incidence data have served to quantify risks of mesothe into account loss due to mortality, would be too high (or lioma and identify specific industries and occupations in low) and the corresponding SIR would be too low (or Sweden where further study may be warranted. The high). Such variations, however, will in most instances identification of specific industries and occupations for affect the SIR by less than a small percent and thus will further study was the main rationale for the study, since not alter the interpretation of the data presented (29). the CER seems most useful as a tool for generating hypotheses about the occupational causes of cancer. Certain limitations of the CER, however, generally REFERENCES preclude making definitive inferences about the role of occupation in cancer etiology. Most notable are the lack of information on risk factors such as smoking, diet, and other environmental exposures that may affect cancer incidence (although for mesothelioma there are no (/) Fraumeni JF Jr, BijOT WJ. Lung and pleura. In: Schouenfeld D, Fraumeni JF Jr, eds. Cancer epidemiology and prevemion. Philadelphia: Saunders, 1982:564-582. (2) McDonald JC, McDonald AD. Mesothelioma as an index ot asbestos impact. Banhury Rep 1981; 9:78-82. established risk factors other than asbestos exposure) and the lack of an occupational history. Since occupational change is not very common in Sweden, the occupational categorization reported in the 1960 census is likely to (3) WiKi.iiND K, Einhorn J, Wknnstrom G, Rapaport E. A Swedish cancer environment registry available for research. Scand J Work Environ Health 1981; 7:64-67. (4) Ahlmark A, Malkkr H. Potentials For epidemiological studies in occupational medicine. Ann Occup Hyg 1981; 24:159-162. represent an individual's usual adult job classification. No linked registry is better than the quality of its underlying registries. In the linkage, only 1% of the cancer cases could not be traced in the census (7). False matching of cancer cases to census information has been (5) Cancer-Miljoregistret-Namnden. Cancer-miljoregisiret uppdaterat. Omfattar nu over 600,000 cancerfall. Lakarlidningen 1985; 80:123-124. (6) National Board of Health and Welfare. Cancer incidence in Sweden, 1980. Stockholm: Liber Forlog, 1983. (7) Mattsson B, Willgren J. Completeness of registration in the estimated to be less than 0.5% (25). There may, of course, be errors in the census classification, but an independent validation of industrial and occupational data by a re interview of a random sample of the 1960 census revealed Swedish Cancer Registry. Stockholm: National Central Bureau of Statistics, 1977 (statistical reports, HS 1977; 15). (8) Socialstyrelsen, Cancer-Miljoregislret: Cancemsk och miljofaktorer. En undersokning baserad pa del nyinrattade CancerMi ljoregistret. Rapport till Arbetarskyddsfonden. Stockholm: close agreement (`=*95%) at the broad 1-digit level of Arbetarskyddsfonden, 1980. coding, with somewhat less agreement for more specific (3-digit) codes (26). The problem of multiple compar isons must be recognized in a review of such a large number of occupational categories. The findings in this (9) Bailar JC III, Ederf.k F. Significance factors for the ratio of a Poisson variable to its expectation. Biometrics 1964; 20:639-643. (10) Statistiska Cenlralbyron och Bosiandsstyrelsen. Folk och Bosladsrakningen. Den I November I960. Redogorelse for Folk-och Bosladsrakningens upplaggning och genomforande. Stockholm: paper, however, are remarkably consistent with previous studies, suggesting underlying validity to our approach. The awareness of employees and their physicians regarding asbestos hazards may differ and, consequently, National Central Bureau of Statistics, 1965. (//) International Labour Office. International standard classification of occupations. Geneva: International Labour Office, 1958. (12) United Nations Statistical Office. International standard industrial classification of all economic activities. New Vork: United diagnostic procedures may differ by occupation. The diagnosis of pleural mesothelioma is often difficult, and both overreporting and underreporting have been sus pected (27). Such a problem seems less likely to have occurred for the occupational cohorts in Sweden, since Nations Statistical Office, 1958 (statistical papers, series M, No 4, Rev 1). (13) McDonald JC, McDonald AD. Epidemiology of mesothelioma from estimated incidence. Prev Med 1977; 6:426-446. (14) Blot WJ, Fraumeni JF Jr. Cancer among shipyard workers. Banbury Rep 1981;9:37-50. JNCI, VOL. 74, NO. I, JANUARY 1985 66 Malker, McLaughlin, Malker, et al. (13) Gardner MJ, Acheson ED, Winter PD. Mortality from meso thelioma of the pleura during 1968-78 in England and Wales. Br J Cancer 198S: 46:81-88. (16) Manguso TF. Mesothelioma among machinists in railroad and other industries. Am J Ind Med 1983; 4:501-513. (17) Tkta JJ, Lkwinsohn HC, Meigs JW, Vioonk RA, Moward LZ, Fi.annery JT. Mesothelioma in Connecticut, 1955-1977. J Occup Med 1983;25:749-756. (18) McDonald AD, McDonald JC. Malignant mesothelioma in North America. Cancer 1980; 46:1650-1656. (19) International Agency for Research on Cancer. Wood, leather and some associated industries. IARC Monogr Eval Carcinog Risk Chem Man 1981; 25:157-197. (20) Malker HR, Malker BK, Bi ot WJ. Mesothelioma among sugar refinery workers. Lancet 1983; 2:858. (21) Steinkck G, CARSTKfj.SKN J, Wiklund K, Eki.und G. Mesothe lioma among sugar refinery workers. Lancet 1983; 2:1503. (22) Newhoi/se ML, Berry C. Patterns of mortality in asbestos factory workers in London..Ann NY Acad Sci 1979; 330:53-60. (23) Fristedt B, Maitsson SB, Schutz A. Talcosis by exposure to. granular talc in rubber industry. Nord HygTidskr 1968;99:6-7. (24) IntemationalAgency for Research on Cancer. The rubber industry. IARC Monogr Eval Carcinog Risk Chem Man 1982; 28:121-147. (23) Eki.und G, Wiklund K. Kvaliteiskoniroll av personideniiteien i cancermiljoregislret. Slatistisk Tidskrift 1979; 5:373-376. (26) Brivkai.ne M. The control study made in connection with the I960 census of population. Statistical reports B. Stockholm: National Central Bureau of Statistics, 1964. (27) McDonald AD, Mesothelioma registries in identifying asbestos hazards. Ann NY Acad Sci 1979; 330:441-454. (28) Wright WE, Shkrwin RP, Dickson EA, Bernstein L, Fromm JB, Henderson BE. Malignant mesothelioma: Incidence, asbes tos exposure, and reclassification of histopathology. Br J Ind Med 1984; 41:39-45. (29) Malker H, Weiner J. The Cancer-Environment Registry 1961 -- 1973. Examples of the use of registry epidemiology in studies of the work environment. Arbete och Halsa 1984; 9:1-107 (English summary only). JNCI, VOl.. 74, NO. I, JANUARY 1985