Document nkL4VKY8dNbo2xXrRbN53KK52

American Journal of Industrial Medicine 23:629-639 (1993) A Case-Control Study of Multiple Myeloma and Occupation Paul A. Demers, PhD, Thomas L. Vaughan, MD, MPH, Thomas D. Koepsell, MD, MPH, Joseph L. Lyon, MD, MPH, G. Marie Swanson, PhD, MPH, Raymond S. Greenberg, MD, PhD, and Noel S. WeiSS, MD, DrPH Lifetime job histories from a population-based, case-control study were analyzed to investigate the relationship between multiple myeloma and employment in various occupations and industries. Interviews were obtained from 89% (692) of eligible incident cases and 83% (1683)of eligible controls. An elevated risk was observed among persons ever employed as painters [odds ratio (OR)= 2.1, 95% confidence interval (CI)= 1.23.61, particularly for those employed for IO or more years (OR=4.1, 95% CI= 1.810.4). A small excess risk was observed among agricultural workers employed for 10or more years (OR = I .3, 95% CI = I .O-2.2). with a higher relative risk observed among farm laborers (OR = I .8,95% CI = 1.O-4.0). Among agricultural workers who reported having been highly exposed to pesticides, the OR was 5.2 (95% CI= 1.6-21.1). Some evidence, based on smaller numbers, was also found to support an association with firefighting and employment in the petroleum- and coal-products manufacturing industries. Little evidence was found to support the previously noted association with wood exposure, and no evidence for an association with employment in the rubber or petroleum refining industries was found. This study lends further support to previously reported associations between multiple myeloma and employment among painters and agricultural workers. 0 1993 Wiley-Liss. Inc. Key words: occupational diseases, painters, agricultural workers, multiple myeloma INTRODUCTION A number of epidemiologic studies have examined the association between multiple myeloma and employment in particular occupations or industries. The strongest evidence has linked employment in agriculture with an increased risk of multiple myeloma [Boffetta et al., 1989; Burmeister et al., 1983; Cantor and Blair, 1984; Cuzick and De Stavola, 1988; Flodin et al., 1987;Gallagher et al., 1983; La Vecchia Department of Epidemiology. University of Washington, Seattle, WA (P.A.D.. T.L.V., T.D.K.,N.S.W.). Program in Epidemiology, Fred Hutchinson Cancer Research Center, Seattle, WA (T.L.V.; T.D.K. N.S.W.). Departments of Family Medicine and Community Medicine, University of Utah Medical Center, Salt Lake City, UT (J.L.L.). Cancer Center, Michigan State University, East Lansing, MI (G.M.S.). School of Public Health. Emory University, Atlanta, GA (R.S.G.). Address reprint requests to Thomas L. Vaughan. Program in Epidemiology (MP-474). Fred Hutchinson Cancer Research Center, I124 Columbia Street, Seattle, WA 98104. Accepted for publication July 15. 1992. 0 1993 Wiley-Liss, Inc. 630 Demers et al. 1 et al., 1989; Milham, 1971; Pearce et al., 1986; Steineck and Wiklund, 19861. Elevated relative risks also have been observed for employment in occupations involving working with wood [Flodin et al., 1987; Milham and Demers, 1984; Milham, I 1976; Nandakumar et al., 1986; Tollerud et al., 19851, painting or paint manufacturing [Adelstein, 1972; Berthwaite et al., 1990; Cuzick and De Stavola, 1988; Friedman, 1986; Lundberg, 19861, and exposure to radiation [Cuzick, 1981; Gilbert et al., 19891. Associations between multiple myeloma and employment in the oil and petrochemical industry [Wong and Raabe, 19891, in rubber manufacturing [Riedel et al., 19911, and as a firefighter [Howe and Burch, 19901 also have been reported. However, results have been inconsistent across studies, and considerable uncertainty remains concerning the role of occupation in the etiology of multiple myeloma [Riedel et al., 19911. In this paper, analyses of lifetime work histories from participants in a large, population-based, case-control study are presented. In an earlier analysis from this study, the relationship between multiple myeloma and self-reported occupational and nonoccupational exposure to a broad range of chemicals and potentially toxic substances was examined [Moms et al., 19861. The goal of the current study was to examine systematically the association between multiple myeloma and employment in various occupations or industries. MATERIALS AND METHODS Data on cases and controls were collected as part of a collaborative study of risk factors for multiple myeloma and chronic lymphocytic leukemia [Koepsell et al., 19871. Cases were identified through tumor registries in four geographic areas: King and Pierce counties in Washington state; Davis, Salt Lake, Utah, and Weber counties in Utah; five counties of metropolitan Atlanta; and the three metropolitan Detroit counties. The four tumor registries are participants in the Surveillance, Epidemiology, and End Results program of the National Cancer Institute. All residents under the age of 80 years who were newly diagnosed with multiple myeloma between July 1, 1977, and June 30, 1981, were approached through their physicians for participation in the study. Controls were selected to be similar in age and sex to the cases but otherwise were representative of the population in the regions. In Washington state, the twocounty region was divided into 175 areas of approximately equal population, within which two sampling units of approximately four households each were selected. All residents between the age of 65 and 79 years, and one out of nine residents between the age of 40 and 64 years were approached for participation. In the three other geographic study areas, random digit dialing methods were used. Control selection methods are more fully described elsewhere [Koepsell et al.. 19871. Cases, or their survivors, and controls were queried by trained interviewers on a broad range of risk factors. Interviews were obtained for 692 (89%) of the 773 eligible cases and 1,683 (83%) of the 2,024 eligible controls. Four hundred seventytwo (68%) cases were interviewed directly. while the remainder of interviews were conducted with a proxy respondent. Lifetime work histories were collected, and occupations and industries were coded according to the 1970 U.S. Census codes. Stratified analysis of occupational histories was performed using a computer program developed by one of the authors [Vaughan. 19891. Mantel-Haenszel odds ratios were cal industries after (white, black, a potential sou jects only. Cor when there WI methods were RESULTS The dem elsewhere [KO( employed in sy all respondent. (OR)= 2.1, 95 (OR=2.5, 95' (OR= 1.7, 95' occupations inc occupations, fi were quite bro veyors (OR= (OR=0.4, 95' CI = o . 1-1.1). economic statu 60 OR for all rt these were bas Controls occupations wt from surrogate duration of 1 I tively less deta cases (Table 1. basis of all res Table I1 1 study participa aircraft and air (OR=2.l, 95' When the anal employed in a; CI = 1.1-4.0) Table 111 years) for occ excluded one ii for the more sp informative. TI manufacturins painters in m a cases and 13 cc firefighters. fa1 Wiklund, 19861. n occupations in*s. 1984; Milham, )r paint manuface Stavola, 1988; ck, 198I ; Gilbert lent in the oil and cturing [Riedel et 'e been reported. :rable uncertainty tultiple mye'oma ipants in a large, nalysis from this occupational and ntially toxic subent study was to and employment itive study of risk [Koepsell et al.. iphic areas: King 1Weber counties ropolitan Detroit ince, Epidemiol1 residents under ma between July :ians for partici- es but otherwise n state, the twoipulation, within ere selected. A11 :sidents between the three other Iontrol selection interviewers on 9%) of the 773 undred seventyinterviews were collected, and Zensus codes. ing a computer 1-Haenszel odds Multiple Myeloma and Occupation 631 ratios were calculated for ever having been employed in particular occupations or industries after adjustment for sex, age (<50, 50-59. 60-69, 70-79 years), race (white, black, other), and study area. To evaluate the use of surrogate respondents as a potential source of bias, the analyses were repeated with the self-respondent subjects only. Cornfield 95% confidence intervals (CI) were calculated except in cases when there were fewer than five exposed cases or controls, in which case exact methods were used [Breslow and Day, 19801. RESULTS The demographic characteristics of the study population have been presented elsewhere [Koepsell et al., 19871. Odds ratios and 95% CI for ever having been employed in specific occupational groups are presented in Table I. In analyses using all respondents, elevated odds ratios were observed among painters [odds ratio (OR)= 2. I , 95% CI = 1.2-3.51, those employed in forestry and logging occupations (OR = 2.5, 95% CI = 1.1-12), and metal and plastic working machine operators (OR = 1.7. 95% CI = 1.1-2.5). Elevated risks also were observed for many other occupations including mathematical and computer scientist, "other"health treatment occupations, firefighters, and insulation workers, although the confidence intervals were quite broad. Decreased risks were observed for engineers, architects, and surveyors (OR =0.3, 95% CI =0.2-0.9), engineering and science technicians (OR =0.4, 95% CI =0.2-1. I ) , and rail and water transportation workers (0.3, 95% CI =0.1-1.1). Additional adjustment for education (as a surrogate measure of socioeconomic status) did not substantially alter the results. For example, only four of the 60 OR for all respondents in column 1 of Table I were altered by greater than 0.2, and these were based on small numbers. Controls and self-respondent cases both reported an average of 3.5 previous occupations whose mean durations were 9 years each, while case interviews obtained from surrogate respondents had an average of 2.7 previous occupations, with a mean duration of 11 years. This may suggest that surrogate respondents provided a relatively less detailed work history. However, restricting the analysis to self-respondent cases (Table I. column 2) did not substantially alter the conclusions drawn on the basis of all respondents. Table I1 presents the results by industry group. When the analysis included all study participants, elevated risks were observed among persons ever employed in aircraft and aircraft parts manufacturing (OR = 1.6, 95% CI = 1.1-2.9) and forestry (OR = 2.1, 95% CI =0.3-13.0), although the latter was based on small numbers. When the analysis was restricted to self-respondent cases, the risk among workers employed in agricultural (OR = 1.6, 95% CI = 1.2-2.6) and mining (OR = 1.8, 95% CI = 1.1-4.0) increased. Table I11 presents analyses by duration of employment (< 10 years or 2 10 years) for occupational groups of a priori interest or whose confidence intervals excluded one in either the full or self-respondent analyses. Results are also presented for the more specific occupations within the grouped categories when the results were informative. The ORs for multiple myeloma among painters in both construction and manufacturing industries increased with duration of employment. The majority of the painters in manufacturing had been employed in the motor vehicle industry (nine cases and 13 controls). Evidence for a duration-response effect was also found among firefighters, farm laborers, farmers, miners, and metal- and plastic-working machine 632 Demers et al. TABLE 1. Relative Risk of Multiple Myeloma by Occupational G r o u p All respondents Self-responding cases only Occupational group Cases Controls OR (95% CI)" Cases OR (95% CI)" Admin. and managenal occupations Engineers. architects, surveyors Mathematical and computer scientists Natural scientists, except chemists Chemists and chemical technicians Health diagnosing occupations Nurses (RNs and LPNs) Other health rreatment occupations Educators. librarians. educ. counselors Social, legal, recreational. religious Writers, entemimrs, athletes photographers, painters. artists Health technicians Engineering, science technicians Other technicians Sales occupations Administrative support occupations Private household service occupations Fire fighting and prevention occupations Law enforcement and guards Food service occupations Health service occupations Cleaning service, except household Personal service occupations Fanners Farm workers and gardeners Forestry and logging occupations Vehicle mechanics Industrial mechanics and maintenance EIectricaUelecmnic equip. repairers Miscellaneous mechanics and repairers Carpenters Electricians Painters Plumbers. pipefitters, and steamfitters Roofers and pavers Insulation workers Other construction occupations Mining occupations Supervisors. production occupations Recision metal workers Precision wood workers Precision textile workers Recision workers. assorted materials Recision food production occupations Inspectors, testers, and calibrators Plant and system operators Metal- and plastic-working machine operators Metal- and plastic-processing machine operators Woodworking machine operator Rinting machine operators Textile machine operators Other machine operators Welders and cutters Other hand working occupations Motor vehicle operators Rail and water transportation workers Material moving equipment operators Handlers, cleaners. and laborers Garage and service station occuoations 109 8 4 2 3 2 IO 5 29 16 14 3 4 6 6 I23 I75 56 5 20 79 25 68 27 26 57 IO 24 6 5 22 17 6 31 6 1 7 IO 18 18 27 2 12 3 15 25 I 42 6 0 4 33 92 -7-7 45 76 4 19 113 I3 309 47 4 4 6 8 35 4 106 49 39 7 IO 47 25 302 449 I22 5 54 200 57 I54 70 55 138 7 72 8 13 68 36 26 40 13 6 0 29 36 44 83 2 27 4 32 94 3 84 13 3 17 89 296 53 I40 I86 23 37 328 4s 0.8 (0.6-1 1 ) 0.3 (0.2-0 9) 2.2 (0.4-13 2 ) 0.9 (0.1-8.1) 1.7 (0.3-8.3) 0.6 (0.1-3.1) 0.6 (0.3-1.4) 2.8 (0.6-15.4) 0.6 (0.4-1. O ) 0.8 (0.5-1.8) 0.8 (0.4-1.7) 1.O (0.2-4.6) 1.1 (0.2-4.2) 0.4 (0.2-1. I 0.5 10.2-1.4) 1.o (0.8-1.4) I .O (0.9-1.4) 1.0 (0.7-1.7) 1.9 (0.5-9.4) 1.1 (0.6-2.1) 0.9 (0.7-1.3) 1.1 (0.7-2.0) 1.0 (0.8-1.6) 0.8 (0.6-1.6) 1.2 (0.8-2.5) 1 .O (0.8-1.6) 2.5 (1.1-12) 0.8 (0.5-1.4) 1.8 (0.5-6.7) 0.7 (0.2-2.4) 0.8 (0.6-1.7) 1.0 (0.5-2.0) 0.7 (0.2-1.8) 2.1 ( I .2-3.6) I .O (0.3-3. I ) 0.6 (0.0-5.0) Inf. (0.I-Inf) 1.0 (0.5-2.3) 1.5 (0.9-3.3) I . I (0.7-2.4) 0.8 (0.5-1.3) 1.9 (0.1-28.8) 1.0 (0.5-2.2) 1.8 (0.2-12.3) 1.3 (0.6-2.6) 0.7 (0.5-1.2) 0.5 (0.0-6.8) 1.6 (1.1-2.5) 1.2 (0.4-3.6) 0.0 (0.0-2.6) 0.6 (0.1-2.0) 0.8 (0.5-1.3) 0.8 10.6-1.1) 1.2 10.7-2.0) 0.9 (0.6-1.3) 1.2 (0.8-1.6) 0.3 (0.1-1.1) I .5 10.8-2 9) 1.0 (0.9- 1.5) 0 8 10.4-1.5) 77 0.9 (0.7-1.3) 3 0.2 (0.0-0.8) 3 2.2 (0.3-14.9) 7 I .5 (0.1-13.0) I 0.8 (0.0-7.3) 2 0.8 (0.1-4.4) 6 0.5 (0.2-1.6) 4 3.5 (0.5-22.9) 23 0.8 (0.5-1.4) IO 0.8 (0.4-1.91 8 0.8 (0.3-1.9) 3 I .5 (0.2-7.1) 4 1.8 (0.3-7.3) 4 0.4 (0.1-1.2) 4 0.5 (0.1-1.7) 83 1 . 1 (0.9-1.6) I14 I . 1 (0.9-1.6) 34 1.0 (0.6-2.0) 4 2.8 (0.5-14.5) 1 1 I .O (0.4-2.0) 57 1 . 1 (0.9-1.8) 20 1.5 (0.9-3.1) 43 1.1 (0.8-1.9) 19 0.9 (0.6-2.0) 19 1.4 (0.9-3.3) 42 I .3 (0.9-2.1) 6 2.3 (0.9-15.4) 16 0.9 (0.5-1.6) 5 2.2 (0.6-8.5) 2 0.4 (0.0-1.9) 13 0.8 (0.5-2.0) 13 1.3 (0.6-2.8) 4 0.7 (0.2-2.3) 22 2.5 (1.3-4.7) 5 1.3 (0.4-4.5) I 1 .O (0.0-8.9) -I Inf. (0.I-Inf) 7 1.2 (0.5-3.1) 14 1.8 (1.1-4.9) I2 1.2 (0.6-2.9) 20 0.9 (0.5-1.7) I 2. I (0.0-46.I ) 7 0.9 (0.3-2.3) 3 2.9 (0.4-20.5) I2 1.8 (0.8-4. I ) 19 0.9 (0.6-1.7) I 0.8 (O.O-I2.0) 27 1.9 (1.1-3.1) 3 I . I (0.2-4.6) 0 0.0 (0.0-3.6) 7 0.5 (0.1-2.4) 23 0.9 (0.5-1.5) 63 1.0 (0.7-1.4) I4 I .3 (0.6-2.6) 26 0.9 (06-1.51 56 I .5 ( I .O-Z.2) 7 0.2 (0.0-1.I ) 12 1.6 (0.R-3.4) 94 1 . 1 (09-1.7) 9 0.8 (0.3-1.7) 'Odds ratio (OR) adjusted for sex, race. age. and study area and 95% confidence intervals (CI). T A B L E 11. Relati! Industry group Agnculture Forestry Mining and extracrinf Construction Food product manufsi Tetile manufacturing Paper product manufa Printing and publishin Chemical manufacturi Petroleum and coal rc Rubber and plastic pr' Leather product manu Lumber and wood pr" Furniture and fixture\ Stone. glass. concrete Priman metal i n d u ~ Metal product manufd Machine? manufactui Electrical product msi Motor vehicle manula Aircraft and aircraft p Ship and boat manufa Miscellaneous manufa Transponation Communications and Wholesale and retail I Financial. insurance. Business services Repair services Personal services Entenainment and rec Health care serviceb Professional and relaic Public administration "Odds ratio (OR)a operators. Whe the overall exct no means ail. c present in those one of these oc The OR : dence limits e x (Table IV). As Patterns of incr agricultural prc nonmetal minir numbers. Whet wood. petroleu or group of occ Self-responding cases onlv ihes OR (95% CI)" 0.9 (0.7-1 3) 0.2 (0.0-0.81 2.2 (0.3-14.9) 1.5 (0.1-13.0) 0.8 10.0-7.3) 0 8 10.1-4.4) 0.5 (0.2-1.6) 3.5 (0.5-22.9) 0.8 (0.5-1 1) 0.8 (0.4-1.9) 0.8 (0.3-1.9) 1.5 (0.2-7. I ) 1.8 (0.3-7.3) 0.4 (0.1-1.2) 0.5 to.1-1.7) I . 1 10.9-1.6) I . I (0.9-1.61 I .O 10.6-2.0) 2.8 (05-14.5) I .o (0.1-2.01 I . I 10.9-1.8) I .S (0.9-3. I ) 1 . 1 (0.8-1.9) 0.9 (0.6-2.0) 1.4 (0.9-3.3) I .3 (0.9-2. I 1 2.3 (0.9-15.4) 0.9 (0.5-1.6) 2.2 (0.6-8.5) 0.4 (0.0-1.9) 0.8 (0.5-2.0) 1.3 (0.6-2.8) 0.7 (0.2-2.3) 2.5 ( I .3-4.7) I 3 (0.1-4.5) 1.0 (0.0-8.9) Inf. (0.I-lnf) I . 2 (0.5-3. I ) I .8 (1.1-3.91 I . 2 (0.6-2.9) 0.9 (0.5-1.7) 2.I (0.0-.6 I ) 0.9 (0.3-2.3) 2.9 (0.1-20.5) 1.8 (0.8-4.1) 0.9 (0.6-1.7) 0.8 (O.O-I'.O) 1.9 (1.1-3.1) I . I (0.2-1.61 0.0 (0.0-3.6) 0.5 (0.1-2.4) 0.9 (0.5-1.5) 1.0 (0.7-1.4) 1.3 (0.6-2.6) 0.9 (0.6-1.5) I .5 ( I .0-1.2) 0.2 (0 0-1. I ) 1.6 (08-3.J) 1.1 (09-1 7) 0 8 (0 3-1 7 ) crvals (CI) Multiple Myeloma and Occupation 633 TABLE I t . Relative Risk of ,Multiple Myeloma by industry Group Industry group Self-responding All respondents cases only - Cases Controls OR (95% C1)" Cases OR (95% CI)" Agriculture Forestrv Mining and extracting Construction Food product manufactunng Textile manufacturing Paper product manufacturing Pnnting and publishing Chemical manufactunng Petroleum and coal refining and manufacturing Rubber and plastic product rnanufactunng Leather product manufactunng Lumber and wood product manufactunng Furniture and fixtures manufacture Stone glass. concrete product manufactunng Primary metal industnes Metal product manufactunng Machinery manufacturing except electnc Electncal product manufactunng Motor \chicle rnanufactunng Aircraft and Aircraft pans manufactunng Ship and boat manutacturing and repair Miscellaneous manufactunng Transportation Communications and utilities Wholesale and retail trade Financial. insurance and real estate Business m-vices Repair services Personal services Entertainment and recreational Health care services Professional and related senices Public administration 84 4 16 17 50 26 6 IO 19 8 9 I 25 9 9 28 39 30 15 I40 50 15 12 73 36 239 65 35 27 I I9 22 58 85 181 181 1.2 (1.0-1 9) 62 1.6 (1.2-2.6) 4 2. I (0.3-13 0 ) 7 1.2 (0.0-14.3) 56 1.5 (0.9-2.81 20 I .8 ( I . 1-4.0) 171 I . I (0.9-1.7) 55 1.4 (1.0-2.1) 132 1.0 (0.7-1.4) 36 1.2 (0.8-1.9) 60 1.0 (0.6-1.71 19 1.2 (0.7-2.3) 23 0.4 (0.2-1 31 3 0.3 (0.1-1.2) 49 0.5 (02-1.2) 7 0.6 (0.2-1.4)- 65 0.7 (0.4-1.2) 15 0.9 (0.5-1.7) 18 1.2 (0.4-3.1) 2 0.4 (0.1-1.9) 30 0.9 (0.4-2.0) 5 0.9 (0.3-2.5) 11 0.2 (0.0-1.61 1 0.4 (0.0-2.7) 43 1.0 (0.6-2.0) 20 1.3 (0.8-2.9) 19 0.8 (0.3-1.9) 5 0.7 (0.2-2.3) 38 0.6 (0.3-1.3) 9 1.0 (0.4-2.3) 96 0.9 (0.6-1.6) 17 I .O (0.6-1.8) 105 1.1 (0.8-1.8) 24 1.2 (0.8-2.1) 93 1 . 1 (0.7-1.8) 19 1.3 (0.7-2.3) 52 0.8 (0.1-1.5) IO 0.8 (0.4-1.8) 467 0.9 (0.7-1.2) 79 I .O (0.7-1.4) 54 I .6 ( I . 1-2.9) 33 I .6 ( I .O-2.9) 19 0.6 (0.4-2. I ) 7 0.4 (0.2-1.5) 21 1.3 (0.6-3.0) IO 1.8 (0.7-4.7) 191 0.8 (0.6-1.2) 51 0.9 (0.6-1.4) 97 0.8 (0.6-1.4) 24 0.8 (0.5-1.5) 644 0.9 (0.8-1.2) 158 0.9 (0.8-1.4) 135 I . I (0.9-1.8) 48 1.3 (1.0-2.2) 80 1 . 1 (0.7-1.8) 19 0.9 (0.5-1.7) 60 1 .O (0.7-1.9) 20 1.2 (0.7-2.5) 288 0.8 (0.7-1.3) 77 0.9 (0.8-1.5) 50 0.9 (0.5-1.7) 12 0.8 (0.4-1.5) 162 0.8 (0.6-1.3) 42 I .O (0.8-1.7) 250 0.8 (0.7-1.2) 53 0.8 (0.6-1.3) 528 0.8 (0.7-12 ) 117 0.8 (0.7-1.3) "Odds ratio (OR) adjusted for sex, race. age. and study area and 95% confidence intervals (CI). * operators. When the individual occupations within the latter group were examined, the overall excess risk seen with increasing duration was present for several, but by no means all, of them. The increased risk among forestry and logging workers was present in those employed for < 10 years (no cases and only one control had been in one of thcse occupations for > 10 years). The OR associated with industries of a priori interest and those whose confidence limits excluded one in Table I1 were also analyzed by duration of employment (Table IV). As in Table 111, more specific categories were included when informative. Patterns of increasing risk with increasing duration were observed for employment in agricultural production, coal mining, petroleum- and coal-product manufacturing, nonmetal mining, and horticultural services, although the latter were based on small numbers. When the occupations of persons who had been employed in the lumber, wood, petroleum, coal, aircraft, and mining industries were examined, no occupation or group of occupations with similar exposures accounted for the majority of cases, 634 Demers et ai. TABLE 111. Relative Risk of Multiple Myeloma by Occupational Group by Duration of Employment Employed < 10 years Employed I O + years Occupational group Cases Controls OR (95% CI)" Cases Controls OR (95% CI)" Farmers Farm workers and gardeners Farm laborers Unpaid farm family members Gardeners Forestry and logging occupations Carpenters Precision wood workers Painters Consuuction and maintenance Manufactured anicles Fire fighting and prevention occupations Mining occupations Metal- and plastic-working machine operators Metal job and die seners Filers, polishers. sanders. and buffers Drill press operatives Grinding machine operatives Punch and stamping press operatives Riveters and fasteners 7 28 21 4 4 IO 7 I 15 II 5 I I2 24 7 I 6 7 3 5 17 1 I (0 5 - 3 . 9 ) 19 74 0.9 10.6-1.6) 29 46 1.1 10.7-2.5) 18 21 0 6 (0.2-1.8) I I IO 0.7 (0.1-2.5) I 6 3.4 (1.3-17.0) 0 18 0.8 (0.3-2.41 IO 2 1.2 (0.0-25.0) I 29 1.4 10.6-2.8) 16 16 1.9 (0.8-4.5) I I 14 0.8 (0.2-2.8) 5 2 0.9 (0.0-22.3) 4 25 1.4(0.7-3.4) 6 52 I .410.8-2.4) I8 4 1.9tO.2-11.9) 4 4 0.4 (0.0-5.1) 7 2.9 (0.8-9.5) 14 1.7 10.6-4.7) 8 1 . 1 10.2-4.6) I 1 6 3 9 1.3 (0.3-4.5) I 38 1.3 (0.8-2.81 64 1 . 1 10.7-2.0) 25 1.8 (1.0-4.0) 30 1.0 (0.5-2.2) 8 0.2 (0.0-1.6) I 0.0 (0.0-29.3) 18 1 . 1 (0.5-3.5) 0 Inf (0.0-In0 I 1 4.1 (1.8-10.4) 8 3.7 (1.3-11.0) 3 4.8 (1.1-28.8) 3 2.9 (0.4-21.6) I I 1.7(0.7-7.1) 32 2.0 ( I . 1-2.9) 4 3.5 (0 7-17.1) I 7.5 10.1-121) 5 0.7 (0.0-7.0) IS 1.4 (0.5-4.0) 2 2.2 (0.3-17.0) 0 Inf (0.l-Inf) a d s ratio (OR)adjusted for sex. race, age. and study area and 95% confidence intervals (CI). and those at excess risk were often employed in administrative as well as blue collar jobs. In an earlier analysis of data from this study, Moms et al. [ 19861 found increased risks associated with self-reported "high" exposure (either at home or at work) to paints and/or solvents (OR = 1.6, 95% CI = 1.1-2.4) and pesticides (OR=2.6, 95% CI= 1.5-4.6). To determine if the excess risks observed among painters and persons employed in agricultural production were due to these exposures, data from the work history and self-reported exposure portions of the interview were combined. The risk observed among painters who had reported "high" exposure to paints and/or solvents (OR=2.3) differed little from painters who did not (OR = 2.0) when all respondents were considered (Table V). In contrast, much of the excess risk of myeloma among persons who had worked in the agricultural industry was concentrated in those who had claimed "high" exposure to pesticides (Table VI). The risk associated with employment in jobs with exposure to radiation could not be assessed because too few workers could be identified on the basis of job title or industry alone. A history of employment as an X-ray technician was reported by one case and one control, but no study participants reported having been employed as radiologists and none who had been utility workers reported having worked in a nuclear power plant. DISCUSSION There are limitations that should be kept in mind when reviewing the results of this study. Occupation is likely a poor surrogate for specific exposures that may influence the occurrence of multiple myeloma. To the extent that occupational and TABLE IV. RI Industry group Agriculture 4gricultural pr Agriculiural \el Honicultural \C - Forestn Lumber and uiw( Mining and ertrdi Metal mining Coal mining Oil and natural Nonmetal min Paper product md Pulp. paper. dn Paperboard con Mi\cellaneous I Petroleum and co Petroleum refin Petroleum- dnd Ruhher- and pia\[ Aircraft and aircr ~ "Odds ratio (OF industrial cat was introduci have the sarr and magnituc it seems like case/control ferences and present exan Selectimultiple myc unlikely that occupations. jobs held for length of jot subjects, wh work history a result of re A last ciations. an< reason. atter suspected. a Associations more cautio1 The re lloyed I O + years mtrols OR ( 9 5 8 CI)" 38 I.z (0.8-2.81 64 1.1 (0.7-2.0) 15 I X (1.0-4.0) 30 I 0 (0.5-2.21 X 0.2 (0.0-1.6) I 0.0(0.0-29.31 I8 1 . 1 10.5-3.5) 0 Inf (0.0-lnf) II 4. I ( I .8-10.4) 8 3.7 (1.3-11.0) 3 4.8 I I . 1-28.8) 3 3.9 (0.4-21.6) I I 1.7 (0.7-7.1) 32 2.0 (I.1-3.9) 4 3.5 (0.7-17.1) I 2.5 (0.1-121) 5 0.7 (0.0-7.0) 15 1.4 (0.5-4.0) 1 2.2 (0.3-17.0) 0 Inf (0.I-tnf) intervals (CI) ell as blue collar 19861 found in:r at home or at and pesticides ibserved among ! to these expoof the interview d "high" expo:rs who did not ast, much of the xltural industry ides (Table VI). radiation could )asis of job title Nas reported by en employed as ig worked in a ,g the results of iures that may cupational and Multiple Myeloma and Occupation 635 TABLE I V . Relative Risk of Multiple Myeloma by Industry Group by Duration of Employment Employed < IO years Employed I O + years Industry group Cases Controls OR (95% CI)" Cases Controls OR (95% CI)" Agnculture Agricultural production Agncultural services Horticultural services Fvrestq Lumber and u d product manufacturing Mining and extracting Metal mining Coal mining Oil and natural gas extraction Non-metal mining and quarrying Paper product manufacturing Pulp. paper. and paperboard mills Paperboard containers and boxes Miscellaneous paper and pulp products Petroleum and coal refining and manufacturing Petroleum refining Petroleum- and coal-product manufacturing Rubber- and plastic-product manufactunng Aircraft and aircraft pans manufacturing 34 29 3 4 4 21 14 4 6 2 3 2 I 0 I 4 I 3 6 30 81 1 . 1 (0.8-2.0) 50 75 1 . 1 (0.7-1.9) 48 5 0.7 (0.1-5.1) 0 7 1.3 (0.3-5.7) 2 4 2.1 (0.3-13.0) 0 32 1.3 (0.7-2.7) 4 34 1.3 (0.6-2.7) I2 I I l.t(O.3-4.6) 4 16 1.2 (0.4-3.5) 6 3 1.6 (0.1-18.5) 0 5 1.9 (0.4-10.0) 2 20 0.2 (0.0-1.0) 4 0.3 (0.0-3.5) 8 0.0 (0.0-1.5) 4 2 2 8 0.4 (0.0-3.0) 0 I2 0.8 (0.2-3.0) 4 6 0.5 (0.0-4.2) 0 6 1.2 (0.2-6.0) 4 19 0.9 (0.3-2.6) 3 40 1.7 (1.0-3.1) 20 100 1.3 (1.0-2.2) 95 1.3 (1.0-2.3) 2 0.0 (0.0-24.6) 4 I .6 (0.2-10.6) 0I I 06(0.1-1.9) 22 1.9 (1.0-5.2)' I2 1 . 1 (0.4-6.2) 7 2.9 (0.8-9.9) 2 0.0 (0.0-26.7) I 5.7 (0.1-184) 3 1.5 (0.2-9.7) 2 0.6 (0.0-7.7) I 4.3 (0.2-263) 0- 6 I .9 (0.4-8.4) 2 0.0 (0.0-21. I ) 4 2.6 (0.5-14.6) I I 0.9 (0.2-3.4) 14 1.3 (0.8-4.5) "Odds ratio (OR) adjusted for sex. race. age. and study area and 95% confidence intervals (CI). industrial categories combined jobs with heterogeneous exposures, misclassification was introduced. Even within specific occupations or industries, all workers would not have the same level or type of exposures. It is not possible to evaluate the direction and magnitude of bias that might be introduced by this misclassification. However, it seems likely that such misclassification would be nondifferential with regard to case/control status and so would likely serve to minimize apparent case-control differences and bias the observed relative risks towards the null value. Tables V and VI present examples of the potential effects of such misclassification. Selective recall of past events is another potential source of bias. Given that multiple myeloma is not widely considered to be an occupational disease, it seems unlikely that cases would selectively recall employment in any one job or group of occupations, although it could be posited that cases may be more likely to remember jobs held for shorter periods of time or farther in the past. However, the number and length of jobs reported by self-respondent cases was quite similar to that of control subjects, while surrogate respondents for the deceased cases provided a less detailed work history. Thus it seems unlikely that the excess risks observed in this study were a result of recall bias. A last limitation is that this study examined a large number of potential associations, and, by chance alone, some may appear elevated or depressed. For this reason, attention in this paper has focused on occupations and industries that were suspected, a priori, to be associated with an elevated risk of multiple myeloma. Associations that have not been observed in previous studies should be considered more cautiously. The results of these analyses provide support for a number of associations 636 Demers et al. TABLE V. Relative Risk of Multiple Myeloma Associated With Employment as a Painter and Self-Reported Exposures to Paint or Solvents Exposure group Ever em.dov* ed as a Dainter No No Yes Yes Reported high exmsure to Daints or solvents ~ ~~~ No Yes No Yes All respondents Self-responding cases only Cases Controls OR (959CI)" Cases OR (959CI)" 1,619 592 1.0(reference) 427 1.0 (reference) 33 60 1.5 (0.9-2.5) 23 1.6(1.0-3.2) 14 19 2.0 (0.8-4.5) 8 1.9(0.6-5.5) 17 21 2.3 (1.1-4.9) 14 3.1 (1.5-7.5) Wdds ratio (OR) adjusted for sex. race, age. and study area and 959 confidence intervals (CI). TABLE VI. Relative Risk of Multiple Myeloma Associated With Employment in Agriculture and Self-Reported Exposures to Pesticides Exposure group Self-responding Ever employed Reponed high exposure All respondents cases only in agriculture to pesticides Cases Controls OR (959CI)" Cases OR (95% CI)" No No 596 1,494 I .O (reference) 402 I .O (reference) No Yes 19 19 2.1 (1.0-4.8) 13 2.I (0.9-5.6) Yes No 68 163 1.1 (0.9-1.8) 49 I .4( 1.0-2.3) Yes Yes 9 7 5.2 (1.6-21.1) 8 7.9(2.4-46.1) "Odds ratio (OR) adjusted for sex, race. age. and study area and 958 confidence intervals (CI) observed by previous investigators. In this study, small excess risks were associated with employment as a farmer or farm worker or in agricultural production or horticultural services, and these risks tended to increase with increasing duration of employment. However, no excess was observed among gardeners and unpaid family members. The magnitude of the increased risk observed in many previous studies also was small, but relatively consistent [Blair and Hoar Zahm, 19911. The agent or agents possibly responsible have not been identified, although attention has focused primarily on insecticides and herbicides [Blair and Hoar Zahm, 1991; Pearce and Reif, 19901. The work history information utilized for the present analysis lacked the detail necessary to shed further light on this issue. However, the excess risk was much higher among agricultural workers who reported that they had been highly exposed to pesticides than for those who did not. Boffetta et al. [ 19891 observed a similar pattern among participants in the American Cancer Society Cancer Prevention Study; among farmers who reported exposure to pesticides. the OR was 4.3, while a much smaller risk was observed among farmers who reported no exposure (OR = 1.7). An excess risk of multiple myeloma among painters also was observed. While solvent exposure is well recognized among painters, paint is actually a complex mixture of pigments, solvents, binders, and additives, and possible exposures include a number of established carcinogens such as chromates, asbestos, and formaldehyde [IARC, 19891. The fact that the risk was elevated for both those who did and did not report high exy io chronic exp' dence also wa employed as f i dustries. In bo: increasing risk an excess risk The situ; clear. While a logging occup: in lumber and al. [1987], wh forestry and I( Associat previous invez includes grind risk for multii groups has be( Potential expc workers in thi fumes. Howe workers and I risks in this a The exc unexpected. ( industry in sol etal. [I9911 t has a diverse not cluster in In sumr studies conce ers and agrici an excess ri: products mar loggers. littlt with wood e` or petroleum analyses desa ACKNOWL This st from the Ni Environmen ronmental P as a Painter and Self-responding cases only es OR (95% CI)" 7 I .O (reference) 3 I .6 ( I .0-3.`) I .9 (0.6-5.5) 4 3.1 (1.5-7.5) intervals (CI). in Agriculture Self-responding cases only es OR (9S% CI)" 2 I .O (reference) 3 2.1 (0.9-5.6) 9 I .4 (1.0-2.3) 8 7.9 (2.4-46.I ) ntervals (CI). were associated luction or hortiduration of em1 unpaid family ious studies also ' agent or agents focused primar:arce and Reif, acked the detail risk was much ghly exposed to Isimilar pattern 1 Study; among a much smaller .7). bserved . While illy a complex iosures include I formaldehyde .iid and did not Multiple Myeloma and Occupation 637 report high exposure to solvents and/or paint may indicate that the association is due to chronic exposures that may not be described by painters as being "high." Evidence also was found to support the existence of an excess risk among persons employed as firefighters and in the petroleum- and coal-products manufacturing industries. In both instances, there was a suggestion, based on very small numbers, of increasing risk with increasing duration of employment. No evidence was found for an excess risk among refinery workers. The situation with respect to wood-related occupations and industries is less clear. While an excess risk was observed among persons employed in forestry and logging occupations, no excess was apparent among carpenters and persons employed in lumber and wood products manufacturing. This agrees with the results of Flodin et ai. [ 19871, who found an excess risk in workers exposed to fresh wood. Workers in forestry and logging may also be exposed to pesticides such as chlorophenols. Associations also were found in this analysis that have not been reported by previous investigators. Metal- and plastic-working machine operators, a category that includes grinding machine, drill press, and stamping press operators, were at higher risk for multiple myeloma. While an excess risk among some related occupational groups has been observed, this has not been a consistent finding of previous studies. Potential exposures that might explain the excess are difficult to determine. Many workers in this group are likely to be exposed to cutting fluids, metal dusts, or metal fumes. However, other categories with similar exposures, such as precision metal workers and metal- and plastic-processing machine operators, did not have elevated risks in this analysis. The excess observed among workers in the aircraft and parts industry also was unexpected. Garabrant et al. [ 19881 did not find an excess risk among workers in this industry in southern California, but a study of aircraft maintenance workers by Spirtas et al. [ 19911 found an elevated risk among workers exposed to solvents. This industry has a diverse set of potential exposures [Garabrant et al., 19881. However, cases did not cluster in any one occupation within the industry. In summary, this study lends further support to the findings of some previous studies concerning associations between multiple myeloma and employment as painters and agricultural workers. Some evidence, based on small numbers, was found for an excess risk among firefighters and persons employed in petroleum- and coalproducts manufacturing. Although an excess risk was observed among foresters and loggers, little other evidence was found to support the previously noted association with wood exposure. No evidence for an association with employment in the rubber or petroleum refining industries was found. The other associations observed in these analyses deserve evaluation in future studies. ACKNOWLEDGMENTS This study was supported in part by Public Health Service grant ROICA-35679 from the National Cancer Institute. P.A.D. was supported by a Training Grant in Environmental Epidemiology and Biostatistics from the National Institute of Environmental Health Sciences (T32 ES07262). I 638 Demers et at. REFERENCES Adelstein AM (1972):Occupational mortality: cancer. Ann Occup Hygiene 15:53-57. Berthwaite PB, Pearce N. Fraser J (1990):Cancer risks in painters: Study based on the New Zealand Cancer Registry. Br J Ind Med 47:742-746. Blair A, Hoar Zahm S (1991):Cancer among farmers. Occup Med State Art Rev 6:335-354. Boffetta P, Stellman SD. Garfinkel L ( 1989): A case-control study of multiple myeloma nested in the American Cancer Society prospective study. Int J Cancer 43:554-559. Breslow NE, Day NE (1980): "Statistical Methods in Cancer Research-The Analysis of Case-Control Studies." Lyon: IARC Sci Pub No. 32. Burmeister LF, Everett GD. Van Lier SF. lsacson P (1983):Selected cancer mortality and farm practices in Iowa. Am J Epidemiol I18:72-77. Cantor KP. Blair A (1984): Farming and mortality from multiple myeloma: A case-control study with the use of death certificates. JNCI 7225 1-255. Cuzick J (1981): Radiation induced myelomatosis. N Eng J Med 304:204-210. Cuzick J, De Stavola B (1988):Multiple myeloma-A case-control study. Br J Cancer 57516-520. Flodin U. Fredriksson M, Penson B (1987): Multiple myeloma and engine exhausts. fresh wood and creosote: a case-referent study. Am J Ind Med 12519-529. Friedman G D (1986): Multiple myeloma: Relation to propoxyphene and other drugs, radiation and occupation. Int J Epidemiol 15:424-426. Gallagher RP. Spinelli JJ. Elmwood JM, Skippen DH (1983): Allergies and agricultural exposure as risk factors for multiple myeloma. Br J Cancer 48:853-857. Garabrant DH, Held J, Langholz B, Bernstein L (1988): Mortality of aircraft manufacturing workers in Southern California. Am J Ind Med 13:683-693. Gilbert ES, Fry SA. Wiggs. LD. Voelz GL. Cragle DL, Petersen GR (1989):Analysis of the combined mortality data on workers at the Hanford Site. Oak Ridge National Laboratory, and Rocky Flats Nuclear Weapons Plant. Radiat Res 120:19-35. Howe GR, Burch JD (1990): Fire fighters and risk of cancer: An assessment and overview of the epidemiologic evidence. Am J Epidemiol 132:1039-1050. IARC (1989): "Some Organic Solvents, Resin Monomers and Related Compounds. Pigments. and Occupational Exposures in Paint Manufacturing and Painting." IARC Monographs on the Eval- uation of the Carcinogenic Risk of Chemicals to Humans, Vol47. Lyon: International Agency for Research on Cancer. Koepsell TD, Daling JR. Weiss NS. Taylor JW. Olshan AF. Lyon JL, Swanson GM, Child M (1987): Antigenic stimulation and the occurrence of multiple myeloma. Am J Epidemiol 126:1051-1062. La Vecchia C, Negri E. D'Avanzo B. Franceschi S (1989):Occupation and lymphoid neoplasms. Br J Cancer 60:385-388. Lundberg I (1986): Mortality and cancer incidence among Swedish paint industry workers with long term exposure to organic solvents. Scand J Work Environ Health 12:108-113. Milham S (1971): Leukemia and multiple myeloma in farmers. Am J Epidemiol 94:307-310. Milham S (1976): Neoplasia in the wood and pulp industry. Ann NY Acad Sci 271:294-300. Milham S, Demers RY (1984): Mortality among pulp and paper workers. J Occup Med 26:844-846. Moms PD, Koepsell TD. Daling JR. Taylor JW. Lyon JL. Swanson GM, Child M, Weiss NS (1986): Toxic substance exposure and multiple myeloma: A case-control study. JNCI 76:987-994. Nandakumar A, Armstrong BK. De Klerk NH (1986): Multiple myeloma in Western Australia: A case-control study in relation to occupation. father's occupation. socioeconomic status and country of birth. Int J Cancer 37:223-226. Pearce N, Reif JS (1990): Epidemiologic studies of cancer in agricultural workers. Am J Ind Med 18:133-148. Pearce N, Smith AH, Howard JK. Sheppard RA. Giles HJ. Teague CA (1986): Case-control study of multiple myeloma and farming. Br J Cancer 54:493-500. Reidel DA, Pottern LM. Blattner WA (1991):Epidemiology of multiple myeloma. In Wiernik PH (Ed): "Neoplastic Diseases of the Blood." New York: Churchill Livingston. Spirtas R. Stewart PA. Lee JS. Marano DE. Forbes CD. Grauman DJ. Pettigrew HM, Blair A. Hoover RN. Cohen JL (1991): Retrospective cohort mortality study of workers at an aircraft maintenance facility. I Epidemiological results. Br J Ind Med 48:515-30. Steineck G. Wiklund K (1986): Multiple myeloma in Swedish agricultural workers. Int J Epidemiol I5:321 - 3 3 , Tollerud DJ. BI eloma ai Vaughan TL ( I ' Ind Med Wong 0. Raab, with a q -57. on the New Zealand 6:335-354. veloma nested in the ysis of Case-Control ry and farm practices ontrol study with the ncer 5 7 5 16-520. 'its, fresh wood and (rugs, radiation and iral exposure as risk .hcturing workers in sis of the combined ry. and Rocky Flats id overview of the ids. Pigments. and gaphs on the Evaliational Agency for 1. Child M (1987): Ol 126:1051-1062. id neoplasms. Br J ters with long term 307-3 IO. 294-300. 4ed 26:844-846. Weiss NS (1986): 76:987-994. ;tern Australia: A status and country I. Am J Ind Med e-control study of Wiernik PH (Ed): Blair A. Hoover craft maintenance Int J Epidemiol Multiple Myeloma and Occupation 639 Tollerud DJ, Brinton LA. Stone BJ. Tobacman JK. Blattner WA (1985): Mortality from multiple my- eloma among North Carolina furniture workers. JNCl 74:799-801. Vaughan TL (1989): Occupation and squamous cell cancers of the pharynx and sinonasal cavity. Am J Ind Med 16:493-510. Wong 0. Raabe GK ( 1989): Cntical review of cancer epidemiology in petroleum industry employees, with a quantitative meta-analysis by cancer site. Am J Ind Med 15:283-310.