Document LK5ZXE9ZEGGzKajppzQyN6G5q

AMERICAN JOURNAL 01W ~ ~ U I O L O C Y Copynght 0 1990 by Tba Jobnr Hopkins Univemty School of Hygiene and Public ? h l t b All rights reserved A Brief Original Contribution NEUROBLASTOMA AND PARENTAL OCCUPATION GRETA R. BUNIN,' ELIZABETH WARD,' SH1.U K M R 3 CHARLOTTE A. RHEE,' AND ANNA T.~ O w S ` Bunin, 0. R. (Children's Hospital of Philadelphia, Philadelphia, PA 19104), E. Ward, S. Kramer, C. A. Rhee, and A. 7'. Meadows. Neuroblastoma and parental occupation. Am J Epidemior 1990,131:776-80. A matched case-control study of neuroblastoma investigated parental occupational risk factors. Cases diagnosed in 1970-1979 were identified through tumor registries in the Greater Philadelphia area. Controls were selected by randomdigitdialing andwere matchedto caseson race, birthdate, andtelephone exchange. Parents of 104 matched pairs were intewkwed by telephone. In contrast to results of a previous study, no significant associations were noted for paternal employment in a job cluster of occupations in electricity, electronics, insulation, utility, and printing; in jobs with electromagnetic field exposure; or in jobs 8s workers in electronics. Six cases and one control, however, had a father or mother who worked in electricalor electronic products assembly. electromagnetics; electronics; neoplasms, embryonal and mixed; neuroblastoma; occupations Parental occupational exposures to mu- tagens or carcinogens have been studied as possible risk factors for childhood cancers (1-17). Such exposures could result in cancer through an effect on the germ cells before the child's conception. Because of the early age at onset in its victims, neuro- Received for publication May 23,1989,and in final form November 3, 1989. Abbreviation: EMF, electromagnetic field(s). Children's Hospital of Philadelphia and Depart- ment of Pediatrics, University of Pennsylvania School of Medicine, Philadelphia, PA. National Institute of Occupational Safety and Health, Centers for Disease Control, Department of Health and Human Services. Cincinnati. OH. `MRiadrgsehCaollmU, Innicv.e,rAsirtdymSocreh,oPoAl .of Medicine, Hun- tington, WV. Reprint requests to Dr. Greta R. Bunin, Children's Hospital of Philadelphia, One Children's Plaza, Room 9093,Philadelphia, P A 19104. This study was supported by Public Health Service grant CA-29275 from the National Cancer Institute. The authors thank Mary Gordon-Rowson, Carol Symmes, and the late Helen Peterson for competently interviewing study subjects; Ira Caplan for computer programming; and Jack Elias forpreparing and editing the manuscript. blastoma is a good candidate for a cancer in which parental occupational exposures may play a role. Only one study, however, has investigated such exposures. In that study, Spitz and Johnson (12) observed significantly increased risk for children of fathers employed in occupations with electromagnetic field ( E M F ) exposure, particularly children of fathers who were workers in electronics. In the present study, we sought to replicate the analysis of Spitz and Johnson (12) using data from a case-control interview study of neuroblastoma, in which other factors were also studied. MATERIALS AND METHODS- Children diagnosed with neuroblastoma between 1970 and 1979 were identified through the population-based Greater Del- aware Pediatric (18) and the tumor registry of the Children's Hospital of Philadelphia. The latter regis- residing Outside the Greater Delaware Valley Pediatric Tumor 776 - NEUROBLASTOMA AND PARENTAL OCCUPATION F area. Eligibility criteria were the culated for potential risk factors, and the : physician's approval to contact corresponding two-sided p values were , an English-speaking biologic computed using McNemar's test. Adjust- available for interview, and histo- ment for possible confounders was per- confirmation of diagnosis. formed by conditional multiple logistic e control per case was selected by tele- regression. Statistical significance was de- random digit dialing. Controls were fined at the 5 percent. level. ases by the area code and the ts of the telephone number at RESULTS h ~ r a c eand date of birth One hundred eighty-one cases of con- 3 years). The random digit firmed neuroblastoma were identified. Of has been described pre- these, 21 cases (12 percent) resided outside the Greater Delaware Valley Pediatric Tu- of cases and controls were inter- mor Registry area. Thirty-two cases were not included for the following reasons: 12 for whom the physician could not be con- tacted, 10 for whom the physician refused job title, and major duties. If the permission, six residing outside the United s not available, the mother pro- States,and four with non-English-speaking roxy interview. Data on nonoccu- parents. Of the 149 cases contacted for factors were also collected. The interview, 19 could not be traced, 18 par- the analysis of these data have ents refused to participate, and eight could ished previously (20). not be interviewed before the study ended. tegorized by the The participation rate for cases invited to d by Spitz and Johnson participate was 70 percent (104 of 149). Of of mothers and fathers those participating, 13 cases (13 percent) ustry and job title accord- and their matched controls resided outside the classification scheme of Hoar et the Greater Delaware Valley Pediatric Tu- mor Registry area. ne of 30 clusters An analysis of the selection of controls for this study has been published previously , as designed by Hsieh et al. (22). (18). The response rate among children f 6 months or more prior to the identified as eligible and whose parents were invited to participate was 57 percent (104 of 181). The educational level of fathers of cases les (ever vs. never empioyed in and controls differed (JJ = 0.05) with more for each of two time periods: control than case fathers having had voca- index child's conception and tional training (seven case vs. 29 control regnancy. For the preconcep- fathers). The distribution of fathers among ere performed using other educational levels was similar for bs and using only the longest cases ar.3 controls. The educational levels he results for all preconception of case and control mothers were almost longest held job were compared identical. the possibility of recall bias con- Among case fathers, 61 percent had blue ort-term jobs held many years collar jobs: 24 percent had professional, interview. The results differed technical, or manageriel jobs; and 14 per- results for the longest held job cent had white collar jobs as their longest unless noted otherwise. held jobs.The longest held jobs for control for matched pairs were cal- fathers were 50 percent blue collar, 30per- I BONIN $T AL. cent professional, technical, or managerial, considered, nonsignificant case and 19 percent white collar. The case- were noted in clusters 5 (stock control differences in job distribution were textile and food produc not statistically significant. and two controls) and Analysis of Hsieh's clusters for paternal packers, four cases and ze occupiltion uncovered no significant asso- case and no control mothers ciations. Nonsignificallt case excesses with cluster 7 (electrical products odds ratios of 2.0 or higher were observed metal product fabrication wo for employment in several clusters (table the case mothers in cluster 7 and 1). case mother worked in electronic Spitz and Johnson (12) observed an as- assembly. sociation with job cluster 7 and with occupations with potential EMF exposure, es- DISCUSSION pecially workers in electronics (12). We The analysis of parental occupation found no associations for these groupings performed to attempt to corroborate (table 2). Using their definitions of EMF findings of Spitz and Johnson (12) exposure, we observed only a small case generate other hypotheses. The associ excess of workers in electrical and elec- observed by Spitz and Johnson tronic products assembly and repair. One Hsieh's cluster 7 (occupations in insula control and four case fathers had jobs as- electricity, electronics, and printing) sembling electrical or electronic products. with workers in electronics and elect . Adjustment for educational level and specifically, was not observed here. source of the father's data (proxy vs. non- and Johnson reported an odds ratio proxy interview) was performed for all most 12 (95 percent confidence in groupings except electrical and electronic 1.4-99) for workers in electronics products assembly workers and did not tricity; the present study had a change the results. Small numbers pre- percent to detect an odds ratio cluded adjustment in the case of the assem- actual odds ratio was close t bly workers. confidence interval of 1.4, our Few mothers worked in industrial set- not have detected the excess tings. When all jobs before conception were excess was observed for a subset of E TABLE 1 Job cluster and exposures of father 5 13 22 27 Description Food product packers and warehouse men Farmers and agricultural workers Some electrical equipment assembly and repair jobs, rubber processing, and painting Chemistry occupations Odds ratio 4.0 3.5 2.0 95% -ranfidenc_e Odds Ratiot interval ratio 4:l 0.4-195 1.5 7.2 0.7-34.6 0.7 4:2 0.3-22.1 0.8 - __ 95% confidence Ratio 7xiCervd 3:2 0.2-18.0 2:3 0.1-5.8 4 5 0.2-3.7 .' t Case exposed-control unexposed pairs:case unexposed-control exposed pairs. NECROBLASTOMA AND PARENTAL OCCUPATIOS 779 TABLE 2 &cupational groupings relevant to the study of Spitz and Johnson (12) results of a mtched case-control study Fathers Job cluster 7 t Occupations with EMF exposure Narrow definition$ --Broad definitions lectrical and electronic products workers lectrical and electronic products assemblers Mothers11 Job cluster 7 Electrical and electronic Odds ratio Ratio' Exact 95% confidence interval 0.3 3:lO 0.05-1.2 Odds ratio . Ratio Exact 95% confidence interval 0.3 3:9 0.1-1.3 1.3 9 7 0.4-4.1 -03-2.5 1.6 8 5 0.5-6.2 4.0 4:1 0.4-195 0.3 3:9 0.1-1.3 0.6 712 0.2-1.6 0.4 3:s 0.1-1.6 0.0 0 1 Not calculable a 2:O Not calculable tility workers, and workers in en,welders, and utility employees. defined by Spitz and Johnson (12), this includes all those in the narrow definition plus electrical ment salesmen and repairmen. e preconception period includes all jobs; no mothers had jobs in these categones during the index CY. {&sed workers, i.e., electrical and elec- #nit products assembly workers. There six cases with either a mother or far who held such a job in the preconcep- iod compared with one control. studies or studies in geographic h more manufacturing of electrical tronic products would be needed to e if the excess was due to chance nonparticipation rates were sub- the other hand, slightly more case than control fathers were blue collar workers, and nearly all the occupations in the job groupings of interest were blue collar jobs. Therefore, one might have expected a bins toward finding the associations previously reported. We did not find the previously reported associations of neuroblastoma with paternal employment in cluster 7 or in an occu- pation with potential EMF exposure. We to imagine a bias that selects jobs. Future studies will face the chpllenge rkers with specificexposures. Se- of investigating narrow occupational socioeconomic status, however, groupings while retaining sufficient power. a study of occupational factors. REPEBENCES I BUNIN ET AL. 2. Hakutinen T, Salonen T,Tepp L. Cancer in the pations. Br J Prev Soc Med 1976;~138-40. 3. Zack M, Cannon S. Layd D, et d. Cancer in children of parenta eqxaed to hydrocarbonrelated industries and occupations. Am J Epidemiol 1980,111:329-36. 4. Kwa SL, Fine W. The asnociationbetween parental occupation and childhood malignancy.J &cup Med 1980:22:792-4. 5. Peters JM, Preston-Martin S,Yu MC. Brain tu- mors in children and occupational exposure of parents. Science 1981;213235-7. 6. Hemminki K, Saloniemi I. Salonen T,et al. Childhood cancer and parental occupation in Finland J Epidemiol Community Health 1981;35:11-15. 7. Sanders BM, White GC, Draper GC.Occupations of fathers ef children dying from neoplasms. J Epidemiol Community Health 1981;35:245-50. 8. Gold EB, Diener MD, Szklo M. Parental occupa- tions and cancer in children. J Occup Med 1982;24:578-84. 9. Hicks N, Zack M, Caldwell CG, et al. Childhood cancer and occupational radiation exposure in parents. Cancer 1984;531637-43. 10. Shaw G, Lavey R, Jackson R, et al. Association of childhood leukemia with maternal age, birth or- der, and paternal occupation:a case-control study. Am J Epidemiol 19&L;119788-95. 11. Wilkins JR, Sinks TH Jr. Paternal occupation and Wilms' tumor in offspring. J Epidemiol Community Health 1984;38:7-11. 12. Spitz MR, Johnson CC. Neuroblastoma and paternal occupation: a case-control analysis. Am J Epidemiol 1985;121:924-9. 13. Olshan AF, Breslow NE. Daling JR, et al. Child- hood brain tumors and paternal occupation in the 17. 18. 19. of random digit dialing in selecting matched trols for a case-control study of pediatric ca Am J Epidemiol 1984;120582-91. 20. Kramer S, Ward E, Meadows AT, et al. and drug risk' factors associated with blastoma: a case-control study. JNCI 797-804. 21. Hoar SK, Morrison AS, Cole P tion and exposure linkage syste occupational carcinogenesis. J 22722-6. 22. Hsieh CC, Walker AM, Hoar cupations according to carcino cupation clustere from an tem. Am J Epidemiol1983;