Document BQJ5wnGbyvq8M3vRn1Eekgqj

|U The Association Between Parental Occupation and Childhood Malignancy Sew-Leong Kwa, M.D., and Lawrence J. Fine, M.D. A case control study was conducted to test Fabia and Thuy's observation that there was an excess of fathers in hydrocarbon-related occupations among children who died of childhood cancer compared to their controls. The study comprised 692 children who were born and died in Massachusetts for the years 1947-1957, and 1963-1967 and a control group of 1,384. No significant association was found between the four major groups of childhood cancer and the three hydrocarbon-related occupations: (1) 'echanics and gas station attendants; (2) machinists; and i3) painters, cleaners, and dyers. However, there were two significant associations: (a) paternal employment as a paper or pulp mill worker was associated with tumors of the brain and other parts of the nervous system (relative odds of 2.81 and(b) paternal employment as a mechanic or machinist was associated with tumors of the urinary tract (relative odds of 2.5). Without strong supporting evidence from other studies, the authors are reluctant to conclude that these associations are causal. A weak association be tween childhood leukemia-lymphoma and parents' ages was observed. Several investigators have studied the relationship of parental occupation to the risk of an offspring developing a childhood malignancy.'11 Fabia and Thuy' proposed that "hydrocarbon-related" occupations of the father at the time of the birth are associated with cancer in the off spring Children who had fathers employed in hydrocar bon-related occupations had a 2:1 relative risk of dying from malignancy before the age of five compared to the non-exposed group Hakulinen et al,' in Finland, and Zack et al,1 in Texas, did similar studies. No significant associa tions were found between the most common types of From the Department of Physiology Harvard School of Public Health (Dr Kt*a, and Dr Fine Assistant Professor of Occupational Medicine) Dr Fme is currently Associate Professor Department of Environmental and Industrial >nalth University of Michigan, Ann Arbor Ml 46109 This investigation was supported in part bv Center Grant No f S-OOOOZ. Na tional Institute of Environment*) Health Sciences 792 childhood cancers and hydrocarbon-related occupations This study has been conducted with similar objectives in Massachusetts, using data from a previous investiga tion of the role of prenatal x-ray in causing childhood cancer4 Material and Method The basic design of this investigation was a case- control study. Deaths from childhood cancer in Massa chusetts were identified with the assistance of the Na tional Center for Health Statistics * In this data set. there was a total of 692 children who were born during the years 1947 to 1957 and 1963 to 1967 and who died of cancer before the age of 15 in Massachusetts Three cases in which either the children were twins or information on the fathers was not accessible due to confidentiality maintained in connection with illegitimacy or adoption were not included in this study. For each case two con trols were selected. These were the children whose birth registration immediately preceded and followed that of the case subject The total control population was 1,384 For analysis, cancer deaths were divided into four categories of malignant neoplasms: leukemia and lym phoma (ICD No. 200, 202, 204), brain and nervous system cancer (ICD No. 193), urinary tract and Wilm s tumor (ICD No 180,181), and all other tumors (ICD No. 140-205) Three groups of occupations of possible high risk of ex posure to hydrocarbons were identified using Fabia and Thuy's classification: (1) motor vehicle mechanics and ser vice station attendants; (2) machinists, miners and lumbermen; and (3) painters, dyers, and cleaners The parental occupations were extracted from the birth certificates and were coded using a code developed by Wegman, which initially contained several hundred oc cupational titles. These codes were combined into nine categories based on the type of chemical exposure, the nature of the work, and the number of cases (Table 1). Statistical analysis was done with the Statistical Package for the Social Sciences.1 OLI 7087 Tablo 1. - Distribution of Occupations ol Fathers lor Casos and Controls Occupation ot Father* Mecnamcs, gas station attendants Machinists Painters, Cleaners, dyers Motor vehicle drivers Printers Baker, restaurant workers Professionals and administrators White-collar workers Blue-collar workers and others Total = 10 59. p = 0 65 Casast No. K 35 71 10 . 42 15 18 172 75 254 692 51 10.3 14 61 22 26 24 6 10 8 36 7 100 0 * Occupation at the time ot birth ot the offspring t No occupation was listed on the birth certificates tor 12 ot the controls Controls t No. K 61 127 24 91 17 27 334 161 503 1372 44 93 17 66 12 20 24 3 11 7 38 6 100 0 Results The results will be presented in three parts, (a) the distribution of occupations among parents of the children who died of cancer and of the controls; (b) the distribution of occupations among parents of the children who died by specific type of malignancy; and (c) the distribution by age of parents of the subjects and of the controls First, the distribution of occupations of fathers does not differ between cases and controls (Table 1). The pro portion of cases with fathers in the three hydrocarbonrelated occupations'J1 is not excessive. Several occupa tions are represented by proportionately more cases than controls; however, when compared with all other occupa tions, these excesses were not statistically significant. For 98% of both the cases and the controls "housewife" was listed as the occupation of the mother. Second, the distribution of specific type of childhood cancer cases and of controls by paternal occupation does not support the hypothesis that paternal occupation at the time of the birth of the offspring is related to the subsequent risk of the offspring dying form leukemia (Table 2). Two associations were statistically significant. There is a weak association between the combined group of blue-collar workers and the diagnosis of central ner vous system tumors (relative odds of 1.48 and p of 0.67 by chi-square). Within the group of blue-collar workers, the strongest association between specific occupation title and central nervous system cancer was for paper and pulp mill workers (relative odds of 2.8 and p of 003) There were six cases of nervous system cancer in this group (Table 3). Likewise, there was an association be tween the combined occupational categories of mechan ics' and machinists1 and the diagnosis of urinary tract tumors (relative odds of 2.5 and p of 0 02). The occupa tional title with the largest number of cases of uripary tract tumors was machinists, with five (Table 3) There were similar distributions of cases to controls when analy sis was restricted to cases under the age of five in a man ner similar to that of Fabia and Thuy Third, there is a statisticaly significant but weak association between maternal and paternal ages and deaths from leukemia or lymphoma (Table 4). Because of the high correlation between the mother's and fathers ages and the weakness of the association between paren tal age groups and leukemia in the offspring, the authors were not able to separate the effects of the mother's age from those of the fathers age upon the incidence of leukemia. Similarly, there was a weak but statistically significant association between being a male child and dying from leukemia or lymphoma (Table 4). Deaths from tumors of the urinary tract or the nervous system were not related to the age of the parents or the sex of the child. Since 97% of the cases and the controls were Caucasian, they did not differ with regard to race Tibi* 2. -- Dlttrtbetlem of Occupations of Fathers at Tima of Birth for Casas and Controls by Diagnosis for Agas < 15 Yaars. Occupation of Fsthor Mechanics and gas station attendants Machinists Painter Printer Motor vehicle driver Baker, cook, restaurant worker Professionals. administrators White-collar workers Blue-collar workers Total leukemia, Lymphoma Cases Neurological Cancer S Casos Urinary Tract Cancer Cataa All Other Cancar % Caaaa K Cantrals 21 4.9 6 49 11 4 9 7 1.6 1 9 2.9 2 28 65 5 45 4 68 6 08 t 15 1 38 2 12 2.8 2 1 5 0 102 23 7 32 24 2 6 53 12.3 12 91 2 149 34.7 63 47 7 12 430 100 132 100 34 11.8 17 6 2.9 29 5.9 0.0 17.6 59 35 3 too 4 7 1 3 7 4 32 8 30 96 4.2 73 10 31 73 42 33 3 83 31 3 100 61 127 24 17 91 27 334 161 530 1372 K 44 93 1.7 12 66 20 24 3 11 7 38 6 100 Journal of Occupational Medicine/Voi. 22, No. 12/December i960 793 Tibia 3 - Aga it Death and rear of Birih of Selected Cases. Occupition* Facer or pulp m:!l Paoer or pulp mnl Paper or pulp mill Paper or pulp mill Paper or pulp mill Paper or pulp mill Machinist Machinist Machinist Machinist Machinist Auto mechanic Auto mechanic Mechanic Mtsc. Mechanic Misc Filer Ag* it Ceith Yiar ot Birth 2 1517 2 1947 2 1949 1 1950 1 1952 ) 1954 1 1947 2 1947 2 1948 1 1953 1 1960 2 1949 2 1958 t 1948 1 1950 1 1947 Ctuto of Otilh 8-a:n cr ff S * 9ram or ,y $ Brain or N S Brain or N S Brain or N S Brain or N .S Kidney Kidney Kidney Kidney Kidney Kidney Kidney Kidney Kidney Kidney 'Paper or pulp mill -- paper, pulp mill worker, machinist -- machinists, type unspecified, auto mechanic -- mechanics and repairmen, type unspecified: tiler -- filers, grinders and polishers, metal cutters f Malignant neoplasm of brain or other parts of nervous system Discussion One of the principal purposes of this study was to evaluate the finding reported by Fabia and Thuy that there is a relationship between an offspring's risk of dying from cancer by the age of five and paternal occupation which involves hydrocarbon exposures Interestingly, there was no evidence in their data for an association of paternal occupation and offspring risk of dying of Wilm s tumor, the major tumor of the urinary tract in children However, the only association the authors found involv ing hydrocarbon-related jobs and childhood cancer is an association between urinary tract tumors (Wilm's tumors) and some paternal occupations involving exposure to hydrocarbons. It is believed that the data in this study are more consis tent with those of Hakulinen and Zack who did not find any relationship between hydrocarbon-related occupa tions and childhood cancer. Because the number of cases for the hydrocarbon-related occupations is small, and because neither Fabia. Hakulinen or Zack found a similar association, it cannot be concluded that paternal employ ment in these occupations is causally related to the development of Wilm's tumors in offspring. Likewise, the authors cannot conclude that paternal employment in paper-pulp mills is causally related to the development of tumors of the nervous system in offspring. It is reassuring that none of these cases occurred after 1954, despite the fact that data were collected from 1947 to 1957 and from 1963 to 1967. This study has several limitations; first, the listing of oc cupations in birth certificates is often not very specific and probably somewhat unreliable. However, this applies equally to records of both subjects and controls. Second, Tibti 4. -- Relation Between Freni s Age. Sex of the Child, and Leukemia. Contfsl Citii with Leukemia ot Lymphoma P* Numoer of subiecis 1364 Age ol lather ( sD) 30.5 6 6 Age of mother ( SO) 27 6 5 7 Percent male 51 3 430 31,5 70 28 6 6.0 59.5 0,007 0 003 0.003 p by 1 - test or chi-square test attempts to investigate the mother's occupation were un successful because for 98% of both the subjects and the controls it was listed as "housewife" Third, the large variety of occupations listed on the birth certificates forces one to combine occupations into groups. This pro cess necessarily results in groups with heterogeneous ex posures. However, with a few occupations it was possible to form groups which were represented by at least ten cases and which were likely to have had similar exposure, such as printers The associations between older fathers and the in cidence of childhood cancer are of interest. Previous authors'1 `T who studied childhood cancers had studied the effect of maternal age as one of their basic variables. Choi et al7 found that children with nervous system cancer tended to have younger mothers. When the data are analyzed by individual types of cancer, more cases than controls in older paternal and maternal age groups are observed, and the mean ages of parents of cases are found to be greater than those of controls. However, the association is significant only for childhood leukemia and lymphoma. The magnitude of the association between the youngest and oldest parental groups is small. The author! wish to thank Drs Richard R Monson and Brim MicMjhon who permitted re-anilysis ot their d*(a for this study. Cindy Rubin and luciHe Poth*f who performed the programming for this Studv; and Dune Ivey who prepared the manuscript end provided editoruf *s$r$t*nce References 1 Fabia I and Thuy TO- Occupation ot father at time of birth of children dying of malignant disease. Sr / Prev Soc Med 28 36-100. 1974 2 Hakulinen T. Salonen T, and Teppo L. Cancer in the offspring of fathers in hydrocarbon-related occupations 8r / Prev Soc Med 3fr138-140, 1976 3 Zack M, Cannon S. Loyd D. et al- Cancer in children of parents exposed to hydrocarbon-related industries and occupations Am / Epidemiol 3 329-336,1980 4 MacMahon 8 and Newill VA Birth characteristics of children dy ing of malignant neoplasms / Naff Cancer Inst 28231-244. 1962 5 Nie NH. Hull OH. lenkins 1C. et al Statistical Package lor the Social Sciences (2nd Ed ) New York: McCraw Hill, 1975 6 Cold E, Cordis L. Tonascia I. and Szklo M Risk factors for brain tumors in children. Am I Epidemiol 109:309-319, 1979 7 Choi NW. Schuman LM. and Cullen WH Epidemiology of primary central nervous system neoplasms, Vol. I(- Case-control Study Am I Epidemiol 91:467-485. 1970. 794 Parental Occupation and Childhood Maiignancy/Kwa and Fin* OLI 7089 m