Document J3Q5Rrk7O89wGKamX4Y8p4M8a

Id. J. Cancer.' 31,407-411 (1983) MORTALITY PATTERNS AMONG EMBALMERS JudyWALRATHand Joseph F. FRAUMENJIR,. Environmental Epidemiology Branch, National Cancer Institute, National Institutes of Health, Bethesda, M D 20205, USA. In view of recent findings of nasalcancer in rats exposed to formaldehyde vapors, we investigatedthe proportionate mortality experience of embalmers lkensed to pracdcl In New York State. Mortality was signifiuntly e l c r a d for cancers of the skin and colon and for arteriovkrotk hwt disease, whereas signifKurt defKib were seen in mortality from respiratory disease and accidents. Respiratory cancer mortality was not excessive and no deaths were attributed to nasal cancer. Mortality was signifkantly elevated for cancers of the skin, kidney, and bnin among those licensed only as embalmers, whereas pattern were unremarkabk among those lktmQd a b as funeral directors (and presumablykuexposed to formaldehyde). ThtK preliminary results indicate the needfor occupationrlcohort studies to clarify the adtogenk potential of formaldehyde. \ Since the turn of the century, formaldehyde has been the main preservative in commercial embalming fluids (Champion Company, 1966). A survey of six funeral homes revealed airborne formaldehyde levels of up to 5.26 ppm, with average concentrations of 0.25-1.39 ppm. Paraformaldehyde particles in formaldehyde vapors were small enough to be deposited in the lungs (Kerfoot and Mooney, 1975). A survey of a mortuary science college by the National Institute for Occupational Safety and Health (1980) reported that the airborne formaldehyde concentrations ranged from 0.20 to 0.91 ppm, but exceeded 3.0 ppm in two samples when the ventilation system was inoperative. This proportionate mortality study compares the causes of death among embalmers with those of the eeneral wpulation in a preliminary search for unusual c 1 1i I 408 WALRATl4 AND FRAUMEN Death certificates were requested for 1,678 em- tality patterns, examined separately, were not balmen from the appropriate state vital statistics of- al. The following results refer to white male e mrpnobRbE fices and were received for 1,263 of the decedents ers except where indicated. (75.3 %). The death Certificate search was more suecessful for those whose names were identified from tioTnaobfleemI bparelmseenrtss atht ethaegteimanedocfadleenadtha.r Fyieratyr dpiesrtrQibn&, the registration file (89.1 %) than for those listed only died before age 65. This relatively young age &trih- in the ledger (65.7 %), since more demographic infor- tion is probably related to the method by wbd deuhmation was available from the registration file. Under- were ascertained. Ninety-five per cent of the tlyhiengrules in oefffdeecat taht wtiamsecoOdf eddeabtyh aanndosaoslsoigginsitnugstWhe eornsewyheoasre yoefalricoefnlsaesttleicrmenisnuarteiown.asTkhneowmneddiiadn wbitihnipb- rubrics of the 8th Revision of the International Clas- year for those in the study group was 1902 and tht sification of Diseases, Adapted (ICDA) (World median year of initial license was 1931. Health Organization, 1967). Deaths observed among the embalmers were compared to expected numbers Table 11 Shows the mortality experience for white computed by applying the age-, race-, and calendar male embalmers according to major causes of death year-specific proportions of deaths for each cause Mortality from all malignant neoplasms combinedw& among the US male population to the total number of elevated (PMR=111)3 butnotsignificantly. T h e n w ~ deaths in the study group by 5-year age and calendar a Significant excess of deaths from arteriosckrotk periods (Monson, 1974). Differences between ob- heart disease (PMR=112) and anon-sitFificant ex- served (OBS) and expected (EXP) numbers of deaths of deaths from Cirrhosis ofthe liver (PMR-133). h. for each specific cause of death category were sum- talitY fromdiseases of the respiratory system was sk. marked as the Proportionate Mortality Ratio (PMR), nificantlY lower than expected (PMR=77), particu- the ratio of the number of deaths observed to that larly for Pneumonia (PMR=@) and emphysema expected multiplied by 100. The statistical significance (PMR=67)- Mortality from accidental deaths was also of each ratio was tested by a chi-square test with one SbificantlY low (PMR-49). degree of freedom (Mantel and Haenszel, 1969). n e distribution of malignant neoplasm Proportionate Cancer Mortality Ratios (PCMR'S) white male embalmers is shown in Table 111. Mortality were also computed utilizing the total number Of cancer deaths as the denominator for calculating the expected number of deaths for each cancer site (Monson, 1974). With the exception of comparisons by latency period, age at first license, and type of license, P M R s and PCMR's were not presented when both the observed and expected numbers of deaths were less than five. fromcancer of the oral cavity and pharynx wm a-to that expected, with no death from of nasopharynx. The number of deaths from tract cancers was close to that expected, there was a significant excess of colon deaths VS. 20.3 expected) and a n0n-sim-t of rectal cancer. Respiratory cancer mortality wpt an- remarkable, and no deaths were ascribed to It was not possible to measure length of employ- cancer (although the expected value was only 0.50). ment or length of licensure, since year of last license Mortality from skin cancer was significantly elevated was not available for decedents who were listed only (8 deaths vs. 3.6 expected); four of the eight casta in the ledger. As an alternative, length of time from were malignant melanoma (PMR-Un), tbret were first license to death was used to approximate length of exposure. RESULTS squamous-cell carcinoma, and one was unspecified. Mortality was slightly elevated for kidney cancer (8 deaths vs. 5.4 expected), leukemia (12 deaths vs. 8.5 expected), and cancer of the brain and nervous system (9 deaths vs. 5.8 expected). The cell types of leukemia There were 1,132 (89.6%) white men and 79 reported on the death certificates were: 6 myeloid (5 (6.3%) non-white men in the study group. Forty-two acute and 1 unspecified) vs. 4.1 expected, 1 acute men (3.3%) whose race was unknown and ten women monocytic vs. 3 expected, 4 lymphatic (2 chronic and 2 were not included in the detailed analysis. Their mor- unspecified) vs. 2.6 expected, and 1 acute leukemia -TABLE I DlSTRlBUnON OF DEATHS AMONG WHITE MALE EMBALMERS BY AGE AND CALENDAR YEAR AT DEATH Age at death ms-n ~~ ~ ~~ 20-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80+ 3 6 6 12 15 19 19 17 13 6 4 All age groups 117 Per cent 10.3 1955-59 4 12 9 24 20 25 27 9 9 12 151 13.3 196044 7 5 12 7 15 22 18 23 10 6 125 11.0 I%-@ 6 7 7 18 40 24 37 32 23 6 2M) 17.7 1970-74 5 6 9 20 29 36 30 36 32 35 238 21.o 1975-80 5 9 10 16 33 32 50 45 30 38 301 26.6 AUcvaelemndar 33 48 61 102 157 162 186 163 115 105 1132 kmt 2.9 4.2 5.4 9.0 13.9 14.3 16.4 14.4 10.2 9.3 MORTALITY AMONG EMBALMERS 409 n -NUMBERS OF DEATHS AND PROPORTIONATE MORTALKY RATIOS (PMRs) AMONG WHlTE MALE EMBALMERS, BY CAUSE OF DEATH 5%. r Cause of death (8th rcnnon, ICDA) Numben of deatbs OBS Exp PMR malignant neoplasms (1 40-209) Circulatory system (390-458) Arteriosclerotic heart disease (410-414) Cerebrovascular disease (430-438) Respiratory system (460-519) Pneumonia (480-486) Emphysema (492) Digestive system (520-577) Gastric and duodenal ulcers (531. 532) C i h o s i s of liver (571) External causes (800-999) Accidents (800-949) Suicide (950-959) Other causes M c~usesof death 'pa.m. 243 654 48 1 94 52 16 12 60 8 34 52 28 21 71 1132 218.9 622.4 430.7 88.9 67.9 25.2 17.9 52.5 8.6 25.5 85.2 57.3 21.1 85.1 1132.0 111 105 112' 106 771 64 67 114 94 133 611 491 99 83 100 -% 1.5 expected. The reported types of brain deaths vs. 14.5 expected). There was significantly high were 6 glioblastomas (67%), 2 astrocytomas mortality from cancers of the larynx and lymphatic1 carcinoma (11 %). In the State of Con- hematopoietic system, with only two and three deaths .3% of the incident cases during the years observed, respectively. Mortality from arteriosclerotic re glioblastoma, 18.0% were meningioma, heart disease was significantlyhigh (33deaths vs. 21.3 astrocytoma, and other histologic types expected). Deaths from diseases of the respiratory 17.6% of the cases (Schoenberg et al., system and from external causes were lower than exfor K M R s was similar to PMR's pected (1 death vs. 4.8 expected for both causes). for cancers of the colon, skin, kid- The PMRs for selected cancer sites among white deficits for cancers of the stomach males were examined by "latency period", defined as the length of time from first license to death (Table Amops the TJ non-white men, mortality from all IV). Since most embalmers died while licensed, the neoplasms combined was elevated (20 length of licensure is assumed to be synonymous with - NUMBERS OF DEATHS FROM MALIGNANT NEOPUSMS WITH PROPORTIONATE MORTALITY RATIOS (PMb) AND PROPORTIONATE CANCER MORTALITY RATIOS (PCMRS), AMONG WHITE MALE EMBALMERS CIure of dcath (8th revision. ICDA) Numkn of deaths OBS EXP PMR PCMR malignant neoplasms (1 40-209) cavity and pharynx (140-149) 243 218.9 111 8 7.1 113 100 103 %Wstive organs and peritoneum (150-159) Esophagus (150) Stomach (151) Colon (153) W m 1154) Liver an;-gail-bladder (155-156) (157) -tory system (160-163) 68 65.2 104 89 5 5.3 95 88 12 13.4 90 71 29 20.3 143l 130 3 7.7 39 32' 5 4.7 106 119 13 12.3 105 95 74 70.7 I05 109 (161) s bh n g and pleura (162, 163) (172, 173) -tc (185) Blrdder (188) -Y (189) B* and central nervous system (191, 192) 2 3.4 72 66.8 108 1113 8 3.6 221 188 15 16.4 91 81 7 7.3 96 85 8 5.4 150 142 9 5.8 156 138 Lymphatic and hematomietic svstem (200-209) 25 20.6 121 -2 Lymphosarcoma aAd reticdosarcoma (206) 5 4.7 108 82 - Hdgkm's disease (201) Other lymphatic cancers (202, 203) Leukemia (204-207) cancers 2 2.3 6 4.9 123 -2 12 8.5 140 119 21 16.8 125 -2 i &~&-b~~acQukl 'EmA ttributneodt btoe Dcaklucurlaaltceadnfcoerrtsh)e. se sites due to limitations in the computer prognmw d . - for cancer of lun&b a d on 410 WALRATH AND FRAUMENI -TABLE IV NUMBERS OF DEATHS AND PROWRTIONATE MORTALITY RATIOS (PMRs) FOR SELECTED W G N M mo- rnAMONG WHlTE MALE EMBALMERS. BY LATENCY PERIOD Cause of death (8th -a, ICDA) 'Latency period c35 yeprs (n = 606) OBS EXP W R 35+ YOBS - ~-(n 526) - PUR All malignant neoplasms (140-209) Respiratory system (160-163) Skin (172, 173) Kidney (189) Brain and central nervous system (191, 192) Lymphatic and hematopoietic system (200-209) 123 113.2 109 39 36.1 108 4 2.3 173 6 3.0 201 6 4.0 150 14 11.6 121 120 105.8 113 35 34.7 101 4 1.3 3082 2 2.4 8 3 1.8 169 11 9.1 121 -'Defined IS length of time from first Iicens-2 to death. ' p 405. latency period. Although the numbers of deaths were small, the proportion dying from skin cancer was greater among those licensed for 35 or more years (4 deaths vs. 1.3 expected) than among those licensed for less than 35 years (4 deaths vs. 2.3 expected). Cancers of the brain and lymphatic-hematopoietic system were also slightlyelevated in both latency groups. The elevated PMR for kidney cancer was limited to em- balmers licensed for less than 35 years. Mortality patterns were examined separately for persons licensed only as embalmers and for those held licenses for both embalming and funeral direct- ing. We assumed that persons licensed only s em- balmers experienced a greater cumulative exposure to formaldehyde than did embalmers who were also funeral directors. As shown in Table VI,the PMR's for cancers of the skin, kidney and brain were si-t- ly elevated for those licensed only as embalmen, -TABLE V NUMBERS OF DEATHS A N D PROPORTIONATE MORTALITY RATIOS (PMRS) FOR SELECTED MALIGNANT NEOpLAsMs AMONG WHlTE MALE EMBALMERS, BY AGE AT FIRST LICENSE Cause of duth (8threvision. ICDA) <XI yeus (n = 665) Age l t lint I*ear -30+ y e l n (n 467) OBS EXP W R OBS E W PMR All malignant neoplasms (140-209) Respiratory system (160-163) Skin (172, 173) Kidney (189) Brain and central nervous system (191, 192) Lymphatic and hematopoietic system (200-209) 153 133.4 115 50 46.1 108 3 2.3 129 4 3.4 118 4 4.1 98 18 13.1 138 90 85.5 105 24 24.6 97 5 1.3 3871 4 2.0 201 5 1.7 2%' 7 7.6 93 ? p <O.M. Mortality by age at first license is shown in Table V. while no unusual patterns were observed among those A significant excess mortality due to skin and brain who held both licenses. The difference in year of cancers was seen among those who began to practice death between the two groups results from the embalming at age 30 or later. Bureau's practice since 1947 of issuing one license TABLE M- NUMBERS OF DEATHS AND PROPORTIONATE MORTALITY RATIOS (PMRs) FOR MALIGNANT h'EOPLASMS AMONG WHITE MALE EMBALMERS. BY TYPE OF LICENSE C a w of death (8th revision,ICDA) Typc of license Embalmer only (n = 546) OBS EXP PMR Both embalmer and funeral director (n = 586) OBS EXP PMR All malignant neoplasms (140-209) Buccal cavity and pharynx (140-149) Digestive organs and peritoneum (150-159) Respiratory system (160-163) Skin (172, 173) Prostate (185) Bladder (188) Kidney (189) Brain and central nervous system (191, 192) Lymphatic and hematopoietic system (200-209) Average age at death Average year of death 105 99.2 106 7 3.5 201 33.5 78 276 28.6 94 5 1.5 326' 6 7.3 83 2 3.5 57 6 2.4 247' 6 2.6 234l 9 9.1 99 63.3 1960.8 138 119.8 115 1 3.6 28 42 31.7 133 47 42.1 112 3 2.1 144 9 9.1 99 5 3.8 132 2 2.9 69 3 3.2 93 16 11.5 139 64.4 1974.2 'p 4.05. M O R T A L m AMONG EMBALMERS 411 which covers both funeral directing and embalming. This study was intended as a preliminary epidemi- Wrtality among those licensed only as embalmers ologic study of the chronic effects of formaldehyde was examined by latency period.Brain cancer mortali- exposure and has several deficiencies. The study ty was significantly elevated among those licensed for group was not exposed solely to formaldehyde but to lem than 35 years (5 deaths vs. 1.9expected) and skin embalming fluids that contain other chemicals (e&, cancer mortality was significantly elevated among tissue moisturizers, antiseptic solutions, dyes, and those licensed for 35 or more years (3 deaths vs. 0.5 deodorizers), partly to offset the adverse reactions of expected). formaldehyde (Champion Company, 1966; Kirk-0th- mer, 1965). In addition, length-of-employment infor- mation was unavailable, the ascertainment of deaths DISCUSSION among retirees was incomplete, and the size of the using the proportionate mortality approach, study group may be insufficient to detect elevated study shows that emba]mers experienced a slightlyele- risks of rare neoplasms such as nasal cancer. Of s p - vatdmortality from Mefiosclerotic heart cancer, disease, aandsiagnsiifgicnaifnitcant of deficit cia1 concern are the limitations of the proportionate mortality method, notably the uncertainty that an ex- of respiratory diseases and accidental deaths. The site- cess propoflion of deaths from a specific reflects specific patterns of cancer mortality were not suffi- a real elevation in mortality or a deficit in the propor- cient to implicate formaldehyde as a carcinogen, but tion of deaths from other Causes (Decoufle er al.* wme variations were intriguing.Skin cancer mortality 1980). Nonetheless, our findings and other prelimi- (including melanoma) was significantly elevated, with nary studies (Harrington and 1975; a greater excess observed among those licensed for 1983; and WOW, 1983) indicate that cohort studies Of more than 35 years and those who began employment several groups of fomaldehYde-exPsed workers are at age u) or later. Also significantly elevated was the I E d e d to further assess the carcinogenic risks Of this proportionate mortality from skin, kidney and brain used CaOcers among those who were licensed only as embdglers. There was, however, no excess mortality ACKNOWLEDGEMENTS h m cancers of the respiratory tract including the na- The cooperation of Mr.Nicholas P.Panepinto and passage. This is noteworthy in view of evidence Mr.Thomas Thorp of the Board of Funeral Directing th@tinhalation of formaldehyde induces squamous- and Embalming, and of Mr. Philip Quickenton from carcinomas of the nasal cavity in rats (Kerns et al., the Division of Vital Statistics of the New York State I=). The mortality patterns in our study, however, Department of Health, is gratefully acknowledged. that further investigation of the carcinogenic We thank Dr. Aaron Blair for helpful suggestions, Of formaldehyde should extend beyond the re- Ms. Hilda Young for preparation of the manuscript, spiratory system. and Ms.Marian Heid for nosology coding. REFERENCES m,R.E., SEUKUMAR, A.R., LASKINS,.,KUSCHNER, M., N-N, N.. and SNYDER. C.A.. Gaseous formaldehvde and hv- chloride induction of nasal cancer in the kt.1. mi. OnCcr ~ m t .a, ,597-603 (1982). KIRK. R.E., and OTHMER,D.F., Embalming fluids. In: Encyclopedia of chemical technology, Vol. 8, pp. 100-102,Interscience Publishers, New York (1965). -1, N., and HAENSZEL,W., Statistical aspects of the analy- BATIELLE COLUMBUSLABORATORIES, F i ~ rleport on a chronic sis of data from retrospective studies of disease. J. nat. Cancer w o n toxicology study in rats and mice exposed to formal- fml., 22, 719-748 (1959). dchyd., Chemical Industry Institute of Toxicology, Research MARSH, G.M., Proportional mortality patterns among chemical 'Wde Park, NC (1981). plant workers exposed to formaldehyde. In: Proceedings of rhe -ION COMPANYE, xpanding encyclopedia of mortuary prachbkshed by the Service and Research Dept. of the Champ- Ion Company, Springfield, Ohio, No. 365,March (1966). b W ,P., THOMAS, T.L., and PICKLE, L.W., Comparison of &. POPortionate mortality ratio and standardized mortality risk measures. Amer. J. Epiderniol., 111, 263-269 (1980). Third CIIT Annual Conference: formaldehyde roxicjty. Hemisphere Publishing Corporation (1983) (in press). MONSON, R.R., Analysis of relative survival and proportional mortality. Cornput. Biorned. Res., 7, 325-332 (1974). NATIONAL INSTITUTE FOR OCCUPATIONAL SAFETY AND HEALTH, Criteria for a recommended standard. Occupational exposure to formaldehyde. DHEW (NIOSH) Publ. No. 77-126. pp. 32-59. h G m N , J.M.. and SHANNON, H.S., Mortality study of US Government Printing Office, Washington, DC (1976). m t b o ~ s t sand medical laboratory technicians. Brit. Med. I . , 4, 329-332 (1975). NATIONAL INS'ITTUTEFOR OCCUPATTONM SAFETY AND HEALTH. Health hazard evaluation determination report No. HE 79-146- O M . , Cancer risk from formaldehyde. Lancet, 2,480- 670, Cincinnati, Ohio (1980). (1980). SCHOENBERG, B.S., CHRISTINE, B.W., and WHISNAKTJ ., P.,T h e O.M., and ANDERSENS, .K., Lung cancer risk from +.b a e h y d e . Lancet, 1, 913 (1982). E.J., and MOONEY,T.F., JR.. Formaldehyde and Plrsfonnaldehyde study in funeral homes. Amer. industr. Hyg. A'. s,553-537 (1975). descriptive epidemiology of primary intracranial neoplasms: T h e Connecticut experience. Amer. 1. Epiderniol.. 104, 499-510 (1976). WONG, 0..An epidemiologic mortality study of a cohort of chemical workers potentially exposed to formaldehyde, with a discussion of SMR and PMR. In: Proceedings of the Third CIIT KEmsW, B..DONOFRIOD, .J.,and PAVKOV, K.L., The chronic Annual Conference:formaldehyde toxicity. Hemisphere Publish- of formaldehyde inhalation in rats and mice: a preliminary ing Corporation, New York (1983) (in press). npon.In: Proceedings of the Third C11T Annual Conference: WORLD HEALTH ORGANIZATION, Manual of the Internarional 'owkhYd toxicity. Hemisphere Publishing Corporation. New (1983) (in press). Statistical Classification of Disease, Injuries. and Causes of Death, 8th Revision. WHO, Geneva (1967).