Document LKDKn5EdLgvnxXwmB6r74Mj5

Corporate Health Physics Corporate Occupational Medicine Corporate Product Responsibility Corporate Toxicology 3M Medical Department 3M Center, 220-2E-02 PO Box 33220 St. Paul, MN 55133-3220 651 733 1110 Mortality Study of Employees at 3M Plant in Decatur, Alabama A retrospective cohort mortality study of employees, who worked at least one year (beginning in March, 1961) at the 3M chemical and film plants in Decatur Alabama, was conducted by University of Minnesota epidemiologists to determine whether the mortality experience of these employees was significantly different from that which would be expected. The cohorts consisted of 1,050 and 1,116 men who were ever employed in the chemical and film plants, respectively. Of these male employees, 485 and 547 were only employed in the chemical and film plants, respectively. A total of 318 female employees were identified (both plants combined). Vital status was searched through December 31, 1991 and determined for 99.7% of the cohort. Of the 74 deceased employees, death certificates were obtained for 72 (97.3%). Standardized Morality Ratios (SMRs) were calculated using U.S., Alabama and a regional Alabama counties for comparisons. There was no significantly elevated SMRs. Investigators recommended the study be updated in 1998 to cover an additional five years of mortality experience. 3M, in conjunction with epidemiologists from the University of Minnesota Division of Occupational and Environmental Health are in the midst of completing this updated retrospective cohort mortality study. Several methodologic improvements have occurred since the original study. Since the completion of the original mortality study which abstracted only the beginning and stop dates of employment from work history 001643 data, 3M has now computerized the entire work history records of employees at the Decatur location. The computerization of this detailed job and department information, in conjunction with information about serum fluorochemical levels acquired from medical surveillance exams and the random sample (described above), will allow for the construction of a potential perfluorooctanesulfonate (PFOS) exposure matrix. This matrix will be used to more thoroughly assess the mortality experience of Decatur employees as it relates to the workers' occupational fluorochemical levels. Estimated date of the final report for this updated retrospective cohort mortality study is November 1, 2000. 001644 FINAL REPORT M ortality Study o f Em ployees at 3M Plant in Decatur, Alabam a Submitted by: Jack S. Mandel, P h D ., MJPJL Rebecca A. Johnson, PhJD., M .PJL Division o f Environmental and Occupational Health School o f Public Health University o f Minnesota Box 807 UMHC 420 Delaware St. SJE. Minneapolis, MN 55455 March 13, 1995 00164S Table of Contents Easfi List of Tables................................................................................................................3 I n tr o d u c tio n ...........................................................................................................................6 Methods....................................................................................................................................... 6 D e fin itio n o f C ohort....................................................................................... 6 Enumeration o f Cohort and Ascertainm ent o fPertinent EmploymentInform ation......... 6 Follow-up o f C ohortand Determination o f V ital Status.................................................... 8 Obtaining and CotBng o fDeath Certificates......................................................................12 S electio n o f C om parison P opulations............................ 13 Study D atabase...................................................................................................................13 A n a lysis................................................................................................... 14 R esults......................................................... 16 Discussion................................................................ ..............................................................40 Recom m endation............................................................................................................... 41 R eferen ces.................................................................................................................................41 A p p en d ices................. 42 2 001646 List of Tables Table 1: Ease Description o f Ineligible Employees in Study Database (n 1,212).............................................................................................................. 8 Table 2: Description o f Final Employment Status and Vital Status for Men (n = 1,639).................................................................................................... 10 Table 3: Description o f Final Employment Status and Vital Status for Women (n - 3 1 8 )................................................................................................. 10 Table 4: Description o f Employees Lost to Follow-up (n = 6)........................................... 11 Table 5: Description o f State o f Residence at Time o f Death for D eceasul Employees for Whom a Death Certificate Was O b tain ed (n = 72)........................................................................................ 12 Table 6: Distribution o f Deaths by Year o f Death (n = 7 4 )................................................ 12 Table 7: C haracteristics o f Men (n = 1,639).............................................................16 Table 8: Characteristics o f Women (n = 318).....................................................................16 Table 9: Distribution o f Men by Age and Year o f Entry into F ollow -up (n = 1,639)............................................. 17 Table 10: Distribution of Women by Age and Year of Entry into Follow-up (n 318).................... 18 Table 11: Distribution o f Person-Years for Men by Age and Calendar Period (n = 1,639)................................................................................................19 3 001647 Table 12: Distribution of Person-Years for Women by Age and C alendar Period (n = 318).........................................................................20 Table 13: Distribution of Absolute Number at Risk by Calendar Period fo r M en (n = 1,639)................................................................................... 21 Table 14: Distribution of Absolute Number at Risk by Calendar Period for Women (n = 318)............................................................................................21 Table 15: Selected Cause-Specific SMRs for Men by Comparison Population (n =* 1,639)......................................................................................... 23 Table 16: Cause-Specific SMRs for Men U sing the U.S. as Comparison Population (n = 1,639).................................................................... 25 Table 17: Cause-Specific SMRs for Women Using the U.S. as C om parison Population (n 318)................................................................ 28 Table 18: Cause-Specific SMRs for Men Ever Employed in Chemical Departmcnt(s) Using die U.S. as Comparison Population (n = 1,050)............................................................................................................ 29 Table 19: Cause-Specific SMRs for Men Only Employed in Chemical Department(s) Using the U.S. as Comparison Population (n = 4 8 5 )............................................................................................................... 32 Table 20: Cause-Specific SMRs for Men Ever Employed in Film Department(s) Using the U.S. as Comparison Population (n = 1,116)............................................................................................................ 35 Table 21: Cause-Specific SMRs for Men Only Employed in Film Department(s) Using the U.S. as Comparison Population (n = 5 4 7 )............................................................................................................... 38 Table 22: Cause-Specific SMRs for Men Using Alabama as Comparison Population (n = 1,639)....................................................................42 4 001648 Table 23: Cause-Specific SMRs for Women Using Alabama as C om parison Population (n * 318).................................. .......................... 45 Table 24: Cause-Specific SMRs for Men Using Alabama Counties as Com parison Population (n = 1,639)................................................................... 46 Table 25: Cause-Specific SMRs for Women Using Alabama Counties as Comparison Population (n = 3 1 8 )..................................................................48 5 001649 M ortality Study o f Employees at 3M Plant in D ecatur, A labam a Introduction A retrospective cohort mortality study of employees who worked at least one year at the 3M Plant in Decatur, Alabama was conducted to determine whether the mortality experience o f these employees was significantly different from that which would be expected in a comparable population. The major elements o f the study were to: (1) define and completely enumerate the cohort; (2) ascertain pertinent employment information; (3) determine vital status as of December 31,1991; (4) obtain death certificates for deceased cohort members and code the underlying causes o f death; (5) select appropriate comparison populations; an! (6) compare the number o f deaths observed to the number of deaths expected, based on the mortality rates of the comparison populations. Methods D efinition o f Cohort The cohort comprised all employees who had worked at least one year at the 3M Plant in Decatur, Alabama and at least one day after March 1,1961. The cohort was followed from March 1,1961 through December 31,1991. March 1,1961 was chosen as the begin date because 3M reported that the plant in Decatur began operations in March o f 1961 and review o f employment records revealed that very few employees started work prior to March 1,1961. December 31,1991 was chosen as the end date for the current follow-up study because National Death Index reporting was complete through that date. Enumeration o f Cohort and Ascertainment o fPertinent EmploymentInform ation The 3M Plant at Decatur provided a computer file and computer-generated pages for 1,263 individuals actively employed anytime since 1986. The computer file only contained information about the most recentjob or employment status, whereas the computergenerated page contained the entire job history, hi addition, 3M provided photocopies of employment records for employees who were: (1) terminated during 1961 - 1971, (2) terminated since 1972, and (3) hired prior to 1986 and currendy working. 6 001650 To identify overlap among the record sources, a label was created for each employee in the computer file. If both a computer record and a photocopied employment record existed for an employee, the label was affixed to, and the computer-generated page was attached to, the photocopied employment record. Subsequently, new records were added to the study database for those employees who had photocopied employment records, but were not in the computer file. A total of 1,906 unique employees were added to the study database. For all employees in the study database, the relevant employment information was abstracted from the photocopied employment records and/or computer generated pages. In those instances where there were minor discrepancies in employment dates between the photocopied employment record and the computer-generated page, the dates on the photocopied employment record were assumed to be correct Lists were sent to 3M for those employees for whom employment information was missing, insufficient and/or inconsistent 3M alto was asked to provide missing demographic information (e.g., sex, date of birth, etc.) or information needed to determine the vital status o f the employee or, if the employee was deceased, to obtain a copy o f the death certificate. W here possible, 3M searched the Decatur Plant personnel records, corporate human resources computer databases, and microfilmed and microfiched corporate payroll records to provide additional information and resolve inconsistencies. However, despite 3M's efforts, a small number o f employees were still missing a start date and/or end date. The date of first employment and the date of last employment are required for determining whether the employee worked for at least one year during the time period from M arch 1,1961 through December 31,1991. For such employees, either the employee was contacted directly and supplied the date(s) or a knowledgeable employee at the 3M Plant in Decatur estimated the date(s). The study databases and employment records were updated using the information provided by the various sources. Any information which remained missing was coded as unknown. There were numerous quality control procedures to ensure the validity o f information in the study database. Information acquired from the computer-generated pages or photocopied employment records, provided by 3M or its employees, or obtained during vital status follow-up was keyed and verified. All sources of information were reviewed to ensure that all eligible employees had been entered in the study database. The study database was examined for duplicate records. When duplicate records were identified, the best information from both records was combined into the first record and the second record 7 001651 was marked as a duplicate. A random sample o f records was selected and key employment information was reabstracted to determine abstracting error rates and to identify any systematic errors which may have occurred. The data were checked in order to identify missing or inconsistent information and 3M was asked to follow-up on this information. A comprehensive error checking program was written to examine w hether (1) the individual values for key variables were acceptable, and (2) the values for key variables were consistent with one another. Errors identified by this program were resolved and the study database was updated. The study database included 3,181 records. O f these, 12 were duplicate records and 1,212 were ineligible employees. A description o f ineligible employees is shown below in Table 1. T ab le 1: D escription of Ineligible Em ployees in S tudy D atabase (n = 1,212) Reason for Ineligibility N W orked Less Than One Year 1,210 D id Not W ork at 3M Plant in Decatur 1 Not 3M Employee TOTAL 1 1,212 Thus, a total o f 1,957 eligible employees (1,639 males and 318 females) were included in the final cohort Follow-up o f C ohort and Determination o f V ital Status Far those employees for whom vital status was not known (including those reported as deceased, but for whom a death certificate had not been obtained), follow-up included computerized death searches via the: (1) National Death Index (NDI), (2) Equifax Death Search, and (3) TRW FACS+ Summary. In addition, some employees who terminated employment prior to 1979 were actively traced through driver's license records, relatives, neighbors, and/or coworkers. A computer file o f employees for whom vital status was not known was submitted to NDI for matching against information for the years 1979 -1991. 8 001652 Equifax has information about deaths reported to the Social Security Administration (SSA) and from sources such as the military, corporate pension plans, insurance companies, etc. A computer file o f employees for whom vital status was not known was submitted to Equifax. TRW FACS+ Summary contains information about deaths and is maintained by the Credit Reporting Division o f TRW. The majority o f death information contained in TRW also is provided by the SSA. TRW was used to search for deaths among employees for whom vital status was not known. Currently, TRW typically provides the name, date of birth, and date of death for individuals who have been reported deceased. Throughout most of the study, however, TRW only provided information that there had been a reported death associated with the social security number, without providing additional identifying information or dead: information. Because all o f die ctcnputerized death searches were dependent upon having the correct social security num ber for an employee, efforts were made to ensure the accuracy o f social security numbers. In addition to comparing social security numbers from the various sources o f information (i.c., die computer file, employment records, etc.), TRW Social Search was used to verify the accuracy o f social security numbers for employees. Some social security numbers were not found in TRW. For these employees, the social security numbers were assumed to be correct because there are a number of valid reasons why a person may not be included in TRW (e.g., a person may ask to be removed from the TRW file or a person may not have any credit activity). Additional tracing was done to resolve discrepancies in social security numbers. Active Gracing o f employees was performed for all employees who terminated prior to 1979 (when NDI was implemented) and for whom vital status was not previously known. Driver's license records frequendy were used to determine vital status. In addition, interactive software was used to access Metronet, a national consumer database, and addresses and/or telephone numbers of relatives, neighbors, and/or form er employees were obtained. Other tracing sources included other 3M employees and law enforcement agencies. Once a tracing lead was identified, telephone tracers followed up to determine, directly or indirectly, the vital status o f the employee. The results o f the vital status follow-up, along with the employment status as o f the study end date, are shown in Tables 2 and 3 for males and females, respectively. 9 001653 T able 2: D escription of Final Em ploym ent S tatus and V ital S tatus for M en (n = 1,639) Final Employment Status and Vital Status N % Currently Employed Retired Alive Dead Unknown Terminated Alive 810 59 10 0 694 4 9 .4 3.6 0 .6 0 .0 42.3 Dead 29 Unknown 6 Died While Employed 31 TOTAL 1,639 * percentages may docadd to 100 doe to rounding 1.8 0 .4 1.9 100.0* Table 3: Description of Final Employment Status and Vital Status _______ _______for Women (n s 318)_____________________ Final Employment Status and Vital Status N % Currently Employed 141 44.3 Retired Alive 9 Dead 0 Unknown 0 Tenninated Alive 164 Dead 1 Unknown 0 Died While Employed 3 TOTAL 318 * percentages may not add to 100 doe to rounding 2.8 0 .0 0 .0 51.6 0.3 0 .0 0.9 100.0* 10 001654 Six (0.3%) employees were lost to follow-up. For these employees, the last known date that the employee was alive is the date of termination of employment A description of the employees who were lost to follow-up is shown in Table 4. Table 4: Study ED 20067 20958 i 21028 21057 21184 21639 Description of Employees Lost to Follow-up (n = 6) M onth/Yearof M onth/Yearof Sex Birth Employment Job Title(s) male 08/49 08/73 - 10/75 general helper, process helper, material handler, resin handler male 12/37 02/65 - 02/71 general helper, process helper, process operator, chemical operator, forem an mal 06/41 01/66-07/68 general helper, process helper, process operator, chemical operator male 01/46 04/66 - 07/67 general helper, process helper, process operator male 11/46 08/66 - 03/69 general helper, process operator malft 08/45 01/69 - 09/77 general helper, process helper, packer, material handler helper, material handler, process operator, w arehousem an, order filler, chemical operator 11 001655 Obtaining and Coding o fDeath Certificates Death certificates were requested from state vital statistics offices. If a death certificate could not be obtained or the death certificate obtained was not for the correct person, additional tracing was implemented. Overall, 72 (97.3%) death certificates were obtained for the 74 deceased employees (70 men and 4 women). A description o f the state o f residence at time o f death for the 72 deceased employees for whom a death certificate was obtained is provided in Table 5. Ninety percent died in Alabama. T able 5: D escription o f S tate o f Residence a t Tim e o f D eath fo r D eceased Em ployees for W hom a D eath C ertificate W as O btained (n 72) State N Alabama R o rid a 65 1 G eorgia Michigan M innesota Ohio 1 1 1 1 Tennessee 1 Texas TOTAL 1 72 The distribution o f deaths by calendar year is shown in Table 6. Table 6: Distribution of Deaths by Year of Death (n = 74) Year of Death N 1967 -1969 3 1970 -1974 9 1975 - 1979 9 1980 - 1984 12 1985 - 89 33 1990-1991 8 TOTAL 74 12 001656 Death certificates were coded by a professional nosologist according to the Eighth Revision of the International Classification of Diseases (ICD 8). The underlying cause of death and any mention of cancer were coded. Selection o f Comparison Populations Three comparison populations were selected: (1) the U.S., (2) the state o f Alabama, and (3) Alabama counties where more than one-half of the county was within 100 miles of Decatur, excluding those counties in which there was a city with greater than 100,000 persons. Counties included in the regional comparison were: (1) Blount, (2) Calhoun, (3) Cherokee, (4) Colbert, (5) Cullman, (6) De Kalb, (7) Etowah, (8) Fayette, (9) Franklin, (10) Jackson, (11) Lamar, (12) Lauderdale, (13) Lawrence, (14) Lim estone, (IS ) Marion, (16) Marshall, (17) Morgan, (18) S t Clair, (19) Shelby, (20) Talladega, (21) Tuscaloosa, (22) W alker, and (23) Winston. Based upon 1990 census figures, the combined population for these counties was 1,387,615 persons. Counties excluded due to a population center o f 100,000 persons or more were: (1) M idison (large city is Huntsville), and (2) Jefferson (large city is Birmingham). Race-, sex-, age-, and time-specific mortality rates, adjusted to ICD &, were obtained for the three comparison populations from die Mortality and Population D ata System (MPDS) at the University o f Pittsburgh. No mortality rates ar available from the MPDS prior to 1950 and only cancer mortality rates are available prior to 1962. The m ost current mortality rates available are for 1989. Study Database The final study database was a combination of data from the following sources: the computer file, computer-generated pages, photocopied employment records, 3M, employees, tracing/Vital status follow-up, and death certificates. The following key information was available in the study database: study ID; first name, middle name, and last name; street address, city, state, and zip code; telephone number; social security number; sex; date of birth; 13 001657  date of first employment at 3M Plant in Decatur, date of last employment at 3M Plant in Decatur, number of lapsed months between date of first employment and date of last employment when not actually working at 3M Plant in Decatur, date first employed in chemical department(s); date last employed in chemical department(s); number o f months worked in chemical department(s) (accounting for any lapses when not actually working in chemical department(s)); flags indicating work in specific department(s) (Le., worked in chemical department(s), worked in film department(s), etc.); final employment status (i.e., currently working, retired, otherwise term inated from employment, died while employed); vital status (Le., alive, presumed alive, deceased, or unknown); date of death o r date last known to be alive (if different from last date o f employment or date o f termination); underlying cause o f death coded to ICD 8; and ucontributing causes o f death or other significant conditions (if cancer) coded to ICD 8. Ih e necessary data files for analysis were prepared using the key information contained in the study database. As part of this process, default values were automatically substituted for any missing components o f employment dates (i.e., 06 for a missing month and 15 for a missing day). If this automatic substitution for missing dates created a date that was chronologically out o f sequence, a date was manually estimated and this estimated date was entered in the study database. A nalysis The method o f analysis was based upon a comparison of the observed to expected numbers o f deaths specific for the cause of death, race, sex, age, and time. The estimate of the expected number o f deaths was calculated by applying cause-, race-, sex-, -- -- age-, and---------------- - tim e- ^ specific rates for the comparison population to the person-years at risk. Follow-up began March 1,1961. Cohort members did not contribute person-years until they had m et the minimum length o f work criterion (i.e., one year or more). Person-years o f follow-up were contributed until death, los^to follow-up, or the end o f the study (December 31, 1991). No information on race was available; thus, employees were assumed to be white. Deceased study members for whom a death certificate could not be obtained only were included in the "all causes of death" and "unknown cause o f death" categories. 14 001658 Standardized mortality ratios (SMRs) were computed using the Occupational Cohort Mortality Analysis Program (OCMAP).1 Mortality rates for whites were used to calculate the expected numbers for all men and all women. For 1961, when mortality rates for noncancer causes were unavailable, the expected numbers of deaths were not calculated (i.e., the expected numbers of deaths were set to zero). There were no observed deaths prior to 1967. M ortality rates for 1989 were used to estimate mortality rates for subsequent years. SMRs were calculated using both OCMAP and the United States Death Rates (USDR) program developed by Monson.2 Results were virtually identical. The overall pattern o f mortality was examined for men and women. For men, the pattern o f mortality by department(s) worked also was examined. For this analysis, the following subgroups o f employees were defined: (1) ever employed in the chemical departm ent^); (2) only employed in die chemical department(s); (3) ever employed in the film department(s); and (4) only employed in the film department(s). Department codes at the 3M Decatur Plant were defined by four-digit codes. The first two digits were '59* for film departments and were 7 6 ' for chemical departments. For 11 employees, at least one department was unknown; and, for 3 of these, all departments were unknown. Thus, these 3 employees could not be included in the subgroup analysis by department(s) worked. 15 0 0 1 6 5 9 RfiSUlia Characteristics of male and female cohort members are given in Tables 7 and 8, respectively. Men contributed 33,108 person-years o f follow-up and women contributed 4,807 person-years. There were 70 (95%) deaths among men and 4 (5%) among women. On average, men entered the cohort at an earlier point in time (1971 versus 1977) and at a slightly younger age than women (25 versus 26). T ab le 7: C h aracteristics o f M en (n = 1,639) Variable Value Number o f Employees 1,639 Num ber o f Person-Years 33,108 Number of Deaths 70 Average Age Started Work 25 Average Year o f Entry 1971 Average Age at Death 47 Average Year o f Death 1984 T able 8: C haracteristics o f W om en (n s 318) Variable Value Number o f Employees 318 Num ber o f Person-Years 4,807 Number of Deaths 4 Average Age Started Work 26 Average Year o f Entry 1977 Average Age at Death 28 Average Year of Death 1980 16 001660 In Table 9, the distribution o f men by age and year of entry into follow-up is shown. Table 10 shows the same distribution for women. Most men entered the cohort during the late 1960s and early 1970s and between the ages o f 20 and 30, whereas m ost women entered the cohort at sim ilar ages, but between 1970 and 1980. Table 9: D istribution of Men by A ge a n d Y ear o f E n try in to Follow -up (n = 1,639) Year of Entry into Follow-up 1960 1965 1970 1975 1980 1985 1990 Age <20 2 0 -2 4 2 5 -2 9 3 0 -3 4 3 5 -3 9 40- 44 4 5 -4 9 5 0 -5 4 5 5 -5 9 60+ TOTAL 1964 18 114 64 21 22 19 3 0 0 0 261 1969 34 238 139 58 25 18 3 1 1 0 517 1974 86 301 65 33 9 4 4 0 0 0 502 1979 31 96 37 19 15 10 5 0 0 0 213 1984 4 28 21 8 6 8 0 2 0 0 77 1989 2 18 14 10 5 2 2 3 0 0 56 1991 TOTAL 0 175 3 798 5 345 4 153 1 83 0 61 0 17 06 01 00 13 1,639 17 001661 Table 10: Distribution of Women by Age and Year of Entry into Follow-up (n = 318) Year of Entry into Follow-up 1960 1965 1970 1975 1980 1985 1990 Age <20 2 0 -2 4 2 5 -2 9 3 0 -3 4 3 5 -3 9 4 0 -4 4 4 5 -4 9 5 0 -5 4 5 5 -5 9 60+ TOTAL 1964 2 4 3 2 1 0 0 0 0 0 12 1969 12 22 4 3 2 0 0 0 0 0 43 1974 18 23 4 7. 3 0 0 0 0 0 50 1979 16 50 47 15 16 8 5 3 1 0 161 1984 4 4 4 1 3 3 1 0 0 0 20 1989 0 9 5 4 3 2 1 0 0 0 24 1991 TOTAL 0 52 1 113 3 70 2 29 2 30 0 13 07 03 01 00 8 318 18 001662 Tables 11 and 12 provide the distribution of person-years by age and calendar period for men and women, respectively. Table 11: Distribution of Person-Years for Men by Age and Calendar Period (n = 1,639) ___________________ Calendar Period_____________ 1960 1965 1970 1975 1980 1985 1990 - Age 1964 1969 1974 1979 1984 1989 1991 TOTAL* <20 9 19 43 23 2 1 0 96 2 0 -2 4 182 583 985 930 244 64 21 3,008 2 5 -2 9 124 861 1,530 1,633 1,171 341 66 5,724 3 0 -3 4 60 463 1,254 1,672 1,728 1,228 208 6,613 3 5 -3 9 50 188 609 1,339 1,716 1,760 636 6,297 4 0 -4 4 35 129 272 645 1,390 1,719 738 4,927 4 5 -4 9 21 91 180 295 665 1,387 630 3,270 5 0 -5 4 0 28 103 184 308 650 456 1,728 55-59 0 5 27 101 185 307 204 827 6 0 -6 4 0 0 5 27 101 174 96 402 6 5 -6 9 0 0 0 5 24 86 47 161 7 0 -7 4 0 0 0 0 5 18 25 47 75-79 0 0 0 0 0 5 1 5 80-84 0 0 0 0 0 0 2 2 85+ 0 0 0 0 0 0 0 0 TOTAL* 479 2,365 5,006 6,853 7,536 7,739 3,129 33,108 totals may not be equal to sum of individual rows or columns due to rounding 19 001663 Table 12: Distribution of Person-Years for Women by Age and Calendar Period (n = 318) ___________________ Calendar Period_______________ 1960 1965 1970 1975 1980 1985 1990 Age 1964 1969 1974 1979 1984 1989 1991 TOTAL* <20 1 9 20 21 12 0 0 63 2 0 -2 4 9 65 138 182 139 38 5 576 2 5 -2 9 7 39 115 253 337 171 34 955 3 0 -3 4 2 19 55 171 376 346 106 1,074 3 5 -3 9 2 14 26 81 235 383 160 900 40-44 2 5 17 48 136 243 143 594 4 5 -4 9 0 2 5 24 70 141 91 332 5 0 -5 4 0 0 2 8 49 70 43 172 55 - 59 0 0 0 3 18 49 19 89 6 0 -6 4 0 0 0 0 5 18 18 41 65-69 0 0 0 0 0 5 7 12 70-74 0 0 0 0 0 0 0 0 75-79 0 0 0 0 0 0 0 0 80-84 0 0 0 0 0 0 0 0 85+ 0 0 0 0 0 0 0 0 TOTAL* 22 153 377 790 1,377 1,463 624 4,807 * totals may not be equal to sum of individual rows or columns due to rounding 20 0 0 1 6 6 4 The distribution o f the absolute number of persons at risk by calendar period is provided in Table 13 for men and in Table 14 for women. T able 13: D istribution of A bsolute N um ber a t Risk by C alen d ar P eriod for M en (n = 1,639) Calendar Period Number at Risk 1960 -1964 261 1965 - 1969 1970 -1974 778 1,274 1975-1979 1,477 1980-1984 - 1,545 1985 -1989 1990-1991 c1L,5s9z1D TOTAL 1,639 T able 14: D istrib u tio n o f A bsolute N um ber a t R isk by C alen d ar P erio d fo r W omen (n = 318) Calendar Period Number at Risk 1960-1964 1965 -1969 1970 -1974 1975 - 1979 1980 - 1984 12 55 105 266 284 1985 -1989 1990-1991 TOTAL 306 314 318 In the tables which follow, the observed number o f deaths, the expected number of deaths, the observed to expected ratio (i.e., the SMR), and the 95% confidence interval (95% Cl) are shown for specific causes o f death. The observed number o f deaths for a specific cause is the number o f deaths from that cause that occurred during die study time period. The expected number o f deaths for a specific cause is the number o f deaths that would have occurred if the employees experienced the same cause-specific mortality rates as observed in the comparison population. The expected number o f deaths was calculated by applying the race-, sex-, age-, and time period-specific rates for the comparison population to the 001665 21 number o f person-years of follow-up contributed by the cohort members. The observed number of deaths for a specific cause was divided by the expected number o f deaths for that cause and multiplied by 100 to obtain the cause-specific SMR. An SMR may be interpreted as follows: (1) an SMR of 100 means that the observed and expected number o f deaths were equal, (2) an SMR less than 100 means that fewer deaths occurred than expected, and (3) an SMR greater than 100 means more deaths occurred than expected. The 95% Cl for the SMR provides an indication of the variability and statistical significance of the estimate o f the SMR. The 95% Q means that there is a 95% chance that the interval includes the true value o f the SMR. If the 95% Cl does not include 100, the SMR is considered statistically significant 22 0 0 1 6 6 6 In Table 15, SMRs for men are presented for selected cause of death categories for each of the three white comparison populations. T ab le 15: Selected C ause-Specific SM Rs fo r M en by C om parison P o p u latio n (n = 1,639) SMR (95% Confidence Interval) Comparison Population Cause o f Death Observed Deaths U .S. Alabama Alabama Counties All Causes o f Death 70 All Malignant Neoplasms 15 Cerebrovascular Disease 1 62.9 (49.0, 79.5) 68.4 (38.3, 112.8) 33.5 54.8 (42.7, 69.3) 60.9 (34.1, 100.4) 27.8 52.0 (40.5, 65.7) 59.9 (33.5, 98.8) 26.5 All Heart Disease Respiratory Disease (0.8, 186.8) (0.7, 155.2) (0.7, 147.5) 17 59.1 53.2 49.1 (34.4, 94.7) (31.0, 85.1) (28.6, 78.6) 0-- -- -- External Causes (0.0, 91.2) (0.0, 80.2) (0.0, 73.9) 25 74.5 59.0 55.0 (48.2, 110.0) (38.2, 87.1) (35.6, 81.3) SMRs based upon the Alabama and Alabama counties comparison populations were similar to, but lower than, those based upon the U.S. comparison population. Since the U.S death rates are more stable than the death rates for the state and regional comparison populations and because the U.S. commonly is used as a comparison population in occupational cohort mortality studies, SMRs based upon the U.S. comparison population will be presented in the results which follow. SMRs for men and women based upon the Alabama and Alabama counties comparison populations are provided in Tables 22 through 25 in the appendices. 23 0 0 1 6 6 7 There were 70 deaths among the men in this cohort Using the U.S. as the comparison population, SMRs for men are shown in Table 16. The SMRs for all causes o f death, heart disease, and respiratory disease (i.e., nonmalignant respiratory disease) were significantly less than 100. The SMR for all causes o f death was 62.9 (95% C l = 49.0,79.5) and the SMR for heart disease was 59.1 (95% C l = 34.4,94.7). No respiratory disease deaths occurred, although 4 deaths were expected. Other deficits occurred for all cancer and eternal causes of death. The SMR for all cancers was 68.4 (95% C l * 38.3,112.8). Most o f the specific cancer SMRs were less than 100, except for cancer o f the bladder and other urinary organs and cancer o f other lymphatic and hematopoietic tissue. For these two causes, the SMRs were based on only one or two deaths and were not statistically significan t F or external causes o f death, the SMR was 74.5 (95% C l - 48.2,110.0). Less than the cr.pected number o f deaths occurred for motor vehicle accidents, suicides, and homicides and other external causes o f death. 24 001668 T able 16: C ause-Specific SM Rs for Men U sing th e U.S. as C om parison P o pulation (n = 1,639) Cause o f Death Observed Expected Deaths Deaths SMR 95% Confidence Interval All Causes o f Death All Malignant Neoplasms Cancer of Buccal Cavity and Pharynx Cancer of Digestive Organs and Peritoneum Cancer o f Esophagus Cancer o f Stomach Cancer o f Large Intestine Cancer o f Rectum Cancer of liv e r and Biliary 70 15 0 1 0 0 1 0 0 111.3 21.9 0.6 4.7 0.5 0.7 1.7 0.4 0 .4 62.9 68.4 21.5 -- -- 60.2 -- --* (49.0, 79.5) (38.3, 112.8) (0.0, 628.2) (0.5, 120.0) (0.0, 722.6) (0.0, 531.8) (1.5, 335.3) (0.0, 1,003.1) (0.0, 919.1) Passages Cancer o f Pancreas Cancer o f All Other Digestive Organs Cancer of Respiratory System Cancer o f Larynx Cancer of Bronchus, Trachea, and Lung Cancer of Other Respiratory Organs Cancer of Prostate Cancer o f Testes and Other Male Genital Organs Cancer of Kidney Cancer o f Bladder and Other Urinary Organs 0 0 7 0 7 0 0 0 0 1 1.0 -- 0.2 -- . 7.6 91.8 0.2 -- 7.3 96.4 0.1 -- 0.5 -- 0.3 -- 0.6 -- 0.3 341.3 (0.0, 379.1) (0.0, 2.442.9) (36.9, 189.2) (0.0, 1,529.7) (38.7, 198.6) (0.0, 3,109.0) (0.0, 694.6) (0.0, 1,123.9) (0.0, 592.5) (8.5, 1,901.6) 25 001669 T able 16: (C ontinued) Cause of Death Cancer of Brain and Other Central Nervous System Cancer of Thyroid and Other Endocrine Glands Cancer of Bone Cancer o f AJ! Lymphatic and Hematopoietic Tissue Lymphoma Hodgkin's Disease Leukemia and Aleukemia Cancer o f O ther Lymphatic and Hematopoietic Tissue Malignant Melanoma of Skin All Other Malignant N eoplasm s Diabetes Mellitus Cerebrovascular Disease All Heart Disease Hypertension Respiratory Disease Ulcer of Stomach and Duodenum Cirrhosis of Liver Nephritis and Nephrosis External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Observed Expected Deaths Deaths 1 1.2 0 0.1 0 0.1 3 3.0 0 0.4 0 0.5 1 1.2 2 1.0 1 0.9 1 1.9 0 1.5 1 3.0 17 28.8 0 0.1 0 4.0 0 0.3 3 4.1 0 0.4 25 33.5 19 20.6 9 11.8 10 8.8 4 7.9 SMR 84.3 ~ -- 101.1 ---- 86.4 203.8 116.8 52.2 -- 33.5 59.1 -- -- -- 73.7 -- 74.5 92.4 76.1 113.6 50.8 95% Confidence Interval (2.1, 470.0) (0.0, 3,749.2) (0.0, 3,035.4) (20.9, 295.5) (0.0, 997.3) (0.0, 804.7) (2.2, 481.4) (24.7, 736.2) (2.9, 650.6) (1.3, 290.7) (0.0, 240.2) (0.8, 186.8) (34.4, 94.7) (0.0, 2,652.5) (0.0, 91.2) (0.0, 1,407.9) (15.2, 215.3) (0.0, 999.6) (48.2, 110.0) (55.6, 144.2) (34.8, 144.5) (54.5, 208.9) (13.9, 130.1) 001670 26 Table 16: (C ontinued) Observed Expected Cause of Death Deaths Deaths SMR Homicides and Other External 2 4.9 40.7 Causes Residual Causes* 7 12.8 54.5 Unknown Causest 2 * all othercauses of death combined t no death certificate obtained; includedonlyin all causes ofdeath category 95% Confidence Interval (4.9, 147.0) (21.9, 112.3) In Table 17, SM Rs are provided for women, usfcg die U.S. as die comparison population. No cause o f death had an SMR significantly different from 100. There were 4 deaths among female employees and the all cause SMB. was 62.6 (95% C l = 17.1,160.4). O f the 4 deaths which occurred, 3 deaths were from external causes (SMR = 213.4; 95% C l = 44.0,623.6). By examining the death certificates for the specific causes o f death and circumstances surrounding the deaths, it was determined that the excess mortality among females due to external causes was not work-related. One of the external causes o f death was due to arsenic poisoning and die other two external causes o f death were due to m otor vehicle accidents. 27 001671 Table 17: Cause*Specific SMRs for Women Using the U.S. as Comparison Population (n = 318) Cause of Death All Causes o f Death All Malignant Neoplasms Cerebrovascular Disease All Heart Disease Respiratory Disease External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Homicides and Other External Causes Observed Expected Deaths Deaths 4 6.4 0 2.3 1 0.3 0 0.9 0 0.3 3 1.4 2 0.8 2 0.6 0 0.3 0 0.4 1 0.2 SMR 62.6 -- 376.0 -- -- 213.4 247.3 364.7 -- -- 449.6 95% Confidence Interval (17.1, 160.4) (0.0, 162.4) (9.4, 2,095.3) (0.0, 413.1) (0.0, 1,344.7) (44.0, 623.6) (29.9, 893.4) (44.1, 1,717.6) (0.0, 1,415.3) (0.0, 992.2) (11.2, 2,505.4) Tables 18 through 21 present SMRs and 95% CIs for men by the departm ents) worked subgroups. Findings were similar for men ever employed in chemical departm ents) (Table 18). A total o f 57 deaths were observed for these men and the all cause SMR was significantly decreased (SMR = 70.0; 95% C l = 53.0,90.6). The SMR for heart disease also was significantly less than 100 (SMR = 48.8; 95% C l = 24.4,87.4). SMRs and 95% CIs for all cancer and external causes o f death were 76.9 (95% Cl = 40.9,131.5) and 90.2 (95% C l =55.1, 139.3), respectively. 28 001672 Table 18: Cause-Specific SMRs for Men Ever Employed in Chemical Department(s) Using the U.S. as Comparison Population (n = 1,050) Cause of Death All Causes of Death All Malignant Neoplasms Cancer o f Buccal Cavity and Pharynx Cancer o f Digestive Organs and Peritoneum Cancer of Esophagus Cancer of Stomach Cancer o f Large Intestine Cancer o f Rectum Cancer o f Liver and Biliary Observed Expected Deaths Deaths (57) 81.5 13 16.9 0 0.5 1 3.6 0 0.4 0 0.5 1 1.3 0 0.3 0 0.3 SMR 70.0 76.9 -- 27.5 -- -- 76.9 -- -- 95% Confidence Interval (53.0, 90.6) (40.9, 131.5) (0.0, 800.6) (0.7, 153.2) (0.0, 913.3) (0.0, 686.1) (1.9, 428.5) (0.0, 1,271.9) (0.0, 1,188.2) Passages Cancer of Pancreas Cancer o f All Other 0 0.8 -- (0.0, 480.6) 0 0.1 -- (0.0, 3,203.2) Digestive Organs Cancer of Respiratory System Cancer o f Larynx Cancer o f Bronchus, Trachea, and Lung Cancer of Other Respiratory Organs Cancer o f Prostate Cancer o f Testes and Other Male Genital Organs Cancer of Kidney Cancer o f Bladder and Other Urinary Organs 7 0 7 0 0 0 0 1 6.1 115.1 (46.3, 237.1) 0.2 -- (0.0, 1,891.9) 5.8 120.7 (48.5, 248.7) 0.1 -- (0.0, 4,214.1) 0.5 -- (0.0, 805.9) 0.2 (0.0, 1,678.5) 0.5 -- (0.0, 768.2) 0.2 415.5 (10.4, 2,315.3) 29 001673 T able 18: (C ontinued) Cause of Death Cancer of Brain and Other Central Nervous System Cancer of Thyroid and Other Endocrine Glands Cancer of Bone Cancer o f All Lymphatic and Hematopoietic Tissue Lymphoma Hodgkin's Disease T.ftiiifwmia and Aleukemia Cancer o f Other Lymphatic and Hematopoietic Tissue Malignant Melanoma o f Skin All Other Malignant N eoplasm s Diabetes Mellitus Cerebrovascular Disease All Heart Disease Hypertension Respiratory Disease Ulcer o f Stomach and Duodenum Cirrhosis of Liver Nephritis and Nephrosis External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Observed Expected Deaths Deaths 1 0.9 0 0.1 0 0.1 2 2.2 0 0.3 0 0.3 1 0.8 1 0.7 0 0.6 1 1.5 0 1.1 1 2.3 11 22.5 0 0.1 0 3.2 0 0.2 3 3.0 0 0.3 20 22.2 16 13.6 8 7.8 8 5.9 3 5.2 SMR 117.2 -- -- 92.9 -- -- ' A20.0 137.2 -- 68.9 -- 43.4 48.8 -- -- -- 100.2 -- 90.2 117.3 102.8 135.7 57.7 95% Confidence Interval (2.9, 653.0) (0.0, 5,113.2) (0.0, 4,368.3) (11.2, 335.5) (0.0, 1,341.4) (0.0, 1,163.3) (3.0, 668.8) (3.4, 764.5) (0.0, 614.7) (1.7, 383.7) (0.0, 323.9) (1.1, 241.7) (24.4, 87.4) (0.0, 3,389.9) (0.0, 116.3) (0.0, 1,796.3) (20.7, 292.7) (0.0, 1,308.2) (55.1, 139.3) (67.0, 190.5) (44.4, 202.5) (58.6, 267.4) (11.9, 168.6) 30 001674 T able 18: (C ontinued) Observed Expected Cause of Death Deaths Deaths SMR Homicides and Other External 1 3.2 31.4 Causes Residual Causes* 7 9.0 77.6 Unknown Causest 2 * all othercausesof death combined t no rfeadi certificate obtained; included only in all causes o f death category 95% Confidence Interval (0.8, 175.1) (31.2, 159.9) Thfcty-two deaths occurred among men only employed in chemical departm ent^) (Table 19). The SMR for all causes o f death was 7 2 5 and was not statistically significant (95% Q =- 4 9 .6 ,1 0 2 4 ). For heart disease, the SMR was 53.4 (95% C l * 21.5, 110.1). The all cancer SMR was 93.6 (95% Cl =*42.8,177.8). The observed and expected numbers of deaths due to external causes were approximately the same, resulting in an SMR o f 95.0 (95% a = 45.6, 174.7). 31 001675 Table 19: Cause-Specific SMRs for Men Only Employed in Chemical Departments) Using the U.S. as Comparison Population (n = 485) 95% Observed Expected Confidence Cause of Death Deaths Deaths SMR Interval All Causes of Death All Malignant Neoplasms (3 ) 9 44.1 72.5 9.6 93.6 (49.6, 102.4) (42.8, 177.8) Cancer o f Buccal Cavity and 0 0.3 -- (0.0, 1,396.4) Pharynx Cancer o f Digestive 0 2.1 mmum (0.0, 175.2) Organs and Peritoneum Cancer of Esophagus 0 0.2 _n (0.0, 1,577.7) Cancer o f Stomach 0 0.3 -- (0.0, 1,195.2) Cancer o f Large Intestine Cancer o f Rectum 0 0 0.8 -- (0.0, 488.5) 0.2 -- (0.0, 2,175.6) Cancer o f Liver and Biliary 0 0.2 -- (0.0, 2,072.6) Passages Cancer of Pancreas Cancer o f All Other 0 0.5 0 0.1 -- (0.0, 827.2) (0.0, 5,692.4) Digestive Organs Cancer of Respiratory System Cancer o f Larynx 5 0 3.6 141.0 (45.8, 329.0) 0.1 -- (0.0, 3,198.4) Cancer o f Bronchus, 5 3.4 147.8 (48.0, 344.9) Trachea, and Lung Cancer o f Other 0 0.1 _ (0.0, 7,763.4) Respiratory Organs Cancer of Prostate 0 0.3 Cancer of Testes and Other 0 0.1 -- (0.0, 1,202.0) (0.0, 3,501.4) Male Genital Organs Cancer of Kidney Cancer o f Bladder and Other 0 1 0.3 __ (0.0, 1,365.0) 0.2 669.9 (16.7, 3,732.7) Urinary Organs 32 001676 Table 19: (C ontinued) Cause o f Death Cancer o f Brain and Other Central Nervous System Cancer of Thyroid and Other Endocrine Glands Cancer of Bone Cancer o f All Lymphatic and Hematopoietic Tissue Lymphoma Hodgkin's Disease Leukemia and Aleukemia Cancer o f Other Lymphatic and Hematopoietic Tissue Malignant Melanoma o f Skin All Other Malignant N eoplasm s Diabetes MelHtus Cerebrovascular Disease All Heart Disease Hypertension Respiratory Disease Ulcer o f Stomach and Duodenum Cirrhosis of Liver Nephritis and Nephrosis External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Observed Expected Deaths Deaths 1 0.5 0 0.0 0 0.0 2 1.2 0 0.2 0 0.2 1 0.4 1 0.4 0 0.3 0 0.8 0 0.6 0 1.3 7 13.1 0 0.1 0 1.9 0 0.1 1 1.6 0 0.2 10 10.5 9 6.5 5 3.6 4 2.9 1 2.5 SMR 222.6 -- -- 174.0 -- -- 226.3 249.5 -- 1 -- -- 53.4 -- -- 62.3 -- 95.0 139.4 138.0 140.2 40.0 95% Confidence Interval (5.6, 1,240.3) (0.0, 9,544.3) (0.0, 8,697.5) (21.1, 628.7) (0.0, 2,466.1) (0.0, 2,346.1) (5.7, 1,260.7) (6.2, 1,390.1) (0.0, 1,200.4) (0.0, 455.3) (0.0, 589.0) (0.0, 274.4) (21.5, 110.1) (0.0, 5,834.4) (0.0, 195.3) (0.0, 3,083.3) (1.6, 347.3) (0.0, 2,346.9) (45.6, 174.7) (63.7, 264.5) (44.8, 322.0) (38.2, 359.0) (1.0, 223.1) 33 0 0 1 6 7 7 Table 19; (C ontinued) Observed Expected Cause of Death Deaths Deaths SMR Homicides and Other External 0 1.5 -- (pauses Residual Causes* 3 4.7 63.4 Unknown Causes* *alMl ochercausesof deat.h com.bi.ne.d 2 *no death certificate obtained; includedonlyin all causes ofdeath category 95% Confidence Interval (0.0, 247.4) (13.1, 185.3) There were 37 deaths among men ever employed in film departm ents) (Table 20). A statistically significant deficit was observed fo r all causes o f death combined (SMR = 58.6; 95% C l - 41.3,80.8). Although nonsignificant, deficits also were seen for heart disease (SMR = 69.9; 95% O = 33J , 128.5), all cancer (SMR = 52.9; 95% Q = 19.4,115.2), ar.d external causes o f death (SMR = 62.9; 95% C l = 34.4,105.5). 001678 34 T able 20: C ause-Specific SM Rs for Men E ver Employe? in Film D e p a rtm e n ts) U sing th e U.S. as C om parison P o pulation (n = 1,116) 95% Cause of Death All Causes of Death All Malignant Neoplasms Observed Expected Deaths Deaths ( y p 63.1 6 11.3 SMR x/ 58.6 52.9 Confidence Interval (41.3, 80.8) (19.4, 115.2) Cancer o f Buccal Cavity and 0 0.3 -- (0.0, 1,248.9) Pharynx Cancer o f Digestive Organs and Peritoneum 1 2.3 43.1 (1.1,2 4 0 .0 ) Cancer of Esophagus 0 0.3 (0.0, 1,465.1) Cancer of Stomach 0 0.4 -- (0.0, 1,043.7) Cancer o f Large Intestine 1 0.8 120.6 (3.0, 672.1) Cancer o f Rectum Cancer o f liv e r and Biliary 0 0 0.2 -- 0.2 -- (0.0, 2,040.6) (0.0, 1,798.6) Passages Cancer o f Pancreas Cancer of All Other 0 0.5 MW (0.0, 767.6) 0 0.1 -- (0.0, 4,625.3) Digestive Organs Cancer o f Respiratory System 2 3.7 54.1 (6.5, 195.4) Cancer of Larynx 0 0.1 -- (0.0, 3,253.4) Cancer of Bronchus, 2 3.5 56.9 (6.9, 205.4) Trachea, and Lung Cancer of Other 0 0.1 _ (0.0, 5,537.8) Respiratory Organs Cancer of Prostate 0 0.2 Cancer o f Testes and Other 0 0.2 -- (0.0, 1,915.6) (0.0, 1,711.8) Male Genital Organs Cancer o f Kidney 0 0.3 WWW (0.0, 1,137.8) Cancer o f Bladder and Other 0 0.1 MW (0.0, 2,877.2) Urinary Organs 001679 35 Table 20: (C ontinued) Cause of Death Cancer of Brain and Other Central Nervous System Cancer of Thyroid and Other Endocrine Glands Cancer of Bone Cancer o f All Lymphatic and Hematopoietic Tissue Lymphoma Hodgkin's D isease Leukemia and Aleukemia Cancer o f O ther Lymphatic and Hematopoietic Tissue Malignant M elanoma o f Skin All Other Malignant N eoplasm s Diabetes MelHtus Cerebrovascular Disease All Heart Disease Hypertension Respiratory Disease Ulcer o f Stomach and Duodenum Cirrhosis o f Liver Nephritis and Nephrosis External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Observed Expected Deaths Deaths 0 0.7 0 0.1 0 0.1 1 1.7 0 0.2 0 0.3 0 0.7 1 0.5 1 0.5 1 1.0 0 0.9 1 1.5 10 14.3 0 0.1 0 2.0 0 0.1 2 0 14 , 10 4 6 2 2.3 0.2 22.3 13.7 8.0 5.7 5.2 SMR -- _ 58.4 T,,-- -- 184.8 191.8 97.6 ,,r 66.3 69.9 -- -- '-- 86.6 -- 62.9 73.2 50.2 104.6 38.6 95% Confidence Interval (0.0, 531.1) (0.0, 6,576.8) (0.0, 4,865.3) (1.5, 325.5) v0.0, 1,785.5) (9.0, 1,276.9) (0.0, 545.8) (4.6, 1,030.0) (4.8, 1,058.8) (2.4, 543.7) (0.0, 434.3) (1.7, 369.6) (33.5, 128.5) (0.0, 5,311.6) (0.0, 187.3) (0.0, 2,826.0) (10.5, 312.7) (0.0, 1,870.8) (34.4, 105.5) (35.1, 134.6) (13.7, 128.7) (38.4, 227.6) (4.7, 139.4) 001680 36 T able 20: (C ontinued) Observed Expected Cause of Death Deaths Deaths Homicides and Other External 2 3.3 Causes Residual Causes* .. . ... ... * all ocher cam es o f death combined 4 7.7 SMR 60.2 52.0 95% Confidence Interval (7.3, 217.6) (14.2, 133.1) Eleven deaths were observed for men only employed in film department(s) (Table 21). Two statistically significant SMRs occurred for this subgroup. The all cause SMR was 41.4 (95% C l 20.7,74.0) and the external causes SMR was 37.3 (95% C l ** 10.2, 95.6). Nonsignificant decreases were observed for all cancer (SMR = 46.9; 95% C3 = 5.7,169.5) and h o iit disease (SMR 96.3; 95% C3 = 31.3,224.8). 001681 37 T able 21: C ause-Specific SMRs for M en O nly Em ployed in Film D ep artm en ts) U sing th e U.S. as C o m p ariso n P opulation (n = 547)_______ Cause of Death All Causes of Death All Malignant Neoplasms Cancer o f Buccal Cavity and Pharynx Cancer o f Digestive Organs and Peritoneum Observed Expected Deaths Deaths & 26.6 2 4.3 0 0.1 0 0.8 SMR 41.4 46.9 95% Confidence Interval (20.7, 74.0) (5.7, 169.5) (0.0, 3,497.2) (0.0, 438.8) Cancer of Esophagus Cancer of Stomach Cancer o f Large Intestine Cancer of Rectum Cancer o f L iver and Biliary o 0 0 0 0 0.1 -- 0.1 -- 0.3 -- 0.1 -- 0.1 -> m j t (0.0, 4,205.1) (0.0, 2,802.6) (0.0, 1,223.0) (0.0, 5,730.7) (0.0, 4,808.4) Passages Cancer o f Pancreas Cancer of AH Other Digestive Organs Cancer of Respiratory System Cancer of Larynx Cancer of Bronchus, Trachea, and Lung Cancer of Other Respiratory Organs Cancer of Prostate Cancer o f Testes and Other Male Genital Organs Cancer of Kidney Cancer o f Bladder and Other Urinary Organs 0 0 0 0 0 0 0 0 0 0 0.2 -- 0 .0 (0.0, 2,170.9) (0.0, 11,962.2) 1.3 -- (0.0, 294.3) 0.0 -- (0.0, 10,015.1) 1.2 r * (0.0, 310.2) 0.0 ' r (0.0, 13,379.9) 0.1 -- 0.1 0.0, 7,688.6) (0.0, 3,598.5) 0.1 -- (0.0, 3,050.4) 0.0 TM""" (0.0, 9,158.4) 38 001682 T able 21: (C ontinued) Cause of Death Cancer of Brain and Other Central Nervous System Cancer o f Thyroid and Other Endocrine Glands Cancer of Bone Cancer of All Lymphatic and Hematopoietic Tissue Lymphoma Hodgkin's Disease Leukemia and Aleukemia Cancer o f Other Lymphatic . and Hematopoietic Tissue Malignant Melanoma o f Skin All O ther M alignant N eoplasm s Diabetes MelHtus Cerebrovascular Disease All Heart Disease Hypertension Respiratory Disease Ulcer o f Stomach and Duodenum Cirrhosis o f Liver Nephritis and Nephrosis External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Observed Expected Deaths Deaths 0 0.3 0 0.0 0 0.0 1 0.7 0 0.1 0 0.1 0 0.3 1 0.2 1 0.2 0 0.4 0 0.4 0 0.6 5 5.2 0 0.0 0 0.7 0 0.1 0 1.0 0 0.1 4 10.7 3 6.6 1 3.8 2 2.7 0 2.5 SMR -- -- -- 137.1 -- -- -- 452.3 427.6 -- -- 96.3 -- -- --. -- -- 37.3 45.8 26.1 73.3 -- 05% C o u id en ce Interval (0.0, 1,233.1) (0.0, 15,776.9) (0.0, 10,720.6) (3.4, 763.8) (0.0, 4,372.5) (0.0, 2,803.9) (0.0, 1,260.7) (11.3. 2,520.4) (10.7, 2,382.5) . (0.0, 920.0) (0.0, 1,056.3) (0.0, 642.5) (31.3, 224.8) (0.0, 14,566.9) (0.0, 509.5) (0.0, 7,745.3) (0.0, 385.9) (0.0, 4,857.8) (10.2, 95.6) (9.5, 133.9) (0.7, 145.4) (8.9, 264.8) (0.0, 147.2) 39 001683 Table 21: (C ontinued) Observed Expected Cause o f Death Deaths Deaths Homicides and Other External 1 1.6 Causes SMR 60.9 95% Confidence Interval (1.5, 339.6) Discussion A retrospective cohort mortality study was conducted of former and current employees who had worked at least one year at the 3M Plant in Decatur, Alabama and at least one day after March 1,1961. V ital status was determined through the end o f 1991. Male and female employ es had less than the expected number of deaths from all causes. For men, this decrease was statistically significant Similar results were seen for the subgroups o f men defined by department(s) worked. For each o f the subgroups, the overall number o f deaths was lesr. than expected. There were a total o f 4 deaths among women and 3 of these deaths were due to external causes resulting in a nonsignificant two-fold increase in risk. Examination o f the death certificates for those women with external causes of death revealed that the excess mortality was not work-related. Follow-up o f the cohort and ascertainment of causes of death for deceased cohort members was virtually complete. Vital status was determined for 1,951 (99.7%) o f die 1,957 cohort members. O f the 74 deceased employees, death certificates were obtained for 72 (97.3%). Quality control measures were implemented throughout the study to ensure the integrity of the data, hi addition, several comparison populations were defined and the results were compared and found to be similar. 3M provided the information used to assemble the cohort Although every effort was made to ensure that the cohort was complete, there was no way to independently verify the completeness o f the cohort In an attempt to assess whether the cohort was complete, 3M was asked to supply year-end counts of employees at the 3M Decatur Plant for each year during the study period (i.e., 1961 -1991). These counts were compared to the counts of employees working at the 3M Decatur Plant in December of each year as determined from 40 0 0 1 6 8 4 the study database. The counts from the study database were nearly the same as, or usually greater than, those reported by 3M Recommendation It is recommended that the mortality experience of this group o f employees be updated in 1998. At that time, the NDI will have death records available for an additional five years, 1^92 - 1 9 9 ^ Analyses with additional years o f follow-up and deaths would allow examination o f whether the mortality profile is consistent over time. References 1. MarshGM, FrciningerM. OCMAP: A user-oriented occupational cohort mortality analysis program. AmerStat 1980,34:245. 2. Monson RR. Analysis ofrelative survival and proportional mortality. Comput Biomed Res 1974;7:325-332. 001685 41 Appendices T able 22: C ause-Specific SM Rs for M en U sing A labam a as C om parison P opulation (n = 1,639) 95% Cause of Death Observed Expected Deaths Deaths SMR Confidence Interval All Causes o f Death 70 127.7 54.8 (42.7, 69.3) All Malignant Neoplasms 15 24.6 60.9 (34.1, 100.4) Cancer o f Buccal Cavity and Pharynx 0 0 .6 (0.0, 649.5) Cancer o f Digestive Organs and Peritoneum 1 4.3 23.1 (0.6, 128.8) Cancer of Esophagus 0 0.5 -- (0.0, 810.9) Cancer o f Stomach 0 0.6 -- (0.0, 662.4) Cancer o f Large Intestine 1 1.6 63.8 (1.6, 355.5) O ncer o f Rectum 0 0.3 -- (0.0, 1,290.0) C.incer o f Liver and Biliary o 0.5 -- . (0.0, 779.9) Passages Cancer of Pancreas 0 1.0 -- (0.0, 381.1) Cancer o f All Other 0 0.1 -- (0.0, 2,904.8) Digestive Organs Cancer o f Respiratory System 7 9.9 70.9 (28.5, 146.1) Cancer of Larynx Cancer o f Bronchus, 0 0.2 -- (0.0, 1,651.8) 7 9.5 73.4 (29.5, 151.3) Trachea, and Lung Cancer of Other 0 0.1 "' (0.0, 3,160.3) Respiratory Organs Cancer o f Prostate 0 0.5 -- (0.0, 728.3) Cancer o f Testes and Other 0 0.3 -- (0.0, 1,309.5) Male Genital Organs Cancer o f Kidney 0 0.6 -- (0.0, 657.5) Cancer o f Bladder and Other 1 0.2 419.9 (10.5, 2,339.6) Urinary Organs 001686 42 Table 22: (C ontinued) Cause of Death Cancer of Brain and Other Central Nervous System Cancer o f Thyroid and Other Endocrine Glands Cancer of Bone Cancer o f All Lymphatic and Hematopoietic Tissue Lymphoma Hodgkin's D isease Leukemia and Aleukemia Cancer o f O ther Lymphatic and Hematopoietic Tissue MalignanfViftlMvwna r tf Sirin A ll Other Malignant N eoplasm s Diabetes MelHtus Cerebrovascular Disease All Heart Disease H y p erten sio n Respiratory Disease U lcer o f Stomach and Duodenum Cirrhosis o f Liver Nephritis and Nephrosis External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Observed Expected Deaths Deaths 1 1.4 SMR 70.7 0 0.1 0 0.1 -- 3 2.9 104.4 0 0.4 -- 0 0.4 -- 1 1.2 84.6 2 1.0 208.7 1 L I 95.2 1 2.6 38.9 0 1.4 -- 1 3.6 27.8 17 32.0 ( g t ? 0 0.2 -- I T 4.6 -- 0 0.3 -- 3 3.4 88.9 0 0.4 -- 25 42.4 ^ 5 9 S y 19 28.2 ~ ~ 6T 3 9 17.1 52.7 10 11.2 89.3 4 7.8 51.3 95% Confidence Interval (1.8, 393.8) (0.0, 3,017.6) (0.0, 3,406.6) (21.5, 305.0) (0.0, 987.9) (0.0, 1,018.5) (2.1, 471.6) (25.3, 754.0) (2 .4 ,5 3 0 .3 ) (1.0, 216.7) (0.0, 259.0) (0.7, 155.2) (31.0, 85.1) (0.0, 2,007.5) (0.0, 80.2) (0.0, 1,256.6) (18.3, 259.7) (0.0, 917.2) (38.2, 87.1) (40.5, 105.1) (24.1, 100.0) (42.8, 164.2) (14.0, 131.5) 001687 43 Table 22: (C ontinued) Observed Expected Cause of Death Deaths Deaths SMR Homicides and Other External 2 6.1 32.6 Causes Residual Causes* 7 13.8 50.6 Unknown Causes'! ** a_ll ot.her causes o -f d-eat-h com.bi.ne. d 2 t no rim th InrliuW t nily In II ram *n f A n irfi nrtjny 95% Confidence Interval (3.9, 117.9) (20.3, 104.2) 00168& 44 T able 23: C ause-Specific SM Rs for W omen U sing A labam a as C om parison Population (n = 318) Cause of Death All Cars is o f Death All Malignant Neoplasms Cerebrovascular Disease All Heart Disease Respiratory Disease External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Homicides and Other External Causes Observed Expected Deaths Deaths 4 6.7 0 2.3 1 0.3 0 1.0 0 0.3 3 1.7 2 1.1 2 0.8 0 0.3 0 0.3 1 0.3 SMR 59.9 -- 357.1 -- -- 179.5 185.8 251.1 -- -- 392.8 95% Confidence Interval (16.3, 153.4) (0.0, 162.5) (8.9, 1,989.6) (0.0, 372.0) (0.0, 1,391.2) (37.0, 524.7) (22.5, 671.2) (30.4, 907.2) (0.0, 1,323.4) (0.0, 1,096.1) (9.8, 2,188.8). 001689 45 f T able 24: C ause-Specific SM Rs for M en U sing the A labam a C ounties as C om parison P opulation (n = 1,639) 95% Cause of Death Observed Expected Deaths Deaths SMR Confidence Interval All Causes o f Death 70 134.7 52.0 All Malignant Neoplasms 15 25.1 59.9 Cancer o f Buccal Cavity and 0 0.5 (40.5, 65.7) (33.5, 98.8) (0.0, 822.6) Pharynx Cancer o f Digestive 1 4.2 23.7 (0.6, 132.3) Organs and Peritoneum Cancer o f Esophagus Cancer o f Stomach Cancer o f Large Intestine Cancer o f Rectum Cancer o f Liver and Biliary 0 0 1 0 0 0.4 -- (0.0, 1,048.0) 0.5 -- (0.0, 679.5) 1.5 64.8 (1.6, 361.3) 0.3 -- (0.0, 1,491.0) 0.5 -- (0.0, 721.6) Passages Cancer o f Pancreas Cancer o f All Other 0 1.0 -- (0.0, 374.8) 0 0.1 -- (0.0, 2,730.1) Digestive Organs Cancer of Respiratory System Cancer o f Larynx Cancer o f Bronchus, 7 0 7 10.2 69.0 0.2 -- 9.8 71.3 (27.7, 142.1) (0.0, 1,623.4) (28.7, 146.9) Trachea, and Lung Cancer of Other 0 0.1 (0.0, 3,493.0) Respiratory Organs . Cancer of Prostate Cancer o f Testes and Other Male Genital Organs Cancer of Kidney Cancer o f Bladder and Other 0 0 0 1 0.5 -- (0.0, 711.6) 0.3 (0.0, 1,409.3) 0.6 -- (0.0, 671.1) 0.3 358.7 (9.0, 1,998.9) Urinary Organs 001690 46 Table 24: (C ontinued) Cause of Death Cancer of Brain and Other Central Nervous System Cancer of Thyroid and Other Endocrine Glands Cancer o f Bone Cancer o f All Lymphatic and Hematopoietic Tissue Lymphoma Hodgkin's Disease Leukemia and Aleukemia Cancer o f Other Lymphatic and Hematopoietic Tissue Malignant Melanoma o f Sirin All Other Malignant N eoplasm s Diabetes MelHtus Cerebrovascular Disease All Heart Disease Hypertension Respiratory Disease Ulcer of Stomach and Duodenum Cirrhosis o f Liver Nephritis and Nephrosis External Causes Accidents Motor Vehicle Accidents All Other Accidents Suicides Observed Expected Deaths Deaths 1 1.3 0 0.1 0 0.1 3 2.9 0 0.3 0 0.4 1 1.2 2 1.0 1 1.0 1 3.0 0 1.3 1 3.8 17 34.6 0 0.1 0 5.0 0 0.3 3 3.2 0 0.4 25 45.4 19 31.2 9 19.4 10 11.9 4 7.5 SMR 78.0 -- -- 102.8 -- -- 81.6 195.5 100.4 33.0 --, 26.5 49.1 -- -- -- 92.8 -- 55.0 60.9 46.5 84.2 53.3 95% Confidence Interval (2.0, 434.7) (0.0, 2,981.2) (0.0, 2,711.5) (21.2, 300.5) (0.0, 1,194.5) (0.0, >,021.6) (2.0, 454.9) (23.7. 706.1) (2.5, 559.5) (0.8, 183.9) (0.0, 275.4) (0.7, 147.5) (28.6, 78.6) (0.0, 2,762.4) (0.0, 73.9) (0.0, 1,212.4) (19.2, 271.3) (0.0, 925.1) (35.6, 81.3) (36.7, 95.1) (21.2, 88.2) (40.4, 154.8) (14.5, 136.5) 001691 47 Table 24: (C ontinued) Observed Expected Cause of Death Deaths Deaths SMR Homicides and Other External 2 6.5 30.9 Causes Residual Causes* 7 14.4 48.7 Unknown Causest 2 * all other causes of death combined t no death certificale obtained; includedonlyin all causes ofdeath category 95% Confidence Interval (3.7, 111.8) (19.6, 100.4) Tabi? 25: Cause-Specific SMRs for Women U sing A labam a C ounties as C om parison Population (n = 318) 95% Observed Expected Confidence Cause o f Death Deaths Deaths SMR Interval All Causes o f Death 4 6.8 58.6 (16.0, 150.1) All Malignant Neoplasms 0 2.3 -- (0.0, 162.4) Cerebrovascular Disease 1 0.3 320.2 (8.0, 1,784.0) All Heart Disease 0 1.1 -- (0.0, 343.9) Respiratory Disease External Causes 0 0.3 -- (0.0, 1,381.0) 3 1.7 172.1 (35.5, 502.9) Accidents 2 1.2 173.1 (20.9, 625.3) Motor Vehicle Accidents 2 0.9 230.5 (27.9, 832.6) AH Other Accidents 0 0.3 -- (0.0, 1,289.2) Suicides 0 0.3 -- (0.0, 1,142.5) Homicides and Other External 1 0.3 382.1 (9.6, 2,129.3) Causes 001692