Document KJGV8D36v4rKM4DKqG58eZwN2

^>n the Decline ir. Coronary lality. Washington, DC- Na Health, 1979:116-23. :ne in hospitalizations for cor and stroke: the Kaiser-Per<- in north California, 197i_ , Feinleib M, eds. Proceedings on the Decline in Coronary tality. Washington, DC: NaHealth, 1979:109-15. : mortality in coronary heart rosis 1981;1:312-25. nt decline in ischemic heart \nn Intern Med 1979;9l:630- 'ouncil B, et al. Cardiovascular atterns in Georgia and North alth Bep 1966;75:841-51. lie patterns in the risk of dying ors, ages 35-74 years. United Rockville, MD: National Cenistics, 1980. (Vital and health no. 18). ic areas of the United States. _>ss of Glencoe, 1961. A, Sauer H, et al. Death rates olitan and other areas. Public >:759-66. ns in heart disease mortality 'lew York State. Public Health t. 1 J. Health consequences of The effect of urbanization on ease mortality in rural resis 1964;17:167-77. ledical care, better health? Vban Institute, 1982. SAS user's guide. Cary, NC: 1979. ry mortality: what is going on? 45-6. rise and fall of ischemic heart 40;243:53-9. t RJ, Thom T. The changing :c heart disease. J Cardiovasc !. Cardiovascular disease death in Georgia. J Med Assoc Ga W. The downward trend in ase mortality. Ann Rev Med a i ,uniriN Journal or Epidemiology Copyright 1985 by The Johns Hopkins University School of Hygiene and Public Health All rights reserved Vol. 122, No. 4 Printed in U S A. MORTALITY OF LEAD SMELTER WORKERS SHERRY G. SELEVAN,' PHILIP J. LANDRIGAN,1 FRANK B. STERN,1 and JAMES H. JONES2 Seievan, S. G. (NIOSH, Cincinnati, OH 45226), P. J. Landrigan, F. B. Stem, and J. H. Jones. Mortality of lead smelter workers. Am J Epidemiol 1985;122:673-83. To examine patterns of death in lead smelter workers, a retrospective analysis of mortality was conducted in a cohort of 1,987 males employed between 1940 and 1965 at a primary lead smelter in Idaho. Overall mortality was similar to that of the United States white male population (standardized mortality ratio (SMR) = 98). Excess mortality, however, was found from chronic renal disease (SMR = 192; confidence interval (Cl) = 88-364), and the risk of death from renal disease increased with increasing duration of employment, such that after 20 years employment, the standardized mortality ratio reached 392 (Cl = 107-1,004). Excess mortality was also noted for nonmalignant respiratory disease (SMR = 187, Cl = 128-264). Eight of 32 deaths in this category were caused by silicosis; at least five workers who died of silicosis had been miners for a part of their lives. An additional 11 deaths resulted from tuberculosis (SMR = 139; Cl = 69 249); in six of these cases, silicosis was a contributory cause of death. Cancer mortality was not increased overall (SMR = 95; Cl = 78-114). An increase, however, was noted for deaths from kidney cancer (six cases; SMR = 204; Cl = 75-444). Finally, excess mortality was noted for injuries (SMR = 138; Cl = 104 179); 13 (23%) of the 56 deaths in this category were caused by mining injuries. The data from this study are consistent with previous reports of increased mortality from chronic renal disease in persons exposed occupationally to lead. - kidney neoplasms; lead; nephritis; nephrosis; silicosis Primary lead smelters are large industrial complexes, whose function is to extract lead from ore. In the traditional smelting pro cess, lead sulfide ore (galena) is first sepa rated from nonmetallic rock by differential flotation in an ore concentrator (1). The ore is then roasted in an oxidizing (sinter ing) furnace, where sulfur is driven off prin cipally as sulfur dioxide (S02). The roasted ore is charged with coke into a blast fur nace, where lead is separated from silica, iron, calcium, copper, and other impurities through regulation of temperatures around the melting point of lead. The resulting lead bullion is further refined to remove trace metals. Lead ingots of greater than 99.9 per cent purity constitute the major final prod uct, but modern smelters may also produce commercial quantities of cadmium, zinc, copper, sulfuric acid, gold, and silver. Occupational exposures to lead in pri- Received for publication June 4, 1984, and in final form January 28, 1985. 1 Division of Surveillance, Hazard Evaluations and Field Studies, National Institute for Occupational Safety and Health, Centers for Disease Control, Cincinnati, OH. 'Division of Physical Sciences and Engineering, National Institute for Occupational Safety and Health, Cincinnati, OH. Reprint requests to Dr. Sherry G. Seievan, Office iof Research and Development, US Environmental Protection Agency, Washington, DC 20460. The authors acknowledge the generous assistance iprovided in this investigation by the following persons: 'Clorinda Battaglia, Edith Dodd, Judy Edelbrock, and itheir colleagues at NIOSH; Cynthia Forbes, Kay 'Woods, Pat Cruz, Don Vickars, and Tom McKenna <of Westat, Inc.; and Janet M. Wick, Chief, Bureau of 'Vital Statistics, Idaho Department of Health and Wel Ifare. 673 674 SELEVAN ET AL. mary smelting are typically heavy. Workers ployed between January 1, 1940 and De are also exposed to zinc, cadmium, and cember 31, 1965 at the Bunker Hill ore arsenic as well as to S02, and in some concentrator and primary lead smelter in departments to airborne free silica. Al Kellogg, Idaho, and who worked there for though lead intoxication occurs with con at least one year. Also included were all siderably lower frequency than in the past male hourly workers employed in the same (2) , cases of lead poisoning continue to be time period for whom data on race were not reported in smelter workers (3). The tissues available (77.5 per cent of the work force); principally affected are those of the hema these workers were considered white for the topoietic system, the brain and peripheral purposes of this analysis, since it was nerves, the kidneys, the gastrointestinal known that in 1975 at least 96 per cent of tract, and the testes. The corresponding the work force had been white. Data on the manifestations of toxicity include anemia study population were derived from person (3) , peripheral neuropathy (4, 5), chronic nel records on 34,124 employees which were nephropathy which may result in renal fail ure (3, 6-9), abdominal colic (3, 10), and possibly decreased male fertility (11). Hy pertension, although not a constant finding in previous studies, has been reported in association with occupational lead expo sure (12-14); it is hypothesized to be a consequence of renal injury. Mortality studies of lead smelter and bat tery workers have consistently found excess deaths from chronic renal disease, particu larly in heavily exposed long-term workers (15-18). Excess deaths from cerebrovascu lar disease (12,16,18) and from other forms of hypertensive vasculopathy (15) have also been noted. Excess mortality from cancer has not been a constant finding (12, 15 17), although lead is a mutagen (19), an animal carcinogen (20-25), and has been associated in two case reports of lead work ers with primary renal carcinoma (26, 27). In this report, we present the results of a retrospective analysis of mortality in a co hort of workers employed at a large lead concentrator and smelter in Idaho. Expo sures to lead at this plant had frequently been above established standards (28, 29), and a previous medical survey had found evidence of lead toxicity (30). We were par ticularly interested in examining mortality due to nephritis, hypertensive vascular dis provided to us by smelter management. This record set included data on concentra tor and smelter workers as well as on work ers in other facilities operated by the com pany; it also included information on a number of short-term workers who were employed for less than one year. Many of the study population lived in towns located within a few miles of the smelter. Those towns were heavily contam inated by airborne heavy metal emissions from the plant, and those workers were therefore at risk of both extraoccupational as well as occupational exposure to lead. Also, a large, although not precisely de fined, number of cohort members had worked for a part of their lives as under ground, hard-rock miners, principally in the lead-zinc-silver mines of the Coeur d'Alene district of northern Idaho. These workers were potentially exposed to the hazards of both mining and smelting. Vital status of each cohort member was ascertained through December 31, 1977 by examination of records maintained by the Social Security Administration, the Vet erans Administration, the Post Office ad dress correction service, state vital statis tics offices, the Internal Revenue Service, and state motor vehicle bureaus. For indi ease, renal cancer, and lung cancer. viduals not located through these sources, Methods other means of ascertainment were em ployed, such as review of city directories. The study population consisted of all Workers with unknown vital status were white male hourly workers who were em considered alive at the end of the study; y 1, 1940 and De2 Bunker Hill ore iry lead smelter in o worked there for > included were all iployed in the same ita on race were not of the work force); udered white for the lysis, since it was least 96 per cent of n white. Data on the derived from personnployees which were nelter management, d data on concentTas as well as on work>perated by the comd information on a i workers who were i one year, population lived in a few miles of the kwere heavily contamlavy metal emissions those workers were oth extraoccupational nal exposure to lead, igh not precisely deohort members had ' their lives as undertiiners, principally in mines of the Coeur lorthern Idaho. These tially exposed to the ing and smelting, ch cohort member was December 31, 1977 by rds maintained by the ministration, the Vet,n, the Post Office ad vice, state vital statisemal Revenue Service, hide bureaus. For indithrough these sources, scertainment were emview of city directories, nown vital status were t the end of the study, MORTALITY OF LEAD SMELTER WORKERS 675 this conservative procedure tends to pro Institute for Occupational Safety and duce a slight underestimation of observed Health (31). Person-years of observation mortality. were tabulated by five-year calendar time Death certificates were obtained from periods and age groups, as well as by ex state vital records offices. Underlying posure and latency intervals. We examined causes of death were coded by a qualified separately the effect of employment in nosologist, according to the revision of the high-lead and high-lead/low-other depart International Classification of Diseases ments. Tabulation of person-years for these (ICD) in effect at the time of death; all two subgroups began with the worker's first codes subsequently were converted to the employment in qualifying departments. In Seventh Revision. the high-lead/low-other group, to eliminate To evaluate in further detail the effects from the analysis the possible effects of on mortality of occupational exposures to exposures to metals other than lead, we lead and to other metals, we developed a "censored" from the life table those workers series of exposure definitions from data who had moved from high-lead/low-other obtained in an industrial hygiene survey departments to other high-lead depart conducted at the smelter in 1975 (28) (table ments; that is, person-years for such work 1). We defined high-lead departments as ers were counted for the high-lead/low- follows: those in which the mean level of other departments only until their first po airborne lead exposure for all jobs exam tential exposure to other agents. After that ined was above the then current standard time, neither their person-years nor their of the Occupational Safety and Health Ad deaths contributed to the high-Iead/low- ministration of 200 *ig/m3; or those in other analysis. For each cause of death, which 50 per cent or more of the jobs ex comparison was made between the number amined had mean airborne lead exposures of deaths observed in the cohort and the greater than twice that standard. number expected on the basis of sex, age, Because workers in some high-lead de race, and calendar time period for the partments were also potentially exposed to United States white male population, ac cadmium, zinc, or arsenic, we divided the cording to data published annually by the high-lead departments into those with and National Center for Health Statistics and without high potential for exposure to other the US Census Bureau. Standardized mor metals. Departments with relatively high tality ratios (SMRs) were computed for exposure to lead and relatively low expo each cause of death. sure to other metals were termed "high- Confidence limits for standardized mor lead/low-other" departments and were dis tality ratios were determined using exact tinguished from departments with mixed methods based on a Poisson distribution exposures, which were defined as follows: (32). Examination of trends in standard those in which workers in 50 per cent or ized mortality ratios was performed using more of the jobs examined were exposed to ax2 trend test with 1 df, as developed by airborne cadmium, zinc, or arsenic in con Breslow et al. (33). Because standardized centrations above any of the then current mortality ratios, calculated by indirect standards (table 1); or those in which work standardization, have different underlying ers in 25 per cent or more of the jobs age distributions and because increasing examined were exposed to airborne cad age is correlated in most instances with mium, zinc, or arsenic in concentrations increases in latency and exposure, these more than twice those standards. trend analyses may conceivably be mea To analyze the mortality of the study suring effects due to interactions of occu- ' population, we employed a modified life pational factors with age as well as effects table program developed by the National due solely to increases in latency or in Current OSHA stan dard 50 Dust 200 5,000 10 t* A lt data on airborne exposures are based on results o f eight-hour time-weighted average (T W A *) samples. Several such sample* were collected fo r each job title (28). NA, not available. CV4 S S! a > % i .t ;i i < XfO/i *E =<z<' i-8 u MORTALITY OF LEAD SMELTER WORKERS 677 duration of exposure. The magnitude of such potential confounding depends on the age distributions in the strata under ex amination. disease (table 4), a tendency which may reflect a decrease in the strength of the "healthy worker effect" with aging of the population. Results The study population consisted of 1,987 white male hourly workers (table 2). At the end of the study period (December 31, 1977), 1,281 (64.5 per cent) of this cohort were alive, 665 (33.4 per cent) were de ceased, and 41 (2.1 per cent) were lost to follow-up. Unfortunately, 64 (9.6 per cent) of the 665 death certificates could not be located at state vital records offices, even after provision of supplementary informa tion by the Social Security Administration as to place of death. These deaths were considered to be of unknown cause, a con servative procedure which produces some underestimation of observed case-specific mortality. The age-adjusted overall mortality of the cohort was approximately equal to that of the United States white male population ("all causes" SMR = 98) (table 3). Despite this absence of overall excess, a positive trend was evident between the all causes standardized mortality ratio and the length of the interval (latency) since first occupa tional exposure to lead (for the total popu lation, p = 0.05; for the high-lead/low-other group, p = 0.007) (table 4). This positive trend appears to reflect an upward trend over time in mortality from cardiovascular Table 2 Cohort vital status. Bunker Hill smelter, Kellogg, Idaho, December 31, 1977 Study cohort members Renal disease Excess mortality was observed for chronic, unspecified nephritis and ne phrosis (nine cases; SMR = 192; confidence interval (Cl) = 88-364) (table 3). When mortality from this cause was examined by duration of occupational exposure to lead and by latency interval since first exposure, positive dose-response relationships were evident (table 4). For those workers with more than 20 years exposure, the standard ized mortality ratio for chronic renal dis ease was 392 (Cl = 107-1,004), and for those with more than 20 years latency, the standardized mortality ratio was 315 (Cl = 127-649). Respiratory disease Elevated mortality was observed for "other nonmalignant respiratory diseases" (32 cases; SMR = 187; Cl = 128-264), a category which includes emphysema and the occupational respiratory diseases (table 3). Eight of the 32 deaths were due to silicosis; in five of these eight cases, mining was listed on the death certificate as the "usual" occupation of the deceased worker. In addition, we found that 11 deaths in this population were caused by tuberculosis (SMR = 139; Cl = 69-249) (table 3). Fur ther examination of these death certificates indicated that silicosis was listed as a con tributory cause of death in six workers; four of these workers had died of silicotuberculosis. Known to be alive Known to be deceased Death certificate obtained Death certificate outstanding Not known to be alive or deceased Total 1,281 665 6on 64 J 41 1,987 64.5 33.4 30.21 3.2 J 2.1 100.0 Vascular disease Cerebrovascular disease mortality was not elevated (SMR = 84) (table 3). Stan dardized mortality ratios for cerebrovascu lar disease, however, tended to increase with increasing duration of employment (p = 0.002; high-lead/low-other group, p = 0.07), and to a lesser extent with increasing I <o -- CM O CO 00 --* -- CM s T> o u t . 1 ^1 r--~l 8 _2 c :5 JC x xs 'uw < --" < ao< 05-2*"* >x A-- Q *o 4k1. 2 o a, II >4) Xa C/3 V.. *40> $d *t . J3 q3 cS-.2s -->. -->a> .-23 u js s- 3 eu s s -a a 5? *C i- a_> _ -3 I c g ^5 & |g5 N-si iE -co oc . S!% = ":=! 5; * 0 *t Q ^ o ^Mo5 1 as i. e a CQ a I "e3 .s 1 0 1 a Is S2 Ii a5. 3 SS 83 ^O *4) c2 MORTALITY OF LEAD SMELTER WORKERS n a> w S^- OQ 3- -- CO 8 cm r5 cm --, --. ^ nlfl J> Jj 2 ~J. --J. C_M1 WI I 00 05 lS r- 8 3 8 5$ 8 S8 - iu5 fS 82 C- 56ft s 679 ^ 04 kO ** t' I 1 ICM C eWo C--M !S S CM O CO CM tC-M 85 CM < CM CM a> at a--c o/s: S3 "I M g o> I 4> is J =E "O I f 44>0 a a> S'g Wo. W ae> Qr 5 2 E< I I V . "" > osK aSX *c o 9! Q i ' W) J S 9 60 t o >n |i S-7 *5 .11 s t 5ic ttsj 55:r frS 5l^|5||3 t s :s; S5-c2 S lap 11J <ge- o I0 D , International Classification <>f Diseases, Seventh Revision.t SMR, standardized m o rta lity ratio (observed/expeeted) x 100; not calculated for observed number o f deaths less than two. } Cl, 95% confidence interval. Because o f changes in ICO coding, these data refer only to deaths occurring after 1950. 680 SELEVAN ET AL. latency (p = 0.07; high-lead/low-other group, p = 0.07) (table 4). Mortality from diseases of the circulatory system occurred at a lower frequency than expected (SMR = 78; Cl = 69-88) (table 3). Mortality from circulatory disease, how ever, tended to increase with increasing latency (p = 0.006) (table 4). Cancer Overall mortality from cancer (all ana tomic sites) was not elevated (SMR = 95). Elevated standardized mortality ratios, however, were noted for kidney cancer (SMR = 204) (table 3). No substantial ex cess was observed in mortality from respi ratory cancer (SMR =111) (table 3). Injuries An excess of deaths due to injuries was observed in this cohort (SMR = 138; Cl = 104-179) (table 3). Twenty-two (39.2 per cent) of the 56 deaths in this category were due to occupational injuries. The majority of these deaths, however, were associated with underground mining and not with em ployment at the smelter. Six were caused by suffocation and smoke inhalation in the Sunshine Mine disaster at Kellogg, Idaho, May 2,1972. Another seven of these deaths were caused by various other mining epi sodes. Discussion This analysis of mortality in workers ex posed to lead at the Bunker Hill smelter was prompted by investigations which had found lead contamination in the environ ment surrounding the smelter (34) and in creased lead absorption, anemia, and neu rologic impairment among children in ad jacent communities (35). The present study was undertaken in conjunction with an in dustrial hygiene evaluation (28) and a cross-sectional medical survey (30). The major finding was the observation of excess mortality from chronic renal dis ease. This effect was most strongly evident in long-term lead workers (table 4). Ne phritis has been recognized as a component of occupational lead intoxication since the late nineteenth century (36, 37). Although the natural history and dose-response re lationships for lead nephropathy are less well understood than those for other as pects of chronic lead poisoning, it appears that nephropathy is a consequence of pro longed, relatively high-dose exposure to lead (3,6-9,37). In previous mortality stud ies, deaths from lead nephropathy have tended to occur in workers with longest duration of exposure to lead (12, 16, 18) and in those who had suffered clinical lead poisioning (17). The epithelial lining cells of the proximal tubules appear to be the tissue in the kid neys most highly sensitive to lead. The first histologically demonstrable effect of lead in the kidneys is the formation in these cells of densely staining intranuclear inclusion bodies (9), which have been shown to be comprised of a lead-protein complex (37). With continuing exposure, irreversible in terstitial renal fibrosis develops (9). The end stage of lead nephropathy is chronic renal failure, characterized pathologically by interstitial fibrosis (38) and by atrophy and cystic dilatation of the tubules with relative sparing of the glomeruli (37). In some instances, lead nephropathy has been reported to be associated with hyperuricemic gout (8), possibly the result of a leadinduced defect in the tubular secretion of uric acid. Lead workers with nephropathy have been shown to have increased quan tities of lead in bone and to excrete in creased amounts of lead in urine following chelation challenge (38, 39), compared with lead workers without nephropathy or with unexposed comparison subjects. The find ing in the present study of increased mor tality from chronic nephritis corroborates the results of four previous epidemiologic analyses of lead workers (15-18). The present study also found increased mortality from renal cancer (six cases; SMR = 204; Cl = 75-444). Lead has been found in experimental studies in rats, mice, i zed as a component oxication since the (36, 37). Although i dose-response rephropathy are less those for other asoisoning, it appears :onsequence of prol-dose exposure to ious mortality studnephropathy have irkera with longest 0 lead (12, 16, 18) uffered clinical lead cells of the proximal le tissue in the kidive to lead. The first able effect of lead in lation in these cells ranuclear inclusion 1 been shown to be otein complex (37). ore, irreversible in- 0develops (9). The opathy is chronic ized pathologically (38) and by atrophy of the tubules with ; glomeruli (37). In phropathy has been ed with hyperuricethe result of a leadtubular secretion of s with nephropathy ave increased quanand to excrete inld in urine following 39), compared with rephropathy or with i subjects. The findly of increased morphritis corroborates evious epidemiologic rs (15-18). also found increased cancer (six cases; -444). Lead has been studies in rats, mice, MORTALITY OF LEAD SMELTER WORKERS 681 and hamsters to produce both chronic renal disease and renal tumors (20-24, 40, 41). Although levels of lead exposure in those studies far exceeded the maximum doses tolerated by man and caused gross morpho logic damage to the kidneys, positive doseresponse relationships were observed be tween lead dose and tumor incidence; also, the interval from onset of exposure to tu mor formation was reduced in the more heavily exposed animals (23). Two recent case reports have noted primary renal tu mors in association with chronic nephritis in long-term lead workers (26, 27). Excess mortality from renal cancer has not been observed in previous epidemiologic studies of lead workers. Pending epidemiologic cor roboration, the present finding must be in terpreted with caution. Death rates from both silicosis and tu berculosis were found to be elevated in this study; in six (55 per cent) of 11 workers who died of tuberculosis, silicosis was also noted to be present, and four of these death certificates recorded a diagnosis of silicotuberculosis. Similar findings have not been reported in previous mortality studies of smelter workers (12, 15-18), and the report by Cooper (15) noted a standardized mortality ratio for tuberculosis (all sites) in smelter workers of only 23. The present findings may reflect occupational exposure to silica in the concentrator and charge preparation departments of the smelter, where limited air sampling detected several exposures to silica dust in concentrations above the Occupational Safety and Health Administration standard (28). A more likely explanation, however, is that many workers in this cohort appear to have had occupational exposure to silica in under ground, hard-rock mining, a common form of employment in northern Idaho. Mining was recorded on death certificates as usual occupation for five of the eight deaths due to silicosis encountered in this study. Un derground miners have long been known to be at increased risk of death from silicosis and tuberculosis (42). In contradistinction to several previous studies of lead workers (12, 16, 17), this analysis did not find increased mortality from cerebrovascular or hypertensive vas cular disease. Only slight upward trends in standardized mortality ratios for cerebro vascular accidents were noted with increas ing duration of employment and with in creasing latency. There is no obvious expla nation for the differences between these results and those of previous studies. A limitation in the present study lies in the definition of exposure. Although we attempted to define exposure by classifying workers according to 1) the duration of their employment, 2) the length of the la tency interval since their first employment, and 3) their assignment to high-exposure departments (28), each of those categori zations is at best a relatively crude surro gate measure of actual exposure. In no way do these measures take cognizance of indi vidual differenes in work practices, in res pirator use, or in extraoccupational expo sures to heavy metals, a form of exposure which could have been quite substantial for workers living in some neighborhoods near this smelter (34, 35). A likely consequence of the insensitivity of our measures of ex posure is that we have underestimated the full range of exposures to lead and to other heavy metals which existed within this co hort. Had we been able to identify more precisely those members of the cohort who had heaviest cumulative exposures, we might have been able to define more pre cisely the effects of such exposures on mor tality. Two further limitations derive from our use of the entire United States white male population as a basis for calculating ex pected mortality. In the first instance, most populations of workers are healthier than the general population, because the latter includes numbers of chronically ill or other wise unemployable persons. Consequently, comparison of the mortality experience of a working population with that of a general population produces an apparent deficit in 682 selevan et al. mortality, the so-called healthy worker ef fect (43). This effect tends to be strongest in the younger age groups and to diminish with increasing age (43). This diminution may reflect erosion in health as a conse quence of hazardous work. Arguably, such erosion may account for the apparent up ward trend for mortality from cardiovas cular disease as well as for overall mortality which was observed in the present cohort with increasing latency since first employ ment. Resolution of the issues surrounding the healthy worker effect will require the development of life table programs which use the mortality experience of blue-collaremployed populations as a basis for com puting expected mortality; such programs are currently under development in a joint effort between the National Cancer Insti tute and the National Institute for Occu pational Safety and Health. The second potential limitation stem ming from our use of the entire US popu lation and a basis of comparison is that this approach may obscure regional differences in mortality and thus distort calculated standardized mortality ratios. To examine this issue, we obtained data on causespecific deaths for the state of Idaho for cancer for the years 1950-1977 and for other causes of deaths for the years 1962 1977. We found that age-adjusted Idaho rates for those years were 27.7 per cent higher than US rates for deaths due to injury and 54.2 per cent lower for deaths due to tuberculosis. However, for deaths due to chronic and unspecified nephritis and for deaths due to kidney cancer, the Idaho and US rates were, respectively, within 5 per cent and 10 per cent of each other. Thus, the US rates provide an ac ceptable basis for calculating expected mor tality due to renal disease in the population of Idaho. The statistical power of this study to detect increases in mortality depended upon the outcome under study and varied widely according to the background fre quency of each condition. Based on the equations of Beaumont and Breslow (44), the power of this study to have detected a doubling in expected mortality for lung cancer was 98; for stroke 99.8; and for nonmalignant respiratory diseases 99.8. The power to detect doubling in mortality, how ever, was much lower for other causes of death: for cancer of the urinary organs 51; for cancer of the bladder 36; for kidney cancer 28; and for nephritis and nephrosis 44. In summary, the major finding of this study was the observation of excess mor tality due to chronic renal disease in long term lead workers. This finding corrobo rates the results of previous mortality stud ies of lead workers (12, 15-18). The study also found an increase in mortality for renal cancer. This finding is consistent with the results of experimental studies in animals, but has not previously been encountered in an epidemiologic analysis. - References 1. Howe HE. Lead. In: Kirk-Othmer encyclopedia of chemical technology. 3rd ed. Vol 14. New York: John Wiley & Sons, 1978:98-139. 2. Hunter D. The diseases of occupations. 4th ed. Boston: Little, Brown & Co, 1969. 3. Baker EL Jr, Landrigan PJ, Barbour AG, et al. Occupational lead poisoning in the United States: clinical and biochemical findings related to blood lead levels. Br J Ind Med 1979;36:314-22. 4. Araki S, Honma T. Relationships between lead absorpf`on and peripheral nerve conduction veloc ities in lead workers. Scand J Work Environ Health 1976;4:225-31. 5. Seppalainen AM, Tola S, Hernberg S, et al. Sub- clinical neuropathy at "safe" levels of lead expo sure. Arch Environ Health 1975;30:180-3. 6. Wedeen RP, Maesaka JK, Weiner B, et al. Occu pational lead nephropathy. Am J Med 1975; 59:630-41. 7. Lilis R, Gavrilescu N, Nestorescu B, et al. Ne phropathy in chronic lead poisoning. Br J Ind Med 1968;25:196-202. 8. Ball GV, Sorensen LB. Pathogenesis of hyperuri cemia in saturnine gout. N Engl J Med 1969; 280:1199-1202. 9. Goyer RA, Rhyne BC. Pathological effects of lead. Int Rev Exp Pathol 1973;12:1-77. . 10. Dahlgren J. Abdominal pain in lead workers. Arch Environ Health 1978;21:156-9. 11. Lancranjan 1, Popescu HI, Gavanescu O, et al. Reproductive ability of workmen occupationally exposed to lead. Arch Environ Health 1975; 30:396-401. 12. Dingwall-Fordyce I, Lane RE. A follow-up study of lead workers. Br J Ind Med 1963;20:313-15. 13. Emmerson BT. Chronic lead nephropathy: the 'I dy to have detected a d mortality for lung oke 99.8; and for nony diseases 99.8. The ing in mortality, how;r for other causes of ;he urinary organs 51; ladder 36; for kidney ,'phritis and nephrosis 3 major finding of this vation of excess morrenal disease in longThis finding corroborevious mortality stud12, 15-18). The study e in mortality for renal is consistent with the tal studies in animals, ly been encountered in ilysis. RENTES virk-Othmer encyclopedia of m3rd ed. Vol 14. New York: 8:98-139. s of occupations. 4th ed. & Co, 1969. ,gan PJ, Barbour AG, et al. isoning in the United States: ical findings related to blood Med 1979;36:314-22. Relationships between lead leral nerve conduction veloc-s. Scand J Work Environ I. la S, Hernberg S, et al. Subit "safe" levels of lead expolealth 1975;30:180-3. a JK, Weiner B, et al. Occuropathy. Am J Med 1975; N, Nestorescu B, et al. Neic lead poisoning. Br J Ind B. Pathogenesis of hypemrigout. N Engl J Med 1969; Pathological effects of lead. 1973;12:1-77. nal pain in lead workers. Arch 8;21:156-9. scu HI, Gavanescu O, et al. . of workmen occupationally trch Environ Health 1975; Lane RE. A follow-up study 1 Ind Med 1963;20:313-15. ronic lead nephropathy: the a I 3 a 1 3 3 3 MORTALITY OF LEAD SMELTER WORKERS 683 diagnostic use of calcium EDTA and the associa tion with gout. Australas Ann Med 1963;12:3-10- 24. 14. Cramer K, Dahlberg L. Incidence of hypertension among lead workers: a follow-up study based on regular control over 20 years. Br J Ind Med 1966;23:101-4. 15. Cooper WC. Mortality in employees of lead bat tery plants and lead-producing plants. Presented at the XXI International Congress on Occupa tional Health, Dublin, Republic of Ireland, Sep tember 1984. 16. Malcolm D, Barnett HAR. A mortality study of lead workers 1925-76. Br J Ind Med 1982;39:404- 10. 17. McMichael AJ, Johnson HM. Long-term mortal ity profile of heavily-exposed lead smelter work ers. J Occup Med 1982;24:375-8. 18. Davies JM. Long-term mortality study of chro mate pigment workers who suffered lead poison ing. Br J Ind Med 1984;41:170-8. 19. DiPaolo JA, Nelson RL, Casto BC. In vitro neo plastic transformation of Syrian hamster cells by lead acetate and its relevance to environmental carcinogenesis. Br J Cancer 1978;38:452-5. 20. Boyland E, Dukes CE, Grover PL, et al. The induction of renal tumours by feeding lead acetate to rats. Br J Cancer 1962;16:283-8. 21. Zawirska B, Medras K. Tumoren und Storungen des Porphyrinstoff-Wechsels bei Ratten mit Chronischer Experimenteler Bleiintoxikation. I. Morphologische Studien. Zentralbl Allg Pathol 1968;111:1-12. 22. Van Esch GJ, Kroes R. The induction of renal tumours by feeding basic lead acetate to mice and hamsters. Br J Cancer 1969;23:765-71. 23. Van Esch GJ, Van Genderen H, Vink HH. The induction of renal tumours by feeding of basic lead acetate to rats. Br J Cancer 1962;16:289-97. 24. Mao P, Molnar JJ. The fine structure and histo chemistry of lead-induced renal tumors in rats. Am J Pathol 1967;50:571-603. 25. Oyasu R, Battifora HA, Clasen RA, et al. Induc tion of cerebral gliomas in rats with dietary lead subacetate and 2-acetylaminofluorene. Cancer Res 1970;30:1248-61. 26. Baker EL, Goyer RA, Fowler BA, et al. Occupa tional lead exposure, nephropathy, and renal can cer. Am J Ind Med 1980;1:139-48. 27. Lilis R. Long-term occupational lead exposure, chronic nephropathy, and renal cancer: a case report. Am J Ind Med 1981;2:293-7. 28. National Institute for Occupational Safety and Health. Final report. Environmental phase. Bunker Hill Study. Prepared by ME Cassady, LB Larsen. Cincinnati, OH: National Institute for Occupational Safety and Health, 1975. Report no. 56.10.5. 29. National Institute for Occupational Safety and Health. An industrial hygiene survey of the zinc cell room, anode room, and zinc casting area: The Bunker Hill Company. Prepared by L Anderson, LB Larsen, W Spear. Cincinnati, OH: National Institute for Occupational Safety and Health, 1971. Report no. 56.10. 30. Johnson BL, Burg JR, Xintaras C, et al. A neurobehavioral examination of workers from a pri mary nonferrous smelter. Neurotoxicology 1980; 1:561-81. 31. Waxweiler R, Beaumont J, Henry J, et al. A modified life table analysis system for cohort stud ies. J Occup Med 1983;25:115-24. 32. Rothman K, Boice J. Epidemiologic analysis with a programmable calculator. Washington, DC: US GPO, 1979. (US DHHS (PHS, NIH), N1H publi cation no. 79-1649). 33. Breslow NE, Lubin JH, Marek P, et al. Multi plicative models and cohort analysis. J Am Stat Aficru* 34. Yankel Aj', von Lindern IH, Walter SD. The Silver Valley lead study: the relationship of child hood lead poisoning and environmental exposure. J Air Pollut Control Assoc 1977;27:763-7. 35. Landrigan PJ, Baker EL Jr, Feldman RG, et al. Increased lead absorption with anemia and slowed nerve conduction in children near a lead smelter. J Pediatr 1976;89:904-10. 36. Emmerson BT. Chronic lead nephropathy. Kid ney Int 1973;4:1-5. 37. Wedeen RP. Lead, mercury and cadmium ne phropathy. Neurotoxicology 1983;4:134-46. 38. Cramer K, Goyer RA, Jagenburg R, et al. Renal ultrastructure, renal function, and parameters of lead toxicity in workers with different periods of lead exposure. Br J Ind Med 1974;31:113-21. 39. Malcolm D. The effects of lead on the kidney. Trans Soc Occup Med 1970;20:50-3. 40. Zollinger HU. Durch chronische Bleivergiftung erzeugte Nierenadenome und-Carcinoma bei Rat ten und ihre Beziehungen zu den entsprechenden Neubildungen des Menschen. Virchows Arch (Pathol Anat) 1953;323:694-710. 41. Roe FJC, Boyland E, Dukes CE. Failure of testos terone or xanthopterin to influence the induction of renal neoplasms by lead in rats. Br J Cancer 1965;19:860-6. 42. Wagoner JK, Miller RW, Lundin FE, et al. Un usual cancer mortality among a group of under ground metal miners. N Engl J Med 1963;269:2849. 43. Fox AJ, Collier PF. Low mortality rates in indus trial cohort studies due to selection for work and survival in the industry. Br J Prev Soc Med 1976;30:225-30. 44. Beaumont JJ, Breslow NE. Power considerations in epidemiologic studies of vinyl chloride workers. Am J Epidemiol 1981;114:725-34.