Document rx5OqrnXxRZ0gv0x9ojYarvEe

'-JT-r o a Jj ui .1 tn MO :: 0. L[ X <r 3 ui tn o- t uk *Z Li1: fot: CI 3 <: 3: o x L <r xu tn in CM 1 0 in rM 0* w X tli c 1 U! N til <: cr>. 1 CD 3: z z Li f > o Cl z hi Ci q, Cl > o xn Lh iI : 1 tM- => CO z in 92 :: _i >f- Ci o c ci IU CD a: 0 1- :> <c x I 2: Jitcv- 111 u o Ll 5 o ot> > X T >0 ct ci c: <: ci u Cl 33 e- <: > l: *. n .1 <: til c: Ci -1 >- ri <t 0: CD 1J C* l1 > h C i. : * <: ils: c 0: 4t 1 ;i Ll I ixc -I O" Ll a: n. 1 Cl c: -- c 1 lj n :: _i <: *: 1 t-- m LI .1 : 1- .1 Id 0: u Ll X h- 3 3: a i cn UJ a a: <c X a i9f C cc Ci z <x D M X zb l-l Q o _J CJ m \ I--I n zo o CD 1X1 a: UJ CD 3 cm o ui xx X0 Cl <x C * J Cbtoo Dii 1973, Vol. -6, pt-. )05--1)B. Pcrgamen Pre*a. Printed in Great Britain MORTALITY AMONG ORCHARD WORKERS EXPOSED TO LEAD ARSENATE SPRAY: A COHORT STUDY W. C. Nelson, M. H, Lykjns, J. Mackey*. V. A- Newill, J. F. FtNKLEA and D. I. Hamsier Division or Health Efleets Research, National Environmental Research Center, Environmental Protection Agency, Research Triingle Park, N.C. 27711, U.S.A.. [JUctntd 6 Jtmt 1972; in finalform 26 Stpttmber 1972) INTRODUCTION Before the discovery and heavy use of DDT in the mid-1940's, lead arsenate insecticide spray was used in huge quantities. Total lead arsenate usage in the United States peaked in 1945 at SO million lb and over 5 million lb were stilt applied on fruit and tobacco in 1968 1,2J. The majority of its use was and is on apple orchards. Both lead and arsenic have long been known to be toxic to humans. Clinical lead poisoning is manifested by severe abdominal colic, headaches, anemia, loss of appetite, fatigue, constipation, motor-nerve paralysis, and encephalopathy. Chronic excessive exposure is characterized by acute symptoms superimposed on a background of progressive renal insufficiency and cerebral incompetence [3-7]. Many children with acute encephalopathy, the most severe clinical symptom of lead poisoning, also have acute renal Injury [8, 9J. Almost exclusively confined to children, lead induced encephalopathy often leaves permanent brain damage after both exposure and acute illness have ceased [10-12]. High lead exposure may also be harmful to repro duction [13, 14] and may aggravate some pre-existing diseases such as cirrhosis of the liver [15]. Whereas the previous effects occur at elevated blood lead concentrations, lead can inhibit the activity or an enzyme controlling heme synthesis even when btood lead levels are within the normal range of 10 to 40 pg per 100 g [16]. Several excellent summary ankles exist on the health effects of lead [2, 8, 17, 18]. Symptoms of acute and chronic arsenic poisoning may include dermatitis, nasal and throat irritation, nasal septum perforation, nausea, vomiting, stomach pain and coughing [19]. Cancer or the skin, lung and liver have been attributed to arsenic exposure white cancers of the mouth, esophagus, and larynx are also suspected [19, 20]. Several studies attempting to show increased prevalence of cancer in indus trially exposed workers and to induce cancer in laboratory animals by arsenic exposure have proven negative. A summary of these studies and other health effects infor mation appears in [21]. *Hcalth Services Division, Washington State Department of Social and Health Services, Seattle, Washington 9S104, U-S.A. 105 STO147447 106 \V. C. Nelson, M. H. Lvkim, 1. Mackey. V. A. Newtu., J. F. Finkua and D. I. Hamsax Most effects studies for both lead and arsenic have explored the acute or chronic morbidity efTects of high exposures. There is a lack of information regarding chronic efTects at lower enviromental levels of exposure, especially at levels that do not pro duce clinical poisoning. As yet, no cases of clinical acute lead poisoning have beta reported in adults with blood lead levels below 80 pg per 100 g or in children with blood lead levels below 50 pg per 100 g. Although it is known that lead accumulates with age in the skeleton, kidneys, liver, aorta, pancreas and lungs, few follow-up studies have been made on the present health status of people exposed to high lead levels 20-50 yr ago [22]. This study of orchard workers exposed to lead arsenate spray fills a portion of this major gap in the knowledge of lead and arsenic health effects by examining the effects of lead arsenate exposure on life expectancy. That excess mortality might be expected is indicated by one study claiming an association between excess deaths from kidney disease and lead exposure among an adult cohort with a history of childhood plumbism [23]. SUMMARY OF THE EARLIER STUDY In 1968-1969, the mortality experience of a cohort of Wenatchee area residents who participated in an earlier morbidity study of the health effects or had arsenate spray exposure was studied. In 1938-1939 an epidemiologic study was conducted in the Wenatchee Valley of Washington by the United States Public Health Service. This prolific apple growing region was selected as the study location because of its long use of large quantities of lead arsenate spray and its isolation from industrial exposures to lead.' and arsenic. In 1938, the 1231 study members were classified into three exposure groups. `Orch- ` ardists' prepared and applied lead arsenate sprays during 1933. 'Consumers* in cluded persons whose occupations did not bring them into contact with lead arsenate. This group included most children and a large number of women. 'Intermediates*' were less homogeneous comprising former orchardists, warehouse workers, and those with infrequent lead arsenate spray exposure. For each intermediate and orchardist, the number of years of spray exposure was recorded. Air monitoring was done to determine exposure associated with various orchard activities. The results are summarized in Table 1. Unfortunately, particle' size, a major determinant of respiratory retention, was not determined. For compari son, nonoccupational ambient air lead levels range from 0.01 to 0.05 mg of lead per 10 m* of air. All orchard operations except sorting and packing involved lead ex posures equal to or greater than those experienced in industrial environments such as battery manufacturing, a hazardous lead trad:. However, orchardists experienced, less annual exposure duration to lead in their seasonal tasks than do constantly employed industrial workers. Urinary lead and arsenic and blood lead measurements were obtained for most participants, though the individual values are no longer available. Samples were collected during most months and showed some seasonal variation. Table 2 displays the average concentrations by exposure group and sex. These values verify an exposure gradient for the three groups. In nearly every case, the consumers had the lowest average values and the orchardists the highest average values. Men consistently had higher levels than women. Mortality Actons Orchard Workers Exposed to Lead Arsenate Spray: A Cohort Study 107 Tails 1. Arsenic and lead concentration in the orchard air in 1938 [24] Insecticide operation Mixing insecticide Burning containers Spraying orchard Thinning fruit Picking fruit Dumping fruit (October) (December) Sonin* and peckin* (October) Length of exposure Hours Weeks per day per year Concentration, milli grams per 10 m* (a) Arsenic Average Range Lead Average Range 1 NA<6) 8-14 8-14 8-14 8-12- 8-12 8-12 NA 8-12 2-3 6-10 8-12 4-12 18.5 166.7 1.4 0.8 3J 0.6 0.1 0.06 0.2-110.7 48.6-261.2 0.4-4.8 0.1-30 2.6-19.0 57.4 35.3 4.5 3.0 29.3 0.1-1.9 0.02-00 1.9 0.3 0.03-0.08 0.16 0.9-4670 10.2-76.5 1.3-14.3 0.4-17.0 7.7-75.2 0.4-6.9 0.01-1.1 0.07-002 (a) This volume of air. 10 m*. is approximately the amount that it inhaled during the working day, although the actual amount will vary with the degree of physical activity. Approximately 20-50 per cent of the total metal inhaled will he retimed, primarily as a function of panicle size. <) NA-not available. Tabu 2 Average concentrations or urinary and blood lead and urinary arsenic in 1918 [241 Urinary lead pg per 1 Blood lead pg per 100* Urinary arseoic pg per 1 Adult ( 1S yi) Consumer* Men Women Adult intermediates Men Women Adult orcharduu Meo Women Children (< 15 yr) Boys Girls a 146 123 102 25 3SS 61 81 65 X 350 274 43.3 27.4 880 46.0 52.9 540 m 148 124 108 27 329 38 17 14 1N 260 140 25.8 121 X 620 560 290 123 21.9 25 71.0 38.0 430 305 1400 43.4 58 97.9 36.9 78 1060 36.1 67 1130 Children's values were higher than consumers and intermediates. To determine if children normally have higher levels than adults, a comparison was made of the urinary lead and arsenic and blood lead levels of Bethesda, Maryland children with Bethcsda adults and Wenatchee children. The Bethesda children were lower than both groups. The high levels of the Wenatchee children were attributed to their playing in the orchards and eating unwashed apples. Although the blood and urine analyses showed a consistent gradient for groups, no significant differences were observed for clinical symptoms. Whereas seven in dividuals in the orchardist group did have a low grade of lead arsenate intoxication, they were not numerous enough to be significant in group comparisons of symptoms. The orebardists showed no tendency for the signs of lead and arsenic intoxication to occur in combination. 108 \Y. C. Nhjov ' i H. L-iVINS. M-cxi". . V. A. NuvtLL, 1. F. Fink" . Lsd D. I. Hammer Despite the ne^itive findings of the morbidity study, this reputation afforded a rare opportunity .'or a follow-up mortality study of individ-als exposed to nontoxic levels of lead and arsenic. FOLLOW-UP METHODS While many records from the 1938 study had been lost, three important infor mation sources were available: the final report of the earlier sre-iy. a set of punched cards (hereafter the 1933 cards) containing for each of 1131 study members the exposure group, exposure duration, age, sex, various clinical findings, and some misoellanous data, and a typed list (hereafter the 1938 list) giving the names or the first 1229 study members, as well as each age, sex, exposure group, and (for orchardists) the exposure duration. The addresses of the study members were unavailable, as were the names rf the last two from the original group. The follow-up was begun in 1968, 30 yr after the original ircJy, with the aim of locating either the original study member himself or as informant who could give information leading to the location of the death certificate. For a given study mem ber, field efforts c tntinuid until there was reasonable certainty that the person loc ated or described by the informant or death record was indeed the participant in the previous study. M rst living study members remembered participating. For others,, occupational background was sought and current age or age a: death was checked for consistency against the age recorded on the 1938 cards. Names had to check with the 1933 list, ?r be obvious deviations. Since addresses were lacking, valuable information for follow-up was obtained from the 1937 Wenatchee city directory, the current Wenatchee telephone boofcf* State vital statistic- records, school records, and local birth and death records. O ANALYSIS METHODS "" The following data for each study member were available for analysis: age ajj 1938, sex, exposure group, exposure duration as of 1938, years of orchard wogf after 1938, age at follow-up or age at death or age last known alive, and primage* and associated causes of death. Death causes listed on the death certificates werp coded according tc the International Classification of Diseases (I CD), 1963 Revision. Primary cause of death was chosen using the Rules for Classification of the World Health Assembly. The Standard Mortality Ratio (SMR) technique was used to assess mortality experience. This ratio is calculated by dividing the observed number of deaths by the expected number of deaths. Washington State was chosen as the standard pop ulation from which expected deaths were calculated. Life tables for Washington were obtained for the years 1939-1941,1949--1951, and 1959-1961. The advantage of the life table method in the computing of expected deaths is thtt division of the total time period into smaller (yearly) intervals for which expected deaths are obtained. All follow-up information is utilized since any post-1938 ex perience, even if only a short time interval, is used in the analysis. For each member of the 1938 cohort, use of the life tables enabled the calculation of each individual's contribution to the vears-at-risk and hence expected death category. Expected deaths were calculated in the following manner. The three time periods (1938-1948, 1948-1958, 195S-196S) are centered on 1943, 1953 and 1963. The Wash- liWMi Mort;ity Among Orchard Workcrj C-'.poicd to Lead Arsenate Spray: A Cohort Study 109 ton life tables were centered on the years 1540,1950 and I960. Linear interpolation as used to obtain life tables for 1943 from the 1940 and 1950 tables and also to itain tables for 1953 from the 3950 and 1960 tables. Since mortality rates have mained essentially unchanged over the Iasi decade, the I960 tables were used >r 1963. This process provided sex-specific tables for 10-yr age groups. Values for the 5+ age groups were extrapolated using a fourth degree polynomial. Individual ge death rales were obtained by a graduation formula which used a fifth degree olynomial (Jenkins fifth difference modified osculatory formula) [25]. [n addition to the above tables for total mortality, similar tables were constructed eparately for the following three specific cause groups: heart disease (ICO 390-29), cancer (ICD 140-209), and stroke (1CD 430-439). The expected death contribution foe an individual of a given age is the death -te for that age (sex and time-specific). For example, the expected death contri bution for a male age 30 in 1947 is found from the male table for 1938-1948 using the death rate for age 30. The exposure group categories never changed. Regardless of subsequent orchard work or retirement from orchard work, these categories were retained at their 1938 designations. It was felt better not to mix people who were in different exposure groups in 1938. Their changing exposure would be reflected better by changing exposure duration categories than by changing group categories. FOLLOW-UP RESULTS By the cut-oT date of 1 October 1969, the current status of 1175 of the 1229 possible Study members was determined. Of these 1175, 452 had died and 723 were located alive. In addition, 26 were located at some time after 1938 but lost before 1968. Only 28 (2 per cent) were not located. This latter group included three members who were located but were excluded from the current study; one because he was not a Wenatchee resident but bad participated in the earlier study while visiting relatives, the other two because of large age discrepancies. A classification of the 1201 study members followed-up appears in Table 3. Upon contact with the study member or informant, the information obtained included name, birthdate, follow-up age, sex and current status (dead or alive). For deaths, follow-up age was age at death. For those followed for a time but lost be fore 1963, follow-up age was the age last known alive. Information on orchard experience after 1938 and the reason for leaving orchard work was also obtained. Death certificates were obtained for 442 of the 452 deaths. Five of the other ten were war deaths. For the other five, the date of death was not specific enough to locate the death certificate. The characteristics of the 28 study members lost to follow-up were investigated. Those people having little (ten years or less) or no orchard exposure were more likely to be lost than those with more exposure (3.0 per cent against 0.7 per cent). Females were more likely to be lost than males (3.3 per cent vs 1.7 per cent). For both sexes, orchardists were lost frequently and consumers most frequently. A similar classification study of the 26 study members who were followed briefly but lost before 1968 yielded almost identical results. It is quite unlikely due to the small number of members lost or withdrawn and the small number of orchardists ST0I4745I Modality Amenc Orchard Workers Exposed lo Lead Arsenate Spray: A Cohort Study 111 Ithin these categories, that these people would have significantly affected the analysis suits bad they been completely followed. ANALYSIS RESULTS For all study members combined, the SMR was 0.70. It is substantially below 00 due to the choice of a standard population, Washington State, which includes ban and disabled populations, both higher-risk mortality groups not included in .e Wenatchee area. No consistent differences were observed in the SMR by exposure group category, he overall SMR was 0.74 for the consumers, 0.78 for the intermediates, and 0.65 , r the orchardists. Orchardists certainly gave no evidence of increased mortality. | Analysis by time period also showed no consistent differences. The overall SMRs ere 0.68 for 1938-1948,0.74 for 1948-1958 and 0.68 for 1958-1968. The time period 'edfic SMRs were very similar even for smaller classes. For example, for males age 5+, the time period ratios were 0.98, 1.02, and 0.94 respectively, for intermediates id 0.59,0.70 and 0.64, respectively, for orchardists. Analysis by exposure duration showed inconsistent differences. There was no idence of increasing mortality risk due to longer exposures. For example, the Uegory males age 55 -f contained 78 per cent of the deaths in the two positive eposure groups. Restricting attention to this age-sex category', the SMR for exosure duration 1-5 yr w-as 1.23, for 6-10 yr was 0.93, for 11-15 yr was 0.92, for 5-20 jt was 1.95 and for 21+ years was 0.65 for intermediales. For orchardists :e SMRs were 0.40, 1.61,0.60, 0.81,0.61, respectively. Analysis by sex gave an overall SMR of 0.69 for males, 0.73 for females. Sexxcific analysis by exposure groups showed that male intermediates had the highest MR, 0.89 compared with 0.62 for consumers and 0.64 for orchardists while female rchardists had the highest SMR, 0.86 (only 18 deaths), compared with 0.79 for rasutners and 0.67 for intermediates. Age-specific analysis added no dramalic insight At the younger ages, 0-54, females *.d exhibit some increased mortality (SMR of 1.03 based on only 19 deaths), primarily : the consumer category. The male SMR for the 0-54 group was 0.51. For ages 5+ the male SMR was 0.72, the female SMR was 0.70. Table 4 summarizes the observed and expected deaths and the SMR for total rortality. The classes shown include consumer (total), intermediate (3 exposure urations and total), orchardists (3 exposure durations and total), 3 age groups, nd total for all groups. The expected deaths were 447; the 5 war deaths were ex uded from the analysis. The only category suggestive of excessive mortality is the .termediates with 11-20 yr of exposure. Tabic 5 shows the SMR for intermediates and orchardists adjusted to the overall jnsumer SMR of 0.74. The p values shown are found from the chi-square statistic me degree of freedom) equal to(observeddeaths --expected deaths)','expecteddeaths, nly the 3 orchardists categories: age 45-64, exposure 21+; age 46-64, all exposures; id total were statistically significant In all three cases the observed deaths were gnificantlv/euer than expected. The intermediate categories of age 65+, 11-20 yr id all ages. 11-20 yr are suggestive of exoess deaths but are not different from -:pected at the 5 per cent significance level. 112 SV. C Nilsos, M. H. Lvkins, J. Mackcy, V. A. Ntwtu, J. F. Finxiia and D. I. Hammm TaSU 4. OSSERVTD AND EXKCTED DEATHS AND SMRl SO* TOTAL MORT OJTY Years of exposure Cohort 0-44 Age 43-64 63- Tout 0 1-10 11-30 Consumers Observed Expected SMR Intermediates Observed Expected SMR Orchardists Observed Expected Shift Intermediates Observed Expected Shift Orchardists Observed Expected SMR 11 10.38 136 3 4.62 0.63 7 9.17 0.76 0 0.31 0.00 2 4.0S (UP 29 38.48 0.75 16 2631 0.62 16 2836 0.36 1 834 0.97 26 4136 033 73 10339 0.70 40 30.13 0.80 26 28.18 - 0.92 36 23.48 132 49 77.04 0.64 113 15236 0.74 CO 80.71 oW 0.73----- 4T* 49 63.71 jr 0.75 cn CO 34 34.03 130 77 122.63 0.63 21+ Intermediates Observed Expected SMR Orchardists Observed Expected SMR 0 0.00 -- 0 0.68 0.00 I 231 0.40 13 30.62 0.42 19 27.46 0.69 82 12131 0.68 20 2937 037 93 15230 032 AD years Intermediates Observed Expected SMR Orchardiia Observed Expected SMR 3 4.93 0.61 9 13.90 0.63 . 35 36.76 038 33 10034 035 S3 10339 0.82 157 22633 039 113 144.7 0.78 221 34037 0.63 All groups Observed Expected SMR 23 29.21 0.79 109 175.79 0.62 313 433.31 0.73 447 634.03 0.70 The analysis was also performed separately for each of three principal causes of death, heart disease, cancer, and stroke. The cause-specific life tables were used with other deaths treated as withdrawals. Tables 6, 7 and 8 summarize these results. It is seen that for classes suggestive of excess mortality for all causes, the excess cannot be attributed to any specific cause. In most cases each of the three specific causes are reasonably consistent. For example, for intermediates, all ages, 11-20 yr exposure, the heart disease SMR is 0.97, the cancer SN1R is 1.6, and the stroke SMR Mortality Among Orchard Worker! Evpo<<4 lo Lead Arsenate Spray: A Cohort Study 113 Tasli 5. Total mortalth SMR adjutted to comvmer death rat* and associated p valves Yeats or exposure Cohort 0-44 Are 43-64 65+ Total t-10 Intermediate SMR 0.83 0.83 1.08 0.9S p value -- 0.47 0.63 0.92 Orehardists SMR p value 1.03 0.76 1.24 ixn 0.97 0.28 027 037 11-20 Intermediates SMR p value 0 -- 1J1 1.38 1J5 0.44 0.10 0.08 Orehardists SMR p value 0.67 0.84 0.86 0.S5 039 039 0.15 21+ Intermediates SMR p value Orehardists SMR p value -- 0 -- 0.33 033 030 -- 0.75 0.64 ost 0.91 0.84 0.04 041 009 AH jean Intermediates SMR p value Orehardists SMR p value 0.81 -- 0.87 0.69 0.92 0.67 0.74 0.02 1.11 0.32 034 0.40 1.05 039 0.87 0.04 All groups SMR p value 1.06 034 038 035 0.76 0.06 0.73 0.29 b 0.65 (only 3 deaths). The overall SMR is 0.65 for heart disease, 0.73 for cancer, and 0.79 for stroke. Less common causes of death, including kidney disease, liver disease, and lung cancer, were also studied but were not found to be unusually pre valent The normal-prevalence measures were obtained from relative cause-specific Vital Statistics from Washington State for 1953, the midpoint in the follow-up period. Data were also collected on reason for leaving orchard work. Responses included health, death, retirement, other employment, and other. This information did not prove useful. The female response rate was too low to allow any use of the reasons. While the response rate was slightly better for males, no additional information on the contribution of lead arsenate spray exposure to mortality was gained. DISCUSSION The main conclusion of this study is that excess mortality does not occur con sistently from exposure to lead arsenate insecticide spray. While the results are suggestive of excess deaths at some lesser exposures, the evidence is that many lives including those with the longest exposures are not being affected. An exhaustive set ofclass-specific analyses did not incriminate the lead arsenate exposure as a consistent factor in the observed deaths. STOI 47455 114 W. C Nelson, M. H. Lykins, J. Mackey, V. A. Newill, J. F. Finklca and D. 1. Hamms Table 6. Observes and expected deaths and SMRs pox heart disease mortality Years of exposure Cohort 0-44 Ate 45-64 65+ Total 0 Consumers Observed 1 8 26 35 Expected 132 1439 44.70 6030 SMR 0.66 036 038 038 1-10 Intermediates Observed Expected SMR Orchardists Observed Expected SMR 1 0.90 1.10 0 1.77 0.00 8 10.29 0.78 4 11.94 034 17 2139 0.79 10 1237 0.81 26 3239 030 14 2&0S 034 11-30 21 + Intermediates Observed Expected SMR Orchxrdists Observed Expected SMR Intermediates Observed Expected SMR Orchardists Observed Expected SMR 0 0.07 0.00 0 035 0.00 0 0.00 -- 0 0.IS 0.00 4 3.44 1.16 9 1839 0.49 0 1 JOB OJOO 8 1340 0.60 10 10.88 0.92 19 34.60 035 10 12.29 031 39 5432 0.72 14 1440 037 28 53.84 032 10 1331 0.73 47 6730 0.69 AD years Intermediates Observed Expected SMR Orchardists Observed Expected SMR l 0.98 1.02 0 2.90 0.00 12 1432 0.81 21 43.63 0.48 37 4436 033 68 10139 0.67 50 6036 033 99 147.82 0.60 All groups Observed Expected SMR 2 5.40 0J7 41 72.73 0.56 131 19036 0.69 174 268.69 0.65 Several caveats are necessary. Information on important covariables such as individual cigarette smoking or lead-arsenic exposure from diet is unfortunately not available. The local water supply was investigated as a possible source of lead or arsenic and found not to be a problem. It is difficult to be sure of the exposure dosage categories. It is unfortunate that the individual blood and urine concentrations were lost. While the group averages Mortality Amc.ij Orchard Workers Exposed to Lead Arsenate Spray: A Cohort Study 113 Tou OBSERVED AND EXPECTED DEATHS AND SMRj rox CANCER KORTAUTV yean of exposure Cohort <M4 A|C 43-64 65- Total 0 Consumers Observed 4 Expected 1.68 SMR 238 9 3 16 9.12 13.68 24.48 0.99 0.22 0.65 1-10 Intermediates Observed Expected SMR Orchardiscs Observed Expected SMR 0 0.62 0.00 0 0.98 0.00 3 537 034 6 5.46 1.10 7 732 1.00 3 4.19 0.71 10 13-21 0.76 9 10.63 0.85 11-20 Intermediates Observed Expected SMR Otchatdisls Observed Expected SMR 0 0X33 0.00 0 0.38 0.00 2 1.47 1.36 9 7.16 U6 6 3.48 1.72 3 10.82 &2B 8 438 i.ro 12 IBJ6 065 21+ Intermediates Observed Expected SMR Orchirdists Observed Expected SMR 0 0.00 -- 0 0.07 0.00 0 0.38 0.00 1 4.87 0-21 3 340 0.88 11 16.20 0.68 3 3.78 0.79 12 21.14 037 All yean intermediates Observed Expected SMR Orchirdists Observed Expected SMR 0 0-65 0.00 0 1.43 0.00 5 7.42 0.67 16 17.48 032 16 1330 1.15 17 31.21 0.54 21 2138 036 33 50.13 0.66 All croups Observed Expected SMR 4 3.76 1.06 30 3432 0.88 36 58.80 0.61 70 9638 0.73 followed the suspected exposure gradient, considerable individual variation was present We cannot be positive who the most exposed study members were. The dosage level is particularly a problem for the intermediates, who have the most heterogeneous exposure. Several speculations which can be neither proved nor disproved are possible. One might claim that orchaidists who were most vulnerable to lead arsenate spray left orchard work before 1938 so were either not in this study or were classed as 116 W. C Nelson, M. H. Lvwss, 1. Macxey, V. A- Niwill, J. F. Fikkua and D.LHammi* Taju 8. observed and expected deaths ano SMRs roe sraou mortality Yean of exposure Cohort 0-44 A*e 45-64 65+ Tout 0 Consurr.en Observed 1 Expected SMR OJJ as 2' 3.61 0.54 23 1149 1X4 26 22X2 1.15 1-10 11-20 21 + Intermediates Observed Expected SMR Orehaidists Observed Expected SMR Intermediates Observed Expected SMR Orchardisu Observed Expected SMR Intermediates Observe Expected SMR Orchirdisti Observed Expected SMR 0 0.16 OJOO 0 0.26 0.00 0 0.01 0.00 0 0.12 aoo 0 0X0 0.00 0 0.02 0.00 1 2X4 0.45 1 2.03 0.49 0 0.71 0X0 t 2.99 0.33 0 0X3 0.00 0 2.41 0X0 3 0.39 0X1 6 4X7 Ml 3 3X0 0.77 10 11.70 0.13 0 4X4 0.00 14 10X1 0.70 6 10.99 0X3 (/> --H 7O 6X3 1X7 jr 3 XT* 4.S0 T_ oxs tn 11 14X2 0.74 0 4.47 aoo 14 2044 060 STOI 47457 All yean Intermediates Observed Expected SMR Orchardists Observed pjTpWlfft SMR 0 0.17 0X0 0 0.41 0.00 1 3.17 0X2 2 7.43 0X7 0 16.72 0.40 30 33X7 0X0 9 20.06 043 32 41X2 0.76 All roups Observed Expected SMR 1 0.93 1X7 5 14.28 0X3 61 6o9.x1s0 67 04.40 079 intermediates. Thus, those who remained in orchard work were a self-screened troop who were relatively resistant to any hannful effects. Similarly, the people most susceptible to the lead arsenate spray might either have died or moved away before 1938 leaving behind a selected group of better mortality risks. This theory would account for the extremely favorable mortality experience of the entire cohort in comparison with the total state. Mortality Amont Orchard Workers Exposed to Lead Anenate Spray: A Cohort Study 117 Another difficulty with interpretation of study results is the relatively small number of study members. Some of the suggestive excess deaths are not significant because of small numbers. Many of the volunteers in I93S are still too young to have reached high risk mortality age. Large numbers of subjects are needed in studies involving subtle effects of environmental exposures rather than mortality, which is a rather crude index. To eliminate the possibility that those most affected by a potential hazard may drop out before a formal study begins, future effects studies should identify and follow all individuals from their first exposure to a possible health risk. In the future, too, the expense of better defining individual exposure is obviously justified. Exposure can be better handled in prospective studies than in follow-up studies. From the beginning of the study, the limitations of the small size of the study group, the paucity of exposure information, and the possible bias of the study pop ulation were apparent. However, the availability of such a highly-exposed group afforded a unique opportunity to obtain a mortality analysis. A large gradient in mortality experience would have overcome these recognized difficulties; that this did not occur only emphasizes the. importance of these limitations for future studies. SUMMARY A follow-up mortality study was conducted for a cohort of 1231 individuals in Wenatchee, Washington, who had participated in a 1938 morbidity survey of the exposure effects of lead arsenate insecticide spray. The individuals were classified by spray exposure and duration as well as by age and sex. An `orchardist* group had highest exposure to the lead arsenate spray while a `consumer' group had no direct spray exposure. A third group had intermediate exposure. Duration of the exposure was categorized into 0,1-10,11-20 and 21+ yr. Over 97 per cent or the original group was located in the follow-up which took place in 196S-I969. There were 432 deaths. A life table method of analysis of the standard mortality ratio (SMR) was used. The standard population for calculating expected deaths was the state of Washington. Rate analysis was done for total deaths and for the specific primary causes of heart disease, vascular lesions, and cancer. For all stud/ members combined, the SMR was 70 per cent, demonstrating that this cohort experienced leas mortality than the Washington average. While several of the analysis sub-categories had significant SMRa, the pattern of these differences was not consistent. Exposure group mortality ratios were not consistent with the exposure gradient. Intermediates had the highest SMR (0.78) and orchardists the lowest (0.6S). The mortality pattern for increasing duration of exposure also was not consistent The highest mortality was experienced by the intermediate group in the 11-20 yr duration category. For the cause-specific analyses, the pattern of mortality was reasonably consistent for each analysis sub-category implying that excess deaths were not solely due to any of these three specific principal causes. The authors wish to thank Ml Janet Deling, Reseucb Analyst, and Dr. Jack AQard, Chief, Office of Laboratories, both of the Washington State Department of Social and Health Sertioca; and Dr. Chula Dobtovolnjr, Division of Pesticide Community Studio, Pesticide Oifise *TM Mr. Frederick Sher, Division of Health Effects Research, Office of Research and Monitoring, both or tbe Environmental Protection Agency, for their invaluable and dedicated efforts on this study. 11$ W. C. Neison, M. H. Lyons, J. Mackey, V. A. Newiu., J. F. Finklea and D. 1. Hammz* REFERENCES I. The Pesticide Review--I96S. U.S. Department of Agriculture, Stabilization and CoaservuJo* Service,'..--liiigtcfi, D.CL, 1968 2. Hnjri RE, Hammer Dl, Horton RJM el ah Environmental lead and Public Health. Eovboo. ' -'I mental Protection Agency, Research Triangle Pari:, North Carolina, 1971 . -,j. 3. Henderson DA: Chronic nephritis in Queensland. Amt An Mad 4:163-177,193S 4. Lane RE: Health control in inorganic lead industries. Arch Environ Hltfa 8; 243-50, 1964 '* ' 5. Byars RJC, Lord EE: Lale effects of lead poisoning on mental development. Am 1 Dls Qiy 66:471-494,1943 6. Moeschlin S: Poboclng: Diagnosis and Treatment First American edition, Crane & Straana,-- c New York and London, pp. 45-71,1965 ' 7. Anb JC rf at: Lead poisoning, Medidne Monographs, VoL 7, Baltimore, WOliaros dt Wnirm. :-i\ 1926 . p;; 1 Chisholm JJ, Leaky MB: Aminoaciduria as a manifestation oT renal tabular injury hi kad intoxication. J Ptdiat SO: 1-17, 1962 9. Chisholm IJ:The use ofebdating agents in the treatment of acuta and ebronic lead inlodcalfon vJt in childhood. JPadlat 73:1-31, 1968 10. Smith HD: Pediatric lead poisoning. Arch Environ Hlth 1:256-266,1964 11. Chisholm JJ: Harrison HE: The exposure ofchildren to lead. Pediatrics 18:943-954,1956 ' * 12. Byets AK: Lead poisoning, review of the literature and report of 45 cases. Pediatrics 23: 585-597, 1959 * tfr 13. Hamilton A: Industrial Toxicology. New York, Hoeber, pp. 43-44, 1934 14. Caatarow A. Trumper M: Lead Poisoning. Baltimore, tMDiams dt Wilkins, pp. 97-98, 1944 ' : 15. Seven MJ, Ed: Metal-Bind!itg In Medkioo. Philadelphia, Llppineott, Chapter 4,1960 16- Hcmberg S, Nikkanen 3. Meltin G, el of: Delta-aminolevulinic acid dehydrate is a measure of - kad exposure. Arch Earlroa Hlth 21: 140-145,1910 17. Grundy R: Toxicologic and epidemiologic hues for air quality criteria: carbon monoxide and - kad. J Air PoBut Cootr Am 19: 729-32. 1969 1 18. Kehoe RA: The metabolism of kad in health and disease. Roy last Public Hlth 34:11-97.' ? 129-143,177-203,1961 16 fevir4...e lUn. TML.t*w ^P i - *------ V.. IA41 21. Sullivan RJ: Preliminary Air Pollution Surrey of Arsenic and Its Compounds. National Air' Pollution Control Admlnistiulon APTD 69-26,1969 22. Schroeder HA, Tipton !H: Homan body burden ofkad. Aish Environ Hlth 17:965-978,1968 '. 23. Hendetsoc DA: The aetiology of chronic nephritis in Queensland. Med J Aust: 377-386,1958 ' , 7! 24. Neal PA rf of: A study of the effect of lead arsenate exposure on oichatdists and consumers of sprayed fruit Public Health Santee BolUtin No. 267, 1941 : -3c 25. Miller D: Ekmenta of Graduation. Philadelphia, Actuarial Society of America, pp. 22-24 KiK oi-4