30f189

ROCKETDYNE WORKER HEALTH STUDY

INTERNATIONAL
EPIDEMIOLOGY
I I 

 

IEI EXECUTIVE SUMMARY

July 13, 2005

4 of 189
Table of Contents
Page
Statement of Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Overall Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Overall Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Overall Radiation Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Overall Chemical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Questionnaire Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Overall Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2
2
2
3
3
4
4

Specific Study Approaches and Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1. Institutional Review Board (IRB) and Other Approvals . . . . . . . . . . . . . . . . . . . 4
2. Identification of the Worker Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3. Population Tracing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
4.
Cause of Death Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.
Radiation Dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
6. Chemical Exposure Assessment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
7.
Study Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
a. Radiation Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
b. Chemical Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
8.
Other Analyses and Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
a. White Males . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
b. External Comparison Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
c. Internal Comparison Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
d. Healthy Worker Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
e. Radiation Dose Lagging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
f. Workers Monitored for Radiation Only at Rocketdyne . . . . . . . . . . . . . . . 10
g. Smoking Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
h. Hourly and Salary Workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
i. Radiation Dose Response by Pay Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
j. Chemical Cohort Tables Excluding Radiation Workers . . . . . . . . . . . . . . . 12
k. Time and Duration Analyses for SSFL and the Other Rocketdyne Workers 12
l. Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
m. Test Stand Workers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
n. Special Groupings of Cancer Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
o. Asbestos . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
p. Beryllium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
q. Trend Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
9.
Comparisons with UCLA Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
a. Radiation Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
b. Chemical Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
i

 5 of 189
10.
11.
12.
13.

Final Comments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Manuscripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Papers Cited . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Summary Charts and Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 1. Vital Status of Rocketdyne Workers . . . . . . . . . . . . . . . . . . . . . . . .
Figure 2. External Radiation Dose Distribution . . . . . . . . . . . . . . . . . . . . . . .
Figure 3. Radiation Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 4. Comparing Radiation Dose Received Only at Rocketdyne with
Total Dose Received at All Facilities . . . . . . . . . . . . . . . . . . . . . . .
Figure 5. SSFL (Chemical) Cohort . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 6. Entire Rocketdyne Workforce Compared to General Population
of California . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 7. Radiation Workers Compared to General Population of California
Figure 8. Radiation Dose Response for All Cancer Excluding Leukemia . . .
Figure 9. Radiation Dose Response for Leukemia . . . . . . . . . . . . . . . . . . . . .
Figure 10. SSFL Workers (Chemical Cohort) Compared to the General
Population of California . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 11. Dose Response for All Cancers Combined by Years Worked
at SSFL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 12. Test Stand Mechanics Compared to the General Population of
California . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 13. Dose Response for All Cancers Combined by Years Worked as
a Test Stand Mechanic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 14. Dose Response for Lung Cancer by Years Worked as a Test
Stand Mechanic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 15. Classification of Potential Exposure to Hydrazines Among Test
Stand Mechanics Based on Job Title & Test Stand . . . . . . . . . . . .
Figure 16. Test Stand Mechanics Potentially Exposed to Hydrazines
Compared to California Population . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 17. Dose Response for All Cancers Combined for Test Stand
Mechanics with Potential Exposure to Hydrazines . . . . . . . . . . . . .
Figure 18. Dose Response for Lung Cancer for Test Stand Mechanics with
Potential Hydrazines Exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 19. Test Stand Mechanics Potentially Exposed to TCE Compared
to the California Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 20. Dose Response for All Cancers Combined for Test Stand
Mechanics with Potential Exposure to Trichloroethylene (TCE) .
Figure 21. Dose Response for Lung Cancer for Test Stand Mechanics with
Potential Exposure to TCE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 22. Dose Response for TCE Suspected Cancers* for Test Stand
Mechanics with Potential Exposure to TCE . . . . . . . . . . . . . . . . . .
14. PowerPoint Presentation 6-8 April 2005 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 1pp. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 2pp. Who was in the study? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ii

15
16
17
18
20
20
21
21
22
22
23
23
24
24
25
25
26
26
27
27
28
28
29
29
30
30
31
33
33

 6 of 189
Figure 3pp.
Figure 4pp.
Figure 5pp.
Figure 6pp.
Figure 7pp.
Figure 8pp.
Figure 9pp.
Figure 10pp.
Figure 11pp.
Figure 12pp.
Figure 13pp.
Figure 14pp.
Figure 15pp.
Figure 16pp.
Figure 17pp.
Figure 18pp.
Figure 19pp.
Figure 20pp.
Figure 21pp.
Figure 22pp.
Figure 23pp.
Figure 24pp.

Figure 25pp.

Figure 26pp.
Figure 27pp.

What were the two types of radiation exposure? . . . . . . . . . . . . . 34
How many people were in the radiation group? . . . . . . . . . . . . . . 34
Potential chemical exposure characterized by years worked . . . . 35
Nine discussion sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
How many SSFL workers were potentially exposed to chemicals
as test stand mechanics? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Worker Groups . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
Rocketdyne workers had a lower risk of death than the general
population of California . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Rocketdyne radiation workers had a lower risk of death than
the general population of California . . . . . . . . . . . . . . . . . . . . . . . 37
Most radiation workers received very low exposures . . . . . . . . . 38
What was the effect of including pre- and post-Rocketdyne
radiation dose? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Interpreting Dose Response Graphs . . . . . . . . . . . . . . . . . . . . . . . 39
No evidence that radiation increased the risk of dying from
cancer (excluding leukemia) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
No evidence that radiation increased the risk of dying from
lung cancer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Suggestive, although not statistically significant, evidence that
radiation increased the risk of dying from leukemia . . . . . . . . . . 40
Radiation Summary Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
SSFL workers (Chemical Group) had a lower risk of death than
the general population of California . . . . . . . . . . . . . . . . . . . . . . . 42
No evidence that working at SSFL increased the risk of dying
from all cancers combined . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
Test stand mechanics had a lower risk of death than the general
population of California . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
No evidence that working as a test stand mechanic increased the
risk of dying from all cancers combined . . . . . . . . . . . . . . . . . . . 43
No evidence that working as a test stand mechanic increased
the risk of dying from lung cancer . . . . . . . . . . . . . . . . . . . . . . . . 44
Classification of potential exposure to hydrazines among test
stand mechanics based on job title and test stand . . . . . . . . . . . . 44
Test stand mechanics potentially exposed to hydrazines had a
lower risk of death overall but slight increased risk of dying from
cancer compared to the general population of California . . . . . . 45
No evidence that test stand mechanics with potential exposure
to hydrazines had an increased risk of dying from all
cancers combined . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Little evidence that test stand mechanics with potential exposure
to hydrazines had an increased risk of dying from lung cancers . 46
Classification of potential exposure to trichloroethylene (TCE)*
among test stand mechanics based on job title and test stand . . . 46
iii

 7 of 189
Figure 28pp. Test stand mechanics potentially exposed to TCE had a lower
risk of death overall but similar risk of dying from cancer
compared to the general population of California . . . . . . . . . . . .
Figure 29pp. No evidence that test stand mechanics with potential exposure
to TCE had an increased risk of dying from all cancers . . . . . . .
Figure 30pp. Chemical Summary Findings . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 31pp. Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 32pp. Strengths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Figure 33pp. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

iv

47
47
48
48
49
49

 8 of 189
STATEMENT OF WORK
The overall objectives outlined in the Statement of Work (12/14/00) were as follows:
“A retrospective cohort mortality study will focus on individuals employed in either
nuclear technology development or in rocket engine testing since 1950 at the following
Boeing (Rocketdyne) facilities: Santa Susana Field Laboratory (SSFL) in the Simi Hills
area of Ventura County, California, Canoga Park and De Soto Avenue. The study will
determine whether mortality rates of cancer and other diseases are elevated among these
workers, and whether mortality varies as a function of length of employment, place of
employment (and or job title) or work with specific chemicals or radiation. Nested casecontrol studies shall be conducted for any type of cancer that appears to show an excess
risk in the cohort (e.g. lung, leukemia or lymphoma).”
“The published reports on Rocketdyne workers (UCLA studies) have recognized
deficiencies, many acknowledged by the authors, and this extended study will
incorporate a more comprehensive and rigorous approach. The observation period shall
be extended, more appropriate comparison populations shall be sought, approaches to
determining vital status shall be expanded, pre- and post-Rocketdyne radiation exposures
shall be ascertained, internal radiation doses shall be determined in a comprehensive
manner, and complete and detailed chemical exposure information shall be sought.
Further, the seller will provide an experienced and highly credible research team
committed to making this project their highest priority during the next five years.”
The objectives in the Statement of Work (12/14/00) were addressed during the four years of
study. Results are summarized in this Executive Summary, in nine booklets prepared for the
seven meetings of the Scientific Committee, and in four manuscripts prepared for publication.
The Executive Summary begins with an Overall Summary and then continues with brief
summaries of specific study activities and issues including Institutional Reviews, Population
Identification, Population Tracing, External and Internal Radiation Dosimetry, Chemical
Exposure Assessment, Study Findings, Auxiliary Analyses, Comparisons with the Previous
UCLA Study, and Final Comments. The PowerPoint presentation for the worker meetings 6-8
April 2005 is included at the end of the Executive Summary.

1

 9 of 189
OVERALL SUMMARY. A retrospective cohort mortality study was conducted of 46,970
Rocketdyne workers employed for at least 6 months in either nuclear technology development or
in rocket engine testing since 1948 at the Santa Susana Field Laboratory (SSFL) and at nearby
facilities, including Canoga Park and De Soto Avenue in California. The Rocketdyne workers
were grouped into three populations: those monitored for radiation (Radiation Cohort), those
who worked at SSFL (Chemical Cohort) and those who worked at all other facilities
(Comparison Cohort). The Radiation Cohort consisted of 5,801 workers monitored for radiation
of whom 2,232 were also monitored for internal radionuclide uptake. The Chemical Cohort
consisted of 8,372 workers at SSFL of whom 1,651 were test stand mechanics assumed to have
the greatest potential for exposure to chemicals such as hydrazines and trichloroethylene (TCE).
The Comparison Cohort consisted of 32,979 workers employed at the other Rocketdyne
facilities. There were 182 workers who during their career at Rocketdyne had been monitored
for radiation and also had worked as test stand mechanics. These workers, 30 of whom were
found to have died, are included in both the Radiation and the Chemical Cohorts.
Methods. The Rocketdyne population was identified from Kardex work history cards, electronic
personnel files and radiation dosimetry records. Other personnel records evaluated included
worker transfer lists, medical record index cards, medical records, and personnel lists (phone
directories). Workers were classified by work location, job title, pay type (hourly or salary), and
whether they were monitored for radiation or held an administrative/scientific position. Lifetime
occupational radiation doses were derived from company records of external and internal
exposures and record linkages with national dosimetry datasets. Bioassay data were evaluated
using current International Commission of Radiation Protection (ICRP) biokinetic models to
estimate annual radiation doses for 16 organs or tissues. The estimation of internal radiation
doses accounted for the type of radionuclides taken into the body and their likely chemical
forms, time of exposure, and excretion patterns. The mortality experience of all workers through
1999 was determined by examination of national, state and company records. Observed numbers
of deaths were compared with the number expected in the general population of California
adjusting for age, gender, race and calendar year. Internal cohort dose-response analyses using
Cox proportional hazards models were conducted to evaluate trends over categories of
cumulative radiation dose and over years of potential exposure to chemicals. For the Radiation
Cohort the comparison group for the internal cohort dose-response analyses was in most cases
Rocketdyne workers who were not monitored for radiation. For the Chemical Cohort the
comparison group for the internal cohort dose-response analyses was in most cases Rocketdyne
workers who did not work at SSFL and who were not monitored for radiation. However, various
other referent groups were used in the analyses and any differences were noted.
Overall Results. Overall, the 46,970 Rocketdyne workers (including both radiation and chemical
cohorts together) accrued 1.3 million person-years of observation (average 27.6 years). Vital
status was determined for 99.2% of the workers: 11,118 (23.7%) had died and only 368 (0.8%)
were lost to follow-up. Cause of death was determined for all but 280 (2.5%) of those who had
died. The overall mortality experience among all Rocketdyne workers was lower than that of the
general population of California, i.e., the ratio of observed to expected numbers of deaths (the
Standardized Mortality Ratio, or SMR) was less than 1.0 (SMR 0.87; 95% CI 0.85-0.88). Low
overall mortality was seen among radiation workers (SMR 0.79; 95% CI 0.75-0.83; n=1,468
deaths), SSFL workers (SMR 0.83; 95% CI 0.80-0.86; n=2,251 deaths) and among the other
2

 10 of 189
Rocketdyne workers (SMR 0.90; 95% CI 0.88-0.92; n=7,429). The observed numbers of cancer
deaths also were slightly below population expectation for all workers (SMR 0.93; 95% CI
0.89-0.96; n=3,189 deaths), radiation workers (SMR 0.90; 95% CI 0.82-0.99; n=456 deaths),
SSFL workers (SMR 0.89; 95% CI 0.82-0.96; n=655) and the other Rocketdyne workers (SMR
0.94; 95% CI 0.90-0.98). The ratios of observed to expected deaths (SMRs) computed using
United States rates were lower than those computed using California rates, whereas county rates
(combined Los Angeles and Ventura Counties) were similar to those computed using California
rates. No cause of death was significantly elevated. There were no notable increases in cancer
deaths over time since first hire, or by duration of employment at SSFL or at the other
Rocketdyne facilities.
Overall Radiation Results. Among the 5,801 radiation workers, the mean dose from external
radiation was 13.6 mSv (maximum 1,000 mSv); the mean lung dose from external and internal
radiation combined was 19.1 mSv (maximum 3,600 mSv). Only 69 workers had career doses
from external radiation greater than 200 mSv, and only 111 workers had lung doses greater than
200 mSv when internal doses were considered. Deaths from all cancers taken together (SMR
0.90; 95% CI 0.82-0.99 , n = 456), all leukemia excluding chronic lymphocytic leukemia (CLL)
(SMR 1.16; 95% CI 0.69-1.84; n = 18), and lung cancer (SMR = 0.89; 95% CI 0.76-1.05; n =
151) were not significantly elevated. Internal cohort dose-response analyses revealed no
significant trends over categories of increasing radiation dose for all cancers taken together,
leukemia, lung cancer or any other cancer. There were no significant associations found among
the 2,232 workers who were monitored for internal radionuclide intakes. For all cancers
excluding leukemia, the RR at 100 mSv was estimated as 1.04 (95% CI 0.86 - 1.26) and for all
leukemia excluding CLL it was 1.32 (95% CI 0.71 - 2.45).
Overall Chemical Results. Overall, 1,651 test stand mechanics were identified and assumed to
have the greatest potential exposure to chemicals associated with the testing of rocket engines.
Compared with the general population of California, test stand mechanics had a lower risk of
dying overall (SMR 0.90; 95% CI 0.82-0.98) and a similar risk of dying from cancer (SMR 1.03;
95% CI 0.88-1.20). The mortality experience of the other male hourly workers at SSFL was
similar to that of the test stand mechanics for all causes (SMR 0.97; 95% CI 0.91-1.03), all
cancers (SMR 0.93; 95% CI 0.82-1.06), and all specific cancers. No cancer of a priori interest
among test stand mechanics was significantly increased: lung (SMR 1.07; 95% CI 0.8-1.4),
esophagus (SMR 1.03; 95% CI 0.3-2.4), kidney (SMR 1.78; 95% CI 0.8-3.5), bladder (SMR
0.98; 95% CI 0.3-2.5), liver (SMR 0.97; 95% CI 0.3-2.5), and non-Hodgkin’s lymphoma (SMR
0.80; 95% CI 0.3-1.9). Among the 315 male test stand mechanics with likely exposure to
hydrazines, there were no significant increases for any cancer and, based on internal cohort
analyses, no evidence of a dose response over years of potential exposure for all causes of death
(SMR 0.89, n=101), all cancers taken together (SMR 1.09, n= 33), lung cancer mortality (SMR
1.45, n=15), or any specific cancer. Among the 1,114 workers potentially exposed to TCE, there
were no significant increases for all causes of death (SMR 0.87; 95% CI 0.78-0.96), all cancers
taken together (SMR 1.00; 95% CI 0.83-1.19) or any specific cancer. Based on internal cohort
analyses, there was no significant dose response over years of potential exposure to TCE for all
cancers combined, lung cancer or any other cancer. Cancer of the kidney was elevated based on
7 deaths (SMR 2.22; 95% CI 0.89-4.57) and there was a suggestion of a dose response over
years of potential TCE exposure, although the trend was not significant. For the three
3

 11 of 189
malignancies most frequently found to be elevated in studies of TCE exposure (i.e., cancers of
the kidney and liver and non-Hodgkins lymphoma), the combined SMR based on 12 deaths was
not significantly increased (SMR 1.09; 95% CI 0.56, 1.90).
Questionnaire Survey. A questionnaire survey of 139 workers indicated that hourly workers
(n=66) were significantly more likely than salaried workers (n=71) to have smoked cigarettes
(61% vs 41%; p=0.02). The smoking prevalences of hourly workers who responded to this
survey were also greater than smoking prevalences in the general population of California, and
indicate the need for caution when interpreting comparisons with the general population for
these subgroups because of the likely differences in tobacco use. All test stand mechanics were
hourly workers. National surveys also indicate that blue collar workers smoke cigarettes to a
greater extent than both white collar workers and people in the general population (Lee et al.
2004; Howard 2004; CDC 2004a, 2004b).
Overall Conclusions. The Rocketdyne workforce overall, including those monitored for
radiation, those employed at SSFL and test stand mechanics potentially exposed to hydrazines or
TCE, did not experience a statistically significant increased mortality for any cancer, including
lung cancer, that could be linked to radiation dose, years of employment at SSFL, years of
employment as a test stand mechanic, or years of potential exposure to hydrazines or TCE. No
statistically significant internal cohort dose-response relationship was seen for leukemia,
lymphoma, or cancers of the esophagus, liver, bladder, kidney or any other cancer over
categories of radiation dose or years of potential chemical exposure. We conclude that radiation
exposure has not caused a detectable increase in cancer deaths in this population and that work at
the SSFL rocket engine test facility or as a test stand mechanic is not associated with a
statistically significant increase in cancer mortality overall or for any specific cancer. A slight
non-significant increase in leukemia (excluding CLL) was seen among radiation workers,
although a similar non-significant increase in CLL (a malignancy not associated with radiation)
was also observed. A slight non-significant increase in kidney cancer and a slight nonsignificant decrease in bladder cancer was also seen among radiation workers. Additional
follow-up would be needed to clarify the inconsistent finding with regard to radiation and kidney
cancer (a cancer not generally found increased in radiation exposed populations) as well as the
non-significant association observed for kidney cancer and potential TCE exposure. Additional
follow-up might also clarify the non-significant elevated risk of lung cancer among workers
potentially exposed to hydrazines when compared with the general population.
SPECIFIC STUDY APPROACHES AND ISSUES
The following sections provide summary details of study approaches, methods and
issues, including those raised by the Science Committee during the conduct of the study.
1.
Institutional Review Board (IRB) and other Approvals. To conduct a study involving
human subjects it was necessary to receive approval from a number of IRBs and other Human
Subjects Review committees. Applications were prepared and approvals were received from the
Boeing Company, Vanderbilt University, Oak Ridge National Laboratory, National Center for
Health Statistics (National Death Index), Social Security Administration, Health Care Financing
Administration (now Centers for Medicare & Medicaid Services), University of Southern
4

 12 of 189
California (Cancer Surveillance Program), Nuclear Regulatory Commission, Department of
Energy, and U.S. Air Force.
2.
Identification of the Worker Population. Sources to identify workers and obtain exposure
information included Kardex job history cards, an electronic personnel file, Radiation Safety
folders, personnel listings (phone directories), medical index cards, medical records, and transfer
lists. Excluded from study were those who had worked less than 6 months (6,601) and those
who were not Rocketdyne employees or who had insufficient identifying information for tracing
(813). A cohort of 46,970 eligible workers was developed (Figure 1). There were 5,801 workers
monitored for radiation, 8,372 workers at SSFL (including 1,651 test stand mechanics of whom
182 were also monitored for radiation) and 32,979 workers at other Rocketdyne facilities.
3.
Population Tracing. Vital status was determined for 99.2% of the worker population.
Mortality information was received from the Social Security Administration, California
Surveillance Program, Health Care Financing Administration (CMS), Compserv Inc., PBI,
Rocketdyne records, state vital statistics departments, and the National Death Index. Individuals
were confirmed alive (35,458, 76%) from Rocketdyne personnel and retirement records, the
Social Security Administration and the Health Care Financing Administration (CMS) databases.
There were 11,144 (23.9%) study subjects who were found to have died. Only 368 workers were
lost to follow-up, i.e., 0.8% of all workers (Figure 1).
4.
Cause of Death Determination. Cause of death was sought for all 11,118 workers who
were found to have died in the United States and all but 265 (2.4%) were obtained. Sources of
cause of death information included death certificates available from the Rocketdyne personnel
files, the California Death Tape, the California Surveillance Program, the National Death Index
and death certificates obtained from individual state departments of vital statistics.
5.
Radiation Dosimetry (Figure 2, Figure 3, Figure 4). Organ-specific doses from lifetime
occupational exposure to external radiation and radionuclide intakes were estimated for the
5,801 Rocketdyne/Atomics International workers monitored for radiation and employed for more
than 6 months between 1948-1999. Radiation-related activities included the operation of ten
nuclear reactors and seven criticality test facilities, nuclear fuel fabrication, reactor disassembly,
spent nuclear fuel decladding, laboratory work and storage of nuclear material. The radiation
workforce was identified from the over 14,000 radiation record folders in the Radiation Safety
(Health Physics) offices. Information in the radiation folders was scanned into machine-readable
images and sent to a central location for abstraction and dose assessment. To obtain prior and
subsequent occupational exposure information, the roster of all workers, including those not
known to be radiation workers, was matched against nationwide dosimetry files after
permissions were received from the Department of Energy, the Nuclear Regulatory Commission,
the Landauer Dosimetry Company, the U.S. Army, and the U.S. Air Force. Requests were also
made to investigators of other worker studies to match their dosimetry files against our roster of
Rocketdyne workers. Computation of organ doses from radionuclide intakes was complicated
by the diversity of bioassay data collected over a 40 year period (urine and fecal samples, lung
counts, whole-body counts, nasal smears, and wound and incident reports) and the variety of
radionuclides with documented intake including isotopes of uranium, plutonium, americium,
calcium, cesium, cerium, zirconium, thorium, polonium, promethium, iodine, zinc, strontium,
5

 13 of 189
and hydrogen (tritium). Over 30,000 individual bioassay measurements, recorded on 11
different bioassay forms, were abstracted. The bioassay data were evaluated using current ICRP
biokinetic models to estimate annual doses for 16 organs or tissues taking into account time of
exposure, type of radionuclide, and excretion patterns. A modification of the ICRP respiratory
model for relatively insoluble material was applied to uranium aluminide, and proposed ICRP
models were used for promethium and cerium. Detailed internal exposure scenarios were
developed and annual internal doses were derived on a case-by-case basis for workers with
committed equivalent doses indicated by screening criteria to be greater than 10 mSv to the
organ with the highest internal dose.
The mean cumulative external dose based only on exposures received while employed by
Rocketdyne was 10.0 mSv and the dose distribution was highly skewed (maximum 500 mSv)
(Figure 2). Only 45 workers received greater than 200 mSv while employed at Rocketdyne.
However, 1,833 (or 32%) of the Rocketdyne radiation workers had been monitored for radiation
at other nuclear facilities and incorporation of these doses increased the mean dose to 13.6 mSv
(maximum 1,005 mSv) and the number of workers with >200 mSv to 69 (Figure 4). For a small
number of workers (n=292), lung doses from internal radionuclide intakes were relatively high
(mean 106 mSv; maximum 3,560 mSv), and increased the overall population mean dose to lung
from 13.6 mSv to 19.1 mSv and the number of workers with lung dose >200 mSv to 111. Nearly
10% of the radiation workers (587) were monitored for neutron exposures (mean 1.2 mSv) at
Rocketdyne and another 2% were monitored for neutron exposures elsewhere. These cumulative
neutron dose levels were small in comparison with other external and internal radiation doses.
Without considering all sources of occupational exposure, however, an incorrect characterization
of worker exposure would have occurred with the potential to bias results. For 604 (10%) of the
Rocketdyne workers, the doses received at other facilities both prior to and after employment at
Rocketdyne were greater than the doses received at Rocketdyne. Similarly, a small number of
workers monitored for internal radionuclides contributed disproportionately to the number of
workers with high lung doses. A manuscript describing the dosimetry approach has been
submitted for publication. Another manuscript describing the unique aspects of internal
exposure to uranium aluminide has also been submitted for publication.
6.
Chemical Exposure Assessment (Figure 5). The potential for chemical exposures at
SSFL from 1948 until 1999 was estimated from job history records and work at specific test
stands. The workforce, particularly the test stand mechanics, were potentially exposed to a wide
range of rocket fuels, oxidizers, exhaust gases, solvents and other chemicals. This potential
exposure to a mixture of substances at rocket engine test areas was evaluated in terms of years of
employment at a test stand. Several patterns of potential exposure to specific chemicals were
identified based on the quantity used and the number of workers exposed. These patterns
included hydrazines used as a fuel in some rocket engines, trichloroethylene (TCE) used to clean
(“flush”) engines and TCE as a utility solvent to clean small metal parts. Since these patterns of
exposure existed only at certain test stands during certain rocket engine tests, individual test
stand mechanics had to be placed at specific test stands during specific calendar years to estimate
their potential for exposure. Because the work history information available on Kardex cards
was not specific enough to do this, historical personnel listings (phone books) were relied upon
to make this placement. Confirmation of these assignments was based on information gleaned
from walkthrough surveys at operating and closed test stands with knowledgeable personnel who
6

 14 of 189
were involved with specific engine tests over the years, discussions with over 100 long-term
employees (both retired and active), and existing medical records which often listed the specific
test stands and specific chemicals that the employees worked with.
Overall, 1,651 test stand mechanics were identified within the SSFL workforce of 8,372 (Figure
5). There were 315 test stand mechanics with likely exposure to hydrazines and 205 with
possible exposure to hydrazines. The terms “likely” and “possible but unlikely” were used to
distinguish two levels of potential exposures to hydrazines: “likely” meant we were able to
assign an individual to a test area where hydrazines were used throughout the year whereas
“possible but unlikely” meant that hydrazines were used only in one of several areas within a
large test area and we were unable to distinguish which test stand mechanics worked in those
areas where hydrazines were used from those who worked in areas where hydrazines were not
used. We estimate that less than about 10% of the 205 workers classified as “possible but
unlikely” actually worked with hydrazine and had potential for exposure. There were 1,114
workers with potential exposure to TCE. There were 182 test stand mechanics who also had
been monitored for radiation. The approach to chemical exposure assessment is included in the
draft manuscript on the mortality experience of Rocketdyne workers who tested rocket engines.
7.

Study Findings.

The entire workforce of 46,970 workers had a lower risk of death from all causes
compared to the general population of California (SMR 0.87; 95% CI 0.85-0.88) (Figure 6).
a. Radiation Cohort (Figures 3, 4, 6-8). Overall, 5,801 workers were monitored for
radiation, including 2,232 monitored for radionuclide intakes (Figure 3). The mean dose from
external radiation was 13.6 mSv (maximum 1,000 mSv); the mean lung dose from external and
internal radiation combined was 19.1 mSv (maximum 3,600 mSv). Only 69 workers had career
doses from external radiation greater than 200 mSv and only 111 workers had lung doses greater
than 200 mSv when external uptakes were considered. Vital status was known for 97.6% of the
workers of whom 25.3% (n = 1,473) had died. The average period of observation was 27.6
years. Radiation workers had a lower risk of death form all causes (SMR 0.79) than the general
population of California (Figure 7). All cancers taken together (SMR 0.90, n = 456) and all
leukemia excluding chronic lymphocytic leukemia (CLL) (SMR 1.16, n = 18) were not
significantly elevated. The most frequent cancer deaths were of the lung (SMR = 0.89, n = 151),
colon (SMR = 1.17, n = 47) and prostate (SMR = 0.93, n = 37). Internal cohort dose-response
analyses revealed no significant trends for all cancers taken together (Figure 8), leukemia (Figure
9), lung cancer or any other cancer over categories of increasing radiation dose. Slight positive
dose-response trends were seen for kidney cancer and slight negative trends were seen for
bladder cancer and for cirrhosis of the liver based on the internal dose-response analyses. For all
cancers excluding leukemia, the RR at 100 mSv was estimated as 1.04 (95% CI 0.86 - 1.26) and
for all leukemia excluding CLL it was 1.32 (95% CI 0.71 - 2.45). There were no significant
modifications in the estimates of radiation risk over categories of attained age, age at exposure,
or time since exposure.

b. Chemical Cohort (Figures 5, 10-20). The Chemical Cohort consisted of 8,372
7

 15 of 189
workers at SSFL of whom 1,651 were test stand mechanics with the greatest potential for
chemical exposures, including 182 test stand mechanics who also were monitored for radiation
(Figure 5). The all cause mortality among SSFL workers (SMR 0.83; 95% CI 0.80-0.86;
n=2,251 deaths) and among the other Rocketdyne workers (SMR 0.90; 95% CI 0.88-0.92;
n=7,429) were lower than the general population (Figure 10, Figure 11). The all cancer
mortality SMRs were similar among SSFL workers (SMR 0.89; 95% CI 0.82-0.96) and the other
Rocketdyne workers (SMR 0.94; 95% CI 0.90-0.98). No cause of death was significantly
elevated. There were no notable increases over time since first hire or duration of employment
at SSFL. Test stand mechanics had a lower risk of dying overall (SMR 0.88; 95% CI 0.81-0.95)
and a similar risk of dying from cancer (SMR 1.00; 95% CI 0.86-1.16) compared with the
general population (Figure 12). No cancer was significantly increased (Figure 13, Figure 14).
The SMRs for cancers of a priori interest among test stand mechanics were: lung (SMR 1.07;
95% CI 0.8-1.4), esophagus (SMR 1.03; 95% CI 0.3-2.4), kidney (SMR 1.78; 95% CI 0.8-3.5),
bladder (SMR 1.14; 95% CI 0.4-2.7), liver (SMR 0.89; 95% CI 0.2-2.3), and non-Hodgkin’s
lymphoma (SMR 0.89; 95% CI 0.3-1.9). There were no significant increases for any cancer
among the 474 male test stand mechanics who worked more than 5 years on a test stand. Among
the 315 male test stand mechanics with likely exposure to hydrazines, there were no significant
increases for any cancer and no evidence of a dose response over years of potential exposure for
all causes of death (SMR 0.89, n=101), all cancers taken together (SMR 1.09, n= 33), lung
cancer mortality (SMR 1.45, n=15) or any specific cancer (Figures 15-17). For those who
worked less than or more than 1.5 years with likely hydrazine exposure, the RRs of lung cancer
were 0.74 and 0.70, respectively (Figure 18). It is noted that the RR of lung cancer for test stand
mechanics not exposed to hydrazines was lower than for those with potential hydrazines
exposure, although the difference is not statistically significant. Among the 1,114 workers
potentially exposed to TCE, there were no significant increases for any cause of death (overall
SMR 0.87; 95% CI 0.78-0.96) or for all cancers taken together (SMR 1.00; 95% CI 0.83-1.19)
(Figure 19). There was no significant dose response over years of potential exposure to TCE for
all cancers taken together, lung cancer or any other cancer (Figure 20, Figure 21). Cancer of the
kidney was elevated based on 7 deaths (SMR 2.22; 95% CI 0.89-4.57), although the increase
was not statistically significant. Non-Hodgkins lymphoma and cancers of the kidney and liver,
combined, were not elevated based on 12 deaths (SMR 1.09; 95% CI 0.56-1.90), and there was
no evidence of a dose-response (Figure 22). These three cancers are those most frequently found
to be elevated in studies of TCE-exposed populations.
8.

Other Analyses and Evaluation.
Additional analyses and evaluations were conducted and are summarized below.

a. White Males. Analyses limited only to white males were conducted for all
Rocketdyne workers, the radiation workers and the chemical (SSFL) workers. The observed to
expected ratios for all 43 causes of death evaluated did not differ from those computed for the
entire population, including women and non-white races. White males constitute the majority
(nearly 75%) of all Rocketdyne workers.
b. External Comparison Populations. Although internal cohort dose-response analyses
based on radiation dose or based on years worked with potential exposure to chemicals was the
8

 16 of 189
primary focus of the evaluation of health risks among Rocketdyne employees, external
comparisons with the general population were also made to evaluate patterns of risk over time
and by duration of employment. There were three general populations that could be used for
comparison purposes: the population of California (which we used), the population of the
United States (which was used by the previous investigators from UCLA) and the population of
persons residing in Los Angeles and Ventura Counties. Many Rocketdyne workers had been
born in states other than California and moved to Los Angeles or Ventura County for
employment. Many Rocketdyne workers also left California for employment elsewhere (e.g.,
Washington State, Missouri, Idaho) or for retirement (e.g., Florida). Approximately 25% of
deaths occurred outside of California, 50% in Los Angeles or Ventura Counties and 25% in other
California counties. It is not known where the majority of workers are currently residing or
whether or when they left California. The observed to expected ratios of deaths (i.e., the SMRs),
were similar when comparisons were made with the general populations of California or with
Los Angeles and Ventura Counties. The SMRs were consistently lower when comparisons were
made with the general population of the United States. None of the external populations is ideal
and there are unknown factors such as differences between workers and the general population in
health, occupational exposures and confounding factors (e.g., tobacco use) that cannot be
accounted for in the analyses. However, patterns of risk over time and by duration of
employment can be informative with regard to revealing the presence of any occupational risks.
The most comparable general population to the Rocketdyne Workforce probably lies midway
between the California and United States populations. However, internal comparisons, described
below, are more appropriate when making inferences about workplace exposures and effects.
c. Internal Comparison Populations. Internal cohort dose-response analyses did not rely
upon an external population but rather compared various groups of Rocketdyne workers over
categories of radiation dose or over categories of years worked with potential exposure to
chemicals. All analyses included adjustments for gender, race, age, pay type (hourly or salary),
and most analyses included an adjustment for duration of employment. These internal cohort
analyses are preferred for causal inferences. There were no internal cohort analyses for which
the test for trend in increasing risks over the categories of exposure were statistically significant,
i.e., no trend p-value was <0.05. For most internal cohort dose-response analyses, all
Rocketdyne workers not monitored for radiation were used as the referent category, but other
groups were used as well. For example, dose-response analyses for the Radiation Cohort used
monitored workers with no measured dose as the referent; none of these analyses produced a
significant trend, although slightly lower relative risks at the highest dose categories were seen.
Similarly for the Chemical Cohort, the different referent groups evaluated included all
Rocketdyne workers not monitored for radiation, and all SSFL workers not monitored for
radiation. For test stand mechanics potentially exposed to hydrazines (or TCE), test stand
workers with no known potential exposure to the chemical being evaluated were also used as
referent. None of these analyses produced a significant trend. Using either the Rocketdyne or
SSFL groups as referent produced essentially the same relative risks at the highest dose
categories of years worked with potential exposure to the chemical/s of interest. Using as
referent the relatively small number of test stand mechanics with no years of exposure to the
chemical/s of interest produced no statistically significant trends and all the confidence limits
about the relative risk estimates became wider. The lung cancer risk among the 205 workers
with “possibly but unlikely” exposure to hydrazine was greater than the lung cancer risk among
9

 17 of 189
the 315 workers with “likely” exposure potential.
d. Healthy Worker Effect. The healthy worker effect usually refers to a type of bias that
results from using a general population for comparison with an occupational group. The general
population differs from a working population in ways that are likely to affect the risk of dying.
The bias is related to selection processes that are in force when a worker enters the workforce
and to the health characteristics that enable a worker to continue on the job for many years.
Workers in general are healthier than the general population and as such are less likely to die at a
young age. These selection factors, however, usually diminish over time, especially for deaths
due to cancer. Analyses were conducted excluding the first 10 years of follow-up after a worker
was hired and, while the SMRs rose in general, none was statistically significant and no different
patterns of risk were seen. Similarly, internal cohort dose-response analyses were conducted
excluding the first 10 years of follow-up and no significant trends were seen over categories of
exposure for any cancer or groups of cancers. To learn whether short-term workers had different
patterns of risk over time than workers of longer duration, SMR analyses were conducted. There
were no material differences in the patterns of risk over time whether a worker had been
employed at Rocketdyne for less than 5 years or whether he had been employed for more than 10
years.
e. Radiation Dose Lagging. There is a certain period of time before damaged cells can
develop and be diagnosed as a leukemia or as a cancer. This minimum latency period is usually
taken as 2 years for leukemia and 10 years for solid cancer, i.e., cancers occurring shortly after
radiation exposure are not likely related to the radiation exposure but to other causes. Analyses
were conducted lagging the dose for two years for leukemia and 10 years for solid cancers, i.e.,
any exposures occurring in these time periods before the diagnosis of cancer or end of follow-up
are excluded. Because most of the high doses occurred in the 1950s and 1960s, lagging doses in
the analyses had little effect on the computations of risk over categories of radiation organ dose
or on the significance of the trend tests.
f. Workers Monitored for Radiation Only at Rocketdyne. To evaluate whether radiation
received while employed at Rocketdyne resulted in any adverse health effects, analyses were
restricted to the 3,968 workers who were monitored for radiation only at Rocketdyne and at no
other place of employment. There were no significant elevations in cancer risk or significant
dose-response relations found. The previous investigation did not exclude workers who where
monitored for radiation elsewhere but did exclude the doses received elsewhere. Not including
the relatively large contribution to radiation dose received by the 1,776 (31%) workers who were
monitored for radiation other than at Rocketdyne could be misleading, as 604 (10%) had
received greater doses elsewhere than at Rocketdyne (Figure 4).
g. Smoking Evaluation. To learn more about the smoking habits of hourly and salaried
workers, a brief questionnaire survey was conducted in 2004. A sample of living workers was
selected and approximately half of those mailed a questionnaire responded (68 hourly and 71
salaried workers). Compared to salaried workers, hourly workers were significantly more likely
to have smoked cigarettes (61% vs 41%), to be current smokers (9% vs 0%), to have started
smoking at a younger age, to have quit at an older age, to smoke for more years (31.4 yr vs 21.1
yr) and to have smoked more cigarettes during their lifetime as measured in terms of
“pack-years”. The number of cigarettes smoked each day and the use of other tobacco products,
10

 18 of 189
such as cigars, did not differ significantly between the two groups. The value of the survey is
limited because only survivors are included and the response rate was only 50%. Distinctions
between SSFL workers and the workers at other facilities by pay type were not informative
because of the small numbers, e.g., there were only 29 salaried workers overall who had smoked
cigarettes. Nonetheless, these distributions are consistent with information obtained from a
sample of over 120 medical records of test stand workers; smoking information was available for
over 60 who had completed questionnaires in the 1960s which included queries on cigarette
smoking habits, i.e., just over 60% of the hourly workers were current or former smokers.
National surveys of smoking habits among hourly (blue collar) and salaried (white collar)
workers also indicate a significantly higher prevalence of tobacco use among hourly workers
compared to salaried workers and hourly workers compared to the general population. These
evaluations indicate that caution should be exercised when interpreting comparisons in cancer
risk between hourly workers and salaried workers and between hourly workers and the general
population because of the differences in smoking habits. This is seen in the SMR analyses in
that hourly workers in general had higher rates of death from lung cancer and other smoking
related causes of death such as heart disease and non-malignant respiratory diseases such as
emphysema. It has been suggested that smoking prevention programs should be considered for
blue collar workers (Howard 2004). While patterns of risk in the observed and expected ratios
can be informative, the internal cohort dose-response analyses comparing hourly workers to
hourly workers and salaried workers to salaried workers over categories of exposure are the most
informative with regard to investigating causal associations.
h. Hourly and Salary Workers. The risk of death is usually found to be different
between hourly and salaried workers in occupational studies. As such, all internal dose-response
analyses included an adjustment for pay type, except for those analyses where only specific pay
types were evaluated. Comparisons with the general population were made for hourly workers
and SMRs were generally higher than 1.0 for smoking-related causes of death in comparison
with the general population of California, whereas there were few elevations when comparisons
were made with the general population of the United States. Salaried workers on the other hand
generally had low SMRs for most causes of death. As indicated above in (g), these differences
may be related to differences in the use of tobacco products, although there may be other
reasons. Because the general population differs in many ways from a worker population, use of
internal comparisons is the more appropriate way to evaluate the exposures of interest. There
was no evidence in these analyses that the risk of death from all cancers taken together or for any
specific cancer among hourly workers (or salaried workers) increased with increasing numbers
of years worked at SSFL, or with increasing level of radiation dose, or with increasing numbers
of years with potential exposure to hydrazines, TCE or work as a rocket engine test stand
mechanic.
i. Radiation Dose Response by Pay Type. Several analyses were conducted with regard
to possible radiation associations and pay type classification. One internal cohort dose-response
analysis evaluated the effect of not controlling for pay type and another evaluated the
dose-response over hourly and salaried workers separately. Similar to the overall analyses, not
adjusting for pay type did not change the pattern of cancer risk over categories of radiation dose
and there were no discernible differences in the internal cohort dose-response analyses that were
restricted to either hourly or salaried workers.
11

 19 of 189
j. Chemical Cohort Tables Excluding Radiation Workers. To be consistent with the
previous investigation conducted by UCLA, analyses for the chemical cohort were conducted
excluding the workers who were monitored for radiation. No material difference was seen, in
large part because the number excluded, only 182, was small.
k. Time and Duration Analyses for SSFL and the Other Rocketdyne Workers. SMR
analyses using California rates as referent were conducted for three durations of employment (<5
years, 5-9 years and > 10 years) by three intervals of follow-up after first hire (< 10 years, 10-29
years, and > 30 years) for selected causes of death. For the SSFL hourly male workers, there
were no noticeable patterns. For the other Rocketdyne hourly male workers, there also were no
apparent patterns although lung cancer and non-respiratory lung disease were significantly
elevated in several subgroups and heart disease was generally elevated. When U.S. rates were
used for comparison, there were no significant elevations for any cause of death within any
subgroup. As discussed previously, caution in interpreting the SMR analyses is warranted when
hourly workers, who apparently smoke more than the general population, are evaluated. The
more valid comparisons are the internal cohort evaluations. Internal cohort dose-response
analyses based on years worked at SSFL or years worked at the other Rocketdyne facilities did
not indicate any increasing trends over categories of years worked.
l. Figures. Graphical representations of many of the internal cohort dose-response
analyses for radiation and chemical exposure are presented in addition to the tabular data found
throughout the study documents. These figures provide a visual representation of the number of
cases involved in the analyses as well as variations in the estimates of relative risk. Several of
these figures have been added to this Executive Summary as were the figures presented at the 68 April 2005 worker meetings.
m. Test Stand Workers. Although test stand mechanics were assumed to have the
greatest potential for exposure to chemicals during the testing of rocket engines, there were other
workers at test stands who had some, but much lower, potential for exposure. These included
inspectors, engineers, and instrument mechanics. Analyses were conducted to see whether
elevated cancer rates were apparent among all test stand workers and among test stand workers
excluding the test stand mechanics. There were no discernible differences and no significant
findings. While all test stand mechanics were hourly workers, many of the other test stand
“workers” were salaried workers. Further, it seems likely that the chemical exposures received
outdoors at a test stand were lower than for indoor circumstances because of the dilution and
dispersion that occurs in the open air.
n. Special Groupings of Cancer Sites. For the radiation analyses, groupings of cancer
sites were made to be similar to the previous UCLA investigation, i.e., aerodigestive sites and all
leukemia and lymphomas combined. These groupings, however, are not typical based on
etiologic considerations, i.e., the causes of leukemia differ from the causes of lymphoma. The
National Cancer Institute SEER (Surveillance Epidemiology and End Results) cancer registries
also do not use such categories. A recent exchange of letters on the issue of lumping leukemias
and lymphomas together appeared in the January 2005 issue of the American Journal of
Epidemiology (Poole et al. 2005; Lee et al. 2005). Regardless, there were no associations with
radiation dose based on internal cohort dose-response analyses for any of these groupings.
12

 20 of 189
o. Asbestos. There was no evidence of significant/heavy excessive exposure to asbestos
for any of the worker cohorts. Observed numbers of mesothelioma deaths and deaths due to
cancer of the pleura and peritoneum did not differ from the expected numbers of deaths in these
categories.
p. Beryllium. There was no evidence of excessive exposure to beryllium for any of the
worker cohorts. Only one death certificate had mention of berylliosis.
q. Trend Tests. Trend tests for all internal cohort radiation dose-response analyses
were conducted by entering the individual cumulative radiation dose as a continuous measure
into a Cox proportional hazards model along with the exact same set of covariates used in the
corresponding categorical dose analysis. This continuous measure of dose was the actual
radiation dose value received by each individual worker. From the Cox model, a single estimate
of risk was calculated for this continuous measure and the p-value from a Wald chi-square test
was presented in the tables as the ‘p for linear trend.’ Thus, the individual dose values and not
group values are used to calculate the trend test. Trend tests were conducted in similar manner
described above for the internal cohort dose-response analyses with years worked taken as the
continuous variable of exposure. However, there was one exception. For the tables with
hydrazines exposure broken down by “potential” and “possible but unlikely”, ordinal values
were used for the independent variable. The ordering was based on a logical ranking of the
potential for hydrazine exposure among workers in each category. Linear trend tests are used in
most of the evaluations and point and interval estimates are also presented for each category in
each interval dose-response table. Use of a linear trend test in radiation studies is standard
procedure, especially in studies of low dose exposures.
9.

Comparisons with UCLA Study.

a. Radiation Cohort. Our radiation cohort differs in several ways from the earlier UCLA
study (Ritz et al. 1999b, 2000; Morgenstern and Ritz, 2001). We included all workers (men and
women) who were hired up to 1999 and followed through December 31, 1999; the previous
cohort accrual stopped December 31, 1993 and follow-up was through December 31, 1994. The
current study included workers employed for at least six months at Rocketdyne whereas the
previous investigation included anyone monitored for radiation, including short-term workers.
We excluded workers not employed at Rocketdyne, i.e., contract workers and visitors. For 617
workers with only a radiation folder and not a Kardex or electronic job history, we were able to
include 332 workers who had both a Rocketdyne serial number and sufficient identifying
information for tracing. Additional data sources that we used to confirm and obtain employment
histories included over 50,000 medical index cards, detailed dosimetry files, worker transfer lists
and employment personnel listings (telephone directories). The previous investigation excluded
workers without a personnel work history or identifying information. These differences resulted
in our radiation cohort being larger by 1,194 (25.9%) workers than the previous study, i.e., 5,801
workers compared to 4,607 workers (Morgenstern and Ritz 2001).
Our study expanded and extended the previous UCLA investigation by five years and did
not find significant associations with radiation dose for lung cancer hemato- and lymphopoietic
cancers or aerodigestive cancers (Morgenstern and Ritz 2001; Ritz et al. 1999b, 2000). The
13

 21 of 189
previous investigators recognized the small size of the population studied and the low
occupational doses received and concluded that their findings would have to be confirmed by
other studies and/or further follow-up of the Rocketdyne workforce (Morgenstern and Ritz 2001;
ATSDR 1999). The differences in findings between the two studies are likely related to the
additional years of follow-up, coupled with differences in study design and the approach to dose
assessment. The number of workers monitored for internal radiation (2,232 vs. 2,297) was
similar but the number of externally monitored workers (5,743 vs 4,563) was appreciably larger
in our investigation. The expanded numbers and longer follow-up (161,605 person-years vs
about 119,100 person-years) resulted in an additional 593 deaths from all causes (a 67.8%
increase) and an additional 198 deaths from all cancers (a 76.7% increase). Another difference
was that the previous investigation was not able to include the occupational doses accumulated
by 1,776 (31%) of the workers at places of employment other than at Rocketdyne. The dose
received elsewhere by 604 (10.4%) workers was greater than the dose received at Rocketdyne.
Further, we were able to compute radiation doses from the intake of radionuclides for specific
organs and did not use lung dose as a surrogate for dose to all organs. Finally, different
analytical methods were used in that the previous analyses used logistic regression whereas we
used Cox proportional hazards methods (Callas et al. 1999).
b. Chemical Cohort. Our SSFL cohort also differs in several ways from the one
previously reported (Ritz et al. 1999a; Morgenstern and Ritz, 2001). We included all workers
(men and women) who were hired up to 1999, whereas the previous cohort included only men
and accrual stopped in 1980. We included test stand workers who worked on the Peacekeeper
missile system from 1979 to about 1999 and who had potential exposure to hydrazines as well as
other chemicals. We included 182 test stand workers who were also monitored for radiation
whereas they were excluded in the previous investigation. We identified additional workers for
study from transfer lists, personnel listings (phone books) and medical record index cards, and
then sought their Kardex work histories. The current study included workers employed for at
least six months at SSFL whereas the previous investigation required that a worker spend at least
two years at any Rocketdyne/Rockwell division with apparently no minimum time restriction for
work at SSFL. These differences resulted in our cohort of SSFL workers being 37.1% larger
than the previous cohort, i.e., 8,372 workers compared with 6,107. In addition, the previous
study did not estimate potential exposure to TCE and assumed all test stand mechanics were
potentially exposed to hydrazines. We were able to make more refined estimates of exposure
potential to both TCE and hydrazines. We determined that the percentage of test stand
mechanics with potential exposure to hydrazines was between 19-33%, depending on how we
classified the workers with regard to likely or “possible not unlikely” exposure potential. Over
65% of the test stand mechanics were unlikely to have been exposed to hydrazines to any
appreciable degree.
Our study, expanded with a larger population and 5-year increase in follow-up, did not
find a significant association between lung cancer and exposure to hydrazines as previously
reported (Morgenstern and Ritz 2001; Ritz et al. 1999a). The previous investigators recognized
the small size of the hydrazine-exposed population studied and concluded that their findings
would have to be confirmed by other studies and/or further follow-up of the Rocketdyne
workforce (Morgenstern and Ritz 2001; ATSDR 1999). Our larger numbers of workers and
longer follow-up (248,849 person-years vs about 171,100 person-years) resulted in an additional
14

 22 of 189
830 deaths from all causes (a 59.6% increase), an additional 243 cancer deaths (a 60.1%
increase), and an additional 66 deaths from lung cancer (a 45.2 % increase) among SSFL
workers. We found little evidence that work as a test stand mechanic during the 1960s was
related to an increased risk of lung cancer. Finally, we did not limit our investigation only to
workers at SSFL, but included the 32,979 workers employed at nearby Rocketdyne facilities as
an additional comparison or referent group, enhancing the statistical power of the internal doseresponse analyses.
The previous investigation also reported an association between hydrazines and all
lymphatic and hematopoietic malignancies taken together (including CLL) based on 41 deaths
which was not seen in our extended follow-up with 67 deaths. Such an aggregated category, as
discussed above in (n), is not generally examined because the component malignancies, i.e.,
Hodgkin’s disease, non-Hodgkin’s lymphoma, multiple myeloma, and myelogenous leukemia,
have such different etiologies (Poole et al. 2005; Lee et al. 2005).
10.

Final Comments.

Every epidemiologic study has strengths and weaknesses and the Rocketdyne Health
Study is not an exception. The limitations of the study include the relatively low exposures to
radiation and chemicals which limits the ability to detect increased risks. The number of cancer
deaths can determine whether a study has the ability to detect a statistically significant increase.
Studies involving small numbers are not as powerful as studies with large numbers. Small
numbers result in estimates of risk that are very imprecise which means that chance often cannot
be ruled out as an explanation for the findings. This does not mean that there was no increase in
risk, just that the ability of the study to detect the risk was limited. Mortality and not incidence
or illness was evaluated. Chemical exposure could be evaluated only as “potential” since few
measurements were made in the early years. Lifestyle factors such as diet and tobacco use were
also not known.
On the other hand the study has several strengths. Multiple data sources were used to
identify the worker population of whom 99.2% were located. Radiation exposure assessment
was comprehensive and included obtaining doses before and after Rocketdyne tenure. The
assessment of organ doses from internal intakes of radionuclides used state-of-the-art
methodologies. Chemical exposure assessment was facilitated by knowledge of test stand
assignment and chemical use which enabled a more accurate assessment of hydrazines and TCE.
Auxiliary analyses were conducted to augment and support the main analyses, including
comparisons with other workers at Canoga Park and other local Rocketdyne facilities.
The radiation dose distribution for workers is relatively low and much lower than in other
studies where effects are clearly evident. The numbers exposed to “high” doses of radiation are
small, as are the numbers of workers “potentially” exposed to hazardous test stand chemicals.
The exposure assessment problem for the chemicals is recognized, which necessitates having to
use “years worked” as a surrogate for actual exposure. Further, chemical exposures at a test
stand occurred outdoors where concentrations were likely diluted. Attempts to improve the
exposure assessment to radiation or to chemicals are unlikely to yield an appreciable
improvement. Additional investigation of potential confounding influences, such as tobacco use,
15

 23 of 189
would likely be unproductive because of the absence of any significant increases over categories
of radiation dose to lung or over categories of years worked as a test stand mechanic. Further,
obtaining accurate and valid smoking information would be difficult for those who have died,
where surrogate responses from spouses or children many years after the fact would have to be
obtained. Finally, the number of cancers for some sites of potential interest, such as kidney, are
small and generally less than 10 and not amenable for meaningful case-control evaluation. Thus,
the small numbers of workers in the study, the relatively low exposures to radiation and test
stand chemicals, and the absence of any significant or consistent excesses argues at this time
against the need for a nested case-control investigation.
The Rocketdyne workers are an aging population with the median age in 1999 being just
over 60 years overall and nearly 70 years for test stand mechanics. Through 1999, 23.7%
(11,118) of the Rocketdyne workforce had died. Based on age and current mortality patterns, an
additional 5,000 workers would be expected to die by the end of 2005, including approximately
700 radiation workers and 1,000 SSFL workers. An additional mortality follow-up through 2005
would result in a much more powerful evaluation of the potential risk from work in nuclear
technology development and work at rocket engine test facilities. Any inconsistencies in the
current data could be resolved with further follow-up and suggestive patterns could be clarified,
such as the non-significant increase in leukemia among radiation workers, the non-significant
increase in lung cancer among hydrazine-exposed workers, and the non-significant increase in
kidney cancer among TCE-exposed workers. Because there were no significant increases seen
in the cohort internal dose-response evaluations, however, there seems little justification to
consider nested case-control studies at this time.
11.

Manuscripts.

Dosimetry Paper
Boice JD Jr, Leggett RW, Dupree BE, Wallace P, Mumma M, Cohen SS, Brill AB, Chadda B,
Boecker B, Yoder RC, Eckerman KF. A comprehensive dose reconstruction methodology for
former radiation workers. Health Phys (Submitted)
Uranium Aluminide Paper
Leggett RW, Eckerman KF, Boice JD Jr. A respiratory model for uranium aluminide based on
occupational data. J Radiol Prot (Submitted)
Radiation Epidemiology Paper
Boice JD Jr, Cohen SS, Mumma MT, Dupree Ellis E, Eckerman KF, Leggett RW, Boecker BB,
Brill AB, Henderson B. Mortality among Rocketdyne/Atomics International workers monitored
for radiation 1948-1999. (In Draft)
SSFL / Chemical Epidemiology Paper
Boice JD Jr, Marano D, Cohen SS, Mumma MT, Blot WJ, Brill AB, McLaughlin JK, Henderson
B. Mortality among Rocketdyne workers who tested rocket engines, 1948-1999. (In Draft)

16

 24 of 189
12.

Papers Cited.

ATSDR. Draft Preliminary Site Evaluation. Santa Susana Field Laboratory (SSFL), Ventura
County, California. CERCLIS NO. CAD074103771, Division of Health Assessment and
Consultation, Agency for Toxic Substances and Disease Registry, Atlanta, Georgia, December 3,
1999. Available at http://www.atsdr.cdc.gov/HAC/PHA/santa/san_toc.html (last accessed 26
October 2004).
Callas PW, Pastides H, Hosmer DW. Empirical comparisons of proportional hazards, poisson,
and logistic regression modeling of occupational cohort data. Am J Ind Med 33:33 47, 1998.
Howard J. Smoking is an occupational hazard. Am J Ind Med 46:161 169, 2004.
Lee DJ, LeBlanc W, Fleming LE, Gomez-Marin O, Pitman T. Trends in US Smoking rates in
occupational groups: the National Health Interview Survey 1987 1994. J Occup Environ Med
46:538 48, 2004.
Lee WJ, Hoppin JA, Blair A, Lubin JH, Dosemeci M, Sandler DP, Alavanja MCR. Re: "cancer
incidence among pesticide applicators exposed to alachlor in the agricultural health study". Am
J Epidemiol 161:102-103, 2005.
Morgenstern H, Ritz B. Effects of radiation and chemical exposures on cancer mortality among
Rocketdyne workers: a review of three cohort studies. Occup Med 16:219-237, 2001.
Poole C, Cullen M, Irons R, Acquavella J. Re: "cancer incidence among pesticide applicators
exposed to alachlor in the agricultural health study". Am J Epidemiol 161:101-102, 2005.
Ritz B, Morgenstern H, Froines J, Moncau J. Chemical exposures of rocket engine test stand
personnel and cancer mortality in a cohort of aerospace workers. J Occup Environ Med 41:903910, 1999a.
Ritz B, Morgenstern H, Froines J, Young BB. Effects of exposure to external ionizing radiation
on cancer mortality in nuclear workers monitored for radiation at Rocketdyne/Atomics
International. Am J Ind Med 35:21-31, 1999b.
Ritz B, Morgenstern H, Crawford-Brown D, Young B. The effects of internal radiation exposure
on cancer mortality in nuclear workers at Rocketdyne/Atomics International. Environ Health
Perspect 108:743-751, 2000.
13.

Summary Charts and Figures
Figure 1.
Figure 2.
Figure 3.
Figure 4.

Vital Status of Rocketdyne Workers
External Radiation Dose Distribution
Radiation Cohort
Comparing Radiation Dose Received Only at Rocketdyne with
Total Dose Received at All Facilities
17

 25 of 189
Figure 5.
Figure 6.
Figure 7.
Figure 8.
Figure 9.
Figure 10.
Figure 11.
Figure 12.
Figure 13.
Figure 14.
Figure 15.
Figure 16.
Figure 17.
Figure 18.
Figure 19.
Figure 20.
Figure 21.
Figure 22.

SSFL (Chemical) Cohort
Entire Rocketdyne Workforce Compared to General Population
of California
Radiation Workers Compared to General Population of California
Radiation Dose Response for All Cancer Excluding Leukemia
Radiation Dose Response for Leukemia
SSFL Workers (Chemical Cohort) Compared to the General22
Population of California
Dose Response for All Cancers Combined by Years Worked
at SSFL
Test Stand Mechanics Compared to the General Population of
California
Dose Response for All Cancers Combined by Years Worked as
a Test Stand Mechanic
Dose Response for Lung Cancer by Years Worked as a Test
Stand Mechanic
Classification of Potential Exposure to Hydrazines Among Test
Stand Mechanics Based on Job Title & Test Stand
Test Stand Mechanics Potentially Exposed to Hydrazines
Compared to California Population
Dose Response for All Cancers Combined for Test Stand
Mechanics with Potential Exposure to Hydrazines
Dose Response for Lung Cancer for Test Stand Mechanics with
Potential Hydrazines Exposure
Test Stand Mechanics Potentially Exposed to TCE Compared
to the California Population
Dose Response for All Cancers Combined for Test Stand
Mechanics with Potential Exposure to Trichloroethylene (TCE)
Dose Response for Lung Cancer for Test Stand Mechanics with
Potential Exposure to TCE
Dose Response for TCE Suspected Cancers* for Test Stand Mechanics
with Potential Exposure to TCE

18

 26 of 189

Vital Status of Rocketdyne Workers

 

 

35,042 Kardex Cards 26,136 Electronic File 14,190 Radiation Folders*
19?1) (1971?99) (1948? 99)

 

 

i *most not monitored for
internal or external radiation

 

54,384
Unique Workers

 

 

6,601 Short
Term 6 mo.)

 

 

 

45,970 813 Insuf?cient Identifying
Eligible Workers Information or Not RD

Employee

46,602 Known 26 Died 368 

Vital Status Not in Us. Lost to Followiup

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 

 

35,458 2 17
Alive Known SSN


 

 

 

 

 

 

 

 

 

 

Figure 1

External Radiation Dose Distribution

 

3,803

>1049 50-99 iDD-igg 2 200

External Radiation Dose [mSvi
1 100 mrem

 

 

 

Figure 2

19

27 of 189

Radiation Cohort

 

 

 

Total in Group: th I 
erna

 and External

Exposure:
2,174

Only External
Exposure: 3,569 

 

 

 

 

 

 

?l

Only Internal
Exposure: 58

 

 

 

 

 

*182 workers in the radiation group also worked on test stands

 

Figure 3

Comparing Radiation Dose Received Only at Rocketdyne
with Total Dose Received at All Facilities

 

Number anorkers
Number ufWurkeIs

<5 5- - so 100 2200
External Dose (mSv) External Dose (mSv)

 

I Dose Only at Rocketdyne

Total Dose, including pre- and 
post-employment at Rocketdyne 1 100 mrem

 

 

 

 

 

Figure 4

20

28 of 189

SSFL (Chemical) Cohort

 

Test Stand Mechanics: 1,651

Potential
Hy drazin 
Exposure



N0 TCE Potential No
Exposure: TCE Hydrazine
537 Exposure: Exposure:
1,1 14 1,127

 

 

Possible
but
Unlikely
Hydrazine
Exposure

 

 

*182test stand workers also included in the radiation group

 

 

Figure 5

Entire Rocketdyne Workforce Compared to General
Population of California
1

11,118
#12816

All Non?Cancer
Causes

NI Cancers

Lung Cancer
El Observed

Liver Cancer

Expected based on
mortality rates for state of
California

Leukemia

4000 6.000 8.000 10.000 12.000 14000
Number of Deaths

 

 

 

Figure 6

21

29 of 189

Radiation Workers Compared to General Population of
California

1

1468
causes H1870

All Non-Can cer Causes

All Cancers

Observed

 

Lung Cancer

I Expected - based on
mortality rates for state
of California

18
16

Leukemia not CLL

200 400 600 800 1000 1200 1400 1600 1800 2000
Observed and Expected Number of Deaths

 

 

 

Figure 7

Radiation Dose Response for All Cancer Excluding Leukemia

3.0

 

 

. Relative Risk 13 cancers

among 100

5 cancers
workers

among 63
workers

95% Con?dence Limits

 

 

 

2?535 cancers ?cancers 93 cancers
amon 41?169 253 cancers
among 3323 among 501 among 949 Scancers
workers workers workers among 150

workers


Relative Risk

Not Monitored 5?9 10?49 50?99 100?199 3200

 

 

 

External Radiation Dose (mSv)

1 mrem

10 year lag

 

 

 

Figure 8

22

Radiation Dose Response for Leukemia

 

2 cancers
1 cancerin the 50?99 range among 334
cancers in the 100 199 range workers

1 cancer in the 200 range

I Relative Risk

95% Con?dence Limits

 

 

1 cancer
among 253
workers

2 cancers
among 535
workers

4 cancers
among 993
workers

Relative Risk

9 cancers
among 3.085
80 cancers workers
among 41,169
workers

 

 

I


 

Not Monitored

Dose to Bone Marrow (mSv)

1 100 mrem
Excludes chronic leukemia

 

 

 

 

Figure 9

SSFL Workers (Chemical Cohort) Compared
to the General Population of California

 

 

2251
All Causes

 

All on?Cancer Causes

All Cancers

Observed

 

Lung Cancer

I Expected - based on
mortality rates for

23 . .
Leukemia 27 state Of California

 

 

15oo 25oo
Dbserv ed and Expected Number of Deaths

 

 

 

Figure 10

23

30 of 189

310f189

Dose Response for All Cancers Combined byYears
Worked at SSFL

 




 

 

I Relative Risk

- - 405 cancers
95% fd Is
on I ence "m among 5,637 204 cancers 

workers among 2,19?r workers

 

 

 

2,086 cancers
among 32,9?9
workers

1:
0


(U





310Referent (Other than 5-14 yr

SSFL Workers
Years Worked at SSFL

 

 

 

Figure 11

Test Stand Mechanics Compared to the General
Population of California

 

1
All Causes

 

All Non?Cancer Causes

All Cancers

Lung Cancer

Observed
Kidney.r Can cer

I Expected - based on mortality
rates for state of California

Leukemia

 

100 200 300 400 500 600 7?00
Observed and Expected Number of Deaths

 

 

 

Figure 12

24

32 of 189

Dose Response for All Cancers Combined by
Years Worked as a Test Stand Mechanic

 

 

35 cancers

among 358
workers 81 cancers 58 cancers

95% Con?dence Limits among EDD among an
workers workers

I Reiative Risk

151 cancers
among 1,598
workers



Relative Risk (All Cancer)

 

 

Referent (SSFL Not a Test 1.4 yr
Stand Mechanic) Years Worked as a Test Stand Mechanic

 

 

 

 

Figure 13

Dose Response for Lung Cancer by Years Worked as a Test
Stand Mechanic

 

 

 

1D cancers 31 cancers 22 cancers
. Ralame Risk among 368 among EDD among 474
workers workers workers

95% Confidence Limits

 

 

 

59 cancers
among 1,598
workers

Relative Risk (Lung Cancer)

 

 

Referent (SSFL Not a Test 1-4 yr 2 5 yr
Stand Mechanic)

 

Years Worked as a Test Stand Mechanic

 

 

 

Figure 14

25

33 of 189

Classi?cation of Potential Exposure to Hydrazines Among Test Stand
Mechanics Based on Job Title Test Stand

 

 

 

 

 

 

Referent (SSFL) None Possible but <15 yr 2 1.5 yr
Not Test Stand Unlikely?
Mechanic Years of Potential Exposure
to Hydrazines

*Most workers (>900fo) did not work with hydrazines but could not be distinguished

 

 

 

Figure 15

Test Stand Mechanics Potentially Exposed to
Hydrazines Compared to California Population

 

 

All Causes

 

All Non?Cancer Causes

All Cancers
Observed

 

Lung Cancer

I Expected - based on
mortality rates for
state of California

2
1

Kidney Cancer

 

20 40 E30 80 100 120
Observed and Expected Number of Deaths

Does not include workers with ?possible but unlikely" hydrazine exposure

 

 

 

Figure 16

26

34 of 189

Dose Response for All Cancers Combined for Test Stand
Mechanics with Potential Exposure to Hydrazines

 

17 r?:an

24 cancers a?101191513
. Relative Risk among 205 
workers 15 cancers

95% Con?dence Limits among 159
workers

 

92 cancers

151 cancers among 920

among 1598

Relative Risk (All Cancers)

 

 

 

Reierert (SSFL) N0 Hydrazines Possible but 1.5 yr 21.5 yr
Not Test Stand Unlikely
Mechanic Years of Potential Exposure to
l-lydrazines

 

 

 

Figure 17

Dose Response for Lung Cancer for Test Stand Mechanics with
Potential Hydrazines Exposure

 

 

19 cancers cancers

among 295 among 156 9 cancers
Relatiye Risk workers workers among 159
workers

 

95% Con?dence Limits

 

 

 

99 cancers
59 cancers among 929
among 1,599 workers
workers



Referent (SSFL) Not No Hydrazines Possible but Unlikely 1.5 yr 2 1.5 yr
Test Stand
Mechanic

 

Years of Potential Exposure to Hydrazines

 

 

 

Figure 18

27

35 of 189

Test Stand Mechanics Potentially Exposed to TCE
Compared to the California Population
391

All Causes

451
270

All Non-Cancer Causes

330
121
121

All Cancers
51
41

Lung Cancer

Kidney Cancer*

7
3

Non-Hodgkins Lymphoma*

1
5

Liver Cancer*

4
3

Observed
Expected - based on
mortality rates from state
of California

12
11

Kidney, Liver, NHL*
0

50

100

150

200

250

300

Number of Deaths

Figure 19

Figure 20

28

350

400

450

500

 36 of 189

Dose Response for Lung Cancer for Test Stand
Mechanics with Potential Exposure to TCE
2.0

59 cancers
among 1,598
workers

7 cancers
among 329
workers

27 cancers
among 695
workers

24 cancers among
416 workers

< 4 yr

= 4 yr

1.0

0.0
Referent (SSFL) Not Test
Stand Mechanic

No TCE

Years of Potential Exposure to TCE

Figure 21

Dose Response for TCE Suspected Cancers* for Test
Stand Mechanics with Potential Exposure to TCE
5.0

Relative Risk (All Cancer

Relative Risk
95% Confidence Limits

4.0

8 cancers
among 695
workers

3 cancers
among 329
workers

4 cancers
among 416
workers

3.0

2.0

13 cancers
among 1,598
workers
1.0

0.0
Referent (SSFL) Not Test
Stand Mechanic

No TCE

< 4 yr

= 4 yr

Years of Potential Exposure to TCE

*Kidney, NHL, Liver

Figure 22

29

 37 of 189
14.

PowerPoint Presentation 6-8 April 2005
Figure 1pp.
Figure 2pp.
Figure 3pp.
Figure 4pp.
Figure 5pp.
Figure 6pp.
Figure 7pp.
Figure 8pp.
Figure 9pp.
Figure 10pp.
Figure 11pp.
Figure 12pp.
Figure 13pp.
Figure 14pp.
Figure 15pp.
Figure 16pp.
Figure 17pp.
Figure 18pp.
Figure 19pp.
Figure 20pp.
Figure 21pp.
Figure 22pp.
Figure 23pp.
Figure 24pp.

Figure 25pp.
Figure 26pp.
Figure 27pp.

Overview
Who was in the study?
What were the two types of radiation exposure?
How many people were in the radiation group?
Potential chemical exposure characterized by years worked
Nine discussion sessions
How many SSFL workers were potentially exposed to chemicals as
test stand mechanics?
Worker Groups
Rocketdyne workers had a lower risk of death than the general
population of California
Rocketdyne radiation workers had a lower risk of death than the
general population of California
Most radiation workers received very low exposures
What was the effect of including pre- and post-Rocketdyne radiation
dose?
Interpreting Dose Response Graphs
No evidence that radiation increased the risk of dying from cancer
(excluding leukemia)
No evidence that radiation increased the risk of dying from lung
cancer
Suggestive, although not statistically significant, evidence that
radiation increased the risk of dying from leukemia
Radiation Summary Findings
SSFL workers (Chemical Group) had a lower risk of death than the
general population of California
No evidence that working at SSFL increased the risk of dying from
all cancers combined
Test stand mechanics had a lower risk of death than the general
population of California
No evidence that working as a test stand mechanic increased the risk
of dying from all cancers combined
No evidence that working as a test stand mechanic increased the risk
of dying from lung cancer
Classification of potential exposure to hydrazines among test stand
mechanics based on job title and test stand
Test stand mechanics potentially exposed to hydrazines had a lower
risk of death overall but slight increased risk of dying from cancer
compared to the general population of California
No evidence that test stand mechanics with potential exposure to
hydrazines had an increased risk of dying from all cancers combined
Little evidence that test stand mechanics with potential exposure to
hydrazines had an increased risk of dying from lung cancers
Classification of potential exposure to trichloroethylene (TCE)*
among test stand mechanics based on job title and test stand
30

 38 of 189
Figure 28pp. Test stand mechanics potentially exposed to TCE had a lower risk of
death overall but similar risk of dying from cancer compared to the
general population of California
Figure 29pp. No evidence that test stand mechanics with potential exposure to TCE
had an increased risk of dying from all cancers
Figure 30pp. Chemical Summary Findings
Figure 31pp. Limitations
Figure 32pp. Strengths
Figure 33pp. Conclusion

31

 39 of 189

Rocketdyne Follow-On Health Study
6-8 April 2005
Overview

Figure 1pp

Who was in the study?
54,384
Rocketdyne Workers
from 1948-1999
6,601 Short
Term (< 6 mo.)
46,970
Eligible Workers

5,801
Radiation
Group *

8,372
Chemical Group
(SSFL)*

813 Insufficient
Identifying Information or
Not Employee

32,979
Comparison Group
(Canoga Park etc)

99.2% of eligible workers as of 12/31/99 were traced
*182 workers included in both groups

Figure 2pp

32

 40 of 189

What were the two types
of radiation exposure?
External

Internal

Uniform dose
Delivered during exposure
Film (TLD) badge reading

Non uniform dose
Protracted in time
Bioassay measurements

Figure 3pp

How many people were in the
radiation group?
Total in Group:
Both Internal
and External
Exposure:
2,174

5,801
Only External
Exposure: 3,569

Only Internal
Exposure: 58
*182 workers in the radiation group also worked on test stands

Figure 4pp

33

 41 of 189

Potential Chemical Exposure
Characterized by Years Worked
• Work at SSFL
• Work as Test Stand Mechanic
– Exposure to “Test Stand Environment”,
including chemical mixture of fuels, oxidizers,
exhaust gasses, solvents and other chemicals
– Hydrazines
– TCE as a “Utility Solvent”
– TCE as a “Flush Solvent”

Figure 5pp

Nine Discussion
Sessions

Figure 6pp

34

 42 of 189

How many SSFL workers were potentially exposed
to chemicals as test stand mechanics?

Figure 7pp

Worker Groups

Number of Workers

40,000

32,979

30,000

20,000
8,372

5,801
10,000

0
Radiation Group Chemical (SSFL)
Group

Figure 8pp

35

Other Workers

 43 of 189

Rocketdyne workers had a lower risk of death than
the general population of California
11,118

All Causes

12,816

All Non-Cancer
Causes

7,929
9,368
3,189
3,448

All Cancers
1,068
1,098

Lung Cancer

Observed

61
90

Liver Cancer

Expected - based on mortality
rates for California

124
126

Leukemia
0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

Number of Deaths

Figure 9pp

Rocketdyne radiation workers had a lower risk of
death than the general population of California

1,468

All Causes

1,870

All Non-Cancer
Causes

1,012
1,365

456
505

All Cancers

151
169

Lung Cancer
Leukemia
not CLL

Observed
Expected - based on mortality
rates for California

18
16
0

200

400

600

800

1000

1200

Number of Deaths

Figure 10pp

36

1400

1600

1800

2000

 44 of 189

Most radiation workers received
very low exposures
3,803
4,000

No. of Employees in Study

3,500
3,000
2,500
2,000
1,500

1012
651

1,000

166

500

100

69

0
<5

5-9

10-49

50-99

100-199

= 200

External Radiation Dose (mSv)
1 mSv = 100 mrem

Figure 11pp

What was the effect of including preand post-Rocketdyne radiation dose?

Number of Workers

180

Dose Only at Rocketdyne

160

Total Dose, including pre- and
post-employment at Rocketdyne

140
120
100
80
60
40
20
0
50-

100-

External Dose (mSv)
1 mSv = 100 mrem

Figure 12pp

37

_ 200
>

 45 of 189

Interpreting Dose Response Graphs
Increasing RR - Noteworthy

Results are presented with the
Confidence Interval:

Comparison
Group

• The confidence interval is the
range of possible Relative Risk
(RR) values.

Flat RR – No Association

• A Confidence Interval that
does not contain 1.0 is
statistically significant.

Comparison
Group

Dose

Dose

Upper Confidence Limit
Decreasing RR

Relative Risk (RR) Value

Comparison
Group

Lower Confidence Limit
Dose

Figure 13pp

No evidence that radiation increased the risk of
dying from cancer (excluding leukemia)
3.0
13 cancers
among 100
workers

Relative Risk

Relative Risk

95% Confidence Limits

5 cancers
among 63
workers

2.0

2,635 cancers
among 41,169
workers

258 cancers
among 3,928
workers

54 cancers
among 601
workers

93 cancers
among 949
workers

8 cancers
among 160
workers

1.0

0.0
Not Monitored
10 year lag
1 mSv = 100 mrem

<5

5-9

10-49

50-99

External Radiation Dose (mSv)

Figure 14pp

38

100-199

= 200

 46 of 189

No evidence that radiation increased the risk of
dying from lung cancer
4.0

5 cancers
among 102
workers

Relative Risk
95% Confidence Limits

Relative Risk

3.0

2.0
917 cancers
among 41,169
workers

96 cancers
among 3,852
workers

17 cancers
among 561
workers

28 cancers
among 976
workers

5 cancers
among 310
workers

5-9

10-49

50-199

1.0

0.0
Not Monitored

<5

= 200

Dose to Lung (mSv)

10 year lag
1 mSv = 100 mrem

Figure 15pp

Suggestive, although not statistically
significant, evidence that radiation increased the
risk of dying from leukemia
14
13

Relative Risk

12

95% Confidence Limits

1 cancer in the 50 – 99 range
0 cancers in the 100 – 199 range
1 cancer in the > 200 range

2 cancers
among 334
workers

11
10

Relative Risk

1 cancer
among 753
workers

9
8

2 cancers
among 636
workers

7

4 cancers
among 993
workers

6
9 cancers
among 3,085
workers

5
4
3
2

80 cancers
among 41,169
workers

1
0
Not Monitored

0

1 mSv = 100 mrem
Excludes chronic lymphocytic leukemia

> 0-4

5-9

Dose to Bone Marrow (mSv)

Figure 16pp

39

10-49

= 50

 47 of 189

Radiation Summary Findings
Radiation exposure has not caused a detectable
increase in cancer deaths among Rocketdyne
workers
– Mean dose was low
– There were no significant trends between radiation
dose and any cancer, including lung cancer
– Suggestive trend for leukemia was based on small
numbers (18 observed v 15.5 expected) and trend
was not statistically significant

Figure 17pp

SSFL workers (Chemical Group) had a lower risk
of death than the general population of California

2,251

All Causes

2,715

All Non-Cancer
Causes

1,596
1,979
655
736

All Cancers
215
241

Lung Cancer

Observed
23
27

Leukemia
0

Expected - based on mortality
rates for California
500

1000

1500

2000

Number of Deaths

Figure 18pp

40

2500

3000

 48 of 189

No evidence that working at SSFL increased the
risk of dying from all cancers combined
2.0

Relative Risk (All Cancers)

Relative Risk
95% Confidence Limits

2,086 cancers
among 32,979
workers

405 cancers
among 5,637
workers

204 cancers
among 2,197
workers

48 cancers
among 538
workers

< 5 yr

5-14 yr

= 15 yr

1.0

0.0
Comparison Group
(Other than SSFL
Workers)

Years Worked at SSFL

Figure 19pp

Test stand mechanics had a lower risk of death
than the general population of California

570

All Causes

650

All Non-Cancer
Causes

396
476

174

All Cancers

174

63

Lung Cancer

60

8

Kidney Cancer

5

Leukemia

7

Observed
Expected - based on mortality
rates for California

8

0

100

200

300

400

Number of Deaths

Figure 20pp

41

500

600

700

 49 of 189

No evidence that working as a
test stand mechanic increased the
risk of dying from all cancers combined
2.0

Relative Risk (All Cancers)

Relative Risk
95% Confidence Limits

81 cancers
among 800
workers

35 cancers
among 368
workers

58 cancers
among 474
workers

151 cancers
among 1,598
workers

1.0

0.0
Comparison Group
(SSFL Not a Test Stand
Mechanic)

< 1 yr

1-4 yr

= 5 yr

Years Worked as a Test Stand Mechanic

Figure 21pp

No evidence that working as a
test stand mechanic increased the
risk of dying from lung cancer
2.0

Relative Risk (Lung Cancer)

Relative Risk
95% Confidence Limits

10 cancers
among 368
workers

31 cancers
among 800
workers

22 cancers
among 474
workers

59 cancers
among 1,598
workers

1.0

0.0
Comparison Group
(SSFL Not a Test Stand
Mechanic)

< 1 yr

1-4 yr

Years Worked as a Test Stand Mechanic

Figure 22pp

42

= 5 yr

 50 of 189

Classification of potential exposure to hydrazines
among test stand mechanics based on job title
and test stand
1,598

No. of Male Employees in Study

1600

920

1200

800

400

205

159

156

0
Comparison
Group
(SSFL Not Test
Stand Mechanic)

None

Possible but
Unlikely*

< 1.5 yr

= 1.5 yr

Years of Potential Exposure to Hydrazines
*Most workers (>90%) did not work with hydrazines but could not be distinguished

Figure 23pp

Test stand mechanics potentially exposed to
hydrazines had a lower risk of death overall but
slight increased risk of dying from cancer
compared to the general population of California
101

All Causes

114

All Non-Cancer
Causes

68
84
33
30

All Cancers
15

Lung Cancer

Observed

10

Expected - based on mortality
rates for California

2
1

Kidney Cancer
0

20

40

60

Number of Deaths

Figure 24pp

43

80

100

120

 51 of 189

No evidence that test stand mechanics with
potential exposure to hydrazines had an increased
risk of dying from all cancers combined

Relative Risk (All Cancers)

Relative Risk

24 cancers
among 205
workers

95% Confidence Limits

2.0

17 cancers
among 156
workers
16 cancers
among 159
workers

92 cancers
among 920
workers

151 cancers
among 1,598
workers

1.0

0.0
No Hydrazines

Comparison Group
(SSFL Not Test
Stand Mechanic)

Possible but Unlikely

< 1.5 yr

= 1.5 yr

Years of Potential Exposure to Hydrazines

Figure 25pp

Little evidence that test stand mechanics with
potential exposure to hydrazines had an increased
risk of dying from lung cancers
3.0
13 cancers
among 205
workers

Relative Risk (Lung Cancer)

Relative Risk
95% Confidence Limits

7 cancers
among 156
workers

8 cancers
among 159
workers

< 1.5 yr

= 1.5 yr

2.0
59 cancers
among 1,598
workers

30 cancers
among 920
workers

1.0

0.0
Comparison Group
(SSFL Not Test Stand
Mechanic)

No Hydrazines

Possible but Unlikely

Years of Potential Exposure to Hydrazines

Figure 26pp

44

 52 of 189

Classification of potential exposure to
trichloroethylene (TCE)* among test stand
mechanics based on job title and test stand

No. of Male Employess in Study

1,598

1600

1200
695

800
416
329

400

0
Comparison Group
(SSFL Not Test
Stand Mechanic)

No TCE

< 4 yr

= 4 yr

Years of Potential Exposure to TCE
*Includes TCE exposure potential from engine flush and use as a utility solvent

Figure 27pp

Test stand mechanics potentially exposed to TCE
had a lower risk of death overall but similar risk of
dying from cancer compared to the general
population of California
391

All Causes

451

All Non-Cancer
Causes

270
330

121
121

All Cancers

51
41

Lung Cancer

Observed
Expected - based on mortality
rates for California

7
3

Kidney Cancer
0

50

100

150

200

250

300

Number of Deaths

Figure 28pp

45

350

400

450

500

 53 of 189

No evidence that test stand mechanics with
potential exposure to TCE had an increased risk of
dying from all cancers
2.0

Relative Risk (All Cancer)

Relative Risk
95% Confidence Limits

28 cancers
among 329
workers

69 cancers
among 695
workers

No TCE

< 4 yr

52 cancers
among 416
workers

151 cancers
among 1,598
workers

1.0

0.0
Comparison Group
(SSFL Not Test Stand
Mechanic)

= 4 yr

Years of Potential Exposure to TCE

Figure 29pp

Chemical Summary Findings
Work at SSFL, as a test stand mechanic or with
specific chemicals, has not caused a detectable
increase in cancer deaths among Rocketdyne
workers
– There were no significant trends or any significant
excesses of cancer among workers at SSFL, or
among test stand mechanics
– Hydrazines were not linked to significant increased
risk of cancer, although lung cancer elevated
compared to general population
– TCE was not linked to any significant increased
risks of cancer
Figure 30pp

46

 54 of 189

Limitations
• Low exposures limit ability to detect
increased risks, if they existed
• Chemical exposure only “potential”
since few measurements made in early
years
• Lifestyle factors such as diet and
tobacco use not known
• Mortality rather than illness

Figure 31pp

Strengths
• Multiple data sources used to identify study groups
– 99.2% of eligible workers traced

• Comprehensive Radiation Assessment
– Doses obtained pre and post Rocketdyne
– Comprehensive estimates of internal radiation doses

• Chemical Exposure Assessment
– Worker assignments to specific test stands
– Accurate assessment of hydrazines and TCE exposure

• Additional analyses conducted
– Including comparisons to other workers at local Rocketdyne
facilities such as Canoga Park

Figure 32pp

47

 55 of 189

Conclusion

The Follow-on Study found no consistent
or credible evidence that employment at
Rocketdyne adversely affected
worker mortality.

Figure 33pp

48