Document jyKjK6eMZGOyXqNrEE4aG8omy
<n>
A U S iirll^
Results to Date from the PFOA Worker Study
Ammonium Perfluorooctonoate: Gross-Sectional Surveillance of Clinical Measures of General Health Status Related to a Serum Biomarker of Exposure and Retrospective Cohort Analyses in a Polymer Production Plant
Robin C. Leonard, Ph.D. Principal Epidemiologist, DuPont Haskell Laboratory for Health and Environmental Sciences
EPA, January 10, 2005
1
EXHIBIT 2
--JJ
x
<HJ>
Agenda
Purpose of Study General methodology Results to date (clinical pathology
parameters) Summary Timeline Further Work
EPA, January 10, 2005
36
Purpose of Study
Primary objectives :
- rDveaelarviateioblonlepsshsstipautgoisgtfeiscseaterludmmboydPeFpOlsreAtvhitoaouthsdeaeanslcitmhribaoeluattchnodeme
worker studies, taking into account potential confounders and effect modifiers. - Conduct retrospective cohort mortality analyses using appropriate stratification based on estimated past exposures to PFOA.
EPA, January 10, 2005
<a&> General Methodology
Voluntary participation across all areas of the plant Cross-sectional design, that is, "snapshot" of both the
exposure marker and the health outcome variables
- Cross-sectional studies address "person and place", but not "time"-- the descriptive triad of epidemiology
- No data collected over time, no baseline
- Cannot determine causality, only identify associations
Logistic, linear, and quadratic regression analyses for modeling
Deciles of exposure used for internal comparisons
EPA, January 10, 2005
4
C f
Results from Clinical Chemistries
1,024 employees participated in the crosssectional health surveillance
- All participants have received their individual serum PFOA levels and medical test results
- 62 parameters have been analyzed
Not all of the questionnaire data have been analyzed, but these analyses are underway
Restrospectiv cohort analyses not completed
- NDI data received Dec. 21, 2004
EPA, January 1052005
5
( J ill)
General Findings
Most of the parameters measured were within normal reference ranges and not associated with serum PFOA levels.
There were statistically significant (p< 0.05) modest increases in some cholesterol fractions (total, LDL) and triglycerides with higher concentrations of serum PFOA.
HDL cholesterol was not associated with serum PFOA levels.
- As expected, age, body mass index, and alcohol consumption were also contributors to increases in lipids.
CRP levels (C-reactive protein) were not
associated with
levels.
6
* > General Findings
There were no consistent relationships between results of liver tests and serum PFOA concentrations.
- Different responses in males and females
- Alcohol consumption was a factor, but inconsistent as to category and between genders
There were statistically significant, but slight, increases in uric acid and iron with higher concentrations of serum PFOA.
- Increased uric acid has been associated with increased lipids.
EPA, January 10, 2005
7
*5 *y Os*?
PB>
Serum PFOA Levels By Work Assignment
Work Assignment
Works in PFOA areas
Serum PFOA (ppb)
Number in Median Group
259 490
Min 17
Max 9550
Previously worked in PFOA 264 areas
Occasionally works in PFOA 160 areas
Never assigned to PFOA areas
342
Total Participants 1025
200 180 110
9 8 5
2590 2070 963
EPA, January 10, 2005
8
3$^
Context for Cholesterol Levels
Male study participants' total cholesterol and LDL levels were compared to U.S. males using NHANES data.
Study participants had a smaller percentage of males with high cholesterol than the U.S. population, except for the top decile of serum PFOA.
EPA, January 10, 2005
Percentage
Overall Comparison of Male Washington Works Study Participants with the General US Population
C ho lesterol'tleve iifmg?dL)
3b
Comparison of Washington Works Workers Compared to US Population
Distribution of total cholesterol level of non-medicated male workers exposed to
different deciles of the-distributionofpfqa
Percentage of workers with
total cholesterol greater
than 240 mg/dL
Sixth Decile
Percentage of workers with total cholesterol greater than 240 mg/dL
oO) O(0 m(0
c*-> ca
oo o2a 0V<1-) <-D
>
anuary 10, 2 0 0 5 '
Cholesterol Level (mg/dL)
37
Excludes those on lipid-lowering meds
Simple charts of mean unadjusted lipid values for exposure decile indicate a small increase in last decile. Modeling that accountsfo r known riskfactors provides more information.
Lipid Fractions Unadjusted: Males
Lipid Fractions Unadjusted: Females --- CHOL LDL HDL y TRIG
Mean adjusted lipid values for serum PFOA levels indicate a modest increase in highest decile (>1000 ppb)
EPA, January 10, 2005
<3?
13
mm
CHOLTOTAL by PFOA Quantile
Bex=M, Heartmeds=N, I4U 2004
Mdn
196.823
Std Dev 32.345
107.016 28.787
205.329 39.434
350 -
Wear Nobs
201267 625000
199.857 31099
300 -
196.306 37.476
204635 32835
200.071 30472
198873 31022
206.097 36.740
C 0
214587 38646
250 -
CHOL TOTAL
60 -
00-1 W1 9
D
m _6
W3 96
W4 129 W5 168 W6 2 0
gipbl
W7 264 W8 360
W_633 WD 963
5
CH O LJO TAL by PFQA Quantile
Sex=F, Heartmeds=N, MU 2004
M*an
190.667
SM D&v 3a 111
171.045 28.550
201727 28944
350 "
Mean 199.839 Nobs 218000
300 -
197.318 40554
195.150 27.500
20a478 $91160
205636 31709
2061045 38367
208.000 43.470
218091 41657
250
-Ii 200 ~
O
XO
50 ~
0 "
C0240
50-1 W1 0
-- )-------- -------- 1-------- 1-------- 1-------- 1
VU2 J S
W3 57 W4 71 W5 _K W6 JJD V.7 JB4
g ip b l
i
W8 193
~i m _246
i
WU4C7
It
LOG_Chol_Total: MALES, WW 2004 Results of Regression Modeling
Where Sex=M, H e a rtm e d s = N O
Source
DF
Model
4
Error
620
Corrected Total
624
SS 0.78955 17.78688 18.57643
MS 0.19739 0.02869
--
FValue 6.88
ProbF <.0001
----
Parameter Estimates
Variable
DF
Estimate
Intercept
1
4.99553
LOG_PFOA
1
0.02307
BMI
1 0.00357
AGE
1 0.00153
ALC6
1 -0.07070
R-Square
0.042503
StdErr 0.06366 0.00623 0.00158 i0 . 0 0 0 7 7 8 4 1 0.03469
tValue 78.47 3.70 2.27 1.97 -2.04
Probt <.0001 0.0002 0.0238 0.0491 0.0420
Sex=M Meds=NO
Where Sex=M, H e a r tm e d s = B
Source
DF
Model
2
Error
778
Corrected Total
780
(B = a ll s u b je c ts ) SS
0.43865 23.98564 24.42429
MS 0.21932 0.03083
--
FValue 7.11
ProbF 0.0009
----
Parameter Estimates
Variable
DF
Estimate
Intercept
1
5.17480
LOG_PFOA
1
0.01921
ALC6
1 -0.05789
R-Square
0.017960
StdErr 0.03196 0.00574
FPA January '
2005
tValue 161.93
3.34 -1.83
Probt <.0001 0.0009 0.0683
Sex=M Meds=B
16
BD
LOG_Chol_Total : FEMALES, WW 2004
Results of Regression Modeling
Source Model Error Corrected Total
DF 4
213 217
SS 1.00313 6.91462 7.91775
MS 0.25078 0.03246
--
FValue 7.73
ProbF <.0001
----
Parameter Estimates : Sex=F, Heartmeds=NO
Variable
DF
Estimate
StdErr
Intercept
1
4.84428
0.08752
LOG_PFOA
1
0.02381
0.01161
BMI
1 0.00413 0.00195
AGE
1 0.00474 0.00141
ALC2
1 0.39119 0.18144
R-Square
0.126694
tValue 55.35 2.05 2.11 3.36 2.16
Probt <.0001 0.0415 0.0358 0.0009 0.0322
Source Model Error Corrected Total
DF 4
238 242
SS 1.04587 7.43958 8.48546
MS <L 26147 0.03126
--
FValue 8.36
ProbF <.0001
----
Parameter Estimates: Sex=F, Heartmeds=B
Variable
DF
Estimate
StdErr
Intercept
1
4.86657
0.07935
LOG_PFOA
1
0.02283
0.01067
BMI
1 0.00475 0.00178
AGE
1 0.00383 0.00127
ALC2 R-Square
1 0.40589 EPA, Jani^-7(7*27)05 0.123255
tValue 61.33 2.14 2.68 3.01 2.28
Probt <.0001 0.0334 0.0080 0.0029 0.0234
LOQ Chol Total vs LOGPFOA : WW 2004, Terms= LogPFQA BMI AGE
Where Sex=M, Heartmeds=NO
5.8
5.7
5.6
5.5
LOG_ChoI_Tot aI
5.4
5.3
5.2
5.1
5.0
4.9
4.8
4.7
234 5678
W it )
L0G_PF0
ALC6
9
RJl L OG _ Ch o I _ T o t o I
LOG Chd Total vs LOG PFOA : WW 2004, Terms=LogPFOA AL06
Where Sex =M, Heartmeds=B
5.8
SLOPE P --VALUE =0 5. 7 R - S Q R = O j02
5.6
Choi of 5. 5 240 mg/dl.....
*
*m
.....:<-- Tfc--- 1 ..%..
5.4
5.3 5.2 5.1 5.0
4.9 4.8 4.7
2
i*
V -
\** 4
.. .
L A '* V
` '
' 7 V } . `
*
.
-1
i f 5? * I t
= .!
* *
%
*t * f #
10th decile of serum
PFOA, >1000 ppb
3456789 LOG PFOA
//<c
LOG _C ho I _ T o t o I
LOG Chol Total vs LOGPFCA : VWV 2004, Terms= LogPFCA BMI AGE ALC2
Where Sex=F, Heartineds=NO,
5.8 5.7 5.6
5.5
5.4
5.3
5.2
5.1
5.0
4 .9
4.8
4 .7
4.6
4 .5
4.4
LOG PFDA
LOG Chol Total vs LOGPFCA : VWV 2004, Terms= LogPFCA BMI AGE ALC2
Mhere 8ex=F, Heartmeds=B
5.8:
: SLOPE P --VALLIE = 0.03
5.7:
*
RX1 L OG _ Chi o I _ T o t a
^7
(gjJPll])
URIC_ACID by PFQA Quantile
Sex=M, Heartmeds=B, UH 2004
Mean Std Dev
2.50 "
2 .2 5 -
1735 0.198
1795 0.209
1.822 0.179
Mean
1805
Nobs 781000
1828 a 183
1820 8177
1.805 0.189
1810 0.205
1816 0.205
1.777 0.180
1.875 0.178
tOO H W1 8
-- !-------- 1------;-- i-------- 1-------- 1-------- 1
I
l
I
W2 67 W3 97 W4 t53 W5 J B 9 W6 210 W7 266 W8J367 W9 537 W10 KMO
(MITO
Mean Std Dev
1410 0213
UHC_AOD by PFOA Quantile
Sex=F, Heartmeds=B, MM 2004
1.314 0.309
1495 0285
1486 0287
1542 0.195
1533 0.192
1568 0275
1552 0242
Mean
1.508
Nobs 243.000
1.594 0.175
1582 0.222
~ll I
I
0499
as -L -- !-------- 1-------- 1-------- 1-------- 1-------- 1-------- 1-------- 1-------- 1-------- 1---
W 15
W2 37 W3_56 W4 74 W5 104 W6 131 W7 166 W9_199 W9_252 W10 488
yiptU
LU C i U K I A U U VS L U H h A : W W 2004, !e rm s= L o g P h U A BMI
Uhfirp Rpy=M, HRnrt.nu>Hs=R
ALC4
2 .G: 2.5: 2.4 :
SLOPE P - VALUE = 0 R --SQ R= 0.16
2.3
2.2 :
2.1 :
* > /
RXU36-URIC-ACID
2.0 1 .9 1.81 1.7 1 .G 1.5 1.4 1.3d
^ W w t f m r m_ M a* tr i
a _ a*
.v K '
r \ \*& * #t . v 4
i y i 1 i* 9
wrf. ** 7 ai i c i
,* v
.. 9
*
1.2
1.1 i ri.i'i.ri.ii iiil i"i i i ii il i i--i--i--i--i.ii iiiii i ii i1"|-
2. 0 2. 5 3. 0 3 . 5 4 . 0 4 . 5 5 . 0 5 . 5 G.O LOG PFOrt
--i--i--i-- r~T~i.i"T ii Ii i ii Iii ii i ii ii i
G. 5 7.0 7.5 8.0 8.5 9.0 9.5
S
LOG URIC ACID vs LO G P FO A : WW 2004, Terms= LogPFOA BMI AG E
Where Sex=F, Heartmeds=B
5/
RX1L0G_URIC_*CID
L O G U R IC A C ID vs LO G P FO A : CLEANED, Terms= LogPFOA BMI AGE
Where Sex=F, Heartmeds=B
5^
LO G JJR ICACID vs LOGPFOA: CLEANED, Terms= LogPFOA BMI ALC4
Where Sex=M, Heartmeds=B
2.6 :
2.5 : 2.4
SLOPE P- VALUE =0 R --SQ R= 0.17
2.3:
2.2 :
2.1 :
f
o iDv~o iyn-oo 11 xy
2 .0 : 1 .9 1.8 1.7 "j
1.6
1.5 1.4 1.3
1.2
.
* V
i - . ' T i p i M 't . i T ' # " .
I
A* #
/. *
1.1 l i ii| iiii ]--n ii |--r
ti l lI~T1.i <T~i-- ii i|i--(i i|i--n --i--f--i--i--i--i--|i--i--r.rj i r i i |i--i ? i ji i i i | iii--t~
2. 0 2. 5 3. 0 3. 5 4 . 0 4. 5 5. 0 5. 5 6.0 6 .5 7 .0 7.5 8 .0 8.5 9.0 9 .5
LOG PFOA
IRON by PFOA Quantile
Sex=M, Heartmeds=B, UU 2004
W1 _8
W2 67
W3 _07
W4 133
WS J09 W6 210 grplftl
W7 266
W8 367
W0 537 W10 1040
5/
IRON by PFOA Quantile
Sex=F, Heartmeds=B, MM 2004
W1 5
W2 37
W3 50
W4 _74
W5 D4 W6 131 gtpifri
W7 0
WS _198 W9 252 WD 488
IRON by PFOA Quantile
Sex=B, Heartmeds=B, UM 2004
W1 5
W2 56
W3 85
W4 1
W5 1 WO 180 gtpioi
W7 234
W8 311
WO 407
WO 838
V
LOGJron vs LOG_PFOA: CLEANED, Terms= LogPFOA BMI
Where Sex=M, Heartmeds=B
6.0
5 .9 -j SLOPE P -V A L U E =0
5.8
5.7 R--SQR= 0.02
5.6
5.51
5.4
5.3
5.21
5.1
Ib .0
4.9
4.8
o
4.7
m
I4 .6
CD
O
4.5
4 .4
X 4.3:
s- -.vzyni'X-ttl4.2-
A
' V
4.1
4.0
3.9:
3.8:
3.7
3.61
m
3.5
3.4
3.3
3.21
3.1
3 . 0 | i i i i | i i i r ] i i i r [ i i-- i i | r..i i
i | i i i i.i | i i i i [ i i i i | i i i i | i i i i'"| i i i i | i i r~t......... j i i i i |~
2.0 2.5 3.0 3.5 4 . 0 4. 5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5
LUli PHJfl
LOGJron vs LOG_PFOA: CLEANED, Terms= LogPFOA BMI
Where Sex=F, Heartmeds=B
p Summary
To date, there are no human health effects known to be caused by PFOA; several statistical observations m erit further study.
- Statistically significant associations are seen with
serum PFOA levels and some serum lipid fractions,
uric acid, and iron.
- These associations were only seen in those study participants with the highest serum PFOA levels, which were equal to or greater than 1000 ppb.
- DuPont, in collaboration with outside experts, is
committed to conducting the studies that are
necessary to understand the significance of these
observations. E P A , January 10, 2005
33
P1J>
Timeline___ 2004
Epidemiology Study Timeline
Design, identification of contractors, approvals from review boards
A M3 3 A S O N D p a u u u e c 0 e IQ 2Q
y n I g P t v c 05 05
Operations and logistics planning, employee communications
Clinical examinations, blood draw
Electronic data capture/transfer/QA
Initiate statistical analyses, draft methods, outline results section, meet with ERB
Draft Final Report (Cross-sectional analyses)
Draft Final Report (Cohort Mortality analyses)
F.PA lannary 10. 7.00 S
GO
<TM i> Plans for Further Work
DuPont Medical, Epidemiology, and Toxicology will work with medical and scientific advisors to design studies to answer remaining questions:
Are these observations reproducible?
Are similar associations seen in other worker populations?
Is there a cause and effect?
Is there a biological basis of these associations?
EPA, January 10, 2005
35
6/
