Document jZxL0gOdZO2DdQ4OJg6pkEZk

TSCA Document Processing Center Attn: Section 8(e)/FYI P-1 August 10,2006 Page 5 of 10 Attachment 2 ^ 226. 37^ Tables and excerpts from the draft protocols that contain information related to C-8 blood levels. Table 1. Descriptive statistics for the community cohort age>=20 at baseline in 2005 (n=23,942) Mean age in 2005 Age <20 Age 20-39 Age 40-49 Age 50-59 Age 60-69 Age 70+ Median Body Mass Index Mean C8 level Median C8 level Percent >=60 yrs age Percent BMI>=30 Percent white Percent male Percent current smokers Percent finished college 46 20% 31% 18% 15% 10% 6% 2 7 .9 130 ppb 48 ppb 20% 36% 98% 46% 26% 13% Table 2. Anticipated distribution of C8 in study participants. C8 median C8 mean C8 <10 ppb 10-49 50-99 100-149 150-199 200-299 300-550 505+ 48 ppb 130 ppb 8% 43% 20% 9% 5% 5% 5% 4% TSCA Document Processing Center Attn: Section 8(e)/FYI Table 1. Descriptive data on subjects>age 20 (n=23,942) Age 20-39 40-49 50-59 60-69 70-79 80+ W hite Fem ale Current smoker School <12 H igh School/G E D Some College College degree C8 median C8 mean C8 <10 ppb 10-49 50-99 100-149 150-199 200-299 300-550 550+ BMI mean BMI median BMI 30+ % prevalent heart disease % prevalent stroke Cholesterol mean Cholesterol median 38% 23% 19% 12% 6% 2% 98% 54% 26% 11% 43% 34% 12% 48ppb 130ppb 8% 43% 20% 9% 5% 5% 5% 4% 28.9 27.9 30% 8.5% 2% 200 mg/dl 196 mg/dl P-2 August 10,2006 Page 6 of 10 TSCA Document Processing Center Attn: Section 8(e)/FYI P-3 August 10,2006 Page 7 of 10 Table 3.2 Blood C8 in participants by water district reported as source of drinking water (last address, ignoring those not reporting one of the six water districts as source). Mean, median and 10th and 90th percentiles. (Units ug/1). District of last address (DW) N mean(c8) med(c8) pl0(c8) p90(c8) 1: C ity o f B e lp re ,O H 2: Tupper Plains 3: Little Hocking Water Association 4: Lubeck Public Service District 2,831 4,807 4,338 5,494 5: Mason County 6: V illage o f Pom eroy 3,522 734 Other 8,903 71.6 63.0 396.9 131.7 19.1 18.6 82.6 45.2 48.9 290.0 90.4 15.5 15.6 29.1 17.6 134.9 18.5 111.7 83.6 820.5 30.9 237.0 7.4 34.2 7.1 32.6 91 138.3 Total 30,629 123.0 47.8 11.7 284.9 Table 3.3 Mean, median and range within some percentile fractions (Units ug/1). N m ean(c8) med (c8) min max Lowest 10% Upper 10% ' Lower 50% Upper 50% 3,117 3,063 15,323 15,306 8.0 664.4 22.9 223.2 8.5 .25 479 285 21.2 .25 113 47.9 11.7 11391.9 47.8 11391.9 TSCA Document Processing Center Attn: Section 8(e)/FYI P-4 August 10,2006 Page 8 of 10 Figure 3.1 Distribution of log c8 in blood * 10 3.3 Power Power is presented in Table 3.4 as the chance o f being able to detect a significant result for a modest increase in RR (1.25) and a substantial increase in RR 2.0 for the simplest comparison between the upper and lower half o f the population (with C8 in blood, mean 223 ug/1 and 23 ug/1 respectively). [With alpha 0.05, two sided. Calculated with 'sampsize' in STATA.] Note that this may underestimate power relative to analyses that consider the full range o f exposure or ordered categories, but ignores more complex models with multiple covariates which can reduce power somewhat. Nonetheless, these estimates convey a good sense of the magnitudes of association we will be capable o f detecting. Table 3.4 summary of power calculations Expected n % Power to detect RRof: 1.25 2.0 Bladder Breast (fern) Kidney Testes 106 559 113 59 18 92 75 100 19 93 11 68 TSCA Document Processing Center Attn: Section 8(e)/FYI P-5 August 10,2006 Page 9 of 10 The Science Panel also reports the following with respect to a trend observed in the C-8 blood levels: 3 The C8 health project data and trends in PFOA levels in the study area [The Science Panel has] a very detailed source o f good quality data on C8 levels at an individual level by virtue o f the large amount of data collected by the Brookmar team. This will be supplemented by a detailed assessment of the patterns of metabolism and clearance o f C8 in the half life study component. This data have been assem bled over nearly a year and an initial exam ination o f it reveals a significant trend in C8 by community, which has im plications for the design o f our study. He following figure shows the pattern o f C8 levels for people classified by the current water district in w hich they live (after excluding those who live outside the catchment area o f the 6 w ater districts). Each point is the average o f a w eeks' collection o f blood samples over 47 weeks o f data collection. In each w ater district they appear to have a downw ard slope. 0 5C 3 60 B Idalew tmaan>c8 -----------Fitted valuas Graphs by if water source wasus public., select district 50 The trends are more readily seen in the log-transformed C8 data: the following graphs illustrate the similar proportional trend in average weekly mean log c8 within each o f the six water districts: TSCA Document Processing Center Attn: Section 8(e)/FYI <6 P-6 August 10,2006 Page 10 of 10 logcij 109c.: ------ FSM vtfUM .JH 1 SO $0 to*c*>10 20 Idattw 40 Fn4 .m m 9Cl_S FK M flftNS FlM vtkM t The magnitude o f the slope is about a 7% drop per month: If the within-individual trend is comparable to the observed slope for the community (i.e. the observed slope is not due to a downward trend in the proportion o f individuals who happen to have relatively high C8 in blood unlikely to have been repeated in each water district) then this would indicate a rather steep decline: the combination of reducing intake and clearance would then appear to be associated with a half life o f less than a year. At this rate the c8 level in the blood has fallen by nearly 60% in a year.