Document rBaXgRVngXgzM445LrBY1xva0
. os i a i TZARS'
*? *
UNITED STATES ENVIRONMENTAL PROTECTION AGENCY WASHINGTON. D.C. 20460
JUN 2 6 S985
OFFICE, OF POUCY. PLANNING AND EVALUATION
Subject: Response to Dupont. Critique of Blood Pressure
Papers
;
From:
Joel Schwarts Office of Policy Analysis
To: David Weil Environmental Criteria and- Assessment Office
The enclosed paper on site effects in the NHANES II blood pressure analysisis in response to the critique submitted by the Dupont Corporation. Please forward it to Richard Royall and James Ware. I would be! glad to meet with them to discuss it further. Dick Landis will be forwarding to you separately his additional analysis, using the Mantel Haensel technique for dealing with the sites, and controlling for age, body mass, and continuous lead exposure.
/v*/' V
JN* x/
N sVs X * V V >
S'
/ ;, f
J <vv
^
V'
/'
/ /jjt, /<&.
Oium)
J
MW*:, A t 1/ ('//" < <* *'
// Ac. tf
/S / V\
fV
\y
*
ufuf m, j
--1iSd ' , ' A - tic
0
/W' *<* "( n
j f A <t-. <*?>j-J1';
w
it
.. /
/1` t,j_-_______f
" *1 , .
J ',
ce'- > y ^
tr.,
S?tY.h*
7 '7'
..Pff * n
h
UJk* U < A*,,`JC v-' ...r..". *T
f
'
j % H ~l,-V Ur-"
O
TEH 0412302
DUP050453354
Blood Lead, Blood Pressure, and Site Effects
This memo addresses the question of whether the relationship between blood lead and blood pressure is confounded by the sampling locations used in NHANES. It consists of several parts. First I discuss some theoretical considerations and results using demographic variables rather than site variables. Second, I discuss the problem of measurement error and raispecification, and the implication of that for the use of site variables. Third, I discuss the results from controlling for individual strata in the analysis, and finally the results using indicator variables for each site.
My first point is that the use of 64 dummy variables for analyzing the NHANES data is unreasonable* There are, depending on'"' which variables are entered in the model about 2260 observations without missing values for adult males (age 20 or over) in NHANES. That is an average of 35 persons per location. Using one dummy variable for each 35 persons absorbs a great deal of the variation of both lead and blood pressure, making it difficult to maintain a significant correlation.
Moreover, location is obviously not a causal factor in blood pressure. At most it stands for some ommitted causal variable which is linked to blood pressure, and varies by site. The most obvious factor is in fact lead. (As my submission to CASAC showed, other strong correlants of blood pressure, such as body mass index, do not vary noticeably between sites.) Given the large number of nutritional, biochemical, personal and demographic factors that we tested, I believe its only reasonable to ask anyone who believes in
TEH 0412303
DUP050453355
N 30984.01
an omitted confounder to at least say what it is, instead of using dummy variables that are known to be confounded with lead.
The most obvious things the site variables stand for, besides lead, are demographic factors. Region of the country, income, race, degree of urbanization, educational level, and genetic background (family history of hypertension) all may vary from site to site. However, forcing all of these factors into the regression model affects neither the magnitude nor the significance of the lead effect, strongly suggesting that the sites are only proxies for lead . , Moreover, using dummy variables for each site presumes the
,'ii `
N sites to be fixed effects, whereas by the design of the NHANES, they are random effects, each being chosen from a large number of candi date PSU's. Thus, Dupont's treatment of sites was incorrect. Ignoring that problem, it is one thing to test whether there are unexplained differences in blood pressure between locations, and another to force terms for those locations into a model, whether or not significant, just to reduce the correlation between blood pressure and lead. Dupont's memo fails to mention that almost all of the site variables are insignificant. I will return to this point when I go over my results using site indicators. One final general point that needs clarification is that, contrary to the claims of the Dupont memo, there are not, in fact, 64 sites in NHANES.- There are 32 strata, with two psuedo-sampling units each, for constructing variance estimates, but some of the
TEH 0412304
DUP050453356
cities visited comprise more than one PSU. There are three sites sampled in the New York SMSA, and two in the Chicago SMSA. More over the two PSU's for Los Angeles are not even geographically distinct parts of the metropolitan area. Rather the clusters of eight households were alternately assigned to the two PSU's for Los Angeles without regard to location. Thus, only 60 county or metropolitan areas were actually sampled.
One major concern over using these locational variables [in the absence of any biologically plausible factor (other than lead) being associated with them] is the issue of measurement error and mispecification of variables. One impressive thing about the robustness of the relationship between blood lead and blood pressure to alternate specifications of biological variables is that both variables are noisy measures. The coefficient of variation of blood lead on the blind quality control pools was about 15% on the low pool and 8% on the high pool. Blood pressure is also a noisy measurement. Noise in the two variables downwardly biases estimated coefficients and significance levels. In addition, the blood lead measurement is only a proxy for the tissue lead concentrations, which is what effects the vascular smooth muscle. While in general blood lead levels parallel tissue lead levels, there are spontaneous fluctuations in the partitioning of lead between blood and soft tissue. Thus, some of the variation in blood lead does not represent Variation in soft tissue lead. The correlation coefficient between blood lead and lead excreted after provocative EDTA testing is about
.0.8
TEH 0412305
DUP050453357
Obviously lead is still significant in all of our regressions despite these variances. However, the use of 64 site dummies in the model can interact with this problem in two ways. Firsb, just the presence of 1 variable for each 35 people can explain a lot of the variance of both blood lead and blood pressure, exacerbating the problem of a significant amount of the variance of the two variables not representing true variance. Secondly, the fact that blood lead levels fell over time and varied seasonally means that a larger proportion of the between site variance in blood lead is *true variance in underlying body burden, and a greater proportion of the within site variance is measurement error and spontaneous fluctuation in partitioning. This is particularly true because each PSU was mesasured over a 6 week period. PSU dummies thus eliminate both seasonal as well as temporal variation in blood lead. Thus, a significant downward bias is introduced by inserting in the model a large number of dummy variables more highly correlated with the true variance in body lead burdens than with the measurement error and mispecification of body lead burden.
Not withstanding these points, I have performed a number of analyses that 1 believe demonstrate that the lead-blood pressure relationship is not due to locational effects. First, as noted above, I have rerun the regression with variables tor region of the country, degree of urbanization, income, educational level, residence in central city, and poverty index. The results, shown on Table 2, indicate that lead has the same coefficient and signi ficance level as it has on Table 1, showing the baseline model.
TEH 0412306
DUP050453358
Secondly, I have included dummy variables for each of the 32 strata in the NHANES II Survey. The NHANES II Survey has 8 sampling strata from each region of the country. Each of the strata -represent counties with similar populations. Besides size and region, the other criteria used to group candidate PSU's into strata were median income, percent below porverty level, rate of growth, and percent hispanic or of races other than white.
There are two PSU's from each strata in the NHANES sample. Thus, the strata differ from each other on region, urbanization and other demographic factors, but the two PSU's in each strata are matched on those factors. The use of strata dummies rather than site dummies should adequately represent the different locational and demographic factors the strata represent while reducing the, collinearity with lead from using's variable for every site. The results are shown on Table 3. As indicated in the table, lead is still significantly related to blood pressure, with only minor changes in the magnitude of its coefficient.
The fact that lead is still significant even after controlling for. all 32 strata suggests that site effects are not confounding lead's relationship to blood pressure. However, as the next step in my analysis, I have created dummy variables for each site, and then used maximum r 2 stepwise regression to select those sites where blood pressure was significantly different. This approach allows site variables to come in when they contribute to R2 and the site blood pressure is significantly different from the intercept. Unlike the Dupont approach, it does not force insignificant site.
vm*** :
w>- .n,\^ It**
TEH 0412307
DUP050453359
-6
variables into the regression model. Essentially, this approach says that some sites may be unusual, but does not assert a priori that they are. It still over controls for sites in that out of 64 PSU's, 3 to 4 will be significant by chance alone. The results of this stepwise regression are shown on table 4. As indicated lead is still significantly related to blood pressure after controlling for all significant site effects.
Finally, sites were also selected by backwards elimination. The results at step o and after all elimination, are shown in tables 5 and 6. They show again that lead was significantly related to blood pressure in the final model, and that lead was in fact significantly related to blood pressure in the intial model with 62 PSU variables., (one PSU is in the intercept, and one PSU was completely collinear with the other variables and had a coefficient of zero. Given a dozen or so nonlocational variables and an average site sample of 35 with some sites only having 15 persons sampled, this is not surprising).. The fact that lead was significant when controlling for 63 sites, and most, of the sites were not significant, should surely settle the matter of the robustness of this relationship. For completeness, I have run similar regressions using just 40-59 year olds. Table 7 shows the results of the basic model. Table 8 shows results using 32 strata as variables. Table 9 shows the results using maximum r 2 stepwise regression and dummies for all PSU's. Table 10 shows the results of forcing all locations in the model. In all cases, lead is significantly related to blood pressure.
TEH 0412308
DUP050453360
DISCUSSION
Several comments occur to me. First, the blood pressure relationship is essentially unchanged when controlling for all of the demographic factors that were insignificant. Second, that at all stages of the analysis, the systolic blood pressure seems less sensitive. This may be because there is less measurement error in systolic blood pressure, which gets back to the mismeasurement: point I raised earlier. It is noteworthy that the 40-59 year old subgroup, where age is less of a confounder, is also less sensitive, there is very little change in the regressions with strata dummies, and the size of the coefficient of lead does not decrease as much when all of the PSU's are entered as it does in the regressions on all adult males. This reminds me of the fact that in primitive cultures (e.g bushmen, pacific islanders) blood pressure does not increase with age. The effects being assigned to age are in fact not due to the natural aging process, but in fact merely represents the cumulative damage done to the cardiovascular system by other, causal factors, in industrialized society. Lead may well be one of these factors that age is standing for.
TEH 0412309
DUP050453361
TABLE 1
BASELINE REGRESSION RESULTS FOR. ALL
>VER 20
SYSTOLIC AND DIASTOLIC BLOOD PRESSURE
NOTE: FAMHYPER means family history of:' hypertension HYPERMED means on hypertensive medication
TEH 0412310
DUP050453362
N 30984.02
DEP VARIABLEJ BHZf'lA
snuh'ce
PF
SUf' OF Si-'UAPES
HOUEL
19
HK0K 222 7
C TOTAL 2296
20041i542 533367139
7337*14621
k g u t i'b PEP' NEAN C.Y.
157,51 62, 345990
lb.79,364
ftEAN SQUARE
! 054 8081.5 2395.0 06.
R-SliUAkE AOJ k-SG
F VALUE 'ia,ou2
0,2731 0,2bt>9
PR0>F 0,0001 -
VA*1 a ^-LE
In THP.CEP AGEYRS AG$U cs * X L0G2 l'-G LHIAUTC ALU3U LSCHUL
CALGI'ih UltTVlTC ZINC MEMtKiLut' TRICfcfH FAriNtTE^ CDP FEE HEINmT HYPf k ' E l> , -R-A-C E L!- ) v....
PAKihETEw UF EST1 'ATE
STANDARD
er r o r
1 91 ,S5`>710 32,73t)31
1 ii, fjfey u 0, L07tft
1 'i ,0 0 c G t 6 9 7 n , 0 o11 ibil*
t f.,734h61 0,075359
1 3o,M7q511
V,o7Sid5
1
1, t S1?25
0,271Caft
1 0,0r'2i 1(j 3o 0 ,.oocP5fti?ie
1 3.60<':15ti 1,11 891.3
1 -0,053137
0,01&fc4 7
1 *0 CO ol> 17 351*2 7,,ft37rit-0 7
1 -0,ulCp56 0 , C o2.7 1 677 l
1 ' ,333o72 0,097705
l 0,125575 0,C'2)7u 7
1 0 .,011794 c-otmi46.j2
1 2,1 j}793b 0,443b5C
l -n, 237405.
0,.07'u2>#l,
1 14,'U *456
3,22u52?
1 5,43972a 0,<<2e72
1 2 ,-*97^5* 0,612e1
i <2.95604sN ue257ei
T FOR HO; PAPArETEksO
2.797 8,119 -7,064 9,749
-4,063 4.247
4,919
3,217 -3.154
2,271 -3,7^2
3,415 5,765 2,866 4,929 -3,37-7 4,477 6,54 0 3,313 4,724
PROP > IT!
O.OOS2 o.oooi 0,0001 0,0001 0,0001 0,0001 0,0001 0,0013 0,0016 0,0232 0,0002 0,000b I/.0O0I 0,0042 O.OCOi 0,0007 0,000J 0,0001 0,U0u9 0,0001
TEH 0412311
DUP050453363
DEP VARIABLE* bP2SYS
SOURCE
OF
SUM OF SQUARES
MODEL *KOR
C t o t al
11 2235 2249
402563*411 131</u 15m37 17215796796
MOOT f`5E OFP MEAN
C v,
2429,32c 126,274
1693,604
M A fSOI) a HE
365967126 5901637
R-SiiUAHE a DJ H-SQ
F VAL'-'E 62,011
0,233b 0,23 OT
PBO0>F o.ooor
VARIABLE Of
PAftA:iPTEK ESTIMATE
5T Af'UARD T FOR HO j tHHOR PAKAi'-ETEHsO
I^TEmCEP A&EYh S
A5LSU flMl LOGZIn C AttiSy
CALCl'JN COFFEE FAfiHrKEB
h eig h t
HYPEi-<tD LL :
vV
1 65,Q31972 1 0,Si>61o 1 0,006*94956
1 1,27*647 1 *o,29n.692 1 0,062ar>37o l ,000^027146o 1 -u,373316 1 3,313669 1 i*,469517 1 ^,7e?-590 l 5,23to73
13,476t65 0,157323
0,091730631 0,096606 2,1**7299
0,0012961*76
00000119754 0,105360 0.6962*0 5,0o3'.`b7 1,29553 ",930371
,o92 2,6f'C
5.139 14,431
-3,664 6.300 2,217
-3 ,5"3 4 ,J59
3,692 6,7*4 5.5*5
pr o s > m
0,0001 0,0042 0,0001 O.OOC1 0,0001 0,0001 0,0235 0,0004 0,0001 o.ouo? 0,0001 0,0001
TEH 0412312
DU PO50453364
TABLE 2 ALL MEN OVER 20 RESULTS WITH. INCOME, EDUCATIONAL- LEVEL, REGION OF THE COUNTRY, AND DEGREE OF URBANIZATION
NOTE: R1 means northeast R2 means midwest R3 means south West is inthe intercept SIZE means degree of urbanization coded 1(cities over 3 million) to 8(rural areas under 2500) ED means educational level ( didn't finish grade school, finished grade school, finished high school, finished college) INCOME (under 5000, 5-15,000, over 15,000)
TEH 0412313
DUP050453365
N 30984.03
PEP VHilAnLEl SP2H1A
SiiU^CE
OF
HPfFU
25
ERROR 2132
C TOTH. 2157
ROUT MSE
DE P M*m c.v.
SUM OF
sq uar es
1909409779 5072222076 69*7691055
1583,342 82,334043
1873,351
MEAN SQUARE
76376791 2301905
R-5'3U ARE ADJ R-S9
F VALUE 32,066
0,2733 0,2647
PRQB>F 0.0001
VAKtA'-^LE
I-j TEKCEP AG-EYKS Ar'?(5
b*-i tnuZI^C L'UjViTC a i.6 39 L^CHUl. 5 'tiKWUM CALCl-JM
DIE T v IT C ZI-MC HFtmULU * 7;jioef'N Fi COT PEE HF IGUT HYPE^^EO
r ac e
LL K1 K? R3 SIZE UCCME ED
PARAMETER OF ESTIMATE
STANPAKO ERROR
t 91,309339 33,409495
1 C,*37379 0.1U064
1 -0,00794721 0,001209621
l 0.77-U03 0,077670
1 -3*.2t>9584
9,240763
1 1,230624 0,276099
1 0,00401397b 0 ,0009611767
1 3.277633 1,135349
1 -0,055426
0,017151
1 ,0000*175909 7.76636E-07
1 -n,010493 0,002743141
1 0.329977 0,099399
1 0,127723 0,022243
l 0,010674 0,0042403b!
1 2,263054 0,451116
1 -0,215501
0,071272
1 15,163679
3,370163
1 5,011146 0,042645
t *424.243 0,860056
1
C?,969793\
0,656125
J
-0'779-3~&S-Q"'
0.640721
1 0,321007
0.622316
1 -0,253b4S
0,626576
1 -0,035966
0,066179
1 *0,359460
0,363202
i -0,489642
0.348652
T FUR HO* PARAMETERS
2,735 7.540 6,569 9,941
3,922 4.457 4,555 2,0fl7
-3,232 2,265
3,b2S 3,3?<J
` 5,742 2,517 5,017
-3,025 a,a*9 5,947 2,821 4,526
-1.236 -0,516 -0,404 0,4*6 -0,990 1,404
PROii > ITI
0,0063 0,0001 0,0001
0,0001 0,0001 0,0001 0.0001 0,0039 0,0012 0.0236 0,0001 0,0009 0.0001 0,0119
0,0001 0,0025 0,0001 0,0001 0,0040 0,0001 0,2165 0,6060 0*6066 0,6832 0.3224
0,1603
TEH 0412314
DUP050453366
DEP VARIABLE! BP2SYS
5'iU^CE
DF
SUM OF SQUARES
HnoeL EWK^K
17 21 49
c t o t a l 2157
HOOT ;<sc l >e p MF'AN c.v.
39*2349530 12455895216 16358244753
2412,574 120 ,,3*5
1H79.177
MEAN SQUARE
229549973 5620512
R-SfJUARE AOJ R-$0
F VALUE 39,436
0.23F6 0,2325
PROB>F 0,0001
V.*.k TA-<LE
3MEWCEP AGEYRS A['Sy a**! LCb/i^c Al.OPU CALCIUM CnfFtE FAflHYPtH Hr IUMT HYPt H"Ef> LL HI
HP K3
S1 IF,
I^cnne
en
PARAMETER DF e s t ima t e
STANDARD ERROR
1 73.513300 13,676346
1 o.'it313
0,167103
l 0,00a45"90d 0,001840373
i 1.25U53 0,090529
1 -T',9'4l73b
2,195370
1 0,0*o44l7So 0,001331242
i .00000275794 ,00*00121252
1 -0.414420
0,106849
!
3,09174 3-
0,704729
1 lb,796596
5,212663
1 e.313200 1,314970
1 v4,649709 1 1V4-1J419
0.9936-71 0,994097
1 3,32<>~y2 0,963625
1 -0.16*257
i 0.301996
0.969&5S 0,136209
i -0,29*310
0,563999
1 -1,129133
0,536051
T FOR HO I Pa HAmEIERc O
5,297 -2.5*3
4,59b 13,909 -4,027
6,341 2.275 -3.879
4,307 3,222 6.322 4,679 1,4 26 3,455 -0.174 2,250 -0.529 -2.106
PROS > ITI
0,0001 0,0124 0,0001 0,0001 0.0001 0,0001 0,0230 0,0001 0,0001 0.0013 0,0001 0,00<n 0,1541 0,0006 0.8623 0,0245 0,5969 0,0353
TEH 0412315
DUP050453367
TABLE 3 ALL MEN OVER 20 CONTROLLING FOR SAMPLING STRATA
TEH 0412316
DUP050453368
N 30984.04
DEP VARIABLE!
0
S3URCE DF
OfiGEEL
SO
ERROR 2196
C TOTAL 2246
t ROOT M3E
DEP MEAN
C.V.
BF2D1A SUM OF
SQUARES
2410661977 4=27102644 7337814621
1497.897
EC.345996
1819.028
MEAN SQUARE 48213240 2243694
R-SQUARE ADJ R-SE
.F VALUE 21 s
0.3285 0.3132
VARIABLE Dr
PARAMETER ESTIMATE
STANDARD T FCF; HOs ERROR FARAM.ETEF.-0
c
(
INTERCEP
ASEYRS AEESQ
SKI LGGIINC LGSVITC AL5SQ LSCrOL SKECNUM CALCIUM
DIETVITC
72 INC
HETCSL2B
TRICEPR FAMHYPER CC"=EE HEIGHT HYFERKEE RACE LOCI LCC2 LCC3 LDC4
L0C5 LCC6 L0C7 LCCS L0C9 LCC10 LOCI 1 LOCI 2 LOCI 3 L0C14
L0C15 LOCI 6 LOCI 7 LOCI 8 L0C19 L0C20
LCC21 LCC22
L0223 L0CC4
LCCI6 LCC27
LDC2E L0C29
1 86.49=50? 31.897330
1
0.33833
C.102089
l -0.00792347 0.001105993
1
0.770175
0.073849
1 -T-.e-v60
S.855553
1
1.21429S
. C.266041
1 0.004=86976 0.0CCS473463 1 2.74=590 1.09=394
1 -0.04=138
0.016497
1 .00000142472' 7.463765-07
1 -0.01 Cl67 0.002653499
1 0.2=0774 0.095351
1
0.114433
0.0216=6
1 0.010348 C.004041507
1 2.053694 0.433107
1 -0.212733
C.068890
1 13.11=232
5.161750
I 5.469599 0.807043
1 2.713463 0.842319
1 -2.204078
1.139634
1' 0.606139
1.160206
l 4.C23SS5 1.J98924
1 3.697475 0.990936
1 -1.140654
1.243662
1 -2.557292
1.345=55
1
0.104=36
1.295=00
l -2.6AC254
l.C41733
1- 3.243974
I.223225
l * 07B7ZZ 1.224830
1
o 522494
1.06S633
1 -2.249564
1.115000
1 4.244303 I -5.342127
1.361645 1 1 --ts Z2
1 3 05!TCww lrcsrsc
1 -4.163947 1 3.127097
1.319138 0.992209
1 2.110825 1.241902
1 0.247288 1.315814
1 -1.3475B1
1.164800
1 6.710=96 1.293417
1 -1.350472
1.125165
1 -1.365455
1.171221
1 -4.24S555
1.658571
1 -4.604S6S
1.115387
1 3.585606 1.30451
1 -1.698953
1 . 1S92S5
1 0.463024 ^ ^t SSv 2
1 0.506636 ' 1.216508
-0.221643
1.245761
-1.031160
1.218031
-2.163910
1.063825
2.217851
0.647621
.E7207E-07
=9 *?
8.217' -7.164 10.429
-3.718 4.3=4 5.331 2.501
***CSvO 1.90S
-3. S24 3. 050
3.275 2.560 4.720 -3.03? 4, 14? 6.777 Z m cm -1.934 0.522 3.381 3.731 -0.917 -1.922 O.OSi -C. 57-5 2.641 1.6=7 0.4S? -2.018 3.117 -4.700 2.465 -3.157
4 S2 1.700 0.188 -1.157
5.18? -1.200 -1.166 -2.562 -4.128
2.914 -1.42?
0.470 0.417 -0.178 -C.S47 -2.034 3.425
FRuS>F
0.0001
FRC& > :t i
0.0067 0.0001 0.0001 0.0001 0.0002 O.COC1 C.0001 C.C125 :. o.OSes 0.0001 0.0023 0.0001 0.0105 O. ?0C 1 0.0020 0.0001 0.0001 0. 00 13 0.C532 C.oOlA C.0007 0. :!<*. OC 0.3392 :. of 0.c355 0.0113 0.0023 0.CS9S 0.=249 0.0433 0.00 19 r. 0. 0138 0.0-016 0.0016 C.OS93 O'.. SCO9 0.2474 0.0001
0:. 2433 0.0105 0.OOOl 0.0-036 0. 1533 0.6383 C.6770 0.8588 O.2973 0.0421 0. 0006
1.0000
TEH 0412317
DUP050453369
DEP VARIABLE!
o
SOURCE
OF
OMODEL
42
ERROR 2204
C TOTAL 2246
0 ROOT MSE
DEP MEAN
C.V.
BP2SYS SUM OF
SQUARES 4721760376
12494036422 17215796798
23B0.924 128.274
1856.128
MEAN SQUARE 112422866 SSoBBOl
*. R-S\1UARE ADO R-SQ
F VALUE 19.832
0.2743 0.2604
PROB>F 0.0001
J*,, i.- . * m --15 VARIABLE
) Of
INTERCEP AGEVRS
: AGESQ BMI
! LOGZINC
ALBSQ
CALCIUM
: CDFFEE
f FAMHYFER
' HEIGHT
j HVrERMED
j LOCI
O
O' L0C2 ' L0C3
* L0C4
1 LOOS
L0C6
L0C7
LDCB
L0C9
L0C10
Jo
LOCH -- LOCI2
. LOCI3 j, LOCI 4
1 LOCI3
, LOCI6
; L0C17
LOCI 8
' LOCI9
, L0C20
o j, i 1
L0C21 L0C22 LOC23 L0C24
L0C25
LQC26
` L0C27
l1 1
PARAMETER OF ESTIMATE
STANDARD T FOR HO: 1 ERROR PARAMETERS FROB > IT!
I 72.651416 13.442201
1 -0.454229
0.155397
I 0,003359766 0.001707302
1 1.256657 0.0S7671
1 -8.893992
2.137958
1 0.00S636816 0.001304348
1 .00000215691 .00000118527
I -0.347059
0.104995
1 3.311628 0.691076
I 17.428677
4.973843
1 8.397995 1.279889
1 -3.651484
1.757160
1 -3.830103
1.826635
1
S.8S9649
1.39S765
1 2.618794 1.562940
I -2.846423
1.959755
1 -4.751148
2.127933
I 4.710306 2.053041
1 -0.46SS97
1.647451
1 2.774705 1.941981
I O.S6E742 1.933437
I -1.0B5997
1.696109
1
1.769556
1.769383
1 5.3IS631 2.143333
I -6.23S360
1.800275
1 -0.948343
1.962503
1 -3.571559
2.092310
1 -0.613658
1.571650
I
5.2481IS'
I.965S1
1
1.833871
2.060980
1 -1.00BSB0
I.845447
1 2.268042 2.034765
I --3.370599
1.750352
I -3.747147
1.843723
1 2.215146
i -2.2S8679
2.628885 1.762717
I 5.201344 1.951021
1 -7.030300
1.887161
1 -3.999159
1.563444
5.405 -2.923
5.189 14.334 ' -4.160
6.622 1.820
o05
4.792 3.504 6.562 -2.078
-2.097 3.102 1.676
-1.452 -2.233
2.294 -0.285
1.429 0.294 -0.640 1.000
2.476 -3.465 -0.483 -1.707 -0.390
2.670 0.890 -0.547 1.115 -1.893 -2.032 0.843 -1.298 2.666 -3.725 -2.558
17*34 FRIDAY. JUNE 21. 1985
5
0.0001 0.0035 0.0001 0.0001 0.0001 0.0001 0.06B9 0.0010 0.0001 0.0005 0.0001 0.0378
0.0361 0.0019 0.0940 0.1465 0.0257 0.0219 O.7760 0.1532 0.7687 0.5221 0.3174 0.0134
0.0005 0.6290 0.0880 0.6962 0.0076 0.3737 0.5348 0.2651 0.0585 0.0422 0.3995 0.1943 0.0077 0.0002 0.0106
SAS
VARIABLE OF
PARAMETER ESTIMATE
STANDARD T FOR HO!ERROR PARAMETERS PROS > !Ti
L0C29 L0C30 L0C31 " ^ '"COBS? \u %
RESTRICT
1 2.065686 1 8.610894 1 0.269317 1 -1.912363
1 4.564323 -1 -.0000033387
1.926394 1.973966 1.928927 1.602934
0.986449
1.072 4.362
0. 140 -1.136
4. 627
0.2837 0.0001 0.BB90 0.2559
0.0001 1.0000
TEH 0412318
DUP050453370
TABLE 4 ALL MEN OVER 20 USING MAXIMUM R-SQUARE STEPWISE REGRESSION TO SELECT SIGNIFICANT PSU'S
TEH 0412319
DUP050453371
N 30984.05
$*S
12110 FRIDAY* "AY 20*- !
MAXIMUM PSQMARE IFR0VEh NT FOR DEPENDENT VARIABLE BP2DIA
LGC36 ENTERED
R SQUARE 0.36153927
,, Ct^J
51,07217816
DF
"EIGmTED $$'
HEAR SJUARE
F PR08>F
REGRESSION ERROR
t o t al
43 220S 22 08
2693522332.15716170 46" 5991042,94539200
7339513375,10255400
61709821,67807352 2125166,00587093
29,0
O.OOOI
b val ue
STD ERROR
TYPE II 88
F PRPB>F
INTERCEPT AGEYRS
AGESO nm L0GZ1NC
LOGVITC ALSO
LSCMOL S"0"NUH
OIETVITC 7 INC HEMOGLOB TRICtPR FARHYPER
COFFEE HEIGHT my PERn ED
RACE toci L0C4
toes LOCI 2 LOCI 3 LDCI4
LOCI 5 LOCI* LOCI 7 LUC 20 L0C21 LUC26 L0C26
L0C31 L0CJ2
L0C35 L0C36 LOCaO LDCai L0C42 L0C55 L0CS6 L0C57 L0C60
L0C63
LL
95.0H40S196 0,71*36075
0,or>7#ai?9 0,7fi4i32i9
-33,63520444
1 ,1*64X345 0,00472567 2,41526564 0,<ia24121 0,009(s9159
0,31263216 0,10123470 0,0 1031693
1.95766806 -0.20710116 11.63676063
S.aSl1S1S4 2,569413264 -2.93331235 3,89735982 -3,73109954
-5,16001637
5.01144962 -10,15976671
3,70569778
-6,13918590 6,92450220
-5,0147lS24
9.67215943 7,06131366 3,77902952 -5,69935596 3,51313037 6,21649088
3,12137175 -4,83491069
0,27284700 4,97799141 -3,15187028
0,33254940 -6.12069187 j - ~ 022773 ( 3.81$940.8,9
7 039931$")
: T"E BEST 43 VA-1 ABLE"oSirE"1p0yND,
0.09923520
0,00107597 0,07178647 6,57086040 0.25627024
0,00002226 1,06676214
0.01598485 0,00258298
0.09227883 C,02098462 0,00392564
0,42174006 0.06670595 3.06541U42 0,78229690
0.7e246140 1.43676242 1.26516443 1,6*226048
1.58519215 1.57379973 1,48936222 1.7697757 1,64311798
1.44416977 1,61515254 1,64807889
1.61703570 1,22290500 1,66806176 1.44053647 1.55459254
1.55207571
1.64341433 1,65028719
2.308300X1 1,55511709
1.67363853 1,48734348 1,73657756 1,76860635 0,61640606
136861410,5399863? 101795153,11394582 2r>3563l76,95916853
3308*444,68327499
43094769,61517487 70194276.54811149
ISB11632.35707536 14961357,56581600 29918638,75090298
24393422.0S71063S 89459470,47064224 14683935,7439066S 45791119,44136249 20464660,40841942
3062525b,78B90515 103187278,05479278
16192250,60654-485 8857632,68117221
I90452F8,84706062 10453941.05213883 22521562.80121089 25126337,68960849
98891 659,39036542 9320575,58391400
29660004,61386860 488577S1,62134628 23884547,47237063 73195395.18135657 35034102.69019053 20294015,48641908
24*09622.68794501 12639586.86053442
3C03976,54093925 6590812,19565778
18343976,65726367 14246510,00593707
923?55.70616249 8729700,84374551 14241468.46847252 35991481,63852198 33165118,90450663 9 E 856 7.9,88326221 16240219.85234127
64,40
47,90
119,31 15,57
20,47
33,03 7.4 7,05
14,08
11.48 23,27
6,91
21,55 9.64
14,41 48,55
7,62 4.17 8.96 4.92 10.60
11, 82 46,53
4,39
13,96 22.99 11.24 34.40
16,86
9.55 11,67
5.95 16,00
4,04 8.66 6.70 4,34
,ti 6,70 16.90 15.61 4,65 7.64
0,0001
o.oooi 0,0001 0,0001 0.0001
0.0001 0,0064
0,0080 0,0002
0.0007 0,0001
0,0086 o.oooi 0,0019
0,0002 0.0001 0,0058 0.0613
0,0028 0,0247
o.oou 0,0006 O.OOOi O.f-i J4 0,0002
0,0001 0,0008 0,0001 0,0001
O,O0?0 0,0006 0,0148
0,0001 0,0445 0.0033 0,0097 0,0372 0,0428 0.0047 0.0001 0.0001 0.0311 0.0058
\f f?
TEH 0412320
DU P 050453372
8*8 J5l0t SATUOAY, MAY
H*XXUH R.SQUARE IMPROVEMENT FOR DEPENDENT VARIABLE BP2SYS
RIAflLE LOC37 ENTERED
R SQUARE a 0,316*88??
CCP3
*2.72535*22
OF
w eig h t ed s s
HEAN SQUARE
F PR08>P
REGRESSION ERROR TOTAL
36 2216 225*
5*78210201.23355700 116313*3661.26021600 1730955*062.*9377*00
152172505.5998210* 533*239,79317*13
28.53
0,0001
9 VALUE
3T0 ERROR
TYPE n SS
F- PROB5F
INTERCEPT AGEYR3 AGE SO BMJ
L0G2INC
al bsq
f akhyper
COFFEE HEIGHT HTPERHED LOCI L0C2 L0C7 LOCH L0C13
LOCI* LQC20
L0C21 L0C26 L0C27 L0C31 IOC 35 L0C37 L0C38
LOC*0 L0C3 L0C*7 L0C*8
L0C*9 L0CS3 L0C55 L0C57 L0CS9 LOC60 -
L0C62 L0C63 LL
70.7207117* -0,52525*t
0.00981728 1,25695879
7.61*50012 0.00938173 2,8961693ft
0,32Sl070 17.*1259997'
8,81779773 -5.05799956 -6.*3573692
7,80*95115 5.73*89391
5.53502*30 -12.9110*987
-7,lft993ft3tt
8,07026059 10.39217117 -9.0*7*6557 -6,5778*567
9.*3078211 -5,8175371* -8 . ftht 2971 *
-6,613*1363 -10.505=3552
-5,8*179363 -8.77963131
-5.51*120*3 -8.69095089 -7.29522917 -5.*0359091
-7,0*115687 -ll.29369fc5 ,----- lJL,2228ft689 ( 5.ft%?6415
6620*13^
0,15008*1*-
0.00165136 0.08*9*613 2.07*75293
0.00125711 0.66633522 0,10155099
*,79208118 1.23211272 2.26926617 2.235*1823
2.5*650116 2.3*530062
2`,*8556572 2.3*817211
2.5*937125 2,60113260 2,67**635* 2.5227*032
2.62600255 2**2857737 2.90*3*613 2.70930687 2.5915*5*8 H.*(17**907 2,75016591
3.*86*8762 2.10323158 3.21790*7* 2.*S217*92 2.33973962
2,83576252 2.7*273921
2.92761063 2.7965817* 0.9*366995
65332223.9019389 1885266S9.0*06692
1*71*92179.9*52*22 718*9302.P1AS928
297093516.6203599
100168633,5099ft37 5*632320,5653969
70*28987.7632*23 273207059.R97ft83
26500815.033529* **213156,8232307
5*110086.8568*16 31B9S269,26*1595
26*52157,320115* i61263'395.1819277
*2192665.P3592U2
513*7697.883**3* 69722266,566903* 66616619,3105*7*
33*69529.51*003* 80*38519,796*0*1 21*01986,1657ft9
5*515752.88*1023 56925797.7900701 96291287,6856232 2*068398.5986187
33825859.2501227 3666*926.2726585 38909957.721*296 *7211*69,6060111 28*51313.992*837 328867*2.5852379
90*39527,361*780 18*610668.269S818
22760886.903*1*2 80330187,218621*
12.25
35.3* 219,62
13.*7 55.70 18.78 10,25 13,20 51,22
,9? 6.29
9.39 5.98 *,96
30.23 7.91 9,63
13.07 12,86
6.27 15,08
,0t 10.22
II. 05 18,*3
*.51
6.3* 6.87 7,29
8,85 5.33
6.17 16.95 3*.61
*.27 15.06
0.0005 0.0001 o.noot 0,0002 0.0001 0.0001 0.001* 0.0003 0,0001 0.0259
0.00*0 0.0022 0,01*6 0.0261 0,0001 0,0050 0 ,,0 01 9' 0.0003' Q',.00 03'
0.0123 0.0001 . 0.0*53 o;.ooi*0,0009
0.0001 0.0338
0.0119
0,0088 0,0070 0,0030 0.0210
0,0131 0,0001 0,0001
0.0390 0.0001
'DEL IS THE BEST 36 VARIABLE MOOEL FOUND.
TTTf H l
TEH 0412321
DUP050453373
TABLE 5 ALL MEN OVER 20 BACKWARDS STEPWISE REGRESSION USING PSU DUMMIES STEP 0 ALL DUMMIES IN
TEH 0412322
DUP050453374
N 30984.06
Y
e*CKw*RD ELIMINATION PROCEDURE FOR PEPEFGEn T VARU6LE 8P2DI4 u s e f v a t io ms d e l e t e o d u e t o h is s in r Va l u e s ,
12110 FKIOAV# "Ay 01
RJARl ES ENTERED
REGRESSION ERROR
t o t al
in t er c ept
AGEYS AGESO AMI LOGZINC LOf.VITC ALHSR LSCHOL SKHKNUM CAtCIUN METVITC ZINC fcEHOGLOS TRICEPS --T? fAXHyPER --I* COFFEE HEIGHT HypR*iED RACE LOCI L0C2 L0C3 LOCR L0C5 L0C6 LOCT
l oce
LOCR L0C1C LOCI I LOCI 2 LOCI 3 LOCt LOCiS L0CI6 U0C17 LOCI# LOCJR LOC20 L0C21 L0C22 L0C23 L0C24 L0C25
R SQUARE * 0.37454826
CtP) * 82,00000000
OF
*EIGHTEt> SB
me a n SQUARE
at 2167
22a
2749001957", 64870830 <159051 1417,5345O0
7339513375,10255400
3393*295.77304084 2110371.67395194
A VALUE
STO CRHQP
TYPE II. as
9t.7il4526
o,7R5aoto -0,OO739t7
0,78656070 *33,71997432
J,160 6 0155' 0,00511053. 3.09465436 0,04452715 0.00000125 -0,01021942 0,30715342 0,10262665 0.01046569
1,91766655 *0.17724012
11.93520411 5.43926007
2,24686253 -1.86147256
1,93266553 1,6*1*4526 4.62554106
-2.51191838 -2,91953016
3.09123966
>0,46067165 3.11754974
1,64296685 1,20631630 -4,20293384 6,37919557 9,41948909
4.73371443 -5,25076043
7,87026897 2,78309191
0.51164212 -4,56640575 10,59041446
0,42854431 1.55910552 0,00000000 -1,68005963
0,09980927 133791011.07669510 0,00108159 100212772,30811022
0.07243925 200345217,16671210
8.68694097 31903838,43016373
0,26102274 43336312,60220083
0,00084392 I .1)8424669
77683353.23663040 17259427,26722727
0,01619151 16020547,33670382
0,00000073
6346829.82306690
o,ro26205 32209042,90630933
0,09344631 22886963,42529907
0.02170577 47540453,91677050
0.0-0398039- *4566771,40270693
0.428381 61: 026G52,86052243
0.06759456' 14504732,70723311
3". 10652691' 31268829.37391137
0,70783049 100074758.59702622
0,65084016 15303619,05250747
2.00451474
1689294,94418625
2,09227058
1807500.68119535
2,43581936
1204386,42690618
1,96309036 11525034.47711854
2.25649308 2,65309403
2625094.16020101 2563652.5B617S83
2,21000521
4141584,82675127
2,03306305
108077,97617729
2.34166567 , 3629671,72395379
2,08857748
1310866.85802038
2,10513054
695611.08001284
2,18511893
7037131,02066478
2,171979 18020125.06867241
2,12511090 2.33069992
01618663,21762163 8730432,75118121
2,22588932 2.06446458
11787981,23296685 30198979.80364389
2.53621187
2550857,48886983
2.5726961$
83848.86924832
2.20525023
9083127,37307610
2.20912005 00900074,40321720
2,12238727
112266.37454559
2.33621612
943467,94329254
0,00000000 2,26334607
o.oooooooe 1167211.689T8538
F 16.02
PHCB>F 0.P001 --;
F
63.'l6 7.31 116.29 15.06 ?0.46 36.67
8,15 7,56 3,00 15.21 10.80 22,44 6,69 20.04 6,88* 1 4,76 47,67
. 7,22 0.90 0,85' 0.60 5.44 1.24
1.21 1.96 0.05 1,71 0,62 0.33 3,70 8.51 19.65 8.13 5,56 14,26 1.20 0,04 4.29 22,17 0,05 0.45 0,00 0,55
PR03>P
ft.POOt 0,0001 0.0001 0,0001' 0,00 01; 0,0 0011 0,0004 0.0068 0,0836 o.nooi 0, POt O' O.OOOU n,0087' n.oooir 0.P0R8& 0,00 0 It o.ooof 0.0072 0.3720' 0.3557 0.4304'
0.0198" 0,2657"* o.2714 0,1622 0.8208 0,1907 0,4316 0.5667 0,0546 0,0036 0.0001 0,0424 0.0184 0,0002 0.2726! 0,0423'* 0.0385* 0,0001' 0.8180 0,5046. 1,0000 0.4580
TEH 0412323
DUP050453375
L0C26 L0C27 L0C28 L0C29 LQC30 L0C31 L0C32LOC33
ioc L0C3S L0C36 L0C37
L0C38 L0C39
l ocoo
L0C1 Locaa L0C03 cocoo LOCOS
L0C06 L0C07 LOCO# L0C09
LOC50 L0C5J. L0CS2L0C53' L0C54 L0C55 L0C56 L0CS7 L0C58 10C59
LCC60 L0C61
LDC62
LDC63 U
5*8 12>10 FRIDAY* RAY
BACKWARD ELIMIMATID^ PROCEDURE EOO d e p e n d e n t v a r ia b l e 8R20IA
8,30099735
2.37008981 ' . 26062750,15086970
12.09
-1,58391870
2,1903*368 *. , 1076019,3*356665
0,51
.70920311
1.90506302 12067157,59557122
5.86
3,79096780
2.21613007
6198923,10592383
2.93
0.67297333
2.27286169
185713,81080908
0.09
-4,6)330057^
2.20208615
8960875,913P6T12
4,23
-2,52377075
2.07709035
3126251,66562093
1.08
-0,27818363
2.37771076
28996,51001109
0,01
0,6?00713
2.51372081
130725,23167610
0.06
7.16305090 0,11600000
3,01056606
2.17309587 2.16351809
2,37997991
23016562.78535103 7668753,38600062 3389672,85801282
10.67 3.62 1.60
- -0,82230073
2,30259160
270191,65422099
0.13
-2,93155828
2.66325010
2566639,80529623
1.21
-3,82903195
2.22613072
6266971,89153571
2,96
5,28208910 6,037*0677
2,23*76559 2,82006590
11816917,08600070 9610886,25976619
5.58 0,50
1,53239608
2.15317200
1100124,26991566
O.S
1.37025389
2.1E6E4003
836566,203*9653
0,39
2,27160563
3.12608938
1118610.76216913
0,53
2.33360509
2.32222255
2139190,00011909
1,01
3,03727601
2,30016597
3816231,66110508
1,80
-0,72162759
2,6792831P
153670,70791953
0,07
0,09770093
2.02599656
85802,88735290
0,00
3.15217222 1,b510170
2.17103585 2.26093929
0065680.59008907 1176785,23290072
2.11 0.56
3.98096213
2,2*301026
6438857.81560035
3.00
3.27320906
2.55209569
3398976,285605*7
1.60
-1.69520030
2.31338460
1021788,66909007
0.67
-2.23001050
2.17257979
2200663,05116615
t ,06
-3.05399306
2,26010687
0930009,03715663
2.33
-6.11185206
2.11505917 12071117,55822123
S.8
1.28307979
2,26300033
680720.51989377
0.32
-0,05560924
2.30570213
1192,27073708
0,00
-5,64006068
2,30070566 13620831,70826513
6.43
2.00046421
2,38759380
1507204.56762722
0,73
1,67361187
2.36720273
1300916,76800407
0.62
/TT, 8-0843 87 8
2,31690781
902*131,86305352
0,05
K 1,3568 7TS9~
0.65620300
9*62332,18832*36
0,07
0,0000 0,0761 0.0156
0,0673 0.7672 0.03*7 0.2206 0,9069 0.8038
o.ooio 0.0572
0.2060 0,7210 0.2711
0,0856 0.0183 0.0333
0.4625 0,5298
0,0675
0.3151 0,1797
0.7B77
0.8005 0.14*7
0,0562 0.0810 0.2050
0.127 0.3038 0.1273 0.0153 0.5709
0.9811
0.0113 0,3929
0.4326 0.0350
0,037
TEH 0412324
DUP050453376
S AS BACKWARD ELIMINATION PROCEDURE FOR DEPENDENT VAPIABLE BP25YS
isjo3 Sa t u r d a y ,-
363 OBSERVATIONS DELETED DUE TO MISSING VALUES,
ILL VARIABLES ENTERED
R SQUARE a 0,32961740
CtPJ
74.00000000
DF
WEIGHTED S3-
MEAN SQUARE
F PR0B5F
REGRESSION ERROR
TOTAL
T3 2181
22S4
5705530176.49210000. 11604923**4,061f>7300 17309554062,49377400
78157947.65057672 5320506,13663534
14.69-
e.ooot
in t er c ept
AGEYPS AGESO BMl
LOGZINC
ALB9Q CALCIUM
Fa mmYPER
COFFEE
h eig h t
HYPtflMED LOCI L0C2 L0C3 LOCR L0C5 L0C6 L0C7 LOCB L0C9
LOCIO
LOCH L0C12 L0C13
L0C14 L0C1S L0C16 LOCI?
LQC1.5 LOCI 9 LOC20
L0C21
L0C22 L0C23 L0C24
L0C25 L0C26 L0C2T
L0C2B L0C29
LOC30 L0CS1 L0C32 CtJCTJ
B VALUE
6P.66527993 -0,55637904
0,01005424
1,26135261 >7,54644441
0.00423317 0,00000207 2.62340741 -0.3164J425 16,40235134 B,56031643 0,07561466 -1.31057653 3.42504684 5,62149276 3.36737S0B H.562B0006
13.06646260 B,36667735 4.14688093 3.71593115
11.02845407 5,02675446
10.67033801 -7.40485444
7.47774743 2,10879164
?,Se946424
4,61732018 1.45S54687 -2.12573556 13.20442516
1.49140175 3.41627030 0,00000000 4.46105234
15.44117655 -3.71546473
3.16602712 8,24578253 4.65140170 -11.38317629 .17349105 0.94039478
STD ERROR
TYPE II SS
0.1S135548
71644916.5972949
0.00166302 194470981.4337965
0,08553166 1157103342,7416376
2,13428701
67410144,5879553
0,00124741. 26946376*.9567338
0,00000115
17034925.7004255
0,67705819
92522662.6015374
0.10314561
50704920.2423381
4.86387318
6P5Ct312.1259693
1.24160395 252910271.9338251
3.27795639 ......... . 2*46,5092051 .........
3.27076207
854239.B2SBJ33
3.82136504
4274266,7123984
3.11043629 3,52506616
17372968.1671197 4655080,8176897
4.17152910
6365393.6983994
3,48200697
75174569.5689692
3.20214854
36322526,0425819
3.75eC384
31552675.4506396
3.275742U
6846503,8945634
3.34447365
57853097.4325694
3.41760209
11519425,3977995
3.43946493
5120663,7253747
3.33857897
3,67311003 3,50336682
29827575.2049695 22050917,1425389
1927742,7677569
3,28599581
28381*81.0946522
4.00023235
307S3224.8207819
4.0333061S
1250739.9693566
3.47857190
1986*70,0064757
3.53371568 742*9723.4674795
3.29860115
1944600.3371839
* 3.64484724
612421.36374BB
0.00000000
0,0000000
3.57899727
10222987.5357586
3.73916052 3,45940999
90733049.1130894 6137269.0594331
3,04200076 3.49453554
5763198.4230455 29983942.9248391
3.56195794
9074786.8565353
3.52101718 3.27452993 3.65088943
821056.3899018 8642790.6472009
353001.0554741
F PROB>F
13.51
36.55 217,48
12,67
50.65
3.20 17.39
9,53 11.37 47,54
0.00 0.16
0.B0 3.2r 0,91
1.20 14.13
6.83
5.93
1.29 10.67
2.17
9.62 5.6! 4,14
0,36
5.33 5.78 0.24
0.37 13.96
0.37
1.15 0.00
1.92 17.05
1.15 1,08 5.64
1.71 0.15
1.62
0.07
0.0002
0.0001 0,0001 0.0004
0,0001 0.0737 0.0001 0.0020
o.oooa
0fc0001
- 1T.981S 0.6887 0,3702 0,0709 0.3396 0.2742
0.0002 0,0090
o.oiso
0,2568
0,0010 0.1413
0,0019
0.0180 0,0419
0.5473 0.0210 0,0163 0.6278 0.5412
0.0002 0.5455 0,2827
1.0000
0.1658 0.0001 0.2829 0.2981 0.0177 0.1917 0,6945 0.2026 0.7668
TEH 0412325
DUP050453377
s* 5-
'lQZU
t(.0C3S f L0C36J(.0C37 `LQC38 L0C39
LUC40 cac4t LQC42 LQC43 LOCUM L0C45 LQC46
L0C47 L0C48
L0C9* L0C5OU0C51 .0CS2_0C53 .0C54
.UC55 .0C56 .0CS7 .0CS6 .DC59 .OC60
,0C6t
.ocsa
,OCt>i
L
8ACKU1R0 ELIMINATION PRPCEOUBE f OR nERENOENT VARIABLE 8P23VS
3,95356696
3,96290358
5215405.3978815
0,98
14,64610695'
3,40422681-
98482824,0688326
18,51
8.13974344
3.39307662
30618721,0913197
5.75
0*66909118
3,74567783
169770.6919515
0.03
-3,20711768
3.61302902
4192178.2789874
0.79
1,27751177
4,20475179
441136.3052976
0,09
-3,fl909iM2
3,50952504
S264043,8996866
0.99
5,79773760.
3,52221323
14415800.7081938
2.71
6.96944008
4.44877124
13057944,1565235
2,45
5,37631303
3,39659674
13320204.3808075
2.50
6,93163813
3.04332687
21560905,7299683
4.05
6,9722976$
4,91005462
10728329,8367545
2,02
6,07170399
3,66047538
14638586,6946077
2.75
0,69908015
3.62664308
197673.9145917
0,04
3,65296365
0.21151933
4002819,9489516
0.75
-0,85162709
3,16639827
108102.1038044
0.02
9,56659367
3.41668249-
41729187,5448203
7.84
8,55811629
3,55133139
30897754,3000946
5.81
8,88880520
3.573BOOMI
32881229,7480363
6,18
-3,37b02516.
8,00329120
3783615.2671272
0,71
0.139531b3
3,63940091
7820.5712274
0.00
2,19689008
3.80914779
2209410,2996625
0.42
6,28675359
3.56546229
16331647.6917072
3.07
-0,31787352
3,32811666
48413.9893740
0,01
8,67887851
3,56204940
9176225,4607176
1.72
1.79667969
3,70030136
1251619.3392221
0,24
-5,93015017
3.63002446
14199244,1924943
2,67
3,11715744
3,75800397
3660631.0237307
0,69
22,27478910
3.76929026 185806535.5314219
34.92
,lT."0T26Ci5t3
3,65657854
48259550,3693793
9.07
V3.233571^
0,99009004
56750332.4079699
10,67
-A*
0.3222 0,0001 0.016S 0.8582 0.3748 0.7613 0,3200 0.0999 0.1174 0.1137 0.0442 0.1557 0,0973 0.8472 0.3858 0,8867 0.005! 0.0160 0,0130 0,3991 9,9690 0,5194 0,0799 0.9240 0.1892
0.6277 0.1025 0,4069 0.0001 ,__fi., 00 26
'-.-.5.0'OT-L.,
\*
TEH 0412326
DUP050453378
TABLE 6 ALL MEN OVER 20 BACKWARDS STEPWISE REGRESSION USING PSU DUMMIES
FINAL MODEL
hOR "Kl'ASTCJL-Fe^SEd&D- PRESSURE - 0UsS=oF-=SSHA5ST0N-
TEH 0412327
DUP050453379
N 30984.07
[ABLE LDC25 REMOVED
r eg r es s io n
er r o r
TOTAL
in t er c ept
AGEYRS AGE SO OMI L0G2INC LOGVITC ALOSU LSCHOL SHOKNUH DIETVITC 2IMC H'10GL09 t r ic EPR Fa h h y PER COFFEE HEIGHT h y p EHHED RACE LOCI LOCO L0C5
t 0C6
LOC12 LOCI 3 LOCI* L0CI6 LOClT LOC20 LOC21 L0C26 L0C2T L0C28 LOC31 " L0C32 LOC35 L0C3*
cocao
LOCO] L0CS4 LOCS5 L0CS6 LUC57 LOC60
LOC6J tv.
Sts 12110 FRIDAY, HAY
b a c k w a r d e l imin a t io n
pr o c edur e
f
ok
r n
^p e n d e n t
v ar iabl e
b p 2o ia
R SUtURE 0.3627409*
C( ?5
48,90856800
OF
WEIGHTEO s s
MEAN SQUARE
P PROB>F
04 2204
220B
266234?J63,5*890460 9677171011.50360900 7339513375.10255900
60507780.99086510 2122128,90812323
26.51
0,0001
B val ue
100,9*401262 0.7P93O177
-0,00739207 0,7706097#
-36,20159081 1,16566121
0,00059153 3,06935119 -0,09798155 -0.00971631 0,31737709 0,11119212 0,01056109
1,91*489872 -0,1*276518 12,69012060
5,91269196 2.00802310 -3.53217991
3.06729905 .07620996
-9,73*89621 -5,87191077
9.72879757 10.85990913
-6,82126863 6.19070813
-6,10692957
8,95907897 6,77267560 -3.91891908
3.19973255 -6.33*69853 ? -9,25113369
5.92811798 -9,55*88269
-5.57607230 3,96613059
-3.65280323 -3.89819256 -5,09667032 -6.76559099 -7,67222959 (-3,81788600
'V 1.60876970-,
STO ERROR
TYPE II SS
0,09916758 0.00107530 0.07168066 8.59193255
0.25502348 0.00061911
1.06996276 0,01598733 0,00256981 0,09238299 0,02099953 0.00392308 0,92137356 0,06675251 3.06255S79 0.78157809 0.76155029 1.93555917 1.28560003
1.68056160 2.16817990 1.5*921830
1.57299600 1.98933992
J.60312131
1.99392882
1.61951959 1.69707199 1.81585135 1.59633*93
1.22275273 1,66620912
1,93957761
1,55922876 2,19778962 1,69296259
1,64979663 1.74339999 1,55999119 1,67352039
1.98615907 1,7360607a
1,84710*93 0.61539603
139936852,55639*29
1002866*3,190**166 25?67944.965*3938
37670313.55069997
,!3 3 J 2459,21500187 65237303.93d9lS12
17644277,31965757 19035196,29665966 299(55963,79759537 28302106.7620*718
59497757,066*5*52 15379265.38717077
47068189,30499111 15*08289.10193612 36151531.89256*95 101783565.05159897 19008615.52153571 12847359,24709089 12238146.19025992 1SO55069.56827486 10141722,89037166
291491*3.302030*8 19190792.00509666 112a33716.87977850 36573178,68431681 3903*664,36237331 30362032.61644423 <2717476.81168893 29521037,74503528
9734139.H100936 14036596.31243235 30721591.72252216
18505934.37088000 25864528.0565*11*
*135003,*5078636 24476480,1257928*
9367570,64*16057 931601*.*88430*6 13345612,92334496 14296321.48*00277
44240424.0465*173 41946227,24333092
9066351,4660*483
F
63.35 47.26 IIS.5* 17.75 20.91 30.74
8.31 6.*7
14.13 13.34 28.09
7.25
22.1* 7.50
17.04 97.*6
6.60 6.05 5.77 7.0* 4,78 13.79 9,04
, *3,IT 17.23 18.39 14,31 29,55 13,91 4,59
6.61 14,48
6.72 12.20
4.30 11.53
4.91 4.39 6.29 9.09 20.85 . 19.53 4.27
PRne>F
C.OOfll 0,0001 0.0001 0.0001 0.0001
0.0001
0.0040 0.0028 o.gone 0,0003 O.OOOI 0.0072 O.OOOI 0.0062
0.0001 0.0001
0.O103 0.0190 0.0169 0.0078 0.0269
0.0002 0,0027
0.0001 0,0001 0.0001
0.0002 O.OOOI 0.0002 0.0323 0.0102 0,0001 0.0032 0.0005 0.0381 0.0007 0.0356 0.0363 0.0122 0.0026 O.OOOI O.OOOI 0.03B9
'** r ruW*>*iuctuR FO<*' JN t k j nonet ARE s ig n if ic a n t a t t h e 0,0500 LEVEL.
***Ot IT v a *i
jT
TEH 0412328
DUP050453380
SA 15*03 SATURDAY. MAY
BACKWARD ELIMINATION PROCEDURE FOR DEPENDENT VARIABLE BP2SYS
BACKWARD ELIMINATION PROCEDURE FOR DEPENDENT VARIABLE H2SYS
TABLE L0C56 REMOVED
R SQUARE * 0,31167100
CCP)
52.386069*7
DF
WEIGHTED 6S
MEAN SOUAKE
F PRCB>F
REGRESSION ERROR TOTAL
in t er c ept
a &EYRS AGESO BMI LOGZINC ALWSQ Fa ^h YPER COFFEE HEIGHT
kyper heo
LOCO L0C7 iocs L0C9 L0C11 L0C13 LOCI* L0C15
l o c it
L0CX6 LUC21 L0C26 LOC29 LOC35 L0CI6 LUC03 LOCOO LOG 50 L0C51 LOC52 LQCpO L0C62 L0C63 LL
33 2221 2250
5399866631.19056100 11910667031.30321300 1730955*062.09377000
163061013,06638060 S36OS50.80705232
B VALUE
STD ERROR
TYPE II 3S
72.20982616 0.52730002
0,00966701
1.28191506
-8.19288700 0.00916608
2,80299053
-0.29907525 16.79858001
8.59580700 0.33098536 11,61710190
6,7023121 7,84557880 9,06781070 9,33629299
-9.05771073 5.93242296
6,07906*50 8.28036026
11.98921092 10.18080003
6,83065165 13.16228070
6.65802703 6,76155512
5.21595603
8.02106030 7.06036720
7.05551905
-7,61015621 / 2O-J-92.193095
V 9,3(j 133216 vx, 3,20100279'
0,15002759 0,00165050
0.08065403
2.07327073 0.00126225 0,66765005 0,10176160 0,80638630 1.23900570 2.02O53993 2.55C55790 2.13625837 2,91266388 2,35018092
2,06959596
2.35313311 2.80230658
2,28380963 3.22702036 2.60532623 2.87990230
2.56283856 2,03369562 2.00628989
2.05118506 2.09685350
2.07129013 2.6*776907 2.67637202 2.70966310 2.93*26037 2.80202935 0,93379105
65926573', 0 693601 {S31057S2.1691326 1230130101,2606088
83770979.0769000 '82863396,(1977691
97212107.8500078 U6336S4U.0585309
65530181.3806267 .'S2o?3738.1819132
20595516.1009731 111290626.7959806
52606*80,1449716 38922667,0009087
87062031.3756729 7576302.091800l d66S8751,9869679 2O0O1778.Su506S8
36007723.4137538
35320661.*311072 113603218.0181303 138100112.3053030
38150788.0919663 156688*31.0509377
39678109,9750917 00820126.8747989
23010758.5970533
56513201.4170752 38103555.0705802 41629022.9952093
410*2001.6617206 272734198.8662520
59860360.9572330 63000103,1427572
30,07
0,0001
F Pftf)B>F
12.29 3o.l0
229.31 15.62 52,73 18.12 8,60
12.22 07.13
0.56 20,75
9,80 7.26 16.23 10.07
16,15 4,08
7.08 6,58 21.18 20.26
7.11 29,25
7.00 7,61 0,36
10,53
7,11 7.76
7,66 50,80 11.16 11.75
.
O.nOoS 0,0001 0.0001
o.ooot 0.0001 0,0001 0.0033 o.oooS 0.0001 0.0320
0,0001 0.0017 0,0071
n.nool o.ooo2 0,0001 0,0300
0.0078 o.ntoo
0,0001 o.dool 0,0077 0.0001 0,0066 0.0059 0,6368
0.0012 0,0077
0.005* 0,0057 0.0001 0.0009
0.0006
THE h q o EL ARE SIGNIFICANT AT THE 0,0500 LEVEL.
TEH 0412329
DUP050453381
T
TABLE 7 MEN 40-59 BASELINE MODEL
NOTE: RECEX means recreational exercise
TEH 0412330
DUP050453382
N 30984.08
4
TEH 0412331
DUP050453383
TABLE 8 MEN- 4.0-59 WITH VARIABLES FOR EACH STRATA
NOTE: THERE ARE ONLY AN AVERAGE 'Or 20 OBSERVATIONS PER STRATA
TEH 0412332
DUP050453384
N 30984.09
-iFCE
-H-'r
r-E_ r;Z'~' 7Z~A.Z
j-2 r-s
r0
F.2DT Z EC, V.
SL:M CF 3GjAr.ES
9': 4425272 :3-M .370 = = 270= ::'=297~
:_65.5CS 2~.e z o ie e
2CS:.i7Z
17EAN S 22 A. RE
21533=4= 7117019
F-EGUAFE a d : f .-s q
F VAL2E 6.909
0.3342 0.2253
F-";GE: ;-F 0, 0 I 01
TEH 0412333
DUP050453385
r" VARIABLE: BP2SYS
JRCE
DF
DEL :.'0R TOTAL
43 577 620
ROOT MSE DEE MEAN C. V.
SUM OF SQUARES
2037304622 43599-S3S44 6397488466
274S.872 12E.969
2171.426
MEAN SQUARE
4738382B 7556298
F.-BQUARE ADJ R-SQ
F VALUE 6.271
0.31S5 0.2677
F'ROB>F 0.0001
7IABLE Dr
TERCEF' ZYRS ESQ I BUM IN
1 1 1 1 1
BRIE'Q 3OLEIC EnC -- EE IEHT ^E^'ED
f a e a me t e r
ESTIMATE
-BO.312400 2.I85243
-0.D13EE0 1.310001 0.8=6=48
V`Ti -4
ri tcs.:~ ~0&C5E
24
--. ar
Co
C
:~17 =
*i. ~r C- C
- r.*
S=37 C-E cr -- r* rr -
34==0=
r* - z. sr " *
C cc. .* r 240194
STANDARD T FOR HO: ERROR PARAMETER*
36.737446
0.021152 0.167006 - * 4L. /
SEE;AS SE=SA2 4437-4 1 E2"*77
-1.419 1.044
-0.731 T w'~r*tr 3 -->E
.: s
PROS
IT!
0.;36e -- . 2=C'S
- ' -r'Z- - 1 v. :* : ji
0 001
~ T . .
OvOl
= ^ . ..
4215=1 19222= BE~ESS 211=64
7630C3 "7 4721 0S6S64 Of -- c
"201 = r = -
.
0=4 4=7
-c
1
.74
--. - '* CT
274 1
- *T "
1 vC3
= : = s
5. 7 -7 .
7:-: 7612
:s
1"H 0412334
DUP050453386
TABLE 9 MEN 40-59 MAXIMUM R-SQUARE REGRESSION WITH DUMMY VARIABLES FOR EACH ................ PSU"
NOTE: DIASTOLIC ONLY SINCE LEAD IS LESS SIGNIFICANT FOR DIASTOLIC
TEH 0412335
DUP050453387
N 30984.1
wr/*4' <* $. . l~orce. d. tuj
VARIABLE L0C26 ENTERED
REGRESSION ERROR TOTAL
in t er c ept
a SETRS AGESQ 8HI POTASS HE.MUGL08
a l b u min
LVC RECEI HEIGHT HYPERHED' L0C6 L0C14 L0C17 LDC21. L0C26 L0C35 L0C57 LL
.* i-oove-e*iT f o p Or-tsee-T. V4,,likBL----
THE FIRST 3 VARIABLES IN EACH MODEL *t INCLUDED VARIABLES.
R SQUARE 0,30924405
CtP)
27,75426056
DF
WEIGHTED SS
HEAR SQUARE
F
10 861006163.69358740 *7633675.76075485=
625
1923222571.80221640
3077156.11488355
643 2784228735.49580360.
IS, .5*
B VALUE
STO ERPOR.
TYPE II S3
F
11.47281798 -0,28123.929' 0.00415042
0,6*246996'
-0.00166657 0,08576666 0,52529679
1,70571254 1.37442616 12.62616933
6,38983766 -6,92974608
-0,2900505 6,61821025 7.78313507
6,23449203 8,93197640 -6.6020*705 3..98330129
1.29407341 0.01308609
0,10393166 0.00034692-
0,036*2593 0.14559608 0,35504319
0.58373132 5.90866370
1.32652*96 2,69963307 2.72378689
3,15530*85 2,8*939928
3,09672376 2.96206426 2.76564613 1.12357019
145339^50564350 309537.06533097
2?l857J15.17565195 71012691.99678090 16690874,16318357
00055270.0B795614 71022973.21468144 17059502,32674835 14051225,51511514
71399568,1B353*69 20275645,67244*71'
28504775.60473393 13537762,78752019
22327677.10879-137 12*72308.87990160 27972467,883*1286 17532777.70949686 38675454.61211582-
0.05 0.10' 72.10
23.0* 5.42'
13.02 23.OB
5.54 a.57
23.20 6.59
9,26 4,40
7.26 4,05 9,09
5,70 12.57
3VE HODEL IS THE BEST 1ft VARIABLE HtlDEL FOUND,.
PR0861
0.000
PR08>!
0.*2*> 0;751 = o.oon 0.000 0,020] 0.000' 0.000 0.01fl: 0.0331 o.ooo: 0.010' 0.002' 0.036' 0.007: 0.044! o.oo2: 0.017: 0.000-
\~^h O L>/~ Ccp^.
F o ratdL t <n ARIABLE I.0C42 ENTERED
REGRESSION ERROR TOTAL
A IIU" P-3QIIARE INPWCVENCNT FOR OCFCn Cn t VARIABLE 8PSDIA
R SOUAPE 0.302*7823
C(P>
27.77826628
WEIGHTED S3
MEAN SQUARE
15 628
643
**** ! I'll o&ct
|941503322,7t6P975C 278422*735.495603BO
56181690.18531370 3091565.60050330
r
18.17
PR08>F 0.0001
INTERCEPT BMI
Po t a s s
HEMDOLOB
ALBUMIN
LVC 6CEX
h y per mED
L0C6 L0C1 LOCI7 L0C21 L0C35
L0C2 L0CS7
LL
8 VALUE
9,85774032 0.88809626 0,00170592 0,08210385 0.9*612761 1,71169697 1,95381030 6,98370056 >6,73166920 8,76951395 6,77357701 7,90896090 9,21958185 9.30906810 -6,S3139t05 3.73210300
d e l IS THE b e s t is VARIABLE MODEL FOUND,
STO ERROR
TYPE II SS
0.10000582 0.00030561
0.036*7555
0,102*9*03
0.35897701 0,58150179
!.31090719 2.7019;*3I 2.72166101
3.15507006
2.88939125 2,95897290 0.063*0007 2.77086901
1,12610736
225015263.10063003 753220*1.05832750 15325800.69317285
3577*665,80502016
71883725,293*1513 19321218.10979101
*7208739,7000067219190200,30072010
52096780,95383639 10205750,76217950
23160619,70698003 30013659.770*7770 13561165,99\17B69 17177019.89723579
33955070,90956801
F
72.91 20,36
0,96 11.57 23.25
6,25 28,21
6.21 10,3*
4.61 7.09
9.71 0,39 5,56 10,98
PR082F
0,0001 0,0001 0.02*3 0.0007 0.0001 0.0127 0.0001 0.0130 0,0013 0.0322 0.0064 0.0019 0.0366 0,0187 0,0010
TEH 0412336
DUP050453388
TABLE 10 MEN 40-59 WITH ALL PSU DUMMIES BOTH RACES AND WHITES ONLY
TEH 0412337
DUP050453389
N 30984.11
4 *-
r^tcsb-
BACKWARD ELIMINATION PROCEDURE FO* DEPENDENT VARIABLE BP2DIA
27 OBSERVATIONS DELETED DUE TO-MISSING VALUES,
ALL VARIABLES ENTERED
R SQUARE a 0,38200238
CtP)
77,00000000
OF
WEIGHTED SS
MEAN SQUARE
F
r eg r es s io n
ERROR TOTAL
76 567
63
1063582002,0*193240 1720646733.41387140 2784228735,49580380
13994500.02739385 3034650.32348125
4,61
B VALUE
STD ERROR
t y p e II SS
F
INTERCEPT tCEYRS
ir.fsa
8 NI
po t ass
h Emo Gl QB
a l b u mi n
L VC / FAHHYPER' RECEX
HEIGHT ^ COFFEE ^
MtPERMEO
LOCI L0C2 LOCI LOCO LOCS L0C6 LOC7 LOCS
r ac e
L0C9
L 0C10
LOCI 1 LOCI2 LOC13
LOCI* LOCJS
LOCJB LOCI 7
LOCI* LOCI** LOC20 L0C21 L0C22 LOC23 LOC2R LOC25 L0C26 LOC27 LOC2B L0C2R LOC30
0.5?28275 -0,48574409
0.00603500
0.900J7R61 -0,00144694
0,07638326
0,51302741
1.S8045271 1,31276676 1,02633604
10,63556085
-0,04363393 5,82838780
-1,86795664
6,0(067026
6,07682353 2,62683322 t.51626556 3.75935696 5,30607639 4,10415034 2,16204678
4,78449883 4,81859349 5,26055324
0.95023592
6,18651602 5,77649415
4,47396307 -0,24694181
9,50496000 O.05060915 >1,6061024a -4,43841262 10,91102216 4,56852308
2,73124233 0,00000000
3,12745094 8,93639767 0,82330405
6,03690673 6.47290332 -3.77154839
1,34130925 0,01356662
0,10836902
0.00038182
0,04120672
0,15830173 0,38003274 0,69680608 0,62889243 6.24208067 0,12152635 1,39917339 5,3624]481 5,14466239 5.52242274
5,35320326 5.10403129
5,02004428
5,62689329 4,85166799
1.79374338 5.19808529
0,80909532 4,96137243 5.06298849 5,18786644
5.02270653 5.50763718
5.7819087$ 5,26346690 5,82999857 5.96135824
5,87964939 5,13229909
5,08923096 5.75296941
0,00000000
5,36524853 5.23715463 5.19353665 5,10831174 5,55437141
5.62721158
397960.S0505190 600509,92256316
209390353.45191064
03700606,14247790
10427226,41879762 31872671.41465214
52484327,70790772 6473879,94987942
6113832,39139423 8809927.36980361
1801499,84466302
52657770.58537715 376159,36074467
4128520.27253775 3674536,84272400
73182S,02711476
?6713,54805451
1701847,71007995 2696557,10056214 2171565,98618686
4408777.77042833 2570956,68011221 3046649,29152956
3384335.97108517 106055,52236057
4318210.19498000 4013855.41309325 2002455,56924847
5535,48193519 9896131,34622148
3268877,67245212 . 220285.71161760
1729263,37909696 13715625,43179490
2045454,19727164
683982,50S51611 0,00000000
1023477,18528437 8835736.13284370
76261,04567581 4238213.22173189 4121326,69786889
1363206.14612696
-
0.13 0,20
69,00 14,40
3,44
10,50 17,30
2.13 2,67 2,90 0.59
17.35 0,12 . 1.36
1.21 0,24 0,09 0,56 0,89
0.72 1.45 0,85 1,00
1.12 0,03 1,42 1.32 0,66 0.00 3.26 1,08 0.07 0,57 4,52 0,61
0.23 0,00 0,34 2.91
0,03 1.40 1.36 0.45
PRC8>F
0,0001
PRCB>F
0,7174 0.6566 0,0001 0,0002 0,0643 0,0013 0,0001 0.1447 0,1026 0,0890 11.4413 0,0001 0,7249 0,2439 0,2716 0.6236 0,7665 0.4542 0,3463 0,3980 0,2286 0,3577 0,3168 0,2914 0,8518 0,2334 0,2506 0,0169 0,9659 0.0715 0,2998 0,7877 00,140530396^/
0,3697 n.6351 1.0000 0,5616 0.0885 0,8741 0.2378 0.24 0.5030
TEH 0412338
DUP050453390
0C31' OC32' QC33 0C38
0C35 0C36 0C37 0C38
0C3 OC80 ocat
0C2 0C83 OC08 ocos oca 6 0C87 0C8OCflO-
OC50 0051 0C52 0C53 ocso
0C553C56 DCS7 OCS* ocso 3C60 3C61 DCb2
3C63
.
i
9*9i .......................................
B*OKI.*ft0 ELIMINATION PRnCEDURE FOR p e p e h d e n t v a r ia b l e b p 2d ia
-6,529U62
6,69214680
2909813.70650525
-2,85013828
5.30756384
875080,60975882
5,70468850
5.18317708
3676036,63064262
1,05745272'
7,15956500
66199,88620316
12.16507430
5.13110342 17057283.95762605
5,6909963
5,31178357
3083009.55931570
6.67305207
5,41106369
7795120,66486085
-2.63522562
5,31561086
745826.65819200
-3,30192524
5,70484256
1016613.32990275
-2.70192746
5.79214015
660355,65706877
8.52778557
5.55939636
7140455.57836950
10.95563094
6,15295611
9620909.02861167
8,41626295
5,52511170
7041098,88958935
3.19820300
5,20931158
1126060.59696621
-1.08418376
5,13623016
135210,98309585
2,56770372
5,50378108
660505.00172529
-1,13267793
5.00201350
133416,70110290
-0,91711028
5,50252832
80299,92806507 .
a,9a 307831
0,92516125
3056769,7720330*
0,70056806
5,33660431
2390656.10300305
8', 15209498
6,12757699
5371178.20003860
a,03701090
5.30300396
1758180.51680590
9 .,97429060
5,95609054
6509255.13160182
1.86881799
5,81582081
361337,60615*17
1,21800050
5.08806308
173*99.62026560
1,75790008
5.11229621
358828,35102870
-3.66573809
5.08168036
1579126.51570013
2,96656878
5.05095621
1059299,83100275
0,43267638
5,60958810
1866156,62668942
-1.16069108
6.19196285
107368,38975175
0,79376863
5.00692095
65802.72396231
1.03986852
5.55899298
106105.79076517
^T6^<70-7563
5.S3553269
2566290,7*230978
_ \^3.261aa5i.
1,27156596 19968151.610*1547
10134 S-TU-OAt,
0,96 0.29
1.21 0,02 5.62 1.15 2.57
0,25 0,38
0.22
2.35 3, IT 2.32 0.37 0.08
0.22 0,04
0,03 1,01 0.79 1.77
0,56 2.80
0.12 0,06
0.12 0.52 0,35 0,62 0,00
0,02 0.03 0,65
6,5*
0.3286 0.5915 0.2715 0,6826 0.0181
0,2800
0,1096
0,6203 0.5630 0.6410 0,1256 0,0755 0.12*2 0,5026 0,8329
0,6410 0.8300
0,8677 0.3160 0,3787 . 0.1839 0.4069
0,0906 0.7302 0.8109 0.7311 0,0710
0,5549
0.0330 0.8509
0,8833 0.8517 0,35*2
0.0106
TEH 0412339
DUP050453391
w !-/-. O*?/^
BACKWARD ELIMINATION PROCEDURE FOR DEPENDENT VAPIA8LE BP2DIA
22 OBSERVATIONS DELETED DUE TO HISSING VALUES,
ALL VARIABLES ENTERED
p s q u a r e 0.40243598
CCPi *
76.00000000
REGRESSION ERROR TOTAL
DF WEIGHTED s s
75 95011792S.04311530
49o 1410798998.14780260
565
2360916923,19091790
s val ue
STD ERROR
MEAN SQUARE 12666239,00057487
2879181.62667307
TYPE II SS
F 4.40
PR08>F 0,0001
F PH0B>F
INTERCEPT AGEYP6 AGESQ 8*1
po t ass
MgMOGLOS ALRUHIN
LVC FAMKTPER RECEX
HEIGHT
COFFEE
h y p e r me o
LOCI L0C2
10C3 10C4
LOCS LOCO
L0C7
L0C8 L0C9
L0C10
LOCU" L0CI2 L0C13
LOCia L0C15 LOCtfc L0C17 LDC18 L0C19 L0C2D
L0C21 U9C22 LOC23 L0C24 L0C25 L0C26 L0C2T L0C26 tOC?* IPCI! kCtll
*15.53115861 0,00021890 0.00060082
1.02706616 *0.00162430
0.09455545 0.62431881 l.O7724650 1,62216652
0,97559232 11,06529403
-0.06932127 5.16223978
3,90364923
-1.46086500
5,61547031 1,50844026 4,11961733 4,22037842
4,06179825 4,44426969
5,20687078
5,18520477 5,78176829
-1,11479501 5.59916664
-5,56131618 4.71077122
-0.13376697
9,59826693 6,20142849 12.49203991 -3.B0235797
U. 25121231 3,79158712 2,34364452 0,00000006 4.31595619 9,37900552
-0.959211J5 6,31285900 7.32959589
3.3*671S?2 -T.3*210390
1.39362464 0,01409861 0.11996267
0.00039468 0.0*396190 0.16684450 0.40370608
0,93648095 0,66679224
6,62404602 0,12165434
1,47063781 5.53271331
7.79439488 5.61763821
5.37156163 5,18821356
5.00237294 5.57690920
4,73523129
5.11683508 4,73050790
4,86649337
5,20683700 5.19290342 4,902*0390
5.37019547 5.64797415 5.12986884 5.69034534
9,23171657 5.95967461 .5,01002082
'4,99169023 5.74074530 0,00000000 5,30513174 5.19353992 5.25554007 5,12009277 5,22CS390 5,49527956 T.*9919385
0,07103492
9342.5c995766 211216971,81240008
<'.6765170.92372740
13315S01,24462181 !8183S67.33982784 38551703.37637411
10900522.01782180
6126654,00526955 8958767,80597756
934860,79152761
35475943.17560251 1500142,67278087 10I137.S736156 3065540.25221091
?2709,70$67375 1615294,64573314
2049366,902040 36
1527279,78881401 2536246.07746409
2481397.08980647 34S276.95778767
4064041.69704815 131980,91423768
3347305. S49f,ft3S6 3704547.70100482 2215505,84949440
1615,03317796
10079574,22411902 34,9599,86844484
5271944,90963170 1172006.35843882 14520736.39544020 1661176,92358605
479861,50212562 0,08000000
1905595,73313799 9369616,34576301
95909.76728394 43768*5.34162304
S?61M. 73178073 l?676*,3r565009 2797S*P.5*137405
o.oo 0.00 73.36 16.94
4,63
6.32 13,39
3,79
2.13
3,11 0,32 12.32 0,52 0.04
1.07 0,08
0.63 0.71 0.53 0,88 1.04
1,20
1.41 0,05 1.16 1,29 0.77 0.00
3.50
1,19 1.83 0.41 5.04
0.58 0.17 0.00 0.66 3.26 0.03 1.52 l.3 0.37 0.47
0,9994
0.9545 0.0001 0.0001 0,0320
0,0123 0.0003 0,0523 _>0,1453 ___> 0.0784 0,5691
0.0005 0,4707 0,8514
0.3011
0.7790 0.4276
0,3993
0.4668 0.3484
0.3094
0,2736 0.2354
0,8306 0.2815 0.2572 0,3808 0.9811 0,0619
0,2763 0,1766 0.5238 0,0252j 0.4479
0,6633 t.OOOOi 0.4J63' 0.0715 0.*553 0,21*2 0.1T7| 0.5628 0.32*8
TEH 0412340
DUP050453392
QC32
0C3S 0C34
0C35 0C 36 0C37 0C38
0C39
ocoo
0C41
0C2 0C03 OCOO
0C4S 0C6 0C07 0C8 0CO9
ocso
0C51 DCS 2 0C53
ocso
0C55 0C56 0C57 0C5B 0C59
OC60 OCM / 0C62
Zth* w - Lvv
: ,
BACKWARD ELIMINATION PROCEDURE POP DEPENDENT VARIABLE 8P2DJA
-2.39163565
5,17407697
615167.02270385
1.86112366
6,05734604
271803.31811939
9,79035801
9.00980235
3399631,62662185
12.53069777
5.06670677 l759616S.65437057
6,35988737
5,20914929
4226539,02660590
5.61037350
5,56359001
2911016.18850998
-1,70270172
5,19193686
309676,16967779
-3,0672240$
5.56057766
870773,56189109
-2.39272704
5.65205798
511687,05656770
9,07899306
5,02663630
8059030,98353789
11,58104670
6,00850974 10696979,22299960
7,37451385
5,66238757
4609205,26305587
3.13493436
5,37951096
977778,24880311
-2,02777066
5,14576002
600892.69930336
6.28273631
5.67822158
3524861.48301037
-0,17287070
5,46166989
2884,20542256
0,00736736
5.69055030
199,4889.0965 . . _ ...
5.08383281
4,8*929100
3112861.56393*83
0,96770550
5.20701113
2620618,13051785
9,68633360
6,80476608
0638977,63010393
0,38646005
5.17712260
2066923.84202778
11.00723274
5.81329SS2 10322395.4*686010
2,37011915
5,28890763
578199.15874103
1,77749695
0,96444305
369309.06720238
2.05282770 3,12523166
3,30308692 4,62014905
0.96510143 0.960S9150 0,93161101 5,50961850
088231.79911430
1142790.0*660017 1323398,66410382
2024637,97840966
-0,99743512
6,90066430
76831.90512597
1,0930329a >^1462770
/6.047S52J9 U, 1191771K
5.33256606 5,42055295 5,06748400 1,37020*01
12096S.92171203 101275,81846951 3497000,32607091 10919371.30018158
0.21 0,09 1.18
6.11 l.7
1.01 0.11 0.36 0.18 2.80
3.72 1.68 0.30
0.22
1.22 0.00 0.00 1.04 0.91
1.61 0.72 3.59 0.20 0.13 0.17 0.40 0.06
0.76 0.03 0.04 0,04
1.21 5.18
0.6401
0.7588 0.2777
0.0138 0.2262 0,3151 0,7031 0,5617 0.6735 0.0950 0.0505 0,1950 0.5663 0.6373 0.2691 0.9706 0.9934
0,2990 0,3005 0,2009
0.3973 0.0589
0.6503 0.7204
0,6807 0.5290 0.0981 0.0021 0.8703 0.8377
0.PSi3 0.2710 0.0233
TEH 0412341
DUP050453393