REPORT OF ASSESSMENT OF A YIELD INVESTIGATION
OF SIX-YEAR SPONTANEOUS ABORTION RATES IN THREE OREGON AREAS
IN RELATION TO FOREST 2,4/5-T SPRAY PRACTICES
Prepared by the Epidaniologic Studies Program
Human Effects l^onitoring Branch
Benefits and Field Studies Division
OPP, OTS, EPA
February 27, 1979
-------
TABLE OF CONTENTS
Page
>- List of Tables ii
^
^ List of Figures iv
^
^ Chapter
p<
^ I Introduction 1
-------
LIST OF TABLES
Table Page
1. Selected Agricultural Statistics for Counties
Including the Study, Control and/or Urban Areas 18
2. Zip Codes Corresponding to the Study,
Urban, and Control Areas 20
3. Total Acres Treated and Pounds 2,4,5-T Used
in Alsea Basin (1972-1977) 22
4. Applications of 2,4,5-rT by Day and Month in the
Alsea Basin (1972-1977) 24
5. Annual Admissions and Births in Hospitals in Study,
Urban, and Control Areas 30
6. Number of Hospitalized Spontaneous Abortion Cases for
Study, Urban, and Control Areas < 33
7. Number of Hospitalized Spontaneous Abortion Cases for
Study, Urban, and Control Areas, Accummulated
by Month 39
8. Hospitalized Spontaneous Abortion Cases by Study,
Urban, Control Area, and Age Group of Patient 40
9. Number and Percent of Spontaneous Abortions According
to Length of Gestation in the Study, Urban and
Control Areas 41
10. Total Number of Births by Months for the
Study, Urban, and Control Areas 42
11. Number of Births Corresponding to the Month of
Spontaneous Abortions in the Study, Urban and
Control Areas 43
ii
-------
LIST OF TABLES
Table Page
12. Monthly Spontaneous Abortion Index for the
Study , Urban, and Control Areas
13 . Abortion Index by Period, Area, and Month ................... 47
14. Analysis of Variance of Abortion Index by
Period, Area, and Month ..................................... 51
15. Parameters of the Model "Index = A + B Sin (2T t-D)"
for the various areas ....................... T ............... 57
16. Study Area Index Corrected for Phase Shifts> Study
minus Control 1 month lag ....... ..... .................... 58
17. Study Area Index Corrected for Phase Shift, Study
minus Urban - 2 month lag ...................... ........... 59
18 . Study Area Index Corrected for Phase Shif t, Study
minus (Urban plus Control) --- 2 month lag .................... 60
19. Number of Births per Month and Unweighted Abortion
Index for the Study, Urban, and Control Areas ............... 64
20. Cross - Correlation of Corrected Study Area
Index with Spray Data ....................................... 63
21. Pounds of 2,4,5-T Applied in the Alsea Basin and
Abortion Index for the Study Area by Month and
Period [[[ 39
22. Cross-Correlation of Study Area Abortion Index
with Spray Data ............................................. 70
23. Study and Urban Area Physician Interviews: Estimated
Numbers of Spontaneous Abortions Treated Annually and
-------
LIST OF FIGURES
Figure Number Page
1 . Oregon
2. Study Area in Lane, Lincoln and Benton
Counties of Oregon
3. Land Cover in Lincoln, Benton, and Lane (Partial)
Counties, Oregon
4 . Control Area in Malheur County, Oregon ...................... 14
5 . Land Cover in Malheur County, Oregon ........................ 15
6 . Urban Area in Benton County, Oregon ......................... 17
7. Acres Sprayed with 2,4,5-T in Alsea Basin
Accumulated by Respective Month, 1972-1977 .................. 25
8. Pounds of 2f4,5-T Sprayed in Alsea Basin
Accumulated by Respective Month, 1972-1977 .................. 26
9. Hospitals in Study and Urban Areas .......................... 23
10 . Hospitals in Control Area ................................... 29
11. Plot of Monthly Spontaneous Abortion Index
for the Study, Urban, and Control Areas ..................... 45
12. Abortion Index for the Study Area by
Month and Period ............................................ 48
13. Abortion Index for the Control Area by
Month and Period ............................................ 49
14. Abortion Index for the Urban Area by
Month and Period ............................................. 50
15. Spontaneous Abortion Index, 1972-1977 For the Study Area
and the Combined Control and Urban Areas ..................... 55
IV
-------
LIST OF FIGURES
Figure Number Page
16. Study and Control Areas: a) Combined Urban and Control Index;
b) Study Area Index; c) Urban and Control Index Shifted Bade
Two Months 56
17. Predicted and Observed Points for the Study Area 61
18. Predicted and Observed Points for the Urban & Control Area 67
19. a) Study Area Abortion Index Corrected for Phase Shifted
Urban Index b) Pounds of 2,4,5-T Sprayed by Month 55
20. Cross-Correlation of Study Area Abortion Index
with Spray Pattern in Total Pounds 2,4,5-T applied
by Month, 1972-1977 66
-------
Chapter I
INTRODUCTION
One of the widely used chlorophenoxy herbicides is the chemical
2,4,5-trichlorophenoxy acetic acid, commonly referred to as 2,4,5-T.
A selective herbicide, 2,4,5-T is especially useful for brush control
on rangeland, along right-of-ways, and in conifer forest habitats.
The parent acid is formulated in a variety of emulsifiable esters
and specific amine salts registered for use in both Canada and the
United States.
Many studies have been conducted to determine the fate of TCED in the
environment. Since TCED is reportedly not mobile in soil, it is not
considered to be a risk in ground water. TCED does not accumulate
in vegetation (1).
In comparison with other pesticide products, the phenoxy herbicides are
relatively non-toxic to mammals, and 2f4,5-T is classified as moderately
toxic with an acute oral LD_ (rat) at 500 mg (acid basis) Ag (2).
However, low levels of TCED have been shown to have oncogenic effects
in rodents and related effects in primates. Fetotoxicity and terato-
genicity have been demonstrated at low levels, with the appearance of
cleft palate and kidney anomalies in rats, mice, and hamsters. Embryo-
toxic effects in avian species have also been reported (1).
-------
In July, 1978, Staff of the EPA Office of Pesticide Programs (OPP) and
Colorado State University epidemiologists met in Oregon with local and
State health officials to begin an investigation into a group of women
living in the vicinity of Alsea who claimed they had experienced
miscarriages because of herbicide spraying in the course of forest
management. The investigation was precipatated in late June, 1978,
when EPA received a letter signed by eight women living in this area
who had experienced 10 miscarriages since 1973. The women claimed
to be surrounded by forest land which has been sprayed for years with
herbicides known to contain dioxin. They also charted their dates of
miscarriage and related dates the forest areas were sprayed. Each of
the women was under a physician's care at the time of miscarriage, and
neither the women nor their doctors could ascertain the reasons for
abortion. The investigators met with four of the women to discuss the
circumstances of their reported exposure and any possible relationship
to their subsequent miscarriages.
Following the visit by EPA and Colorado State University staff, an
extensive health questionnaire (Appendix A) was designed jointly by
OPP's Epidemiologic Studies Programs (ESP's) in Miami, Iowa, South
Carolina and Colorado, in which detailed questions on self and family
pregnancy and medical histories, environmental aspects, diet, occu-
pational and household pesticide usage were included. During the first
week of August, 1978, an epidemiologist from Colorado's ESP admin-
istered the questionnaire to nine women who had experienced 13
confirmed miscarriages from May, 1973 to March, 1978.
-------
During late August and early September, the health questionnaire and
related data (spray application, vital statistics, etc.) were evaluated
independently by 10 experts in the fields of obstetrics, gynecology,
epidemiology, biostatistics, reproductive endocrinology, and perinatal
medicine. The consensus of the reviewers was that:
1. the spontaneous abortions did appear to follow a seasonal
pattern (two of the reviewers noted the seasonal relationship
but drew no conclusions);
2. Good Samaritan Hospital records for Benton County women
for the years 1975-77 showed numerically higher rates of
spontaneous abortions per live births during Jan., Feb., Mar.,
and Oct., Nov., and Dec., of each year than during April
through Sept. (these records were based on stillbirths of
terms greater than 20 weeks). Conversely, 10 of the 13 miscar-
riages reported by the Alsea participants occurred during the
months of April through September (1973-1978);
3. there was a high numerical incidence of March to June
miscarriages (nine of 13) among the Alsea participants. However,
there was concern that the reports might comprise a biased
sample (albeit unintended) of all miscarriages that occurred
within the area and years under investigation;
-------
4. a causal relationship between forest herbicide spraying
and reproductive wastage had not been demonstrated from the data
presented. Opinions ranged from "... no evidence of a
causal relationship . . ." to "I cannot support or refute a
cause and effect relationship ..." Reviewers either stated
or inferred that there was no real evidence of an epidemic
based on the data presented.
An analysis of the data by staff of OPP's Human Effects Monitoring
Branch (HEMB) identified:
1) the possibility of a relationship between time of
spraying and conception and subsequent abortions among
the Alsea women; and
2) 'the fact that/ while State and county records of spontaneous
abortions are given for terms of 20 weeks or greater/ 12 of
the 13 miscarriages experienced by the women in Alsea were
for terms of less than 20 weeks.
Based upon the comments of the reviewers and HEMB staff, OPP undertook
to develop data on spontaneous abortions of 20 weeks duration or less
in the Study area and in a comparable control population.
-------
The current study of 6-year spontaneous abortion rates in three Oregon
areas was initiated by the Human Effects Monitoring Branch, Office of
Pesticide Programs, in October of 1978. The study was accomplished
under contract to the Epidemiologic Studies Program (ESP) projects in
Colorado, Florida and Idaho. Scientists from the Colorado Project,
under the direction of Dr. Eldon Savage, organized and conducted the
field investigations, developed the data including the spontaneous
abortion index, and prepared much of the report. Idaho project staff
assisted in collection of hospital data in Malheur County. Statistical
analysis and the interpretative narrative were developed by Drs. Robert
Duncan and Thomas Keefe of the Florida and Colorado projects, respect-
ively.
The following scientist were instrumental in the successful development
and conduct of the study:
Dr. Eldon P.. Savage, Director, Colorado ESP, Colorado State
University
Colorado ESP Staff: Drs. Thomas Keefe, Robert Zimmerman and
Richard Hayes; Messrs. William Wheeler, Lawrence Mounce and Jerry
Rench; Ms. Lois Cox and Ms. Barbara Stevens.
Idaho ESP Staff (Dr. Charles Brokopp, Director): Ms. Jill Wyatt
and Ms. Pamela Smith.
Dr. Robert C. Duncan, Director, Florida ESP, School of Medicine,
University of Miami.
HEMB Staff: Drs. Jack Griffith and Charles Miller, Mr. Robert
Heath and Ms. Mary Frankenberry.
This report was organized and edited by:
Jack Griffith, Ph.D.
Robert Heath, M.S.
Mary Frankenberry, B.A.
-------
Chapter II
PURPOSE AND SCOPE
The purpose of this study was to assess the rates of spontaneous abor-
tion occurring in a forested region of Oregon's Coastal Range, centered
about the Alsea basin, where 2,4,5-T has been cuiuiunly used in forest
management, and to compare those rates with rates occurring in a compa-
rable control area.
Specific objectives of the study were to test the following hypotheses:
a. whether or not differences in spontaneous abortion rates exist
between the study and control populations;
b. whether or not seasonal variations in rates exist within the
study and control copulations;
c. whether or not such variations, if they exist, can be as-
sociated with time and concentration of spray applications in
the study area.
The data are limited to in-patient records of women hospitalized for
spontaneous abortions of less than 20 weeks term, based upon:
a. the observation that 12 of the 13 miscarriages reported by the
original nine participants were for terms of less than 20
weeks duration;
-------
b. the assumption that a possible chemical effect would be most
likely during the first trijnester of pregnancy and could
therefore be masked in abortion data spanning the entire nine
months of term.
-------
Chapter III
METHODS
The methods used in this study consisted of defining the study area,
developing/confirming a 2,4,5-T use history in that area, researching
U.S. Postal ZIP code boundaries, selecting a control area, defining an
urban area, abstracting spontaneous abortion data from the hospitals
in the three areas, interviewing area physicians, retrieving county
birth data from Oregon computer tapes, and collecting various descript-
ive data ancillary to the study.
Description of the Study Area
The Study Area comprises approximately 1600 square miles of Oregon's
forested Coastal Range (Figures 1 and 2). It was selected so as to be
centered 'around the "Alsea basin", an area of approximately 400 square
miles. The Study area includes the western half of Benton County,
northwestern Lane County and all but the northern and northwestern
reaches of Lincoln County. It is bounded on the west by approximately
70 miles of the Pacific Coast entending from Lincoln City southward to
Florence, and extends inland for distances ranging from 10 to 35 miles.
(Exact boundaries conform to U.S. Postal ZIP boundaries.) The Study
area includes all but the northern and southern reaches of the Siuslaw
National Forest. Interspaced throughout one numerous commercially owned
and Bureau of Land Management forested acreages (Figure 3). fountain
elevations of approximately 1,000 feet are not uncommon; peak elevation
is slightly more than 4,000 feet.
-------
Figure 1. OREGON
.',6
> ,
SCAIC- STATUTE MILES
J 10 20 30 5O 60
-------
Figure 2.
Study Area in Lane, Lincoln and Benton Counties of Oregon
STUDY AREA
30 Ull«»
SCALE
10
-------
Figure- 3. Land Cover in Lincoln, Benton, and Lane (partial) Counties Oregon
FOREST
CROP &/or RANGE
-------
Drainage is primarily westward; principal rivers include the Siletz,
Alsea, Yaquina and the Siuslaw. Eastern fringes of the area drain
eastward into the Willamette Valley. Maximum runoff is reached gen-
erally during the winter months as the result of storms off the Pacific
occurring usually as rain. There is little snow accumulation.
The Study area is predominantly rural. The four hospitals in the area
are located in the four largest towns: Newport (pop. 5,200); Lincoln
City (4,200); Toledo (2,800); and Florence (2,250). Each of these hospi-
tals was utilized in the study. With the exception of Philomath
(1,700), all other towns/villages have populations of less than 1,000.
Alsea has a population of 400 (1970 census).
All of the nine women who participated in the first phase of the investi-
gation resided, at the time of pregnancy, in rural residences located
within 12 miles of Alsea. All but one of the women resided in the Alsea
River watershed; the ninth resided southwest of Philomath in the
Corvallis watershed.
Description of the Control Area
After careful review and consultation with staff of the Oregon State
Health Department, the Control area was selected in Malheur County,
Oregon. Selection was based on the following criteria:
1. The area had little or no use of 2,4,5-T.
2. The area is primarily rural, as is the Study area.
12
-------
3. Physician practices and hospital facilities were expected to
be similar to those in the Study area.
4. The area bore topographic similarities to the study area,
being of similar elevation and, although not mountainous, having
rugged terrain (escarpments, rolling hills, arroyos, canyons).
The Control area comprises four contiguous postal ZIP code zones in the
northeastern part of Malheur County (See Figure 4). The area covers
approximately 1,000 square miles and is bounded on the east and north-
east by the Snake River, which there forms the Oregon-Idaho boundary.
Several creeks drain the area eastward into the Snake River. Approxi-
mately 90 percent of the area is classified as rangeland, sagebrush
being the dominant vegetation. Cropland accounts for a small but
important percentage of the area along stream and river courses.
Twenty-one percent.of the land in the county is in private ownership,
75 percent is Federal and the remainder (5%) is State, county, or
local governments. (See Figure 5)
The two hospitals in the Control area are located in the two largest
towns: Ontario (pop. 8,200) and Nyssa (pop. 2,900). Both hospitals
were utilized in the survey. The area also includes the town of Vale
(1,850) and the villages of Harper and Cairo (pop. less than 250).
13
-------
Figure 4. Control Area in Malheur County Oregon
CONTROL AREA-
30 Mll«t
SCALE
14
-------
Figure 5. Land Cover in Malh-eur County Oregon
FOREST
CROP &/or RANGE
15
-------
Personnel from the Malheur County Cooperative Extension Agent's office
had no specific figures to report on pesticide use in the county;
however, it was stated that there is very, very little use of 2,4,5-T
in the county, particularly in the cultivated areas. Personnel from
the area office of the BLM, which has the responsibility for management
of public land within Malheur County, reported that 2,4,5-T has not
been applied on BLM lands in the county since 1972 by either BLM or
ranchers with grazing permits. Additionally, BLM personnel stated that
no pesticides of any kind had been applied to BLM grazing land since
1968. A sagebrush control program used 2,4-D but not 2,4,5-T in Malheur
County.
Description of the Urban Area
The Urban area is comprised of the two connecting Postal ZIP zones that
encompass the cities of Corvallis and Albany, Oregon. Both sites are
located in the agricultural non-forested Willamette Valley. The
Corvallis ZIP zone is contiguous with the east-central boundary of the
Study area and the Albany zone connects the Corvallis zone at its north-
east corner (see Figure6). The populations of Corvallis and Albany are
over 37,000 and 21,000 (1970 census).
16
-------
Figure 6.
Urban Area in Benton County, Oregon
URBAN AREA
3O Mil*.
SCALE
17
-------
TABLE 1
Selected Agricultural Statistics for Counties Including
the Study, Control and/or Urban Areas
Oregon, 1972 - 1977
All Farms, Number
Land in Farms, Acres
Approximate Land Area, Acres
Percent in Farms
Forest Production, Number
of Farms
Forest. Product ion, Dollars
Crops, Number of Farms
Crops, Dollars
Livestock, Number of Farms
Livestock, Dollars
Malheur
Control
1,357
1,360,195
6,309,760
21.6
5
1,278
935
23,040,163
991
20,063,573
County and
Lincoln
Study
258
47,390
631,104
7.5
38
208,726
59
495,703
214
1,028,121
Area
Lane
Study
1,840
270,587
2,913,280
9.3
158
553,042
900
11,780,769
1,152
9,319,416
Benton
Study &
Urban
575
129,034
427,520
30.2
58
457,813
301
5,317,112
423
2,667,955
U.S. Bureau of the Census, Census of Agriculture, 1969 \fol. 1 area
Report Part 41. Oregon Section 2. County Data. U.S. Government
Printing Office, Washington, D.C. 1972.
18
-------
Wbmen in both cities use Qsod Samaritan Hospital in Corvallis for
gynecological and/or obstetrical care, as do various women throughout
the Study area. All spontaneous abortion records for terms of less than 20
weeks were obtained from the Hospital, first, to derive additional
data for the Study area.- and, second, to permit comparison of seasonal
spontaneous abortion patterns and frequencies in an unsprayed urban
area adjacent to the Study area.
The abortion data for the Urban area are considered to be of limited
utility in this study because of an apparent tendency for first-
trimester abortions to be frequently handled in urban clinical
facilities of a type that do not exist in the Rural Study and Control
areas. The data are useful, howeverr in providing a measure of
monthly and seasonal trends on patterns in abortion frequencies.
Research of ZIP Cede Boundaries
Tb facilitate an identification of the boundaries of the Study area,
ZIP code maps were developed with the cooperation of personnel in
local post offices. The boundaries of the Study area, which coincide
with the zip code delivery routes, remained unchanged during the study
time frame of 1972-1977 (See Table 2).
IS
-------
TABLE 2
ZIP Codes Corresponding to the Study, Urban,
and Control Areas
Oregon, 1972 - 1977
Area Zip Codes
Urban 97321, 97330
Control 97906, 97913, 97914, 97918
Study 97324, 97326, 97341, 97343
97357, 97365, 97366, 97367
97369, 97370, 97376, 97380
97388, 97390, 97391, 97394
97439, 97453, 97480, 97498
20
-------
Spray Data
In the study of the Alsea area women, spray data on the use of 2,*. ,5-7
were collected and plotted for the immediate area referred to as the
"Alsea Basin". This information was supplied u-y the following major
organizations that used the chemical: USFS-Siuslaw National Forest;
USDI-Bureau of Land Management, Alsea Resources Area; Willamette Indus-
tries, Inc., Pilotnath, Oregon; and Starker Forest, Philomath, Oregon.
The supplied data consisted of the date(s) of application, rate of
application, formulation, .'lumber of acres treated, and the location of
the treated land.
The locations of the sprayed areas were plotted to quarter-section on
township maps. In this manner the perimeter of the "Alsea Easin" coulrj
then he defined as one covering approximately 400 square miles or
355,OOf! acres. During the 6-year period frcn 1972-1977, a total of 7121
acres was treated with 0215 pounds of 2,4,5-T. The poundage and acreage
varied from year to year (Table 3). The areas treated during this
period represent approximately 3% of the total acreage within the "Alsea
Sasin".
As mentioned previously, the predominant feature of the study area is
the forested Coastal Range. Since the Coastal Range extends fron north
err. California into Washington, it was considered necessary to establish
that: 1) topography and vegetation are similar throughout Oregon; and 2)
forest management practices in the Msea basin are representative of the
21
-------
Table 3
Total Acres Treated and Pounds 2,4,5-T Used in Alsea Basin
Oregon, 1972-1977
Year
1972
1973
1974
1975
1976
1977
Acres Treated
88
98
219
63
468
444
25
469
207
80
287
223
239
90
552
1619
1259
2875
1946
444
90
2480
Treatment
0.5 Ib/A
1.0 Ib/A
2.0 Ibs/A
3.0 Ibs/A
1 Ib/A
2 Ibs/A
1 Ib/A
.3 Ibs/A
1 Ib/A
2 Ibs/A
3 Ibs/A
1 Ib/A
2 Ibs/A
1 Ib/A
2 Ibs/A
3 Ibs/A
Total 2,4,5-T (Ibs.)
44
98
438
189
769
444
50
494
207
240
447
223
478
270
971
1619
2512
4131
1946
888
270
3104
Total
7,131
9,916
22
-------
entire forested area. Personal communications with representatives of
the U.S. Forest Service and commercial tree farm operators substan-
tiated that the ecological characteristics of the Coastal Pange were
consistent in the Oregon Coastal Range and that the chemical, 2,4,5-T,
is used as a cannon forest management tool in this region.
The herbicide 2,4,5-T is applied almost exclusively by helicopter at an
average rate of two pounds per acre for control of undesirable vege-
tation such as red alder, vine maple, salmonberry, and Thimbleberry.
Certain weather factors such as wind and precipitation dictate time of
application, but in general the compound is used in the spring (March,
April, or May) with a second application made/ if needed, in middle to
late summer (July and/or August). These seasonal usage patterns are
shown in Table 4 and Figures' 7 and 8.
To avoid contamination of water sources prior to 1978, the general appli-
cation policy was to avoid spraying near homes and provide for a single
swath of 30 to 60 feet on each side of any major stream. In September,
1978, the Oregon Forest Practices Act provided extended guidelines.
These guidelines required that no spraying was to be made within 500
feet of an inhabited residence nor within 200 feet on either side of a
Class A stream. Class A streams are defined as major streams with fish
and/or ones that are used for domestic water supplies.
Thus, the data from the Alsea basin illustrate a pattern which may be
considered as representative of the Study area. In general, greater
amounts of the chemical were applied during the spring than
23
-------
Table 4
Applications of 2,4,5-T by Day and Month in the Alsea Basin
Oregon, 1972-1977
Year
1972
1973
1974
1975
1976
1977
Dates
March 17, 20, 31
April 4
July 31
August 1, 19, 23, ?
May 5,6,10,13,14
August 15
April 26, 27, 29
May 4
July 29
August 2
April 9, 16
May 25, 27, 29
July 31
August 20
April 3-10
May 6
March 12-15, 19
March 24 - April 14
Total Days
3
1
1
4_
9
5
1
6
3
1
1
1
6
2
3
1
1
7
8
1
9
4
22
26
Acres treatment
148
121
48
151
468
444
25
469
180
27
48
_32_
287
239
202
16
95
552
2840
35
2875
534
1946
2480
Amount Applied (Ibs)
296
no
144
219
769
444
50
494
180
27
144
96
447
478
282
16
195
971
4096
35
4131
1158
1946
3104
-------
Figure 7.
Acres Sprayed with 2,4,5-T in AT sea
Basin Accumulated by Respective Month, 1972 through 1977
5-
4-
2-
M
25
-------
Figure 8.
Pounds* of 2,4,5-T Sprayed in AT sea
Basin Accumulated by Respective Month, 1972 through 1977
5-
4-
o
o
23-
x
2
1
M
MONTH
*Active ingredient
26
-------
during the summer treatments. Additionally, the spray program is not
a month-long operation. Usually it spans only a few days' time. The
duration of the spraying depends on the number of acres to be treated
and on the weather conditions.
Selection of Hospitals and Abstraction of Spontaneous Abortion Data
Spontaneous abortion data were abstracted from all five hospitals
located in the Study and Urban areas (See Figure 9) and from the two
hospitals located in the Control area (See Figure 10). These hospitals
were viewed as the primary source of health care delivery for the
respective areas. The seven Oregon hospitals contacted are located
in Corvallis, Lincoln City, Newport, Toledo, Florence, Nyssa, and
Ontario. Those hospitals and the number of admissions and live births
that each recorded in 1975 are found in Table 5. All field work was
completed during November and December, 1978.
Most of the hospitals were reluctant to participate in a record search
with non-hospital personnel because of patient confidentiality. Three
hospitals permitted a complete record search by field epidemiologists,
while at the other four hospitals the data were abstracted by hospital
personnel. Patients' names and addresses, excepting ZIP codes, were not
recorded.
For each of the spontaneous abortions the following information was
provided or abstracted: H-ICDA diagnosis code, age of patient, date of
27
-------
Figure 9.
Hospitals in Study and Urban Areas
Lincoln/
City /x
1NCOI
COUNTY!
Newport/x
Toledo
x/Corvallis
Florence,
E
BENTON
COUNTY /
LANE
COUNTY
23
-------
Figure -10. Hospitals in Control Area
MALHEUR
COUNTY
-------
TABLE 5
Annual Admissions and Births in Hospitals in
Study/ Urban, and Control Areas
Oregon 1972 - 1977
Hospital
location
Urban Area
Study Area
Control Area
Hospital
Name
Good Samaritan
Western Lane
North Lincoln
Pacific Communities
New Lincoln
Malheur Memorial
Holy Rosary
City
Gorvallis
Florence
Lincoln City
Newport
Toledo
Nyssa
Ontario
Admissions
1975
7589
802
1367
1159
875
710
4097
Births
1975
986
87
77
48
103
110
474
3U
-------
spontaneous abortion, gestation period, and ZIP code of patient's resi-
dence. Cnly records of spontaneous abortions that occurred from 1972
through 1977 were sought. This time period coincided with the temporal
miscarriage pattern of the nine original Alsea women.
From 1972 to 1977/ two ICDA texts were used by the hospitals. In 1973
H-ICDA was published and spontaneous abortions were listed by the codes
643.0, 643.1, 643.2, and 643.9. The 643.9 code includes any spontaneous
abortions not listed as induced or spontaneous. Prior to 1973, hospitals
used the 8th revision of ICDA. Spontaneous abortions were coded as
643.0, 643.1, 643.2, and 643.9. The code number 644- was also used for
1972, since this code approximated H-ICDA code number 643.9.
Physician Interviews
A list of private physicians who practice at each of the five hospitals
in the Study and the Urban areas was compiled from information pro-
vided by the hospitals and by local health officials. From the list,
30% of the physicians were randomly selected for interview. However,
since physicans tended to practice in groups, doctors not randomly
selected were also interviewed whenever available. In the Study area
19 of 27 (70%) of the physiciansall general practitionerswere
contacted. In Corvallis all eight of the obstetricians/gynecologists
were interviewed as well as 20% (five) of the general practitioners and
10% (two) of the internists.
31
-------
When possible, interviews were carried out in the doctor's office.
Each physician was asked to estimate the number of spontaneous
abortion cases of terms less than 20 weeks that he or she had treated
per year during the 1972 to 1977 time period. Each was also asked
to estimate the percentage of those cases which had been hospitalized.
Because of similarities in medical facilities and rural population
distribution in the Control and the Study areas, physician practices
were assumed to be similar in the two areas. Therefore, it was
decided not to conduct physician interviews in the Control area.
Data Preparation
The data on spontaneous abortions of less than 20 weeks from the seven
study hospitals were edited, coded, keypunched, and computer edited.
The number of monthly hospitalized spontaneous abortions during 1972-
1977 were tabulated for each of the three areas. These data were also
tabulated and cross-tabulated according to the several variables of
interest. For example, the data on hospitalized spontaneous abortions
were cross-tabulated according to area (Study, Urban, and Control) and
gestation period (less than 4 weeks, 5-8 weeks, 9-13 weeks, 14-17
weeks, and 17-20 weeks).
Birth certificate data for 1972 through 1977 were obtained in the form
32
-------
of computer tapes from the Vital Statistics Section of the Oregon
State Health Department. These computer tapes were used to obtain the
number of births per month in the Study, Urban, and Control areas for
1972-1977.
The Spontaneous Abortion Index: In order to make comparisons among
areas and among the months of the year, especially within areas, the
data on spontaneous abortions need to take into account the number of
births in the three areas during each of the twelve months. The index
described below is basically the ratio of the number of hospitalized
spontaneous abortions to the number of births corresponding to the
spontaneous abortions, based on the residence ZIP Code of the women
contributing to each event. Thus, the ratio is not a true rate but
rather an index of the hospitalized spontaneous abortion experience of
the women residing in the three areas.
In order to describe the spontaneous abortion index, the following
notation is required:
Y^ * the number of hospitalized spontaneous abortions in the
i^"- area (i = 1,2,3...) in the >="- month (j = 1,2,...,
12) during 1972-1977;
the number of births in the il area and in the j
month during 1972-1977.
33
-------
An index that has been used in similar studies is sinply the ratio
^ij^ij. The inherent problem with such an index is that the
.numerator involves women of one conception period whereas the
denominator involves women of a different non-overlapping conception
period.
The spontaneous abortion index developed here is Z. » ^i-i/^in
where C. is a five-month moving-average of the Xj. 's which have
been appropriately shifted. In particular, C^- £ W^X^ j+3+fc
where the weights (V^, ..., W ) represent the proportions of
abortions of varying gestation. In the above definition of C^, the
second subscript on the number of births is taken modulo 12; for
example, j+3-Hc = 13 refers to January births. The rationale for the
use of C^ rather than X^. is that a woman who has a spontaneous
abortion in month j could have delivered between month j+4 and month
j+8, depending on her length of gestation. In averaging the number of
births in the five months-between month j+4 and month j+8f one could
use a simple average (i.e., W^ = W- = ... = W^ = .20). Instead,
a weighted average of the number of births was applied with estimated
weights given by the proportions of abortions for the five gestation
periods.
Index values were calculated, by month and area, for the aggregated
six-year data and for the two three-year periods 1972-1974 and 1975-
1977.
34
-------
Statisticial Procedures
Statistical evaluation of the data is based on the following analyses:
1. Analysis of variance of the abortion index by area and
month, deriving a residual error term by dividing the data
into two 3-year periods.
2. Frequency table analyses, by chi-square, to test monthly
variation in the cumulated number of spontaneous abortions
analyzed as sinple frequencies and as expected monthly
frequencies of spontaneous abortions calculated from the
frequencies of corresponding live births.
3. Tests of cyclic trends including:
(a) A power-spectrum analysis of the abortion index over
months to test for cyclic trends in the monthly data.
(b) Adjustment of the abortion index data to account for
phase differences between the Study area and the Con-
trol and Urban areas.
(c) Pit of a sine-wave model to the abortion index data.
(d) Cyclic analysis of the raw abortion data (unadjusted
for births).
35
-------
(e) Examination of birth data from Miami, Florida (Jackson
Memorial Hospital, 1976-77) for cycli behavior.
(f) A test for cyclic trends after recalc Lating the
abortion index by raw numbers of birti s as the
denominator rather than the five-mont! moving average.
4. Cross-correlatioh analyses between Study a ea abortion
indices and spray patterns.
(a) Cross-correlation analyses, parametric and non-parmetric,
between the Study area abortion index ind the monthly
spray pattern of 2,4,5-T in pounds sp: iyed per month.
(b) Cross-correlation analysis with the me ithly spray pat-
tern after adjusting the abortion indt c data for dif-
ferences in cyclic trends.
(c) Cross-correlation analyses of Study ai ;a abortion index
and spray patterns for the periods 19: '.-74 and 1975-77.
5. Evaluation of physician interview data.
36
-------
Chapter IV
RESULTS
The annual numbers of hospitalized spontaneous abortion cases in the
three areas appear in Table 6. There was a total of 477 cases: 188
in the Study area, 109 in the Control area and 180 in the Urban.
Table 7 depicts these cases accumulated by month of the year for the
1972 to 1977 time period. The same cases are also represented accord-
ing to the age of the patient in Table 8. Table 9 presents, for each
area, the cases by weeks-of-term categories for those cases (456) for
which term was known.
As mentioned previously, the number of births in these three areas
needs to be taken into account in any comparisons made among the
areas with respect to the number of spontaneous abortions. Table 10
presents the number of births per calendar month for the Study, Urban
and Control areas; the five-month moving average of the number of
births corresponding to the month of spontaneous abortion is given in
Table 11 for each area, where the weights were obtained from the
tabulation of spontaneous abortions according to length of gestation
(Table 9).
Table 12 presents the monthly spontaneous abortion index for each of
the three areas in the study; the monthly spontaneous abortion index
for these areas is also displayed graphically in Figure 11. An
obvious feature of this graph is the elevated index of the Study area
for the month of June.
37
-------
TABLE 6
NUmber of Ffospitalized Spontaneous Abortion Cases for Study,
Urban, and Control Areas
Oregon, 1972 - 1977
Year
1972
1973
1974
1975
1976
1977
total
Study Area
37
34
23
31
33
30
188
Urban Area
44
40
38
19
20
19
180
Control Area
9
19
27
27
8
19
109
38
-------
TABLE 7
Number of Hospitalized Spontaneous Abortion Cases for Study/
Urban, and Control Areas, Accumulated by Month
Oregon, 1972 - 1977
Month
January
February
March
April
May
June
July
August
September
October
November
December
Total
Study Area
10
17
18
11
16
24
20
17
9
15
16
15
188
Urban Area
25
17
15
16
17
15
5
11
17
19
7
16
180
Control Area
12
4
7
14
9
6
7
10
11
7
8
14
109
39
-------
TABLE 8
Hospitalized Spontaneous Abortion Cases by Study, Urban,
Control Area, and Age C^oup of Patient
Oregon, 1972 - 1977
Age
Group
10-14
15-19
20-24
25-29
30-34
35-39
40-44
45-49
Total
Study
Area
3
38
69
44
18
9
6
1
188
Percent
1.6
20.2
36.7
23.4
9.6
4.8
3.2
0.5
100.0
Urban
Area
0
12
68
62
26
7
5
g_
180
Percent
0
6.7
37.8
34.4
14.4
3.9
2.8
0_
10.0
Control
Area
0
14
38
24
21
9
3
g_
109
Parcent
0
12.8
34.9
22.0
19.3
8.3
2.8
0_
100.1
-------
TABLE 9
Number and Percent of Spontaneous Abortions According to
Length of Gestation in the Study, Urban, and Control Areas
Oregon, 1972 - 1977
Study
Gestation Period Area
1. 18 - 20 weeks 8
(4.6%)
2. 14 -17 weeks 24
(13.8%)
3. 10 - 13 weeks 68
(39.1%)
4. 5 - 9 weeks 62
(35.6%)
5. 4 weeks or less 12
(6.9%)
Total 174
Urban
Area
12
(6.7%)
15
(8.3%)
83
(46.1%)
60
(33.3%)
10
(5.6%)
180
Control
Area
7
(6.8%)
14
(13.7%)
48
(47.1%)
28
(27.5%)
5
(4.9%)
102
Overall
27
(5.9%)
53
(11.6%)
199
(43.6%)
150
(32.9%)
27
(5.9%)
456
41
-------
TABLE 10
Total Number of Births by Months for the Study,
Urban/ and Control Areas
Oregon 1972 - 1977
Study Area
January
February
March
April
May
June
July
August
September
October
November
December
Six-Year Total
194
188
202
189
201
228
203
212
204
170
172
181
2344
Urban Area
338
353
344
335
357
378
315
351
342
343
337
327
4120
Control Area
124
124
126
131
155
145
156
131
154
136
155
129
1666
-42
-------
TABLE 11
NuraDer or t
Month
January
February
March
April
May
June
July
August
September
October
November
December
Ttotal
sirens uorresponaing a
in the Study, Urban
Oregon,
Study Area
208.8
206.8
191.8
177.8
177.9
184.0
189.7
193.0
195.6
196.7
208.7
213.4
2344
3 me romm or spon^ane
, and Control Areas
1972 - 1977
Urban Area
338.3
345.0
341.5
340.4
334.8
334.3
342.0
345.7
342.5
346.9
358.1
350.7
4120
sous Aoorcions
Control Area
146.3
142.6
145.6
143.6
142.3
130.5
126.5
125.4
129.0
138.7
147.3
148.1
1666
Based on five-month moving average.
43
-------
TABLE 12
Monthly Spontaneous Abortion Index for the Study, Urban, and
Control Areas
Oregon, 1972 - 1977
Month
January
February
March
April
May
June
July
August
September
October
November
December
Average
Study Area
48.1
82.2
93.8
61.9
89.9
130.4
105.4
88.1
46.0
76.2
76.7
70.3
80.8
Urban Area
73.9
49.3
43.9
47.0
50.8
44.9
14.6
31.8
49.6
54.8
19.6
45.6
43.8
Control Area
82.0
28.1
48.1
97.5
63.2
46.0
55.3
79.8
85.3
50.5
54.3
94.5
65.4
Average
68.0
53.2
61.9
68.8
68.0
73.8
58.4
66.6
60.3
60.5
50.2
70.1
63.3
Ihe spontaneous abortion index is defined as the ratio of the number of
hospitalized spontaneous abortions to the corresponding number of live
births based on a five-month moving average, and is expressed as
abortions/1,000 births.
-------
Figure 11. Plot of Monthly Spontaneous Abortion Index
for the Study, Urban, and Control areas
Oregon, 1972-1977
-STUDY AREA
........, URBAN AREA
ro-r'-.CONTROL AREA
V
\
m
\
\
gp
* *
/ *
/ \
** -*
/ 1
1
/
I
JFMAMJJASOND
45
-------
Statistical Analyses
In order to arrange the data in such a way that proper error terms
for hypothesis tests could be calculated and to see whether seasonal
spontaneous abortion patterns were consistent over time, the data
were gathered into two 3-year periods as. shown in Table 13. The
plots of these data are shown in Figures 12, 13, and 14.
From the graphical representations it is seen that the time
(seasonal) patterns within each area are remarkably similar for the
two periods.
The analysis of variance appropriate for the data in Table 13 is
given in Table 14. Although the three-year time periods could have
been viewed as a blocking factor, it was decided to test the various
interaction terms for possible significance. Clearly, the only
significant variation is among the three areas.
The mean values for the three areas are: Study area = 80.8; Urban
area - 43.8; and Control area = 65.4 (see Table 12). By the New
Duncan's Multiple Range Test, all three means are significantly
different from each other (p<.05, two tailed).
From Table 13 and Figure 14 it is seen that there is a decrease in the
overall Abortion Index for the Urban area during Period II. Analysis
showed that this decrease is significant (p<.0002). The mean
-------
TABLE 13
Portion Index" by Period, Area, and Month
Oregon, 1972-1977
PERIOD I
(1972 - 3 - 4)
January
February
March
April
May
June
July
August
September
October
November
December
Average
*Hospitalized
gestational «
Study
Area
48
77
94
45
112
98
95
76
61
71
105
84
80.5
Urban Control
Area Area
95
58
59
76
54
72
29
52
76
63
22
57
59.
68
28
14
84
84
46
63
96
140
29
68
95
4 67.9
PERIOD II
(1975 - 6 - 7)
Study Urban
Area Area
48
87
94
79
67
163
116
104
31
81
48
56
81.2
spontaneous abortions/1,000 live births adjusted
ige distribution of observed abortions. The comp
53
41
29
18
48
18
0
12
23
46
17
34
28.3
for
uted
Control
Area
96
28
82
111
42
46
47
64
31
72
41
95
62.9
monthly births for the six-year interval were divided equally between
the two periods.
47
-------
Figure 12 Abortion Index for the Study Area By Month and Period
CO
X
01
t!
o
180
160
140
120
100
80
60
40
20
0
f 1 -4-
a F M
~» 1 fe"
H i a
S 0 H D
-------
Mgure 13 Abortion Index for the Control Area By Month and Period
16(U
10
X
01
o
c
o
I
4J
O
3
140
120
100
80
60
40
20
M
-I f,
H
J J
Month
0 N
D
-------
Figure 14 Abortion Index for the Urban Area $y Month, and Period
en
o
X
0)
TJ
c
t
o
0 N D
Month
-------
TABLE 14
Analysis of Variance of Abortion Index by Period, Area, and Month
(See T^ble 13)
Source of
Variation
Periods
Areas
Months
Periods x Areas
Periods x Months
Areas x Months
Non-additivity
Residual
Degrees of
Freedom
1
2
11
2
11
22
1
21
Sum of
Squares
2,485.13
16,475.53
3,377.82
3,498.08
7,749.38
25,156.81
0.34
12,738.58
Mean
Square
2,485.13
8,237.76
307.07
1,749.04
704.49
1,143.49
0.34
606.60
F-Ratio
4.10
13.58*'
<1
2.88
1.16
1.89
<1
EX.0002
51
-------
abortion index of the Urban area for Period I is 59.4. When this is
compared to the six-year mean for the Study area (80.8) and that for the
Control area (65.4) using the New Duncan's Multiple Range Test it is
found that both the Urban area and the Control area differ from the
Study area (p<.05f two tailed), but not from each other. This seems
a more conservative analysis since the reason for the precipitous
decrease of the abortion index during Period II is presently unknown.
Figure 12 is noteworthy for two reasons. First, there is striking
similarity in the patterns, especially the "shoulders" on the curves
following the peaks for each period. Second, the peak for the second
period corresponding to a greatly increased pesticide usage is
higher than that for the first period.
Frequency Table Analyses
Uniform Hypothesis; Since the adjusted numbers of births used for the
denominators in calculating indices appear sensibly flat (see Table 11)
monthly variations in the cumulated number of spontaneous abortions
were analyzed as simple frequencies (see Table 7).
Under the hypothesis that monthly accumulated spontaneous abortions
were uniformly distributed throughout the year, chi-square analyses
were performed for each area. The urban area showed a significant
variation among months due to a high number of spontaneous abortions in
Eecembar and low numbers in July and November. The other two areas did
52
-------
not shew significant variation among months, but in the Study area the
month of June had a significant contribution to the overall chi-square
value. Further, in each area, various months showed large
contributions to the overall chi-square.
Expected Frequencies Based on Live Births: When the expected monthly
frequencies of spontaneous abortions were calculated from the
frequencies of corresponding live births from liable 11, the chi-square
analyses were virtually the same as when based on the actual number
of spontaneous abortions.
Cyclic Trends
Spectral Analysis; The Analyses of Variance and the chi-square
analyses, together with a close study of Figure 11, strongly suggested
the possibility of cyclic trends in the data.
Power-spectrum analysis of the abortion index over months showed that
for each area there was only one cycle of significance and it had a
period of about four months in each area.
Adjustment For Cyclic Trends; The inportance of indentifying cyclic
trends is two-fold. First, their presence might have sore biological
meaning relevant to this study. Second, a month-to-roonth comparison
between the Study area and the Control areas should take into account
phase differences which might be related to fertility patterns or some
other as-yet-unknown phenomenona.
53
-------
The need for this type of adjustment is shown in Figure 15. Since the
Study area and the combined Urban and Control areas seem to be almost
exactly 180° out of phase, any month-by-roonth differences could be
falsely amplified. Clearly/ if the Control data are shifted back in
time as shown in Figure 16, monthly differences could possibly reflect
excess abortions in the Study area if they exist.
The Sine Wave Model: Since the power-spectrum analysis identified
only one frequency for each set of data, it was decided to fit the
sirrplest cyclic model:
Index = A + B Sin (jj£ t - D) where t is in months. Table 15 shows
quite clearly that this model adequately represents the data in the
Urban and Control areas. Because of the peak in the Study area data
around June and July, the model does not fit the Study area as well.
The phase difference between the Study and the Urban and Control
areas is on the order of two months.
Tables 16, 17, and 18 show the Study area data adjusted for the Control
and the Urban area and for both areas combined. Figures 17 and 18 show
the sine wave models fitted to the abortion index data for the Study
area and the Control and Urban Areas.
54
-------
< FIGURE 15)1 SPONTANEOUS ABORTION INDEX. 1972 - 77 FOR THE STUDY AREA AND THE
| COMBINED CONTROL AND URBAN AREAS.
(]40|J-
|J20-
ui
f?0"-
I- STUDY AREA I
I-"URBAN AND CONTROL)
fj |F M A! M |J: J
(Month"
S' Of N D
-------
FIGURE 16: STUDY AND CONTROL AREAS: (a) COMBINED URBAN AND CONTROL INDEX,
(b) STUDY AREA INDEX, (c) URBAN AND CONTROL INDEX SHIFTED BACK
TWO MONTHS
TO fa-
(a) CONTROL AND URBAN AREA
A
50s-
40 i-
30 -
'V
~ "I 1 I 1
V
1 1 1
' V
i i i i r
P M
M
N
140
120
5)100
0!
80.
60
40
A/
(b) STUDY AREA
w
I I I I I
I I
M
M J J A
N D
80
70
X
ill
O
3
SO
40
30
i_
M
(c) URBAN AND CONTROL AREA
WITH 2-MONTH PHASE SHIFT
M
N
\
J -F
56
-------
TABLE 15
Parameters of the Model "Index = A + B Sin t - D)
For the Various Areas T
A (Mean)
B (Anplitude)
T (Period in Months)
D (Phase Shift in Months)
Adequacy of the Model
( Correlation)
Urban +
Study Urban Control Control
83.1 43.4 65.5 56.0
15.6 15.2 24.7 15.6
3.6 4.1 4.1 4.0
1.1 3.1 2.2 2.8
0.49 0.78* .92** .84**
^* 0.05
-------
TABLE 16
Study Area Index Corrected for Phase Shifts
Study minus Control 1 month lag
Study
48.1 (J)
82.2 (F)
93.8 (M)
61.9 (A)
89.9 (M)
130.4 (J)
105.4 (J)
88.1 (A)
46.0 (S)
76.2 (0)
76.7 (N)
70.3 (D)
Control
28.1 (F)
48.1 (M)
97.5 (A)
63.2 (M)
46.0 (J)
55.3 (J)
79.8 (A)
85.3 (S)
50.5 (0)
54.3 (N)
94.5 (D)
82.0 (J)
Difference
20.0
34.1
-3.7
-1.3
43.9
75.1
25.6
2.8
-4.5
21.9
-17.8
-11.7
58
-------
TABLE 17
Study Area Index Corrected for Phase Shift
Study minus Urban 2 month lag
Study
48.1 (J)
82.2 (F)
93.8 (M)
61.9 (A)
89.9 (M)
130.4 (J)
105.4 (J)
88.1 (A)
46.0 (S)
76.2 (0)
76.7 (N)
70.3 (D)
Urban
43.9 (M)
47.0 (A)
50.8 (M)
44.9 (J)
14.6 (J)
31.8 (A)
49.6 (S)
54.8 (0)
19.6 (N)
45.6 (D)
73.9 (J)
49.3 (F)
Difference
4.2
35.2
43.0
17.0
75.3
98.6
55.8
33.3 .
26.4
30.6
2.8
21.0
59
-------
TABLE 18
Study Area Index Corrected for Phase Shift
Study minus (Urban plus Control) 2 month lag
Study
48.1 (J)
82.2 (F)
93.8 (M)
61.9 (A)
89.9 (M)
130.4 (J)
105.4 (J)
88.1 (A)
46.0 (S)
76.2 (0)
76.7 (N)
70.3 (D)
Urban + Control
46.0 (M)
72.3 (A)
57.0 (M)
45.5 (J)
35.0 (J)
55.8 (A)
67.5 (S)
52.7 (0)
37.0 (N)
70.1 (D)
78.0 (J)
58.3 (F)
Difference
2.1
9.9
36.8
16.4
54.9
74.6
37.9
35.4
9.0
6.1
-1.3
12.0
60
-------
130
120
110
100
3
"2 i BO
80
I 70
60
50
FIGURE 17 PREDICTED AND OBSERVED POINTS FOR THE STUDY AREA
STUDY AREA
Sina Wave Modal
M
M
J J
Month
N
-------
80
FIGURE 18 PREDICTED AND OBSERVED POINTS FOR THE URBAN + CONTROL AREAS
Urban + Control ATMS
75
66
60
en
to
1
65
60
46
40
-------
Cyclic Analysis of the Numbers of Abortions; The raw numbers of
abortions showed the same time behavior as did the indices. In order
to find some rationale for cyclic variation in spontaneous abortions,
a physician board-certified in obstetrics and gynecology was
consulted. He was aware that births might show some.high frequency
cyclic behavior superimposed on a yearly cycle. The data for 1976-77
for Jackson Memorial Jtospital in Miami, Florida are shown in
Appendix B« A plot of these data suggests cycles of two-months
duration. Plots of the Oregon birth data in this report suggest
cycles of two-to-three-months duration. Clearly, these patterns need
investigation since they might be viewed as "noise" in the data and
thus obscure important relationships.
Simple Number of Births as the Denominator for the Abortion Index; In
order to see if any unanticipated algebraic relationships involved in
computing the denominator for the index could be causing a problem,
the raw numbers of births were used to compute abortion indices. The
results of the analyses did not change from the above. Table 19 shows
the distribution of births by month and the computed indices.
Cross-Correlation Between Study Area Abortion Indices and Spray
Pattern
Cross-Correlation Analysis; A cross-correlation was computed between
the abortion index for the Study area and the monthly pattern of
spraying 2,4,5-T in pounds sprayed by month. This
63
-------
TABLE 19
Number of Births per Month and Unweighted Abortion Index
for the Study, Urban, and Control Areas
Oregon 1972 - 1977
Month
January
February
March
April
May
June
July
August
September
October
November
December
Total/Average
Study Area
194
188
202
189
201
228
203
212
204
170
172
181
2344
51.5
90.5
89.1
58.2
79.6
105.3
98.5
80.2
44.1
88.2
93.0
82.9
80.1
Urban Area
338
353
344
335
357
378
315
351
342
343
337
327
4120
74.0
48.1
43.6
47.8
47.6
39.7
15.9
31.3
49.7
55.4
20.8
48.9
43.6
Control Area
124
124
126
131
155
145
156
131
154
136
155
129
1666
96.8
32.3
55.6
100.9
58.1
41.4
44.9
76.3
71.4
51.5
51.6
108.5
65.8
64
-------
FIGURE 19 (a) STUDY AREA ABORTION INDEX CORRtCTED FOR PHASE-SHIFTED
URBAN INDEX; (b) POUNDS OF 2,45-T SPRAYED BY MONTH
Ok
1001
CORRECTED
INDEX
0
4 J
2.4.5-T
(POUNDS) 3
XI,000
2
1 ~
0 ~J
7*v'
/ V
/ ^^a. ^^
(b)
J
FMAMJ
i
JASOND
-------
FIGURE 20 CROSS-CORRELATION OF STUDY AREA ABORTION INDEX WITH
SPRAY PATTERN IN TOTAL POUNDS OF 2,4,5-T APPLIED BY
MONTH. 1972-77.
f
01
cc
QC
O
O
OC
u
.8
.6
.5
.4-
.1 -
.'73 -
0.61"
0.10
0.10
-0.09
* P<0.035
[fill
43210
Months Shift in Spray Data
-------
analysis showed that the abortion index for the Study area was
significantly correlated (p<.01) with the spray pattern after a lag
of two (r*.70) to three (r=.76) months (see Figure 20).
To avoid problems with a non-normal distribution of the spray data,
a Spearman rank correlation corrected for ties was computed. The two-
month lag was not quite significant (r=.46), but the three-month lag
was (r=0.58, p<.05).
Cross-Correlation Between the Adjusted Study Area Index and the Spray
Pattern
Removal of the Cyclic Trend: The differences shown in Tables 16, 17,
and 18 now appear to be free of cyclic trends. This is shown in
Figure 19. Thus, the adjusted indices over time can be related to the
spray pattern over time in a straight-forward way. This is a more con-
servative approach than to use the unadjusted differences between the
Study area and the Control or Urban areas when correlating with the
spray pattern.
Cross-correlation of Spray Pattern with Adjusted Abortion Index; From
Figure 19 it appears that the peak in the corrected abortion index
for the Study area follows the peak in the spray pattern by two to
three months. This is borne out in Table 20 which shows significant
correlations (p<.01) after two months.
67
-------
TABLE 20
Cross - Correlation of Corrected Study Area Index with
Spray Data
Study vs. Urban
Study vs. Control
Study vs. Urban +
Control
lag (months)
2 3
-.07
-0.22
0.10
0.46
0.35
0.48
0.88*
0.84*
o.ai*
0.42
0.33
0.42
-0.04
-0.23
0.11
p<.01
68
-------
TABLE 21
Pounds of 2,4,5-T Applied in the Alsea Basin and Abortion Index
For the Study Area by Month and Period
Oregon, 1972-1977
January
February
March
April
May
June
July
August
September
October
Navember
December
PERIOD
(1972 - 3
Abortion
Index
48
77
94
45
112
98
95
76
61
71
105
84
I
- 4)
Pounds
2,4,5-T
0
0
296
290
471
0
288
365
0
0
0
0
PERIOD
(1975 -
Abortion
Index
48
87
94
79
67
163
116
104
31
81
48
56
II
6-7)
Pounds
2,4,5^T
0
0
1158
2990
317
0
16
195
0
0
0 .
0
69
-------
TABLE 22
Cross-Correlation of Study Area Abortion Index with Spray Data
Period I
Period II
0 1
0.19 -0.06
0.32 -0.24
Lag (months)
2 3
0.27 0.31
0.31 0.66**
4
0.00
0.00
Spearman Rank Correlation corrected for ties
0.01
-------
Abortion Index/Spray Pattern Cross-Correlation Analyses for Periods
I and II
Table 21 shews the pounds of 2,4,5-T used by month and the abortion
index by month for Periods I and II (1972-74 and 1975-77). Cross-
correlation analyses were completed between the abortion indices for
each period and the respective monthly patterns of spraying 2f4,5-T in
pounds sprayed by month in the "Alsea Basin." Since the spray data
are clearly not approximately normally distributed, a Spearman Rank
Correlation corrected for ties was done to find the cross-correlation
between spray patterns and the abortion index in the Study Area.
The results are shown in Table 22. These correlations show the same
pattern as those for the six-year aggregate of data (see Table 20),
although the correlation does not reach significance for the Period I
data.
Physician Interviews
Table 23 lists the physicians' estimates of the number of spontaneous
abortion cases of terms less than 20 weeks treated per year (1972-77)
and the percentage that were hospitalized.
Of the 19 doctors contacted at the four hospitals in the Study area,
four refused to provide estimates and four indicated they did not
handle such cases. Hie remaining 11 physicians provided positive
71
-------
responses on number of cases handled and/or percentage hospitalized.
Assuming that the four who refused to respond do handle such cases,
79% (15 of 19) of the physicians contacted have treated spontaneous
abortions during the study period.
From the physician interviews, it is estimated that approximately
70% of the spontaneous abortion cases are hospitalized among those
treated by physicians practicing at one of the Study area's four
hospitals. The percentage is calculated by weighting the individual
hospitalized percentage estimates by the respective number of cases
treated by each responding physician.
It is concluded, therefore, that the abortion data presented for the
Study area are, indeed, representative of that area. It is further
assumed that because of the aforementioned medical/population
similarities between the Study and the Control areas, a similar
representative percentage of spontaneous abortions are hospitalized in
the Control area.
The percentage of spontaneous abortion cases hospitalized in the Urban
area is estimated to be approximately 30%. (The estimate is adjusted
for general practitioners not selected for interview). Ihe percentage
is clearly less than that for the Study area, apparently due to those
cases treated at the clinic and medical center facilities not
available in the Study area. The data are considered to be reflective
72
-------
of seasonal spontaneous abortion patterns; however, Urban area index
values are undoubtedly biased downward relative to those for the Study
and Control areas.
73
-------
TABLE 23
Study and Urban Area Physician Interviews: Estimated Numbers
of Spontaneous Abortions Treated Annually and Percentages
Hospitalized, 1972 - 1977
Annual Mean Number
Hospital
Pacific Communities
Newporte, OR
(9 Physicians)
Western Lane Hospital
Florence, OR
(& Physicians)
New Lincoln Hospital
Toledo, OR
(4 Physicians)
Contacted
Physician
A
B
ci
D2
£3
F
G
H
I
J4
K
L
M
N
Type of
Patients
GP
GP
GP
GP
GP
GP
GP
GP
GP, Surgery
GP
GP
GP
GP
GP
of Spontaneous
Abortion Cases,
1972 - 1977
*
0
0
10.0
0
1.5
6
1.5
15
7.1
*
0
0.3
Percent
Hospitalized
75-80
95
20-30
Unknown
25
95
75
50
100
74
-------
TABLE 23 (Cbntinued)
Hospital
North Lincoln Hospital
Lincoln City, OR
(8 Physicians)
Good Samaritan Hospital
Corvallis, OR
Contacted
Physician
O
P
Q
R
S
T
U5
V6
w
X
Y
Z
AA
BB
CC
DD
Type of
Patients
GP
GP
GP
GP
GP
GP
GP
GP
GP
Internist
OB/GYN
OB/GYN
OB/GXN
OB/GXN
Internist
GP
Annual Mean Number
of Spontaneous
Abortion Cases,
1972 - 1977
5.0
5.0
*
*
*
5.0
1
3.9
*
0
50.0
50.0
50.0
50.0
0
11
Percent
Hospitalized
80
80
80
50
30
10
10
10
10
0
75
-------
TABLE 23 (Continued)
Hospital
Good Samaritan
Corvallis/ OR
(continued)
Contacted
Physician
Hospital EE
FF
GG
HH
Type of
Patients
OB/GYN
OB/GYN
OB/GXN
OB/GYN
Annual Mean Number
of Spontaneous
Abortion Cases/
1972 - 1977
50.0
0
50.0
50.0
Percent
Hospitalized
10
10
10
Refused to release data.
^Physician C has practiced in the community since January 1978.
Physician D has practiced in the community since 1975.
^Hiysician E has practiced in the community since 1977.
Physician J has practiced in the community since 1976.
^Physician U has not taken any obstetric cases in past three years.
Physician V has practiced for 14 months.
76
-------
Chapter V
SUMMARY AND CONCLUSIONS
The objectives of the study were to test the following hypotheses:
(a) whether or not differences in spontaneous abortion rates
exist between the study and control population;
(b) whether or not seasonal variations in rates exist within
the study and control populations;
(c) whether or not such variations, if they exist, can be
associated with time and concentration of spray applications
in the Study area.
The statistical analyses of the spontaneous abortion and spray data
presented herein to test these hypotheses have demonstrated that:
1. The 1972-77 abortion rate index for the Study area is
significantly higher than those for either the Control or
the Urban area.
2. There is a statistically significant seasonal cycle in the
abortion index in each of the areas with a period of about
4 months. In particular, there is an outstanding peak in
June in the Study area.
77
-------
3. There is a statistically significant cross-correlation
between the Study area spontaneous abortion index and
spray patterns in terras of pounds applied by months in the
Alsea basin, 1972-77, after a lag time of 2 or 3 months.
The results based on the six-year aggregate of abortion and spray data
are confirmed and enhanced by the analysis of the two 3-year
aggregates. The cyclic time patterns are consistent over time
periods; the Study area shows an elevated abortion index over the
Control and the Urban areas; and there is a peak in the abortion index
which correlates positively with the spray pattern in the Alsea basin
after a lag time of 2 to 3 months.
For all its complexity, however, this analysis is a correlational
analysis, and correlation does not necessarily mean causation.
78
-------
REFERENCES
1. Rebuttable Presumption Against Registration and Continued
Registration of Pesticide Products Containing 2,4,5-T.
Environmental Protection Agency, Office of Pesticide Programs,
Washington, D.C., Federal Register [6560-01] Vol. 43, No. 78 -
Friday, April 21, 1978.
2. Farm Chemicals Handbook - 1978. R. L. Meister, Jr. (ed.),
Meister Publ. Co., Willoughby, Ohio, pp. 250-251, 1978.
79
-------
80
APPENDIX A
QUESTIONNAIRE: Oregon Miscarriage Investigation
NAME DATE
ADDRESS
Prior addresses:
From 19 to 19.
From 19 to 19.
From 19 to 19
~CTTT COUNTY STATE
Family Physician:
NAME
ADDRESS PHONE
Date of Last Complete Physicial Examination
CITY
CITY
COUNTY
COUNTY
STATE
STATE
-------
PREGNANCY DATA
Please list jill_pregnanclea If more Chan 5, add pages. If the answer to any
question is "yes" please specify time In weeks of gestation and describe fully.
PKEGNANCIES
Month ami yuar of conception
Kuaidenct! during pregnancy:
(City, County, State)
Judlcutc live births
(give date)
(1) Full Term
(2) Premature
Still births/Spontaneous
abort IOUH ({'jive date)
liuluced nbortions
(t;lve date)
Wtitjkn of f'ctitation
ill rtli weight (lb. , oz.)
Any li trdi defects, mental
duf Ir Icncy, or other
coiiKuni tal Jiiipairutcnt?
(Aiuiwer j^o, Yes (deacribe)
or unknown)
1st
2nd
3rd
4th
5th
oo
-------
PREGNANCY DATA
CONTINUED
PREGNANCIES
l-'or llvehlrths, did you
experience spotting during
pregnancy?
('in in lucuri: luges and
:;L 1 1 llil rrli.-i, was there a
lung period oC illness
and spotting, or was
mlur.arrluge sudden?
Weight change during
prttgnuncy (11>) .
Did your doctor tell
you thut your weight
gain WUH excessive?
Did your doctor Indicate
that you hud high blood
pressure during
pregnancy?
Dlil your doctor Indicate
that you hud kidney
Infection during
preRnuney't (Weeks of
1st
h
2nd
3rd
4th
5th
CO
ro
-------
PREGNANCY DATA
CONTINUED
PREGNANCIES
Olil tic prescribe any
moil UutLions for weight
control, high blood
pruKHi«re, or kidney
Infections? If yes
please Bpeclfy
mud lent lona.
Did you have uuy
reap lira tory infections
during pregnancy?
(wc-uks of ^eatatiun)
Did you use any of cha
follow Inn medication
prior to or during
prei'iiuucy? If yes,
Indlcutu how often.
1. Oral contraceptives
2. Hormone a
3. Ant lunuseunta
ft. Ant Ih laLiimlntia
5. Aupirln
f>. AnLikiotLcB (Specify)
7. Sletsplnil pills
8. VI Lamina
9. Tranquiiiztirs
(spiici fy)
ID. Other ined lent lona
(apecify)
1st
2nd
3rd
4th
5th
00
-------
PREGNANCY DATA
CONTINUED
PREGNANCIES
Did you use any of the
following prior Co or
during pregnanee? (Specify
kind, amounts, frequency
and when)
1. Tobacco
2. Alcohol
3. HarL.iuunn
4. "Drugs" (coccalne,
LSI), etc.)
Did you suffer any physical
iiijuriea during pregnancy?
(If yuu, allow kind, weeks
of pregnancy, wau physicians
caru required.)
Were you X-rayed during
pr^i'mmcy? (If yes,
upuelfv tvpii ">H\ \\ou o£ten)
1st
2nd
3rd
4th
5th
CO
-------
PREGNANCY DATA
CONTINUED
PREGNANCIES
What household chemicals
(paints, garment cleaners,
solvents, pesticides) were
yon ex nosed to prior to or
during |ire^nuncy? (Specify
types and how often)
tJiiu your pregnancy confirmed
l>y u physician i f an early
mltu:arrlai;e?
Was youc miscarriage
iU>r.umi'nLed by a physician?
1st
2nd
3rd
4th
T^-~ *« .... -.1. .11.1.1 ,-..
5th
CD
Ul
-------
86
PREGNANCY OUTCOME OF SISTERS
For each of your sisters, please list pregnancy outcomes.
Total number of pregnancies
City of Present Full Spontan. Induced
Residence Age term Premature Abortions Abortion!
1.
2.
3.
4.
5. __ .
6.
-------
87
MEDICAL HISTORY
A. SUBJECT'S:
Have you had any of the following diseases or disorders in you
lifetime?
1. Measles:
Regular
German
2. Infectious heptatitis
3. Pneumonia, other
respi ratory i nfecti ons
4. Allergies (asthma, hay
fever, drug reactions)
5. Blood disorders, amemia
6. Heart disease
7. High blood pressure
8. Kidney disease; cystic
nephritis, etc.
9. Diabetes
10. Thyroid disorders
11. Stomach disorders, ulcer
12. Small bowel disorders,
ileitis
13. Large bowel disorders,
colitis
14. Nervous system disease,
central, peripheral
15. Epilepsy
16. Muscle system disorders,
paralysis
17. Skeletal (bone)
disorders
18. Skin disease
19. V.D.
20. Congenital defects
21. Mental illness
22. Other (Please List)
NO YES If yes, when/how long?
-------
88
MEDICAL HISTORY (CONTINUED)
B. Subject's hospitalizations, excluding pregnancies:
APPROXIMATE LENGTH OF
DATE PLACE REASON STAY
C. MENSTRUAL INFORMATION:
1. Are your menstrual periods regular? Mo, Yes.
2. Would you characterize your menstrual period as (check one):
a. Probably light
b.- Probably average _____
c.' Probably heavy _____
0. SUBJECT'S FAMILY
1. Did subject's father, or father's brother(s) or sister(s) suffer
from any congenital defect? No, Yes, Not sure.
If yes, please list:
Father
His brother(s)
His sister(s)
2. Did subject's mother, or mother's brother(s) or sister(s) suffer
from any congenital defect? No, Yes, Not sure.
If yes, please list:
Mother
Her brother(s)
Her sistar(s)
-------
89
MEDICAL HISTORY (CONTINUED)
3. Mother's pregnancies: Total number .
How many were full term ?
How many miscarriages ?
If known, during which week(s) of pregnancy ?
4. Did any of subject's brothers suffer from any congenital defect?
No, Yes, Not sure.
If yes, list type of defect:
5. Major illnesses or diseases of brothers:
6. Did any of subject's sisters suffer from any congenital defect?
No, Yes, No sure.
If yes, list type of defect:
7. Major illnesses or diseases of sisters;
-------
ENVIRONMENTAL ASPECTS
PREGNANCIES
How many miles is (was) your
residence or place of work
from the nearest area sprayed
with herbicides during each
pregnancy?
1st
2nd
3rd
4th
5th
Could you ever smell the
cliemlcul at the time It
was applied to the forest?
(IT yes, specify during
WJlltll pregnancies.)
Uluit Is (was) the source
of your water supply?
Ol1!!!lull']91 "el I y type)
Ueie you ever aware of a
change hi I lie tasilu of
y.iur waler during or
alter herhlcJde
IcutlonV
I Jure your garden &
flowers, trees, shrubs
ever damaged when
'herbicide was applied
in I he forest?
-------
ENVIRONMENTAL ASPECTS
CONTINUED
PREGNANCIES
Old you liuvu any pets during
your |itu^iiiiiicics? No,
Yea. If yes, please
Hut (d»B, cut, bird, etc.)
lot
2nd
3rd
4th
5th
your wntur supply ever been tested for pesticide reaidues?
II: yea, do you know when ? By whom
No,
Yea,
Not sure.
vo
Do you know the results? (list)
II not, do you know where they may be obtained
-------
92
FOOD SUPPLY AND DIET
A. What percentage of your meat and poultry products are raised
locally? *.
B. What percentage of your fruits and vegetables are raised
locally? %.
C. Does your milk come from local cows? No, _____ Yes,
Some(X).
D. Do you have a home garden? No, Yes.
E. Do you eat game taken from local forest areas that have been
treated with herbicides? No, Yes. Kind?
Approximate number of meals per year .
-------
93
HOUSEHOLD PESTICIDE USAGE
1. Has a commercial applicator treated your residence for pests within
the past five (5) years? Yes, No, _ Unknown.
2. Has your residence ever been treated for termites? ^ No, Yes,
_ Unknown.
3. Within the past five (5) years has your household used no-pest strips?
Yes, No, Unknown.
4. Within the past five (5) years has your household used any of the
following pesticides on pets?
Insecticide Collar Yes, No, NA
Insecticide Shampoo Yes, No, NA
Insecticide Powder Yes, No, NA
Other _
5. Within the past five (5) years have you used moth balls, crystals,
flakes, or aerosols? Yes, No, Unknown.
6. Within the past five (5) years have pesticides been used in the:
House Yes, No.
Garden Yes, No, NA.
Yard Yes, No, NA..
7. Are any pesticides stored on the premises? Yes, No,
____ Unknown.
3. Within the past five (5) years have you used any disinfectants?
Yes, No, Unknown.
-------
-
-
!
:
_
-
1
\
^W
1
I
^^
:
;
-
H
^^
:
m^
^^
-
-
i
-
»
-
^
P«*
i
n What are names of all pesti-
H cides used by the household
n the past five years and the
5 names of all non-used pestle
w currently stored on the pren
> Also, show EPA registration
PS number if known.
^N
0 tl
O <-n m
3 i-n O M
K- I- OH
«< o a M
*-' 18 MO
M
a a
en n
(D
COMMERCIAL
HOUSEHOLD HEAD - MALE
HOUSEHOLD HEAD - FEMALE
OTHER ADULT->18 vrs old
CHILD -<18 vrs old
LIQUID Is this pesticide
SOLID solid fora? (Aero
Who used this
specific pesti
cide?
in liquid or
sols are liauid
How many ounces of this pesticide have ;
or any member of your household used in
the oast 12 months?
How many ounces of this pesticide are
currently stored on the premises?
HOUSE Where on the pre-
GARAGE nises has this
LAWN pesticide been
YARD (other than lawn) used in the past
VEGETABLE GARDEN 5 years.
FLOWER BED
OTHER
How many times has this pesticide been
used in the aast 12 months?
BOOTS What type of
precautions are
GLOVES taken when using this
MASK pesticide?
WASH HANDS
AREA OFF-LIMITS TO CHILDREN & PETS !
OTHER
NONE
YES Is this pesticide stored
NO under lock and key?
NOT STORED
YES Is this pestii
NO its original
NOT STORED
KITCHEN-UNDER SINK]
KITCHEN-OTHER I Where
;ide stored in
jontainer?
on the premises
UTILITY ROOM is this oesticide
GARAGE stored?
SHED
BACK PORCH !
3ASE1ENT
OTHER
MOT STORED
-------
95
OCCUPATIONAL HISTORY OF SUBJECT
Please list any employment:
Employer Describe Beginning and ending
(location) Type of Business Job Title Dates
Do you launder your husband's work clothes with family wash?
Yes No
OCCUPATIONAL HISTORY OF HUSBAND
Employer Describe Beginning and ending
(location) Type of Business Job Title Dates
1.
2.
3.
4.
5.
6.
-------
THESE QUESTIONS REFER TO HUSBAND'S OCCUPATION:
Was This Job
Considered
Dangerous
Was Protective
Equipment or
Clothing Available
Did He Inhale
Chemical Solvents
Dust or Other Fumes?
Did Chemical Solvents
Oils, Dusts, etc. Get
on His Skin/Clothes?
(1)
Yes
No
If yes,
why?
(2)
Yes
No
If yes,
why?
(3)
Yes
No
If yes,
why?
(4)
Yes
No
If yes,
why?
(5)
Yes
No
If yes,
why?
Yes
No
If yes, what
was it?
Yes
No
If yes, what
was it?
Yes
No
If yes, what
was it?
Yes
No
If yes, what
was it?
Yes
No
If yes, what
was it?
Yes
No
If yesTHst if
known:
Yes
No
If yes, list if
known:
Yes
No
If yes, list if
known:
Yes
No
If yes, list if
known:
Yes
No
If yes,. list if
known:
Yes
No
If yes, list
if known:
Yes
No
If yes, list
if known:
Yes
No
If yesTlist
if known:
Yes
No
If yes, list
if known:
Yes
No
If yes, list
if known:
-------
97
THESE QUESTIONS REFER TO HUSBAND'S OCCUPATION CONTINUED:
Was This Job Was Protective Did He Inhale Did Chemical Solvents
Considered Equipment or Chemical Solvents Oils, Dusts, etc. Get
Dangerous Clothing Available Dust or Other Fumes? on His Skin/Clothes?
(6)
Yes
No
If yST
why?
Yes
No
If yes, what
was it?
Yes
No
If yesTlist if
known:
Yes
No
If yes, 1
if known:
1st
1. Did your husband serve in Vietnam?
service.
No,
Yes,
Months of
2. If yes, was he directly involved with spraying of Agent Orange Mo,
Yes.
3. Was he otherwise exposed to Agent Orange during his military activities?
None; Some; Often.
4. Has your husband ever had a sperm examination?
If yes, were abnormalities suspected? No,
No, Yes.
Yes.
5. Has your husband ever had V.D.?
No knowledge.
6. Has your husband ever used marijuana?
No know!edge.
No, Yes (specify),
No, Yes,
7. Has your husband ever used "drugs" such as coccaine, LDS, etc?
No, Yes, No knowledge.
-------
98
EDUCATIONAL BACKGROUND
Did you graduate from high school? No, _Yes, Class of 19_
If not, what was the highest grade completed? .
Did you graduate from college? _____ No, _____ Yes,
degree's and year's m.
If not, did you attend college? No, Yes, Years.
Did your husband graduate from high school? No, Yes,
Class of 19 .
If not, what was the highest grade he completed? .
Did your husband graduate from college? No, Yes.
List degree's and year's .
If not, did your husband attend college? No, Yes, Years,
-------
I, A i O P. ROOM STATISTICS
JaeJtsan Me=srial Hospital APPENDIX B
1977-1973
OCT..
Missions
76-77 534
77-78 1673
'.al Daliveriss
7S-77 *70
^"«
13
515
570
6
564
411
7
47?
472
3
469
95
6
39
98
3
95
4
12
42
59
1
1
DSC.
614
So5 '
CC4
in
*s>e.
S4S
14
SV,
A?
39
S3?
30
553
543
14
529
507
21
4c5
453
4
4-4
SI
9
72
90
S
35
o
71
12
JAJI.
*
570
539
««
7d
*T>
543^
a
534
AS
86
550
34
526
'S17
9
464
24
440
442
5
437
95
10
85
95
3
92
s
6
63
r>*
FZ3.
461
«o
TT
<;> «
«7
.I,-
33
409
1370
21 '
1149
72
12
60
6
43
HAJ9..
563
*7«
1.1
405
47
577
23
1439
423
24
1399
9
90
1
4
54
^ap-rr
491
if
'ft?
22
445
390
12
373
*
77
10
57
2
JULY
574
<7<.
71
<;/^7
71
.
srn
23
t;cs
4is
15
397
105
1101
1
4
62
AUG.
655
«c-»
1 7
57A
7-r
57 S
16
553
475
14
451
105
2
1C2
~
1
/ j.
SST.
S1 5
sa;
51
S77
7/1
«£>
23
^-s
432
14
459
109
9
135
5
5s
'CTAl FO?. Y'£.'
6730
(=547
in7
«ni7
f-)3
"S-JS7
107
5140
220
4520
1122
57
1025
O '
/ ±±
99
-------
APPENDIX B - Continued
.tiliborr.s
76-77
VAGTKA1.
Private
Staff
c/s
Private
Staff
77-78
VAGZJSI.
Private
Staff
C/S
Private
Staff
oortiar.s
76-77
77-73
ibor Poaa
asarear. Section
76-77
77-75
ssarear. Section^
76-77
?rir-iry
?.a?eat
77-7S
?rir.ary
?-ss-ai
CCT.
10
0
10
0
0
5
0
4
0
]_
21
28
61
85
S3'
55
33
102
65
37
HCV.
8
0
7
. 0
1
o
0
7
ft
}
23
45
73
95
97
76
. 21
93
73
2S
£Z.
6
0
a
0
0
2
0
2
0
0
25
31
62
35
32
54
2s
^i"-
s?
M
JXN,
'§
0
4
0
2
5
q
5
0
0
22
31
32
3.6
57
60
37
95
60
35
Fi3.
7
0
Q
0
1
32
50
75
56
19
Ml^J
Win***
7
0
6
1
0
44
30
99
73
26
A?RJL
2
0
2
0
0
35
S3
93
63
25
KA^
6
0
5
0
1
29
SO
L33
73
30
JVXZ
7
0
7
0
0
50
67
77
51
25
JULY
4
0
4
0
0
34
92
107
74
33
AUG.
14
1
13
0
0
43
92
104
75
25
sr?T.
»
.0
2
0
1
35
,.
53
10?
30
25
TCCAL JO?. 2SA
3n
_
' 1
72 ;
_i
1 i
5 ;'
.
__,,;
_.,
:
_^
^^
411 ;'
^J
I
?
905 _J
;
1135 _*
755 ;
340 _j
^j
_j
-"
100
------- |