‘. . ' a. t 4M ' - ‘
:-.a,z§e¢..:.a.,,-:.<.:'
I a 0' -‘ r' -

THE SEXUAL ORIGIN OF BLOODSTAINS:
SPECIFIC DETERMINATION OF TESTOSTERONE
USING RADIOIMMUNOASSAY

Thesis for the Degree of M. S.
MICHIGAN STATE UNIVERSITY
FRANK EDWARD SCHEHR
1976

"-"t-

.
..\
I.
~!

‘. “ISL-"W 1.1 1.1

94‘

4 ‘32.".1‘03141‘:

 

b
" Lillufifi

 

 

III” II II IIIIIIIIIZIIIIIIIIII

3 129

WWW-Q TE.

THE SEXUAL ORIGIN OF BLOODSTAINS:
SPECIFIC DETERMINATION OF TESTOSTERONE

USING RAD IOIMMUNOAS SAY
by

Frank E. Schehr

AN ABSTRACT

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

School of Criminal Justice

II

Chairman - Prof. Ra Ip h Turner - School of Criminal Just1ce

LIME—LC W

Member - Victor Strecher, Ph. D. - School of Criminal Justice
1" 7 ’3 . . A
' ‘1 .‘g .a ’41- C / - r/' a I
Member - Har‘ld D. Hafs, Ph.| /- Department 0 Phy31ology

@“WEA-Wzg .

Member - LaurenEe Simson,IK:D. - Pathologist

1976

 

ABSTRACT

I THE SEXUAL ORIGIN OF BLOODSTAINS:
SPECIFIC DETERMINATION OF TESTOSTERONE
USING RADIOIMMUNOASSAY

BY

Frank E. Schehr

Three experiments were performed using radioim-
munoassay (RIA) for the determination of the sexual
origin of bloodstains. It was also necessary to measure
a second determinant in a bloodstain to determine the
volume of blood represented by the stain so that
testosterone could be expressed per unit volume of blood.
The first experiment dealt with the RIA of testosterone
concentrations in whole blood. Values for the con-
centrations of testosterone found closely approximated
the volumes found in the literature for serum or plasma.
This indicated that differences between males and females
might be feasible when whole blood is shed and then found
in the dried condition.

The second experiment dealt with the application
of the RIA technique on dried bloodstains. Extraction

efficiencies calculated from the recoveries found in the

Frank E. Schehr

radioactive tracer stains were low and postulated to be an
error in preparation of the radioactive tracers onto the
stains. Overlap between males and females in this experi-
ment initiated a change in procedure.

The third experiment included not only the improved
extraction technique, but also a blind study trial. Ex-
traction efficiency was enhanced by preparing radioactive
tracers in the fluid whole blood stage before drying the
blood onto the substrate material. Results of the assays
showed a significant difference between male and female
origin of bloodstains. The blind trial study was con-
ducted on eighteen bloodstains cut from cotton fabric,
and this investigator's judgment was 90 percent correct
for discrimination between male and female origins.

The determination of hemoglobin by use of
cyanmethemoglobin method was studied to determine the
amount of blood present in an unknown stain. No corre-
lation was found between the measurement of hemoglobin
extracted from a stain and the amount of testosterone
extracted from the same stain. Difference in sexual
origin were still detected, however, when the stain ap-

proximated the size of a 0.1 m1 stain of whole blood.

THE SEXUAL ORIGIN OF BLOODSTAINS:
SPECIFIC DETERMINATION OF TESTOSTERONE

USING RADIOIMMUNOASSAY

BY

Frank Edward Schehr

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

School of Criminal Justice

1976

ACKNOWLEDGMENTS

The author would like to extend special thanks
first to Detective First Lieutenant Donald Bennett for
affording the time to complete the laboratory portion of
this work.

Appreciation is extended to Professor Ralph Turner
for the personal interest and guidance so freely given
while preparing for this degree.

Sincerest appreciation to Dr. Harold Hafs, James
W. Schwalm, and Ms. Jean Horshuh of the Dept. of Dairy
Science, Michigan State University,for their continued
guidance and technical assistance in the procedural de-
tails of this research, the use of the Dairy Physiology
facility, and the constructive review of this manuscript..

Special acknowledgment also to Dr. Laurence Simson,
Pathologist at E. W. Sparrow Hospital,for the generous
time given by his laboratory in the collection of blood
samples.

The author wishes to extend his deepest and most
sincere gratitude to his wife Jeanne for her encouragement
and assistance during the research and preparation of this

manuscript.

ii

TABLE OF CONTENTS

LIST OF TABLES O O O O C O O O O O 0

LIST OF FIGURES O C O O O O O C O O 0

Chapter

1. INTRODUCTION . . . . . . . . . .

2. REVIEW OF THE LITERATURE . . . . . .

SEX DETERMINATION ON DRIED BLOODSTAINS
PRINCIPLE OF RADIOIMMUNOASSAY . . .
HUMAN TESTOSTERONE CONCENTRATIONS . .

3. RESEARCH DESIGN AND PROCEDURES . . . .
MATERIALS I O O O O O O O
WHOLE BLOOD TESTOSTERONE ASSAYS . . .

BLOODSTAIN TESTOSTERONE ASSAYS . . .
HEMOGLOBIN DETERMINATION . . . . .

4. ANALYSIS OF THE DATA . . . . . . .

CALCULATIONS . . . . .
RELIABILITY CRITERIA OF THE ASSAYS .

EXPERIMENT l--WHOLE BLOOD ASSAY RESULTS
EXPERIMENT 2--BLOODSTAIN ASSAY RESULTS
EXPERIMENT 3--BLOODSTAIN ASSAY RESULTS

5. DISCUSSIOIq O O O O O O O O O O

6. SUMMARY . . . . . . . . . . .

BIBLIOGRAPHY . . . . . . . . . . . .

APPENDIX

iii

Page

iv

N00 ab H

22
24
26
28
34

34
34

44
46

53
62

64
72

Table

1.

LIST OF TABLES

Mean Serum or Plasma Testosterone Levels in
Normal Women and Men . . . . . . . . .

Comparison of Testosterone Levels for Direct
Extraction and Column Purification of Whole
Blood Extracts . . . . . . . . . . .

Percent Binding, Sensitivity, and Crossreactions
of MSU-74 Rabbit Antitestosterone Sera . . .

Mean Whole Blood Testosterone Levels in Males
and Females--Experiment l . . . . . . .

Mean Bloodstain Testosterone Levels in Males
and Females--Experiment 2 . . . . . . .

Mean Bloodstain Testosterone Levels in Males
and Females--Experiment 3 . . . . . . .

Results of Blind Trial Study . . . . . . .

iv

Page

13

39

40

42

44

46
50

Figure

1.

LIST OF FIGURES

Diagram Showing the Various Components of
RIA I O O O O O O I O O O O O

Pathways of Testosterone Biosynthesis in
the Leydig Cell . . . . . . . . .

Flowsheet for Radioimmunoassay of
Testosterone in Bloodstains . . . . .

Semilog Plot of Standard Curve for Deter-
mination of Hemoglobin . . . . . . .

Log Transformation of Data for RIA Standard
Curve . . . . . . . . . . . .

Log-Log Transformation of Data and Regression
Analysis for Fitting the Standard Curve .

Percent Occurrence in Males and Females of
Concentrations of Testosterone. Whole
Blood Assays: Experiment 1 . . . . .

Percent Occurrence in Males and Females of
Concentrations of Testosterone. Bloodstain
Assays: Experiment 2, Experiment 3 . .

Distribution of Testosterone Levels in Males
and Females of Different Age Groups.
Results of Bloodstain Assay--Experiment 3

Page

ll

17

29

32

35

36

43

45

58'

Chapter 1
INTRODUCTION

In recent years, the determination of the sexual
origin of bloodstains has occupied much of the research
in forensic serology. This research has included studies
of the differences in chromosomes between males and
females, the counting of female Barr bodies, and the
identification of menstrual blood. The research pre-
sented in this paper centers upon the fact that males
and females differ substantially in the amounts of
androgens (sex hormones) in the bloodstream, and pre-
sents the method of radioimmunoassay (RIA) for detecting
these differences.

The choice of an analytical procedure for the
detailed laboratory evaluation and interpretation for
comparative purposes of a particular item of evidence
depends on many factors, including the nature of the
particular crime and likely identifying characteristics.
Y—chromosome detection, Barr body counting, and menstrual
blood identification, suffer from the absence of clearly
defined and distinguishable standards, either from test
samples, reference collections, or statistical background

data. Radioimmunoassay, however, lends itself readily to

quantitative study, especially for steroid hormones. Much
background data is available on the differences in the
concentration of androgens in blood from males and females,
and in recent years most of this data comes from the tech-
nique of radioimmunoassay.

A major hormone difference between the sexes is
testosterone. Therefore this research centers upon the
quantitation of testosterone through radioimmunoassay
and the applicability of this testing to dried blood-
stains. This research describes an attempt to differen-
tiate a bloodstain with regard to the sex of the person
depositing it. The first phase of the work involved
detecting the level of the hormone in whole blood ex-
tracts, principally because much of the background work
involved serum or plasma. This was undertaken for the
reason that one does not lose only serum or plasma from
a wound or cut, but whole blood. The second phase of
the research involved applying this method to dried
bloodstains, both with known amounts of blood on a stain
and a blind trial on unknown amounts. Since this is a
quantitative study, the amount of blood present in a
stain is important.. Therefore the third phase of the
research deals with a method of estimating the amount
of blood in a stain, using hemoglobin determination as

the internal standard.

It is expected that this paper will produce other
avenues of research for the application of radioimmunoassay
to forensic serology, and suggest solutions for overcoming
difficulties arising with the application of clinical

methods to forensic science.

Chapter 2
REVIEW OF THE LITERATURE

For purposes of review, this chapter is separated
into three divisions: (1) sex determination on dried
bloodstains, (2) principle of radioimmunoassay, and

(3) testosterone concentration in humans.
SEX DETERMINATION ON DRIED BLOODSTAINS

Methods of nuclear sex determination by means of
staining techniques are based on the structural differ-
ences of the sex chromosomes. In normal females these
chromosomes occur as an homologous pair called the x-
chromosomes. These sex chromosomes take a less active
part in cell division and their constituent chromatin
(heterochromatin) shows a greater degree of simplification
with respect to the DNA base pairs than does the euchromatin
of the remaining chromosomes. The Y-chromosome is composed
of a large number of consecutive adenine-thymine pairs
(1” this forms the basis of the differential staining
behavior of the Y-heterochromatin of the male cells. The
sluggish behavior of the X-chromosomes during cell division
led to its discovery and subsequently to the discovery by

Barr (2) of the sex chromatin or Barr body (a derivative

4

of one or more X-chromosome displaying positive
heteropyknosis). This is the basis for the several
attempts at sex determination on dried bloodstains
utilizing Y-chromosome flourescense and Barr body
counting (3,4,5).

Two staining techniques are employed in the
detection of the characteristics of the heterochromatins.
The first is based on the affinity of quinacrine
dihydrochloride for adenine-thymine pairs. This is a
flourescent stain and its combination with the higher
concentration of these pairs results in a higher degree
of flourenscence of the Y-chromosome (3). This appears
as a bright spot in the nucleus of the cell; the second
is based on the presense of a denser area of fused X-
chromatin taking up an aceto-orcein dye and appearing
as a dark red spot on the inside of the nuclear membrane
in cells from females.

Philips and Webster (4) simplified the method
devised for Y—chromosome flourescence for all types of
cells by introducing magnesium ions to the staining
procedure. These ions have been shown to cause the DNA
helix to supercoil, bringing the sites into closer
proximity and thus increasing the level of flourescence.

Philips and Gitsham (7) applied this new method
in a series of blind trials and evaluated the results.

The percentage of negative nuclei (not exhibiting

Y-chromosome flourescence) was plotted against the per-
centage of positive nuclei (exhibiting flourescence)
summed with half the positive-negative score. In each
trial, the combined positive and positive-negative scores
of male stains cover a wide range, often overlapping
with the female region. The differentiation between
male and female stains was no better if negative nuclei
were plotted against positive nuclei alone. The trials
were investigated by both researchers, with Philips'
score significantly higher than Gitsham. Clearly, each
person's judgment needs to be "calibrated," and different
minimal scores for the varying degrees of certainty of
maleness would be needed. It was also unfortunate that
the complementary test for femaleness, which would have
been provided by the identification of the Barr body in
polymorphonuclear lymphocytes, proved unreliable in their
hands (8). Other studies performed on these drumsticks
or Barr bodies showed exceptions to the fact that one
chromatin body can be found in normal females and none

in normal males(6). Low counts of the chromatins, af-
fecting the reliability of the differentiation, are
common when working with stains. Many factors contri-
bute to this effect, but the age of the specimen and its
general condition may be important. Due to the good

stability of steroid hormones, it is hoped that this

factor will be negated in the tests with radioimmunoassay
for testosterone.

Methods presently used for the identification of
menstrual blood are the assay of soluble fibrinogen in
bloodstain extracts and the characterization of lactic
dehydrogenase isoenzymes in stains (9,10,11). The demon-
stration that a bloodstain is of menstrual origin may be
of great importance in forensic investigations of sexual
assault. One of the characteristics of menstrual blood

1 the

is that it is fluid (12) and cannot be made to clot;
fluidity of menstrual blood has formed the basis for a
number of earlier suggestions for the identification of
menstrual bloodstains. Popielski states that the absence
of fibrin threads in a bloodstain is consistent with it
being due to menstrual blood (13).' This method, however,
is satisfactory only when heavy bloodstains, not soaked
into the substratum, are studied. Other investigators
have tried to utilize the high fibrinolytic activity of
menstrual blood to breakdown "fresh fibrin obtained by
defibrination" and to study the products by paper

electrophoresis (14) but with limited success. The work

by Whitehead and Divall (9,10) utilized a hemagglutination

 

1Clot-like material in menstrual discharge con-
sists of red cells aggregations to mucoproteins and
glycogen, and is not composed of fibrin.

inhibition technique to measure soluble fibrinogen.
Their results were consistent with finding the presence
of high proportions of fibrin degradation products
coinciding with present theories on menstrual bleeding.
However, they also found that the fluidity and noncoagu-
lability of menstrual blood is a property shared with
certain postmortem bloods (15). On this basis alone it
is not possible to distinguish between menstrual blood-
stains and postmortem bloodstains. The circumstances
of a case can, however, exclude the possible contamination
of bloodstained clothing, etc. with postmortem blood.

This same problem also arises with the method of
lactate dehydrogenase (LDH) determination. Using electro-
phoretic methods, Asano.0ya, and Hayagawa (ll) relied on

3 of this

the concentration of the LDH-4 and LDH-S bands
enzyme, for the identification of menstrual blood.
Comparison was made between these isoenzymes in menstrual
blood, normal venous blood, cadaveric blood, and blood
shed during partuition. The marked increase of the bands

LDH-4 and LDH—5 (measuring by densitometry) was used for

differentiation of menstrual blood from other "types of

 

2"Soluble fibrinogen" is a collective term for
fibrinogen and fibrin degradation products.

3Lactate Dehydrogenase (LDH) is separated into
five bands on electrophoresis: LDH-l, LDH-2, LDH-3,
LDH-4, and LDH-S. The enzyme patterns are characteristic
for each organ and tissue of the body.

blood." Increases in cadaveric blood and blood during
partuition were also noted. Time studies on LDH isoenzymes
in menstrual bloodstains showed a slight decrease in
comparison with the fresh stains used in earlier studies.
The data indicate that this method can differentiate

menstrual blood stains if examined within one week.

PRINCIPLE OF RADIOIMMUNOASSAY

The principle of radioimmunoassay has been applied
to many organic compounds of biologic interest. One of
the requirements for the application of radioimmunoassay
to the measurement of a compound is that this compound
must be antigenic and must elicit high affinity antibodies.
The smallest compound which has been shown to be immunogenic
is the hormone vasopressin (molecular weight 1080) (16).
Usually compounds with molecular weight below 1000 are
not immunogenic per se. Since steroids have molecular
weights much below 1000, they are not immunogenic. How-
ever, when coupled to protein carriers, they become
immunogenic. Reviews covering this subject can be found
(l7,l8,l9,20).

All RIA procedures are based on the original
observation by Berson and Yalow (21) that low concentra-
tions of antibodies to the antigenic hormone, insulin,
could be detected by their ability to bind radiolabeled

insulin. Unknown concentrations of antigen may be

10

determined by taking advantage of the observation that
the radiolabeled hormone molecules, or tracers (Ag*I)
compete physiochemically with the nonlabeled hormone
molecules (AgII) for binding sites on the antibodies
(AbIII) (see Fig. 1). Moreover, the assay requires that
behavior between the standard and unknown antigen be
identical in their ability to displace labeled antigens
from a labeled antigen-antibody immune complex (Ag*AbIV),
but not identical behavior between the tracer and the
unknown or standard antigen. When increasing amounts
of unlabeled antigen (AgII) are added to the assay, the
limited binding sites of the antibody ABIII are pro-
gressively saturated and the antibody can bind less of
the radiolabeled antigen (Ag*I). The antibody solution
or antiserum is diluted to allow about 50% of the tracer
dose of Ag*I to be bound in the absence of unlabeled
standard or unknown antigen. If there is no nonspecific
inhibition of the immunochemical reaction, a diminished
binding of labeled antigen offers evidence for the
presence of unlabeled antigen. After an incubation of
the three essential Components (Ag*I, A911, and AbIII),
the antigen-antibody complexes, or bound antigens (Ag*AbIV
and AgAbVI) are separated from the free antigens (AgV and
Ag*VII) and the radioactivity of either or both are

measured.

11

 

 

 

Bound Free
Ag*Ab Ag
IV V
Ag* + Ag + Ab
I II III
Ag Ab Ag*
VI VII
I Ag* Labeled Antigen
II Ag Unlabeled Antigen
III Ab Antibody to Ag (and Ag*)
IV Ag*Ab Labeled Antigen Bound to Antibody
V Ag Free Unlabeled Antigen
VI AgAb Unlabeled Antigen Bound to Antibody
VII Ag* Free Labeled Antigen

Source: Yalow, R. S., and Berson, S. A. Assay of plasma
insulin in human subjects by immunological
methods. Nature 184, 164 (1959).

Figure l.--Diagram showing the various components of RIA.

12

The criteria for acceptance of an assay includes
reliability and practicability. The reliability of an
assay depends on its specificity, sensitivity, accuracy,
and precision (22). All of these criteria were judged
at different points of this research, and will be pre-

. sented in the appropriate discussion chapter.
HUMAN TESTOSTERONE CONCENTRATIONS

The purpose of this section is to review briefly
a fairly substantial body of knowledge about the con-
centration of testosterone in human plasma or serum, in
normal and abnormal conditions. For the most part, this
review will be directed to testosterone in the male;
however, testosterone concentration in the female will
be discussed, particularly when it may be appropriate to
compare and contrast them between the two sexes.

In the normal, sexually mature, adult male, there
is good agreement that the concentration of testosterone
in peripheral plasma is between 3.5 and 10.5 ng/ml.
These data, which are not generally disputed, have been
derived from many different laboratories using for the
most part RIA (see Table 1). In round numbers, the values
for males are about 10 times those found in the female.
Plasma concentrations of testosterone were measured in
oligospermic and azoospermic males (45), range 20-51

years of age. The only significant difference was that

13

Table 1

Mean Serum or Plasma Testosterone Levels

in Normal Women and Men.

 

 

Method ng/ml Female Male Ref.
a range .22-.45 3.65-8.15

RIA mean .321.07 5.90:1.49 23
n(age) 12(25-40) 13(25-40)
range .25-.80 4.50-ll.00

RIA mean .54i.22 7.90il.43 24
N9age) 22(25-40) 30(25-40)
range .l8-.51 3.18-8.10

RIA mean .311.10 5.14:1.19 27
n(age) 10(21-65) 20(21-65)
range .30-.85 3.80-9.65

RIA mean .571.l7 6.8611.96 26
n(age) 20(21-65) 24(21-65)
range .09-.25 2.50-8.80

RIA mean .221.07 4.9011.60 25
n(age) 18(19-55) 19(21-60)
range .09-.51 l.67-lO.89

RIA mean .261.15 4.6211.99 28
n(age) 11(20-32) 22(19-35)
range .10-.81 3.10-8.21

RIA mean .351.12 5.3111.70 29
n(age) 53(20-30) 14(20-28)
range .10-.58 .15-.90

RIA mean .281.15 .461.21 29
n(age) 14(58-79) 11(61-80)

b range .20-l.20 4.40-13.00

DID mean .701.30 7.4012.60 30
n(age) 10(21-55) 15(19-40)
range .04-.70 2.80-14.40

DID mean .351.22 6.7012.30 31
n(age) 20(21-65) 21(21-35)

 

aRIA Radioimmunoassay.

bDID Double isotope derivative technique.

14

testosterone was lower in the oligospermic male, but that
this concentration never fell to that of the female.
Blood concentrations of testosterone reflect the
momentary balance between entry into and removal from the
blood. In the male, about 1000 liters of plasma are
cleared of testosterone each day (32). Although the
liver is undoubtedly the principal organ responsible
for the metabolic removal of testosterone from plasma,
extra hepatic metabolism of testosterone certainly occurs
(33,34), but the precise contribution from skin, muscle,
and other tissue has not been precisely determined.
With plasma concentration of testosterone between 3.5
and 10.5 ng/ml., and clearance rates of between 900 and
1200 liters/day, the rate of entry of testosterone into
the blood may be calculated, assuming a steady state, to
be between 3 and 13 milligrams/day. But, is the state
steady, and is secretion synonymous with rate of entry?
It seems likely that the state is not entirely
steady. There has been some controversy as to whether
the level of this steroid changes within the day or
between days (34,36,37). The weight of evidence would
now suggest that within the day there is a variation, With
higher'morning and lower. evening values. Plasma testosterone
was measured by a competitive protein binding procedure
at 10-20 minute intervalinn eight normal adult men (37).

Testosterone appeared to be released episodically with

15

increasing blood concentrations throughout sleep and ap-
peared to relate to other hormone increases. Testosterone
ranged from approximately 7414 ng/ml of plasma throughout
the time period, with no consistent rhythms across eight
subjects for any particular of time. However, investi-
gators found little evidence that there is such systematic
variation in plasma concentration of testosterone over a
period of five weeks; in fact, values were found to range
widely in an apparently random fashion (38). Even with
this variation though, concentration of testosterone in
males always exceeds that in females (39).
Infancy, Childhood, Adolescence,
and Old Age

At the extremes of life there are distinct differ-
ences from the normal adult male pattern of testosterone
secretion. In the immediate neonatal period, plasma
testosterone is elevated in both.male and female infants,
and there is no significant sex difference (39,40).
Immediately after birth plasma testosterone falls to
values that are at, or below, the normal female range,
and remain at this level until puberty. Umbilical cord
plasma concentrations of testosterone have been measured
by radioimmunoassay in both male and female fetuses at
12 to 22 weeks of gestation and at term (39,40). Testos-
terone concentration from 12 to 18 weeks average 249

ng/lOOml and 29 ng/lOOml for male and female fetuses,

16

respectively. But at term there were no significant
differences in testosterone concentrations between sexes,
averaging 84 ng/lOO ml and 77 ng/lOO ml for male and
female fetuses, respectively. Increased testosterone
concentration in male fetuses paralleled leydig cell
hyperplasia and the presence of 3 B-hydroxysteroid
dehydrogenase in human fetal testis (39).

Frazier and Horton (41) have shown that during
this period the ratio of plasma levels of androstenedione
to testosterone ratio is the order of two. It should be
noted that as much as 5% of peripheral testosterone is
derived as a conversion product mainly from androstenedione
(42). This other steroid hormone, androstenedione,is
secreted from the leydig cells and is represented in
Figure 2. This progression and activity in the body is
important for this analysis, and each steroid is cross-
checked with the testosterone antibody for possible
interference or cross-reactivity. During this time before
puberty, about 40% of testosterone is derived from
androstenedione by conversion.

As puberty approaches in the male there is a
gradual increase in the concentration of testosterone
in plasma. These measurable chemical changes precede by
many months the pubertal anatomical changes. Plasma
testosterone concentrations in male children through

puberty have been determined with several methods yielding

17

Acetate
I

I
Cholesterol
CH3

 

Progesterone l7-hydroxyprogesterone Androstenediol

 

O

Androstenedione Testosterone

Figure 2.--Pathways of testosterone biosynthesis in the
leydig cell.

18

similar results. Testosterone measured by a double
isotope derivative technique in boys from 7 to 14 years
showed an increase from 5.4 to 19 ng/100 ml up to 250

to 482 ng/lOO ml (43). Adult male concentrations were
reported to be 625 i 104 ng/lOO ml (25,26). Serum
testosterone measured by competitive protein binding
showed an increase from 21 ng/lOO ml for 8 years olds to
52 ng/lOO ml for 16-year-old girls. Plasma testosterone
in girls measured by RIA ranged from 5.9 to 25.5 ng/lOO
ml for ages 6 and 13 respectively (43). The finding of
plasma testosterone in young males above the prepupertal
level but below the adult male range usually indicates
that puberty is likely to proceed normally (43).

At the other extreme of life, an interesting
study by Kent and Acone (44) showed that the clearance
of testosterone from blood fell progressively with
advanced years; mean plasma testosterone was lower in

older men, than the normal adult values for males.

Abnormalities of Testosterone Secretion

In the male, an apparent deficiency of androgen
secretion is more commonly encountered clinically than
androgen excess. Low plasma testosterone can result
from disorders that primarily and directly affect the
testis and Leydig cell function.

Following castration, the concentration of

testosterone in plasma falls to low levels within two

19

or three days, and although testosterone can be measured
in the plasma of these individuals, there is reason to
believe that it is derived by conversion from
androstenedione (45) (see Figure 2).

Numerous studies have been made on the con-
centration of plasma testosterone in Klinefelter's
syndrome, the most common cause of male hypogonadism
(46,47,48). These have shown that a fairly wide range
of values may exist among people with this disease, from
those found in the normal female to those observed in
the adult male. Attempts by a number of workers to
correlate testosterone concentration with chromatin
patterns have not yielded meaningful results. Hudson
et al. (49) attempted to correlate plasma concentration
with the number of interstitial cells, but did not find
this to be helpful. Paulsen et al. also pointed out
the difficulty of correlating plasma levels of testos-
terone with the degree of secondary sexual development
(50). It is of interest in the few studies that have
been made, plasma androstenedione appears to be normal
in Klinefelter's syndrome.

It would seem relevant to refer to a number of
studies that have been undertaken on males with an XYY
chromosomal complement. Although these subjects may
show distinctive anatomical and psychological features,

abnormalities of testicular function do not appear to

20

' be present. Hudson et a1. investigated several such
subjects and found blood testosterone concentration to
be normal (51,52).

The testes may fail to mature because of primary
gonadotropic hormone failure or following surgical
removal of the gonadotropic stimulus before puberty.

In such individuals plasma testosterone is usually lower
than normal, the ultimate level depending upon the extent
and duration of the failure of the gonadotropic secretion
(53).

Plasma testosterone was abnormal in a variety of
disorders that are not normally regarded as testicular
in origin. Plasma levels of testosterone have been found
to be low in perhaps the best recognized of these dis-
orders, liver cirrhosis,in which alcohol seems to play
a part (54). Other studies showed this same result, but
as an increase in testosterone metabolism in the liver,
and as the data indicated there is no compensatory in-
crease in the biosynthesis of testosterone under these
conditions (55). Kolodny has extended research into the
effects of Marihuana on testosterone. Continued use can
interfere with the production of this hormone, suppressing
the supply of testosterone. According to his results,
concentration of testosterone had decreased by an average

of one-third (56).

21

In females, serum androgens vary randomly during
individual menstrual cycles. When the cycle was divided
into three equal periods, androstenedione and testos-
terone were highest during the middle period (1932 and
342 pg/ml, respectively) as compared to the first (1689
and 289 pg/ml, respectively) or the last third (1553 and
252 pg/ml, respectively) of the cycle (43).

Recent animal experiments have suggested that
vasectomy may be followed by transitory testicular
dysfunction and transitory reductions in serum testos-
terone levels (57). Studies by Smith et al. and Kobrinsky
et a1. (58): however, showed that mean hormone levels
stayed within normal adult male ranges.. Plasma samples
were obtained from 50 men 1 month before and 1 month after
vasectomy. Both studies found significant changes in
individual values, but the new values were within the

normal range of variation for adult males.

Chapter 3
RESEARCH DESIGN AND PROCEDURES

The background study and subsequent research was
conducted at the Dairy Physiology Laboratory, Dairy
Science Department, in Anthony Hall, Michigan State
University. All materials and methods were those of
this laboratory. There were some slight modifications
introduced. For clarity, this chapter is separated
’into four divisions under the headings of (1) materials,
(2) experiment 1: whole blood testosterone assays,

(3) experiment 2 & 3: bloodstain testosterone assays,

and (4) hemoglobin determination.
MATERIALS

Solvents and reagents. Benzene (Nanograde, Fisher
Scientific Co.) was twice distilled. Hexane (Fisher) was
glass distilled commercially. A solution of Benxene:
Hexane (1:2) was prepared and stored at room temperature
in a dark bottle. To separate bound and free testosterone,
0.5 ml of 1.0% dextran T—70 (Pharmacia, Uppsula, Sweden)
and a 0.50% carbon decolorizing neutral norit (washed
twice with methanol, Fisher Scientific Co.) in glass

distilled water was kept at 4°C.

22

23

Glassware used in routine analysis was disposable.

 

Steroids and antiserum. Nonlabeled standard

 

testosterone was obtained from Sigma Chemical Co. 3H-l,
2,6,7-testosterone repurified by column chromatography
using Sephadex LH-20 (Pharmacia Co.), 85 c/mM, was
obtained from the New England Nuclear Co., Boston, Mass.
Anti-testosterone (MSU-74) was prepared in rabbits against
3-oxime testosterone conjugated to human serum albumin.
Antibody was diluted to l/50,000 phosphate (0.01 M)
buffered (pH 7.4) Saline with 0.1% gelatin (Knox Gelatin)

added.

Instruments. Counting of radioactive bound antigens

 

was performed utilizing Nuclear Chicago's Mark I liquid
scintillation spectrometer or their ISOCAP/300 liquid
scintillation system. Liquid scintillation fluid (3a70B
Preblend scintillation cocktail, Research Products
International Corp.,.Illinois) was used to dilute the
supernatant solution to 5 ml before placing in the
instrument.

Materials for analysis. Blood samples were drawn
from apparently normal males and females, ranging from
age 17 years to 79 years. Samples were drawn under the
authority of Dr. Laurence Simson, Pathologist, at E. W.
Sparrow Hospital, Lansing, Michigan, into heparinized
tubes to prevent coagulation. Portions of the samples

were frozen to lyse the cells for the whole blood

24

determinations. Another portion of the sample was used
for the bloodstain preparation. These were made on

cotton weave fabric (cleaning cloths).

4
WHOLE BLOOD TESTOSTERONE ASSAYS

Experiment 1

Duplicate aliquants of whole blood (0.1 ml) were
placed in disposable culture tubes. To account for pro-
cedural losses, 2700 cpm of 3H-l,2,6,7-testosterone was
added to a third aliquant from a representative number
(6-8 within each assay) of unknowns. These samples of
whole blood plus the 3H-l,2,6,7-testosterone (tracer)
were vortexed for 5 seconds and endogenous and labelled
hormone are allowed to equilibrate. The tracer is used
for recovery analysis of testosterone. This then is
compared to the total counts obtained from the radio-
active hormone itself. Comparison gives the extraction
efficiency > for the unknown sample, be it fluid blood
or dried stain. The duplicate samples of whole blood
were extracted by vortexing with 2.0 ml of Benzene:hexane
(1:2) for 30 seconds. Tubes were then stored at -20°C
for at least 1 hour to freeze the whole blood. With
precautions taken to avoid thawing the whole blood, the

organic solvent extracts destined for radioimmunoassay

 

IdProcedure of J. Mapes and H. D. Hafs, Jan. 1975.
See also Reference 59 (Smith and Hafs).

25

were decanted into 12 x 75 mm disposable test tubes and

those for procedural losses were decanted into scintillation
vials. Radioactivity of these extracts was averaged to
determine a single correction factor to account for pro-
cedural losses of testosterone in all unknowns (59).

Standard testosterone was pipetted from a stock
solution of 10 ng/ml and at least two sets (0.0,0.01,0.02,
0.05,0.l,0.25,0.50,l.0,l.5, and 2.0 ng/ml) were included
in each assay. Standard testosterone and the whole blood
extracts were dried under a stream of air.

Antibody (0.2 ml) in 0.01 M phOSphate buffered
saline containing gelatin was added to each tube. Then,
3H-l,2,6,7-testosterone (13000 cpm), in 0.2 ml of 0.1%
gelatin and 0.01 M phosphate buffered saline, was added
to each tube. The contents were vortexed for 5 seconds
and incubated at 4°C for 12-16 hours.

To separate bound and free testosterone, 0.5 ml
of 1.0% dextran T-70 and 0.50% carbon decolorizing neutral-~
norit in glass distilled water was added to each tube.
This technique is performed as quickly as possible to
avoid the stripping action of the bound testosterone by
dextran. The contents were vortexed for 5 seconds and
then centrifuged at 2,900 g for 15 minutes. After
centrifugation, a 0.5 m1 aliguant of the supernatant fluid

was diluted to 5 ml with the liquid scintillation fluid

for quantification in the instrument.

26

BLOODSTAIN TESTOSTERONE ASSAYS

Experiment 2

Bloodstains were prepared on cotton weave fabric
and on wool fabric. To achieve this, 0.1 m1 from the
samples of whole blood was pipetted onto the substrate
material. For this study, extraction of testosterone
from the stains for assay was performed between 36-48
hours after preparation.

The testosterone radioimmunoassay procedure was
essentially that used for the whole blood assays with
these modifications. Duplicate samples of the stained
substrate material were placed in 16 X 125 mm disposable
culture tubes. A third sample was prepared by drying
0.01 ml of 3H-l,2-testosterone onto the stained substratum,
and placed into a culture tube for procedural losses.

At this time, 1.0 ml of glass distilled water was added

to each unknown and the tracers. The stains were allowed
to extract in the distilled water for 1 hour. To each

tube was added 5 m1 of benzene:hexane (1:2), and the
samples extracted by vortexing for 30 seconds. An emulsion
was created in the tube by the presence of the stained
substrate material, and to overcome this problem, the

tubes were centrifuged at 2500 g for 10 minutes, stored

at -20°C for 1 hour to freeze the aqueous layer, and

then recentrifuged for 10 minutes at 2500 g. The samples

27

were again stored at -20°C for 1 hour, to freeze the
aqueous phase for ease of decanting. Extracts destined
for radioimmunoassay were decanted into 12 X 75 mm dis-
posable culture tubes and those tracers for procedural
losses into scintillation vials. The procedure then

followed that of the whole bloods.

Experiment 3

 

This experiment was conducted to improve the
extraction efficiency of the entire assay on bloodstains
and to introduce some time saving into the extraction
process.

Stains were prepared by adding .1 m1 of the whole
blood samples to cotton weave fabric. A blind trial was
also run using stains collected from casework samples
collected over the past nine months. The stains from
casework with an unknown amount of blood deposited on
them were also submitted to hemoglobin determinations,
to give an estimate of the concentration of blood dried
onto the substrate.

The procedure changed slightly from that outlined
in the 2nd Experiment, for purposes of better extraction
efficiency the tracer stains for procedural losses were
prepared by first adding the 3H-l,2,6,7-testosterone
(3H-l,2,6,7-testosterone) to the whole blood sample before

drying on the fabric. This would allow the endogenous

28

and labelled testosterone in the sample to equilibrate
before drying. The samples were extracted in distilled
water to extract the bloodstain. The solvent was added,
the tubes were stored at -50°C, for one-half hour in a
freezer, then centrifuged for 10 minutes at 2500 g. This
step essentially removed the emulsion present in the tube
after vortexing. The tubes were then refrozen at this
temperature for one-half hour before decanting. It was
found that the lower temperature broke the emulsion
better, thus eliminating the need for extra centrifuging.
The procedure then followed the prescribed assay (see
Figure 3).

The standard testosterone amounts were also changed
for this study. Two sets of the following amounts were
pipetted for the standard curve points: 0.01,0.02,0.05,
0.08,0.l,0.25,0.5,0.8,1.0, and 1.5 ng/ml. This was
included because of the extracted amounts found previously
for bloodstains, which were found to be lower than the

amount for whole blood assay in Example 1.
HEMOGLOBIN DETERMINATION

Since the amount of blood on the substrate material
may vary with the individual, the rate of bleeding, and
the conditions of drying it was necessary to measure a .
second determinant in the stain to determine the volume

of blood represented by the stain so testosterone could

2

TRACERS

STAIN

| Add .01 ml. 3H-l,2-testosterone

EX RACT IN WATER

EXTRACT WITH SOLVENT
freeze
centrifuge

freeze

DECANT

.1 ml.

5 ml.

Drabkins
CYANOMETHEMOGLOBIN
DETERMINATION

EVAPORATE +

 

AQUEOUS LAYER———————

9

UNKNOWNS

SUBSTRATE AREA APPROXIMATING
.1 ml. OF WHOLE BLOOD

STAIN
EXTRACT IN 1 m1. DIST. WATER

EX RACT WITH 5 m1. BENZENE/HEXANE (2/1)
freeze
centrifuge at 2500 rpm, 15 min.
freeze

DECANT SOLVENT

 

EVAPORATE ORGANIC SOLVENT

prepare standard testosterone
(10 ng/ml.) for standard curves

RADIOIMMUNOASSAY.‘

.2 m1. of 0.01 M PBS with .l% Knox
gelatin for antibody dilution

(1/50,000)

.2 ml. of 0.01 M P883 1% Knox
gelatin containing H-l,2,6,7-
testosterone (30,000dpm)

INCUBATE
2-3 hrs.

4° C.
with

room temp. or 16 hrs. at
.5 ml. 1.0% dextran T-70
.50% carbon neutral norit

 

CENTRIFUGE AT 2900 RPM FOR 15 MIN. AT
5° C.
SCINTILLATION

DILUTE .5 ml. WITH 5 ml.

FLUID

COUNT BOUND (4 MIN.) LIQUID SCINTILLATION

 

ADD 5 ml. SCINTILLATION LIQ

SPECTROMETER (MARKI, NUCLEAR CHICAGO
CORP.)

RESULTS IN NG/UNITS CYNHEM

Figure 3.--Flowsheet for radioimmunoassay of testosterone

in bloodstains.

30

be expressed per unit volume of blood. Since hemoglobin
is relatively constant in the bloodstream, and only
varys slightly in normal adults (values between 11-16
grams/deciliter), it was felt this might be an adequate
internal standard.

The cyanmethemoglobin method was chosen for
determining the hemoglobin level of a particular stain.
This method measures the sum of oxyhemoglobin, methemo-
globin, and carboxyhemoglobin present in a sample. On
bleeding and drying, as when a bloodstain is formed,
the oxygenation of the blood proceeds virtually to com-
pletion and then another and much slower process inter-
venes. This process is the oxidation by atmospheric
oxygen of the iron atom in the hemoglobin molecule from
the ferrous to the ferric state, to form the compound
methemoglobin (58). Since the method involves this
measuring of methemoglobin, it would be included in the
total absorption for cyanmethemoglobin at 540 nanometers
on a spectrometer.

Accordingly, a Drabkin's solution was prepared
from the following reagent grade chemicals: Sodium
bicarbonate-*2 grams; Potassium cyanide--.05 grams:
Potassium ferricyanide--.2 grams; and brought up to a
volume of 1 liter with distilled water. Hycel
Cyanmethomoglobin standard (80 mg/ml, Hycel Co., Houston,

Texas) was used for the standard curve.

31

The following amounts of cyanmethemoglobin standard
and Drabkin's solution were pipetted into 13 mm X 105 mm
matched cuvettes, and mixed by inversion:

Standard (ML) 0.5 1.0 1.5 2.5 3.0 3.5

Drabkin's (ML) 5.5 5.0 4.5 3.5 3.0 2.5
The % transmittance (measured at 540 nm) of each tube was
recorded against a blank of Drabkin's solution using a
Bausch and Lomb Spectronic 70 spectrometer. These values
were plotted on semi-log paper as % tramittance vs. con-
centration in grams/deciliter of hemoglobin.

The equivalents for the hemoglobin standard used
'in this determination were calculated from the factor
received from the amount of extract used in this deter-
mination. The factor in this study was 1/501 volume,
and the equivalents of hemoglobin was ascertained at 40
grams/deciliter.

The procedure for the bloodstain extracts was as
follows. 0.1 ml of the aqueous layer left over from the
benzene:hexane extraction was added to the 13 mm X 105 mm
cuvettes, and then 5 m1 of the Drabkin's solution was added
and let stand for 5 minutes. The cuvette was then placed
in the spectrometer and the % transmittance read at 540 nm.
Percent transmittance was then converted to grams/deciliter
by use of the standard scale (60) (see Figure 4). This

calculation was then compared to the assayed value of

T%

32

10---

9---

8---

7---

5---

5—--

4---

3---

2---

10 20 30 40
grams/deciliter (hemoglobin)

Figure 4.--Semilog plot of standard curve for determination
of hemoglobin.

33

testosterone and comparisons made on the amount of

testosterone in nanograms per units of hemoglobin.

Chapter 4
ANALYSIS OF THE DATA
CALCULATIONS

Counting data for samples and testosterone stand-
ards were analyzed by computer programs developed by
James W. Schwalm at MSU (Dept. of Dairy Science)
for calculations of the radioimmunoassay data. The
use of such programs facilitated computation of assay
data, enabling corrections for recovery, sample aliquants,
and nonspecific counts to be rapidly performed, and
giving dupliCate estimates for each sample. In this
system, the standard curve is linearized by log-log
transformation; calculated by fitting data to a poly-
nomial equation via multiple regression analysis pro-
cedures; and yields a correlation of all duplicate points

on the curve (see Figures 5 and 6).

RELIABILITY CRITERIA OF THE ASSAYS

Standard Curve
The standard curve transformation data is shown
in Figure 5 for assays performed in Experiment 2. The

regression analysis accounted for 99% of the variation

34

35

TEMPLE REP TIME 6905? GP" CORRECTED TIHE QDJUSTED HER" “DJUETED
(HTN' COUNTS ‘09 BECKGQUUND TO 10009 CPH COUNTING TIRE

 

STRNDQRD CURVE ‘09 DIRECT EMTQRCTION RSSRY ----

COUNTER REG.

1 4.00 153. 0.0 0.0000
3 4.00 154. 0.0 0.0000
3 4.00 155. 0.0 0.0000
0.0 0.0000
21040590 00. 1
< .010 M5 , ML.
1 4.00 50:7 1331.4 3.1307
3 4.00 531:. 1353.3 7.3155
3 4.00 5407. 1311.4 7.5353
1353.7 7.3533
3104 «:0 Mn. 3
1 .020 MG ML.
1 4.00 5011. 1313.4 3.3450
3 4.00 4313. 1154.4 5.5350
3 4.00 5153. 1343.3 3.0053
1303.? 5 3735
21400400 00. 3
( .050 MG ' ML'
1 4.00 4505. 1110.9 0.0015
3 4.00 4734. 1140.7 5.3553
3 4.00 4235. 10 3.7 9.5743
1035.1 9.131?
:1400400 no. 4
- .030 N6 ' ML-
1 4.00 4003. 550.3 10.4140
3 4.00 3330. 357.3 10.4435
3 4.00 3354. ‘ 343.3 10.5457
555.3 10.4534
.TWWUHJU HO. 5
c . 00 55 fit:
1 4.00 353: 557.5 11.5313
3 4.00 3513 554.4 11.5535
3 4.00 3773 304.4 11.0559
373.0 -11.37r5
:7400400 00. 5
v .350 05 - Mu-
1 4.00 3731. 533.3 15.5330
3 4.00 3433. 555.3 17.5333
3 4.00 3400. 533.3 17.1373
535.3 15.7551
STHHURPD NU. ’
t .500 05 ' wt-
1 4.00 1500. 333.3 35.1555
3 4.00 1445. 330.4 31.3034
3 4.00 1753. 400.3 34.5335 .
357.5 37.3043
21000400 N0. 5
4 .300 05 , wt.
1 4.00 1335. 305.3 33.5335
3 4.00 1305. 351.3 3.3337
3 4.00 1355 375.7 35.1445
331.3 35.5555
ZTHWDQRD M0. 9
a 1.000 05 , ML!
1 4.00 1054. 335.7 44.3131
3 4.00 1034. 330.7 43.3535
3 4.00 1017. 313.9 45.7473
323-4 44.7594
31500090 no. 10
1 1.500 05 , ML'
1 4.00 557. 301.4 49.5433
3 4.00 340. 131. 53.1577
3 4.00 94?. 195.4 50.9133
153.9 53.554?

Figure 5.--Log transformation of data for RIA standard
curve.

-fl. . 36
R SQQD I .990
IND Vflfi REBfiESSIOfl COEFFICIEHVS

l .503305405403 ‘IGURE F.Lon-Lnn TQANQ‘ORHATION at

a -.34943393E+03

3 .833109445401 DATA I=0" annnnsss 0‘ 'HT 0'
INFERCEFI - -.3479333’E403

QTAHhADh cuovr.

L06 L05
STD CURVE 9T. QCTURL Y fiCTURL Y RCTUfiL X RCIUHL K ESTIHQIED V K - ESTIHRTED X

1 7.3923 .3967 .0100 -3.0000 .013363 -.003363

. 2 2.2736 .9177 .0?00 -1.6990 .013343 .001035
3 3.1317 .9606 .0500 '1.3010 .035736 .014364

4 10.46?4 1.0199 .0300 '1.0?69 .JTB’EO .006330

3 11.3776 1.0561 .1000 -1.0000 .105?39 —.005339

6 15.7691 1.2343 .3500 -.60?1 .303344 -.053344

7 35.5646 1.4076 .5000 -.3010 .19936? .000731

3 35.3556 1.5509 .3000 -.0?6? .6?9031 .110976

9 44.7394 1.6509 1.0000 0.0000 1.030354 -.0?035$

10 3?.6547 .7314 1.3000 .1761 1.571153 -.071133

I: 4 9L0? 0F quHDQPD CUPVE 13 DESIREDu TVPE --1-‘- IE HUI TYPE -°0--.
01

THI? I: q LOG-LOG DLOT 0E IIHE (Y RWI§) 73. CONE. (K fiKIS‘.
EtERIE NOTE. --0.00-- CONE. VRLUES REE NOT PLUTTED.

IV
IU
no
w
J\

2.0660

1.91?5

§
4
O
4
4
O-
4
4
f
O
4
1|-
O
4
4
O
4
4
4
I 1"“? §
§
4 4
O
4 4
1.5333 +
6
4 4
.-
.-
1.479/ 4
O
4 4
4
0'
1 33°? +
’ 0
4
0-
4 4
- 1.1506 +
O
O
O-
- 4
1.0320 4 4
f
4 4
4
4 4
.3301 4 4
4 4 - - - - 4 - - - - 4 — - - - 4 - - - - 4 - - — - 4 ----- 4 - -
10 -2.0356 -l.4637 -.300? -.33?3 .3253 .7331 1.3011.

7115193

Figure 6.--Log-Log transformation of data and regression
analysis for fitting the standard curve.

37

(R2 = 0.990). The between assay variation of the standard
curve points was significant (r = .988) for each point on
the standard curve. A comparison was made between the
log-log transformation of five replicate standard curves

of assays run on different days.

Recovery and Precision

The mean recovery of tritiated (3H-testosterone)
from the whole blood samples of Experiment 1 was 86 t 4.8
(S.D.) %. The mean recovery of triated testosterone for
Experiment 2 of the bloodstain assays was 61 t 8.0 (S.D.)
%. The mean recovery for Experiment 3 of the bloodstain
assays was 84 t 6.9 (S.D.) %. The accuracy and precision
of the assay were estimated by repeated analysis of pooled
whole blood from males and pooled stains from females.
The interassay precision was 5-9% (coefficient of variation)
for the pooled whole blood and 8-11% (coefficient of

variation) for the pooled stains.

Specificity

The specificity of the method was determined by
that of the antibody employed in the assay. It was
tested by determining cross-reactions with various
steroids and sterols, and by the comparison of RIA of
testosterone in direct extracts of whole blood with
that of testosterone isolated from the same extracts by

column chromatography on Sephadex LH-20.

38

The chromatographic validation was performed to
verify the specificity of the antibody used to determine
testosterone in human whole blood. Ten male and ten fe-
male whole bloods were selected. The testosterone con-
centration of each was determined by the RIA procedure
outlined in the methods chapter (see page 24) on direct
extracts. In addition, on extracts of each of the samples
chromatography was performed on Sephadex LH-20. In this
procedure, 28 cc. Glaspak mini-columns were packed with
Sephadex. After evaporation of the organic solvent
extract, 0.2 ml of benzene:methanol (90:10) was added
to each extract, and this was placed on the respective
column. Benzenezmethanol (90:10) was then eluted through
the column, letting pass the first 0.8 ml of solvent and
collecting the next 2.0 ml. This fraction of the eluent
containing testosterone was then submitted to the RIA
procedure. Table 2 shows the comparison of the direct
extraction of whole blood with those isolated by
chromatography. As the results show, the values are cor-
related (r=0.988) for the separate techniques.

The antibody used (MSU-74) was tested against
other steroids and sterols found in human blood; it
cross-reacted significantly only with dihydrotestosterone
(28%) (see Table 3). Since the concentration of dihydro-

testosterone in males and females is only 0.65 i 0.24

Comparison of Testosterone Levels for Direct
Extraction and Column Purification of

39

Table 2

Whole Blood Extract.

 

 

Direct Column
Sex Age ng/ml ng/ml
Male 30 4.067 5.782
Male 34 4.064 3.823
Male 54 7.430 7.299
Male 49 4.628 4.967
Male 62 3.645 2.306
Female 44 .321 .386
Female 28 1.214 .574
Female 55 .691 .868
Female 20 .671 .631
Female 34 .205 .220
Male 20 5.174 4.668
Male 48 7.169 7.743
Male 45 5.818 7.648
Male 68 .643 .654
Male 24 6.662 6.167
Female 27 .399 .408
Female 45 .489 .301
Famale 76 .303 .342
Female 42 1.270 1.325
Female 55 .205 .226

 

40

Table 3

Percent Binding, Sensitivity, and Crossreactions
of MSU-74 Rabbit Antitestiosterone Sera.a

 

 

Steroid % Crossreactionc
Dihydrotestosterone 28
Androstenedione 0.9
Androstadiene-3,17-dione 0.04
Epiandrosterone 0.03
Dihydroepiandrosterone 0.02
Estradiol 0.02
Estrial 0.02
Estrone 0.02
Progesterone 0.02
l7-a-hydroxyprogesterone 0.01
Cortisol 0.01
Cholesterol 0.001
% Binding in zero tube 39b
% Binding reduction by .01 ng. 26b

 

a1/50,ooo dilution.

bBased on three determinations.

cPooled data from antibodies 2,15,17,19 and 21
comprising MSU-74.

41

(S.D.) ng/ml and 0.29 i .09 (S.D.) ng/ml respectively,
these amounts would not seem to interfere in this type

of study.

Sensitivity

Sensitivity was estimated by including background
tubes in the counter, and by including zero tubes (no
standard testosterone added) in with the assays. In the
bloodstain assays, blanks were controlled by including

unstained substrate material throughout the determination.
EXPERIMENT l--WHOLE BLOOD ASSAY RESULTS

Two hundred whole blood samples were obtained for
testosterone analysis. Of these, a portion of 170 were
frozen for RIA analysis. Twenty-six aberrant observations
were found in the study, mainly from errors in the tech-
nique. Of these, several whole bloods were unsuitable
for preparation into a dried stain, either from lysing
due to storage problems or contamination. Table 4 shows
the data for the 144 samples. Figure 7 shows the distri-
bution of the samples over the range of testosterone
measured.

Although the values for males are somewhat lower
than the values reported in the literature for serum or
plasma, the age-range in the Table 4 is much larger. In

this particular study of whole blood, for males in the

42

Table 4

Mean Whole Blood Testosterone Levels in
Males and Females-—Experiment 1.

 

 

(ng/ml)
Age 17-60 61-79
Range .727-1l.27 .313-4.95
Males Mean+S.D. 3.75 i 2.48 1.39 i 1.26
n 63 16
Range .113-l.274 .103 - .691
Females Mean+S.D. .421 i .22 .361 f .18
n 53 14

 

age-range 22-45, it was shown that the testosterone con-
centration was 1.136 - 10.126 ng/ml, Mean+S.D. to be
4.37 i 2.20 ng/ml. This is consistent with the values
reported in the literature for this age range.

Since the sample used was whole blood, the cellular
constituents of blood were not separated as they would be
in the plasma or serum testosterone estimates. This could
be compared to the levels of glucagon in the blood, when
measurements with whole blood as compared to serum or
plasma were different to 30 percent, based on the hemato-
crit or red cell count in the body, which comprises
approximately 30 percent of venous blood. This study
indicates that the steroid hormone levels do not seem to
be dependent on this observation.

Differences between males and females were studied

statistically using a paired t-test, and significant levels

43

.mnmwm mblam moon .mmHmEmm can moame How mcoflumuucmocoo mopnaocHn

fluency moamamm Amhucv moamzm
.H ucofiwummxm "mammmm cooHn maonz n.msoumumoumou
mo mcowumnucooaoo mo mmmamamm can moama cw monouusooo usmoummll.> musmwm
AHE\mzv maoumumoumma
.. ... ..

 

. o . 9.2.: 2.... ,. . o . J
. D 5

. mm
on
o
o
m
1 cm .1
E
u
o
a
moamsom “ammo r mm

MOHMS u @HHOM

 

 

 

r 00H

44

of difference were found at the p=.01 level, df=51. How-
ever, males aged 61-79 had testosterone concentrations
equivalent to that in females 17-60 (df = 15). Therefore,
testosterone alone does not differentiate between males

and females of these age groups.
EXPERIMENT 2--BLOODSTAIN ASSAY RESULTS

Seventy-one stains were studied in this experiment
over several different assays. A11 stains were prepared
in the laboratory and were tested 36-48 hours after
drying. The results of these assays are shown in Table 5

and in Figure 8.

Table 5

Mean Bloodstain Testosterone Levels in Males
and Females--Experiment 2.

 

 

(ng/ml)
Age 17-60 61-79
Range 2.04-7.325 2.11-4.67
Males MeaniS.D. 3.61 i 1.17 3.43 i .70
n 28 10
Range 1.85-3.67 .92-2.81
Females MeaniS.D. 2.57 i .48 2.18 t .54
n 22 ll

 

Extraction efficiency based on the tracer stains
prepared was in the range of 47-56 percent. The extraction

efficiency is used to give a correction factor to approximate

.mumwm mhuam momm .moamawm can mmHmE mo mcofiumuuamocoo mopsHOGHm

.m ucosflnmmxm
Aeouuonv .N ucmeummxm Amouv "mammmm cfimumpooam .mcouwumoumou

mo mcowumuucmoqoo mo wmmamamm 0cm moame.:H mocmnusooo unmoummnl.m madman

AHE\mqv mcoumumoumma
m.m o.m m.¢ o.¢ m.m o.m m.m o.m m.H o.a m.o

 

45

 

5&8:

 

 

 

 

 

.3
F

 

 

 

 

 

 

 

%
.2: m
o.m m.-~ o.~ m4 04 m.o m
E
g s E a.
3
9
IL
.3

mmHmEmm "ammo
moan: ucfiaom

 

1.00..”

46

the expected values received in the whole blood assays.

As can be seen from the results, the correction factor
only increased the overall values for female concentrations,
which could be misleading. Although the difference between
male and female stains was significant using a paired t-
test, a different method was incorporated to increase

the extraction of testosterone from dried blood. Experi-

ment 3 was conducted to look at this problem.
EXPERIMENT 3--BLOODSTAIN ASSAY RESULTS

Two assays consisting of 60 stains were utilized
for this experiment, as outlined in the design and pro-
cedure chapter. The results are shown in Table 5 and

distribution for the total assay are shown in Figure 8.

Table 6

Mean Bloodstain Testosterone Levels in Males
and Females--Experiment 3.

 

 

(ng/ml)
Age 17-60 61-79
Range .775-3.275 .539-l.116
Males MeaniS.D.' 1.68 1 .54 .769 t .25
n 27 8
Range .439-1.281 .313-.724
Females MeaniS.D. .590 i .17 .550 t .19

n 29 6

 

47

The extraction efficiency for this experiment
(84 t 69 [S.D.] %) gave a lower correlation factor (11.0)
than in Experiment 2 (61 i 8.0% extraction efficiency)
(17.2 correction factor). The results show a lowered
value for male testosterone concentration, even with
the correction factor, which suggests a systematic error.
One explanation is the drying process onto cotton sub-
strate material. The results may be explained by postu-
lating that during the drying process of a normal blood-
stain, i.e., venous blood, clotting proceeds and fibrin
threads are formed that become attached to the material
substratum and are not soluble under the conditions of
extraction reported here. Besides fiber adhering to the
substratum, loss of sensitivity is because crusts have
not completely dissolved to obtain a complete suspension
of blood in solution. Structurally, the cellular con-
stituents of blood are composed of a lipoprotein envelope
enclosing a red, iron-containing protein known as hemo-
globin.

Since most steroids are bound to plasma proteins
(61), solvent extraction is less effective if the plasma
proteins bind the steroids with high affinity. This
binding of the steroid could be enhanced by the affinity
of the protein to the material substratum. Therefore,
if extraction is carried out soon after addition of

radioactive tracer (one-half hour in this experiment),

48

the added tracers are still unbound and extract easily
and efficiently, whereas the steroids already present
(up to 48 hours on the substrate material) are more
tightly bound and resist extraction. The end result is
a higher recovery of tracers than of endogenous steroids
and when correction for recovery is made, the endogenous
level of these samples is under estimated.

If plasma and cellular proteins also have high
affinity for the substrate material, as has been postu-
lated for fibrin, then the endogenous steriods also would
be more tightly bound to the substrate material. How-
ever, this does not explain the higher extraction efficiency
in Experiment 3 than Experiment 2. If anything, the radio-
active tracer being dried onto the already prepared stain
(as was done in Experiment 2) would be bound only to the
outside crust of blood on the substrate. Possibly, this
layer of crusted dried blood on the outside of the stain
was undissolved in the aqueous layer, causing it to be
untouched by the solvent extraction.6

The results between males and females were com-

pared using a paired.t-test and found to be significantly

 

6One test was carried out where the stains were not
extracted first in distilled water, Benzene:hexane was
added directly to the stained substrate material. The
extraction efficiency was calculated at .6%. The drying
process therefore must encapsulate in someway the constitu-
ents of blood, making them impervious to organic solvents.
This has been found to be true in extracting with other
solvents except those of a very polar nature, close to
that of water.

49

different at the p=.05 level, df=26. However, no differ-

ence was found between males 61-79 and females 18-60.

Blind Trial Study

Eighteen stains were used for this study and
extraction was performed as in Experiment 3. The stains
were prepared from casework whole bloods received up to
fourteen months earlier in the lab. The aqueous layer
was submitted to the cyanmethemoglobin determination
after extracting with the organic solvent. Prior to
the blind study trial, the 0.1 m1 stains prepared for
Experiments 2 and 3 were submitted to this determination.
These bloodstains fell within the range of 65-75 percent
transmittance (4.6 - 7.5 gms/dl). Although age affects
the amount of hemoglobin detected (62), it was felt that
the amount present in the size of cut substrate would be
fairly consistent, regardless of the age of the dried
stain. .The results of the study are shown in Table 7
and the distribution across the concentration of testos-
terone measured is shown in Figure 8.

As the results point out, there seems to be no
relationship between the amount of testosterone present,
the hemoglobin level after organic solvent extraction,
and the age of the stain. Stains No. 2, 4, and 5 illus-
trate this. All stains cut from cotton fabric approximated

the size of a 0.1 m1 stain. The stains fell within two

50

Table 7

Results of Blind Trial Study.

 

 

Testosterone Age of
. a Detected Hgb Stain
Stain # Sex ng/ml gms/dl (months)
1 F .043 35.2 2
2 M .914 35.8 4
3 F .353 4.8 7
4 M 3.046 17.6 5
5 M 1.584 18.6 6‘
6 M 1.239 9.3 9
7 M .980 2.0 13
8 F .122 2.3 11
9 M .849 15.8 3
10 F .066 20.0 1
11 M .237 4.8 14
12 F .081 7.4 5
13 F .137 7.2 4
14 M 1.868 7.0 2
15 M 1.322 7.4 15
16 F .063 11.8 5
17 M 1.306 9.3 5
18 F .033 6.3 4

 

aThe unknown stains used were received from
caseload whole bloods in the laboratory. Age range for
these reported were between 22 and 61 years.

51

months of being prepared, but the hemoglobin level in
grams per deciliter calculated from the cyanmethemoglobin
method were not consistent with the concentration of
testosterone. Differences do exist for males and females
regardless of the hemoglobin level, but the spurious
results detract from the overall method--to express
testosterone concentration in a particular volume of
blood in a stain.

It is known that during the drying process,
especially on absorbent material, that cellular components
dry in the center of a stain while serum migrates to the
periphery of the stain (63). One explanation for the
inconsistent results could be the portion of a dried stain
that is used for analysis. It was found that there were
undissolved crusts in the aqueous solution, which might
prohibit the complete dissolving of hemoglobin in the
suspension. In this regard, the presence of cellular
constituents from the center of a stain could "blanket"
the cyanmethemoglobin spectrum along with coextracted
degradation products from the denaturation of blood during
the drying process.

However, it must be noted that differences in
males and females were detected, and this investigator
judged sixteen correct out of the eighteen stains tested,
a 90 percent proficiency. Stain No. 11 was a male stain,

age 61, which could account for the lowered testosterone

52

concentration. If one were to make 1.25 ng/ml the
discriminating level, where concentration above this
certain amount in ng/ml would mean the stain was of
male origin, then for this study six of the ten male
stains would have been judged correctly. This investi-
gator though allowed for the results of the entire assay
to enter into the judgment. This observation strengthens
the idea that proper controls and reference stains should
be included within every assay attempted.

The results do show that some other method for
relating testosterone to the volume of blood present in

a stain will have to be investigated.

Chapter 5
DISCUSSION

The purpose of this study was to devise a radio-
immunological method for the specific determination of
testosterone in bloodstains. Fortunately, there was
available a selective antisera with carefully checked
specificities. Results from the validation study between
direct extracts of known levels of testosterone in whole
blood against column chromatographic separations of
testosterone showed the antisera to be suitable for human
studies. The correlation coefficient between the direct
extraction and the chromatographic separation was signifi-
cant (r=.986) (see Table 3). The properties of the
antibody had been characterized in the Dairy Physiology
Laboratory, Michigan State University, and the specificity
obtained from the chromatographic validation gave con-
siderable reliability to the data.

With regard to the cross-reactivity of dihydro-
testosterone, a number of reports have appeared in recent
years on its concentration in the serum of normal men
and women (64,65). There is a considerable variation in
the values reported in different studies, probably because

there are no anti-dihydrotestosterone antibodies, as of

53

54

yet,and measurement is usually achieved in samples with
the use of anti-testosterone itself. Therefore, different
antibodies of testosterone would react with variable
results in these studies.

The difference in serum dihydrotestosterone
concentration between males and females is highly sig-
nificant (P=.001). The ratio of testosterone to dihydro-
testosterone is much lower in serum from females than
from males, being 1.4 and 5.2 respectively. Moreover,
the concentration of dihydrotestosterone in serum is
higher in the luteal than in the follicular phase of
the menstrual cycle. This is analagous to serum testos-
terone concentrations in the female (66), and might
reflect a secretion of dihydrotestosterone by the corpus
luteum, or an increase of its precursor testosterone
during the latter half of the menstrual cycle. One female
in this study with the whole blood assays, and included
in the validation study, was found to have 1.23 ng/ml
testosterone in direct extraction assays, while the
separation on Sephadex LH-20 showed a testosterone con-
centration of 0.57 ng/ml. Because of the lowered amounts
of the male values in the bloodstain assays, this type of
cross-reaction might be a difficulty in differentiation
probabilities. Without the use of chromatography though,
the false increases in apparent serum or plasma testos-

terone would be approximately 10-15 percent in men, but

55

closer to 30-40 percent in women, with the use of this
antisera. To alleviate this problem of the female during
the menstrual cycle and the cross reactivity of dihydro-
testosterone, a chromatographic separation should be
introduced into the procedure.

Testosterone data from the whole blood assays
of Experiment 1 were consistent with those in the litera-
ture for blood serum or plasma. Therefore, the samples
used seemed not to be endocrinologically abnormal. The
twenty-six samples discarded were not suitable for blood-
stain preparation because of storage problems of the whole
blood causing the sample to 1ySe.

The bloodstain analysis of Experiment 2 illus-
trated the problem with faulty preparation of the tracer
stains for procedural losses. The extraction difficulties
encountered with the tracer stains may be explained by
the drying process. Adding the labelled testosterone
suspended in an organic solvent to a crust of dried blood
on a particular substrate did not allow the endogenous
steroids in the stain to equilibrate with the added steroid.
Fixation of stains by organic solvents is a routine prac-
tice in forensic science, for it prevents the dissolving
of hemoglobin and other water soluble material. This
phenomenon is consistent with the lower extraction effi-
ciency of Experiment 2. Unfortunately, the causes of the

particular behavior of dried blood are not completely

56

clear at this time. Possibly, the lowering of sensitivity
is because the crusts have not completely dissolved in
distilled water to obtain a complete suspension of stroma.
Perhaps a slight change of pH of the aqueous solution,
similar to the use of ammoniacal extracts deve10ped for
aged bloodstains will offer a solution to this problem.

Extraction efficiency in Experiment 3 was higher
than that in Experiment 2. Lower values for males were
observed, with no significant change in the female con-
centration as compared to the whole blood results of
Experiment 1. Lower values would also be expected in
the females, and this adds to the reason for introducing
a chromatographic separation to reduce cross-reacting
steroids. Moreover, when dealing with a lower concentra-
tion of testosterone, as in females, the standard curve
used should be changed. Since the values fall on the
low end of the curve, sensitivity is not as great as for
the male values. Possibly, this could only be alleviated
when, once a female is indicated, that a second RIA pro-
cedure be run with a lower concentration of testosterone
used for the standard curve.

But the lower values for males (although dis-
tinguishable from the females) is an inherent problem
in the technique. The higher extraction recovery in the
tracer stain was contrary to the expected extraction

recovery in the unknowns. Perhaps the use of a bloodstain

57

tracer prepared one-half hour before extraction caused
this problem. The equilibrating for only a half-hour of
endogenous and labeled testosterone may not have bound
the tracer to the products of the whole blood as much

as the testosterone already present. One solution might
be the storage of tracer whole bloods or tracer stains
over several different periods, so that if a stain were
to be analyzed, the tracer used would be more approximate
in age to the stain being tested. Extraction efficiency
then would more closely approximate the extraction of
endogenous steroid. It might also be possible to elimi-
nate the need for tracer analysis altogether, since
differentiation of the sexual origin of stains is depend-
ent only on the level or concentration of a particular
steroid.

The differentiation between sexes was the ulti-
mate goal of this research. Figure 9 shows the distri-
bution of testosterone levels of bloodstains analyzed
in Experiment 3. The graph shows that approximately 90
percent of the male testosterone concentrations fall
above the level of 1.25 ng/ml. For forensic purposes,
this level would be the differentiation point for deter-
mining whether or not a stain was in fact of male origin.
At this time, the measurement of testosterone levels below
this point would be inconclusive, as was shown in the

t-test for differentiating males of age 61-79 years and

58

 

 

 

 

 

 

AMales
0 Females
4A
3.0.
A» 40
A A
2.0.
46
A A
A» A. (5
A» 4A
1.5. A
4AA: 13€B
a, 4A
)- ------------------------------------------------
0 £3 £3
‘23 A
Ng/ml. 1 [L
013‘3
O 00 A O
0
£5
0 o 0 AA
.5. 0 °0 0 0° °
00 O
o o 0
<0 0 o o
o O
O O
0
17-25 26-45 46-60 61-79
Age

Figure 9.--Distribution of testosterone levels in males
and females of different age groups. Results
of bloodstain assay--Experiment 2.

59

females of age 17-60 years. The differences in levels
of older males and older females are discrete, but the
probability of differentiation at these levels is
impractical.

The results of the blind study trial indicate
that this method of radioimmunoassay is a valid technique,
even for unknown amounts of blood in a stain. The 90
percent proficiency would be acceptable in lieu of the
points in Figure 9. On a practical basis, the statement
that data points below 1.25 ng/ml would be inconclusive,
increases the validity of such a determination. It is
unfortunate that the level of testosterone present could
not be correlated with the hemoglobin content calculated
with the cyanmethemoglobin method. The method may have
several drawbacks. In addition to the age of a stain,
the portion of a stain that is tested seems to influence
the measurement. It is felt that some other method of
measuring the amount of blood present in a stain should
be studied. This study also indicated no difference in
analytical characteristics in testosterone levels in
fresh stains and those up to fourteen months in age.

At the time that this research was undertaken,
other investigators at the FBI laboratory and in Pittsburgh
were looking at the problem of the amount of blood pre-
sent in a stain as an index to the testosterone concen-

tration. This work is not completed, but preliminary

60

studies show results for testosterone levels consistent
with those presented in this paper.

. Brown and associates at the FBI Laboratory suggest
a different approach to the amount of blood present in a .
stain (67). Their method includes the measurement of
estradiol and progesterone in the stain and comparing
these with the concentration of testosterone in the same
stain. Testosterone to estradiol ratios would be higher
in the male, and both would increase on a commensurate
basis with the amount of blood present on the substrate.
Progesterone would aid as an index to the female in the
late luteal phase of the menstrual cycle, which would
account for a higher testosterone estimate.

Shaler, et a1. (68) at the University of Pitts-
burgh are also looking into the sexual origin of blood-
stains at this time. Besides measuring estradiol as the
internal standard against testosterone, they are at-
tempting to fractionate the organic solvent extract from
the bloodstain in order to identify the enzyme Peptidase
A (an isoenzyme of peptidase present in blood which has
a racial population frequency) and to separate another
fraction for toxicological analysis. This type of
analysis should find merit in the area of forensic science.

For the present, the relative ease of the RIA
assay of testosterone using specific antibody gives a

fairly reliable indication of the sexual origin of a

61

bloodstain. Although the amount of blood present in a
stain for relating back to testosterone concentrations
without the use of another assay presents difficulties,
it appears that radioimmunoassay of steroid hormones
offers the convienience of practicability and objective
interpretation that other methods of sex determination
on bloodstains lack. The RIA method for testosterone

is seemingly neither dependent on the age of a stain

nor hopefully on the conditions of drying, i.e., bacterial
contamination, temperature, or putrefaction. Again,
this is based on the relative stability of the steroid
hormones. These conditions should be given considerable

attention, though, by investigators.

Chapter 6
SUMMARY

Radioimmunoassay of testosterone in whole blood
and bloodstain extracts was applied for the development
of a technique for the determination of the sexual origin
of a bloodstain. To account for the amount of blood
present on a stained substrate material, a cyanmethemo-
globin method was utilized for hemoglobin determinations
as an internal standard to relate to the testosterone
concentration. Testosterone in whole blood for males and
females averaged 3.75 i 2.48 (S.D.) and 0.421 f 0.22 (S.D.)
ng/ml respectively for the age range 17-60 years. For
the age range 61-79 years, the values were 1.39 i 1.2
(S.D.) and 0.361 t 0.18 (S.D.) ng/ml for males and fe-
males respectively.

An improved extraction technique for bloodstains
indicated average testosterone concentrations of 1.68 t
0.54 (S.D.) ng/ml for males and 0.590 1 0.17 (S.D.) ng/ml
for females over the age range 17-60 years. Values for
older males (0.769 f 0.25 [S.D.] ng/ml) and older females
(0.550 t 0.19 [S.D.] ng/ml) were also noted. Males aged
17-60 had higher (p=.05) testosterone concentrations than

females, ages 17-60, but no significant difference was

62

63

found between females 17-60 years and males 61-79 years.
The discriminating probability of the two latter groups
would then approach random error (p=.5).

A blind study trial showed 90 percent proficiency
for this investigator, judging the sexual origin on 16
out of 18 unknown bloodstains. However, no correlation
was found between the amount of blood present in a stain
as determined by the cyanomethemoglobin method and the
concentration of testosterone. However, the differentia-
tion of males from females on a stain approximating 0.1
m1 of dried blood is possible, regardless of the amount
of blood present.

The major conclusion of this study is that radio-
immunoassay is a reliable and viable technique for deter-
mining the sexual origin of bloodstains in the analysis

of physical evidence for forensic purposes.

BIBLIOGRAPHY

64

10.

BIBLIOGRAPHY

Ellison, J. R..and Barr, H. J. Quinacrine Flourescence
of Specific Chromosome Regions. Chromasoma (Berl),
1973, 36, 375.

Barr, M. L. and Bertram, E. G. A Morpholoogical
Distinction Between Neurons of the Male and
Female, and the Behavior of the Nucleolar
Satellite during Accelerated Nucleoprotein
Synthesis. Nature (1949), 163, 676.

Philips, A. P.; Gaten, E.; and Webster, D. F. The
Detection of Sex Chromatins in Bloodstains and
Solid Substrata. C.R.E. Report (1971), 56.

Philips, A. P. and Webster, D. F. Improved Y-
Chromosome Flourescence in the Presence of
Magnesium Ions. g. Forens. Sci. Soc. (1972),
12, 2.

Sanderson, A. R. and Stewart, J. S. Nuclear Sexing
with Acetoorcein. B.M.J. (1961), 5259, 1065.

Barr, M. L. and Carr, D. H. Nuclear Sex. In
Chromosomes in Man, Ch. 6, Ed. by J. L. Hamerton.
Laverham Press, Ltd. (1965).

 

Philips, A. P. and Gitsham, C. The Identification
of Male Bloodstains by Y-chromosome Flourescence.
J. Forens. Sci. Soc. (1974), 13, 47.

Philips, A. P.; Gaten, E.; and Webster, D. August,
1972, Home Office Central Research Establishment,

Report 69.

Whitehead, P. H. and Divall, G. B. Assay of "Soluble
Fibrinogen" in Bloodstain Extracts as an Aid to
Identification of Menstrual Blood in Forensic
Science: Preliminary Findings. Clin. Chem.
(1973), 19, 762.

Whitehead, P. H. and Divall, G. B. The Identification
of Menstrual Blood--The immunoelectrophoretic
Characterisation of Soluble Fibrinogen from
Menstrual Bloodstain Extracts. Forensic Science
(1974), 4, 53.

65

11.

12.

13.

14.

15.

l6.

l7.

18.

19.

20.

21.

22.

66

Asano, M.; Oya, M.; and Hayakawa, M. Identification
of Menstrual Bloodstains by the electrophoretic
pattern of Lactate Dehydrogenase Isoenzymes.
Forens. Sci. (1972) 1.330.

 

Bellar, F. K. Observations on the Clotting of
Menstrual Blood. Amer. J. Obstet. Gynecol.
(1971), 111, 535.

Popielski, B. A study of menstrual blood traces.
Bull. Acad. Pol. Sci. Lett. Cl. Med. (1949)
275.

 

 

Culliford, B. J. Identification of Menstrual Blood
in dried stains by an electrophoretic method.
g. Forens. Med. (1959), 6, 112.

 

Mole, R. H. Fibrinolysin and the fluidity of blood
post mortem. J. Pathol. Bacteriol. (1948), 60,
413.

Roth, J.; Glick, S. M.; Klein, L. A.; and Petersen,
M. J. J. CLIN. Endocrinol. Metab. (1966), 26,
671.

Lieberman, 8.; Erlanger, B. F.; Beiser, S. M.; and
Agate, F. J. Recent Progr. Hormone Research.

(1959) , 1_5, 165-200.

Thorneycroft, I. H.; Tillson, S. A.; Abraham, G. E.;
Scaramuzzi, R. J.; and Caldwell, B.V. Preparation
and purification of antibodies to steroids.
Immunologic Methods in Steroid Determination,
Appleton-Century-Crofts, 63-86.

Goodfriend, L. and Sehon, A. Early Approaches to
Production Analysis, and use of Steroid-~Specific
Antisera. Immunologic Methods in Steroid Deter-
mination, Appleton-Century-Crofts, 15-39.

Abraham, G. E. Radioimmunoassay of Steroids in
Biological Materials. Acta. Endo. S pp. (1974),
15, 183.

Yalow, R. S. and Berson, S. A. Assay of Plasma
Insulin in Human Subjects by Immunologic Methods.
Nature. (1959), 184, 164.

Abraham, G. E. Radioimmunoassay of Steroids in
Biological Fluids. Clin Biochem.7. (1974), 3, 196.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

67

Furujama, S.; Mayes, D. M.; and Nugent, C. A. A
Radioimmunoassay for Plasma Testosterone.
Steroids. (1970), 16, 415.

Dufau, M. L.; Catt, K. J.; Tsuruhara, T.; and Ryan,
D. Radioimmunoassay of Plasma Testosterone.
Clin. Chem. (1972), 31, 109.

 

Coyotupa, J.; Parlow, A. F.; and Abraham, G. E.
Simultaneous Radioimmunoassay of Plasma Testosterone
and Dihydrotestosterone. Anal. Lett. (1972), 5,
329.

Anderson, P. H.; Fukushima, K.; and Shiller, H. S.
Radioimmunoassay of Plasma Testosterone, with
use of Polyethylene Glycol to Separate Antibody-
Bound and Free Hormone. Clin. Chem. (1975),
21/6, 712.

Barbaria, J. M. and Thorneycroft, I. H. Simultaneous
Radioimmunoassay of Testosterone and Dihydro-
testosterone. Steroids. (1974), g}, 757.

Apter, D.; Janne, 0.; Karvonen, P.; and Vihko, R.
Simultaneous Determination of Five Sex Hormones
in Human Serum by Radioimmunoassay after
Chromatography on Lipidex 5000, Clin. Chem.

(1976), 31L, 37.

Auletta, F. J.; Caldwell, B. V.; and Hamilton, G. L.
Androgens: Testosterone and Dihydrotestosterone,
in Methods of Hormone Radioimmunoassay, Eds. by
B. M. Jaffe and H. R. Behrman. New York:
Academic Press, 1974, 368.

Kirschner, M. A.; Lipsett, M. B.; and Collins, D. R.
Plasma Ketosteroids and Testosterone in Man: A
Study of the Pituitary-Testicular Axis. g. Clin.
Invest. (1965), 44, 657.

Gandy, H. M. and Peterson, R. E. Measurement of
Testosterone l7-Ketosteroids in Plasma by the
Double IsotOpe Dilution Technique. g. Clin.
Endocrinol. Metab. (1968), 28, 949.

Horton, R. and Tait, J. F. Androstenedione Production
and Interconversion Rates Measured in Peripheral
Blood and Studies on the Possible Site of Its
Conversion to Testosterone. g, Clin. Invest.
(1966), 45, 301.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

68

Gomez, E. C. and Hsia, S. L. In Vitro Metabolism of
Testosterone-4l C and D4-Androstene-3,17-dione 4
14C in human skin. Biochemistry. (1968), 1, 24.

Wilson, J. D. and Walker, J. D. The Conversion of
Testosterone to Sa-Androstan-l7b-ol-3-one
(Dihydrotestosterone) by Skin Slices of Man.

g. Clin. Invest. (1969), 48, 371.

 

Southren, A. L.; Gordon, G. G.; and Tochimoto, S.
Plasma Production Rates of Testosterone in Normal
Adult Men and Women and in Patients with the
Syndrome of Feminizing Testes. g. Clin. Endocrinol.
(1965), 25, 1441.

Resko, J. A. and Eik-Nes, K. B. Diurnal Testosterone
Levels in Peripheral Plasma of Human Male Subjects.
J. Clin. Endocrinol. (1966), 26, 573.

 

Rubin, Robert T.; Gouin, Paul R.; Lubin, Ardie:
Poland, Russell E.; and Pirke, Karl M. Nocturnal
Increase of Plasma Testosterone in Man: Relation
to Gonadotropins and Prolactin. J. Clin.
Endocrinol. Metab., 40: 1027-1033, 1975.

 

Burger, H. G.; Brown, J. B.; Catt, K. J.; Hudson, 3.:
and Stockigt, J. S. In Protein and Polypeptide
Hormones. Ed. by M Margoulies. Excerpta Medica
Foundation International Congress Series,
Amsterdam (1968) 161, 412.

Abramovich, D. R. and Rowe, P. Foetal Plasma
Testosterone Levels at Midpregnancy and at Term:
Relationship to Foetal Sex. g. Endocrinol. (1973),
56, 621-622.

August, G. P.; Thachuk, M.; and Grumbach, M. M.
Plasma Testosterone Binding Affinity and Testos-
terone in Umbilical Cord Plasma, Late Pregnancy,
Prepupertal Children and Adults. g. Clin.
Endocrinol. Metab. (1969), 29, 891.

 

Frazier, S. O. and Horton, R. Androgens in the
Peripheral Plasma of Prepupertal Children and
Adults. Steroids. (1966), 8, 777.

Korenman, S. G. and Lipsett, M. B. Is Testosterone
Glucuronoside Uniquely Derived from Plasma
Testosterone. ,1. Clin. Invest. (1964), 43, 2125.

43.

44.

45.

46.

47.

48.

49.

50.

69

Winter, J. S. and Faiman, D. Pituitary-Gonadal
Relations in Male Children and Adolescents.
Pediat. Res. (1972), 6;, 126. See also,
Sizonenko, P. C. and Paunier, L. Hormonal
Changes in Puberty: III. Correlation of Plasma
Dehydroepiandrosterone, Testosterone, FSH, and
LH with Stages of Puberty and Bone Age in Normal
Boys and Girls and in Patients with Addisons
Disease or Hypogonadism or with Premature or
Late Adrenarche. J. Clin. Endocrinol. Metab.
(1975), 4;, 894.

 

 

Kent, J. R. and Acone, A. B. In Androgens in Normal
and Pathological Conditions. Eds. A. Vermuelen,
and D. Exley. Excerpta Medica Foundation Inter-
national Congress Series, Amsterdam (1966), JQJ,
33.

Purvis, K.; Brenner, P. F.; Landgren, B. M.; Cekan,
C.; and Diczfalusy, E. Indices of Gonadal Function
in the Human Male. I. Plasma Levels of Unconjugated
Steroids and Gonadotropins under Normal and
Pathological Conditions. Clin. Endocrinol.
(1975), 4, 237.

 

Lipsett, M. B.; Davis, T. E.; Wilson, H.; and Canfield,
C. J. J. Clin. Endocrinol. (1965), 25, 1027.

 

 

Paulsen, C. A.; Gordon, D. L.; Carpenter, R. W.;
Gandy, H. M.; and Drucker, W. D. Klinefelters
Syndrome and its Variants, A Hormonal and
Chromosomal Study. Recent Progr. Hormone Research.
(1968), 33, 321.

Lipsett, M. D.; Wilson, H.; Kirschner, M. A.; Korenman,
S. G.; Fishman, L. M.; Sarfaty, G. A.; and Bardin,
C. W. Studies on Leydig Cell Physiology and
Pathology: Secretion and Metabolism of Testos-
terone. Recent Progr. Hormone Research. (1966),
22, 245.

Hudson, B.; Burger, H. G.; and de Kretser, D. M. In
The Human Testis. A Advances in Experimental
Medicine and Biology, Vol. 10, eds. E. Rosenbert,
and C. A. Paulsen. New York: Planum Press, 1970.

 

Paulsen, C. A. and Gordon, D. L. Klinefelters
Syndrome and its Variants a Hormonal and
Chromosomal Study. Recent Progr. Hormone Research.
(1968), 34, 321.

51.

52.

53.

54.

55.

56.

57.

58.

59.

60.

61.

62.

70

Wiener, S.; Sutherland, G.; Bartholomew, A. A.; and
Hudson, B. XYY Males in a Melbourne Prison.
Lancet. (1968), J, 150.

Hudson, B.; Burger, H. G.; Weiner, S.: Sutherland,
G.; and Bartholomew, A. A. Plasma Testosterone
and Luteinizing Hormone in XYY Men. Lancet.
(1969), J, 699.

de Kretser, D. M. Studies on the Structure and
Function of the Human Testis. M.D. Thesis.
Monash University, 1968.

Coppage, W. S. and Cooner, A. E. Testosterone in
Human Plasma. New Eng. J. Med. (1965), 273, 902.

Rubin, E. and Lieber, C. S. Prolonged Ethanol
Consumption Increases Testosterone Metabolism
in the Liver. Science. (1976), 191, 564.

Medicine Column. "Pot and Sex." Time, September
29, 1975, 54.

Nickell, M. D.; Russel, R. L.; Fahim, Z.; and
Fahim, M. S. Long Term Effect of Vasectomy on
Reproduction and Liver Function. {24. Proc.
(1974), 44, 531.

Smith, K. D.; Tcholakian, R. K.; Chowdhury, M.; and
steinberger, E. An Investigation of Plasma
Hormone Levels Before and After Vasectomy; also
Kobrinsky, N. L.; Winter, J. S.; Keyes, F. 1.; and
Faimen, C. Endocrine Effects of Vasectomy in Men.
Fertility and Sterility. (1976), 27/2, 145 and
152.

Smith, 0. W. and Hafs, H. Competitive Protein
Binding and Radioimmunoassay for Testosterone
After Injection of LH or Prolactin in Rabbits.
Proc. Soc. Explt. Biol. M44. (1973), 444, 804.

 

Wintrobe, M. Clinical Chemistry, Principles and
Procedures. Chapter 23, 1131.

 

Abraham, G. E. Radioimmunoassay of Steroids in
Biological Fluids. Clin. Ciochem. (1974), 7/3,
194.

Kleihauer, B.; Stein, G.; and Schmidt, G. Beitrag
zur alterbestimmung von blutflecken. Arch.
Kriminol. (1967), 140, 84-94.

63.

64.

65.

66.

67.

68.

71

Kind, S. 8.; Patterson, D.; and Owen, G. W. Esti-
mation of the Age of Dried Bloodstains by a
Spectrophotometric Method. Forens. Sci. (1972),
1:, 27-54.

Ito, T. and Horton, R. Dihydrotestosterone in
Human Peripheral Plasma. J. Clin. Endocr.
(1970), §lI 362.

Bauminger, S.; Kohen, F.; Lindner, H. R.; and
Weinstein, A. Antiserum to 5-a-dihydrotestosterone:
Production, Characterization, and use in
Radioimmunoassy. Steroids. (1975), 34, 237.

Janne, 0.; Apter, D.; and Vihko, R. Assay of
Testosterone Progesterone and l7-a-hydroxypro-
gesterone in Human Plasma by RIA after Separation
on Hydroxyalkoxypropyl Sephadex. ‘J. Steroid.
Biochem. (1974), 5, 155.

Dr. Barry Brown, personal communication, February 18,
1976.

Shaler, R.; Mortimer, C. F.; and Stuver, W. C.
Sexing of Bloodstains. Paper presented at meeting
of American Academy of Forensic Sciences,
Washington, D.C., February, 1976.

APPENDIX

The following computer printouts are compilations
of the raw data from the assays outlined in the Procedures
chapter. The computer program is written in BASIC lan-
guage. The original data sheets are indexed and will be
kept in the library at the Scientific Laboratory Section V,
Michigan State Police Headquarters, 714 S. Harrison Ave.,
East Lansing, Michigan. .

72

STAND ALONE STATISTICS PROGRAM

'PUN

cu

READY PUNCH: CR TO START

10
20
25
30
'40
50
60.
65
'70
80
90

PRINT "STATIST ICAL PROGRAM"
PRINT "ENTER DATA AND TERMINATE WITH 0"
LET N80 '
INPUT X

IF X=0 THEN 90

LET S=S+X

LET S2=S£+(X*X)

LET N=N+l

LET M=SIN

GOTO 30

LET V=<N#52-S*S)/N/(N-l)

100 LET D=SQR(V)

110
I20
I30

LET P=.6745*D
LET E=SQR(V/N)
LET CaD/M

135 PRINT
136 PRINT
137 PRINT

140
150
160
170
180
I90
200
210
220
TYP

*

PRINT "NUMBER"1."SIM": "SUM- SQUARES"
PRINT N05152
PRINT
PRINT "MEAN": "VAR IANCE": "STD. DEV."
PRINT M: V: D
PRINT .
PRINT "PROBO ERR-“a"STDo DEV. MNo"o"COEPPo VAR."
PRINT POEJC '
END '
E

73

*LI

10
20
30
31
32
40
50
58
59
60
’61
62
63
64
65
66
70
71
72
73
74
75
76
77
‘100
115
116
120
130
140
150
I60
I70
180
I90
200
210
220
230
240
245
250
260
270

t

74

ST 10:270

DIM X(IOO):Y(2S)
PRINT"ENTER PROG. I"
PRINT"I=STAT."
PRINT"2=LST. SQ."
PRINT"3=GLASS"

INPUT Z

0N Z GO TO 60:70:100011900
PRINT "N0 SUCH OPTION"
GO TO 20

PRINT"ENTER X‘Sa 0=END"
FOR I=l T0100'

INPUT X(I)

IF X(I)=0 THEN 65

NEXT I

GOSUB 100

END

PRINT"HOU MANY PAIRS 7"}

'INPUT NI

PRINT"ENTER XaY"

FOR I=l T0 N1

INPUT XCI):Y(I)
NEXT 1

(30sz 300

END

PRINT

PRINT "STATISTICS PROGRAM"
PRINT

LET 18=l

LET S=52=0

FOR I=l T0 100

IF X(I)=0 THEN 200

LET s=s+x<11

LET S2=52+X(I)*X(I)

LET M=SII

NEXT I

LET 1:1-1

LET V=(I*S2*S*S)/I/(I-l)
LET D=SQR(V) ‘

LET E=SQR(V/I)

LET C=D/M

11“ 18:2 THEN 270
PRINT"N0- "a "SUM"; 'MEAN": "STD. DEV. "
PRINT IJSaMoD

RETURN

'75

STATISTICAL PROGRM
ENTER DATA AND TERMINATE WITH 0
47.974

7.95! -
2.208 WHOLE BLOOD EXP 1

7.223
7.453 FEMALES

7.454 17 _ 60

2.428
?.383 NG/ML
2.382
?.373
?.295
?.210
7.567
?.368
7.558
2.545
?.617
2.956
?.410
?.6I9
?.l97
?.415
2.266
?.291
2.320
2.383
2.235
?.691
2.113
2.242
7.187
2.454
2.393
?.806
?.378
?.361
2.701‘
7.350
?.219
2.345
?.!46
?0208
?.701
2.543
7.155
?O 14'
7.472
?.961
?.417
?.325
?.467
2.194
?0

NUMBEP SUM SUM-SQUARES
53 , 22.336 11.9883

MEAN VAPIANCE STD. DEV.
0.421433 4.95218E-02 0.222535

PPOB. EPP. STD. DEV. MN. COEFF. VAR.
0.150! 3.05675E-02 0.528043

76

52421521041. "0mm
ENTER DATA nun 2zanxn43: VITH o
2.980 ’
28.498 ' ‘ WHOLE BLOOD - EXP 1
28.232

22.193 MALES
21.482

TO'SOQBB _ ' 17 - 60
22.821

24.940 ‘ NG/ML
22.521

22.220

25.201

23.338

22.918

21.998

24.228

21.328

24.081

73.794

25.022

22.455

21.138

22.288

22.204

21.591

22.208

23.449

24.181

71.100

71.269

22.141

21.423

22.023

71.751

21.200

21.529

210.128

211.221

21.008

25.481

76.533

71.463

24.218

24.210

23.028

74.090

21.992

71.918

21.892

74.864

74.719

72.703

21.113

23.851

28.959

7.727

71.184

22.542

77.377

24.893

74.648

24.048

22.915

20

NUMBER SUM SUM-SQUARES
63 236.507 1269.01

MEAN VARIANCE STD. DEV.
3.75408 6.14758 2.47943

PROB. ERR. STD. DEV. MN. COEFF. VAR.
1.67238 0.312379 0.660464

RUN

STATISTICAL PROGRAM

ENTER DATA AND TERMINATE WITH 0

72.494
7.421
73.051
7.456
7.776
?4.951
72.059'
71.031
71.539
7.478
7.446
7.313
71.351
?1.787
7.660
?oll°'506
70

NUMBER
16

MEAN
1.39494

PROB. ERR.
0.849644

*

SUM
22.319

'VARIANCE

1.58675

77

MALES
61 - 79
NG/ML

SUM-SQUARES
54.9349

STD. DEV.
1.25966

STD. DEV. MN. COEFF. VAR.

0.314916

0.903026

WHOLE BLOOD - EXP 1

78

RUN

STATISTICAL PROGRAM _
ENTER DATA AND TERMINATE WITH 0 WHOLE BLOOD EXP 1

7.873
7.1”5 FEMALES

7.183 ' 61 _ 79

2.308
7.569

7.034 NG/ML
2.820

7.691

7.387

7.411

2.440

2.222

2.199

7.162

20

NUMBER SUM SUM-SQUARES
14 ' 4 5.082 ‘ .2.2724

~MEAN VARIANCE ‘ STD. DEV.
1 09361571 3943945’02 00185456

PROB. ERR. STD. DEV. MN. COEFF. VAR.
0.12509 4.95652E-02 0.512917

*

STATISTICAL PROGRAM

79

‘ENTER DATA AND TERMINATE WITH 0

75.266
73.475
72.861
73.843
72.368
74.368
73.312
73.396
73.090
-?3.200
73.874
73.433
73.533
75.213
73.789
72.795
72.776
72.382
73.677
72.045
73.126
73.899
73.377
72.138
72.615
74.039
77.325
75.933
70

NUMBER
28

MEAN

3.61243

PROB.

‘ 0.789659

*RUN

101.148

VARIANCE
1.37061

0.221248

BLOODSTAINS - EXP 2
MALES
l7 - 60
NG/ML

SUM-SQUARES
402.396

STD. DEV.
1.17073

MN. COEFF. VAR.
0.324085

80

STATISTICAL PROGRAM
ENTER DATA AND TERMINATE WITH 0

‘73o475
72.903
.22.923
72.897
72.903
73.032
472-513
72.724
72.634
72.427
72.266
72.408
72.238
72.048
.729065
72.089

71.899 ‘

,22.321
22.848
22.454
23.877
20

NUMBER
22

MEAN

2.57495

PROB.

0.325881

*RUN

BLOODSTAINS - EXP 2
FEMALES
17 - 60
NG/ML

SUM ' SUM-SQUARES
58.849 150.221

VARIANCE STD. DEV.
0.233429 0.483145

STD. DEV. MN. COEFF. VAR.
0.103007 - 0.187632

STATISTICAL PROGRAM ,
ENTER DATA AND TERMINATE WITH 0

-72.713
73.291
749673
739414
73.280
73.235
72.119

.73o641
73.939
73.997

70

NUMBER
10

MEAN

3.4302

PROB. ERR.
0.476044

*RUN

- SUM

34.302

VARIANCE
0.498116

STD. DEV. MN.
0.223185

81

SUM-SQUARES
122.146

STD. DEV.
0.705773

COEFF. VAR.
0.205753

BLOODSTAINS - EXP 2
MALES
61 - 79
NG/ML

 

STATISTICAL PROGRAM
ENTER DATA AND TERMINATE WITH 0

72.465
72.814
72.716
71.835
72.354
72.717
72.115
72.104
71.794
7.923
702.218
70

NUMBER
11

MEAN
2.18682

PROB. ERR.
0.366231

4‘

SUM
24.055

VARIANCE
0.294813

STD. DEV. MN.

0.163711

82

BLOODSTAINS - EXP 2
FEMALES
61 - 79
NG/ML

SUM-SQUARES
55.552

STD. DEv.
0.542982

COEFF. VAR.
0.248291

83

*RUN

STATISTICAL.PROGRAM
ENTER DATA AND TERMINATE wITH o BLOODSTAINS 7 EXP 3

71.861
21.478 . MALES

22.740

23.275 17 ' 60
31.333 NG/ML

21.398

71.782

21.597

21.848

71.437
71.637

_ 71.877

71.468
71.299
71.236
72.133
71.649
71.232
72.806 '
71.798
71.532
7.775
71.301
71.355
71.590
7.949
70

NUMBER SUM ' SUM-SQUARES
27 45.402 - 84.1818

MEAN VARIANCE STD. DEV.
1.68155 0.300613 0.548282

PROB. ERR. STD. DEVH MN. COEFF. VAR.
0.369816 0.105517 0.326056

84

*RUN

STATISTICAL PROGRAM
ENTER DATA AND TERMINATE VITH

7.504
7.75, FEMALES

7.748 17 _ 60

7.436
7.744
7.508.
7.586
7.458
7.639
71.281
7.644
7.550
7.568
7.638
7.563
7.536
7.475
7.741
7.524
7.601
7.443
7.689
7.578
7.439
7.399
7.613
7.634
7.381
7.462
70

0 BLOODSTAINS - EXP 3

NG/ML

NUMBER , SUM SUM-SQUARES
29 17.133 10.9366

MEAN VARIANCE . , STD. DEV.
0.590793 2.90916E-02 0.170562

PROB. ERR. STD. DEV. MN. COEFF. VAR.
0.115044 3.18727E-02 0.288701

85

RUN
BLOODSTAINS - EXP 3

STATISTICAL PROGRAM .
ENTER DATA AND TERMINATE VITR 0 MALES

- 2.539

7.661 ' 51 ‘ 79
71.116
7.762 ' . NG/ML
, 2.888 ' -

"7.724

2.488
2.448
20_

NUMBER SUM SUM-SQUARES
8 ‘5.402 3.98825

MEAN VARIANCE STD. DEV.

.PROB. ERR. ' STD. DEVu MN. 'COEPP. VAR.
‘ 0.143888 7.54219E-02 0.31592

0

86

RUN

STATISTICAL pROGRAM BLOODSTAINS - EXP 3

ENTER DATA AND TERMINATE WITH 0 FEMALES
7.274

70199 ' _
7.162 61 79
7.724
2.428 NG/ML
7.446.

7.686 -

70
NUMBER SUM SIM-SQUARES

7 2.969 1.56309
MEAN VARIANCE STD. DEV.

0.424143 'S.063558-02 0.225023

PROB. ERR. STD. DEV. MN. COEFF. VAR.

0.151778 8.505078-02 0.530536

1.:

11111111171111.11711111711111“

 

214992