THE RELATIONSHIP OF BLOOD
PRESSURE LEVELS YO ‘i'HE CHRGNEC
ENHEEETEQN QF AGGREfiEGN

Thesis far We Degree of MI. D.
MICHIGAN STATE UNIVERSITY

Robert George Meyer
1967

Thaw:

0-169.

This is to certify that the
thesis entitled

THE RELATIONSHIP OF BLOOD PRESSURE
LEVELS TO THE CHRONIC INHIBITION
OF AGGRESSION

presented by

Robert George Meyer

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in Clinical'Psychology

.

Major professor

 

 

 

ABSTRACT

THE RELATIONSHIP OF BLOOD PRESSURE
LEVELS TO THE CHRONIC INHIBITION
OF AGGRESSION

by Robert George Meyer

Several investigators have suggested that the inhibition
of aggression may be related to chronic high blood pressure. Stud-
ies of the etiology of essential hypertension also lend credibility
to such a hypothesis. Recent investigations support this hypoth—
esis in regard to short-term effects, and have in addition indicated
that guilt over the expression of aggression is a relevant personality
factor related to high blood pressure.

Anokhin asserts that the depressor impulses that act
to lower blood pressure can only fire consecutively for approxi-
mately 25 minutes, and if felt aggression is suppressed beyond
that time, the vascular system slowly adjusts itself to a higher
level of blood pressure. The cortex, as well as the hypothalamus
and reticular formation, play important parts in his theory, and
fit with hypotheses using higher-level constructs such as "guilt"

and the "suppression of aggression. "

Robert George Meyer

This study is naturalistic-statistical in approach in
order to most effectively study the chronic effect of the non-
expression of aggression in human subjects. As mentioned,
while other studies have shown short-term effects, chronic
effects have yet to be clearly demonstrated. For this reason,
observations were made on inmates of the State Prison of South-
ern Michigan who had served at least two years time previous to
the initiation of this study. Reliable records on relevant vari-
ables were available.

Inmates aged 23 to 35 were included only if the offense
resulting in the most recent incarceration had involved direct
physical aggression. These inmates were categorized into low
and high aggression groups, depending upon their behavior pattern
during their institutionalization. They were also matched on the
relevant factors of age, weight, race, and prior medical history.

Sixty-eight subjects (thirty-four in each group) were
examined. Ten minutes after a preliminary blood pressure was
taken, two blood pressure measures were taken fifteen seconds
apart, and the Mosher Incomplete Sentences Test was then admis-
tered to assess the amount of guilt over the eXpression of aggres-
sion. The Cattell 16 PF was also administered in order to obtain

both anxiety and intelligence scores.

Robert George Meyer

An average of the last two blood pressures taken on
each subject was used, in order to increase the validity of the
measure. The MIST responses were rated for aggression-guilt.
The reliability of such rating has been previously established.
Anxiety and intelligence scores were assessed according to the
instructions in the Cattell 16 PF Manual.

The primary hypothesis, that subjects who for the
18 months prior to the study expressed little aggression would
have significantly higher diastolic pressures than subjects who
continued to aggress substantially, was supported at the . 01 level
of confidence. The second main hypothesis, that within the high
aggression group, those who manifested high guilt over the expres-
sion of aggression would have significantly higher diastolic pres-
sures than those who showed low guilt, was significant at the . 025
confidence level.

A similar though not significant trend was found when
systolic measures were employed.

The two groups were not found to be significantly dif-
ferentiated by any Cattell scale, including anxiety or intelligence
scores. Though Negro subjects had a slightly higher mean dias-
tolic blood pressure than whites, this difference was not signif—
icant.

The results led to the following conclusions:

Robert George Meyer

A continued state of the non-expression of aggression,
in those men who have at one time eXpressed direct
physical aggression, results in higher levels of dias-
tolic blood pressure than in those men who continue
to aggress against the environment.

Within that group of men who continue to aggress
against the environment, those who have guilt over
such expression develop higher blood pressure than
those who do not have such guilt.

Self-reported anxiety, or other Cattell personality
dimensions, are not related to the raising or lower-
ing of diastolic blood pressure levels, or to the ex-
pression of aggression.

THE RELATIONSHIP OF BLOOD PRESSURE LEVELS

TO THE CHRONIC INHIBITION OF AGGRESSION

By

Robert George Meyer

A THESIS

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

DOCTOR OF PHILOSOPHY

Department of Psychology

1967

ACKNOWLEDGMENTS

I wish first of all to sincerely thank Dr. Albert 1.
Rabin, my dissertation chairman, for his advice, support, and
guidance in this endeavor. His criticisms were always both fair
and meaningful, and his interest has always been consistent and
very supportive.

My thanks are also extended to Dr. Bill Kell for his
criticism and support in this work, and for his help in the latter
years of my graduate studies. A special note of thanks is extended
to Dr. Joseph Reyher for his assistance in this project and while
serving as my advisor over the years, and to Dr. Bertram P.
Karon, who has been an unfailing source of constructive advice
and vital encouragement throughout all my years of graduate study.

My appreciation is very strong toward those inmates
who gave of their time to help in this study. Most especially, I
wish to thank Mr. Robert Hearn, who ran the subjects and as-
sisted in the statistical computation, and Mr. Harvey Webb, who

typed the drafts and revisions.

ii

TABLE OF CONTENTS

ACKNOWLEDGMENTS
LIST OF TABLES
LIST OF DIAGRAMS
LIST OF APPENDICES
Chapter

I. INTRODUCTION

Psychosomatic Background
Physiological Considerations
Anatomical Considerations
Prominent Theories

Psychological Factors and Hypertension

Theoretical Orientation .
Aggression .
Factors in Essential Hypertension

II. PROBLEM .
Hypotheses

III. METHOD .
Subjects
The Sample .
Measurements .

Procedure . . .
Treatment of the Data

iii

Page

ii

vi

vii

H

«loci-t

15
21
28

33
35
37
37
4O
46

50
54

Chapter
IV. RESULTS

Primary Hypotheses

Additional Findings .

V. D ISCU SSION

VI. SUMMARY .

REFERENCES

APPENDIX .

iv

Page
56

56
59

65
73
76

82

n

 

LIST OF TAB LES

Table ‘ Page

1. Basic comparisons of High and Low A

groups on mean age, intelligence,

andrace...................42
2. Distribution of High and Low A groups

on Cattell intelligence scores . . . . . . . . . 43
3. The number of Ss incarcerated for a

particular crime in both the Low

and High Aggression groups . . . . . . . . . . 44
4. Systolic and diastolic pressures for 10 83,

to establish E' s reliability in taking

blood pressures . . . . . . . . . . . . . . . 53
5. Mean blood pressures for the High and

LowAgroups . . . ............. 55
6. Mean blood pressures for the high and

low guilt groups within the High

Aggression group. . . . . . . . . . . . . . . 56
7. Mean anxiety scores for the High and

Low Aggression groups . . . . . . . . . . . . 57
8. Cattell factor descriptions, Low and

High A group and overall group mean
sten scores, and the percentage of
general population adults attaining

each score. . . . . . . . . . . . . . . . . . 61
9. Correlations of systolic blood pressure

indices with Cattell factor scores,

using the overall sample . . . . . . . . . . . 62

LIST OF DIAGRAMS

Diagram Page

1. Schema of cardiovascular regulation. . . . . . . . 9

vi

LIST OF APPENDIC ES

Appendix Page
_ A. The 25 MIST items employed in this

study, taken from the original 50
items employed by Mosher. . . . . . . . . . .81

vii

CHAPTER I

INTRODUCTION

Psychosomatic Background

 

This paper is concerned with the possible effects that
psychological factors, such as the chronic inhibition of aggression
and guilt over the expression of aggression, exert on blood pressure
levels. Studies carried out on the psychosomatic disorder desig-
nated "essential hypertension" will also be examined since chronic
high blood pressure is the primary sympton of this disorder.

The use of the term "psychosomatic" has increased
rapidly throughout the last several decades. It is not only a core
concept of current psychological research, but has infiltrated
with varying degrees of accuracy. into contemporary intellectual
discussions. A proliferation of research has paralleled this grow-
ing familiarity with the concept.

Two converging factors have paved the way for this
quick rise. The acceptance of the psychosomatic approach was
in part a consequence of World War II, a factor of generative

import for many areas of psychological research. During World

War II, the psychosomatic effects of high and prolonged anxiety
were clearly apparent to many men under stress, as well as to
those attending them (Groen, 1964).

The second and more important factor was an incre-
ment in sophisticated research literature. The landmark work
was that of W. B. Cannon (1932), which provided a theoretical
basis for much of the research in psychosomatics.

A basic assumption as to what a psychosomatic dis-
order consists of has gradually evolved. The consensus is that
the term "psychosomatic" should be limited to those disorders
in organs or systems controlled by the autonomic nervous system
in which chronic emotional maladjustment is the primary factor
and somatic dysfunction is the result or by-product. The term
"somato-psychic" is then employed for those disorders in which
the bodily change is primary to the psychological one, e. g. , as
in tuberculosis.

Groen (1964) is most emphatic in distinguishing the
two categories; however, there is little consensus as to which
disorders fall into which category. For instance, while Groen
sees the somatic dysfunction as primary in rheumatoid arthritis
and asthma, many people would postulate the psychological aspects
as more important in these disorders. There are important and

relevant anatomical and physiological considerations that should

be mentioned concerning psychosomatics in general and, as such,

are pertinent to the particular focus of this paper.

Physiological Considerations

 

The general concept of "psychosomatic" refers to somatic
disorder in organs or systems controlled by the autonomic nervous
system, and thence subordinated to the higher brain areas, e. g. ,
the cerebral cortex. The autonomic nervous system enervates
all the nonstriated muscle and the glands, which are not under the
voluntary control of the cerebral cortex. It is usually referred
to as the involuntary or vegetative nervous system, and is func-
tionally defined as that system of motor nerves which oversees
the interacting internal affairs of the organism (White and Smith—
wick, 1941). These authors state that it was Claude Bernard, the
great French physiologist, who first pointed out that it is the
mammals' constant preservation of a "milieu interieur" which
is a determining factor in a free and an independent life from the
general environment. The autonomic nervous system is the basic
functional factor providing this independence.

The autonomic system is basically divided into the
parasympathetic and sympathetic sub-systems, a division based
on anatomical and physiological factors. There is no fundamental

difference between the sympathetic and parasympathetic systems

in terms of either neuron and ganglionic structure. The important
anatomical distinction is the location of the ganglia, the ganglia of
the sympathetic division lying in closer proximity to the Spinal
cord.

Physiologically, the sympathetic system acts in a more
general and diffuse manner than the parasympathetic division.
Gellhorn and Loofbourrow (1963) comment that partial sympathetic
discharges can occur, although the basic fact remains that irradia-
tion to a number of effector organs from one stimulation is much
more common in the sympathetic system than in the parasympa—
thetic system. Parasympathetic discharges are physiologically
Specific.

The sympathetic system is functionally concerned with
mobilizing the resources of the body for use in work or emergen-
cies. It is an emergency protective mechanism, which, while
often dormant, _is always ready to take action to remedy various
circumstances such as hemorrhage, as well as intense emotional
or psychic trauma. An important point in this regard is Cannon' 3
(1932) discovery that once the sympathetic system has been aroused,
it is very important that some form of activity be carried out, to
complete the system's cycle once it has been begun. Therefore,
one must either permit excitement to find an appropriate physical

mode of expression in order to obviate the harmful effects of

continued arousal. Cannon has emphasized that if the emotion is
once set in motion, the best thing that can be done is to engage
in strenuous physical exercise if it is not to be allowed expression.

The function of the parasympathetic division is the
conservation of bodily resources. It dominates the bodily functions
during the time when the organism is recuperating. For example,
it slows the heart rate, thereby allowing rest for cardiac muscle.

These are not truly separate divisions of the autonomic
nervous system; rather they both act in terms of a balance effect.
The organism is operating optimally only when these two divisions
interact in tune with each other. For example, if the organism
is to react effectively after a strong emotion has activated the
sympathetic system, a rise in parasympathetic activity must soon
follow the sympathetic discharge.

This balanced opposition and dual enervation under-
lies Cannon' 8 (1933) very important concept of "homeostasis. "
Homeostasis is the designation for the coordination of the body
by means of autonomic adjustments through which changing con-
ditions in the environment, both external and internal, are dealt
with. It frees the individual from having to pay detailed attention
to the very routine and minute aspects of everyday life.

Psychosomatic abnormality is then a breakdown of the

normally efficient homeostatic mechanism. This breakdown can

in some instances be termed "normal. " Chronically cold feet

or moist hands are examples of a breakdown of homeostatic con-
trol in the normal person. In psychosomatic conditions, the dis-
ruption of the homeostatic balance occurs in a more abrupt or
definite manner than in the examples cited above.

The cerebral cortex is the agent providing the chronic
inhibitory effect that interrupts this normally efficient autonomic
cycle. White and Smithwick (1941) refer to several studies and
conclude from them that chronic vasoconstriction is often produced
through the interruption of the autonomic cycle by inhibitory im-
pulses from the cortex to those vasoconstrictor centers in the
hypothalamus and the medulla.

Likewise, a local stimulation of the paraventricular
nuculei in the walls of the third ventricle of the hypothalamus
results in a widespread discharge of sympathetic activity with
such characteristic responses as a rise in blood pressure, car-
diac acceleration, erection of hairs, etc. The hypothalamus
connects with cortical areas, as well as with the pituitary gland,
mainly the posterior lobe, and with the reticular activating sys-
tem.

The hypothalamus is central, anatomically, in this
conceptualization. However, the reticular activating system is

of primary functional importance since, as Gellhorn and Loofbourrow

(1963) have commented, it is primary to the continued activation
of a system and this continued activation is precisely that which
eventually upsets the autonomic balance, leading to a ”higher level

of balance. "

The hypothalamus provides the anatomical mediation,
the reticular activating system (RAS) provides for the continued

arousal, and the cerebral cortex is the agent for disruption of

the cycle.

Anatomical Considerations

 

There are several psychosomatic-anatomical aSpects
relevant to blood pressure hemodynamics. Brod (1961) points
to a prime hemodynamic difficulty in conceptions of essential
hypertension, as well as in other cardiovascular stress reactions.
An increase in cardiac output and a low total peripheral vascular
resistance are found in acute emotional stress, whereas in high
blood pressure a virtually normal cardiac output with increased
total peripheral vascular resistance is typical.

Brod found the hemodynamic response to emotional
stress in all subjects was similar to the hemodynamic reSponse to
severe muscular exercise, and stated there is a similar signifi-
cance for both, i. e. , preparing the organism for "fight or flight. "
He concludes that the normal cardiac output and increased periph-

eral vascular resistance in essential hypertension are a result

of chronic stress, representing a state of fixed preparedness for
a severe muscular process that never takes place.

The internal radius of the blood vessel, as well as
the vessel wall thickness, should not be conceived of as important
only when "abnormally defective, " but rather normal, genetically-
inherited aSpects that do influence blood pressure. However,
these factors are neither primary nor effective, as Peterson
(1961) states, unless there is a concurrent and primary influence
of emotional factors. The vessel wall itself naturally resists being
stretched, although it is in a state of at least minimal stretch at
all times. The stretching acts on the baroreceptors, the nerve
ends which are embedded in the vessel wall, as e. g. , the carotid
sinus. Stretching the wall in turn stretches the carotid sinus
which relays impulses through the autonomic system and initiates
a parasympathetic discharge which in turn acts on heart rate. In
In this sense, the carotid sinus acts as a strain receptor as opposed
to the common conception of it as a "pressure" receptor. It becomes
the effector organ within the blood vessel itself. The general me-
chanics of this cardiovascular regulation process are diagramed
on the following page. (See Diagram 1. )

As the strain in the vessel wall increases, the nerve
impulse traffic increases, and as the impulses from the various

receptors increase, the parasympathetic activity affecting the

heart and the blood vessels likewise increases. The center for
the integration is hypothesized as being the hypothalamus (Gell-

horn and Loofbourrow, 1963).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

> / A
P L \ > C. N S
/ \ __
1'
E R _____._.__ _ __ _ N
P P e
Heart
Blood
Pressure IL "Systemic" vessels K———
Diagram 1. -- Schema of cardiovascular regulation.

The circle represents a blood vessel whose wall contains receptors
such that the vessel wall varies with intravascular pressure (P).
Afferent nervous activity (N ) from these receptors is transmitted
to the central nervous system (CNS). Efferent nervous activity is
labeled N . Dotted line represents efferent nervous activity which
effects changes in the vessel walls. From L. Peterson (1961),

p. 306.

Prominent Theories

There are two general camps of theorists on the genesis
of chronic high blood pressure. The first hypothesizes physiolog-
ical factors as prominent, with psychological factors occurring
in an epiphenomenal sense. The other is psychological-emotional
in emphasis, high blood pressure being primarily determined by
the psychological factor. The cause in the latter approach is typ-

ically considered to be emotional stress per se.

10

Cottrell (1961) points out that conceptualizations of
essential hypertension as deriving from physiological factors have
fallen by the wayside much as the peeling of leaves off of an arti-
choke. Cottrell sees the first leaf to be pulled off as that of uni-
lateral renal disease, as promulgated by Harry Goldblatt. The
other leaves would be those of primary aldosteronism, 1 aortic
coarctation, 2 and chronic pyelonephritis, 3 leaving the conception
of essential hypertension not as an entity, but rather as a hypo-
thetical construct.

The one physiological theory which has always received
attention, and which still does, contrary to Cottrell, is the renal
or reno-humoral theory which has consistently been linked with
the name of Harry Goldblatt. All theorists would agree that some
cases of hypertension are secondary effects of renal failure. The
distinguishing mark of Goldblatt' s theory is his assertion that all
cases of hypertension are primarily derived from renal factors.
As Stamler in his discussion of Selye and Bayusz' s (1961) paper
has asserted, several points must be established before the renal
theory can be accepted. First, the problem of the etiology and

pathogenesis of nephro-sclerosis would need to be solved, and

 

1Chronic hypertension of the adrenal cortex.
2Constriction of the aorta.

3Chronic inflammation of the kidney.

11

this is a virtually unexplored problem. Secondarily, it must be
proven in humans that the renal vascular disease process is prior to
the hypertension. Anokhin (1961) has stated that in the majority of
cases it is clearly the other way around.

The term hypertension therefore covers two separate
entities. The first is renal hypertension, where the hypertension is a
secondary outcome of a renal arterio-occlusive disease. In the dis-
cussion of Pickering' s (1961) paper, Corcoran has estimated that the
frequency of renal arterio-occlusive disease in an unselected hyper-
tensive population is somewhere along the order of one percent. While
certain authors claim this figure should be something like 25 percent,
he notes that these samples are usually composed of referrals on that
small portion of symptomatic hypertensives who came for treatment,
were selected for aortograms, and who showed positive findings.

The hemodynamic reSponse to a Specific stress includes
a blood pressure rise via a strong rise in the total peripheral vas-
cular resistance in the kidneys and skin. This explains the occur-
rence of renal failure in longstanding cases of hypertension as a
secondary phenomenon. It is the result rather than the cause.

Psychological Factors and
Hypertension

 

 

There have been various approaches throughout the

years which have linked high blood pressure to personality types or

12

differing modes of emotional expression. The first published
research on psychological factors and blood pressure in American
psychology has been credited by McGinn, fl. (1964) to W. M.
Marston (1917). Marston, a Massachusetts lawyer rather than a
psychologist, was interested in the physiological changes accompany-
ing the process of deception. After he gave the subjects one or
two facts, they were asked to make up lies surrounding these facts.
These "lies" were to be mixed in with "truths. " Systolic blood
pressure measures were taken and were correlated with the sub—
ject' 5 report of whether he was telling the truth or attempting to
deceive the experimenter at that particular time. Marston found
that it did not matter what the actual facts were, as long as the
subjects thought they were deceiving him, their blood pressures
rose. When they thought they were telling the truth, their blood
pressures decreased.

Several psychoanalysts have correlated interviews
involving stressful or conflict-laden material with blood pressure
readings. The most notable here is that by Saul (Alexander and
French, 1948), who reported observations of blood pressures before
and after psychoanalytic sessions. He found that whenever unper-
ceived resentments were coming to the fore, though not yet in
consciousness, the blood pressure levels would fluctuate up to

30 or 40 points. This rise would persist as long as there was

13

any material associated with these unverbalized conflicts. It was
only after a time in which the person could discuss freely these

unsatisfied resentments that the blood pressure would subside.

Psychological test data

 

Other psychological indices such as objective and pro-
jective tests have been used to study the correlates of high blood
pressure. Cattell and Scheier (1959) attempted to relate various
Cattell 16 PF factors with changes in systolic blood pressure.

They took for their sample 18 University of Illinois undergraduate
males, the top ten percent from the 180 who originally took an IPAT
Anxiety Scale. They tested this obtained sample on 103 variables,
including 33 questionnaire scales. Those who had high systolic
blood pressure along with the high anxiety were found on the 16 PF
to be sociable, group dependent, sophisticated, practical or conven-
tional, and submissive interpersonally.

Brower (1947) studied the relationship of blood pressure
to performance on the Rorschach test. He individually administered
the Rorschach to 38 undergraduates, as well as the Mirror Drawing
Test and an MMPI. He also measured their pulse rate and the
diastolic blood pressure, asserting that the systolic blood pressure
is a composite score and therefore inaccurate. He found that those

with a higher diastolic blood pressure had a lower FC score on the

l4

Rorschach, which he interprets as the degree of emotional adjust-
ment to outer reality. The results are "suggestive, " to use his own
term. Other researchers (Matarazzo, 1954; Neiberg, 1957) have
employed the TAT, the Rosenzweig Picture-Frustration Study, etc. ,
with little or no significant results in clarifying the variables in
chronic high blood pressure, at least as regards the psychological
test results.

Hamilton (1942) administered self-rating scales and
peer-rating scales, as well as a Rorschach and a Murray Picture
Test, to 130 subjects. They correlated these scales and tests with
extensive physiological measurements and life history data that
were gathered as a matter of routine procedure for admission to
the educational institution involved. The study employs fairly
rigorous controls, but they used only the systolic pressure for
comparison to other measures. Their reasoning for not employing
the more relevant diastolic pressure, i. e. , statistical ease, seems
very inadequate. They did not find those with high systolic pres—
sure to be neurotic, unstable, or physically hyperactive; although
they stated these high systolic pressure subjects were more easily
aroused to anger. Yet, on all rating scales they came out Signifi-
cantly more submissive than normals in Spite of their low threshold

for anger arousal.

15

There are, of course, studies (Oken, 1960; Storment,
1951; Innes, Millar, and Valentine, 1959) which have found no dif—
ferences in scores on personality tests or reactions to stress inter-
views between high and low blood pressure samples. AS ‘McGinn,
3133.1. (1964) conclude, these studies do not refute the hypothesis
of consistent personality correlates to blood pressure changes,
as there is much reason to question the theoretical methodology

and/ or statistical approaches employed.

Theoretical Orientation

 

The quantity of stress theory

 

This section will attempt to integrate the various find—
ings and principles discussed in the previous sections of this paper,
in order to present a meaningful theory of the development of chronic
high blood pressure. A prior section of this paper verified the
feasibility of a theory based on emotional factors. However, this
trend of thought needs further refinement. Is it the quantity of
emotion that leads to high blood pressure, a particular emotion, or
a mode of reaction to emotional arousal?

A prime psychosomatic theory of essential hypertension
holds that it is the quantity of stress per se that is causal (Brod,
1961). Most support for this view has come from animal experi-

ments. Selye and Bayusz (1961) have experimentally developed

16

hypertension in rats by exposing them to loud noises at frequent
intervals over a long period of time. John Calhoun (1962), in a
rigorously-defined procedure, found that crowded living conditions
as a stress situation produced cardiovascular pathology in rats.
These studies do not definitely support a "hyperten-
sion results from stress per se" hypothesis. One still wonders
whether it is the stress or the emotional reaction induced by such
a stress which is the causative factor. Results from other studies
(Shapiro and Horne, 1955) would, in fact, argue against the idea
that stress in a quantitative sense directly correlates with the de-

velopment of high blood pressure.

Specificity

 

Stress as a quantity is not the important factor; rather
it is the organism' S handling of the stress. The next question:
which emotion is causal? This implies a prior logical point, the

problem of "specificity. "

This refers to the question whether emo-
tional states can, in fact consistently be differentiated, i. e. , along
physiological lines, or whether particular persons react similarly
regardless of the particular eliciting stimulus. In regard to this
paper, are circulatory response patterns specific to the particular

individual in question regardless of the emotion elicited, or is the

circulatory pattern more related to the particular emotional state?

17

As Gambaro (1966) implies, the former notion follows
Cannon' s idea that one cannot discriminate the various emotion
states in terms of the physiological changes. For example, these
physiological aspects'would be the same for both fear and anger

under this hypothesis. Lacey (1952) and Lacey, Bateman and Van
Lehn (1953) are modern proponents of this view that response pat-
terns are differentiated in terms of the particular responding indi-
vidual rather than the eliciting stimuli.

McGinn, _e_t_al. (1964), Buss (1961), and Gambaro (1966)
all cite evidence from which they conclude there is significantly
stronger support for the hypothesis that the circulatory response
patterns are directly and consistently related to the particular
emotional state which is elicited. The two studies which offer the
most definitive support in this regard are those by Ax (1953) and
Schachter (1957).

Granted that specific emotional reaction patterns can
be differentiated, which particular emotion is the factor in high
blood pressure? Aside from anger, fear is the only emotion to
have either face validity or theoretical backing as the Specific
underlying emotion. Ax' s (1953) and Schachter' s (1957) results
on diastolic blood pressure contraindicate fear and support anger

as the underlying emotion. Evidence from such diverse sources as

18

Saul (1948) and Groen (1964) lend support to an anger or aggression-
based hypothesis.

Directly relevant is a series of studies carried out by
Hokanson and his colleagues (Hokanson, 1961; Hokanson and Burgess,
1962 [a] ; Hokanson and Burgess, 1962 [b]; Hokanson, Burgess and
Cohen, 1963), focusing on the expression of aggression. In this
series of studies, 80 students were typically involved in various
factorial designs. Several variables were manipulated, and the
variations in frustration involved the type of harassment employed
while subjects attempted to count backwards from 100 to 0. These
experimenters repeatedly found the frustrating condition to cause
a Significant rise in systolic blood pressure. An important finding
was that aggression against the frustrator himself significantly
lowered blood pressure, whereas aggression against a "substitute"
was not significantly effective, and blood pressure remained some-
what elevated.

There are two main weaknesses with these studies.
The first, which should have been eliminated, was the exclusive
use of the systolic pressure measure, as opposed to the more
relevant and reliable diastolic measure. The other problem vari-
able was the use of such an elicited, explicitly "socially-sanctioned"
aggression Situation, where the subject was given the message

that he was helping the experimenter (E) study the effect of pain

19

on guessing behavior by shocking him when E was wrong in guess-
ing a number. This unique situation for the expression of aggres-
sion obviously limits the generalization of results.

The central factor in this inhibition of aggression was
hypothesized by Gambaro (1966) to be guilt over the expression of
aggression. As mentioned previously, Gambaro (1966), using the
Mosher Incomplete Sentences Test (MIST) (Mosher, 1961), found
that both high and low guilt groups showed Similar blood pressure
increases in an anger-inducing situation. Also, only those Ss with
low guilt showed a significant decrease in diastolic blood pressure
when allowed to aggress. Since aggression will be the main inde-
pendent variable, a later section of this paper will examine it in
more detail.

The following will present physiological theory and
data necessary to support an aggression-inhibition hypothesis.

A rise in blood pressure is quickly communicated in the form of
nervous information to the central nervous system from the sensi-
tive baroreceptors in the aorta and the carotid sinus. Anokhin (1961)
has demonstrated that a rise in afferent signals from the barorecep—
tors in the aortic arch occurs immediately upon a rise in blood pres-
sure in the aorta, and that these impulses can increase by several
orders in magnitude. These impulses then converge on the vaso-

motor center where the tone of the vascular system is determined.

20

Anokhin has also Shown that if systolic blood pressure
rises above 170 mm. Hg. from some form of stimulation which is
then immediately removed, thus allowing the pressure to fall to
something like 140 mm. Hg. , there is complete inactivity of the
baroreceptors, even though this level would have previously pro-
duced a heavy train of impulses. This explains how the barore-
ceptors adapt to long term rises in blood pressure. There is an
entire range of threshold sensitivities to pressure over the indi-
vidual baroreceptors which at the same time adapt organismically
as a whole range to a chronically-held blood pressure. In sustained
arousal, this successive and continuous rise of afferent signals
results in the organism being unable to maintain a physiological
force field stabilized at the lower level over a period of time. Like
other "plastic constancy" systems, such as body temperature,
there is an upward resetting of the reactivity threshold.

Anokhin (1961) asserts that in order to understand the
pathogenesis of sustained high blood pressure, one must hypothe-
size nervous impulses from the cerebral cortex, which fits well
with an emotional origin theory of chronic high blood pressure. He
cites work by several Russian physiologists to support his assertion.

The sustained arousal is then mediated by the reticular
formation. Much evidence has linked the reticular activating system

I

(HAS) to subserving the attributes of "consciousness,' or, more

21

Specifically, of organismic arousal (Magoun, 1950). This ”arousal"
reaction has been related to the ascending fibers of the RAS leading
to the cerebral cortex, and several sensory paths also connect the
RAS with the hypothalamus.

Direct stimulation of the RAS produces EEG changes
seemingly identical with activating changes observed in EEG' s
that were taken upon awakening from sleep, and damage to the
reticular formation produces a sleeplike state. This makes all
the more understandable the ability of chlorpromazine to reduce
chronic high blood pressure, since this substance blocks the ad-
renergic substrate of the RAS, and in this way counteracts the
excitatory effects of the cerebral cortex.

Anokhin points out that this ”cortico-reticular” system
keeps the organism in a state of arousal well beyond the twenty or
thirty minutes of time during which the depressor impulses will
fire in order to lower the blood pressure. After this timethese
impulses lose their continuous character on which the inhibitory
effect of the vasomotor center depends, and an upward resetting of
the stimulus threshold of the baroreceptors occurs, the stage where

chronicity of high blood pressure becomes a physiological fact.

Aggression

 

This paper so far has focused on what will be the depen-

dent variable in this study, 1. e. , the quantitative measure of blood

22

pressure. This section will examine the main independent variable,
the amount of aggression manifested by each subject. One of the
most confusing things in any study which either directly or indirectly

involves aggression is the conceptual murkiness associated with

H H l l

the terms "hostility, anger, ' and "aggression. ' More often than
not, these concepts are used interchangeably to designate the same
phenomenon, or one concept is used to designate qualitatively dif-
ferent phenomena. Kaufmann (1965), as well as others (Buss, 1961),
have tried to clarify the usage of these concepts. Extrapolating
from these articles, one comes to the conclusion that these are
really three separate and distinct constructs.

Hostility, probably the most broadly and often inappro-
priately used of the three terms, is a motivational construct. It
typically is measured through inference or through the introspective
report of a reliable subject. Buss (1961) has insisted that one must
clearly distinguish hostility, the motivational aspect, from two other
factors which are relatively equivalent to hostility. The first is a
tendency to view the environment as inirnical, a perceptual variable.
The second is aggression. A logical conclusion following from the
conceptualization of hostility as a motivational construct is that it

is more specific and particular than either aggression or anger.

It is always transitive to a particular stimulus object.

23

Anger, while involving psychological aspects, is typi-
cally differentiated from hostility and aggression via physiological—
anatomical factors. It is a pattern of autonomic arousal serving a
preparatory function for subsequent aggression. It does not neces-
sitate aggression, and in fact the facial and postural changes that
are components of anger may likewise be inhibited. The facial and
postural elements of an angerresponse can, of course, be observed;
but the label "anger" is more reliably applied only after some fairly
extensive physiological measures are taken. The general consensus
of what an anger state involves physiologically has been adequately
described by both Buss (1961) and Kaufmann (1965).

' is a behavioral one.

The third construct, "aggression, '
It is measured by observing the acting-out of the organism on the
environment. Some theorists do attempt to define aggression in
strictly behavioristic terms, as e. g. , the number of organism-to-
organism contacts (Kaufmann, 1965). However, Feshbach (1964)
has insisted on the important point that one must include motiva-
tional factors in a conceptualization of aggression. Feshbach' S
assertion is not surprising since most of his work has been done on
the motivational aspects of the phenomenon via the TAT.

By following a strictly behavioristic hypothesis, one

begins to define aggression in terms of the other organism' S

reaction, or in terms of the strength of contact made by one

24

organism with another, leading to logical absurdities (Kaufmann,
1965). Behavioral acts per se cannot consistently be labeled ag-
gression, unless inferences about motivation or theindividual' s
expectation regarding outcome are made. As the independent vari-
able in this paper, aggression will be used to denote behaviors
(Kaufmann, 1965) which are transitive, i. e. , are directed against
an object. It will denote behaviors where the attacker forcibly
removes a person from his goal path and/ or imparts obnoxious
stimuli to a person. This includes responses where there is a
high probability of either of these conditions being achieved were
it not for the action being blocked or redirected. This categoriza-
tion of responses as aggressive is then not defined strictly as be-
havioral, but rather as a hypothetical construct.

What is the origin of aggression? One theoretical
View invokes genetic or hormonal factors to explain aggressive
behavior. Beeman (1947) attributed increased aggressive behavior
in mice to the administration of male hormones. Konrad Lorenz
(1966) is undoubtedly the most influential proponent of the aggression-
instinct hypothesis, though his definition of aggression is behavioral
and parallels the definition used in this study, in that he considers
it as behavior calculated to inflict pain on another or the alteration

of conditions in favor of the aggressor. His observations in this

25

area have been for the most part made on ducks, from which he
makes a theoretical leap to parallel explanations in man.

Lorenz' data and his conclusions have both been severely
- taken to task by Ashley Montagu (1966)., the noted anthropologist.

He argues that Lorenz in his comment that aggression in man has
been "inherited from his anthropoid ancestors" is simply inaccurate.
Other such inaccuracies in Lorenz' supporting data are cited by
Montagu, but his most vehement critique is reserved for Lorenz'
process of inference. He asserts that Lorenz makes a totally
unjustified and unsupported leap from a minimal amount of data sup-
porting an aggression—instinct hypothesis in animals to stating that
human aggression behavior basically derives from an instinct. In

a Similar vein, Clark and Birch (1945), as well as Kaufmann (1965),
conclude that physiological explanations are of little or no value in
accounting for even moderately complex behavior, let alone aggres-
sive behavior. They point out that while hormonal factors may be
of some importance in rats and mice, similar hypotheses relevant
for animals above this level have not been substantiated.

The other main proposal to explain the origin of aggres-
sive behavior is an environmental one, the frustration-aggression
hypothesis (Dollard, Doob, Miller, Mowrer, and Sears, 1939).
These authors defined frustration as ”that condition which exists

when a goal response suffers interference (p. 11), " and aggression

26

as "an act whose goal response is injury to an organism (p. 11). "
The authors hypothesized a one-to-one relationship between frus—
tration and aggression, although this aggression is not always
observable, a point often overlooked. There are two basic prob-
lems with the original formulation of the frustration-aggression
hypothesis.

The first is that while this hypothesis, as originally
stated, does contain a "catharsis hypothesis, " the latter does not
deal substantially with the issue of autonomic arousal. This has
resulted in a theoretical short-circuiting of the actual sequence
involved. By focusing wholly on the stimulus and reSponse vari-
ables, which are indeed important, the organism variables involved
have been disregarded. Both the autonomic arousal variables and
hostility variables have been largely ignored, the basis for Kaufmann' s
criticism of the frustration-aggression hypothesis as being circular
in definition. The hypothesis, in orderto be more accurate, should
read the frustration-anger-hostility-aggression hypothesis. This
refers to hostile aggression, as opposed to either instrumental or
accidental aggression (Feshbach, 1964).

This sequence does not always occur in actuality, and
this makes the hypothesis of sustained high blood pressure as a
result of the non—expression of aggression an understandable one.

When this sequence is frustrated, a higher level of felt or

27

psychsomatically manifested anxiety can also be expected (Fesh-
bach, 1964; Grown, 1964).

Strictly behavioral interpretations of the frustration-
aggression hypothesis have contributed to the lack of knowledge as
to which personality factors lead to an interruption or inhibition
of the full sequence noted previously. Theorists have hypothesized
fear of expressing aggression as the underlying variable in high
blood pressure, yet studies involving the experimental arousal of
the organism have typically not attempted to be more specific as
to what is inhibiting the sequence. Sal Gambaro (1966), in a doctoral
dissertation carried out at Michigan State University, has investi-
gated the variables of guilt over the expression of hostility as a
major determinant in the inhibition of the expression of aggression
subsequent to anger arousal.

Gambaro used the Mosher Incomplete Sentences Test
(MIST) (Mosher, 1961) as a measure of guilt over the expression
of aggression in his "anger manipulation" experiment. He found
that both "low guilt" subjects and "high guilt" subjects Showed ap-
proximately the same blood pressure increases in the anger manip-
ulation phase. He also found that subjects who were allowed to
aggress physically against their frustrator did not ShOW a signifi—

cant decrease in diastolic pressure. Obviously, this

28

variable of "aggression—guilt" should at least be considered in any
subsequent work on blood pressure levels.

To summarize, the sequence of frustration-anger-
hostility-aggression is an ideal sequence of actual occurrences,
and can be inhibited at various stages, this inhibition also result-
ing in such manifestations as anxiety. Kaufmann (1965) cites work
stating that anger arousal is not a sufficient condition for aggres-
sion, and that one can yet on other occasions make a sound infer-
ence to hostility without the actual expression of aggression. How-
ever, the hypothesis holds that the sequence would always occur if
it were not inhibited or suppressed and in this sense it is a one-to—

one response sequence (Kaufmann, 1965; Berkowitz, 1964).

Factors in Essential Hypertension

 

This paper has presented and h0pes to support an aggres—
sion-inhibition theory of chronic high blood pressure. Yet many
other factors contribute to this development. This section will
look at these factors and consider their relevance for this paper.
Procedures for controlling same will be in the methodology section.
Age, sex, and intelligence level are three variables that immediately
come to mind as relevant in virtually any study. There is no strong
evidence, however, that intelligence level is an important factor in

the development of high blood pressure, except where it is

29

epiphenomenal to such things as race or age. Miall (1961) con-
cludes there is overwhelming evidence that blood pressure tends

to consistently increase with age in the general population, and

is one of those variables that must be controlled. McDermott
(1960) found a tribe of Navajo Indians in the southwest United States
whose blood pressure does not increase with age, yet such a find-
ing is rare.

Miall further notes that, in most studies, consistently
replicable results on sex differences have been found. Blood pres—
sure was found to be lower in young women than in young men in
all of the populations studied, and higher in old women than old
men. Blood pressure rises faster with age in women than in men
at all adult age points. The crossover to a higher level of hyper—
tension for women than men appears to occur at the end of the
typical reproductive period.

All indications are that race is a factor which should
be considered in a study on blood pressure levels. Schroeder
(1961), an acknowledged expert in this area, implies that this is
not so much because there is clear-cut evidence to support the
notion that race is a definite primary factor. Rather, the majority
of evidence is extremely confounded by other factors, and no con-

clusive decision can be made at this point.

30

The papers presented by both Schroeder (1961) and by
Miall (1961) at the Prague Symposium and the round table discus-
sions following same suggest two conclusions. First, findings of
high blood pressure levels in various populations probably stem
from cultural-emotional factors. Secondly, seldom has anyone
clearly delineated which factor is operating.

Schroeder (1961) has also offered a most interesting
finding: a consistently significantly negative correlation of the
hardness of finished public water supplies with rates of hyperten-
sive heart disease; thus, the "softer" the water, the higher the
epidemiological incidence of hypertension. He has presented a
most fascinating physiological-anatomical theory as to why this
occurs. While fascinating, it ignores the fact that softness of
water is usually associated with the degree of civilization of a
population, a factor usually positively correlated with the degree
of suppression of hostility. This latter line of thinking, as well
as directly contradictory evidence to the hypothesis, was offered
during the discussion of Schroeder' 5 paper.

Height, weight, diet and smoking are factors that have,
on occasion at least, been hypothesized as contributory to the de-
velopment of high blood pressure. They may at times be correlated
with contributing factors such as age or sex, and thus on occasion

will appear to be important in themselves. The paper by Schroeder

31

(1961) and a proceeding discussion of this paper suggest that there is
no substantial reason to assume these factors are contributory in
even a slight manner. When controlled studies are carried out to
investigate the effects of these factors, e. g. , as by Damon (1961),
no relation to chronic high blood pressure has been found.

The genetic variables must be considered in any study
potentially involving physiological factors. The hypothesis of a
genetic factor as directly causal to hypertension underlies the
theories promulgated by Goldblatt and his colleagues. As mentioned
earlier, recent research mirrors a growing general disfavor with
this theory. This has tended at the same time to suggest that there
is no genetic factor operating.

Yet, to hold the genetic hypothesis in total disfavor
because of the inadequacy of Goldblatt' s theorizing is to throw
out the baby with the bath. Sir George Pickering (1961) has taken
a different approach than Goldblatt. He accumulates large statisti-
cal samples of hypertensives, normotensives, and their close
relatives. He has found higher mean blood pressure values for
relatives of hypertensives as opposed to relatives of normals. How-
ever, just as in the constant debate over whether there is a genetic
factor in schizophrenia or not, this type of approach cannot sub-

stantiate a genetic factor in high blood pressure. To interchange

32

the concepts of "genetic factor" and "family history" is a logical
and theoretical error.

The consensus of most modern thinking is that one can
at most only hypothesize a possible genetic predisposition to high
blood pressure. As mentioned earlier, this could be a "normal"
inherited factor such as the strength or size of the artery, which
in a large pOpulation over a long period of time could play a small

part in a quantitative assessment of blood pressure levels.

CHAPTER II
PROBLEM

Prior parts of this paper deal with certain areas and
issues that are relevant to a study of aggression and blood pressure.
A theory positing emotional rather than physiological factors as
primary was presented. The concept of "specificity" was examined
Since emotion per se is not seen as a causative factor. Rather it
is the inhibition of aggression, apparently considered to be basically
a suppressive pattern (Feshbach, 1964; Groen, 1964), which leads
to either felt and/ or psychosomatically manifested anxiety. This
anxiety, if it is experienced, should be manifested on a question-
naire measure. Specific emphasis is also placed on the variable
of aggression-guilt, i. e. , guilt over the expression of aggression.
This aggression-guilt was found to be a factor by Gambaro (1966),
and an attempt to verify it as a significant factor is an important
part of the present study.

Empirical attempts to confirm a hypothesis positing
emotional factors as primary have been of two types. First,

laboratory studies (Shapiro and Horn, 1955; Oken, 1960; Hokanson,

33

34

1961) have been undertaken where varying degrees of rigor have
been exerted on the control of contributing variables. They have
generally supported the hypothesis. Grinker (1953) and Groen
(1964) point out the basic problem with this approach. They note
that short-term effects can be clearly demonstrated, yet the cru-
cial factor in the development of a high blood pressure level is the
chronicity of the elevated pressure. It is then important to estab-
lish the chronicity aspect of the independent variable, in this case
the inhibition of aggression. Laboratory studies have not been set
up to correlate elevated blood pressure with a chronic pattern of
inhibiting aggression, at least not on human subjects.

The other avenue is to study differences shown by
various observed and categorized samples by means of an
observational-statistical approach. For example, Miasnikow
(1961) found two to three times as many cases of essential hyper-
tension among employees of the Moscow Telephone Exchange as
in control groups. As in most studies of this nature, there was a
control for only age and sex variables.

The naturalistic-statistical study has the advantage of
being able to observe the effect of chronicity, but in such studies
(Miasnikow, 1961; Hood, 1961), a majority of the relevant variables
typically go unconsidered, leaving the ”true" correlation an open

question. The present study is basically observational-statistical

35

in method. To definitively assess the chronicity variable in a
study of blood pressure in the experimental laboratory would
require extensive facilities and time. Another problem occurs
if the study uses human SS, since actual tissue change is an inte-
gral part of the theory. As this is a naturalistic study, it has to
adequately control all relevant variables to improve on previous

studies which have used this approach.

Hypotheses

 

The basic problem for this study is to investigate the
effects of the chronic inhibition of aggression upon blood pressure.
First of all, samples of subjects (Ss) are obtained who significantly
differ from each other in the amount of aggression expressed in
the prison environment.

It is hypothesized that:

1. Ss who expressed interpersonal aggression prior to
incarceration and who in the past 18 months while
incarcerated have expressed little or no aggression
have significantly higher diastolic pressures than those
Ss who likewise expressed aggression prior to incar-
ceration, but who have since continued to aggress sub-
stantially against the environment.

2. Within the high aggression group of Ss, those who show

high guilt over the expression of aggression have

36

significantly higher diastolic pressures than those who
show low guilt over such expression.

58 who expressed interpersonal aggression prior to
incarceration and who now suppress or inhibit such
expression exhibit more anxiety on an anxiety scale
than SS who prior to. incarceration expressed inter-
personal aggression and continue to do so after incar-

ceration.

CHAPTER III

METHOD

Subjects

The selection of subjects is most important when, as
in the present study, it provides crucial prior control of relevant
variables.

The subjects (SS) were inmates of the State Prison of
Southern Michigan (SPSM) at Jackson, Michigan. The original
selection procedures were directed at the main record folders of
the some 2, 000 inmates presently at SPSM. Each file includes
every piece of written material produced concerning the individual
inmate from the time of his pretrial social history to the present.

The selection procedure focused on the control of these
variables:

1. The first was a time variable, more specifically, the
length of time Spent in the institution from the time of
incarceration for the present offense up to the time of
selection. Only folders on inmates who have been in

prison for at least two years immediately prior to this

37

38

selection procedure were retained for further perusal.
This eliminated measuring phenomena that were actually
reactions to the chaos and threatening situations that
the new prison inmate encounters, especially in the
first three months of incarceration.

Files were retained for further processing only when
the inmate had been incarcertated for a crime involving
direct physical aggression. This includes murder,
manslaughter, felonious assault and also such crimes
as rape and armed robbery when perusal of the file
indicated that direct aggression was also involved. SS
were thereby comparable on the variable of having
expressed aggression immediately prior to incarcera-
tion and the differences in subsequent expression within
the prison population are rendered more meaningful

for this study.

While the sex variable was automatically controlled
since SPSM is all male, a consideration must be made
for age. The age range of 23 to 35 years was employed.
Hood, in a discussion of Pickering' 3 results (Pickering,
1961), states there is only a slight blood pressure rise
over time in the age range of 21 to 35 years, relative

to later age ranges of similar length, pointing to

39

the reasonableness of the limits used in this

study.

Evidence of a direct relationship between race and

high blood pressure is at best weak and contradictory
(Miall, 1961). Only main files of the few non-Negro,
non-Caucasian inmates of the prison were eliminated.
For two reasons the files on subjects who are medically-
established cases of hypertension were eliminated.

The first is so that the small percentage of cases in
whom hypertension is of a physiological origin was not
involved. Secondly, one avoids studying SS in whom
suppression of affect could be a result of being labeled
"hypertensive" (e. g. , where advised to "take it easy”).
A fortunate mode of control is available in this instance.
Less than a year earlier, the Upjohn Drug Company
searched the prison for cases of essential hypertension
to be employed in a drug study. They gathered all
possible clues from the medical records of all patients
to obtain as many cases as possible. The list was
availableand, of course, covered any inmate who had
been in the institution for two years.

An earlier section of this paper pointed out that obesity

has received no substantial support as a primary factor

40

in chronic high blood pressure ( Miall, 1961). How-

ever, considering the possibility that it may be con-

tributory in extreme cases, and thus be overlooked
in a statistical approach, those men diagnosed as
obese or substantially overweight in their medical
history file were eliminated.

Diet and the hardness of finished water supplies
(Schroeder, 1961) have been suggested as contributing or even
primary factors in essential hypertension. These factors, as
well as many others along the same line, were controlled by a
combination of the homogeneous prison environment and by only
using inmates who had been in SPSM for two years or more and
who were not medically—established cases of essential hyperten-

sion.

The Sample

 

An original population was delineated by the above
procedure. The population obtained consists of males, 23-35 years
of age, who are neither mentally defective, overweight, nor
medically-established cases of essential hypertension. Each was
incarcerated for a crime invOlving actual physical aggression,
had been in prison for at least two years prior to this study, and

is either Negro or Caucasian.

41

A sample of 111 SS was selected from this population
via a table of random numbers. These SS were assesSed for the
amount of aggression expressed during the 18 months immediately
prior to this study. Thirty-nine 88, approximately the top one-
third of the sample who were lowest .on the expression of aggression,
were designated the low aggression (Low A) group. This allowed
for SS who might not be tested, and was also a natural break in the
variable measured. That is, these 39 SS were the only 39 not
involved in any aggressive incidents, nor did any of these 83 have
a report which even suggested an incident of aggression while in
prison. The middle group was composed of 33 Ss who had only
one definite report of acting-out aggression. The high aggression
(High A) group was composed of 39 inmates who definitely acted-
out aggressively two or more times. The average number of
definite reports for this group was 2. 91, and, in addition, this
group had been involved in many potentially-aggressive incidents.
The range was from two to five definite reports of aggression.

This is a typical officer' 3 report judged as aggressive:

At 2:10 p. m. today, I observed Inmate X talking loudly to
Inmate Y; however, I am pretty sure they didn' t see me.
As I started to walk up to them, Inmate X jumped at In-
mate Y and started hitting him with his fist on his neck and
head. Inmate Z helped me pull them apart. As I took In-
mate X to the Hall Office, he laughed and said, "I' m glad

I hit that black son-of—a-bitch. " He didn' t resist coming
along.

42

A typical report judged as non-aggressive read:

At 2:45 this afternoon, I saw a man who I didn' t know walking
around the kitchen area. I asked his name and what he was
doing here. He gave me his name and said he was just walk-
ing through. I shook him down and found four pork chops.

He would not tell where he got them, and said it was none of
my business. I am reporting him for "skating, " possession
of contraband, and insolence.

All Ss in the high and low aggression groups were then
called in for testing, and 35 SS in each group actually completed
the testing. Two of those four Ss in the High A group were not
tested because they had recently been transferred to the Marquette
Branch (maximum security) of the Michigan Prison System. This
furnishes extra validity to the criteria employed, since a transfer
to Marquette is used when an inmate has come to be considered so
consistently and immediately dangerous to civilians and inmates
that he must be placed in virtual isolation. The other two Ss not
tested simply refused to participate and would give no reason for
this.

Four of the low aggressive SS were likewise not tested.
As in the high group, two men refused to be involved and would
give no explanation. Another man had been given a special dis-
charge due to an exemplary record before he could be tested. The
fourth was found to be mentally deficient. Also, two completed

protocols, one from each aggression group, mysteriously disap-

peared from the examiner' s desk shortly after they had been

43

completed. Thus, the processing of the results was carried out
upon 34 Ss each in both the low and high aggression groups.

The groups were assessed as to possible significant
differences on other relevant variables. While several variables
have been controlled wholely or partially by the population selec-
tion procedures, a check must be made on age, and a check was
also made on intelligence. The samples do not differ significantly
on any of these three variables. The intelligence scores were on
a 1 to 10 unitary scale based on the 16 PF test (Cattell, Saunders,
and Stice, 1957). (See Table 1.)

Table 1. -- Basic comparisons of High and Low A groups on mean
age, intelligence (Cattell scale), and race.

 

 

 

 

Race 16 PF
Age Intelligence
Negro White Scale
HighA 28.1 18 16 3.44
LowA 27.8 17 17 3.32

 

 

 

 

 

A t-test produced no significant differences between the
two groups in age. A Chi-Square yielded like results for race. A
Kolmogorov-Smirnov test, for samples less than N = 40 (Siegel,
1956), indicated no significant differences between the high and low
aggressive groups in Cattell 16 PF scores (Cattell, Saunders, and

Stice, 1957).

44

Both groups adhere very closely to the distribution of
the general prison population with respect to race and intelligence.
The prison population is approximately 50% Negro, as in this sample.
Approximately 88% of the present sample falls in the low average-
average range of intelligence, and the SPSM Reception-Diagnostic
Center reports they get closely [similar results within the general
prison population. Four men in the Low A group and two in the
High A group were above average in intelligence, though not one of
these men was in the superior range. (See Table 2.)

Table 2. -- Distribution of High and Low A groups on Cattell

intelligence scores. (The percentage of general p0pulation
adults attaining each score is in Table 8. )

 

 

Sten Score . 1 2 3 4 5 6 7 8 9 10 Total

 

High A 1 13 2 10 6 0 2 0 O O 34

LOW A 5 8 7 4 6 3 1 0 O O 34

 

The literature (Miall, 1961) suggests that over the age
range employed in this study (23—35 years), one would only expect
an overall average blood pressure rise of 5 mm. Hg. Also, there
were no significant differences between the two groups in age. In
any case, it appeared worthwhile to investigate whether indeed
there was a correlation between blood pressure and age within the

two groups. In the High A group, a correlation coefficient of -. 19

45

was obtained between these two variables. A correlation of +. 35
was the result in the Low A group, a result more in the expected
direction. Neither result was close to significance at even the

. 10 level of confidence.

In addition to the similarities in the variables men-
tioned above, there were also no substantial differences in the
type of crime involved in the two groups. (See Table 3.) For
example, the Low A group had four crimes involving rape and the
High A group included five such offenses. Obviously, if there had
been a significant difference between the two groups in crimes
involving rape, closer scrutiny would be necessary.

Table 3. -- The number of SS incarcerated for a particular crime
(last offense) in both the Low and High Aggression groups.

 

 

 

Crime High A Low A
Rape, attempted rape (with assault) 5 3
Murder, first degree 8 11
Murder, second degree 4 3
Manslaughter 4 1
Robbery, armed, with assault 7 10
Assault to murder, felonious assault 6 5
Sodomy, with assault 0 1

 

TOTAL A 34 34

 

 

 

46

A post hoc decision was made to see whether there
was a significant difference between groups in the amount of con-
secutive time (in months) spent in the institution prior to being
included in this study. Due to transfers and discharges, data was
available on only 29 SS in each group, out of the original 34 Ss
tested. The mean number of months was 36. 1 in the High A group,
and the range was from 24 to 60 months. In the Low A group, the
mean was 35. 2 months, and the range was from 24 to 65 months.
The t-test on this data was not even close to being significant, and
the F-test indicated homogeneous variances. A conclusion that
these two groups are quite Similar on all relevant variables appears

to have strong justification.

Measurements

 

1. Blood pressure was measured by means of a sphygmo-

manometer, the most common measure employed. Gambaro (1966)

 

described the procedure as follows:

This method of measuring blood pressure is called the aus-
cultatory technique because the examiner employs a listening
device, a stethosc0pe, in order to take his measurements.

To take a blood pressure reading, the examiner encircles

the subject' 3 upper arm (usually his left arm) with a rubber
tube that can be inflated with air. The amount of air pressure
in the cuff is measured by a column of mercury (Hg. ). Hence,
blood pressure readings are usually stated in millimeters
(mm.) of mercury. The examiner inflates the cuff so as to
completely stop the flow in the main, or brachial, artery

of the arm. He then releases pressure in the cuff until he

 

47

can hear a beating sound through the stethosc0pe, which is

applied to the artery below the cuff. The height of the

mercury column at the time when he first hears the sound

is the systolic blood pressure. The examiner then releases

more air from the cuff until he can no longer hear the beat—

ing sound. The height of the mercury at that somewhat
indetermined time when the beating sound can no longer be

heard is the diastolic blood pressure. (p. 27)

2. Aggression-Guilt was measured by means of the Mosher
Incomplete Sentence Test (MIST), or, more specifically, by the
fifteen sentence stems that refer to aggression-guilt. Ten extra
MIST items were also administered so as to cushion a blatant
"aggression" pull in the scale. The test is actually a fifty item
sentence completion sheet devised and then validated by Mosher
(1961) as a measure of three types of guilt: aggression-guilt, sex-
guilt, and morality-conscious guilt. His method for scoring the
sentence completions used referents for guilt largely derived from
psychoanalytic theory. The MIST allegedly measures relatively
suppressed feelings of self-criticism for failure to live up to the
person' s internalized ideals.

The reliability between judges in the categorization
of reSponses on the aggression-guilt continuum of the MIST has
been established in two previous studies (Mosher, 1961; Gambaro,
1966). Mosher obtained an inter-rater reliability of . 90, while

Gambaro reported a reliability of . 86. The same instructions,

etc. , were used in this study as in the two previous ones, with

48

one difference. One MIST aggression-guilt stem was typed "I
have . . . " rather than "I hate . . . ;" and since this was not
noticed until the completion of the study, only 14 responses were

available on each subject for rating.

Construction of the MIST

 

The stems for the sentence completion test were col-
lected by Mosher from three psychologists, and were derived mainly
from ideas drawn from psychoanalytic theory. The two hundred
stems collected were divided into parallel forms and administered
to 38 VA Hospital patients. Stems were rated on degree of informa-
tion elicited, and those selected were rated into the three afore-
mentioned guilt classes. The final form then consisted of fifty of
these agreed—upon items.

In his dissertation, Mosher (1961) predicted that sub-
jects high on hostile-guilt, as determined by his test, would not
condition as easily to hostile content as SS scoring low on hostile-
guilt. This hypothesis, reflecting the degree of construct validity,
was substantiated. As was mentioned earlier, Gambaro (1966)
found guilt over aggression, as measured by the MIST, to be a
contributing factor in the subject' 8 reaction to his experimental
manipulation. Mosher' S and Gambaro' 8 use of median cutoff points
for dichotomizing into high and low guilt groups was employed in

this study also.

49

The Cattell 16 PF Test

 

The Cattell 16 Personality Factor Test is a question-
naire containing 187 items. Cattell, Saunders, and Stice (1957)
assert that measurements on 16 unitary personality factors are
obtained by this test, and that each of the 16 factors corresponds
to a primary personality factor established outside the question-
naire realm, such as in ratings of real life behavioral situations
or in social response patterns. Split half consistency coefficients
on each of the 16 factors range from 0.71 to 0. 93.

There are three forms of this test, which are consid-
ered to be equivalent when eXpressed in standard or sten scores.
Form A was employed in this study. The sten scores are obtained
by referring to norms that the authors have gathered on various
populations such as male and female college students, or on males
or females in the general population, etc. Three alternative an-
swers to each test item are provided for, since the authors feel a
"two-choice" item frustrates genuine attempts to give accurate
answers and can produce a general disinclination to respond to
the test.

After the items were first screened for factor validity,
and for a good distribution of their alternative responses, they
were chosen so that on each factor just as many positive or nega-

tive responses contributed to the total score, thereby avoiding

50

score distortion through mental sets. Also, the items have been

arranged so that responses for one factor do not bunch together.

Procedure

 

The selected SS were classified on the aggression vari-
able, using the files of the Deputy Warden of SPSM. The Deputy' s
office keeps an extensive file on every acting-out incident in which
any inmate is involved. An occurrence is reported and filed, whether
it is punished in some way or not; whether it involved inmates I
alone, or both custody personnel and inmates; and also if it was
the threat of aggression that did not get beyond the verbal stage.

SS were ranked by the number of incidents that involved acting—out
aggression, or where this was apparently prevented only by the
interference of custody officers.

A reliability check had to be carried out on the placing
of officer' 8 reports from the Deputy' s files into aggressive and
non-aggressive categories. Ten reports were randomly selected
from those previously seen as aggressive by the investigator. Ten
reports that had been seen as non-aggressive were also randomly
selected. These twenty reports were presented to a psychiatric
social worker, who is employed by the Psychiatric Clinic. He
was told to simply judge whether the report involved direct physi-
cal aggression or not. He was presented with a pile of approxi-

mately 50 slips of paper, the top 20 of which were the retyped

51

reports, the last 30 being only blank sheets of paper. This alle-
viated the problem of biasing his last several responses from a
potential attempt to even up the response piles, etc. There was
agreement between the two judges (investigator and social worker)
on 19 of the 20 incidents. One of the reports categorized as
aggressive by the primary judge was seen as non-aggressive by
the second judge. Using a 2 by 2 table, a Chi-Square of 16. 36
was obtained, high enough to reject a null hypothesis of no rela-
tionship between judges at the . 001 level of confidence. The con-
tingency coefficient of correlation on this data was . 67, quite high
since, as Siegel (1956) asserts, the upper limit of this correlation
coefficient for a 2 by 2 table is .707.

A short letter was written to the selected inmates,
requesting them to come to the Psychiatric Clinic when they re-

ceive a call. The letter covered these points:

1. It emphasized that the particular inmate contacted
had been randomly selected from the general popula-
tion.

2. Each inmate would receive two packs of cigarettes

for participation in a research project, and the time
required would be from thirty to fifty minutes. (Two
packs of cigarettes is equivalent in cost to the highest
pay a prisoner can receive for an eight hour job in
SPSM. The typical pay is only 35 to 40 cents a day
for an inmate' 8 work.)

3. The study was in no way connected with the Psychiatric
Clinic, except that the person running the study works
in the clinic. The affiliation with Michigan State Uni—
versity was noted.

52

4. For each person who participated, a letter will be
placed in his record file stating only that he very
generously and c00peratively participated in research
conducted by one of the Psychiatric Clinic Staff. The
high value of such a commendation is well known to
each inmate for its relevance in parole processing.

5. Anonymity will be retained and all information will
be available only to the investigator and his assistant.
Information as to the outcome of the study, etc. , will
be available to all participants, upon request, at the
end of the study. They will be notified when this is
available.

These SS were put on call to appear in the Psychiatric
Clinic at a specified time. If a man did not Show up, a letter was
written indicating that he had been put on call for a specific time.
(At times, calls are simply not relayed by the civilian clerks. )
This second letter re-emphasized the points made in the first let-
ter, stated what time the S would be placed on call for, and stated
that if he did not receive this call he should call himself or send
a note to the experimenter in the Psychiatric Clinic and he would
be called up immediately, a common procedure. When he reported
to the clinic, the following occurred:

1. He was escorted to an office by E. 1 A conversation
of approximately five minutes was held to allay any
anxieties or answer any questions S might have and
simply make him more comfortable in the situation.

2. E then showed S the sphygmomanometer, explained

what it is, took a blood pressure reading, and then
recorded it.

 

1E, the prison psychology clerk, carried out the pro-
cedure under the direct supervision of the author.

53

3. Following this reading, he told S to relax, and that
he (E) would return in a few minutes or so.

4. In ten minutes E did return and took a second blood
pressure reading, then waited exactly 15 seconds,
and took a third reading.

5. Immediately after taking the third reading, E gave S
a pencil and a copy of the MIST, telling him to answer
the items as honestly as possible and that his answers
would remain anonymous and confidential. He was
told to let E, who read a book at the other end of the
office, know when he finished.

6. At that point E gave S Form A of the Cattell 16 PF
Test, with the appropriate test instructions, and also
repeated the instructions in step five.

7. At the completion of the test, S was thanked and asked
not to divulge to anyone what went on in the experiment
until notified by mail that the entire test was completed.
An average of the second and third blood pressure

readings (step 4) was used as the index of the inmate' S blood pres-
sure level. Previous to the running of 85, blood pressure readings
were taken on ten randomly selected inmates by both the psychology
clerk (E) and Mr. Alfred Wigley, an R. N. with twenty years of
experience. (See Table 4.) The complete procedure of putting the
cuff on, etc. , was carried out by both examiners. A Pearson
product-moment correlation of . 97 was attained on the diastolic
pressure readings, the focal variable in this study. (A correlation
of . 67 was attained on the more unstable systolic pressures. )

Each single response to the MIST was retyped on a

single slip of paper, coded, and presented for judging to a psychology

54

graduate student. 1 The responses were rated on a one to five scale
for the degree of guilt over the expression of aggression. Ss total
scores were tallied, and SS were then ranked in order of indicated
guilt (Mosher, 1961).

Table 4. -- Systolic (upper number) and diastolic (lower

number) pressures for 10 S8, to establish E' s
reliability in taking blood pressures.

 

 

Examiner 1 2 3 4 5 6 7 8 9 10

 

124 130 124 132 122 126 116 130 130 118

E 66 7o 88 78 62 7o 60 7o 80 7o

 

 

 

126 130 120 136 120 126 114 126 126 120

R' N' 68 70 86 80 66 76 60 72 78 7O

 

 

 

 

Two hundred responses were drawn from these responses,
using a table of random numbers, and presented to the investigator
for judging. An inter-rater reliability coefficient of . 86 was obtained,
comparable to the . 90 and . 86 obtained by Mosher (1961) and Gambaro
(1966) respectively. Responses from this sample can apparently be
reliably scored. The anxiety factor on the Cattell 16 PF was scored
as directed in the Cattell Manual (Cattell, Saunders, and Stice, 1957),

a combination of six of the sub-scales in that test.

Treatment of the Data

 

T—tests may be employed to test the significance of all

three hypotheses. The sten scores of the 16 PF test are reported

55

by the Cattell Manual (Cattell, Saunders, and Stice, 1957) to be
equal-interval scales. Also, the variables are essentially normally
distributed, using sten scales, in either a college or general pOpu-
lation. The general population sten norms were used in this study.
(T-tests and non-parametrics have been employed to assess whether
the groups differed significantly from each other on such variables

as race, intelligence, etc.)

CHAPTER IV

RESULTS

P rimary Hypothe se 8

 

After the reliabilities of the various measurements had
been established, and the two samples were found to be derived
from the same p0pulation pool, the results relevant to the hypoth-

eses were then examined.

Hypothesis I

 

The first hypothesis states:

SS who expressed interpersonal aggression prior to incar-
ceration and who in the past 18 months while incarcerated
have expressed little or no aggression have significantly
higher diastolic pressures than those SS who likewise
expressed aggression prior to incarceration but who have
since continued to aggress substantially against the environ-

 

 

 

ment.
Table 5. -- Mean blood pressures (using the average of the
83' s second and third readings as the item score)
for the High and Low A groups.
ngh A LOW A "t" d. f. P Direction
Mean Mean
Systolic 116.2 120. 9 1. 39 66 2-tailed
Diastolic 71. 1 76. 8 2. 43 66 . 01 1-tailed

 

 

 

 

 

 

 

56

57

A t-test on the diastolic blood pressures between the
High and Low A groups yielded a "t" of 2. 43. With 66 d. f. , and
the 1-tailed confidence limit decided upon earlier, the null hypoth-
esis was rejected at the . 01 level of confidence. Thus, the hypoth-
esis may be accepted that within the group of men who were incar-
cerated for aggression, those who continued to aggress while
institutionalized have significantly lower diastolic blood pressures
than those men who ceased to aggress physically while institu-
tionalized. Using the systolic pressure, a significant result was
not obtained, though it was in the same direction as found using

diastolic pressures.

Hypothesis II

 

The second hypothesis states:

Within the high aggression group of Ss, those who Show
high guilt over the—expression of aggression have signifi-
cantly higher diastolic pressures than those who show low
guilt over such expression.

Table 6. -- Mean blood pressures for the high and low
guilt groups within the High Aggression group.

 

 

 

High Low
Guilt Guilt "t" d. f. P Direction
Mean Mean
Systolic 120. 2 1 12. 7 1. 77 32 2-tailed
.Diastolic 75. 0 67. 35 2.41 . 32 . 025 1-tailed

 

 

 

 

 

 

 

58

Measures of such guilt were compounded from the MIST
ratings as described. earlier and the HighA group dichotomized at
the median MIST score, as per Mosher (1961) and Gambaro (1966),
into high and low guilt groups. A t-test on the blood pressure
indices between these two groups resulted in a "t" of 3. 17. USing
a two-tailed test with 32 d. f. , the result is significant at the . 05
level of confidence. Thus, (the null hypothesis is rejected. Within
those inmates who continued to aggress substantially after incar-
ceration, those with higher guilt over the expression of this aggres-
sion showed significantly higher diastolic blood pressures. Using
systolic pressure measures, the result was not significant, yet is

again in the same direction as obtained using diastolic pressures.

Hypothesis III

 

The third main hypothesis reads:

SS who expressed interpersonal aggression prior to incar-
ceration and who now suppress or inhibit such expression

exhibit more anxiety on an anxiety scale than SS who

prior to incarceration expressed interpersonal aggression
and continued to do so after incarceration.

Table 7. -- Mean anxiety scores (derived according to
Cattell, Saunders, and Stice, 1957) for the
High and Low Aggression groups.

 

 

 

ngh A LOW A "t" d. f. P Direction
Mean Mean
Cattell
Anxiety 7. 82 7. 62 . 449 66 2-tailed
Score

 

 

 

 

 

 

 

59

Anxiety scores were derived from the Cattell 16 PF
protocols, as described earlier. A t-test on these anxiety scores
between the High and Low A groups yielded a "t" of .449. The
result was not close to Significant; therefore, the null hypothesis

was accepted.

Additional Findings

 

While not proposed earlier, there is an apparent cor-
ollary hypothesis to Hypothesis II. That is, those Ss in the Low A
group should have significantly lower guilt scores as measured by
the MIST than the High A group. A t-test on the guilt scores between
the High and the Low A groups resulted in a "t" of 1. 68. This was
extremely close to significance at the . 10 level of confidence using
a two-tailed test and 66 d. f. However, an F-test between the vari-
ances of the Low A (S = 22. 4) group and the High A (S = 45. 9) was
significant, indicating heterogeneous variances. Using one-half
the degrees of freedom to compensate for this heterogeneity, the
"t" did not come close to significance. The trend, however, was
in the expected direction.

From the results on Hypothesis II, pointing to the
importance of a guilt factor in the deveIOpment of high blood pres-
sure within the High A group, the question arises whether this

guilt factor operates throughout the sample. A product-moment

60

correlation was run between the blood pressure indices and the

MIST guilt scores over the whole group. A positive correlation
of . 106 was obtained, a trend in the expected direction, though

certainly not statistically significant.

Each of the Cattell 16 PF factors was graphed to show
the distribution of scores within both the High and Low A groups,
as well as the overall distribution. The Cattell sten scores were
designated as the ordinate variable and the blood pressures were
ordered along the abscissa. Only the data on Factors A and F
indicated the possibility of a significant difference between groups.
A high score on factor F, according to the Cattell Manual, indicates
a happy-go-lucky personality, whereas a low score reflects a sober,
prudent, and serious individual. This experimenter' s first hypoth-
esis would have been that the High A group would score higher on
this factor, yet the Low A group was higher. The results on both
factors, however, were not significant, using t-tests.

Other than this, the scores of the two groups are re-
markably Similar. Two explanations are feasible in this regard.
First, the similar Cattell scores could be construed as supporting
earlier assertions that these samples are truly comparable and
from the same population pool. Otherwise, it suggests that the
Cattell 16 PF is not powerful enough to differentiate these two

groups that do indeed exhibit different behavior. The means of

61

the Cattell sten scores of both groups, as well as the overall mean,
are presented in Table 8, along with a description of each 16 PF
factor and the percentage of the general population adults usually
attaining each score.

Taking the sample as a whole, the overall mean
deviated two and one-half or more sten scores from the general
p0pulation mean on factors B, C, M, and 0. (See Table 8.) This
indicates that the sample mean of these factors occurred at a point
beyond which only 24% of the general adult pOpulation would fall.

One could speculate from this that the present sample tends to be
less emotionally stable (C), less intelligent (B), more imaginative
or bohemian rather than conventional (M), more suspicious (L), and
more apprehensive than placid or serene (0), when compared to the
general population.

The literature suggests possible correlations between
blood pressure levels and Cattell 16 PF factors. Cattell and Scheier
(1959) claim to have found positive correlations between high systolic
blood pressure and factors A+ (extroversion), Q2- (group depen-
dency), N+ (shrewdness), M- (practicality), and E+ (assertiveness),
yet no further description of subjects, procedures, or results was
provided. Harburg, Julius, McLeod, McGinn, and Hoobler (McGinn,

et a1. , 1964), however, find a correlation of high systolic pressure

62

Table 8. -- Cattell factor descriptions, Low and High A group
and overall group mean sten scores, and the percentage of
general p0pulation adults attaining each sten score.

 

 

Low Sten Score High Sten Score Low A High A Overall

Factor

 

 

 

 

 

 

 

Description Description Mean
Reserved, .
A detached Outgo1ng, warm 6. 35 7. 20 6. 77
B Less intelligent More abstract . 32 .44 . 38
. Emotionally

C Eas11y upset stable . 71 . 94 . 82

E Humble, mild Assertive 4. 41 4. 85 4. 63

F Sober, serious Happy-go-lucky 4. 55 3. 91 4. 22

G Expedient Conscientious 3. 94 3. 97 3. 96

H Shy, timid Socially bold 5. 23 4. 67 4. 95

I Tough-minded Sensitive 6. 67 7. 23 6. 95

L Trusting Suspicious 7. 17 7. 58 7. 37

M Conventional Imaginative 7. 91 7. 17 7. 54

N Forthright Sentimental 5. 08 5. 20 5. 14

O Placid, serene Apprehensive 7. 61 7. 88 7. 74

Q1 Traditional Experimenting 5. 52 5. 41 5. 46
Group- . .

Q2 dependent Self suff1c1ent 6. 23 6. 35 6. 29
F 011°“ own Controlled .05 .82 .93
urges
Relaxed, torpid Tense . 29 . 17 . 23

A sten of 1 2 3 5 6 7 10 is obtained

by about 2.3 4.4 9.2 15.0 19.1 19.1 15.0 9.2 4.4 2.3% of adults.

 

63

with the minus end of the E scale (submissiveness), and also with
the I+ scale (sensitivity).

Using the overall group of 68 SS, correlations were run
between the average of the last two systolic measures on each S and
the factors cited above. The results are presented below in Table 9.
No significant correlation was found between high systolic blood
pressure and any of the Cattell 16 PF factors Suggested above.

Table 9. -- Correlations of systolic blood pressure indices

with Cattell factor scores, using the overall sample.
(See Table 8 for Cattell factor descriptions.)

 

 

 

 

Factors
A E I M N Q2
corrélétmn +.068 +.095 +.058 -.056 -.125 -.030
Coeff1c1ents

 

 

 

 

 

 

 

While the groups did not differ significantly in race, it
was decided to find out whether Negroes and whites in the overall
group differed significantly in blood pressure, using the diastolic
pressure indices. The mean blood pressure for the 35 Negro SS
from the overall group was 74. 8, and the mean for the 33 whites
was 73.0. A "t” of 1. 03 was obtained, and using degrees of free-
dom and a two-tailed test, the result would not be significant at

better than the . 30 level of significance. Nevertheless, the trend

64

for Negroes to exhibit higher diastolic blood pressures is in line
with the majority of evidence (Miall, 1961).

At one point, there was a question whether the assess-
ment of differences between groups regarding changes in blood
pressure from the first to the second reading might yield interest-
ing data, especially in regard to systolic pressure, since it is
significantly more reSponsive to short-term attitudinal changes.
However, it became apparent, partially in retrospect, that results
would be highly contaminated by several factors. For example,
several SS had a chance to Sit and rest 15 minutes or more prior
to being met by E and escorted to an office, while others were
seen immediately, often after a long walk to the clinic. This
would certainly affect the systolic pressure results, eSpecially
on the first administration of same, and there was no control for
this or other like variables. An inspection of the relevant data
indicates that indeed there were no significant differences in the

scores between groups anyway.

CHAPTER. V

DISCU SSION

The main hypotheses of this study were apparently
substantiated. Yet, before these could be meaningfully examined,
several other aspects had to be established. Most importantly,
the measuring instruments used have to be reliable. The use of
a sphygmomanometer for recording blood pressure was obviously
appropriate, and in a reliability study between the examiner and
an experienced registered nurse, reliability was established for
this procedure. Mosher (1961) thoroughly established the reli-
ability and validity of the MIST, and the reliability of aggression-
guilt items was further supported by Gambaro (1966). Constant
re-evaluations for both reliability and validity have established
the Cattell 16 PF test as a solid questionnaire instrument for re-
search purposes.

The categorization of incident reports as aggressive or
non-aggressive was supported as reliable. A second judge simi-
larly assessed 19 of the 20 decisions made by the investigator,
enough to reject the null hypothesis of no relationship at the .001

level of confidence.
65

66

The validity of the samples used must also be ascer—
tained. The results strongly suggested that the samples were indeed
random samples of a larger population pool, and did not differ from
each other, exclusive of the hypothesized variables. The samples
were similar in such potentially within-sample differentiating vari-
ables as age, race, and intelligence. The high similarity in intel-
ligence scores, the limited range of intelligence scores, and the
fact that no more than approximately 10% of the Ss have a high
school diploma or more suggests the samples are homogeneous
socio-economically. Important also in the establishment of sample
validity were the low and non-significant correlations of blood
pressure and age within each sample. The high similarity of both
the High and the Low A groups on the Cattell Scales, as shown in
Table 8, likewise tends to support the assertion that these samples
were comparable and from the same population pool.

The primary hypotheses (I and II) were strongly sup-
ported, at the . 01 and the . 025 levels of confidence respectively.
This adds depth to the research and theorizing cited earlier by
providing an "in vivo" study on chronicity in humans. These re-
sults dovetail nicely with those studies carried out by Gambaro
(1966), Hokanson, 919:1: (1961; 1962, a, b, 1963), and others.

Those authors solidly documented an immediate rise

of blood pressure upon the presentation of aggression-eliciting

67

stimuli. Calhoun (1962) has demonstrated that a long-term condi-
tion of such stimuli may induce hypertensive-like states in rats.
Groen (1964) reports the non-expression of aggression in those
diagnosed as "essential hypertension" cases. Though this paper
does not focus on hypertension cases, one could logically expect
that in human SS, a long—term condition of not expressing felt ag-
gression would result in a state of chronically higher blood pressure
than in those SS who did express aggression in a similar environ-
ment. This was the object of the first hypothesis, and was so sup-
ported.

The second hypothesis focuses on the personality factors
related to high blood pressure. Little previous work has been done
in this area to indicate whether the factor involved might be anxiety
over fear of retaliation, guilt, or what. Gambaro' s (1966) paper,
focusing on Short-term effects, would suggest guilt over the expres-
sion of aggression as the relevant factor. This was supported and
amplified in the present study as pertinent to chronic effects.

No significant results were found when systolic pres-
sure indices were substituted for diastolic indices in the first and
second hypotheses, though results were in the same direction.
Actually, this experimenter did not expect significant results using
systolic pressure Since, first of all, it is a very unstable physio-

logical measure. Some evidence of this was noted in the inability

68

of the psychology clerk and registered nurse to get highly reliable
systolic blood pressure measures, whereas they did obtain such
results on the diastolic measures. Wilson (1961) cites another
reason why significant results with systolic pressures should not
be expected. He asserts that the establishment of persistent high
diastolic pressure is a prelude to a consistent and stable rise in the
systolic blood pressure, and is substantially prior to the deveIOp-
ment of pressures high enough to be considered essential hyperten-
sion. It is in this latter condition that the systolic pressures also
remain consistently high.

The non-significant results obtained in regard to Hypoth-
esis III indicates that anxiety, as measured by the Cattell, is not
related to chronic high blood pressure. Two research trends offer
a rationale as to why this is not so. Groen (1964) has most clearly
stated the first point. He notes that in persons who incur a psycho-
somatic disorder, there is little or no consciously—expressed
anxiety. In fact, he argues that anxiety, if it is present, is totally
repressed rather than suppressed. Much of what is tapped by the
Cattell Scales is consciously-expressed anxiety, or at most, anxiety
which is only thinly-disguised and suppressed.

From a different viewpoint, the absence of an anxiety
state accompanying the aggression-high blood pressure syndrome

could be inferred. The experiments of Ax (1953) and Schachter

69

(1957), noted earlier, did not consistently find a long-term diastolic
blood pressure rise accompanying anxiety. Yet, the first reason
given, the denial of anxiety by psychosomatic-prone individuals,

has the most immediate relevance for this study considering the
measuring instrument used. A strictly physiological measure of
anxiety might have produced different results, but many would then
argue that a strictly physiological measure could not properly be
termed "anxiety. "

A strict generalization from the results of this study
can be made only to the population from which the SS are drawn.
This would then limit the results to SPSM inmates who were incar-
cerated for a crime involving physical aggression. Yet, there is
no reason to suppose that the general finding is not valid throughout
the general population. Further discussion will be in light of the
probable relevance of these results to the general population.

The crucial role that guilt over the expression of aggres-
sion plays in the development of chronic blood pressure is quite
consonant with the hypothesizing referred to in the theoretical
section of this paper. Anokhin (1961) is most typical and lucid in
elaborating this viewpoint. The crux of his theory involves the
cortically—generated inhibition which upsets the autonomic-sympathic
balance, especially in regard to blood pressure. This cortical

inhibition, mediated by the reticular activating system, keeps the

7O

organism in a state of arousal well beyond the approximately twenty-
five minutes of time during which depressor impulses will fire.

Guilt, since it is closer to a strictly cognitive construct,
fits better with Anokhin' s standpoint than with such emotion-oriented
constructs as "fear" or "anxiety. " These latter constructs could
fit comfortably with a physiological theory of chronic high blood
pressure that refers exclusively to lower brain functions. However,
the cognitive aSpects of the guilt construct demand a theory which
considers the role of the cerebral cortex, i. e. , the type of theory
supported by the present data.

The findings of this paper are also more pertinent to
Anokhin' s physiological data than a theory positing only stress per
se as the decisive factor. Stress alone could be fitted more parsi-
moniously with a theory referring only to lower and mid—brain
structures. However, where guilt is involved, even under the aegis
of the strictest parsimony, reference to cortical processes is neces-
sitated.

Another assertion is quite clearly implied in the previous
statements. That is, these results support an emotion-origin based
theory of chronic high blood pressure or "essential hypertension"
over any variant of the physiological disease approach such as Harry
Goldblatt' S renal—humoral theory, considered earlier in this paper

(Cottrell, 196 1) .

71

In this study, Negro SS manifested a higher mean
diastolic pressure index (74. 8) than white SS (73.0). This dif-
ference was not significant and therefore offers little substantial
support for the majority of research which suggests Negroes have
higher blood pressures than whites (Miall, 1961), though it cer-
tainly does not contradict this research.

The research literature (Cattell and Scheier, 1959;
McGinn, _et_a1. , 1964) also suggested that several Cattell scales
(A+, Q-, I+, N+, M-, and E+ or -; see Table 8) might be corre-
lated with high systolic blood pressure. For the overall sample
employed in the present study, not one of these scales was found to
have anything near a significant correlation with high systolic pres-
sure. Neither were the High and Low A groups found to differ
Significantly from each other on any of the Cattell scales, even
when the two groups differed from the general population mean.
Apparently in this study the Cattell 16 PF was more responsive to
the gross variables of age, sex, and intelligence than to those
variables related to the differing behaviors manifested by these two
groups.

The primary hypotheses of this paper were strongly
and clearly supported. However, in order to substantiate any
generalizations made to the general population, further studies

are going to be needed. One interesting feature Should be examined

72

in these further studies. That is, it would appear very relevant to
determine what types of personality factors allow one group of
people to act-out aggressively in spite of guilt, whereas another
group is inhibited from such action, apparently as a result of this

guilt.

CHAPTER VI
SUMMARY

Research on high blood pressure and theories as to how
"essential hypertension" develops have suggested that the suppres-
sion of aggression may be a primary agent contributing to these
disorders. Such people as Groen (1964), Saul (Alexander and French,
1948), and Hokanson and his colleagues (1961; 1962, a, b; 1963) have
all, in one way or another, supported this hypothesis. Gambaro
(1966) has indicated that guilt over the expression of aggression is
a vital factor in this regard. If guilt is a factor, a supporting
physiological theory must take account of cortically-generated in-
hibition. Anokhin' s (Cort, e_til. , 1961) theorizing on the genesis
of chronic high blood pressure, which was presented in this paper,
takes account of such factors.

A population of SPSM inmates was selected on the basis
of several criteria. First, only inmates were used where the offense
resulting in incarceration had involved direct physical aggression.
They were also selected on such relevant factors as age, weight,

race, and prior medical history. These inmates were categorized

73

74

into Low and High Aggression groups, depending upon their behavior

pattern since being institutionalized.

Sixty—eight SS (thirty—four in each group) were pro-—
cessed. After taking a preliminary blood pressure, two blood
pressure measures were taken fifteen seconds apart, and the
MIST (Mosher, 1961) was administered to assess the amount of
guilt over aggression. The Cattell 16 PF test was administered

in order primarily to obtain both an anxiety and intelligence score.

An average of the last two blood pressures taken on
each subject was used, in order to increase the validity of the
measure. The MIST responses were rated for aggression-guilt.
The reliability on rating the MIST has been established by Mosher
(1961), Gambaro (1966), and by the investigator and an independent
judge in the present study. Anxiety and intelligence scores were

assessed per instructions in the Cattell 16 PF Manual.
The results led to the following conclusions:

1. A continued state of the non-expression of aggression
in those men who have at one time expressed direct
physical aggression results in chronically higher
levels of diastolic blood pressure than in those men
who continue to aggress against the environment.

2. Within that group of men who continue to aggress
against the environment, those who have guilt over
such expression develop chronically higher blood
pressure than those who do not have such guilt.

sented.

75

Self-reported anxiety is not related to the raising or
lowering of diastolic blood pressure levels, or to the
eXpression of aggression.

Several suggestions for future research were pre-

REFERENCES

Alexander, F., and French, T. Studies in Psychosomatic Medicine.
New York: The Ronald Press, 1948.

Anokhin, P. "The Physiological Basis of the Pathogenesis of
Hypertensive States, " Symposium on the Pathogenesis
of Essential Hypertension. Edited by J. Cort, V.
Fencl, Z. Hejl, and J. Jirka. Prague: State Medical
Publishing House, 1961, pp. 199-222.

 

 

Ax, A. F. "The Physiological Differentiation between Fear and
Anger in Humans, " Psychosom. Med. , 1953, 15, 433-442.

 

Beeman, E. A. "The Effect of Male Hormones on Aggressive
Behavior in Mice, " Physiol. Zool. , 1947, _2_0_, 373-405.

 

Berkowitz, L. "Aggressive Cues in Aggressive Behavior and
Hostility Catharsis, " Psych. Rev. , 1964, 11, 104-122.

 

Best, C. H., and Taylor, N. B. The Livirg Body. New York:
Henry Holt and Company, 1958.

 

Brod, J. "Haemodynamic Response to Stress and Its Bearing on
the Haemodynamic Basis of Essential Hypertension, "
Symposium on the Pathogenesis of Essential Hyperten-
sion. Edited by J. Cort, V. Fencl, Z. Hejl, and J.
We. Prague: State Medical Publishing House, 1961,
pp. 256-264.

Brower, D. "The Relation between Certain Rorschach Factors and
Cardiovascular Activity Before and After Visuo-motor
Conflict," J. Gen. Psychol., 1947, g, 93—95 (a).

 

Buss, A. H. The Psychology of Aggression. New York: Wiley,
1961.

Calhoun, J. "POpulation Density and Social Pathology, " _S_c_:i_.
American, 1962, 206, 139-148.

 

76

77

Cannon, W. B. The Wisdom of the Body. New York: Norton,
1932.

 

Cattell, R. B., Saunders, D. R., and Stice, G. Handbook for
the 16 Personality Factor Questionnaire. Champaign,
Illinois: Institute for Personality and Ability Testing,
1957.

 

 

Cattell, R. B., and Scheier, 1. "Extension of Meaning of Objective
Test Personality Factors: Especially into Anxiety,
Neuroticism, Questionnaire, and Physical Factors, "
J. Gen. Psychol., 1959, _6__1_, 287-315.

 

Clark, G. , and Birch, H. C. "Hormonal Modifications of Social
Behavior. 1. The Effect of Sex Hormone Administra-
tion on the Social Status of a Male Castrate Chimpanzee, "
Psychosom. Med. , 1945, 1, 321-329.

 

Cottrell, J. "Opening Address, " Symposium on the Pathogenesis
of Essential Hypertension. Edited by J. Cort, V. Fencl,
Z. Hejl, and J. Jirka. Prague: State Medical Pub-
lishing House, 1961, pp. 15-19.

 

 

Cranston, R. , Chalmers, J., Taylor, H., Henschel, A., and
Keys, A. "Effect of a Psychiatric Interview on the
Blood Pressure ReSponse to Cold Stimuli, " Federat.
Proc. , 1949, 8, 30.

Damon, A. "Constitution and Smoking, " Science, 1961, 34, 339-
340.

Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., and
Sears, R. R. Frustration and Aggression. New Haven:
Yale, 1939.

 

Edwards, A. L. Experimental Design in Psychological Research.
New York: Holt, Rinehart, and Winston, 1962.

 

Fejfar, Z. , and Widimsky, J. "Juvenile Hypertension, " Symposium
on the Pathogenesis of Essential Hypertension. Edited
by J. Cort, V. Fencl, Z. Hejl, and J. Jirka. Prague:
State Medical Publishing House, 1961, pp. 33-42.

 

 

Feshbach, S. "The Function of Aggression and the Regulation of
Aggressive Drive, " Psych. Rev. , 1964, 11, 257-272.

 

78

Fisher, S. , and Cleveland, S. "Relationship of Body Image to
Site of Cancer, " Psychosom. Med. , 1956, 1_8_, 304-309.

 

Gambaro, S. "Blood Pressure Reactions to Expressed and Un-
expressed Anger in Low Guilt and High Guilt Subjects. "
1966. (In press.)

Gellhorn, E. , and Loofbourrow, G. Emotions and Emotional Dis-
orders. New York: Harper and Row, 1963.

 

Grinker, R. Psychosom. Research. New York: W. W. Norton
and Co., 1953.

 

Groen, J. J. Psychosom. Research. New York: The MacMillan
Co., 1964.

 

Hamilton, J. A. "PsychOphysiology of Blood Pressure, " Psycho-
som. Med., 1942, 4, 125-133.

 

Hokanson, J. "Vascular and Psychogalvanic Effects of Experi-
mentally Aroused Anger, " J. Pers. , 1961, _2_9, 30-39.

Hokanson, J. , and Burgess, M. "The Effects of Three Types of
Aggression on Vascular Processes, " J. Abnorm. Soc.
Psychol., 1962 (a), 64, 446-449.

 

Hokanson, J. , and Burgess, M. "The Effects of Status, Type of
Frustration and Aggression on Vascular Process, "
J. Abnorm. Soc. Psychol., 1962 (b), 6_5, 232-237.

 

Hokanson, J., Burgess, M. , and Cohen, M. "Effect of Dis-
placed Aggression on Systolic Blood Pressure, "
J. Abnorm. Soc. Psychol. , 1963, 61, 214-218.

 

Hood, B. "Diagnostic Problems in Hypertension: The Changing
Picture of the Disease after a Decade of Active Treat-
ment, " Symposium on the Pathogenesis of Essential
Hypertension. Edited by J. Cort, V. Fencl, Z. Hejl,
and J. Jirka. Prague: State Medical Publishing House,
1961, pp. 21-32.

Innes, G., Millar, W. M., and Valentine, M. "Emotion and Blood
Pressure, " J. Ment. Sci., 1959, 105, 840-851.

 

Kaufman, H. "Definitions and Methodology in the Study of Aggres-
sion, " Psych. Bull. , 1965, 64, 351-364.

 

79

Lacey, J. I. , and Van Lehn, Ruth. "Differential Emphasis in
Somatic ReSponse to Stress, " Psychosom. Med. , 1952,
14, 71-81.

 

Lacey, J. I., Bateman, Dorothy E., and Van Lehn, Ruth. "Auto-
nomic ReSponse Specificity: An Experimental Study, "
Psychosom. Med., 1953, 1_5_, 8-21.

 

Lorenz, K. On Aggression. New York: Harcourt, Brace, and
World, 1966.

 

Magoun, H. W. "Caudal and Cephalic Influences of the Brainstem
Reticular Formation, " Physiol. Rev. , 1950, E, 459-
474.

 

Marston, W. M. "Systolic Blood Pressure Symptoms of Deception, "
J. Exp. Psychol., 1917, g, 117-163.

 

McDermott, W., Deuschle, K., Adair, J., Fulmer, H., and
Loughlin, B. "Introducing Modern Medicine in a
Navajo Community, " Science, 1960, 131, 280-287.

Matarazzo, J. D. "An Experimental Study of Aggression in the
Hypertensive Patient, " J. Pers. , 1954, 22, 423—447.

McGinn, N., Harburg, E., Julius, S., and McLeod, J. "Psycho-
logical Correlates of Blood Pressure, " Psych. Bull. ,
1964, 61, 209-219.

 

Miall, W. ”The Epidemiology of Essential Hypertension, "
Symposium on the Pathogenesis of Essential Hyper-
tension. Edited by J. Cort, V. Fencl, Z. Hejl, and
J. Jirka. Pregue: State Medical Publishing House,
1961, pp. 88-102.

 

Miasnikov, A. "The Significance of Disturbances of Higher Nervous
Activity in the Pathogenesis of Hypertensive Disease, "
Symposium on the Pathogenesis of Essential Hyperten-
sion. Edited by J. Cort, V. Fencl, Z. Hejl, and
fiirka. Prague: State Medical Publishing House,
1961, pp. 153-162.

 

Milton, C. A. "Changes in Essential Hypertension Over Time: A
Factor Analytic Case Study, " J. Clin. Psych. , 1961,
_11, 322—326.

 

8O

Montagu, A. Aggressive Behavior of Man. Chicago: Chicago
Tribune, 1966, p. 10.

Mosher, D. L. "The Development and Validation of a Sentence
Completion Measure of Guilt. " Unpublished doctoral
dissertation, Ohio State University, 1961.

Neiberg, N. A. "The Effects of Induced Stress on the Management

of Hostility in Essential Hypertension, " Dissertation
Abstr. , 1957, H, 1957.

 

 

Oken, D. "An Experimental Study of Suppressed Anger and Blood
Pressure, " A.M.A. Arch. Gen. Psychiat., 1960, _2_,
441-456.

Peterson, L. "The Mechanical Properties of the Blood Vessels
and Hypertension, " Synflmsium on the Pathogenesis of
Essential Hypertension. Edited by J. Cort, V. Fencl,
Z. Hejl, and J. Jirka. Prague: State Medical Pub-
lishing House, 1961, pp. 295-313.

Pickering, G. "The Concept of Essential Hypertension as a Quanti—
tative Disease, " Symposium on the Pathogenesis of
Essential Hypertension. Edited by J. Cort, V. Fencl,
Z. Hejl, and J. Jirka. Prague: State Medical Pub-
lishing House, 1961, pp. 43-54.

Saul, L. "Effects of Psychoanalytic Therapy, " Studies in Psycho-
somatic Medicine. Edited by F. Alexander and T.
French. New York: Ronald Press, 1948, pp. 91-100.

 

Schachter, J. "Pain, Fear and Anger in Hypertensives and Normo-
tensives," Psychosom. Med., 1957, 19, 17-29.

 

Schroeder, H. "Influence of Finished Water Supplies on the
Epidemiology of Cardiovascular Disease, " Symposium
on the Pathogenesis of Essential Hypertension. Edited
by J. Cort, V. Fencl, Z. Hejl, and J. Jirka. Prague:
State Medical Publishing House, 1961, pp. 81-87.

 

Selye, H. , and Bayusz, E. "Adrenal Cortex in Essential Hyper-
tension, " Symposium on the Pathogenesis of Essential
Hypertension. Edited by J. Cort, V. Fencl, Z. Hejl,
and J. Jirka. Prague: State Medical Publishing House,
1961, pp. 365-372.

81

Shapiro, A. P., and Horn, P. W. "Blood Pressure, Plasma
Pepsinogen, and Behavior in Cats Subjected to Experi-
mental Production of Anxiety, " J. Nerv. Ment. Dis. ,
1955, _1_2_2_, 222-231.

 

Siegel, S. Non-parametric Statistics. New York: McGraw—Hill
Co., 1956.

 

Storment, Charlyne Townsend. "Personality and Heart Disease, "
Psychosom. Med., 1951, 1_3_, 304-313.

 

White, Robert. The Abnormal Personality. New York: The Ronald
Press Company, 1956.

 

White, J. , and Smithwick, R. The Autonomic Nervous System.
New York: MacMillan Co. , 1941.

 

Wilson, C. "The Kidney and Essential Hypertension, " Symposium
on the Pathogenesis of Essential Hypertension. Edited
by J. Cort, V. Fencl, Z. Hejl, and J. Jirka. Prague:

- State Medical Publishing House, 1961, pp. 405-417.

 

 

APPENDIX A

These are the 25 MIST items employed in this study,
taken from the original 50 items used by Mosher. Those items
with the number underlined were those rated in order to obtain

the aggression-guilt score.

1. When I tell a lie .

2._ The idea of murder is . . . .

3'. When someone swears at me . . . .
4. Most women who curse . .

_5_. After a childhood fight, I felt . .

6. My father. . . .

_7_. After an argument . . . .
8. When I have sexual dreams . . . .
9. To kill in a war . .

10. Arguments leave me feeling . . . .
11. Most women .

12. If I robbed a bank . . . .

13. When anger builds inside me . . .

14. Amother. . . .

82

83

If I felt like murdering someone . . . .

I have .

I should have been punished for . . . .

After an outburst of anger . . . .
One Should not . . . .

When I was younger, fighting . . . .
A guilty conscience . . . .
Capital punishment. . . .

If I hated my parents . . . .

If I committed a homosexual act . . . .

If I killed someone in self-defense . . . .