Vasovagal Fainting: A Diphasic Response
DAVID T. GRAHAM, M.D., J D KABLER, M.D,,
and LEWIS LUNSFORD, Jr., M.D.
VASOVAGAL fainting, also called vaso- diograms made during a faint have been
depressor fainting, is a common and dra- published.
matic reaction frequently occurring as a re- The literature contains little discussion
sponse to an emotionally disturbing event. of an apparent paradox: Situations which
The pathophysiology of the faint itself has provoke fainting seem in general to be
been recently reviewed by Edholm7 and by those likely to arouse anxiety, yet the heart
Sharpey-Schafer.19 The cardinal manifesta- rate and blood pressure in anxiety are ex-
tions are low blood pressure and slow heart pected to be raised rather than lowered, a
rate. change exactly opposite to that seen in the
The cardiovascular events which precede faint. Wolf21 alludes to this difficulty.
the faint have received very little attention, This report may resolve the paradox, by
however, and their importance has not been showing that vasovagal fainting is a dipha-
recognized. The few published protocols sic response. From casual observation and
showing prefaint events are concerned with from occasional newspaper accounts of
such special maneuvers as postural changes, faints, it appears that a faint occurs most
hypoxia, or significant blood loss. Except commonly immediately after a threatening
for the Greenfield11 and Engel-Romano8" experience has ended. Thus it may be that,
reports, there are apparently no published in response to a threat, an emotional state
records of data obtained at short intervals properly called anxiety occurs, of which
preceding the onset of emotionally induced rapid heart rate and rising blood pressure
fainting. Even the physiology of the faint are a part; and with cessation of the threat,
seems not to be as familiar as would be ex- there is relief from anxiety, with the fall in
pected from its frequency. Many physicians heart rate and blood pressure which are the
are unaware that asystole is common and chief manifestations of vasovagal fainting.
that when it occurs convulsions may also The data which follow are evidence that
develop. It is also not widely known that such a diphasic response is, in fact, charac-
cardiac arrhythmias may occur as part of teristic of vasovagal fainting.
the faint, perhaps because few electrocar-
Plan of the Experiment
From the Department of Medicine, University of
Wisconsin Medical School, Madison 6, Wise. Fainting was studied: (1) in persons giv-
Supported by research Grant M-2014 from the ing blood for a hospital blood bank; (2)
National Institute of Mental Health, U. S. Public in persons having a simple venipuncture
Health Service, and by a grant from the Universal without significant blood loss; and (3) in
Match Foundation, St. Louis, Mo.
We wish to thank Olivia J. Pautler for technical patients undergoing pneumoencephalogra-
assistance. phy. Particular care was taken to obtain
Received for publication Sept. 26, 1960. measurements of cardiovascular variables
VOL. XXIH, NO. 6, 1961
494 VASOVAGAL FAINTING
at intervals not longer than about 3 min. comparisons between fainters and non-
for at least 10 min. preceding the onset of fainters, blood donors who seemed most
fainting. likely to faint were selected. This pro-
The variables recorded in all three cedure tends to minimize differences be-
phases of the study were systolic and dia- tween fainters and nonfainters among the
stolic blood pressure and electrocardiogram selected subjects; differences were never-
changes. The blood pressures were obtained theless found. Various more or less intui-
by means of a recording sphygmomanome- tive criteria were used at first in the choice
ter, which responds to sounds from the of subjects, but as the companion study of
brachial artery by means of a microphone fainting prediction10 progressed, improved
placed in the antecubital fossa. Blood pres- criteria for selection were used. In general,
sure readings were obtained about every those selected were young, with rapid heart
21/2-3 min. This determination sometimes rates and an appearance of anxiety. For the
took as long as 30 sec, so that during very sake of greater homogeneity, only males
rapid changes, such as those that occur at were used. The study was continued until
the beginning of a faint, it was not always 10 faints had been recorded.
possible to obtain both systolic and dia-
stolic readings. In general, the electrocai'- Procedure
diographic tracing was standard lead 2, in
order to obtain most satisfactory evidence Prospective donors came into the room
of changes in heart rhythm. This was usu- where blood was drawn, and lay down on
an examining table. Pulse rates and blood
ally recorded on a portable clinical electro-
cardiograph, although in a few instances pressures were taken by the usual methods,
ot venipuncture fainting, the tracing was after which the venipuncture and blood
recorded on an eight-channel polygraph, drawing proceeded. Differences from the
which was also being used for recording of usual routine introduced by this study for a
respiratory rate and skin temperature. It selected donor consisted of attaching the
faint seemed imminent, the electrocardio- sphygmomanometer cuff and microphone
graph was run continuously; otherwise, and the three electrocardiographic elec-
short records were obtained every 2i/2-3 trodes.
min. Thirty-two donors were studied. Of
these, 10 fainted, according to the criteria
For the purposes oi this investigation, a
vasovagal taint was defined as a sudden given above. None of the donors who did
drop in blood pressure and pulse rate, ac- not faint reported any significant symp-
companied by a report by Lhe donor ol toms. It, therefore, appears, for this and
some disturbance ol consciousness, ex- other reasons, that nearly all the so-called
picssed in such words as "dizzy," "light- reactions occurring in blood donors are in
neaded," and '"woozy." Some ot the sub- fact vasovagal faints, although there may
jects lost consciousness completely, but itbe admixtures of other phenomena, such as
hyperventilation. A survey of the litera-
is unrealistic to insist on complete loss of
consciousness as a criterion of fainting, ture, as well as experience with other "re-
s.nce it is clear, as will be shown below, actors" who were not as intensively fol-
that tne vasovagal faint is not an all-or-lowed as those in the present group, sup-
nothing reaction but occurs in various de- ports this conclusion.
grees. After the apparatus was in place, a con-
trol period of 6—12 min. elapsed before the
actual venipuncture. The time was always
Blood-Donor Faints long enough to permit a minimum of three
Since the primary purpose was to study control readings at intervals of about 3
the course of fainting rather than to make min. before the donor was again touched
PSYCHOSOMATIC MEDICINE
GRAHAM ET AL. 495
in any way. Longer periods were either the mean changes in systolic, diastolic, and
result of technical difficulties with the ap- mean arterial pressures and in the heart
paratus or necessitated by the demands of rate for subjects in both these groups. The
the routines of the blood-donor room. mean blood pressure was estimated by the
After the venipuncture had been performed formula
and blood had begun to flow, records con- „ ,. systolic — diastolic
tinued to be made at about 3-min. intervals Diastolic + -
(or more frequently if the donor appeared 3
about to faint), until the donation had It is clear that there is a distinct differ-
been completed. ence between the two groups in the dia-
The donor was allowed to sit up a few stolic and mean blood pressure changes.
minutes after the end of the donation if The fainters showed a rapid rise in dia-
there were no signs of a u ction. A final stolic pressure in the 8 min. preceding the
set of readings was taken after he had sat venipuncture; the mean rise for the group
up, and if there were still no evidences of of 10 was 11.1 mm. Hg. This differs sig-
reaction, the apparatus was disconnected. nificantly (P cfSrpnTus situation, comparisons be-
VOL. xxm, NO. 6, 1961
500 VASOVAGAL FAINTING
FINGER
STICK
Warning Needl* "1 f M l faint"
1 i 150
1 Fig. 4. Systolic and diastolic
pressures and heart rate in a male
subject before, in, and during re-
\j 140
covery from a faint induced by
venipuncture with puncture of fin-
130
ger for blood count.
v
-^ ,/l20
1
no
100
1 /
90
80
n-
^""~*N^^—^ 70
\i hj /
60
50
\w / r
Blood Presiure
Heart Rate
45
30
\ /
0 10 20 30
Time in Minutes
tween fainters and nonfainters are harder more important in terms of the definition
to make than with the blood donors. In of fainting, they did not have any general-
the latter situation, the subject's attention ized symptoms. The occurrence of these
was not focused on the manipulations in- mild changes demonstrated again that defi-
cidental to the procedure; with investigative nition of fainting must be somewhat arbi-
venipuncture, the manipulations necessary trary.
for physiological measurement assume cru-
cial importance, and the entire course be- Pneumoencephalography Faints
comes less predictable.
Systematic comparisons of fainters with Disturbance of consciousness is a com-
nonfainters in the simple venipuncture mon occurrence with diagnostic pneumo-
situation have, therefore, not been made. encephalography. To determine if this was
A number of those subjects who did not vasovagal fainting, and if so, whether a di-
faint showed a drop in diastolic pressure phasic cardiovascular pattern was present,
and pulse or heart rate after venipuncture, blood pressure records and electrocardio-
but not one of sufficient degree to justify grams were obtained in 15 patients during
calling the response a vasovagal faint, and this procedure. Six of these, including 2
PSYCHOSOMATIC MEDICINE
GRAHAM ET AL. 501
women, had reactions that included some also factors in the manipulation of the pa-
disturbance of consciousness, bradycardia, tient that contributed to the difficulty of
and lowering of the blood pressure, and standardization.
therefore satisfied the criteria for vasovagal
fainting. Results
All of the faints showed the same dipha-
Procedure sic pattern seen in the other 2 situations.
Even more than with the simple veni- Figure 5 shows the readings obtained from
puncture faints, standardizing this investi- a 31 -year-old white man, who showed the
gation was almost impossible. Pneumo- most striking response of any of the group,
encephalography, of course, is a compli- partly because it was delayed long enough
cated procedure and there are at least three to permit good prefaint records. There
critical points from the patient's point of were 2 slight drops in diastolic pressure,
view: (1) when procaine is injected into after the completion of the spinal puncture
the skin prior to lumbar puncture; (2) and after the injection of air, respectively.
when the lumbar puncture is performed; However, the faint occurred only after he
and (3) when the injection of air is com- had been told that the procedure was over.
pleted. Times for the various stages were
extremely variable. The requirements of Comment
the clinical procedure sometimes interfered Although it is impossible to exclude the
with obtaining readings at the best times presence of air in the cerebrospinal fluid
for the purposes of the study. There were space as an important contributing factor
Blood Pressure
Pulse
10 20 30
Tim* in Minutes
Fig. 5. Cardiovascular variables during a faint occ.irring with pneumoencephalography. "Pulse" in this
figure represents heart rate.
VOL. xxiii, NO. 6,1961
502 VASOVAGAL F A I N T I N G
in the reaction, nevertheless the data sug- ized convulsive movements occurred; con-
gest that the reactions of the pneumoen- vulsions were never seen without asystole.
cephalography patients were vasovagal Restoration of heart beat occurred during
faints and that the same emotional factors the convulsion on each occasion. Figure 6
were operating as in the other 2 situations. shows 18 sec. of asystole in a blood donor;
It is not critical that the reaction in this Fig. 7, interference dissociation and nodal
situation be accepted as emotional; the ma- beats in another blood donor, and Fig. 8,
jor point is that it is vasovagal and that it asystole in a simple-venipuncture fainter.
is the second limb of a diphasic response. Respirations were counted in 4 of the
However, it is true that the pre-lumbar venipuncture faints by means of a strain
puncture cardiovascular changes observed gauge attached to an elastic belt around
were like those seen prior to venipuncture the chest. This is not very satisfactory, since
in the other situations. Thus, the same rea- changes in the mechanics of breathing, and
sons for believing that anxiety is involved irregular respiratory and other skeletal
apply here also. muscle movements, such as occur in anxious
subjects, may make interpretation of the
Special Characteristics of Faints record difficult. However, in 2 of the veni-
In many instances, the slowing of the puncture fainters, there were periods of 10
heart rate represented more than simple and 12 sec. respectively, coinciding roughly
sinus bradycardia. with the periods of cardiac arrest, during
Five of the 10 blood-donor fainters which there were no definite respiratory
showed nodal rhythm, interference dissocia- excursions. It, therefore, appears that
tion, or both. Three of the fainters with apnea may be part of a severe faint.
simple venipunctures had the same dis- It was not necessary for the subject to be
turbances, as did 4 of the pneumoen- sitting or standing to faint. The most
cephalography fainters. severe faint seen occurred in a supine sub-
More striking was the occurrence of ject.
asystole for periods ranging from 4y2 to 18
sec. Three instances of asystole were seen Discussion
with simple venipuncture, none with pneu- Human reactions are such that irregulari-
moencephalography, and 2 with blood ties in either the rising or falling phase of
donations. In 4 of these instances, general- a diphasic reaction may make it difficult
Fig. 6. Electrocardiogram
showing 18 sec. of asystole in a
faint during blood donation.
-• The faint and asystole were
terminated by convulsive move-
ments. Subject remained supine
throughout the procedure.
1 i T,! 1
•f
T
*
PSYCHOSOMATIC MEDICINE
GRAHAM ET AL. 503
Fig. 7. Electrocardiogram showing interference dis ocialion and A. V. nodal beats in a faint dtiring blood
donation.
to determine whether one is observing any- the needle enters the vein or the time at
thing more than random variations. There which it is removed is critical. Since the
are always fluctuations in the graphs of the former seemed to be more commonly the
variables for any subject, so that, by arbi- case, it was taken as the reference point for
trarily selecting the segment of the graph the blood donors. This procedure under-
dealt with, it is possible to "produce" a di- states the significance of the results, since
phasic effect. The problem can be solved it does not allow, for instance, for the
by finding: (1) an obvious "turn-around donor who continues to show the first-phase
point," at which the change from the rising pattern throughout the period of blood
to the falling phase occurs (the falling phase drawing.
being essentially uninterrupted as it pro- In several instances, the onset of the
ceeds into the faint) or (2) a point that second, or fainting, phase began at the
can be designated for all subjects as the criti- moment when the subject's arm was first
cal point. The latter was the case in our re- touched in preparation for the venipunc-
cordings from blood donors and, usually, in ture. It is possible to consider such in-
those from simple venipunctures. stances as evidence that the proposed in-
The hypothesis guiding this investiga- terpretation of the diphasic response is
tion states not only that the fainting re- wrong and that the second phase should
sponse is diphasic, but also that the turn- be looked on simply as anxiety which has
around point occurs when the threat dis- passed some critical point of intensity. This
appears; in other words, at the time the alternative interpretation would fit some of
subjects's reaction changes from anxiety to the faints seen in this study and some seen
relief from anxiety. In venipuncture or in daily life. A frequently cited observa-
blood donation, it is natural to assume tion is that of the man waiting in line for
that, in general, either the time at which an injection who faints when watching
VOL. xxm, NO. 6,1961
VASOVAGAL FAINTING
Fig. 8. Electrocardiogram showing asystole in a faint during venipuncture. Obvious muscle tremor was
present before the faint. Asystole and faint were terminated by convulsive movements.
someone ahead of him receive one. It is cephalography patient who fainted only
sometimes felt that such an occurrence can after he was told the procedure was over, as
only be understood as the consequence of well as those of the blood donation or veni-
mounting anxiety, since there has been no puncture subjects who had the first evi-
opportunity for relief of anxiety to occur. dence of fainting when the procedure
However, it may be that the fainter identi- ended, are hardly compatible with any view
fies with the other person, so that the except that offered by our hypothesis. One
carrying out of the threat to the latter is could, therefore, invoke two different ex-
reacted to as if the event were happening planations, one for the apparently aberrant
to himself. There are many observations of cases, and another for the majority. On
this kind of reaction, e.g., the fainting of the grounds of parsimony, the use of one
medical students at surgical procedures explanation for all is preferable.
where there is no real threat to the fainter The biological interpretation of vaso-
at all. vagal fainting offered by Engel,8 who
One of the venipuncture subjects fainted studied it intensively, has aroused much in-
when the blood pressure cuff was applied terest. He proposed that such fainting is
and before there had been a threat of veni- to be understood as a result of preparation
puncture. In this more deviant instance, for running when, in fact, running is not
one can suppose that the application of the carried out. This interpretation is based
cuff represented to him the carrying out in part on the demonstration by Barcroft
of a threat. et al.,2 that the mechanism of the drop in
Faints, such as that of the pneumoen- blood pressure is vasodilatation in muscle,
PSYCHOSOMATIC MEDICINE
GRAHAM ET AL. 505
mediated by sympathetic vasomotor nerves; effect, saying to himself, "I'm dead." Al-
thus, it appeared logical that such vaso- though this seems plausible where the
dilatation would be important in increas- threat is actually carried out, it is not so ap-
ing the blood flow to the muscles used in plicable to faints occurring when a threat is
running. suddenly removed without being carried
There are several objections to Engel's out. Many everyday observations support
interpretation in its simplest form. (1) It this concept of "fainting with relief."
takes no account of the bradycardia, a We suggest that the faint can be under-
reaction that cannot be regarded as desir- stood, on the physiological level, as the con-
able for running; (2) the occurrence of sequence of the sudden cessation of those
fainting after the threat has either been physiological processes which support the
carried out, or can be seen to be safely hyperdynamic first phase. These almost
passed, is hard to call preparation for certainly call forth antagonistic reflexes
running (Even if all vasovagal fainting is which act to prevent the pulse rate and
not a postthreat occurrence, there are too blood pressure from rising without control.
many cases in which it is clearly such for Then, if the hyperdynamic processes cease
the preparation theory to be generally abruptly, the opposing reflex mechanisms
valid.); (3) blood flow is not necessarily will be suddenly unopposed, and fainting
increased in the tense muscles of an organ- will occur. In the case of heart rate, vagal
ism preparing for, but not carrying out effects which have previously only modified
action. (It is decreased if contraction is sus- the tachycardia will now, without sympa-
tained at 20 per cent or more of maximal thetic opposition, produce bradycardia or
contraction.)2 Nevertheless, if the entire re- even asystole. Except for the reports of
sponse of two phases is considered the in- Dermksian and Lamb, 5 ' e little attention
hibited running theory may be applicable has been paid to disturbances of heart
to the first phase rather than the second. rhythm in fainting. They suggest that
Engel does comment that "sometimes" a respiratory stretch reflexes are important
diphasic response may be seen, but his in in mediating the arrhythmias but neither
terpretation of it is different from ours their work nor ours indicates that these re-
We suggest that the diphasic response is flexes are always of primary importance.
always present. Relevant reflex cardiovascular responses
The diphasic nature of the process mus( have been discussed by other authors.1' 4-12
be recognized if meaningful interpretations It is striking that asystole, fall in blood
are to be made. It resolves the dilemma pressure (not purely a consequence of the
mentioned by Wolf21 of relating fainting to asystole), and apnea present a picture indis-
anxiety and simplifies the physiological tinguishable from death. There is reason
questions raised by Edholm7 in classifying to think that vasovagal fainting may some-
the symptoms and signs of fainting. times be fatal.8' ^ The many sudden catas-
trophes developing in connection with
It is not clear whether some special psy- anxiety-inducing medical procedures may
chological or psychodynamic interpretation be examples of this phenomenon. Death
should be given to the faint itself. Clearly, with surgical anesthesia, where there has
the first phase deserves the name "anxiety," been a stormy induction, probably fits this
and is a response to a threat. It is possible diphasic pattern. In this case the "relief"
to look on the faint as simply the physio- may be provided by pharmacologic sup-
logical expression of the sudden cessation pression of the activity of the centers medi-
of anxiety. An alternative is that the stim- ating the hyperdynamic phase.
ulus which induces the second phase (the
faint) means more than simply the end of The observations of Richter18 on sudden
a threat. It might, for instance, be a sym- vagal death in wild rats swimming in cold
bol that all is lost, so that the subject is, in water or held in the hand appear to fit
VOL. xxin, NO. 6, 1961
506 VASOVAGAL FAINTING
very well with the ideas here presented, al- may be that the diphasic pattern will be
though not enough details of the protocol found in many other responses and diseases.
are given to be certain. It is possible, how-
ever, that his animals were first frightened Summary and Conclusions
by hair-dipping or other procedures and
that the subsequent immersion in cold wa- 1. Vasovagal (vasodepressor) fainting
ter may have been a stimulus situation for was studied in three different situations: in
persons donating blood, in persons having
them tantamount to insertion of a veni- a simple venipuncture, and in patients un-
puncture needle for a man. dergoing pneumoencephalography. Ten
There are features of interest in the faints were observed in the first situation,
physiology of the first phase, the phase we 7 in the second and 6 in the third.
have considered to be part of anxiety. The 2. In all instances, the faint itself, char-
pattern of rising diastolic pressure, rapid acterized by low blood pressure and brady-
heart rate, and steady or even falling sys- cardia, was the second phase of a diphasic
tolic pressure appears to be what is referredresponse; the first phase was characterized
to by Laurellir> as orthostatic arterial ane-by rapid or rising heart rate, and by rising
mia, and more recently discussed by Hick- blood pressure, especially diastolic.
ler e.t al.'13 Although these authors describe 3. The data are compatible with the
this picture as induced only by standing, it view that the first (hyperdynamic) phase
appears not to differ from that seen in some is a reflection of anxiety, while the second
of our subjects while they were lying-down, phase begins with the sudden cessation of
and in whom it appears to have been fre- anxiety. It is suggested that physiologically
quently a prelude to vasovagal fainting. the faint reflects the action of reflex mecha-
Protocols of faints involving significant nisms activated by the first phase, and then
blood loss and those induced by hypoxia,- left suddenly unopposed. The other psy-
indicate that these also fit the diphasic pat-
chobiological interpretations exhibit incon-
tern, and this may be true of all vasovagal sistencies because they fail to recognize the
fainting, however induced. diphasic nature of this response.
2
Barcroft ct al. have presented data in- 4. Fainting and dying resemble each
dicating that the cardiac output does not other closely, and there is reason to think
fall during a faint, although it may pro- that some vasovagal faints are fatal. Car-
gressively drop before a faint. This drop diac asystole is not rare in faints and is
corresponds to what Laurell15 reports, and often associated with convulsions. Apnea
it may be peculiar to situations in which was observed in 2 of our subjects.
the subject is upright. However, it is obvi- 5. Disturbances of cardiac rhythm, in-
ous that in severe faints with asystole the cluding A. V. nodal rhythm, interference
cardiac output must fall to zero. The find- dissociation, and asystole (in 1 instance for
ings of Barcroft ct al. may be due to diffi- 18 sec.) were observed. Generalized con-
culty in measuring cardiac output at such vulsions were seen on four occasions, al-
short intervals. Inspection of their graph ways in association with asystole.
suggests that they did not determine the University Hospitals
output until after the recovery from the 1300 University Ave.
faint had begun. Madison 6, Wise.
The data presented in this paper indi-
cate that the vasovagal fainting response References
consists in its entirety of two successive 1. AVIADO, I). \f., JR., and SCHMIDT, C. F. Reflexes
phases, of opposite direction. It has been from stretch receptors in blood vessels, heart
Physio}. Rev. .J.5.-247,
suggested that a pattern of this kind is in- 2. and lungs. H.. and SWAN, H. J. C. 1955.
BARCROFT, Sympathetic
volved in the reactions leading to mi- Control of Human Blood Vessels. Arnold,
14 16
graine," asthma, and acne vulgaris. It London, 1953.
I -\ : i:os-)MM!C MEDICINE
GRAHAM ET AL. 507
BRIODEM, W., HOWARTII, S., and SHARPEY- GRUHZIT, E. L., I'REYBURGER, W. A., and MOF.,
S< HAFKR, E. P. Postural changes in the periph- G. T . T h e nature of the reflex vasodilation
eral blood-flow of normal subjects with obser- induced by epinephrine. J. Pharmacol, &
vations on vasovagal fainting reactions as a re- Exper. Therap. 7/2.138, 1954.
sult of tilting, the lordotic posture, pregnancy HICKLER, R. B., WF.LLS, R. F.., JR., TVI.FR,
and spinal anesthesia. Clin. Sc. 9:19, 1950. R. H., and HAMLIN, J. T., III. Plasma catecho-
DAWES, G. S., and COMROF, J. H., JR. Chemore- lamine and electroencephalographic responses
flexes from the heart and lung. Physiol. Rev. to acute postural change. Am. J. Med. 26:410,
J-/.167, 1954. 1959.
DERMKSIAN, G., and LAMB, L. E. Syncope in KNAPP, P. H., and NF.MK.TZ. S. J. Acute bron-
a population of healthy young adults. J.A.M.A. chial asthma: I. Concomitant depiession and
76S.12OO, 1958. excitement, and varied antecedent patterns in
DKRMSKIAN, G., and LAMB, L. E. Cardiac 406 attacks. Psychosom. Med. 22:42, 1960.
arrhythmias in experimental syncope. JAM.A. LAUKF.I.L, H. Die "orthostatische arteriellc
/6.S.-1623, 1958. Anamie," ein gewohnliches aber oft fehlesje-
EDHOI.M, O. G. Physiological changes during deutetes Krankheitsbild. Fortschr. Geb. Ro'nt'
fainting, In Ciba Foundation Symposium on gensttahlen 53:501, 1936.
Visceral Circulation. Little, Boston, 1957. LORF.NZ, T . H., GRAHAM, D. T., and WOI.F, S.
ENGFX, G. L., Fainting, Physiological and Psy- The relation of life stress and emotions to hu-
chological Considerations. Thomas, Spring- man sebum secretion and to the mechanism of
field, 111., 1950. acne vulgaris. / . Lab. ir Clin. Med. 41:11, 1953.
. ENCF.L, G. L., and ROMANO, J. Studies of syn- POLES, F. C , and BOYCOTT, M. Syncope-'in
cope: IV. Biologic interpretation of vasodepres- blood donors. Lancet 2:531, 1942.
sor syncope. Psychosom. Med. 9:288, 1947. RICHTF.R, C. P. On the phenomenon of sudden
GRACE, W. J., and GRAHAM, D. T. Relation- death in animals and man. Psychosom. Med.
ship of specific attitudes and emotions to cer- 79:191, 1957.
tain bodily disease. Psychosom. Med. 14:243, Sn\RPFY-Scn\FER, E. P. Syncope. Brit. M. J.
1952. 1:506, 1956.
GRAHAM, D. T . Prediction of fainting in blood SIMPSON, K. Deaths from vagal inhibition.
donors. Circulation 2.?:901, 1961. Lancet 2:558, 1949.
GREENFIELD, D. M. An emotional faint. Lancet WOLF, S. Cardiovascular reactions to symbolic
I:'M)'Z, 1 9 5 1 . stimuli. Circulation /A':287, 1958.
VOL. XXIII, NO. 6, 1961