Cortisol Excretion During the Defense Reaction

in Humans
M A T S FREDRIKSON, P H D , O R J A N S U N D I N , P H D ,
AND M A R I A N N E F R A N K E N H A E U S E R , P H D

Twelve subjects with specific phobias were exposed to slides with a phobic or a neutral content
while cortisol excretion, electrodermal activity, and distress-ratings were studied. Six subjects
fearful of blood and mutilation, and six of snakes or spiders, were presented with two sets of
ten different slides (phobic or neutral; 8-sec exposures, separated by 50-sec intervals) with the
order of presentation balanced between days. Before and after each session, subjects rated
feelings of distress, and urine samples were obtained for the determination of cortisol by
radioimmunoassay. Electrodermal activity was recorded before and during slide presentation.
Compared to neutral exposures, phobic slides elicited larger cortisol excretion, higher distress
ratings, and greater skin-conductance responses with slower recovery. No differences between
animal and blood and mutilation phobics were observed. Thus, humans having specinc phobias
exhibit pituitary-adrenal cortical arousal during the defense reaction elicited by slides of their
phobic objects.

Phobics confronted with their phobic

object report fear and display heart-rate
acceleration and skin conductance responses with a slow recovery (15). This
has been interpreted as an example of the
defense reaction described by Hilton (cf.
6) and Sokolov (7) and linked to the sympathetic-adrenal medullary axis with increased excretion of epinephrine and norepinephrine (8). In nonphobic subjects,
states of fear or anxiety have been associated with increased activity also of the
pituitary-adrenal cortical axis, reflected
in high cortisol levels (9). In cats (10) and

From the Psychology Division, Department of Psychiatry and Psychology, Karolinska Institute, and Department of Psychology, University of Stockholm
(M.F.; M.F.) and the Department of Clinical Physiology, University Hospital, Uppsala, Uppsala, Sweden (M.F.; O.S.).
Address reprint requests to: Mats Fredrikson, Department of Psychiatry and Psychology, Karolinska
Hospital, Box 60500, S-104 01 Stockholm, Sweden.
Received November 4,1983; revision received October 2, 1984.

monkeys (11), the defense reaction elicited by implanted hypothalamic electrode

stimulation is accompanied by increased
plasma cortisol levels. The electrodermal
defense reaction is characterized by skin
conductance responses with a slow recovery (12) that also habituate slowly. In the
present study, we used half-time recovery,
as described by Edelberg (12), and skinconductance habituation rate to index the
defense reaction. We asked whether the
defense reaction elicited in phobics confronted by slides of their feared object is
accompanied by elevated urinary cortisol
levels.

MATERIALS AND METHODS

Subjects
Twelve phobics were recruited by means of advertisements in a local paper. Three men and three
women were phobic of blood and mutilation, three
women were phobic of snakes, and three woman were
phobic of spiders. They were paid for their partici-

Psychosomatic Medicine Vol. 47, No. 4 (July/August 1985)

Copyright 1985 by the American Psychosomatic Society, Inc.
Published by Elsevier Science Publishing Co., Inc
52 Vanderbllt Ave., New York, NY 10017

313
0033-3174/8553.30

M. FREDRIKSON, O. SUNDIN, and M. FRANKENHAEUSER

pation in the study and were also offered a fading

treatment. The mean age (SD) was 26.7 [6.7) and
the mean duration of their phobias was 21.0 (8.3)
years. The group was selected using Swedish standardizations (13) of the Snake, Spider, and Mutilation
Questionnaires (SNAQ, SPQ, and MQ, respectively)
(14). Using the Swedish norms, 11 phobics scored in
the top 5% of SNAQ, SPQ, or MQ, and 1 scored in
the top 25% of MQ. All subjects included in the present study had previously visited the same laboratory
four times as part of another study (15).

Procedure and Recording Methods

Subjects attended two sessions between 8 A.M and
9:30 AM on two separate days. Each day they were
shown either ten different neutral slides (flowers) or
ten different slides with a content relevant to their
phobias. Thus, each set of ten slides was shown on
a separate day with the order of presentation balanced between days. Slides were projected by a Prestinox projector to an image size of 60 x 95 cm on a
screen 2 m in front of the subject. Each slide was
shown for 8 sec, with successive presentations separated by a mean intertrial interval of 50 sec, varied
between 45, 50, or 55 sec. The timing was controlled
by an Akai FM-tape player. Subjects were informed
about the nature of these pictures but not informed
whether they would see phobogenic or neutral slides
during their first visit. Thus, all subjects knew the
content of slides exposed the second session. Upon
arrival at the laboratory subjects reported time of last
prior voiding, were asked to void urine (pretest sample), and then served a standardized breakfast consisting of a slice of white bread and hot water. Feelings of distress were rated using an instrument
previously described by Lundberg and Frankenhaeuser (16). Subjects were seated in a comfortable annchair and Beckman skin conductance electrodes (8
mm in diameter, filled with an isotonic paste, 0.4
KCl/dl H2O) were applied to the second phalanges
of the first and second fingers of the subject's right
hand. A Hagforsconstant voltage conductance bridge
was used to measure skin conductance level read
each minute 5 min before and each minute during
an 8-min stimulation period. Skin conductance levels reported are means of readings before and during
stimulation. Skin conductance response magnitude
was scored as a change in skin conductance exceeding 0.05 ^.rnho initiated in the interval 14 sec after
stimulus onset. Recovery rate was scored using halftime recovery, as described by Edelberg (12). Skin
conductance responses were analyzed as magni314

tudes, that is, trials with no responses were entered

as zeros into the analyses. Half-time recovery was
analyzed only with trials having responses entered
into the analyses. In addition, habituation was measured as number of trials to reach a response criterion
of two successive response failures. Before and after
each session subjects were asked to rate feelings of
distress, using an instrument previously described
by Lundberg and Frankenhaeuser (16). After slide
presentation subjects were asked to rate feelings of
distress (16), and were then asked to void urine (posttest sample). Urine samples were stored at - 18C for
later analysis by radioimmunoassay, using a kit manufactured by New England Nuclear (17). Due to technical mishaps pretest cortisol excretion is missing
for one subject in the phobic condition; hence cortisol data from 11 subjects only were analyzed for
this condition. The mean ( SEM) collecting time
for pre- and posttest cortisol samples was 151 (44.7)
and 55 (4.3) min. Pretest diuresis was 1.15 ( + 0.60)
and 0,98 (0.24) ml/min and t(ll) < 1 for phobic
and neutral stimulation, respectively. Posttest diuresis was 1.54 ( 0.56) and 0.97 ( 0.41) ml/min for phobic and neutral stimulation, respectively [t(10)
< 1]. Differences between pre- and posttest diruesis
was nonsignificant both for neutral [t(ll) < 1) and
phobic slide exposure [t(10) = 1.22 NS).

RESULTS

Table 1 gives means ( SEM) and paired

t tests for pretest cortisol excretion and
excretion during exposure to neutral and
phobic slides. Neutral slides did not affect
cortisol excretion, whereas phobic slides
induced a significant increase from rest to
stimulation. Expressed as percentages of
pretest values, phobic slides elicited significantly greater cortisol excretion than
did neutral slides [t(10) = 2.25, p = 0.02,
one-tailed probability] (see Fig. 1). No differences were observed between blood-andmutilation phobics and animal phobics in
cortisol excretion during pretest or stimulation. Ten of eleven phobics had larger
cortisol excretion after than before exposure to phobic slides, and nine of these
showed a greater increase to phobic than

Psychosomatic Medicine Vol. 47, No. 4 (July/August 1985)

DEFENSE REACTION AND CORTISOL EXCRETION

TABLE 1.

Means ( SEM) and Paired f-Tests of Cortisol Excretion Before and After Exposure to
Neutral or Phobic Slides
Before Exposure
(pmol/min)

857.3 (275.8)
559.0 (288 8)

Neutral
Phobic

After Exposure
(pmol/min)
747.3
1024.0

(172.8)
(420.0)

1(10) = 3.42"

'p < 0.01

to neutral slides. Table 2 gives means

( SEM) and paired i tests for pretest skin
conductance levels (SCL) and SCLs during
exposure to neutral and phobic slides.
Neutral slides did not affect SCLs, whereas
phobic slides elicited an increase. Expressed as percentages of pretest values,
phobic slides elicited greater SCLs than
did neutral slides [t(ll) = 3.69, p = 0.002,
one-tailed probability] (see Fig. 2). No dif-

400 1

Q-

300-

2 200 -

100--

NEUTRAL
STIMULI

PHOBIC
STIMULI

Fig. 1. Means ( SEM) for cortisol excretion in percentage of pretest during exposure to neutral
or phobic slides.

ferences in SCL were observed between

different types of phobias during pretest
or stimulation. SCL increased in 9 of 12
subjects during phobic stimulation, and 11
of 12 showed a greater increase (or no decrease) to phobic compared to neutral
stimulation. Ratings of distress obtained
after slide presentations were expressed as
percentages of pretest values, and ratings
obtained during neutral slides were subtracted from those obtained during exposure to phobic cues. Phobic slides elicited
a greater increase in distress than did neutral slides [t(10) = 2.913, p = 0.008, onetailed probability) (see Fig. 3).
Table 3 gives means ( SEM) and paired
t tests for skin conductance responses and
half-time recovery to phobic and neutral
slides. Compared to neutral slides phobic
slides elicited greater skin conductance responses with slower half-time recovery. In
addition, responses to phobic slides habituated slower than those to neutral slides.
No differences were observed between different types of phobias for any electrodermal measure. Not in any case did neutral slides elicit greater or more slowly
habituating skin conductance responses.
One subject showed slower half-time recovery to neutral than to phobic slides.
No significant correlations between
SNAQ, SPQ, and MQ or ratings of distress
and cortisol excretion or skin conductance
variables emerged during phobic exposure. However, during neutral slide pre-

Psychosomatic Medicine Vol. 47, No. 4 (July/August 1985)

315

M. FREDRIKSON, 6 . SUNDIN, and M. FRANKENHAEUSER

TABLE 2. Means (SEM) and Paired f-Tests of Skin Conductance Levels Before and After Exposure to
Neutral or Phobic Slides
After
Exposure
((imol/min)

Before
Exposure
(|xmol/min)
Neutral
Phobic

3.94
4.55

(0.74)
(1.01)

3.91
5.23

(0.81)
(1.14)

((11) < 1
((11) = 2 96a

p < 0.05.

sentation distress ratings were positively

correlated with amount of change in cortisol (rxy = 0.63, p = 0.03), this correlation disappeared when changes were expressed as percentage of pretest
(rxy = 0.02).
DISCUSSION
The aim was to study cortisol excretion
in humans during the defense reaction

elicited by exposure to phobic material.

Both animal phobics and blood-and-mutilation phobics reacted with increased
urinary cortisol levels, skin conductance
levels, and distress ratings when exposed
to phobic slides but not when exposed to
neutral slides. Skin conductance responses to phobic as compared to neutral
slides were greater, had longer recovery
times, and habituated slower. No differences were observed between animal phobics and blood-and-mutilation phobics. The

~ 1201

400 1

300110200 -

PRETEST LEVEL

100

100

0J

NEUTRAL

PHOBIC

STIMULI

STIMULI

Fig. 2. Means ( + SEM) for skin conductance level

in percentage of pretest during exposure to
neutral or phobic slides.

316

J/)

NEUTRAL
STIMULI

PHOBIC
STIMULI

Fig. 3. Means (SEM) for distress ratings in percentage of pretest during exposure to neutral
or phobic slides.

Psychosomatic Medicine Vol. 47, No. 4 (July/August 1985)

DEFENSE REACTION AND CORTISOL EXCRETION

TABLE 3. Means ( SEM) and Paired f-Tests of Skin Conductance Responses, Half-Time Recovery, and
Number of Trials to Reach a Criterion of Two Successive Response Failures During Exposure to Phobic
or Neutral Slides
Phobic
Slides

Neutral
Slides
Response magnitude ((xmho)
Half-time recovery (sec)
Trials to criterion

0.50
8.48
2.40

(0.07)
(2.13)
(1.05)

0.89 (0.09)
14.00 (1.24)
5.70 (1.50)

= 3.13a
= 2.89b
= 2.746

p < 0.01.
p < 0.05.

electrodermal response pattern is consistent with previous findings (cf. 1] and indicates that phobic, but not neutral slides
elicit the defense reaction. The increase in
cortisol during phobic, but not neutral,
stimulation is consistent with increased
plasma cortisol levels observed during direct hypothalamic stimulation that induces the defense reaction in animals (10,
11) and with the cortisol increase elicited
by fear-producing conditions in normal
human subjects (9). In contrast, flooding
procedures produced only moderate elevations of plasma cortisol levels in four
out of six subjects phobic of snakes, birds,
and insects (18). Even though the present
study and Curtis et al. (18) differ in type
of phobics included this might not be crucial for the different results obtained, as
the present study found no differences in
cortisol excretion between animal and
blood and mutilation phobics. However,
the present study and that of Curtis et al.
(18) differ in several respects that might
account for the discrepancy: 1) Curtis et
al. measured plasma cortisol levels,
whereas we used urine levels. Sampling
effects are probably not crucial since Curtis et al. (18) sampled every 20 min and
the half-life of plasma cortisol approximates 1 hr. In addition, plasma and urinary cortisol excretion are positively correlated (19). 2) Venipuncture, used by Curtis

et al. (18), might raise cortisol-levels and

thereby mask any further increases due to
feaT, anxiety, or apprehension. 3) Curtis et
al. (18) provided continuous exposure to
phobic objects, whereas we used intermittent stimulation. Natelson et al. (11)
presented data indicating that plasma cortisol levels were increased in monkeys with
intermittent stimulation, whereas Mason
et al. (20), using continuous stimulation,
observed habituation of plasma cortisol
levels. We suggest that the cortisol response habituates to continuous but not to
intermittent stimulation, and that procedural differences between us and Curtis et
al. (18) accounts for the discrepancy in results.
In the present study, phobic slides elicited higher distress ratings than did neutral slides. The lack of correlation between
fear questionnaire scores and physiologic
variables probably reflects the restricted
range of scores on SNAQ, SPQ, and MQ.
The lack of correlation between subjective
distress and cortisol excretion or skin conductance activity during exposure to phobic slides could be interpreted as reflecting
nonparallelism or "desynchrony of fear"
(21). Curtis et al. (18) found poor correlations between subjective distress and
cortisol activation and Natelson et al. (11)
found a poor group correlation between
defensive behavior and plasma cortisol,

Psychosomatic Medicine Vol. 47, No. 4 (July/August 1985)

317

M. FREDRIKSON, O. SUNDIN, and M. FRANKENHAEUSER

even though individuals showed a good

correlation. Thus, cortisol increase does
not seem to be consistently related to other
aspects of fear, such as subjective distress
(18) or behavioral excitement (11). These
might be examples of "desynchrony of fear"
(21), probably due to differences in conditions regulating different autonomic and
behavioral systems involved in producing
an integrated fear response.

This research was supported by grants

from the Swedish Council for Research in
the Humanities and Social Sciences, from
the Swedish MedicaJ Research CounciJ
(Project No. 997), and from the KaroJinska
Institute. We are indebted to Tomas Berggren and Ulf Dimberg for comments on an
earlier version, and to Tomas Danielssons, Lars HoJmberg, and flagnar Levi for
technical assistance.

REFERENCES

1. Fredrikson M: Orienting and defensive reactions to phobic and conditioned fear stimuli in phobics
and normals. Psychophysiology 18:454463, 1981
2. Hare RD: Orienting and defensive responses to visual stimuli. Psychophysiology 10:453464, 1973
3. Hare RD, Blevings G: Defensive responses to phobic stimuli. Biol Psychol 3:1-13, 1975
4. Klorman R, Wiesenfeld AR, Austin ML: Autonomic responses to affective visual stimuli. Psychophysiology 12:553-560, 1975
5. Klorman R, Weissberg RP, Wiesenfeld AR: Individual differences in fear and autonomic reaction to
affective stimulation. Psychophysiology 14:45-51, 1977
6. Abrahams VC, Hilton SM, Zybrozyna A: Active muscle vasodilation produced by stimulation of the
brain stem: Its significance in the defense reaction. J Physiol 154:491-513, 1960
7. Sokolov EN: Perception and the Conditioned Reflex. Oxford, Pergamon, 1963
8. Frankenhaeuser M: The sympathetic-adrenal and pituitary-adrenal response to challenge: Comparison
between the sexes. In T Dembroski, TH Schmidt, G Blumchen (eds), Biobehavioral Bases of Coronary
Heart Disease. Basel, Karger, 1982, pp 91-106.
9. Henry JP, Stephens PM: Stress, Health and the Social Environment. A Sociobiologic Approach to
Medicine. New York. Springer, 1977
10. Folkow BP, Hedner B, Lisander B, Rubinstein E: Release of cortisol upon stimulation of the hypothalamic
"defense area" in cats. In Levi L (ed), Emotional Stress. New York, Elsevier, 1967, pp 114-119
11. Natelson BH, Smith GP, Stokes PE, Root AW: Plasma 17-hydroxycorticosteroids and growth hormone
during defense reactions. Am J Physiol 226:560-567, 1974
12. Edelberg R: The information content of the recovery limb of the electrodermal orienting response.
Psychophysiology 6:527-529, 1970
13. Fredrikson M: Reliability and validity of some specific fear questionnaires. Scand J Psychol 14:331-334,
1983
14. Klorman R, Weerts TC, Hastings JE. Melamed BG, Lang PJ: Psychometric description of some specific
fear questionnaires. Behav Ther 5.401^09, 1974
15. Fredrikson M, Danielssons T, Iremark H, Sundin O: Autonomic nervous blockades and phobic fear
responses. Submitted.
16. Lundberg U, Frankenhaeuser M: Pituitaryadrenal and sympathetic-adrenal correlates of distress and
effort. I Psychosom Res 24:125-130, 1980
17. Ruder HJR, Guy RL, Lipsett A: A radioimmunoassay for cortisol in plasma and urine. J Clin Endocrinol
Metab 35:219-224, 1972

318

Psychosomatic Medicine Vol. 47, No. 4 (July/August 1985)

DEFENSE REACTION AND CORTISOL EXCRETION

18. Curtis GC, Nesse R, Buxton M, Lippman D: Anxiety and plasma cortisol at the crest of the circadian
cycle: Reappraisal of a classical hypothesis. Psychosom Med 40:368378, 1978
19. Lundberg U, Hansson U, Andersson K, Eneroth P, Frankenhaeuser M, Hagenfeldt K: Hirsute women
with elevated androgen levels: Psychological characteristics, steroid hormones and catecholamines. J
Psychosom Obstet Gynaecol 2:86-93, 1983
20. Mason JW, Maher JT, Hartley LH, Mougey E, Perlow MJ, Jones LG: Selectivity of corticosteroid and
catecholamine responses to various natural stimuli. In Serban G (ed), Psychopathology of Human
Adaptation. New York, Plenum, 1976, pp 147-172
21. Rachman S, Hodgson R: Synchrony and desynchrony in fear and avoidance. Behav Res Ther 134:311-318,
1974

Psychosomatic Medicine Vol. 47, No. 4 (July/August 1985)

319