Biofeedback and Self-Regulation, VoL 1, No.

1, 1976

Case Reports and Training Techniques

Effects of an E M G B i o f e e d b a c k Relaxation

P r o g r a m on the Control of Diabetes 1

A Case Study

Jane E. Fowler
University of Denver
Thomas H. Budzynski 2 and Richard L. VanflenBergh
University of Colorado Medical Center

It was hypothesized that EMG biofeedback relaxation training, applied to a

diabetic patient, wouM result in a decreased level o f insulin with fewer epi-
sodes o f ketoacidosis. A 20-year-aM female, diabetic since age nine, kept
daily records o f insulin doses and rated herself on an emotionality and a
diabetic scale. A full-semester baseline was taken. This was followed by a
semester-long training program during which the patient practiced relaxing
her frontalis muscle with a portable EMG feedback unit which produced a
geiger-counter-like click feedback. A cassette-tape series was used along
with the portable EMG. The patient was encouraged to practice twice each
day and to attempt to maintain a relaxed state even when not in the practice
situation. The daily use o f the portable unit was terminated at the end o f the
semester. In addition, the patient ceased practicing twice daily with the
cassette tape. Daily insulin averaged 85 units f o r the six-week baseline and
59for the final six weeks o f the training period. Moreover, at the end o f the
training period the average close had reached 43 units. During the training
period the patient rated herself as decreasing in emotionality and in diabetic
fluctuations.

It has been well documented that the course and control of diabetes mellitus
and the levels of blood glucose can be altered directly and indirectly by emo-
tional stress (Baker, Barcai, Kaye, & Haque, 1969; Hinkle, 1956; Vanden-
'Supported by the National Institute of Mental Health Grant MH-15596.
2Requests for reprints should be sent to Dr. Thomas H. Budzynski, University of Colorado
Medical Center, 4200 East Ninth Avenue, Box C250, Denver, Colorado 80220.
105
© 1976 Plenum Publishing Corporation, 227 West 17th Street, New Y o r k , N.Y. 10011. No
part o f this publication may be reproduced, stored in a retrieval system, or transmitted, In
any form or by any means, electronic, mechanical, photocopying, microfilming, recording,
or otherwise, without written permission of the publisher.
106 Fowler, Budzynski, and VandenBergh

Bergh, Sussman, & Vaughn, 1967; VandenBergh, Sussman, & Malburg,

1968; VandenBergh, Sussman, & Titus, 1966; Weller, Linder, Nuland, &
Kline, 1961). If patients could be taught to reduce the effects of emotional
factors then the course and control of the disorder might improve.
Certain biofeedback techniques have been employed successfully for
the reduction of symptoms in a number of stress-related disorders, includ-
ing chronic muscle contraction headache (Budzynski, Stoyva, & Adler,
1970; Budzynski, Stoyva, Adler, & Mullaney, 1973), migraine headache
(Sargent, Green, & Waiters, 1973), anxiety (Budzynski & Stoyva, 1973), and
essential hypertension (Montgomery, Love, & Moeller, 1974). All of the
biofeedback procedures used in these studies resulted in some degree of
voluntary control over some aspect of physiological functioning formerly
not under voluntary control. That control enabled the patients to modify
their physiology so as to reduce the symptoms of the disorders. Electromyo-
graphic (EMG) feedback was in fact used in all of the above studies, with
the exception of the migraine research in which peripheral temperature
feedback was used. The application of a training program primarily based
on EMG feedback has been described by Stoyva and Budzynski (1974). The
training, designated "cultivated relaxation," allows the individual to
modify his reaction to stressful situations so that a more optimal physiolog-
ical pattern is produced.
With regard to diabetes it was hypothesized that in certain patients
this "cultivated relaxation" ability would, through daily application,
produce a decrease in emotional lability. This decrease would in turn have a
stabilizing effect on the diabetes, thus resulting in a lessened insulin require-
ment.

METHOD

Patient's History

The patient, a 20-year-old woman, has had diabetes since the age of 9.
Control of her diabetes was excellent until the age of 15 years, when she de-
veloped thyroiditis.
At age 16, her diabetic control became very poor, and she was hospi-
talized for 14 months. During this time, physical or emotional stress ap-
peared to produce severe ketoacidosis. Even "positive" emotions such as
excitement caused her diabetes to go out of control. Insulin doses of several
hundred units per day were common, as were massive swings in blood
sugar. Episodes of ketoacidosis were often followed by hypoglycemic reac-
tions.
When she was 17, the patient's therapy consisted of the administra-
tion of MJ 1999 (Sotalol), a fl-adrenergic blocker, and family psychological
EMG Biofeedback and Diabetes 107

therapy. There was a marked improvement in diabetic control at this time,

as she was able to live at home and attend classes. Although ketoacidosis
was still rather frequent and often associated with infections, the severity of
the ketoacidosis was lessened. The MJ 1999 was discontinued after 15
months.
During her first quarter at college, the patient was hospitalized four
times for ketoacidosis. Infections were present during each episode, but
other factors, including emotional upsets, also appeared to be involved. She
was hospitalized four more times during her freshman year; during the first
half of her sophomore year she was hospitalized three times. Baseline obser-
vations were begun in the winter quarter of her sophomore year. At this
time the patient was fully informed of the details of the study and consented
to the procedures.

Baseline Period (Winter Quarter: January to mid-March, 1973)

The primary dependent measure was the daily dose of insulin units.
Urine testing (Clinitest and Acetest, Ames) was done two to four times a
day. Blood sugar (Folin-Wu, and Dextrostix, Ames) was measured when re-
quested by the patient's physician. These measures were made at 2-4-week
intervals. A daily diary was used to record the above information plus such
factors as diet, exercise, medication, infection, accidents, and emotional
conflicts. Additionally, two subjective measures were taken: a diabetic state
assessment and an assessment of emotional state (see descriptions below).
The adjustment of the daily insulin dose typically was made by the
patient, based on urine-test and blood-sugar-test results. The advice of her
physician was secured whenever possible.
Because the Diabetes Scale was a subjective estimate of diabetic state,
the patient made the estimate before performing each urine or blood-sugar
test. The ratings ranged from --4 (severe hypoglycemia), to +4 (severe
ketoacidosis). The rating descriptions are given in detail in Table I. Ratings
were recorded four times a day. In the event of extreme conditions the
ratings (--4 or + 4) were made after the fact.
The Emotional Scale was a subjective measure feeling state--primar-
ily reflecting general tension and psychological conflict (or lack of it). Emo-
tional Scale ratings were recorded at the same time as the diabetic ratings.
The Emotional Scale descriptions are given in Table II.

Training Period (Spring Quarter: late March to early June, 1973)

All clinical testing and self-rating continued during this period. The
training involved twice a day practice periods of 30-40 minutes each. A
108 Fowler, Budzynski, and VandenBergh

Table I. Self-Rated Diabetic Scale

Rating Description
-4 Severe hypoglycemia
Unconscious
Convulsions
Little or no recall of events
Hospitalization
-3 Moderate hypoglycemia
Some lapse in memory
Insulin reaction requiring large amounts of glucose, etc.
Strange behavior, i.e., acting drunk
Nervousness, anxiety, and/or apprehension
Confusion, difficulty in thinking
Difficulty in focusing eyes
Poor coordination
Speech difficulties
-2 Mild hypoglycemia
Hunger
Sweating
Cold, clammy hands
Cold, shivering
Tachycardia
Able to recall all events
Shaking
Drowsy
Weak
-1 Feel drop in blood sugar, but no insulin reaction
Tired
Hunger
Mild weakness
0 Feel fine: No signs of hypo- or hyperglycemia
+1 Mild symptoms of hyperglycemia and/or acetone
Weakness
Thirsty
Easily fatigued during mild exercise, e.g., walking
+2 Moderate symptoms
Increasing nausea
Weakness
Starting to become dehydrated
Headache
+3 Serious symptoms
Extreme weakness
Dehydrated
Dizziness
Blurred vision
Drowsy, tired
Difficulty in concentrating
Severe nausea, and/or vomiting
Kussmaul breathing
Aches (muscular, joint)
Difficulty in extending arms, legs, fingers
+4 Ketoacidosis requiring hospitalization
EMG Biofeedback and Diabetes 109

Table II. Self-Rated Emotional Scale

Rating Description

0 Very relaxed
Completely free of problems and worries, got mind completely off
work and other areas of life where problems abound
Relaxed
Very few problems, little pressure or conflict; things handled very
easily
Average Amount of Tension
Some problems, some tension, coping adequately
Tense
Concerned, uncomfortable, really working at resolving problems or
conflicts
Extremely Tense
The level of tension as high as it ever gets. Very uncomfortable, anx-
ioux, close to losing control

portable electromyograph (EMG) feedback unit, PE-2 (Biofeedback Sys-

tems, Inc., Boulder, Colorado), was used in conjunction with a series o f
cassette tapes containing relaxation instructions (Biofeedback Systems,
Inc.).
The EMG feedback unit was operated in the following manner. Three
surface sensors were applied over the forehead (frontalis) area. Visual feed-
back and quantification were provided by a meter. The auditory feedback
was a series of clicks (similar to a geiger counter) presented through the
headphones. The click rate was proportional to the level of tension in the
muscle. While listening to the tapes, the patient would try to relax as much
as possible using the click rate as an indication of degree of muscle relaxa-
tion.
The cassette tapes contained a progression of relaxation instructions
for relaxing the major muscle groups, as well as stabilizing and slowing
respiration. As the patient became more adept at relaxing during the train-
ing periods, she was asked to attempt to use this new skill in everyday situa-
tions, i.e., she would try to relax during stressful situations.

Follow-up Period (Winter Quarter: mid-December,

1973 to late January, 1974)

At the end of the training period the patient discontinued practice

with the E M G unit and the cassette tapes, although she continued to use her
relaxation ability in stressful situations. A follow-up period was begun ap-
proximately one year from the start of the baseline period.
110 Fowler, Budzynski, and VandenBergh

Weekly Mean and Standard Error of the Mean of Daily Insulin

120 == .,=
o

g ~ ~ nN
I012 ~ .~_
L

•-~ 8C E

6C

~ 4o

o 20
Baseline Training Follow-up

Week

F i g . 1. W e e k l y m e a n a n d s t a n d a r d e r r o r o f d a i l y insulin.

RESULTS

Daily Insulin

Daily insulin intake is presented in Figure 1. The weekly mean of daily

insulin fluctuated somewhat but peaked during the stressful week of final
exams. The sixth week of the baseline period was actually spent at home on
vacation. The average of the six week baseline period was 85 units daily.
Appreciable results began to appear in the fourth week of training.
The average insulin intake for the last 6 weeks of the training period was 59
units. As contrasted to the relatively high level during the baseline final-
exam week (103 units), the average of 44 during the training-period exam
week is striking. In addition, the patient reported that the training and
follow-up periods were much more academically demanding than the base-
line period.
The follow-up was initiated roughly 6 months after the end of the
training period. This matched the baseline period for the time of year and
covered the same duration of time. The average daily dose of insulin was 52
units during the follow-up period.
It is evident from Figure 1 that the variation measure (standard error
of the mean) decreases as training progresses and remains small during the
follow-up period. This lends support to the hypothesis that the training
stabilizes the insulin intake as well as lowering the quantity required.
EMG Biofeedback and Diabetes 111

Self-Rated D&betic and Emotional Scales

The self-ratings on the Diabetic (D) correlated positively with insulin

dose (r = ,77, p < .01). The daily ratings of the scale changed from the
typically positive (hyperglycemic) during the baseline to typically negative
(hypoglycemic) during the training phase. This change was signaled by the
occurrence of frequent and sometimes severe insulin reactions. It was
apparent that the patient's need for insulin decreased more rapidly than the
rate at which the daily dose was actually adjusted.
The Emotional Scale showed a positive, but not significant correla-
tion (r = .35, p < . 1) with daily insulin. The E-scale readings were consider-
ably higher during the baseline (average = 2.11) than during the training
phase (average = 1.74).
The D and E scale readings correlated positively and significantly (r =
.49, p < .05).

DISCUSSION

The training period, carried out through the Spring quarter of 1973,
was the first quarter since starting college that the patient was not hospital-
ized. Her urine sugars were not nearly as high during training and follow-up
as during the baseline period and only rarely was acetone spilled. This seems
especially significant in light of the fact that the training and follow-up
school quarters were more demanding academically than that of the base-
line quarter. In spite of the heavier schedules, the patient found that her
anxiety and tension levels had decreased. She reported that she does not let
things upset or worry her as much as before and she is able to relax fairly
well when she feels herself becoming tense.
The data from this patient indicate, however, that this sort of training
regime may well result in a rapid decrease in insulin requirements, which
could lead to frequent and severe hypoglycemic reactions.
Although this was only a study of one case, the results obtained, in
light of the lengthy baseline combined with the follow-up, lend support to
the hypothesis that training in "cultivated relaxation" is an effective proce-
dure for helping certain diabetics control their disorder. The training pro-
gram is relatively simple to administer and can be combined with existing
forms of therapy for diabetics.

REFERENCES

Baker, L., Barcai,A., Kaye,R., & Haque, N. Beta adrenergicblockadeand juvenilediabetes:

Acute studies and long-termtherapeutictrial. Journal of Pediatrics, 1965, 75, 19-29.
112 Fowler, Budzynski, and VandenBergh

Budzynski, T. H. Biofeedback procedures in the clinic. Seminars in Psychiatry, 1973, 5,

537-547.
Budzynski, T. H., & Stoyva, J. M. Biofeedback techniques in behavior therapy. In N. Bir-
baumer (Ed.), Neuropsychologie der Angst. Reihe Fortschritte der Klinischen Psy-
chologie. Bd. 3. Munchen, Berlin, Wien: Verlag Urban & Schwarzenberg, 1973. Repub-
lished in D. Shapiro et al. (Eds.), Biofeedback and self-control: 1972. Chicago: Aldine
Publishing Company, 1973.
Budzynski, T. H., Stoyva, J. M., & Adler, C. S. Feedback-induced muscle relaxation: Appli-
cations to tension headache. Behavior Therapy and Experimental Psychiatry, 1970, 1,
205-211.
Budzynski, T. H., Stoyva, J. M., Adler, C. S., & Mullaney, D. EMG biofeedback and tension
headache: A controlled outcome study. Psychosomatic Medicine, 1973, 35, 484-496.
Hinkle, L. E. The influence of the patient's behavior and his reaction to his life situation upon
the course of diabetes. Diabetes, 1956, 5, 406.
Montgomery, D. D., Love, W. A. Jr., & Moeller, T. A. Effects of electromyographic feedback
and relaxation training in blood pressure in essential hypertension. Paper presented at
the Biofeedback Research Society Annual Meeting, Colorado Springs, Colorado,
February, 1974.
Sargent, J., Green, E., & Waiters, E. D. Preliminary report on the use of autogenic feedback
training in the treatment of migraine and tension headaches. Psychosomatic Medicine,
1973, 35, 129-135.
Stoyva, J. M., & Budzynski, T. H. Cultivated low arousal--An anti-stress response? In L. V.
DiCara (Ed.), Recent advances in limbic and autonomic nervous systems research. New
York: Plenum, 1974.
VandenBergh, R. L., Sussman, K. E., & Malburg, B. J. Effects of the stress of final examina-
tions on university students with diabetes mellitus, 1968. Unpublished manuscript.
VandenBergh, R. L., Sussman, K. E., & Titus, C. C. Effects of hypnotically induced acute
emotional stress in carbohydrate and lipid metabolism in patients with diabetes mellitus.
Psychosomatic Medicine, 1966, 28, 383.
VandenBergh, R. L., Sussman, K. E., & Vaughn, G. D. Effects of combined physical-antici-
patory stress on carbohydrate lipid metabolism in patients with diabetes mellitus. Psy-
chosomaticMedicine, 1967, 8, 16.
Weller, C., Linder, M., Nuland, W., & Kline, M. V. The effects of hypnotically-induced emo-
tions on continuous uninterrupted blood glucose measurements. Psychosomatics, 1961,
2, 375.

(Original received June 9, 1975)