Efficacy of biofeedback in the treatment of migraine and tension type headaches.
ABSTRACT Biofeedback is an established non-pharmacologic technique commonly used in the treatment of migraine and tension type headaches. Multiple published studies have suggested that biofeedback is effective in reducing the frequency and severity of headaches, often allowing patients to decrease their dependence on medication. Studies have also suggested that biofeedback may effect a decrease in medical utilization.
Assess the efficacy of biofeedback in reducing the frequency and severity of migraine and tension type headaches.
Randomized, prospective, single blind, single center controlled trial.
Sixty-four patients with migraine with or without aura and/or tension type headaches, by ICHD-1 criteria, age 18 to 55, who had suffered from headaches for more than one year, were entered into the study. Patients were randomly assigned to receive biofeedback in addition to the basic relaxation instruction or relaxation techniques alone. All patients received instruction in pain theory. Biofeedback training consisted of 10 50-minute sessions utilizing standard EMG feedback from the frontalis and trapezius muscles and temperature from the third finger of the dominant hand. Visual and auditory feedback was provided. Thirty-three patients were assigned to receive biofeedback plus the relaxation techniques and 31, the relaxation techniques alone. All patients were asked to respond to periodic questionnaires for 36 months. The primary analysis was an intention-to-treat (ITT) analysis. The subsidiary analyses were not and the 11 subjects (7 in the relaxation alone and 4 in the biofeedback group) who received no treatment at all were analyzed and the results were qualitatively the same.
Patients who completed the program with education in pain theory and relaxation techniques showed a statistically significant decrease in the frequency and severity of the headaches in the first 12 months that continued to 36 months. Biofeedback provided no additional benefit, specifically no change in the frequency or severity of the headaches. After 3 months 48% of those in the relaxation group reported fewer severe headaches, while 35% of those in the biofeedback group reported fewer severe headaches; after 6 months, 52% of those in the relaxation group reported fewer severe headaches as compared with 57% reporting fewer severe headaches in the biofeedback group. The number of medications used by the patients and the utilization of medical care decreased in both groups over 36 months suggesting a regression to the mean.
Compliance was an issue throughout the study. Patients dropped out from the outset and that increased over time. Recovery of questionnaires was difficult and fewer were completed at each 3-month interval. Lack of a large control group who did not receive biofeedback or instruction in relaxation techniques.
Biofeedback is an extremely costly and time-consuming treatment modality that, in our study, provided no additional benefit when compared to simple relaxation techniques alone, in the treatment of migraine and tension type headaches in adults.
- [Show abstract] [Hide abstract]
ABSTRACT: In this article, we hope to summarize current understanding of pediatric headache. We discuss epidemiology, genetics, classification, diagnosis, outpatient, emergency and inpatient treatment options, prevention strategies, and behavioral approaches. For each section, we end with a series of questions for future research and consideration.Headache The Journal of Head and Face Pain 04/2014; · 2.94 Impact Factor
Dataset: Pompili 2010
Article: Migraines and CAM[Show abstract] [Hide abstract]
ABSTRACT: Migraines are a disabling condition affecting millions of adults and children worldwide. Complementary, alternative, and integrative medical approaches have been used for thousands of years to treat not only the migraine but also to help in prevention, reduction, or severity of future migraines. Many migraineurs use pharmacological treatments prescribed by their physician; many find these treatments do not provide relief, loose their effectiveness over time, or would rather use a nonpharmacological therapy. Although there has been research into a variety of complementary, alternative and integrative approaches to managing migraines, little extensive and reproducible research on any one approach has been conducted. The author investigated some of the current research and clinical thinking about complementary, alternative and integrative therapies, as well as general internet information that is easily accessible to all.Journal of Consumer Health on the Internet 10/2013; 17(4):432-438.
Background: Biofeedback is an established non-pharmacologic technique commonly used in the
treatment of migraine and tension type headaches. Multiple published studies have suggested that
biofeedback is effective in reducing the frequency and severity of headaches, often allowing patients
to decrease their dependence on medication. Studies have also suggested that biofeedback may
effect a decrease in medical utilization.
Objective: Assess the efficacy of biofeedback in reducing the frequency and severity of migraine
and tension type headaches.
Design: Randomized, prospective, single blind, single center controlled trial.
Methods: Sixty-four patients with migraine with or without aura and/or tension type headaches,
by ICHD-1 criteria, age 18 to 55, who had suffered from headaches for more than one year, were
entered into the study. Patients were randomly assigned to receive biofeedback in addition to the
basic relaxation instruction or relaxation techniques alone. All patients received instruction in pain
theory. Biofeedback training consisted of 10 50-minute sessions utilizing standard EMG feedback
from the frontalis and trapezius muscles and temperature from the third finger of the dominant
hand. Visual and auditory feedback was provided. Thirty-three patients were assigned to receive
biofeedback plus the relaxation techniques and 31, the relaxation techniques alone.
All patients were asked to respond to periodic questionnaires for 36 months. The primary analysis
was an intention-to-treat (ITT) analysis. The subsidiary analyses were not and the 11 subjects (7 in the
relaxation alone and 4 in the biofeedback group) who received no treatment at all were analyzed and
the results were qualitatively the same.
Results: Patients who completed the program with education in pain theory and relaxation
techniques showed a statistically significant decrease in the frequency and severity of the headaches
in the first 12 months that continued to 36 months. Biofeedback provided no additional benefit,
specifically no change in the frequency or severity of the headaches. After 3 months 48% of those
in the relaxation group reported fewer severe headaches, while 35% of those in the biofeedback
group reported fewer severe headaches; after 6 months, 52% of those in the relaxation group
reported fewer severe headaches as compared with 57% reporting fewer severe headaches in the
The number of medications used by the patients and the utilization of medical care decreased in both
groups over 36 months suggesting a regression to the mean.
Limitations: Compliance was an issue throughout the study. Patients dropped out from the outset
and that increased over time. Recovery of questionnaires was difficult and fewer were completed at
each 3-month interval. Lack of a large control group who did not receive biofeedback or instruction
in relaxation techniques.
Conclusion: Biofeedback is an extremely costly and time-consuming treatment modality that, in
our study, provided no additional benefit when compared to simple relaxation techniques alone, in
the treatment of migraine and tension type headaches in adults.
Key words: Biofeedback, relaxation techniques, tension type headache, migraine, pain program,
non-pharmacologic treatment of headache
Pain Physician 2009; 12:1005-1011
Efficacy of Biofeedback in the Treatment of
Migraine and Tension Type Headaches
From: Department of
Neurology, Harvard Vanguard
Medical Associates, Brigham
and Women’s Hospital,
Harvard Medical School,
Dr William J. Mullally
Department of Neurology,
1153 Centre Street, Suite 47,
Boston, MA 02130
Disclaimer: This study was
supported by a grant from the
Harvard Community Health
Conflict of interest: None.
Revised manuscript received:
Accepted for publication:
Free full manuscript:
William J. Mullally, MD, Kathryn Hall MS, RNCS, ANP-BC, and Richard Goldstein, PhD
Pain Physician 2009; 12:1005-1011 • ISSN 1533-3159
Pain Physician: November/December 2009:12:1005-1011
of headache sufferers. The purpose of this study was
to determine if the noninvasive, non-pharmacologic
treatment of biofeedback was unequivocally effective
in reducing the frequency and severity of migraine
and tension type headaches and as secondary effects,
decreasing the need for pharmacologic intervention
and direct medical care.
This was a randomized, prospective, single blind,
single center study. The study was approved and dili-
gently monitored by the Harvard Community Health
Plan Foundation Clinical Research Committee.
Sixty-four patients were entered into the study af-
ter obtaining written informed consent. Participants
ranged in age from 18 to 55 and 52 were female. All
of the participants suffered from migraine with or
without aura and/or tension type headache, by ICHD-
1 criteria, for more than 12 months occurring 2 to 5
times/month (23). Patients with a history of psychosis
were excluded. None of the participants suffered from
hemiplegic migraine. The groups were similar in age
distribution and sex.
All of the participants were referred to the Har-
vard Community Health Plan Comprehensive Pain Pro-
gram, a modified group day treatment program of 6
weeks duration for chronic pain of non neoplastic ori-
gin. The program emphasized education and training
in pain theory including headache. Pain management
methods included instruction in relaxation techniques,
meditation, self hypnosis, cognitive therapy, and art
and movement therapy with a pain clinician. Head-
ache patients received a modified program with em-
phasis on education and relaxation techniques. From
the Pain Program patients were assigned at random to
receive biofeedback in addition to the basic relaxation
techniques or to just complete the basic relaxation in-
struction. All patients were followed by a neurologist
who had no knowledge of who received the biofeed-
back treatment. Biofeedback training consisted of 10
50-minute sessions utilizing standard EMG feedback
from frontalis and trapezius muscles and temperature
from the third finger of the dominant hand. Both visu-
al and auditory feedback was provided. Biofeedback
Since that time multiple studies have suggested that
EMG biofeedback training administered alone or
in combination with relaxation techniques reduced
tension type headaches by 40% to 60% (2-6).
Subsequent studies using both EMG and temperature
biofeedback revealed the techniques to be extremely
effective in the treatment of migraine (7-10). In
1984 Seymour Diamond found that 68% of 395
patients with migraine and tension type headaches
reported improvement in the severity, duration, and
frequency of their headaches using biofeedback
and 65% were able to maintain those gains (11).
Biofeedback is a technique that combines modern
technology and psychology with the precepts of
Eastern self-discipline. The basis for biofeedback is
operant conditioning. Through biofeedback training,
patients develop a physiologic response to certain
stimuli. They then acquire a certain degree of control
over physiologic functions that contribute to the
genesis of head pain. EMG biofeedback promotes a
general sense of relaxation of the entire body using
an informational feedback system where the patient
hears a tone through headphones with the frequency
proportional to the EMG activity in the muscle (most
often the frontalis in headache) being monitored. As
relaxation is achieved the tone decreases. Temperature
biofeedback trains the patient to increase finger
and hand temperature, in theory, stabilizing the
vasomotor responses of migraine. The biofeedback
monitor is only a temporary facilitating device and
patients who practice the techniques should be able to
effect the same changes after they are weaned from
the monitor, ultimately decreasing the frequency and
severity of their headaches (12). Migraine headaches
affect approximately 12% of the population, directly
impair the quality of life, and result in reduced work
capacity and social activity (13-17). The direct medical
and indirect costs of migraine total billions of dollars
each year (18). Tension type headaches occur in at
least 40% of the population and the impact on health
care utilization and decreased productivity is marked
as well (19-22). Medication remains the mainstay of
treatment for all types of headaches and vast amounts
of prescription and over-the-counter medications are
used. Side effects frequently occur with medication
and at times can be life threatening. The medications
themselves often contribute to the reduced productivity
n 1970 Budzynski, Stoyva, and Adler introduced
the use of electromyographic (EMG) biofeedback
in the treatment of tension type headaches (1).
Biofeedback in the Treatment of Migraine and Tension Type Headaches
for all patients was administered by a licensed psychol-
ogist with extensive experience in the use of the tech-
niques for the management of headaches. Thirty-three
patients were entered into the biofeedback plus Pain
Program and 31 into the Pain Program alone.
Patients were asked to complete a questionnaire
upon entering the study then at 3, 6, 9, 12, 24, and 36
months. The questionnaire established the total num-
ber of headaches, the number of severe headaches
(grade 7-10/10) and the number of mild to moder-
ate headaches (grade 1-6/10), in the prior 3 months.
Patients who received biofeedback were questioned
about the effectiveness of biofeedback techniques in
preventing or lessening severe and mild to moderate
headaches on a scale of 0 to 4 where 0 was ineffective
and 4 completely effective. They were also asked to
rate the biofeedback experience using the same scale.
Headache medications and the number of medical vis-
its, specifically for headaches, were monitored in the
electronic medical record.
Descriptive tables and graphs are presented
showing the changing means for the 4 outcome vari-
ables over time. The main analysis; however, is a form
of longitudinal regression for each of the 4 outcomes.
All analyses were done in Stata (various versions) (24).
The question of interest in all cases was whether there
was a consistent pattern over time (e.g., was there a
general trend showing improvement over time) and
whether this pattern differed by treatment group.
The null hypothesis in each case was that there was
no difference between the groups. The structure of
the data was that each person was represented by
a row of data on each occasion on which they were
assessed. This means that rows of data were not inde-
pendent since people who were assessed more than
once were represented by more than one row. This
lack of independence would bias the standard errors
unless we corrected for this; we used Stata’s built-in
correction to obtain cluster-adjusted standard errors
and those are shown in the table (25). The main analy-
ses were based on intention-to-treat (ITT); the subsid-
iary analyses, which were not ITT, were repeated after
dropping the 11 subjects who received no treatment
at all and the results were qualitatively similar. The
sample size was too small for multiple imputation.
Subject Accounting and Demographic
Of the 100 potential subjects who were screened,
64 entered the study and were randomized between
the 2 groups (31 into the pain program group and
33 into the biofeedback group). The ages of the sub-
jects ranged from 18 to 55 and 52 were female. All
had had migraine and/or tension type headaches for
at least 12 months. As Table 1 shows, not all subjects
responded to all questions in the survey and some
subjects dropped out over time. One patient was lost
to follow-up immediately after entering the study
and 4 other patients did not complete any of the
questionnaires. Note that 2 of the main outcomes
(number of headaches and number of severe head-
aches, each in the preceding 3 months) were mea-
sured via the periodic survey. On the other hand,
the other 2 outcomes (number of medications and
number of medical visits (each over the preceding 12
months) are based on data collected from electronic
medical records and virtually no data is missing (with
the exception of the one patient who was lost to fol-
low-up at the outset).
Figures 1–4 are graphs showing the mean val-
ues at each time point by group. Each graph shows a
clear trend toward improvement over time and each
Table 1. Means, standard deviations, and frequencies of the
number of headaches/month in the prior 3 months.
Pain / Bio
4.46 ± 0.92
4.21 ± 1.14
4.05 ± 1.21
3.84 ± 1.01
3.71 ± 1.10
3.91 ± 1.04
3.00 ± 1.41
4.00 ± 1.14
4.71 ± 0.74
4.46 ± 0.99
4.08 ± 1.13
3.68 ± 1.43
3.71 ± 1.45
3.36 ± 1.65
4.59 ± 0.83
4.34 ± 1.06
4.06 ± 1.16
3.76 ± 1.24
3.71 ± 1.27
3.60 ± 1.41
2.95 ± 1.36
4.00 ± 1.23
36 2.92 ± 1.38(13)
4.00 ± 1.31
Pain Physician: November/December 2009:12:1005-1011
shows little difference between the groups. We do not
think that the differences that are shown are clinically
We start with descriptive information for each of
the 4 outcomes of interest. Each table shows, in the
rows, the time since the start of the subject in the
study; the columns are the group: relaxation only (RL)
and relaxation plus biofeedback (BF). In each cell are
3 numbers: the mean number of headaches, the stan-
dard deviation, and the number of people reporting
results. As expected the means at time 0 (start of the
study) are very similar. But note that the means con-
tinue to be similar at the other time points.
Table 2 shows the mean number of severe head-
aches/month in the prior 3 months. The means start
Fig. 1. Change in mean number of headaches over time.
Fig. 2. Change in mean number of severe headaches over
Fig. 3. Change in mean number of medications over time.
Fig. 4. Change in mean number of visits over time.
at the same level and, again, continue to be very
Table 3 shows the number of medications in
the preceding 12 months. The table shows that the
means differ throughout. At baseline, the biofeed-
back group were using more medications. This con-
tinued through month 24; however, at 36 months,
the remaining people in the biofeedback group were
using fewer drugs than the remaining people in the
Table 4 demonstrates the number of visits in the
preceding 12 months.
Finally, our regression results are summarized in
Table 5. The first number in each cell is the coeffi-
cient; the P-value is shown in parentheses. The rows
Biofeedback in the Treatment of Migraine and Tension Type Headaches
are interpreted as follows: biofeedback group gives
the difference between the 2 groups, shown as bio-
feedback minus relaxation; thus, for example, the
0.24 under number of headaches says that the mean
number of headaches in the biofeedback group was
0.24 higher than the mean for the relaxation group.
The other rows show the various time points — each
is compared to the baseline measure (i.e., at random-
ization and before treatment); all the coefficients are
negative showing that each measure declines over
time compared with the baseline value. Ninety-five
percent confidence intervals are shown for the group
predictor (which was coded 0 for “relaxation only”
group and 1 for those who were in the “biofeedback
also” group). Note that medications and visits are only
measured at annual assessments and thus the rows for
the 3, 6, and 9-month values are blank in these col-
umns. In addition, we checked for an interaction be-
tween time and group (pain program or biofeedback)
for each of the 4 outcomes; in no case was there a
statistically significant interaction, with P-values rang-
ing from > 0.1 to > 0.9. Thus, there is no evidence that
there was any difference either over time or at any
single time between the 2 groups.
The results are similar in structure for each of
the 4 outcomes: there is a general trend showing
improvement over time and this trend does not dif-
fer by treatment group. Recall that each coefficient
Table 2. Means, standard deviations and frequencies of
number of severe headaches/month in prior 3 months.
Pain / Bio
3.22 ± 1.22
2.83 ± 1.43
2.59 ± 1.26
2.32 ± 1.016
2.52 ± 1.08
2.36 ± 1.12
2.11 ± 1.45
2.66 ± 1.25
3.32 ± 1.25
2.96 ± 1.48
2.19 ± 1.20
2.05 ± 1.09
2.14 ± 1.20
1.86 ± 0.95
1.69 ± 0.95
2.45 ± 1.32
3.28 ± 1.23
2.90 ± 1.45
2.38 ± 1.23
2.17 ± 1.07
2.33 ± 1.14
2.08 ± 1.04
1.86 ± 1.17
2.55 ± 1.29
Table 3. Means, standard deviations and frequencies of
number of medications in prior 12 months.
Pain / Bio
6.80 ± 8.35
6.37 ± 12.33
3.68 ± 6.29
2.61 ± 3.82
5.03 ± 8.61
8.76 ± 12.04
7.09 ± 9.16
5.22 ± 9.09
1.89 ± 4.79
5.90 ± 9.50
7.83 ± 10.41
6.75 ± 10.70
4.50 ± 7.88
2.22 ± 4.36
5.49 ± 9.09
Table 4. Means, standard deviations, and frequencies of
number of visits in prior 12 months.
Pain / Bio
4.93 ± 4.91
3.13 ± 5.49
1.93 ± 3.31
1.22 ± 1.44
2.92 ± 4.40
5.61 ± 6.37
4.21 ± 4.90
2.50 ± 3.71
1.07 ± 1.90
3.44 ± 4.86
5.29 ± 5.69
3.70 ± 5.18
2.23 ± 3.51
1.14 ± 1.70
3.20 ± 4.65
shows the predicted mean change in outcome when
comparing people in a row to relaxation people; for
the time oriented rows (e.g., 3 months, 6 months),
the comparison is to the time 0 (baseline) measure.
For example, the coefficient of 0.24 for biofeedback
under number of headaches means that those in the
biofeedback group suffered, on average, 0.24 more
headaches during the prior 3 months than did the
relaxation group; similarly, they suffered approxi-
mately 0.2 fewer severe headaches, use almost one
more medication, and just over half an additional
visit when compared to the relaxation group. On
the other hand, all the coefficients for the time vari-
ables are negative meaning that headaches, severe
headaches, medications, and visits all decreased
Pain Physician: November/December 2009:12:1005-1011
Finally, P values calculated by unequal variance 2-
sample t-tests are as follows. Number of headaches:
0.2678; number of severe headaches: 0.7584; number
of medications: 0.4532; number of visits: 0.6390. Each
of the 4 P-values is larger than 0.05 so we cannot re-
ject the null hypothesis.
We initiated this study to provide scientific evi-
dence validating the use of biofeedback techniques
in the treatment of migraine and tension type head-
aches. Biofeedback is virtually free of untoward side
effects and if effective for preventative and abortive
treatment of headaches would obviously be prefer-
able to the use of medication. Our study, however,
failed to reveal a beneficial effect. Compliance with
this type of study is always a limitation and the pa-
tients were asked to complete multiple questionnaires
so that a sufficient number would be returned allow-
ing us to complete a valid statistical analysis. We were
able to track medical utilization in all but one patient
over the 3-year period using an electronic medical re-
cord. The data that we obtained allowed us to reach
our goals and clearly showed that completion of the
modified pain program, which incorporated educa-
tion in pain theory and instruction in simple relax-
ation techniques, resulted in a statistically significant
decrease in the frequency and severity of the migraine
and tension type headaches in the first 12 months that
continued to 36 months. Biofeedback provided no ad-
ditional benefit and did not satisfy the primary end-
point which was, specifically, an additional reduction
in the frequency and severity of the headaches. The
patients who received biofeedback training reported
a subjective improvement in their headaches but the
objective data did not support their perception.
The secondary endpoints were a reduction in the
total number of medications used and the utiliza-
Table 5. Summary of regression results.
tion of medical care. Patients in both the relaxation
and biofeedback groups showed a decrease with re-
gard to both parameters over the 36 months. Elev-
en patients either did not commence treatment or
dropped out before completing treatment and their
results, with regard to medication use and medi-
cal utilization, did not differ from the groups that
received the treatment modalities. One alternative
hypothesis that cannot be ruled out without addi-
tional data is whether this general trend toward im-
provement is an artifact of regression to the mean.
Given that 11 people received no treatment of any
kind, and that the results do not differ to a statisti-
cally significant extent between those who received
and those who did not receive treatment, this alter-
native must be taken seriously. However, since only
11 people received no treatment, this comparison
has little statistical power and we cannot distinguish
between regression to the mean and improvement
based on treatment.
Headaches are a major health problem in the
United States because of direct medical costs, in-
creased health care dependency, and loss of produc-
tivity (14,17,18,20). Optimal treatment for headaches,
and in fact any disorder, would be effective in reduc-
ing or eliminating symptoms with no or few side ef-
fects. Non-pharmacologic treatments have been uti-
lized by practitioners for many years in the hope of
reaching those goals. However, controlled studies to
demonstrate efficacy of those treatment modalities
are lacking (26). Biofeedback has been a mainstay
of headache treatment for decades and is offered at
many of the major pain and headache clinics through-
out the United States. It is expensive and time consum-
ing and while it may have a beneficial effect in the
treatment of migraine and tension type headaches,
our study found it to be no more effective than simple
CI: -.49, .54
CI: -.73, .34
CI: -2.85, 4.75
CI: -1.22, 2.38
Biofeedback in the Treatment of Migraine and Tension Type Headaches
In conclusion, non-pharmacologic treatment,
specifically simple relaxation techniques and educa-
tion in pain theory, should remain an integral part
of the treatment program for migraine and tension
type headaches. Instruction in relaxation can be ac-
complished in a few visits and does not require a
prolonged course of treatment (27). Biofeedback,
however, should no longer be universally accepted as
standard treatment for these disorders in adults as the
addition of this treatment modality, in our study, did
not provide additional therapeutic benefit.
1. Budzynski T, Stoyva J, Adler C. Feed-
back-induced relaxation: Application to
tension headache. Journal of Behavior-
al Therapeutic Experimental Psychiatry
2. Budzynski TH, Stoyva JM, Adler CS, Mul-
laney DJ. EMG biofeedback and tension
headache: A controlled outcome study.
Psychosomatic Medicine 1973; 35:484-
3. Blanchard EB, Andrasik F, Ahles TA, Ted-
ers SJ, O’Keefe D. Migraine and tension
headache: A meta-analytic review. Be-
havioral Therapy 1980; 11:613-631.
4. Blanchard EB, Kim M, Hermann CU, Stef-
fek BD. Preliminary results of the effects
on headache relief of perception of suc-
cess among tension headache patients
receiving relaxation. Headache Quar-
terly, Current Treatment and Research
5. Bogards MC, ter Kuile MM. Treatment of
recurrent tension headache: A meta-an-
alytic review. The Clinical Journal of Pain
6. Rokicki LA, Holroyd KA, France CR, Lip-
chik GL, France JL, Kvaal SA. Change
mechanisms associated with combined
relaxation/EMG biofeedback training
for chronic tension headache. Psycho-
physiology and Biofeedback 1997; 22:
7. Friar LR, Beatty J. Migraine: Manage-
ment by trained control of vasoconstric-
tion. Journal of Consultative and Clinical
Psychology 1976; 44:46-53.
8. Blanchard EB, Theobald DE, William-
son DA, Silver BV, Brown DA. Tempera-
ture biofeedback in the treatment of mi-
graine headaches. Archives of General
Psychiatry 1978; 35:581-588.
9. Blanchard EB, Applebaum KA, Radnitz
CL, Morrill B, Michultka D, Kirsch D,
Guarnieri P, Hillhouse J, Evans DD, Jac-
card J A controlled evaluation of ther-
mal biofeedback and thermal biofeed-
back combined with cognitive therapy
in the treatment of vascular headache. J
Consult Clin Psychol 1990; 58:216-224.
10. Rains JC, Penzien DB, McCrory DC, Gray
RN. Behavioral headache treatment:
History, review of the empirical liter-
ature, and methodological critique.
Headache 2005; 45:S92-S109.
11. Diamond S, Montrose D. The value of
biofeedback in the treatment of chron-
ic headache. A four year retrospective
study. Headache 1983; 24:296-304.
12. Schwartz MS and Associates. Biofeed-
back, A Practitioner’s Guide, 2nd edi-
tion. The Guilford Press, New York,
13. Scher AL, Stewart WF, Lipton RB. Mi-
graine and headache: A meta-analytic
approach. In: Crombie. IK, Kroft PR, Lin-
ton (eds.) Epidemiology of Pain. IASP
Press, Seattle, 1999, pp. 159-170.
14. Lipton RB, Scher AL, Kolodner K, Liber-
man J, Steiner TJ, Stewart WF. Migraine
in the United States: Epidemiology and
patterns of health care use. Neurology
15. Bigal ME, Lipton RB, Stewart WF. The
epidemiology and impact of migraine.
Current Neurology, Neuroscience Re-
ports 2004; 4:98-104.
16. Bussone G, Usai S, Grazzi L, Rigam-
onti A, Solari A, D’Amico D. Disability
and quality of life in different primary
headaches: Results from Italian Stud-
ies. Neurologic Science 2004; Suppl 3:
17. Lipton R. Stewart W, Diamond S, Dia-
mond M, Reed M. Prevalence and bur-
den of migraine in the United States:
Data from the American Migraine Study
II. Headache 2001; 41:646-657.
18. Hu X, Markson L, Lipton R, Stewart W,
Berger M. Burden of migraine in the
United States: Disability and econom-
ic costs. Archives of Internal Medicine
19. Schwartz BS, Stewart WF, Simon D, Lip-
ton RB. Epidemiology of tension type
headache. JAMA 1998; 279:381–383.
20. Rasmussen BK. Epidemiology of head-
ache. Cephalalgia 1995; 15:45-68.
21. Rasmussen BK, Olesen J. Epidemiology
of migraine and tension type headache.
Current Opinion in Neurology 1998;
22. Lavados PM, Tenhamm E. Epidemiolo-
gy of tension type headache in Santia-
go Chile: a prevalence study. Cephalal-
gia 1998; 8:552-558.
23. Headache Classification Committee of
the International Headache Society.
Classification and diagnostic criteria for
headache disorders, cranial neuralgias,
and facial pain. Cephalalgia 1988;8
24. Stata: Statacorp LP, College Station, TX.
25. Williams RL. A note on robust variance
estimation for cluster-correlated data.
Biometrics 2000; 56:645-646.
26. Andrasik F. What does the evidence
show? Efficacy of behavioral treatments
for recurrent headaches in adults. Neu-
rologic Science 2007; 28:S70-S77.
27. Warner G, Lance JW. Relaxation therapy
in migraine and tension type headache.
Medical Journal of Australia,. 1975;