Clinical Rehabilitation 2003; 17: 624–630

The effects of exercise and exercise-related

changes in blood nitric oxide level on migraine
headache
S Osün Narin, L Pinar School of Physical Therapy and Rehabilitation, Dokuz Eylül University, Izmir, D Erbas Gazi
University , Faculty of Medicine, Department of Physiology, Ankara, V Oztürk and F Idiman Dokuz Eylül University,
Faculty of Medicine, Department of Neurology, Izmir, Turkey

Received 6th August 2002; returned for revisions 2nd December 2002; revised manuscript accepted 14th March 2003.

Objective: To observe the effects of moderate aerobic exercise on migraine

headache, to assess exercise-related changes in blood nitric oxide (NO) levels,
and to examine the impact of such changes on migraine attacks.
Design: Controlled clinical trial.
Setting: School of Physical Therapy and Rehabilitation.
Subjects: Forty women with general migraine attending the Neurology
Department of the Faculty of Medicine Faculty of Dokuz Eylül University.
Intervention: Patients were assigned alternately into two groups: exercise
group undertaking 1 hour aerobic exercise three times weekly, and a control
group.
Main outcome measures: Patients were assessed before and after treatment
using three clinical scales – visual analogue scale for headache, Pain Disability
Index and Quality of Life Scale – and chemiluminescence analysis for plasma
nitric oxide.
Results: After the eight-week therapy period, patient complaints concerning
the intensity, frequency and duration of pain had decreased signiŽcantly in
both groups; however, visual analogue scale scoring showed better pain relief
in the exercised group than in the controls (from 8.8 ± 1.7 to 4.0 ± 1.4 and
from 8.5 ± 0.8 to 7.0 ± 0.9 respectively). Quality of life measures also
revealed better migraine relief in the exercised women than in those who
received medical treatment only. Blood NO rose signiŽcantly from pre- to
post-therapy in the exercised group, but the change was not signiŽcant in the
control group.
Conclusion: The study showed that regular long-term aerobic exercise
reduced migraine pain severity, frequency and duration possibly due to
increased nitric oxide production.

Address for correspondence: Selnur Osün Narin, Fizik Tedavi

ve Rehabilitasyon Yüksekokulu, Inciralti, 35340, Izmir,
Turkey. e-mail: selnur62@hotmail.com. or lamia@deu.edu.tr

© Arnold 2003 10.1191/0269215503cr657oa


Introduction
There are many theories about the pathophysi-
ology of migraine attacks. Most of these involve
hereditary1,2 biological states that may cause
increases in free fatty acids and blood lipids,
increased platelet aggregation, decreased sero-
tonin levels and increased prostaglandin levels.3
Such changes can cause the vasodilatation that
precedes migraine headache.1,3 Before the pain
attack, cerebrovascular spasm induced by factors
such as menstruation, lack of sleep, skipped
meals, allergic reaction, and physical or mental
stress, among others,1 initially causes ischaemia
which, in turn, produces prodromal symptoms,2
such as nausea, photophobia or phonophobia.4
Headache experts have proposed that regular,
moderate aerobic exercise improves cardiovascu-
lar Žtness and helps to reduce the frequency,
severity and duration of migraine attacks.5,6 Dur-
ing physical exercise, cardiac output increases
and blood is widely distributed to the working
muscles. This increased tension in the muscle vas-
culature facilitates the production of nitric oxide
(NO).7 NO is an endothelium-derived relaxing
factor (EDRF)-like substance8,9 and relaxes vas-
cular smooth muscle in the systemic and cerebral
circulation.10,11 Jungersten et al. found that both
acute exercise and physical Žtness enhance NO
formation.12,13 Regular exercise has been shown
to upregulate the expression of the endothelial
nitric oxide synthase (NOS) gene in vascular tis-
sue.12,14–16 Repeated exercise also helps NO main-
tain higher blood levels between exercise
sessions.16 Increased production of this vasodilat-
ing agent might partially explain how physical Žt-
ness beneŽts cardiovascular health,5,12,13,17,18
However in apparent contrast to the above-
mentioned vascular beneŽts, some case studies
have suggested that exercise may precipitate
migraine attacks due to excess NO produc-
tion.7,9,19,20
In this study, we explored the effects of aero-
bic exercise with respect to migraine severity and
the triggering of migraine attacks. We also
assessed exercise-related changes in blood NO
levels, and the association of such changes with
changes in migraine attacks.
Effects of exercise on migraine 625
Methods
The study included 40 female participants among
the 320 patients with headache in both sexes who
had applied to the headache outpatient clinic of
the Neurology Department of Dokuz Eylül Uni-
versity Faculty of Medicine from October 1998 to
January 1999. We decided to select the patients
with common migraine from the patients having
other types of migraine and headache. Seventy of
them had been diagnosed with common migraine
headache without aura. The criteria for selection
of the subjects among the other migrainous
patients were according to the deŽnition of the
International Headache Society of ‘common
migraine without aura’.21 According to this deŽ-
nition Žve attacks, each lasting 4–72 hours and
having two of the following four pain character-
istics: unilateral location, pulsating quality, mod-
erate-to-severe intensity, and aggravation by
routine physical activity, were needed to have
‘common migraine’.
Twelve men with similar symptoms mentioned
above were excluded from the study in order to
standardize the group to women. Six women did
not accept the therapy with exercise. Fifty-two
women agreed to participate in the study. Forty
of them were found to Žt the inclusion criteria
mentioned below and were selected for the study.
All participants gave their informed consent.
The University’s Health Science Ethics Commit-
tee approved the protocols.
The inclusion criteria were as follows:
1) Diagnosis of migraine with at least four
attacks per month before therapy
2) No systemic disease that contraindicated
exercise
3) Age between 20 and 50 years
4) Not having a history of sports (not being a
sports women)
5) Willing to participate in the exercise group
6) Receiving similar medical treatment (ergota-
mine, simple analgesics, antiemetics, suma-
triptan) with others during the study.
Each individual was assigned to one of two
groups according to their registration sequence.
Alternate patients joined the exercise group or
the control group. The 20 controls received only
medication for eight weeks. The other 20 women
626 SO Narin et al.
were placed on a moderate aerobic training pro-
gramme in addition to medical treatment for the
same period. The programme involved 1 hour of
exercise per day three times weekly.22 All exer-
cise was performed in the Žtness unit of the
School of Physical Therapy and Rehabilitation at
Dokuz Eylül University, and was supervised by a
physiotherapist. The mean age in the exercise
group was 35.20 ± 10.23 years, and that in the
control group was 40.0 ± 8.3 years. Both groups
were assessed at baseline, and then again after
the eight-week period. They all followed up the
therapies.
The exercise protocol was:
• 5 minutes warm-up exercises
• 10 minutes cycling with constant speed
• 10 minutes walking on treadmill
• 5 minutes stepper
• 10 minutes training upper extremities at the
power station
• Neck and postural exercises with 10 repeti-
tions
• Rowing with 10 repetitions
• 5 minutes cooling exercises.
Each participant completed a questionnaire that
revealed her age, height, weight, occupation and
smoking habits, as well as the severity, frequency,
duration and location(s) of her headaches. The
survey provided information on physical charac-
teristics, existing risk factors, quality of life and
current migraine status.
As noted above, each woman was evaluated at
baseline and after treatment. The features of the
headache were assessed using the visual analogue
scale (VAS), and quality of life was assessed
using the Pain Disability Index (PDI) and a
severity scale.23,24 Direct measurement of NO was
detected by means of a chemiluminescence NO
analyser.25
Pain severity
The VAS has been shown to be reliable for
measuring pain.23 Each subject gauged her
headache intensity on a 100-mm VAS that
ranged from no pain (0) to very severe pain
(100).
Disability
The PDI is a practical and quick self-reporting
instrument that is used to measure how much a
patient’s chronic pain interferes with normal
daily functioning. To complete the PDI, the
respondent uses an 11-point scale that ranges
from 0 (no problems due to pain) to 10 (total dis-
ability due to pain) to rate the degree of inter-
ference in the following seven areas: family/home
responsibilities, recreation, social activities, occu-
pation, sexual activity, self-care and life support
activity. The results for the seven items are then
added together to give a total PDI score, which
ranges from 0 to 70.23
Grading the severity of chronic pain
Quality of life scale was used to grade the
severity of chronic pain. This scale uses scores for
disability and pain parameters to form a hierar-
chical scale.24 The parameters are time since
onset, days in pain, characteristic pain intensity,
disability score and number of disability days.
Combining the disability and pain Žndings, the
severity of chronic pain is categorized as I, II, III
or IV, as follows: I = low disability–low intensity
pain (<50); II = low disability–high intensity pain
(³50); III = high disability–moderately limiting
(3–4 disability points); and IV = high disabil-
ity–severely limiting (5–6 disability points).
For pain intensity, patients marked a number
on a 0–10 scale where 0 is ‘no pain’ and 10 is ‘pain
as bad as could be’. First item was about at the
present time. Second item was about the worst
pain experienced in the past six months. The
third item was about the average pain intensity
in the past six months. The next four items exam-
ined the pain-related disabilities in the past six
months. Patients marked a number on a 0–10
scale where 0 is ‘no interference’ and 10 is
‘unable to carry on any activities’ for being kept
from usual activities (work, school or house-
work), daily activities, taking part in recreational,
social and family activities. Scoring was calcu-
lated from the ‘characteristic pain intensity’,
which is a 0–100 score derived from questions
1–3: mean (pain right now, worst pain, average
pain) ´ 10.
Disability score was a 0–100 score derived from
questions 5–7: mean (daily activities, social activ-
ities, work activities) ´ 10. For disability points,
 Effects of exercise on migraine 627

the indicated points for disability days (question
4) and for disability score were added.
Nitric oxide analysis
Blood samples from both groups were col-
lected at the same time in heparin tubes in the
mornings. Plasma was immediately separated by
centrifugation for 10 minutes at 1500 ´ g. Plasma
samples were stored at –20°C until they were
analysed. Because nitrate is rapidly reduced after
heating to 80–90°C to NO in acidic medium con-
taining vanadium (III), the total amount of
nitrate in each sample of plasma was determined
using a modiŽed version of the procedure
described by Braman and Hendrix.25 The purge
system of a Sievers Instruments Model 280A
Nitric Oxide Analyzer (Ionics Instruments,
Boulder, Co, USA) was used for the measure-
ments. First, the sample was deproteinized with
chilled 96% ethanol (plasma/ethanol = 1/2
vol/vol). A saturated solution of vanadium III
chloride (VCl3) in 1 M HCl (800 mg/dL) was pre-
pared and Žltered. Then 5 mL of this reagent
were added to the purge vessel and purged with
nitrogen gas for 10 minutes. The purge vessel was
equipped with a cold-water condenser and a
water jacket to permit heating of the reagent to
95°C in a circulating water bath. The HCl
vapours were removed by a gas bubbler contain-
ing 15 mL of 1 M NaOH, and the rate of gas ow
into the chemiluminescence detector was con-
trolled with a needle valve. Each sample was
injected into the purge vessel to react with the
VCl3/HCl reagent, which converted the existing
nitrate to NO. The resultant NO was then
removed from the reaction chamber, and the con-
centration was detected by ozone-induced chemi-
luminescence in the chemiluminescence detector.
Findings were recorded in parts per billion (ppb),
and the NO level in each sample was calculated
from the peak values of the records. A standard
curve was constructed using various concentra-
tions of nitrate (10–100 mM). The nitrate level in
each sample was calculated using this standard
curve. All results of measures were expressed as
mean ± SD.25 Analysis of NO was performed in
Physiology Department of Medicine Faculty of
Gazi University, blindly.
Results
Both groups had similar measures at baseline.
There were no signiŽcant differences between the
two groups before intervention except disability
points, which change due to personal features.
After intervention there were statistically sig-
niŽcant reductions in the frequency of pain, the
Pain Disability Index, the disability points in days
and in the nitrous oxide level in both group (all
p less than 0.05). In comparison, the exercise
group exhibited signiŽcantly more pain relief
than the nonexercised controls (p < 0.05). Table
1 shows the relationships between intensity
(according to VAS) duration and frequency
(according to the questionnaire forms’ score) of
pain and NO levels in the two groups before and
after treatment.

Table 1 Mean (SD) levels of pain, disability and plasma NO levels before and after therapy

Exercise group (n = 20) Control group (n = 20)

Variable Before After Before After

Pain intensity (VAS) 8.8 (1.7) 4.0 (1.4) 8.5 (0.8) 7.0 (0.9)
Pain duration (hours) 34.7 (22.1) 11.4 (7.3) 40.3 (26.1) 27.2 (16.9)
Pain frequency (days) 7.4 (2.9) 3.6 (1.6) 8.9 (3.3) 7.0 (2.4)
Pain Disability Index 46.1 (7.9) 20.4 (4.9) 49.2 (6.7) 41.3 (6.5)
Quality of life scale
Pain intensity 84.2 (8.1) 32.6 (20.2) 84.5 (6.9) 75.5 (6.0)
Disability score 72.6 (5.2) 26.2 (7.1) 84.5 (6.9) 75.0 (6.1)
Disability points 4.4 (4.9) 0.3 (0.7) 1.6 (1.5) 0.7 (1.1)
Nitric oxide
Nitric oxide (ng) 13.52 (3.62) 19.63 (5.28) 16.2 (6.03) 13.17 (6.00)
628 SO Narin et al.
Discussion
Exercise seems to reduce the frequency and
severity of migraine headache. During our study
period, the subjects with migraine declared that
exercise brought tranquility, relaxation and a
good mood in addition to the decrease in pain
attacks in every session. Migraine headaches are
triggered by distension of cerebral arteries, large
veins and venous sinuses, which is caused by the
various biological factors noted above.7,19 Exper-
iments have demonstrated that aerobic exercise
signiŽcantly improves cardiovascular Žtness, reg-
ulates vessel tone, and decreases migraine
attacks.22,26,27 However, it is also known that rel-
atively intense exercise can provoke migraine
headache.6,7,28 Some investigators have suggested
that this might be linked to an acute rise in the
blood NO level.29 Sarchielli demonstrated
increased L -arginine/NO pathway activity in
platelets from patients who suffered from
migraine with or without aura, and this effect was
particularly striking during the attacks.30 In a
study that included athletes and nonathletes,
Jungersten et al. showed that 2 hours of physical
exercise was associated with an 18% rise in
plasma nitrate (a major stable end-product of NO
metabolism) above resting levels.12 Other authors
have also reported a transient rise in blood
nitrate concentration after exercise.12,31
The main limitation of our study is the alter-
nate and nonrandom allocation of patients. How-
ever we do not believe that this has introduced
bias: the fact that all patients expressed a desire
to undertake exercise reducing the chance of
Clinical messages
• Regular submaximal aerobic exercise may
reduce the pain and disability associated
with migraine.
• Exercise might have its beneŽcial effect
through increasing production of nitric
oxide.
• However, this study was limited to women
with common migraine and needs replicat-
ing.
selection bias. And the two groups were similar
at entry.
Our study assessed the effects of combining
moderate-level aerobic exercise with medical
therapy in 20 migrainous women. The 20 controls
received only medical treatment. The VAS
results at baseline showed similar migraine fea-
tures in both groups. After eight weeks of ther-
apy, we observed signiŽcant improvements in
both groups (Table 1). However, the groups’
mean VAS scores at the end of the treatment
period revealed that exercise combined with
medical therapy was more effective at relieving
migraines than medical therapy alone.
Similarly, the Žndings for the PDI and quality
of life scale, both of which reect pain-related
limitations on daily living, showed signiŽcant
decreases in severity of pain, monthly pain fre-
quency and incapacity due to pain after eight
weeks. As with the VAS results, the improve-
ments in PDI and GS were more marked in the
exercise group than in the nonexercise controls
(Table 1).
Although exercise alone cannot prevent or
cure disease, it can relax muscles, improve circu-
lation and help the headache sufferer feel better
in general.22 As explained previously, exercise
partially restores ow-mediated dilatation
(FMD) in the vessels by increasing the expres-
sion of endothelial NOS.16 We determined the
subjects’ plasma NO levels at rest before and
after the eight-week treatment period. The base-
line values for the exercise and control groups
were similar (13.52 ± 3.62 ng versus 16.2 ± 6.03
ng, respectively; p > 0.05), but the exercise group
had signiŽcantly higher post-therapy levels than
the controls (19.63 ± 5.28 ng versus 13.17 ± 6.0
ng, respectively; p < 0.05). The signiŽcantly
higher plasma NO levels in the exercised group
(Table 1) were not associated with more intense
or more frequent migraine attacks during the
therapy period. In fact, the women who were on
the combined programme reported that they felt
better at the end of therapy. We did not prescribe
strenuous exercise, which would have caused a
sudden rise in NOS production. Rather, we used
a moderate, submaximal exercise model for an
extended period, which keeps NO levels ele-
vated.
Similar to our results, Lockett found that a six-

week aerobic exercise programme enhanced the
overall well-being of 11 subjects with common
migraine.5 The author showed that exercise
improved cardiovascular Žtness and decreased
the severity, frequency and duration of headache
pain. Considering the possible mechanisms at
work, a group of Japanese investigators reported
that regular aerobic exercise causes endothelium-
dependent vascular relaxation in both hyperten-
sive and normotensive individuals by enhancing
long-term NO release.32 This maintenance of
plasma NO at elevated levels might partially
explain the beneŽcial effects of physical Žtness on
cardiovascular health.18
In addition to increasing NO, regular exercise
may also protect the endothelium by reducing the
concentration of plasma norepinephrine32 and
preventing the production of vasoconstrictor
prostanoids and free radicals in vessel walls.16
These NO effects may protect against the trig-
gering of migraine attacks by vasospasm of the
cerebral arterioles.
Large population-based studies have docu-
mented that individuals with migraine consis-
tently report lower levels of mental, physical and
social well-being than unaffected controls. It
seems logical that migraine patients would be
depressed because of their pain, and migraine
and depression also have speciŽc effects on qual-
ity of life.33 Exercise tones the blood vessels that
are involved in migraine attacks and improves
overall health. Such changes can lead to a longer,
healthier happier life. The additional beneŽts of
exercise include reduced stress, improved sleep-
ing and more rapid recovery from illness.27
References
1 Alexandrea J. Managing migraines. Diabetes
Forecast 1998; 51: 23.
2 Lance JW, Goodsby PJ. Mechanism and
management of headache, sixth edition. Oxford:
Butterworth-Heinemann, 1998: 40–45.
3 Bic Z, Blix GG, Hopp HP, Leslie FM. In search of
the ideal treatment for migraine headache. Med
Hypotheses 1998; 50: 1–7.
4 Sadovsky R, Evaluation and management of
headache symptoms. Am Fam Phys 1998; 58:
523–24.
5 Lockett DM, Champbell JF. The effects of aerobic
exercise on migraine. Headache 1992; 32: 50–54.
Effects of exercise on migraine 629
6 Diamond S. Managing migraines in active people.
Physician Sportsmed 1996; 24: 1–11.
7 Thomsen L, Olesen J. Nitric oxide theory of
migraine. Clin Neurosci 1998; 5: 28–33.
8 Stefanovic-Racic J, Stadler J, Evans CH. Nitric
oxide and arthritis. Arthritis Rheum 1993; 36:
1036–43.
9 Ashina M, Lassen LH, Bendtsen L, Jensen R,
Olesen J. Effect of inhibition of nitric oxide synthase
on chronic tension type headache: a randomised
crossover trial. Lancet 1999; 353: 256–57.
10 Erba D. Nitric oxide. Gazi Med J 1998; 9 (suppl 1):
1–11.
11 Bredt DS. Endogenous nitric oxide synthasis:
biological functions a pathophysiology. Free Radic
Res 1999; 31: 577–96.
12 Jungersten L, Ambring A, Wall B, Wennmalm A.
Both physical Žtness and acute exercise regulate
nitric oxide formation in healthy humans. J Appl
Physiol 1997; 82: 760–64.
13 Bode-Boger SM, Boger RH, Schroder EP, Frolich
JC. Exercise increases systemic nitric oxide
production in men. J Cardiovasc Risk 1994; 1:
173–78.
14 Bradley WG, Daroff RB, Fenichel GM, Marsden
CD. Neurology in clinical practice: principles of
diagnosis and management, second edition, Volume
2. Oxford: Butterworth-Heinemann, 1996:
1695–701.
15 Olesen J, Moskowitz MA. Experimental headache
models. Philadelphia: Lippincott-Raven, 1995:
207–11.
16 Varin R , Mulder P, Richard V, Tamion F. Exercise
improves ow-mediated vasodilatation of skeletal
muscle arteries in rats with chronic heart failure:
role of nitric oxide, prostanoids and oxidant stress.
Circulation 1999; 99: 2951–57.
17 Sessa WC, Pritchard K, Seyedi N, Wang J, Hintze
TH. Chronic exercise in dogs increases coronary
vascular nitric oxide production and endothelial cell
nitric oxide synthase gene expression. Circ Res 1994;
74: 349–53.
18 Bronwyn AK. Nitric oxide-mediated metabolic
regulation during exercise: effects of training in
health and cardiovascular disease. FASEB J 2000;
14: 1685–96.
19 Bartleson JD. Treatment of migraine headache.
Mayo Clinic Proc Jan 1999; 1–5.
20 Ashina M, Bendtsen L, Jensen R et al. Possible
mechanisms of action of nitric oxide synthase
inhibitors in chronic tension type. Brain 1999; 122:
1629–35.
21 Headache ClassiŽcation Committee of the
International Headache Society. ClassiŽcation and
diagnostic criteria for headache disorders, cranial
neuralgia and facial pain. Cephalalgia 1988; 8 (suppl
7): 1–96.
22 Chillot R. Stop headache . . . like magic. Prevention
630 SO Narin et al.
1998; 50: 98–108.
23 Tait RC, Chibnall JT, Krause S. The pain disability
index: psychometric properties. Pain 1990; 40:
171–82.
24 Kroff MV, Ormel J, Keefe FJ, Dworkin SF. Clinical
section. Pain 1992; 50: 133–49.
25 Braman RS, Hendrix SA. Nanogram nitrite and
nitrate determination in environmental and
biological materials by vanadium (III) reduction
with chemiluminescence detection. Anal Chem 1989;
61: 2715–18.
26 Darling M. The use of exercise as a method of
aborting migraine. Headache 1991; 31: 616–18.
27 Swain RA, Kaplan B. Diagnosis, prophylaxis and
treatment of headache in the athlete. South Med J
1997; 90: 878–88.
28 Indo T, Takahashi A. Swimmer’s migraine.
Headache 1990; 30: 485–87.
29 McCrory P. Recognizing exercise-related headache.
Physician Sportsmed 1997; 33–39.
30 Sarchielli P, Alberti A, Russo S et al. Nitric oxide
pathway, Ca2+, and serotonin content in platelets
from patients suffering from chronic daily headache.
Cephalalgia 1999; 19: 810–16.
31 Silberstein SD. Comprehensive management of
headache and depression. Cephalalgia 1998; 18:
50–55.
32 Higashi Y, Sasaki S et al. Clinical abstracts: vascular
disease. Modern Med 1999; 67: 30.
33 Weber W. Neurologists determine the effect of
migraine on quality of life. Lancet 2000; 356:
1007–1009.
34 Badyak M, Blumenthal JA, Herman S et al. Exercise
treatment for major depression: maintenance of
therapeutic beneŽt at 10 months. Psychosom Med
2000; 62: 633–38.