Effectiveness of jogging, eye movement exercises, and diaphragmatic breathing in patients with migraine: a randomized controlled trial

Background: The recurrent and progressive presence of maladaptive stress response-related activities or electrophysiological dysresponsiveness are apparent in a brain with migraine. Objectives: Various methods have been suggested to regulate stress response-related activities in a brain with migraine, including jogging, eye movement exercises, and diaphragmatic breathing. Methods: A three-group, double-blind, and randomized design with pre-test, post-test, and follow-up assessments was used to test the effectiveness of jogging, eye movement exercises, and diaphragmatic breathing in the treatment of patients with migraine. International Classication of Headache Disorders 3rd edition (ICHD-3) was used to diagnose the patients with migraine, who were randomly assigned to one of the three study groups to perform either 12 consecutive weeks of (a) jogging plus eye movement exercises (n = 22) ;(b) jogging plus diaphragmatic breathing (n = 19); or (c) treatment as usual (TAU), receiving prescribed medication (n = 22) group. Results: The results of a series of MANCOVA showed that patients in the two experimental groups showed signicant (p < .05) reductions in the frequency, duration, and intensity of their migraine attacks compared to the TAU group. Moreover, the frequency of menstrual cycle-dependent headache attacks and over the counter drugs use were reduced, and the quality of sleep and drinking water were improved at post-test and a 12-month follow-up. Conclusion: The results suggest that jogging plus eye movement exercises or jogging plus diaphragmatic breathing can be used as effective alternative interventions in the treatment of migraine. to a brain with migraine. synchronizing


Introduction
Alterations in response to daily life stressors and neuronal excitability hint the brain with migraine. 1,2 Examples of such stressors could be observed as maladaptive stress responses in both peripheral and central nervous systems. 3,4 The migraine pain also depends on the activation and sensitization of rstorder or cranial neurons and sensitization and altered habituation of second-order or cortical neurons. The activation or sensitization is determined by the quantity, duration, or presence of other elements in the brain (e.g., pituitary adenylate cyclase-activating peptide and calcitonin gene-related peptide) that can lead to (a) genetic modulation; (b) modulating cells communications (e.g., transcription factors); and (c) cerebral blood ow (CBF) or vascular tone alteration (i.e., vasodilation) . 4 For instance, eye-brain electrochemical transduction and retinal nerve ber alterations, structural and functional abnormalities in the endothelial progenitor cells, interlinks between mitochondrial energy production, reactive oxygen species (ROS) availability (e.g., Nitric Oxide-NO), and ion homeostasis (e.g., calcium in ux) seem to play key initiating roles in a brain with migraine. 4-8 Therefore, regulation and availability of signaling biomolecules (e.g., NO and brain-derived neurotropic factor-BDNF) and regulation of signaling pathway (e.g., cAMP-response element binding protein-CREB) are essential during health and illness, including migraine attacks, as stressful insults to the brain. 9,10 Although non-steroids anti-in ammatory drugs (NSAIDS), antidepressants, triptans, gepants, and anti-epileptic drugs are the frequently used medications for migraine attacks, most of these medications have low therapeutic gain, low resolution of symptoms, and undesirable side effects. 11,12 For example, in the presence of triptans or morphine, NO boosts the activation of trigeminal pathways or increases the risk of medication overuse headache (MOH). 13,14 Nevertheless, the chronic administration or high doses of antidepressants suppress the expression of the signalling biomolecule genes. 15 Moreover, the manipulation of BDNF gene expression may contribute to the development of other disorders or diseases including anxiety, depression, or chroni cation of migraine. 16 Considering the complications, developing novel, safe, multidimensional, and non-pharmacological interventions appears to offer a promising method for decreasing the load of the stress insults in a brain with migraine. 17 The effects of regular aerobic exercises or progressive relaxation techniques on pain have been documented in previous studies. [18][19][20] Therefore, the present study takes the assumption that regular aerobic exercises in adjunct to visual exercises, exercise-induced, nasal, or diaphragmatic breathing maintain the brain homeostasis, improve the brain function, protect the brain against various-induced toxicities, and relieve pain experience. Inhibition of excessive presence of N-Methyl-D-Aspartate (NMDA) receptors or regulation of serotonin or glutamate excitatory signaling, and promotion of BDNF gene expression are the expected positive outcomes in the present study. [21][22][23][24][25][26][27][28][29][30][31] That is, through regulation of local and cortical endothelium releasing factors (e.g., NO and BDNF), regular aerobic exercises, nasal breathing, and visual exercises facilitate and re-arrange cortical and oscillatory feedback signaling between cranial motor nerves, brain stem, thalamus, hypothalamus, motor, and sensory cortices, all of which are closely related to a brain with migraine. [32][33][34][35][36][37] Methods

Study population
First, a power analysis was conducted the sample size needed for the present effectiveness study.
Considering the similar studies in the eld and the novelty of both experimental interventions, a priori effect size of .175 (midpoint between small and medium effect size for MANOVA models) deemed suitable for the present study. In the G*Power software, Power was entered as . 80 (Cohen, 1992), alpha α = .05, number of groups k = 3, and response variables = 3 (i.e., pain frequency, duration, and intensity), which yielded a total sample size of n = 45. Because in a study with a 12-month follow up attrition rate could be as high as .40% then 63 participants deemed necessary to be included at the baseline assessment. All participants in the nal sample were right-handed 38  All the methods were performed in accordance with the Guidelines of the International Headache Society for controlled trials of preventive and prophylactic treatments of migraine in adults. 44,45 Experimental design The present study (3×3 factorial design; three groups and three dependent variables) is a randomized, three-armed, and double-blind controlled trial. Although it was an ethical necessity for the participants to know about the nature of the study, neither the participants nor the research assistants responsible for interventions and data collection were not aware of their group allocation. The study aimed to investigate whether jogging plus eye movement exercises or jogging plus diaphragmatic breathing could be effective in alleviating migraine symptoms compared to a control group that used prescribed medications for migraine (i.e., treatment as usual).
All participants had to meet the inclusion criteria and be exible about the terms of the study after their diagnosis was con rmed by two consultant neurologists at a local state hospital based on ICHD criteria 42 . The entire sample was tested over a 46-week period between August 2016 and July 2017. By using a paper-based, baseline headache diary, the sample was screened for frequency, duration, and intensity of attacks in the last three months. 46,47 Next, using the random digit table, participants were randomly (1:1:1) assigned to one of the experimental groups, i.e., eye movement exercises (n = 22) or diaphragmatic breathing (n = 19) or to the control group (n = 22). All the data were recorded by three research assistants who were blind to the study design. The data analysis was also conducted by one of the co-authors, who was blind to the study design and its hypotheses.

Randomization procedure
Because the present study was based on a double-blind design, clinicians, experimenters, and research assistants who were working with the patients were not aware of the nature of the study design and procedures. Participants were informed only about what they were expected to do but not about their group allocation and other participants in the study, and the study design and goals. To reduce the risk of prediction for a sample size of n > 50, we used telephone (secure) allocation of participants by an independent person in line with instructions for simple random method 48 . Before proceeding to the next stage, the inclusion and exclusion criteria of the present study were monitored and checked. For example, if a participant was on a prescribed or non-prescribed medication, they were recommended to stop their medication in a tapering manner (i.e., over a 6-week period). Moreover, patients were asked and screened for any types of smoking, however, no participant were deemed to have smoking habit.
One experimenter was randomly assigned to each group to execute the participants' enrollment and interventions' assignments in each group during study. By using a paper-based diary, experimenters were instructed and scheduled to monitor and record medications, menstrual cycles, sleeping, drinking water, and exercise intolerance behaviors in the experimental and control groups. The participants were told that their participation was voluntary, and they could withdraw from the experiment at their will. They were told that they are part of a study on the migraine.

Interventions
Jogging plus eye movement exercises or jogging plus diaphragmatic breathing instruction There were two experimental groups. The rst experimental group performed jogging plus eye movement exercises. The second experimental group performed jogging plus diaphragmatic breathing instruction.
Jogging required circling around the outer ring of a local public park for 3.2 km (≈ 25-min). During the jogging, they were instructed to follow a breathing rhythm that required two relaxed, deep nasal inhalations, a repose, and two deep oral exhalations. The jogging was exercised every other day (between 5:00 p.m. and 7:00 p.m.) over a period of 12 weeks.
The eye movement exercises included horizontal (dextroversion / levoversion) and vertical (sursumduction / deorsumduction) conjugated eye movements 49 . Participants performed a 5-min eye movement exercise 30-min before having their breakfast every morning and 30-min before going to bed at night for 12 weeks. The 5-min bedtime phase was initiated by 15-round movements of the right-hand index nger, followed by 15-round left-hand index nger movements in the horizontal visual eld, which was followed by the same procedure in the vertical visual eld with a front-forward, xed head with the eyes following the nger in a constantly controlled manner ( Figure 1). The 5-min morning phase was initiated, sustained, and terminated with the same procedure, but instead of moving the index nger in the visual eld, the head was moved horizontally (30 rounds laterally) and then vertically (30 rounds up and down) with xed eyes on a xed index nger in the visual eld ( Figure 2). 50, 51 Figures 1 and 2 show participants' hands and head-guided eye movement exercises in the experimental groups.
The diaphragmatic breathing practice required a deep nasal inhalation, a two-second repose, and a long, deep oral exhalation three times a day around 7:00 a.m., 2:00 p.m., and 9:00 p.m. for 5-min for 12 consecutive weeks. Participants were persuaded to focus on their breathing while inhaling and exhaling. 52,53 During the eye movement exercises and diaphragmatic breathing practice, participants were asked to sit in an upright position and plant their feet on the ground at a 90-degree angle to match the extension of their shoulders. Moreover, they were instructed to x their head horizontally, leveled with their body. They were also asked to arch their back completely and slouch forward. The position was followed by rolling shoulders back and dropping them down. The preparation phase was ended by a deep breath and relaxation ( Figure 3). 54 Tutorial sessions After the randomization and group allocation, two research assistants who were blind to the objectives and nature of the study were instructed to teach the participants how to perform eye movement exercises and diaphragmatic breathing, in two face-to-face sessions (each session = 30-min). Although jogging was taught in a separate and a face-to-face session (30-min) by another a research assistant, for everyone, for each study group, jogging was accompanied, performed, and monitored every session for each experimental group by two speci c research assistants.

Control group
The control group was receiving medical treatment as usual, which solely consisted of common prescribed medication to help with the migraine symptoms. Because there were a variety of prescribed prophylactic medications (e.g., Propranolol, Imipramine) and their effects on the migraine symptoms were not a goal for the present study, it was not necessary to keep a record of the type of medication that participants were using over the course of the study. Therefore, like the experimental groups, they were monitored regularly to reduce the attrition rate, and completed all the study measures at the baseline, post-test, and the follow-up assessments.

Procedure
To measure the duration, frequency, and intensity of pain, the study measures were administered at the baseline, post-test, and a 12-month follow-up-by using a baseline headache diary over three months. Moreover, weekly telephone follow-ups were made to enquire about the participants' experiences of pain, use of prescribed or non-prescribed medication, menstrual cycle, sleeping pattern, drinking water, or other related behaviors for both experimental and control groups. The telephone calls also helped minimize the participants' dropouts. Participants were also reassured that they could have control visits during the study in case of experiencing migraine attacks or complications that could affect the intake of regular prescribed or non-prescribed medication. However, none of the participants in the experimental groups reported referring to their physician or resuming the use of migraine prescribed or non-prescribed medication at any of phone calls or follow-up assessments in the experimental groups. Therefore, there was no data on prescribed or non-prescribed medication relapse for the experimental groups. The same experimenters administered all study measures at the baseline, post-test, and the follow-up assessments.

Measures
Headache diary A paper-based and baseline headache diary was used 46,47 to record the frequency, (i.e., one attack per four weeks, one attack in two weeks, two or three attacks per week, or more than three attacks per week), duration (i.e., 4, 4-24, or 24-72 hours per attack), and the intensity of pain (i.e., moderate (1-3), severe (4-7), and worse possible case (8-10)) per migraine attack during the past four weeks-over three months. The measure has desirable reliability and validity indices 46,47 .

Compliance with intervention
Based on recorded data and the present study protocols, which included 36  Sleeping hours included sleeping less than or more than seven hours per night (between 10:00 p.m. and 7:00 a.m.) 57,58 . In addition, drinking water was also recorded for the participants 59 . However, any incidents of muscle cramps, fatigue, chronic dizziness, and vomiting, which were related to exercise intolerance were recorded. The data was recorded every other day during the experiment and every week during the 12-month follow-up for experimental and control groups. The rationale for recording and analyzing this data was to test whether the interventions in uenced such variables beside migraine pain characteristics.

Data analysis
To test the relative effect of jogging plus eye movement exercises or jogging plus diaphragmatic breathing, three multivariate analyses of covariance (MANCOVA) were conducted. To ensure that there were no missing data, initial data evaluation was performed for experimental and control groups. Next, assumptions of homogeneity, linearity, and normality of variance-covariance matrices were calculated and a p-value ≥ .05 was set for all analyses. The results of models' tness tests did not reveal any violation of underlying assumptions. In each model, group (three levels) was entered into the model as factor, pain characteristics (i.e., frequency, duration, and intensity of pain) at baseline were entered as covariates, and pain characteristics at the post-test and follow-up were entered into the model as dependent variables. Moreover, as secondary analyses, Cochran's tests were calculated to measure the effects of the two types of interventions on menstrual cycle, OTC, sleep, and drinking water as behavioral outcomes from baseline to post-test and the 12-month follow-up. Table 2 shows means and SDs of frequency, duration, and intensity of pain in each study group at baseline, post-test, and the 12-month follow-up assessments.

Primary data analyses
In (2.91)) of the pain at the post-test and follow-up assessments. Pairwise comparisons showed that, compared to the control group, the experimental groups showed signi cant improvements in all pain characteristics at the post-test and follow-up assessments (Table 3). There was no signi cant difference between the two experimental groups on any of the pain characteristics. The effect sizes for all models were greater than Cohen's d = .80 for a large effect size (Table 3). Figure 4 shows changes in the migraine pain across assessment points for each group. Figure 4. Migraine pain (frequency; ranges 1-9 attacks, duration; ranges 1-8 hours/attack, and intensity; ranges a scale of 1-7/attack during the last four-week) changes in each study (EM = eye movement exercises, DB = diaphragmatic breathing, control) group across three assessment points (pre-test, pot-test, and follow-up).
The results of a series of Cochran's tests were used to test changes in the patients' menstrual cycle, OTC, sleep pattern, and drinking water across the assessment points and showed signi cant improvements for the experimental groups but not for the control group (

Discussion
The present study tested the effectiveness of jogging, eye movement exercises, and diaphragmatic breathing in patients with migraine. The rst experimental group received jogging plus eye movement exercises and the second experimental group received jogging plus diaphragmatic breathing, whereas the control group only received treatment as usual. The results showed signi cant effects of the interventions on the patients' symptoms in the experimental groups ( Figure 4 and Table 3). Therefore, regular aerobic exercises (i.e., jogging) and practicing body-mind awareness (i.e., eye movement exercises and diaphragmatic breathing) appear to elucidate the negative effects of stressful insults in the brain, on the body muscles, respiratory rate and depth, and the ow of biomolecules and hormones (i.e., Nitric Oxide (NO), brain-derived neurotrophic factor (BDNF), and adenosine monophosphate-activated protein kinase (AMPK)). These effects could stem from various possible mechanisms, including (a) brain-gut and neuroimmune axis (e.g., hypothalamus-pituitary-adrenal (HPA); (b) brain-brain reciprocal interlinks (e.g., hippocampus-anterior cingulate cortex and medial-prefrontal cortex (mPFC) oscillations); and (c) circadian disruption and biomolecular dysregulation (e.g., oxidative stress and neurogenic in ammation), each of which may play a key initiating role in spurring migraine-related symptoms [60][61][62][63] .
Although the results of a meta-analysis 20 suggested that the effect of aerobic exercise on patients with migraine is only related to the frequency of the attacks, in a recent study 64 , aerobic exercises are assumed to in uence all aspects of migraine pain including frequency, duration, and intensity. The paradoxical ndings in the literature on the effectiveness of aerobic exercise could be related to the following reasons. First, based on the International Headache Society (IHS) recommendations 45,47 , one major drawback of previous studies is heterogeneity in applying headache-related outcome measures. Second, some studies have failed to adequately address issues related to blinding, sample size, designing, and randomization. Third, no study has compared the outcomes of aerobic exercises on patients with migraine without aura (MwoA) and patients with migraine with aura (MwA). Forth, evidence was scarce for comparing the outcomes of aerobic exercises and pharmacological treatment of migraine pain. Fifth, in no study so far aerobic exercise has been complemented with another non-medication intervention; therefore, this was the rst study that, based on ndings from the brain with migraine studies, combined jogging with eye movement exercises or diaphragmatic breathing to decrease kinesiophobia or sensitivity to movements in patients with migraine 65,66 .
As a diagnostic criterion, patients with migraine have a fear of movement or kinesiophobia [67][68][69] . To develop readiness for and potentiate neural responsiveness to a full-body exercise and to improve body awareness 52, 53 , the present study provided the patients with eye movement exercises or diaphragmatic breathing [70][71][72] . Considering the neural or Hebbian learning, stimulating brain on a regular basis adjusts brain functions at the molecular level (e.g., signaling mechanisms and memory formation) 73 . Eye movement exercises are identi ed to activate widespread, interconnected cortical and subcortical networks such as superior colliculus, oculomotor network (ON) in the brain stem, dorsolateral prefrontal brain cortex, basal ganglia, subthalamic structures, substantia nigra pars reticulata, visuo-motor, parietal, and posterior cingulate cortices [74][75][76][77][78] . Each of these networks plays an excitatory or inhibitory role in pain processing 79,80 . Therefore, performing eye movement exercises on a regular basis stimulates or potentiates almost all major parts of the brain that are important in the experience of the pain. The corneoretinal potential (CRP) transduction by eye movement exercises is a bioelectrical signal, which is produced by two modes of polarities during eye movement exercises: a positively charged end (cornea) and a negatively charged end (retina) 81 . The conjugated eye movement exercises (i.e., from right to left or from top to bottom and the vice versa) transduces sequential negative or positive electrical charges on the retina 82 . These types of training-induced transduction should potentiate synaptic plasticity and facilitate myelination mechanisms by (a) regulating signal transmission (i.e., neuromodulation of ion channels and neurotransmitters); (b) synchronizing oscillations (i.e., frequency, amplitude, or phase); and (c) cortical reorganization or enhancement of timing within related cortical-cortical or cortical-subcortical brain networks (i.e., enhancement of memory, visuospatial accuracy, decision-making, or task switching) [83][84][85][86][87] .
In the present study, another essential guided behavior was diaphragmatic breathing. It should be noted that jogging itself may increase exercise-induced stress insults in brain with migraine; however, diaphragmatic breathing helps with promoting stress-response control indices and potentiating cortical readiness prior to jogging 88,89 . Diaphragmatic breathing can improve brain function via the following mechanisms: (a) chemical, which involves the regulation of biomolecules such as adenosine, melatonin, orexin, or calcitonin-gene-related peptide (CGRP); (b) mechanical, which involves enhancing the rate, length, or intensity of breathing at the cortical level; and (c) cortical-subcortical control, which is related to shared breathing centers and headache-related networks such as brainstem and medullary centers [90][91][92][93] . Together, jogging, eye movement exercises, and diaphragmatic breathing may induce mitochondrial oxidative phosphorylation, electrical transduction, ionic, and biomolecule homeostasis that can explain primary and secondary outcomes of the present study.
All participants in the present study were female patients with migraine. This may limit the generalizability of our ndings to male patients with migraine. Subsequently, the same study should be also conducted with male migraine patients. Moreover, heterogeneity in the rate of prescribed and OTC medication use in the control group was another limitation of the present study. Moreover, we did not ask patients in the control group to keep a journal of any medication that they were taking for their pain because, there is no single or speci c medication with consistent or long-term effect for migraine, hence even recording the rate of medication use over a two-year by the patients in the control group deemed an unviable goal and could not help with interpreting the study outcomes. Furthermore, given that the present study was conducted in a Middle Eastern country with a different environment and culture, future studies can replicate the study with patients from different cultural, racial, and ethnical backgrounds. Future studies can also examine the outcomes of the present interventions for other types of headaches, migraine-related comorbidities (e.g., epilepsy, tinnitus, chronic dizziness, or vertigo (e.g., phobic postural vertigo), sleep apnea, or restless leg syndrome), and even for psychopathologies like depression and anxiety.

Conclusion
The primary outcomes of the present study can be summarized as positive effects of jogging plus eye movement exercises or jogging plus diaphragmatic breathing on decreasing the frequency, duration, and intensity of pain in patients with migraine. The secondary outcomes could be summarized as decreased headache attacks associated with menstrual cycle and use of OTC medication, and on improving sleep and drinking water. Together, interventions used the present study offer promising prophylactic and therapeutic outcomes for patients with migraine.

Clinical implications
Performing 12-week jogging (every other day ≈ 25-min) either with eye movement exercises (60 rounds early in the morning / 60 rounds before the bedtime) or diaphragmatic breathing (3 times a day, each time for ≈ 5-min): can signi cantly reduce the measured aspects of migraine pain characteristics.
offers a safe and convenient intervention.
produces no short-term or long-term undesirable side effects.

Consent for publication
All subjects provided written informed consent for publication prior to the present study enrolment.

Competing interests
The authors declare that they have no competing interests.  Note. EME = Eye movement exercises; DB = Diaphragmatic Breathing; Frequency = headache days for four weeks-over three months; Duration = 4-72 hours per attack; Intensity = 1-10 from a visual analog scale (VAS).  Note. EME = Eye movement exercises; DB = Diaphragmatic breathing   Shows participants eye movement exercises for experimental groups: rst row shows vertical headguided eye movement exercises; second row shows horizontal head-guided eye movement exercises.

Figure 3
Shows participants eye movement exercises for experimental groups: rst row shows vertical headguided eye movement exercises; second row shows horizontal head-guided eye movement exercises.

Figure 4
Participants posture during eye movement exercises or diaphragmatic breathing: a = upright back; b= xed, leveled, and high-up head; c = rolled backed shoulders; d = planted feet on the ground at a 90-degree angle to the extension of shoulders; e = slouched forward breast.

Figure 5
Migraine pain (frequency; ranges 1-9 attacks, duration; ranges 1-8 hours/attack, and intensity; ranges a scale of 1-7/attack during the last four-week) changes in each study (EM = eye movement exercises, DB = diaphragmatic breathing, control) group across three assessment points (pre-test, pot-test, and followup).