Breath holding index in episodic primary headaches

Background/Aim. Examination of cerebrovascular reactivity in patients with primary headaches is focused mainly on migraine, while the smaller number of studies deals with tensiontype and cluster headache, or comparison of cerebral haemodynamic in migraine and tension-type headache (TTH). In this study, we hypothesized that cerebrovascular reactivity differs among different types of episodic primary headaches. In order to prove that we aimed to compare the interictal cerebrovascular reactivity in patients with the episodic form of the three most common types of primary headaches using the breath holding test. Methods. Examination was performed in 243 patients, 100 migraineurs with aura (group I), 70 migraineurs without aura (group II), 38 patients with episodic tension-type headache (group III), 35 patients with episodic form of cluster headache (group IV) and 35 healthy controls (group V). The Doppler instrument was used for transcranial doppler (TCD) sonography and breath-holding test performance. Blood flow mean velocities (MV), pulsatility indices (PI) and breath-holding index (BHI) for middle cerebral artery among these groups were analyzed. Results. The mean velocities and pulsatility indices were not different in 4 groups of headache patients and controls. The BHI was found to be significantly greater in the migraineurs with aura (1.668 ± 0.269) compared with the patients with migraineurs without aura (1.411 ± 0.358, p = 0.005), tension type headache (1.401 ± 0.428, p = 0.035), cluster headache (1.203 ± 0.311, p < 0.01) and controls (1.195 ± 0.269, p < 0.01) showing an exaggerated reactivity to hypercapnia in patients with migraine with aura. Conclusion. In conclusion, our finding support the literature data that increased cerebrovascular reactivity is a feature of migraine with aura. Result of unchanged cerebrovascular reactivity in migraine without aura, cluster headache and tension-type headache is expected, still, it is possible that in future, using different technique, we will be able to put more light on vascular changes that are following different headache disorders.

Examination of cerebrovascular reactivity in patients with primary headaches is focused mainly on migraine, while the smaler number of studies deals with tension-type headache, or comparison between cerebral haemodynamic in migraine and tension-type headache (TTH) [6][7][8][9][10][11] .
3][14][15][16][17][18] .The key point of those differences was whether the vasoconstrictor or vasodilator stimuli was used.The results of several studies performed with vasoconstrictor stimuli indicated an increased cerebrovascular reactivity in patients with migraine.These conclusions had been limited by results showing an increased cerebrovascular reactivity exclusively in patients with migraine with aura.The results of the studies using vasodilator stimuli are contradictory, probably due to differences in methodology and patients selection.
Cerebrovascular reactivity in patients with tension-type headache has been the subject of a significantly smaller number of researches 8,11,14 .The difference in interictal cerebrovascular reactivity in patients with migraine without aura and episodic TTH was not found 9 .That kind of difference did not show neither by comparison of patients with TTH and healty controls 13,14 .
In patients with cluster headache cerebrovascular reactivity had been examined during the cluster period, after the inhalation of 100% oxygen 19 .In comparison to migrainers, patients with cluster headache had increased response.Comparative data about cerebrovascular reactivity in more than 2 types of episodic primary headaches does not exist.
In this study, we hypothesized that cerebrovascular reactivity differs among different types of episodic primary headaches.In order to prove that, we aimed at comparing the interictal cerebrovascular reactivity in the patients with the episodic form of the 3 most common types of primary headaches using the breath holding test.

Methods
Examination was performed at the Headache Center and Ultrasound Laboratory in the Neurology Clinic, Clinical Center of Serbia, Belgrade, over 5 groups of subjects were treated for the period of two years: migraineurs with aura (group I), migraineurs without aura (group II), patients with episodic TTH (group III), patients with episodic form of cluster headache (group IV) and healthy controls who had no history of headache (group V).Exclusion criteria were cardiovascular, cerebrovascular, or pulmonary disease, arterial hypertension, therapy with beta-adrenoceptor blockers or calcium antagonists in the last three months, comorbidity of 2 types of primary headaches.
All patients and healthy control subjects gave their informed consent to participate in the study and the study was approved by the Ethics Committee of Neurology Clinic, Clinical Center of Serbia, Belgrade.
The diagnosis of episodic primary headache was based on the International Classification of Headache Disorders criteria 20 .The Doppler instrument, RIMED Digi-Lite, a dualchannel transcranial Doppler (TCD) system, was used for TCD sonography and breath-holding test performance.Insonation was performed interictaly, throughout the temporal acoustic bone windows according to a standard approach using 2 MHz transducers to display flow through the middle cerebral artery (MCA).Bilateral monitoring of the MCA, from a depth of 45 mm to 65 mm, was performed with each probe held in place over the temporal bone by the head frame.
Cerebrovascular reactivity has been examined by breath-holding test, based on vasodilatator effect of hypercapnia resulted after 30 seconds of breath holding 21,22 .
Breath-holding index (BHI) was calculated for each MCA, using the formula 15

Results
A total of 243 patients were studied, including 100 migraineurs with aura (group I), 70 migraineurs without aura (group II), 38 patients with episodic tension-type headache (group III), 35 patients with episodic form of cluster headache (group IV) and 35 healthy controls (group V).Demographic features of examined groups are presented in Table 1.
The mean velocities and pulsatility indices were not different in 4 groups of headache patients and controls (Tables 2 and 3).
Mean velocities for MCA in examined groups are presented in Table 2.
Pulsatility indices for MCA in examined groups are presented in Table 3. BHI was higher in patients with migraine with aura than in migraine without aura, episodic TTH, cluster headache and healthy controls (Table 4).No difference was found among the other groups regarding BHI.

Discussion
Diferent results of cerebrovascular reactivity in migraine might be caused by the diferences in patient selection and methodology.Beside heterogenous data, the result of higher cerebrovascular reactivity in migraine with aura has remained stable over decades of research 23 .Results of our study show that BHI is higher in patients who have migraine with aura than in patients with other types of primary episodic headaches, migraine without aura, episodic   TTH and cluster headache.These data confirm results of previous studies reporting an increased cerebrovascular reactivity exclusively in patients with migraine with aura in comparison with migraineurs without aura 12,14 .Also, our data do not confirm literature reports of increased vasodilatatory response in migraine without aura 12 .This difference is just one among other, epidemiological and clinical differences between these two entities imposing the question older more than thirty years, whether the migraine with and without aura are two kinds of headache disorder 24 .In our group of patients with migraine with aura, one third of them had migraine without aura as well.Beside that, the cerebrovascular reactivity was significantly higher in this group of patients.
Potential explanation for increased cerebrovascular reactivity in migraine, particulary with aura, lies in hypersensitivity and impaired habituation to stimuli 23,25 .Literature data, in accordance with our results, suggest that the autoregulation disorder leading to inadequate, increased response of intracranial arteries to metabolic stimuli could be the key feature for increased cerebrovascular reactivity 25 .According to neurovascular coupling theory 26,27 cerebral blood flow varies due to local cortical activity.Intraictally impaired cerebrovascular reserve could point to disfunction of vascular elements of neurovascular unit, meaning pericits, muscle and endothelial cells contained in the wall of small vessels 28 .Endothelial dycfunction in migraine is the new question arrised just few years ago, with increasing number of ongoing researches dealing with it 29 .
To our best knowledge, this is the first study on cerebrovascular reactivity covering and comparing, at the same time, the episodic forms of the 3 most common types of primary headaches.
There are studies showing no difference in cerebrovascular reactivity in patients with migraine without aura in comparison to patients with tension-type headache and healthy controls 14 .Our study showed the same result, overcoming the limitation of broad overlap between migraine and TTH data that were presented in a report of Arjona et al. 14 , with strict patient selection.
The results considering interictal cerebrovascular reactivity in patients with cluster headache, being significantly lower in comparison to migraine with aura, and showing no difference to migraine without aura, tension-type headache and healty subjects, are in accordance with conclusions of other authors 19 who found significant difference in vasomotor reactivity between CH patients and controls in response to hypocapnia only during the headache phase, with difference disappearing 30 min after the attack.Recent study showed that the BHI mesured after the oxygen inhalation is significantly higher in the cluster patients compared to the migraine patients which is the conclusion that our study could not support.Opposite to CO2, a powerful vasodilatatory stimulus, oxygen is a powerful vasocontrictor and its inhalation just before the testing of cerebrovascular reactivity by breath-holding test, without doubt affects the test results.Beside vasoconstriction, with direct impact on vessels, 100% oxygen influence the cerebral blood flow indirectly, by inhibition of neurons in the trigeminocervical complex.This "oxygen inhibition" of neuronal activation in the trigeminocervical complex is shown on an animal model for cluster headache, developed in order to reveal therapeutic effect of oxygen in cluster headache 30 .

Conclusion
In conclusion, our finding support the literature data that increased cerebrovascular reactivity is a feature of migraine with aura.Result of unchanged cerebrovascular reactivity in migraine without aura, cluster headache and tension-type headache is expected, still, it is possible that in future, with different technique, we will be able to put more light on vascular changes that follow different headache disorders.

R E F E R E N C E S
velocities (MV), pulsatility indices (PI) and breath-holding index (BHI) for the middle cerebral artery among these groups were analyzed.Statistical analysis was performed using the SPSS software version 17.0.Distrubution of parameters was assessed by Kolmogorov-Smirnov test.For multiple comparisons among the groups, ANOVA and Kruskal-Wallis test were used, with Tukey honest significance difference (HSD) test and Mann-Whitney test applied in post hoc analyses.The significance level was set at 5% (p < 0.05).Podgorac A, et al.Vojnosanit Pregl 2018; 75(4): 347-351.

Table 1 Demographic features of the examined groups
*MA -migraine with aura; MO -migraine without aura; TTH -tension-type headache; CH -cluster headache.