Abstract
While facial autonomic signs are prominent during cluster headache (CH) attacks, cardiovascular autonomic changes have been described in few CH patients. Cardiovascular autonomic function tests (AFT) can be used to assess general autonomic function in CH patients in different stages of the disease. We aimed to assess whether general autonomic function is changed in CH patients during a cluster period. AFT was performed both during a cluster period, but outside an actual attack, and outside a cluster period in 18 patients. Heart rate variability was studied at rest, during deep breathing, after standing up and during a Valsalva manoeuvre. Blood pressure (BP) changes were recorded at rest, during standing up and during sustained handgrip. Measurements during and outside the cluster period were compared using the paired t-test. AFT measurements revealed no significant differences between the two measurements, except for diastolic BP in rest, which was higher during the cluster period [80.3 (SD 12.2) vs. 74.8 (SD 9.0), P = 0.04]. Autonomic dysfunction during a cluster period, but outside an attack, does not include systemic cardiovascular control.
Introduction
Cluster headache (CH) is characterized by attacks of very intense pain in the (peri)orbital region, which usually lasts for 15–180 min and may occur up to eight times per day. Attacks typically occur in so-called ‘cluster periods’ lasting weeks to months, alternated by symptom-free periods of months to years (episodic CH) (1). Involvement of the autonomic nervous system during attacks is evident through the distinct local ipsilateral autonomic symptoms, including conjunctival injection, tearing, eyelid oedema, rhinorrhoea and Horner's syndrome which indicate transient parasympathetic hyperactivity and sympathetic hypofunction. ECG disturbances have been reported during CH attacks (2, 3), suggesting that autonomic dysfunction in CH is not limited to the periorbital area. The source of autonomic impairment may be found in the hypothalamus, as this structure is not only the main regulator of autonomic function but also a key site in CH pathophysiology (4).
We investigated cardiovascular autonomic function in CH patients, focusing on differences between cluster periods and remission.
Methods
The study was approved by the local ethics committee. After obtaining informed consent, cardiovascular autonomic functions (5) were tested during and outside a cluster period in 18 patients [13 males, mean age 47.2 (SD 10.0) years] with episodic CH. The mean duration of a cluster period was 9.1 weeks (SD 5.6). Median time between measurements was 10 weeks. Patients who were on prophylactic medication for CH were excluded from the study. All measurements were performed in the early afternoon. Heart rate (HR) was measured from ECG and blood pressure (BP) was measured with a sphygmomanometer. Heart rate was assessed during a rest period, during forced breathing at 6–8/min, following active standing up, during a Valsalva manoeuvre and during sustained force (handgrip). For the Valsalva manoeuvre, the subject was instructed to keep up a pressure of 40 mmHg by blowing into a manometer mouthpiece for 15 s. For the handgrip test the subject had to squeeze a dynamometer at 30% of maximum voluntary force. Minimum–maximum ratios and differences were calculated using 120 successive beats. (BP was measured during a resting period, 1 and 2 min after standing up and during 5 min of sustained handgrip or until the subject gave up. The minimum and maximum systolic and diastolic pressures after standing up were noted and both ratios and differences between starting and minimum values were calculated. For sustained handgrip, maximal systolic and diastolic BP and BP during rest were used for ratios and differences.
Data were compared using the paired t-test, with P < 0.05 indicating significant differences.
Results
No significant differences between measurements during and outside attack periods were found, except for resting diastolic BP before the sustained handgrip test, which was higher during the cluster period (P = 0.04) (Table 1). A comparison with HR and BP data of a study in 75 healthy volunteers (6) revealed no important differences.
Cardiovascular autonomic tests during and outside a cluster period
Data are expressed as means (SD).
P = 0.04.
Discussion
The present study did not reveal systematic cardiovascular autonomic functional changes in CH patients during a cluster period, outside the actual attack. The one exception was that we found a higher diastolic BP during the cluster period during rest. This probably represents stress of anxiety concerning a subsequent attack during the cluster period (7).
Cardiovascular autonomic disturbances have previously been described in CH in association with attacks, including bradycardia in one CH case (6), and variable heart frequency in a series of CH patients (2, 3). These were suggested to be due to central autonomic dysfunction. Local autonomic function studies in CH, in particular pupillometric studies, suggested lesions of the sympathetic tract (8, 9), although no conclusions were drawn about a possible central or peripheral site of such a lesion.
In our study, no differences were found between patients during or outside a bout in a study of respiratory sinus arrhythmia in CH, although lower HR ratios were reported during a deep breathing test in all CH patients compared with controls (10, 11). However, in those studies individual patients were not all measured both during and outside a bout and therefore a change of autonomic function within patients could not be assessed.
We conclude that autonomic dysfunction in CH is for practical purposes a local affair, and does not affect systemic cardiovascular autonomic control.
Footnotes
Acknowledgements
J.A.v.V. is supported by the Asclepiade foundation.