Carotid sinus massage (CSM) is commonly performed as a bedside test for determining the type and sometimes also the mechanism of different rhythm disturbances, or for routine investigation of older patients who experience syncope, dizziness, or unexplained falls.1 Recently, it has been reported that the response to CSM is affected by the patient's position, and in patients with unexplained syncope or drop attacks, upright CSM is recommended if initial supine CSM is not diagnostic.2 ,3 On the other hand, time of day at which massage is performed is usually not taken into consideration.

Circadian rhythm, which is the variability of physiology and biochemistry of humans in a predictable fashion during a 24 hour period, may also have an effect on response to CSM as well as the patient's position. There has been no study, to our knowledge, that has prospectively evaluated the relation between circadian rhythm and response to CSM. For this reason, we conducted such a study.

A total of 120 consecutive patients (mean (SD) age 56.9 (12.4) years, 76 men and 44 women) who were in sinus rhythm were included in the study. Patients with a history or clinical findings consistent with cerebrovascular disease and who had murmurs on carotid arteries were excluded. Also, patients who suffered from serious arrhythmia or ischaemia and had symptoms suggestive of carotid sinus syncope were not included. Clinical characteristics of the patients were: hypertension in 72 (60%) patients; diabetes mellitus in 12 (10%); valvar heart disease in six (5%); ischaemic heart disease in 46 (38.3%); and compensated heart failure in 11 (9.1%) patients. Two thirds of the study population were not taking any drugs at the time of hospitalisation. The remainder of the study population were on one or more of the following medications; nifedipine, indapamide, nitrate, aspirin, or β blocker. All medications were discontinued for at least 48 hours before the CSM.

CSM was performed at 0600, 1200 (noon), 1800, and 2400 (midnight) on each patient on both right and left carotid bulb for five seconds when the patients were in the supine position in a quite room while they were at rest. Massage was done three times on each side with 30 second intervals, and continuous rhythm ECG was taken during the massages. CSM was stopped prematurely only if asystole longer than three seconds resulted. The ECG that showed the longest RR interval during the massages for each side was studied.

In the ECGs basal RR interval (RR_basal), the absolute changes in RR interval (RR_max−RR_basal), and the maximal change in RR interval (%) were measured. The following formula was used for maximal change in RR interval (%)4:

The Friedman test was used for the comparison of the maximal change in RR interval at four different times within a 24 hour period. The absolute change (ms) and the maximal change (%) in RR interval between two different times was compared with the multiple comparison test. A probability value of p < 0.05 was considered significant.

All of the patients underwent CSM. No neurological complications or exaggerated response suggesting hypersensitive carotid sinus syncope were seen during and after CSM. Standard deviation and mean of the basal RR interval, the absolute changes, and the maximal changes in RR interval were measured for each time (table 1). The basal RR value (ms) among the four different time points did not differ, although there was an tendency for an increase over the 24 hour period from 0600 to 2400 (mean (SD) 820.4 (151.0) ms, 829.3 (146.3) ms, 834.8 (150.1) ms, 857.6 (214.4) ms, respectively for the four time points). There were significant differences among the four different times with respect to the maximal changes and absolute changes in RR interval (p < 0.001 for both). The significant differences were found between 0600 and 1800, and between 0600 and 2400 (p < 0.05 and p < 0.001 for both absolute changes and maximal changes in RR interval, respectively) in multiple comparison test. This means that the absolute changes and the maximal changes in RR interval were at their minimum at 0600 and at their maximum at 2400. There were no significant differences between 0600 and 1200, 1200 and 1800, and 1200 and 2400 (p > 0.05).

Despite the ubiquitous influence of diurnal cycles on the cardiovascular system, we know relatively little of the clinical significance of the circadian rhythm. Through a number of clinical trials and epidemiological studies, it has become evident that the levels of disease activity of a number of clinical disorders have a pattern associated with the body's inherent clock set according to the circadian rhythm.

On the other hand, circadian rhythm may also have an important effect on response to CSM as well as the activity of a number of disorders. Therefore, time of day at which massage is performed should be taken into consideration while assessing the response.

A normal response to CSM is a transient decrease of the sinus rate and slowing of atrioventricular nodal conduction. However, the present study showed that this response could not be the same if it was performed at different times within a day. In this small prospective series of patients, the response to CSM assessed by the maximal change in RR intervals (%) was found to be at a minimum at 0600 and at a maximum at 2400. The reasons for this differing response are not yet understood. One explanation may be that the efferent limb of the reflex arcus, which reaches the sympathetic and parasympathetic nervous system of the heart and the peripheral vasculature, is affected by the sympathetic activity of the body. The plasma concentrations of adrenaline and noradrenaline (epinephrine and norepinephrine) in man, which reflect the sympathetic tone, display significant daily variations which are greatest in the morning hours and least at night.5 Therefore, decreased activity of the sympathetic tone at night may be responsible for the enhancement of the baroreflex gain and may heighten the response to CSM.

In summary, our findings support the proposal that the diagnostic and therapeutic value of CSM in patients with unexplained syncope or different rhythm disturbances may vary according to the time interval when that massage is performed.