Influence of κ-rhythm when assessing drowsiness

INTRODUCTION

An earlobe reference electrode is sometimes activated by spontaneous electrical potential occurring in the temporal lobe. The most frequent cause of activation is κ-rhythm, which appears at the highest amplitude on a bipolar T3–T4 derivation, and its frequency range is 6–12 Hz (predominantly 7–10 Hz), which is similar to α-waves. It does not disappear during sleep or upon waking,^1,2 and its polarity is opposite between the right and left hemispheres. The κ-rhythm causes serious problems when assessing drowsiness because it obscures the disappearance of α-waves, ripple waves, etc., which are the characteristic features of drowsiness. The present study examined the influence of κ-rhythm when assessing drowsiness, using both earlobes as references.

METHOD

A total of 129 inpatients and outpatients (mean ± SD 57.7 ± 10.1 years; range 38–75 years; 80 men and 49 women) who showed κ-rhythm on a clinical electroencephalogram (EEG) examination and then fell asleep were included in the study. The EEG was recorded by a digital EEG recorder (2514; NEC Medical Systems, Tokyo, Japan), and the data displayed on a PC monitor with a SynaViewer ES5003 (NEC Medical Systems). Sleep stages were scored manually with a unipolar C3–A2 or C4–A1 derivation, and with a C3–A1 or C4–A2, separately. Bipolar T3–T4 and A1–A2 derivations were also displayed simultaneously on a PC monitor for checking the presence of κ-rhythm. When the earlobe reference was activated, source derivation (SD) methods, in which the potentials of four active electrodes around an active electrode were averaged and used as a reference, were used. An O1–SD or O2–SD derivation was displayed for checking α-waves. For those subjects who showed a strong κ-rhythm, a C3–SD or C4–SD derivation was used for sleep scoring.

RESULTS

Figure 1 shows the polysomnogram (PSG) of a representative subject. A prominent κ-rhythm appeared on T3–T4 and A1–A2 derivations. On a C4–A1 derivation, α-like waves appeared continuously, but on C4–A2 and C4–SD1 derivations α-like waves sometimes disappeared, indicating drowsiness. We judged drowsiness using both earlobes as references in 129 subjects who showed a κ-rhythm on the T3–T4 derivation. When the contra-lateral earlobe was used as a reference, drowsiness could be judged properly in 26.4% of subjects. Conversely, drowsiness could be judged properly in 90.0% of subjects using the ipsi-lateral earlobe. It was impossible to assess drowsiness in 10.1% of subjects using either earlobe as a reference. Using the ipsi-lateral earlobe was significantly better than the contra-lateral earlobe (P < 0.0001, χ² test) for assessment of drowsiness. When we used the SD method, drowsiness could be judged in all subjects (Fig. 1), but at least 19 active electrodes were required.

DISCUSSION

In the present experiment, assessment of drowsiness was improved significantly by using an ipsi-lateral earlobe. When the amount of κ-rhythm increases, it is necessary to counterbalance the rhythm, which activates not only an earlobe but also active electrodes, by using the ipsi-lateral earlobe. We have reported previously that was observed in 21.6% of 440 subjects (mean age 48.4 years).⁴ Furthermore, the presence rates of κ-rhythm were 9.8% for subjects aged in their 20s, 6.6% in their 30s, 15.1% in their 40s, 24.2% in in their 50s, 39.5% in their 60s, and 43.8% in their 70s. Kennedy et al. have reported that κ-rhythm was observed in 30% of healthy adults.³

The present study’s results suggest that the bipolar T3–T4 derivation should be monitored and checked for the presence of κ-rhythm, and that the ipsi-lateral earlobe or the SD method should be used as a reference, especially in the elderly. Using these methods, amplitude was reduced by approximately 5–15% and has very little influence on judging drowsiness.