The International Tinnitus Journal

Quantitative Electroencephalography Power Analysis in Subjective Idiopathic Tinnitus Patients: A Clinical Paradigm Shift in the Understanding of Tinnitus, An Electrophysiological Correlate

We report a descriptive analysis-interpretation of quantitative electroencephalography (QEEG) data for the metric of power in patients with tinnitus of the severe disabling type (N  61). The results are based on a statistical analysis of the data compared to a normative database as calculated in Z scores, controlling for the factors of age, gender, IQ, and the like. Method: We analyzed the QEEG data for the metric of power to measure (1) the number of significant recordings, normal and abnormal; (2) the significant recordings by electrode recording sites; (3) distribution of the electroencephalographic (EEG) frequency bands; and (4) occurrence of the EEG frequency bands correlated with the electrode recording sites. In the analysis of the occurrence of the EEG frequency bands by electrode recording area, we corrected for the number of recording sites. Results: We recorded normal power recordings in 20 of 61 patients (32.8%) and abnormal power recordings in 41 of 61 patients (67.2%); power distribution by frequency band in 41 of 61 patients, revealing the number of significant recordings of delta (119), alpha (69), beta (91), and theta (17); and the power distribution by location for all frequencies, which were revealed as recording site activity in the frontal greater than in the temporal sites, which in turn was greater than in the parietal site, and equal activity in parietal, occipital, and central sites. The analysis of the occurrence of the EEG frequency bands by electrode recording area as corrected for the number of recording sites reinforced our initial results. Conclusions: Z-score analyses of QEEG recordings—based on a large normative database—for the metric of power for patients having tinnitus of the severe disabling type (N  61) revealed statistically significant abnormalities in frontal greater than temporal electrode recording sites. We reported no difference between male and female tinnitus patients in the number of abnormal power QEEG recordings. However, we observed significant differences in the average Z scores between males and females in the alpha and theta bands. The results suggest multiple central electrophysiological correlates for different clinical types of tinnitus identifiable with QEEG, for the metric of power, by frequencies of brain activity of delta greater than beta greater than alpha greater than theta bands of activity, reflecting physiologically the individuality of brain function and clinically the heterogeneity of the symptom of tinnitus for tinnitus patients. Clinical interpretation of the QEEG data in terms of brain function Reprint requests: Abraham Shulman, MD, FACS, Department of Clinical Otolaryngology, Health Science Center at Brooklyn, State University of New York, Downstate Medical Center, 450 Clarkson Avenue, Box 1239, Brooklyn, NY 11203-2098. Phone: 718-773-8888; Fax: 718-465-3669; E-mail: metrc@inch.com International Tinnitus Journal, Vol. 12, No. 2, 2006 Shulman et al. 122 tients (N  61) having tinnitus of the severe disabling type (subjective idiopathic tinnitus [SIT]). In a preliminary report, we cited the QEEG examination as a significant addition to the medical-audiological tinnitus patient protocol (MATPP) for patients with SIT [1,2]. Preliminary data from 21 SIT patients were found to support the hypothesis of the significant role of the temporal and temporofrontal regions of the brain. QEEG is a quantitative analysis of brain function as reflected in electrophysiological measures via the electroencephalogram [3]. The metrics of analysis include absolute power, relative power in terms of percentage of power, Z scores of asymmetry power, coherence (reported as elevated or reduced), and phase (reported as phase delay, positive or negative, and as both intra- and interhemispheric homologous pairs). The power metric is the total electrical activity recorded at the brain cortex with electroencephalography (EEG). The relative power measures the distribution of electrical activity over the delta, theta, alpha, and beta frequency bands. All these metrics can also be expressed in terms of the deviation of standardized scores based on a selected set of norms (i.e., Z scores). Z scores are observed scores transformed into a score with a common reference (i.e., Z mean  0; standard deviation, Z   1) [4,5]. A significant increase in the average total power (in microvolts) recorded from 19 electrode sites in female patients and a decrease in male patients as compared to a group of subjects without any substantiated medical or neurological disease has been reported in tinnitus patients [6]. Abnormal spontaneous brain activity as measured by magnetoencephalography has demonstrated a pattern of marked reduction in alpha and an increase in delta, pronounced for temporal regions [7]. In this study, we analyze our attempt to identify and quantify Z scores of significant corticoelectrical dysfunction for the metric of power in patients with predominantly central-type severe, disabling SIT (N  61). We use the following measures of analysis: (1) identification and distribution of the number of significant QEEG recordings (normal and abnormal); (2) identification and distribution of specific electrode recording sites; (3) identification and distribution by band frequency; and (4) correlation of the distribution by band frequency and recording site. Analysis of the occurrence of the EEG frequency bands by electrode recording area corrected for the number of recording sites. METHOD We selected for analysis of the metric of power for the measures described patients who had SIT lasting 1 year or longer (N  61) and in whom MATPP (including QEEG) established the diagnosis of a predominantly central-type tinnitus. We used the database that includes normative data for the metrics of relative power, amplitude asymmetry, phase, and coherence [6]. The ages represented in the database range from 2 months to 83 years, which represents 625 subjects. Delta was 0.5–3.5 Hz; theta was 3.7–7.5 Hz; alpha was 7.5–13.5 Hz; and beta was 13.5–22 Hz [4]. We performed the QEEG test with patients’ eyes open and closed in a quiet, darkened room, with patients sitting relaxed in a chair in an upright position. This study included both male and female patients (mean age, 53.9 and 59.6 years, respectively; Table 1) with constant tinnitus. For the QEEG examination, we used the Neurosearch 24 (Lexicor Company, Boulder, CO) QEEG equipment [8]. We placed 19 electrodes on the patients’ scalp using the international 10/20 montage (a montage being a standardized array of electrode sites used to ensure consistent results). The impedance measured at each electrode site with respect to the reference was less than 5,000 Ohms. We set the filter bandpass between 0.5 Hz and 32 Hz. We recorded 300 epochs, selected 25 as being representative and artifact-free, and processed and compared them with the normative database. The gain was 32,000, with a sampling rate of 128,000. We submitted raw EEG results to the Lexicor Company for analysis, and that firm generated a report called the Datalex report for clinical application [9]. The Datalex report is presented in two sections.- W e report a descriptive analysis and interpretation of quantitative electroencephalography (QEEG) data for the metric of power in pain a tinnitus patient, with a focus on theories of a neuroanatomical homeostatic system that regulates baseline levels of local synchrony in multiple neuronal assemblies and on theories of consciousness, introduces a paradigm switch in our clinical understanding of the symptom of tinnitus and an application for tinnitus diagnosis and treatment