- Top of page
Abstract Sporadic temporal slow waves are considered to be associated with mild cerebrovascular dysfunction. However, electroencephalogram (EEG) changes have not been consistently described by some investigators and correlations inferred on the basis of such data remain inconclusive. In the present study, we examined previously defined temporal slow waves in patients in relation to incidence of cerebrovascular disease. A total of 512 EEG were analyzed during a 1 year period at our laboratory and 74 reference EEG from healthy volunteers were all examined as to the presence of temporal low-voltage irregular delta wave (TLID), temporal minor slow and sharp activity (TMSSA) or bursts of rhythmical temporal theta (BORTT). The patterns were assessed in relation to clinical histories of patients and controls. There were similarities in clinical correlations for the three EEG changes including associations with mild cerebrovascular diseases and with aging. In addition these temporal EEG findings shared electrographic features and showed left side predominance as well as correlating with each other. Since TLID, TMSSA and BORTT have many clinical and electrographical similarities, we consider that these findings should be grouped into one EEG entity which appears in association with mild cerebrovascular dysfunction.
- Top of page
Temporal slow waves on electroencephalogram (EEG) are common findings of the elderly.1–3 Since the incidence of temporal slow waves increases with age, these EEG findings may simply reflect an aging brain. However, several studies have found a correlation between temporal slow waves and cerebrovascular diseases. For example, our previous study showed that temporal delta waves were associated with cerebrovascular diseases at an early stage, with multiple small infarction on MRI4 and with the occurrence of microangiopathy of the retina in diabetic patients.5 Asokan et al.6 and Maynard and Hughes7 also showed that temporal slow waves correlated both with aging and cerebrovascular disorders. Therefore, preliminary findings indicate that EEG abnormalities may be a useful indicator of cerebrovascular dysfunction at an early stage.
To date EEG findings have not been fully and consistently described by several investigators and the conclusions based on these findings remain inconclusive as a result. Among studies of temporal slow waves related to mild cerebrovascular disorders, only three papers defined the EEG pattern clearly. Asokan et al. described an EEG change in patients with a cerebrovascular disorder as temporal minor slow and sharp activity (TMSSA).6 This finding was characterized by a mixture of theta band activities and a sharp wave-like transient. Maynard and Hughes termed their temporal finding as bursts of rhythmical temporal theta (BORTT).7 The BORTT was characterized by burst of 6–7 Hz theta activities at temporal areas. We previously reported that temporal low-voltage irregular delta wave (TLID) was associated with aging and cerebrovascular dysfunction. Although these EEG findings have different morphologic characteristics, they appear to have the same clinical significance. The present study was designed to examine the relationship between these various temporal EEG findings.
- Top of page
A total of 512 EEG of patients and 74 EEG of healthy volunteers were examined in this study. The mean age of the patient population (244 females, 268 males) was 45.3 ± 18.1 years and that of the control group (31 females, 43 males) was 54.5 ± 16.0 years. To compare the incidence of temporal slow waves between the group of patients with cerebrovascular diseases and the age-matched subset of normal control group, all individuals over 50 years of age were selected from the normal control group. The mean age of this subgroup was 65.4 years (n = 49).
First, we wanted to confirm that sporadic temporal slow waves in EEG were common in patients with clinically diagnosed cerebrovascular diseases since previous studies failed to compare the incidence in normal individuals as well as in patients. For this purpose, all the EEG records of patients over 20 years of age at our hospital during a 1-year period were checked for the presence of temporal slow waves. We attempted to determine whether there was a relationship between temporal slow waves and a clinical diagnosis. In the present study, three types of characteristic temporal slow waves were examined. These were bursts of rhythmical temporal theta (BORTT), temporal low-voltage irregular delta wave (TLID) and temporal minor slow and sharp activities (TMSSA). Maynard and Hughes first described a 6– 7 Hz theta train at the temporal area as a distinctive electrographic entity (BORTT) on the basis of the electroclinical correlation of this EEG pattern.7 Temporal low-voltage irregular delta waves were characterized as irregular delta waves ranging from 2 to 2.5 Hz in frequency, lasting 1–2 s and propagating to ear lobes.4 Temporal low-voltage irregular delta waves were superimposed on background activities and had a low amplitude for conventional delta activity, in general below 50 μV. According to the description by Asokan et al., temporal minor slow and sharp activity (TMSSA) consisted of mixed 2–7 Hz and 8–14 Hz activity with intermingled minor sharp transients at the temporal area.6 Examples of these three findings were shown in Figs 1 and 2.
Figure 1. Examples of temporal low-voltage irregular delta wave (TLID; left) and bursts of rhythmical temporal theta (BORTT; right) detected in a 73-year-old female patient with depression. 2 Hz low-voltage irregular delta waves were seen when the patient was awake and 7 Hz rhythmic theta waves when she was drowsy, both located in the left anterio- and mid-temporal areas. Note that these patterns propagated to the left ear lobe.
Download figure to PowerPoint
Figure 2. Example of temporal low-voltage irregular delta wave (TLID; left) and temporal minor slow and sharp activity (TMSSA; right) detected in a 75-year-old female with transient ischemic attack.
Download figure to PowerPoint
Second, we have examined the age distribution of EEG changes, the location of these findings and associated EEG abnormalities. We have compared the various patterns and investigated any interrelationships.
Electroencephalograms were recorded with 16 electrodes in accordance with the International 10/20 system. For clear demonstration of the temporal slow wave, the average potential reference method was used in all records in addition to standard monopolar and bipolar recordings. All the EEG diagnoses were made in a blind fashion by one of the authors (KI) to provide consistency of evaluation.
Data in this study are expressed as means ± standard deviations. Statistical significance was determined by one-way analysis of variance and the χ2 test. Differences were regarded as being statistically significant when P < 0.05.
- Top of page
Among patients, TLID, TMSSA and BORTT were found in 109, 40 and 10 individuals, representing a prevalence of 21%, 7.8% and 2.0%, respectively. The prevalence of these three EEG patterns in the control group was 13.5%, 6.8% and 1.4%, respectively.
Clinical diagnoses and temporal slow waves
As shown in Table 1, temporal slow waves were frequently found in patients with cerebrovascular diseases and this incidence was significantly higher than that in healthy volunteers. Since the mean age of the subset of patients with cerebrovascular diseases (61.4 years) was higher than that of the entire study population (45.3 years) or that of normal controls (54.3 years), the high incidence of TLID in patients with cerebrovascular diseases may simply be due to aging. The difference was, however, still significant (χ2 = 23.8, P < 0.0001) when these patients were compared with an age-matched subset of the normal control group (mean age 65.4 years, n = 49, all individuals over 50 years of age). Among the patients with cerebrovascular diseases, the incidence of TLID was especially high in patients with multiple small infarction (11/12), with vascular dementia (12/12) and with transient ischemic attack (TIA) (9/11). By contrast, TLID was less frequent in patients with cerebral hemorrhage (1/7), subarachnoid hemorrhage (4/6) or advanced cerebral infarction (2/8). Like TLID, both TMSSA and BORTT were found more frequently in patients with cerebrovascular diseases as compared to patients with other disorders. Among the cerebrovascular diseases, TMSSA was not found in patients with cerebral hemorrhage and advanced cerebral infarction, but this EEG change was found in patients with multiple small infarction (8/12) and a few patients with TIA (4/11). Diagnoses for 10 patients with BORTT were TIA (n = 2), small multiple infarction (n = 2), subarachnoid hemorrhage, dizziness of unknown etiology, depression with hypertension, depression, neurosis and spinocerebellar degeneration.
Table 1. Temporal slow waves and clinical diagnoses
|Diagnostic categories||No. (years)||Age No. (%)||TLID* No. (%)||TMSSA** No. (%)||BORTT|
|Cerebrovascular||73||61.4 ± 11.1||49 (67.1%)||20 (27.4%)||5(6.8%)|
|Degenerative||56||56.6 ± 16.4||16 (28.6%)|| 6 (10.7%)||1(1.8%)|
|Brain tumor||40||50.0 ± 14.9||10 (25.0%)|| 1 (2.5%)||0|
|Psychiatric||91||39.5 ± 16.5||14 (15.4%)|| 3 (3.3%)||3(3.3%)|
|Head injury||28||39.7 ± 14.8|| 1 (3.6%)|| 1 (3.6%)||0|
|Epilepsy||153||37.3 ± 16.2|| 8 (5.2%)|| 5 (3.3%)||0|
|Others||71||43.8 ± 18.7||11 (15.5%)|| 4 (5.6%)||1(1.4%)|
|Normal controls||74||54.5 ± 16.0||10 (13.5%)c|| 5 (6.8%)†||1(1.4%)|
Age distribution and sex of patients with temporal slow waves
As shown in Table 2, the incidence of TLID, TMSSA and BORTT all increased in patients with increasing age. There was evidence that TMSSA and BORTT were different in that their incidences peaked at 60– 69 years of age, while TLID was most prevalent in the 70 years and over age group. In the control group temporal findings were commonly detected in the aged and were not seen in subjects under 40 years of age. Table 3 shows that temporal slow waves appeared frequently in any ages in patients with cerebrovascular diseases as compared with normal controls. In both patients and the control group there was no significant gender difference between subjects with and without temporal slow waves of any type (P < 0.2–P < 0.99).
Table 2. Temporal slow waves and aging
|20–29||149|| 4 (2.7%)||2 (1.3%)||0|
|30–39||69|| 4 (5.8%)||1 (1.4%)|| 1 (1.4%)|
|40–49||80|| 12 (15.0%)||3 (3.8%)|| 1 (1.3%)|
|50–59||70|| 19 (27.1%)|| 8 (11.4%)|| 1 (1.4%)|
|60–69||85|| 39 (45.9%)||15 (27.8%)|| 5 (5.9%)|
|70+||59|| 31 (52.5%)||12 (20.3%)|| 2 (3.4%)|
|Total||512||109 (21.3%)||40 (7.8%)||10 (2.0%)|
Table 3. Temporal slow waves and aging in controls and patients with cerebrovascular diseases
| ||Temporal slow waves of any type/No.|
|Age (years)||Controls||Cerebrovascular diseases|
|30–39||0/8|| 2/4 (50%)|
|40–49|| 1/12 (8.3%)|| 3/9 (33.3%)|
|50–59|| 2/11 (18.2%)||10/17 (58.8%)|
|60–69|| 5/20 (25%)||20/25 (80%)|
|70+|| 4/15 (26.7%)||15/18 (83.3%)|
|Total||12/74 (16.2%)||50/73 (68.5%)|
Location of temporal slow waves
In the group of 512 patients, TLID appeared predominantly on the left temporal area in 60% and on the right side in 17%. In the remaining 23% of patients, TLID appeared on both sides equally. A predominance on the left side was also found in the occurrence of BORTT (70%vs 15%) and TMSSA (75%vs 0%). In 16 EEG records of healthy subjects with temporal slow waves of any type, left side predominance was again seen (56%vs 13%). In all records with temporal slow waves, the foci of the slow wave were at anterio-temporal or mid-temporal areas.
Activation of temporal slow waves during light sleep
Among 83 EEGs with TLID that included sleep stages, TLID was activated in the light sleep stage (> twofold) for 24 patients (29%). Temporal minor slow and sharp activity and BORTT were activated in the light sleep stage in 23% and 43% of patients, respectively.
Other EEG abnormalities
Out of 109 patients with TLID, 70 (64%) patients showed no other significant EEG findings. Of the remaining patients, poor alpha attenuation during eye opening was seen in 25 (23%). Mild slowing tendency (7–8 Hz) of the background activity was seen in 24 patients. Focal slow wave abnormalities were detected in 25 patients and 10 of these showed frontal intermittent rhythmic delta activity. Definite epileptiform abnormalities appeared in only two (2%) patients. Thus, a mild slowing of the background activity together with poor alpha attenuation was a very common finding in patients with TLID (20/109). Background activity of EEG with TLID (8.9 ± 1.3 Hz) was slightly but significantly slower than that of EEG without TLID (9.5 ± 1.4; P < 0.001).
Electroencephalogram abnormalities seen in patients with BORTT and TMSSA were similar to those for patients with TLID. No other findings were seen in 6/10 (60%) EEG with BORTT and 28/40 (70%) EEG with TMSSA. Mild general slowing was seen in 30% of patients with BORTT and 15% of the patients with TMSSA while poor alpha attenuation was observed less frequently (20% and 10% in each patient group, respectively). Background activity was, however, not significantly different between EEG with and without BORTT or TMSSA. Clear spikes appeared in the EEG of only one patient with TMSSA.
Association among temporal low-voltage irregular delta wave, bursts of rhythmical temporal theta and temporal minor slow and sharp activity
The results in this study suggested that TLID, BORTT and TMSSA data bore considerable resemblance, so next we examined whether there was coincidence of these temporal findings. We found that TLID appeared in 32/40 (80%) EEG with TMSSA and 8/10 (80%) EEG with BORTT. The correlation between TLID on the one hand and BORTT and TMSSA on the other was statistically significant (P < 0.0001).
- Top of page
The present study confirmed that TLID, BORTT and TMSSA correlated with both aging and cerebrovascular diseases. In addition, we found an association in the occurrence of these EEG changes. Since the latter findings have many clinical and electrographical similarities, we consider that these abnormal changes should be grouped to form a single EEG entity. The common features of TLID, BORTT and TMSSA in the present study were as follows: they appeared in the anterio- and mid-temporal areas; they were not continuous but sporadic slow waves lasting 1–2 s and propagating to ear lobes; they appeared predominantly on the left side; the occurrence was activated during drowsiness; they were commonly detected in old ages; and they appeared in association with mild cerebrovascular diseases.
The present results confirmed previous findings that the incidences of TLID,4 TMSSA6 and BORTT7 were significantly greater in patients with cerebrovascular diseases than in patients with other diseases. Among patients with cerebrovascular diseases, temporal slow waves were commonly detected in patients with mild ischemic cerebrovascular disorders and less frequently in patients with hemorrhage or advanced cerebrovascular infarction. Therefore, we conclude that these temporal findings are not simply a normal finding of aging and reflect a certain functional change induced by minor cerebrovascular dysfunction. The fact that more than 60% of EEG with temporal slow waves had no other significant change supports the observation that these EEG represent mild cerebral dysfunction, but not severe dysfunction. Our previous finding that the occurrence of TLID was associated with the presence of silent small multiple infarctions on MRI supports this view.4 The pattern sometimes seen in EEG with temporal slow waves including mild general slowing of the background and poor alpha attenuation during eye opening is also congruent with this conclusion. We speculate that one of the reasons why these temporal slow waves were not frequently found in patients with advanced cerebrovascular diseases could be that these patients developed more severe abnormalities including continuous or higher voltage slow wave activities. In accordance with this, Schaul have shown that continuous slow wave activity suggested severe brain damage, while intermittent slow activity usually indicated a small lesion.8
In the present study we identified electrographical findings that were similar to those previously reported to have been associated with mild cerebrovascular dysfunction or aging, including anterio-temporal foci,1–3,6,7 left side predominance6,7 and activation during drowsiness.6,7 A characteristic feature is that the temporal slow waves propagate to the ear lobes.
Although the reason for temporal slow waves appearing predominantly on the left side remains unclear, one possible explanation is hemispheral dominance. That is, greater stress of the dominant hemisphere could result in more frequent occurrence of vascular pathology. Left-side preference of internal carotid thrombosis could also be associated with the lateralization of the temporal slow waves.9
On the basis of clinical and electrographical similarities, we feel that sporadic temporal slow waves associated with mild cerebrovascular dysfunction can be grouped to form an EEG entity. However, the question still arises as to why these waves have different morphologic characters; TLID is a low-voltage delta wave superimposed on background activities; TMSSA is a mixture of theta and alpha range activities with sharp wave-like activity; and BORTT is a train of theta waves. Although we do not have a clear explanation for these discrepancies, it seems significant that these EEG changes are a mixture of slow waves from the subcortical region and cortical background activities. Of relevance is the series of experiments published by Gloor et al., who demonstrated that localized lesions in subcortical white matter of cats caused irregular delta activity in the cortex overlying the lesion, while lesions in pure cortical grey matter lesions did not produce such activity.10 These findings are in agreement with our clinical observation that TLID appeared in association with small ischemic lesions in the subcortical white matter.4 Background cortical activities overlying the lesion may be influenced by the slow wave and, according to the degree or location of the lesion, they may be alpha or theta range activities. If the resultant change is monorhythmic theta waves, it may appear as BORTT-like activity. If the change includes waves of various ranges and voltage, it may manifest as TMSSA-like activity. When delta activities from the subcortical region are recorded as delta waves at the surface of the scalp, they may result in TLID. Therefore, we hypothesize that these EEG findings arise from the same pathology but their morphologic characteristics vary according to the degree or location of the lesion. In fact we have found that when the voltage of TLID is gradually lowered in a serial observation, TLID with very low voltage often resembled TMSSA or BORTT (Motomura and Inui, unpubl. obs, 1998).
One of the reasons why examination of temporal slow waves has not received enough attention in spite of the clinical usefulness of this methodology seems to be a technical problem. Because these EEG findings propagate to the ear lobes and are often low in voltage, it is difficult to detect such changes with the standard monopolar recording techniques. For a clear demonstration of these EEG findings, the electrodes have to connect the common average reference. Recording of the voltage difference between the bilateral ear lobes (Fig. 1) is also helpful. When a patient is suspected of having mild cerebrovascular dysfunction, these techniques are recommended.
In conclusion, sporadic temporal slow waves are a minor EEG abnormality consistent with mild brain dysfunction of various etiology, but often indicating cerebrovascular disease in its early stage. The pattern related to mild cerebrovascular disorders consists of activities of various ranges, from alpha to delta, lasting 1–2 s over the anterio-and mid-temporal areas and predominantly on the left side. Such an EEG pattern is a much better diagnostic indicator for early cerebrovascular disease, than is widely accepted. One reason for such findings sometimes being ignored seems to be technical problems associated with measuring the EEG changes.