Provocative and inhibitory effects of a video-EEG neuropsychologic protocol in juvenile myoclonic epilepsy


  • Mirian Salvadori Bittar Guaranha,

    1. Department of Neurology and Neurosurgery, Universidade Federal de São Paulo, Escola Paulista de Medicina, São Paulo, Brazil
    Search for more papers by this author
  • Patrícia Da Silva Sousa,

    1. Department of Neurology and Neurosurgery, Universidade Federal de São Paulo, Escola Paulista de Medicina, São Paulo, Brazil
    Search for more papers by this author
  • Gerardo Maria De Araújo-Filho,

    1. Department of Neurology and Neurosurgery, Universidade Federal de São Paulo, Escola Paulista de Medicina, São Paulo, Brazil
    Search for more papers by this author
  • Katia Lin,

    1. Department of Neurology and Neurosurgery, Universidade Federal de São Paulo, Escola Paulista de Medicina, São Paulo, Brazil
    Search for more papers by this author
  • Laura Maria Figueiredo Ferreira Guilhoto,

    1. Department of Neurology and Neurosurgery, Universidade Federal de São Paulo, Escola Paulista de Medicina, São Paulo, Brazil
    Search for more papers by this author
  • Luís Otávio Sales Ferreira Caboclo,

    1. Department of Neurology and Neurosurgery, Universidade Federal de São Paulo, Escola Paulista de Medicina, São Paulo, Brazil
    Search for more papers by this author
  • Elza Márcia Targas Yacubian

    1. Department of Neurology and Neurosurgery, Universidade Federal de São Paulo, Escola Paulista de Medicina, São Paulo, Brazil
    Search for more papers by this author

Address correspondence to Mirian Salvadori Bittar Guaranha, Rua Botucatu, 740 – Vila Clementino, São Paulo, SP, CEP 04023 900 Brazil. E-mail:


Purpose: Studies suggest that higher cognitive functions could precipitate seizures in juvenile myoclonic epilepsy (JME). The present study aimed to analyze the effects of higher mental activity on epileptiform discharges and seizures in patients with JME and compare them to those of habitual methods of activation.

Methods: Seventy-six patients with JME (41 female) underwent a video-EEG (electroencephalography) neuropsychologic protocol (VNPP) and habitual methods of activation for 4–6 h.

Results: Twenty-nine of the 76 (38.2%) presented provocative effect, and inhibition was seen in 28 of 31 (90.3%). A mixed effect was observed in 11 (35.5%), and 30 patients (39.5%) suffered no effect of VNPP. Action-programming tasks were more effective than thinking in provoking epileptiform discharges (23.7% and 11.0% of patients, respectively, p = 0.03). Inhibitory effect was observed equally in the various categories of tasks, except in mental calculation, which had a higher inhibitory rate. Habitual methods of activation were more effective than VNPP in provoking discharges. Anxiety disorders were diagnosed in 24 of 58 patients (41.4%); anxious patients had greater discharge indexes and no significant inhibitory effect on VNPP.

Discussion: Praxis exerted the most remarkable provocative effect, in accordance with the motor circuitry hyperexcitability hypothesis in JME. Inhibitory effect, which had no such task specificity, might be mediated by a widespread cortical–thalamic pathway, possibly involving the parietal cortex. The frequent inhibitory effect found under cortical activation conditions, influenced by the presence of anxiety, supports nonpharmacologic therapeutic interventions in JME.

Juvenile myoclonic epilepsy (JME) is the most common age-related idiopathic generalized epilepsy (IGE), corresponding to 5–11% of all epilepsies (Panayiotopoulos et al., 1991). It is characterized by myoclonic jerks, generalized tonic–clonic seizures (GTCS), and absences (Commission on Classification and Terminology of the International League Against Epilepsy, 1989). Bilateral spike or polyspike-wave complexes of 4–6 Hz, more often asymmetric, on a normal background, are the typical electroencephalography (EEG) findings (Panayiotopoulos et al., 1991).

Patients with JME frequently recognize that sleep deprivation, fatigue, alcohol intake, stress, and flashing lights may act as precipitant factors for their seizures (Clement & Wallace, 1988; Panayiotopoulos et al., 1991; Oguni et al., 1994; Pedersen & Petersen, 1998; Waltz, 2000; Da Silva Sousa et al., 2005a). Beyond these general factors, some patients report other precipitant factors: being sensitive to situations in which they are obliged to consider complicated spatial tasks in a sequential fashion, specifically with the intention of decision making; and responding practically by using a part of their bodies under stressful circumstances. These were conceptualized as praxis induction (Inoue et al., 1994) and include ideation and execution of elaborated movements involving a sequential spatial process such as arithmetic, playing cards and sequential games, drawing, writing, and finger manipulation in more elaborated tasks or those of constructive character. In addition, reflex seizures have been identified during reading and speaking as perioral myoclonia, especially in patients with JME (Mayer et al., 2006).

Neuropsychological methods of EEG activation have been used by groups from Japan (Matsuoka et al., 2000), Germany (Mayer et al., 2006), Italy (Chifari et al., 2004), and Greece (Karachristianou et al., 2004) as an auxiliary method to identify specific seizure patterns in various epileptic syndromes. Some studies suggest that JME would be the most sensitive epileptic syndrome to this form of cognitive activation (Matsuoka et al., 1988, 2000; Senanayake, 1992; Chifari et al., 2004; Karachristianou et al., 2004; Da Silva Sousa et al., 2005b; Mayer et al., 2006).

The aim of this study was to assess the effect of a video-EEG neuropsychological protocol (VNPP), whether precipitant or inhibitory, on epileptiform discharges and seizures in patients with JME, and to compare the effect of this protocol with those of habitual methods of activation.


Seventy-six patients (41 female) with JME underwent VNPP at the Epilepsy Unit of the Hospital São Paulo, Universidade Federal de São Paulo, São Paulo, Brazil. After ethical committee approval, advantages and risks for participation were explained and written informed consent was obtained. Inclusion criteria were age over 12 years, clinical and electroencephalographic features of JME, and a minimum of 4 years of formal education. Clinical signs of antiepileptic drug (AED) intoxication, occurrence of a GTCS, and use of intravenous AED within the last 72 h were exclusion criteria.

Fifty-eight of the 76 patients were submitted to psychiatric evaluation. To assess patients older than 18 years, Schedule Clinical Interview for DSM-IV, Axes I and II (SCID-I and SCID-II) were performed, and the Brazilian version of the Schedule for Affective Disorders and Schizophrenia for School-Aged Children (K-SADS-PL) was used to assess those 18 years old or younger. Because of its high prevalence and the potential role of stress as a trigger of seizures in JME, only anxiety disorders were investigated, and patients were divided in to anxious and nonanxious groups. This division was based upon current anxiety disorders, as opposed to lifetime history of anxiety disorders. All patients presented normal physical and neurologic examinations, as well as normal routine blood tests. Forty patients had a 1.5T magnetic resonance imaging (MRI) of the brain, and all had normal results.

Video-EEG was recorded on a 32-channel digital equipment (Ceegraph software, Bio-Logic Systems Corp., Mundelein, IL, U.S.A.) using the 10–20 International Electrode System, in addition to perioral and deltoid electrodes.

Our protocol was based on those reported by Matsuoka et al. (2000) and Mayer and Wolf (2004) and will be referred to as VNPP. After having slept for at least 6 h, all patients were submitted to 30-min awake video-EEG recording starting at 7 a.m. Medications were maintained in all treated patients. Sixty-seven were treated with AED at the time of the examination and nine were not. Among the treated patients, therapeutic scheme was considered appropriate in 49 (73.1%) and inappropriate in 18 (26.9%). As appropriate treatment we included valproate, phenobarbital, benzodiazepines, lamotrigine, and topiramate, in monotherapy or in different combinations. Treatment with carbamazepine, oxcarbazepine, or phenytoin was considered inappropriate.

The VNPP and habitual methods of activation were performed as described in Table 1 and lasted 4–6 h. Video-EEG was registered during lunchtime and postprandial sleep. Upon awakening, patients were submitted to 5-min hyperventilation (HV) and intermittent photic stimulation (IPS).

Table 1.   Video-EEG protocol
Recording of background activity, awake, for 30 min
Eyes opened/closed (5 min)
Reading a Portuguese text (patients read the same sentences aloud that they had read silently); this was a medical text describing seizures
 10 min silently
 10 min aloud
Reading an English text
 10 min silently
 10 min aloud
Speaking aloud for 5 min (patients described their seizures, their lives, and the impact of epilepsy)
Writing for 5 min (patients were asked to write about their seizures)
Mental calculation: subjects responded aloud with answers to four arithmetic problems (18 − 7, 23 + 46, 11 × 11, 125 ÷ 5); when calculation was difficult, an easier problem was presented
Written calculation: patients responded in writing to one arithmetic problem (15 × 67 × 23 × 48)
Drawing: patients were instructed to draw a family, a house, and a clock showing quarter to four
Spatial construction:
 Patients performed a sequence of tasks, for 10 min each: Rubik’s cube, Jenga tower, Conundrum’s cube, Pyramid puzzle, Hanoi tower
 For 5 min they reproduced a figure with matchstick pattern and for 1 min, plastic models
Photic stimulation for 5 min (flashes 1, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 24, 30, 33, 50, and recovered)
Hyperventilation for 5 min

When a task induced epileptiform discharges or seizures, reproducibility was confirmed by retrial of the same or a modified task. According to the effect of induction or inhibition of discharges, four groups were identified: (1) provocative effect, when there was induction of discharges; (2) inhibitory effect, when there was reduction; (3) mixed effect, when there was provocative effect in a determined category of task and inhibitory effect in another; and (4) no effect.

The same criteria defined by Matsuoka et al. (2000) were used in the EEG analysis: When no discharge was found on the awake EEG, “a provocative effect” meant that one or more tasks induced paroxysmal discharges and its reproducibility was confirmed by retrial. No discharges either in the awake or in the task EEG was judged as “no effect.” When epileptiform discharges were found on the awake EEG, the discharge index was calculated for each task as follows: The number of discharges per recording time (number/minute) during each task condition was divided by the frequency (number/minute) during awake recording. Discharge indexes greater than 2.0 were considered as “provocative effect”; below 0.5 as “inhibitory effect”; and between 0.5 and 2.0 as “no effect” (Matsuoka et al., 2000). Reproducibility was always mandatory to exclude contamination of incidental seizures or discharges.

In VNPP, the different modalities of tasks were categorized according to Matsuoka et al. (2005) in two groups: action-programming (reading aloud, speaking, writing, written calculation, drawing, and spatial construction) and thinking (reading silently and mental calculation). In each of these groups, we considered tasks to be related to spatial (mental and written calculation, drawing, and spatial construction) or linguistic (reading aloud and silently, speaking, and writing) functions.

When both inhibitory and provocative effects were observed in different tasks within the same group, a mean of the discharge indexes was calculated. The same methodology was used for habitual methods of activation. The provocative effect of eyes opening and closure and IPS was classified as present or absent.

Data were analyzed using SPSS for Windows, version 10.0 statistical software (SPSS Inc., Chicago, IL, U.S.A.). The comparison between awake (considered as 1) and mean of discharge indexes on each task was performed using the z-test for single means. Indexes of anxious and nonanxious patients were compared using Student’s t-test. The criterion for statistical significance was p < 0.05.


At the time of VNPP the mean age was 24.3 ± 8.33 years (range 12–53 years) and duration of epilepsy 11.3 ± 8.85 (range 1–38); 54 patients (71.0%) had a positive family history of epilepsy, being JME in 9 (11.8%). Demographic and clinical data are in Table 2.

Table 2.   Demographic and clinical data of 76 patients with juvenile myoclonic epilepsy
DataCategoryn (%)
GenderMale35 (46.1)
Female41 (53.9)
SchoolingElementary27 (35.5)
Secondary40 (52.6)
University level9 (11.8)
Antiepileptic therapyWithout antiepileptic drugs9 (11.8)
Appropriate treatment49 (64.5)
Inappropriate treatment18 (23.7)
Type of seizuresOnly myoclonia1 (1.3)
Myoclonia + generalized tonic–clonic41 (54.0)
Myoclonia + generalized tonic–clonic + absences32 (42.1)
Myoclonia + absences2 (2.6)
Seizure controlControlled23 (30.3)
Noncontrolled53 (69.7)

Video-EEG neuropsychological protocol effect

Awake EEG was normal in 45 (59.2%) and showed epileptiform discharges in 31 (40.8%), consisting of spike and polyspike-wave complexes. Seizures were observed in 31 patients (40.8%): myoclonia in 23 (74.2%), absences in 16 (51.6%), and GTCS in 4 (12.9%). Myoclonia were observed during VNPP in 13 patients, HV in 7, and IPS in 5. Among those patients with myoclonic seizures, four presented perioral reflex myoclonia during speaking or reading aloud, in two of them accompanied by reflex limb myoclonia during action-programming tasks. Absences were observed in 16, during HV in 12, VNPP in 3, and IPS in one. GTCS were observed in four patients: one during resting, two during habitual methods of activation (somnolence and HV), and the last during spatial construction tasks of the VNPP.

Twenty-nine of the 76 patients (38.2%) presented provocative VNPP effect in at least one task. The rate of provocative effect was similar in patients with paroxysms in the awake EEG (14 of 31, 45.2%) and in those without (15 of 45, 33.3%; p = 0.34). Among the former, 11 (35.5%) also had inhibitory effect in another task, thus considered as having a mixed effect. Inhibitory effect was seen in 28 of 31 (90.3%) patients with epileptiform discharges on awake EEG. Thirty of the 76 patients (39.5%) had no effect (Table 3).

Table 3.   Video-EEG neuropsychological protocol effects
VNPP effectProvocativeMixedInhibitoryNoneTotal
  1. Values expressed in parentheses are %.

  2. DA, discharges in awake EEG; MDI, mean of discharge indexes (mean ± SD).

  3. *Z-test for single means = 1 (one-tailed), representing comparison between the mean and a preestablished value (μ0), in this case μ0 = 1.

DA absent15 (33.3)30 (66.7)45
DA present3 (9.7)11 (35.5)17 (54.8)0 (0.0)31
MDI1.01 ± 1.174 (p = 0.483)* 

Regarding treatment there was no difference among appropriately, inappropriately, and nontreated patients in VNPP effects, both in those with discharges in awake EEG and in the group without (Table 4). The mean discharge rate in awake EEG was similar in the three treatment groups: 0.21 ± 0.25, 0.17 ± 0.25, and 0.23 ± 0.33 discharges/min, respectively (p = 0.91).

Table 4.   Effects of Video-EEG neuropsychological protocol according to treatment and seizure control. The mean of discharge indexes in each subgroup is compared to the awake value equal to 1 and to each other
VNPP effectProvocativeMixedInhibitoryNoneTotalp-value
  1. Values expressed in parentheses are %.

  2. DA, discharges in awake EEG; MDI, mean of discharge indexes (mean ± SD);

  3. MDR, mean discharge rate *z-test for single means = 1 (one-tailed).

 DA absent
  Without antiepileptic drugs2 (40.0)3 (60.0)50.676
  Appropriate treatment 10 (30.3)23 (66.7)33
  Inappropriate treatment3 (42.9)4 (57.1)7
 DA present
  Without antiepileptic drugs1 (25.0)0 (0.0)3 (75.0)0 (0.0)40.224
  MDI0.46 ± 0.683 (p = 0.106)*
  Appropriate treatment 2 (12.5)5 (31.3)9 (52.2)0 (0.0)16
  MDR1.12 ± 1.393 (p = 0.364)*
  Inappropriate treatment0 (0.0)6 (54.5)5 (45.5)0 (0.0)11
  MDI1.04 ± 0.973 (p = 0.444)*
 DA absent
  Controlled 5 (25.0)15 (75.0)200.230
  Uncontrolled 10 (40.0)15 (60.0)25
 DA present
  Controlled 1 (33.3)1 (33.3)1 (33.3)0 (0.0)30.308
  MDI 1.21 ± 1.051 (p = 0.380)*
  Uncontrolled2 (7.1)10 (35.7)16 (57.2)0 (0.0)28
  MDI0.99 ± 1.203 (p = 0.478)*

As for seizure control, VNPP effects were similar for groups with and without persistent seizures (Table 4). There was no difference in discharge rate in awake EEG for these two groups: 0.10 ± 0.42 and 0.25 ± 0.37 discharges/min, respectively (p = 0.13).

Video-EEG neuropsychologic protocol effect according to categories of tasks

The effects obtained in the two categories of tasks (action-programming and thinking), subdivided into four subcategories (action-programming linguistic, action-programming spatial, thinking linguistic, and thinking spatial) are depicted in Fig. 1.

Figure 1.

Video-EEG (electroencephalography) neuropsychological protocol effects according to task category.

Action-programming tasks were more effective than thinking tasks in provoking discharges (23.7% and 11.0% of patients, respectively, p = 0.03). No difference was observed between linguistic and spatial tasks in both categories (18.7% and 21.1% of patients, respectively, p = 0.43).

Inhibitory effect was more frequently observed than provocative effect. However, no difference was observed either between action-programming and thinking (p = 1.00) or between spatial and linguistic categories (p = 0.31). When comparing the four subcategories, thinking spatial (represented by mental calculation task) more often exerted inhibitory effect than the other three subcategories (p = 0.02). This subcategory showed inhibitory effect in 18 of 27 (66.7%) of the patients (mean discharge index 0.57 ± 1.07).

Video-EEG neuropsychologic protocol effect according to individual tasks

In Fig. 2, individual tasks are shown in decreasing order of inhibitory effects. Mental and written calculations were the most inhibitory tasks (66.7% and 64.0%, respectively). Among all tasks, those involving manual praxis, such as spatial construction 15 of 75 (20.0%), written calculation 12 of 60 (20.0%), and writing 9 of 62 (14.5%) were the most provocative ones. They were followed by reading aloud (Portuguese 7 of 70, 10.0%; and English 10 of 69, 14.5%), speaking 6 of 65 (9.2%), drawing 5 of 65 (7.7%), reading silently (Portuguese 4 of 63, 6.3%; and English 5 of 64, 7.8%), and mental calculation 5 of 66 (7.6%).

Figure 2.

Video-EEG neuropsychological protocol effects according to individual tasks. Effect of tasks represented in decreasing order of inhibitory effect (blue column). On the right, effect observed in patients with discharges in awake EEG; on the left, in patients without discharges in awake EEG.

Habitual methods of activation

The mean discharge rate during sleep (0.73 ± 1.43) was higher than that during awake EEG (0.20 ± 0.39; p = 0.001). Sleep activated discharges in 36 of 62 (58.0%) and inhibited in 2 patients.

HV was effective in activating discharges in 27 of 70 (38.6%; discharge rate mean 0.58 ± 1.14; p = 0.001) and in inhibiting in 2 patients. Photosensitivity was present in 19 of 64 patients (29.7%). Eye closure sensitivity was observed in 13 patients (21.3%), 8 of whom (14.6%) were also photosensitive (Table 5).

Table 5.   Discharge indexes in habitual methods of activation. The mean of discharge indexes in sleep and hyperventilation was compared to the awake value; photosensitivity and eye closure sensitivity were considered present or absent
Habitual methods effectsProvocativeInhibitoryNoneTotal
  1. Values expressed in parentheses are %.

  2. DA, discharges in awake EEG; MDI, mean of discharge indexes (mean ± SD).

  3. *Z-test for single means = 1 (one tailed).

 DA absent25 ( 64.1)14 (35.9)39
 DA present11 (47.8)2 (8.7)10 (43.5)23
 MDI7.78 ± 14.881 (p = 0.018)* 
 DA absent9 (22.5)39 (77.5)40
 DA present18 (60.0)2 (6.7)10 (33.3)30
 MDI4.08 ± 4.353 (p < 0.001)* 
 Photosensitivity19 (29.7)45 (70.3)64
 Eye closure sensitivity13 (21.3)48 (78.7)61
 Photo and eye closure sensitivity8 (14.6)35 (64.8)54

Comparison between video-EEG neuropsychological protocol and habitual methods of activation effects

Discharge indexes were higher in habitual methods of activation than in VNPP tasks (sleep p = 0.018, HV p < 0.001), as shown in Table 5 and Fig. 3. No difference was observed in VNPP effect between photosensitive and nonphotosensitive patients (Table 6).

Figure 3.

Mean of discharge indexes in habitual methods and in video-EEG (electroencephalography) neuropsychological protocol categories: action-programming, thinking, spatial, and linguistic tasks. *Discharge indexes statistically different from the awake value considered as 1 (z-test). Inhibitory effect was represented by the inverse of discharge indexes mean.

Table 6.   Effects of video-EEG neuropsychological protocol according to photosensitivity.
 VNPP Effect
  1. Values expressed in parentheses are %.

  2. DA, discharges in awake EEG; MDI, mean of discharge indexes (mean ± SD).

  3. *Z-test for single means = 1 (one tailed).

 DA absent
  Yes3 (42.9)4 (57.1)70.678
  No10 (33.3)20 (66.7)30
 DA present
  Yes1 (8.3)5 (41.7)6 (50.0)120.549
  MDI0.73 ± 0. 675 (p = 0.093)*
  No2 (13.3)9 (60.0)4 (26.7)15
  MDI1.04 ± 1. 406 (p = 0.462)*
Eye closure sensitivity
 DA absent
  Yes3 (75.0)1 (25.0)40.287
  No12 (37.5)20 (62.5)32
 DA present
  Yes1 (11.1)6 (66.7)2 (22.2)90.266
  MDI0.63 ± 0. 772 (p = 0.093)*
  No3 (18.7)5(31.3)8 (50.0)16
  MDI1.38 ± 01. 419 (p = 0.153)*

Photosensitive patients were younger than nonphotosensitive. The mean age for each group was 20.4 ± 6.32 and 26.4 ± 8.98 years, respectively (p = 0.01). On the other hand, a correlation between VNPP effects and age was not observed. Patients who had provocative effect of VNPP (22.4 ± 6.92 years) were of a similar age when compared to those without VNPP effect (23.1 ± 7.04) and to patients with inhibitory effects (28.1 ± 11.36, p = 0.09).

Psychiatric evaluation and discharge indexes

Anxiety disorders were diagnosed in 24 of 58 patients (41.4%). Twenty-three had generalized anxiety disorder and one had obsessive-compulsive disorder. Regarding the presence of discharges in the awake EEG, there was no difference between anxious (11 of 24, 45.8%) and nonanxious patients (11 of 34, 32.4%; p = 0.41). In addition, there were no differences between these two groups in the mean of discharge indexes in VNPP. However, in all tasks except writing, there was a tendency to greater discharge indexes in anxious patients (Fig. 3).

Although the anxious group had no significant inhibitory VNPP effect in any task, the nonanxious group had inhibition in mental calculation, a thinking spatial task (0.19 ± 0.56; p = 0.001), and in action-programming spatial tasks (0.62 ± 0.64, p = 0.04). Considering individual tasks, the nonanxious group had significant inhibitory effect when doing mental calculation (0.19 ± 0.56; p = 0.001), speaking (0.28 ± 0.56; p = 0.003), spatial construction (0.31 ± 0.65; p = 0.003), and reading English silently (0.63 ± 0.60; p = 0.035). Mental calculation was the only VNPP task exerting inhibitory effect when both groups of JME patients were analyzed together (0.57 ± 1.07; p = 0.02).


Reflex seizures and epileptiform discharges induced by higher mental function in JME patients have been reported in a few studies (Matsuoka et al., 1988; Senanayake, 1992; Matsuoka et al., 2000; Karachristianou et al., 2004; Mayer & Wolf, 2004; Matsuoka et al., 2005; Mayer et al., 2006). As photosensitivity and eye closure sensitivity, they suggest the presence of regional hyperexcitability and raise questions about the strict concept of JME as an IGE (Wolf & Mayer, 2000; Inoue & Zifkin, 2004; Inoue, 2007).

Among cognitive tasks, linguistic operations and decision-making associated with visuospatial manipulation are the best-characterized triggers (Ritaccio et al., 2002). Matsuoka et al. (2000) reported induction of discharges by cognitive activities almost exclusively in IGE (36 of 38 patients), particularly JME (22 of 36).

We found a VNPP provocative effect in at least one task in 29 of 76 patients (38.2%), a rate much inferior to that described in smaller series such as 21 of 25 (84%) of treated (Matsuoka et al., 1988) and 23 of 30 (76.6%) of nontreated (Karachristianou et al., 2004) JME patients. However, it was closer to 22 of 45 (48.8%) of JME patients reported by Matsuoka et al. (2000).

The rate we found was neither dependent on treatment nor on seizure control. The lack of influence of these factors over the provocative effect might indicate that it is a true reflex trait. According to this idea, two JME patients were reported to remain sensitive to neuropsychological provocative tasks for more than 20 years, indicating that this feature was not transitory (Matsuoka et al., 2002). Indeed, we found no correlation between sensitivity to VNPP provocative effect and age. This did not hold true in relation to photosensitivity, as the mean age was lower in the photosensitive group.

In our series, action-programming tasks were more effective than thinking in exerting provocative effect. This is in accordance with previous reports of Matsuoka et al. (2000, 2005) who also found action-programming as the most crucial provocative task category. In their series, among 38 patients who had EEG discharges activation by neuropsychological tasks, 32 (84.2%), had this effect by action-programming and only 4 (10.5%) by thinking.

On the other hand, in our series, the provocative effect of linguistic and spatial tasks was similar (p = 0.84). Eighteen patients had activation of discharges by action programming tasks, linguistic in 14 (77.8%), and spatial in 15 (83.3%). These numbers also equal those of Matsuoka et al. (2005), who, within the action-programming category, found 33 of 36 activations (91.7%) by linguistic subtype and 30 of 36 (83.3%) by spatial tasks.

Among individual tasks, those related to manual praxis were the most provocative. Although in a much higher rate, Matsuoka et al. (2000) also identified as significant triggers tasks involving praxis, such as writing in 26 of 38 (68.4%), spatial construction in 24 (63.2%), and written calculation in 21 (55.3%). In contrast, mental calculation exerted provocative effect in only three (7.9%) and reading in two of their patients (5.3%). In a smaller series of 25 patients with JME, Mayer et al. (2006) did not consider the role of praxis as pivotal in inducing discharges, since in their series reading exerted provocative effect in 20% and praxis and speaking in 16% each.

Considering calculation, in our series, the provocative effect of written was twice that of mental calculation. This has also been a matter of discussion: whereas in the series of Matsuoka et al. (2000), written calculation exerted provocative effect in 55.3% versus only 7.9% in mental calculation, in the report of Mayer et al. (2006) the effect of calculation was irrelevant (1 of 25 JME patients). In the present series, mental calculation was a powerful inhibitor of discharges (18 of 27, 66.7%). The same was seen in written calculation that exerted inhibition in 16 of 25 (64%) of the patients. However, the last also elicited discharges in 12 of 60 (20%) of patients. Wolf (2005) stated that the same function could sometimes be provocative or inhibitory depending on the state of cortical activation; when resting, a task could precipitate discharges and seizures, whereas when already firing, the activation of a nearby network could inhibit its activity. The parietal cortex would be the structure involved in this process from where the spatial thinking processes would or would not be transformed into voluntary movements (Goossens et al., 1990; Senanayake, 1992; Striano et al., 1993; Inoue et al., 1994). When parietal cortex processes involve the frontal region, for example, in written calculation, it could activate discharges; when there is no recruitment of the motor network, as in mental calculation, the parietal cortex might act as the nearby region, inhibiting the hyperexcitable motor cortex, or on the contrary, promoting discharges on it in the dependency of the state of the network at that moment.

In this series, although tasks involving motor networks exerted excitation, those restricted to thinking had the opposite effect, regardless of spatial or linguistic character. Praxis induction would be the preponderant trigger in JME, implying the role of the motor network in the epileptogenic process (Inoue & Zifkin, 2004). Its hyperexcitability determines the clinical hallmark of this syndrome, the motor manifestations and the great amount of evidence of frontal dysfunction by EEG (Janz, 1985), neuropsychological studies (Devinsky et al., 1997; Pascalicchio et al., 2007; Piazzini et al., 2008), quantitative MRI (Woermann et al., 1999), functional imaging such as positron emission tomography (PET) (Koepp, 2005), and MR spectroscopy (Savic et al., 2000) and, finally, anatomopathologic findings (Meencke & Janz, 1984; Meencke, 1985). Therefore, current evidence suggests that JME is a frontal lobe variant of a multiregional, thalamocortical epilepsy “network” (Koepp, 2005). However, the participation of other cortical areas in this network, in particular the parietal, needs to be clarified.

In addition to specificity, the complexity of a function should be considered. Action-programming might recruit a more extensive amount of hyperexcitable cortex than thinking activities. The sum of tissue involved in a network would explain the fact that the more complex the task, the greater the probability of exerting a provocative effect (Wilkins et al., 1982; Ferlazzo et al., 2005). In our series, comparing drawing and spatial construction, both related to visuospatial functions, similar proportions of provocative and inhibitory effects were observed; however, spatial construction, a task involving a more complex, three-dimensional manipulation was more effective when compared to drawing, a two-dimensional task. Different to what was reported in reading epilepsy, in which the provocative effect is enhanced by increasing the difficulties of the reading material (Wolf & Inoue, 2002), we failed to demonstrate this effect in reading different languages.

As a diagnostic tool, the comparison between habitual methods of activation and VNPP favored the first. The awake EEG recordings in our JME patients resulted in 59.2% of normal tracings. This percentage was higher than previously described, at between 19% and 56% (Canevini et al., 1992; Panayiotopoulos et al., 1994; Murthy et al., 1998; Pedersen & Petersen, 1998), probably due to the fact that 64.5% of the patients in our series were appropriately treated. In our sample, sleep was the most effective activation method, followed by HV and IPS. In JME patients, Matsuoka et al. (1988) considered the neuropsychological tasks as the most effective provocative method, being observed in 21 of 25 (84%), followed by light drowsiness in 17 (68%), HV in 10 (40%), and IPS in 9 (36%). Inoue and Kubota (2000) found photosensitivity in 47 (22.0%) and praxis sensitivity in 27 (12.7%) among 213 JME patients. In a series of selected JME patients who had complained of perioral myoclonia, Mayer et al. (2006) reported myoclonia induced by linguistic and other neuropsychological tasks in 9 of 25 (36%) followed by HV in 7 (28%) and IPS in 5 (20%). In our unselected series, only four (5%) patients presented perioral reflex myoclonia.

Inoue and Kubota (2000) observed that no patient in the praxis-sensitive group showed EEG photosensitivity and stated that “photosensitivity and praxis sensitivity seem to stand in clear contrast.”Mayer et al. (2006) found photosensitivity in both groups—three of their six patients with praxis-induced discharges also showed photosensitivity. We also found praxis sensitivity in photosensitive (4 of 19, 21%) and nonphotosensitive patients (9 of 45, 20%). This indicates that these two traits may coexist in the same patient.

Another interesting application of VNPP is related to the possibility of inhibition of epileptiform discharges through cognitive activity, a fact that beyond helping to clarify physiopathologic aspects may be explored as a therapeutic tool. Unexpectedly, inhibitory (28 of 31, 90.3%) was more prevalent than provocative effect (29 of 76, 38.2%). In a population of 480 patients with epilepsy, Matsuoka et al. (2000) reported 63.9% of inhibitory versus 7.9% of provocative effect. However, the authors decided not to deal with inhibitory effect, since the discharge index under this circumstance did not show a peak clearly separated from that under “no effect” condition. Because inhibitory effect was defined in a very strict manner (between 0 and 0.5), in order to provide a better demonstration of this effect, we decided to consider the inverse of each index.

The stronger inhibitory effect exerted by the subcategory of spatial mental task could be biased, given that this subcategory was composed of only one task—mental calculation—of short duration. Alternatively, the activation of parietal cortex without motor involvement could explain the inhibitory effect by a cortical mechanism as already mentioned. However, considering the great categories of tasks, inhibitory effect was not related to any one of them. This may suggest that the inhibitory cognitive effect might be mediated by a more widespread inhibitory pathway, such as the cortical-thalamic. Normally cortical-thalamic feedback loops regulate the flow of sensory information to the cortex through activation of inhibitory neurons in the reticular nucleus, gating thalamic output to the cortex (Kostopoulos, 2001; Zikopoulos & Barbas, 2006).

In addition to influence of subcortical structures, many authors have emphasized the role of stress in precipitating reflex seizures independent of the epileptic syndrome (Matsuoka et al., 1988; Goossens et al., 1990; Inoue & Kubota, 2000; Inoue, 2007). Furthermore, it was found in JME patients a correlation between the presence of anxiety, lack of seizure control, and antecedent of more than 20 lifetime GTCS (De Araújo Filho et al., 2006). We confirmed this suggestion, once the presence of anxiety determined a less expressive inhibitory effect and a tendency to higher discharge indexes. In addition, a high prevalence of anxiety in JME patients might indicate a common physiopathology of the two morbidities. Recent neuroimaging studies in animal models of anxiety have indicated the participation of the amygdala-prefrontal circuitry in the manifestation of this symptom (Bishop, 2007).

In conclusion, we advise habitual methods of activation for EEG diagnoses in JME. However cognitive tasks, mainly those involving hand praxis, may be useful on an individual basis and for research purposes. The understanding of all these dynamic interactions, including the stress contribution, would help to clarify the physiopathologic mechanisms of this prototype of IGE. At last, it could facilitate the design and development of successful nonpharmacologic therapeutic interventions for the treatment of JME patients as cognitive therapy or antistress programs, as already done in an empirical form (Dahl et al., 1985, 1987; Martinovic, 2001; Wolf, 2002).


We would like to thank Prof. Peter Wolf and Dr. Thomas Mayer for having kindly forwarded their valuable expertise, including handing us their protocol. We would also like to thank Patricia Guilhem de Almeida Ramos for her valuable help with statistics.

CAPES and FAPESP from Brazil and DAAD from Germany supported this study.

We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Disclosure: The authors have no conflicts of interest to declare.