Sage Journals: Discover world-class research

Abstract

Introduction: Photic driving is believed to be increased in migraineurs and has been interpreted as a sign of cortical hyperexcitability. However, most previous studies have included patients in various phases of the migraine cycle. The results are, therefore, difficult to interpret as neurophysiological abnormalities tend to accumulate close to the attack in migraineurs.

Subjects and Methods: We recorded steady state visual evoked EEG-responses (SSVEPs) for 6, 12, 18 and 24 Hz flash stimuli from 33 migraineurs without aura, eight migraineurs with aura and 32 healthy controls. Interictal recordings were compared pair-wise with recordings before, during and after attack, as well as with EEGs from healthy controls. Driving power was also correlated with sensory hypersensitivity and severity of migraine.

Results: Between attacks, driving responses to 18 Hz and 24 Hz were attenuated in migraineurs without aura. Driving power of 12 Hz increased before the attack. Attack trigger sensitivity, photophobia, pain intensity and a family history of migraine were related to decreased and/or symmetric photic driving.

Conclusions: Earlier results may have overestimated the driving response in migraine due to inclusion of recordings during the preictal interval and/or habituation among controls. Abnormal photic driving may be related to the pathophysiology of clinical sensory hypersensitivity.

Keywords

Migraine SSVEPs trigger photophobia EEG visual excitability

Introduction

Increased photic EEG driving amplitudes, i.e. steady-state visual evoked potentials (SSVEPs), are reported in migraineurs (1–10). Golla and Winter (9) were the first to find prominent photic driving to medium stimulation frequencies (18–24 Hz) in headache patients compared to declining responses from 18 Hz and above in headache-free subjects, the so-called ‘H-response’. Also, with the use of quantitative EEG (QEEG), flash rates above 20 Hz seemed to evoke the largest responses in migraineurs (2–7,10). A replication of former results is warranted for two reasons. First, QEEG studies should preferably be done blindly to avoid selection bias (11), but only two studies describe a blinded design (4,5). Second, in later years, many studies have revealed a larger response to stimuli in migraineurs just before the attack with a variety of neurophysiological methods (12–19). Results from the preictal period have usually been merged with those from the true interictal period in most photic response studies (2,7–10,20–24). However, Nyrke et al. (20) suggested that increased photic driving could be explained by the inclusion of EEGs from the preictal phase (20), but no data about photic driving in the preictal period in migraine seem to be available.

Accordingly, our aim was to perform a blinded study of the EEG-driving pattern to intermittent photic stimulation frequencies above and below 20 Hz in migraineurs, analyzing results from the different ictal phases separately. We also explored the relationship between photic driving and trigger factors, photophobia and severity of migraine.

Subjects and methods

Subjects

We included 41 migraine patients and 31 healthy controls. Patients were recruited from the general population by a newspaper advertisement. Initially, a study nurse trained in headache research did a telephone interview. A neurologist subsequently examined eligible subjects. The study group consisted of patients aged 18–65 years suffering from migraine with aura (MA) or without aura (MwoA) according to The International Headache Society’s Classification of Headache Disorders, 2nd Edition (25), codes 1.2.1 and 1.1 (MA, 8 subjects and MwoA, 33 subjects, respectively). They had 2–6 attacks per month and did not use prophylactic medication. Demographic and clinical data are described in Table 1. We recruited the controls (Ctr) from blood donors. They had a semi-structured interview before inclusion. Exclusion criteria for both migraineurs and controls were any other painful disorder (in particular, co-existing frequent episodic or chronic tension-type headache), neurological disease or connective tissue disorder. In addition, persons with cancer, previous craniotomy or cervical spine surgery, cardiopulmonary or cerebrovascular disease, hypertension, ferromagnetic implants or pregnancy were excluded. Use of neuroleptics or medication for acute or chronic pain, depression or epilepsy as well as alcohol or drug abuse was not allowed.

Table 1.

Background data for the group studied in the interictal interval^a

	Migraine without aura (n = 20)	Migraine with aura (n = 5)	Controls (n = 18)
Age (years)	35.3 (11.2)	39.6 (14.6)	38.5 (11.1)
Sex (women)	18	5	16
Coffee cups/day^b	2.89 (2.5)	3.60 (3.6)	3.22 (1.3)
Smoking (yes)	1	1	3
Alcohol consumption (0–5)^c	2.75 (0.9)	2.40 (0.9)	2.78 (1.1)
Days since last menstruation	10.5 (9.4)	14.8 (9.7)	12.56 (11.9)
Attack duration (h)^d	17.6 (18.3)	20.8 (30.1)	–
Attack frequency (0–4)^d	2.0 (0.6)	2.1 (0.6)	–
Disease duration (years)	17.85 (9.3)	23.80 (11.0)	–
Pain intensity (1–4)^d	2.65 (0.5)	2.20 (0.4)	–
Photophobia (0–2)^d	1.35 (0.8)	1.40 (0.5)	–
Trigger score (0–4)^e	2.4 (1.2)	2.6 (0.9)	–
Family history (0–4)^f	1.4 (1.0)	1.6 (0.9)	–
Triptan users during attacks (n)	18	4	–

Values are mean (SD).

>72 h from attack. 25 migraineurs had at least one EEG more than 72 h from the past and future attack. From the control group an EEG from an age- and sex-matched control with the same recording order was chosen (n = 18).

n = 19, 5, 18 in the three groups due to incomplete data.

0, no drinking; 1, less than once a month; 2, once a month; 3, 2–4 times a month; 4, 2–3 times a week; 5, ≥4 times per week.

During the last 3 months.

Headache triggered by light, sound, smell and physical activity, each given one point and summed.

Children, parents, siblings or grandparents with migraine each given one point and summed.

No significant differences (Wilcoxon Signed Ranks Test, Mann–Whitney U-test, or Fisher’s exact test) were found between groups.

Study protocol

We collected demographic data from all included subjects with questionnaires. Migraineurs also reported data on the clinical presentation of their headache; i.e. what triggered their headache, disease duration, attack duration, attack frequency, headache intensity, and photophobia. In addition, they completed detailed headache diaries, starting at least 2 weeks before and ending at least 2 weeks after the test period. It included entries on pain characteristics for each headache attack, accompanying symptoms and consequences for work and leisure. The subjects had three EEGs recorded on different days. The detailed information from the headache diaries during the test period enabled us in retrospect to pinpoint the exact temporal relation of the EEG recording to the headache attack. The recordings were classified as preictal or postictal if obtained within a time window of 72 h before or after the attack respectively. We defined EEGs during attack or between two closely spaced attacks as ictal, and those recorded more than 72 h before and after an attack as interictal. Twenty-five migraine patients had an interictal recording. To ensure optimal comparability of the control EEG recordings (n = 18), each control for a patient with an interictal migraine recording was matched for age and sex. The EEG recording was the same number in the sequence as that of the patient. This matching was done by a computer. This procedure enabled us to compare preictal, postictal and ictal recordings with interictal recordings in a paired analysis.

Data loss occurred in two patients who were not willing to undergo all three EEG recordings because headache worsened after the tests. One of them did not complete the diary. For group comparisons and correlations we used interictal recordings. EEG recording and analysis were done in a blinded manner without knowledge of disease status. The study was carried out according to the Helsinki Declaration and approved by The Regional Ethics Committee.

EEG recording and stimulus procedure

The EEG was recorded with eyes closed for approximately 30 min. The first 5 min was undisturbed; thereafter a motor test and the photostimulation trains followed. Only the latter is reported in the present paper. Spectral EEG data (26–28), as well as other neurophysiological investigations (12–14) from the same study sample have been reported previously. For a more detailed description of the EEG recording and the inclusion process, see Bjørk et al (28).

Twenty-one scalp electrodes were positioned according to the 10–20 system supplemented by lateral anterior temporal electrodes (29) and channels recording horizontal and vertical eye movements and ECG. EEG was recorded digitally with Nervus 3.0. The sampling rate was 256 per second. The EEG was recorded with a common cephalic reference between F4, Fz, C4, and Cz (FC2 according to the 10–10 electrode system (30)), but re-referenced before analysis with the average reference protocol in Nervus 3.0. The band pass filter setting was 0.5–70 Hz. The subjects were lying down in a dimly lit room. Stimulation was given with the standard photostimulator of Nervus 3.0 at maximum intensity (step 8). Flash power was 1.0 J. The stroboscope was positioned 25 cm (1st train) and 15 cm (2nd train) from the eyes. We measured the illumination to be 20 lux s/pulse at 20 cm with an integrating photometer with autoreset and a Philips 3355 oscilloscope. Intermittent photic stimulation was presented for 10 s for each frequency with a 10 s break in between. The stimulation procedure included the following frequencies 6-12-18-24-6-12-18-24-30-36-40-30-36-40. This flash train was delivered two times. Hence we obtained four 10-s segments for each flash rate. Thirty, 36 and 40 Hz stimuli were applied because they belong to a standard EEG protocol for evaluation of epileptiform photoparoxysmal activity. The latter responses were not analyzed because we intended to focus on the frequencies that earlier has been found to be abnormal in migraineurs (2,10,20) and because these responses are of a very low amplitude. In addition we found it important to restrict the number of variables as a large number of analyses would increase the likelihood of coincidentally significant results. For these reasons we evaluated 6 Hz, 12 Hz, 18 Hz, and 24 Hz photic responses in the present study.

Quantitative analysis

We manually selected the EEG segments during visual stimulation. We restricted the analysis to the occipital channels (O1 and O2) because the strongest sources of the SSVEP signals are in the occipital cortex (31,32). If not contaminated by artefacts, the entire stimulus period (10 s) was sampled. In the case of occasional movement artefacts, we used the longest artefact-free part of the segment. The Fast Fourier transformation application included in the Nervus 3.0 software yielded power per Hertz (µV²/Hz) spectra with the following settings: 0.25 Hz frequency resolution, half-overlapping blocks and a hamming window applied over the entire 4-s epoch. The FFT spectral curves from the four segments were averaged to reduce variance. We estimated the narrow-band driving power for each of the stimulus frequencies in question (X) ± 0.25 Hz. According to Nyrke et al. (7,20), the contribution to the evoked signal from outside the X ± 0.25 Hz band is negligible. Band power values from O1 and O2 were averaged for the main analysis. Asymmetry was calculated as the sign-free difference between left and right electrode: ABS (O1–O2).

Statistical analysis

We used SPSS v17.0 for statistical computations. Where possible non-parametric tests were used because the power variables were not normally distributed and the MA group was small. Before parametric tests all variables were Ln-transformed to obtain normal distribution. Group effects on the response pattern in response to increasing flicker frequencies were evaluated with repeated measures analysis of variance (rANOVA). We used frequency (6-12-18-24 Hz) as the within-subject factor, driving power for each frequency as the dependent variables and diagnostic group as the between subject factor (MA, MwoA and Ctr). Thereafter three post-hoc two-group rANOVAs were performed to determine if any group was significantly different from the others. Sphericity was ascertained for most of the tests; if needed the Greenhouse–Geisser correction was applied.

Driving power for each flash frequency in MA, MwoA and Ctr was separately compared with the non-parametric Kruskal–Wallis test. Contrasts were explored with Mann–Whitney U-tests. In migraineurs we used the paired Wilcoxon signed ranks test to compare differences between the interictal interval and the preictal, ictal and postictal period and Spearman’s rho to explore correlations between trigger sensitivity score (scale 0–4; headache triggered by light, sound, smell and physical activity, each given one point and summed), attack duration, headache intensity (1–4), family history of migraine (scale 0–4; where children, parents, siblings or grandparents with migraine each was given one point and summed), photophobia (0–2), headache history (number of years with headache) and headache frequency (1–4). Correlations and paired tests were performed on the combined migraine group to preserve statistical strength. Fisher exact test was used for categorical variables. To avoid excessive type II errors (33–35) statistical significance was kept at 0.05.

Results

Background data were similar between migraineurs and controls (Table 1). There tended to be fewer days since menstruation in the attack period compared to in the interictal interval (P = 0.12). Apart from this, there were no differences among the migraine subgroups that entered the paired analyses of the preictal, postictal and attack recordings (Table 2).

Table 2.

Background data on different subgroups used in paired comparisons

	Pre-attack (n = 13)	Attack (n = 13)	Post-attack (n = 10)
Sex (women)	11	12	9
Migraine without aura/migraine with aura (n)	10/3	11/2	8/2
Age (years)	33.8 (12.4)	36.8 (12.6)	39.5 (10.5)
Days since last menstruation	12.8 (10.6)	4.5 (6.1)^a	11.3 (7.7)
Disease duration (years)	18.8 (8.8)	20.5 (9.3)	22.6 (11.3)
Pain intensity(1–4)^b	2.7 (0.5)	2.5 (0.5)	2.6 (0.5)
Attack duration (h)^b	18.1 (22.1)	17.5 (17.7)	13.6 (21.6)
Attack frequency (0–4)^b	2.0 (0.6)	2.0 (0.6)	1.9 (0.6)
Photophobia (0–2)^b	1.6 (0.7)	1.2 (0.8)	1.5 (0.7)
Trigger score (0–4)^c	2.5 (1.2)	2.2 (0.9)	2.6 (1.3)
Family history (0–4)^d	1.3 (1.0)	1.4 (1.0)	1.8 (1.2)

Values are mean (SD).

Pre-attack subgroup: migraine patients with EEGs both interictally (>72 h after/before previous/next attack) and within 72 h of next attack. Attack group had EEGs in the attack phase and in the interictal period. Post-attack group had EEGs in the interictal phase and within 72 h after termination of the previous attack.

Non significantly fewer days since last menstruation at the attack recording compared to the paired interictal recording. Interictal data not tabulated (mean 9.2 days, SD 9.1, P = 0.12; Wilcoxon Signed Ranks Test).

During the last 3 months.

Headache triggered by light, sound, smell and physical activity, each given one point and summed.

Children, parents, siblings or grandparents with migraine each given one point and summed.

Effect of diagnosis on photic driving

The driving pattern to flicker stimuli of increasing frequencies was different in MwoA and Ctr (Figure 1, Tables 3 and 4). For MwoA patients the response curve appeared raised for 6 Hz driving power, but thereafter decreased for higher frequencies (driving power versus frequency interaction; P = 0.03).

Figure 1.

Mean driving power in migraineurs with aura (MA), migraineurs without aura (MwoA) and Controls (Ctr) for 6, 12, 18 and 24 Hz flicker stimulation. The response patterns are different between the three groups (P = 0.028) and the main difference is between MwoA and Ctr (P = 0.019, see Table 4). When driving power is separately compared between groups, MwoA patients have significantly attenuated 18 Hz and 24 Hz power compared to Ctr (Mann–Whitney tests, P = 0.007 and P = 0.021, respectively).

Table 3.

Interictal (>72 h from the attack) driving power for 6, 12, 18 and 24 Hz flash stimulation

	Driving power in µV² (SD)
	Migraineurs without aura (n = 20)	Migraineurs with aura (n = 5)	Controls (n = 18)	P-value^a
6 Hz	38.8 (35.7)	62.5 (37.3)	29.6 (19.5)	0.22
12 Hz	55.9 (52.4)	43.9 (27.7)	68.5 (58.3)	0.54
18 Hz	13.8 (11.9)	28.9 (28.8)	33.5 (30.7)	0.03 ^b
24 Hz	9.2 (9.3)	29.1 (34.2)	14.8 (10.0)	0.04 ^c

P-values < 0.05 in bold.

Kruskal–Wallis test.

Migraineurs without aura versus controls P = 0.007.

Migraineurs without aura versus controls P = 0.021.

Table 4.

Driving power related to group and frequency

	Diagnostic group difference (main effect)	Frequency pattern × diagnostic group interaction
	F (P)	F (P-value)
MA vs MwoA vs Ctr	2.1 (0.14)	2.5 (0.03)
MwoA vs Ctr	3.5 (0.07)	3.5 (0.02)
MA vs Ctr	0.6 (0.80)	2.1 (0.09)
MwoA vs MA	1.8 (0.2)	1.3 (0.29)

Results from four repeated measures ANOVAs. P-values < 0.05 in bold.

MA, migraine with aura; MwoA, migraine without aura; Ctr, controls.

The driving amplitude for each separate flash frequency was suppressed for 18 Hz and 24 Hz stimuli in MwoA compared to Ctr (Table 3). The combined migraine group also had less 18 Hz driving power than controls (mean driving power [SD]: 16.9 µV² [17.0] vs 33.5 µV² [30.7]; P = 0.014). However, photic responses were similar between groups for 6 Hz and 12 Hz flicker frequencies.

The 18 Hz driving power was more symmetric in MwoA compared to both Ctr and MA (mean asymmetry [SD]: 5.6 µV² [9.2] vs 15.6 µV² [16.6] and 19.6 µV² [28.7]; P = 0.006 and 0.035, respectively).

Effect of the migraine cycle on photic driving

The driving amplitude to 12 Hz flicker was greater within 72 h before the attack compared to the interictal period in paired analysis (Table 5). Of the 13 patients with a preictal recording within 72 h before the attack, 11 had raised 12 Hz power (Figure 2). Power also appeared to be slightly raised before the attack for 18 Hz and 24 Hz flicker frequencies but the difference did not reach statistical significance (Table 5). Driving power during or after the attack was not different from interictal values (Table 5).

Figure 2.

The 12 Hz driving power in migraineurs between attacks (interictal) and before attack. Preictal power increased in 11 of 13 cases.

Table 5.

Mean fundamental driving power before, after and during attack

		Before attack (n = 13)	P-value^a	Under attack (n = 13)	P-value^a	After attack (n = 10)	P-value^a
6 Hz	Attack-related response	39.3 (35.4)	0.463	44.0 (43.0)	0.650	17.5 (14.6)	0.169
	Paired interictal response	40.2 (40.3)		40.4 (36.1)		33.8 (33.9)
12 Hz	Attack-related response	49.2 (36.8)	0.023	59.9 (64.4)	0.507	71.8 (80.6)	0.241
	Paired interictal response	38.2 (32.3)		54.0 (51.5)		45.9 (49.3)
18 Hz	Attack-related response	16.9 (21.5)	0.311	16.4 (15.8)	0.600	17.7 (18.9)	0.285
	Paired interictal response	12.9 (13.6)		15.8 (18.9)		14.8 (15.5)
24 Hz	Attack-related response	9.4 (12.9)	0.221	12.3 (17.5)	0.311	10.1 (11.0)	0.445
	Paired interictal response	7.9 (10.6)		15.9 (22.9)		10.2 (12.6)

Values (SD) are in µV².

Wilcoxon signed ranks test. P-values < 0.05 in bold.

Driving power symmetry did not change through the migraine cycle (P > 0.075, Wilcoxon signed ranks test).

Effect of clinical variables on photic driving

Trigger sensitivity score correlated negatively with 6 Hz driving power and asymmetry (rho = −0.65, P = 0.00005, Figure 3, and rho = −0.59 P = 0.002, respectively) and 12 Hz driving power (rho = −0.42, P = 0.04, Figure 3). In order to check the consistency of this result we repeated the analyses on the entire migraine group (regardless of temporal distance to attack, n = 41) in separate analyses of test 1, test 2 and test 3. The correlation results above were reproduced in all three tests (rho = −0.3 to –0.6, P = 0.03 to 0.0001).

Figure 3.

Trigger sensitivity score (0–4, y-axis) correlated negatively with 6 Hz (dashed line) and 12 Hz (filled line) driving power (Ln-transformed power-data, x-axis).

Self-reported photophobia correlated negatively with 6 Hz driving power (rho = −0.52, P = 0.008). A post-hoc comparison of patients with little (0–1 on a 0–2 scale, n = 12) versus much (2 on a 0–2 scale, n = 13) photophobia revealed significantly attenuated 6 Hz driving power in patients with severe symptoms (P = 0.014, Mann–Whitney U-test, Figure 4). Photophobia also correlated negatively with 6 Hz driving power asymmetry (rho = −0.44, P = 0.027).

Figure 4.

The 6 Hz driving power in patients with slight (n = 12) versus severe (n = 13) self-reported photophobia (0–2) and controls (n = 18). Patients with severe photophobia had significantly less 6 Hz driving power than patients with slight photophobia (P = 0.014, Mann–Whitney U-test). One outlier is shown (asterisk).

Pain intensity correlated negatively with 24 Hz driving power asymmetry (rho = −0.41, P = 0.04, without outlier rho = −0.37, P = 0.078). Patients with high-intensity attacks (grade 3–4, n = 14) had less 24 Hz driving power asymmetry than those with milder pain (grade 1–2, n = 11, P = 0.044). Family history of migraine correlated negatively with 12 Hz driving power (rho = −0.43, P = 0.033). Headache attack duration, disease duration and attack frequency did not correlate with the driving frequency response.

Discussion

Our main results are decreased interictal photic driving in migraineurs without aura. Driving amplitude increased before attack and correlated negatively with trigger sensitivity and photophobia. Photic driving has been interpreted as increased in migraineurs the last 50 years and regarded as a sign of increased cortical excitability (5,36). Our results challenge this notion and may accordingly provide indirect support to the theory of hypoactive sensory cortices between attacks in migraine (37).

There are several possible reasons why our results are different from former work. Most of the earlier studies seem to have merged recordings from the preictal period and the true interictal period (1–3,7–10,21–24,38–41). Table 6 gives an overview of former SSVEP and H response studies. Column 5 shows that the pre/post attack intervals, i.e. the 72 h before and after attack, have been included in most studies of interictal photic responses. This might have led to an overestimation of driving power. We have only found two studies excluding the preictal interval (Table 6) (4,6). One reported increased 3 Hz and 6 Hz power in frontocentral and parietal areas. However 6 Hz power in occipital areas and 9 Hz power in all regions were similar to controls (6), in agreement with our results. The other study reported increased mean frequency driving (15–30 Hz) in MwoA compared to controls (5). The study also found depressed mean frequency photic driving during attacks (5). We could not confirm this, as driving power appeared depressed both during and between attacks in our data. Older studies have not found changes in photic driving during attack either (8,9).

Table 6.

Summary of SSVEP and H-response studies in migraine

Reference	Study group (classification)	Study parameters	Channels	Pre-/post-attack interval excluded^a	Stimulation mode	Stimulation protocol^b	Results	Blinding
Golla and Winter (1958) (9)	113 Mix mig, 50 N (none)	H-response^c	Te-oc	No/No	EC. Illumination 88000 cd	6, 8→24 (10/0 s, i.e. 100 s continuous freq train)	H response: Mix mig: 108, N 8. N response: Mix mig: 5, N 40. A response: Mix mig: 0, N: 2. Interictal vs attack →	?
Smyth & Winter (1964) (8)	202 Mix mig ; 144 T ; 66 N (D & W)	H response^c	8 channels	No/No	?	6, 8→24 (10/0 s, i.e. 100 s continuous freq train)	H response in 95% (191/202) of Mix mig, 14% of T (18/144), 14% (9/66) of N. Interictal vs attack →	?
Nyrke et al. (1982) (7)	30 Cm, 23 N, (WFN)	F ‘habituation’	2 cm above inion	No/No	EO. Sinusoidal light panel. Constant illumination 400 cd/m²± 30%	10, 14, 16, 18, 20, 22, (36(4)^f/40)	(Cm vs N) ↑ F16–F22, → F10–F14. Relative ‘habituation’ from first 4 s compared to the last 16 s for F18–F22 in both groups	?
Simon et al. (1982) (23)	11 Mix mig, 11 N, (AHC)	F group classification	O1–2, P3–4	No/No	?	10, 12→40 (12/12)	F24, F34 and F38 used successfully in discriminate analysis but no significant group differences^d	?
Nyrke et al. (1989) (20)	30 Cm^e, 20 Cl, 49 N, (AHC)	F asymmetry, ‘habituation’	2 cm above inion	No/Yes (<24 h)	EO Sinusoidal light panel. Constant illumination 400 cd/m²± 30%	10, 16, 18, 20, 22, 24 (12(4)^f/40, 36/40 for 22 Hz)	(Cl vs Cm) ↓ F18, F 20, F 22. (Cl vs N) ↓ F20. (Cm vs N) ↑ F16, F18, F20, F22. (Cl vs Cm and N) ↑ F18-24 asymmetry. (Cl vs N) ↓ ‘habituation’: first 4 s vs last 8 s (trend). First 4 s vs 20–36 s: F 22 Hz ‘habituation’ for all groups, strongest for Cm	?
Nyrke et al. (1990) (66)	21 Cl, 24 N, (AHC)	Asymmetry	2 cm above and 5 cm to the left/right of inion	Yes (<24 h)/Yes (<24 h)	EO Sinusoidal light panel. Constant illumination 400 cd/m²± 30%	10, 16, 18, 20, 22, 24 (12/40)	(Cl vs N) ↑ F18, F 24 asymmetry. ↑ asymmetry from 37% interictally to 50–100% close to attack	?
Puca et al. (1992) (2)	20 MA, 30 MwoA, 20 N, (IHS 88)	F	Fp1–2, F3–4, Fz, T3–6, C3–4, Cz, O1–2, P3–4	No/Yes (<48 h)	EC. Flash power 0.2 J	12, 15, 18, 21, 24, 27, (20/20)	(MA vs N): ↑ F18 ce-te-pa, ↑ F21 fr, ↑ F27 fr-ce-te-pa. (MwoA vs N) ↑ F18 ce-te-pa, ↑ F21 fr-ce-te-pa, ↑ F24 pa-oc, ↑ F27 fr-ce-te-pa-oc. (MwoA vs MA): ↑ F24Hz te-pa-oc	?
Genco et al. (1994) (10)	15 MA^g, 25 MwoA^g, 11 N^g (IHS 88)	F	Fp1–2, F3–4, Fz, T3–6, C3–4, Cz, O1–2, P3–4	No/Yes (<48 h)	EC. Flash power 0.2 J	12, 15, 18, 21, 24, 27, (20/20)	(MwoA vs N): ↑ F21 entire scalp. ↑ F24 fr-ce-te, ↑ F27 fr. (MA vs N): ↑ F15 fr-ce, ↑ F18 fr-pa-oc, ↑ F21 entire scalp, ↑ F24 ce-te. MA = MwoA	?
Puca et al. (1996) (3)	43 MwoA, 15 MA (29 teens), 60 N (30 teens), (IHS 88)	F_(mean) ^h	Fp1–2, F3–4, Fz, T3–6, C3–4, Cz, O1–2, P3–4	No/Yes (<48 h)	EC. Flash power 0.2 J	21, 24, 27, (20/20)	(Mix vs N)↑ F_(mean) oc-pa-te. (MA vs MwoA) ↑ F_(mean) O2, P4, Cz	?
De Tommaso et al. (1998) (5)	20 MwoA, 20 N, (IHS 88)	F attack vs interictal	Fp1–2, F3–4,7–8, T3–6, C3–4, O1–2, P3–4	?/Yes (<72 h)	EC. Flash power 0.2 J	27 (20/20)	(MwoA vs N) Interictal: ↑ F27: oc, te, pa, ce, fr. Attack: → F27. (Attack vs interictal) ↓ F27 te, pa, ce, fr, oc. No relation to headache side	Y
De Tommaso et al. (1999) (4)	120 MwoA, 64 T, 51 N (IHS 88)	F_(mean) ^h	Fp1–2, T3–6, C3–4, O1–2, P3–4	Yes (<72 h)/Yes (<72 h)	EC. Flash power 0.2 J	15, 16,→30 (20/20)	(MwoA and T vs N): ↑ F_(mean) Fp1, C3, C4, P4, O1 and O2. (MwoA vs T) → F_(mean)	Y
Chorlton and Kane (2000) (1)	11 MwoA, 11 MA, 11 basilar, 40 N (IHS 88)	H response^c	C3–O1, C4–O2	No/No		6, 8→24 (10/0 s, i.e. 100 s continuous freq train)	H response = 9 MwoA, 10 MA, 0 basilar,1 N. N response = 2 MwoA, 1 MA, 9 basilar, 28 N. A response = 0 MwoA, 0 MA, 2 basilar, 11 N. 86.4% sens and 97.5 spes (P < 0.001)	?
De Tommaso et al. (2003) (6)	15 MwoA, 15 N, (IHS 88)	F habituation, variability	Fp1–2, F7–8, T3–6, C3–4, O1–2, P3–	Yes (<48 h)/Yes (<72 h)	EC. Flash power 0.2 J	3, 6, 9 (40/20)	(MwoA vs N) ↑ F3: fr-ce, ↑ F6: fr-pa. (MwoA vs N) ↑ 3 Hz response variability. No clear habituation in controls	?

Results for the fundamental component, driving asymmetry, driving response pattern/the H responses and habituation are reported. Only studies which have provided measures or group comparisons for the fundamental driving component (F1) are included.

Mix mig = migraineurs with and without aura, N = normal controls, MA = migraine with aura, MwoA = migraine without aura, Cl = Classic migraine (corresponds to MA), Cm = common migraine (corresponds to MwoA), T = tension headache, D & W = Dow & Witty (1947) (67), AHC = Ad hoc Committee on Classification of Headache (1962) (68), WFN = World Federation of Neurology 1969 (69), IHS 88 = International Headache Society Headache classification 1988 (70), F = fundamental component (F1), ↑ = increased, → similar, ↓ decreased, Ce = central region, Te = temporal region, Fr = frontal region, Pa = parietal region, Oc = occipital region. EC eyes closed, EO = eyes open, freq = frequency, incl = included.

Regarding interictal registrations. Preattack and Postattack interval = the interval before/after attack. Number of hours before/after attack in brackets.

Flash frequencies used. Duration of flicker stimuli/break in between flicker trains in brackets.

H response: photic driving >14 Hz defined as s/r > 0.33, t/s > 0.5 where r = alpha peak frequency abundance/power, s = 18 Hz driving abundance/power and t = 14–16 Hz (trough) abundance/power. N response: declines from 14 Hz (s/r < 0.33), A response: declines from 14 Hz, rises from 16 Hz (s/r > 0.33, t/s < 0.5).

Our computations based on numbers reported in the paper: (23) P = 0.51, P = 0.95 and P = 0.12, respectively, Mann–Whitney U-test.

Cm group gathered for earlier study (7).

Only the first 4 s used for some fundamental component calculations.

Teenagers = 10–17/18 years old.

Mean of fundamental component for F21, F24 and F27.

Possibly, the discrepancy between studies is due to the different stimulus train durations used. Column 7 in Table 6 shows the duration of the stimulation period/resting period (brackets) in each study. In the present study we used 10-s flicker durations, which is half (2–5,10) or one quarter (6) of the durations used in most recent studies. This might explain the discrepancy from former results, as the duration of stimulus may have a strong effect on recorded SSVEPs (32). Habituation is defined as a progressive response decrement as a result of repeated stimulation, occurring more rapidly at higher frequencies (42). In cats marked habituation occurs to intermittent photic stimulation, especially for frequencies above 18–20 Hz (43). It is possible that long-lasting flicker trains in other studies (up to 100 s) could have introduced seemingly higher power in the migraine group when the total stimulus period is averaged due to a possible habituation phenomenon in the controls. Migraineurs probably lack habituation to other kind of sensory stimuli (44–53). Unfortunately, we have not found studies on habituation as such to medium frequency flash stimulation in controls or migraineurs, and it was not investigated in the present study. Common migraine patients (equivalent to MwoA with current classification (25)) showed a habituation-like response to sine-modulated light in a way similar to healthy controls when driving power in the first 4 s was compared with the last 8 s. However, classic migraineurs (equivalent to MA) did not show the same power reduction as the controls (20), possibly reflecting a trend towards habituation deficit. When possible habituation phenomena were abolished by only using the first 4 s of stimulation, the authors did indeed find lower driving power than controls in this group for 16–22 Hz stimulation (though only reaching significance for 20 Hz) (20).

The depressed photic driving in MwoA may indirectly support the notion of low cortical pre-activation in migraine between attacks (54). A tendency towards decreased amplitude for the first evoked potentials epoch has been seen in migraineurs in some studies; (44,45,55,56) according to the theory of low pre-activation, the decreased initial responses at the start of stimulation as well as the deficient habituation in migraineurs could be explained by hypo-activation of sensory cortices (37). The low pre-activation level then allows for a wider range of suprathreshold responses later in the course of the stimulation period (54). The hypo-activation may be caused by inefficient thalamocortical drive caused by changed activity in state setting brain stem nuclei (37,57,58). In animal studies, brain stem as well as thalamic stimulation does indeed increase cortical photic driving amplitudes to frequencies above, but not below, 6 Hz (43). Hence, it is plausible that decreased subcortical stimulation would decrease medium frequency (but not 6 Hz) SSVEPs. The increased driving symmetry between the hemispheres found interictally may possibly reflect thalamocortical synchronization and further support this hypothesis. Interestingly, two studies of alpha-phase synchronization during photic stimulation have found increased synchrony in migraineurs (59,60). De Tommaso et al. (59) attributed this to decreased pre-excitability in thalamocortical pathways and suggested that a state of cortical hypo-activity favoured an aberrant response to repetitive flash stimulation. A fluctuating dysregulation of cortical excitability causing excessive swings in either direction might explain the change of driving amplitude close to the attack in the present study, as well as the deviating results from other neurophysiological studies of evoked responses and transcranial magnetic stimulation (TMS) (61).

We found lower photic drive and symmetric hemispheric activity in patients who easily got attacks after exposure for triggers, i.e. light, sound, smell and physical activity. Photophobia during attack correlated in the same way. These results strengthen the assumption that the sensitivity to environmental stimuli is coupled to aberrant sensory processing in migraine, possibly to low cortical pre-activation. Interestingly, recent TMS data also suggested that patients with more intense pain had lower visual cortex excitability (62) in line with our results. The biological trigger mechanism for environmental stimuli in migraine is unknown. Recently, brain stem nuclei dysregulation has been hypothesized to be central for the pathogenesis (63,64). One may speculate that a well-regulated photic driving response serves to protect against intense stimuli, whereas a dysregulated, attenuated response predisposes to sensory hypersensitivity. To our knowledge, trigger factor sensitivity has not been correlated to visual responses or other neurophysiological response in humans before.

Other explanations for our results should be considered. We did not randomize the order in which each stimulation frequency was presented, partly for practical limitations in the Nervus program and partly in order to minimize the variance in group comparisons. However, provided that any sequence-effect is similar between groups, the comparison between groups will not be biased. If, on the other hand, group differences were influenced by stimulation order due to a possible habituation deficit in migraineurs, we should expect a low response to 6 Hz and higher responses to the subsequent higher frequencies in migraineurs compared to controls. However, our results showed the opposite: migraineurs had a (non-significant) increased response to 6 Hz whereas the response to the subsequent higher frequencies was decreased. Hence, it is not likely that our main results were biased by the fixed order.

Illumination intensity could theoretically have produced differences between studies, as strong intensity dependence is found in migraineurs (47,65). Illumination intensity during flash light is hard to measure reliably, and is often ambiguously reported (or not mentioned, available technical illumination details in former studies are depicted in column 6 in Table 6). It is accordingly hard to make inferences about the effect of flash intensity between studies. However it seems unlikely that low flash intensity could explain low amplitudes in our study as our flash power was 1.0 J as opposed to 0.2 J in some studies which have found increased driving amplitudes in migraineurs (2–6,10). However, low stimulation intensity increases the tendency towards habituation in healthy people (42), making it more probable that differences between our results and the low intensity studies are caused by habituation deficit in migraineurs in the latter. Other methodological differences could also account for differences, like the lack of blinding (Table 6) or artefact detection in most studies. The present study seems to be the only one where migraineurs were included from the general population and not from hospital or out patient clinics.

Conclusions

The depressed photic driving reported here is not in agreement with most former studies. This discrepancy may be explained by inclusion of migraineurs in the preictal phase as well as habituation deficit in migraineurs during long intervals of stimulation in earlier investigations. Our results should be reproduced with other blinded, controlled studies which account for the preictal interval in the analyses. Habituation to steady-state visual stimulation should also be investigated in migraineurs.

Footnotes

Acknowledgements

The authors are grateful for the invaluable technical assistance of Grethe Helde, Marit Stjern and Gøril Gravdahl. Grants were provided from The Norwegian Research Council and Norwegian University of Science and Technology. None of the authors had any conflicting interests.

References

Chorlton

Kane

. Investigation of the cerebral response to flicker stimulation in patients with headache. Clin Electroencephalogr 2000; 31: 83–87.

Puca

de Tommaso

Savarese

Genco

Prudenzano

. Topographic analysis of steady-state visual evoked potentials (SVEPs) in the medium frequency range in migraine with and without aura. Cephalalgia 1992; 12: 244–249, discussion 185–6.

Puca

de Tommaso

Tota

Sciruicchio

. Photic driving in migraine: correlations with clinical features. Cephalalgia 1996; 16: 246–250.

de Tommaso

Sciruicchio

Guido

Sasanelli

Puca

. Steady-state visual-evoked potentials in headache: diagnostic value in migraine and tension-type headache patients. Cephalalgia 1999; 19: 23–26, discussion 1.

de Tommaso

Sciruicchio

Guido

Sasanelli

Specchio

Puca

. EEG spectral analysis in migraine without aura attacks. Cephalalgia 1998; 18: 324–328.

de Tommaso

Stramaglia

Schoffelen

Guido

Libro

Losito

. Steady-state visual evoked potentials in the low frequency range in migraine: a study of habituation and variability phenomena. Int J Psychophysiol 2003; 49: 165–174.

Nyrke

Lang

. Spectral analysis of visual potentials evoked by sine wave modulated light in migraine. Electroencephalogr Clin Neurophysiol 1982; 53: 436–442.

Smyth

Winter

. The EEG in migraine. Electroencephalogr Clin Neurophysiol 1964; 16: 194–202.

Golla

Winter

. Analysis of cerebral responses to flicker in patients complaining of episodic headache. Electroencephalogr Clin Neurophysiol 1959; 11: 539–549.

10.

Genco

de Tommaso

Prudenzano

Savarese

Puca

. EEG features in juvenile migraine: topographic analysis of spontaneous and visual evoked brain electrical activity: a comparison with adult migraine. Cephalalgia 1994; 14: 41–46.

11.

Sand

. Electroencephalography in migraine: a review with focus on quantitative electroencephalography and the migraine vs. epilepsy relationship. Cephalalgia 2003; 23(Suppl 1): 5–11.

12.

Sand

Zhitniy

Nilsen

Helde

Hagen

Stovner

. Thermal pain thresholds are decreased in the migraine preattack phase. Eur J Neurol 2008; 15: 1199–1205.

13.

Sand

Zhitniy

White

Stovner

. Brainstem auditory-evoked potential habituation and intensity-dependence related to serotonin metabolism in migraine: a longitudinal study. Clin Neurophysiol 2008; 119: 1190–1200.

14.

Sand

Zhitniy

White

Stovner

. Visual evoked potential latency, amplitude and habituation in migraine: a longitudinal study. Clin Neurophysiol 2008; 119: 1020–1027.

15.

Judit

Sandor

Schoenen

. Habituation of visual and intensity dependence of auditory evoked cortical potentials tends to normalize just before and during the migraine attack. Cephalalgia 2000; 20: 714–719.

16.

Fritzer

Strenge

Göder

Gerber

Aldenhoff

. Changes in cortical dynamics in the preictal stage of a migraine attack. J Clin Neurophysiol 2004; 21: 99–104.

17.

Kropp

Gerber

. Prediction of migraine attacks using a slow cortical potential, the contingent negative variation. Neurosci Lett 1998; 257: 73–76.

18.

Siniatchkin

Gerber

Kropp

Vein

. How the brain anticipates an attack: a study of neurophysiological periodicity in migraine. Funct Neurol 1999; 14: 69–77.

19.

Bowyer

Aurora

Moran

Tepley

Welch

. Magnetoencephalographic fields from patients with spontaneous and induced migraine aura. Ann Neurol 2001; 50: 582–587.

20.

Nyrke

Kangasniemi

Lang

. Difference of steady-state visual evoked potentials in classic and common migraine. Electroencephalogr Clin Neurophysiol 1989; 73: 285–294.

21.

Lehtonen

. Visual evoked cortical potentials for single flashes and flickering light in migraine. Headache 1974; 14: 1–12.

22.

Simon

Zimmerman

Sanderson

Tasman

. EEG markers of migraine in children and adults. Headache 1983; 23: 201–205.

23.

Simon

Zimmerman

Tasman

Hale

. Spectral analysis of photic stimulation in migraine. Electroencephalogr Clin Neurophysiol 1982; 53: 270–276.

24.

Tsounis

Varfis

. Alpha rhythm power and the effect of photic stimulation in migraine with brain mapping. Clin Electroencephalogr 1992; 23: 1–6.

25.

Headache Classification Committee of the International Headache Society. Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 2004; 24(Suppl 1): 1–160.

26.

Bjork

Sand

. Quantitative EEG power and asymmetry increase 36 h before a migraine attack. Cephalalgia 2008; 28: 960–968.

27.

Bjork

Stovner

Engstrom

Stjern

Hagen

Sand

. Interictal quantitative EEG in migraine: a blinded controlled study. J Headache Pain 2009; 10: 331–339.

28.

Bjork

Stovner

Nilsen

Stjern

Hagen

Sand

. The occipital alpha rhythm related to the ‘migraine cycle’ and headache burden: a blinded, controlled longitudinal study. Clin Neurophysiol 2009; 120: 464–471.

29.

Jasper

. Report of the committee on methods of clinical examination in electroenchephalography. Electroencephalogr Clin Neurophysiol 1958; 10: 370–375.

30.

American Clinical Neurophysiology Society. Guideline 5: guidelines for standard electrode position nomenclature. Am J Electroneurodiagnostic Technol 2006; 46: 222–225.

31.

Pastor

Valencia

Artieda

Alegre

Masdeu

. Topography of cortical activation differs for fundamental and harmonic frequencies of the steady-state visual-evoked responses. An EEG and PET H215O study. Cereb Cortex 2007; 17: 1899–1905.

32.

Vialatte

Maurice

Dauwels

Cichocki

. Steady-state visually evoked potentials: focus on essential paradigms and future perspectives. Prog Neurobiol 2010; 90: 418–438.

33.

Perneger

. What’s wrong with Bonferroni adjustments. BMJ 1998; 316: 1236–1238.

34.

Feise

. Do multiple outcome measures require p-value adjustment? BMC Med Res Methodol 2002; 2: 8–8.

35.

Schulz

Grimes

. Multiplicity in randomised trials I: endpoints and treatments. Lancet 2005; 365: 1591–1595.

36.

Gronseth

Greenberg

. The utility of the electroencephalogram in the evaluation of patients presenting with headache: a review of the literature. Neurology 1995; 45: 1263–1267.

37.

Coppola

Pierelli

Schoenen

. Is the cerebral cortex hyperexcitable or hyperresponsive in migraine? Cephalalgia 2007; 27: 1427–1439.

38.

Slatter

. Some clinical and EEG findings in patients with migraine. Brain 1968; 91: 85–98.

39.

Jonkman

Lelieveld

. EEG computer analysis in patients with migraine. Electroencephalogr Clin Neurophysiol 1981; 52: 652–655.

40.

Nyrke

Kangasniemi

Lang

Petersen

. Steady-state visual evoked potentials during migraine prophylaxis by propranolol and femoxetine. Acta Neurol Scand 1984; 69: 9–14.

41.

Mariani

Moschini

Pastorino

Rizzi

Severgnini

Tiengo

. Pattern-reversal visual evoked potentials and EEG correlations in common migraine patients. Headache 1988; 28: 269–271.

42.

Rankin

Abrams

Barry

Bhatnagar

Clayton

Colombo

. Habituation revisited: an updated and revised description of the behavioral characteristics of habituation. Neurobiol Learn Mem 2009; 92: 135–138.

43.

Steriade

Demetresco

. Phenomena of reticular dynamogenesis on various levels of the optic tracts during intermittent light stimulation. J Physiol (Paris) 1960; 52: 224–225.

44.

Schoenen

Wang

Albert

Delwaide

. Potentiation instead of habituation characterizes visual evoked potentials in migraine patients between attacks. Eur J Neurol 1995; 2: 115–122.

45.

Afra

Cecchini

De Pasqua

Albert

Schoenen

. Visual evoked potentials during long periods of pattern-reversal stimulation in migraine. Brain 1998; 121: 233–241.

46.

Ozkul

Bozlar

. Effects of fluoxetine on habituation of pattern reversal visually evoked potentials in migraine prophylaxis. Headache 2002; 42: 582–587.

47.

Wang

Timsit-Berthier

Schoenen

. Intensity dependence of auditory evoked potentials is pronounced in migraine: an indication of cortical potentiation and low serotonergic neurotransmission? Neurology 1996; 46: 1404–1409.

48.

Ambrosini

Rossi

De Pasqua

Pierelli

Schoenen

. Lack of habituation causes high intensity dependence of auditory evoked cortical potentials in migraine. Brain 2003; 126: 2009–2015.

49.

Ozkul

Uckardes

. Median nerve somatosensory evoked potentials in migraine. Eur J Neurol 2002; 9: 227–232.

50.

Kropp

Gerber

. Contingent negative variation during migraine attack and interval: evidence for normalization of slow cortical potentials during the attack. Cephalalgia 1995; 15: 123–128, discussion 78–9.

51.

Evers

Bauer

Suhr

Husstedt

Grotemeyer

. Cognitive processing in primary headache: a study on event-related potentials. Neurology 1997; 48: 108–113.

52.

de Tommaso

Valeriani

Sardaro

Serpino

Fruscolo

Vecchio

. Pain perception and laser evoked potentials during menstrual cycle in migraine. J Headache Pain 2009; 10: 423–429.

53.

Valeriani

de Tommaso

Restuccia

Le Pera

Guido

Iannetti

. Reduced habituation to experimental pain in migraine patients: a CO(2) laser evoked potential study. Pain 2003; 105: 57–64.

54.

Schoenen

. Deficient habituation of evoked cortical potentials in migraine: a link between brain biology, behavior and trigeminovascular activation? Biomed Pharmacother 1996; 50: 71–78.

55.

Ambrosini

de Noordhout

Sandor

Schoenen

. Electrophysiological studies in migraine: a comprehensive review of their interest and limitations. Cephalalgia 2003; 23(Suppl 1): 13–31.

56.

Afra

Proietti Cecchini

Sandor

Schoenen

. Comparison of visual and auditory evoked cortical potentials in migraine patients between attacks. Clin Neurophysiol 2000; 111: 1124–1129.

57.

Mesulam

. Large-scale neurocognitive networks and distributed processing for attention, language, and memory. Ann Neurol 1990; 28: 597–613.

58.

Steriade

Gloor

Llinas

Lopes de Silva

Mesulam

. Report of IFCN Committee on Basic Mechanisms. Basic mechanisms of cerebral rhythmic activities. Electroencephalogr Clin Neurophysiol 1990; 76: 481–508.

59.

de Tommaso

Stramaglia

Marinazzo

Guido

Lamberti

Livrea

. Visually evoked phase synchronisation changes of alpha rhythm in migraine. Correlations with clinical features. Neurol Sci 2004; 25(Suppl 3): S283–S284.

60.

Angelini

de Tommaso

Guido

Ivanov

Marinazzo

. Steady-state visual evoked potentials and phase synchronization in migraine patients. Phys Rev Lett 2004; 93: 038103–038103.

61.

Charles

. Stimulating studies of visual cortical function in migraine. Cephalalgia 2010; 30: 643–645.

62.

Lum

Fook-Chong

Cui

Siow

. Clinical correlates of phosphene perception in migraine without aura: an Asian study. J Neurol Sci 2008; 264: 93–96.

63.

Lambert

Zagami

. The mode of action of migraine triggers: a hypothesis. Headache 2009; 49: 253–275.

64.

Chakravarty

. How triggers trigger acute migraine attacks: a hypothesis. Med Hypotheses 2010; 74: 750–753.

65.

Connolly

Gawel

Rose

. Migraine patients exhibit abnormalities in the visual evoked potential. J Neurol Neurosurg Psychiatry 1982; 45: 464–467.

66.

Nyrke

Kangasniemi

Lang

. Transient asymmetries of steady-state visual evoked potentials in classic migraine. Headache 1990; 30: 133–137.

67.

Dow

Whitty

CWM

. Electro-encephalographic changes in migraine. Lancet 1947; 2: 52–54.

68.

Ad Hoc Committee on Classification of Headache of the National Institute of Neurological Diseases and Blindness. Arch Neurol 1962; 6: 173–176.

69.

World Federation of Neurology. Definition of migraine. Hemicrania 1969; 1: 3–3.

70.

Classification Committee of the International Headache Society. Headache Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 1988; 8: 1–96.

Photic EEG-driving responses related to ictal phases and trigger sensitivity in migraine: A longitudinal,controlled study

Abstract

Keywords

Introduction

Subjects and methods

Subjects

Study protocol

EEG recording and stimulus procedure

Quantitative analysis

Statistical analysis

Results

Effect of diagnosis on photic driving

Effect of the migraine cycle on photic driving

Effect of clinical variables on photic driving

Discussion

Conclusions

Footnotes

Acknowledgements

References