Transcranial static magnetic field stimulation (tSMS) of the visual cortex decreases experimental photophobia

Abstract

Background

Transcranial static magnetic field stimulation (tSMS) reduces cortical excitability in humans.

Methods

The objective of this study was to determine whether tSMS over the occipital cortex is effective in reducing experimental photophobia. In a sham-controlled double-blind crossover study, tSMS (or sham) was applied for 10 minutes with a cylindrical magnet on the occiput of 20 healthy subjects. We assessed subjective discomfort induced by low-intensity and high-intensity visual stimuli presented in a dark room before, during and after tSMS (or sham).

Results

Compared to sham, tSMS significantly reduced the discomfort induced by high-intensity light stimuli.

Conclusions

The visual cortex may contribute to visual discomfort in experimental photophobia, providing a rationale for investigating tSMS as a possible treatment for photophobia in migraine.

Introduction

Photophobia describes an experience of discomfort or pain in response to light exposure, which can be caused by ophthalmological disorders or neurological conditions, most commonly migraine (1–3). Similar visual discomfort can be induced in healthy subjects by controlled exposure to high-intensity light (experimental photophobia). In migraine, photophobia is related to functional alterations of the trigeminal system, the thalamus, and the visual cortex (4–10). The relation between the excitability of the thalamo-cortical visual system and photophobia remains unclear.

Non-invasive brain stimulation (NIBS) techniques offer an attractive paradigm to modulate cortical excitability (11,12). These techniques typically employ either electrical currents or changing electromagnetic fields (13). More recently, transcranial static magnetic field stimulation (tSMS) was introduced as a new NIBS technique that reduces cortical excitability (14–18), changes ongoing cortical activity (19,20) and affects behavioral performance (19,20). Specifically, tSMS applied to the occipital cortex produces a focal and frequency-specific increase in the power of alpha oscillations (19), associated with reduced visual performance in humans (19) and primates (21).

The aim of this study was to assess whether tSMS applied over the occipital cortex can reduce experimental photophobia. This would provide evidence that the visual cortex is important, directly or indirectly, in the origin of visual discomfort induced by high-intensity light exposure in healthy subjects. We consider this experiment an important intermediate step in motivating a possible clinical use of tSMS in migraine.

Methods

Subjects

Twenty healthy volunteers participated in this study (11 females; age 30.1 ± 9.2 years; range 19–46 years), recruited by local advertisement. Exclusion criteria were significant medical or psychiatric illness, pregnancy, and concurrent use of neuroactive drugs. Exclusion criteria also comprised currently suffering any type of headache, individuals with pacemakers, brain stimulators, medication pumps, or any type of metal object in the head including eyes (except for dental appliances or fillings), which might pose a physical hazard during tSMS. Six subjects wore glasses during the experiment, due to either myopia or astigmatism. The study was approved by the local ethical committee. Informed consent was obtained from all subjects.

Experimental design

This was a real-sham, double-blinded randomized crossover study. The sample size and crossover design were chosen to provide greater statistical power than that of our previous behavioral studies using tSMS (19,20). A power analysis with alpha 0.05 and power 0.80 shows that n = 20 allows a paired t-test to detect significant differences when the effect size Es is greater than 0.67, with Es = (mean1–mean2)/sqrt(SD1² + SD2² – 2*Rho*SD1*SD2). The intervention was tSMS (or sham) applied for 10 minutes on the occiput (Figure 1(a)). The effect of the intervention (real vs. sham) was tested in two experimental sessions at least one week apart, at approximately the same time of the day (Figure 1(b)). In each session, subjects were seated in a comfortable armchair in a dark room, facing the wall, and received six 5 minute blocks of low-intensity and high-intensity visual stimuli (eight stimuli per block; two blocks in baseline, two during the intervention and two post-intervention). After each visual stimulation, subjects were asked to rate visual discomfort, using a visual analog scale (VAS), with 0 for no-discomfort and 10 for very-severe- discomfort. Low-intensity stimuli were used as a control to exclude the possibility that our intervention could, in itself, induce visual discomfort. In each session, subjects were not aware whether sham or real tSMS was used. At the end of each session, subjects were given a forced choice question about whether magnet or sham was received to exclude any difference in perception.

Figure 1.

tSMS of the visual cortex decreases experimental photophobia. (a) Schematic drawing of the helmet, incorporating tSMS and the visual stimulation equipment. (b) Experimental protocol. Each subject repeated the protocol twice, at least one week apart, exchanging the tSMS intervention (sham vs. real). (c) Effect of tSMS on visual discomfort – as assessed by a visual analog scale (VAS; y axis) – induced by low-intensity (lower lines) and high-intensity (upper lines) visual stimuli over time (x axis): at baseline (B1, B2), during the intervention (I1, I2), and post-intervention (P1, P2). tSMS significantly reduced the discomfort induced by high-intensity stimuli. *p < 0.1, **p < 0.01 (Tukey). Error bars represent SEM (standard error of the mean).

Double-blinding

To achieve double blinding, two experimenters participated in each experimental session: Experimenter 1 was not aware whether the intervention was real or sham and she had to record all the answers given by the subjects; experimenter 2 was responsible for choosing randomly whether the experimental session was with the real tSMS or with the sham intervention, and for changing the real or sham magnet during the session (see below). Experimenter 2 was not involved in data analysis.

Visual stimulation

Visual stimulation equipment comprised a modified BMX adjustable helmet, an Arduino Uno board, a 68 ohm resistor and a light-emitting diode (T-1 3/4 LEDs). The emitted light had a wavelength of 475 nm (the blue part of the spectrum). The Arduino Uno board was placed on the top of the helmet, attached to an aluminum platform, to control the pre-programmed light timing and the illumination intensity of the LED. Two metal rods were screwed to the platform, to support the LED at eye level, at 12 cm distance from the nasion. For each 5 minute block, the visual stimulation was presented pseudo-randomly, using two intensities (high intensity = 235 lux, low intensity = 1.6 lux). The light intensity was measured in a darkened room using a commercial luxometer (ISO-TECH LUX-1335 digital light meter). For each block, visual stimulation was applied for 1 second, every 39 seconds, for eight trials (four low light intensity and four high light intensity). Two blocks were performed during baseline, two during the intervention and two post-intervention. After each block, the cylinder was manipulated – sham was replaced by another sham in the sham group, real/sham was replaced by sham/real in the real group – to avoid the subjects receiving only sham becoming aware of being in the sham group. The replacement of the real or sham magnet lasted about 10 seconds.

Intervention

The magnet (or sham cylinder) was placed in a hole made in the helmet over the occipital cortex (centered over the Oz position of the EEG 10–20 system). To deliver tSMS, we used a cylindrical neodymium magnet (MAG60r, Neurek SL, Toledo, Spain). A non-magnetic cylinder of the same size, weight and appearance was used for sham tSMS (MAG60s, Neurek SL, Toledo, Spain). The cylinders were held in place on the helmet with a leather strapping system (Neurek SL, Toledo, Spain). The tSMS was applied for 10 minutes. The baseline and post-intervention blocks were always performed using a sham cylinder (to have a fixed weight for the helmet). After the baseline, the sham cylinder was removed and substituted (by experimenter 2) with another sham or with a real magnet. After the intervention, the sham/real cylinder was removed and substituted (by experimenter 2) with a sham cylinder.

Statistical analyses

We analyzed the effects of tSMS on visual discomfort evoked by low-intensity and high-intensity visual stimuli. VAS measurements at each block were averaged separately for high and low intensity. Since low-intensity data were zero-bounded, they were entered into non-parametric Friedman analysis of variance with TIME (six time points) as within-subjects factor, separately for the real and sham sessions. High-intensity data were verified to be unbounded, squared-root transformed, tested for normality (Kolmogorov-Smirnov, p > 0.2) and entered into a repeated-measures analysis of variance (ANOVA), with INTERVENTION (Real or Sham) and TIME (six time points) as within-subjects factors, using a multivariate approach to bypass the assumptions of compound symmetry and sphericity. In case of significant effects, Tukey’s post hoc tests were applied. Results were considered significant at p < 0.05. Data are reported as mean ± SEM (standard error of the mean).

Results

The experimental procedure was well tolerated. None of the subjects needed to interrupt or terminate the session due to side effects. Three subjects reported mild headache at the end of the experiment (two sham and one real). Subjects were not able to identify any difference between the magnet and sham sessions (Pearson chi-squared = 0.42, p = 0.52).

For low-intensity visual stimuli, the visual discomfort was small (VAS < 1, Figure 1(c)) and did not change with time either in the real session (Friedman chi-squared = 1.6, p = 0.90) or the sham session (chi-squared = 4.3, p = 0.51).

For high-intensity visual stimuli, sensitization was observed. Visual discomfort increased with repetition (TIME: F(5,15) = 10.6, p = 0.0002). Critically, tSMS induced a significant behavioral effect (INTERVENTION × TIME: F(5,15) = 2.9, p = 0.0484). Specifically, the mean VAS was lower with real tSMS compared to sham during the second intervention block (Tukey: p = 0.0012) and displayed a tendency to be lower also in the first post-intervention block (p = 0.06; Figure 1(c)).

Discussion

We found that occipital tSMS reduces the visual discomfort evoked by high-intensity visual stimuli in healthy subjects.

The double-blind sham-controlled design rules out the possibility of our results being due to the awareness of tSMS, experimenter bias, arousal decay or fatigue. The placement of the magnet over the occipital cortex makes it unlikely, due to physical distance (22), that our results could be explained by a direct effect of the static magnetic field on the eyes, optic nerves, or subcortical structures of the visual system. It is also unlikely – although we cannot fully exclude – that the reduction of visual discomfort might have originated in extra-visual structures more related to pain and/or emotions.

We previously demonstrated that tSMS over the occipital cortex increases alpha EEG oscillations (19). Thus, the behavioral changes induced by tSMS may be due to a change of the cortical state of the occipital cortex. In the visual cortex, it is well established that higher alpha oscillatory power is associated with reduced cortical excitability (23,24) and impaired behavioral performance on visual tasks (18,24–26). We therefore suggest that the reduction of the visual discomfort observed here may reflect a reduction of the excitability of the visual cortex. Since neuromodulatory techniques such as rTMS and tDCS modify functional coupling between the thalamus and cortex (28–30), changes in thalamocortical coupling may contribute to our results. Additional mechanisms can obviously not be excluded.

Using high-intensity visual stimuli, visual discomfort increased over time, suggesting a sensitization mechanism that resembles the one observed with somatic pain (27,28). The reduction of visual discomfort induced by tSMS can thus be explained not only by a reduction of responsiveness of the visual cortex, but also by a partial reversal of the sensitization process. The impact of tSMS on sensory sensitization and habituation (19) deserves further investigation.

We conclude that focal static magnetic fields can interfere with sensitization to high-intensity visual stimulation. This result provides motivation for applying this novel NIBS technique, which is both portable and inexpensive, to reduce photophobia in healthy humans. As photophobia seems to play a role in migraine (9), our study provides a rationale for the possible clinical application of tSMS in this clinical condition.

Article highlights

The visual cortex contributes to visual discomfort in experimental photophobia.

tSMS over the occipital cortex modulates visual cortex excitability.

tSMS over the occipital cortex reduces the discomfort induced by high-intensity light stimuli.

This study provides a rationale for investigating tSMS as a possible treatment of migraine.