Test-retest reliability of visual-evoked potential habituation

Objective Habituation of visual-evoked potentials (VEPs) is typically described as deficient interictally in migraine patients, supposedly indicating altered cortical excitability. Use of this parameter for monitoring changes over time, e.g. under treatment, requires demonstration of test-retest reliability. Methods VEPs were recorded interictally in 41 episodic migraine patients and 40 controls. N75-P100 amplitudes were measured over six consecutive blocks of 75 VEPs each. Amplitude regression slopes and block ratios were used to quantify VEP habituation. Test-retest reliability was assessed over 15 minutes and two to three weeks. Results Controls showed significantly more negative VEP habituation slopes than migraine patients (-0.210.40 vs. 0.04 +/- 0.46 mu V/block, p<0.05). Results were similar for block ratios, though, in the migraine group, VEP habituation significantly increased from test to two- to three-week retest (p<0.05). In addition, VEP habituation test-retest correlations were mostly poor both in migraine patients and controls (intraclass correlation coefficients, 15 minutes: -0.13 to 0.30, two to three weeks: 0.07 to 0.59). Conclusions Deficient VEP habituation in migraine was confirmed. However, the test-retest reliability of VEP habituation was rather weak. Therefore, we suggest that VEP habituation should be used for evaluation of cortical excitability under treatment only at the group level and only when a control group with sham treatment is included.