In our study, we have found a correlation between higher seizure frequency and a higher clustering coefficient and small world index. We also found, however, a higher seizure frequency to be correlated with the history of epilepsy. This could be expected because data were collected in the clinical setting of an epilepsy surgery program. Surgery is preferred as the epilepsy burden (of which seizure frequency is an important factor) becomes subjectively unacceptable; surgical intervention might therefore be evaluated earlier in the disease course when patients suffer from seizures more frequently. Seizure frequency must be considered as possible confounder in this study. However, a Bonferroni correction was performed and our results regarding small world index remained significant, whereas the local clustering reached a p-value that was the same as the cut-off point for significance. We have found our results to be evidence of a robust correlation between epilepsy duration and these network characteristics. Apart from that, seizure frequency is reported as an extrapolation of the number of seizures patients suffered from at time of surgery, which does not truly represent the
