Background
Previous studies investigated cold‐evoked potentials (CEPs) for the assessment of the integrity of cold‐mediating A‐delta fibres and the spinothalamic tract. Nevertheless, several methodological questions remained unanswered to proceed to clinical application. How do latencies and amplitudes vary between CEPs and contact heat‐evoked potentials (CHEPs)? Are there differences between variable and fixed thermode positions or between glabrous and hairy skin? Are CEPs recordable in patients with abnormal cold processing?
Methods
A total of 16 healthy subjects were tested with CEPs and CHEPs at the face, hand and foot. Variable and fixed thermode positions, hairy and glabrous skin were compared. Three patients with abnormal cold processing were tested with CEPs and quantitative sensory testing.
Results
Compared to CEPs, CHEPs latencies were significantly longer at all locations, amplitudes were significantly larger at the face and the hand whilst comparable at the foot. CEPs and CHEPs did not differ significantly between variable and fixed thermode positions using inter stimulus intervals of 8–12 s. CEP latencies were increased by around 20% at the glabrous skin. Patients with known abnormal cold processing (central pain, polyneuropathy, Fabry's disease) showed increased N2 latencies as compared to normal controls.
Conclusions
Inter stimulus intervals of 8‐12 s allow the use of a fixed thermode position for reliable CEPs/CHEPs recording. Hairy skin stimulation results in faster latencies as compared to glabrous skin, without influencing EP amplitudes. In patients with abnormal cold processing, CEPs are recordable and increased latencies may be expected as compared to healthy controls and the healthy contralateral side.
Significance
Cold‐evoked potentials are an innovative, non‐invasive technique to assess cold detection and processing objectively. This study shows that CEP can be recorded from the hairy and glabrous skin, regardless of using fixed or variable thermode positions. Loss of A‐delta fibre function leads to an increased CEP latency, consistent with loss of cold detection in the QST.