Test-retest reliability of dual-recorded brainstem versus cortical auditory-evoked potentials to speech

Background: Auditory-evoked potentials have proven useful in the objective evaluation of sound
encoding at different stages of the auditory pathway (brainstem and cortex). Yet, their utility for
use in clinical assessment and empirical research relies critically on the precision and test–retest repeatability
of the measure.
Purpose: To determine how subcortical/cortical classes of auditory neural responses directly compare in
terms of their internal consistency and test–retest reliability within and between listeners.
Research Design: A descriptive cohort study describing the dispersion of electrophysiological measures.
Study Sample: Eight young, normal-hearing female listeners.
Data Collection and Analysis: Werecorded auditory brainstemresponses (ABRs), brainstem frequencyfollowing
responses (FFRs), and cortical (P1-N1-P2) auditory-evoked potentials elicited by speech sounds
in the same set of listeners. We reassessed responses within each of four different test sessions over a
period of 1 mo, allowing us to detect possible changes in latency/amplitude characteristics with finer detail
than in previous studies.
Results: Our findings show that brainstem and cortical amplitude/latency measures are remarkably stable;
with the exception of slight prolongation of the P1 wave, we found no significant variation in any
response measure. Intraclass correlation analysis revealed that the speech-evoked FFR amplitude
and latency measures achieved superior repeatability (intraclass correlation coefficient .0.85) among
the more widely used obligatory brainstem (ABR) and cortical (P1-N1-P2) auditory-evoked potentials.
Contrasting these intersubject effects, intrasubject variability (i.e., within-subject coefficient of variation)
revealed that while latencies were more stable than amplitudes, brainstem and cortical responses did not
differ in their variability at the single subject level.
Conclusions: We conclude that (1) the variability of auditory neural responses increases with ascending
level along the auditory neuroaxis (cortex.brainstem) between subjects but remains highly stable within
subjects and (2) speech-FFRs might provide a more stable measure of auditory function than other conventional
responses (e.g., click-ABR), given their lower inter- and intrasubject variability