To assess the detectability of amplitude-modulation following response
(AMFR) elicited at different carrier (CFs) and modulation frequencies
(MFs) during sleep, as well as the usefulness of AMFR in evoked respo
nse audiometry, AMFR was examined in 10 adults with normal hearing whi
le sleeping. The stimulus was a 50 dBnHL sinusoidally amplitude-modula
ted (SAM) tone with a modulation depth of 95%. The MF of the stimulus
tone was varied from 20 to 120 Hz in 20 Hz steps and CFs were 500, 100
0, 2000 and 4000 Hz. An SAM tone with a CF of 20,000 Hz was used for s
timulation in 5 subjects to confirm that the response did not contain
any artifact. Response was determined by phase spectral analysis. The
component synchrony measure of AMFRs at MF of 40 Hz (40-Hz AMFR) was h
igh al lower CFs, but 40-Hz AMFRs at higher CFs were unreliable. The d
etectability of 80-Hz AMFR was high for ail CFs. It was confirmed that
the response waveform was not contaminated by any electromagnetic art
ifact. Eighty-Hz AMFR detected by phase spectral analysis should thus
be useful for predicting frequency-specific hearing thresholds during
sleep.