Age-related differences in auditory cortex activity during spoken word recognition

Project description
===================

In the current study we investigated spoken word processing in young and
older adult listeners in the absence of background noise. We compared
paradigms requiring words to be repeated with "attentive listening" (no
motor response required).

Supplemental materials (including presentation scripts, stimuli,
analysis scripts, extracted data, and figures) can be found at
<https://osf.io/vmzag/>.

## Materials

Stimuli for this study were 375 monosyllabic consonant-vowel-consonant
(CVC) words. The auditory stimuli were spoken versions of the above word
sets recorded at 48,000 Hz using a 16-bit Digital-to-Analog converter
with an Audio Technica 2045 microphone in a quiet room. Words were
spoken by a female speaker with a standard American dialect.
Root-mean-square (RMS) amplitude of the stimuli was equated across
recordings. Out of the full set of words, 75 words were vocoded using a
single channel with white noise as a carrier signal using jp_vocode.m
from <http://github.com/jpeelle/jp_matlab>. These stimuli were used for
an unintelligible baseline “noise” condition. The remaining 300 words
were divided into 5 lists of 60 words, using MATCH software (Van
Casteren and Davis, 2007), and were balanced for word frequency (as
measured by the log of the Hyperspace Analogue to Language dataset;
citation), orthographic length, concreteness, and familiarity. One of
these lists was combined with 15 of the noise vocoded words and used for
word repetition task practice outside of the scanner. The remaining four
lists of 60 words served as the critical items inside the scanner, with
half of the lists used for passive listening and the other half for word
repetition. Word lists were counterbalanced participants such that each
word was presented in both “listen” and “repeat” conditions across
participants.

## Participants

Two groups of participants (young and older adults) were recruited for
this study. The young adults were 29 healthy, right-handed adults, aged
19-30 years (M = 23.8 , SD = 2.9), and were recruited via the Washington
University in St. Louis Department of Psychology Subject Pool. Older
adult participants were 31 healthy, right-handed, aged 65-81 years (M =
71.0 , SD = 5.0). Where available, pure-tone thresholds are given for the
left and right ear (L_250 = threshold for left ear at 250 Hz in dB, and so on).

## Procedure

Prior to scanning participants were taken to quiet room. During that
time participants provided informed consent, completed demographic
questionnaires, and a subset had their hearing tested using pure-tone
audiometry. Participants were then instructed for the two tasks they
would perform in the scanner: passive listening and word repetition.
During passive listening, participants were asked to stay alert, still,
and keep their eyes focused on a fixation while listening to a sequence
auditory sounds including words, silence, and burst of noise
(single-channel white noise vocoded words). During word repetition,
participants were asked to do the same as in passive listening, with the
addition of repeating the word they just head aloud. Participants were
instructed to only repeat the words following the volume acquisitions
after each word. Participants were told if they ever could not
understand a word to give their best guess. Participants practiced a
simulation of the word repetition task until the experimenter was
confident that the participant understood the pacing and the nature of
the task.

Functional MRI scanning took place over the course of four scanning
blocks, where participants alternated between blocks of passive
listening and word repetition. The order of blocks was counterbalanced
such that participants were equally likely to begin with a word
repetition or passive listening block. During word repetition,
participants' spoken responses were recorded using an in-bore Fibersound
optical microphone. These responses were scored offline manually.

## MRI data acquisition

MRI data were acquired using a Siemens Prisma scanner (Siemens Medical
Systems) at 3 T equipped with a 32-channel head coil. Scan sequences
began with a T1-weighted structural volume using an MPRAGE sequence
[repetition time (TR) = 2.4 s, echo time (TE) = 2.2 ms, flip angle = 8°,
300 × 320 matrix, voxel size = 0.8 mm isotropic]. Blood oxygenation
level-dependent (BOLD) functional MRI images were acquired using a
multiband echo planar imaging sequence (Feinberg et al., 2010) [TR =
3.07 s, TA = 0.770 s, TE = 37 ms, flip angle = 37°, voxel size = 2 mm
isotropic, multiband factor = 8]. We used a sparse imaging design in
which there was an 2.3 second delay in between scanning acquisitions,
where the TR was longer than the acquisition time to allow for minimal
scanning noise during stimulus presentation and audio recording of
participant responses (Edmister et al., 1999; Hall et al., 1999).

## NeuroVault

Result images can be found at: <https://neurovault.org/collections/6735/>