Abstract
Purpose:
This study examined how cognitive load (CL) influences vocal motor control by integrating acoustic measures with biomechanical trajectories derived from a model-based approach. The aim was to determine whether speakers with vocal hyperfunction (VH) exhibit distinct modulation patterns in laryngeal control compared with healthy control (HC) speakers.
Method:
Recordings from female speakers (63 with VH and 62 HC speakers) performing a Stroop task were analyzed. Trajectories of fundamental frequency (
f
o
) and sound pressure level (SPL) were extracted and used as inputs to a computational model that estimated cricothyroid (
a
CT
) and thyroarytenoid (
a
TA
) muscle activations, as well as subglottal pressure (
P
S
). Acoustic and model-derived measures were segmented and analyzed using mixed-effects models.
Results:
Significant condition effects were found across acoustic and biomechanical features, with parallel patterns in
f
o
–
a
CT
and SPL –
P
S
across segments and groups. No consistent differences were observed for
a
TA
. The strongest effects occurred in SPL, which increased under CL, especially among speakers with VH, with
P
S
showing a similar but weaker trend. For
f
o
and
a
CT
, opposite group patterns emerged at task onset: HC speakers showed slight decreases, whereas those with VH showed increases.
Conclusions:
The correspondence between acoustic and estimated biomechanical modulations under CL supports the utility of this model-based approach for physiologically interpretable inference of laryngeal motor behavior. Speakers with VH showed greater modulation of
a
CT
and
P
S
control, with no systematic changes in
a
TA
, suggesting that arousal-related laryngeal tension primarily affects more superficial control dimension.