Each participant completed three trials of each of the
two swallow conditions—typical swallow and forceful lingual
swallow—for a total of six measured swallows. The
order of the two swallow conditions was randomized across
participants using a computer randomization technique.
Prior to presenting each bolus, the participant was told
which condition to perform, and the clinician provided the
instructions for that condition. When residue remained in
the pharynx after the swallow, the participant was asked
to clear the residue by completing dry swallows. However,
to be consistent between trials and study participants, data
were collected from the first swallow attempt with each
bolus.
Dependent Measures
The manometric swallow pressures at the base of tongue,
hypopharynx, and UES were measured in millimeters
of mercury (mmHg) on the KayPENTAX Digital Swallowing
Workstation. Each sensor location was measured
individually, as described below (see Figure 4 for visual
depiction).
Sensor 1. The dependent measure is peak amplitude
of the BOT wave during the swallow.
Sensor 2. The dependent measure is peak amplitude
of the hypopharyngeal wave during the swallow.
Sensor 3. The dependent measure is UES nadir pressure
during the swallow. This wave is seen on Sensor 3 (the
most distal sensor) while the UES is relaxing to allow bolus
passage during the swallow. This wave is sometimes referred
to as the “M” wave due to the characteristic pressure signal
generated during the swallow (Huckabee et al., 2005;
Huckabee & Steele, 2006). As the swallow is initiated, the
pressure wave increases from baseline, followed by a decrease
during the swallow, and another increase at the end
of the swallow before returning back to baseline.
Results
Data Reduction
Cronbach’s alpha was utilized to measure internal consistency
across the three repeated trials for each dependent
variable—BOT peak, hypopharynx peak, and UES nadir—
in both swallow conditions. The values obtained by this
analysis revealed high internal consistency (see Table 3), with
all alpha coefficients greater than .90. Therefore, data points
were averaged across swallows within participant, at each
condition, for each variable.
Descriptive Statistics
Data were examined for distribution characteristics.
Box plots, histograms, and measures of skewness and kertosis
were examined to identify outliers and to assess distribution
features. There were no extreme outliers for any
dependent variable of interest in this study. Assessment of
normality was conducted by visual examination of histograms