Skip to main content
. 2022 Oct 29;38(4):1059–1071. doi: 10.1007/s00455-022-10526-1

Table 1.

Relationships between swallowing outcomes and HNL via regression modelling

Response variables (possible range) PAS (1–2 vs. 3–8)a FOIS (1–7)b MASA-C (40–200)2 VHNSS solids (0–50)2 VHNSS liquids (0–20)2 VHNSS nutrition (0–40)2
Explanatory variables Odds ratio p β p β p β p β p β p
External HNLc 3.5 <0.01 − 0.7 0.04 − 9.4 <0.01 4.5 <0.01 1.8 0.04 3.7 0.05
Max internal HNL 5.2 <0.01 − 0.4 0.01 − 5.6 <0.01 2.1 0.19 − 0.3 0.44 0.4 0.76
Sum internal HNL 1.2 <0.01 − 0.1 0.01 − 0.7 <0.01 0.4 0.02 0.0 0.46 0.1 0.35
Number of internal sites 1.6 0.04 − 0.2 <0.01 − 1.7 <0.01 1.1 <0.01 0.3 0.04 0.7 0.03
Individual internal sites
 Base of tongue 1.2 0.57 0 0.95 − 1.5 0.27 − 1.0 0.41 − 0.3 0.17 0.2 0.86
 Posterior pharyngeal wall 2.1 0.07 − 0.2 0.24 − 2.4 0.12 1.2 0.43 − 0.6 0.15 − 0.5 0.70
 Epiglottis 2.5 <0.01 − 0.3 0.10 − 3.6 0.02 2.3 0.07 − 0.3 0.53 − 0.2 0.87
 Pharyngoepiglottic folds 2.4 <0.01 − 0.4 0.02 − 4.5 0.01 2.5 0.02 0.1 0.87 0.7 0.49
 Aryepiglottic folds 4.3 <0.01 − 0.6 <0.01 − 6.6 <0.01 3.0 0.02 0.5 0.22 1.7 0.12
 Interarytenoid space 2.3 0.01 − 0.5 0.01 − 5.5 0.01 3.1 0.01 0.5 0.22 1.2 0.18
 Cricopharyngeal prominence 3.3 <0.01 − 0.5 0.01 − 5.9 <0.01 3.4 <0.01 0.6 0.09 1.5 0.14
 Arytenoids 3.4 <0.01 − 0.4 0.03 − 5.2 0.01 3.4 0.01 0.6 0.12 1.6 0.25
 False vocal folds 2.3 0.05 − 0.3 0.09 − 4.9 0.01 2.8 0.02 0.5 0.12 1.2 0.34
 True vocal folds 1.9 0.33 − 0.8 0.28 − 13 0.18 4.1 0.61 4.0 0.28 − 3.6 0.49
 Anterior commissure 2.8 −  −  0.52 − 3.5 0.10 0.2 0.86 0.3 0.56 − 0.1 0.96
 Valleculae 1.8 0.14 − 0.2 0.15 − 3.1 0.02 − 0.5 0.62 − 0.5 0.16 0.7 0.45
 Pyriform sinus 3.9 <0.01 − 0.5 0.01 − 5.8 <0.01 3.2 <0.01 0.4 0.29 0.7 0.41

PAS penetration-aspiration scale, FOIS functional oral intake scale, MASA-C mann assessment of swallowing ability—Cancer, VHNSS vanderbilt head and neck symptom survey, p = p-value

Bold type indicates statistical significance p < 0.05

aLogistic regression with response PAS=1 or 2, with standard errors clustered on participant

bLinear regression, with standard errors clustered on participant

cDetermined by the MDACC Scale rating only