Background: Parkinson disease (PD) has detrimental effects on swallowing function. Treatment options are largely behavioral; thus, patients would benefit from an earlier start to therapy. Early swallowing changes in PD are not well-known, so patients do not typically receive swallowing treatment until later in the progression of PD.
Objective: We used predictive modeling to determine what quantitative swallowing variables best differentiate individuals with early to mid-stage PD from healthy controls.
Methods: Participants included twenty-six individuals with early to mid-stage PD and 26 healthy, age- and sex-matched controls. Swallowing was evaluated by simultaneous high-resolution manometry and videofluoroscopy as well as the Sydney Swallow Questionnaire (SSQ). Binomial logistic regression was performed on 4 sets of data: 1) high-resolution manometry only; 2) videofluoroscopy only; 3) SSQ only; and 4) all data combined.
Results: A model from a combined data set had the highest accuracy in differentiating individuals with PD from controls. The model included maximum pressure in the velopharynx (soft palate), pressure variability in the velopharynx, and the SSQ item concerning difficulty with swallowing saliva. No significant models could be generated using the videofluoroscopy data.
Conclusions: Individuals with PD show quantitative changes in pressure generation and are able to self-assess aspects of swallowing function in the early and mid-stages of PD, even in the absence of swallowing changes seen on videofluoroscopy. A multimodal approach for the assessment of swallowing may be more accurate for determining subtle swallowing changes that occur in the early stages of PD.
Keywords: Parkinson disease; deglutition; high-resolution manometry; videofluoroscopy.