This research was approved by the Certified Review Board of Hiroshima University (Hiroshima, Japan) (approval number: CRB6180006) and conducted in accordance with the national government’s regulations following the Helsinki Declaration of 1964. The study was registered with jRCT (jRCTs062220013). Written informed consent was obtained from all participating patients.

This study comprised cross-sectional exploratory research that is part of the project in which we investigated the effect of cervical percutaneous electrical interferential current stimulation on swallowing. The project protocol has been published previously [6]. This study was conducted at the Hiroshima University Hospital. We enrolled patients who were clinically diagnosed with probable or confirmed PD according to the Movement Disorder Society criteria [5] and possessed the ability to provide informed consent. Additionally, they were required to be at Hoehn and Yahr stages 2–4 at the time of enrollment and provide written informed consent. The inclusion criteria were as follows: receiving consistent levodopa dose for > 1 month and age > 19 years < 86 years. Individuals meeting any of the following criteria were not considered for the study: having a pacemaker or implantable defibrillator; undergoing deep brain stimulation; currently pregnant or attempting to conceive; and a diagnosis or history of head or neck cancer, active pneumonia, or past swallowing rehabilitation. Participants’ motor symptoms and swallowing dynamics were examined using baseline data collected before an intervention.

An X-ray imaging device (Ultimax-i; Canon Medical System Corporation, Tochigi, Japan) was used to conduct the tests while the patients were seated. The trials involved 3 and 10 mL of water mixed with 30%w/w barium contrast medium (Barytester A240 Powder ®, Fushimi Pharmaceutical Co. Ltd., Kagawa, Japan). The patients were instructed to ingest this mixture, delivered via a syringe directly to the floor of the mouth. Employing an X-ray system, we captured images from various angles: forward, towards the lips, backward to the pharyngeal wall, upward to the nasal cavity, and downwards to the upper esophageal sphincter. The recordings were obtained at a rate of 30 frames/s and stored on a DVD. To prevent fatigue phenomena, these investigations outlined here were conducted at the beginning of the videofluoroscopy, with breaks taken to prevent consecutive swallows.

Three experienced dentists (A Hiraoka, A Haruta, and MY), blinded to patient information, evaluated the videofluorographic recordings. Their expertise was in assessing the presence or absence of laryngeal penetration/aspiration and identifying the clearance or prevalence of residue in the oral cavity, vallecular area, and pharynx following a single swallow.

Additionally, they measured the time the bolus took to pass through specific anatomical landmarks and performed a temporal analysis. Two key parameters were calculated: laryngeal elevation delay time (LEDT) and pharyngeal delay time (PDT). LEDT refers to the duration between the bolus tip reaching the vallecula and the peak laryngeal elevation. An established LEDT threshold was set at 0.32 s [10]. In contrast, PDT, is the duration from when the bolus tip reaches the intersection of the lower border of the mandible and the base of the tongue until laryngeal elevation begins. In healthy adults, this interval ranges from 0 to 0.2 s [11].

Furthermore, we assessed whether there was a delay in the swallowing reflex. This was defined as the liquid remaining in the pyriform sinuses for more than 0.1 s (equivalent to 3 frames) before being swallowed [12]. The three observers engaged in thorough discussions and reached a consensus for each observation and measurement.

Two neurologists (MN and HY) performed the clinical evaluation and diagnosis. The recorded data included body mass index (BMI), grip strength, disease duration, alcohol consumption, smoking habits, MDS-UPDRS score [9], medication, Functional Oral Intake Scale (FOIS) score, Eating Assessment Tool-10 (EAT-10) score, and blood test values [13]. Previously reported methods were used to investigate cough reflex, tongue pressure, and peak expiratory flow [14,15,16]. The levodopa equivalent daily dose (LEDD) was calculated according to a recent study [17]. All evaluations were conducted in the On state. By referring to a previous study [18], data were extracted and calculated separately for the following components from the MDS-UPDRS Part 3 to analyze motor symptom subtypes: the mean value of the rigidity (3.3 a–e), tremor (3.15–3.18), postural instability and gait difficulty (PIGD) (3.9–3.13), and limb (3.4–3.8) scores.

Data are expressed as mean ± standard deviation or median (minimum, maximum) for continuous variables and frequencies and percentages for discrete variables. Univariate analyses were conducted using methods such as single regression analysis, contingency tables, analysis of variance, and logistic regression analysis to investigate the data related to the findings of VF for swallowing 3 and 10 mL of water, as well as temporal analyses. In this process, we included the mean values of rigidity scores, tremor scores, PIGD scores, and limb scores, and as mentioned, UPDRS Part 3, in the analyses. Subsequently, we explored the associations between various VF findings and the corresponding UPDRS subscores. Initially, we conducted univariate analyses and extracted factors with p-values < 0.10. These factors were then subjected to a multivariate analysis. Statistical analyses were performed using JMP statistical software version 16 (SAS Institute Inc., Cary, NC, USA). Appropriate statistical tests, such as the χ2 test, Mann–Whitney U test, or unpaired t-test, were employed to assess intergroup variances. Statistical significance was set at p < 0.05.