The study was approved by the ethics committee of our hospital, and informed consent was obtained from each participant. From January 2014 to April 2015, healthy controls and a consecutive series of PD patients with dysphagia were enrolled in this study prospectively. Inclusion criteria for the PD group were as follows: patients 1) with a confirmed diagnosis of PD by a neurologist, 2) with subjective clinical complaints of dysphagia, including slow eating; food, liquid and tablet swallowing difficulties; food sticking in the throat; cough and wet voice after solid/liquid intake; nocturnal sialorrhea; or slow eating [16], 3) with stable disease, 'on state', with the same medications for more than 2 months, and 4) able to perform 3-step commands. Exclusion criteria for the PD group were as follows: participants 1) with severe cognitive impairment, 2) with other neurologic disorders that can induce dysphagia, such as stroke, traumatic brain injury, or motor neuron disease, 3) with a history of other conditions associated with oropharyngeal structures, such as thyroidectomy, vocal cord palsy, or cancer, and 4) with difficulty sustaining the evaluation process or with difficulty applying the US probe to the neck due to a Levin tube or tracheostomy. Medical records data, including age, gender, body mass index (BMI), Mini-Mental Status Examination (MMSE), and Modified Barthel Index (MBI) were recorded in both groups. Duration of PD and a modified Hoehn-Yahr (HY) scale score were also recorded [17].

An ultrasound machine (Accuvix XQ; Medison, Seoul, Korea) was used with a 2–6 MHz curved-array transducer. The evaluations were performed by one investigator (EH Oh), and were repeated three times for all participants. Throughout the process, the least pressure with a large amount of coupling agent was applied on the probe to minimize the effect of the transducer on swallowing movements. The participants were examined in a comfortable upright sitting position with a neutral head position. The transducer was placed in the midsagittal plane on the submental area (Fig. 1A, 1B), which showed a longitudinal view from the tongue base to the thyroid cartilage on a monitor. The participants were asked not to move their head or neck and to maintain a relaxed state, and then the maximum tongue thickness was measured (Fig. 1C). Two anatomic references, the hyoid bone and thyroid cartilage, were used for measuring the approximate hyoid-thyroid distance [10]. This can be visualized as two calcifications, each with an acoustic shadow, indicating ossification of the hyoid bone and thyroid cartilage (Fig. 1C). The distance between the two calcifications is the hyoid-thyroid approximation. The investigator asked the participants to swallow 0.5 mL of water, and simultaneously began to record a video loop with a program for the US equipment. Three swallows were recorded for each participant with intervals of 1 minute. The shortest hyoid-thyroid approximation distance for each swallow was determined by tracing the recorded images and then measuring with an electronic caliper in the US equipment (Fig. 1D). In addition, the frame counts between the initiation of tongue movement and the shortest hyoidthyroid approximation were investigated and calculated in milliseconds, and this was considered the time to hyoid-thyroid approximation.

Of 24 patients with PD, 15 were evaluated by VFSS in an upright seated position. Fluoroscopy (Sonialvision Versa; Shimadzu, Kyoto, Japan) was performed with a lateral view and recorded on VHS cassettes in real time. The VFSS protocol included the following: 1) water intake, 2) puree intake, and 3) semisolid intake (cookies with 5 mL of 20% semisolid or liquid containing barium material) [18].

For VFSS, the focus of our evaluation was on penetration and aspiration, which both occur during the pharyngeal phase. Penetration refers to when the intake bolus remains above the true vocal folds, inferior to the epiglottis and spilling into the laryngeal vestibule. Aspiration refers to materials that remain below the level of the vocal folds before the initiation of swallowing. If the food did not pass into the true vocal fold, it was considered to be penetration, and if the food passed into the true vocal fold, it was considered to be aspiration [18]. The PD patients were then divided into two subgroups based on the presence of penetration or aspiration during VFSS.

The intra-rater reliability of all US parameters, including tongue thickness, shortest hyoid-thyroid approximation, and time to hyoid-thyroid approximation, were analyzed by determining the intraclass correlation coefficient (ICC) of 3 trials for each parameter. An independent t-test was used to compare the baseline characteristics (age, gender, BMI, height, weight, MMSE score, and MBI score) of healthy controls and PD patients, and to compare the shortest hyoid-thyroid approximation distance and tongue thickness between the two groups. Time to hyoidthyroid approximation was analyzed with the Mann-Whitney U test. Spearman rank correlation analysis was performed in order to assess the correlation between US parameters and disease severity, as measured by PD duration, MBI scores, MMSE scores, and the HY scale scores; and also to assess the correlation between each of the US parameters. In addition, PD patients were divided into two subgroups: patients with or without penetration or aspiration during VFSS. US parameters in the two subgroups were compared with the Mann-Whitney U test. All statistical analyses were performed using SPSS ver. 17.0 for Windows (SPSS Inc., Chicago, IL, USA). The level of statistical significance was set at a p-value less than 0.05.