Dysphagia is a common symptom of Parkinson’s disease [
The purpose of this study is twofold: to determine the reliability of timed as well as spatial parameters in videofluoroscopy of swallowing necessary to interpret the study outcome and to determine the differences in swallowing function between patients with Parkinson’s disease and healthy control subjects, matched for age and gender. Qualitative assessment of swallows was based on visuoperceptual evaluation of videofluoroscopic signs by an expert panel, whereas quantitative assessment was obtained by means of a specialized computer software application that allowed capturing and digitizing frames in videofluoroscopic recordings. Using the manual input of anatomical references determined by an expert panel, parameters could be measured.
Patients with idiopathic Parkinson’s disease (diagnosed by a neurologist) having subjective clinical complaints of dysphagia were recruited from several neurological departments in the Netherlands. All patients reported dysphagic complaints ranging from mild to severe, for example, these included slow eating, oral or pharyngeal passage disorder, coughing while drinking, and a diminished quality of life as a direct consequence of dysphagia. None of them suffered from any neurological disease except Parkinson’s. All patients were able to perform a swallow. Other exclusion criteria were deep brain stimulation (DBS), a Mini Mental State Examination (MMSE) score below 23, severe dyskinesia of head and neck (resulting in problems with VFS recording), mental depression, head and neck cancer, severe cardiopulmological disease, speech therapy during the past six months, or surgery of the swallowing mechanism or the central nervous system [
Subject characteristics.
Number of matched pair of subjects | Sex (female/male) | Patients with Parkinson’s disease | Healthy controls | |||
Hoehn and Yahr scalea (H&Y) | Duration of Parkinson’s disease (years) | Functional oral intake scaleb (FOIS) | Age (years) | Age (years) | ||
1 | M | I | 5 | 7 | 70 | 68 |
2 | F | II | 7 | 7 | 64 | 63 |
3 | M | II | 7 | 7 | 50 | 46 |
4 | M | III | 6 | 7 | 80 | 81 |
5 | F | II | >5 | 7 | 73 | 74 |
6 | M | III | 5 | 7 | 57 | 53 |
7 | F | III | 7 | 7 | 62 | 60 |
8 | M | II | 7 | 5 | 70 | 67 |
9 | M | III | 16 | 7 | 70 | 70 |
10 | M | I | >5 | 7 | 66 | 68 |
a(H&Y): the range of scores is one to five, indicating, respectively, unilateral involvement usually with minimal or no functional disability, and confinement to bed or wheelchair unless aided [
b(FOIS): the range of scores is one to seven, indicating nothing by mouth to total oral diet with no restrictions [
All patients underwent a detailed clinical examination by an experienced laryngologist and a speech and language pathologist. Before the VFS examination the presence and severity of dysphagia were assessed using the Functional Oral Intake Scale (FOIS) and a fiberoptic endoscopic evaluation of swallowing (FEES) [
Swallows were analyzed using a specific software application (Image & Physiology SL, Barcelona, Spain) to capture, digitize, and measure the videofluoroscopic swallowing sequences [
Measurements in videofluoroscopy.
Method of measurement | Parameters | ||
Name | Definition | Rating scale or units | |
Temporal parameters | GPJo | Moment of separation of the tongue and soft palate. | Seconds |
GPJc | Moment of contact of the tongue and palate after bolus propulsion. | ||
GPJd | |||
VPJc | Moment of first contact of the soft palate against the posterior pharyngeal wall. | ||
VPJo | Moment of separation of the soft palate and the posterior pharyngeal wall with reentry of air in the retrolingual space from the nasopharynx. | ||
VPJd | |||
LVc | Moment when laryngeal elevation results in making contact between the arytenoid cartilages and the underside of the epiglottis. | ||
LVo | Moment of separation of the arytenoid cartilages and the underside of the epiglottis with reentry of air in the laryngeal vestibule. | ||
LVd | |||
UESo | Moment of opening of the esophagus with entry of either air or barium. | ||
UESc | Moment of closure of the esophagus after bolus transport. | ||
UESd | |||
GPJo-LVc | |||
GPJo-UESo | |||
GPJo-UESc | |||
Aspiration-penetration | Moment of aspiration or penetration | ||
Spatial and temporal parameters (hyoid motion) | Horizontal hyoid motion | Maximum horizontal (anterior) motion during swallowing act. | Millimeters |
Vertical hyoid motion | Maximum vertical motion during swallowing act. | Millimeters | |
Duration vertical hyoid motion | Duration between initiation of swallow and moment of maximum vertical motion. | Seconds | |
Duration horizontal hyoid motion | Duration between initiation of swallow and moment of maximum horizontal (anterior) motion. | Seconds | |
Visuoperceptual parametersa | Preswallow anterior spill | Preswallow loss of bolus from the lips | Five-point scale (0–4) |
Preswallow posterior spill | Preswallow loss of bolus into the pharynx | Five-point scale (0–4) | |
Lingual pumping | Preswallow involuntary repetitive tongue movements | Five-point scale (0–4) | |
Swallow hesitancy | Delayed onset oral transport | Three-point scale (0–2) | |
Piecemeal deglutition | Sequential swallowing on the same bolus | Five-point scale (0–4) | |
Delayed initiation pharyngeal reflex | Delayed onset pharyngeal triggering | Three-point scale (0–2) | |
Postswallow oral residue | Postswallow pooling in the oral cavity | Five-point scale (0–4) | |
Postswallow vallecular pooling | Postswallow pooling in the valleculae | Three-point scale (0–2) | |
Postswallow pyriform sinus pooling | Postswallow pooling in the pyriform sinuses | Three-point scale (0–2) | |
Penetration aspiration scale (PAS) [ | Penetration and/or aspiration | Eight-point scale (1–8) |
aLower scores refer to normal functioning whereas higher scores refer to more severe disability.
Quantitative measurements were determined for each swallow by two experienced raters with extensive training in the analysis of VFS studies of normal and disordered swallowing. Kahrilas et al. used more physiologic references to calculate timed variables in contrast with fixed anatomical landmarks, like the point where the mandible crosses the tongue base [
Movement patterns of the hyoid bone as described by Logemann et al. were used as spatial variables to analyze the swallowing function [
Single frame of the videofluoroscopic recording showing the landmarks used for spatial measurements. Extent of movement of the hyoid bone in the
The following visuoperceptual parameters to evaluate videofluoroscopic signs were scored: preswallow anterior and preswallow posterior spill; lingual pumping; swallow hesitancy; piecemeal deglutition; delayed initiation of the pharyngeal reflex; postswallow oral residue; postswallow vallecular pooling; postswallow pyriform sinus pooling [
Following consensus training, two experts assessed all quantitative and qualitative variables independently at varying speed, ranging from normal to slow motion to frame-by-frame viewing. The raters were blinded to the diagnosis (Parkinson versus healthy subjects) and swallow trials were scored in randomized order. The consensus training in visuoperceptual evaluation was accompanied by a manual including strict, well-defined guidelines to rate these ordinal variables. The exact interpretation per level of each of the three or five point scales was trained during five separate sessions with intervals of one week. During these intervals the expert raters had to accomplish test trials separately that were discussed the next session.
To obtain the intrarater reliability each rater performed repeated measurements of all temporal, spatial, and visuoperceptual variables in all swallows within a period of two weeks.
Statistical analysis was conducted stepwise. First, Intraclass Correlation Coefficients (ICC) and Cronbach’s alphas (>0.65) were computed for all quantitative parameters to determine the reliability and the degree of intra- and interrater (absolute) agreement. For all ordinal parameters Cohen’s Kappa index of agreement was used. Second, for each reliable temporal, spatial, and visuoperceptual variable, group comparisons were performed over the averaged data of the expert raters. Differences between the Parkinson and the healthy control populations were tested for significance by means of the Mann-Whitney
Ten mentally competent dysphagic patients (3 women, 7 men) with a diagnosis of idiopathic Parkinson’s disease were included. The H&Y staging scale ranged from mild to moderate (median II). Table
In Tables
(a) Reliability of the temporal variables (quantitative assessment), intraclass correlation coefficients (ICC)a. (b) Reliability of the spatial-temporal variables (hyoid motion), ICCa. (c) Reliability of the visuoperceptual variables (qualitative assessment), Cohen’s Kappa.
Temporal variable | Intrarater test-retest reliability (ICC) | Interrater reliability | |
Rater 1 | Rater 2 | ||
GPJ opening | — | — | — |
GPJ closure | 0.4 | 0.8 | — |
GPJ duration | 0.4 | 0.8 | — |
VPJ closure | 0.6 | 0.8 | 0.7 |
VPJ opening | 1.0 | 1.0 | 0.7 |
VPJ duration | 0.9 | 1.0 | 0.7 |
LV closure | 0.9 | 0.9 | 0.9 |
LV opening | 0.9 | 0.6 | 1.0 |
LV duration | 1.0 | 0.7 | 0.9 |
UES opening | 0.4 | 0.5 | — |
UES closure | 0.6 | 0.8 | 0.8 |
UES duration | 0.7 | 0.8 | −0.1 |
GPJo_LVc | 0.9 | 0.9 | 0.9 |
GPJo_UESo | 0.4 | 0.9 | — |
GPJo_UESc | 0.6 | 0.8 | −0.02 |
Aspiration or penetration | — | — | — |
aSingle measures ICC, Cronbach’s alpha >0.65.
Spatial-temporal variable | Intrarater test-retest reliability (ICC) | Interrater reliability | |
Rater 1 | Rater 2 | ||
Maximum horizontal (anterior) hyoid motion. | 0.9 | 0.6 | — |
Maximum vertical hyoid motion. | 0.8 | 0.9 | 0.8 |
Duration between initiation of swallow and moment of maximum horizontal (anterior) hyoid motion. | 0.9 | 0.9 | 0.9 |
Duration between initiation of swallow and moment of maximum vertical hyoid motion. | 0.6 | 0.7 | 0.9 |
aSingle measures ICC, Cronbach’s alpha >0.65.
Visuoperceptual variable | Intrarater test-retest reliability | Interrater reliability | |
Cohen’s Kappa | Cohen’s Kappa | ||
Rater 1 | Rater 2 | Rater 1-2 | |
Preswallow anterior spill | — | — | — |
Preswallow posterior spill | 1.0 | 1.0 | — |
Lingual pumping | 1.0 | 1.0 | 0.6 |
Swallow hesitancy | 1.0 | — | — |
Piecemeal deglutition | 1.0 | 1.0 | 0.9 |
Delayed initiation pharyngeal reflex | 0.8 | 1.0 | 0.3 |
Postswallow oral residue | 0.6 | 0.7 | — |
Postswallow vallecular pooling | 0.9 | 0.9 | 0.6 |
Postswallow pyriform sinus pooling | 0.8 | 1.0 | 0.6 |
Penetration aspiration [ | 1.0 | — | — |
Tables
(a) Descriptive statistics of temporal data (quantitative assessment) and level of significance (
Temporal parameters | Parkinson’s disease | Healthy controls | Statistical analyses | ||
Median | 25′, 75′ perc. | Median | 25′, 75′ perc. | ||
VPJ closure | 0.08 | 0.02, 0.12 | 0.03 | −0.02, 0.06 | 0.02 |
VPJ opening | 0.76 | 0.64, 1.00 | 0.71 | 0.66, 0.78 | 0.31 |
VPJ duration | 0.70 | 0.55, 0.94 | 0.66 | 0.64, 0.74 | 0.78 |
LV closure | 0.13 | 0.10, 0.26 | 0.12 | 0.10, 0.20 | 0.58 |
LV opening | 0.91 | 0.72, 0.98 | 0.84 | 0.78, 0.91 | 0.51 |
LV duration | 0.69 | 0.60, 0.80 | 0.66 | 0.60, 0.78 | 0.78 |
UES closure | 0.86 | 0.74, 1.14 | 0.79 | 0.74, 0.86 | 0.13 |
GPJo-LV closure | 0.13 | 0.10, 0.26 | 0.12 | 0.10, 0.20 | 0.58 |
Spatial-temporal parameters | Parkinson’s disease | Healthy controls | Statistical analyses | ||
Median | 25′, 75′ perc. | Median | 25′, 75′ perc. | ||
Maximum vertical hyoid motion (mm) | 13.68 | 9.26, 20.02 | 14.68 | 13.26, 19.50 | 0.35 |
Duration between initiation of swallow and moment of maximum anterior/horizontal hyoid motion | 1.64 | 1.01, 2.62 | 1.12 | 0.86, 2.16 | 0.29 |
Duration between initiation of swallow and moment of maximum vertical hyoid motion | 1.72 | 1.04, 2.09 | 1.28 | 0.88, 2.04 | 0.48 |
Visuoperceptual (ordinal) parameters | Parkinson’s disease | Healthy controls | |||||||||||||
Lingual pumping | 0 | 1 | 2 | 3 | 4 | 0 | 1 | 2 | 3 | 4 | 0.35 | ||||
76% | 17% | 7% | 0% | 0% | 83% | 17% | 0% | 0% | 0% | ||||||
(23) | (5) | (2) | (0) | (0) | (25) | (5) | (0) | (0) | (0) | ||||||
Piecemeal deglutition | 0 | 1 | 2 | 3 | 4 | 0 | 1 | 2 | 3 | 4 | 0.23 | ||||
31% | 41% | 24% | 4% | 0% | 39% | 54% | 7% | 0% | 0% | ||||||
(9) | (12) | (7) | (1) | (0) | (11) | (15) | (2) | (0) | (0) | ||||||
Postswallow vallecular pooling | 0 | 1 | 2 | 0 | 1 | 2 | 0.16 | ||||||||
48% | 42% | 10% | 45% | 55% | 0% | ||||||||||
(14) | (12) | (3) | (13) | (16) | (0) | ||||||||||
Postswallow pyriform sinus pooling | 0 | 1 | 2 | 0 | 1 | 2 | 0.06 | ||||||||
83% | 17% | 0% | 59% | 41% | 0% | ||||||||||
(25) | (5) | (0) | (17) | (12) | (0) |
Before interpreting the present study outcome, the intrarater and interrater reliability was determined. Despite the thorough consensus training using well-defined guidelines as well as the strict methodological protocol for repeated measurements by the expert raters to calculate the intra- and interrater reliability in the present study poor reliability was observed for several parameters. Diverse situations may have contributed to this finding. First, despite their high expertise, the raters may have lacked consensus on a definition of, for example, delayed initiation of the pharyngeal reflex or maximum horizontal (anterior) hyoid motion. Second, patients were found to be rather homogeneous with respect to several of the variables being measured; that is, some of them, such as preswallow anterior spill and preswallow posterior spill, always scored zero. Finally, it cannot be excluded that the lack of internal consistency may have resulted from the nature of the measurement scales and/or instruments; for instance, visuoperceptual variables were scored on three or five point scales. As worldwide no strict guidelines exist, some scope for subjectivity was left to the rater in the discrimination of these ordinal variables. Consequently, these variables were less reliable compared to the temporal or spatial variables that were defined frame by frame using specific software [
Despite the presence of a disturbed swallow physiology in Parkinson patients, few significant differences were found between the two pilot groups. Velopharyngeal junction closure (VPJc) was different, being significantly delayed in the Parkinson group. The remaining temporal, spatial, or visuoperceptual variables did not show significant differences or tendencies between the groups. The lack of significant group differences can be the result of the nature and sensitivity of the assessment tool (VFS). As described in the study of Ertekin et al. conflicting results obtained from radiological and manometric studies in Parkinson patients were found [
In order to design therapy effect studies for dysphagia in Parkinson’s disease, it is necessary to understand the pathophysiology of swallowing in this patient group. In previous studies, several authors investigated pathological mechanisms of oral-pharyngeal dysphagia in Parkinson patients [
In this pilot study, few restrictions were placed on the kind and number of parameters to use. Per swallow (
Despite the special attention paid to methodology in this present study, insufficient reliability for fifteen out of thirty variables (temporal, spatial, and visuoperceptual) was found, particularly for the visuoperceptual variables. The reliable variables could only reveal very few significant differences between both pilot groups. Swallowing seems relatively preserved in the early stages of Parkinson’s disease. Recommendations for future research on pathophysiological aspects of swallowing in Parkinson’s disease may include study designs with larger numbers of dysphagic patients and healthy control subjects. It may be interesting too to include patients with a more severe degree of Parkinson’s disease (higher scores on the Hoehn and Yahr scale) to determine differences between such a patient population and the patients who visit speech therapists as described in this study. The present study has found relatively low intrarater and interrater reliability for many of the variables used despite thorough training and high level of the raters’ expertise. In the literature, information on reliability is usually lacking. After training, raters should have maximum consensus about the exact definition of the variables included and be familiar with the rating scales being used as well as with their levels, preferably anchored in detailed descriptions. Any swallowing study should provide information on training and the intrarater and the interrater reliability in order to allow accurate interpretation of the study outcome. Maybe the suggestion that data described in earlier studies may lack sufficient reliability and, therefore, may not be useful in determining therapy outcome, has been rather unexpected. Still, the problems with reliability in outcome studies such as described in this paper may also be the most interesting as well as the most important finding. Newly developed research will need to take this issue into account so that in the near future, in research on the physiology of swallowing in Parkinson’s disease, the complementary benefits of using qualitative as well as quantitative variables in videofluoroscopy can be studied more thoroughly.