Influences of Motor Speech Impairments on the Presentation of Dysphagia in Progressive Supranuclear Palsy

Abstract

Purpose:

The purpose of this study was to examine whether differences in motor speech features are related to presentations of dysphagia in Progressive Supranuclear Palsy (PSP) given the sparsity of data examining this relationship.

Method:

Motor speech disorder (MSD) type and severity along with specific swallowing variables were analysed to obtain insights among these relationships in 73 participants with PSP.

Result:

Results revealed that most participants (93%) had dysarthria, with 19% having co-occurring apraxia of speech (AOS). Greater MSD severity was related to more severe pharyngeal phase impairments (CI: −0.917, −0.146, p=0.008). While certain motor speech and swallowing scores varied minimally across participants, incremental changes in these functions were more likely to occur when specific MSD features were present. A trend for participants with spastic dysarthria and/or AOS to exhibit more severe dysphagia was observed.

Conclusion:

This study points to the need for thorough neurological evaluation, with inclusion of speech-language pathology consultation, in the standard of care for PSP. Comprehensive assessment of both motor speech and swallowing functions can inform differential diagnosis and assist patients/families facing decisions regarding modalities for communication and nutrition in the setting of neurodegenerative disease. Additional research may yield greater insights about relevant assessment and intervention considerations in PSP.

Introduction

Progressive supranuclear palsy (PSP), a neurodegenerative disease with deterioration across a number of motor and cognitive domains, commonly presents with slow vertical saccades or vertical supranuclear gaze palsy, postural instability, and falls. These are usually evident within the first year of symptom onset. Various PSP phenotypes exist, with the most recent clinical diagnostic criteria proposed by Hoglinger et al. (2017). Each unique phenotype has overlapping parkinsonian features including rigidity, akinesia, and cognitive deficits. Motor speech and language disorders can also present among the core clinical features, with individuals shown to have progressive apraxia of speech (AOS) or progressive nonfluent aphasia (Boeve et al., 2003; Josephs et al., 2005; Josephs et al., 2006). Dysarthria has also been noted, with mixed hypokinetic and spastic often (Collins et al., 1995; Rusz et al., 2015; Kluin, Foster, Berent, & Gilman, 1993), and less frequently ataxic (Kluin et al., 1993), features being described. A recent large cohort study of PSP (Clark et al., 2021) confirmed a high prevalence of dysarthria, predominantly hypokinetic followed by hypokinetic-spastic dysarthria; across all dysarthria types, one-fifth also had co-occurring AOS. In other PSP studies, motor speech disorders (MSDs) are either not specified or vaguely described (Goetz, Leurgans, Lang, & Litvan, 2003; Golbe, Davis, Schonberg, & Duvoisin, 1988; Litvan et al., 1996a, 1996b, 1997; Muller et al., 2001; Warnecke et al., 2010).

Dysphagia may also be a core feature of PSP. While some studies ambiguously note dysphagia signs or symptoms (e.g., Goetz et al., 2003; Golbe et al., 1988; Litvan et al., 1996a, 1996b), others specify abnormalities in oral (Clark et al., 2019; Johnston et al., 1997; Leopold & Kagen, 1997; Litvan et al., 1997; Sonies, 1992; Warnecke et al., 2010), pharyngeal (Borders et al., 2023; Clark et al., 2019; Johnston et al., 1997; Leopold & Kagen, 1997; Litvan et al., 1997; Warnecke et al., 2010), and uncommonly esophageal phases (Clark et al., 2019; Leopold & Kagen, 1997). Dysphagia is observed three-to-four years post disease onset (dell’Aquila et al., 2013) or occasionally sooner (Muller et al., 2001; Warnecke et al., 2010). Correlations between dysphagia severity and disease severity (Clark et al., 2019) and between dysphagia severity and disease duration (Warnecke et al., 2010) have been identified in PSP. Shorter survival rates occurred for individuals with early onset of dysphagia (dell’Aquila et al., 2013; Litvan et al., 1996b) as well as early cognitive deficits, and older age at disease onset (dell’Aquila et al., 2013). For those demonstrating PSP for less than three years, dysphagia symptoms consisted solely of liquids prematurely spilling into the pharynx, with risk of aspiration before swallow initiation (Warnecke et al., 2010); with disease progression, penetration and aspiration occurred during and after swallow initiation. Thus, onset of dysphagia, particularly pharyngeal symptoms, may signal the need to closely monitor swallowing in PSP with disease progression (Clark et al., 2019).

While dysphagia has been documented as a symptom of PSP for some time and while dysarthria and AOS in PSP have also been described, the connection between dysphagia and MSDs in PSP is elusive, with vague reports of co-occurrence early in the disease (Litvan et al., 1996a, 1996b), dysarthria preceding dysphagia (Golbe et al., 1988; Muller et al., 2001), or finding (Litvan et al., 1997) or not finding an association between the two (Warnecke et al., 2010). Limited data (e.g., Muller et al., 2001) show that onset of dysphagia or dysarthria within a year of PSP onset is a distinguishing feature of atypical parkinsonian disorders compared to Parkinson’s disease (PD), though for which atypical disorders this is applicable to remaining undetermined. Data are also lacking on the co-occurrence of dysphagia and MSDs in PSP, with factors contributing to their relationship being unknown. To date, only one study (Clark et al., 2021) evaluated MSD and communication changes across PSP variants. Given the sparsity of data on the relation among motor speech and swallowing functions in PSP, the purpose of this study was to determine whether differences in motor speech features had an effect on presentations of dysphagia in PSP.

For this exploratory study, we hypothesised there would be a relationship between overall severity of MSD and dysphagia. We further hypothesised that MSD type would influence the nature and/or severity of the swallowing impairments observed. Specifically, hypokinetic dysarthria (associated with parkinsonism) was hypothesised to be associated with dysphagia (Kalf, de Swart, Bloem, & Munneke, 2012; Michou, Harris, & Hamdy, 2013; van Hooren, Baijens, Vos, et al., 2016; Takizawa, Gemmell, Kensworthy, & Speyer, 2016), though the presence of spastic dysarthria to be associated with more severe dysphagia. This was based on known disruption of swallowing in early ALS (Hillel & Miller, 1989; Solazzo, Monaco, Vecchio, et al., 2014) in the setting of mixed spastic-flaccid dysarthria (Strand, Miller, Yorkston, & Hillel 1996). Finally, we hypothesised that AOS would further negatively impact swallowing outcomes.

Methods

Participants

Seventy-three participants (41 males, 32 females) diagnosed with probable or possible PSP (median age: 70 years) were included. They presented to the Department of Neurology, Mayo Clinic, with motor speech features suspected to be due to a degenerative disease. Participants were recruited by the Neurodegenerative Research Group and were enrolled into a large National Institutes of Health-funded study. The study was approved by the Mayo Clinic Institutional Review Board, with informed consent obtained from all participants. Inclusion criteria were participants over age 40 with the ability to independently provide evaluation of their function, meet criteria for possible or probable PSP, be less than six years since symptom onset, able to stand unassisted, agreeable to undergo neuroimaging, and without concurrent neurological illnesses that could account for their motor deficits. Participants’ diagnosis of PSP was based on the current international clinical diagnostic criteria by Hoglinger et al. (2017). Table 1 summarizes participant characteristics related to age, education, disease duration, and disease severity.

Table 1.

Participant characteristics (n=73)

Participant characteristics	Mean	Median	Standard deviation	Minimum	Maximum
Age at exam (years)	69.5	70	7.08	54	86
Education (years)†	15.3	16.0	2.74	8	20
Disease duration (years)	4.45	4.0	2.51	1.50	12.0
Progressive Supranuclear Palsy Rating Scale	39.0	37	10.5	17	67

^†Years of education was missing for one male participant.

Procedures

Participants underwent thorough clinical examination of motor speech and clinical and instrumental (videofluoroscopy also referred to as modified barium swallow study) examination of swallowing. Deficits, when present, were diagnosed by expert speech-language pathologists (HMC, JAGS). Thorough neurological examination was conducted by neurologists (KAJ, HB, FA) specializing in behavioral and movement disorders, and included testing of motor, behavioral, functional, and cognitive performance.

Participants were assessed to determine presence, type, and severity of dysarthria and AOS based on perceptual ratings of speech characteristics during standard tasks: conversational speech, repetition of sentences and words, oral reading, and alternating and sequential motor tasks (Darley et al., 1969; Duffy, 2020). Motor speech impairments affecting movement execution and control were diagnosed as one or more dysarthria types (e.g., spastic, ataxic, hypokinetic) and/or problems with voluntary planning/programming of sensorimotor commands indicating AOS. MSD severity was determined based on the 10-point Motor Speech Disorders Severity Rating (MSDSR) scale (1=nonvocal, 10=normal speech) (adapted from Hillel et al. 1989) previously described in Clark et al. (2021). The Apraxia of Speech Rating Scale-Version 3 (Duffy et al., 2023) was also used to quantify the severity of AOS. Additional procedural details for differential diagnosis as well as intra and inter-rater reliability data for MSD judgements are reported in Clark et al. (2021).

The presence, nature, and severity of swallowing impairments were documented using the Penetration-Aspiration Scale (PAS) (Rosenbek, Robbins, Roecker, Coyle, & Wood, 1996), Modified Barium Swallow Impairment Profile (MBSImP) oral total sum scores (OTSS) and pharyngeal total sum scores (PTSS) (Martin-Harris et al., 2008), and Functional Oral Intake Scale (FOIS) (Crary, Mann, & Groher, 2005). The PAS was judged from videofluoroscopic images to determine occurrence and depth of penetration, aspiration, and corresponding residue of consistencies (0=contrast does not enter the airway; 8=contrast passes the glottis, with subglottic residue remaining and no reflexive response, e.g., coughing, clearing throat). As greater variability was present in PAS scores for thin liquids than for other consistencies, these scores were included for data analysis. All but two participants had a PAS score of 1 for puree and solid textures, with two participants scoring no higher than 2 for each of those textures. The MBSImP, a standardized protocol for conducting videofluoroscopy, includes a scoring system applied to 17 oral, pharyngeal, and esophageal swallowing parameters using a 3- to 5-item rating scale corresponding with observed individual performance. Consistencies were administered per the MBSImP protocol (Martin-Harris et al., 2008). MBSImP OTSS (range 0 to 22) and PTSS (range 0 to 29) provide severity of overall oral and pharyngeal phase impairments. Additionally, the 7-point FOIS (1=nothing by mouth; 7=total oral diet, no restrictions) provides outcome measures for clinically determining safe levels of intake for nutritional/hydration needs. Videofluoroscopic procedures and intra and inter-rater reliability findings for swallowing function are reported in Clark et al. (2019).

Determining the severity of the underlying neurologic diagnosis of PSP included the use of the PSP Rating Scale (PSPRS; Golbe & Ohman-Strickland, 2007), completed by the neurologist who evaluated each participant. This qualitative measure consists of 28 items addressing six domains (history/daily activities, mentation, bulbar, ocular motor, limb motor, and gait) (range 0 to >70) specific to PSP.

Data Analysis

For this cross-sectional study, univariate and multivariate linear regression analyses were conducted to examine potential relationships between MSDs and dysphagia. Univariate linear regression analyses were used to explore relationships between presence and severity of dysphagia features (i.e., PAS liquid severity, MBSImP OTSS and PTSS, and FOIS), presence and type of dysarthria (e.g., spastic, hypokinetic), presence of AOS, and severity of the MSD (MSDSR). Multivariate linear regression analyses were performed to control for age, education, gender, and disease severity (PSPRS) to determine whether relationships remained statistically significant after correcting for these variables. Disease severity and age were specifically controlled for as they have been associated with dysphagia onset and severity in PSP (Clark et al., 2019; dell’Aquila et al., 2013; Litvan et al., 1996b; Warnecke et al., 2010; Whitwell et al., 2019). For all measures analysed, higher scores indicated greater impairment except for the MSDSR and the FOIS, for which lower scores denoted greater impairment.

Descriptive statistics were used to examine the frequency and types of MSDs present across participants and to evaluate how motor speech and dysphagia outcome measures were displayed across these variables using median and interquartile ranges (IQR) or proportions (%). Wilcoxon tests were used to make comparisons between groups. This included a secondary analysis to determine whether there were differences among dysarthria types for the four variables we adjusted for and disease duration.

A p value of < 0.05 was considered significant for all unadjusted and adjusted comparisons. All statistical analyses were performed using Jamovi (The jamovi project, 2021) and R statistic (R Core Team, 2020; see also Garbett, 2018).

Results

Descriptive Statistics

Motor Speech

Of the 73 study participants, 93.2% (68) had dysarthria. Table 2 provides group comparisons for MSD severity across participants with and without dysarthria. As expected, the severity of motor speech impairment, based on MSDSR scores, was significantly different between those with and without dysarthria (median 6 and 9, respectively) (p < 0.01). Of those with dysarthria, 64.7% (44 of 68) exhibited hypokinetic dysarthria (monopitch/monoloudness, reduced stress, rapid rushes of speech, variable speech rate, repeated phonemes, inappropriate silences), 13.2% (9) demonstrated spastic dysarthria (strained-strained phonatory quality, slow rate, low pitch), and 22.1% (15) had mixed hypokinetic-spastic dysarthria (Supplementary Material A). Given the few number of participants per group, three individuals with ataxic dysarthria were excluded.

Table 2.

Group comparisons of motor speech and swallowing scores across participants with versus without dysarthria as well as dysarthria with and without co-occurring AOS

Groups	Motor Speech Disorder Rating Scale	Penetration-Aspiration Scale	Oral Total Sum Score	Pharyngeal Total Sum Score	Functional Oral Intake Scale
No dysarthria group (n=5)	9.0 (9.0, 10.0)	1.0 (1.0, 2.7)	7.0 (4.0, 9.0)	1.0 (1.0, 2.7)	7.0 (7.0, 7.0)
Dysarthria group (n=68)	6.0 (5.4, 7.6)	2.0 (1.0, 3.6)	7.0 (5.0, 9.0)	2.0 (1.0, 4.6)	7.0 (6.0, 7.0)
p-value	p<0.01*	p=0.51	p=0.73	p=0.19	p=0.16
Dysarthria without AOS (n=55)	6.0 (6.0, 8.0)	2.0 (1.0, 2.8)	7.0 (7.0, 9.0)	2.0 (1.0, 4.0)	7.0 (5.0, 7.0)
Dysarthria with AOS (n=13)	4.0 (2.7, 7.0)	2.0 (1.0, 4.0)	8.0 (4.7, 9.7)	4.0 (2.7, 7.0)	7.0 (7.0, 7.0)
p-value	p<0.01*	p=0.43	p=0.42	p=0.01*	p=0.19

^*Significant values (p<0.05) for Wilcoxon test

Note: Data are shown as medians and interquartile ranges (IQR).

Table 3 summarizes MSDSR across dysarthria types; figure 1 displays medians and ranges of these scores across participants. Overall, participants had a median MSDSR score of 6 (IQR 5.75, 7.25). Figure 2 depicts MSDSR medians and ranges of scores by dysarthria type. Wilcoxon tests revealed significant differences in MSDSR scores between dysarthria types (p = 0.01). Specifically, participants with spastic dysarthria had a significantly worse MSDSR than those with hypokinetic dysarthria (median 4 (IQR 3, 7) and 6 (IQR 6, 8), respectively; p = 0.04). No differences in MSDSR were found for hypokinetic compared to mixed hypokinetic-spastic dysarthria or spastic compared to hypokinetic-spastic dysarthria (p > 0.05).

Table 3.

Group comparison of motor speech and swallowing severity scores across dysarthria types

Motor speech and swallowing scores	N	Dysarthria type		p-value
		Hypokinetic (n=44)	Spastic (n=9)
Motor Speech Disorder Rating Scale	53	6.0 (6.0, 8.0)	4.0 (3.0, 7.0)	p=0.04*
Penetration-Aspiration Scale	53	2.0 (1.0, 2.6)	2.0 (1.0, 4.0)	p=0.47
Oral Total Sum Score	53	7.0 (5.0, 9.0)	6.0 (3.7, 9.0)	p=0.63
Pharyngeal Total Sum Score	53	2.0 (1.0, 3.0)	4.0 (2.7, 5.7)	p=0.01*
Functional Oral Intake Scale	53	7.0 (6.0, 7.0)	7.0 (7.0, 7.0)	p=0.21
		Hypokinetic (n=44)	Hypokinetic-spastic (n=15)
Motor Speech Disorder Rating Scale	59	6.0 (6.0, 8.0)	6.0 (5.0, 7.0)	p=0.08
Penetration-Aspiration Scale	59	2.0 (1.0, 2.6)	2.0 (1.0, 5.3)	p=0.67
Oral Total Sum Score	59	7.0 (5.0, 9.0)	7.0 (6.2, 9.0)	p=0.40
Pharyngeal Total Sum Score	59	2.0 (1.0, 3.0)	4.0 (2.0, 5.0)	p=0.05
Functional Oral Intake Scale	59	7.0 (6.0, 7.0)	7.0 (5.0, 7.0)	p=0.36
		Spastic (n=9)	Hypokinetic-spastic (n=15)
Motor Speech Disorder Rating Scale	24	4.0 (2.7, 7.3)	6.0 (5.0, 7.0)	p=0.21
Penetration-Aspiration Scale	24	2.0 (1.0, 4.0)	2.0 (1.0, 5.3)	p=0.83
Oral Total Sum Score	24	6.0 (3.7, 9.0)	7.0 (6.2, 9.0)	p=0.40
Pharyngeal Total Sum Score	24	4.0 (2.7, 5.7)	4.0 (2.0, 5.0)	p=0.44
Functional Oral Intake Scale	24	7.0 (7.0, 7.0)	7.0 (5.0, 7.0)	p=0.09

^*Significant values (p<0.05) for Wilcoxon test

Note: Data are shown as medians and interquartile ranges (IQR).

Figure 1.

Histograms and density plots of motor speech and swallowing scores across participants with dysarthria (A-E).

Note. This figure shows the absolute proportion of values (bar lines) and relative densities (shaded area) across the range of variables analysed. The vertical dash lines represent the median values across participants with dysarthria.

Figure 2.

Box plots summarising motor speech and swallowing severity across specific dysarthria types (A-E).

Note. Data are shown as medians (solid and bold horizontal lines) and interquartile ranges (IQR; top and bottom lines of the box plots) by dysarthria type (hypokinetic, n = 44; spastic, n = 9; and hypokinetic-spastic, n = 15). Vertical lines above and below the box plots represent the full range of scores.

AOS was observed in 17.8% (13) of all 73 participants and in 19.1% of those with dysarthria. All participants who had AOS also had dysarthria. Seven (53.8%) participants with AOS exhibited spastic dysarthria, with the rest split equally between having either hypokinetic (23.1%) or hypokinetic-spastic (23.1%) dysarthria. Wilcoxon tests revealed significant differences between MSDSR scores for those with dysarthria plus AOS and those with dysarthria only (median 4 (IQR 2.7, 7) and 6 (IQR 6, 8), respectively; p < 0.01) (table 2).

Swallowing

Table 2 summarizes group comparisons for swallowing scores across participants with and without dysarthria. There were no significant differences in swallowing indices between these groups (p > 0.05).

Figure 1 depicts medians and ranges of scores for swallowing variables across participants with dysarthria; figure 2 displays these scores based on dysarthria type. Table 3 also provides swallowing scores across dysarthria types. Participants with dysarthria had a median PAS of 2 (IQR 1, 3.25). Those with hypokinetic dysarthria had a median PAS of 2 (IQR 1, 2.25) as did those with spastic (2, IQR 1, 4) and hypokinetic-spastic dysarthria (2, IQR 1, 4). There were no significant differences for scores across dysarthria type (p > 0.05).

For MBSImP total scores, participants with dysarthria had a median OTSS of 7 (IQR 5, 9) and a median PTSS of 2 (IQR 1, 4.25). Examining dysarthria types, the hypokinetic and hypokinetic-spastic groups each had a median OTSS of 7 (IQR 5, 9 and 6.5, 9, respectively); similarly, the spastic group had a OTSS of 6 (IQR 4, 9). While no significant differences in OTSS were found across dysarthria types (p > 0.05), significant differences were found for PTSS across types (p < 0.01). Wilcoxon tests revealed significantly worse/higher median PTSS for participants with spastic compared to hypokinetic dysarthria (i.e., 4 (IQR 2.7, 5.7) and 2 (IQR 1, 3), respectively; p = 0.01); those with hypokinetic-spastic dysarthria showed a trend toward significantly worse median PTSS compared to those with hypokinetic dysarthria (i.e., 4 (IQR 2, 5) and 2 (IQR 1, 3), respectively; p = 0.05). There was no difference in PTSS for spastic compared to hypokinetic-spastic dysarthria (p > 0.05). Of interest was that the median PTSS was significantly higher for those with dysarthria plus AOS than those with dysarthria only (i.e., 4 (IQR 2.7, 7.0) and 2.0 (IQR 1, 4), respectively; p = 0.01). There were no significant differences between groups for other swallowing variables (p > 0.05).

Individuals with dysarthria had a median FOIS score of 7 (IQR 6, 7). These scores were consistent across dysarthria types, with minimal or no variation in IQRs and no significant differences among types (p > 0.05).

Linear Regression Analyses for Motor Speech and Swallowing

Table 4 summarizes relevant linear regression analyses for motor speech and swallowing variables, with significant multivariate relationships corresponding to those variables depicted in Supplementary Material B. While there was a significant multivariate linear relationship between disease severity, a control variable, and PAS (Estimate: 0.068; 95% CI: 0.021, 0.115; p = 0.005), no other statistically significant relationships were found between MSD type and severity and PAS (p > 0.05) after adjusting for age, education, gender, and disease severity.

Table 4.

Relevant linear regression analyses for motor speech and swallowing variables

Swallowing measures (dependent variables)	Independent variables (covariate/factor)	Estimate (lower & upper confidence intervals)	Multivariate p-value
Penetration-Aspiration Scale	Dysarthria type: hypokinetic vs hypokinetic-spastic	0.675 (−0.514, 1.865)	0.261
	Dysarthria type: hypokinetic vs spastic	0.556 (−0.898, 2.010)	0.448
	Dysarthria type: hypokinetic-spastic vs spastic	−0.120 (−1.790, 1.551)	0.886
	AOS present	0.103 (−1.121, 1.328)	0.866
	Motor Speech Disorder Severity Rating	−0.125 (−0.459, 0.208)	0.455
Variables controlled for	Age at exam	−0.009 (−0.079, 0.061)	0.798
	Gender	−0.850 (−1.811, 0.111)	0.082
	Education	0.107 (−0.069, 0.283)	0.228
	Disease severity (PSPRS)†	0.068 (0.021, 0.116)	0.005*
Oral Total Sum Score	Dysarthria type: hypokinetic vs hypokinetic-spastic	0.436 (−0.819, 2.054)	0.592
	Dysarthria type: hypokinetic vs spastic	−0.819 (−2.360, 1.595)	0.700
	Dysarthria type: hypokinetic-spastic vs spastic	−0.819 (−3.091, 1.453)	0.474
	AOS present	0.739 (−0.902, 2.380)	0.371
	Motor Speech Disorder Severity Rating	−0.272 (−0.718, 0.175)	0.228††
Variables controlled for	Age at exam	−0.030 (−0.124, 0.065)	0.535
	Gender	−0.549 (−1.845, 0.748)	0.401
	Education	−0.131 (−0.368, 0.105)	0.272
	Disease severity (PSPRS)	0.072 (0.008, 0.136)	0.027*
Pharyngeal Total Sum Score	Dysarthria type: hypokinetic vs hypokinetic-spastic	0.782 (−0.647, 2.211)	0.278
	Dysarthria type: hypokinetic vs spastic	1.534 (−0.212, 3.280)	0.084
	Dysarthria type: hypokinetic-spastic vs spastic	0.752 (−1.254, 2.758)	0.456
	AOS present	1.644 (0.209, 3.078)	0.025*
	Motor Speech Disorder Severity Rating	−0.532 (−0.917, −0.147)	0.008*
Variables controlled for	Age at exam	0.112 (0.026, 0.198)	0.011*
	Gender	−0.634 (−1.807, 0.540)	0.285
	Education	0.156 (−0.059, 0.370)	0.151
	Disease severity (PSPRS)	0.017 (−0.041, 0.074)	0.572
Functional Oral Intake Scale	Dysarthria type: hypokinetic vs hypokinetic-spastic	−0.090 (−0.594, 0.414)	0.723
	Dysarthria type: hypokinetic vs spastic	0.446 (−0.170, 1.062)	0.153
	Dysarthria type: hypokinetic-spastic vs spastic	0.535 (−0.172, 1.243)	0.135
	AOS present	0.378 (−0.136, 0.892)	0.147
	Motor Speech Disorder Severity Rating	0.010 (−0.133, 0.153)	0.890
Variables controlled for	Age at exam	−0.009 (−0.039, 0.021)	0.540
	Gender	0.192 (−0.218, 0.602)	0.353
	Education	−0.042 (−0.117, 0.032)	0.262
	Disease severity (PSPRS)	−0.023 (−0.043, −0.002)	0.030*

^†PSPRS=Progressive Supranuclear Palsy Severity Scale

^††While values reached significance (p<0.05) for the univariate analysis, they did not for the multivariate analysis

^*Significant values (p<0.05)

A significant univariate linear relationship was found between MSDSR and OTSS (p = 0.038) though not maintained after adjusting for age, education, gender, and disease severity. There was also a significant multivariate linear relationship between disease severity, a control variable, and OTSS (Estimate: 0.072; 95% CI: 0.008, 0.136; p = 0.027). No other statistically significant relationships were found between MSD type and severity and OTSS (p > 0.05).

After adjusting for age, education, gender, and disease severity, a significant multivariate linear relationship was found between MSDSR and PTSS (Estimate: −0.531; 95% CI: −0.917, −0.146; p = 0.008), and between presence of AOS and PTSS (Estimate: 1.64; 95% CI: 0.209, 3.078; p = 0.02). There was also a significant multivariate linear relationship between age and PTSS (Estimate: 1.643; 95% CI: 0.209, 3.078; p = 0.02). The significant univariate linear relationship found between the presence of spastic dysarthria and PTSS (p = 0.03) was not maintained after adjusting for confounding variables (p > 0.05).

While there was a significant multivariate linear relationship between disease severity, a control variable, and FOIS (Estimate: −0.022; 95% CI: −0.042, −0.002; p = 0.03), no other statistically significant relationships were found between MSD type and severity and FOIS (p > 0.05) after adjusting for confounding variables.

Secondary analyses showed no statistically significant differences between dysarthria types for age, education, gender, and disease severity/PSPRS or for disease duration (p > 0.05).

Discussion

Little is known about the co-occurrence of MSDs and dysphagia in PSP despite these impairments being among the presenting signs/symptoms of PSP or manifested with disease progression. The purpose of this study then, was to investigate whether differences in motor speech features influenced the presentation of dysphagia in PSP. To our knowledge, this is the first study to characterize MSD and dysphagia features simultaneously in a large PSP cohort.

Our findings revealed that most participants exhibited dysarthria, with hypokinetic, spastic, or a mixed hypokinetic-spastic dysarthria being represented, and nearly one-fifth of participants having co-occurring AOS. Our results are similar to previous findings about dysarthria types present in PSP (Collins et al., 1995; Kluin et al., 1993; Rusz et al., 2015; Skodda et al., 2011) yet differ on which features predominated (e.g., often spastic or mixed hypokinetic and spastic, whereas in our sample hypokinetic dysarthria predominated), which as noted by Clark et al. (2021) is likely due to presence of AOS in PSP variants not being accounted for. Also, earlier studies did not specify or vaguely described MSDs in PSP (Goetz et al., 2003; Golbe et al., 1988; Litvan et al., 1996a, 1996b,1997; Muller et al., 2001; Warnecke et al., 2010). Understanding and specifying such features has improved recently and should continue to be explicitly described as this has implications for disease evolution and management considerations.

Examining dysarthria severity in PSP revealed that, participants had to implement behavior modification with communication interactions. Those with hypokinetic dysarthria had to occasionally repeat information, particularly in adverse listening conditions, whereas those with spastic dysarthria faired significantly worse, relying on augmentative communication (i.e., writing, a spokesperson) with intelligibility problems. Age, education, and/or disease severity did not account for these differences in our study, lending support that the nature and impact of the speech changes can be attributed to the dysarthria type and associated underlying motor impairments. Similarly, worse/higher MSDSR scores were also found for AOS with co-occurring dysarthria (just over half had spastic dysarthria) compared dysarthria alone. This revealed a greater detriment to communication, limiting participants’ use of verbal output and increasing reliance on alternative communication methods. Considered together, the findings highlight the importance of differential diagnosis for treatment planning.

A previous study on MSDSR in ALS identified critical intervention periods corresponding to specific scores (Yorkston, Miller, Strand, & Britton, 2013) (e.g., scores of 7 or 8 signaling the need to implement environmental modifications to maximize communication interactions, and scores of 5 or 6 involving training individuals to use behavioral strategies such as change speaking rate and manage communication breakdowns) to enhance speech intelligibility. Since the MSDs accompanying PSP and ALS are distinct and the diseases have different progression rates, unique critical periods may correspond to speech severity in PSP that a longitudinal study could elucidate, along with potential implications for swallowing.

Previous studies show that oral phase impairments in PSP include abnormal tongue movements (back/forth rocking) that disrupt bolus containment, impaired clearance of residue, and slow and inadequate bolus manipulation and propulsion (Clark et al., 2019; Johnston et al., 1997; Leopold & Kagen, 1997; Litvan et al., 1997; Sonies, 1992; Warnecke et al., 2010). Pharyngeal impairments include velar function abnormalities, delayed swallow trigger, pharyngeal residue, and penetration and aspiration (Clark et al., 2019; Johnston et al., 1997; Leopold & Kagel, 1997; Litvan et al., 1997; Warnecke et al., 2010), with a recent study also showing more frequent and deeper airway invasion of thin liquids (based on the PAS) in PSP than PD (Borders et al., 2023). Further, videofluoroscopic data have shown more significantly impaired oral relative to pharyngeal phase findings in PSP (Clark et al., 2019; Johnston et al., 1997). This aligns with our results of participants with dysarthria having a OTSS of 7 and a PTSS of 2, suggesting worse oral compared to pharyngeal function. It was therefore unexpected that MSD severity was more closely related to PTSS than OTSS. Collectively, participants with dysarthria had a MSDSR of 6 and a PTSS of 2. Though this PTSS is within normal range based on recent findings in healthy adults (Garand et al., 2022), it was estimated to be 0.531 times higher/worse for each point that the MSDSR decreased/worsened. Future work could examine MSDSR and the specific MBSImP components that may contribute to OTSS and PTSS.

Our results also showed a relationship between spastic features and PTSS, though not statistically significant when controlling for confounding variables. Moreover, we found significantly worse pharyngeal scores for spastic and mixed spastic-hypokinetic compared to hypokinetic dysarthria types. Future studies could explore potential associations between specific dysarthria features (e.g., strained-strangled voice) and pharyngeal impairments (e.g., laryngeal vestibule closure), recognizing some features would likely co-occur (e.g., hypernasality and impaired soft palate elevation).

Also unexpected was AOS having a stronger relationship to PTSS than OTSS. PTSS was estimated to be 1.643 times higher/worse when AOS occurred. This finding is not easily explained. Although nonverbal oral apraxia (NVOA) may occur in up to 60% of individuals with AOS having neurodegenerative disease, by definition NVOA impacts volitional movements, leaving vegetative and reflexive movements intact. Thus, if present, NVOA should be associated with more volitional oral phase than patterned pharyngeal phase responses. Further surprising is that, anecdotally, most of our participants who exhibited AOS as their primary speech impairment at onset did not develop dysphagia until after developing dysarthria. Unfortunately, because this study did not include AOS participants without dysarthria, we could not address the nature of that relationship.

Further unexpected was finding no significant relationships between MSDSR and PAS or FOIS scores. PAS scores for participants with dysarthria were normal, consistent with Garand et al. (2019) (median PAS=1 (range 1–5) for healthy adults age 60 and over), Robbins et al. (1999) (PAS ≤ 3) and Steele and Grace-Martin (2017) (PAS 1, 2, 4) data, and did not vary by dysarthria type. Similarly, FOIS scores were normal (7) for all dysarthria types, likely reflecting this sample of PAS and FOIS scores being too homogeneous to reveal significant relationships with MSD severity. Relatedly, there were significant differences in MSDSR scores for those with versus without dysarthria, though not when examining swallowing indices. Potential contributing factors include the few participants without dysarthria not yielding enough power to make that distinction, dysphagia needing to be worse than it was for significant group differences to manifest, or some participants having had abnormal motor speech versus swallowing aspects that were difficult to discriminate. Future studies following dysphagia and dysarthria progression may reveal relationships not detected in this cross-sectional sample.

Further, we expected age, gender, education, and disease severity to have confounding effects on PSP variables of interest. Indeed, disease severity was a significant predictor variable of PAS, OTSS, and FOIS scores. Increased difficulties with oral abilities, airway protection, and safely tolerating oral intake was indicative of worsening underlying etiology. Thus, swallowing outcomes are expected to worsen as a sign of disease progression. Age was also a significant predictor of PTSS. As age increased, the severity of pharyngeal impairments worsened. This is consistent with Whitwell et al.’s³³ findings that some with PSP present with speech-language deficits, with older age seeming to be a predictive factor in this variant. Shorter survival rates were found for participants demonstrating dysphagia early in the disease (dell’Aquila et al., 2013; Litvan et al., 1996b), with that sample at disease onset being older with early cognitive deficits (dell’Aquila et al., 2013). While our results are consistent with recent normative data for age-related trends in MBSImP scores for healthy adults age 60 and older (i.e., higher average OTSS and PTSS: 5.5 and 6, respectively) than their younger counterparts (Garand et al., 2022), our current findings, similar to previous studies, suggest individuals with PSP presenting with greater disease severity and increased age would benefit from more frequent monitoring of swallowing.

Clinically, an important implication of our study is that changes in both motor speech and swallowing should be evaluated and monitored over time in PSP. This is because both functions have the same underlying etiology and given that the greater the MSD severity, the greater the likelihood of dysphagia, potentially more severe at times irrespective of disease severity. Another contribution is that when individuals with PSP present with more severe MSDs, especially spastic dysarthria and when AOS co-occurs, clinicians are to provide counseling and intervention emphasizing augmentative compensation, preparing patients/families/caretakers for progressive speech changes. Finally, introducing compensatory swallowing strategies early on and modifying recommendations as dysphagia progresses are essential. This would include counseling about costs/benefits of oral/nonoral intake options given the most common cause of death in PSP being pneumonia (e.g., Litvan et al.,1996b), a symptom of the underlying etiology.

Comprehensive evaluation by neurologists specializing in behavioral and movement disorders and speech-language pathology consultation are imperative in PSP standard of care for differential diagnosis and management decisions given expected progression of MSDs and dysphagia. Being competent in thoroughly examining cranial nerves corresponding to speech and swallowing and evaluating motor speech functions is crucial for clinicians working with patients who have neurologically-based communication and swallowing disorders. Focusing on swallowing at the exclusion of motor speech, or vice versa, is unacceptable given that standard of care calls for assessing both. Moreover, clinicians are to make the connection that when motor speech and swallowing are affected in an individual, impairments often stem from the same underlying cause, with patterns of functional impairment revealing clues about underlying neuropathophysiology (e.g., hypokinetic speech and swallowing features reflecting reduced range of movement and rigidity due to basal ganglia control circuit impairments). Thus, examining both motor speech and swallowing functions is essential for gaining insights about nervous system organization and differential diagnosis.

Study limitations include the following. Our data were taken from initial participant work-up; thus, impairments may not reflect a representative scope of disease progression. There was also some overlap between these participants and those in the Clark et al. (2021) study. Further, we acknowledge the possibility that using the worst MBSImP and PAS scores may present a bias for overidentifying impairments inherent in the applications of these measures. As our participants tended to have relatively mild impairments, this likely did not influence the interpretation of our findings. Also, three participants with ataxic features were excluded as a few participants per group would not have yielded enough power to run the analyses. Ideally, future studies will include sufficient representation of all dysarthria profiles observed in PSP and go beyond our exploratory efforts to determine what statistical power would be needed.

Conclusion

Findings from this large cohort study provide insights about the relationship between MSDs and swallowing function in PSP. Greater MSD severity was associated with more severe pharyngeal phase impairments. Though motor speech and swallowing scores varied minimally across participants, incremental changes in these functions co-occurred and were more likely when certain MSD features were present. A trend for participants with spastic dysarthria and/or AOS to exhibit more severe dysphagia was observed. This study supports including speech-language pathology consultation as part of the neurological work-up in the standard of care for patients with PSP. Understanding salient impairments in motor speech and swallowing functions can facilitate with differential diagnosis, provide insights about clinical practice and theoretical models relevant to underlying disease processes, and inform quality patient-centered intervention for PSP. Further research is needed on MSDs and dysphagia in PSP to reveal additional factors that influence their co-occurrence and greater insights about assessment and intervention considerations.