Prevalence of Speech and Swallowing Dysfunction and Intervention Among Individuals With Myasthenia Gravis

Authors: Tyler J. Gallagher, Chloe Santa Maria, Michael M. Johns

Abstract

Objective

To examine the prevalence of dysphonia and dysphagia among individuals with myasthenia gravis (MG) and the frequency of utilization of speech and swallow evaluation and therapy.

Study Design

Retrospective cohort study.

Setting

TriNetX US Collaborative Network.

Methods

The cohort includes adults (≥18 years) with and without MG diagnosis. Propensity score matching was used before generating odds ratios (OR) of receiving speech and swallow diagnoses including dysarthria, dysphonia, dysphagia, and aspiration pneumonia. Subsequently, rates of individuals with MG receiving voice and swallow evaluation and therapy were analyzed.

Results

In this cohort, the prevalence of voice, speech, and swallowing pathology was higher in those with MG than without, including aspiration pneumonia (3.7% vs 0.7%), dysarthria (9.2% vs 0.5%), dysphonia (4.4% vs 1.2%), and dysphagia (24.9% vs 4.7%). ORs for diagnosis of each type of pathology were higher after MG diagnosis, including for aspiration pneumonia (OR: 2.69 [95% confidence interval, CI: 1.95‐3.70]), dysarthria (OR: 9.28 [95% CI: 6.67‐12.92]), dysphonia (OR: 2.63 [95% CI: 1.97‐3.53]), and dysphagia (OR: 4.18 [95% CI: 3.62‐4.83]). Finally, the rate of receiving a swallow and/or voice evaluation and/or therapy was low among all individuals with MG (16.9%) but higher among those with MG with any speech or swallow symptom (47.0%), MG with dysarthria and/or dysphonia (56.2%), and MG with dysphagia and/or aspiration pneumonia (49.0%).

Conclusion

Rates of dysarthria, dysphonia, dysphagia, and aspiration pneumonia were far more common among those with MG than without, but formal speech and/or swallow evaluation was not routine. These findings suggest a need for more routine speech‐language pathology evaluation among individuals with MG.

Myasthenia gravis (MG) is a relatively rare condition for which the prevalence is estimated to be between 1.5 and 17.9 per 100,000 individuals worldwide, ^1^ including approximately 80,000 individuals in the United States. ^2^ MG is a type of neuromuscular junction disorder, which comprises a group of disorders that ultimately lead to muscle weakness. In the case of MG, muscular weakness is caused by antibodies that attack the postsynaptic membrane, thus impairing neuromuscular transmission. ^3^ The mainstay of treatment is a combination of immunotherapy and acetylcholinesterase. ^4^

Historically, MG has been primarily treated by neurologists. While the pathogenesis is clearly neurologic, MG has a variety of otolaryngologic manifestations. The most common among these otolaryngologic symptoms include dysarthria and dysphagia. ^5^ , ^6^ While rare, there are many reports of individuals with MG presenting with exclusively otolaryngologic complaints. ^7^ , ^8^ , ^9^ , ^10^ The variable presentation of MG is likely due to its variable immunologic pathenogenic origins, which has caused MG cases to be divided into subgroups based on immunologic and clinical profiles. ^11^

Although speech and swallow symptoms are common in MG, knowledge of the prevalence of these symptoms, particularly in comparison to individuals without MG, is limited, and results have been inconsistent. Furthermore, among those with such symptoms, it is unknown how frequently such symptoms are evaluated or addressed outside of general neurologic treatment for MG, such as through speech and swallow evaluation or treatment. Thus, we seek to understand the prevalence and relative frequency of dysphonia, dysarthria, dysphagia, and aspiration pneumonia in individuals with MG. Further, among individuals with MG, we seek to explore how frequently individuals receive formal speech and swallow evaluation and therapy. We hypothesize that individuals with MG will be at much higher risk of experiencing dysarthria and dysphagia, though formal diagnosis, evaluation, and treatment of the otolaryngological components of the disease may be infrequent.

Methods

This study received exempt approval from the University of Southern California Institutional Review Board (UP‐23‐00652).

Study Population

This was a retrospective cohort study utilizing TriNetX, a globally federated health research network providing access to consistently updated deidentified electronic medical records. ^12^ Thus, TriNetX serves as a unifying organization of electronic medical record data for clinical trial recruitment information and secondary research. We utilized the US Collaborative network, which provided access to deidentified patient data from 64 health care organizations, accounting for greater than 110 million individuals. The cohort is primarily composed of tertiary care centers within the United States. TriNetX, LLC is compliant with the Health Insurance Portability and Accountability Act, as information is supplied only in aggregated counts of 10 or greater with deidentified information only.

Two primary cohorts were created for analysis. Each cohort contained only individuals who were ≥18 years of age. Cohort 1 included individuals who were diagnosed with MG (International Classification of Disease‐10 [ICD‐10]: G70.0) and had a documented virtual visit. Cohort 2 was defined by individuals who had a documented virtual visit and have not been diagnosed with MG. Virtual visits were defined as the first time an individual had a medical appointment via telecommunication (eg, via phone or video conference). Virtual visit was included in both cohorts for multiple reasons, including (1) to preserve similarity between the patient populations, (2) to provide an index event for subsequent comparison in propensity score matching, and (3) to maximize the number of individuals in each group while allowing for propensity score matching on multiple relevant variables. For additional analysis, 4 more groups were created, including (1) all individuals with MG, (2) individuals with MG and aspiration pneumonia, dysarthria, dysphonia, and/or dysphagia, (3) individuals with MG and dysarthria and/or dysphonia, and (4) individuals with MG and dysphagia and/or aspiration pneumonia. These 4 groups did not include virtual visits as a filter, as each group already had an index event (diagnosis of MG), but did include other age and time period criteria. Each group was defined by diagnosis or hospital visit in the time frame of June 6, 2004 to June 6, 2024. Analysis was completed on June 6, 2024.

Outcomes of Interest

Two sets of outcomes were of interest. First, the prevalence of dysarthria, dysphonia, dysphagia, and aspiration pneumonia were analyzed and compared between the 2 cohorts. Second, the utilization of speech‐language pathology services, including voice evaluation, swallow evaluation, voice therapy, and swallow therapy, was analyzed among the cohort with diagnosis of MG. ICD‐10, Current Procedural Terminology (CPT), and Healthcare Common Procedure Coding System (HCPCS) codes utilized for the outcomes listed above were informed through a thorough literature review and are documented in Supplemental Table S1, available online. ^13^ , ^14^ , ^15^ ICD‐10, CPT, and HCPCS codes were selected on the basis of their potential role as a component of the disease state of MG as well as their specificity for the symptom characterized. Most codes utilized for speech and swallow evaluation and therapy were specific to speech‐language pathology. TriNetX utilizes ICD‐9 to ICD‐10 mapping when necessary for codes before ICD‐10 implementation.

Statistical Analysis

TriNetX advanced analytics were utilized for all analyses. First, baseline demographics of the cohorts were collected. Prevalence of speech and swallow dysfunction in the cohort were then collected. Then, propensity score matching was performed for age, sex, race, ethnicity, and gastroesophageal reflux disease (ICD‐10: K21; included due to association with dysphagia and other laryngeal symptoms), and the odds ratios (ORs) of developing dysarthria, dysphonia, dysphagia, or aspiration pneumonia was compared between a cohort with MG and a cohort without. Propensity score matching is a statistical method used to control for potential confounding factors between groups by selecting individuals from each cohort to create comparable groups. This method ensures that the groups have similar rates of confounding variables, with the primary difference being the independent variable of interest (MG, here). ^16^ Standardized differences before and after matching were calculated to demonstrate adequacy of matching. Finally, the rate at which a variety of groups received voice and/or swallow evaluation and therapy was examined. The groups compared included (1) all individuals with MG, (2) individuals with MG and any speech or swallow symptom, (3) individuals with MG and dysarthria and/or dysphonia, and (4) individuals with MG and dysphagia and/or aspiration pneumonia.

Results

Characteristics of the primary cohorts utilized for comparison are described in Table 1. The cohort with MG includes 3724 individuals, and the cohort without MG contains 2,261,194 individuals. After propensity score matching, each group contained 3724 individuals. Standardized differences between each characteristic of the cohorts (including current age, age at index event, sex, race, ethnicity, and history of gastro‐esophageal reflux disease) were less than 0.02 for each comparison.

The prevalence of various diagnoses within each cohort is described in Figure 1. For each symptomatic diagnosis (eg, dysphagia and dysarthria), there was a higher rate of diagnosis in the MG cohort compared to the cohort without a diagnosis of MG. In the MG cohort, 3.7% (n = 139) had a diagnosis of aspiration pneumonia compared to 0.7% (15,334) in the no MG cohort. 9.2% (n = 341) in the MG cohort had dysarthria compared to 0.5% (n = 10,305) in the no MG cohort. Dysphonia was also more common in the MG cohort at 4.4% (n = 164) compared to the no MG cohort with 1.2% (n = 27,015). Finally, dysphagia was more common in the MG cohort (24.9%; n = 926) compared to the cohort without MG (4.7%; n = 105,066).

To compare the ORs for being given various speech and swallow dysfunction diagnoses after diagnosis of MG, ORs were calculated after propensity score matching between these groups (Figure 2). First, individuals were more likely to receive a diagnosis of aspiration pneumonia after MG diagnosis, with an odds ratio (OR) of 2.69 [95% confidence interval, CI: 1.95‐3.70] and a risk difference of 2.31% [95% CI: 1.59%‐3.03%] between groups. OR was higher for diagnosis of dysarthria after MG at an OR of 9.28 [95% CI: 6.67‐12.92] and risk difference of 8.08% [95% CI: 7.10%‐9.07%]. OR of dysphonia was also significant at 2.63 [95% CI: 1.97‐3.53] and a risk difference of 2.69% [95% CI: 1.91%‐3.47%]. Finally, the odds of a diagnosis of dysphagia were increased among those with MG at an OR of 4.18 [95% CI: 3.62‐4.83] and a risk difference of 17.54% [95% CI: 15.91%‐19.16%].

Finally, the relative frequency of utilization of voice evaluation, swallow evaluation, and voice therapy were analyzed for all individuals with MG (n = 65,310), a group with MG and any described speech or swallow issue (n = 18,701), a group with MG and dysarthria and/or dysphonia (n = 7072), and a group with MG and dysphagia and/or aspiration pneumonia (n = 16,947; Figure 3). In the cohort of all individuals with MG, 83.1% received no swallow or voice evaluation or treatment, 3.6% had voice evaluation, 12.6% had swallow evaluation, 2.0% had voice therapy, and 5.6% had swallow therapy. In a cohort with MG and any speech or swallow dysfunction, 53.0% had no evaluation or therapy, 8.7% had voice evaluation, 33.4% had swallow evaluation, 5.8% had voice therapy, and 16.2% had swallow therapy. Among those with MG and dysarthria and/or dysphonia, 43.8% had no evaluation or therapy, 15.2% had voice evaluation, 35.6% had swallow evaluation, 10.1% had voice therapy, and 19.0% had swallow therapy. Finally, among those with MG and dysphagia and/or aspiration pneumonia, 51.0% had no evaluation or therapy, 8.4% had voice evaluation, 35.5% had swallow evaluation, 5.5% had voice therapy, and 17.4% had swallow therapy.

Discussion

Our study leveraged a large, national database to explore the frequency and relative odds of having dysphonia, dysarthria, dysphagia, and aspiration pneumonia in a sample of adult individuals with MG. We hypothesized that individuals with MG would have greater odds of speech and swallow dysfunction though formal diagnosis, evaluation, and treatment may be low. In this study, we found far greater odds of such diagnoses among individuals with MG, including aspiration pneumonia, dysarthria, dysphonia, and dysphagia. Despite these higher odds, rates of evaluation and treatment for these symptoms were relatively low, even among those diagnosed with 1 or more types of speech or swallow dysfunction, likely related to their neuromuscular disease.

While several studies have demonstrated cases of isolated speech or swallow symptoms among MG patients, ^17^ , ^18^ , ^19^ , ^20^ the overall rates of speech and swallow dysfunction among all individuals with MG have been less frequently reported. Studies assessing the electrophysiological characteristics of individuals with MG have consistently found significant impairments in phonation and swallowing function among MG patients compared to healthy individuals. ^21^ , ^22^ Previous estimates of rates of reported swallowing and voice difficulties have ranged from 20% to 40% or greater, depending on the definition of these symptoms. ^23^ , ^24^ , ^25^ Rates of speech and swallow diagnoses among individuals diagnosed with MG are relatively similar in our study, as almost 25% were diagnosed with dysphagia and 9% with dysarthria. Furthermore, we quantified the increased odds and increased percent chance of having such symptoms compared to the general population after propensity score matching, demonstrating strikingly increased odds of dysarthria (OR: 9.28; risk 8.08%) and dysphagia (OR: 4.18; risk difference 17.54%). Despite the high rates of these diagnoses, these may still be underreported, as objective studies have demonstrated that even among those without reports of dysphagia, some level of swallowing dysfunction may be present. ^26^ , ^27^

To our knowledge, previous studies have not explored the rate of formal assessment and isolated treatment for voice and swallowing dysfunction in this population, making this a novel component of our study. Overall, we found very low rates of formal evaluation and treatment for speech and swallow. There is potential for underestimation of such rates due to our methods, which only include formal assessment with a clear purpose of evaluation, leaving out some codes that otolaryngologists may use for speech and swallow evaluation as well as other reasons (eg, laryngoscopy). Thus, formal assessment and treatment of voice and swallowing analyzed in this study is primarily via speech‐language pathologist investigation. Though evaluation and treatment were higher among groups diagnosed with speech and swallow disorders, rates of no evaluation or treatment remained above 40% in these groups. Interestingly, even among those diagnosed with dysphagia and/or aspiration pneumonia, rates of swallow therapy were low (17.4%) demonstrating low utilization of this modality among individuals with MG and swallowing dysfunction.

Given the significant potential for impaired quality of life and complications related to speech and swallow dysfunction in this population, coupled with the potential underreporting of such pathology, it is reasonable that speech‐language pathology examination should be a routine part of care for individuals with MG. We suggest this should include evaluation of voice with a laryngoscope when symptomatic with consideration of routine swallow studies given the relatively high rates of silent aspiration in this population. ^28^ Current guidelines on management of MG do not discuss the potential of physical therapy or speech‐language pathology in the treatment of the condition, in part due to a lack of evidence to support physical therapy for this condition. ^29^ , ^30^ Despite little research in the area, a recent study demonstrated that instructed chin positioning can improve pharyngeal clearance in individuals with MG. ^31^ Furthermore, a survey study of individuals with MG demonstrated that while most were unaware of speech‐language pathologist services, many were interested in exploring the evaluation and treatment of their speech and swallow problems in this setting. ^32^ Together, this suggests significant underutilization of a potentially important resource for individuals with MG.

Individuals diagnosed with MG require early evaluation by a wide variety of providers, potentially including neurology, immunology, ophthalmology, and pulmonology, we believe that early speech and swallow evaluation is also important. Dysarthria and dysphagia both have significant impacts on social and psychological quality of life, ^33^ , ^34^ potentially contributing significantly to the burden of disease caused by MG. Furthermore, given that dysphagia can result in aspiration and pulmonary compromise (the primary cause of myasthenic crisis ^35^ ), early identification of these symptoms of MG is critical. Early identification can allow for early interventions, including attempting swallowing rehabilitation, dietary modifications, and medication adjustment as necessary to prevent unnecessary complications.

While this study contains a variety of novel findings, there are notable limitations. Although a thorough literature review was utilized to locate all relevant medical coding, this study is limited by the use of medical record coding, which may be incomplete and inconsistent. Furthermore, only procedures completed at institutions participating in TriNetX are included in the analysis, so evaluation and intervention may be underreported in this study. It is important to consider the results of this manuscript as diagnosed pathology rather than rates of symptoms among this cohort, as actual rates of symptoms are likely to be higher. While the high power of this study allows to the comparison of rates of speech and swallow dysfunction, the cohort is primarily composed of tertiary care centers in the United States and may not be applicable to other populations. While we adjusted for a variety of potential confounders in calculations of ORs, we cannot rule out additional confounding. Future studies should seek to prospectively assess the rate of speech and swallow dysfunction among individuals with MG, including among those who do and do not report such symptoms, in order to understand the need for speech and swallow evaluation in this population.

Conclusion

Speech and swallow symptoms, including dysarthria, dysphonia, dysphagia, and aspiration pneumonia, were frequently diagnosed in individuals with MG, and the odds of diagnosis were far greater compared to a group without MG diagnosis. However, formal speech and/or swallow evaluation was not routine among this group, even when considering individuals diagnosed with those symptoms. These findings suggest a need for more routine speech‐language pathology evaluation among individuals with MG.

Author Contributions

Tyler J. Gallagher, conception and design of work, data acquisition, data analysis, interpretation of data, drafting of manuscript, critical revision; Chloe Santa Maria, conception and design of work, data analysis, interpretation of data, drafting of manuscript, critical revision; Michael M. Johns III, conception and design of work, data acquisition, data analysis, interpretation of data, drafting of manuscript, critical revision.

Disclosures

Competing interests

None.

Funding source

None.

Supporting information