Customize your JAMA Network experience by selecting one or more topics from the list below.
Bertelsen C, Zhou S, Hapner ER, Johns MM. Sociodemographic Characteristics and Treatment Response Among Aging Adults With Voice Disorders in the United States. JAMA Otolaryngol Head Neck Surg. 2018;144(8):719–726. doi:10.1001/jamaoto.2018.0980
What are the sociodemographic characteristics and treatment characteristics and response among adults 65 years or older with voice disorders?
In this cross-sectional database study, 4.2 million of 41.7 million adults 65 years or older (10.1%) reported voice disorders; of these, only a small percentage sought treatment, and a minority of individuals were evaluated by either an otolaryngologist or a speech language pathologist. Adults who seek treatment are significantly more likely to report improvement in their symptoms; however, fewer adults 65 years or older who were treated for a voice disorder improved relative to adults younger than 65 years.
Many older adults with dysphonia are not treated by the appropriate health care practitioners; outreach efforts should be directed at increasing access to, and awareness of, services available from otolaryngologists and speech language pathologists.
Aging adults face unique barriers to care and have unique health care needs with a high prevalence of chronic conditions. A high proportion of individuals in this group have voice disorders, in part due to age-related changes in laryngeal anatomy and physiologic features. These disorders contribute significantly to health care costs and remain poorly understood.
To describe sociodemographic characteristics and response to treatment among aging adults with voice disorders.
Design, Setting, and Participants
A cross-sectional study using the 2012 National Health Interview Survey was used to evaluate adults who reported voice disorders in the past 12 months. Self-reported demographics and data regarding health care visits for voice disorders were analyzed. Statistical analysis was conducted from March 1, 2017, to February 1, 2018.
Main Outcomes and Measures
Self-reported voice disorders, whether or not treatment was sought, which types of professionals were seen for treatment, and whether or not the voice disorder improved after treatment.
Among 41.7 million adults in the United States 65 years or older, 4.20 million (10.1%; 2 683 199 women and 1 514 909 men; mean [SE] age, 74.5 [0.3] years) reported having voice disorders. Of those with voice disorders, 10.0% (95% CI, 8.3%-11.7%) sought treatment. Of individuals seeking treatment, 22.1% (95% CI, 7.9%-36.3%) saw an otolaryngologist and 24.3% (95% CI, 10.6%-38.0%) saw a speech language pathologist. By controlling for race/ethnicity, income, sex, and geography, it was found that men were less likely than women to report voice disorders (36.1% [95% CI, 31.7%-40.5%] vs 63.9% [95% CI, 59.5%-68.3%]; odds ratio, 0.70; 95% CI, 0.57-0.86). Race/ethnicity, income, and geography were not significantly associated with the likelihood that an individual 65 years or older reported voice disorders. A greater percentage of elderly adults seeking treatment than not seeking treatment reported improvement in symptoms (32.4%; 95% CI, 17.9%-47.0% vs 15.6%; 95% CI, 10.4%-20.8%). Among adults treated for a voice disorder, a lower proportion of adults 65 years or older reported improvement in symptoms with treatment compared with adults younger than 65 years (32.4%; 95% CI, 17.9%-47.0% vs 56.0%; 95% CI, 42.5%-69.6%).
Conclusions and Relevance
A small percentage of older adults with voice disorders seek treatment; even fewer are treated by an otolaryngologist or a speech language pathologist. A greater percentage of those who undergo treatment experienced symptomatic improvement compared with those who did not undergo treatment. These trends highlight the need for greater access to and awareness of services available to older adults with voice disorders.
The population of US adults older than 65 years is rapidly increasing and, by 2060, will comprise 25% of the US population.1 Given this trend, it is of paramount importance to understand and meet the health care needs of elderly adults. Aging adults have complex health care requirements owing to a high prevalence of chronic conditions.2 Furthermore, this population faces unique barriers to health care, including costs of care, difficulties with transportation, a lower educational level than the population as a whole, and the perception that physicians lack responsiveness to their concerns.3
Voice disorders are one of many chronic conditions that affect a significant proportion of elderly adults.4,5 However, dysphonia in this population is poorly understood. The larynx undergoes several age-related physiologic changes including laryngeal muscle atrophy,6 disorganization of collagen fibrils,7 and decreased synthesis of hyaluronic acid and other extracellular matrix components.8 These changes often result in vocal fold bowing, a spindle shaped glottic gap, and constriction of the laryngeal vestibule,6 all of which may contribute to dysphonia but may be targets for treatment of voice disorders in this age group.
Given the changes to the phonatory system that occur with age, it follows that possible causes and treatment of voice disorders in the elderly differ from those in the general population. These differences foster a general lack of understanding in the aging population regarding the nature of voice disorders and treatment options; this problem is exacerbated by the various barriers to care faced by elderly patients. Studies have shown that between 25% and 50% of patients with dysphonia believe that their voice disorder is a normal part of aging.5,9 Although it is true to a degree that voice disorders can be a normal part of aging owing to the physiologic changes mentioned previously, it is striking that as many as 50% of patients with a voice disorder or swallowing disorder are unaware that treatment options exist.5 These findings are not benign considering that voice disorders can significantly affect quality of life5,10 and that costs of health care associated with dysphonia increase with age.11
In light of this evidence, continuing to elucidate the features of voice disorders in elderly adults is critical. Sociodemographic variables are likely associated with patterns of seeking care among elderly adults, as well as their subjective experience of health care. Possessing a better understanding of these issues informs and enhances our ability to provide effective and compassionate care to these patients. The objective of this study is to describe the socioeconomic characteristics of elderly adults with voice disorders and to describe their responses to treatment.
This study is a cross-sectional study using data on adults collected from the 2012 National Health Interview Survey (NHIS) from the Minnesota Population Center and State Health Access Data Assistance Center’s Integrated Health Interview Series. The data from the NHIS include information about the health of the civilian noninstitutionalized US population and were taken from 108 131 survey respondents.12 Because the data were already deidentified and are publicly available, the Institutional Review Board of the Keck School of Medicine of the University of Southern California waived the need for approval of this study.
The variable “Voice problem past 12 months” from the voice, speech, and language supplement of the NHIS was used to identify a subset of survey participants who have voice disorders. The study population was identified from survey participants 18 years or older who answered “yes” to this question and were 65 years old or older. The main independent variable was age. The main outcome variable was a self-reported voice disorder. Data on whether an individual received any treatment (medical, surgical, or behavioral) for the voice disorder and whether their voice disorder improved after treatment were also collected. There were 23 survey respondents who were excluded because their responses to this question were unknown. The data were divided among 7359 participants 65 years or older and 27 121 adults younger than 65 years whose responses to this question were known.
Socioeconomic variables were also collected. “Self-reported race” was joined with “Hispanic ethnicity, dichotomous” to create the categories of black, Hispanic, non-Hispanic white, and other. The “other” group included Asian, American Indian or Alaskan Native, and multiracial individuals.
Age and educational level were grouped into categories as represented in Table 1. Employment status was grouped into categories of “working” (working with or without pay at a job or business) and “nonworking” (participants with a job or business but who were not at work, looking for work, or not working or looking for work). Individuals with both public and private insurance were categorized as having private insurance. Finally, poverty levels were defined using the NHIS income variable, categorizing respondents as being at less than 100% of the federal poverty level (FPL), 100% to 199% of the FPL, 200% to 399% of the FPL, and 400% or more of the FPL. The poverty variable contained a significant degree of missing data. Thus, the Integrated Health Interview Series provides multiply imputed poverty variables to account for the missing data. In total, there are 5 multiply imputed poverty variables that were joined to correctly assess statistical error when analyzing the data.
The complex, multistage probability sampling used by the NHIS incorporates stratification, clustering, and oversampling of some racial subpopulations (black, Hispanic, and Asian). To account for this complex sample design, the NHIS provides sampling weights to produce representative estimates of the total US population. The sampling weights are created based on the survey participants’ demographic information. The final weighted data set that is analyzed is the result of applying sample weights to the unweighted data. This sample weighting adjusts the survey data to represent the population from which the survey participants are drawn.
Statistical analysis was conducted from March 1, 2017, to February 1, 2018. The annualized data were then imported into R, version 3.3 (http://www.r-project.org), on the RStudio, version 1.0.136, development environment using the publicly available SAScii package, version 1.0, and analyzed using the survey package, version 3.32-1. The multiply imputed files were analyzed using the mitools package, version 2.3. The absolute number of NHIS respondents 65 years or older was extrapolated to reflect the size of the US population. The survey package was used to create a survey object based on the primary sampling unit, stratum (for estimating variance), and sample person weights, which resulted in a total weighted sample of 41.7 million US adults older than 65 years. Of these, 4.20 million (95% CI, 3.81-4.59 million) (10.1%; 95% CI, 9.3%-10.9%) reported having voice disorders. This number represented the final sample size used in the study for individuals who experienced the primary outcome of interest.
Multivariate logistic regression (syyglm function) was used to analyze the population and its subgroups. This analysis included a subset of the demographic variables to avoid overfitting the data; it also allowed calculation of odds ratios that better reflect the overall population from which the survey responses were drawn while still taking into account possible sociodemographic confounders. The svymean function in the survey package was used to perform 95% CI calculations. The micombine function from the package mitools was used to pool multiply imputed regression results. Results were considered significant if the 95% CIs did not overlap or if the 95% CIs of odds ratios did not overlap with 1.
Of adults 65 years or older reporting voice disorders, 63.9% (95% CI, 59.5%-68.3%) were female and 36.1% (95% CI, 31.7%-40.5%) were male (2 683 199 women and 1 514 909 men; mean [SE] age, 74.5 [0.3] years). Significant differences were observed between the numbers of adults 65 years or older reporting voice disorders in the last year and those who did not report voice disorders when separated by sex, region, and income level. Our study found that a significantly greater percentage of those with voice disorders who were 65 years or older were female and from the Midwest. Table 1 summarizes the sociodemographic characteristics of adults 65 years or older who reported voice disorders.
Table 2 displays the odds ratios of the factors influencing whether an adult 65 years or older reported voice disorders as obtained from multivariate regression analysis. In this model, men were significantly less likely to report voice disorders than were women (odds ratio, 0.70; 95% CI, 0.57-0.86). Adults 65 years or older in the Midwest were significantly more likely to report voice disorders than were those in other regions (odds ratio, 1.51; 95% CI 1.08-2.11). Race/ethnicity, insurance status, and income were not associated with the probability that an individual 65 years or older would report voice disorders.
Table 3 displays a comparison between adults younger than 65 years and adults 65 years or older with voice disorders. Significant differences were observed between the numbers of adults 65 years or older and the number of adults younger than 65 years who reported voice disorders when stratified by sex, income, educational level, employment status, and insurance status. Relative to adults younger than 65 years, a higher proportion of adults 65 years or older with voice disorders were white (83.0%; 95% CI, 79.8%-86.2% vs 70.2%; 95% CI, 68.0%-72.4%), and fewer adults 65 years or older with voice disorders were black (7.6%; 95% CI, 5.4%-9.8% vs 12.4%; 95% CI, 10.8%-14.0%) or Hispanic (5.6%; 95% CI, 3.7%-7.5% vs 11.4%; 95% CI, 9.8%-13.0%). Relative to adults younger than 65 years, there were fewer adults 65 years or older in both the lowest (under the FPL) and highest (>400% of FPL) income strata. There were also fewer adults 65 years or older with a college degree and more with less than a high school graduate education. A higher proportion of adults 65 years or older with voice disorders were not currently employed compared with adults younger than 65 years with voice disorders (86.7%; 95% CI, 83.5%-89.9% vs 38.4%; 95% CI, 35.8%-41.0%). Finally, a higher percentage of adults younger than 65 years with voice disorders had private insurance compared with those 65 years or older (62.3%; 95% CI, 59.7%-64.9% vs 54.9%; 95% CI, 50.4%-59.4%), whereas more adults 65 years or older with voice disorders had public insurance. Of adults with voice disorders who had public insurance, 98.8% (95% CI, 97.5%-10.0%) had Medicare.
Table 4 depicts the distribution of comorbidities among adults with voice disorders, both for those 65 or older and for those younger than 65 years. Adults 65 years or older with voice disorders were significantly more likely to have received previous diagnoses of asthma, chronic bronchitis, or heart disease; a head or chest cold within the previous 2 weeks; rheumatologic disease; a history of sinus disease; or a previous stroke. All of these comorbidities were also more common among adults younger than 65 years who reported voice disorders than among those who did not. In addition, adults younger than 65 years with voice disorders were more likely to have received diagnoses of coronary artery disease, diabetes, and hypertension.
Of adults 65 years or older with voice disorders, only 10.0% (95% CI, 8.3%-11.7%) sought treatment. Of those who sought treatment, 24.3% (95% CI, 10.6%-38.0%) were treated by a speech language pathologist (SLP), 22.1% (95% CI, 7.9%-36.3%) by an otolaryngologist, 32.8% (95% CI, 15.4%-50.2%) by a primary care professional (PCP), and 28.7% (95% CI, 14.2%-43.2%) by another type of practitioner. Of adults younger than 65 years with voice disorders, 17.5% (95% CI, 7.2%-27.8%) were treated by an SLP, 24.8% (95% CI, 13.5%-36.1%) by an otolaryngologist, 37.1% (95% CI, 25.0%-49.2%) by a PCP, and 38.3% (95% CI, 25.0%-51.6%) by another type of practitioner. (The total of these percentages exceeds 100% owing to respondents who may have seen multiple professionals for their voice disorders.) These percentages did not differ between adults 65 years or older and adults younger than 65 years with voice disorders. A greater percentage of adults 65 years or older who sought treatment achieved symptomatic improvement compared with those who did not seek treatment (32.4%; 95% CI, 17.9%-46.9% vs 15.6%; 95% CI, 10.4%-20.8%). A lower proportion of adults 65 years or older who sought treatment reported improvement in symptoms with treatment compared with those younger than 65 years who sought treatment (32.4%; 95% CI, 17.9%-46.9% vs 56.0%; 95% CI, 42.5%-69.5%). However, these differences did not reach statistical significance at the 95% CI level. Comparisons of responses to treatment between adults 65 years or older and adults younger than 65 years are depicted in Table 5.
In this study, we describe population characteristics and responses to treatment for adults 65 years or older who are experiencing voice disorders. This study also highlights key similarities and differences between those 65 years or older and those younger than 65 years with voice disorders. Female sex and residence in the Midwest were associated with an increased probability that an adult 65 years or older would report voice disorders. This finding is consistent with work by others13,14 examining the demographic characteristics of all US adults with voice disorders. Sex is an established risk factor for having voice disorders in the population as a whole13 as well as in young adults.15 Although differences in laryngeal anatomy and phonatory physiologic features may explain why women report voice disorders at a higher rate than men, it is also possible that age-related changes in laryngeal physiologic features affect women differently than men. Regional variations in the prevalence of dysphonia among aging adults are likely multifactorial and may reflect regional variations in factors affecting vocal function, such as patterns of tobacco use16 and prevalence of various allergens.17 Further studies are needed to provide insight into the potential causal relationships between these factors.
Our study also found that adults 65 years or older with voice disorders were more likely to be white, have middle-range income, and have public insurance; they were less likely to have a college degree or to be employed. These differences are explained in part by characteristics of the population 65 years or older, specifically by the fact that those 65 years or older qualify for Medicare and thus have improved access to public health insurance. However, it is nevertheless important for clinicians who care for these patients to understand and take these differences into account. The lower educational level among this group is particularly relevant in light of evidence that lower educational level is associated with poorer health literacy and self-care.18 Other studies have found that elderly adults with voice disorders do not seek treatment owing to various misconceptions—beliefs that dysphonia is a normal part of aging,5,19 that no treatment options exist,5 and that voice disorders will resolve without any treatment.19 These findings accentuate the need for patient advocacy on the part of clinicians and caregivers to ensure that elderly patients express their voice-related concerns and are treated appropriately.
Adults 65 years or older with voice disorders were also more likely than those without voice disorders to have several different medical comorbidities, which is consistent with the findings of others.20-22 Although the characteristics of dysphonia related to some clinical entities, such as cerebrovascular accidents, have been relatively well studied,23 dysphonia related to other diseases is less well understood. Further studies evaluating possible causal relationships between medical comorbidities and voice disorders should be prioritized, especially in light of evidence24 that the overall burden of comorbid disease correlates with response to voice therapy.
The possible causes of voice disorders among elderly adults also likely play a role in the patterns of seeking, and responses to, treatment. Roy et al21 recently performed a database study describing the characteristics of adults 65 years or older who were seeking treatment for voice disorders. They reported that the prevalence of nonspecific dysphonia and vocal fold paralysis increased with age, whereas the prevalence of benign laryngeal pathologic conditions and laryngitis decreased with age. That study, however, only drew on information from the minority of elderly adults who sought treatment for their voice disorder. Our study complements the insights of this group by describing the entire population of adults 65 years or older with voice disorders, including those who do not seek treatment. More important, in our study, we found that only a very small percentage of those who reported voice disorders sought treatment. Furthermore, a minority of those who sought treatment were evaluated by either an otolaryngologist or an SLP. This finding may reflect a lack of availability of, or access to, otolaryngologists or SLPs. It also might be explained by a lack of awareness among both patients and PCPs regarding the services and treatment options available to adults with voice disorders. Given our finding that elderly adults with voice disorders generally have a lower educational level compared with younger adults with voice disorders, the issue of awareness about health care options is especially relevant to the elderly. This issue is compounded by the fact that PCPs frequently do not ask patients about voice disorders, which further decreases the number of patients who are referred and who receive an appropriate diagnosis.19 Furthermore, patient concerns about adequate insurance coverage may be an additional barrier to seeking care for voice disorders.25 Outreach efforts should be directed toward improving the understanding among patients and PCPs of the symptoms of voice disorders. Similarly, both patients and physicians should be educated regarding the range of voice care services available from otolaryngologists and SLPs.
Another important finding of our study was that a greater percentage of adults 65 years or older who sought treatment for voice disorders noticed an improvement in their symptoms than those who did not seek treatment. Although this difference did not reach statistical significance at the level of the 95% CI, this outcome likely results from the paucity of individuals who sought treatment, creating larger 95% CIs and increasing the likelihood of false-negative findings. Although there is overlap in the 95% CIs, the absolute percentages do represent clinically significant differences that warrant further study.
In addition, while a majority of adults younger than 65 years noted improvement in symptoms with treatment, the percentage of adults 65 years or older was smaller. The causes of this discrepancy are likely multifactorial. The lower educational attainment and lower health literacy among elderly adults, as well as a low use of specialty care as already described, likely explain part of the lower rate of improvement with treatment. Furthermore, as detailed by Roy et al,21 the possible causes of voice disorders in elderly adults differ from those in younger adults. Disorders found to be more common in elderly adults, such as unilateral vocal fold paralysis, are more likely to require subspecialty evaluation and may be optimally addressed by both an SLP and an otolaryngologist working together. This finding underscores the need for an early otolaryngologic evaluation in ensuring a prompt and accurate diagnosis for these patients, as well as initiating appropriate treatment plans.
Finally, we found that, among patients 65 years or older, a higher proportion of those who were treated than those who were untreated reported worsening of their voice disorder. This finding might be the result of a higher proportion of patients with progressive or recalcitrant disease in the group seeking treatment Another possibility is that voice disorders may not only be underrecognized and undertreated but may also be inappropriately treated. Because a minority of patients with voice disorders were treated by either an otolaryngologist or an SLP, it might be that management by nonspecialists was inadequate. It is also possible that management by specialists was inadequate or that subjectively negative experiences influenced patient-reported results.
The limitations of this study are as follows. First, this cross-sectional database study did not allow for an assessment of the duration or severity of voice disorders owing to the large amount of missing data associated with these variables or the lack of these measurements. The database also provides information regarding whether an individual sought treatment for a voice disorder but did not specify what type of treatment was received. In addition, the prevalence estimate of voice disorders is dependent on patient-reported symptoms and does not require an objective diagnosis. Survey responses may be affected by recall bias; respondents who had recently experienced a voice disorder may be more likely to report the disorder than those whose symptoms occurred much further in the past. In querying respondents regarding the presence of a voice disorder in the last 12 months, the survey may have captured respondents who experienced dysphonia related to an upper respiratory tract infection or other self-limiting infectious cause. This possibility would overestimate the proportion of elderly adults with chronic voice disorders for whom otolaryngology referral would be appropriate. Third, the analysis of many variables was limited by the low number of observations, which complicated further elaboration of the possible causes of voice disorders and multivariate analysis of the treatment of voice disorders. Despite this issue, the variables and assessments described in this study were included only after rigorous examination to ensure that there was an adequate number of observations. Finally, the results of the study are generalizable only to the United States.
A significant proportion of elderly adults experience voice disorders. However, only a small percentage of this population seeks treatment for their voice disorder, and even fewer are evaluated by an otolaryngologist or an SLP. A greater percentage of elderly adults who sought treatment for their voice disorder noted an improvement in symptoms compared with those who did not seek treatment. However, compared with younger adults, a smaller proportion of elderly adults with voice disorders improved with treatment. This finding is likely associated with differing causes of voice disorders in elderly adults vs younger adults, as well as the low use of specialty care. Outreach efforts should be targeted toward improving the understanding of voice disorders among patients and PCPs as well as increasing awareness of the services available via otolaryngologists and SLPs.
Accepted for Publication: April 1, 2018.
Corresponding Author: Michael M. Johns III, MD, University of Southern California Voice Center, Caruso Department of Otolaryngology–Head and Neck Surgery, University of Southern California, 1540 Alcazar St, Ste 204M, Los Angeles, CA 90033 (email@example.com).
Published Online: July 12, 2018. doi:10.1001/jamaoto.2018.0980
Author Contributions: Drs Bertelsen and Johns had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: All authors.
Acquisition, analysis, or interpretation of data: Bertelsen, Zhou.
Drafting of the manuscript: Bertelsen, Johns.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Zhou.
Administrative, technical, or material support: Hapner.
Supervision: Bertelsen, Johns.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Meeting Presentation: This study was presented at the Annual Conference of the Fall Voice; October 13, 2017; Washington, DC.
Create a personal account or sign in to: