Integrating Advances in Personality Science to Re-Examine the Trait Theory of Voice Disorders

Abstract

Purpose:

Voice disorders associated with vocal hyperfunction are some of the most common vocal pathologies. Certain personality traits are thought to be a risk factor for developing these disorders. The trait theory of voice disorders (TTVD) provided a unified framework to understand these relationships. This study re-examines the TTVD by adopting current theories and methods from personality science.

Method:

This cross-sectional study recruited individuals diagnosed with primary muscle tension dysphonia (MTD) or a diagnosis associated with phonotraumatic vocal hyperfunction (PVH), that is, benign bilateral lesions of the lamina propria and bilateral or unilateral vocal fold polyp(s). Participants completed a contemporary personality battery. Data were analyzed via structural equation modeling and compared to vocally healthy controls.

Results:

Several significant differences existed between vocally healthy controls (n = 416) and participants with MTD (n = 71) or PVH (n = 38). Compared to the controls, participants with MTD reported lower levels of Stability, Plasticity, Conscientiousness, Openness/Intellect, and the corresponding aspects of Industriousness and Intellect, respectively. Conversely, the MTD cohort was significantly higher in Neuroticism and its two corresponding aspects, Withdrawal and Volatility. Likewise, when compared to controls, the PVH reported significantly lower levels of Stability, Agreeableness, its aspect Politeness, and the aspect of Industriousness. Finally, compared to the MTD cohort, individuals with PVH were higher on Extraversion, specifically the aspect of Assertiveness, and lower on the aspect of Politeness.

Conclusions:

The current study largely replicates the initial TTVD studies and updates them with a modern theory of personality. These results provide a strong foundation for future investigations to continue to study the relationships between personality traits and voice disorders.

Supplemental Material:

https://doi.org/10.23641/asha.28843052

In the United States, approximately 30% of people will experience voice problems at some point in their lives, and an estimated 8% of adults currently report difficulties with using their voice (Bhattacharyya, 2014; Roy et al., 2005). People with voice disorders report lower productivity at work and increased activity impairment and may require a short-term disability leave of absence, annually costing the U.S. economy $179–$295 million (L. Allen & Hu, 2022; S. M. Cohen et al., 2012a, 2012b; Meyer et al., 2013). Moreover, they report higher levels of depression, anxiety, and decreased quality of life (S. M. Cohen, 2010; Dietrich et al., 2008; Marmor et al., 2016; Martinez & Cassol, 2015; Mirza et al., 2003), as well as higher levels of suicidal ideation and suicide attempts (Kim et al., 2023).

Two of the most common voice disorders include primary muscle tension dysphonia (MTD) and benign phonotraumatic lesions of the lamina propria, also known as phonotraumatic vocal hyperfunction (PVH). These voice disorders comprise approximately 40% and 10%–15% of voice disorder clinic caseloads, respectively (Altman et al., 2005; Coyle et al., 2001; Van Houtte et al., 2010). Unlike other voice disorders, MTD and PVH are thought to result from suboptimal respiratory and phonatory coordination, that is, vocal hyperfunction—in the absence of an alternative etiology. When this vocal hyperfunction causes dysphonia in the absence of any structural changes or neurological impairment, it is considered MTD (also known as non-PVH). When this vocal hyperfunction results in benign tissue changes to the lamina propria (i.e., vocal fold nodules [VFNs] or polyps with or without reactive lesions), it is considered PVH (Hillman et al., 2020). Current theoretical frameworks also consider non-intubation-related vocal fold granulomas as possibly belonging to this PVH category.

Despite the negative economic, social, and psychological impacts of these voice disorders, the mechanisms underlying the pathogenesis and maintenance of MTD and PVH continue to not be well understood. Stress appears to play a key factor in MTD, as laboratory studies have demonstrated increased intrinsic and extrinsic laryngeal muscle activation during acute stressors in vocally healthy women (Dietrich et al., 2019; Dietrich & Abbott, 2014; Dietrich & Verdolini Abbott, 2012; Helou et al., 2013, 2018, 2020). However, millions of people routinely experience stress but manage to not develop MTD. Likewise, many people engage in high vocal doses and/or engage in phonotraumatic behaviors but do not develop phonotraumatic lesions. Essentially, these behaviorally driven voice disorders result from an otherwise healthy individual who, in a relatively short period of time, begins to engage in suboptimal phonatory behaviors that result in dysphonia, warranting medical intervention.

Several lines of evidence suggest possible etiological factors that may contribute to MTD and PVH. Hillman et al. (2020) provide an excellent summary of these different lines of evidence, which are reviewed briefly here. First, individuals who develop PVH might be more susceptible to vocal fold tissue alteration/damage due to irritants (e.g., tobacco use) or genetic factors. Although this idea is accepted clinically, little to no empirical evidence exists to demonstrate conclusive findings. Second, individuals with MTD and PVH may engage in aberrant phonatory biomechanics (Espinoza et al., 2017; Hillman et al., 1990) and/or increased intrinsic laryngeal muscle tension (Heller Murray et al., 2017). These findings support the idea that individuals with MTD and PVH use “inefficient” phonatory behaviors that contribute to developing/maintaining these disorders.

Additional evidence suggests that individuals with MTD and PVH might have anomalistic sensorimotor functioning, as they exhibit atypical responses to pitch-shift paradigms (Stepp et al., 2017; Ziethe et al., 2019). However, what is considered a “typical” response to a pitch-shift paradigm remains unclear, as no known factors predict whether a person will oppose or follow the pitch change (e.g., Burnett et al., 1998; Franken et al., 2018). Although these findings suggest contributions from altered auditory feedback, more work is needed to better understand if altered auditory feedback is causal in nature. Finally, and most germane to the present study, is the recognition that individual psychological differences may increase the risk of developing these voice disorders.

Relative to PVH, clinicians and researchers have generally considered MTD's etiology to be more psychological in nature. This perspective is apparent given the various names used to describe MTD in the literature, including psychogenic dysphonia, conversion dysphonia, and hysterical dysphonia. Early research described MTD as the result of an internal personality conflict or as a neurotic voice disturbance (Moore, 1939; Wyatt, 1941). Notably, these studies relied on early personality scales at a time when the field of personality psychology lacked a broad consensus on personality trait structure. Subsequent voice literature has suggested a psychoanalytic notion of “conflict over speaking out” (Baker, 2003; House & Andrews, 1988). Many individuals with these disorders, especially women, have been considered “emotionally unstable” and/or thought to have histrionic or hysterical personality traits (House & Andrews, 1987, p. 484, 1988).

Conversely, Yano et al. (1982) were among one of the first studies that examined personality traits in patients diagnosed with vocal fold polyps or nodules (i.e., PVH). These authors found that individuals with PVH exhibited higher Extraversion. Later, Green (1989) reported that children with VFNs displayed significantly higher scores on traits of Acting Out, Distractibility, Disturbed Peer Relations, and Immaturity.

While these early studies found meaningful relationships between personality traits and voice disorders, they lacked a comprehensive framework to understand these relationships. Roy and Bless (2000) integrated then-contemporary ideas in personality psychology, including Eysenck's (1967) three-factor model of personality traits (i.e., Neuroticism, Extraversion, and Psychoticism) and Gray's neuropsychological model of the conceptual nervous system (Gray, 1975, 1987). Roy and Bless (2000) posited that people with VFNs were “neurotic extraverts” and that people with MTD were “neurotic introverts.” This landmark article and subsequent investigations (Roy et al., 2000a, 2000b) have since become known as the trait theory of voice disorders (TTVD) and have influenced much of the research into personality traits and voice disorders in the decades since.

Generally, studies demonstrate that people with VFNs are higher in Extraversion and Psychoticism while lower in Constraint and Agreeableness (Mattei et al., 2017; Roy et al., 2000a, 2000b). Although Roy et al. (2000a) hypothesized that people with PVH would score higher on Neuroticism, these findings have been less robust, ¹ with the most significant relationship between PVH and Neuroticism measured in children (J. M. Lee et al., 2021). Broadly, these studies converge on the idea that people with PVH are more social (and thus have higher vocal demand) but also may tend to be more socially dominant, aggressive, impulsive, and/or less agreeable. Researchers speculate that these combinations of traits lend individuals to be in the social limelight or to speak over others in “an aggressive verbal style” (Roy et al., 2000a, p. 762). These behaviors and implied “vocal abuse” could then lead to the tissue changes associated with PVH.

Roy et al. (2000a, 2000b) also found that people with MTD were more introverted (low Extraversion) and scored higher on Neuroticism/Negative Emotionality and their facets, particularly Stress Reaction. These findings were consistent with the TTVD's prediction that people with MTD would score higher on Neuroticism and lower Extraversion. Additionally, Roy et al. (2000b) found that individuals with MTD scored higher on Constraint (inversely related to Eysenck's Psychoticism scale), although not all studies have replicated these differences (van Mersbergen et al., 2008).

Other researchers have tested the hypothesized relationships of the TTVD in different populations as well. Toles et al. (2021) reported that female singers with VFNs scored higher in the lower order traits of Social Potency and Control compared to vocally healthy singers. Research into children with VFNs has found they exhibit higher levels of Extraversion (Verduyckt et al., 2019) and higher levels of Neuroticism and lower levels of Agreeableness and Conscientiousness (J. M. Lee et al., 2021).

While the exact findings differ across studies, collectively, these studies support the idea that certain personality traits may serve as risk factors for developing MTD or PVH. The TTVD was seminal for its time and has guided and informed research in the decades since. However, in its current form, the TTVD is inconsistent with contemporary theories and methods in robust personality science. If researchers seek to better understand these relationships, then it is essential to adopt theories and methods used in modern, rigorous personality science. Herein we identify several theoretical and methodological discrepancies that could be improved upon to better understand the relationship between personality traits and voice disorders.

First, personality scientists have overwhelmingly adopted a five-factor model (FFM) of personality (John, 2021). These five broad personality trait domains (the “Big Five”) are Neuroticism, Agreeableness, Conscientiousness, Extraversion, and Openness/Intellect. ² While slight variations exist, similar factor structures have largely been replicated across numerous languages (Ashton et al., 2004; Hofstee et al., 1997; John, 2021) and linked to various neurophysiological processes (T. A. Allen & DeYoung, 2016; DeYoung, 2010; DeYoung et al., 2021).

Investigations into personality traits and voice disorders in adults have overwhelmingly used a three-factor model of personality. Amir et al.'s (2023) study is one of the few studies using an FFM in adults, but they examined a wide variety of voice disorders not previously associated with personality traits (e.g., sulcus, paresis). Research into children with PVH using an FFM has found that these children have higher levels of Neuroticism and lower levels of Agreeableness and Conscientiousness (J. M. Lee et al., 2021) or only higher levels of Extraversion (Verduyckt et al., 2019).

Second, adopting a comprehensive theory of personality may prove beneficial. Jeffrey Alan Gray, a student of Hans Eysenck, proposed biological mechanisms to explain Eysenck's (1947) two-factor model of personality, Extraversion and Neuroticism. ³ This biopsychological theory of personality later evolved into the reward sensitivity theory (RST; Gray, 1975, 1978, 1981). The RST asserted that the behavioral inhibition system, behavioral activation system, and the fight–flight–freeze system (FFFS) ⁴ are responsible for individual differences to reward and punishment and thus may be viewed as a biological basis for personality traits. Roy and Bless (2000) integrated the RST to the TTVD, providing a mechanistic explanation for the relationships between personality traits and voice disorders.

In the decades since, the RST has been revised, and the mechanisms associated with personality traits are now inconsistent with the original TTVD framework (Gray & McNaughton, 2000). Although Gray's work helped to pioneer the biological basis of personality, neither the RST nor its revision are a comprehensive theory of personality. One theory that meets this criterion is the Cybernetic Big Five Theory of Personality (CB5T). The CB5T is arguably one of the first “comprehensive, synthetic, and mechanistic” personality theories; this theory seeks to “explain not only how individuals differ from each other in their persisting patterns of emotion, motivation, cognition, and behavior, but also why” (DeYoung, 2015, p. 33, emphasis in original). As a result, the CB5T is “the most recent biologically based theoretical account of the Big Five” personality traits (John, 2021, p. 70). DeYoung acknowledges that room for improvement in this theory exists. This room for improvement notwithstanding, to date, the CB5T represents the most comprehensive, biologically motivated, and mechanistic explanation for personality that integrates empirical research from neuroscience, genetics, and observational studies (T. A. Allen & DeYoung, 2016; DeYoung, 2010, 2015; DeYoung et al., 2010, 2021). Framing our understanding of personality and voice disorders within such a theory will allow future research to generate novel, testable hypothesis regarding physiological mechanisms that may contribute to certain voice disorders.

Like many other approaches to understanding personality, the CB5T leverages a hierarchically arranged FFM of personality traits (see Figure 1). These traits are arranged from the most narrow traits (facets), intermediate-level traits (aspects), the broad “Big Five” traits (domains), and the highest order/broadest traits (metatraits). The CB5T defines personality traits as “probabilistic descriptions of relatively stable patterns of emotion, motivation, cognition, and behavior, in response to classes of stimuli that have been present in human cultures over evolutionary time” (DeYoung, 2015, p. 35). Importantly, personality researchers have not reached a consensus on the number or nature of the lowest order traits, facets, and these traits will vary across different batteries.

Figure 1.

The personality trait hierarchy in DeYoung's (2015) Cybernetic Big Five Theory of Personality. Traits are hierarchically arranged from the broadest traits (top) to the narrowest (bottom). Facets, the narrowest of the traits, are shown here for illustrative purposes, as researchers have not yet reached a consensus on these traits. Neuroticism and its lower order traits are inversely related to Stability.

In addition to adopting a comprehensive and mechanistic theory of the “Big Five” traits, careful consideration should be given to the psychometric properties of the instruments used to measure these traits. Most of the research examining the TTVD has used either the Eysenck Personality Questionnaire (EPQ; Eysenck & Eysenck, 1975) or the MPQ or its corresponding brief form (MPQ-BF; Patrick et al., 2002; Tellegen & Waller, 2008). The EPQ requires participants to respond “yes”/“no” to a series of 90 statements, while participants respond “true”/“false” to a series of 300 statements for the MPQ (155 items for the MPQ-BF). These dichotomous response options have mathematical limitations when using factor analysis. Current evidence suggests using at least a 3-point response option but ideally a 5- or 7-point option to improve psychometric precision (K. Lee & Ashton, 2007; Simms et al., 2019).

An ideal FFM personality battery would balance participant burden with internal consistency and use polytomous response options. The Big Five Aspect Scales (BFAS; DeYoung et al., 2007) meets these criteria and is a well-validated personality battery rooted in the CB5T. The BFAS measures the personality trait hierarchy at the levels of the aspects, domains, and metatraits shown in Figure 1.

Finally, adopting analytic methods used by personality researchers may also advance our understanding of these relationships. Roy et al. (2000a) categorized individuals as “high” or “low” on a trait via a median split. Median splits were once a common approach when analyzing data via analyses of variance. However, artificial dichotomization reduces statistical power and effect sizes, attenuates correlations, and introduces measurement error via a nonlinear transformation of a continuous variable (J. Cohen, 1983; Decoster et al., 2011; McClelland et al., 2015). Additionally, Roy et al. (2000a, 2000b) used a stepwise logistic approach to examine group differences across personality traits. Stepwise regression has been highly criticized and shown to inflate Type I errors and spuriously identify “significant” predictor variables (Derksen & Keselman, 1992; Smith, 2018). These analytic limitations pose substantial threats to internal validity. Subsequent studies into personality and voice disorders have often replicated these methods, which may, in part, contribute to the inconsistent findings reported across these various studies.

One analytic technique to address these limitations is structural equation modeling (SEM). This method is a highly flexible, gold-standard analytic framework in rigorous multivariate research. SEM is a family of methods that allows researchers to model “the mechanisms presumed to give rise to observed variability, covariation, and patterns in the data” (Hoyle, 2023, p. 3). With SEM, researchers can use latent variables, theoretically “purer” measures of the construct of interest that account for measurement error in the observed variables. Additionally, unlike regression-based approaches that are mathematically just-identified, SEM allows for mathematical overidentification, permitting researchers to impose model constraints and test their theoretical model.

To summarize, decades of research have demonstrated meaningful relationships between psychological components (e.g., personality) and certain voice disorders. The TTVD revolutionized this endeavor by providing a mechanistic explanation for these relationships. The current study builds upon the TTVD by integrating theoretical and methodological advances from contemporary personality science while adopting an updated framework of vocal hyperfunction. Addressing these discrepancies is a necessary first step to establish a solid foundation for future, more rigorous research.

Hypotheses

The current study has two main goals—to replicate Roy et al.'s (2000a, 2000b) initial TTVD studies and to do so within a contemporary, mechanistic, and comprehensive theory of personality. To achieve these goals, we propose a series of hypotheses to translate between the three- and five-factor personality structures. We base these hypotheses both on the original tenets of the TTVD (Roy & Bless, 2000), as well as the results of the empirical investigations into the TTVD (Roy et al., 2000a, 2000b). Figure 2 displays a schematized version of the below hypotheses to aid with readability.

Figure 2.

A schematization of the hypothesized personality trait differences across the groups. MTD = primary muscle tension dysphonia; PVH = phonotraumatic vocal hyperfunction.

First, compared to controls, we hypothesize that individuals with MTD will score higher on Neuroticism and its two aspects, Withdrawal and Volatility. Second, compared to controls, we hypothesize that the MTD group will score lower on Extraversion and its two aspects, Enthusiasm and Assertiveness. These first two hypotheses interrogate the TTVD's premise that people with MTD will score high on Neuroticism and low on Extraversion.

Third, we hypothesize that the PVH cohort will score higher on the aspect of Volatility compared to controls. Although the TTVD asserted that people with PVH will have higher levels of Neuroticism, Roy et al. (2000a, 2000b) did not find strong evidence for this relationship. Roy et al. (2000b) found that these individuals exhibited higher levels of Aggression, a facet of the MPQ's Negative Emotionality. In the FFM, MPQ's Aggression most closely correlates with Hostility (Church, 1994), a facet that loads onto Volatility (DeYoung et al., 2007). It stands to reason that if the TTVD's claim that people with PVH are higher in Neuroticism, then this difference may be due to the lower order trait of Volatility.

Fourth, we hypothesize that the PVH group will score significantly higher on Extraversion and its two aspects, Assertiveness and Enthusiasm, compared to controls. The TTVD proposes that people with PVH will score higher on Extraversion. The evidence for this relationship is mixed, with some studies supporting this association (Yano et al., 1982), while others do not (Roy et al., 2000a, 2000b). However, Roy et al. (2000b) reported that the PVH group was higher on Social Potency, a facet of MPQ's Positive Emotionality. Social Potency most closely correlates with Assertiveness in an FFM (Church, 1994; DeYoung et al., 2007).

Currently, no evidence exists suggesting that this group would score higher on Enthusiasm. This trait is characterized by “the tendency toward gregarious social interaction” (DeYoung, 2015, p. 42); Roy et al. speculated that people with PVH may have “a predilection for socializing” (2000b, p. 538). Based on this speculation, we hypothesize that the PVH group will score higher on Enthusiasm compared to controls. Collectively, the third and fourth hypotheses interrogate the TTVD's assertion that people with PVH will score high on both Neuroticism and Extraversion.

Fifth, we hypothesize that the PVH group will score significantly lower on Conscientiousness and its two aspects, Industriousness and Orderliness. Roy et al. (2000a, 2000b) found that individuals with PVH reported significantly higher scores on the EPQ's Psychoticism and lower scores on the MPQ's Constraint. MPQ's Constraint most strongly correlates with the FFM's Conscientiousness domain (Church, 1994). Roy et al. (2000b) also found that the PVH group reported significantly lower scores compared to vocally healthy controls on Control, a facet of MPQ's Constraint. MPQ's facet of Control correlates with the facets Order, Dutifulness, Achievement, and Deliberation (Gaughan et al., 2009), which load onto both Industriousness and Orderliness (DeYoung et al., 2007).

Finally, our last hypothesis concerns the difference between the PVH and MTD groups. The TTVD asserts that the PVH and MTD groups will differ only by higher and lower scores on Extraversion, respectively. While Roy et al. (2000a, 2000b) did not find significant differences in Extraversion in the PVH group compared to controls, the PVH group was higher on Extraversion and the facets of Positive Emotionality compared to the MTD group. We seek to replicate this difference and hypothesize that, compared to the MTD group, the PVH cohort will be higher on Extraversion and its aspects, Enthusiasm and Assertiveness.

Method

This study was reviewed and approved by the University of Pittsburgh's institutional review board (STUDY23010171). We partnered with eight specialized voice disorder clinics around the country to recruit a diverse sample of participants receiving a first-time diagnosis of a voice disorder.

Participants

All participants were community-dwelling individuals who sought out medical intervention for a voice complaint at their respective interdisciplinary voice disorder clinic. All participants underwent each clinic's own voice evaluation protocols, which included auditory perceptual evaluations and laryngeal imaging. Participants were evaluated with an interdisciplinary voice care model (typically a voice-specialized speech-language pathologist and a board-certified laryngologist or laryngology physician assistant overseen by a board-certified laryngologist). While the specific procedures may have differed across clinics, all clinic recruitment sites were chosen for their reputations of providing gold-standard voice assessment and treatment.

When a patient being evaluated by the health care team met the inclusion and exclusion criteria (see Table 1), a clinician would inform them of the study and provide them with a flyer. We adopted Hillman et al.'s (2020) definition of PVH, that is, a benign bilateral mid-membranous lesion, with or without a reactive lesion, specifically diagnoses of VFNs or unilateral/bilateral polyp(s). Although Hillman et al. (2020) acknowledge that non-intubation-related granulomas might also exist within this PVH category, we refrained from recruiting these individuals given the uncertainty around including this diagnosis in the PVH category. Additionally, clinicians at the voice clinics were instructed to recruit classic, unambiguous presentations of the vocal pathologies. We aimed to minimize/avoid recruiting patients who presented with ambiguous or complex presentations that did not fit neatly in the diagnostic categories.

Table 1.

Inclusion and exclusion criteria for the voice disorder cohorts.

Inclusion criteria	Exclusion criteria
Between ages 18 and 60 years Demonstrated fluency with written and spoken English Receiving a first-time diagnosis of MTD or a diagnosis related to PVH (i.e., benign bilateral mid-membranous lesion[s], with or without a reactive lesion)	Outside the 18–60 years age range The presence of a concomitant voice or upper airway diagnosis (except for laryngopharyngeal reflux, which was allowed due its high concomitance with MTD and PVH) A history of a previous voice/upper airway disorder A history of receiving voice/speech therapy for a voice/upper airway disorder (e.g., articulation therapy as a child would not exclude someone)

Note. MTD = primary muscle tension dysphonia; PVH = phonotraumatic vocal hyperfunction.

All data were collected using Research Enterprise Data Capture (REDCap; support from the National Institutes of Health through Clinical and Translational Sciences Institute at the University of Pittsburgh Grant UL1-TR-001857). Individuals who were interested used the flyer to access the secure online REDCap survey and participate in the study. All participants completed the study on their own personal electronic device. Ideally, participants were recruited at the time of their evaluation after receiving their diagnosis. However, due to slow recruitment, we also allowed clinicians to recruit individuals at the time of their first voice therapy appointment in case the patient did not receive a recruitment flyer at the time of their evaluation. Individuals were ineligible to participate if they were not recruited by their first voice therapy appointment.

After consenting to participate, the participants completed a series of online questionnaires. Participants were encouraged to complete the study in one sitting; however, participants could quit or resume the study at any point. Participants were prevented from going back and changing answers on previous pages of the survey. If a participant did not finish the survey, any data provided were analyzed. Participants who successfully completed the survey were offered an electronic gift card to their choice of a major retailer. Initially, reimbursement was set at $10. After a period of slow recruitment, we increased reimbursement to $20 and finally to $40. Most participants who completed the survey received a $40 gift card.

The data for the vocally healthy controls were collected from a separate research study. Briefly, in that study, we collected both personality and perceived vocal handicap (i.e., Voice Handicap Index–10 [VHI-10]) data from online users who agreed to listen and rate speech samples. From this previous listening study, we collected data from 1,843 individuals. For the current study, we kept individuals only between the ages of 18 and 60 years who did not have any missing data (N = 1,811). Because these individuals did not receive a voice evaluation, we retained data from participants with a VHI-10 sum score ≤ 3 based on published norms indicating that the average VHI-10 score for people without voice disorders is 2.83 (SD = 3.93; Arffa et al., 2012). With this conservative cutoff, we use the personality data from these individuals as a “vocally healthy” control group (n = 415).

Measures

After completing a brief screening questionnaire, participants read and completed the online consent form. Participants filled out a brief demographic questionnaire and were asked a series of general voice questions (e.g., voice use, dysphonia symptoms). Participants were also asked to indicate their diagnosis.

We employed two methods aimed at improving the quality of the data collected. First, participants were asked to “commit to providing accurate and honest information.” Participants answered whether they do or do not commit to providing accurate and honest information. Regardless of their answer, participants were allowed to continue with the survey. These types of commitment questions have been shown to improve the accuracy and quality of participants' responses for surveys in person and online (Hibben et al., 2022). We also embedded three “traditional” attention check items (e.g., “Select ‘Strongly Agree’ for this item”). Participants who missed the first two attention check items were prompted to “pay close attention to the questions and provide as much accurate and honest information as possible.”

Participants filled out a series of well-validated self-report measures of personality, perceived vocal handicap, depression, anxiety, and perceived stress. Participants also completed an abbreviated measure of internalizing and externalizing psychopathology. The current study only uses the measures of perceived voice handicap and personality. The other measures collected are reported in Supplemental Material S1.

For the current study, participants reported their perceived voice handicap using the VHI-10, a 10-item self-report measure of perceived vocal handicap (Rosen et al., 2004). The VHI-10 contains a series of statements with a 5-point Likert response scale from never (0) to always (4). Participants' responses are summed to yield a score from 0 to 40, with higher scores indicating higher levels of perceived vocal handicap. Prior research suggests a score greater than 11 should be considered “abnormal” and indicative of a possible vocal impairment (Arffa et al., 2012).

The BFAS (DeYoung et al., 2007) is a rigorously tested personality battery measuring traits at three separate levels of abstraction, that is, aspects, domains, and metatraits. Each personality trait serves a specific theoretical function in the evolved cybernetic mechanism of personality (DeYoung, 2015). The BFAS requires individuals to use a 5-point Likert scale to respond to a series of statements from strongly disagree (1) to strongly agree (5). Certain items are reverse scored. Items corresponding to each trait are averaged to yield a mean trait score between 1 and 5.

Data Processing and Cleaning

All data from participants recruited from voice clinics were imported and processed in R (V.4.0.2; R Core Team, 2020) via REDCap's secure application programming interface. Data from vocally healthy controls were imported from a previously cleaned csv file. Packages used to process and analyze the data included: tidyverse (V.1.3.0; Wickham et al., 2019), semTools (V.0.5-6; Jorgensen et al., 2022), and lavaan (V.0.6.14; Rosseel, 2012).

Participants were automatically grouped into their diagnostic cohort based on the flyer their clinician handed them. We designed this step to ensure that participants were sorted into the correct group based on the clinician's decision, rather than solely relying on self-reported diagnosis from the participants. However, we still asked participants to report their diagnosis. Whenever a discrepancy existed (e.g., a participant who completed the MTD survey but indicated they had a lesion), the first author (B.W.) contacted the participant for clarification and manually coded them to the appropriate group.

Statistical Analysis

We initially sought to use a multigroup SEM to model personality traits and other psychological processes (e.g., depression) across groups. We conducted an a priori power analysis via Monte Carlo simulation, which indicated that adequate model fit could be achieved with 100 people per group. More information about this power analysis is available in Supplemental Material S1. However, due to slower-than-anticipated recruitment, we abandoned this initial plan and examined personality differences between the groups using a multiple indicator and multiple causes (MIMIC) modeling approach.

We tested our hypothesized group differences in via three MIMIC SEM analyses (Joreskog & Goldberger, 1975), one for each level of the personality trait hierarchy. Like multigroup SEM, MIMIC models allow for group comparisons. However, group mean differences are expressed as regression coefficients instead of latent variable means, allowing for accurate parameter estimates with smaller sample sizes compared to multigroup SEM methods.

All three MIMIC models and their fully standardized solutions were estimated via maximum-likelihood estimation and contained the same exogenous predictors (i.e., age, sex, and two dummy variables—MTD vs. controls and PVH vs. controls). ⁵ Age and sex were included as covariates to control for known personality differences across the life span and by biological sex (Caspi et al., 2005; Roberts et al., 2006, 2008; Roberts & Delvecchio, 2000; Soto et al., 2011; Weisberg et al., 2011). ⁶

We fixed the latent variances to 1 for identification and used random parcel allocation for the respective manifest indicators. Parceling is a procedure where multiple items are averaged together to serve as an indicator for a latent variable. This method helps to increase power by reducing the number of indicators needed in the model and more closely approximates a continuous, normal distribution (Sterba, 2011; Sterba & MacCallum, 2010; Sterba & Rights, 2023). Random parcel allocation iteratively repeats this process, randomly assigning items across corresponding indicators and estimating the model to each allocation. The results of each model are pooled using multiple imputation (Rubin, 1987). We chose 100 random allocations for each model. The metatrait model included six parcels each for Stability and Plasticity, while the domain and aspect models included four parcels for each respective latent variable.

Although the metatraits are theoretically orthogonal (DeYoung, 2015), the metatraits demonstrate inflated correlation when measured only via self-report (Change et al., 2012). As a result, we allowed their residuals to covary. Additionally, the latent variables' residuals were allowed to covary based on previous evidence that these traits are intercorrelated (DeYoung et al., 2002, 2007; Digman, 1997).

We used the following global fit indices and their commonly used cutoffs to help aid with interpreting model fit. We assessed model fit via the χ², comparative fit index (> .95), Tucker–Lewis index (> .95), root-mean-square error of approximation (< 0.06), and the standardized root-mean-square residual (< 0.08; Hu & Bentler, 1999; West et al., 2023). We set our α to .05 to determine statistical significance.

Results

Participants

Data from the participants with voice disorders were collected from September 8, 2023, to September 26, 2024 (384 days). All participants voluntarily committed to providing accurate and honest information. After removing a participant who failed the three traditional attention checks (n = 1) and a repeat entry from an individual who filled out the survey twice (n = 1), a total of 110 individuals were recruited from voice disorder clinics to participate in the research study. One participant completed the survey and indicated they were a healthy control (a clinician at one of the voice centers). Of the 109 participants with a voice disorder that consented to the study, 102 completed the study through the BFAS personality battery (100 completed the study in its entirety). Because the models were estimated via maximum likelihood, we retained all data, including those who did not complete the entire survey.

Measures

Table 2 contains the demographic data of the participants, and Table 3 contains the sum score of the VHI-10 and the mean score for each of the BFAS personality traits. Figure 3 contains the density distribution of participants' mean scores of each of the BFAS personality traits across groups.

Table 2.

Participants' demographic information.

Variable	MTD n = 71	PVH n = 38	Control n = 416	Total N = 525
Age: M (SD)	38.59 (11.69)	38.29 (13.62)	37.31 (10.23)	37.55 (10.70)
Female: n (%)	57 (80.28)	29 (76.32)	286 (68.75)	372 (70.86)
Cisgender: n (%)	66 (92.96)	36 (94.74)	413 (99.28)	515 (98.10)
Race: n (%)
White	51 (76.92)	30 (76.92)	309 (78.83)	390 (74.29)
Black/African American	12 (16.00)	4 (10.26)	51 (13.01)	67 (12.76)
Asian	6 (8.00)	2 (5.13)	16 (4.08)	24 (4.57)
Prefer not to say	5 (6.67)	2 (5.13)	3 (0.77)	10 (1.90)
Other	0	0	7 (1.79)	7 (1.33)
American Indian / Alaska Native	1 (1.33)	1 (2.56)	5 (1.28)	7 (1.33)
Pacific Islander	0	0	1 (0.26)	1 (0.19)
Ethnicity: n (%)
Not Hispanic / Latino/e	65 (91.55)	36 (94.74)	391 (93.99)	492 (93.71)
Hispanic / Latino/e	3 (4.23)	1 (2.63)	25 (6.01)	29 (5.52)
Unknown / not reported	3 (4.23)	1 (2.63)	0	4 (0.76)
Highest level of education attained: n (%)
No high school education	0	0	1 (0.24)	1 (0.19)
Some high school education	0	2 (5.26)	2 (0.48)	4 (0.76)
High school diploma	8 (11.27)	6 (15.79)	46 (11.06)	60 (11.43)
Trade or technical certificate	4 (5.63)	0	14 (3.37)	18 (3.43)
Some college	15 (21.13)	4 (10.53)	81 (19.47)	100 (19.05)
Associate degree	6 (8.45)	0	41 (9.86)	47 (8.95)
Bachelor's degree	20 (28.17)	14 (36.84)	126 (30.29)	160 (30.48)
Master's degree	8 (11.27)	10 (26.32)	88 (21.15)	106 (20.19)
Doctoral/terminal degree	8 (11.27)	2 (5.26)	16 (3.85)	26 (4.95)
Prefer not to say	2 (2.82)	0	1 (0.24)	3 (0.57)

Note. MTD = primary muscle tension dysphonia; PVH = phonotraumatic vocal hyperfunction.

Table 3.

Sum scores for the Voice Handicap Index–10 (VHI-10) and the mean scores for each of the personality traits on the Big Five Aspect Scales.

Variable	MTD	PVH	Control
VHI-10 sum score: M (SD)	18.99 (8.67)	17.89 (9.72)	1.38 (1.14)
Personality traits mean score: M (SD)
Stability	3.54 (0.38)	3.59 (0.32)	3.74 (0.42)
Neuroticism	2.90 (0.75)	2.65 (0.56)	2.50 (0.69)
Withdrawal	3.02 (0.76)	2.64 (0.60)	2.55 (0.73)
Volatility	2.78 (0.83)	2.65 (0.64)	2.46 (0.79)
Agreeableness	3.95 (0.38)	3.82 (0.41)	4.00 (0.52)
Compassion	4.08 (0.56)	4.04 (0.51)	4.01 (0.66)
Politeness	3.83 (0.36)	3.61 (0.44)	4.00 (0.55)
Conscientiousness	3.57 (0.54)	3.60 (0.52)	3.71 (0.55)
Orderliness	3.69 (0.60)	3.67 (0.56)	3.68 (0.61)
Industriousness	3.44 (0.69)	3.53 (0.60)	3.73 (0.66)
Plasticity	3.62 (0.45)	3.80 (0.37)	3.76 (0.48)
Extraversion	3.47 (0.50)	3.78 (0.46)	3.57 (0.60)
Enthusiasm	3.62 (0.59)	3.83 (0.56)	3.62 (0.70)
Assertiveness	3.31 (0.62)	3.73 (0.52)	3.52 (0.70)
Openness/Intellect	3.76 (0.51)	3.83 (0.45)	3.95 (0.51)
Openness	3.83 (0.59)	3.79 (0.57)	3.89 (0.60)
Intellect	3.70 (0.65)	3.87 (0.58)	4.00 (0.57)

Note. MTD = primary muscle tension dysphonia; PVH = phonotraumatic vocal hyperfunction.

Figure 3.

Density distribution and box plots of the mean scores of personality traits across groups. MTD = primary muscle tension dysphonia; PVH = phonotraumatic vocal hyperfunction.

Model Parameter Estimates

Each model measured one level of the personality trait hierarchy, that is, metatraits, domains, and aspects. The parameter estimates reflect the pooled values of the 100 randomly allocated item parcels. All three models successfully converged with admissible solutions for each of the 100 estimations. All three models demonstrated acceptable model fit. Modification indices were not examined for any of the models, as any model modifications would not align with the intended goal of examining group differences. Table 4 contains the global fit indices for the models examining each level of personality trait hierarchy. Table 5 contains the regression parameter estimates for the MIMIC models used to analyze the data. This table is a combined and abbreviated version of the full model output to aid with readability. The complete model results are available in Supplemental Material S1. Figure 4 displays the dot-and-whisker plots of the parameter estimates for the effects of age and sex across the models; similarly, Figure 5 displays the dot-and-whisker plots of the parameter estimates for the personality trait group differences controlling for age and sex.

Table 4.

Global model fit indices.

Model	df	χ2	p	CFI	TLI	RMSEA	SRMR
Metatraits	93	1.263	1.000	1.000	1.266	0	.005
Domains	220	1.993	1.000	1.000	1.335	0	.008
Aspects	815	7.297	1.000	1.000	1.550	0	.008

Note. CFI = comparative fit index; TLI = Tucker–Lewis index; RMSEA = root-mean-square error of approximation; SRMR = standardized root-mean-square residual.

Table 5.

Pooled regression parameter estimates for the metatrait, domain, and aspect-level models.

Regression parameter	Estimate	SE	95% CI [LL, UL]	p
Metatraits
Stability
Age	−.096	.105	[−.303, .110]	.361
Sex	.010	.004	[.002, .019]	.020
MTD vs. controls	−.470	.139	[−.744, −.197]	< .001
PVH vs. controls	−.492	.192	[−.869, −.115]	.011
PVH vs. MTD	−.022	.222	[−.457, .413]	.922
Plasticity
Age	−.171	.103	[−.374, .031]	.098
Sex	−.007	.004	[−.015, .002]	.140
MTD vs. controls	−.297	.138	[−.568, −.027]	.031
PVH vs. controls	.075	.192	[−.301, .451]	.694
PVH vs. MTD	.373	.224	[−.065, .811]	.095
Domains
Neuroticism
Age	−.011	.005	[−.020, −.002]	.012
Sex	−.379	.103	[−.580, −.178]	< .001
MTD vs. controls	.522	.138	[.253, .792]	< .001
PVH vs. controls	.311	.189	[−.060, .683]	.100
PVH vs. MTD	−.211	.221	[−.644, .223]	.340
Agreeableness
Age	.005	.005	[−.004, .014]	.265
Sex	−.673	.109	[−.887, −.459]	< .001
MTD vs. controls	−.089	.155	[−.392, .216]	.568
PVH vs. controls	−.488	.207	[−.894, −.082]	.018
PVH vs. MTD	−.400	.232	[−.854, .055]	.085
Conscientiousness
Age	.005	.005	[−.004, .014]	.256
Sex	−.144	.107	[−.355, .066]	.179
MTD vs. controls	−.291	.142	[−.569, −.012]	.041
PVH vs. controls	−.285	.196	[−.668, .099]	.145
PVH vs. MTD	.006	.227	[−.440, .451]	.980
Extraversion
Age	−.008	.005	[−.017, .001]	.082
Sex	−.228	.105	[−.435, −.021]	.031
MTD vs. controls	−.162	.140	[−.437, .113]	.247
PVH vs. controls	.368	.194	[−.013, .749]	.058
PVH vs. MTD	.531	.227	[.086, .945]	.019
Openness/Intellect
Age	−.003	.005	[−.012, .006]	.484
Sex	−.062	.106	[−.270, .147]	.561
MTD vs. controls	−.364	.142	[−.642, −.085]	.010
PVH vs. controls	−.283	.196	[−.667, .101]	.149
PVH vs. MTD	.081	.229	[−.368, .529]	.724
Aspects
Withdrawal
Age	−.020	.005	[−.030, −1.114]	< .001
Sex	−.324	.107	[−.534, −1.137]	.003
MTD vs. controls	.649	.145	[.365, .932]	< .001
PVH vs. controls	.264	.197	[−.123, .651]	.181
PVH vs. MTD	−.385	.231	[−.837, .068]	.096
Volatility
Age	−.002	.005	[−.011, .007]	.641
Sex	−.385	.103	[−.586, −.183]	< .001
MTD vs. controls	.360	.138	[.090, .630]	.009
PVH vs. controls	.321	.191	[−.054, .695]	.093
PVH vs. MTD	−.039	.223	[−.475, .397]	.860
Compassion
Age	.001	.005	[−.008, .010]	.748
Sex	−.598	.107	[−.807, −.389]	< .001
MTD vs. controls	.100	.140	[−.175, .375]	.477
PVH vs. controls	−.053	.195	[−.436, .329]	.785
PVH vs. MTD	−.153	.228	[−.600, .294]	.502
Politeness
Age	.009	.005	[−.001, .019]	.085
Sex	−.612	.123	[−.854, −.370]	< .001
MTD vs. controls	−.346	.214	[−.766, .074]	.107
PVH vs. controls	−1.000	.258	[−1.506, −.493]	< .001
PVH vs. MTD	−.654	.258	[−1.159, −.149]	.011
Industriousness
Age	.011	.005	[.001, .020]	.023
Sex	.059	.106	[−.148, .266]	.576
MTD vs. controls	−.447	.143	[−.728, −.167]	.002
PVH vs. controls	−.399	.197	[−.786, −.012]	.043
PVH vs. MTD	.048	.230	[−.403, .499]	.835
Orderliness
Age	−.003	.005	[−.012, .007]	.576
Sex	−.350	.113	[−.571, −.128]	.002
MTD vs. controls	−.045	.150	[−.340, .250]	.766
PVH vs. controls	−.076	.207	[−.481, .330]	.714
PVH vs. MTD	−.031	.237	[−.496, .434]	.896
Enthusiasm
Age	−.006	.005	[−.015, .003]	.197
Sex	−.409	.109	[−.622, −.196]	< .001
MTD vs. controls	−.002	.144	[−.284, .281]	.990
PVH vs. controls	.294	.199	[−.096, .685]	.139
PVH vs. MTD	.296	.232	[−.158, .750]	.201
Assertiveness
Age	−.008	.005	[−.017, .001]	.080
Sex	.161	.105	[−.190, .222]	.878
MTD vs. controls	−.271	.142	[−.549, .007]	.056
PVH vs. controls	.350	.167	[−.035, .735]	.075
PVH vs. MTD	.621	.230	[.170, 1.073]	.007
Openness
Age	−.008	.005	[−.018, .001]	.083
Sex	−.279	.110	[−.496, −.062]	.012
MTD vs. controls	−.122	.147	[−.410, .165]	.405
PVH vs. controls	−.210	.204	[−.610, .190]	.303
PVH vs. MTD	−.088	.238	[−.554, .378]	.711
Intellect
Age	.002	.005	[−.007, .011]	.640
Sex	.159	.107	[−.052, .369]	.139
MTD vs. controls	−.524	.145	[−.809, −.239]	< .001
PVH vs. controls	−.304	.200	[−.696, .088]	.129
PVH vs. MTD	.220	.233	[−.237, .678]	.345

Note. This table contains an amalgamated and abbreviated results from three separate statistical models. The complete model outputs are available in Supplemental Material S1. SE = standard error; CI = confidence interval; LL = lower limit; UL = upper limit; MTD = primary muscle tension dysphonia; PVH = phonotraumatic vocal hyperfunction.

Figure 4.

Dot-and-whisker plot of the regression estimates across the hypothesized models for the effects of age and sex. Black squares represent p < .05; solid lines = 95% confidence intervals. MTD = primary muscle tension dysphonia; PVH = phonotraumatic vocal hyperfunction.

Figure 5.

Dot-and-whisker plot of the combined regression estimates across the hypothesized models examining personality trait differences across the groups. Black squares represent p < .05; solid lines = 95% confidence intervals. MTD = primary muscle tension dysphonia; PVH = phonotraumatic vocal hyperfunction.

Hypothesized Group Differences

These results support our first hypothesis that the MTD group would score higher on Neuroticism and its aspects compared to controls. The MTD group reported significantly higher levels of Neuroticism (γ = .522, 95% CI [.253, .792], p < .001), Withdrawal (γ = .649, 95% CI [.365, .932], p < .001), and Volatility (γ = .360, 95% CI [.090, .630], p = .009). These results do not support our second hypothesis. Compared to the controls, the MTD group did not differ significantly on Extraversion (γ = −.162, 95% CI [−.431, .113], p = .247), Enthusiasm (γ = −.002, 95% CI [−.284, .281], p = .990), or Assertiveness (γ = −.271, 95% CI [−.410, .165], p = .405).

Similarly, these results did not support our third hypothesis that the PVH group would score higher on Volatility compared to the controls. The PVH group had a higher but not significantly different score on Neuroticism (γ = .311, 95% CI [−.060, .683], p = .100), Volatility (γ = .321, 95% CI [−.054, .695], p = .093), and Withdrawal (γ = .264, 95% CI [−.123, .651], p = .181). These data also do not support the fourth hypothesis that the PVH group would report higher levels of Extraversion and its aspects compared to the controls. However, the PVH group did report higher but not significantly different scores on Extraversion (γ = .368, 95% CI [−.013, .749], p = .058), Enthusiasm (γ = .294, 95% CI [−.096, .685], p = .139), and Assertiveness (γ = .350, 95% CI [−.035, .735], p = .075).

These results partially support our fifth hypothesis. While the PVH group did not score significantly lower on Conscientiousness (γ = −.285, 95% CI [−.668, .099], p = .145) or Orderliness (γ = −.076, 95% CI [−.481, .330], p = .714), they did score significantly lower on Industriousness (γ = −.399, 95% CI [−.786, −.012], p = .043), compared to controls. These results partially supported our final hypothesis that the PVH group would score significantly higher on Extraversion and its aspects compared to the MTD group. The PVH group reported significantly higher Extraversion (γ = .531, 95% CI [.086, .945], p = .019) and Assertiveness (γ = .621, 95% CI [.170, 1.073], p = .007), compared to the MTD group, but not on the aspect Enthusiasm (γ = .296, 95% CI [−.158, .750], p = .201).

Discussion

The current study re-investigated the TTVD by examining personality trait differences in adults with MTD and PVH. We sought to not only replicate the TTVD but to do so by integrating contemporary theories, perspectives, and evidence in voice and personality science. Specifically, we adopted Hillman et al.'s (2020) Updated Framework of Vocal Hyperfunction to investigate both non-PVH (i.e., MTD) and PVH (i.e., benign phonotraumatic lesions of the lamina propria). Additionally, we leveraged a comprehensive personality theory and its corresponding personality battery to examine a hierarchically arranged FFM of personality using SEM, a gold-standard analytic approach in the psychological sciences.

Age and Sex Differences

Personality traits have known differences across the sexes and lifespan. The current study sought to control for these confounding variables and investigate group differences above and beyond the effects of sex and age. Because these results are not central to the main purpose of the article, we will briefly review these findings.

Personality traits are relatively stable over time but continually change across the lifespan in relatively predictable patterns (see Roberts & Nickel, 2021, for a review). The current results are generally consistent with previous age and personality research (Roberts et al., 2006). We also largely replicated previously reported sex differences. Weisberg et al. (2011) studied sex differences on personality traits with the BFAS in 2,643 people (892 men, 1,751 women). The current results only differ in that we did not observe that men were significantly higher on Assertiveness and Intellect. Otherwise, the current results replicate their previous findings in a large separate sample.

Previous investigations into voice disorders and personality traits primarily recruit (presumably cisgender) women. This decision is often motivated by the fact that voice disorders are more common in women, with a preponderance of 3:1 for MTD and 1.8:1 for PVH (Aronson & Bless, 2009; Herrington-Hall et al., 1988). Yet, to uncover the veridical relationships between personality traits and voice disorders, it may be more beneficial to recruit a large diverse sample of participants and control for known confounding factors like age and sex.

Metatraits

The metatraits are the highest order traits in the CB5T trait hierarchy and relate to self-regulation and characteristic adaptations (i.e., interpretations, strategies, and goals; see DeYoung, 2015, for a full review). To our knowledge, this investigation represents the first attempt to study these highest order traits in patients with voice disorders. Mechanistically, Stability is associated with global levels of serotonin, while Plasticity is associated with global levels of dopamine (T. A. Allen & DeYoung, 2016; DeYoung, 2010, 2013; DeYoung et al., 2021).

Interestingly, both the MTD and PVH groups reported significantly lower Stability compared to controls. Higher levels of Stability relate to more behavioral restraint and maintaining goal-directed psychological functioning (DeYoung, 2015; Hirsh et al., 2009). Conversely, low Stability is “associated with impulsivity and lack of self-control, with various forms of distress, and, perhaps most relevantly, with a shaky sense of identity, direction, and social role” (Wilmot et al., 2016, p. 345). Wilmot et al. (2016) also suggest that low Stability relates to lower levels of self-monitoring. This association is consistent with previous literature demonstrating that people with MTD or PVH display aberrant responses to sensorimotor tasks (Stepp et al., 2017; Ziethe et al., 2019).

Compared to the controls, the MTD group also reported significantly lower Plasticity. Higher levels of Plasticity relate to more exploration, sensation seeking, and adaptability (DeYoung, 2015; DeYoung et al., 2002; Hirsh et al., 2009; Wilmot et al., 2016). This finding that the MTD cohort was significantly lower in Plasticity is consistent with previous portrayals of MTD that conceptualized the underlying processes of this disorder as fear of novelty and passive avoidance (Roy & Bless, 2000). To our knowledge, these findings represent the first investigation into group differences at the highest level of the personality trait hierarchy. These observed differences warrant further investigation and may provide novel insight into the processes underlying these disorders.

Domain and Aspect Traits

Neuroticism, Withdrawal, and Volatility

These results supported our hypothesis that people with MTD would report higher levels of Neuroticism, Withdrawal, and Volatility. These findings further corroborate previous studies that implicate stress, in part, as a contributing factor to the pathogenesis and maintenance of MTD. Individuals higher in Neuroticism are more likely to experience negative emotions; higher levels of Withdrawal are associated with anxiety and depression, and higher levels of Volatility relate to increased emotional lability (DeYoung et al., 2016). Higher stress levels and/or inadequate stress recovery has been theoretically linked to MTD (Dietrich & Abbott, 2008). Likewise, vocally healthy individuals exhibit increased laryngeal tension across physical (Helou et al., 2013), psychological (Dietrich & Abbott, 2014; Dietrich & Verdolini Abbott, 2012; Helou et al., 2018, 2020), and cognitive (Dahl & Stepp, 2023) laboratory stress paradigms.

Consistent with Hillman et al.'s (2020) Updated Theoretical Framework for Vocal Hyperfunction, it appears that increased laryngeal muscle activation (i.e., vocal hyperfunction) is a normal physiological response to stress. However, for people who are predisposed to more frequent/severe stress and negative emotions, they may exhibit increased laryngeal tension more frequently. This increased frequency may become habituated over time, resulting in chronic vocal hyperfunction (e.g., MTD).

Although the TTVD posits that people with PVH will also report higher levels of Neuroticism, empirical evidence for this claim is less robust. Our results demonstrate that the PVH group reports elevated levels of Neuroticism, Withdrawal, and Volatility, but none of them are significantly higher compared to the controls. Neither Toles et al. (2021) or Yano et al. (1982) reported significantly higher levels of Neuroticism for people with PVH.

While Roy et al. (2000a, 2000b) did not find that the PVH group reported higher levels of Neuroticism, the PVH group was significantly higher on the facets of Stress Reaction and Aggression (Roy et al., 2000b), specifically when using age-adjusted scores. Although the MPQ's facet of Stress Reaction correlates with Neuroticism in the FFM, the facet of Aggression is more closely related to low Agreeableness in a five-factor framework (Church, 1994; Gaughan et al., 2009). These results may indicate that a different constellation of traits than first proposed by the TTVD may better differentiate the PVH and control groups in a FFM framework. Alternatively, it is possible that the current PVH sample size (n = 38) is underpowered to detect a statistically significant effect.

Agreeableness, Compassion, and Politeness

Our results did not find any significant differences between the MTD group and the controls on Agreeableness or its two aspects, Compassion or Politeness. However, compared to the controls, the PVH group was significantly lower on Agreeableness and Politeness. Compared to the MTD group, the PVH group scored significantly lower on Politeness. Although we did not initially hypothesize any differences for these traits, these differences align with Roy et al. (2000b) when translating between the MPQ's and FFM's trait structure. With the MPQ, Roy et al. found that the PVH group scored significantly higher on the traits of Aggression and Social Potency. Viewed through an FFM lens, both Aggression and Social Potency exhibit negative correlations with Agreeableness and Politeness (Church, 1994; DeYoung et al., 2007; Gaughan et al., 2009). Thus, our current results are consistent with Roy et al.'s (2000b) findings for the PVH group through an FFM.

Conscientiousness, Industriousness, and Orderliness

We hypothesized that the PVH group would score lower compared to the controls on these traits based on the differences observed by Roy et al. (2000a, 2000b) on Psychoticism, Constraint, and Control, as these traits closely correlate with Conscientiousness (Church, 1994; Gaughan et al., 2009). The current results partially support this hypothesis; compared to the controls, the PVH group was only significantly lower on Industriousness, not Conscientiousness or Orderliness. Likewise, compared to the controls, the MTD group was significantly lower on Conscientiousness and Industriousness. These findings were unexpected, as we are unaware of any evidence linking Conscientiousness and/or Industriousness with MTD.

Although these differences with MTD were unexpected, leveraging a modern, mechanistic theory of personality like the CB5T has the potential to provide novel insight into possible mechanisms underlying these voice disorders. In personality neuroscience, Conscientiousness is associated with the central goal priority network and lower gray matter volume of the insula (T. A. Allen & DeYoung, 2016; DeYoung, 2010; DeYoung et al., 2021; Rueter et al., 2018). While no consensus exists for neurophysiological mechanisms for the lower order aspects, current thinking suggests that the goal priority network allows people to detect relevant stimuli and integrate emotional, motivational, and interoceptive information to influence behavioral outputs. Likewise, Conscientiousness is inversely related to insula volume, as the insula is thought to generate potentially distracting impulses (T. A. Allen & DeYoung, 2016; DeYoung et al., 2021; Rueter et al., 2018).

Several converging lines of evidence implicate self-monitoring with MTD and PVH that may help to explain these unexpected group differences. As previously mentioned, people with these voice disorders exhibit nontypical responses on sensorimotor tasks (Stepp et al., 2017; Ziethe et al., 2019). Likewise, individuals with MTD report significantly lower scores on a measure of interoceptive attention regulation (Smeltzer et al., 2023). Compared to controls during phonation, women with MTD exhibited decreased activation of the superior temporal gyrus, part of the goal priority network (Rueter et al., 2018). Similarly, the women in that study also exhibited increased insula activation during phonation compared to controls (Kryshtopava et al., 2017). Collectively, these separate findings are consistent with empirical investigations into the neurophysiological mechanisms underpinning Conscientiousness and may help to explain the unexpected group differences observed. Yet, it remains to be seen why both groups were significantly lower specifically on Industriousness. Future personality research may provide better insight into mechanisms associated with the aspects of Conscientiousness.

Extraversion, Enthusiasm, and Assertiveness

The current results do not support the hypothesized group differences for Extraversion or its aspects, Enthusiasm and Assertiveness. Based on the original tenets of the TTVD, compared to the controls, we hypothesized that the MTD group would be significantly lower on Extraversion and its aspects, while the PVH group would be significantly higher on these traits. While significant group differences did not exist compared to the controls, modest effects were in the hypothesized directions for MTD and PVH, respectively. Despite not being statistically significant, the MTD group reported generally lower Extraversion, specifically on the trait of Assertiveness. Likewise, the PVH group reported elevated Extraversion, Enthusiasm, and Assertiveness.

Although these traits did not differentiate these groups from the controls, the PVH group was significantly higher on Extraversion and Assertiveness compared to the MTD group. These findings support the notion that, in the context of chronic vocal hyperfunction, individuals higher on these traits may be more likely to develop PVH over MTD but that higher extraversion itself may not serve as a risk factor for developing chronic vocal hyperfunction. Alternatively, it may also be possible that Extraversion does serve as a risk factor, especially for PVH, and the small sample size of this group is underpowered to detect a statistically significant effect. This latter point would align with previous studies that have found significantly higher levels of Extraversion and corresponding lower order traits in people with PVH compared to controls (Mattei et al., 2017; Toles et al., 2021; Yano et al., 1982).

Overall, these findings again mirror Roy et al.'s original investigations. Roy et al. (2000a) found that the PVH group was significantly more extraverted than the MTD group, but not significantly higher on Extraversion compared to the controls. Likewise, Roy et al. (2000b) reported that the PVH group was significantly higher on the domain and lower order traits of Extraversion compared to the MTD group. Social Potency was the only lower order facet that the PVH group was significantly higher on compared to the controls. Social Potency most strongly correlates with Assertiveness but also has a strong negative correlation with Agreeableness and Politeness (Church, 1994; DeYoung et al., 2007; Gaughan et al., 2009). Compared to the controls in the current sample, the PHV group reported elevated (but not significant) levels of Assertiveness and significantly lower levels of Agreeableness and Politeness.

Openness/Intellect, Openness, and Intellect

Likewise, we did not expect to see any differences on Openness/Intellect or its aspects. The MPQ's Absorption scale most closely approximates the Openness/Intellect domain (Church, 1994; Gaughan et al., 2009), and it was initially developed to measure hypnotic susceptibility (Tellegen & Atkinson, 1974). Given its initial purpose, the Absorption scale was not discussed in the TTVD or Roy et al. (2000b). When it has been reported, no differences have been found in voice disorders (Toles et al., 2021; van Mersbergen et al., 2008). Notably, Amir et al. (2023) found that the voice disorder cohort was significantly lower on the domain Openness to Experience compared to the nondysphonic sample; however, directly comparing this trait across different personality batteries poses challenges. ⁷ Additionally, the heterogeneous sample of voice disorders in Amir et al. (2023), many of which have not been previously associated with personality traits, further limits our ability to compare findings. Mechanistically, Openness/Intellect and its aspects have been associated with dopamine, the dorsolateral prefrontal cortex, and the frontoparietal control network (T. A. Allen & DeYoung, 2016; DeYoung et al., 2005, 2021). However, we are unaware of any investigations into MTD demonstrating that these mechanisms relate to phonatory muscle pattern activation.

It is also important to acknowledge that the current study is among one of the first to show that the MTD group reported lower scores on Openness/Intellect and Intellect. These differences may reflect natural sampling variability and need to be replicated in separate, larger samples before concluding that these traits meaningfully differ in individuals with MTD. Without more rigorous work and/or sufficient theoretical rationale for these differences, we strongly caution against any interpretation or insinuation that individuals with MTD are less intelligent.

Limitations and Recommendations for Future Research

The current study integrated advances in voice and personality science to interrogate the relationships between voice disorders and personality traits. While we aimed to address several important limitations in the extant literature, the current study is not without its own. We hope future work will build off the current study to address the limitations of this study, some of which are described below.

First, the current sample size of participants with voice disorders is small. Although this sample size is relatively large for the field of voice disorders, samples with hundreds of participants yields more precise estimates. This small sample size should be kept in mind when interpreting these results, as we may be underpowered to detect meaningful differences with a small effect size. The current estimates are based on a relatively small sample size and therefore may change with larger samples, and while we attempted to recruit a large diverse sample of participants by partnering with multiple voice disorder clinics, 61.47% of the participants were recruited from one clinic. This limitation is an important consideration as meaningful personality differences exist across geographical locations (Allik & McCrae, 2004; Ebert et al., 2022; Ren et al., 2020). Future work would benefit from recruiting large diverse samples to better understand these relationships. Alternatively, future investigations with smaller samples may benefit from leveraging Bayesian frameworks, which are better suited for smaller samples when using appropriate priors (McNeish, 2016).

Another limitation is the cross-sectional design of this study. Although, logically, personality traits have temporal precedence before the presence of a voice disorder, it is possible that experiencing dysphonia may impact someone's personality. Additionally, a cross-sectional design only provides a single “snapshot” of someone's personality, which will be influenced by uncontrolled contextual factors (e.g., the argument they had on the phone before filling out the survey). Future work would benefit from adopting ecological momentary assessment designs that obtain measurements multiple times across a variety of situations. These longitudinal designs are necessary to better understand and model these psychological processes.

Conclusions

The current study re-investigated the TTVD, a seminal theory about the role of personality traits as potential risk factors for developing MTD or PVH. To our knowledge, this study is among one of the first to examine a hierarchically arranged FFM of personality that is rooted in a contemporary, comprehensive, and biologically motivated theory of personality. Likewise, this study is also among one of the first to investigate these relationships using SEM, a gold-standard analytic framework used in personality psychology research.

Although several differences exist, overall, the current study largely replicated many of the previous findings in Roy et al. (2000a, 2000b), but viewed through a FFM of personality. The MTD cohort reported higher levels of Neuroticism and its lower order traits compared to the controls, suggesting that stress may, in part, contribute to the pathogenesis and maintenance of MTD. Conversely, the PVH cohort reported lower levels of Agreeableness and Politeness compared to the controls and higher levels of Extraversion and Assertiveness compared to the MTD group. These findings align with previous suggestions that these individuals may experience higher vocal demands and/or speak in an aggressive or combative manner, resulting in phonotraumatic lesions.

Notably, both groups reported lower levels of Industriousness, an aspect of Conscientiousness, a trait that has been associated with attentional processes. This finding aligns with previous work suggesting that people with vocal hyperfunction may have aberrant sensorimotor and/or interoceptive attention regulation. While more work is necessary to fully understand personality's role as a possible risk factor for these voice disorders, these findings empirically contextualize our understanding of the TTVD through current personality science and provide novel questions for future research.

Data Availability Statement

The de-identified data set generated during the current study will be shared publicly in the Open Science Framework repository (https://osf.io/p3m8b/) after subsequent analyses for other manuscripts have been completed. In the meantime, the de-identified data set is available from the corresponding author on reasonable request.

Funding Statement

This project was funded, in part, by the Council of Academic Programs in Communication Sciences and Disorders PhD Scholarship and the American Speech-Language-Hearing Foundation New Century Scholars Doctoral Scholarship (both awarded to Brett Welch), as well as the National Center for Advancing Translational Sciences of the National Institutes of Health under Award Number TL1TR001858 (awarded to Brett Welch through the University of Pittsburgh's Institute for Clinical Research Education). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.