Vocal Fatigue: Current Knowledge and Future Directions

INTRODUCTION 

Vocal fatigue presents a significant challenge to contemporary research and clinical practice. Prevalence data suggest that vocal fatigue is particularly common among the teaching, singing and acting professions;1., 2. however, the definition, critical identifying features, and underlying mechanisms of this condition remain either uncertain or unproven. Much of the experimental literature attempting to induce vocal fatigue in human participants has yielded varied and inconsistent results. Further, no research exists specifically addressing the treatment of vocal fatigue.

An evaluation of progress in the area of vocal fatigue is needed to summarize the complexity within current knowledge and to identify key directions for future research. This article reviews the literature pertaining to the nature, underlying processes, and salient features of vocal fatigue. First, vocal fatigue is defined, its major symptoms are discussed, and hypotheses concerning its primary physiological and biomechanical mechanisms are considered. Second, studies of experimentally induced vocal fatigue in humans are evaluated. Third, research investigating the clinical and occupational manifestations of vocal fatigue is discussed. Fourth, directions for ongoing research in this area are offered.

NATURE AND UNDERLYING PROCESSES 

Definition and symptomatology 

The field of voice research currently lacks a universally accepted definition of vocal fatigue.2 Perhaps for this reason, vocal fatigue has been conceptualized in a variety of ways. Several commentators describe vocal fatigue as a symptom of a voice disorder with either a functional or an organic etiology.3., 4., 5. Others imply that vocal fatigue and associated compensatory behaviors may predispose phonotrauma and the development of laryngeal pathology.6., 7. Further, vocal fatigue as an isolated phenomenon has been considered in a number of empirical studies.8., 9., 10., 11., 12. In short, while a link between vocal fatigue and other laryngeal pathologies is plausible, it is unclear whether vocal fatigue primarily contributes to, results from, or exists independently of other voice conditions.

The definition of vocal fatigue employed here has been adapted from that offered by Scherer et al.9., 10. Vocal fatigue is used to denote negative vocal adaptation that occurs as a consequence of prolonged voice use. Negative vocal adaptation is viewed as a perceptual, acoustic, or physiologic concept, indicating undesirable or unexpected changes in the functional status of the laryngeal mechanism. In focusing on the short-term implications of prolonged phonation, this definition separates vocal fatigue from its possible relationship with other pathological voice conditions. This separation is deliberate, directed by the lack of data illuminating the precise position of vocal fatigue with respect to the overall spectrum of voice disorders.

Vocal fatigue has been described symptomatically by several authors. With few exceptions2., 10. reports are based on clinical anecdotes. Kostyk and Rochet,7 summarizing information provided by Titze13 and Scherer et al,10 listed 18 primary symptoms of vocal fatigue. These are presented in Table 1. Kitch and Oates2 used a retrospective questionnaire to investigate the perceptual experiences of actors and singers when vocally fatigued. Actors primarily reported an increased difficulty achieving adequate voice projection accompanying fatigue, whereas singers primarily reported a reduction in the pitch and dynamic range of the voice. Both groups indicated increased tension and discomfort in the throat, neck, and jaw regions. Interpreted as a whole, current literature suggests that vocal fatigue symptoms manifest in the following areas: voice quality, dynamic range, pitch range, respiratory support for phonation, level of muscular and structural tension/discomfort, vocal mechanism control, and level of vocal effort.

Table 1. Symptoms of Vocal Fatigue

	Hoarse/husky vocal quality	Running out of breath while talking
	Breathy vocal quality	Unsteady voice
	Loss of voice	Tension in neck/shoulders
	Pitch breaks	Throat/neck pain
	Inability to maintain typical pitch	Throat fatigue
	Reduced pitch range	Throat tightness/constriction
	Lack of vocal carrying power	Pain on swallowing
	Reduced loudness range	Increased need to cough/throat clear
	Need to use greater vocal effort	Discomfort in chest, cars, or back of neck

Reported from Kostyk and Rochet.7 © 1988 The Voice Foundation. Used with permission.

EXPERIMENTALLY INDUCED VOCAL FATIGUE IN HUMANS 

The majority of empirical studies addressing vocal fatigue have adopted a paradigm of measuring changes in vocal function under induced conditions of fatigue. Typically, participants with no history of vocal impairment are studied. This experimental approach has potential merit in helping to illuminate the onset and progression of fatigue as it affects the normal voice, and in determining which measures of vocal change are most indicative of fatigue. Pertinent studies are reviewed below. Findings are grouped into three categories: auditory perceptual and acoustic findings, laryngoscopic and videostroboscopic findings, and aerodynamic findings.

PATHOLOGICAL VOCAL FATIGUE AND OCCUPATIONAL VOICE DEMAND 

Each of the studies reported above has attempted to induce vocal fatigue in participants with no history of voice difficulties by systematically manipulating factors such as pitch, intensity, and duration of phonation. A central limitation of this experimental paradigm is its artificiality. While the variables investigated appear to place some degree of loading upon the vocal mechanism, it is unlikely that the multiplicity of factors that contribute to the clinical phenomenon of vocal fatigue are adequately captured within such an approach. Consequently, research methodologies that attempt to induce vocal fatigue in either the trained or untrained voice user require more realistically challenging experimental tasks, in addition to more sensitive indices of vocal change. In response to this need, a number of investigators have begun to address experimental fatigue questions in a more ecologically valid manner. This has included studying individuals with symptomatic vocal fatigue complaints, in addition to conducting experiments under realistic occupational demands, such as those presented by the teaching workday or vocal performance period.

Eustace et al39 reported acoustic, aerodynamic, and videostroboscopic data on 88 patients seen over a 4-year period with a primary complaint of vocal fatigue. Compared with normative data, participants were characterized by elevated airflow rates, reduced maximum phonation times, and the presence of anterior chinks, anterior and posterior chinks, and spindle-shaped vibratory closure patterns. Data collected on F0, F0 range, jitter, and airflow volume were all within normal limits. Buekers40 investigated the usefulness of a voice endurance test in discriminating patients with symptoms of vocal fatigue from participants with negative voice histories. Patients with vocal fatigue had been observed videostroboscopically to have mild vocal fold edema, posterior chinks, hour-glass, and spindle-shaped vibratory closure patterns, and in some instances vocal nodules. Nevertheless, following 30 minutes of excessive voice use (loud reading, coughing, singing in alternate registers, imitating voices, making animal noises), the two participant groups were not significantly different on electroglottographic or acoustic perturbation measures. It was concluded that short periods of excessive and unnatural phonation may not be representative of the prolonged vocal load typically experienced in the occupational setting.

Studies of vocal endurance conducted under various occupational conditions have typically revealed negative voice changes following prolonged phonatory demand. Rantala et al41 collected acoustic vowel tokens from 10 teachers across a typical working day. Analyses revealed significant changes in several spectral parameters of the vocal signal (reduced spectral tilt and increased high-frequency energy concentration). These changes were significantly correlated with the participants' self-report data. In addition, studies investigating the impact of prolonged occupational voice use in teachers with and without symptoms of vocal fatigue have documented significant differences between these groups using both auditory perceptual and aerodynamic methods.1., 7. Thus, it appears that the investigation of vocal function deviations within the occupational environment may be a useful approach in further understanding fatigue phenomena.

Investigations into the impact of vocal performance on the professionally trained voice user are scarce. Because of the physical and artistic demands of professional performance,14 and several anecdotal reports testifying to voice deterioration in individuals with poor vocal technique,42., 43. this is surprising. Froeschels44 documented incomplete glottal closure in singers following performance. As stated by Kitch et al,45 however, Froeschels' observations cannot be confidently attributed to singing demands as participants' larynges were not visualized before performance. Novak et al46 utilized acoustic measures to study vocal change in 45 professional actors following a stage performance. Measures of habitual F0 and long-term average spectra failed to consistently reflect changes in participants' self-reported level of vocal fatigue.

Kitch et al45 reported the only empirical study in the literature investigating the impact of performance on the singing voice. Vocal function data were collected from 10 tenors before and after a live choral performance. Participants sang the same repertoire simultaneously under identical performance conditions. Acoustic analyses (jitter, shimmer, HNR, F0 range, dynamic range) indicated negative vocal adaptation for the majority of participants, although several individuals demonstrated positive vocal adaptation following the performance. Interestingly, the most experienced vocal performer demonstrated the most consistent positive vocal adaptation across the evaluative tasks. HNR was reported to be the most sensitive index of vocal change. In contrast with the objective vocal function data, participants' self-ratings consistently indicated vocal improvement following performance.

The exploratory data reported by Kitch et al45 suggest that individuals may respond idiosyncratically to the demands of a vocal performance task. While negative vocal adaptation was documented for several performers, this trend was not consistent for all participants. As noted, individual differences in vocal training and performance experience may have contributed to the reported findings. Other potentially influencing factors, indicated by Kitch et al,45 were performance anxiety and the nature and extent of voice use prior to the performance. While participants were questioned regarding their voice use in the 24 hours prior to the performance, regulation of behavior over this period was not employed in the study. Additionally, while participants completed the performance task simultaneously, the collection of vocal function data was staggered over a 30 minute period following the performance. Because of this, participants may have experienced different degrees of vocal recovery before the postperformance measurements were taken.

SUMMARY AND CONCLUSIONS 

Titze's hypotheses13., 14. suggest that several mechanisms contribute to vocal fatigue. Of these, issues related to a potential increase in vocal fold tissue viscosity have received the most attention experimentally. Many of the mechanisms described by Titze are difficult to study in vivo, particularly in humans. While the relative importance of each purported mechanism is unknown, it is possible that individual anatomical and physiological idiosyncrasies, as well as task variables, may be equally significant in influencing the fatigue process.

Studies that have attempted to experimentally induce vocal fatigue in humans have yielded various and often conflicting results. While some trends appear evident, research in this area has generally been unable to document consistent responses to prolonged phonatory tasks. In some instances, methodological differences across studies may be responsible. Also significant may be a lack of sensitivity in the vocal function measures employed, as well as idiosyncratic physiological and psychological responses across participants when placed under an imposed vocal load. It has been hypothesized that select individuals may be able to make subtle functional adjustments in their vocal styles in order to counter the initial onset of fatigue.12 Whatever the underlying experimental constraints, the nature, progression, and salient identifying characteristics of vocal fatigue remain clouded.

Regardless of the discrepancies among studies, several guarded conclusions may be drawn from the human experimental literature. First, while inconsistent across participants, changes in glottal vibratory closure pattern appear to manifest after as little as 15 minutes of excessive phonatory behavior. Second, an elevation in habitual F0 has been noted following loud phonation prolonged over a period greater than one hour. Other acoustic data are difficult to reconcile across studies. Third, measurements designed to reflect phonatory threshold appear to shift negatively after at least 2 hours of extended voice use. Fourth, it appears that individuals with vocal training may be less negatively affected by prolonged excessive phonation than untrained individuals. As noted by Gelfer et al,8 more sensitive evaluative tasks may be required to track the onset of vocal fatigue in this population.

Occupational vocal loading research appears useful in exploring the real-life manifestation of vocal fatigue. There is a particularly critical shortage of data concerning the nature of vocal function changes following singing or acting performance. The clinical and pedagogical implications of research in this area are highly significant, especially considering the high prevalence of fatigue-related complaints among the performing population.2 Critical questions requiring investigation include the following: What amount/nature of voice use during a performance can be considered safe? Are negative voice changes across a performance inevitable, or can they be avoided or attenuated with appropriate vocal training and technique? Can individual susceptibility to vocal fatigue vary as a function of performance style or mode? Can the presence of a warm-up effect be isolated from a fatigue effect within performance? Further, can appropriate vocal warm-up attenuate or delay the onset of vocal fatigue?

DIRECTIONS FOR FUTURE RESEARCH 

Vocal fatigue remains a fertile area for ongoing research. First, careful clinical research is required to unravel the precise relationship between fatigue and other voice disorders. Second, laboratory studies are needed to further examine the physiological and biomechanical mechanisms that appear to underlie fatigue. Titze's13., 14. hypotheses should continue to be evaluated. Some mechanisms, such as reduced blood circulation and non-muscular tissue strain, may be best explored using physical and computer modeling. Others, such as neuromuscular fatigue and changes in vocal fold viscosity, are more amenable to investigation using animal models and humans. Third, clinical and mechanistic studies should be complemented by occupational research conducted under ecologically valid conditions. While methodologically challenging, research of this nature is closely representative of everyday vocal demand.

Research questions and experimental designs should accommodate probable individual differences in response to occupational and other voice demands. By further understanding the factors that promote individual susceptibility to vocal fatigue, clinicians can better advise in issues relating to voice care. Variables of interest include age, gender, nature and degree of voice training, nature and degree of occupational voice demand, vocal hygiene, and history of voice problems. Issues related to gender and voice training have begun to be explored.8., 9., 10., 47., 48.

Psychological aspects of vocal fatigue have received little attention in the literature, however progress in this area is required to gain a complete understanding of the vocal fatigue phenomenon. Important areas for research include the relationship between psychological fatigue and vocal fatigue, variations in sense of effort with respect to vocal fatigue onset and progression, and the importance of different psychological pacing strategies in achieving a vocal endurance target. Improved understanding of the psychological constructs and processes involved in vocal fatigue may help to partially explain individual differences in fatigue susceptibility.

Besides examining the underlying mechanisms and salient features of vocal fatigue, research must address treatment issues. No empirical research exists specific to the treatment of vocal fatigue. While Solomon and DiMattia's11 work provides some support for hydration as a useful management strategy, it is unknown what forms of voice therapy or vocal endurance training might function to delay, attenuate, or prevent fatigue; or indeed how intervention should be tailored for different individuals. Progress in this area is critical to the appropriate management of all voice users experiencing vocal fatigue difficulties.