J. FLUENCY DISORD. 10 (1985), 317-324
STUTTERING AND THE VALSALVA MECHANISM: A HYPOTHESIS IN
NEED OF INVESTIGATION
WILLIAM D. PARRY
Member of the Philadelphia, Pennsylvania Bar and the National Stuttering
Reprinted from the December 1985 issue of the Journal of Fluency
Copyright © 1985 by Elsevier Science Publishing Co.,
Inc., 52 Vanderbilt Ave., New York, NY 10017.
This note presents the hypothesis that the excessively forceful closure of
the mouth or larynx associated with certain types of stuttering may involve
a neurological confusion between speech and the human body's Valsalva mechanism,
which is designed to increase pulmonary pressure by forceful closure of the
upper airway to assist in many types of physical effort. It is suggested
that such behavior may result from excessive neuromotor tuning of the Valsalva
mechanism prior to speech, especially in situations where the stutterer
anticipates the need to "try hard" to speak properly. Such tuning may both
interfere with phonation and render the Valsalva mechanism overexcitable
to triggering stimuli, such as the increase in subglottal pressure that
accompanies the start of articulation.
I am an attorney who has spent most of his life attempting to resolve a severe
stuttering problem. Over the years I have tried a wide range of treatments,
including elocution lessons, attitude therapy, psychotherapy, psychoanalysis,
metronome training, biofeedback, and airflow therapy - none of which provided
satisfactory long-term relief. Since my termination of therapy several years
ago, I continued to explore the problem on my own, through extensive review
of the existing literature, personal experimentation, and, more recently,
consultations with professional researchers in the fields of laryngology
and speech pathology.
My stuttering was primarily characterized by some form of excessively forceful,
prolonged, or repeated closure of the upper airway, at the level of either
the mouth or the larynx. The stuttering of initial vowel sounds involved
tight closure of the larynx, including both the vestibular and the vocal
folds, such as occurs during effort closure. These laryngeal closures were
subsequently observed through a nasally inserted, fiberoptic nasopharyngoscope
by Dr. Joseph P. Atkins, Jr., Chief of Otolaryngology at the Pennsylvania
Hospital in Philadelphia, and video recorded.
During severe blocks, my mouth or larynx would clamp shut, while my abdominal
and thoracic muscles would contract, as if straining to force air through
the closure, which, in turn, seemed to tighten more than ever. As I studied
my behavior more carefully, I noticed that my stuttering (even when less
severe) was preceded and accompanied by tension in certain muscles in and
around the larynx, in the abdomen, and in the rectum, as well as difficulties
or delays in initiating phonation. When I consciously relaxed my abdominal
and rectal muscles, phonation became easier and stuttering was substantially
As a result of my personal research, I have formulated a hypothesis as to
my own problem, based on a possible relationship between stuttering and the
Valsalva mechanism - a powerful alliance of laryngeal, respiratory, abdominal,
and rectal muscles that are often activated to create pulmonary pressure
to assist in many types of physical effort, as well as defecation. As shall
be discussed below, the involvement of this mechanism in certain types of
stuttering behavior might help to explain a stutterer's involuntary use of
excessive force in speaking, and why stuttering is more severe in situations
where the person feels the need to "try hard" to "force out" the words.
Although I have found this hypothesis to be my own key to fluency, I recognize
that its general validity has yet to be established. Indeed, the possible
involvement of the Valsalva mechanism seems to have been totally ignored
by the existing literature. Because I do not have the necessary laboratory
facilities at my disposal, I am therefore writing this note for the limited
purpose of stimulating empirical research into this question.
THE VALSALVA MECHANISM
The term Valsalva mechanism (after the Italian anatomist Antonio Maria Valsalva,
1666-1723) refers to those muscles that are neurologically coordinated for
simultaneous contraction in the performance of a Valsalva maneuver, which
may occur during many kinds of physical exertion (Carlson, Johnson and Cavert,
1961; Ardran and Kemp, 1967; Fink, 1973, 1975; Fink and Demarest, 1978).
Usually such a maneuver involves tight closure of the vocal and vestibular
folds of the larynx (called effort closure) for the purpose of sealing off
the respiratory tract, accompanied by contraction of the parietal muscles
of the abdomen and thorax (Fink, 1975; Fink and Demarest, 1978).
The resulting increase in pulmonary pressure can serve two important functions:
(1) bracing and stiffening the trunk of the body, forming a base for the
limbs to perform strenuous tasks and (2) permitting abdominal pressure to
be used more efficiently to expel visceral contents, as in defecation, urination,
and childbirth (Ardran and Kemp, 1967; Fink, 1975; Fink and Demarest, 1978;
Dickson and Maue-Dickson, 1982). When the straining effort is not for the
purpose of defecation, the anal sphincter is closed, and closure is reinforced
by contraction of the puborectalis muscle, to help prevent evacuation (Fink,
The various elements of the Valsalva mechanism appear to be linked neurologically
so that they tend to contract simultaneously. For example, electromyographic
studies have indicated that the thyrohyoid muscle is automatically excited
or activated in proportion to the contraction of the abdominal wall muscles
and the resulting increase in sublaryngeal pressure (Fink, 1975; Fink and
Demarest, 1978). Abdominal straining is always associated with complete closure
of the larynx in young, normal people (Ardran and Kemp, 1967). Likewise,
contraction of the rectal muscles will also stimulate a reflexive closure
of the larynx-a fact that could be visually observed during the fiber optic
study performed on me by Dr. Atkins. This is not meant to imply that straining
and defecation cannot occur without effort closure, or that persons cannot
voluntarily refrain from closing the larynx in such instances (Ardran and
THE VALSALVA HYPOTHESIS
The hypothesis suggested by this note may be briefly summarized as follows:
1. Certain types of stuttering behavior may be preceded by excessive
"tuning" of the neural pathways to the muscles of the Valsalva
Research into muscular activity has indicated that the onset of movement
is preceded by preliminary tuning of the excitability of all participating
sensory and motor elements in accordance with the motor program for the intended
act (Zimmerman, Smith, and Hanley, 1981). Such tuning either promotes or
inhibits the excitability of the various motoneuron pools, and thereby enables
groups of muscles to act as functional units, called coordinative structures,
which are activated by the appropriate triggering signals. In the case of
a stutterer, the degree of such tuning might be influenced by psychological
and emotional factors, and would more likely occur in those instances when
the stutterer feels a particular need to "try hard" to speak.
2. The tuning of the Valsalva mechanism would increase the excitability
of those muscles to stimuli that would trigger a Valsalva maneuver.
Such stimuli might (but not necessarily) include the abrupt increase in
subglottal pressure that usually occurs at the beginning of articulation
(Ladefoged, 1968). The pressure increase would be sensed by mechanoreceptors
found in the subglottic portion of the laryngeal mucous membrane (Wyke, 1971)
and transmitted to the central nervous system. These stimuli might then be
misinterpreted as indicating the beginning of a Valsalva maneuver, resulting
in a responsive triggering of the Valsalva mechanism. The reported reduction
of stuttering by methods that emphasize a "gentle onset" of voice and
articulation (e.g., Perkins, 1973; Webster, 1974) might therefore be attributed
to the avoidance of abrupt increases in subglottal pressure.
3. The triggering impulses might be sent to the muscle groups comprising
the Valsalva mechanism (as well as to the oral structures that caused the
closure of the upper airway) to stimulate their simultaneous contraction,
so as to continue the increase of subglottal pressure.
These contractions might include effort-type closures of the larynx during
the stuttering of initial vowel sounds, as were observed in the fiber optic
study performed on me by Dr. Atkins. Van Riper (1973) has also described
such closures as consisting of occlusion at the level of the larynx with
both the vocal folds and ventricular folds closed tightly. However, effort
closure of the larynx would not necessarily occur when closure of the upper
airway is already being accomplished by the various oral structures, such
as by the lips and the tongue in the articulation of certain consonants.
Such lack of effort closure would be consistent with the behavior of the
larynx observed in the fiber optic study when I used my lips for upper airway
closure during a Valsalva maneuver. (Of course, mere visual observation might
not reveal the full extent of laryngeal muscular activity during such a
maneuver.) Therefore, the validity of the Valsalva hypothesis would not require
visible effort closure of the larynx during stuttering that involves oral
My hypothesis proposes, instead, that the oral structures (primarily those
whose closure had initiated the rise in air pressure), might be neurologically
recruited to assist the Valsalva mechanism. During both stuttering and straining,
I have found the overall sensation to be the same, regardless of whether
closure is at the oral or laryngeal level. Wyke (1971) has stated that an
increase in subglottal air pressure may reflexively induce spasm of the laryngeal
and oral musculature.
4. As subglottal pressure rises, the oral or laryngeal structures
blocking the upper airway would be stimulated to close with proportionately
greater force, so as to resist the increased pressure.
In this way, the increased subglottal pressure would serve as a continuing
stimulus for the contraction of the Valsalva muscle groups, creating a vicious
cycle that would render fluent speech impossible for as long as the maneuver
continues. The stutterer may also adopt various secondary behaviors in an
attempt to terminate or avoid such blocks, including inspiratory gasps and
attempts to begin speech with little or no available air in the lungs (Van
Riper, 1971, 1973; Adams, 1974). Such behaviors may themselves interfere
with phonation in other ways, such as by failing to provide sufficient airflow
to sustain vibration of the vocal folds.
VALSALVA TUNING AND PHONATION
The possible relationship between difficulty in phonation and stuttering
has been suggested by several researchers (e.g., Wingate, 1969, 1976; Wyke,
1971; Adams, 1974; Perkins, et al., 1976). This difficulty might be explained,
at least in part, by the possibility that the neuromotor tuning of the Valsalva
mechanism may interfere with the larynx's readiness to produce phonation,
by preparing it instead to take part in a Valsalva maneuver.
EMG studies have observed neuromotor tuning occurring in the intrinsic laryngeal
muscles some 0.05 to 0.55 sec prior to audible phonation (Wyke, 1971). However,
tuning of the Valsalva mechanism might supersede the normal prephonatory
tuning, because the laryngeal muscles would become set for their role in
effort closure rather than phonation. That is, the laryngeal muscles might
not be prepared to place the vocal ligaments in the proper position, to give
them the necessary tension, or to permit them to vibrate in the manner required
for phonation (Fink and Demarest, 1978). This might help to explain, at least
in part, the delay in vocal reaction time and voice onset that has been observed
in stutterers (Adams and Reis, 1971; Adams, 1974; Starkweather, Hirschman,
and Tannenbaum, 1976; Starkweather, Franklin, and Smigo, 1984).
Because a Valsalva maneuver does not normally occur during inspiration, a
stutterers' vocal reaction time might be expected to improve when inspiratory
phonation is used. In a study by Reich, Till, and Goldsmith (1981), the
stutterers' average reaction time was 208.2 msec for inspiratory phonation
and 236.8 for expiratory phonation of a vowel sound. The corresponding reaction
times for nonstutterers were 182.1 and 177.0 msec, respectively. Although
the stutterers' reaction times were somewhat slower than those of nonstutterers
as to every task, whether manual or laryngeal, the stutterer-nonstutterer
difference was more than twice as great for expiratory phonation than for
If tuning of the Valsalva mechanism does, in fact, interfere with the phonatory
function of the larynx, then, conversely, it would seem to follow that tuning
of the larynx for phonation might tend to reduce the excitability of the
Valsalva mechanism in general. This effect may help to explain the reported
reduction of stuttering when vocalization is emphasized (Wingate, 1969, 1970,
1976; Weiner, 1978).
As Henry Freund (1966) has written, one of the primary experiences of stuttering
is that of "an obstacle which needs force to overcome it"; however, such
effort by the stutterer "only increases the force of the closure." In my
own case, I have found the Valsalva hypothesis to be an effective solution
to this paradox. I believe it highly plausible that, as a 4-yr-old child
experiencing anxiety, emotional conflict, and perhaps some simple disfluency
in expressing himself, I may have unconsciously sought help from the Valsalva
mechanism, which had previously served me well in defecation and other types
of effort, and thereby established my stuttering behavior.
This behavior may have been reinforced by my perception that the increased
physical effort was ultimately successful in forcing the words out (Freund,
1966). Activation of the Valsalva mechanism may have achieved certain
psychological gains as well. For example, I may have felt that this display
of effort would demonstrate to adults "how hard I was trying" to please them,
thereby reducing the danger of punishment. The Valsalva hypothesis would
seem to provide a common nexus between the psychological and the physiological
components of stuttering. It would seem to explain why stuttering is more
frequent at the beginnings of words and on stressed syllables (Froeschels,
1961; Soderberg, 1966), instances in which abrupt increases in subglottal
air pressure are probably most likely. More important, it would help to explain
some of the psychosocial and linguistic variables in stuttering. As summarized
by Starkweather (1982), stuttering is more frequent in content words and
meaningful speech than in function words and non-meaningful speech, and is
more likely to occur at points of high uncertainty or high information load,
precisely those instances when a speaker might tend to use increased effort
THE NEED FOR RESEARCH
A determination of the validity of the Valsalva hypothesis in explaining
certain types of stuttering behavior will require broad-based empirical research.
Therefore, I would urge researchers to consider the question of Valsalva
mechanism involvement when designing future experiments.
For example, it may be helpful if future EMG studies of muscular activity
during stuttering would include some of the significant laryngeal, abdominal,
and rectal muscles associated with the Valsalva maneuver. Biofeedback experiments
might also test whether stuttering is affected by relaxation of muscles in
various parts of the Valsalva mechanism.
I wish to express my most sincere thanks to Dr. Joseph P. Atkins,
Jr., of the Pennsylvania Hospital Department of Otolaryngology, for his advice
and assistance in this matter, and to the Journal of Fluency Disorders
for allowing me the opportunity to share these ideas.
Adams, M.R. A physiologic and aerodynamic interpretation of fluent and stuttered
speech. Journal of Fluency Disorders, 1974, 1, 35-47.
Adams, M.R. and Reis, R. The influence of the onset of phonation on the frequency
of stuttering. Journal of Speech and Hearing Research, 1971, 14,
Ardran, G.M. and Kemp, F.H. The mechanism of the larynx: 11. The epiglottis
and closure of the larynx. British Journal of Radiology, 1967, 40,
Carlson, A.J., Johnson, V., and Cavert, H.M. The Machinery of the Body.
Chicago: University of Chicago Press, 1961.
Dickson, D.R. and Maue-Dickson, W. Anatomical and Physiological Bases
of Speech. Boston: Little, Brown, 1982.
Fink, B.R. The curse of Adam: Effort closure of the larynx.
Anesthesiology, 1973, 39,325-327.
Fink, B.R. The Human Larynx: A Functional Study. New York: Raven
Fink, B.R. and Demarest, R. Laryngeal Biomechanics. Cambridge, MA:
Harvard University Press, 1978.
Freund, H. Psychopathology and the Problems of Stuttering. Springfield,
IL: Charles C. Thomas, 1966.
Froeschels, E. New viewpoints on stuttering. Folia Phoniatrica,
1961, 13, 187-201.
Ladefoged, P. Linguistic aspects of respiratory phenomena. Annals of
New York Academy of Sciences, 1968, 155, 141-151.
Perkins, W.H. Replacement of stuttering with normal speech: 11: Clinical
procedures. Journal of Speech and Hearing Disorders, 1973, 38, 295-303.
Perkins, W., Rudas, J., Johnson, L., and Bell, J. Stuttering: Discoordination
of phonation with articulation and respiration. Journal of Speech and
Hearing Research, 1976, 19, 509-522.
Reich, A., Till, J., and Goldsmith, H. Laryngeal and manual reaction times
of stuttering and nonstuttering adults. Journal of Speech and Hearing
Research, 1981, 24, 192-196.
Soderberg, G. The relations of stuttering to word length and word frequency.
Journal of Speech and Hearing Research, 1966, 9, 584-589.
Starkweather, C.W. Stuttering and laryngeal behavior: A review. ASHA
Monographs, 1982, 21, 1-45.
Starkweather, C.W., Franklin, S., and Smigo, T.M. Vocal and finger reaction
times in stutterers and nonstutterers: Differences and correlations.
Journal of Speech and Hearing Research, 1984, 27, 193-196.
Starkweather, C.W., Hirschman, P., and Tannenbaum, R.S. Latency of vocalization:
Stutterers vs. nonstutterers. Journal of Speech and Hearing Research,
1976, 19, 481-492.
Van Riper, C. The nature of stuttering. Englewood Cliffs, NJ: