EFFECTS OF EMG-ACTIVATED ALARMS
ON NOCTURNAL BRUXISM
JEFFREY E. CASSISI
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
The author would like to thank his wife, mother, father
and entire family for their unwaivering support,
encouragement, and confidence. This work is dedicated to
Emil F. Ersay, M.D., who continues to inspire a long line of
This research could not have taken place without the
electrical-engineering expertise of James "Buddy" Lee. He
is one of the University of Florida's finest assets and
illustrates more than anyone that knowledge and
understanding are their own reward.
The author also expresses appreciation to his chairman,
F. D. McGlynn, Ph.D. This research could not have taken
place without him either. His knowledge and experience in
research design and methodology were indispensable. The
other members of the supervisory committee were extremely
helpful too. They contributed many suggestions which
enhanced the execution and interpretation of this
The Dental Occlusion and Facial Pain Center under the
direction of Parker E. Mahan, D.D.S., Ph.D, has greatly
facilitated many aspects of this research. The staff in
that clinic are truly in the business because they want to
help people. Michael Henry, D.D.S., was always ready to
examine a new potential subject and he gave each one
Charles Gibbs, Ph.D. answered many questions about
muscle physiology early in the study. Scott Vrana and John
Steele helped with computer programming. Dale Belles and
Larry Kapel assisted with editing. Marc Wruble asked some
Many people are due thanks for being there when they
were needed: Robert and Connie Head, Michael Jamerson, Rade
Musulin, Carol Cornell, Laura Mee, Lynn Harllee, Charles
Bichajian, Alan and Elizabeth Dougherty, Lynn Cornell, and
Tom and Iris Spikes.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS .................................... ii
LIST OF TABLES ................... ....... ......... vi
LIST OF FIGURES ...................................... viii
ABSTRACT .............................................. ix
INTRODUCTION .......................................... 1
ETIOLOGICAL THEORIES OF BRUXISM ....................... 3
Dental Theories ..................................... 5
Psychological Theories .............................. 6
Interactive Theories ............... .............. 7
Sleep Theories ...................................... 8
THE ASSESSMENT OF BRUXISM ............................ 10
Dental Signs ....................................... 10
Intraoral Telemetry ................................. 11
Audible Grinds .................................... 12
Bruxscore Monitor ................................... 13
Electromyography ................................... 14
AVAILABLE TREATMENTS FOR NOCTURNAL BRUXISM ............ 18
Dental Therapies .................................... 18
Psychological Therapies ............................. 20
EMG-ACTIVATED ALARM THERAPY ........................... 24
Early Research ..................................... 24
Contemporary Research ............................... 25
PURPOSES OF THE EXPERIMENT ........................... 43
METHODS ............................................... 47
Subjects ............................................ 47
Instrumentation ................................... 51
Measures .......................... ...... ......... 55
Procedure and Experimental Protocal ................ 60
RESULTS ................................. ............ .
Impact of Treatment .................................
Evidence Concerning Rebound Effects .................
Self-Monitored Versus Automated Alarm Count .........
Effects of Treatment on Vigor, Fatigue,
Sleepiness, and Pain ............................
Impact of Treatment ................................
Evidence Concerning Rebound Effects .................
Self-Monitored Versus Automated Alarm Count .........
Effects of Treatment on Vigor, Fatigue, and
Sleepiness ....... .. ... ..... .......... ...........
Effects of Treatment on Facial Pain .................
Theoretical Explanations for Alarm Therapy
Effects ................................ ..
A SUBJECT RECRUITMENT NOTICE .............
B STANDARDIZED DENTAL SCREENING EXAM .....
C SCREENING QUESTIONNAIRE ................
D INSTRUCTIONS FOR USING THE BRUX-MONITOR ..
E SCHEMATIC OF ALARM/MONITOR .............
F MORNING QUESTIONNAIRE ..................
G NIGHTLY ALARM RECORD .....................
H NIGHTTIME QUESTIONNAIRE ..................
I INSTRUCTIONS FOR USING THE BRUX-ALARM ....
J SUMMARY TABLES FOR ALL ANOVAs ............
K GRAPHS OF NIGHTLY BRUX RESPONSE FREQUENCY PER
HOUR FOR SUBJECTS 1 THROUGH 10 .................
REFERENCES ................................... ...
BIOGRAPHICAL SKETCH .................................
o o o
.... .. .
LIST OF TABLES
1 Subject Demographics ........................... 49
2 Means and Standard Deviations for the Nightly
Brux Responses per Hour by Group and Phase ... 67
3 Mean Self-Monitored Alarm-Count Versus
Automated Alarm-Count ...................... 71
4 Means and Standard Deviations for Daily
Ratings of Fatigue, Vigor, Sleepiness,
Tension, and Pain by Group ................... 72
J-1 Summary Table for the 2 x 3 ANOVA for Hourly
Bruxing Frequencies within the Screening ..... 108
J-2 Summary Table for the 2 x 2 x 14 ANOVA for
Hourly Bruxing Frequencies across 28 Nights .. 109
J-3 Summary Table for the 2 x 14 ANOVA for Hourly
Bruxing Frequencies within Phase 1 ........... 110
J-4 Summary Table for the 2 x 14 ANOVA for Hourly
Bruxing Frequencies within Phase 2 .......... 111
J-5 Summary Table for the 2 x 14 ANOVA for Hourly
Bruxing Frequencies for both Treatment
Phases ....................................... 112
J-6 Summary Table for the 2 x 14 ANOVA for Hourly
Bruxing Frequencies for Group 1 .............. 113
J-7 Summary Table for the 2 x 14 ANOVA for Hourly
Bruxing Frequencies for Group 2 .............. 114
J-8 Summary Table for the 2 x 14 ANOVA for Hourly
Bruxing Frequencies for both Baselines ....... 115
J-9 Summary Table for the 2 x 14 ANOVA for Method
of Alarm Count ............................... 116
J-10 Summary Table for the 2 x 2 x 14 ANOVA for
Vigor Ratings Across 28 Nights ............... 117
J-11 Summary Table for the 2 x 2 x 14 ANOVA for
Fatigue Ratings Across 28 Nights ............. 118
J-12 Summary Tables for the 2 x 2 x 14 ANOVA for
SSS Ratings for 28 Days ...................... 119
J-13 Summary Table for the 2 x 2 x 14 ANOVA for
Facial Pain/Discomfort Ratings for 28
Nights ..................................... 120
J-14 Summary Table for the 2 x 2 x 14 ANOVA for
Anxiety Ratings for 28 Nights ............... 121
LIST OF FIGURES
1 Flow-chart of the Alarm/Monitor ................. 53
2 Overview of the Experimental Design ............. 62
3 Brux Frequency by Condition and Group ........... 68
K-1 Brux Episode Frequency Per Night
For Subject 1 ............................... 123
K-2 Brux Episode Frequency Per Night
For Subject 2 ................................ 124
K-3 Brux Episode Frequency Per Night
For Subject 3 ............................... 125
K-4 Brux Episode Frequency Per Night
For Subject 4 ......................... ..... 126
K-5 Brux Episode Frequency Per Night
For Subject 5 ......................... ........ 127
K-6 Brux Episode Frequency Per Night
For Subject 6 ....................................... 128
K-7 Brux Episode Frequency Per Night
For Subject 7 ................................. 129
K-8 Brux Episode Frequency Per Night
For Subject 8 ............................... 130
K-9 Brux Episode Frequency Per Night
For Subject 9 ............................... 131
K-10 Brux Episode Frequency Per Night
For Subject 10 .............................. 132
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
EFFECTS OF EMG-ACTIVATED ALARMS
ON NOCTURNAL BRUXISM
JEFFREY E. CASSISI
Chairperson: F. Dudley McGlynn
Major Department: Clinical Psychology
Nocturnal bruxism is nonfunctional grinding and/or
clenching of the teeth during sleep. This behavior has been
directly implicated in various facial pain and
temporomandibular joint (TMJ) disorders. The present study
examines the effectiveness and contraindications of one
psychological treatment for bruxism, namely EMG-activated
Modification of bruxism is attempted ideally by a
dentist-psychologist team. Therefore, selected theories
related to dental and psychological practice are first
reviewed. Then alternative strategies for the assessment
and treatment for the problem are identified.
The overview of alarm treatment suggests that it might
be a relatively beneficial approach. The present study
tested the effectiveness of an alarm therapy protocol in
which alarm termination required wakeful operation of a
push-button switch. Ten heavy bruxers, as determined by a
dental screening and ambulatory monitoring, volunteered for
the 28 day study and were paid for participation. They were
divided into two groups that differed in sequencing of
baseline and treatment phases.
Comparisons both within and between groups revealed
that treatment significantly reduced bruxes per hour. Some
evidence suggested that treatment had an enduring effect.
Unobtrusive counts of alarm soundings during the night
were compared with subjects self-monitored tallies and they
were found to be significantly higher. This finding
suggests that self-monitoring is not an adequate method for
assessing nightly alarm sounding frequency.
Ratings of anxiety, fatigue, vigor, and sleepiness,
were made each evening. No evidence was found that alarm
therapy affected fatigue or sleepiness ratings. No evidence
was found that alarm therapy decreased rated vigor.
However, a significant increase in anxiety ratings was
detected during treatment phases.
The conclusion is offered that alarm therapy requiring
manual alarm termination is an effective treatment approach.
The conclusion is offered also that informed consent
requires information concerning the possible side-effect of
increased psychological tension. Future research avenues
Recently there has been growth in the health-care
liaison between the professions of dentistry and psychology.
The dental community has encouraged the liaison by granting
support for behavioral science research in dentistry, by
publishing results from such research in dental journals,
and by hiring psychologists to participate in dental
education (Cohen, 1977; Sachs, Eigenbrode & Kruper, 1979).
Psychologists, in turn, have facilitated the liaison by
undertaking behavioral research related to dentistry, by
publishing findings in both dental and psychological
journals, and by adding content as well as methods to dental
curricula (Cohen, 1977; Page, 1978).
The continuing relationship between dentistry and
psychology has resulted in the emergence of "behavioral
dentistry," a special field of subject matter devoted to
areas such as promoting oral-hygiene self-care behaviors
(cf. Claerhout & Lutzker, 1981; Iwata & Becksfort, 1981;
McGlynn, Mings, Marks, & Goebel, 1985), and evaluating and
treating fear of dental settings (cf. Kleinknecht, McGlynn,
Thorndike, & Harkavy, 1984; Melamed, 1979). Substantial
research effort within behavioral dentistry also has been
devoted to evaluating and treating the various disorders of
craniomandibular articulation, e.g. the so-called
temporomandibular joint (TMJ) and myofascial pain
dysfunction (MPD) "syndromes" (Mealiea & McGlynn, 1986;
Moss, Garrett, & Chiodo, 1982; Rugh & Solberg, 1985; Scott,
1980). A closely related problem that has received renewed
research interest is nocturnal bruxism (Glaros & Rao, 1977;
McGlynn, Cassisi, & Diamond, 1985).
The experiment reported here is concerned with the
treatment of bruxism and facial pain. In the experiment an
attempt was made to reduce the frequency of nocturnal
bruxist behavior with the use of a portable biofeedback-like
device called a wake-up alarm. By way of introducing the
experiment, the following narrative overviews the various
theories concerning the etiology of nocturnal bruxism as
well as various assessment and treatment approaches for the
problem. Following this, the narrative reviews research
using EMG-activated wake-up alarms. The early sections of
the introduction are adapted from various sources to which
the author has contributed including Cassisi, McGlynn, and
Belles (1986) and McGlynn et al. (1985).
ETIOLOGICAL THEORIES OF BRUXIST BEHAVIOR
Bruxism is defined most frequently as a class of
oral-motor behaviors that includes nonfunctional or
parafunctional clenching, grinding, and gnashing of the
teeth (cf. Dubner, Sessle, & Storey, 1978; Scharer, 1974).
Bruxism is significant to health-care professionals because
it is widespread (cf. Scharer, 1974) and because its
symptomatic correlates frequently prompt help seeking.
Common among these symptomatic correlates are abnormal tooth
wear, damage to the temporomandibular joint (TMJ), various
facial pains, referred pains, and headache (Alling & Mahan,
1977; Ramfjord & Ash, 1983).
Frohman (1931) first used the English word bruxism. It
was derived from the French "la bruxomanie" used by Marie
and Pietkiewicz (1907). Bruxism has subsumed several
nonfunctional behaviors of the masticatory system and
uniform standards for defining and detecting bruxism have
not been used. Hence, there is extreme variability in
existing estimates of the incidence of the behavior (Glaros
& Rao, 1977). Some amount of bruxing or nonfunctional tooth
contact is normal during sleep (Powell, 1965) and many
authorities believe that all people engage in transient
bruxing at some time.
The various health-care disciplines involved with
bruxist patients have undertaken etiological research and
have brought to bear their own concepts in explaining the
behavior. Therefore, a large literature on empirical
correlates of bruxism exists and there are several "theories
of bruxist behavior." An exhaustive review of the various
theories is not presented here and no attempt at theoretical
arbitration is made. This follows from several
considerations. First, a comprehensive summary of the very
complex theoretical state of affairs is beyond the present
scope. Second, the empirical knowledge is limited and
support of particular theoretical formulations is weak.
Third, activities of diagnosis and treatment do not derive
from theoretical considerations, except at a general level.
Nadler (1957) arranged the correlates of bruxism into
four categories that included local/dental,
psychological/emotional, systemic, and occupational factors.
Glaros and Rao (1977) organized theories of bruxism into
three categories that subsumed local/mechanical,
psychological, and systemic/neurophysiological factors.
Scharer (1974) organized theories of bruxism into three
categories that involved neural, internal, and external
factors. These expository schemes impose categories that do
not exist as actually separate clusters of causes. Bruxism
is a behavior that results from multiple, interacting causes
that cut across these domains of variables (Rugh, 1976).
The categories below were used solely for purposes of
Some early dentists wrote that bruxism is caused by
anatomical factors such as missing or elongated teeth.
Contemporary views have shifted toward occlusal interference
as an etiological factor. There are various opinions about
which kinds of occlusal interference produce bruxing
(Carlsson & Droukas, 1984), but most dentists would agree
that "any occlusal interference may trigger bruxism"
(Ramfjord & Ash, 1983, p. 181).
There are reports of experiments in which various
occlusal interference were produced experimentally while
bruxing was monitored concurrently. For instance, Barghi,
Rugh, and Drago (1981) recorded nocturnal masseter activity
among five adult humans before, during, and after 10-15
days, during which a lateral deflection of the mandible was
produced with a crown. However, the results from studies of
this type have been inconclusive. Barghi et al. (1981), for
example, found that the nocturnal EMG activity of two
subjects dropped, one increased, and two remained unchanged
when the crown was in place. Hence, the consensus among
dentists that occlusal interference trigger bruxing is
Psychological theories of bruxism are derived from two
intellectual traditions. One is the tradition of
personology or personality theories. The other is the
tradition of psychosomatics.
Psychoanalytic writers have suggested that bruxism is a
result of unconscious mental forces. For example, bruxism
might reflect repressed oral aggression (Forsberg, 1956;
Pond, 1968; Ramadan, 1970; Shapiro & Shannon, 1965; Walsh,
1965), or attempted gratification of oral needs (Frohman,
1931, 1932; McCartney, 1951; Ross, Bentley, & Greene, 1953;
Sumner, 1949). It is not scientifically productive to view
bruxing as related to unconscious mental forces, because the
psychodynamic theories are not disconfirmable.
There is a large literature on personality traits among
patients who suffer from craniomandibular disorders related
to bruxing. Some authors have described these patients as
passive-dependent and frustrated, while others have
described them as perfectionistic, domineering, and striving
to appear normal (Lefer, 1972; Moulton, 1955, 1966). Recent
investigations with the Minnesota Mutiphasic Personality
Inventory (MMPI) suggest that many patients with
craniomandibular disorders show conversion or depression
profiles (cf. Eversole, Stone, Matheson, and Kaplan, 1985).
For at least three reasons, however, the value of
personality-trait measures in understanding bruxism remains
in doubt. There is no agreement in the literature regarding
the nature of abnormal personality as it occurs in bruxist
populations. Studies have not found consistent personality
differences between craniofacial pain patients and "normal"
controls (Reding, Zepelin, & Monroe, 1968; Solberg, Flint, &
Brantner, 1972). Relevant studies have used subjects with
various craniomandibular disorders and only some of them
Bruxism can be conceptualized as a psychophysiological
or stress-related disorder. The following evidence is
required to confirm this view: (a) that stress and negative
emotions can lead to exaggerated activity in the muscles of
mastication; (b) that bruxers manifest greater masseter EMG
under stress than do nonbruxers; and (c) that there is a
correlation between stressful events and bruxing.
All three requirements are supported by some empirical
evidence. Yemm (1968, 1969, 1971, 1972) has demonstrated
that anxiety and frustration can produce increased masseter
and temporalis muscle activity. Rao and Glaros (1979) showed
that bruxists had higher baseline masseter EMG levels and
greater masseter EMG response to experimental stress than
did nonbruxists. Rugh and Solberg (1976) have used
ambulatory EMG monitoring devices to show that daytime
stress correlates with nocturnal bruxing.
While many dentists believe that occlusal interference
can produce bruxing (Christensen, 1970; Lindquist, 1972;
Ramfjord, 1961), it is known that some bruxers have no
interference and that some patients with interference do
not brux (Olkinura, 1969; Robinson, Reding, Zepelin, Smith &
Zimmerman, 1969; Schwartz, 1958). Accordingly it has been
argued that occlusal interference participate as
interacting factors in complex multifactor etiologies
leading to the behavior. Ramfjord and Ash (1983) for
example, suggested that "a variable tolerance level to
occlusal interference exists and that this tolerance level
may be altered by psychic stress affecting the tonus
activity in the jaw muscles" (p.81).
Only recently have researchers attempted to study
interference in interaction with other variables that
influence bruxing. Evidence from one controlled experiment
suggests that occlusal factors can interact with situational
stress to influence diurnal clenching (Bichajian, 1984).
However, the multifactor theory of bruxism needs massive
empirical fleshing out to be of significant value.
Laboratory research has provided several hypotheses
about relationships between bruxing and sleep activities.
Reding, Rubright, Rechstaffen, and Daniels (1964)
hypothesized that bruxing is associated with dreaming.
Satoh and Harada (1973) hypothesized that bruxing is
associated with arousal during sleep and with transitions
toward lighter sleep stages. There are several reports
suggesting that bruxing is associated with periods of body
movement and rapid-eye-movement or REM sleep stages (Powell,
1965; Reding, Zepelin, & Robinson, 1968).
As with the other etiological literatures on bruxing,
much work remains before the roles of sleep activities will
be known confidently. The bulk of evidence does suggest,
however, that at least some nocturnal bruxing is associated
with arousal phenomena (Rugh & Ware, 1986).
THE ASSESSMENT OF BRUXISM
The occurrence of bruxism can be evaluated by
monitoring and recording the behavior itself and by
measuring its effects on the intraoral environment.
Typically psychologists have evaluated bruxing behavior and
dentists have evaluated its stomatognathic consequences.
Various methods are reviewed here.
The most obvious dental signs of bruxism are unique
wear patterns on the teeth. These are flat and highly
polished occlusal surfaces termed bruxofacets (Xhonga,
1977). The formation of wear facets is facilitated by the
presence during grinding of extremely small particles of
enamel that actually function as a gritty polish (Ramfjord &
Ash, 1983). Bruxing sometimes produces wear through the
enamel and can lead to exposure of the tooth pulp. Faceting
due to bruxism sometimes takes the form of sharp cuspal
edges and sometimes appears as cupping of the teeth or as
Extreme wear of the teeth can result in occlusal
disharmony. Among the features of occlusal disharmony that
dentists assess are unilateral mastication, bite
discrepancies, locked bite, and sounds emanating from the
TMJ (Alling & Mahan, 1977).
Persons who brux excessively might exhibit well
developed or bulging masseter muscles. Nocturnal bruxing
can involve extended periods of masseteric hypertonicity
that, in turn, lead to the observed muscular hypertrophy
(Meklas, 1971; Posselt, 1968).
Hyperactivity of the masticatory muscles can result
also in various facial pains. Bruxist patients report
sometimes that the elevator muscles feel tired or painful,
especially in the morning (Alling & Mahan, 1977). Affected
teeth might be tender to percussion or hyperresponsive to
thermal changes (Dubner et al., 1978).
Brewer and Hudson (1961) were able to quantify patterns
of tooth contacts by using miniaturized radio transmitters
fitted into complete dentures. Later, intraoral telemetry
systems were developed that used radio transmitters small
enough to fit into parts of bridges that substitute for
absent teeth (Glickman, Haddad, & Roeber, 1971).
In a telemetry system the transmitter consists of an
oscillator circuit, a battery power supply, and a
multilayered switch. The unit is embedded in a pontic
replacing a molar, and is activated when the switch makes
contact with a pinpoint gold inlay on the cusp of an
opposing tooth. Signals are emitted at frequencies, that
vary according to the part of the switch that makes contact.
The signals then are picked up by an antenna, and fed
through receivers and an oscillograph (c.f. Glickman,
Pameijer, Roeber, & Brion, 1969).
Miniaturized radio transmitters were developed to study
occlusal relationships. Scott and Ash (1966), however,
developed transmitters that were capable also of registering
the force of tooth contacts. Powell (1965) succeeded in
studying tooth contacts during sleep in seven nonbruxing
patients. Hence intraoral telemetry seems to be adaptable
to the study of bruxing.
Patients who brux report frequently that they were
unaware of the habit until a spouse or roommate told them of
grinding sounds. These sounds have been used by researchers
to measure bruxing behavior. In one approach grinding
sounds are microphonically recorded, amplified, and hand
scored on a polygraph record (Reding, Zepelin, Robinson,
Zimmerman, & Smith, 1968).
The second approach utilizes an automated tape
recorder. In one variation of this approach the recorder is
constructed to record at preset intervals. For instance, it
may record for 1.5 minutes every 15 minutes. In another
variation of this approach the recorder is activated by a
sound operated relay. In both cases the tapes are later
replayed and scored for brux frequencies (Heller & Strang,
Using audible grinding sounds to measure bruxing is
problematic. Automated tape recordings that sample the
nights at preset intervals are potentially unrepresentative
and time consuming to score. Sound operated relays on tape
recorders do not provide for real time analysis of bruxing
episodes (McGlynn et al., 1985). A major problem is that
audible sounds are associated only with ecentric bruxism.
Centric bruxism does not appear to generate a characteristic
noise (Reding, Zepelin, Robinson, Zimmerman, & Smith, 1968).
The bruxscore monitor (Forgione, 1974) is a thin
plastic plate with imbedded microdots. It is molded to
conform to the occlusal surfaces and is worn at night.
Bruxing is quantified by counting missing dots.
The original appliance was composed of four differently
colored plastic sheets laminated to a total thickness of
0.02 inches. Microdots were printed in edible ink on each
plastic surface. The dots were 1/180 inches in diameter,
fitting 120 dots per inch in a pattern of 14,400 dots per
square inch. The appliance is worn by the subject during
sleep then examined and scored under a microscope.
Mejias and Mehta (1982) developed an index of bruxism
for use with the bruxscore monitor. In brief, the numbers
of dots ground away after one week are tallied. For example
if two dots are ground away from one plastic surface, then a
score of 2 would be recorded. Depending on where the score
falls within the author's norms, he or she is classified as
a non-bruxer, a mild-bruxer, a moderate-bruxer, or a
There are several problems with the bruxscore index in
its current form. For example, the procedure used to
develop norms for the index has not been published.
Additionally, the exact process used to construct the device
itself has never been described. Finally, use of the
bruxscore monitor raises the issue of reactivity of
measurement. A device that is placed in the oral cavity can
influence the very behavior that the device is intended to
Notwithstanding the problems just described,
investigators have used the bruxscore monitor in several
treatment outcome studies. Heller and Forgione (1975)
compared massed negative practice (below) with automated
relaxation training (below) and found that neither technique
reduced bruxism significantly. Mejias and Mehta (1982)
studied short-term splint therapy and found that bruxing
decreased immediately after therapy.
During any muscle fiber contraction a minute electrical
potential is discharged and dissipated into the surrounding
tissue. If surface electrodes are attached properly over a
muscle, then the cumulative electrical activity of the
muscle bundles can be quantified in terms of total voltage.
Typically this voltage is on the order of microvolts or
thousandths of volts. Measuring the electrical activity of
muscles is called electromyography (EMG).
Currently EMG recording is the most common method of
assessing bruxism. This approach is possible because there
is a high correlation between muscle activity and EMG levels
and because muscle activity during bruxism is generally
greater than is muscle activity during mastication. In
fact, EMG recording can be used to measure bruxist behavior
at microvolt levels completely outside the ranges found for
functional oral motor behaviors (Dubner et al., 1978;
Solberg, Clark, & Rugh, 1975).
It is possible to employ EMG assessment procedures in a
variety of settings including the laboratory, clinic, home,
or work environment. Some procedures for naturalistic
recording are discussed here. There are several EMG devices
on the market that amplify, filter, integrate and record EMG
activity. Each device should be evaluated for usefulness
based on the criteria of expense, portability, and ease of
operation (Burger & Rugh, 1978).
Solberg and Rugh (1972) developed the first portable
device for EMG monitoring. It is about the size of a pack
of cigarettes. A tone sounds via an earphone each time the
subject's EMG level exceeds a preset level. The patient
writes down the time and the activity or situation when
he/she hears the tone.
The first portable EMG recorder is the same size and
shape as the monitor (Burger & Rugh, 1983). Unlike the
monitor, the recorder actually stores electrical activity
above a minimum (e.g. 20 microvolt) threshold. A digital
reading is noted by the subject at the end of a recording
period and written on a log. The digit is a function of the
signal amplitudes, frequencies, and durations. The EMG
recorder is much more sophisticated than the monitor. Even
so it has limitations. The integrated digital data fail to
discriminate an intense short duration response from a less
intense long duration response in terms of
microvolt/seconds. In brief, the single value is vague.
A portable EMG recorder that might overcome some of the
limitations of earlier equipment is currently being
developed by James G. Lee, F. D. McGlynn, and the author. In
general, the frequencies and amplitudes of electrical
impulses above an adjustable microvolt threshold are
continuously recorded on minicassette tapes (Dowdell,
Clarke & Kardachi, 1976). In addition, a high precision
crystal clock marks the tape at intervals so as to permit
real time recording. The amplifiers and clock are housed in
the standard minicassette player case.
At the beginning of each recording period the subject
simply inserts a tape into the player, attaches the
electrodes over the belly of the masseter, records the time,
and activates the player. The subject later returns the
collection of tapes and the recorded activity on each is
played into wave form analysis peripherals contained within
an IBM-PC computer. The data are analyzed and stored in the
Data obtained with the unit provide for fine-grain
assessment of the number of separate bruxing episodes per
night, the duration of individual episodes, and the average
EMG amplitude over individual episodes. In addition, the
typical times bruxing episodes occur during the night for an
individual are known (McGlynn et al., 1985; Cassisi,
McGlynn, & Mahan, in press).
AVAILABLE TREATMENTS FOR NOCTURNAL BRUXISM
The professions of clinical psychology and dentistry
both have methods for treating bruxist patients. Therapy by
professionals from either or both groups might be called for
in any individual patient. Dental and psychological
approaches to the problem are reviewed generally in the two
narratives that follow. Then uses of nocturnal wake-up
alarms are reviewed in more detail.
The benzodiazepines, and other muscle relaxants, are
used frequently for controlling extreme cases of bruxism.
Typically a small dosage is prescribed before bed (Alling &
Mahan, 1977). The precise mechanisms of action for these
drugs are unknown. Rugh (1978) reported a representative
evaluation of diazepam effects on masseter muscle activity
before, during, and after use of a 5 mg. dose for seven
nights. Nocturnal EMG activity decreased dramatically
during the drug period, but overactivity returned soon after
discontinuation of drug therapy. Similar findings have been
reported in the literature consistently (Ramfjord & Ash,
Equilibration is defined as the permanent adjustment of
occlusion with grinding techniques. Equilibration follows a
more or less standard course of events. The procedure is
explained to the patient. Next, impressions and casts are
made of the patient's teeth. The upper and lower casts are
then mounted on an adjustable articulator, and interference
to smooth gliding contact between them are noted as the
"jaws" of the articulator are moved about. Finally, the
patient's teeth are marked and the interference are ground
down in a manner so as to maximize remaining tooth surface.
Dentists report high rates of success using
equilibration and some laud occlusal therapies as highly
successful treatments for bruxism (Ramfjord & Ash, 1983).
However, replicable treatment effects on bruxing have not
been shown, because researchers have not characterized
fundamental aspects of the behavior accurately. In
addition, the possibility exists that clinically beneficial
effects from equilibration derive to an important degree
from placebo or nonspecific influences in the overall
treatment context (Greene & Laskin, 1972; Goodman, Greene, &
Laskin, 1976). Hence, there is a clear need for further
investigation (cf. Clark, Beemsterboer, Solberg, & Rugh,
Occlusal appliances (variously called bite splints,
night guards, occlusal splints, etc.) are used frequently in
the treatment of bruxist patients. The appliances are
fabricated in acrylic from articulator-mounted castings of
the patient's dentition and are retained in the mouth by
wire or hard acrylic clasps. The purpose of splints include
functionally removing occlusal interference and preventing
occlusal contacts between maxillary and mandibular teeth
(Kass & Tregaskes, 1978)
Patients treated with splints report clinical
improvement in 70 to 90 percent of the cases. However,
methodological shortcomings are apparent in the clinical
research reported to date (McGlynn & Cassisi, 1985; Okeson,
Moody, Kemper, & Calhoun, 1983). In addition, the research
shows mixed results for therapeutic efficacy, including some
reports describing all possible outcomes (Cassisi, McGlynn,
& Mahan, in press).
Diverse psychological treatments for bruxism have been
reported. They range from insight-oriented psychotherapy
and hypnosis to biofeedback and behavioral conditioning.
Components of psychological therapies for which some
meaningful treatment-outcome data exist are reviewed below.
Muscular Relaxation Training
Progressive relaxation training (Jacobson, 1938)
teaches patients the skill of voluntary-muscle relaxation.
It entails: (a) subdividing the skeletal musculature into a
number of discrete muscle sub-groups, (b) presenting
instructions which will produce easily detectable tension
alternating with relaxation in each of the muscle
sub-groups, and (c) suggesting that the patients focus
attention on how differing sensations arise from tense
versus relaxed muscles (see Bernstein & Borkovec, 1973).
Muscular relaxation has been used as a component
treatment in several multiple-treatment experiments having
to do with bruxing. Heller and Forgione (1975), for
example, compared relaxation training to massed negative
practice (below) as a treatment among 27 patients clinically
and radiographically diagnosed as bruxers. Relaxation had
no impact on a bruxscore monitor definition of bruxing.
Moss, Hammer, Adams, Jenkins, Thompson, and Haber (1982),
for another example, used relaxation training as one
component in a relaxation-feedback treatment of bruxing for
a 29-year-old female patient with a 3-year bruxing history.
Relaxation had no effect on an EMG definition of bruxing.
Extensive reports in the literature attest to the
tension-reducing effect of relaxation training when it is
done correctly (cf. Jacobson, 1938). Progressive relaxation
is used frequently to treat conditions such as insomnia,
tension headache, and spasm-related pain. No application of
relaxation to bruxist patients involved clinically sound
relaxation training procedures. Therefore, the utility of
progressive relaxation in some cases of bruxism has not been
Massed Negative Practice
Dunlap (1932) suggested that repetitive, voluntary
practice to the point of fatigue might reduce habits such as
nailbiting and tics. Yates (1958) presented a
neobehavioristic learning theory account of how such an
effect might be achieved. The first use of "massed negative
practice" as a treatment for bruxism was reported at about
the same time.
Successful use of massed practice in individual case
studies of bruxist patients has been reported by Wolpe
(1958) and Ayer and Gale (1969). Each patient was taught to
perform six series of daily trials in which five 1-minute
periods of voluntary clenching were alternated with 1-minute
periods of rest. Patient- and spouse-reports of bruxing
cessation were used as the measure of treatment success.
Ayer and Levin (1973) treated 14 nocturnal bruxists
using massed-practice clenching. A 14-day massed-practice
protocol prompted six daily series of self-monitored
voluntary clenchings in which five, 5-second clenches were
separated by 1-minute rest periods. Self- and
spouse-reports of grinding were used as measures of bruxism.
Eleven of 14 patients stopped bruxing within 9 days. Ayer
(1976) used the same massed practice protocol to treat 33
nocturnal bruxists. He found approximately 25 percent of
patients to have ceased the behavior through a 1-year
follow-up. Self-reports of grinding were used to measure
As noted previously, Heller and Forgione (1975)
compared massed practice to muscular relaxation training as
treatment among 27 patients clinically and radiographically
diagnosed as bruxers. The protocol of Ayer and Levin (1973)
was replicated. Massed practice had no impact on the
behavior as assessed by a bruxscore monitor.
Rugh (1976) compared massed-practice clenching to
controlled gum chewing and to placebo exercises as
treatments for bruxing among 20 TMJ patients. Eight of the
patients used a 10-14 day massed-practice protocol that
entailed six daily sessions of vigorous clenching. Massed
practice did produce significant reductions in nocturnal
bruxing according to an EMG measure.
Various conceptualizations of massed-practice effects
on nocturnal bruxing have been offered by Rugh (1976).
These conceptualizations include a combination of aversive
punishment, stress reduction, homeostasis, and "response
sensitization." Thus therapeutic effects from massed
practice protocols can be explained by theories of differing
systematic origins. However, the weak literature on massed
practice treatment for bruxing and the antiquated theories
concerning massed practice mechanisms call for caution in
endorsing the approach.
EMG-ACTIVATED ALARM THERAPY
As noted, the basic purpose of the present research was
to evaluate EMG-based feedback alarms for the treatment of
nocturnal bruxing. Hence, the present section considers in
detail the research reported to date in which response
produced signals have been used to modify nocturnal bruxist
behavior. Emphasis is placed on research that has made use
of signals triggered by facial EMG activity.
DeRisi (1970) studied three bruxist subjects in their
own homes. Equipment was developed that recorded bruxing
episodes with a pressure transducer that was implanted in a
silastic mouthpiece and that presented a loud tone
contingent upon rhythmic or lengthy pressure registration.
The alarm did not produce enduring behavior change according
to three intrasubject (ABA) experiments. However, the use
of the mouthpiece raised an important methodological issue.
The masticatory system is reactive to disruptions in
the relationships between tooth surfaces. As described
earlier, some dentists have suggested that the modification
of gliding contacts between maxillary and mandibular teeth
can increase or decrease rates of bruxing activity.
Therefore, any effects from the nocturnal alarm in DeRisi's
(1970) experiments might have been confounded in unknown
ways by the presence of the mouthpiece itself within the
Heller and Strang (1973) treated one bruxist patient in
her home. In this case, equipment was developed via which
audible grinding sounds activated a voice operated relay
that sounded an alarm. Use of the alarm reduced audible
grinding according to an intrasubject (ABAB) design. More
importantly, a second methodological problem was raised.
Descriptions of the various topographies of nocturnal
bruxing have not been adequate. Self-report data (Glaros,
1981) and the author's clinical experience suggest, however,
that a large proportion of nocturnal bruxing takes the form
of prolonged clenching. Since clenching produces no
transducible sound, alarm contingencies based on audible
grinding will be incomplete for some patients. Heller and
Strang (1973) noted also that some grinding sounds were
recorded below the trip threshold of their voice-operated
The basis for modern bruxing alarm systems was
established by Solberg and Rugh (1972) who used recordings
of facial EMG activity to define bruxing. There are three
reasons why this approach is beneficial. First, as noted
earlier, the muscular forces associated with bruxing are
generally above those associated with functional oral-motor
behaviors such as swallowing (Dubner et al., 1978; Solberg
et al., 1975). Therefore, bruxing should be relatively easy
to identify electromyographically. Second, surface
recording does not entail any incursion into the oral
cavity, hence eliminating confounds from the effects of
occlusal changes. Third, EMG recording identifies silent
clenching as well as audible grinding provided that suitable
instrumentation is used. The remainder of this section
reviews the literature in which facial EMG-triggered alarms
have been used to modify nocturnal bruxing.
Clinical Trials of Feedback Alarms
Kardachi and Clarke (1977) used an EMG-correlated tone
to treat a series of nine nocturnal bruxist patients 19 to
38 years of age. Bruxing feedback was provided by an
audible tone that varied in pitch according to the intensity
of muscular activity. EMG data were recorded with a small
tape player that was activated and driven by bruxing
activity (Dowdell et al., 1976). The tapes provided a
measure termed the total count which was a function of
bruxing duration and intensity.
Bruxing was recorded for seven baseline nights for most
subjects. This was followed by the use of biofeedback for
an equal period. Nights were not consecutive in either the
baseline or feedback phases. Presentation of the mean
bruxing units recorded for each subject during baseline and
feedback phases showed reduced bruxing in eight of the nine
patients. Reductions were fairly uniform across seven of
these eight patients. The authors suggested that the counts
reflected reduced durations of bruxist episodes, not reduced
Rugh and Johnson (1981) described the application of an
EMG-activated tone in treating five nocturnal bruxist
patients 20-39 years of age. Bruxing was defined by
unilateral masseter EMG signals of at least one second
duration with 20 microvolts (peak to peak) or greater
activity. Bruxing episodes were counted with a small chart
recorder fitted into a briefcase to render it more portable.
The charts were hand scored so as to yield frequency and
duration data. Bruxing feedback consisted of a 700 Hz, 300
mW tone temporally contingent on bruxing. Data were
collected for at least 10 baseline nights. Feedback was
provided subsequently for at least eight nights.
Presentations of the mean bruxing frequencies and durations
for each subject showed obvious reductions in bruxing
durations as a function of auditory bruxing feedback. As
was reported by Kardachi and Clarke (1977) there were no
reductions in bruxing frequencies.
Some have interpreted the absence of feedback effects
on bruxing frequencies to mean that feedback paradigms are
devoid of "learning." In turn, this consideration has
prompted efforts to enhance the durability of feedback
effects by awakening patients during bruxing episodes. The
first report of work along these lines was presented briefly
by Beemsterboer, Clark, and Rugh (1978). Recordings of
cumulative masseteric EMG activity above 20 microvolts were
obtained from six subjects for 12 nights before and 10
nights after a two-week period of nocturnal alarm feedback.
During the feedback period subjects were told to become
fully awake following alarm soundings. They were given a
performance task that required three to five minutes of
wakefulness. Uniformly significant decreases in each
subject's nightly mean suprathreshold masseter activity from
before to after treatment were noted. The authors concluded
that nocturnal auditory feedback must be combined with an
arousal task in order to promote long term reductions in
Auditory feedback coupled with an arousal task was
studied and reported in more detail by Clark, Beemsterboer,
and Rugh (1981). Cumulative recordings of masseteric EMG
activity above 20 microvolts were obtained from 10 subjects
for 10-12 nights before and 7-14 nights after a two-week
period of alarm treatment. During the treatment each
subject received auditory feedback upon emitting a
moderately hard biting force which was defined separately
for each subject. The auditory signal served further to
prompt subjects to get out of bed, cross the room, and
record the time and sleep quality at the time of wakening.
Comparisons before and after treatment showed significant
decreases in nightly mean suprathreshold masseter activity
among nine of 10 patients. Marked reductions in the numbers
of self-monitored awakenings during treatment were noted
also by the authors. These data appeared to indicate that
the treatment package effectively reduced bruxing for up to
two weeks after its withdrawal.
The reports just-reviewed suggest that auditory
feedback of bruxing during sleep has differential effects
depending on the use of a contingent arousal task. However,
this observation is based on comparisons between trials in
which the variables of feedback and of feedback-plus-arousal
instructions were studied separately. There are two reports
describing work in which the effects of feedback and of
feedback-plus-arousal instructions were studied within the
The first is a report by Funch and Gale (1980) who
modeled their experiment on that of Rugh and his colleagues
with respect to experimental variables such as bruxing
feedback and arousal task requirements. A 27-year-old
female patient was used in two baseline-treatment-baseline
experiments that were separated by two months and were
followed by a final baseline assessment one month later.
Baseline and treatment phases lasted for 10 consecutive
nights. In the first experiment feedback was provided
following Rugh and Solberg (1975). In the second experiment
feedback was paired with arousal task instructions following
Beemsterboer et al. (1978). The results reported by Funch
and Gale (1980) differed from earlier results in three ways.
First, durations of bruxist activity following the
feedback-alone phase increased markedly over baseline
levels. Second, durations of bruxist activity following
feedback-plus-arousal instructions increased over baseline
levels as did bruxing durations following feedback alone.
Finally, the feedback conditions brought about reduced
frequencies as well as reduced durations of bruxing during
The second report comparing alarm treatment with and
without arousal instructions was provided by Moss et al.
(1982). They described the treatment of a 30-year-old
female with a seven year history of nocturnal bruxing. This
study manipulated feedback-plus-arousal and feedback alone
by sometimes requiring and sometimes not requiring manual
termination of the feedback signal. In addition the
variable of feedback termination was crossed sequentially
with the variable of quality/loudness of the feedback signal
itself. Feedback in the form of a soft tone had no
meaningful impact on bruxing frequencies or durations either
with or without the manual reset contingency. Loud buzzer
feedback that required manual termination demonstrably
reduced both the frequencies and durations of bruxist
episodes. The effects of loud buzzer feedback alone were
evaluated for only two nights.
Funch and Gale (1980) noted that intrasubject research
is clearly problematic from the standpoint of generalizing
results to other patients. Variability in the data reported
by Moss et al. (1982) rendered some of their phase durations
too short to meet the requirements of intrasubject
replication design logic. Finally, a number of potentially
important variables differed between these two intrasubject
experiments and the other studies reviewed in this section
of the narrative. On the whole the larger literature about
clinical trials suggests that alarms per se reduce bruxing
durations, while alarms plus arousal task instructions
reduce both durations and frequencies.
Laboratory Studies of Feedback Alarms
Three relevant studies have been conducted in sleep
laboratory settings. Wagner (1981) combined home based
recording with laboratory work to evaluate nocturnal alarm
treatment. The patient was a 26-year-old female with a
lengthy history of nocturnal grinding. Bruxing frequency in
the home setting was quantified for 60 nights by spouse
recordings of nightly grinding sounds on an 11-point scale
that ranged from 0 (no grinding) to 10 (grinding throughout
the night). A Coulbourn Instrument Modular System was used
to amplify, filter (90 Hz low cut, 500 Hz high cut), and
integrate ENG activity in the laboratory. Bruxing
frequency, intensity, and duration were quantified for six
nights with this device. The resulting signal was sent to a
threshold trigger, and to a recorder that cumulated seconds
of suprathreshold activity, as well as to the polygraph.
Bruxing was defined as activity above five microvolts
(integral average). Three seconds of quiescent EMG
recording were required to tally separate bruxing episodes.
When bruxing occurred during treatment, the threshold
trigger sounded an alarm until bruxing ceased. The alarm
was "loud enough to disturb the subject's sleep, but not so
loud as to startle or frighten the subject" (Wagner, 1981,
p. 88). The subject was told that relaxing her jaw would
stop the tone.
The protocol began with 24 nights of home baseline
recording and two nights of laboratory baseline recording.
These 26 nights were followed by a single night of alarm
treatment in the laboratory. The single treatment night was
followed by 14 nights of home recording that, in turn, were
followed by a second treatment night in the laboratory.
Finally there were 26 nights of "follow-up" recording, 24 in
the home and two in the laboratory.
Despite having only two nights of alarm treatment in
the laboratory, four nights of EMG monitoring demonstrated
reductions of 50 to 65 percent between baseline and
"follow-up" for bruxing frequencies and durations,
respectively. The spouses' ratings for 60 nights showed
reduced variability and average reductions in the 11-point
rating scale from 3 before treatment to 0.5 during
Piccione, Coates, George, Rosenthal and Karzmark
(1982) used a nocturnal alarm coupled with a task requiring
wakefulness to treat two bruxist patients both at home and
in a sleep laboratory. The first subject was a 50-year-old
female who had been bruxing 29 years. The second subject
was a 44-year-old female who had been experiencing morning
jaw fatigue for three years and who had often awakened with
her teeth clenched tightly. Bruxing rates and durations in
the home were assessed using the ambulatory equipment and
parameters developed by Rugh and his colleagues (described
previously). Bruxing rates and durations in the laboratory
were assessed with a Grass Model 7D polygraph that amplified
and filtered (10 Hz low cut, 75 Hz high cut) bilateral
masseter EMGs. There were three criteria for defining
bruxism; amplitude, duration and rhythmicity. The amplitude
criterion was at least 20 microvolts. The duration
criterion was 0.5-1.5 seconds. The rhythmicity criterion
was three incidents of criterional amplitude and duration
separated by return to baseline EMG for 2.5 seconds or less.
Bruxing episodes were scored for number of bruxes and for
total suprathreshold duration. Bruxes during treatment
phases produced a contingent alarm and arousal-task.
Instructions were identical to those reported by
Beemsterboer et al. (1978).
Both subjects were exposed to alternating baseline and
treatment phases (ABAB) of two to three weeks duration. One
subject lagged behind the other by approximately a week.
Each subject slept in the laboratory for up to three nights
per week during baseline and treatment phases. The
investigators were also interested in the effects of alarm
treatment on sleep, so laboratory assessment of EMG was
accompanied by polysomnographic characterization of sleep
stages using standard methods and definitions (Rechtschaffen
& Kales, 1968).
The results did not replicate prior research using
feedback alarms plus arousal task instructions. One
subject's bruxing was not reduced significantly during the
first treatment period. Subsequently, there was a marked
rebound during the "return-to-baseline" period and the
second treatment period served only to reduce the rebound to
just above the original baseline levels. The other
subject's, bruxing increased slightly from the original
baseline through the treatment period. Subsequently, there
was a further increase during the "return-to-baseline" phase
and it dropped to the original baseline level during the
second period of treatment. These results were similar for
bruxing frequency and duration measures. In sum, the alarm
treatment protocol failed to affect bruxing favorably. In
addition, there were signs of negative effects on sleep
duration, at least initially during treatment. The authors
discussed the possibility that sleep deprivation could have
changed the depth and duration of sleep stages as treatment
Purzycki, Harsh, and Badia (1986) conducted a
laboratory study of nocturnal bruxing that compared feedback
alone to feedback-plus-arousal instructions. They also
characterized the effect of feedback on selected sleep
variables. Two subjects were recruited via a newspaper
advertisement. The first was a female 22 years of age with
a lengthy history of nocturnal bruxing and of unsuccessful
intervention by dentists. The second was a female 26 years
of age with a lengthy bruxing history.
A Narco six-channel physiograph and a Coulbourn Hi-Gain
Bioamplifier/Coupler were used to quantify bruxing. The
amplified signal was filtered with a 150 Hz low cut, 1.0 kHz
high cut bandpass filter. Bruxing events were detected via
a bipolar comparator that served also to trigger feedback
during treatment periods. Bruxing events were defined as
EMG above 30 microvolts for 0.5 seconds or longer that
followed prior events by at least one second. Bruxing
frequency and duration data were digitized and printed on a
continuous tape. Subthreshold EMG events from 20-29
microvolts were handled similarly.
The subjects were exposed to baseline, "feedback,"
"feedback/awake," and baseline phases for 17 to 20 nights.
There were at least three nights per phase. Phase shifts
were made following true intrasubject replication design
logic (Johnston & Pennypacker, 1980). The "feedback"
condition consisted of diurnal pre-training to produce and
escape a 65 db tone via increasing then decreasing
masseteric activity. Next the 65 db tone was presented
contiguously with bruxing during sleep. The
"feedback/awake" condition consisted presenting an 87 db
tone contingent upon nocturnal bruxing. The 87 db tone
could be stopped only by manually closing a microswitch that
was located above the bed. The subjects were told to stay
awake for two minutes after alarm soundings and intercom
instructions were used to assist in maintaining wakefulness.
Bruxing was differently affected by the "feedback"
condition across the two subjects. For the first subject
the 65 db feedback reduced suprathreshold EMG to 50 percent
of baseline duration while increasing subthreshold EMG
durations nearly fivefold. For the second subject the 65 db
feedback did not reduce suprathreshold EMG durations as
compared to baseline levels.
The "feedback/awake" condition affected bruxing by the
two subjects differently also. For the first subject the
manually terminated 87 db tone reduced dramatically the
duration and frequency of suprathreshold EMG and failed to
change subthreshold values from baseline levels. For the
second subject the manually terminated tone was associated
with a significant increase in suprathreshold EMG duration.
Since the tone was failing to produce two-minutes of
wakefulness for the second subject, the experimenter began
entering the sleep area at tone terminations and turning on
a lamp. He also prompted the subject to prop herself on her
elbows for two minutes. This additional procedure produced
bruxing decreases similar to those displayed by the first
subject during her last treatment period.
In the last baseline, the first subject demonstrated
brux frequencies equal to and durations longer than those
observed in the original baseline assessment. The second
subject demonstrated similar results. Unlike the earlier
report by Piccione et al. (1982), there were no signs that
alarm therapy affected sleep staging. The percentages of
time in REM and NREM stages were equivalent across all
In summary, the protocol used by Wagner (1981) appeared
to have awakened the subject and to have reduced both
durations and rates of bruxing. The alarm-plus-arousal
approach reported by Piccione et al. (1982) altered the
EEG-defined sleep staging of two subjects but did not reduce
their bruxing durations or rates. The alarm-plus-arousal
paradigm used by Purzycki et al. (1986) did not appear to
change sleep staging but did reduce bruxing durations and
frequencies when the two subjects were awakened fully.
These results are less consistent than the results of
clinical trials. Despite the "laboratory" context, these
investigations also were less methodologically rigorous than
were some of the clinical studies.
Comparisons of Feedback Alarms
With Other Treatment Approaches
Three experiments have been published in which
EMG-based alarms were compared directly to other means of
modifying nocturnal bruxing. Kardachi, Bailey, and Ash
(1978) compared EMG-correlated auditory feedback to
noncontingent tones, to occlusal adjustment, and to mock
occlusal adjustment as treatments for 16 nocturnal bruxers
18-39 years of age.
Bruxing was monitored for all groups as reported by
Kardachi and Clarke (1977). Masseter EMG was recorded
unilaterally. Activity above 100 microvolts was amplified
then stored on a small tape recorder in the form of duration
and intensity counts.
For two groups, bruxing during sleep was recorded for
seven nights before and seven nights after occlusal or mock
occlusal treatments. For two other groups, bruxing was
recorded during sleep for seven nights before, during, and
after auditory feedback or noncontingent tone treatments.
In addition, bruxing among subjects who received either
occlusal adjustment or true EMG feedback, was recorded for
several nights approximately three months after the
experiment. True auditory feedback of masseter activity was
contingent on activity above a 100 microvolt threshold.
This treatment continued nightly for one week. The
noncontingent tone treatment consisted of auditory signals
from three to seven seconds in duration presented six to
eight times per hour. This control "treatment" also
continued nightly for one week. Occlusal adjustment was
performed in accordance with standard dental practice. Mock
occlusal adjustment consisted of grinding nonopposing and
nonsupporting tooth surfaces minimally. Four subjects who
were identified clinically as nonbruxists were evaluated
also for seven nights.
True auditory feedback of suprathreshold
masseter/temporalis EMG activity produced a 70 percent
average reduction in bruxing during treatment. However,
this reduction was followed by an immediate return to
baseline levels among three of the four subjects when
treatment was withdrawn. The noncontingent or false
feedback produced no meaningful change from baseline bruxing
activity among four subjects. The effects of occlusal
adjustment were more complex. Two subjects showed immediate
increases in bruxing followed by short-term reductions to or
below the baseline levels. Two subjects showed immediate
decreases in bruxing followed by short-term returns to
baseline levels. Mock equilibration produced slight
reductions in bruxing for all four subjects. The four
nonbruxist subjects bruxed much less than did the other 16
subjects. The three-month follow-up of bruxing demonstrated
that one subject who had received occlusal adjustment and
one who had received auditory feedback showed 50 percent
less activity than had been shown before treatment. The
authors suggested that the effect of auditory feedback in
reducing suprathreshold EMG activity appeared to be more
consistent than was the effect of occlusal adjustment.
Casas, Beemsterboer, and Clark (1982) compared
nocturnal alarms to stress-management training, to nocturnal
alarms plus stress-management training, and to no treatment.
Sixteen bruxers who averaged 29 years of age served as
subjects. Bruxing was monitored as in the study reported by
Clark et al. (1981). The protocol entailed recording
bruxing during sleep for approximately ten nights before and
ten nights after each experimental treatment.
Four subjects received the nocturnal alarm treatment
described by Clark et al. (1981). This treatment included
their arousal task instructions and it continued nightly for
Four subjects received stress management training that
had components of both cognitive behavioral (Beck, 1976) and
clinical biofeedback (Budzynski & Stoyva, 1973) approaches.
There were four, weekly, individual treatment sessions of 60
minutes in duration. Each had the threefold purpose of
teaching the subjects (1) to identify stressful situations
and correlated cognitions, (2) to counter the aversive
cognitions with private self-efficacy statements, and (3) to
relax with EMG assisted biofeedback training of the arm and
Four subjects received the four-week stress management
training protocol to which the nocturnal
alarm-plus-arousal-task treatment was added during the third
week. A wait-list control group of four subjects was used
Differences in mean suprathreshold EMG activity from 10
nights before to 10 nights after treatment showed
significant bruxing reductions in all three groups of
treated subjects. These changes were not different but were
consistently larger than changes among the four untreated
subjects. Casas et al. (1982) concluded that stress
management and nocturnal alarm approaches both are
The third comparison of nocturnal alarms with other
approaches was reported by Moss et al. (1981). The patient
was a 29-year-old female who reported a three year bruxing
history. The effects of loud buzzer feedback requiring
manual termination were compared with those of muscular
relaxation training and with those of baseline constraints.
The feedback contingency was associated with dramatically
reduced bruxing frequencies and durations, while muscular
relaxation training had no effect.
The results of the three treatment comparisons suggest
that nocturnal alarm approaches are competitive with the
other approaches that are available for managing nocturnal
bruxing. This conclusion is not surprising given the modest
effects that have been reported in the larger literature on
dental approaches to the problem (cf. McGlynn et al., 1985;
Okeson et al., 1983).
PURPOSES OF THE EXPERIMENT
The exhaustive review of the literature provided above
suggests that alarm therapy is a promising approach for the
treatment of bruxism. There remain significant research
questions at all levels of inquiry. Below are the questions
that were asked in this experiment along with the rationales
for asking them.
One conclusion reached in previous studies is that full
arousal following alarm soundings increases the efficacy of
therapy. However, different methods to promote wakefulness
have been used. At one extreme, investigators have merely
instructed the patients to record the time and quality of
sleep following alarm soundings (Beemsterboer et al., 1978).
At the other extreme, investigators have actually entered
the sleep environment and tapped on the shoulder of the
subject following alarm soundings (Purzycki et al., 1986).
Neither approach is entirely satisfactory.
Two previous studies have included a manipulation that,
in principle, provides a practical solution to the problem
of requiring arousal following alarm soundings (Moss et al.,
1981; Purzycki et al. 1986). This is the inclusion of a
manual reset switch on the alarm itself. Using this
approach, the tone sounds continuously following a brux
episode until a motor escape response is emitted.
The manual escape manipulation has only been studied in
three subjects to date. The primary purpose of this study
was to evaluate the efficacy of EMG-activated alarm therapy
with manual reset as a treatment/management package for
reducing rates of nocturnal bruxing.
Several studies have reported that bruxism levels
following alarm protocols have actually increased above
those found before treatment (Funch & Gale, 1980; Piccione
et al., 1982). If one likened this treatment to operant
punishment (DeRisi, 1970), such effects would be viewed as
mirroring a behavioral contrast effect (Reynolds, 1961).
A second purpose of this study was to evaluate the
occurrence of post-treatment rebound or "contrast" following
As noted, one method of promoting wakefulness following
alarm soundings has been to instruct subjects to monitor
data such as time of night. Clark et al. (1981) have used
these records of alarm soundings as an adjunct dependent
measure. This raises a third question of interest. Since
the sleep records are obtained as a matter of course, might
such records be used to monitor the course of brux frequency
across successive nights? In the present experiment nightly
diaries were kept in addition to alarm counts made
automatically and unobtrusively in the apparatus. The
third purpose of the study was to compare self-monitored
alarm counts with automatically recorded alarm counts so as
to evaluate the former as a data acquisition mode.
The potential risks and contraindications of alarm
therapy procedures have rarely been discussed in prior
reports. This is surprising given that the treatment
disrupts sleep and, while controversial, there is a very
large literature arguing about the adverse effects of sleep
fragmentation (Bonnet, 1985; Pearlman, 1982). No previous
study has included measures of diurnal fatigue in patients
undergoing nocturnal alarm treatment. The fourth purpose of
this study was to evaluate the impact of alarm therapy on
reports of daily sleepiness, fatigue and vigor.
The fifth and final purpose of this study was to
examine relationships between bruxism and facial pain or
discomfort. This purpose is ancillary because facial pain
patients were not used as subjects. Nonetheless, an attempt
was made to evaluate the controversial relationship between
nocturnal bruxism and subclinical muscular
Previewed briefly, subjects were recruited via public
notices and by word of mouth. Ten heavy bruxers were
selected from 18 individuals who presented. Subjects were
evenly divided into two groups whose experimental
participation differed only in the sequence of 14 day
no-treatment and treatment phases. Nightly brux-episode
frequency, and the daily ratings of vigor, fatigue,
sleepiness, anxiety, and facial pain were taken throughout
the study. In addition, nightly self-monitored alarm counts
and automated alarm-counts were recorded during treatment
Individuals who suspected they were bruxing were
solicited and screened. These included various health
center employees, students, and acquaintances. They were
recruited with notices that were placed on various health
center bulletin boards, and in the mail boxes of the third
and fourth year dental students. (See Appendix A for copy
of notice.) During recruitment, individuals were asked if
their dentist had indicated a diagnosis of bruxing, if their
roommates had reported overhearing grinding sounds during
the night, and if they had experienced facial pain/muscular
soreness. If there was an affirmative answer to any of
these questions, the individual was given an informed
consent document. After the potential subjects signed the
informed consent document they entered the screening phase.
In return for their participation during the screening
phase, individuals were paid $10.00. In return for
participation in the study, subjects were paid $50.00, and
given the opportunity to purchase an EMG-activated alarm at
Activities during the screening phase included a
structured dental exam (Appendix B), three nights use of the
EMG monitor (described below), and completion of a Screening
Questionnaire (Appendix C). Criteria for inclusion in
the study included confirmation by the dentist of
characteristic wear facets, no observed dental infection,
and demonstrable bruxing during the three night screening
phase. While not criterional, masticatory muscle soreness
on palpation and hypertrophy of the masseter muscles were
noted during the dental exam.
Twelve of the 18 individuals who presented met the
inclusion criteria. Of the 12, one subject dropped-out
following the death of a close family member. Another
subject completed the study, but was excluded because it was
learned that she had taken muscle-relaxant medication at
various times. All subjects were requested to refrain from
other treatments during screening and during the study.
The dental exam and the Screening Questionnaire also
were structured so as to allow a comprehensive description
of the subjects in terms of demographic variables and dental
status. Six females and four males served. The average age
of the male subjects was 25. The average age of the female
subjects was 32. On average the subjects reported having
been aware of the habit for 7.5 years. Three of the males
and five of the females reported facial pain/discomfort.
Two of the subjects had received an occlusal splint
treatment for the problem. The demographic data are
presented in Table 1.
I .- 0)
I (-0 O \
U -4.0 -) 41 M) 4) 0 4) 0 4) w1
x ( I >4 >4 >4 >4 z a > a > >4
I 0 0P40 04 0
1 0 0 0 r 0 0 01
14 r X CM Cm tC C P: c mm m cm
Ir0 t -4 HO -d 0 0 0 0 4.) 0 0 0
I3 C C A C C C C
I 3 4B1 41 I
10 00 iO i Mr-4CO (A r- 00 -4 04O r 01 r-AI
10 OO- C e O -- r-g I i
fth 1 V V V 4J+ (0 rVPJ V 4
I (* a) I (d ) o (D fa 4)i M > ) fo a 4 ) 4
o ( QS a M 0 wQ c to- E-l ZJ W 0 E 0-4 0 u I
M I -44-
1 I0 HO CO CO CO N CO QI
r-4 I V0 I
O0 C U) ln a S 0 0 a I
0 w u
9 0 V
0 o o m 0 0H 0 M 0 i
>> > z > >4-) > >Oa0 > *1
a) 4) '-I-) rc( '- 01i m 5 a W I
0 0 0 r I
No (4 h 4 C4 o- -O O
to 0 0 +J -'-40 I *r4- e I-
42 4 N- N in N N
H H H H H H H H H I
H H H I
O O O -H O 4 0 r4 I
Z c+ 0 2 0 u I
in rH N N N N eq I
eq in f. N 0 I
(N n 'r u) ~ oo
EMG activity was recorded via three electrodes (two
reference, one ground) with 40 mm leads. Bard pregelled, 10
mm diameter, disposable, ECG electrodes were used (reorder
number #160340). Exact placement of the electrodes on the
face followed the standard masseter site described in
Lippold (1967). The electrodes are fitted with
surgical-tape adhesive collars. In addition, strips of 25
mm wide Micropore Hypoallergenic Surgical Tape were placed
over the attached electrodes. One was placed over the two
active electrodes, one was placed over the ground, and one
over the wire leads on the subject's neck. During pilot
work it was determined that electrode stability was enhanced
when the leads were looped over the ear-lobe prior to taping
them down. Appendix D illustrates the electrode placement
and contains the written instructions given to each subject
during monitoring phases.
A detailed description of the EMG-activated
alarm/monitor is in preparation. The device can function as
a portable EMG-activated nocturnal alarm, or by disabling
the alarm, the device can function as a portable EMG
response-episode monitor. The following information about
the device is available.
The alarm/monitor is contained within a 12 cm x 20 cm x
30 cm plastic box that weighs 411 gm. The internal
components of the alarm/monitor are an amplifier, a
gain/threshold control, a comparator, a Schmitt trigger, an
alarm disabling jumper switch, a piezo buzzer (75dB/meter
mini), an EMG response counting chip, and an alarm response
counting chip. External components are an on/off switch, an
alarm reset switch, a momentary data display push button,
and an LED data display. A flow-chart of the system is
presented in Figure 1, and a schematic diagram of the device
is presented in Appendix E.
The gain/threshold adjustment was set so that a 100
microvolt (integral average) signal of .25 seconds duration,
would trigger the alarm. In order to set and calibrate the
device, a constant independently measured criterion
microvolt value was inputted through the electrode leads.
The recessed threshold screw was then turned
counterclockwise until the alarm sounded.
The following method was used to produce independently
measured criterion values with which to validate the
apparatus. A voltage output of preset frequency was
produced by a signal generator (Wavetek 4 MHz Function
Generator Model 182A). This equipment allowed precise
adjustment to the 150 Hz frequency that was used for all
tests except the bandpass. However, because the signal
generator did not provide scaling of voltage, its output was
interfaced with a bioamplifier and digital integrator.
A one volt output from the signal generator was stepped
down by a factor of 1000 by using a simple two resistor
Figure 1: Flow-Chart of the Alarm/Monitor
network (100,000 ohm in series with 100 ohm). Appendix E
illustrates the configuration schematically. The output
from the resistor network (now on the order of microvolts)
was then split and one signal was input into a bioamplifier
(Autogen 1700 Feedback Myograph, bandpass 20 Hz to 1 KHz),
while the other was input into the alarm/monitor. The
instantaneous output from the bioamplifier was connected to
an Autogen 5100 Digital Integrator/Wave Form Analyzer which
allowed digital display of the microvolt amplitude (1 sec,
integral average) that was being input to the alarm/monitor.
Using this system, the frequency of the input was
varied and the functional bandpass of the alarm/monitor was
determined. From 10 Hz to 550 Hz the device sounded
reliably to a 100 microvolt, 2 second signal.
After the threshold of the alarm/monitor was adjusted
using the method described above, the signal generator was
adjusted to several values above criterion. The output from
the signal generator was set to the gated mode. One hundred
2-second trials at each of five suprathreshold levels (105,
107, 109, 111, and 113 microvolts) were inputted to the
alarm/monitor and resulted in 100 percent correct alarm
soundings. Then one hundred 2-second trials at each of five
subthreshold levels (95, 93, 91, 89, 87 microvolts) were
inputted to the alarm/monitor. These trials resulted in 100
percent correct rejection. Taken together, the above
results indicate that the alarm/monitor has adequate
performance characteristics for experimental use.
No differences in the performance of the alarm/monitor
were detected when it was tested in and out of the alarm
mode. The device was tested out of alarm mode by depressing
the read button and observing increments in the
response-episode count function.
The response-episode counting function of the device is
activated whenever the device is turned-on. It tallies a
response-episode after EMG/microvolt activity exceeds the
preset threshold for .2 seconds or more and returns to
subthreshold levels for at least one second. The
response-episode count is read from the second of two
columns of ten LED lights located on the face of
alarm/monitor. The count is recorded by the subject in the
morning before detaching and turning the device off. The
subject activates the display by pushing a button marked
"read." Both columns of the light array are then activated.
The lights that are illuminated are recorded by the
subject on the Morning Questionnaire (see Appendix F)
contained within the Daily Supply Book (see below). The
response-episode count is encoded into a binary numbering
system such that illuminated lights represent "l's" and
nonilluminated lights represent "O's." Therefore, a
continuous response-episode count is available up to 1024.
The recorded response count was later decoded with the
aid of the calculating function of the IBM PC-standard
computer software Sidekick, version 1.56 by Borland
International Inc., Scotts Valley, California. This
function provides for binary to decimal transformations with
a single key stroke.
Alarm Frequency Count
The first column of ten LED lights on the monitor
displays the count of alarm soundings. This count
increments by one at the onset of an alarm sounding. This
count is also encoded, as is the response-episode count to a
limit of 1024. When the contingent alarm function of the
device is disabled and the device is in the monitoring mode,
this column of lights is left free to vary with incidental
contact of the manual reset switch. Additionally, subjects
were not informed about the meaning of the LED until the
debriefing at the end of the study. They were initially
instructed that the entire light array is the measure of
"how much they brux."
Self-Monitored Alarm Frequency Count
During treatment subjects were instructed to record on
a Nightly Alarm Record (Appendix G) the times when the alarm
sounded. They were instructed to place the record in an
area that could be reached conveniently from the bed and to
make sure that the chosen place was flat, and close to a
light. They were instructed also to acquire an illuminated
clock. Pretreatment instructions explained that
consistently timing the alarm soundings was important to the
project. The instructions explained also that individuals
who "wake-up fully following alarm soundings appear to
respond much better to treatment."
Sleepiness, Vigor, and Fatigue Measures
Daytime sleepiness was assessed via self-reports on the
Stanford Sleepiness Scale or SSS (Hoddes, Dement, & Zarcone,
1972; for review see Carskadon, 1982). This seven-item
scale sound psychometrically and it correlates significantly
with behavioral measures of sleepiness such as the Sleep
Latency Test (Magee, Harsh, Badia, Revis, & Purvis, 1986).
The SSS was taken in the evening as one of the Nighttime
Questionnaires (see Appendix H) and in the morning as part
of the Morning Questionnaire.
Fatigue and vigor were measured with the corresponding
subscales from the Profile of Mood States or POMS (McNair,
Lorr, & Droppleman, 1981). The standard POMS form available
from the Educational and Industrial Testing Service, San
Diego, California, was page 2 of the Nighttime
Questionnaire. At least six independent factor analytic
studies have been reported that identify fatigue and vigor
as mood factors on the POMS (Eichman, 1978; McNair, Lorr, &
Droppleman, 1981). The vigor factor is defined by adjectives
suggesting a mood of vigorousness and high energy. The
fatigue factor represents a mood of weariness, inertia and
low energy. While negatively related, these two subscales
appear do not appear to be opposite extremes of a single
bipolar factor. The POMS has been used to investigate the
effects of sleep deprivation in several studies (Freidmann,
Globus, Huntley, Mullaney, Naitoh, & Johnson, 1977; Johnson,
1982, p. 129).
The instructions of the POMS were modified to read "HOW
HAVE YOU BEEN FEELING DURING TODAY," instead of "HOW HAVE
YOU BEEN FEELING DURING THE PAST WEEK INCLUDING TODAY."
McNair et al. (1981) reported replicating the the original
factor structure with a similar instructional change.
The POMS has a tension-anxiety factor among the six
factors produced reliably by factor analytic work. This is
a factor related to heightened musculoskeletal tension and
observable motor manifestations of anxiety. Since the POMS
provides information on this construct and it is
theoretically related to bruxism, the tension-anxiety factor
was scored and used in subsequent exploratory analyses.
Facial Pain/Discomfort Ratings
Facial pain/discomfort ratings were made on an
adaptation of the visual analog scale (VAS) developed by
Scott and Huskisson (1976). The VAS in this study was a 10
cm line which is taken to represent the continuum of facial
muscle comfort to severe facial pain. One extreme was
defined as "Comfortable/No Pain". The other extreme was
defined as "Pain As Bad As It Could Be." The subject was
asked to mark the line at a point corresponding to the
severity of his or her pain at the moment. The VAS ratings
were made on the Morning Questionnaire (Appendix F).
The distance of the mark from the end of the scale was
taken to represent the subjects pain severity. This method
has proven to be a simple, robust, sensitive, and reliable
approach to assessing pain (Huskisson, 1983).
Daily Supply Book
The Daily Supply Book was a one-inch looseleaf
notebook. The front pocket contained instructions for
either the Brux-Monitor, or the Brux-Alarm (see Appendix I).
These instructions reviewed the daily procedures. Behind
these instructions was the original informed consent form,
which overviewed the experiment as a whole. Two vinyl
pencil/pen pouches were contained in the front of the ring
binder of the notebook. The first pouch contained seven
disposable 70 percent isopropyl alcohol prep pads. The
second contained seven reclosable (Ziplock) sandwich bags
each containing three disposable electrodes. Following the
two pouches on the ring binder were the forms that had to be
completed daily. Each day's supply was separated by a
plastic tabbed manila notebook divider. The first sheets
for all days were the Nighttime Questionnaires (containing
space to write bedtime, the SSS, and instructions to
continue on the next page where the POMS was located). This
was to be completed before retiring.
During alarm treatment, the next page contained
instructions and the Nightly Alarm Record form for recording
alarm soundings. This form contained numbered spaces to
record the time for up to 93 alarm soundings.
The last form for all subjects was the Morning
Questionnaire. This contained instructions and spaces to
record the monitor readings from the previous night. It
contained also spaces to record the time of awakening, the
VAS, and the SSS.
Procedure and Experimental Protocol
In the first meeting, or telephone contact, with each
potential subject the investigator said: "Psychologists
often successfully treat bruxism by providing patients with
an alarm that they wear when asleep. If they brux, the
alarm sounds a warning tone that wakes the person up and
signals them to stop. We want to determine the
effectiveness of this type of therapy for bruxism. First we
want to determine if you are a good prospect for this type
of therapy. Thus, we will need for you to wear a monitoring
device for three nights. It will measure how much you
clench and grind your teeth. You will be paid $10.00 for
this screening. If we can use you in the study, and if you
do volunteer, it will involve up to four weeks of your time
and quite a bit of effort from you. To repay you for the
effort, we will provide you with the alarms free of charge
for the period of the study. Also, on completion of the
study we will pay you $50.00. If the alarms turns out to be
helpful, you will be provided one at the end of the study at
a 25 percent reduction in cost, or $150.00. If you are
interested, please read this informed consent form. It
tells you about the details."
After signing the informed consent form, the subjects
were trained to use the monitoring function of the the
alarm/monitors, and to use a three day version of the Daily
Supply Book. All individuals were instructed in skin
preparation with alcohol prep pads, and electrode placement.
Each subject was required to demonstrate proper application
and use of the monitor on two successive occasions before
the start of the screening phase.
During the three day screening period all potential
subjects were evaluated by a Dentist in the University of
Florida Dental Occlusion and Facial Pain Center. (All
patients were seen by Michael Henry, D.D.S., a second year
pain-center resident.) He followed an abridged version of
the clinic's standardized examination that incorporated the
three criteria mentioned previously. On the basis of this
exam he did or did not make a diagnosis of bruxism.
Individuals who met the criteria during screening and
during the dental exam were considered as subjects. They
were assigned randomly to groups with an equal group-size
There were two groups that differed only in the
sequencing of no-treatment and alarm-treatment conditions.
The first group received 14-days of baseline, followed by
14-days treatment. The second group received 14-days of
treatment, followed by 14-days of baseline. Figure 2
overviews the experimental design.
Group 1: Alarm Therapy No-Treatment
EMG (EMG Monitoring)
Group 2: No-Treatment
(EMG Monitoring) Alarm Therapy
3 days 14 days 14 days
Figure 2: Overview of the Experimental Design
During the first meeting following screening, subjects
either were shown how to use the alarm function of the
alarm/monitors or they were told to continue monitoring as
in the screening phase. Subjects were then given a
seven-day version of the Daily Supply Book appropriate to
their condition. An appointment for three to four days
later was then made in order to change batteries in the
alarm/monitor. In addition, subjects were informed that the
investigator would contact them in two days in order to
ensure that the procedures had been explained properly. The
investigator called subjects at least twice weekly
throughout the study to prompt adherence and to troubleshoot
The second meeting was brief. Batteries were changed,
and an appointment to take place seven-days after the first
meeting was made.
At the third meeting, the first Daily Supply notebook
was collected and new ones provided. Batteries were changed
again, but this time all devices were reevaluated with the
calibration procedure described above. No threshold drift
was detected. A fourth meeting three to four days later was
held in order to change batteries.
Seven-days following the third meeting, a fifth meeting
was held to collect materials and check calibration. During
this meeting the alarm mode was switched to that appropriate
for the opposite experimental condition. Some subjects
received instructions in operating the alarm mode of the
alarm/monitor. For other subjects, the monitoring
procedures were reviewed.
Over the following 14-day period four meetings were
held exactly as in the first phase, to change batteries,
check calibration, troubleshoot, and give the seven-day
Daily Supply Book. Additionally, during the last meeting,
subjects were debriefed about the meaning of the
light-display, and shown a graph of their total nightly brux
episodes throughout the study. They were paid at this time
Data for one to two nights in one or both phases were
missing for four subjects. This happened for several
reasons. In some cases the subject forgot, or was unable to
comply. In other cases, the batteries failed, or the
electrodes fell off during the night. Phases were extended
for appropriate numbers of nights for these four subjects in
order to compensate for the missing data.
The questions of interest are reviewed below. The
overview is followed by the relevant statistical analyses
1.) The average nightly brux-episode frequency during
the no-treatment and treatment conditions were compared
within and between groups. The question of interest was
whether EMG-activated alarm therapy would reduce the
frequencies of bruxing episodes. Based on previous
literature the directional prediction was made that
brux-episode frequencies during alarm-treatment phases would
be lower than during no-treatment (baseline) phases.
2.) The average nightly brux-episode frequencies
during a phase preceding treatment were compared with those
during a phase following treatment. Based on the literature
the prediction was made that bruxing following alarm therapy
would exceed bruxism preceding alarm therapy. Because this
is a between-group comparison and only five subjects were
included in each group, this was a tentative prediction.
The prediction was tentative also because "rebounds"
following alarm therapy have not been observed uniformly.
3.) The nightly self-monitored alarm count and the
automated alarm count during treatment phases were compared
for all subjects during the alarm phases. The prediction
was that self-monitored alarm frequency would be lower than
automatically recorded alarm frequency. This is an informal
prediction based on anecdotal reports (Piccione et al.,
1982) that subjects sometimes sleep through alarm soundings.
4.) The average daily sleepiness, fatigue, and vigor
ratings during alarm treatment and baseline phases were
compared for both groups. This is exploratory work that
involves no prediction. A neutral stance is adopted because
the untoward effects of sleep deprivation are not
overwhelming in the literature (Johnson, 1982) and because
there is no guarantee that the present alarm-therapy package
will work sufficiently to constitute a de facto sleep
5.) A comparison was made of morning facial pain and
discomfort ratings during the treatment and no-treatment
phases. The tentative prediction was that alarm therapy
would reduce the intensity of facial pain insofar as it did
reduce bruxing rates. Therefore ratings on the VAS would be
lower for subjects while they were undergoing alarm therapy.
Impact of Treatment
The group mean hourly brux frequencies during screening
are presented in Table 2. The 30 nightly brux frequency
values from the screening phase were subjected to a 2
(groups) x 3 (days) repeated-measures Analysis of Variance
(ANOVA; BMDP Statistical Software Manual, 1983). This
analysis yielded no significant effects when tested at the
Table 2. Means and Standard Deviations for the Nightly Brux
Responses Per Hour by Group during Screening,
Phase 1, and Phase 2.
Screening Phase 1 Phase 2
M SD M SD M SD
Group 1 2.3 1.6 2.2 2.6 0.9 0.8
Group 2 3.4 2.4 0.7 0.8 1.5 2.1
.15 level. Hence there was no evidence that the groups
differed before treatment and no effort was made to
incorporate pretreatment differences into the logic of
subsequent analyses. This finding also justified
comparisons between groups. The ANOVA summary is presented
in Table J-l. Comparisons within groups were made as well,
and they are described following the between-groups
Average hourly brux frequency for each group during
each phase are shown also in Table 2 and in Figure 4. The
280 nightly brux frequency values were subjected to a 2
(groups) x 2 (phases) x 14 (days) repeated-measures ANOVA.
The ANOVA summary is presented in Table J-2. This analysis
yielded significance for the Group x Phase Interaction
(F=6.18, df=1/8, p<.05). This finding justified additional
post-hoc analyses to identify where the significance
00 0 0
Figure 3: Brux Frequency By Condition and Group.
First, the 140 nightly brux frequency values from Phase
1 were subjected to a 2 (groups) x 14 (days)
repeated-measures ANOVA. The ANOVA summary is presented in
Table J-3. This analysis yielded significance for the Group
Main Effect (F=14.19, df=1/8, p<.01). Inspection of the
means in Table 2 indicates the direction of the significant
difference. The group receiving alarm therapy displayed
lower brux frequencies than did the group who monitored
without the alarm.
Next the 140 nightly brux frequency values from Phase 2
were subjected to a 2 (groups) x 14 (days) repeated-measures
ANOVA. The ANOVA summary is presented in Table J-4. This
analysis failed to yield significant effects. Hence the
significant Group x Phase Interaction in the initial ANOVA
mirrors the finding that significant between-group
differences occurred during Phase 1, but not during Phase 2.
The 140 nightly brux frequency values from the two
treatment phases were subjected to a 2 (groups) x 14 (days)
repeated-measures ANOVA. This ANOVA summary is presented in
Table J-5. The analysis failed to yield significance for
The 140 nightly brux frequency values during no
treatment and during treatment within Group 1 were subjected
also to a 2 (conditions) x 14 (days) repeated-measures
ANOVA. This ANOVA summary is presented in Table J-6. It
yielded significance for the Condition Main Effect (F=11.63,
df=1/8, p<.01). Table 2 indicates directionality. Subjects
bruxed less frequently during treatment than during no
treatment. Visual inspection of the graphs of nightly brux
frequency per hour for subjects 1 through 5 (see Appendix K)
indicates that four of the five subjects replicated the
group effect. Only Subject 3 failed to demonstrate a
convincing response to treatment.
The 140 nightly brux frequency values during treatment
and during no treatment within Group 2 were subjected also
to a 2 (conditions) x 14 (days) repeated-measures ANOVA.
This ANOVA summary is presented in Table J-7. It failed to
yield significance for any effect. Visual inspection of the
graphs of nightly brux frequency per hour for subjects 6
through 10 (see Appendix K) indicates that four of the
subjects remained at brux frequency levels observed during
treatment. Subject 8 appears to have returned to
pretreatment levels during the no-treatment period.
In order to examine the data for evidence of a rebound
effect following a period of treatment the 140 nightly brux
frequency values from the two no-treatment phases were
subjected to a 2 (groups) x 14 (days) repeated-measures
ANOVA. This ANOVA summary is presented in Table J-8. It
failed to yield significance for any effect. Therefore, no
evidence for a "rebound" or a relatively high posttreatment
bruxing frequency was found.
Automated Alarm Count
The mean self-monitored alarm-count and the mean
automated alarm-count for each subject are reported in Table
3. The 140 nightly self-monitored alarm-counts and 140
Table 3. Mean Self-Monitored Alarm-Count (SMAC) and
Automated Alarm-Count (AAC) for the 14 day
treatment phase by Subject.
Subject # SMAC AAC % of AAC
1 .9 3.8 23
2 1.6 3.9 41
Group 1 3 1.8 5.3 34
4 3.1 7.9 39
5 6.0 13.3 45
6 1.2 1.2 100
7 3.6 4.4 82
Group 2 8 6.9 8.6 80
9 0.3 0.9 33
10 2.4 8.0 30
automated alarm-counts were subjected to a 2 (methods) x 14
(days) repeated-measures ANOVA. This ANOVA summary is
presented in Table J-9. This analysis yielded significance
for the Method Main Effect (F=4.52, df=l/18, p<.05).
Collapsing across all subjects the mean self-monitored count
was 50 percent of the automated-count. Inspection of the
average for all subjects on Table 3 reveals that
self-monitored alarm-count varied from 23 percent of the
automated-count in subject 1, to 100 percent in subject 6.
Effects of Treatment on
Vigor, Fatigue, Sleepiness,
Anxiety and Pain
Group mean vigor, fatigue, sleepiness, anxiety, and
pain ratings by phase and group are displayed in Table 4.
Table 4. Means and Standard Deviations for the Daily
Ratings of Fatigue, Vigor, Sleepiness, Tension,
and Pain by Group during Phase 1 and Phase 2.
Group 1 VIG
Group 2 VIG
Note. VIG=Vigor subscale from the Profile of Mood States
(R=0-32); FAT=Fatigue subscale from the Profile of Mood
States (R=0-28); SSS=Stanford Sleepiness Scale (R=0-7);
ANX=Anxiety subscale from the Profile of Mood States
(R=0-36); PAIN=Visual Analog Scale of Facial Muscle
Visual inspection revealed only small differences in
means between groups and phases. However, visual inspection
pointed also to small variances. Therefore, the various
data were subjected to separate ANOVAs. A repeated-measures
Multivariate Analysis of Variance approach was rejected as
unduly complex, given the appearance of nondifferent values
in Table 4.
Initially, 2 (groups) x 3 (days) repeated measures
ANOVAs were performed on each of the five sets of 30 values
acquired from the 10 subjects during the three-day
screening. No significant effects were produced by any of
the five analyses. Hence, pretreatment values for vigor,
fatigue, sleepiness, anxiety, and pain were viewed as
Next, separate 2 (groups) x 2 (phases) x 14 (days)
repeated-measures ANOVAs were done within each of the five
sets of self-report data. For the variables of vigor,
fatigue, sleepiness, and pain these analyses produced no
significant effects (see Appendix J-10, J-ll, J-12, and
The ANOVA summary for anxiety ratings is presented in
Table J-14. This analysis yielded a significant Group x
Phase Interaction (F=6.77, df=1/8, p<.05). As is shown in
Row 5 and Row 11 of Table 4 lower mean anxiety ratings were
obtained during baseline phases than during treatment.
The narrative below discusses the results of the
experiment as they are related to the experimental purposes
described earlier. In each section an attempt is made to
delineate implications of the results and directions for
future research. Also included is a narrative about
theoretical explanations of alarm-therapy mechanisms.
Future research is discussed in that connection as well.
Impact of Treatment
The experiment reported here supports the conclusion
from prior research that EMG-activated alarm therapy can
reduce nocturnal bruxism (Beemsterboer et al., 1978; Casas
et al. 1982; Clark et al., 1981; Kardachi & Clarke, 1978;
Rugh & Solberg, 1975). Specifically, it strengthens the
literature (Moss et al., 1981; Purzycki et al., 1986) in
which manual termination of a nocturnal alarm was included
in intrasubject replication protocols. This conclusion is
based on several comparisons. The comparison between groups
during the first phase shows that subjects who received
alarm therapy had significantly lower rates of bruxing than
did untreated subjects. In addition, the comparison across
phases within Group 1 showed that subjects had significantly
higher bruxing rates during no treatment than during
subsequent alarm therapy. Therefore, both between- and
within-group comparisons support the efficacy of the alarm
therapy protocol. Finally, graphic data for each subject in
Group 1 showed the overall effect in four of the five
Notwithstanding the encouraging results, additional
research is needed to define the boundary conditions of
successful alarm therapy packages with bruxist individuals.
Larger samples of subjects are required. Classifications of
bruxist subjects according to differing bruxing etiologies
will be valuable as etiological data become available.
Variations in the fine grain of alarm-therapy protocols will
be of interest also as the "mechanisms" of alarm therapy
effects become known.
As noted earlier, several studies have found that
post-treatment bruxing actually increased above
pre-treatment bruxing when alarm therapy was used (Funch &
Gale, 1980; Piccione et al. 1982). No similar effect was
seen in the present results even though the experimental
design afforded several opportunities for it to occur.
The research reported by Piccione et al. (1982) is weak
methodologically in several respects (Mealiea & McGlynn,
1986). Funch and Gale (1980), in turn, used only one
subject. Hence, relatively greater confidence can placed in
the data reported here. However, the issue should be viewed
as open to continuing investigation.
Beemsterboer et al. (1978) proposed originally that
wakefulness following alarm soundings should increase
maintenance of treatment effects. Neither Piccione et al.
(1982) nor Funch and Gale (1980) provided for wakefulness
following alarm soundings. Alarm-contingent wakefulness
could be an important determinant of rebound phenomena
associated with alarm therapy protocols. Thus, alarm
features such as the manual reset switch constitute one
important domain for further study. Research might begin by
pitting manual-reset alarm therapy against the more common
paradigm in which instructions for wakefulness are provided
Automated Alarm Count
The validity of some previous conclusions about alarm
therapy effects is doubtful owing to the assumption that
subjects self-monitor accurately (Clark et al., 1981).
Hence, an important purpose of the present study was to
characterize the degree to which subjects were able to
self-monitor alarm sounds during the night. The comparison
between self-monitored and automated counts of alarm
soundings revealed very large differences.
The above finding suggests that instructions to
self-monitor alarm soundings during the night are an
insufficient means of tracking the course of alarm therapy.
It draws into question those reports of successful alarm
therapy in which self-monitored alarm records constituted a
mode of treatment evaluation (Clark et al., 1981).
Additional contingencies are needed to provide more reliable
self-monitoring. In the meantime, research should make use
of automated alarm records.
Effects of Treatment on
Vigor, Fatigue, Sleepiness,
As noted earlier, no discussion of the risks or
contraindications of alarm therapy exists in the literature.
This is surprising given that the treatment disrupts sleep
intentionally. Some laboratory research has investigated
the effect of alarms on sleep staging. One study has
suggested that alarm therapy does influence sleep staging
(Piccione et al., 1982), while another found no effect on
sleep staging (Purzycki et al., 1986). However, no previous
study has examined diurnal sleepiness or mood ratings in
patients undergoing alarm treatment.
No evidence was found here that ratings of sleepiness,
vigor, or fatigue were influenced by alarm treatment. This
benign effect occurred even though the alarm protocol did
produce frequent arousals during the night. However, the
present treatment lasted only two weeks. The very real
possibility exists that maladaptive effects on sleepiness
and/or mood reports would be seen with continuation of
treatment. Hence, a high priority area for future research
is the effects of longer therapy durations on this domain of
Caution is prompted also by the present findings
concerning self-reported anxiety on the POMS. Subjects
reported higher levels of anxiety during treatment phases
than during baseline phases. Even though anxiety ratings
for all subjects remained far below normatively significant
levels, they should not be ignored. Several post hoc
explanations for this effect exist, but they need not be
explored until the result is replicated.
For the time being studies should proceed with added
caution. Informed consent should require statements about
the possibility of moderately increased anxiety during
Effects of Treatment
on Facial Pain
No reductions in daily ratings of facial pain were
found during alarm treatment. The trend was in the
direction of lowered ratings during and following treatment,
but it was not significant. While eight of the subjects
reported a history of facial pain and headaches, only two
had sought treatment. Therefore, they constitute a
subclinical pain population.
The question of whether alarm therapy can reduce some
facial pains remains to be answered by research. Only two
studies have used alarms with facial pain patients (Rugh &
Solberg, 1977; Clarke & Kardachi, 1977) and neither
quantified pain ratings or bruxist behavior systematically.
Theoretical Explanations of
Detailed versions of various theories concerning alarm
effects have not been presented because clinical outcome
research has not derived from theoretical considerations.
However, future research could benefit from a better
integration of theoretical assumptions with experimental
problems. Hence, a theoretical narrative is offered in
DeRisi (1970), and Heller and Strang (1973) proposed
two orthodox theoretical explanations for why alarm therapy
reduces bruxing. DeRisi (1970) described his work as an
extension of Azrin's work on avoidance conditioning within
the operant paradigm (Azrin, 1958, Azrin & Holz, 1966).
Heller and Strang (1973), in turn, briefly cited a
connection with the work of Mowrer and Mowrer (1938) and the
Alarm therapy is relatively straightforward when
conceptualized within the operant paradigm. If the
probability of a bruxist response decreases after contingent
tone presentation, then the tone is a punishing stimulus.
Procedures via which a patient learns to first escape then
avoid a punishing stimulus are referred to as avoidance
An operant analysis of bruxing accesses a large
literature on punishment parameters and predicts the
posttreatment return of bruxing in the absence of a
reinforced competing response (Walters & Grusec, 1977).
Also, such an analysis suggests potentially beneficial
modifications to alarm therapy protocols. One obvious
modification is to schedule alarm soundings intermittently
so as to enhance the durability of punishment effects
(Finley, Rainwater, & Johnson, 1982).
A specific problem for an operant analysis of alarm
therapy effects on bruxing involves whether to construe them
as reflecting punishment or passive avoidance. In general,
competing conceptualizations and research with operant
analyses are likely to be relatively unproductive insofar as
the operant analysis generically is incomplete.
The classical conditioning model of alarm therapy
effects is somewhat more complicated. The alarm serves as
an unconditioned stimulus (UCS) for the response (UCR) of
awakening. The UCS is presented repeatedly in temporal
proximity with masseteric muscle tension. In time,
therefore, the muscle tension becomes a conditioned stimulus
(CS) for awakening (CR).
There are several shortcomings with the Pavlovian
model. In general respondent conditioning is a procedural
paradigm, not an explanatory structure. Thus if
alarm-therapy effects are said to result from respondent
conditioning, then the task becomes that of explaining
respondent conditioning. Not only is the task of explaining
respondent acquisition complex (Delprato & McGlynn, 1984),
but also it draws attention away from the problem of
interest; namely describing variables that influence
alarm-therapy outcomes. Furthermore, the Pavlovian model of
the Mowrer and Mowrer (1938) work is now recognized as
incomplete (Lovibond, 1964, 1972). Hence, drawing parallels
between alarm-therapy for bruxing and the bell-and-pad
treatment of enuresis does not strengthen Pavlovian
interpretations of alarm therapy effects generally.
There are specific problems with a respondent analysis
as well. For example, the CS (bruxing) is also the response
to be eliminated. For the CS to elicit awakening, the
problematic response must occur. In a sense, this model
predicts that the response will never be totally eliminated.
The claim that alarm therapy is analogous to the Mowrers'
(1938) treatment for nocturnal enuresis is specifically
flawed as well. The Mowrers' argued that the bell and pad
worked because the alarm (UCS) was paired with bladder
fullness (CS). After repeated pairings, the CS alone would
be sufficient to elicit arousal from the subject.
Problematically, there is no analogous antecedent for
bruxism. Bladder fullness might sometimes be part of an
arousal/bruxing pattern of behavior. But, as noted earlier,
the Pavlovian view of alarm therapy effects is not
applicable even in such a situation because it does not
account satisfactorily for the events of bell-and-pad
Research within the Pavlovian construction of
alarm-therapy effects would identify precursors to bruxing
that reliably occur within an ideal conditioning window of
.2 to 2 seconds before the response (Kimble, 1961). Then
the alarm would be paired with such precursors. Probably
more effective treatments could be based on work along these
lines. Nonetheless, excessive or sole reliance on a
Pavlovian view of the problem should be avoided for the
various reasons mentioned.
To varying degrees the nocturnal alarm paradigms
represent the larger continuum of biofeedback treatment.
The most representative are procedures in which EMG activity
initiates and terminates actual analog feedback of
suprathreshold values (Kardachi & Clarke, 1977). The least
representative are procedures in which suprathreshold EMG
activity initiates a constant signal that is terminated by
some psychomotor performance that is not related to bruxing
itself (Purzycki et al., 1986). There are a half dozen or
so theoretical models of how biofeedback treatments result
in efficacious outcomes (for a review see Raczynski,
Thompson, & Sturgis, 1982). Despite some overlap (Black,
Cott, & Pavloski, 1977) these models can supplement the
traditional learning approaches such as those just discussed
in suggesting experimental variables and interpreting
experimental results (Mealiea & McGlynn, 1986). Allowances
can be made for unique features and procedural nuances in
the nocturnal alarm paradigms.
Notwithstanding the experimental guidance that is
afforded by learning and biofeedback theories, a
comprehensive program of research on nocturnal
feedback-alarm effects should be guided also by etiological
theories of bruxism itself. In this connection, a major
domain of interest has to do with variables related to
There is a literature comprised of a dozen or so
studies in which nocturnal bruxing was studied in relation
to EEG-defined sleep staging. The most impressive work in
the area suggests that bruxing is differentially associated
with REM sleep (Clarke & Townsend, 1984) and with
transitions between sleep stages (Satoh & Harada, 1972,
1973). Some recent work suggests that bruxing during REM
and NREM sleep might differ fundamentally (Rugh & Ware,
1986). In any event, relatively stable relationships
between nocturnal bruxing and sleep phenomena probably do
exist. Hence research on the effects of alarm-feedback
paradigms on sleep variables is likely to be beneficial, as
is research on treatment efficacy as a function of sleep
phenomena. The work of Piccione et al. (1982) and Purzycki
et al. (1986) constitutes preliminary effort. A high
priority area for research in the near future should entail
use of manual reset feedback alarms under conditions of
night long polysomnographic recording.
SUBJECT RECRUITMENT NOTICE
You Grind Your
eth While Yu'-
Is Your Face Sore When You Wake Up?
Have Roommates Told You That They
Hear You Grinding Your Teeth When
Have Dentists Told You That Your
Teeth Are Worn?
Excessive grinding and clenching is called bruxism.
Bruxism may lead to serious problems such as tooth wear,
facial pain, and damage to the jaw joints. We are
studying a treatment for bruxism and we are looking for
individuals who have this problem. In addition to free
treatment, volunteers will earn money for participating.
For more information call Mr. Jeff Cassist
at 377-2883 between 5:00 p.m. and 10:00 p.m.,
or stop by room D6-11 in the dental tower
STANDARDIZED DENTAL SCREENING EXAM
DENTAL OCCLUSION AND FACIAL PAIN CENTER U. OF FLORIDA
COLLEGE OF DENTISTRY
street city county state zip
TELEPHONE: AGE: DATE OF BIRTH: SEX: M F
OCCUPATION/EMPLOYER: BUS. PHONE:
name street address
city state zip
Tobacco: Yes_ No Cigs_ Pipe_ Smokeless_ Amount
Alcohol: Yes_ No_ Type/Amount
Accident or trauma: Yes No When What
Arthritis: Yes No Comment
Head, neck swelling: Yes No Facial Asymmetry: Yes_ No
EAR:Hearing loss Hyperacusis_ Tinnitus_ Pain_ Stuffy_ Drainage_
SPEECH PROBLEMS: (such as persistent horseness) Yes_ No
HEADACHE: Yes No VascularMuscular Migraine Cluster_ AM_ PM
Right_ Left_ Hours per day_ Comment
ORAL PATHOLOGY: Yes No Comment:
TONGUE: Abnormal: Yes No
GINGIVA: Abnormal: Yes No
LYMPHNODE: Abnormal: Yes No
JOINT LAXITY: 1 tight; 2 = normal; 3 = loose
Wrist: Rt. Lft. Elbow: Rt. Lft. Other:
----------------- ---------- ---------
Right TMJ Left PAIN SCALE Right Musculature Left
0 = NO PAIN
Pain, lat. 1 MILD Masseter, super.
Pain, post. 3 SEVERE Masseter, deep.
Pain, vert. Masseter, trig. pt.
Click, open Masseter, insert.
Click, close ___ Temporalis, ant.
Click, ipsilat __ Temporalis, post.
Click, contra ___Temporalis, insert.
Crep., vert. __Trapezius, ant.
Crep., ipsilat ___ Trapezius, insert.
Crep., contra Splenius Capitis
ROM: open with pain mm. SCM, insert.
open w/o pain mm. SCM, middle
deviation Digastric, ant.
R mm lat. move L mm. Digastric, post.
protrusive mm. pain:Yes_ No ___ Lateral pterygoid
Right VASCULAR Left COMMENT:
SIGNIFICANT DENTAL WORK: Ortho
-Surg Perio Endo
PREVIOUS TREATMENT: Medical Occ adj _TMJ Surg Splints
BRUX, CLENCH, other habit: Yes No_ Day_ Night_ Other
OCCLUSAL FUNCTION: Angle Class: Tooth code: 11-18 21-28
Teeth missing: Replaced with:
Wear facets: Mobile teeth#, class:
Crossbite: R L Teeth# Crossover#
Slide: Lat. Dev. R L Vert. Dev. Ant. Dev.
Work<------------- Balance------------->Work Prot--------V--------
Right Left Right Left
DIAGNOSIS: Noxious occlusion
DEPARTMENT OF CLINICAL PSYCHOLOGY
DENTAL OCCLUSION AND FACIAL PAIN CENTER
__ State Zip
Phone (where to
reach you during the day and evening):
We are asking you to help us get to know you and your problem
better by filling out this questionnaire at your first appointment.
This information will be held in strict confidence and will not be
released to others without your specific written permission. We know
this form is long and will take time, but the information is very
important in helping us determine the correct diagnosis and factors
that may contribute to the problem. It will be used to help provide
you with the best and most comprehensive care. Thus, it is to your
ultimate benefit to answer the questions by yourself as accurately and
sincerely as possible.
Please read each question carefully and answer by writing
clearly, circling carefully, or marking distinctly. We appreciate the
time you will spend in answering these questions carefully.
1.) HOW DID YOU FIND OUT ABOUT THIS TREATMENT PROGRAM?
2.) YOUR PRESENT OCCUPATION
3.) SPOUSE'S PRESENT OCCUPATION
4.) GENDER: MALE
FEMALE (5.) BIRTHDATE _/_/
6.) HEIGHT (feet,inches)
7.) WEIGHT (pounds)
8.) RACE (circle:) White Black Indian Asian
9.) MARITAL STATUS (circle:) Single Married Widowed Divorced