8/12/2019 Dub et al.,2004

1/6

INTRODUCTION

The oral device known as the occlusal splint (OS) is frequently used in themanagement of sleep bruxism (SB) to protect teeth from damage (e.g.,wear or fracture) resulting from forceful jaw muscle contractions or to

reduce concomitant orofacial pain, if present (Pierce et al., 1995; Okeson,

2003). However, the efficacy of the OS in reducing jaw muscle activity

remains controversial. Some studies reported a reduction in SB motoractivity (electromyographic [EMG] recording) when comparisons were

made with recordings from the baseline night, whereas others showed no

effect (Solberg et al., 1975; Okeson, 1987; Rugh et al., 1989; Okkerse et al.,

2002; Sjholm et al., 2002). It should be noted that these studies did not

include a palatal-control device (PCD), which covers the palatal area

without protecting the tooth. The absence of a PCD prevents any conclusion

that occlusal tooth coverage explains OS action. The use of a PCD, in a

limited number of SB subjects, was reported to reduce or have no effect on

SB motor activity (Cassisi et al., 1987; Hiyama et al., 2003).

SB is a parasomnia, an excessive motor activity with tooth-grinding,

that intrudes upon a subject's otherwise normal sleep (Thorpy, 1997;

Lobbezoo and Naeije, 2001). Evidence from recent controlled studies

suggests that most SB episodes are secondary to a cascade of physiologicalevents related to sleep arousal (Okura et al., 1996; Macaluso et al., 1998;

Kato et al., 2001, 2003). The predominant sequence is as follows: a transient

(3-10 sec) brain and heart activation, a rise in muscle tone of jaw openers-

suprahyoid muscles, then rhythmic contractions of jaw-closer muscles with

occasional tooth-grinding. The incidence of sleep arousals in SB subjects is

within the normal range ( 14 arousals/hr of sleep) (Mathur and Douglas,

1995; Boselli et al., 1998; Lavigne et al., 2001a). However, SB episodes

associated with sleep arousals are characterized by a rapid onset of

tachycardia and an important rise in electroencephalographic or

electromyographic activities (Kato et al., 2001). The influences of OS on

sleep arousal are unknown.

Sleep apnea (i.e., cessation of breathing in sleep with hypoxemia and

risk of hypertension, daytime sleepiness) is a health hazard found twice as

often in the general population reporting tooth-grinding than in the normalpopulation (Krieger, 2000; Ohayon et al., 2001). The safety of using OS in

subjects with SB and sleep apnea needs to be assessed. In a recent

preliminary study, it was noticed that, out of 10 subjects with a clear

diagnosis of sleep apnea, the use of OS aggravated respiratory disturbances

(e.g., from a lower to a more severe diagnostic category) in four of them

(Gagnon et al., 2004). However, it was reported by others that OS had no

effect on the mean index of respiratory disturbancesperhour of sleep. Since

individual subject variation was not shown in these studies, we do not know

if some of them had an aggravation (Sjholm et al., 1994; Mehta et al.,

2001; Gotsopoulos et al., 2002).

The objective of the present study was to assess, by the use of a short-

ABSTRACTThe efficacy of occlusal splints in diminishing

muscle activity and tooth-grinding damage

remains controversial. The objective of this study

was to compare the efficacy and safety of an

occlusal splint (OS) vs. a palatal control device

(PCD). Nine subjects with sleep bruxism (SB)participated in this randomized study. Sleep

laboratory recordings were made on the second

night to establish baseline data. Patients then wore

each of the splints in the sleep laboratory for

recording nights three and four, two weeks apart,

according to a crossover design. A statistically

significant reduction in the number of SB episodes

per hour (decrease of 41%, p = 0.05) and SB

bursts perhour (decrease of 40%, p < 0.05) was

observed with the two devices. Both oral devices

also showed 50% fewer episodes with grinding

noise (p = 0.06). No difference was observed

between the devices. Moreover, no changes inrespiratory variables were observed. Both devices

reduced muscle activity associated with SB.

KEY WORDS: sleep bruxism, tooth grinding, bitesplint, randomized controlled study.

Received May 14, 2003; Last revision November 26, 2003;

Accepted March 2, 2004

Quantitative PolygraphicControlled Study on Efficacy

and Safety of Oral Splint Devicesin Tooth-grinding Subjects

C. Dub1,2, P.H. Rompr1,2,C. Manzini1,2, F. Guitard1,2,P. de Grandmont1, and G.J. Lavigne1*,2

1Dpartement de Restauration, Prosthodontics Postgraduate

Program, Facult de mdecine dentaire, Universit deMontral, C.P. 6128, Succursale Centre-ville, Montral(Qubec) H3C 3J7, Canada; and 2Centre d'tude dusommeil, Hpital du Sacr-Cur de Montral, Canada;*corresponding author, Gilles.Lavigne@Umontreal.ca

J Dent Res 83(5):398-403, 2004

RESEARCH REPORTSClinical

398

8/12/2019 Dub et al.,2004

2/6

J Dent Res 83(5) 2004 Splint Devices in Sleep Tooth-grinders 399

term controlled-random design,

whether OS reduced SB motor

activity, influenced sleep variables

(e.g., duration and quality of sleep,

number of arousals), and are safe

with regard to respiratory

parameters (e.g., apnea/hypopnea,

snoring) in young healthy SB

subjects.

MATERIALS & METHODS

PopulationFive young women and four men

(mean age + SEM, 23.7 + 0.9 yrs;

range, 20-29 yrs) with a history of

tooth grinding were selected for this

study. All participants signed a

consent form and received financial

compensation for inconvenience

related to the study. The institutional

ethics committee approved the study.Subjects were recruited by

referrals from clinicians and by

advertising on the University

campus. A history of tooth-grinding

events occurring 3 times or more a

week, as reported by the patient's sleep partner over the preceding

6 mos, was the main criterion for selection (Lavigne et al., 1996;

Thorpy, 1997; Lobbezoo et al., 2001). The presence of tooth wear

ranging from class 2 through class 4 (Johansson et al., 1993) on at

least 3 occlusal surfaces and/or masseter muscle hypertrophy upon

voluntary clenching and/or symptoms of morning orofacial jaw

muscle fatigue were also noted when all subjects were examined.

To be eligible to participate in the study, SB subjects were

required to be between 18 and 45 years of age, have a good

comprehension of French, be able to sign a consent form, and

agree to spend at least 4 nights at the sleep research laboratory.

The first night was for habituation and was not included in the

statistical analysis. The second night was used to record jaw

muscle activity and tooth-grinding sounds to establish baseline

levels and to rule out other sleep disorders. At least 4 phasic (3

muscle contractions at a frequency of 1 Hz) or mixed (phasic and

tonic contractions) episodes of SBperhour of sleep with 2 audible

tooth-grinding events per night had to be present to confirm a

subject's eligibility to participate in the study (Lavigne et al.,

2001a,b). During baseline recording, patients who showed signs of

other sleep disorderssuch as periodic leg movements during

sleep (> 10 events per hour of sleep), electroencephalographic(EEG) epileptiform activity, sleep apnea (> 5 apnea or hypopnea

events per hour of sleep)and snoring were excluded. Also

excluded were patients reporting pain, those who had been treated

with any type of oral device in the preceding 6 mos, those wearing

a partial denture, missing more than 2 posterior teeth (third molars

excluded), presenting gross malocclusion, or taking medication or

alcohol on a regular basis. Finally, a negative history of medical,

neurological, motor, or psychiatric disorders was required for

subjects to be included in the study.

Experimental Procedure and Occlusal Splint FabricationThis crossover study evaluated two oral devices (Fig. 1): a hard

acrylic U-shaped occlusal splint and a palatal device (e.g., not

interfering with the occlusion in any mandibular movements). The

OS was used as the treatment and the PCD as the active control.

The technician scoring sleep and oromandibular activity data was

blind to the type of device used.

Maxillary and mandibular arch impressions were made with

alginate, and models were cast in artificial stone. The centric tooth

relation was taken with a blue wax waffle. A face bow was used to

mount the models on a semi-adjustable articulator. The two oral

devices were made on the maxillary models and then inserted and

adjusted. The OS was adjusted in centric relation with the use of a

32-m articulation paper. Only the points corresponding to contact

between the lower buccal cusp and the splint were preserved. We

adjusted lateral guidance and protrusion by eliminating any contact

other than with the canine in lateral or incisor in anterior-posterior

mandibular movements. The OS was 1-2 mm thick over the incisor

tooth area. The PCD was adjusted for maximum tooth

intercuspation, and any tooth contact upon mandibular movement

was eliminated (Fig. 1). The same operator (CD) provided the

treatments, and each patient was given the same instructions. To

prevent bias toward the design of the oral devices, and since most

subjects expected tooth protection, subjects were told that bothsplints had been reported to be beneficial and that one of the study

goals was to test the efficacy and comfort of both devices. This

"goal" was reinforced with a questionnaire given at the end of the

night and another at the end of the study assessing sleep quality,

oral device comfort, preference, and efficacy.

The first night of sleep laboratory recording was for

habituation. The second night was used for sleep disorders

diagnosis and to establish baseline data. A computer-generated

sequence then randomly assigned which of the two oral devices

was to be worn first by each patient. Patients were given two

weeks to get used to the splint. The subjects then spent a third

night at the laboratory, wearing their first splint, for the collection

Figure 1. Photographs of the occlusal splint (a,b) and palatal control device (c,d) on model and inmouth, respectively.

8/12/2019 Dub et al.,2004

3/6

400 Dub et al. J Dent Res 83(5) 2004

of polygraphic data. The second splint was given on the next

morning and was worn by the subject for two weeks. Further

laboratory recordings were made on the fourth night, with subjects

wearing their second splint. Patient compliance was checked on an

irregular basis by a 'phone call to the patient to ensure that he/she

was using the oral device as requested.

Polysomnographic Recordings and Scored Variables

Sleep recordings were made on each of the 4 nights from 10:30p.m. to 7:00 a.m. The setting of the recordings has been described

elsewhere (Lavigne et al., 1996; Lobbezoo et al., 1997). In

summary, the following surface electrodes were used: 2

electroencephalograms (C3A

2, O

2A

1), one electrocardiogram

(EKG) and bilateral electro-oculograms (EOGs), and

electromyograms (EMGs) from the masseter, sternocleidomastoid,

anterior tibialis, and one site for chin/suprahyoid activities. Data

were collected and amplified with a sampling rate of 128 Hz and

kept for further scoring with the use of sleep recording and scoring

software (Harmonie, Stellate System, Montral, Canada). Audio

and video signals were recorded in parallel. Information on sleep

quality, total duration, efficiency, percentages of stage duration,

number of micro-arousals per hour, number of awakenings per

hour, and sleep latency was calculated. Moreover, the frequency of

SB episodes perhour of sleep, the number of bruxism bursts per

hour, and the number of episodes with sounds were estimated. A

detailed analysis of SB muscle activity was also performed for the

right masseter. For each SB episode, the total episode duration,

number of bursts, number of bursts/sec, mean amplitude of the

bursts (RMS calculation), sum of burst duration, mean burst

duration, and mean interval between bursts were calculated

(Lavigne et al., 1997). The sleep scoring was done according to the

standard criteria of Rechtschaffen and Kales (1968), and the final

diagnosis of SB was made according to previously published

criteria (Lavigne et al., 1996).

Respiratory function was assessed by nasal airflow measures

through a thermistor sensor (Thermocouple, Protech, Woodville,

WA, USA) and a thoracic and abdominal belt. The number of

apnea-hypopnea events per hour of sleep was computed. The

presence of swallowing events was estimated indirectly with the

use of video signals and laryngeal movements as recorded over the

thyroid cartilage with a piezoelectric sensor (Opti-Flex, Newlife

Technologies, Midlothian, VA, USA). This method is a valid and

non-invasive technique currently used in sleep medicine

(Miyawaki et al., 2003). An index of the number of swallowing

eventsperhour was computed based on data from the piezoelectric

sensor.

Statistical AnalysisWe used repeated-measures ANOVA to evaluate treatment effects.

The baseline data were then compared with data from either the

occlusal or palatal nights by paired comparisons. Friedman two-

way ANOVA followed by Wilcoxon signed-ranks tests for paired

comparisons were used when the data distribution was not normal.

We performed sign tests to evaluate whether subjects did or did not

improve with the splints.

RESULTSThe influence of the oral

devices on sleep variables

was that both reduced the

percentage of time that

subjects spent in deep

non-REM sleep (stages 3and 4, Table). However,

simple contrast analysis

revealed trends only

when baseline recordings

were compared with

those with OS (14.9% to

10.6%; p = 0.057) and

PCD (14.9% to 11.0%; p

= 0.085). Although the

duration of stages 3 and 4

was slightly lower during

the nights with oral

devices, no other sleep

variables (e.g., efficiency,

sleep latency, incidence

of micro-arousals or

awakenings) differed

among the 3 recorded

conditions. The presence

of the splints did not

induce an increase in the

respiratory variables,

apnea and hypopnea

index, which remained

low for the whole study.

A non-statistically signif-

Table. Sleep and Bruxism Variables (means + SEM) during Baseline, OS, and PCD Nights

p ValuesVariable Baseline (B) OS PCD overall B-OS B-PCD

Sleep% Stage 1 4.2 + 0.7 6.7 + 1.3 6.4 + 1.5 0.17 0.12 0.13

% Stage 2 56.9 + 2.1 60.2 + 1.9 57.9 + 1.7 0.41 0.25 0.73% Stages 3 & 4 14.9 + 2.4 10.6 + 1.6 11.0 + 1.9 0.038 0.057 0.085% REM 24.0 + 2.1 21.4 + 2.6 24.7 + 2.0 0.40 0.34 0.81Sleep efficiency % 95.8 + 1.5 93.3 + 2.7 95.7 + 1.4 0.54 0.44 0.99Sleep latency (min)a 8.3 [1.0-69.3] 4.3 [0.7-25.3] 3.7 [0.7-32.0] 0.37 0.44 0.52Micro-arousals/hr 9.7 + 2.1 8.0 + 1.9 7.6 + 1.7 0.50 0.38 0.37

Awakenings/hr 3.1 + 0.8 3.8 + 1.0 3.4 + 1.1 0.62 0.40 0.52Apnea + hypopnea/hra 0.4 [0.0-3.8] 0.8 [0.1-2.3] 0.4 [0.0-2.7] 0.92 1.00 0.89Swallowing/hr 7.3 + 1.4 12.2 + 3.4 8.8 + 2.4 0.15 0.12 0.40

Bruxism

OvernightEpisodes/hra 6.3 [3.7-10.5] 3.7 [0.2-8.2] 3.7 [2.8-7.9] 0.016 0.051 0.051

Episodes with noise 22.1 + 4.9 10.9 + 3.9 10.0 + 4.1 0.057 0.058 0.054Bursts/hr 48.4 + 5.7 26.4 + 6.6 28.2 + 5.6 0.026 0.048 0.046% Episodes in stages 1 & 2 80.7 + 3.4 77.7 + 10.2 83.3 + 4.7 0.80 0.78 0.63

Within an episodeTotal episode duration (sec) 18.1 + 2.1 18.6 + 2.1 17.0 + 1.4 0.82 0.87 0.66Bursts 6.4 + 0.3 4.6 + 0.5 4.4 + 0.5 0.007 0.02 0.003Bursts/sec 0.44 + 0.09 0.24 + 0.01 0.25 + 0.01 0.027 0.067 0.060Burst amplitude (V) 26.0 + 2.7 31.1 + 4.7 27.5 + 1.5 0.48 0.33 0.62Total burst duration (sec) 6.8 + 0.5 5.6 + 0.4 5.1 + 0.3 0.002 0.019 0.004Mean burst duration (sec) 1.6 + 0.2 2.0 + 0.3 1.8 + 0.2 0.18 0.10 0.093Interval between bursts (sec) 0.65 + 0.08 0.93 + 0.06 0.91 + 0.04 0.018 0.063 0.018

a Median [min-max].

8/12/2019 Dub et al.,2004

4/6

J Dent Res 83(5) 2004 Splint Devices in Sleep Tooth-grinders 401

icant increase in the number of

swallowing events per hour was

observed with the OS (67%; p =

0.12).

The median number of SB

episodes pe r hour of sleep was

lower compared with the baseline

(Table, Fig. 2a) when OS and PCD

were used (41% reduction; p =

0.051). This result occurred in eight

of the nine SB subjects (Fig. 3; sign

test, p = 0.04). The number of SB

episodes with tooth-grinding

sounds was decreased by 51% and

55% with OS and PCD,

respectively (Table, Fig. 2b; p