Passive tactile sensibility in edentulous subjects treated with dental

implants: A pilot study

Ali M. El-Sheikh, BDS, MSD, MSc, PhD,a John A. Hobkirk, BDS, PhD, MD,b Peter G. T.
Howell, BSc, BDS, PhD,c and Mark S. Gilthorpe, BSc, PhDd
University College London, London, United Kingdom; University of Leeds, Leeds,
United Kingdom; Tanta University, Tanta, Egypt

Statement of problem. Edentulous patients treated with implant-supported prostheses have shown increased
passive tactile sensibility compared with those using conventional complete dentures. This is thought to be due
to the close mechanical coupling between the implant and bone via the osseointegrated interface, yet the
phenomenon has received little attention.
Purpose. The purpose of this study was to measure passive tactile sensibility in a group of edentulous subjects
treated with dental implants, and to relate the measured sensibility to a range of factors thought to be of possible
relevance, namely, patient age, gender, time since implant placement, implant length, and implant separation.
Material and methods. Twenty edentulous subjects successfully treated with 2 or more Nobel Biocare dental
implants in the anterior mandible were studied. The inclusion criteria were : (1) age of less than 50 years, (2) a
period of at least 12 months since implant placement, (3) implant length of at least 10 mm and of standard
diameter (excluding narrow and wide platform designs), and (4) implant separation of at least 18 mm. Using a
computer-controlled custom-made device, pushing forces (2.1, 2.4, 2.7, and 3.0 N/s) were applied directly and
perpendicular to the long axes of the implant abutments until the subjects felt the first sensation of pressure. The
magnitude of these forces was measured with an integral transducer. The applied force had a ramped staircase
pattern, and force application rates were varied between 2.1 and 3.0 N/s. Multilevel modeling was used to
analyze the collected data (␣⫽.05).
Results. The threshold values of passive tactile sensibility ranged between 3.1 and 15.7 N (mean 10.9; SD 3.9).
Analysis failed to show any significant association between passive tactile sensibility and the variables studied.
Conclusion. Within the limitations of this study, which included a small sample size, no relationship was found
between passive tactile sensibility associated with long-standing implants and any of the variables studied (age,
gender, time since implant placement, implant length, and implant separation). (J Prosthet Dent 2004;91:
26-32.)

CLINICAL IMPLICATIONS
This pilot study suggests that the decisions made by the dentist on implant length and separation
may have no significant impact on the oral sensations tested in both male and female patients,
irrespective of age.

L ongitudinal studies have shown that implant-sup-

ported mandibular prostheses can be a successful substi-
sibility, such as the detection threshold of pressure, de-
pends largely on the presence of the periodontal ligament
tute for missing teeth in edentulous patients.1,2 Re- receptors.4 As a result, there is a reduction in passive tactile
search, although only in small groups, supports the sensibility in patients with partial or complete loss of the
clinical view that patients wearing implant-supported periodontal ligament. What sensation remains may result
complete dentures have a lower passive tactile threshold from the remaining receptors in the alveolar bone,5 inner
than do those wearing conventional complete dentures.3 ear,6 jaw muscles,7 and temporomandibular joints,8 which
In dentate and partially dentate subjects, passive tactile sen- might contribute part of the normal exteroceptive func-
tion. Periodontal receptors may not be totally destroyed
a
Lecturer, Department of Prosthetic Dentistry, Faculty of Dentistry, and can still evoke a response in the mesencephalic nucleus
Tanta University. when stimulated.9,10 Receptors in the bone11 and soft tis-
b
Professor and Head, Unit of Prosthetic Dentistry, Eastman Dental sue surrounding the implants may increase the mechano-
Institute, University College London.
c
receptive sensitivity.
Senior Lecturer, Unit of Prosthetic Dentistry, Eastman Dental Insti-
tute, University College London. The phenomenon of passive tactile sensibility and its
d
Reader in Statistical Epidemiology, and Head, Biostatistics Unit, comparison with the sensibility of natural teeth have
School of Medicine, University of Leeds. been studied by several groups,12,13 and it has been

26 THE JOURNAL OF PROSTHETIC DENTISTRY VOLUME 91 NUMBER 1

EL-SHEIKH ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

hypothesized that a number of different factors influ- Table I. Details of test group and related variables
ences tactile perception.14,15 Among these are the dura-
Time since Length of
tion of the stimulus exerted, rate of application and di- implant right Implant
rection of the forces, and sensory fatigue. Jacobs and van placement implant separation
Subject Gender Age (y) (mo) (mm) (mm)
Steenberghe3 reported that no relationship was found
between tactile perception and the subject’s age and 1 F 62 110 13 23
implant function time, whereas Keller et al16 found little 2 M 63 12 15 24
change in tactile perception during the healing phase, in 3 F 63 13 18 28
contrast to the findings of Lundqvist and Haraldson.17 4 M 78 40 15 32
Muller18 reported age-related effects on tactile percep- 5 M 72 25 13 36
tion, and Hammerle et al13 found gender-related differ- 6 F 69 14 18 25
7 F 50 12 18 18
ences. While there are no reports on the effects of im-
8 M 82 36 15 29
plant length and separation on tactile perception, it
9 F 59 58 13 28
could be postulated that because these are associated 10 M 61 15 15 20
with different strain patterns in the jaw when loaded, 11 F 62 52 15 26
altered patterns of perception may occur as different 12 F 53 102 15 27
groups of nerve endings are stimulated. For mechanical 13 F 69 50 15 28
force application, pushing forces are preferred to tap- 14 M 68 13 18 25
ping forces in order to discriminate receptors in the 15 M 61 12 15 29
peri-implant environment from more distant recep- 16 F 53 17 10 28
tors.19 17 F 76 68 18 25
The tactile threshold of passive tactile sensibility is re- 18 F 62 12 18 24
19 F 75 66 15 22
ported as being significantly lower for teeth than for im-
20 M 63 36 15 20
plants, in partially dentate or edentulous patients with im-
plant-supported fixed prostheses or overdentures.20,21
Few data are available on passive tactile sensibility in eden-
tulous patients restored with implant-supported mandibu- Table I. Because the right and left implants had the same
lar prostheses. length, the length of the right implant for each subject
The aim of this study was to measure passive tactile was selected as representative of the overall trends. The
sensibility in edentulous subjects treated with dental im- subjects had no evidence of local or systemic disease that
plants, using pushing forces applied directly and perpen- might have influenced the outcome of the study, and all
dicularly to the long axes of the implant abutments. gave informed consent to participate in the study, which
Variables—namely, age, gender, time since implant had institutional approval for the use of human subjects.
placement, implant length, and implant separation— Test loading was accomplished with a precision cus-
that might affect the threshold value of passive tactile tom-made device (Fig. 1) that consisted of 2 square-
sensibility were also considered. section titanium cranked arms, 22 cm long and 9 mm
thick. These were joined together in a precision box
joint that restricted the opening and closing movements
MATERIAL AND METHODS to a single plane. Each arm had two 2.3-mm holes in the
The test group consisted of 20 edentulous subjects free end parallel with the long axis of the arm. The holes
treated at least 12 months previously with 2 or more were used for mounting 2 steel rods, 24 mm long and
dental implants (Nobel Biocare, Gothenburg, Sweden) 2.3 mm in diameter, to act as loading probes, which
in the anterior region of the mandible. The treatment were secured in position with screws. This configuration
was successful, based on the criteria suggested by Al- provided 3 different offsets to change the position of the
brektsson et al.22 Twelve subjects were women and 8 rods according to implant separation. Each steel rod had
were men, with a mean age of 65.1 ⫾ 8.3 years and a a 1.5-mm– diameter radial groove 3 mm from its end to
range of 50 to 82 years (Table I). All patients used ensure reproducible location on the abutments. The end
implant-supported mandibular prostheses that occluded of each steel rod was covered with a plastic sleeve, 8 mm
with maxillary conventional complete dentures. The se- long and 1.6 mm in diameter, to cushion the contact
lected subjects had at least 2 implants of the same length. between the rods and the implants during load applica-
The transmucosal abutments (TMAs) (Nobel Biocare) tion. The plastic sleeves were changed for each subject.
on the top of these 2 implants were parallel to each other The movement of the loading probe was operated by
within ⫾5 degrees to ensure that test forces would be a computer-controlled stepper motor. This consisted of
applied perpendicularly to the long axes of the implant a programmable stepper motor control board (Stock
TMAs. Details regarding time since implant placement, No. 440-098; RS Components Ltd, Corby, Northamp-
implant length, and implant separation are shown in tonshire, UK) controlling a bipolar stepper motor drive

JANUARY 2004 27
THE JOURNAL OF PROSTHETIC DENTISTRY EL-SHEIKH ET AL

Fig. 1. Precision custom-made loading device (left) and loading probes (right) with strain gauge (arrow) for sensing applied
loads.

board (Stock No. 255-9065; RS Components) that in

turn controlled a stepper motor (Stock No. 318-705;
RS Components) via control software (Stock No. 440-
105; RS Components) installed in a personal computer
(Tandon 486 SX25; Tandon Computers, Worcester-
shire, UK). A strain gauge (2N-120-PC11-A; TSM Ltd,
Coleraine, Londonderry, Northern Ireland) was
mounted longitudinally on the surface of the left rod
perpendicular to the horizontal plane to serve as a sen-
sor. This was connected in a quarter bridge configura-
tion to a strain-gauge amplifier (Microlink 3052 16
channel, Microlink Products; Biodata Ltd, Manchester,
UK) mounted in a frame (Microlink 3000; Biodata).
This also housed a Microlink 3042 16 bit A-D converter
(Biodata) that enabled transmission of data via an
IEEE488 interface to a personal computer (Optiplex; Fig. 2. Loading probes in position between notches of im-
Dell Computer Corp, Berkshire, UK) using computer pression copings.
software (Microsoft Windows 98; Microsoft Corp, Red-
mond, Wash). Data were captured and subsequently
analyzed using proprietary software packages (Windmill
Data Acquisition Software, and FAMOS Data Analysis to the long axes of the impression copings. Contact
Software; Biodata). between the lips and any part of the device was avoided
The experiment was carried out in a quiet room. Each to prevent triggering of the remote receptors. The push-
subject was comfortably seated in a dental chair in an ing forces applied directly between the abutments had a
upright position with his or her head resting against the ramped staircase pattern. The selected rates of force ap-
head support. The maxillary denture and the implant- plication were of 2.1, 2.4, 2.7 and 3.0 N/s. The force
retained superstructure were removed and 2 square im- application rates were separated from each other by 2
pression copings (#DCB026; Nobel Biocare) were seconds. The force application rates were the most suit-
placed onto the abutments. The force sensor was ad- able rates based on a pilot study developed as a part of
justed to ensure that electrical zero occurred before the this study. An interval of 2 seconds was used between
grooves of the loading probes were brought into light each stimulus in order to avoid fatigue of the receptors,
contact with the notches of the impression copings (Fig. because van Steenberghe and de Vries15 reported that
2). The device was carefully held by hand at right angles the fatigue occurred with an interval of 1 second be-

28 VOLUME 91 NUMBER 1
EL-SHEIKH ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

tween each stimulus. Fatigue is due to incomplete and tercept (which when evaluated is equivalent to the mean
temporary recovery of the receptor’s excitability after a outcome value) and variance terms, 1 for each level of
stimulus. The force amplitude was increased in discrete the defined hierarchy. The VC model was derived to
steps with a tip velocity of 30 ␮m/s until the subject felt partition the total modeled variance across all levels of
the first sensation of pressure, which was indicated using the defined hierarchy, in order to establish that there was
a finger-operated microswitch. The time lag of the sub- a significant and sizeable degree of variation at each
ject’s reflex was not measured. The applied force was level. The VC model takes the form:
continuously measured with the force sensor and re-
corded electronically for subsequent analysis. y ij ⫽ ␤ 0 ⫹ e ij ⫹ u j
Before the assessment of threshold values, each sub-
ject had a training session to become accustomed to the where yij is the outcome (threshold value) on measure-
loading device and to practice detection of the threshold ment occasion i (i⫽1...5) for subject j (j⫽1...20); ␤0 is
forces. The subject was then allowed to rest for a quarter the intercept (mean threshold) for all measurements
of an hour before 5 measurements were made at inter- across all subjects; eij is the random variation across re-
vals of 2 minutes. The loading device was calibrated peated measurement occasions (often thought of as
before and after each recording session to confirm its measurement error and/or biological variation), which
linearity and reproducibility. The tip of the force sensor has a mean of 0 and is assumed to be normally distrib-
was loaded with a series of weights (0.2-1.8 kg) hanging uted with variance (square of the standard deviation)
vertically at a right angle to it and the outputs recorded. ␴e2; and uj is the random variation across all subject’s
This procedure was repeated 10 times for each weight. mean threshold scores, which also has a mean of 0 and is
These data were then used to calculate the forces applied again assumed to be normally distributed with variance
by the loading device. The mean of the 5 repeated mea- ␴u2.
surements was calculated and used as the threshold value Initially, the VC model was explored for the specified
of the passive tactile sensibility for that subject. The 5 2-level structure. In turn, each covariate was added to
repeated measurements also were used for the statistical the VC model and all other covariates excluded. This
analysis with multilevel modeling (MLM).23 approach assessed the degree to which each covariate
was associated with the outcome, while ignoring the
potential influence of other covariates. The next model
Statistical analysis explored the association between outcome and all co-
The outcome of this study was the determination of variates simultaneously, accounting for the combined
tactile threshold, as detected by edentulous subjects ex- influence of all covariates. This approach assessed the
periencing pushing forces on dental implants. Factors degree to which each covariate was associated with the
which were considered potentially to influence this out- outcome, while accounting for the potential influence of
come were age, gender, time since implant placement, all other covariates simultaneously. The effect of re-
implant length, and implant separation. The data col- peated measurements on the tactile threshold was mod-
lected had a complex structure, comprising repeated eled through the inclusion of time-varying covariates
measurements (5 recordings made consecutively) that centered about repeat 3. For each covariate considered
were clustered within each of 20 edentulous subjects. separate, the model would be:
Although several statistical methods could have been
considered for the analysis of such data, the sophisti- y ij ⫽ ␤ 0ij ⫹ ␤ 1 x j
cated statistical modeling technique, multilevel model-
ing (MLwiN, Version 1.10; Multilevel Models Project, where yij is the outcome as before; ␤0ij ⫽ ␤0 ⫹ eij ⫹ uj
Institute of Education, University of London, UK), de- contains the intercept and random terms for variation at
veloped to analyze clustered data, was used.24,25 This each level of the hierarchy; and ␤1 is the coefficient of the
allowed explicit determination of the magnitude of ef- covariates x, which is selected in turn to be 1 of the
fects for each factor being considered. The data within possible choices of covariate. The full model would thus
this study form a natural hierarchy of observations, with be:
each upper level (for example, subject level 2) contain-
ing a cluster of lower-level units (for example, repeat y ij ⫽ ␤ 0ij ⫹ ␤ 1 age j ⫹ ␤ 2 gender j ⫹ ␤ 3 [Link]
level 1). ⫹ ␤ 4 [Link] ⫹ ␤ 5 [Link] ⫹ ␤ 6 [Link]
The analysis began with the simplest model that con-
sidered no explanatory variables (covariates), and then where yij and ␤0ij are as before; and ␤n (n⫽1...6) are
more complex models were explored in a logical se- the 6 coefficients of the covariates for age, gender,
quence to gain maximum insight into factors associated time since implant placement, implant length, im-
with the tactile threshold. The simplest model, known as plant separation, and the linear term for the repeated
variance components (VC), therefore comprised the in- measurement occasion (coded ⫺2, ⫺1, 0, 1, 2).

JANUARY 2004 29
THE JOURNAL OF PROSTHETIC DENTISTRY EL-SHEIKH ET AL

Table II. Variance components model

Coefficient (SE) P value % variance

Fixed effects
Intercept (mean), ␤0 10.881 (0.932) ⬍0.001
Random effects variances
Subject (level 2), ␴u2 17.378 (5.483) 0.002 99.8
Repeat (level 1), ␴e2 0.030 (0.005) ⬍0.001 0.2
Total modeled variance 17.408 — 100.0

Table III. Fixed part of multilevel models assessing Table IV. Multilevel model assessing association between
association between tactile threshold and each covariate tactile threshold and each covariate, accounting for all
separately other covariates simultaneously

% of Coefficient (SE) P value

explained
Covariate Coefficient (SE) P value variance Fixed Part Covariates
Intercept, ␤0 9.955 (1.598) ⬍0.001
Age 0.59 (0.114) 0.607 0.00* Age, ␤1 0.027 (0.156) 0.862
Gender 1.402 (1.926) 0.467 0.00* Gender, ␤2 2.423 (2.587) 0.349
Time since implant 0.032 (0.032) 0.306 0.35 Time since implant placement
placement (mo), ␤3 0.045 (0.043) 0.294
(months) Implant length (mm), ␤4 ⫺0.173 (0.607) 0.775
Implant length (mm) ⫺0.314 (0.449) 0.485 0.00* Implant separation (mm), ␤5 ⫺0.094 (0.298) 0.752
Implant separation 0.013 (0.231) 0.956 0.00* Repeat—linear, ␤6 ⫺0.023 (0.012) 0.052
(mm) Random Part Variances
Repeat—linear ⫺0.023 (0.012) 0.052 0.18 Subject level, ␴u2 20.106 (6.326) 0.001
*Total variance increased slightly for these models, suggesting that inclusion of Repeat level, ␴e2 0.030 (0.005) ⬍0.001
these non-significant covariates did not yield an improved model. Total modeled variance 20.136

RESULTS considered simultaneously. Furthermore, it can be seen

The threshold values of the passive tactile sensitivity that the total variation in the tactile threshold increased
associated with dental implants ranged between 3.1 and slightly from 17.408 (Table II) to 20.136 (Table IV).
15.7 N. The mean threshold value was 10.9 ⫾ 3.9 N and
DISCUSSION
was not different from the multilevel VC model esti-
mate, ␤0 ⫽ 10.88 (SE 0.93 N). The VC model demon- Since endosseous implants are firmly anchored in the
strated that variation at all levels of the specified hierar- jaw bone, it is possible that the receptors in the bone11
chy significantly differed from 0, indicating that lower- and soft tissues surrounding the implants may increase
level outcomes were not strictly independent (Table II); the mechanoreceptive sensitivity. This study was de-
analyzing these data while ignoring the inherent hierar- signed to apply pushing forces directly and perpendicu-
chical structure would therefore be erroneous. Approx- larly to the long axes of the implant abutments. A direct
imately 99.8% of the outcome variation was attributable contact between the loading device and the implants was
to differences between subjects (␴u2⫽17.38), whereas achieved to avoid impact forces that could easily be
very little (⬍0.2%) arose from differences between re- heard by the subjects. The principal weakness of this
peated measurements (␴e2⫽0.03). investigation was the relatively small number of subjects
Models were then developed which explored the im- involved. This reflects the difficulties of obtaining clini-
pact of each covariate in turn, in isolation of all other cal data when patient selection is based on strict criteria.
covariates. The fixed part of these models, that is, the It was for these reasons that the statistical technique of
covariate fixed effect, is summarized in Table III. From multilevel modeling analysis was employed. This study
the family of models, none of the covariates operating at nevertheless may assist in selecting the most appropriate
the subject level (age, gender, time since implant place- areas for further investigation, perhaps using a multi-
ment, implant length, and implant separation) demon- center design to maximize the amount of data poten-
strated a significant association with the tactile thresh- tially available.
old. The repeat (linear) covariate operating at the repeat The subjects’ reaction times would have resulted in
level was borderline significant (P⫽.052). Table IV pro- an overestimate of the threshold because this was re-
vides evidence that no covariates had any association corded at the moment of the subject’s response. Using a
with the tactile threshold, even when all covariates were value of 220 milliseconds for reaction time as reported

30 VOLUME 91 NUMBER 1
EL-SHEIKH ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

by Mattes et al,12 it is possible to calculate the threshold variation between subjects. The effect of repeated mea-
error. This would be at its most unfavorable value with surements remained consistent when considered sepa-
the highest load rate and lowest threshold value (21.2% rately from and simultaneously with all other covariates.
estimated error) and at its least with the lowest load rate Nevertheless, at borderline significance, it suggests that
and highest threshold value (2.9%). For an average on average there was a pattern of decreasing threshold
threshold value, the estimated error would have been across the 5 repeated measurements.
4.2 and 6.0% for the lowest and highest load rates re- This study showed that no covariates had any associ-
spectively. This factor is common to all studies that use ation with the tactile threshold, even when all covariates
subjects’ responses as a component of threshold detec- were considered simultaneously, although moderate ef-
tion. fects might be important but could be overlooked be-
Hammerle et al13 studied the threshold for tactile cause the study size did not provide sufficient power to
perception of ITI dental implants and natural teeth in conclusively detect the influences of these variables. The
the same patients. The mean value for the implants slight increase in the variance when all covariates were
ranged from 13.2 to 189.4 g (0.132 to 1.894 N), while considered simultaneously suggests that the inclusion of
for natural teeth, it ranged from 1.2 to 26.2 g (0.012 to nonsignificant covariates did not yield an improved
0.262 N). The difference in the thresholds for tactile model but only confused any assessment of explanatory
perception was about 10-fold. Mericske-Stern et al20 factors influencing the tactile threshold.
reported that the tactile sensation threshold of ITI den- The present study confirmed the earlier finding3 that
tal implants in edentulous subjects varied from 148 to there was no relationship between passive tactile sensi-
1300 gm (1.48 to 13 N) in the vertical direction, bility and the subject’s age and time that an implant was
whereas in the horizontal direction, the forces varied in function. Furthermore, Keller et al16 concluded that
from 118 to 1200 gm (1.18 to 12 N). The reported the absolute threshold of tactile perception with dental
passive tactile sensibility associated with dental implants implants during the phase of osseointegration was not
in the present study was somewhat higher than those of affected by bone and soft tissue healing taking place
previous reports. This may be attributed to the different during the time period.
implant systems used for assessment. Another factor in It is known that there are residual nerves in bone
this difference might be the experimental arrangement following surgery11 and that there is nerve regrowth into
in which triggering of the remote receptors was avoided. the area around an implant.26 Consequently, it has been
Van Steenberghe and de Vries15 found a linear rela- hypothesized that long implants may be associated with
tionship between the rate of force application and the lower threshold values of passive tactile sensibility than
threshold of tactile perception. Slower rates resulted in short ones, since they might be expected to lie closer to
higher threshold values. Slow rates of force application larger numbers of nerve endings. This study failed to
were used in the present study, since a separate pilot demonstrate such an association.
investigation using the same loading device had shown Similarly, no link was found between passive tactile
that slower rates resulted in higher threshold values. sensibility and the subject’s age, time since implant
Mericske-Stern et al21 reported that the horizontal force placement, or the implant separation. The variables of
direction yielded a tendency toward a higher threshold age and gender were included because they have been
than did the vertical force direction when measuring on known to be associated with changes in percep-
implants, whereas with natural teeth no difference was tion.3,13,18 The literature on time-related changes in tac-
found. This could be the reason that higher thresholds tile susceptibility is equivocal in its findings16,17 and this
were found in the present study, because horizontal matter requires further investigation.
forces were applied.
The variance components model of the statistical CONCLUSIONS
analysis indicates the need for adopting a statistical ap-
Within the limitations of this pilot study, there was a
proach that respects the inherent data clustering. Simply
variation between and within subjects when measuring
ignoring the hierarchy would be equivalent to assuming
passive tactile sensibility. Passive tactile sensibility asso-
negligible variation at the subject-level. Using the arith-
ciated with implants in function for more than 12
metic means of repeated measurements and analyzing
months was not found to be related to the variables
them as independent observations (which would be
studied, namely, age, gender, time since implant place-
standard practice if the hierarchy was circumnavigated)
ment, implant length, and implant separation.
would lose some of the valuable information collected.
MLM was the preferred technique for this study, al-
though some other statistical techniques can take into REFERENCES
account the within-subject variation. The repeat-level 1. Albrektsson T. A multicentre report on osseointegrated oral implants. J
Prosthet Dent 1988;60:75-84.
variation was very small (⬍0.2%), signifying that the 2. Geertman ME, Boerrigter EM, van Waas MAJ, van Oort RP. Clinical
outcome measures were very consistent with respect to aspects of a multicentre trial of implant-retained mandibular overdentures

JANUARY 2004 31
THE JOURNAL OF PROSTHETIC DENTISTRY EL-SHEIKH ET AL

in patients with severely resorbed mandibles. J Prosthet Dent 1996;75: 18. Müller F. The role of exteroception in oral prosthetic rehabilitation. In:
194-204. Jacobs R, editor. Osseoperception. Leuven: Catholic University; 1998. p.
3. Jacobs R, van Steenberghe D. Comparison between implant-supported 187-197.
prostheses and teeth regarding passive threshold level. Int J Oral Maxillo- 19. van der Glas HW, de Laat A, Carels C, van Steenberghe D. Interactive
fac Implants 1993;8:549-54. periodontal and acoustic influences on the masseteric post-stimulus elec-
4. Jacobs R, van Steenberghe D. Role of periodontal ligament receptors in the tromyographic complex in man. Brain Res 1988;444:284-94.
tactile function of teeth: a review. J Periodontal Res 1994;29:153-67. 20. Mericske-Stern R, Hofmann J, Wedig A, Geering AH. In vivo measure-
5. Jacobs R, Schotte A, van Steenberghe D, Quirynen M, Naert I. Posterior ments of maximal occlusal force and minimal pressure threshold on
jaw bone resorption in osseointegrated implant-supported overdentures. overdentures supported by implants or natural roots: a comparative study,
Clin Oral Implants Res 1992;3:63-70. Part 1. Int J Oral Maxillofac Implants 1993;8:641-49.
6. Siirilä HS, Laine P. The tactile sensibility of the parodontium to slight axial 21. Mericske-Stern R, Assal P, Mericske E, Bürgin W. Occlusal force and oral
loading of the teeth. Acta Odontol Scand 1963;21:415-29. tactile sensibility measured in partially edentulous patients with ITI im-
7. Christensen J, Morimoto T. Dimension discrimination at two different plants. Int J Oral Maxillofac Implants 1995;10:345-53.
degrees of mouth opening and the effect of anaesthesia applied to the 22. Albrektsson T, Zarb G, Worthington P, Eriksson AR. The long-term efficacy
periodontal ligaments. J Oral Rehabil 1977;4:157-64. of currently used dental implants: a review and proposed criteria of
8. Klineberg I. Influences of temporomandibular articular mechanoreceptors success. Int J Oral Maxillofac Implants 1986;1:11-25.
in functional jaw movements. J Oral Rehabil 1980;7:307-17. 23. Goldstein H. Multilevel Statistical Models. London: Edward Arnold; 2002.
9. Linden RW, Scott BJ. The effect of tooth extraction on periodontal ligament 24. Gilthorpe MS, Maddick IH, Petrie A. Introduction to multilevel modeling
mechanoreceptors represented in the mesencephalic nucleus of the cat. in dental research. Community Dent Health 2000;17:222-26.
Arch Oral Biol 1989;34:937-41. 25. Gilthorpe MS, Griffiths GS, Maddick IH, Zamzuri AT. The application of
10. Bonte B, Linden RW, Scott BJ, van Steenberghe D. Role of periodontal multilevel modeling to periodontal research data. Community Dent
mechanoreceptors in evoking reflexes in the jaw-closing muscles of the Health 2000;17:227-35.
cat. J Physiol 1993;465:581-94. 26. Wang Y-H, Kojo T, Ando H, et al. Nerve regeneration after implantation
11. Lambrichts I. Histological and ultrastructural aspects of bone innervation. in peri-implant area. A histological study on different implant materials in
In: Jacobs R, editor. Osseoperception. Leuven: Catholic University; 1998. dogs. In: Jacobs R, editor. Osseoperception. Leuven: Catholic University;
p. 13-20. 1998. p. 3-11.
12. Mattes S, Ulrich R, Mühlbradt L. Detection times of natural teeth and
endosseous implants revealed by the method of reaction time. Int J Oral Reprint requests to:
Maxillofac Implants 1997;12:399-402. DR J. A. HOBKIRK
13. Hammerle CH, Wagner D, Bragger U, et al. Threshold of tactile sensibility DEPARTMENT OF PROSTHETIC DENTISTRY
perceived with dental endosseous implants and natural teeth. Clin Oral EASTMAN DENTAL INSTITUTE
Implants Res 1995;6:83-90. 256 GRAY’S INN ROAD
14. Anderson DJ, Hannam AG, Matthews B. Sensory mechanisms in mam- LONDON/WC1X 8LD
malian teeth and their supporting structures. Physiol Rev 1970;50:171-95. UNITED KINGDOM
15. van Steenberghe D, de Vries JH. Psychophysical threshold level of peri- E-MAIL: [Link]@[Link]
odontal mechanoreceptors in man. Arch Oral Biol 1978;23:1041-49.
16. Keller D, Hammerle CH, Lang NP. Thresholds for tactile sensitivity per- 0022-3913/$30.00
ceived with dental implants remain unchanged during a healing phase of Copyright © 2004 by The Editorial Council of The Journal of Prosthetic
3 months. Clin Oral Implants Res 1996;7:48-54. Dentistry.
17. Lundqvist S, Haraldson T. Oral function in patients wearing fixed pros-
thesis on osseointegrated implants in the maxilla. Scand J Dent Res
1990;98:544-49. doi:10.1016/[Link].2003.10.015

Availability of JOURNAL Back Issues

As a service to our subscribers, copies of back issues of The Journal of Prosthetic Dentistry for
the preceding 5 years are maintained and are available for purchase from the publisher, Elsevier Inc.,
until inventory is depleted. Please write to Elsevier Inc., Subscription Customer Service, 6277 Sea
Harbor Dr, Orlando, FL 32887, or call 800-654-2452 or 407-345-4000 for information on avail-
ability of particular issues and prices.

32 VOLUME 91 NUMBER 1