Journal of Dentistry 70 (2018) 104–109

Contents lists available at ScienceDirect

Journal of Dentistry
journal homepage: www.elsevier.com/locate/jdent

Periodontal status of tooth adjacent to implant with peri-implantitis T

a,⁎ a a b c
Cheng-En Sung , Cheng-Yang Chiang , Hsien-Chung Chiu , Yi-Shing Shieh , Fu-Gong Lin ,
Earl Fud
a
Department of Periodontology, School of Dentistry, Tri-Service General Hospital and National Defense Medical Center, Taipei, Taiwan, ROC
b
Department of Oral Diagnosis and Pathology, School of Dentistry, Tri-Service General Hospital and National Defense Medical Center, Taipei, Taiwan, ROC
c
School of Public Health, National Defense Medical Center, Taipei, Taiwan, ROC
d
Department of Dentistry, Tzu Chi General Hospital – Xindian Branch, New Taipei City, Taiwan, ROC

A R T I C L E I N F O A B S T R A C T

Keywords: Objectives: To evaluate the relationship between peri-implantitis and the periodontal health of the adjacent
Dental implant tooth, the periodontal status of the teeth adjacent and contralateral to the implants with and without peri-
Peri-implantitis implantitis.
Peri-mucositis Methods: Fifty-three subjects with existing dental implants and chronic periodontitis were examined in this
Risk factor
cross-sectional study. Seventy implants were categorized into peri-implantitis (n = 42) and healthy/mucositis
Periodontitis
(n = 28) groups. The periodontal and peri-implant status, including probing depth (PD), clinical attachment
level (CAL), and gingival recession (GR) were measured at 6 sites around the implants and the teeth adjacent and
contralateral to those implants. In total 560 sites of the 70 teeth/implant sets, the association between the
periodontal status at the near and away sites of the teeth (according to implant) and the implant status (without/
with peri-implantitis) was examined.
Results: A significantly different mean PD (5.01 ± 1.69, 4.42 ± 1.8, 3.55 ± 0.88, and 3.71 ± 1.07 mm,
p < 0.001) and CAL (6.02 ± 2.36, 4.89 ± 2.04, 4.35 ± 1.11, and 4.35 ± 1.5 mm, p < 0.001) were noted at
the near sites of the teeth adjacent to the implants with peri-implantitis when compared with the away sites of
adjacent and contralateral teeth and the near sites of contralateral teeth. With generalized estimating equation
(GEE), the presence of peri-implantitis (β = 1.041 mm, confidence interval = 0.646–1.435, and p < 0.001; β
= 0.857 mm, confidence interval = 0.279–1.434, and p < 0.004) and tooth location (β = 0.65 mm, confidence
interval = 0.4–0.9, and p < 0.001; β = 0.682 mm, confidence interval = 0.34–1.024, and p < 0.001) were
significantly associated with the values of the PD and CAL of the teeth. Moreover, the factor of examining sites
(i.e. near and away sites of the tooth) was significantly associated with CAL (β = 0.304 mm, confidence in-
terval = 0.019–0.588, and p = 0.036) and GR (β = 0.136 mm, confidence interval = 0.02–0.252, and
p = 0.022).
Conclusion: The existence of peri-implantitis, the tooth location, and the examining site are significantly asso-
ciated with the periodontal measurements of the remaining teeth.
Clinical significance: Peri-implant health is related to the periodontal health of the natural teeth close to the
dental implant.

1. Introduction surrounding an implant [3]. Peri-implant mucositis affects the soft

Implant therapy was introduced to dentistry 50 years ago and has
become one of the routine procedures for replacing the missing tooth.
In spite of a relatively high long-term survival rate of dental implants,
the peri-implant disease may still lead to some potential problems such
as discomfort of the patients, additional surgical or nonsurgical treat-
ments, and the associated costs [1,2]. Peri-implant disease is a collec-
tive term describing the inflammatory reactions in the tissue
tissue around dental implants without signs of loss of supporting bone,
whereas the peri-implantitis causes loss of supporting bone. Peri-im-
plant disease is a multifactorial disease with an origin of infection. Both
periodontal and peri-implant diseases share many risk factors, in-
cluding poor oral hygiene, smoking, and diabetes [4].
Periodontitis is an important risk factor for the development of peri-
implant disease [4]. Many studies have shown that the individuals with
a history of periodontitis had a higher risk of developing peri-

⁎
Corresponding author.
E-mail address: peter71130000@hotmail.com (C.-E. Sung).

https://doi.org/10.1016/j.jdent.2018.01.004
Received 6 April 2017; Received in revised form 5 January 2018; Accepted 6 January 2018
0300-5712/ © 2018 Elsevier Ltd. All rights reserved.
C.-E. Sung et al. Journal of Dentistry 70 (2018) 104–109

implantitis [4–7]. Periodontally compromised patients may experience
more implant loss, more bone loss,and more peri-implantitis [5]. A
recent cross-sectional study showed that the peri-implant conditions
were significantly related to the general periodontal conditions of the
remaining natural teeth [8]. However, the correlation of periodontal
condition of the tooth/site next to or away from the implant to the peri-
implant condition has never been assessed. There is also no information
available on the above mentioned correlation in patients with existing
periodontitis. Therefore, this study aimed to 1) evaluate whether the
periodontal status of the tooth/site adjacent or contralateral to an im-
plant could be correlated to the peri-implant condition of the implant in
the patient with periodontitis, and 2) further elucidate the confounding
factors of the periodontal status of the teeth adjacent or contralateral to
the implants.
2. Material & methods
2.1. Collection of patients
All the subjects were new patients presented to the Periodontal
Division, Dental Department of Tri-Service General Hospital, Taipei,
Taiwan for the needs of periodontal treatments from July 2009 to
September 2013. The project and protocol were approved by the in-
stitutional review board of Tri-Service General Hospital, National
Defense Medical Center (1-104-05-118). A total of 53 patients, 25 fe-
males and 28 males, with chronic periodontitis (having > 30% of teeth
with BoP, ≥4 mm probing pockets, and the presence of bone loss as
evidenced by radiographic images) were included [9].
The age of the subjects ranged from 42 to 74 years old with the
mean age being 58.6 years old (Table 1). Inclusion criteria were: 1) had
received implant surgery at least 1 year ago, 2) had at least one implant
in the posterior edentulous ridge, 3) there were at least one adjacent
natural tooth and one contralateral tooth around the dental implant,
and 4) had full mouth periodontal charting data (including probing
depth, gingival recession, and clinical attachment loss). The exclusion
criteria were: 1) the opposing arch was full edentulous, 2) had history
of periodontal surgical treatment, and 3) had third molar teeth.
probing (BoP) and/or suppuration [10]. When there were two teeth
adjacent to the implant, both teeth mesial and distal to the implant
were evaluated and the tooth with more periodontal bone loss and at-
tachment loss was then selected as the adjacent tooth (Adj-tooth). The
contralateral tooth (CL-tooth) was further chosen from the opposite side
of the jaw by using the Adj-tooth as the index (Fig. 1). Therefore, the
implant, the Adj-tooth,and the CL-tooth were gathered as one set for
each sample. In this study, there were 70 sets obtained, in which 36 sets
were from 36 patients and another 34 sets were from 17 patients (2 sets
from each patient). All examined natural teeth were categorized as
having normal-mild (PD ≤ 4 mm), moderate (4 mm < PD < 7 mm),
or advanced (PD ≥ 7 mm) periodontitis according to the measurement
of the initial PD using a method modified from a previous study [11].
2.3. Periodontal and peri-implant status of tooth/implant
Periodontal and peri-implant examinations were performed around
the teeth and the implants. PD was measured using a Michigan‘O’ probe
with Williams markings and the probing tip was inserted into the gin-
gival sulcus parallel to the long axis of the tooth until slight resistance
was met. Gingival recession (GR) was measured for the distance from
the gingival margin to the cement-enamel junction (CEJ) at each ex-
amination site. Clinical attachment loss (CAL) was recorded as PD plus
GR. The assessment of the above parameters was measured at six as-
pects (i.e. mesio-buccal, buccal, disto-buccal, disto-lingual, lingual, and
mesio-lingual sites) of the tooth and implant,and presented as the mean
for each tooth and implant. The tooth mobility and furcation involve-
ment of natural teeth were also recorded at the same time. In order to
further evaluate the influence of the periodontal status of a specific site
of the tooth on the peri-implant conditions, the mesio-buccal, disto-
buccal, disto-lingual, and mesio-lingual sites of each Adj/CL tooth were
divided into near site and away site according to the location related to
the implant (Fig. 1). In each near or away site, an average of the per-
iodontal parameters on the buccal and lingual surfaces of the tooth was
used. All measurements were made to the nearest millimeter.
2.4. Statistical analysis

2.2. Selection of implant and teeth In order to compare each measurement of the periodontal/peri-
implant status (i.e. PD, CAL and recession) among the selected im-
In this study, the teeth gathered were according to the selected plants, Adj-teeth and CL-teeth, one-way analysis of variance with
implants that were grouped by the conditions of peri-implantitis and Tukey’s post-hoc tests and Fisher's exact tests for categorical variables
healthy/mucositis. Peri-implantitis was defined when an implant was were used. The independent-samplet tests were used to evaluate the
presented with radiographic bone loss of > 3 mm following implant risk of peri-implant condition (with and without peri-implantitis) based
integration, with the probing pocket depth (PD) ≥4 mm, bleeding on on the each measurement of periodontal status for Adj-teeth and CL-
teeth. The one-way analysis of variance with Tukey’s post-hoc tests
Table 1 were also used to compare each measurement at different sites. To
Demographic of the study subjects and distribution of study implants. examine the impacts of the confounding factors and adjust the within
factors (the subjects and the teeth), generalized estimating equation
Patients Implants obtained
(GEE) method was selected and used to analyze the periodontal status
examined
obtained from 560 sites of the 70 teeth/implant sets according to the
Age 58.66 ± 8a peri-implant status. In this study, the peri-implant condition, the ex-
Gender: Male/Female 28/25(53/ 35/35(50/50%) amining sites (near/away site of the Adj/CL teeth), the buccal/lingual
47%)
examining surfaces, gender, age, implant(s)/person, and the period-
No. of implant per person: 1/2 36/17(68/ –
implant(s) 32%) ontal measurements of the teeth associated with the implants were
Year(s) after functional loading – 5.71 ± 1.87 a (3–10 b) selected as the confounding factors. All statistical analysis was per-
Side of implant inserted: Right/left – 31/39(44/56%) formed by SPSS for Windows (PASW Statistics, version 18.0, SPSS) and
Jaw of implant located: Maxilla/ – 22/48(31/69%) the level of statistical significance was set at p < 0.05.
Mandible
Peri-implantitis: with/without 35/18(66/ 42/28(60/40%)
34%) 3. Results
Status of Periodontitis: n(%)of 12(23%) –
patients with > 30% 3.1. Periodontal status of the implants and the teeth adjacent and
pockets > 4 mm
contralateral to the implants
Total (n) 53 70

a
mean ± SD. The distribution of significant difference in PDs was recorded
b
range of years for loading. among the implants, Adj-teeth and CL-teeth (Table 2). The mean PD for

105
C.-E. Sung et al. Journal of Dentistry 70 (2018) 104–109

Fig. 1. Illustration of the categorized teeth in the panoramic radiography.

White star (*) indicates the Adj-tooth and white cross (†) indicates the CL-
tooth. White (↑) and black arrows (↑) indicate the near and away site of the
Adj-tooth, respectively. White (△) and black arrow head (▲) indicate the
near and away site of the CL-tooth, respectively.

Table 2 adjacent to the implants with peri-implantitis when compared with

The peri-implant status of 70 implants obtained and the periodontal status of teeth ad- those without peri-implantitis were also observed (2.90 ± 0.61
jacent and contra-lateral to implants (Adj-/CL-teeth). (PD: probing depth, CAL: clinical
vs.4.31 ± 1.34 mm, p < 0.001 and 3.83 ± 1.07 vs.5.15 ± 1.94 mm,
attachment loss, GR: gingival recession, and data shown as mean ± SD).
p = 0.001). For the CL-teeth, a different mean PD was noted in the teeth
Implants Adjacent teeth Contralateral P value contralateral to the implants with the peri-implantitis when compared
teeth with those without (2.93 ± 0.62 vs.3.36 ± 0.74 mm, p = 0.011)
(Table 3).
PD 4.37 ± 1.61a 3.76 ± 1.29b 3.25 ± 0.85c < 0.001*
CAL 4.91 ± 1.81a 4.64 ± 1.78a,b 4.1 ± 1.28b 0.014*
GR 0.54 ± 0.88 0.88 ± 1.08 0.85 ± 0.92 0.069 3.3. Periodontal status at the near and away sites of teeth related to the
Periodontal 0.005# implants
severity
Normal/mild 28 (40.0) 39 (55.7) For the implants without peri-implantitis, similar periodontal mea-
(N [%])
surements (including PD, CAL or GR, with p = 0.702, 0.812, and 0.874,
Moderate (N 21 (30.0) 25 (35.7)
[%])
respectively) were observed at the four sites (i.e. the near and away
Advanced (N 21(30.0) 6 (8.6) sites of the Adj-teeth and CL-teeth) (Table 4). For the implants with
[%]) peri-implantitis, however, a significantly different mean PD or CAL
Furcation 0.177 (p < 0.001) was noted at the near sites of the Adj-teeth when com-
involvement pared with any other sites, including the away site of the Adj-teeth, and
0 (N [%]) 51 (72.9) 57 (81.4) the near and away sites of the CL-teeth (the mean PDs between the near
I (N [%]) 14 (20) 9 (12.9)
and away sites of the Adj-teeth didn’t show statistical significance).
II (N [%]) 2 (2.9) 4 (5.7)
III (N [%]) 3 (4.3) 0 (0)

Tooth mobility 0.166

0 (N [%]) 50 (74.1) 58 (82.9)
I (N [%]) 11 (15.7) 10 (14.3)
II (N [%]) 6 (8.6) 1 (1.4)
III (N [%]) 3 (4.3) 1 (1.4)
* significant difference at p < 0.05 by one-way ANOVA, whereas a, b and c designate
significantly distinct data subsets with the Tukey’s post-hoc analysis.
#
significant difference, at p < 0.05 by Fisher’s test.
implants (4.37 mm) was significantly different from that for Adj-teeth
(3.76 mm), which was also significantly different from that for CL-teeth
(3.25 mm) (p < 0.001). An overall significant difference in CALs
among implants, Adj-teeth, and CL-teeth were also observed
(p = 0.014), although the statistical significance was only observed
between that for implants (4.91 mm) and CL-teeth (4.10 mm). GRs
around the implants, Adj-teeth and CL-teeth did not show any statistical
difference (0.54, 0.88, and 0.85 mm for implants, Adj-teeth and CL-
teeth, respectively, p = 0.069). Moreover, the distribution of furcation
involvement and tooth mobility shows no significant difference be-
tween Adj-teeth and CL-teeth (p = 0.117) (Table 2).
3.2. Periodontal status of the Adj-teeth and CL-teeth according to the
presence of peri-implantitis
The implants with peri-implantitis presented a significantly dif-
ferent mean PD (5.17 ± 1.57 vs. 3.15 ± 0.63 mm, p < 0.001) and
mean CALs (5.72 ± 1.70 vs. 3.62 ± 1.12 mm, p < 0.001), but not a
mean GR, when compared with those without peri-implantitis
(Table 3). Similar findings of a different mean PD and CAL in the teeth
3.4. Factors influencing the periodontal status of teeth
With generalized estimating equation (GEE), PD was significantly
influenced by the presence of peri-implantitis, the tooth location (CL-
tooth vs. Adj-tooth), and the gender of the patients (Table 5). PD was
increased in the condition of peri-implantitis (β = 1.041 mm, con-
fidence interval = 0.646–1.435, and p < 0.001) when compared with
that without peri-implantitis, and was greater in the tooth adjacent to
the implant than that of the contralateral tooth (β = 0.65 mm, con-
fidence interval = 0.4–0.9, and p < 0.001). For CAL, similar con-
founding factors were observed as with PD, except a significantly in-
creased CAL was noted at the near sites when compared with that of the
away sites (β = 0.304 mm, confidence interval = 0.019–0.588, and
p = 0.036). For GR, the probing site (near vs. away site) was the only
confounding factor that showed statistical influence on the measure-
ment of GR (β = 0.136 mm, confidence interval = 0.02–0.252, and
p = 0.022).
4. Discussion
This cross-sectional study aimed to evaluate whether the period-
ontal status of the tooth/site adjacent or contralateral to an implant
could be correlated to the peri-implant condition of the implant in the
patient with chronic periodontitis, and further elucidate the con-
founding factors. In the present study, all the subjects had chronic
periodontitis with 66% (35/53) of the subjects and 60% (42/70) of the
implants exhibiting peri-implantitis (Table 1). The prevalence of peri-
implantitis varied depending on the studies. For instance, Zitzmann and
Berglundh reviewed 6 experiments and found that the prevalence of

106
C.-E. Sung et al. Journal of Dentistry 70 (2018) 104–109

Table 3
Periodontal statuses of the adjacent and the contralateral teeth according to the implants with and without peri-implantitis. (PD: probing depth, CAL: clinical attachment loss, GR:
gingival recession, PI: peri-implantitis, and data shown as mean ± SD).

The implants The teeth related to implants

(n = 70) Adjacent teeth (n = 70) Contralateral teeth (n = 70)

Without PI (n = 28) With PI (n = 42) P value Without PI (n = 28) With PI (n = 42) P value Without PI (n = 28) With PI (n = 42) P value

* *
PD 3.15 ± 0.63 5.17 ± 1.57 < 0.001 2.90 ± 0.61 4.31 ± 1.34 < 0.001 2.93 ± 0.62 3.36 ± 0.74 0.011*
CAL 3.62 ± 1.12 5.72 ± 1.70 < 0.001* 3.83 ± 1.07 5.15 ± 1.94 0.001* 3.77 ± 1.09 4.20 ± 1.16 0.126
GR 0.47 ± 0.68 0.55 ± 0.96 0.684 0.92 ± 1.07 0.84 ± 1.07 0.754 0.85 ± 0.97 0.84 ± 0.89 0.97

* significant difference at p < 0.05 by the independentt test.

Table 4 portion of 1.7 mm, is somehow different from what was measured for
Periodontal statuses at the near and away sites of the adjacent and the contralateral teeth the periodontal gingiva [19]. Parpaiola and colleages also reported that
according to the implants with and without peri-implantitis. (PD: probing depth, CAL: the height of the peri-implant mucosa was about 1 to 1.5 mm greater
clinical attachment loss, GR: gingival recession, and data shown as mean ± SD).
than that of the healthy gingiva [20].
Adjacent teeth Contralateral teeth P Uncontrolled periodontitis has been known to be a risk factor for the
development of peri-implant diseases [14]. Some other risk factors such
Near site Away site Near site Away site as poor oral hygiene, lack of supportive periodontal therapy, history of
Implants without peri-implantitis (n = 28)
periodontitis, diabetes, and smoking have also been identified as risk
PD 2.96 (0.80) 3.18 (0.78) 3.02 (0.73) 3.13 (0.69) 0.702 factors [12]. It has been shown that the subjects who presented ≧5%
CAL 3.95 (1.27) 3.95 (1.20) 3.70 (1.31) 4.00 (1.31) 0.812 sites with PD ≧ 4 mm and ≧30% sites with BoP in the dentition showed
GR 0.98 (1.07) 0.77 (1.15) 0.79 (0.92) 0.88 (1.17) 0.874 significantly poorer peri-implant conditions [8]. Another study has as-
Implants with peri-implantitis (n = 42) sessed peri-implant conditions of 74 partially dentate patients who had
PD 5.01 (1.69)a 4.42(1.80)a,b 3.71(1.07)b,c 3.55(0.88)c < 0.001* been diagnosed with and treated for periodontitis. Implants in those
CAL 6.02 (2.36)a 4.89(2.04)b 4.35(1.5)b 4.35(1.11)b < 0.001* patients with residual PD of ≧6 mm had increased PD and bone loss
GR 1.01 (1.36) 0.48(0.80) 0.63(0.9) 0.80(0.94) 0.1
compared with the implants placed in patients without residual pock-
* significant difference at p < 0.05 by one-way ANOVA, whereas a, b and c designate eting [21]. However, the information of the relationship between peri-
significantly distinct data subsets with the Tukey’s post-hoc analysis. implant disease and periodontal status around the nearby and con-
tralateral natural teeth is still limited. The present study was the first
peri-implantitis ranged from 28% to 56% for subjects and 12% and 43% study examining the relationship between peri-implantits and the per-
for implant sites [3], whereas Atieh and co-workers summarized 9 iodontal status of the teeth adjacent to the implants in patients with
studies with a meta-analysis and described that the prevalence of peri- periodontitis. In Table 3, a relatively poorer periodontal status of the
implantitis was 18.8% for subjects and 9.6% for implants [12]. The Adj-teeth and CL-teeth were found when the corresponding implants
differences in the prevalence might be due to the ethnic population, the had peri-implantitis. A significantly increased mean PD was observed
study design (retrospective or prospective), as well as the definition of for the teeth when the implants had peri-implantitis compared with the
peri-implantitis [12,13]. The exact reasons for having a high incidence condition without peri-implantitis. The mean PD was increased in both
of peri-implantitis in this study are still uncertain; however, our in- Adj-teeh and CL-teeth when the corresponding implant had peri-im-
clusion criteria of the periodontitis patients might be attributed to the plantitis, compared with the teeth without peri-implantitis implants;
high prevalence and might introduce potential bias [12,14]. however, the increase of CAL only occurred in Adj-teeth, not in CL-
In this study, the set of teeth/implant was selected as the analysis teeth, when the corresponding implants had peri-implantits (Table 3).
unit to examine the periodontal status of the teeth adjacent to the im- These increased PDs around Adj-teeth and CL-teeth might be as a result
plants with and without peri-implantitis at first (Tables 2–4). The lim- of tissue inflammation [22]. The finding of increased susceptibility for
itation is that this methodology assumed and treated the 70 sets, ob- peri-implantitis in periodontitis patients is also consistent with previous
tained from 53 patients, as independent to one another. However, literatures and suggests that increased susceptibility might be at the
observations from the same subject may not behave independently but patient level.
are clustered within that subject. This assumption would increase the Increasing evidence suggests that some specific bacterial pathogens
Type I error if significance was found. In order to adjust the within may play an important role in the initiation of cases of peri-implantitis
factors for the repeated measurements, including subjects and teeth, the with similar microbiota rich in gram-negative bacteria having been
generalized estimating equations model was selected for the final ex- identified in peri-implant diseases and chronic periodontitis. The
amination, using the sites as the analysis unit (Table 5). This method composition of the microflora associated with peri-implantitis closely
could produce reasonably accurate standard errors and confidence in- resembles what is found in chronic periodontitis [23]. However, the
tervals in the presence of correlation between the observations [15]. host response to the bacterial challenge in peri-implantitis elicits a more
In Table 2, our results showed that the mean of the implant PD was pronounced inflammatory response in peri-implant mucosal tissues
4.37 mm, which was significantly different than the PD of the Adj-teetj than that in the dentogingival tissue as a result of structural differences
and CL-teeth (3.76 mm and 3.25 mm, respectively). The increased (such as vascularity and fibroblast-to-collagen ratios) [24]. Some of the
probing depths around implants might be due to the structural differ- peri-implantitis lesions appeared to have periods of rapid progression,
ence between tooth gingival and peri-implant soft tissues. In animal that were more pronounced than that observed in cases of chronic
studies, the cuff of soft tissue that surrounds implants is found to be periodontitis.
more voluminous than the gingiva of contralateral teeth [16]. This In this study, there was no obvious difference in the periodontal
finding was also supported by the results from probing depth mea- condition between Adj-teeth and CL-teeth when there was no peri-im-
surements of teeth and implants in human studies [17,18]. In the study plantitis observed (Table 4). In contrast, there was a significant differ-
of Tomasi and co-workers, the morphogenesis of peri-implant mucosa, ence in the near and away sites of the Adj-teeth and CL-teeth. The near
including a barrier epithelium of 1.9 mm and a connective tissue sites in the Adj-teeth had the greatest mean CAL (6.02 mm) where the

107
C.-E. Sung et al. Journal of Dentistry 70 (2018) 104–109

Table 5 Table 5 (continued)

Effect of the factors of peri-implantitis, tooth location (adjcent/contralateral tooth), ex-
aming site (near/away site), buccal/lingual surface, gender, age, and implant set (implant β Standard Confidence p value
(s)/person) on the periodontal measurements, including PD, CAL and GR, of 140 teeth error interval
(560 sites) associated with implants, analyzed by multiple regression model with GEE
method adjusting the within variables, including the implant set(s) per person, the tooth Male 0.310 0.2336 −0.148 to 0.184
location (Adj or CL tooth), the examing site (near/away site), and the buccal/lingual 0.768
surface. Age −0.001 0.0155 −0.031 to 0.968
0.030
β Standard Confidence p value Implant set:
error interval 1 (reference)
2 −0.002 0.2023 −0.399 to 0.991
Probing Depth 0.394
Peri-implantitis:
Without (reference) * significant difference at p < 0.05 by GEE.
With 1.041 0.2013 0.646–1.435 < 0.001*
Tooth location:
implant had peri-implantitis. This result implied the tooth location and
Contralateral tooth
(reference) examining site relative to the implant might play a role in the period-
Adjacent tooth 0.65 0.1276 0.400–0.900 < 0.001* ontal condition. After adjusting other confounding factors, our results
Examining site: showed that the existence of peri-implantitis deteriorated the period-
Away site (reference)
ontal condition of natural teeth, not only in PD (β= 1.041, p < 0.001)
Near site 0.164 0.119 −0.069 to 0.167
0.398
but also in CAL (β = 0.857, p < 0.004) (Table 5). Moreover, the PD of
Observing surface: the Adj-tooth increased 0.65 mm compared to the CL-teeth and the CAL
Buccal (reference) of the near site increased 0.304 mm compared to the away site. It is
Lingual 0.064 0.0833 −0.099 to 0.440 possible that the dental implants would create a site-specific local en-
0.228
vironment aggravating the periodontal destruction with involvement of
Gender:
Female (reference) the Adj-teeth and near sites. Therefore, the periapical radiographs of
Male 0.441 0.2151 0.019 to 0.863 0.040* the Adj-teeth revealed greater CALs at the near sites (Fig. 2) and can be
Age −0.007 0.0113 −0.029 to 0.533 explained by the common presentation of peri-implantitis associated
0.015
bone loss as a circumferential ‘trough’. In the study by Mombelli and
Implant(s)/person:
1 (reference)
et al., failed osseointegrated titanium implants had also showed site-
2 0.177 0.2097 −0.235 to 0.400 specific infection with remarkably similar ecosystems to those en-
−0.588 countered in periodontal disease [10]. However, results in the present
Clinical Attachment loss cross-sectional study should be interpreted carefully due to lack of
Peri-implantitis: longitudinal investigation to explain the relationship of “cause and ef-
Without (reference) fect” between natural teeth and dental implants. Most studies empha-
With 0.857 0.2946 0.279–1.434 0.004*
sized that subjects with periodontitis may have increased susceptibility
Tooth location:
CL-tooth (reference) to peri-implant disease, but it may also be a mutual effect between the
Adj-tooth 0.682 0.1744 0.340–1.024 < 0.001* natural teeth and dental implants and needs more studies to analyze
Examining site: their relationship.
Away site (reference)
Near site 0.304 0.1451 0.019–0.588 0.036*
Observing surface:
Buccal (reference) 5. Conclusion
Lingual 0.011 0.1101 −0.205 to 0.922
0.226 This is the first study to compare the periodontal status of the teeth
Gender:
adjacent to and contralateral to the implants with and without peri-
Female (reference)
Male 0.748 0.3065 0.147 to 1.348 0.015* implantitis in order to explore the relationship of the peri-implantitis
Age −0.007 0.0188 −0.044 to 0.694 with the periodontal health of the tooth adjacent to that implant. It is
0.029 necessary to mention that all our patients had chronic periodontitis.
Implant set: Our results showed: 1) significant different PD and CAL of the Adj-teeth
1 (reference)
were observed in peri-implantitis group from that without peri-im-
2 0.170 0.3214 −0.460 to 0.596
0.800 plantitis group, but the CALs of CL-teeth between the two groups were
not different (at the tooth/implant level); 2) PD and CAL increased at
Gingival Recession
Peri-implantitis: the near sites (compared to away sites) of teeth adjacent to the implants
Without (reference) with peri-implantitis, but these findings were not observed for those
With −0.180 0.2015 −0.575 to 0.373 teeth without peri-implantitis (at the site level); and 3) the presence of
0.215 peri-implantitis and the tooth location relative to the implant sig-
Tooth location:
CL-tooth (reference)
nificantly influenced the PD and CAL after the adjustment. With lim-
Adj-tooth 0.036 0.1161 −0.192 to 0.758 itations of the study, we suggest that the peri-implant health is related
0.263 to the health of periodontium of the tooth adjacent to dental implant.
Examining site:
Away site
Near site 0.136 0.0592 0.020–0.252 0.022*
Observing surface: Conflict of interest and source of funding
Lingual (reference)
Buccal 0.057 0.0886 −0.116 to 0.519 This research did not receive any specific grant from funding
0.231 agencies in the public, commercial, or not-for-profit sectors. The au-
Gender:
Female (reference)
thors have stated explicitly that there are no conflicts of interest in
connection with this article.

108
C.-E. Sung et al.
Author contributions
CE Sung, E Fu designed the experiment; CE Sung, CY Chiang, and
HC Chiu collected data and performed experiments; FG Lin and CE Sung
interpreted and analyzed data; E Fu, CY Chiang, and HC Chiu approved
of article; YS Shieh and CE Sung prepared the draft.
Acknowledgements
The study was self-funded by the authors and their institution. The
authors acknowledge Dr. Martin Ming-Jen Fu (Adjunct Assistant
Professor, Department of Periodontology, School of Dentistry, Tri-
Service General Hospital and National Defense Medical Center, Taipei,
Taiwan.), Dr. Wan-Chien Cheng (Assistant Professor, Department of
Periodontology, School of Dentistry, Tri-Service General Hospital and
National Defense Medical Center, Taipei, Taiwan.), and Jonathan Yao
for their help in manuscript editing.
References
[1] D. Cochran, S. Froum, Peri-implant mucositis and peri-implantitis: a current un-
derstanding of their diagnoses and clinical implications, J. Periodontol. 84 (2013)
436–443, http://dx.doi.org/10.1902/jop.2013.134001.
[2] D. Busenlechner, R. Fürhauser, R. Haas, G. Watzek, G. Mailath, B. Pommer, Long-
term implant success at the academy for oral implantology: 8-year follow-up and
risk factor analysis, J. Periodontal Implant Sci. 44 (2014) 102–108, http://dx.doi.
org/10.5051/jpis.2014.44.3.102.
[3] N.U. Zitzmann, T. Berglundh, Definition and prevalence of peri-implant diseases, J.
Clin. Periodontol. (2008) 286–291, http://dx.doi.org/10.1111/j.1600-051X.2008.
01274.x.
[4] S. Renvert, A. Aghazadeh, H. Hallström, G.R. Persson, Factors related to peri-im-
plantitis – a retrospective study, Clin. Oral Implants Res. 25 (2014) 522–529,
http://dx.doi.org/10.1111/clr.12208.
[5] S. Renvert, I. Polyzois, Risk indicators for peri-implant mucositis: a systematic lit-
erature review, J. Clin. Periodontol. 42 (2015) S172–S186, http://dx.doi.org/10.
1111/jcpe.12346.
[6] L.J.A. Heitz-Mayfield, Peri-implant diseases: diagnosis and risk indicators, J. Clin.
Periodontol. (2008) 292–304, http://dx.doi.org/10.1111/j.1600-051X.2008.
01275.x.
[7] P.R. Klokkevold, T.J. Han, How do smoking, diabetes, and periodontitis affect
outcomes of implant treatment? Int. J. Oral Maxillofac. Implants 22 (Suppl) (2007)
173–202, http://dx.doi.org/10.1038/bdj.2007.841.
[8] R. Wang, W. Zhao, Z.H. Tang, L.J. Jin, C.F. Cao, Peri-implant conditions and their
relationship with periodontal conditions in Chinese patients: a cross-sectional
study, Clin. Oral Implants Res. 25 (2014) 372–377, http://dx.doi.org/10.1111/clr.
12114.
109
Journal of Dentistry 70 (2018) 104–109
Fig. 2. Alveolar bone levels of the Adj-tooth. (a-c)
Representative periapical radiographs showing the alveolar
bone levels of the Adj-teeth from 3 different subjects. White
(↑) and black arrows (↑) indicate the near and away site of the
Adj-tooth, respectively.
[9] L. Canullo, D. Peñarrocha-Oltra, U. Covani, D. Botticelli, G. Serino, M. Penarrocha,
Clinical and microbiological findings in patients with peri-implantitis: a cross-sec-
tional study, Clin. Oral Implants Res. 27 (2016) 376–382, http://dx.doi.org/10.
1111/clr.12557.
[10] A. Mombelli, M.A. van Oosten, E. Schurch, N.P. Land, The microbiota associated
with successful or failing osseointegrated titanium implants, Oral Microbiol.
Immunol. 2 (1987) 145–151, http://dx.doi.org/10.1111/j.1399-302X.1987.
tb00298.x.
[11] W.B. Kaldahl, K.L. Kalkwarf, K.D. Patil, J.K. Dyer, R.E.J. Bates, Evaluation of four
modalities of periodontal therapy, J. Periodontol. (1988) 783–793, http://dx.doi.
org/10.1902/jop.1988.59.12.783.
[12] M.A. Atieh, N.H.M. Alsabeeha, C.M. Faggion, W.J. Duncan, The frequency of peri-
implant diseases: a systematic review and meta-analysis, J. Periodontol. 84 (2012)
1–15, http://dx.doi.org/10.1902/jop.2012.120592.
[13] D.P. Tarnow, Increasing prevalence of peri-implantitis, J. Dent. Res. 95 (2016) 7–8,
http://dx.doi.org/10.1177/0022034515616557.
[14] B.R. Chrcanovic, T. Albrektsson, A. Wennerberg, Periodontally compromised vs.
periodontally healthy patients and dental implants: a systematic review and meta-
analysis, J. Dent. 42 (2014) 1509–1527, http://dx.doi.org/10.1016/j.jdent.2014.
09.013.
[15] A. Hannigan, C.D. Lynch, Statistical methodology in oral and dental research: pit-
falls and recommendations, J. Dent. 41 (2013) 385–392, http://dx.doi.org/10.
1016/j.jdent.2013.02.013.
[16] T. Berglundh, J. Lindhe, Dimension of the periimplant mucosa, J. Clin. Periodontol.
23 (1996) 971–973, http://dx.doi.org/10.1111/j.1600-051X.1996.tb00520.x.
[17] M. Chang, J.L. Wennström, P. Odman, B. Andersson, Implant supported single-tooth
replacements compared to contralateral natural teeth. Crown and soft tissue di-
mensions, Clin. Oral Implants Res. 10 (1999) 185–194, http://dx.doi.org/10.1034/
j.1600-0501.1999.100301.x.
[18] M. Chang, J.L. Wennström, Soft tissue topography and dimensions lateral to single
implant-supported restorations. A cross-sectional study, Clin. Oral Implants Res. 24
(2013) 556–562, http://dx.doi.org/10.1111/j.1600-0501.2012.02422.x.
[19] C. Tomasi, F. Tessarolo, I. Caola, J. Wennström, G. Nollo, T. Berglundh,
Morphogenesis of peri-implant mucosa revisited: an experimental study in humans,
Clin. Oral Implants Res. 25 (2014) 997–1003, http://dx.doi.org/10.1111/clr.
12223.
[20] A. Parpaiola, D. Cecchinato, M. Toia, E. Bressan, S. Speroni, J. Lindhe, Dimensions
of the healthy gingiva and peri-implant mucosa, Clin. Oral Implants Res. 26 (2015)
657–662, http://dx.doi.org/10.1111/clr.12359.
[21] J. Cho-Yan Lee, N. Mattheos, K.C. Nixon, S. Ivanovski, Residual periodontal pockets
are a risk indicator for peri-implantitis in patients treated for periodontitis, Clin.
Oral Implants Res. 23 (2012) 325–333, http://dx.doi.org/10.1111/j.1600-0501.
2011.02264.x.
[22] S. Renvert, I.N. Polyzois, Clinical approaches to treat peri-implant mucositis and
peri-implantitis, Periodontology 2000 (68) (2015) 369–404, http://dx.doi.org/10.
1111/prd.12069.
[23] A. Mombelli, F. Décaillet, The characteristics of biofilms in peri-implant disease, J.
Clin. Periodontol. (2011) 203–213, http://dx.doi.org/10.1111/j.1600-051X.2010.
01666.x.
[24] L.J.A. Heitz-Mayfield, N.P. Lang, Comparative biology of chronic and aggressive
periodontitis vs. peri-implantitis, Periodontology 53 (2010) (2000) 167–181,
http://dx.doi.org/10.1111/j.1600-0757.2010.00348.x.