| |

Received: 26 June 2019    Revised: 29 November 2019    Accepted: 13 January 2020

DOI: 10.1111/jcpe.13257

S Y S T E M AT I C R E V I E W

Levels of smoking and dental implants failure: A systematic

review and meta-analysis

Roohollah Naseri1  | Jaber Yaghini2  | Awat Feizi3

1
Dental Research Center, Department of
Periodontics, Dental Research Institute, Abstract
School of Dentistry, Isfahan University of Aim: The present systematic review and meta-analysis was performed to investi-
Medical Science, Isfahan, Iran
2 gate if there was a significantly enhanced risk of dental implant failure due to the
Dental Implants Research Center,
Department of Periodontics, Dental increased number of cigarettes smoked per day.
Research Institute, School of Dentistry,
Materials and Methods: Four databases, including PubMed, Embase, Web of Science
Isfahan University of Medical Sciences,
Isfahan, Iran and Scopus, were searched until January, 2019. The search terms “dental implant,
3
Department of Epidemiology and oral implant, smoking, smoker, tobacco, nicotine and non-smoker” were used in com-
Biostatistics, School of Health, Isfahan
University of Medical Sciences, Isfahan, Iran
bination to identify the publications providing data for dental implant failures related
to the smoking habit. Publications were excluded if the quantity of cigarettes con-
Correspondence
Jaber Yaghini, Department of Periodontics,
sumed per day was not reported. Fixed- or random-effects meta-analyses were used
School of Dentistry, Isfahan University of to pool the estimates of relative risk (RR) with 95% confidence intervals (CI).
Medical Science, Hezar Jarib St., Isfahan,
Iran.
Results: Having additional information supplied by the authors, 23 articles were se-
Email: [email protected] lected for final analysis. The meta-analyses based on implant- and patient-related data
showed a significant increase in the RR of implant failure in patients who smoked >20
cigarettes per day compared with non-smokers (implant based: p = .001; RR: 2.45; CI:
1.42–4.22 and patient based: p < .001; RR: 4; CI: 2.72–5.89).
Conclusion: The risk of implant failure was elevated with an increase in the number
of cigarettes smoked per day.

KEYWORDS

dental implants, implant failure rate, meta-analysis, smoking

1 | I NTRO D U C TI O N (Bornstein, Cionca, & Mombelli, 2009; Palma-Carrió, Maestre-Ferrín,

Peñarrocha-Oltra, Peñarrocha-Diago, & Peñarrocha-Diago, 2011).
Today's dental implants are highly successful (Levin, Laviv, & Previous studies have proved the detrimental effects of smoking
Schwartz-Arad, 2006; Schwartz-Arad, Herzberg, & Levin, 2005). on oral health. Oral precancerous lesions, oral cancers and periodon-
However, some factors might make patients susceptible to a greater tal diseases are some of its effects on the oral cavity (Calsina, Ramón,
risk for implant failure. A better understanding of the factors associ- & Echeverría, 2002; Johnson & Bain, 2000). Smokers also have a
ated with the failure of implants provides an insight into predicting higher rate of tooth loss than non-smokers and the demands for
the dental implant outcomes. Clinical studies have introduced sev- dental implants in smoking patients have gradually been increased
eral risk factors that may affect the short- and long-term implant (Jansson & Lavstedt, 2002). Some recent meta-analyses have re-
success, including quality and volume of bone, jaw location, implant ported a higher risk of dental implant failure in smokers (Chen, Liu,
dimensions and augmentation procedures, as well as systemic and Xu, Qu, & Lu, 2013; Chrcanovic, Albrektsson, & Wennerberg, 2015;
environmental conditions, such as diabetes mellitus and smoking Moraschini, 2016; Strietzel et al., 2007). Smoking has also been

|
518     © 2020 John Wiley & Sons A/S. wileyonlinelibrary.com/journal/jcpe J Clin Periodontol. 2020;47:518–528.
Published by John Wiley & Sons Ltd
NASERI et al. |
      519

shown to be a risk factor for peri-implantitis (Heitz-Mayfield, 2008;

Renvert & Quirynen, 2015). Smokers have shown a significant in-
Clinical Relevance
crease in the marginal bone loss around implants as compared with
non-smokers (Qian, Wennerberg, & Albrektsson, 2012). Therefore, Scientific rationale for the study: Recent meta-analyses have
identification of smoking patients before implant therapy seems to reported a higher risk of dental implant failure in smok-
be necessary. Current patient smoking status is the least informa- ers, but the association between the number of cigarettes
tion that should be recorded (number of cigarettes smoked per day). smoked and implant failure is still unknown.
The association between the number of cigarettes smoked and im- Principal findings: The findings showed the more cigarettes
plant failure is still unknown, but heavy smokers may show a higher smoked daily, the more was the probability of dental im-
incidence (Alsaadi, Quirynen, Komarek, & van Steenberghe, 2007; plant failure.
van Steenberghe, Facobs, Desnyder, Maffei, & Quirynen, 2002; Practical implications: It seems that implant failure due to
Twito & Sade, 2014). A new critical review addressed this issue but cigarette smoking operates along a continuum. Smoking
failed to determine a relationship between the quantity of cigarette more than one pack/day can be considered a risk factor
consumption (in terms of smoking dose) and its effect on dental for implant failure.
implant failure (Barzanji, Chatzopoulou, & Gillam, 2018). However,
this review was only able to include eight studies according to their
inclusion criteria. In addition, there are some discrepancies in de- operators: (“dental implant” OR “oral implant”) AND (“smoking” OR
fining the heavy smokers in various studies. Hence, we performed “smoker” OR “tobacco” OR “nicotine” OR “non-smoker”) (see the
a systematic review and meta-analysis concerning the relationship search strategies based on each database-specific filter in Appendix
between the quantity of smoking and dental implant failure. This S1-Supporting Information). The reference part of the included stud-
research sought to determine if there was a significantly enhanced ies (cross-referencing) and previous systematic reviews was also
risk of implant failures in heavy smokers compared with light smok- searched for further papers. Smoking has been rarely considered
ers. Publications have reported the dental implant failure rates on the main focus in dental implant studies and has been investigated
an implant- and/or patient-related basis. The patient-related data as a complicating factor involved in the implant success. Therefore,
are based on the assumption that patients are independent from a broad-based search strategy was adopted to seek the articles that
each other, but implants within a patient mouth may be correlated studied the effect of smoking on implant failure.
with each other (Herrmann, Lekholm, & Holm, 2003; Herrmann,
Lekholm, Holm, & Karlsson, 1999). This systematic review provided
both implant- and patient-related data for implant failure in smoker 2.2 | Selection criteria
subgroups.
A literature search was performed to seek randomized-controlled
clinical trials, cohort studies or case–control studies, and case series.
2 |  M ATE R I A L S A N D M E TH O DS The exclusion criteria were animal studies, in vitro studies, case re-
ports, finite element analysis studies and reviews. Articles providing
The methodology of this study was prepared in accordance data on dental implant failures related to the smoking habit were
with the PRISMA statement (Moher, Liberati, Tetzlaff, & Altman, considered eligible in the first analysis. Publications were excluded
2009). The study has been registered with PROSPERO, number if the quantity of cigarettes consumed per day was not reported and
CRD42019121556. The purpose of the present systematic review smoking status was not categorized at least in two subgroups (i.e.
and meta-analysis was to assess the null hypothesis that there would light smokers vs. heavy smokers). In this article, implants were con-
be no difference in the implant failure rates by increasing the quan- sidered failures if they fulfilled the failure criteria presented in Table
tity of cigarettes smoked per day. The focused question according to S1. To calculate the patient-level failure, if one or more implants were
the PICO format (Patient, Intervention, Comparison and Outcome) failed in a patient, the treatment outcome was defined as failure.
was in patients undergoing dental implant placement, are heavy
smoker patients versus light smokers at a higher risk for implant fail-
ure, on an implant- and/or patient-related basis? 2.3 | Study selection

The titles and abstracts of the searched results were screened ini-
2.1 | Search strategy tially by two independent authors (R. N. and J. Y.). Publications were
included for full-text evaluation if they met the inclusion criteria in
An electronic search was performed, with no time or language the first analysis, or if insufficient information was provided in the
restrictions, in the following electronic bibliographic databases: title and abstract to enable a decision to be made. Following full as-
PubMed, EMBASE, Web of Science and Scopus, up to January sessment, studies were either selected for inclusion or rejected. Any
2019. The following search model was accomplished using Boolean disagreement between the authors was resolved by discussion with
 |
520       NASERI et al.

a third review author (A. F.). In the papers that included inadequate 2.5 | Data extraction
or limited information about dental implant failure in the smoking
subgroups, the corresponding authors were contacted via email for The following data, if available, were extracted from the studies se-
clarification and requesting the missing data, and a reminder email lected for inclusion by one of the reviewers: year of publication,
was sent twice after. country, study design, follow-up period, age of the subjects, number
of patients, smoking status, implant system, implant surface modifica-
tion, dental implant failure definition and failed and placed implants in
2.4 | Quality assessment the smoking subgroups based on implant- and/or patient-related data.

Quality assessment of the included observational studies was per-

formed using the Newcastle–Ottawa scale (NOS) (Peterson, Welch, 2.6 | Statistical analysis
Losos, & Tugwell, 2011). The NOS includes three categories, and al-
locates a maximum of nine points, as follows: four points for selection, Because cigarette consumption was presented in different categories
two points for comparability and three points for outcome. Articles in the articles and meta-analyses were performed separately in each
scoring six points or more were considered to be of high quality. category, for articles with different categories of smoking, if possible,

F I G U R E 1   Flow diagram of the selection process

NASERI et al. |
      521

the odds ratios (ORs) of implant failure were also synthesized into 2005; Noguerol, Munoz, Mesa, de Dios Luna, & O'Valle, 2006;
other smoking classifications to perform the meta-analysis. In cases Peñarrocha, Guarinos, Sanchis, & Balaguer, 2002; Sanchez-Perez,
where no events were observed in each smoking subgroup, 0.5 was Moya-Villaescusa, & Caffesse, 2007; van Steenberghe et al., 2002;
considered the significance value of these subgroups because calcula- Testori, Weinstein, Taschieri, & Del Fabbro, 2012; Twito & Sade,
tion of an OR was undefined (Fleiss, Levin, & Paik, 1981). Statistical 2014) were included in this systematic review and meta-analysis
heterogeneities were assessed using Cochran's Q-statistics (Lipsey & (31,129 implants).
Wilson, 2001), and the heterogeneity was quantified using I2 statistical Detailed characteristics of the 23 included studies are pre-
test (Higgins & Thompson, 2002). I2 values ranging from 0% to 100%, sented in Table S2. In four papers, the patients received im-
and I2 values of 25%, 50% and 75% were indicated low, moderate plants in areas with sinus lifting procedures (Brizuela et al., 2014;
and high levels of heterogeneity, respectively (Higgins & Thompson, Franceschetti et al., 2014; Kan et al., 1999; Testori et al., 2012);
2002). If a statistically significant (p < .05) heterogeneity was found, three studies reported the results of immediate implant loading
a random-effects model was used. A fixed-effect model was applied (Agliardi et al., 2010; Cercadillo-Ibarguren et al., 2017; Hinze et al.,
if no statistically significant heterogeneity was observed. The funnel 2009), one study recruited patients with history of periodontal dis-
plot and Begg's and Egger's tests were used to assess the publication ease (Correia et al., 2017), and one paper studied implant placement
bias. Sensitivity analysis was conducted to explore the extent to which in the HIV-positive patients (Gherlone et al., 2015). Depending on
inferences might depend on a particular study or number of publica- definition of patient smoking status, the included studies were clas-
tions. Statistical analyses were conducted using Stata version 11.2 sified into four categories according to the number of cigarettes
(Stata Corp). p < .05 was considered statistically significant. smoked per day; studies categorizing patients into three sub-
groups: non-smokers, <10 or >10 cigarettes per day (Agliardi et al.,
2010; Balderas Tamez et al., 2017; Bornstein et al., 2008; Brizuela
3 |  R E S U LT S et al., 2014; Correia et al., 2017; Gherlone et al., 2015; Hinze et
al., 2009; Peñarrocha et al., 2002; Strietzel & Reichart, 2007; van
The flow diagram of the selection process is drawn in Figure 1. The Steenberghe et al., 2002), and none, <15 or >15 cigarettes per day
search strategy yielded 5,035 papers, from which 2,782 articles re- (Franceschetti et al., 2014; Guido Mangano & Ghertasi Oskouei,
mained after elimination of the duplicate records. After assessment 2018; Habsha, 2000; Kan et al., 1999; Testori et al., 2012), and
of titles and abstracts, 2,280 articles were omitted. Full text evalua- none, <20 or >20 cigarettes per day (Agliardi et al., 2010; Arora et
tion was performed on the remaining 502 articles. A total of 479 pa- al., 2017) and studies categorizing patients in four subgroups (none,
pers were excluded as they did not conform to the inclusion criteria. 1–10, 10–20 or >20 cigarettes per day) (Alsaadi et al., 2007, 2008;
In 397 studies, the number and/or quantity of cigarettes was Cercadillo-Ibarguren et al., 2017; Nitzan et al., 2005; Noguerol et
not reported. A total of 109 studies reported the quantity of cig- al., 2006; Sanchez-Perez et al., 2007; Twito & Sade, 2014). The NOS
arette smoking. Of them, 47 studies had excluded some smoking score of the one relevant article was less than six points (Gherlone
subgroups (Light smokers or heavy smokers). Two studies were et al., 2015). The results of the quality assessment of the studies
published with the same patient sample (further results) (E. F. are summarized in Table S3.
Gherlone et al., 2015; Enrico F Gherlone et al., 2016). In 41 papers,
the corresponding authors were contacted via email for clarifica-
tion of the missing data, 11 emails were not delivered to the corre- 3.1 | Meta-analysis based on implant-related data
sponding authors because the email addresses were not valid, and
23 authors did not respond to the additional information or could The results of the meta-analyses of implant failure/success were al-
not retrieve the requested data. In one article, information on im- located into four subgroups according to the definition of patient
plant failure in smoker subgroups was extracted from the chart smoking status in studies:
(van Steenberghe et al., 2002). Finally, 23 publications, including
one clinical trial (Agliardi, Clericò, Ciancio, & Massironi, 2010), 6 I (None, <10 or >10): fourteen papers were in this subgroup. A
prospective studies (Brizuela, Martín, Fernández, Larrazábal, & higher implant failure rate was found in the patients who smoked
Anta, 2014; Cercadillo-Ibarguren, Sánchez-Torres, Figueiredo, & <10 cigarettes/day (cigarettes per day) than in non-smokers;
Valmaseda-Castellón, 2017; Franceschetti et al., 2014; Gherlone (p = .046) (Table 1; Figure 2). There was a statistically significantly
et al., 2015; Hinze, Thalmair, Bolz, & Wachtel, 2009; Strietzel & increased risk of failure rates in patients who smoked >10 cig-
Reichart, 2007) and 16 retrospective studies (Alsaadi et al., 2007; arettes/day than in non-smokers (p < .001) (Table 1; Figure 3).
Alsaadi, Quirynen, Komarek, & van Steenberghe, 2008; Arora Furthermore, patients who smoked >10 cigarettes/day showed
et al., 2017; Balderas Tamez, Neri Zilli, Fandiño, & Guizar, 2017; higher implant failure rates than those with a smoking rate
Bornstein, Halbritter, Harnisch, Weber, & Buser, 2008; Correia, <10 cigarettes/day (p < .001) (Table 1; Fig. S1).
Gouveia, Felino, Costa, & Almeida, 2017; Guido Mangano & Ghertasi II (None, <15 or >15): four papers were categorized in this sub-
Oskouei, 2018; Habsha, 2000; Kan, Rungcharassaeng, Lozada, group. The pooled estimates of studies demonstrated no statis-
& Goodacre, 1999; Nitzan, Mamlider, Levin, & Schwartz-Arad, tically significant difference between the patients who smoked
 |
522       NASERI et al.

TA B L E 1   Meta-analysis of the smoking subgroups based on the implant-related data

Heterogeneity
Fixed- or Random- 95% Confidence p Value for
Subgroups I 2% p Value effects model Relative risks Interval overall effect

I None versus <10 12.4 .317 Fixed 1.28 (1–1.64) .046

None versus >10 53.8 .009 Random 2.4 (1.71–3.36) <.001
<10 versus >10 35.4 .092 Fixed 1.69 (1.31–2.17) <.001
II None versus <15 37.9 .185 Fixed 1.29 (0.77–2.17) .335
None versus >15 82.1 .001 Random 2.82 (1.01–7.9) .048
<15 versus >15 23 .273 Fixed 1.54 (0.89–2.69) .125
III None versus <20 5.7 .386 Fixed 1.46 (1.22–1.76) <.001
None versus >20 71 .001 Random 2.51 (1.47–4.28) .001
<20 versus >20 70.9 .001 Random 1.6 (0.89–2.86) .118
IV None versus 1–10 0 .524 Fixed 1.31 (1.01–1.7) .045
None versus 10–20 34.4 .166 Fixed 1.53 (1.22–1.93) <.001
None versus >20 74.3 .001 Random 2.48 (1.46–4.22) .001
1–10 versus 10–20 0 .644 Fixed 1.18 (0.86–1.62) .313
10–20 versus >20 76.2 <.001 Random 1.74 (1.33–2.26) <.001

F I G U R E 2   Forest plot for the event “implant failure between the patient who smoked <10 cigarettes/day and non-smokers” based on the
implant-related data

<15 cigarettes/day and non-smokers (p = .335) (Table 1; Fig. S2). There was no statistically significant difference between the pa-
Patients who smoked >15 cigarettes/day showed higher implant tients who smoked <15 cigarettes/day and those who smoked
failure rates than non-smokers (p = .048) (Table 1; Figure S3). >15 cigarettes/day (p = .125) (Table 1; Figure S4).
NASERI et al. |
      523

F I G U R E 3   Forest plot for the event “implant failure between the patient who smoked >10 cigarettes/day and non-smokers” based on the
implant-related data

III (None, <20 or >20): eight articles were included in this subgroup. A 3.2 | Meta-analysis based on patient-related data
higher implant failure rate was found in the patients who smoked
<20 cigarettes/day than in non-smokers; (p < .001) (Table 1; Figure The results of meta-analysis based on the patient-related data were
4). Similarly, the patients who smoked >20 cigarettes/day had allocated into three subgroups:
a higher rate of implant failure than the non-smokers (p = .001)
(Table 1; Figure 5). There was no statistically significant difference I (None, <10 or >10): ten papers were in this subgroup. The pooled
between the patients who smoked <20 cigarettes/day and those estimates of studies demonstrated no statistically significant
who smoked >20 cigarettes/day (p = .118) (Table 1; Figure S5). difference between the patients who smoked <10 cigarettes/
IV (None, 1–10, 10–20 or >20): seven studies were categorized in this day and the non-smokers (p = .956) (Table 2; Figure S11). There
subgroup. A higher implant failure rate was found in the patients was a statistically significantly increased risk of failure rate in the
who smoked <10 cigarettes/day than the non-smokers (p = .045) patients who smoked >10 cigarettes/day than the non-smokers
(Table 1; Figure S6). Patients who smoked 10–20 cigarettes/ (p = .002) (Table 2; Figure S12). Moreover, there were not statis-
day had a higher rate of implant failure than the non-smokers tically significant differences in implant failure rates between the
(p < .001) (Table 1; Figure S7). There was also a statistically sig- patients who smoked <10 and those who smoked >10 cigarettes/
nificantly increased risk of failure rate in the patients who smoked day (p = .154) (Table 2; Figure S13).
>20 cigarettes/day than the non-smokers (p = .001) (Table 1; II (None, <15 or >15): three papers were included in this sub-
Figure S8). There was no statistically significant difference be- group. There was no statistically significant difference between
tween the patients who smoked <10 cigarettes/day and those the patients who smoked <15 cigarettes/day and non-smok-
who smoked 10–20 cigarettes/day (p = .313) (Table 1; Figure S9). ers; (p = .192) (Table 2; Figure S14). Patients who smoked
Finally, the patients who smoked >20 cigarettes/day had a higher >15 cigarettes/day showed higher implant failure rates than
rate of implant failure than those who smoked 10–20 cigarettes/ non-smokers (p = .002) (Table 2; Figure S15). There was no statis-
day (p < .001) (Table 1; Figure S10). tically significant difference between the patients who smoked
 |
524       NASERI et al.

F I G U R E 4   Forest plot for the event “implant failure between the patient who smoked <20 cigarettes/day and non-smokers” based on the
implant-related data

<15 cigarettes/day and those who smoked >15 cigarettes/day an increase in the RR of implant failure in the implants placed in
(p = .285) (Table 2, Figure S16). the grafted sinuses in the patients who smoked >15 cigarettes/day
III (None, <20 or >20): There were four studies in this subgroup. A (Table S5).
higher implant failure rate was found in the patients who smoked A meta-regression analysis considering the mean follow-up pe-
<20 or >20 cigarettes/day than non-smokers (p < .001) (Table 2; riod as a covariate was performed for each smoking subgroup. No
Figures S17 and S18). There was no statistically significant dif- significant relationship was found between the follow-up time and
ference between the patients who smoked <20 cigarettes/day implant failure rate in any of the smoking subgroups (Table S6).
and those who smoked >20 cigarettes/day (p = .172) (Table 2;
Figure S19).
3.4 | Publication bias

3.3 | Subgroup meta-analysis In all meta-analyses, neither inspection of the funnel plot nor for-
mal assessment using Egger's test showed any evidence of publica-
Subgroups meta-analysis based on different implant surfaces was tion bias. The result of Egger's test for each subgroup is given in
performed in each smoking subgroup. In implant-related data, sta- Table S7.
tistical significance implant failure relative risks (RRs) were observed
for implants with micro-rough surfaces in comparison with other im-
plant surfaces in patients who smoked <10, >10 and <20 cigarettes/ 4 | D I S CU S S I O N
day than non-smokers. In subgroup who smoked >10 cigarettes/day,
significant differences in implant failure rates were also obtained in The ability to predict the treatment outcomes is a substantial point
both turned and rough surfaces than non-smokers. The RR for each in implant treatment settings. Among the variety of conditions
subgroup examined is given in Table S4. considered to affect the outcome of dental implants, increasing
Subgroups meta-analyses of studies with implants placed in the attention has been concentrated on the patient-related risk fac-
grafted sinuses were also performed in groups who smoked <15 cig- tors (systemic and environmental conditions, such as smoking)
arettes/day and those who smoked >15 cigarettes/day. There was (Moy, Medina, Shetty, & Aghaloo, 2005). Recent meta-analyses
NASERI et al. |
      525

F I G U R E 5   Forest plot for the event “implant failure between the patient who smoked >20 cigarettes/day and non-smokers” based on the
implant-related data

TA B L E 2   Meta-analysis of the smoking subgroups based on the patient-related data

Heterogeneity
Fixed- or random- 95% Confidence p Value for
Subgroups I 2% p Value effects model Relative risks interval overall effect

I None versus <10 0.0 .512 Fixed 0.99 (0.63–1.56) .956

None versus >10 7.4 .373 Fixed 1.56 (1.18–2.06) .002
<10 versus >10 0.0 .489 Fixed 1.44 (0.87–2.36) .154
II None versus <15 0.0 .702 Fixed 1.64 (0.78–3.46) .192
None versus >15 0.0 .941 Fixed 2.73 (1.42–5.24) .002
<15 versus >15 0.0 .660 Fixed 1.65 (0.66–4.12) .285
III None versus <20 0.0 .427 Fixed 2.82 (1.87–4.25) <.001
None versus >20 7.9 .354 Fixed 4 (2.72–5.89) <.001
<20 versus >20 0.0 .845 Fixed 1.35 (0.88–2.09) .172

demonstrated a significant relationship between smoking and the subgroups than in non-smokers. Significantly enhanced failure
risk of implant failure (Chrcanovic et al., 2015; Moraschini, 2016). risks were obtained among all smoker subgroups compared with
To the best of our knowledge, this is the first systematic review and non-smokers except in the subgroups who smoked <15 cigarettes/
meta-analysis that evaluates the association between the heaviness day. The findings showed the more cigarettes smoked daily, the
of smoking and implant failure. A new critical review has addressed more probable was the dental implant failure.
this issue but has failed to determine a cut-off point in terms of the Comparing the implant failure rate between the smoker sub-
quantity of daily cigarette consumption for dental implant failure groups considering different ranges of cigarettes smoked per day,
(Barzanji et al., 2018). there was a statistically significant risk of implant failure between
Meta-analyses based on implant-related data, as expected, the smokers who smoked <10 and those who smoked >10 ciga-
demonstrated an increase in the RRs of implant failure in all smoker rettes/day. There were no significant differences between the other
 |
526       NASERI et al.

subgroups who smoked <15 versus >15 and <20 versus >20 ciga- of the implant-based analyses. In the patient subgroups smoking
rettes/day. Based on these findings, smoking <10 cigarettes/day <15 or >15 cigarettes/day, almost similar results were obtained in
can be recommended with caution as a safe level for dental implant implant- and patient-based meta-analyses. However, it should be
failure. The success rate of dental implants dropped considerably noted that the number of studies included in the patient-based
when the patients smoked >10 cigarettes/day (Kullar & Miller, meta-analyses was less than that of the studies included in the
2019). However, some authors believe no level of cigarette con- implant-based meta-analyses. Many studies only reported the fail-
sumption can be considered “safe” with respect to the risk of the ure outcome based on the effects of smoking on the implant level,
disease (Husten, 2009). which may not have taken into account the patients receiving
Detailed meta-analyses on the publications that categorized the multiple implants. It is recommended that dental implant failure
patients into four subgroups (None, 1–10, 10–20 or >20) revealed a be reported on both implant- and patient-related basis in further
significantly enhanced implant failure risk among the patients who studies.
smoked 10–20 compared with the patients who smoked >20 ciga- In the selected articles, there was no uniform classification
rettes/day. No statistically significant difference was found between for the quantification of smoking considering the number of ciga-
the subgroups who smoked <10 and those who smoked 10–20 cig- rettes smoked per day. We found four different classifications ac-
arettes/day. These findings cautiously confirm that smoking more cording to the patient's cigarette consumption levels. There were
than one pack/day can be considered a risk factor for implant fail- some discrepancies in the definition of heavy smokers in various
ure. The domain of 10–20 cigarettes smoked per day is an uncer- studies. This heterogeneity made it difficult to compare the data
tain range for implant failure. It seems that implant failure operates of the included studies. A patient self-reporting method was used
along a continuum with no apparent threshold of smoking level. It to determine smoking status in all studies. Using other accurate
is recommended that cigarette consumption data be analysed as a methods may be necessary to determine the actual patient smok-
continuous rather than as a categorical variable in the future studies ing status to investigate the true impact of smoking on implant
(Husten, 2009). failure.
To eliminate the cumulative effects of the patient risk factors, Within our limitations, the results of this study should be inter-
the patient-related data were analysed separately. Similar to the me- preted with caution. In most selected studies, smoking was rarely
ta-analyses based on the implant-related data, there was an increase considered the main focus and was investigated as a complicat-
in the implant failure RRs in all smoker subgroups (except those ing factor involved in the implant success. Therefore, most of the
who smoked <10 cigarettes/day) than in the non-smoker group. studies included in the analysis were of the cohort type, which
Significant differences were not found between the patients who could increase the probability of biases (Higgins & Green, 2008).
smoked <10 and <15 cigarettes/day and the non-smokers. A statisti- The present study might have been exposed to all of the inherent
cally significant difference was obtained between the patients who confounding factors in the included articles and we might not have
smoked >10, >15, <20 and >20 cigarettes/day and the non-smok- been able to control all of these restricting factors. The method-
ers, which confirms the results of the implant-based meta-analyses. ological diversity in the selected papers could have affected the
The results highlight the risk of dental implant failure as a result of outcomes of the meta-analysis. For example, sample size and fol-
consuming more than one pack of cigarettes a day. Comparing the low-up periods of studies, presence of medical history, site of im-
implant failure between the smoker subgroups considering different plant placement, type of implants, presence of advanced surgeries
ranges of cigarettes smoked per day—despite the increase in the RRs like GBR, and open sinus lift, might have impacted the results of
of implant failure—there were no significant differences between analysis.
the subgroups who smoked <10 versus >10, <15 versus >15 and <20 Half of the papers included in this meta-analysis had a short-
versus >20 cigarettes/day. This confirms that the probability of im- term follow-up period, lower than 2 years. Most of them assessed
plant failure due to smoking is continuously increased by increasing the early implant failure, up to abutment connection. This may have
the smoking level. distorted the results of our study. However, no significant relation-
It should be noted when implant failure is calculated with the ship was found between the follow-up time and implant failure rate
patient as the statistical unit, the statistical methodology can in any of the smoking subgroups in this study. Previous meta-analy-
overestimate the outcome, resulting in a more negative outcome ses have reported that the implant failure rates do not increase lin-
for patients with multiple implants because when one implant is early with an increase in the follow-up time, indicating a higher risk
failed in a patient with multiple implants, the patient's treatment of early versus delayed implant failure due to smoking (Moraschini,
outcome is considered a failure. In the patient-based meta-analy- 2016; Strietzel et al., 2007).
ses on the patient subgroups smoking <20 or >20 cigarettes/day Concerning the subgroup meta-analyses for the different implant
compared with non-smokers, almost twice RRs of implant failure surfaces, increased RRs of implant failure were more highlighted in
rate were obtained compared to the results of our implant-based the micro-rough implants than in other implant surfaces. In a recent
analyses, though surprisingly the RRs of the patient-based implant meta-analysis, Chrcanovic et al. (2015) reported that implants with
failure in the patient subgroups smoking <10 or >10 cigarettes/ roughened surfaces had a higher implant failure risk ratio in compar-
day compared with non-smokers were nearly half of the values ison with the turned implants in smokers. This is in contrast with the
NASERI et al. |
      527

studies reporting an association between smoking and implant fail- in the specialty prosthodontics and implantology at Universidad of
La Salle Bajio. Revista Espanola de Cirugia Oral y Maxilofacial, 39(2),
ure among the turned surfaces not the new implant surfaces (Balshe,
63–71. https​://doi.org/10.1016/j.maxilo.2016.02.001
Eckert, Koka, Assad, & Weaver, 2008). Balshe, A. A., Eckert, S. E., Koka, S., Assad, D. A., & Weaver, A. L. (2008).
The subgroup meta-analysis also suggested an increase in the RR The effects of smoking on the survival of smooth-and rough-surface
of implant failure for the implants placed in the grafted sinuses in dental implants. International Journal of Oral & Maxillofacial Implants,
patients who smoked >15 cigarettes/day. The association between 23(6), 1117-1122.
Barzanji, A., Chatzopoulou, D., & Gillam, D. (2018). Impact of smoking
smoking and implant failure in sites with grafted sinuses has been
as a risk factor for dental implant failure: A critical review. BAOJ
demonstrated in a previous meta-analysis (Chambrone et al., 2014). Dentistry, 4(3), 4-10.
Our results may suggest an increased risk of implant failure in the Bornstein, M. M., Cionca, N., & Mombelli, A. (2009). Systemic conditions
grafted sinus sites in the heavy smokers. and treatments as risks for implant therapy. International Journal of
Oral and Maxillofacial Implants, 24(Suppl), 12–27.
In conclusion, within the limitation of this study, the results sug-
Bornstein, M. M., Halbritter, S., Harnisch, H., Weber, H.-P., & Buser, D.
gested that the risk of implant failure was elevated with an increase (2008). A retrospective analysis of patients referred for implant
in the number of cigarettes smoked per day. Smoking more than one placement to a specialty clinic: Indications, surgical procedures, and
pack/day can be considered a risk factor for implant failure. Further early failures. The International Journal of Oral & Maxillofacial Implants,
23(6), 1109.
studies are suggested to investigate smoking as a continuous vari-
Brizuela, A., Martín, N., Fernández, F. J., Larrazábal, C., & Anta, A. (2014).
able rather than a categorical one. Osteotome sinus floor elevation without grafting material: Results
of a 2-year prospective study. Journal of Clinical and Experimental
AC K N OW L E D G E M E N T S Dentistry, 6(5), e479–e484. https​://doi.org/10.4317/jced.51576​
Calsina, G., Ramón, J. M., & Echeverría, J. J. (2002). Effects of smoking on
We would like to thank Dr. Aritza Brizuela, Dr. Paolo Cappa,
periodontal tissues. Journal of Clinical Periodontology, 29(8), 771–776.
Dr. Francisco Correia, Dr. Massimo Del Fabbro, Dr. Devorah https​://doi.org/10.1034/j.1600-051X.2002.290815.x
Schwartz-Arad, Dr. Frank Peter Strietzel, Dr. Leonardo Trombelli Cercadillo-Ibarguren, I., Sánchez-Torres, A., Figueiredo, R., & Valmaseda-
and Dr. Agurne Uribarri who provided us with some supplemen- Castellón, E. (2017). Bimaxillary simultaneous immediate load-
ing of full-arch restorations: A case series. Journal of Clinical and
tary data about their studies, and Dr. Marco Esposito, Dr. Sergio
Experimental Dentistry, 9(9), e1147–e1152. https​://doi.org/10.4317/
García-Bellosta, Dr. Effrat Habsha, Dr. Henri Tenenbaum, Dr. jced.54172​
Tobias Thalmair, Dr. Maurizio S. Tonetti who kindly responded to Chambrone, L., Preshaw, P. M., Ferreira, J. D., Rodrigues, J. A., Cassoni,
our email although it was not possible for them to provide the re- A., & Shibli, J. A. (2014). Effects of tobacco smoking on the survival
quested data. rate of dental implants placed in areas of maxillary sinus floor aug-
mentation: A systematic review. Clinical Oral Implants Research, 25(4),
408–416. https​://doi.org/10.1111/clr.12186​
C O N FL I C T O F I N T E R E S T Chen, H., Liu, N., Xu, X., Qu, X., & Lu, E. (2013). Smoking, radiotherapy,
The authors declare that there are no conflicts of interest in this diabetes and osteoporosis as risk factors for dental implant failure:
study. A meta-analysis. PLoS ONE, 8(8), e71955. https​://doi.org/10.1371/
journ​al.pone.0071955
Chrcanovic, B. R., Albrektsson, T., & Wennerberg, A. (2015). Smoking
ORCID and dental implants: A systematic review and meta-analysis.
Roohollah Naseri  https://orcid.org/0000-0002-9252-6565 Journal of Dentistry, 43(5), 487–498. https​ ://doi.org/10.1016/j.
Jaber Yaghini  https://orcid.org/0000-0003-1775-3343 jdent.2015.03.003
Correia, F., Gouveia, S., Felino, A. C., Costa, A. L., & Almeida, R. F. (2017).
Awat Feizi  https://orcid.org/0000-0002-1930-0340
Survival rate of dental implants in patients with history of periodon-
tal disease: A retrospective cohort study. International Journal of Oral
REFERENCES and Maxillofacial Implants, 32(4), 927–934. https​://doi.org/10.11607/​
Agliardi, E., Clericò, M., Ciancio, P., & Massironi, D. (2010). Immediate jomi.3732
loading of full-arch fixed prostheses supported by axial and tilted Fleiss, J., Levin, B., & Paik, M. (1981). Statistical methods for rates and pro-
implants for the treatment of edentulous atrophic mandibles. portions (2nd ed.). New York, NY: John Wiley & Sons.
Quintessence International, 41(4), 285-293. Franceschetti, G., Farina, R., Stacchi, C., Di Lenarda, R., Di Raimondo,
Alsaadi, G., Quirynen, M., Komarek, A., & van Steenberghe, D. (2007). R., & Trombelli, L. (2014). Radiographic outcomes of transcrestal
Impact of local and systemic factors on the incidence of oral implant sinus floor elevation performed with a minimally invasive technique
failures, up to abutment connection. Journal of Clinical Periodontology, in smoker and non-smoker patients. Clinical Oral Implants Research,
34(7), 610–617. https​://doi.org/10.1111/j.1600-051X.2007.01077.x 25(4), 493–499. https​://doi.org/10.1111/clr.12188​
Alsaadi, G., Quirynen, M., Komarek, A., & van Steenberghe, D. (2008). Gherlone, E. F., Cappare, P., Tecco, S., Polizzi, E., Pantaleo, G., Gastaldi, G.,
Impact of local and systemic factors on the incidence of late oral im- & Grusovin, M. G. (2015). A prospective longitudinal study on implant
plant loss. Clinical Oral Implants Research, 19(7), 670–676. https​://doi. prosthetic rehabilitation in controlled HIV-positive patients with
org/10.1111/j.1600-0501.2008.01534.x 1-year follow-up: The role of CD4+ level, smoking habits, and oral
Arora, A., Reddy, M. M., Mhatre, S., Bajaj, A., Gopinath, P. V., & Arvind, hygiene. Clinical Implant Dentistry and Related Research, 4(10), 12370.
P. (2017). Comparative evaluation of effect of smoking on survival Gherlone, E. F., Capparé, P., Tecco, S., Polizzi, E., Pantaleo, G., Gastaldi,
of dental implant. Journal of International Oral Health, 9(1), 24–27. G., & Grusovin, M. G. (2016). Implant prosthetic rehabilitation in
https​://doi.org/10.4103/0976-7428.201094 controlled HIV-positive patients: A prospective longitudinal study
Balderas Tamez, J. E., Neri Zilli, F., Fandiño, L. A., & Guizar, J. M. (2017). with 1-year follow-up. Clinical Implant Dentistry and Related Research,
Factors related to the success or failure of dental implants placed 18(4), 725–734. https​://doi.org/10.1111/cid.12353​
 |
528       NASERI et al.

Guido Mangano, F., & Ghertasi Oskouei, S. (2018). Low serum vitamin D Noguerol, B., Munoz, R., Mesa, F., de Dios Luna, J., & O'Valle, F.
and early dental implant failure: Is there a connection? A retrospec- (2006). Early implant failure. Prognostic capacity of Periotest((R)):
tive clinical study on 1740 implants placed in 885 patients. Journal Retrospective study of a large sample. Clinical Oral Implants Research,
of Dental Research, Dental Clinics, Dental Prospects, 12(3), 174–182. 17(4), 459–464. https​://doi.org/10.1111/j.1600-0501.2006.01250.x
https​://doi.org/10.15171/​joddd.2018.027 Palma-Carrió, C., Maestre-Ferrín, L., Peñarrocha-Oltra, D., Peñarrocha-
Habsha, E. (2000). Survival of osseointegrated dental implants in smok- Diago, M. A., & Peñarrocha-Diago, M. (2011). Risk factors associated
ers and non-smokers. Canada: University of Toronto. Retrieved with early failure of dental implants. A literature review. Medicina
from https​://tspace.libra​r y.utoro​nto.ca/bitst​ream/1807/14487/​1/ Oral, Patología Oral y Cirugía Bucal, 16(4), e514–517.
MQ534​13.pdf Peñarrocha, M., Guarinos, J., Sanchis, J. M., & Balaguer, J. (2002). A ret-
Heitz-Mayfield, L. J. (2008). Peri-implant diseases: Diagnosis and risk in- rospective study (1994–1999) of 441 ITI® implants in 114 patients
dicators. Journal of Clinical Periodontology, 35, 292–304. https​://doi. followed-up during an average of 2.3 years. Medicina Oral, 7(2),
org/10.1111/j.1600-051X.2008.01275.x 144–155.
Herrmann, I., Lekholm, U., & Holm, S. (2003). Statistical outcome of ran- Peterson, J., Welch, V., Losos, M., & Tugwell, P. (2011). The Newcastle-
dom versus selected withdrawal of dental implants. International Ottawa scale (NOS) for assessing the quality of nonrandomised studies
Journal of Prosthodontics, 16(1), 25-30. in meta-analyses. Ottawa, ON: Ottawa Hospital Research Institute.
Herrmann, I., Lekholm, U., Holm, S., & Karlsson, S. (1999). Impact of implant Qian, J., Wennerberg, A., & Albrektsson, T. (2012). Reasons for marginal
interdependency when evaluating success rates: A statistical analysis bone loss around oral implants. Clinical Implant Dentistry and Related
of multicenter results. International Journal of Prosthodontics, 12(2). Research, 14(6), 792–807. https​://doi.org/10.1111/cid.12014​
Higgins, J., & Green, S. (2008). Cochrane Collaboration Cochrane Renvert, S., & Quirynen, M. (2015). Risk indicators for peri-implantitis. A
handbook for systematic reviews of interventions. Chichester, UK: narrative review. Clinical Oral Implants Research, 26, 15–44. https​://
Wiley-Blackwell. doi.org/10.1111/clr.12636​
Higgins, J. P., & Thompson, S. G. (2002). Quantifying heterogeneity in a Sanchez-Perez, A., Moya-Villaescusa, M. J., & Caffesse, R. G. (2007).
meta-analysis. Statistics in Medicine, 21(11), 1539–1558. https​://doi. Tobacco as a risk factor for survival of dental implants. Journal
org/10.1002/sim.1186 of Periodontology, 78(2), 351–359. https​ ://doi.org/10.1902/
Hinze, M., Thalmair, T., Bolz, W., & Wachtel, H. (2009). Immediate load- jop.2007.060299
ing of fixed provisional prostheses using four implants for the reha- Schwartz-Arad, D., Herzberg, R., & Levin, L. (2005). Evaluation of long-
bilitation of the edentulous arch: A prospective clinical study. The term implant success. Journal of Periodontology, 76(10), 1623–1628.
International Journal of Oral & Maxillofacial Implants, 25(5), 1011–1018. https​://doi.org/10.1902/jop.2005.76.10.1623
Husten, C. G. (2009). How should we define light or intermittent smok- Strietzel, F. P., & Reichart, P. A. (2007). Oral rehabilitation using Camlog
ing? Does it matter? Nicotine & Tobacco Research, 11(2), 111. https​:// screw-cylinder implants with a particle-blasted and acid-etched mi-
doi.org/10.1093/ntr/ntp010 crostructured surface. Results from a prospective study with special
Jansson, L., & Lavstedt, S. (2002). Influence of smoking on mar- consideration of short implants. Clinical Oral Implants Research, 18(5),
ginal bone loss and tooth loss–a prospective study over 20 years. 591–600. https​://doi.org/10.1111/j.1600-0501.2007.01375.x
Journal of Clinical Periodontology, 29(8), 750–756. https​ ://doi. Strietzel, F. P., Reichart, P. A., Kale, A., Kulkarni, M., Wegner, B., &
org/10.1034/j.1600-051X.2002.290812.x Kuchler, I. (2007). Smoking interferes with the prognosis of den-
Johnson, N., & Bain, C. (2000). Tobacco and oral disease. British Dental tal implant treatment: A systematic review and meta-analysis.
Journal, 189(4), 200–206. https​://doi.org/10.1038/sj.bdj.4800721 Journal of Clinical Periodontology, 34(6), 523–544. https​ ://doi.
Kan, J. Y. K., Rungcharassaeng, K., Lozada, J. L., & Goodacre, C. J. org/10.1111/j.1600-051X.2007.01083.x
(1999). Effects of smoking on implant success in grafted maxillary Testori, T., Weinstein, R. L., Taschieri, S., & Del Fabbro, M. (2012). Risk
sinuses. Journal of Prosthetic Dentistry, 82(3), 307–311. https​://doi. factor analysis following maxillary sinus augmentation: A retrospec-
org/10.1016/s0022-3913(99)70085-5 tive multicenter study. International Journal of Oral and Maxillofacial
Kullar, A. S., & Miller, C. S. (2019). Are there contraindications for plac- Implants, 27(5), 1170–1176.
ing dental implants? Dental Clinics, 63(3), 345–362. https​ ://doi. Twito, D., & Sade, P. (2014). The effect of cigarette smoking habits on
org/10.1016/j.cden.2019.02.004 the outcome of dental implant treatment. PeerJ, 2, e546. https​://doi.
Levin, L., Laviv, A., & Schwartz-Arad, D. (2006). Long-term success of org/10.7717/peerj.546
implants replacing a single molar. Journal of Periodontology, 77(9), van Steenberghe, D., Facobs, R., Desnyder, M., Maffei, G., & Quirynen,
1528–1532. https​://doi.org/10.1902/jop.2006.060005 M. (2002). The relative impact of local and endogenous pa-
Lipsey, M. W., & Wilson, D. B. (2001). Applied social research methods tient-related factors on implant failure up to the abutment stage.
series. Practical meta-analysis. (vol. 49). California: SAGE publications Clinical Oral Implants Research, 13(6), 617–622. https​ ://doi.
Inc. org/10.1034/j.1600-0501.2002.130607.x
Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred re-
porting items for systematic reviews and meta-analyses: The PRISMA
statement. Annals of Internal Medicine, 151(4), 264–269. https​://doi. S U P P O R T I N G I N FO R M AT I O N
org/10.7326/0003-4819-151-4-20090​8180-00135​
Additional supporting information may be found online in the
Moraschini, V. (2016). Success of dental implants in smokers and
non-smokers: A systematic review and meta-analysis. International
Supporting Information section. 
Journal of Oral and Maxillofacial Surgery, 45(2), 205–215. https​://doi.
org/10.1016/j.ijom.2015.08.996
Moy, P. K., Medina, D., Shetty, V., & Aghaloo, T. L. (2005). Dental implant How to cite this article: Naseri R, Yaghini J, Feizi A. Levels of
failure rates and associated risk factors. International Journal of Oral smoking and dental implants failure: A systematic review and
& Maxillofacial Implants, 20(4), 569-577.
meta-analysis. J Clin Periodontol. 2020;47:518–528.
Nitzan, D., Mamlider, A., Levin, L., & Schwartz-Arad, D. (2005). Impact
of smoking on marginal bone loss. International Journal of Oral and https​://doi.org/10.1111/jcpe.13257​
Maxillofacial Implants, 20(4), 605–609.