Success of dental implants in smokers and non-smokers: a systematic review and meta-analysis


The purpose of this review was to test the null hypothesis of no difference in marginal bone loss and implant failure rates between smokers and non-smokers with respect to the follow-up period. An extensive electronic search was performed in PubMed, Web of Science, and the Cochrane Central Register of Controlled Trials to identify relevant articles published up to February 2015. The eligibility criteria included randomized and non-randomized clinical studies. After an exhaustive selection process, 15 articles were included. The meta-analysis was expressed in terms of the odds ratio (OR) or standardized mean difference (SMD) with a confidence interval (CI) of 95%. There was a statistically significant difference in marginal bone loss favouring the non-smoking group (SMD 0.49, 95% CI 0.07–0.90; P = 0.02). An independent analysis revealed an increase in marginal bone loss in the maxilla of smokers, compared to the mandible (SMD 0.40, 95% CI 0.24–0.55; P < 0.00001). A statistically significant difference in implant failure in favour of the non-smoking group was also observed (OR 1.96, 95% CI 1.68–2.30; P < 0.00001). However, the subgroup analysis for follow-up time revealed no significant increase in implant failure proportional to the increase in follow-up time ( P = 0.26).

Dental implants have a high rate of survival and success ; however, certain local and systemic conditions are known to cause dental implant failures. Factors such as low insertion torque (e.g. poor quality bone, poor surgical skill, inaccurate drilling), peri-implant disease, smoking, diabetes, and bisphosphonate use have been reported previously as some of the possible causes of dental implant failure.

Some studies have shown that the nicotine absorbed by the oral mucosa in smokers can negatively affect periodontal tissue healing and peri-implant health. The risk mechanisms involving tobacco intake are yet to be fully elucidated; however, it is believed that the negative influence of tobacco is linked to the adverse effects on fibroblast function, reduced collagen production, and an increase in vascular problems. Moreover, tobacco can exert a negative effect on immune function, interfering with the chemotaxis and phagocytosis mechanisms of polymorphonuclear neutrophils, and decreasing immunoglobulin production and functioning of lymphocytes. There is still no consensus on the number of cigarettes smoked and the relationship with implant failure, however heavy smokers may exhibit a higher incidence.

Some longitudinal studies have reported a higher rate of implant failure in smoking patients. However, these failures are subject to many factors, such as a history of periodontitis, hormonal or metabolic diseases that affect bone turnover, diseases that favour infection, characteristics of the implant used (shape and surface treatment), and prosthetic factors, such as the loading protocol and type of occlusion. It is rather difficult to interpret the data and reduce the bias, as all of these factors interact with each other.

The goal of this meta-analysis was to compare marginal bone loss and the failure of implants during different follow-up periods between smokers and non-smokers.

Materials and methods

The methodology of this study was adapted from the PRISMA statement (Preferred Reporting Items for Systematic Reviews and Meta-analyses). The clinical questions were divided and categorized according to the PICO strategy (Patient, Intervention, Comparison, and Outcome).


The purpose of this review was to test the null hypothesis of no difference in marginal bone loss or implant failure rates between smokers and non-smokers, depending on the follow-up period.

Search strategy

A comprehensive electronic search was conducted, with no date or language restrictions, in PubMed, Web of Science, and the Cochrane Central Register of Controlled Trials, up to February 2015. The search strategy and the PICO tool are presented in Table 1 . In addition, references of the studies included (cross-referencing) was also performed to obtain further new studies.

Table 1
Systematic search strategy (PICO strategy).
Search strategy
Population #1 (edentulous partially[MeSH] OR edentulous[MeSH] OR edentulous jaw[MeSH] OR edentulous maxilla*[all fields] OR edentulous mandible*[all fields] OR smoke*[all fields] OR smokers*[all fields] OR nonsmokers*[all fields])
Intervention #2 (implant*[all fields] OR dental implant[MeSH] OR single implant*[all fields] OR multiple implant*[all fields] OR immediate loading*[all fields] OR conventional loading*[all fields])
Comparisons #3 (smokers*[all fields] OR nonsmokers*[all fields])
Outcomes #4 (survival*[all fields] OR implant survival*[all fields] OR dental implant survival*[all fields] OR bone loss[MeSH] marginal bone loss*[all fields] OR implant bone resorption*[all fields] OR dental implant bone loss*[all fields])
Search combination #1 AND #2 AND #3 AND #4
Database search
Language No restriction
Electronic databases MEDLINE/PubMed, Web of Science, and Cochrane Central Register of Controlled Trials

Selection criteria

This review sought prospective and retrospective cohort studies, as well as randomized controlled trials (RCTs) that compared the marginal bone loss and implant failure rates between smokers and non-smokers. For this review, implant failure was regarded as the absolute loss of the implant. The exclusion criteria included animal studies, in vitro studies, and case series, case reports, and reviews. In addition, studies conducted on volunteers with unbalanced metabolic diseases, or with periodontal disease without prior treatment, were excluded.

Screening process

The research and screening processes were conducted by both reviewers (V.M.F. and E.P.B.). The titles and abstracts were first analyzed. The second step involved the selection of full papers for careful reading; these were analyzed according to the eligibility criteria (inclusion/exclusion) for future data extraction. Disagreements between the reviewers were resolved through careful and detailed discussions. Agreement between the two reviewers was evaluated statistically with Cohen’s kappa ( κ ) test. The authors of the studies were contacted via e-mail for clarification of any points, when required.

Quality assessment

The analysis of quality of the non-randomized studies (prospective and retrospective cohort studies) included in this review was performed using the Newcastle–Ottawa scale (NOS). For the categories of ‘selection’ and ‘outcome’, studies may obtain a star/point for each item. For the ‘comparability’ category, two stars/points may be awarded. The highest score that could be assigned to a study according to the NOS was nine stars/points (highest scientific evidence). Studies scoring six stars/points and above were considered to be of high quality.

Data extraction

The following data were extracted from the studies included (when available): authors, publication year, follow-up period, number of subjects, gender and age of the subjects, smoking status, number of implants placed, implant system, implant length and diameter, healing period, days of antibiotic prophylaxis, use of mouth rinse, marginal bone loss, implant survival, failed and placed implants, P -value for implant failure rate, and the number of drop-outs. This review regarded anyone who consumed any quantity of tobacco at any time during the surgery and recovery period to be a smoker.

Statistical analysis

Binary and continuous variables from the studies included were analyzed through meta-analysis when the same type of data was assessed by at least two studies. For binary outcomes (e.g. implant failure), the estimate of the intervention effect was expressed in the form of an odds ratio (OR) with a confidence interval (CI) of 95%. For continuous outcomes (e.g. marginal bone loss), the average and standard deviation (SD) were used to calculate the standardized mean difference (SMD) with a 95% CI. The results were pooled using the fixed-effects model (Mantel–Haenszel–Peto test) or random-effects model (DerSimonian–Laird test). The I 2 statistical test was used to express the percentage of heterogeneity in the studies. Values up to 25% were classified as indicating low heterogeneity, values of 50% as indicating medium heterogeneity, and values of ≥70% as indicating high heterogeneity. The results of the random-effects model were validated when significant heterogeneity was observed ( P < 0.10). The fixed-effects model was considered when low heterogeneity was observed. The level of statistical significance was set at P < 0.05.

Publication bias was explored graphically through a funnel plot. Asymmetry in the funnel plot may indicate possible publication bias.

All data were analysed using the statistical software Review Manager version 5.2.8 (The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark; 2014).


Literature search

The initial search yielded 965 titles from Medline/PubMed, 85 titles from the Cochrane Central Register of Controlled Trials, and 240 titles from the Web of Science. After the initial evaluation, 29 full papers were selected. Fourteen studies were excluded after careful reading, as they did not conform to the eligibility criteria of this review. Therefore, 15 studies published between 1993 and 2013 were included in this meta-analysis. The selection process and the grounds for exclusion of studies are presented in Fig. 1 .

Fig. 1
Flow diagram (PRISMA format) of the screening and selection process.

The κ value for agreement between the reviewers for the inclusion of potential studies (titles and abstracts) was 0.79 and for selected articles was 0.85, demonstrating a substantial agreement between the reviewers for inclusion of potential articles and an ‘almost perfect’ concordance for studies selected, according to the criteria proposed by Landis and Koch.

Study characteristics

The characteristics of the studies included are presented in Table 2 . Five prospective studies and 10 retrospective cohort studies were included. The number of participants in the studies ranged from 60 to 1727, and the average age was 52.5 years. The follow-up period ranged from 8 to 240 months. The number of implants installed in smokers was 5840 and in non-smokers was 14,683. The number of implants installed in the groups was not clearly reported in one article. Implants were installed without surface treatment in four studies. The Brånemark system (Nobel Biocare AB, Göteborg, Sweden) was the most commonly used implant system. In one study, all patients received implants in areas of bone regeneration ; another study included patients susceptible to periodontal disease (which was clinically controlled).

Table 2
Main characteristics of the studies selected.
Authors (year) Study design
Follow-up time in months (mean or range)
No. of subjects
No. per group
Age range
Mean age
Smoking definition No. of implants Implant brand
Implant size (mm) (diameter × length)
Bain and Moy (1993) Retrospective
229 M/311 F
Smoker and non-smoker 2194 Brånemark
NR × 7, 10, 13, 15, 18, 20
Haas et al. (1996) Retrospective
Up to 108
107 (G1)/314 (G2)
171 M/250 F
Smoker and non-smoker 1366 Brånemark, Friatec
Machined, rough
Kan et al. (1999) Retrospective
16 (G1)/44 (G2)
27 M/33 F
Low consumption (<15 cigarettes/day); high consumption (≥15 cigarettes/day) 228 NR
Lambert et al. (2000) Prospective
Smoker and never smoker 2887 NR
Kumar et al. (2002) Prospective
72 (G1)/389 (G2)
Smoker consisted of patients who smoked half a pack or more cigarettes a day 1183 Straumann
Schwartz-Arad et al. (2002) Prospective
89 (G1)/172 (G2)
Non-smokers; mild smokers (up to 10 cigarettes/day); heavy smokers (>10 cigarettes/day) 959 NR
Nitzan et al. (2005) Prospective
9.4–86.6 (mean 45.5)
59 (G1)/102 (G2)
Non-smokers; mild smokers (up to 10 cigarettes/day); heavy smokers (>10 cigarettes/day) 646 NR
DeLuca and Zarb (2006) Prospective
Up to 240
54 (G1)/146 (G2)
Smoker and non-smoker 1539 Brånemark
Sánchez-Pérez et al. (2007) Retrospective
40 (G1)/26 (G2)
Non-smokers; light smokers (<10 cigarettes/day); moderate smokers (10–20 cigarettes/day); heavy smokers (>20 cigarettes/day) 165 Biotech
Balshe et al. (2008) Retrospective
199 (G1)/1299 (G2)
637 M/861 F
Smoker and non-smoker 4607 Brånemark, Nobel Biocare
Machined, rough
3.3, 3.75, 4, 5 × 7, 8.5, 10, 11.5, 13, 15, 18, 20
Sverzut et al. (2008) Retrospective
76 (G1)/574 (G2)
Smoker and non-smoker 1628 NR NR
Cavalcanti et al. (2011) Retrospective
549 (G1)/1178 (G2)
702 M/1025 F
Smoker and non-smoker 5843 Biomet 3i, Astra Tech, Camlog, Friadent-Dentsply, Nobel Biocare, Straumann, Sweden and Martina, Zimmer Dental
Vandeweghe and De Bruyn (2011) Retrospective
41 (G1)/288 (G2)
141 M/188 F
Smoker and non-smoker 712 Southern Implants
3.5, 3.75, 4, 4.3, 5, 6 × 8.5, 10, 10.5, 11.5, 12, 13, 13.5, 15, 16.5, 18
Vervaeke et al. (2012) Retrospective
65 (G1)/235 (G2)
114 M/186 F
Smoker and non-smoker 1093 NR
3.5, 4, 4.5, 5 × 8, 9, 11, 13, 15, 17
Sayardoust et al. (2013) Retrospective
40 (G1)/40 (G2)
38 M/42 F
Smoker and non-smoker 80 Brånemark; Nobel Biocare
Healing period for loading (months) Antibiotics/mouth rinse (days) Marginal bone loss (mm) (mean ± SD) Implant survival rate (%) Failed/placed implants P -value (for implant failure rate) Drop-outs
Bain and Moy (1993) 6 (maxilla)
3 (mandible)
NR NR 88.7 (G1)
95.2 (G2)
44/390 (G1)
86/1804 (G2)
<0.001 NR
Haas et al. (1996) 3 to 7 NR 2.7 ± 1.87 (G1)
1.58 ± 1.42 (G2)
NR NR/366 (G1)
NR/1000 (G2)
Kan et al. (1999) NR NR NR 65.3 (G1)
82.7 (G2)
12/70 (G1)
11/158 (G2)
0.027 0
Lambert et al. (2000) NR NR NR 91.1 (G1)
94 (G2)
85/959 (G1)
115/1928 (G2)
Kumar et al. (2002) 1 to 3 NR NR 97 (G1)
98.3 (G2)
8/269 (G1)
15/914 (G2)
<0.05 NR
Schwartz-Arad et al. (2002) NR 7/NR NR 96 (G1)
98 (G2)
15/380 (G1)
12/579 (G2)
<0.05 NR
Nitzan et al. (2005) NR NR 0.15 ± 0.09 (G1)
0.04 ± 0.04 (G2)
NR/271 (G1)
NR/375 (G2)
DeLuca and Zarb (2006) 6 (maxilla)
3 (mandible)
NR 0.07 ± 0.26 (G1)
0.04 ± 0.12 (G2)
NR/494 (G1)
NR/1045 (G2)
Sánchez-Pérez et al. (2007) NR NR 2.41 ± 1.46 (G1)
3.13 ± 1.59 (G2)
84.2 (G1)
98.6 (G2)
15/95 (G1)
1/70 (G2)
<0.001 NR
Balshe et al. (2008) NR NR NR 91.2 (G1)
95.2 (G2)
NR NR 23
Sverzut et al. (2008) NR NR NR 96.6 (G1)
97.1 (G2)
7/197 (G1)
43/1431 (G2)
0.5994 NR
Cavalcanti et al. (2011) 0 to 9 5/14 NR 94.5 (G1)
97.1 (G2)
107/1961 (G1)
112/3882 (G2)
0.003 250
Vandeweghe and De Bruyn (2011) NR NR 1.56 ± 0.53 (G1)
1.32 ± 0.38 (G2)
95.2 (G1)
98.8 (G2)
5/104 (G1)
7/608 (G2)
0.007 NR
Vervaeke et al. (2012) NR NR 0.53 ± 0.92 (G1)
0.29 ± 0.54 (G2)
96.7 (G1)
98.7 (G2)
8/244 (G1)
11/849 (G2)
0.025 0
Sayardoust et al. (2013) 3 to 4 NR 1.39 ± 1.57 (G1)
1.01 ± 1.09 (G2)
89.6 (G1)
96.9 (G2)
4/40 (G1)
1/40 (2)
<0.05 0
M, male; F, female; NR, not reported; SD, standard deviation; G1, smokers; G2, non-smokers.

Ten studies reported a statistically significant difference in the average number of implant failures between smokers and non-smokers ; the difference was not statistically significant in only one study. With regards to post-surgical care, two studies reported the use of antibiotic prophylaxis, while only one provided information related to the prescription of chlorhexidine mouthwash.

Quality assessment

Only two studies obtained a score of less than six stars. The scores for each study are summarized in Table 3 .

Table 3
Quality assessment of the studies using the Newcastle–Ottawa scale.
Authors (year) Selection Comparability Outcome
Representativeness of the exposed cohort Selection of external control Ascertainment of exposure Outcome of interest not present at start Comparability of cohorts on the basis of the design or analysis a Assessment of outcome Was follow-up long enough for outcomes to occur? b Adequacy of follow-up of cohorts Total 9/9
Bain and Moy (1993) 0 * * * * 0 * * 0 6/9
Haas et al. (1996) 0 * * * * 0 * * 0 6/9
Kan et al. (1999) 0 * * * * 0 * * * 7/9
Lambert et al. (2000) * * * * * 0 * * 0 7/9
Kumar et al. (2002) 0 * * * * 0 * 0 0 5/9
Schwartz-Arad et al. (2002) 0 * * * * 0 * * 0 6/9
Nitzan et al. (2005) 0 * * * * 0 * * 0 6/9
DeLuca and Zarb (2006) 0 * * * * 0 * * 0 6/9
Sánchez-Pérez et al. (2007) * * * * ** * * 0 8/9
Balshe et al. (2008) 0 * * * * 0 * * * 7/9
Sverzut et al. (2008) 0 * * * * 0 * 0 0 5/9
Cavalcanti et al. (2011) * * * * * 0 * * * 8/9
Vandeweghe and De Bruyn (2011) 0 * * * ** * * 0 7/9
Vervaeke et al. (2012) 0 * * * ** * * 0 7/9
Sayardoust et al. (2013) 0 * * * ** * * 0 7/9
Only gold members can continue reading. Log In or Register to continue

Jan 16, 2018 | Posted by in Oral and Maxillofacial Surgery | Comments Off on Success of dental implants in smokers and non-smokers: a systematic review and meta-analysis
Premium Wordpress Themes by UFO Themes