Current Evidence on Dental Implants Outcomes in Smokers and Nonsmokers: A Systematic Review and Meta-Analysis

Implant-retained prostheses have been shown to be associated with successful long-term outcomes.^1–6  However, numerous local and systemic factors have been hypothesized to interact to various degrees with implant success.^7–9  Such factors include implant placement in type III and IV bone qualities (Lekholm and Zarb classification); compromised initial implant stability, especially with an immediate loading protocol; implant placement in the maxilla and in the posterior region; postradiation therapy sites; drug and alcohol consumption; and smoking.

The carbon monoxide released during cigarette smoking lowers oxygen tension in tissues by displacing oxygen from hemoglobin.¹⁰  Nicotine, which has been found in ample concentrations in saliva (96 ng mL⁻¹ to 1.6 mg mL⁻¹)^11,12  and crevicular fluid (5961 ng mL⁻¹),¹³  has been reported to have a negative impact on the bone regeneration process.^14,15  Moreover, the viability of polymorphonuclear neutrophils and the phagocytosis of opsonized Candida albicans are reported to be significantly lower in smokers than in nonsmokers.¹⁶ 

Many studies on the correlation between smoking and implant outcomes have been published. Some support the hypothesis that smoking increases implant failure, but others did not reach the same conclusion.^17–20  To date, no consensus has been reached, and no evidence-based guidelines have been developed, to help clinicians make informed clinical decisions regarding the use of dental implant treatment in smokers. These deficiencies can be attributed to several factors, one of which is variability in the design, quality and, therefore, findings of previous studies. The existence of such heterogeneity has precluded the performance of a direct comparison across studies. Moreover, there is a lack of recent reviews assessing only randomized clinical trials (RCTs), controlled clinical trials (CCTs), and prospective clinical trials.

Therefore, this systematic review aimed to identify the impact of smoking on the clinical outcomes of implants using the best available scientific evidence. Accordingly, the objective of this study was to answer the following PICO (Participants, Intervention, Comparison, and Outcome) question: Does smoking increase the rates of implant failure and peri-implant marginal bone loss in patients with dental implants?

The null hypotheses were as follows:

• 1.

  There is no difference between smokers and nonsmokers in terms of the rates of dental implant failure.

• 2.

  There is no difference between smokers and nonsmokers in the extent of peri-implant marginal bone loss.

This study is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guidelines.^21–23 

This review aimed to identify all RCTs, CCTs, and prospective studies that reported outcomes for both smoking and nonsmoking patients who were treated with removable and/or fixed prostheses supported by dental implants and who had a postinsertion follow-up of at least 1 year. The elements of the PICO questions were as follows:

• •

  P (Types of participant): Adult patients (≥ 18 years) who were treated with osseointegrated root-formed dental implants.

• •

  I (Types of intervention): Implant-supported prostheses in smokers.

• •

  C (Control intervention): Implant-supported prostheses in nonsmokers.

• •

  O (Types of outcome measure): Cumulative implant failure rate and the extent of peri-implant marginal bone loss.

Controlled/randomized clinical trials and prospective cohort/case series studies evaluating implant-supported prostheses in adults (>18 years old) over a study duration of at least 1 year were included; only studies published in English were eligible for inclusion.

Studies were not included in this review if they were retrospective; did not separately report outcomes for smokers and nonsmokers; were poorly controlled for confounding variables; included patients with congenital/familial medical conditions or uncontrolled systemic diseases; did not obtain ethical approval or written informed consent; included fewer than 10 patients in each group; and used old implant technology (ie, studies published prior to 1985). Letters to the editor, case reports, and commentaries were also excluded.

To identify studies eligible for inclusion in this review, detailed search strategies were developed for each of the searched databases. These approaches were based on the search strategy developed for Medline (OVID) but revised appropriately for each database. The search strategy incorporated a combination of the following controlled vocabulary and free text terms:

• 1.

  exp Dental Implants/

• 2.

  exp Dental Implantation/ or dental implantation

• 3.

  exp Dental Prosthesis, Implant-Supported/

• 4.

  ((osseointegrated adj implant$) and (dental or oral))

• 5.

  dental implant$

• 6.

  (implant$ adj5 dent$)

• 7.

  (((overdenture$ or crown$ or bridge$ or prosthesis or restoration$) adj5 (Dental or oral)) and implant$)

• 8.

  “implant supported dental prosthesis”

• 9.

  (“blade implant$” and (dental or oral))

• 10.

  ((endosseous adj5 implant$) and (dental or oral))

• 11.

  ((dental or oral) adj5 implant$)

• 12.

  “smoke$”

• 13.

  “Tobacco”

• 14.

  “cigarette$”

• 15.

  OR/1-14

The Cochrane Central Register of Controlled Trials (CENTRAL; the Cochrane Library 2017, Issue 1), PubMed, Medline (OVID), Embase, and Web of Science databases were searched to identify articles published up to January 2017. Peer-reviewed dental journals were hand searched for additional relevant studies. Moreover, the bibliographies of all identified studies and relevant review articles were assessed to identify other studies that were potentially missing from hand-searched journals. Personal references were also searched. When necessary, the authors of eligible studies were contacted via e-mail to inquire about missing data.

The titles and abstracts of all reports identified through the electronic searches were scanned. Full texts were obtained for studies appearing to meet the inclusion criteria or for which there were insufficient data in the title and abstract to make a clear decision regarding inclusion, and these texts were assessed. All included studies underwent quality assessment and data extraction. The studies rejected at this or subsequent stages and the reasons for their exclusion were recorded.

Quality assessment of the included trials was undertaken as part of the data extraction process. The “quality assessment of cohort studies-8 items” form, which combines the MOOSE (Meta-analysis of Observational Studies in Epidemiology), STROBE (Strengthening the Reporting of Observational Studies in Epidemiology), and PRISMA statements, was used to evaluate the prospective studies.^24,25  For RCTs, 10 main methodologic quality criteria were examined. Studies scoring 4 pluses or higher were considered to be of high quality. Further quality assessments were carried out to assess sample-size calculations, definitions of exclusion/inclusion criteria, and comparability of control and test groups at study entry.

Data were extracted using a specially designed electronic extraction form that was piloted and modified as required before use. For each trial, the following data were recorded: year of publication, country of origin, source of study funding, details regarding the participants' demographic characteristics, details regarding the type of intervention, details regarding the outcomes reported (including method of assessment), and whether the groups were comparable in terms of important prognostic factors at study entry.

For dichotomous outcomes, the effect estimates for smoking are expressed as odds ratios (ORs) and 95% CIs. For continuous outcomes, weighted mean differences (WMDs) and 95% CIs are presented.

Odds ratios and mean differences were combined for dichotomous data and continuous data, respectively, using fixed or random-effects models. Odds ratios were selected because the studies were prospective in nature.

The numbers of implants placed in smoking and nonsmoking patients were obtained from each study, as were the number of failed implants in each group. If reported, the mean marginal bone loss and its standard deviation were also obtained. Both fixed-effects and random-effects models were constructed, and the level of heterogeneity across studies was assessed using the Cochrane test for heterogeneity and the I-squared test, which was used to determine the percentage of variation across studies that was due to heterogeneity. When the test results were statistically significant (P < .05), the studies were different in terms of effect size; thus, random-effects models were used. Otherwise, the studies were considered to be similar and could be pooled using fixed-effects models.

Publication bias was examined visually by constructing funnel plots and checking for the symmetry of effect vs sample size (standard error value). Modified Harbord and Egger tests were used to quantitatively evaluate the presence of publication bias in studies reporting implant failure and marginal bone loss, respectively.

The studies exhibited some differences in the proportions of males/females, the average age of the study participants, and the number of years of follow-up. To identify differences between smokers and nonsmokers in terms of demographic characteristics across the included studies and to examine their relationships with the identified effect sizes (ORs for implant failure and WMDs for bone loss), meta-regression analyses were performed. Significant results indicated that the differences observed between smokers and nonsmokers were associated with other study characteristics.

A total of 674 titles were identified during the initial searches. Following the initial screening, 53 full-text articles were evaluated, of which 43 were excluded and 10 were included.^6,17,26–33  The search strategy used for the identification of eligible studies is presented in Figure 1.

Tables 1 and 2 present the characteristics of the 10 included studies. The number of participants in these studies ranged from 50 to more than 800, and the average age of participants was 54 years. The average follow-up ranged from 3 to 8.3 years, with a mean follow-up of 4.6 years. In total, 2296 implants were placed in smokers and 4854 in nonsmokers. Studies usually included more female patients, with the proportion of males ranging from 29.8% to 59.6%.

Seven studies reported the number of failed implants in smokers and nonsmokers (Figure 2).^{17,26–28,30–32}  Of the 2001 implants placed in smokers, 166 failed, representing an 8.3% failure rate. Of the 4298 implants placed in nonsmokers, 183 failed, representing a 4.3% failure rate.

Given the level of heterogeneity observed across studies (I² = 65.8%, X²  (6) = 17.54, P = .007), a random-effects model was used for the meta-analysis. The overall failure OR (for smokers vs nonsmokers) was 2.92 (95% CI, 1.76-4.83), demonstrating a statistically significant difference in favor of the nonsmoking group (P < .001, Figure 2). Meta-regression analyses were also conducted to determine if these results changed after controlling for the demographic characteristics of the patients. The regression models indicated that the percentage of males in the study (P = .38), the average age of the study participants (P = .61), and the number of years of follow-up (P = .17) did not significantly impact the results.

Marginal bone loss was measured from the implant platform to the most crestal contact point between the implant and the bone. Six studies reported the amount of marginal bone loss.^{6,26,27,29,30,33}  Five studies reported overall bone loss, and 1 study reported bone loss during the first year of follow-up in addition to overall loss.²⁹  Annual marginal bone loss data were theoretically estimated using the following equation: Bone Loss (per year) = (Bone Loss(study overall)) / (Years of followup).

The extent of marginal bone loss during the first year in the group of smokers ranged from 0.02 to 0.45 mm, while in the nonsmoking group, this figure ranged from −0.08 to 0.42 mm (Figure 3). A random-effects model was used because of the heterogeneity observed across the studies (I² = 98.3%, X² [5] = 291.21, P < .001). Nonsmokers had significantly less bone loss during the first year, as indicated by the overall WMD of 0.11 mm (95% CI, 0.03–0.19; P = .009).

Meta-regression analyses were also conducted to determine if the difference in the extent of bone loss between subgroups changed after controlling for the demographic characteristics of the patients. The regression models indicated that the percentage of males included in the study (P = .76), the average age of the study participants (P = .06), and the number of years of follow-up (P = .09) did not significantly impact the results observed. The effect of age approached level of statistical significance, potentially indicating that smoking and nonsmoking older patients demonstrated a greater difference in bone loss (B = 0.02 mm per year). With regards to the number of follow-up years, the difference in the extent of bone loss between smokers and nonsmokers also widened (B = 0.04 mm per follow-up year).

Significant heterogeneity was observed across the studies (I² = 98.3%, X² [5] = 291.19, P < .001), and, therefore, the analysis of marginal bone loss between the 3 groups was performed using the random-effects model. An overall WMD) of 0.11 mm (95% CI, 0.03–0.19) was observed, indicating that nonsmokers lost significantly less bone than did smokers (P = .009, Figure 4).

Publication bias was explored by analyzing the funnel plots and examining their symmetry (Figure 5). The Harbord modified test for small study effects was used to test publication bias in studies reporting implant failure. The results of this test were not significant (P = .114), indicating the absence of small study effects (that is, the absence of publication bias). The Egger test for small study effects was used for studies of bone loss to evaluate the presence of publication bias. The P value for this test was 0.175, suggesting the absence of publication bias in the studies of bone loss. The results of the quality assessment of the studies are presented in Table 3.

This meta-analysis of CCTs, RCTs, and prospective clinical studies was conducted to determine the effect of smoking on the rate of implant failure and peri-implant marginal bone loss. Three RCTs and 7 prospective studies were included in the final analysis.

Smokers experienced significantly more implant failure and marginal bone loss relative to nonsmokers (P < .001). These findings are in accordance with those of another review conducted by Moraschini et al.³⁴  The mechanism by which tobacco affects the osseointegration process remains unknown. However, some of the chemicals found in tobacco have been shown to reduce the vascularity of the peri-implant tissues, which, in turn, may compromise the bone healing process.³⁵  Approximately 3 mg of nicotine and 20–30 mL of carbon monoxide are inhaled for each cigarette smoked.³⁶  Nicotine has been shown to increase platelet aggregation and impede the function of fibroblasts, red blood cells, osteoblasts, and macrophages.^37–41  Additionally, because its affinity for hemoglobin is greater than that of oxygen (200-fold greater), carbon monoxide will convert hemoglobin into carboxyhemoglobin rather than oxyhemoglobin, which is formed when oxygen binds to hemoglobin. The formation of carboxyhemoglobin decreases oxygen transportation, resulting in reduced oxygen tension in the tissues, which is known as hypoxia.^42–46 

While the achievement of smoking cessation and sustained abstinence well before dental intervention should be the ultimate goal, nicotine dependence has been scientifically acknowledged as being a chronic relapsing disorder that is usually characterized by multiple failed quitting attempts.⁴⁷  Nonetheless, several studies have suggested that adjunctive measures may minimize the negative effects of smoking on the survival of dental implants.^48,49  Abstaining from smoking for 1 week before and 8 weeks after implant placement was reported to improve the success rate associated with the Brånemark implant.⁴⁸  Opting for a delayed loading protocol with implants submersion during the healing process may minimize the accumulation of pathogenic biofilms and the diffusion of some of the nearly 4000 chemicals present in cigarette smoke.^34,50  Recombinant human parathyroid hormone (PTH 1-34), an anabolic agent approved for the treatment of patients with osteoporosis that preferentially stimulates osteoblasts function over osteoclast function,⁵¹  has been reported to increase bone volume around implants in the presence of cigarette smoke in animal models.⁵² 

The greater difference in marginal bone loss observed between smokers and nonsmokers in association with aging (B = 0.02 mm per year) may be explained by an amalgamation of the depleting effect of the tobacco chemicals on bone vascularity and the slow, progressive, age-related phase of bone loss in trabecular and cortical bone.^53–55 

When a cause-and-effect relationship is being investigated, a precise definition of the potential cause is imperative. This is particularly crucial when the effect is anticipated to be frequency- and dose-dependent. In medicine, smoking has been established to exert a dose-dependent effect on the extent of bone loss and, consequently, the risk of fracture.⁵⁶  Interestingly, however, smaller doses of nicotine have been found to stimulate the growth of osteoblasts.⁵⁷  Wide variations may be observed in the definition of smoking in the dental literature in terms of smoking duration, amount of cigarettes consumed per day, and categorization of previous smokers, and these variations preclude a detailed analysis of the predictability of implant outcomes.

Diverse patient- and implant-mediated confounding factors have been shown to impact the clinical outcomes associated with implant treatment.^7–9  However, these factors were not always taken into consideration in the included clinical trials. Only 1 study reported implant diameters in smokers and nonsmokers.²⁹  None of the included studies reported data for smokers and nonsmokers separately in terms of the average age, implant length, implant type and surface characteristics, and number of implants placed in the maxilla and mandible. Consequently, well-designed, long-term RCTs are needed. These studies should include a sufficient number of patients (with sufficient power) to identify a true difference, proper allocation concealment, independent outcome assessors, and a sufficient duration of assessment (≥5 years); additionally, these studies should adhere to the reporting guidelines of the Consolidated Standards of Reporting Trial (CONSORT).⁵⁸ 

In light of the findings of this meta-analysis, the conclusion is that smoking may be associated with significantly increased rates of implant failure and marginal bone loss. Research exploring various preventive and interventional measures that may limit the adverse effect of tobacco on implant outcomes is strongly encouraged. Additionally, the potential adverse effects of smoking on treatment outcomes should be discussed with the patient before treatment, and the dentists' clinical decisions should be case-specific.

Abbreviations

CCT

controlled clinical trial

HA

hydroxy apatite

ISFP

implant-supported fixed prosthesis

ISP

implant-supported prosthesis

MBL

marginal bone loss

NR

not reported

OR

odds ratio

PICO

Participants, Intervention, Control, Outcome

RCT

randomized clinical trial

RSR

radiographic success rate

WMD

weighted nean difference

This research project was supported by a grant from the Research Centre of the Female Scientific and Medical Colleges, Deanship of Scientific Research, King Saud University. Many thanks to Mr Anton Svendrovski for assisting with the statistical analysis.

The author states explicitly that there are no conflicts of interest in connection with this article.