Journal of Dental Research

http://jdr.sagepub.com/

Association between Marginal Bone Loss around Osseointegrated Mandibular Implants and Smoking Habits: A 10-year Follow-up Study
L.W. Lindquist, G.E. Carlsson and T. Jemt J DENT RES 1997 76: 1667 DOI: 10.1177/00220345970760100801 The online version of this article can be found at: http://jdr.sagepub.com/content/76/10/1667

Published by:
http://www.sagepublications.com

On behalf of:
International and American Associations for Dental Research

Additional services and information for Journal of Dental Research can be found at: Email Alerts: http://jdr.sagepub.com/cgi/alerts Subscriptions: http://jdr.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://jdr.sagepub.com/content/76/10/1667.refs.html

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

J Dent Res 76(10): 1667-1674, October, 1997

Association between Marginal Bone Loss around Osseointegrated Mandibular Implants and Smoking Habits: A 10-year Follow-up Study
L.W. Lindquist, G.E. Carlsson*, and T. Jemt'
Department of Prosthetic Dentistry, Faculty of Odontology, Goteborg University, Medicinaregatan 12, S-41390 Goteborg, Sweden; and 'The BrAnemark Clinic, Public Dental Health Service, Goteborg, Sweden; *to whom correspondence and reprint requests should be addressed

Abstract. While many factors are conceivable, occlusal loading and plaque-induced inflammation are frequently stated as the most important ones negatively affecting the prognosis of oral implants. Currently, little is known about the relative importance of such factors. The aim of this study was to analyze the influence of smoking and other possibly relevant factors on bone loss around mandibular implants. The participants were 45 edentulous patients, 21 smokers and 24 non-smokers, who were followed for a 10-year period after treatment with a fixed implant-supported prosthesis in the mandible. The peri-implant bone level was measured on intraoral radiographs, information about smoking habits was based on a careful interview, and oral hygiene was evaluated from clinical registration of plaque accumulation. Besides standard statistical methods, multiple linear regression models were constructed for estimation of the relative influence of some factors on peri-implant bone loss. The long-term results of the implant treatment were good, and only three implants (1%) were lost. The mean marginal bone loss around the mandibular implants was very small, about 1 mm for the entire 10-year period. It was greater in smokers than in nonsmokers and correlated to the amount of cigarette consumption. Smokers with poor oral hygiene showed greater marginal bone loss around the mandibular implants than those with good oral hygiene. Oral hygiene did not significantly affect bone loss in non-smokers. Multivariate analyses showed that smoking was the most important factor among those analyzed for association with peri-implant bone loss. The separate models for smokers and non-smokers revealed that oral hygiene had a greater impact on peri-implant bone loss among smokers than among non-smokers. This study showed that smoking was the most important factor affecting the rate of peri-implant bone loss, and that oral hygiene also had an influence, especially in smokers, while other factors, e.g., those associated with occlusal loading, were of minor importance. These results indicate that smoking habits should be included in analyses of implant survival and peri-implant bone loss.
Key words: bite force, cigarette smoking, dental implants, occlusal loading, oral hygiene.
Received August 26, 1996; Revised January 22, 1997; Accepted April 8, 1997

Introduction
It has often been stated that plaque-induced inflammation and occlusal loading are among the most important factors influencing the prognosis for oral implant treatment (Schnitman and Schulman, 1979; Lindhe et al., 1992; Quirynen et al., 1992). This might be a qualified guess based on animal experimentation, but evidence from human clinical studies is mainly anecdotal (Carlsson, 1996). Many other factors are certainly also of importance, including, for example, strict adherence to the treatment protocol, which was strongly emphasized in the early presentation of the osseointegration concept (Branemark, 1983). Today, much research is focused on implant material aspects, and the design and surface structure of the implants have been shown to be decisive for the establishment of reliable osseointegration (Henry, 1995; Wennerberg, 1996). In the long-term perspective, general health factors are most probably of significance (Weyant, 1994). The adverse effects of tobacco smoking on general health and biological tissues are undisputed (US Department of Health and Human Services, 1989). It is well-recognized that cigarette smoking is associated with impaired wound healing after surgical treatment in the oral cavity (Meechan et al., 1988; Preber and Bergstrom, 1990), as well as in other locations of the human body (Mosely et al., 1978). Reduced alveolar bone height in smokers compared with non-smokers has frequently been reported in different periodontal studies (Arno et al., 1959; Bergstrbm and Eliasson, 1987; Bolin et al., 1987, 1993). In an epidemiological study, it was also shown that tobacco smoking is a major risk factor for tooth loss in elderly subjects (Osterberg and Mellstrbm, 1986). However, not until the last decade has it been clearly documented that there is a direct association between smoking and periodontal bone loss and that smoking should be considered a major risk factor, separated from other factors, such as oral hygiene (Haber et al., 1993; Haber, 1994). Osseointegrated implants have been extremely successful in the treatment of total and partial edentulousness (Branemark, 1983; Lindquist and Carlsson, 1985; Albrektsson et al., 1986; Jemt et al., 1989; Zarb and Schmitt, 1990; Quirynen et al., 1991). Failures sometimes occur, however, and much interest has been directed in the
1667

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

1668

Lindquist et al.

j Dent Res 76(10) 1997

after a healing time of 4 mos. Within 3 wks after abutment connection, a full-arch 12-unit fixed mandibular prosthesis was fabricated in Type III gold alloy with resin teeth and with posterior bilateral cantilever sections of an average length of 15 mm (Lindquist et al., 1987). For all patients, the clinical prosthetic work was performed by one experienced prosthodontist and the technical laboratory work by one senior dental technician. After the patients had received their mandibular OIFPs, a follow-up program with annual clinical check-ups for the following 10 yrs was begun. During the follow-up period, 13 patients were also treated with fixed implant-supported prostheses in the maxilla (Carlsson and Lindquist, 1994). The remaining 32 patients had removable maxillary complete dentures throughout the observation period. The design of the study was approved by the Ethical Research Committee of the Faculties of Medicine and Odontology, Goteborg University. The patients were informed of the aims and methods of the study, and all consented to participate.

last few years to analyzing the causes of implant failure and other complications of implant treatment (Friberg et al., 1991; Jemt, 1994). Smoking has been found to be one important factor related to the loss of and soft tissue changes around implants, according to some studies (Bain and Moy, 1993; De Bruyn and Collaert, 1994; Gorman et al., 1994; Weyant, 1994). Bone loss around implants in relation to smoking does not seem to have been studied systematically. Because the titanium implant is attached directly to the alveolar bone, and no intermediate layer of connective tissue exists between the bone and the implant, as it does between the bone and a natural tooth, direct measurement of alveolar bone loss is possible in longitudinal studies of oral implants. Since the outer surface of the implant is scored by threads of an identical dimension, they serve as a measuring stick when a sensitive radiographic technique is used (Hollender and Rockler, 1980; Strid, 1985). Thus, conditions for evaluating, with a high degree of precision, the alveolar bone level surrounding the implant at different times after its insertion are optimal (Lindquist et al., 1988, 1996). The aim of this study was to measure and compare marginal bone loss in smoking and non-smoking implant patients and to analyze further the role of poor oral hygiene and other factors of possible relevance for marginal bone loss in these two groups.

Registrations
Bone loss. The marginal bone level around the implants was documented with apical intra-oral radiographs according to a previously described technique (Hollender and Rockler, 1980; Strid, 1985). The radiographic examinations for this investigation were performed at the time of insertion of the fixed prosthesis and then 1, 3, 5 to 6, and 10 yrs thereafter. The radiographs were examined with respect to osseointegration and bone loss around the implants. The marginal bone level (i.e., the height of the bone anchorage zone) of each implant was measured on the radiographs to the nearest 0.3 mm in relation to the reference point on the mesial and distal sides of the implant (Lindquist et al., 1988). Based on these measurements, a mean bone level was calculated for each implant as well as for the individual patient.

Materials and methods

This longitudinal study was comprised of 45 edentulous patients, 13 men and 32 women, under 65 yrs of age (range, 33 to 64) at the time of implant placement. All patients could be followed for the entire 10-year period except for one woman who died after 6 yrs. The composition of the study population with respect to age, gender, smoking habits, and other descriptive statistics is given in Table 1. All the patients had worn complete dentures in both jaws for at least 1 yr before consulting or being referred to the Dental School, Goteborg University, for possible treatment with fixed prostheses on osseointegrated implants (OIFPs) because of severe difficulties with the removable dentures. The selection, treatment principles, and examination procedures have previously been described in detail (Lindquist and Carlsson, 1985; Lindquist, 1987; Lindquist et al., 1987); only a brief description is given here. After initial examinations and recordings, all the patients were fitted with optimized complete dentures, mainly so that we could judge whether they could adapt to the improved removable prostheses, before installation of the mandibular implants was begun. Implant installation was carried out by three experienced surgeons during the period 1978 to 1982, according to principles described elsewhere (Adell et al., 1985). A total of 266 standard Branemark implants (Nobel Biocare AB, Goteborg, Sweden) was inserted. All implants had a length of 10 mm, the longest available at that time. Two of the implants were lost before abutment connection, and 5 were left unconnected ("sleeping"). All but eight patients received 6 implants in the lower jaw. Of the remaining patients, five had 5 and three had 4 implants connected. In each patient, two of the implants were placed bilaterally in the premolar region anterior to the mental foramina, the others in the incisor area. Standard abutment cylinders were connected to the implant

Oral hygiene. Plaque accumulation on the titanium abutments was recorded at the annual follow-ups during the observation period. The level of oral hygiene was evaluated on a three-point scale: 0 = no plaque; 1 = local plaque accumulation (< 25% of the visible abutment area); and 2 = general plaque accumulation (> 25%). We calculated an oral hygiene index retrospectively for each patient by scrutinizing the records and then adding the scores for the annual visits and dividing by the number of visits. The values for the index thus ranged from 0 (good) to 2 (poor).

Smoking habits. Information about smoking habits for the entire observation period was based on the patients' answers to an
interview at the 10-year follow-up and notes in the patient records. Non-smokers were subjects who, at the start of the 10year follow-up, had never smoked. Patients were also considered non-smokers if they had ceased smoking 1 yr or more before implant treatment. Current smokers were those who had been smoking cigarettes 1 yr or more before implant treatment and during the entire observation period of 10 yrs. Smoking exposure was expressed in cigarettes per day, and the median value of cigarette consumption for the subjects during the 10-year observation period in the smoking group was 20 cigarettes per day (range, 8 to 30) during the time of the investigation. None of the subjects used snuff tobacco.

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

J Dent Res 76(10) 1997

Smoking and Bone Loss around Implants

1669

Table 1. Descriptive statistics of the patients with respect to gender and smoking habits Bone Implants Maxilla n Aged Height 6 5 CD FP OH Non-smokers
male
5 58 2
3a

Bone Loss Mean SD

0.6 0.1

1.3

female Smokers male female

a b

19
8

55
46

3
2

16a b
7

3
1

7
1

12
7

0.5
1.3

0.7
1.4

0.3
0.7

13c

53

13

0.5

1.2

0.4

One implant failed before abutment connection. One implant failed after connection of the prosthesis. One patient died 6 yrs after the start of implant treatment. Age: Median age at insertion of implants, in yrs. Bone height: median value of the height of the mandibular residual ridge at time of implant placement according to a scale of 1 to 3. Implants: distribution of the patients according to number of implants connected. Maxilla: prosthetic situation in the maxilla at the 10-year follow-up. CD = complete denture; FP = fixed implant-supported prosthesis; OH = median value of oral hygiene index according to a scale 0-2. Bone loss: mean and standard deviation (SD) of peri-implant bone loss after 10 yrs, in mm.

Other recordings. At each follow-up, clinical examination was also performed of variables other than oral hygiene, such as mucosal conditions around the implants, prosthodontic complications and stability, occlusion, occlusal wear, and changes in function and esthetics of the prostheses (Lindquist, 1987; Lindquist et al., 1987). These observations were entered into the patients' records. We performed function tests to measure bite force using an apparatus with strain gauges mounted in a bite fork and chewing efficiency using a sieving method. A visual analog scale was used to record the patients' own evaluation of their chewing ability. The length of the posterior cantilever units from the distal side of the most posterior implant was measured in millimeters. The degree of bone resorption in the mandible prior to implant surgery was estimated according to a three-point scale by means of measurements on profile radiographs: 1 = little, 2 = moderate, and 3 = advanced resorption. Details of the recording methods have been presented previously (Lindquist and Carlsson, 1985; Lindquist, 1987).

0 to 3), oral hygiene (from 0 to 2), and smoking (yes, no, and number of cigarettes per day).

Results
The mean marginal bone loss around the mandibular implants was very small, about 1 mm for the entire 10-year period. It was significantly greater (P < 0.001), however, in smokers than in non-smokers. The mean values were already about twice as large for the smokers 1 yr after the implant treatment, and the ratio remained the same throughout the observation period. The difference between smokers and nonsmokers had reached a mean of about 0.6 mm after 10 yrs. The bone loss was smaller around the more posterior implants than around those in the mesial incisor region. In smokers, the difference in bone loss between implants in the incisor and premolar regions was already strongly significant (P < 0.001) after 1 yr and remained so throughout the period. In non-smokers, this difference was smaller and did not become statistically significant until after 5 to 6 yrs (Table 2). When both smoking and non-smoking groups were analyzed together, patients with poor oral hygiene were found to have greater marginal bone loss around the mandibular implants than those with good oral hygiene. The difference was evident after 1 yr but did not become significant (P < 0.05) until after 3 yrs. When the smoking and non-smoking groups were analyzed separately, no significant differences in bone loss between those with good and poor oral hygiene could be seen among the nonsmokers. Among the smokers, those with poor oral hygiene had a significantly greater bone loss (P < 0.001) than those with good oral hygiene at all examinations (Table 3, Fig.). When we analyzed the bone loss in relation to cigarette exposure by dividing the patients into three groups (nonsmokers, smokers smoking < 14 cigarettes a day, and smokers smoking > 14 cigarettes a day), we found significant differences among the means of all three groups. The non-smokers had less bone resorption than either of the smoking groups, and the smokers with low cigarette consumption had less bone loss than those with high consumption.

Statistical methods We first analyzed the influence of various factors on marginal bone loss by dividing the material into subgroups. Since no significant difference was found between men and women with respect to marginal bone loss, the subjects were pooled in all succeeding analyses. The significance of differences between two groups was tested by Student's t test for paired and unpaired observations, when applicable, and covariance analysis when differences among three groups were tested. We analyzed bivariate linear correlations by calculating Pearson's
coefficient of correlation in smokers and non-smokers separately. The influence of some factors on bone loss around the implants was studied in multiple linear regression models. Separate models were constructed for bone loss after 1 and 10 yrs after the start of implant treatment in smokers, nonsmokers, and the entire group. The following independent variables were tested: age (yrs), period of edentulousness before implant treatment (yrs), height of the mandibular ridge before implant treatment (low, medium, high), length of the cantilever sections (mm, mean of left and right sides), maximal bite force (N, measured at the one-year examination), occlusal wear (from

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

1670

Lindquist et al.

j Dent Res 76(10) 1997

Multiple linear regression models In the models for the group as a whole, smoking was the most important factor associated with bone loss. In the one-year model, smoking was the only significant factor, and the model had an explanatory value of 42% (R2 = 0.42). In the 10-year model, not only smoking but also oral hygiene was significant (Table 5). The separate models for smokers and nonsmokers revealed that oral hygiene had a greater influence on peri-implant bone loss among smokers than among non-smokers. Bite force was the second most important factor among the smokers, while length of cantilevers and pre-treatment ridge height were significant factors among the nonsmokers. The predictive (R2) values were higher in the smoking than in the nonsmoking group, and the models were not significant for the non-smokers (Table 5).

Table 2. Marginal bone loss (mean and standard deviation, in mm) around i,mplants inserted into different regions of the mandible in 21c smokers and 24 non-smokers 1, 3, 5/6, and 10 yrs after implant treatment Obs. Time Smokers Non-smokers Region Mean SD (yrs) Mean SD pd ***b Mesial incisors 0-1 0.86 0.62 0.40 0.30 Anterior 0-1 0.78 0.55 0.33 0.30
pap

*NSa
0.34 0.84 0.73 0.40 0.77 0.69 0.42 0.94 0.83

Premolar

0-1 0-3 0-3

pap

0.49

0.29 0.59 0.48 NS 0.38 0.76 0.64 0.43 0.90 0.78

0.47

0.27

Mesial incisors Anterior

Premolar Mesial incisors Anterior Premolar Mesial incisors Anterior Premolar
a b

1.28 1.07
0.59

0.34 0.34
0.30
0.39 0.39 0.32 0.46 0.45

0-3

0-5/6 0-5/6

1.45 1.24
pap

0-5/6
0-10 0-10 0-10
pap

0.71 1.71 1.51

0.89

0.50

0.35

NS = not significant. *** P < 0.001. C At the 10-year follow-up, the number of smokers was 20. d P denotes difference between smokers and non-smokers (righ t column) and between implants in the anterior and premolar regions (PaP).

Discussion

Of the 266 inserted implants, 3 (1%) were lost during the observation period. Two of the failures occurred before abutment connection and one 6 yrs after connection of the fixed prosthesis. The losses occurred in two patients, neither of whom was a smoker. In spite of the implant failures, both these patients, as well as all other patients, had stable fixed mandibular implant-supported prostheses at the last follow-up examination.
Bivariate correlations The correlation analyses gave different results for the smokers and non-smokers. Oral hygiene showed a moderate (r = 0.5-0.6) and significant association with bone loss at all examinations in the smokers, while it was not significant in the non-smokers (Table 4). In the smokers, correlation coefficients of similar strength (r = 0.5-0.6) were also found between the number of cigarettes per day and bone loss. The correlation between number of cigarettes smoked per day and oral hygiene was 0.56 (P < 0.01), indicating that high consumers had poorer oral hygiene than low consumers. One more variable, duration of edentulousness before implant treatment, showed significant, negative correlation to bone loss during the first part of the follow-up period. In the non-smokers, no significant correlations between bone loss and the other variables were found. In the entire group, smoking and oral hygiene but no other factors were significantly correlated to the average bone loss. The correlation between the bone loss up to 1 yr and from 1 to 10 yrs was r = 0.52 (P < 0.001) in the entire group, but differed between smokers (r = 0.43) and non-smokers (r = 0.15).

The outcome of this prospective study has clearly shown the extremely favorable results of treatment with osseointegrated implants supporting fixed prostheses in the mandible. However, the significant influence of smoking on mandibular peri-implant bone loss was also evident. This is not surprising, considering the negative effects of smoking in other areas of the human body as well as in the oral cavity (US Department of Health and Human Services, 1989). The effect of cigarette smoking on tissues in the oral cavity has also been well-established by many studies. Most of them have focused on the periodontal tissues, and, in general, smokers seem to have more marginal bone loss than nonsmokers. In the light of current knowledge of the effects of smoking, the finding in this study that high consumers had more bone loss than low consumers who had more than nonsmokers was a logical one. The underlying mechanisms, however, are not yet completely understood (Thomson et al., 1993). Many factors are involved in marginal bone loss in the natural dentition. Long-term longitudinal studies on marginal bone loss in humans are quite rare (Bolin et al., 1987). Such studies are said to require an observation period of at least 10 yrs for the results to be reliable, due to the relatively slow progression of the alveolar bone loss. In the present implant study, however, the effect of smoking on the rate of bone loss was already evident at the one-year followup. This indicates that smoking may have an influence on wound healing, as reported previously (Moseley et al., 1978; Preber and Bergstrom, 1990), and perhaps on the initial phase of osseointegration. Our long-term observations found an ongoing increase in the difference in bone loss between smokers and non-smokers. It should be noted, however, that the bone loss, although markedly greater in smokers, did not lead to any implant loss over the 10-year period in this group of patients. On the average, the bone loss was extremely small considering the long follow-up period. The only

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

j Dent Res 76(10) 1997

Smoking and Bone Loss around Implants

1671

implant failure that occurred after Table 3. Marginal bone loss (mean and standard deviation, SD, in mm) around implants prosthesis connection was probably with respect to smoking habits and oral hygiene caused by trauma. Oral Obs. Time Smokersa Non-smokers Information about smoking habits Hygiene pb (yrs) Mean SD n Mean SD n was recorded at a careful interview 0-1 0.50 0.21 10 0.37 0.16 14 NSd performed by the first author at the 10- Good ** Poor 0-1 0.85 0.35 11 0.25 0.23 10 year examination. Notes on smoking phc * NS habits had previously been entered into Good 0-3 0.67 0.23 10 0.47 0.20 14 *f patients' records, but not always Poor 0-3 1.09 0.36 11 0.41 0.26 10 * systematically. This was the reason ph NS smoking was not included in the earlier 0-5/6 0.85 0.26 10 0.59 0.25 14 * reports of this material (Lindquist, 1987; Good Poor 0-5/6 1.26 0.39 11 0.54 0.23 10 * Lindquist et al., 1988). The available ph *** NS notes, however, were now used to check Good 0-10 0.99 0.35 10 0.69 0.31 14 * the answers to the interviews. All Poor 0-10 1.61 0.45 10 0.65 0.27 10 * patients included in the study showed ph *** NS consistent smoking habits during the observation period. Cigarette smoking a Smokers and non-smokers are divided into two groups: those above (poor) and those was the dominant mode of tobacco use. b P below (good) the median value of the oral hygiene index, ranging from 0 to 2. the difference between smokers and non-smokers (right column). Only minor fluctuations over time in the c phdenotes denotes the difference between those with good and those with poor oral hygiene. number of cigarettes smoked per day d NS = not significant. were reported. Two subjects occasionally e ***p < 0.001. smoked cigarillos during the observation f * 0.01 < P< 0.05. period but only to a minor extent. This oral hygiene, according to the multivariate analysis. Patients was not considered a problem in the analyses. with the combination of smoking and poor oral hygiene had The importance of oral hygiene factors and the effect of about three times greater bone loss after 10 yrs than nonmicrobiota on peri-implant tissues have been intensively smokers. This suggests that when the importance of oral studied during the last few years, but the results have not been hygiene on peri-implant tissues is studied, smoking should be consistent (Adell et al., 1986; Lekholm et al., 1986; Lindquist et included as a factor in the analysis. al., 1988; Lindhe et al., 1992; Quirynen, 1993; Leonhardt, 1996). Although convincing documentation of the negative effects of Bone poor oral hygiene on implant Go- Smokers poor oral hygiene loss success is scarce, most authors Smokers good oral hygiene -a-mm and textbooks stress the - Nonsmokers good oral hygiene importance of good oral hygiene - -0- - Nonsmokers poor oral hygiene in implant patients. In a study covering 6 yrs after implant treatment, poor oral hygiene was found to be associated with 1.5 increased peri-implant bone loss (Lindquist et al., 1988). When smoking habits were included in the analysis of the same patient material, now followed for 10 yrs, 1 the influence of poor oral hygiene was less marked, and in nonsmokers the effect of poor oral : : : : :6 hygiene was insignificant (Fig.). 4E) Another observation indicating 0.5 -0- " that poor oral hygiene might be of less importance is the fact that oral hygiene was much better among the women than the men, 1 Years but there was no significant I I I I I I I difference in bone loss between 3 5 10 the genders (Table 1). On the other hand, the most important Figure. Mean marginal bonLe loss (mm) around implants with respect to smoking habits and oral factor for increased peri-implant hygiene. Smokers and non-si,mokers are divided into two groups: those above (poor) and those below bone loss in smokers was poor (good) the median value of t]:he oral hygiene index, ranging from 0 to 2.
...
. -

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

1672

Lindquist et al.

j Dent Res 76(10) 1997

Occlusal loading is another factor often stated to be of importance for implant survival. Theoretically, it is easy to show the unfavorable biomechanical effects of, for example, heavy loading, unstable occlusion, and long cantilevers (Skalak, 1983; Rangert et al., 1989; Brunski and Skalak, 1992). These statements are primarily supported by anecdotal observations, and systematic long-term clinical studies showing the consequences of such Non-smokers no significant correlations factors are scarce, perhaps partly due to the a 0.01 <P < 0.05. fact that it is difficult for overload to be b 0.001 < P < 0.01. determined in the clinical setting (Quirynen et al., 1992; Carlsson, 1996). In a three- to sixLittle has been published about the association between year report of the data from this study, there smoking and implant failure and between smoking and was some evidence that factors indicating heavy or bone loss around dental implants. In fact, a search of the unfavorable occlusal loading (reported tooth clenching, Medline system revealed only four studies concerning increased dental wear, long cantilever extensions) were dental implants and smoking between 1992 and May, 1996, associated with increased peri-implant bone loss (Lindquist and no publications before 1992. Two retrospective studies et al., 1988). The dominating role of smoking found in this 10(Bain and Moy, 1993; De Bruyn and Collaert, 1994) showed a year follow-up may have masked the effects of other factors, connection between cigarette smoking and failures of dental e.g., biomechanics and loading. The recording of increased implants, while another investigation (Weyant, 1994), which dental wear at the first examinations was certainly associated used a statistical logistic regression model on a large with great occlusal forces, but the recording at the last multicenter patient population, reported that patients' use of examination might not truly reflect long-term occlusal tobacco negatively affected peri-implant soft tissue health loading, since several of the prostheses including the but not implant failure. Implant survival was associated artificial teeth had been adjusted or replaced during the with the medical status of the patient, the surface coating observation period. Para-functional habits are known to material of the implant, and surgical and healing fluctuate over time and had obviously done so among our complications after implant installation. Another patients. In smokers, however, not only oral hygiene but also multicenter study, including more than 2000 implants, bite force was a significant factor. In non-smokers, the analyzed implant survival at stage 2 surgery. It was lengths of the cantilever sections were of some importance concluded that smoking is detrimental to implant success according to the multivariate analyses. Nevertheless, the (Gorman et al., 1994). results showed convincingly that smoking had a greater influence than factors indicating occlusal loading on peri-implant bone loss. Table 5. Significant associations with bone loss around implants ac cording to The length of the pre-treatment period of multiple regression models using 7 independent variables edentulousness and the height of the edentulous ridge were negatively correlated to the periGroup. Period Pb Sign. Factor R2c Model pd implant bone loss (Tables 4 and 5). This means that subjects who had extensive ridge resorption Whole Samplea 0-1 yr Smoking 0.004 0.42 0.01 after long periods of denture-wearing at the time 0-10 yrs Smoking 0.001 0.57 0.001 of implant treatment lost less bone around the Oral hygiene 0.03 implants than those with better-preserved ridges. Patient age was not significantly Smokers 0-1 yr Oral hygiene 0.004 0.74 0.03 correlated to peri-implant bone loss. Bite force 0.03 A large patient population and multivariate 0-10 yrs Oral hygiene 0.002 0.78 0.02 statistical methods are probably necessary for Bite force 0.06 analysis of the relative importance of the Non-smokers numerous factors that influence the long-term 0-1 yr Length of cantilever 0.01 0.50 0.08 prognosis of osseointegrated implants. This Oral hygiene 0.03 study has shown that smoking, which 0-10 yrs Pre-treatment ridge height 0.04 0.38 0.28 previously was seldom included in systematic Length of cantilever 0.09 analyses of implant survival, was the most factor of those correlated with important a cSenaratp (-1 nv%A ranctr.+mtr+ for hnan Ineer A1l-rinczr aaLtC mntilfNl LlUUtlb wPrP Wl-C LU>LlfULt:U n-1n lUF VUUtI JUSS alLg U-1 aCi U-1U yrs for the increased peri-implant bone loss. A clinical whole sample and for smokers and non-smokers separately. b P value for bivariate correlation. implication would be that smoking habits C R2 exploratory value for the multiple linear regression model. should be recorded and evaluated in the d P value for the multiple linear regression model. examination of candidates for implant
Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

Table 4. Significant (P < 0.05) correlation coefficients between mean ma:rginal bone loss around implants after various periods and 7 other factors in smokers arad non-smokers (there were no significant correlations among the non-smokers) Follow-up periods in years 0-1 1-10 0-3 0-5/6 0-1t0 Smokers 0.54a Oral hygiene 0.53a 0.57b 0.55b 0.64b Number of cigarettes 0.49a 0.52a 0.56b 0.45a 0.5i6b Length of edentulism -0.46a -0.46a Length of cantilevers 0.49a

j Dent Res 76(10) 1997

Smoking and Bone Loss around Implants

1673

treatment. Since good oral hygiene seemed to be able to reduce the negative influence of smoking, the recommendation that good oral hygiene be maintained in implant patients is corroborated. In conclusion, this study on edentulous patients provided with osseointegrated mandibular implants has shown extremely successful long-term results, with a loss of only 1% of the implants and a small mean peri-implant bone loss of about 1 mm over a 10-year period. The marginal periimplant bone loss was greater in smokers than in nonsmokers and correlated to the amount of cigarette consumption. In smokers, it was also greater in those with poor than in those with good oral hygiene. Multivariate analyses showed that smoking was the most important factor for peri-implant bone loss. Poor oral hygiene also had an influence, especially in smokers, while other factors analyzed, e.g., those related to occlusal loading, were of minor importance. These results indicate that smoking habits should be included in analyses of implant survival and peri-implant bone loss.

Acknowledgments
This study was supported by a grant (no. 9531) from Swedish Match-Svenska Tobaks AB. We are indebted to Dr. Tommy Johnsson, biostatistician, for help with the statistical analyses.

References
Adell R, Lekholm U, Branemark P-I (1985). Surgical procedures. In: Tissue-integrated prostheses. Osseointegration in dinical dentistry. Branemark P-I, Zarb GA, Albrektsson T, editors. Chicago: Quintessence, pp. 211-232. Adell R, Lekholm U, Rockler B, Branemark P-I, Lindhe J, Eriksson B, et al. (1986). Marginal tissue reactions at osseointegrated titanium fixtures. A 3-year longitudinal prospective study. Int J Oral Maxillofac Surg 15:39-52. Albrektsson T, Zarb GA, Worthington P, Ericsson AR (1986). The long-term efficacy of currently used dental implants: A review and proposed criteria for success. Int J Oral Maxillofac Implants 1:11-25. Arno A, Schei 0, Lovdal A, Waerhaug J (1959). Alveolar bone loss as a function of tobacco consumption. Acta Odontol Scand 17:3-10. Bain CA, Moy PK (1993). The association between the failure of dental implants and cigarette smoking. Int J Oral Maxillofac Implants 8:609-615. Bergstrom J, Eliasson S (1987). Cigarette smoking and alveolar bone height in subjects with a high standard of oral hygiene. J Clin Periodontol 14:466-469. Bolin A, Lavstedt S, Frithiof L, Henriksson CO (1987). Proximal alveolar bone loss in a longitudinal radiographic investigation: IV. Smoking and some other factors influencing the progress in individuals with at least 20 remaining teeth. Gerodontics 3:43-46. Bolin A, Eklund G, Frithiof L, Lavstedt S (1993). The effect of changed smoking habits on marginal bone loss. Swed
Dent J 17:211-216.

Branemark P-I (1983). Osseointegration and its experimental background. J Prosthet Dent 50:399-410.

Brunski JB, Skalak R (1992). Biomechanical considerations. In: Advanced osseointegration surgery: Applications in the maxillo-facial region. Worthington P, BrAnemark P-I, editors. Chicago: Quintessence Publishing Co., pp. 15-39. Carlsson GE (1996). Factors influencing bone resorption around osseointegrated implants in a long-term perspective. Proceedings of the 2nd World Congress of Osseointegration, Oct 3-6, 1996, Rome, pp. 42-47. Carlsson GE, Lindquist LW (1994). Ten year longitudinal study on masticatory function in edentulous patients treated with fixed complete dentures on osseointegrated implants. Int I Prosthodont 5:448-453. De Bruyn H, Collaert B (1994). The effect of smoking on early implant failure. Clin Oral Impl Res 5:260-264. Friberg B, Jemt T, Lekholm U (1991). Early failures in 4,641 consecutively placed BrAnemark dental implants: A study from stage 1 surgery to the connection of completed prostheses. Int J Oral Maxillofac Implants 6:142-146. Gorman LM, Lambert PM, Morris HF, Ochi S, Winkler S (1994). The effect of smoking on implant survival at second-stage surgery. Implant Dent 3:165-168. Haber J (1994). Smoking is a major risk factor for periodontitis. Curr Opin Periodontol 12-18. Haber J, Wattles J, Crowley M, Mandell R, Joshipura K, Kent RL (1993). Evidence for cigarette smoking as a major risk factor for periodontitis. J Periodontol 64:16-23. Henry PJ (1995). Implant hardware-Science or commodity development? J Dent Res 74:301-302. Hollender L, Rockler B (1980). Radiographic evaluation of osseointegrated implants of the jaws. Dento-Maxillo-Fac Radiol 9:91-95. Jemt T (1994). Fixed implant-supported prosthesis in the edentulous maxilla. Clin Oral Impl Res 5:142-147. Jemt T, Lekholm U, Adell R (1989). Osseointegrated implants in the treatment of partially edentulous patients: A preliminary study on 876 consecutively placed fixtures. Int J Oral Maxillofac Implants 4:211-217. Lekholm U, Adell R, Lindhe J, Branemark P-I, Eriksson B, Rockler B, et al. (1986). Marginal tissue reactions at osseointegrated titanium fixtures. A cross-sectional retrospective study. Int J Oral Maxillofac Surg 15:53-61. Leonhardt A (1996). Studies on bacterial relation to titanium (dissertation). Goteborg, Sweden: Goteborg University. Lindhe J, Berglundh T, Ericsson I, Liljenberg B, Marinello C (1992). Experimental breakdown of peri-implant and periodontal tissues. A study in the beagle dog. Clin Oral Impl Res 3:9-16. Lindquist LW (1987). Prosthetic rehabilitation of the edentulous mandible. A longitudinal study of treatment with tissue-integrated fixed prostheses. Swed Dent J Suppl 48:1-39. Lindquist LW, Carlsson GE (1985). Long-term effects on chewing with mandibular fixed prostheses on osseointegrated implants. Acta Odontol Scand 43:39-45. Lindquist LW, Carlsson GE, Glantz P-O (1987). Rehabilitation of the edentulous mandible with a tissueintegrated fixed prosthesis. A six-year longitudinal study. Quintessence Int 18:89-96. Lindquist LW, Rockler B, Carlsson GE (1988). Bone resorption around fixtures in edentulous patients treated

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010

1674

Lindquist et al.

j Dent Res 76(10) 1997

Implants Res 3:104-111. Rangert B, Jemt T, Jorneus L (1989). Forces and moments on Branemark implants. Int J Oral Maxillofac Implants 4:241-247. Schnitman PA, Schulman LB (1979). Recommendations of the consensus development conference on dental implants. J Am Dent Assoc 98:373-377. Skalak R (1983). Biomechanical considerations in osseointegrated prostheses. J Prosthet Dent 49:843-848. Strid K-G (1985). Radiographic results. In: Tissue-integrated prostheses. Osseointegration in clinical dentistry. Br'anemark P-I, Zarb GA, Albrektsson T, editors. Chicago: Quintessence Publ. Co., Inc., pp. 317-327. Thomson MR, Garito ML, Brown FH (1993). The role of smoking in periodontal diseases: A review of the literature. J Western Soc Periodontol, Periodont Abstr 41:5-9. US Department of Health and Human Services (1989). Reducing the health consequences of smoking: A report of the Surgeon General. US Public Health Service, Office on Smoking and Health. DHHS Publication No. (PHS) 81-50156, p. 269. Wennerberg A (1996). On surface roughness and implant incorporation (dissertation). Goteborg, Sweden: Goteborg University. Weyant RJ (1994). Characteristics associated with the loss and peri-implant tissue health of endosseous dental implants. Int J Oral Maxillofac Implants 9:95-102. Zarb GA, Schmitt A (1990). The longitudinal clinical effectiveness of osseointegrated dental implants: The Toronto study. Part II: The prosthetic results. J Prosthet Dent 64:53-61.

with mandibular fixed tissue-integrated prostheses. J Prosthet Dent 59:59-63. Lindquist LW, Carlsson GE, Jemt T (1996). A prospective 15year follow-up study of mandibular fixed prostheses supported by osseointegrated implants. Clin Oral Impl Res 7:329-336. Meechan JG, Macgregor ID, Rogers SN, Hobson RS, Bate JP, Dennison M (1988). The effect of smoking on immediate post-extraction socket filling with blood and on the incidence of painful socket. Br J Oral Maxillofac Surg 26:402-409. Mosely LH, Finseth F, Goody M (1978). Nicotine and its effect on wound healing. Plast Reconstr Surg 61:570-575. Osterberg T, Mellstrom D (1986). Tobacco smoking: A major risk factor for loss of teeth in three 70-year-old cohorts. Community Dent Oral Epidemiol 14:367-370. Preber H, Bergstrom J (1990). Effect of cigarette smoking on periodontal healing following surgical therapy. J Clin Periodontol 17:324-328. Quirynen M (1993). Tissue response to loading and microbiota. In: Osseointegration in oral rehabilitation. An introductory textbook. Naert I, van Steenberghe D, Worthington P, editors. London: Quintessence Publ. Co., pp. 171-180. Quirynen M, Naert I, van Steenberghe D, Schepers E, Calberson L, Theuniers G, et al. (1991). The cumulative failure rate of the Branemark system in the overdenture, the fixed partial, and the fixed full prostheses design: A prospective study on 1273 fixtures. J Head Neck Pathol 10:43-53. Quirynen M, Naert I, van Steenberghe D (1992). Fixture design and overload influenced marginal bone loss and fixture success in the Branemark system. Clin Oral

Downloaded from jdr.sagepub.com at International Association for Dental Research on July 20, 2010